The Political Mobilization of the European Left, 1860-1980: The Class Cleavage (Cambridge Studies in Comparative Politics)

THE POLITICAL M O B I L I Z AT I O N O F T H E E U R O P E A N L E F T, 1 8 6 0 – 1 9 8 0 In an in-depth comparative an...

Author: Stefano Bartolini

19 downloads 864 Views 5MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

THE POLITICAL M O B I L I Z AT I O N O F T H E E U R O P E A N L E F T, 1 8 6 0 – 1 9 8 0 In an in-depth comparative analysis, Stefano Bartolini studies the history of socialism and working-class politics in Western Europe. While examining the social contexts, organizational structures, and political developments of thirteen socialist experiences from the 1880s to the 1980s, he reconstructs the steps through which social conflict was translated and structured into an opposition, as well as how it developed its different organizational and ideological forms and how it managed more or less successfully to mobilize its reference groups politically. Bartolini provides a comparative framework that structures the wealth of material available on the history of each unit and allows him to assess the relative weight of the complex explanatory factors. Stefano Bartolini is Professor of Comparative Political Institutions at the European University Institute in Florence. He has contributed articles to numerous journals and has edited and written several books, including Identity, Competition, and Electoral Availability (Cambridge University Press, 1990), for which he won the UNESCO Stein Rokkan Award.

CAMBRIDGE STUDIES IN COMPARATIVE POLITICS General Editor

PETER LANGE Duke University Associate Editors

ROBERT H. BATES Harvard University ELLEN COMISSO University of California, San Diego PETER HALL Harvard University JOEL MIGDAL University of Washington HELEN MILNER Columbia University RONALD ROGOWSKI University of California, Los Angeles SIDNEY TARROW Cornell University OTHER BOOKS IN THE SERIES Carles Boix, Political Parties, Growth, and Equality: Conservative and Social Democratic Economic Strategies in the World Economy Catherine Boone, Merchant Capital and the Roots of State Power in Senegal, 1930–1985 Michael Bratton and Nicolas van de Walle, Democratic Experiments in Africa: Regime Transitions in Comparative Perspective Valerie Bunce, Leaving Socialism and Leaving the State: The End of Yugoslavia, the Soviet Union, and Czechoslovakia Ruth Berins Collier, Paths Toward Democracy: The Working Class and Elites in Western Europe and South America Donatella della Porta, Social Movements, Political Violence, and the State Gerald Easter, Reconstructing the State: Personal Networks and Elite Identity Roberto Franzosi, The Puzzle of Strikes: Class and State Strategies in Postwar Italy Geoffrey Garrett, Partisan Politics in the Global Economy Miriam Golden, Heroic Defeats: The Politics of Job Loss Frances Hagopian, Traditional Politics and Regime Change in Brazil J. Rogers Hollingsworth and Robert Boyer, eds., Contemporary Capitalism: The Embeddedness of Institutions Ellen Immergut Health Politics: Interests and Institutions in Western Europe Torben Iversen, Contested Economic Institutions Torben Iversen, Jonas Pontusson, and David Scoskice, eds., Unions, Employers, and Central Banks: Macroeconomic Coordination and Institutional Change in Social Market Economies continued after last page of index

TH E POL I TI CA L M O BI LI Z ATI ON OF T HE E URO PE A N L E F T , 1860–1980 The Class Cleavage

STEFANO BARTOLINI Department of Political and Social Sciences European University Institute Florence, Italy

CAMBRIDGE UNIVERSITY PRESS Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo Cambridge University Press The Edinburgh Building, Cambridge CB2 2RU, UK Published in the United States of America by Cambridge University Press, New York www.cambridge.org Information on this title: www.cambridge.org/9780521650212 © Cambridge University Press 2000 This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 2000 This digitally printed first paperback version 2007 A catalogue record for this publication is available from the British Library Library of Congress Cataloguing in Publication data Bartolini, Stefano. The political mobilization of the European left, 1860–1980 : The class cleavage / Stefano Bartolini. p. cm. – (Cambridge studies in comparative politics) Includes bibliographical references. ISBN 0-521-65021-6 (hardback) 1. Socialist parties – Europe, Western – History. 2. Socialism – Europe, Western – History. I. Title. II. Series. JN94.A979B37 2000 324.2'17'094 – dc21 99-33725 CIP ISBN-13 978-0-521-65021-2 hardback ISBN-10 0-521-65021-6 hardback ISBN-13 978-0-521-03343-5 paperback ISBN-10 0-521-03343-8 paperback

a Bianca e Duccio, per il nostro tempo perduto

CONTENTS

List of Figures and Tables Acknowledgments

page xiii xxiii

Introduction 1

2

3

THE CLASS CLEAVAGE: CONCEPTUAL AND METHODOLOGICAL FRAMEWORK Conceptual Framework Methodological Choices THE EUROPEAN LEFT: SIZE, IDEOLOGICAL ORIENTATION, AND ORGANIZATIONAL COHESION Size and Electoral Development Ideological Orientation and Radicalism Organizational Fragmentation and Communist Strength A Synthetic Map INDUSTRIALIZATION, URBANIZATION, AND LABOR’S RESPONSE Conceptual Framework Working-Class Constituency Formation and Socialist Mobilization Urbanization and Left Mobilization Social Mobilization Model Conclusion: The Limits of the Social Mobilization Model

1

9 11 35

54 54 66 97 120

122 122 130 163 167 174 ix

x

4

5

6

CONTENTS

CULTURAL HETEROGENEITY Dimensions of Cultural Heterogeneity Cultural Segmentation Cognitive Stratification Cultural Heterogeneity and Left Electoral Development Conclusion: Social Mobilization and Cultural Heterogeneity

180 180 184 192 199

ENFRANCHISEMENT The Role of the Franchise Earliness Tempo Reversals Turnout Enfranchisement and the Development of the European Left Role of the Franchise in the Model Enfranchisement Pattern, Social Mobilization, and Cultural Heterogeneity

206 206 209 215 220 221

ORGANIZATIONAL STRUCTURING AND MEMBERSHIP MOBILIZATION Early Organizational Consolidation Membership Mobilization Organizational Structuring, Membership Mobilization, and Electoral Development Organizational Consolidation, Membership Development, and the General Model

200

225 232 235

240 241 261 290 307

7

POLITICAL INTEGRATION State Response Responsible Government Fair Representation Access to Executive Power Stateness, Institutional Openness, and the Class Cleavage Hostility Toward the State and Ideological Orientation

312 313 335 348 358 391 405

8

CLEAVAGE STRUCTURES Opportunities for Social and Political Alliances Early Political Mobilization and the Role of Liberalism and the Bourgeoisie

411 412 416

CONTENTS

Mobilization of the Religious Cleavage and the Formation of Denominational Parties The Peasant Issue: Mobilization of the Peasantry Conclusions: The Resulting Class Cleavage 9

10

THE COMMUNIST SPLIT: UNITED AND DIVIDED LEFTS Interpretation of Communist Success or Failure Communism, Socialism, and Patterns of Social Mobilization and Economic Development Communism, Left Radicalism, and the Organizational Features of the Socialist Movement Institutional and Political Integration Conclusion: A Syndrome Model

xi

454 466 486 502 502 508 522 532 537

THE MACROCONSTELLATION OF CLASS CLEAVAGE STRUCTURING The Macroconstellation Three Models Concluding Consideration

546 547 558 569

DATA APPENDIX Party Composition of the Class Left by Country Data Files Notes Concerning the Socioeconomic Data Cultural Heterogeneity Data Sources of Party Membership Figures Sources of Trade Union Membership Figures Index of Coalition Potential Class Cleavage Inclusiveness and Distinctiveness List of Variables in the Election Data File

573 574 575 581 588 589 590 590 593

References Index

601 633

FIGURES AND TABLES

FIGURES 1.1. 1.2. 2.1. 2.2. 2.3. 2.4. 2.5. 2.6. 3.1. 3.2. 3.3. 3.4. 3.5. 3.6. 4.1. 4.2. 4.3. 5.1.

Dimensions of cleavage structuring Four modes of analysis combining cross-time and crossspace variation Boxplot of total left vote by country (1880–1989) Electoral development: lowest fit lines by period National electoral developments (1880–1989) A map of early socialist ideological orientation Left electoral fragmentation by decade Electoral development of the internal components of the left Thresholds of mature industrial society in Europe Industrial working-class and left vote: regression line by country and period Types of working-class structures and left vote Patterns of industrialization and urbanization Social mobilization and the left vote Structure of the social mobilization model data Cultural heterogeneity dimensions Cultural heterogeneity and the left vote (scattergram) Scattergram of the left vote and the sociocultural context Role of the franchise

page 26 46 57 59 64 87 98 112 137 152 157 170 173 176 183 201 203 210

xiii

xiv

F I G U R E S A N D TA B L E S

5.2.

Franchise, turnout, and left vote: national developments 5.3. Left vote and the combined index of social mobilization, cultural heterogeneity, and enfranchisement pattern 6.1. Socialist party membership as a percentage of the electorate 6.2. Communist party membership as a percentage of the electorate 6.3. All trade union members as a percentage of the dependent labor force 6.4. Left trade union members as a percentage of the dependent labor force 6.5. Non-left trade unionization by country 6.6. Modeling the organizational consolidation process 6.7. Association among partisan, corporate, and electoral mobilization 6.8. Correlation by decade between left votes and trade union density 6.9. Organizational density of the class cleavage (1918– 1985) 7.1. A theoretical framework for the analysis of governmental power 7.2. Socialist size and coalition potential at their cabinet entry 7.3. Coalition potential and governmental power 8.1. Alignment of social groups and political movements 8.2. Left electoral mobilization by timing and type of religious mobilization 8.3. Scatterplot of religious and class voting indices (1950– 1960s) 8.4. Inclusiveness and distinctiveness of the class cleavage 9.1. Class polarization in the countryside 9.2. Rank ordering of institutional integration and the socialist and communist vote (1918–1985) 10.1. The electoral mobilization model 10.2. The ideological orientation model 10.3. The organizational division model

216

239 266 269 279 280 282 292 298 299 304

381 382 384 413 492 494 497 519 535 561 565 569

F I G U R E S A N D TA B L E S

xv

TABLES 1.1.

Mobilization of the class left: ordering of influencing variables 2.1. European left average vote by country (% of valid votes, 1880–1989) 2.2. Average European left vote development (five-year means in thousands) 2.3. Average electoral strength of the national left by period and differences between periods 2.4. Industrial conflict levels (N strikers * N strikes) before and after World War I 2.5. Configurational aspects of the “War and Revolution” crisis of the socialist movement 2.6. Electoral presence of left parties other than Communist and socialist parties 2.7. Electoral size of socialist and Communist parties and of the total left (period averages) 2.8. Synthetic map of the European left experiences 3.1. Dimensions and indicators of industrialization 3.2. The European sector transformation (1880–1970) 3.3. Levels of manufacturing output per capita (triennial averages except for 1913) 3.4. Timing and length of industrial society and left electoral mobilization 3.5. Correlation between active population in the economic sector and electoral socialist mobilization 3.6. Status transformation and electoral mobilization 3.7. Global working-class constituency (workers in all sectors as % of the active population) 3.8. Industrial working-class size and left political mobilization 3.9. Internal composition of the working-class and left electoral mobilization 3.10. Mean left electoral mobilization by type of working class 3.11. Employers/occupied ratio in Western European countries

31

55 55 61 94 96 118 119 120 128 133 135 138 141 144 148 149 154 155 159

xvi

F I G U R E S A N D TA B L E S

3.12. Working-class concentration and left electoral mobilization 3.13. Correlation between mean left electoral size and mean rates of change in industrialization measures by periods 3.14. Urbanization rates in Western Europe 3.15. Electoral mobilization and urbanization levels 3.16. Patterns of historical social mobilization and left vote 4.1. Religious and linguistic heterogeneity in Western European countries 4.2. Mean left electoral support by levels of religious and linguistic fragmentation 4.3. Mean left vote by types of cultural fragmentation 4.4. Religious and linguistic heterogeneity and left vote (Pearson’s correlation) 4.5. Illiteracy rate (% unable to write and read) 4.6. Illiteracy and left electoral mobilization 4.7. Cultural heterogeneity and left vote 4.8. Mean left vote in different sociocultural contexts 4.9. Regression analysis of the sociocultural model 5.1. Rates of growth per decade in the male electorate 5.2. Comparative enfranchisement: Timing and tempo 5.3. Mean turnout by levels of the enfranchised electorate 5.4. Major increases in the electorate and corresponding variations in turnout 5.5. Pattern of enfranchisement and left vote 5.6. Electorate and turnout levels and size of the left vote 5.7. Differentials in electorate, turnout, and left vote before and after male and female enfranchising elections 5.8. Electorate levels and left vote levels: Association with lagged variables (1880–1917) 5.9. Association between social mobilization and left vote, controlling for electorate level 5.10. Enfranchisement pattern: Individual and combined index association with the left vote 5.11. General determinants of the left vote 6.1. Type of interlinkage between electoral and corporate mobilization 6.2. Socialist parties and trade union formation 6.3. Socialist party foundation, franchise, and social mobilization levels (cases rank-ordered according to the earliness of national party foundation)

160 164 165 167 171 186 188 189 191 195 198 200 203 204 219 221 223 224 227 227 230 233 234 237 237 242 246

251

F I G U R E S A N D TA B L E S

6.4. 6.5. 6.6.

6.7. 6.8. 6.9. 6.10. 6.11. 6.12. 6.13. 6.14. 6.15. 6.16.

7.1. 7.2. 7.3. 7.4. 7.5. 7.6. 7.7. 7.8. 7.9.

Trade unions and parties: Organizational consolidation (1880–1940) Average left partisan density (1918–1989) Trade union and left trade union membership as a percentage of the dependent labor force: Country rank ordering Development of collective bargaining in European countries Timing and type of development of collective bargaining Levels of trade union density by timing of institutionalization of collective bargaining Trade union density and development of collective bargaining Pattern of organizational consolidation (1880–1920) and levels of political mobilization (1900–1940) Relative weight of organizational consolidation factors (1900–1940) Levels of corporate, partisan, and electoral mobilization by decade (as % of national electorate) Types of left movements by organizational and electoral mobilization Regression of organizational variables on left vote (beta coefficients) Social mobilization, cultural heterogeneity, enfranchisement, organizational model, and left vote (beta coefficients) Modalities of state formation in Europe Dimensions of stateness (1880s–1920s) (average standardized scores) Press, association, and strike freedoms: Repression and harassment Timing of responsible government development Representational inequalities in the lower chambers Disrepresentation by party and by period (% of seats– % of votes) Mean misrepresentation of the left by country, period, and electoral system Executive entry of socialists and communists Socialist party presence in cabinets (1918–1966)

xvii

262 271

284 286 288 289 290 293 296 300 305 309

310 316 318 321 349 353 355 356 360 365

xviii

F I G U R E S A N D TA B L E S

7.10. Percentage of all cabinets including the socialists by number of parties in cabinet (1918–1966) 7.11. Cabinets with socialists, by type and by total months in office 7.12. Ideological composition and duration of cabinets with socialist parties 7.13. Dimension of governmental power 7.14. Socialist parties’ governmental experiences (countries ranked according to governmental power index) 7.15. Stateness, strength of commercial/industrial interests, and repression 7.16. Socialist party foundation and democratization sequences (1880–1918) 7.17. Institutional integration, socialist ideological orientation, and electoral mobilization (1880–1917) 7.18. Social mobilization, cultural heterogeneity, enfranchisement, organizational model, institutional integration, and left vote (beta coefficients) 7.19. The institutional integration syndrome and predominant ideological orientation 8.1. Types of churches 8.2. Patterns of religious mobilization in Western Europe 8.3. Independent peasantry in Western Europe: Basic features (mean values for the 1900 and 1939 censuses; family workers excluded) 8.4. Productivity of the agrarian world (1960s data) 8.5. Weight of agricultural workers within the European working class 8.6. Self-representational mobilization of the agrarian world 8.7. Electoral strength by religious composition 8.8. Proportion of left support by language and religion and by sense of identity in Switzerland 9.1. Timing and length of industrial society and Communist support 9.2. Social structure and Communist/socialist electoral support (1918–1980 elections) 9.3. Working-class composition and support of left parties 9.4. Rate of socioeconomic change and Communist strength 9.5. Socialist movement organization in the early twentieth century

367 369 374 376 378 394 399 400

404 408 456 467

470 473 476 477 490 491 511 511 513 515 527

F I G U R E S A N D TA B L E S

Correlation between the vote for different components of the left and organizational density indicators (1918–1985) 9.7. Pre–World War I (1900–1920) socialist electoral and organizational density and post–World War I Communist vote 9.8. Pre–World War I organizational density of the socialist movement 9.9. Institutional integration and vote of components of the left 9.10. A syndrome for Communist political support 10.1. The macroconstellation of class cleavage structuring 10.2. Macroprocesses and variation in outcomes

xix

9.6.

529

531 531 534 539 549 559

To predict the presence of such movements was simple; to predict which ones would be strong and which ones weak, which ones unified and which ones split down the middle, required much more knowledge of national conditions and developments and a much more elaborate model of the historical interaction process. Stein Rokkan, 1970

ACKNOWLEDGMENTS

Almost no part of this work has appeared in published form. Several points in the ﬁrst chapter are drawn from ‘‘On Time and Comparative Research,’’ Journal of Theoretical Politics, 5 1993: 131–167. A few others refer to parts in Identity, Competition and Electoral Availability (Cambridge University Press, 1990), and I am grateful to Peter Mair for his permission to use and elaborate on them. An early version of part of Chapter 6 was published as a Working Paper of the Juan March Institute in Madrid. Part of Chapter 5 has appeared as a Working Paper of the Institute of Political Science of the University of Barcelona. The empirical research on which this book is based was made easier by the ﬁnancial support of the European Union allocated by the Research Council of the European University Institute (Grant No. 36, 1995 and 1996). I want to thank Simon Dubbins, Gianfranco Miglio, and Stefano Bocconi for their research assistance in data collection; Niki Outhram for her revision of my English; and Mogens Pedersen and Peter Mair for their valuable comments on parts of the manuscript. Hans Peter Kriesi deserves special mention for his thorough reading of and comments on the entire manuscript The same applies to the anonymous readers of Cambridge University Press. I also want to express my grateful thanks to all the colleagues and members of the three institutions in which this book was written between 1990 and 1998: the Department of Political Sciences of the University of Trieste, the Department of Political Science of the University of Geneva,

xxiii

xxiv

ACKNOWLEDGMENTS

and the Department of Social and Political Sciences of the European University Institute of Florence. In the Florentine institution, two friends, whom I have known for a long time, deserve a ﬁnal and special wholehearted ‘‘thank you’’: Maureen Lechleitner and Marie-Ange Catotti. Maureen offered help that went beyond that usually provided by a secretary; she transformed herself into a research assistant. Marie-Ange, during the three years I served as Head of Department, shielded me from most problems, including those I myself created for her.

INTRODUCTION

ith the passing of time, what was regarded in the last quarter of the twentieth century as a troubling problem of limited economic growth will appear as a period of turbulence and social dislocation caused by the transition away from the two-century-old cycle of the Industrial Revolution. This cycle opened with the slow erosion of the agricultural population and culminated in a similar erosion of the industrial working class, because, by the 1970s, in the early industrialized countries, pure manufacturing activities were no longer able to increase employment. A distinct historical period seems to be coming to an end, even if it will take some time before its full effects will be visible everywhere. Even then, however, this is likely to take much less time than it once took for industrialization to spread from the early to the late arrivals in the Western world. In its various ideological streams, its political organizations, and its social and political battles, the history of socialism is linked closely to this two-century cycle. Indeed, its story, many would argue, now comes to an end with the closing of this period. The passing away of working-class politics, and of class politics as such, leaves open the question of how many of its aims have actually been achieved. This study takes a historical view of socialism and aims to reconstruct the steps through which a social conﬂict was translated and structured into a political opposition, how it developed its different organizational and ideological forms, and how it managed more or less successfully to mobilize politically its putative reference groups. The theme of this book is the variation in the electoral size, social cohesion, and organizational strength of the left in Western

W

1

2

THE POLITICAL MOBILIZATION OF THE LEFT

Europe. The topic is hardly new; indeed, histories of European socialism and the left are too numerous to be cited. What is original about this study, however, is the way in which the theme is approached. The analysis is organized in a strictly comparative framework aimed at the systematic appreciation of the social contexts, the organizational structures, and the political developments of thirteen Western European socialist experiences from the 1880s to the 1980s. In this sense, the goals of the study are methodological as well as substantive. The methodologically aware comparative social scientist is often subject to fears and doubts and is frequently not self-assured and sometimes even unhappy scholar. Unlike the case-study analyst, the comparativist is not protected by the self-conﬁdence that is generated by an in-depth knowledge of the topic at hand; by the reliance on a large body of literature sharing the same conﬁguration; or by the surrounding conventional wisdom that states that what is done makes sense. She or he needs to argue about diversity without resorting to the sorts of ad hoc hypotheses relating to one case that constitute the forte of the case specialist. Constantly worrying about the combination of an esprit de finesse with an esprit de ge´ome´trie, the comparativist must balance both in the full knowledge of the risk of falling under Pascal’s dictum: ‘‘. . . les esprits faux ne sont jamais ni ﬁns, ni ge´ome`tres.’’ Comparativists, in short, must always justify their research choices. I take it for granted that large-scale cross-country comparative research is unlikely to make any new discoveries in the history of the units analyzed. The purpose is, ﬁrst, to provide a comparative framework capable of structuring the wealth of material available on the history of each single unit; second, and on the basis of such a theoretical framework, to arrive at comparative judgments concerning the relative rank ordering of cases along the dimensions of the framework; and third, to assess the relative weight of single or complex explanatory factors. The comparative structure is thus the essential instrument for processing information. The yardstick of relative weight is the speciﬁcity of the exercise and the essence of the comparative research design. The testing of existing theses and generalizations, and sometimes the advancement of new ones, is obviously the ultimate goal of this work. In the study of one country, or even when dealing with a small number of countries, it is possible to refer to a large number of important variables that are relevant to each particular context. Indeed, some would argue that a case-centered strategy is actually the preferred option, and that it is typical of the work of classic authors such as Weber and Tocqueville. However, although case-oriented comparisons attempt to aim at the rep-

INTRODUCTION

3

resentation of a complex set of interrelations among key variables in different individual settings, I am still not convinced of the merits of this approach. In fact, I am also quite skeptical about the possibility of a truly comparative analysis that is not variable-oriented but rather case-oriented. In any case, I feel unable to carry out such an analysis, because I believe that cases cannot be compared to one another any more than individuals can be compared to one another. For in any comparative statement whatsoever, what lies at its heart is not a case, but rather the deﬁnition of a common dimension along which values or status may be attributed to cases. Hence, what we actually compare are the status and values of individual cases along a theoretically constructed common dimension. In this sense, the activity of comparison, to my mind, is entirely dependent on the intellectual and conceptual scheme that guides it. The point here is not to reopen the enduring debate between nominalists and realists, but simply to argue that a strategy of comparison based on cases rather than variables – that is, a strategy of comparison that assumes that it is possible to compare cases without common abstract properties in the form of variables – is, for me, difﬁcult to imagine and therefore impossible to perform. All of this serves to explain why a great deal of energy is devoted in this work to the development of a set of macrovariables that provides a common framework of comparison and to the collection of comparable data and information for each of the thirteen Western European countries included in the analysis. In the course of this work on the historical mobilization of the class cleavage, I have processed a wealth of historical material, and in order to utilize it fully, it would, of course, simply be easier to deal with one unit at a time, treating similar aspects and problems in the analysis of each unit. Indeed, to do so would allow the wealth of material that has been accumulated regarding each unit to be fully presented and, at the same time, for the ﬁrm control of each case to be maintained. In my opinion, however, this practice is unsatisfactory, because it leaves too much room for the possibility of sliding into description, for the temptation of introducing ad hoc aspects, and for the risks of resorting to systematic justiﬁcations about why particular cases and particular experiences should be regarded in one way or another. I have therefore chosen to follow a different strategy, and, as a result, I have found that even more energy has been required in order to maintain this framework and its dimensions. In other words, it has been difﬁcult to avoid adding information, hypotheses, and sheer guesswork about one country or case when the argument involved could not be generalized conceptually or empirically to the universe of cases. In consequence, I have often had to sacriﬁce the

4

THE POLITICAL MOBILIZATION OF THE LEFT

richness of the material at my disposal, for whenever it proved impossible to reconcile particular pieces of information with the overall macrocomparative framework, or to generalize conceptually across the countries those aspects and factors that were fundamental to just one or only some of them, I opted for exclusion. This strategy is costly, but it also has some advantages, several of which are discussed in the chapter that opens this book. The second major aim of this study is to reach what I have called a ‘‘comparative relative weighting’’ of factors. It is overwhelming to realize how much of the national literature is replete with arguments and statements that are unquestionably comparative in nature but that are expressed, at the same time, without making any explicit comparisons. Arguments concerning the ‘‘very radical nature’’ of a movement; the ‘‘early industrialization’’ of a country; the ‘‘high inclusiveness’’ of a regime; the ‘‘late enfranchisement’’ of another; and so on; as well as terms such as ‘‘high,’’ ‘‘low,’’ ‘‘early,’’ ‘‘late,’’ ‘‘sudden,’’ ‘‘gradual,’’ ‘‘strong,’’ and ‘‘weak’’ are all inherently comparative. The problem here is the following: Can the comparativist take the judgment of the national expert, who knows inﬁnitely more about the unit in question, as a solid and grounded comparative assessment? What is in question is obviously not the competence of the scholar, but rather the inevitable need in national accounts to utilize a terminology that is comparative without at the same time having a comparative framework of analysis. In these cases, the comparative judgment is normally arrived at not through any explicit comparative exercise, but rather by reference to the accumulated (and implicit) common knowledge and/or wisdom of a discipline. The task of the large-scale comparativist is to relativize this information within a comparative perspective. Because, however, explicit comparative research does not exist on many of the relevant concerns, the comparativist must also search for information that is often scattered and unsystematic, organize these materials along various comparative dimensions, and ﬁnally come to the sort of rank-ordering or quantitative or qualitative evaluation that is suggested by this analysis. In my experience, this effort has proved to be both arduous and unrewarding. The more we attempt to discover about a set of countries, the more the dimensions of the comparison increase and the more difﬁcult it is to come to a proper evaluation of the comparative position of each country. A factor that is regarded as important by country experts needs to be generalized as an inclusive comparative dimension along which all other cases must be located, even if, for these other cases, this element has not been seen as important and has therefore not been dealt with extensively by the respective country experts.

INTRODUCTION

5

In principle, this particular work has required the comparative analysis of a very large number of theoretically relevant dimensions: rates and levels of police repression; legislation concerning the recognition of political parties; comparative ﬁgures on the average size of industry, types of land tenure, and average land holdings; ﬁgures on the territorial organizational strength of parties; and so on. In some cases, this information exists in the form of important monographs or source books. In others, it can be collected only with the enormous effort of trawling through several scattered sources whose level of reliability and standardization is unknown. In still other cases, only indirect qualitative judgments and estimates are available in national accounts. Indeed, this has been the main problem that I have encountered in dealing with the enormous existing historical literature on the parties of the left and the working class: Only rarely does this literature address those questions that appear crucial from a comparative perspective, and even when these questions are considered, the insistence on a national framework of reference makes it difﬁcult to evaluate the real comparative signiﬁcance of the results. For instance, what may appear to be a strong organizational network when seen within the context of a particular country’s history may well turn out to be very weak when placed within a comparative perspective. In other words, the comparative panorama requires a constant reevaluation of national expert judgments and a constant reassessment of what has otherwise been considered with reference to national development to be strong or weak, early or late, or whatever. The core of this book is the attempt to devise a common comparative framework and to construct a net that allows the maximum available information to be organized as a common dimension of cross-national comparison. Within this general framework, descriptive accounts of the different national experiences are organized along common lines; wellknown hypotheses are speciﬁed and linked to one another in terms of mutual speciﬁcations; ﬁnally, and whenever the necessary evidence is available, a comparative test is attempted. This research lies at the crossroads of two traditions: (1) that of the imprint of stratiﬁcation and social class, and of ‘‘social factors’’ in general, on mass political behavior and (2) that of the independent effect of speciﬁc political and institutional variables. The study of electoral mobilization requires the combination of both perspectives. It invariably turns out, however, that the political data that are required (regarding organization, propaganda, and so on, that is, the list of precisely those political variables that the late Stein Rokkan continuously reminded us we should ﬁll with data) are unavailable; or that they need to be reconstructed and standardized from hundreds of different sources; or that they are scattered and

6

THE POLITICAL MOBILIZATION OF THE LEFT

unsystematic. Conversely, and as usual, the socioeconomic data are more easily available. The result is that inevitably, if unwittingly, we end up with a ‘‘sociology of politics’’ rather than a ‘‘political sociology,’’ with the socioeconomic aspects as the input and the political and institutional aspects as the output. For this reason, I have devoted a great deal of attention to the political, organizational, and institutional comparative aspects. Of the four major parts into which this work is divided, the ﬁrst views the political mobilization of labor mainly as a response to patterns of economic development, industrialization and urbanization; the second part relates these responses to the degree of homogeneity of the national cultural environment; the third part relates them to organizational models and developments; and the fourth part relates them to the patterns of European political and institutional integration. In the concluding section, I have attempted to describe a model of interaction between socioeconomic, cultural, organizational, political, and institutional variables in order to combine these different perspectives into a general interactive model. The structure of the chapters of this book closely follows the aspects and theoretical framework just sketched. Chapter 1 discusses the general conceptual framework of the book and the key methodological problems it poses. Chapter 2 describes the variance in the historical development of the European left from the point of view of electoral size, organizational cohesion, and ideological orientation. Chapter 3 concerns the structuring of the class cleavage in light of the socioeconomic transformations and conditions that shaped its putative social constituency. Comparative patterns of industrialization and urbanization are related to the cross-country and over-time development of electoral support for the left. Chapter 4 investigates the level of cultural homogeneity of the class environment. Social mobilization and cultural homogeneity are seen as the two major long-term determinants of the environmental context of class cleavage structuring. Chapter 5 discusses the historical development of the franchise and its ambiguous role in terms of integration and representation. Chapter 6 deals with the organizational dimension of cleavage structuring, comparing the conditions of early organizational consolidation as well as the levels of electoral, corporate, and partisan organizational density of different socialist movements. Chapters 7 and 8 move to the last step of institutional and political integration, discussing respectively the openness/closure of the institutional sys-

INTRODUCTION

7

tem and the opportunities for political alliance offered to socialist movements by the cleavage structure of their countries. Chapter 9 reconsiders all these aspects in speciﬁc reference to the issue of organizational cohesion and, in particular, to the question of the strength of communism after World War I and World War II. The problem deserves special attention because communism is both the most conspicuous instance of organizational split and the most important ideological innovation and radical version of the post–World War I socialist movement. Chapter 10 brings together the different threads of the argument, viewing the different conﬁgurations of class cleavage structuring as results of particular preconditions, historical sequences, and crucial junctures. Finally, the notes concerning the data sources and the data problems of this research are presented in the Appendix. A ﬁnal word about the choice of the spatial and temporal range of this research: thirteen countries across a century. More will follow about the methodological problems of this kind of comparison in the next chapter. Here I want to explain the reasons that pushed me into embarking on a project that I will never consider undertaking again in my future work. In my early doctoral studies, a distinguished economic historian and demographer told me a story. He had asked his professor if he could do a thesis dissertation on the history of world population and was given the demographic history of a small Romagna village during one decade of the seventeenth century to study. The moral of this story is that eventually he became a mature scholar by writing one of the most authoritative histories of world population.1 The fact that I am now approaching a similarly broad question regarding the mobilization of the left in Western Europe should not mislead the reader. Although a fairly keen reader of history, I am not a professional historian and I have never made a detailed analysis of the mobilization of the left in a small village in my native Tuscany, say from 1870 to 1892. In fact, the story is relevant because in my case it worked the other way round: I actually began in the second half of the 1970s by studying socialism (and communism) in Italy and in Tuscany in particular. I later thought that a comparison with an equivalent French region would be useful, and ﬁnally I opted for a general comparison of the socialist–communist split in both of these countries. I probably read most of the available literature on the topic, and I also spent some time in the 1

This was Carlo Cipolla in 1978, telling his personal story to an eager audience of Italian students who were hosted – in the early days of the European University Institute life – by another distinguished Italian scholar, Mauro Cappelletti.

8

THE POLITICAL MOBILIZATION OF THE LEFT

Archives d’Etat in Paris and Rome searching for information about the early penetration of socialism in rural France and Italy. Several years were spent on this project; three boxes of notes were collected; the structure of the book was ﬁnally ready, and a few chapters were drafted. However, the book was never completed. The point here is that while most professional historians would react to a French–Italian comparison by feeling uneasy about overgeneralization, the lack of detailed knowledge, the need to go deeper into local speciﬁcities, and so on, I ended up by experiencing the opposite. The more I compared socialism and communism in Florence and Lucca in Tuscany, the more I uncovered variables whose weight could not be evaluated except by broadening the scope of comparison. As yet more information was revealed about the differences between Tuscany and the Veneto in Italy, the more it seemed that these hypotheses needed to be related to the sorts of broader questions that could not be studied at this level in order for them to be validated. Moreover, as the process continued, I began to question whether French and Italian differences and similarities in both explanans and explanandum should not really be tested in an even broader context, a context in which these differences and similarities could take on a different shape. Was it possible to explain the strength of communism without also considering communist movements that were weak or had failed? Why did reformist socialism fail in both countries? Was there something common to France and Italy that could not be perceived without also considering other countries? This endless chain of questions eventually generated the need for a more general exploration and interpretive study of socialism and communism at a wider level. If anybody feels that in order to understand the history of the world you must ﬁrst be very familiar with and conﬁdent about the history of a village, then in my case the reverse came to pass: In order to be aware of the key question that needs to be asked in your village, you must ﬁrst start with a broad understanding of the history of the world.

1

THE CLASS CLEAVAGE: CONCEPTUAL AND METHODOLOGICAL FRAMEWORK o deﬁne what is ‘‘left’’ within the European cultural and political experience, scholars have come up with a variety of focuses.1 The connotation of ‘‘left’’ may be independent of any school or doctrine and may identify a position of loyalty to the original programs, to the statu nascenti doctrine, or to the spirit of the original creed. This conception justiﬁes such terms as ‘‘dynastic left’’ (liberal parliamentarians who installed Luis Philippe [1830] on the throne but later opposed Guizot in the name of the original manifesto); fascismo di sinistra (the original radical corporativist and anticapitalist spirit of the fascist movement, as expressed in the San Sepolcro manifesto of 1919 or in the punti di Verona); and ‘‘Catholic left’’ (linked closely to the original evangelical message of solidarity and egalitarianism and to the social doctrine stemming from it). More typical of philosophical analysis is to view ‘‘left’’ (and ‘‘right,’’ of course) as referring to patterns of thought and behavior that are embedded profoundly and permanently in human nature: ‘‘to become’’ versus ‘‘to be’’; change versus conservation; the ontological opposition between a righthanded and a left-handed cosmology. Another tradition is to search for the permanent value, or constant guide, of the left – the general principle that it embodies and that differentiates it from any other current of political thinking. The emphasis is most frequently placed on the value of ‘‘equality,’’ although this is deﬁned in different ways.2 Finally, in a more histori-

T

1

2

For general overviews, see Laponce (1981) and Cofrancesco (1990: 4th ed., vol. 18, 883– 884). For this line of search, see Bobbio (1994) and Lukes (1995). 9

10

THE POLITICAL MOBILIZATION OF THE LEFT

cally deﬁned context, ‘‘left’’ is a spatial location, originally linked to the position within the parliamentary hemicycle. This conception of the left has little to do with issues and principles; it is relational and changes over time, just as it also depends on what else deﬁnes the spatial continuum. All these meanings attempt to deﬁne the left independently of the name of its historical actor. My aim here is not so ambitious and the left I will be speaking of is identiﬁed with a speciﬁc set of ideas and political and social organizations stemming from the Industrial Revolution: socialism. From a broad historical perspective, the patterns of industrialization and democratization resulted in the structuring of working-class movements that developed either into a single socialist party or, alternatively, into a socialist party plus a communist party. Later on, this picture was complicated by the emergence of other parties that resulted from splits within the two main tendencies. The class conﬂict is mostly responsible for the similarity of ‘‘party landscapes’’ across Europe. It was the only social conﬂict to be politically mobilized in every European country, contributing to the standardization of party systems. The ubiquitous presence of socialist and communist parties is indeed the most visible common feature of European party systems, while most of the variance among systems is accounted for by other nonindustrial or preindustrial cleavages, the decisive contrasts of which shaped the individual constellation of each system.3 This is a historical identiﬁcation of the left with no ambition to being a theoretical deﬁnition, and I am aware of its shortcomings. Grouping all these parties4 since the end of the nineteenth century into one category such as the ‘‘class left,’’ on the assumption that they all represent and hinge on the class cleavage, is a daring task from ideological, political, and social points of view. First, ‘‘other left’’ parties existed, which were neither ofﬁcial socialists nor ofﬁcial communists and which constituted a far from homogeneous category, including small parties of the extreme left as well as right-wing socialist groups with a more humanitarian or radical ﬂavor. Second, the electoral combination of these parties does not necessarily justify their political combination, as they were often strongly opposed on many issues. Finally, it is equally audacious to regroup these parties according to their connection with the working or lower classes. The classi3 4

Rokkan (1970b: 130). For a list of the parties that have been included in the left, the reader can consult the Appendix. Beyond the ofﬁcial socialist and communist parties, there are a number of other left parties whose inclusion or exclusion may be more controversial. As indicated in the text, I have included such parties that were originally splinter groups or wings of the historical socialist and communist organizations.

CLASS CLEAVAGE: FRAMEWORK

11

ﬁcation of parties that appeal to, are supported by, and represent such social groups may have little in common across countries. However, these parties are unquestionably part of the genetic process of lower-class enfranchisement and early political mobilization; they had and have maintained closer contact with the trade unions than any other political family, are generally regarded as being part of one tendance,5 and the literature is rich in hypotheses concerning this political construct, used as a meaningful term of reference for long-term electoral changes. The solution I have chosen is to subsume the three elements of the social constituency (support), the ideological orientation (appeal), and the organizational structures (representation) within the general historical process of class cleavage structuring. By means of a genetic approach, I hope to avoid having to resort to an implicit ‘‘class theory’’ of political representation. The goal of this chapter is to discuss the nature of the class cleavage that stemmed from the Industrial Revolution and political democratization. In the ﬁrst section, the concept of cleavage will be discussed and deﬁned; then the peculiarities of the class cleavage compared to other cleavages will be speciﬁed; and ﬁnally, the macroconstellation of historical processes within which the the class cleavage was structured will be identiﬁed. In this process, I will also deﬁne my dependent and independent variables more accurately, presenting thereby the general framework for this study. In the second section, the methodological choices that underlay this research are brieﬂy discussed and justiﬁed.

CONCEPTUAL FRAMEWORK POLITICAL MOBILIZATION AND CLASS CLEAVAGE STRUCTURING ‘‘Mobilization’’ is an ambiguous term. Imported from totalitarian theory and analysis and later used in all sorts of different contexts, it now conveys the meaning of a complex process of self-mobilization and heteromobilization – of ‘‘being mobilized’’ and of mobilizing.6 Mobilization was a multifaceted process of citizen involvement in the (post-)national and 5

6

The rather telling introduction of this term by Andre´ Siegfried should be noted: ‘‘Mainly preoccupied with the reality, I therefore concentrate less on parties – superﬁcial and continuously changing categories – than on the basic tendencies’’; Siegfried (1913: xxiv). See Neumann (1956b: 395–421) on ‘‘integration’’ parties addressing themselves to speciﬁc social groups that they deliberately try to mobilize, integration being at one and the same time the result of self-interest and organizational enterprise.

12

THE POLITICAL MOBILIZATION OF THE LEFT

industrial phases of modernization. Citizens were progressively mobilized in various nonpolitical spheres: by capitalism and industrialization in the economic sphere through media such as exchange and money; through the extension of the market; geographical and labor mobility; the imposition of tariffs, credit, and capital procedures and techniques; and the availability of services and goods. They were also mobilized by the military and administrative machine of the state, as soldiers, as subjects of administrative agencies, and through traveling and residential restrictions and/or liberalization. They were also mobilized culturally through scripts and other mass media into ideological, religious, and ethnolinguistic movements by socializing agencies of the nationally dominant culture, as well as by dissident intellectuals, missionaries and messages, news, and so on.7 In the Western experience of the last two centuries, mobilization appears to have acquired a self-sustaining impulse, with spillover effects between one sphere and others. Once started, it became an ongoing process whereby change concerned quantitative growth – new recruits and generation turnover – as well as qualitative structural modiﬁcations in the main forms and agencies. Different phases or waves of (economic/administrative/cultural) political mobilization can be distinguished. However, it was the ﬁrst wave that was of paramount importance because it not only opened the door to successive waves, it also set the original opportunity structure within which those that followed had to be accommodated. The ﬁrst political mobilization was the process by which former subject individuals were initially recruited as active participants in forms of nationwide organizational and electoral activities for the purpose of influencing political decision making;8 in order for this to take off, a minimum level of the other forms of mobilization had to be reached. Vertical and downward-ﬁrst political mobilization was not necessarily monopolized by new actors such as political parties and interest 7

8

On the relationship between political and other types of mobilization see Nettl (1967: 115–122). An analytical discussion of the concept of political mobilization is presented in Nedelmann (1987: 181–191, 199), where the distinction between the three dimensions of mobilization as ‘‘formation of interest,’’ ‘‘management of emotions,’’ and ‘‘development of instrumental capacities’’ is developed. Nedelmann’s general deﬁnition of political mobilization as ‘‘the actors’ attempts to inﬂuence the existing distribution of power’’ is too broad for our purposes. Such a deﬁnition is introduced as the result of dissatisfaction with others who limit political mobilization to the processes of authority legitimization (Nettl 1967, cited in footnote 7) or attribute excessive emphasis to the dimension of instrumentality, in the sense of mobilization as ‘‘resource control’’ (as utilized by Tilly [1978]). The emphasis on ‘‘recruitment of citizens in active political participation’’ implies a reference to the three dimensions listed by Nedelmann. However, the main concern and empirical data refer here to the development of instrumental capacities.

CLASS CLEAVAGE: FRAMEWORK

13

organizations. Political and even electoral mobilization was performed by governments, state bureaucracy, charismatic leadership, and so on. In these cases, parties and interest organizations competed with other established agencies that preceded them.9 However, they remained the most important early mobilization actors. Special attention should be paid to the specific interaction between the formation of electoral alternatives in the political arena – that is, the formation of speciﬁc political organizations for the mobilization of the vote – and the structuring of mass organizations in the corporate channelof-interest organization. The set of electoral and corporate organizations and their interaction depended on the structure of political opportunity and on the strategic choice of mobilizing actors during the formation and politicization of a given cleavage line. Drawing on Rokkan’s rich contribution to this ﬁeld,10 this process can be summarized in a set of analytical steps. I have slightly modiﬁed the terminology, as I believe the term ‘‘cleavage’’ should be kept for the politicized dividing line and not applied to the original functional, cultural, or territorial conﬂict. 1. The initial generation of oppositions due to differences of interest and/or Weltanschauung generated by the macroprocesses of modernization: monetarization, urbanization, secularization, cultural standardization, industrialization, administrative control, and centralization. 2. The crystallization of opposition lines into conflicts over public policy once (and if) the centralization of political decision making became established. 3. The emergence of alliances of political entrepreneurs engaged in mobilizing support for one set of policies against others. 4. The choice of mobilization strategy made by such entrepreneurs. a. Action through and reliance on preestablished community and other association networks. b. Action through and reliance on the development of purpose-speciﬁc membership organizations. 5. The choice of arena for the confrontation of mobilized resources. a. Aggregation of votes/members for political/electoral contests. b. Direct action (strikes, pressure through public demonstrations, revolt, revolution, etc.). 9

10

H. Daalder (1966a: 43–77) directly tackles the issue of the extent to which and the conditions in which parties more or less successfully monopolized such a role vis-a`-vis other agencies. A very large part of Rokkan’s work was devoted to the problem of cleavage formation. The classic reference is (1970b: 72–144).

14

THE POLITICAL MOBILIZATION OF THE LEFT

Historically, different alliances of entrepreneurs have chosen dissimilar strategies, both in terms of organizational choice (point 4) and in terms of confrontation-area choice (point 5). Some have relied more intensively on preexisting networks of association, such as occupational, cultural, and religious groups; others have engaged forcefully in the development of their own speciﬁc organizational weapons, as distinct and autonomous as possible from others. Some have preferred to concentrate their efforts for demand satisfaction in purely political and electoral strategies; others have resorted to more-direct actions in the market and society. For parties and groups, different strategies in different contexts have yielded different payoffs, ranging from gaining public recognition and legitimacy to substantial success through speciﬁc legislation, agreements, and package deals with the state and/or other political forces. In the analysis of historical cases, ﬁnal payoffs are of great importance. Their evaluation by the actors implied feedback reactions, while any dissatisfaction with existing payoffs and arrangements involved changes in organizational and confrontationarena strategy. In his general classiﬁcation of historical conﬂicts arising in the course of European modernization, Rokkan simpliﬁes the analysis by concentrating on broad fronts of conﬂict in the national histories. He distinguishes four critical cleavages that he broadly attributes as the consequences of two revolutions: the national and the industrial. The processes of formation of the nation-state usually provide the potential for two fundamental conﬂict lines: (1) the dominant cultural group and the nation builders’ elite versus the ethnically, linguistically, or religiously distinctive subjected population (i.e., external and internal conﬂicts concerning cultural and religious identities) and (2) the attempt by the nation-state to centralize, standardize, and mobilize versus the church and its traditional encroachments and privileges in society (e.g., secular versus clerical control of mass education). The Industrial Revolution also produced two lines of conﬂict: (1) the ﬁrst between the predominantly rural landed interests and the emerging classes of commercial and industrial entrepreneurs (mainly over tariff policies: free trade versus the protection of agricultural products); (2) the second, which split the urban milieu, between the owning classes and the tenants and workers (free enterprise versus state control; rights of workers versus rights of property owners). The age of Reformation and Counter-Reformation signaled the climax of the conﬂicts between the center and the periphery. The result of this was generally the strengthening and consolidation of territorial boundaries, as well as that of the linked issue of the cultural and religious identity of

CLASS CLEAVAGE: FRAMEWORK

15

the state within these boundaries. In the new polities resulting from national secessions, similar conﬂicts were postponed and reemerged only later in the nineteenth century. The colossal political mobilization that occurred during the French Revolution and its spread with the Napoleonic wars also determined the emergence of the potential conﬂict between the mobilization efforts of the state – with its need for system support and legitimacy – and the resistance of the church. The key issue quickly became the control of the growing mass-education and welfare provisions. The structural conditions for the urban–rural economic conﬂict developed later. This was the result of the growth in world trade and industrial production during the nineteenth century, and it determined the conﬂict between the landed and urban interests regarding tariff problems, which in turn were linked to issues of the maintenance of acquired status and the recognition of achievement. These conﬂicts were not necessarily translated into politics everywhere, and, indeed, they were stronger or weaker according to their contextual situations. In Rokkan’s reconstruction, the roots of these cleavages all predate the roots of the class cleavage. This point is often overlooked due to the fact that, even if these conﬂicts had already been present for a long time, their electoral mobilization tended to coincide historically with that of the class cleavage. That is, although they were the result of very different and longer lasting conﬂicts, they were all mobilized at exactly the time that suffrage was extended to the lower classes. Therefore, to fully appreciate the historical sequence of conﬂict crystallization, one should not concentrate exclusively on electoral politics, but rather on preelectoral forms of organizational developments and political conﬂict. This clariﬁes the marked historicity in the formation of basic conﬂict lines. This aspect of the historical sequence of conﬂict formation and cleavage structuring is crucial, given that the alliances among insiders at each given historical moment tended to reduce the alliance opportunities for later newcomers. THE THREE DIMENSIONS OF CLEAVAGE STRUCTURING Despite his historically creative use of the concept of cleavage, Rokkan never gave a clear deﬁnition of what he meant by the term. Its meaning has therefore remained loose,11 and the concept has been and is used in reference to all sorts of divisions and conﬂicts. Various authors refer to 11

See Zuckermann (1975: 231–248) and (1982: 131–144).

16

THE POLITICAL MOBILIZATION OF THE LEFT

cleavages as ‘‘political’’ and deﬁne them in terms of political attitudes and behavior, depriving the concept of any link to social structural variables.12 Others use the term ‘‘social cleavages’’ to indicate nothing more than the divisions implied by social stratiﬁcation. Still others have identiﬁed ‘‘cultural cleavages,’’ assuming that ‘‘it is a set of beliefs rather than any demographic attribute that deﬁnes one’s location along the cleavage.’’13 Contrary to this reduction of the concept of cleavage to one predominant dimension, more-complex conceptualizations are formulated in Eckstein’s concept of ‘‘segmental cleavages’’ and in the distinction made by Allardt and Pesonen between ‘‘structural’’ and ‘‘nonstructural’’ cleavage.14 The essential problem with the concept of cleavage lies in its intermediary location between the two main approaches of political sociology: that of social stratiﬁcation and its impact on institutions and political behavior, on the one hand, and that of political institutions and their impact on social structure and change, on the other.15 At the theoretical level, the synthesis of these different approaches is difﬁcult; for this reason, the concept of cleavage is often either reduced down to that of social cleavage or raised up to that of political cleavage. To solve this problem, the concept of cleavage has to be regarded as a link between social structure and political order, and – in much the same way as the Marxist concept of class – it should not be considered as a descriptive concept aimed at the identiﬁcation of a particular reality. Within such a perspective, the concept of cleavage can be seen to incorporate three dimensions:16 an empirical element, which identiﬁes the 12 13

14

15 16

Cf. Dahl (1966b: 448–486). See Inglehart (1984: 25–69) and Dalton, Flanagan, and Beck (1984: 3–22). Contrasting traditional with new cleavages on these grounds, this line of reasoning neglects the normative-ideological – and therefore cultural – attributes of traditional cleavages such as class. Indeed, by treating a cleavage such as class as primarily socioeconomic – or demographic – such a perspective neglects the fact that ideological or cultural factors lie at the very root of any group deﬁnition or sense of self-awareness. Eckstein (1966) distinguishes ‘‘segmental cleavages’’ from ‘‘cultural disagreement’’ and ‘‘speciﬁc disagreement.’’ Although in this case the term ‘‘cleavage’’ refers to the link between social structure and political order, the distinction raises problems similar to those of ‘‘cultural cleavages’’ in that the normative element embodied in the second category is in fact also typical of the ﬁrst. The deﬁnition of ‘‘structural cleavages’’ by Allardt and Pesonen (1967: 325–366) is restricted to differentiated social groups that are also characterized by cohesion and solidarity. As such, in addition to the social attributes (differentiation) on which they focus, they also implicitly introduce organizational (cohesion) and cultural (solidarity) attributes, a combination that strikes me as offering the most fruitful basis on which to build a deﬁnition. See E. Allardt (1968: 66–74). For this three-dimensional characteristic of theoretical concepts, see the treatment of the

CLASS CLEAVAGE: FRAMEWORK

17

empirical referent of the concept and which we can deﬁne in sociostructural terms; a normative element, that is, the set of values and beliefs that provides a sense of identity and role to the empirical element and reﬂects the self-awareness of the social group(s) involved; and an organizational/ behavioral element, that is, the set of individual interactions, institutions, and organizations, such as political parties, that develop, as part of the cleavage.17 The term ‘‘cleavage’’ should, I suggest, be restricted to the indication of a dividing line in a polity that refers to and combines all three aspects, and alternative terms should be adopted when referring to objective social distinctions or to ideological, political, and organizational divisions per se. Because the sociostructural, normative/cultural, and political/organizational elements play an inextricable role of mutual reinforcement in shaping individual attitudes and behavior, cleavages can be considered as only one particular kind of division rather than as a concept that exhausts the realm of all possible divisions. For example, it is clear that important lines of social stratiﬁcation may exist that cannot be identiﬁed as cleavages; similarly, there may exist political and ideological divisions that also, however important, cannot be identiﬁed as cleavages. Differences in occupation qua occupation, or differences in language qua language, do not produce the respective cleavage, but rather the nature and intensity of the emotions and reactions that can accompany membership in these occupational or linguistic groups, as well as the social and political bonds that organizationally unite the individuals who belong to them. Once achieved, these positions become ﬁrmly established, and it is the endurance of this entrenched position in group terms that produces the image of a

17

concept of class by Aron (1964: chapter 3) and the concept of nation by Lepsius (1985: 43–64). Katznelson (1986; 14–22) has reviewed the problem of the deﬁnition of class and class formation, suggesting that class formation is identiﬁed by four levels: (1) the structure of capitalist economic development, i.e., class as a constitutive element of every capitalist structure and capitalist transformation, a concept distant from empirical reality; (2) social organization of society as lived by actual people – workplace social relationships, etc., i.e., class-based ways of life; (3) class as formed groups sharing ‘‘dispositions’’; and (4) collective action. I do not see clear advantages with respect to the structural, normative and behavioral components of the classic distinction between class position, class consciousness, and class action. These three dimensions of a cleavage should not be confused with the three ‘‘types’’ of cleavage mentioned by Rae and Taylor: ‘‘ascriptive’’ cleavages or ‘‘trait’’ cleavages such as race or caste; ‘‘attitudinal’’ or ‘‘opinion’’ cleavages such as ideology or preference; and ‘‘behavioral’’ or ‘‘act’’ cleavages such as those elicited through voting and/or organizational membership. These authors regard them as mutually exclusive classes of different cleavages rather than as different constitutive aspects of every cleavage. See Rae and Taylor (1970: 1–3).

18

THE POLITICAL MOBILIZATION OF THE LEFT

cleavage becoming stable over and beyond the individuals involved, creating a speciﬁc cultural background and a varying propensity to collective action.18 From the point of view of its consequences, a cleavage has to be considered primarily as a form of closure of social relationships.19 The concept of cleavage is therefore clearly at quite a remove from any deﬁnition of the sociostructural base that provides its reference point, and this approach clariﬁes the deﬁnition of the boundaries and the differences between the two. More concretely, what distinguishes class from the class-based cleavage; religiosity or the community of religious people from the religiousbased cleavage; the ethnic group from the ethnolinguistic cleavage? Three key differences are evident, albeit not always to the same degree in each type of cleavage. First, the sociostructural reference of a cleavage and the cleavage itself are products of different historical phenomena: The former emerges from the processes of state and nation formation and from the development of capitalism and industrialization; the latter emerges by the coupling of these processes with those of politicization, electoral mobilization, and democratization. The social basis of the class cleavage originates in the social stratiﬁcation produced by industrialization and capitalist development, and these processes establish the structural conditions for group distinctiveness;20 but it is also clear that the class cleavage derives its special character only in relation to the other, more strictly political processes. A similar argument holds true for religious-based cleavage, whose structural basis is set by the process of the breaking down of Christian European unity in the sixteenth century and the later process of state centralization and secularization. In the same vein, ethnolinguistic groups emerged from the long process of linguistic differentiation, migration, and state boundary creation in European history, but speciﬁc cleavages of an ethnolinguistic nature develop only in response to the modern nation builders’ attempts to effect cultural and linguistic standardization and when the opportunities to express dissent and to organize opposition become available. It is only with the development of the modern nation-state and with 18 19 20

See Schumpeter (1951). Cf. M. Weber (1978: 43–47). The best conceptual reference for the result of such a process is indeed the Weberian idea of ‘‘social class’’ as a group of class situations that are made homogeneous by the existence of a ‘‘common possibility of mobility either within the career of the individuals or in the following of successive generation’’; cf. Weber (1978: 302–305) and Giddens (1973: 47–48).

CLASS CLEAVAGE: FRAMEWORK

19

the integration of different groups into the central sphere of society that the conﬂicts between these groups have become centralized. State building brings about a consolidation of military, administrative, and economic boundaries that reduces the possibility of exit for individuals and groups. Nation building and attempts at national cultural standardization bring about, in varying degrees, an ethnically and culturally homogeneous context that sets the stage for the articulation of emerging functionaleconomic differentiation and conﬂicts. It is this process – one of ‘‘unifying potentially opposing camps, facilitating their society-wide organization, their becoming symbols of social and political identiﬁcation and their making demands on central political institutions’’21 – that gives rise to the enduring relations between speciﬁc social groups, organizational networks, and ideologies. It is, ﬁnally, through the historical process of mobilization, politicization, and democratization that voice options are articulated and organized within a bounded territory and any speciﬁc cleavage acquires its distinctive normative proﬁle and organizational network. In short, cleavages initially develop on the basis of a social stratiﬁcation that sets the structural conditions for group identity; only later do they become fully political, particularly with the development of mass democracy. If a cleavage is regarded as a conﬂict line or a division line translated into politics, the translation is what historically constitutes the linkage between social condition, consciousness, and action. Conﬂicts and oppositions may not be translated into politics – either repressed, depoliticized, or deﬂected versus other divisions or channeled into politics in various ways; they may even be generated by politics, activated and reinforced by political processes and institutions. Translation points to the crucial role of the translators and of the mechanisms and conditions of translation, and it implies that there is variance in the capacity to translate the basic preconditions. Therefore, the nature and magnitude of the sociostructural basis of the conﬂict must be viewed as basic conditions facilitating – to a greater or lesser degree – the translation. The transition from the early period of politicization to the establishment of a hierarchy of superimposed cleavages, as well as to the creation of close links between cleavage systems and party systems, is complex. The ‘‘rationalistic’’ bias of Marxist class theory was based on a long chain of reasoning. Disciplined by the same conditions and circumstances of their work, brought together in a great multitude where they experienced the community of their status and deprivation, the workers would develop a collective class consciousness. In this phase, Marx expressed and trans21

Eisenstadt (1966: 22).

20

THE POLITICAL MOBILIZATION OF THE LEFT

formed into a general philosophy of history the grievances and claims of a segment of the politically aware Western European workers, at that time only a very modest proportion of the proletariat. These grievances were then expressed in the form of the class struggle, that is, as total opposition to the existing system. In this second step, using as a base the psychology of these workers in the transition period, Marx theorized their rejection of capitalism but not of industrialism. Workers would have to realize that, although their chains were those of the new industrial system, such a system would have a benevolent long-term effect for them once its capitalist structure was destroyed. Finally, the workers should not be distracted from this realization by other nationalistic, religious, ethnic, economic, or other considerations. Only then would social grievances transform themselves into class action. However, at each of these critical junctures, alternatives were possible, and other actors were interested in making these available and credible. From this point of view, class or class conﬂict has no special or dominant guaranteed role in politics. Class conﬂict has dominated the scene of sociology, and its almost mechanical reverberation in politics has led to the idea that class political alignments are the modern and normal form. By the same token, other forms of political division are seen as somehow artiﬁcially superimposed on the real and important divisions in the economic functional domain. Sometimes this belief is accompanied by the idea that this is a deliberate attempt by dominant economic circles or classes to conceal the base of domination; sometimes by the idea of false consciousness, that is, the limited capacity of workers and lower classes to see their real interests. This perspective (1) underestimates the ideological nature of the class cleavage itself, (2) exaggerates the simplicity of the political translation of class conditions and identities in class political action, and (3) indulgently disregards the late and often residual nature of such a dividing line with respect to others. The tendency is thus often to see modern mass politics in the light of the class cleavage. Second, because the relations of individuals to the social basis of a cleavage are deﬁned by certain attributes that can change with relative ease, this social basis represents a grouping and a set of social relationships that are normally much more ﬂuid than those constituted by the cleavage itself. The cleavage, on the contrary, thanks to its behavioral and organizational dimension, is a social relationship that implies a level of external closure that is always more pronounced than that of the social group. In this sense, it should be noted that there is an important difference between class cleavages, on the one hand, and other cleavages – which may develop on

CLASS CLEAVAGE: FRAMEWORK

21

the basis of ethnolinguistic or peripheral communities or even religious identity – on the other. Class conditions are a social stigmata that can be modiﬁed by individual or group mobility or by emigration. Ethnic or religious identities are based on characteristics that lead more easily to a wholly closed relationship. This difference, in turn, helps to explain why these kinds of cleavages demonstrate such an impressive capacity to survive over time and to encapsulate their respective communities. Even in the case of social class, it would be a mistake to underestimate the extent to which members of class organizations – both parties and unions – have sought to improve their position through a monopolization of this social relationship and through its closure toward the exterior world. As an illustration of this, one need only recall the debates within the Second International on the ‘‘peasant question’’: the resistance of the skilled working class to the incorporation into the movement of unskilled workers or even the more recent problems experienced by immigrant workers. The process of political translation was even more complex for class than for other traditional identities. This is because class had left parties to mobilize their constituency in opposition to the individuals’ traditional ties of a local, ethnic, and/or religious nature. In every case, they had to induce them to act in a way that was in most cases contrary to and in contradiction with the norms and roles dominant in their geosocial milieu and, in general, against the authority of the consolidated social hierarchy at the local level. The competition among forms of representation based on territorial representation, cultural defense, or functional-economic interest does not easily ﬂow to the advantage of the latter. Territorial forms of representation impose limits on the capacities and possibilities of party conﬂicts within the localities and tend to reduce or transform politics into a question of external representation, as the whole community is represented to the external world. By contrast, the functional/economic emphasis implies and reﬂects a type of alliance that crosses local geographical units and undermines the established leadership structure within the community, introducing into it elements of direct-interest conﬂict. Moreover, territorial and cultural defense need not be linked together, as they tend to be in Rokkan’s work. There might be prevalent forms of territorial representation without any cultural defense – as examples from France and the United States in the nineteenth and early twentieth centuries show – as well as a cultural-defense mobilization without a strong territorial basis, which cuts across territory and nationalizes politics, and a cultural defense guaranteed through territorial defense – as the Swiss Catholic cantons’ experience exempliﬁes. Thus, in certain circumstances, functional-class in-

22

THE POLITICAL MOBILIZATION OF THE LEFT

terests may be trapped in an uneasy environment in which both cultural and territorial defenses coalesce to make it difﬁcult to appeal to functional identities across territories and across cultural traits. This problem may be less important for mobilization efforts along other lines, such as for instance religion. Here, there may be less emphasis on abstract ideals and remote collective groups against the local power structure; instead, the emphasis is on well-established identities and on authorities that offer a sense of protection against external authorities. In this sense, the task of mobilizing, already difﬁcult for working-class parties when they had to cut individuals away from these traditional ties, became even more difﬁcult when and where these traditional ties were already mobilized and were the source of very speciﬁc political identities. Third, the social basis of a cleavage is essentially unorganized. In the case of the class cleavage, for example, the workers’ parties, the trade unions, and other agencies of this kind are not the organizations of the social class per se, but are the institutional components of the cleavage. In other words, it can be argued that the institutional nature of the class cleavage, in terms of both its social membership and its organizational form, is historically and country specific precisely because it does not depend exclusively on social class. Of course, the model of interaction between the class cleavage and other cleavages is also largely country speciﬁc, and it is the sheer complexity of the manner in which the various cleavages become interlinked and superimposed on one another that makes it particularly difﬁcult to disentangle the process leading from the development of structural prerequisites to the ideological and organizational patterning of the politicization phase. Moreover, once cleavages become established and organizationally institutionalized, they develop their own autonomous strength and, in turn, act as an inﬂuence on social, cultural, and political life. Thus, not only do cleavages become more stabilized than social classes or groups, but they are themselves a means of political stabilization, providing individuals with a constellation of preexisting alternatives for their own social and political integration. Parties, and in general the organizations of the cleavage line, also have the job of creating opportunities and ‘‘spaces’’ in which the feeling of emotional belonging among the members can be created and reinforced. This is more or less necessary, depending on the conditions of environmental hostility that surround the parties. Such spaces become more necessary the more the party has to face severe forms of repression and opposition. Thus, the building of a large, isolated subcultural network of ancillary organizations has often primarily served the function of members’ socialization and defense in periods of environmental hostility. Moreover, party

CLASS CLEAVAGE: FRAMEWORK

23

subcultural organizational networks reinforce emotional solidarity as well as socializing new members. The way in which such subcultural networks are established in several areas of social life depends on the speciﬁc nature of the cleavage and, in particular, on its social basis. Emotional feelings of solidarity are probably more easily shaped and organized by peripheral and ethnolinguistic minorities or by religious groups that are sharply opposed to dominant groups or are simply discriminated against and/or repressed. The homogeneity of the working class, in terms of division of labor, housing conditions, educational levels, workplace, and so on, creates those areas of equality that facilitate the establishment of such subcultural milieus and the development of strong emotional ties within the group. By contrast, such changes as the increasing internal differentiation of labor, workplace heterogeneity, disappearance of speciﬁc industrial communities, increasing spatial mobility, and the separation of workplace and residence, as well as growing dependence on impersonal contacts with the party organization and the reduction of daily personal contacts with other members of the group (e.g., through increasing resort to private means of transport), create conditions that weaken the emotional ties of a group. At the same time, however, these subcultural networks are a result not only of favorable social conditions, but also of deliberate efforts by parties to create them and to ﬁght against unfavorable conditions. Once in place, these networks can help to serve the goal of maintaining and even reinforcing group emotional ties with great effectiveness even in periods of rapid change in the earlier structural conditions of similarity. However difﬁcult the comparative study of these processes might be, it is here that we ﬁnd the key to understanding the development of different cleavage systems characterized by varying levels of organizational fragmentation and social homogeneity across European countries. For example, it is impossible to grasp the difference between the fragmented and socially heterogeneous social basis of the class cleavage in the Southern European countries, on the one hand, and the cohesive and relatively socially homogeneous basis of the class cleavage in the Scandinavian North, on the other, other than by reference to the interplay between the structural, ideological, and organizational aspects of the class cleavage itself and between the class cleavage and other cleavages. It follows from this that cleavages have histories of their own that differ substantially from the histories of their social bases and from the histories of their original mobilization and politicization. Such histories, and the strength and hold of traditional or emerging cleavages, can be assessed only in terms of all three constitutive elements, that is, changes in social stratiﬁcation, changes in the corresponding cultural systems, and changes in sociopolitical orga-

24

THE POLITICAL MOBILIZATION OF THE LEFT

nizational forms (not only political parties, but also the networks of social, professional, and other organizations expressive of the cleavage). It also follows that although we deﬁne the class cleavage in relation to its genetic origin, this expression does not indicate that the workers, all the workers, or only the workers represent its social constituency. The social membership of the class cleavages may vary considerably over time and across countries. Any attempt to translate this line of reasoning into an empirical inquiry regarding the modalities of class cleavage structuring necessitates a major attempt to devise indicators and collect data on social structure, attitudes, and organizations. Such an inquiry also needs to be crossnationally comparative and to have a long-term historical dimension. It is not surprising that it has rarely been attempted at a comparative level. Rather, most research22 usually analyzes and privileges a single dimension of cleavage structuring (or destructuring) at the national or cross-national level. Some analysts concentrate mainly on the relationship between the social stratiﬁcation of European societies and patterns of voting. Through survey data, attempts are made to ascertain the closeness of the relationship between the two, measuring the extent to which social group membership proves an effective predictor of partisan choice and, more broadly, investigating the social, cultural, and psychological characteristics of the leftwing voter.23 These studies invariably concentrate on the social homogeneity of the class cleavage membership. Another research tradition focuses primarily on the changing attitudes and beliefs of the mass public, arguing that the parties of the traditional left face an increasingly complex pattern of competition along a number of other issue dimensions and that a wide selection of new concerns – ranging from neolocalism to environmentalism to welfare problems and law and order – impact on different sections of the population in different ways and thus tend to cut across moretraditional ties based on occupation, income, and status and to undermine the cultural distinctiveness of any class alignment.24 Finally, a separate tradition challenges the view that the political parties of the left reflect the identity, interests, and consciousness of the social group. Assuming that 22

23

24

I am not concerned in this context with that tradition of analysis that treats cleavages as an independent variable to account for systemic properties such as stability and democratic performance. The starting point of this literature is Simmel (1956). For other key analyses see Eckstein (1966), Nordlingler (1972), and Lijphart (1968: 3–44). The literature in this area is too large to be quoted in full: Representative of this line of interpretation are Alford (1963); Rawson (1969), Abramson (1971), Goldthorpe, Lockwood, Bechhofer, and Platt (1971: 11–13), and Lipset (1981). See the sources on cultural cleavage cited in footnote 13. For a good synthesis see Hildebrandt and Dalton (1978: 69–96).

CLASS CLEAVAGE: FRAMEWORK

25

the role of political organization is essential to the creation of the subjective class, this perspective stresses that attention should be focused on the fate of the major left political organizations – the trade unions and other agencies and organizations that help to create, reinforce, and transmit group attitudes and attachments. The fate of the class left is thus linked to its capacity to adapt to the new conditions in which it is obliged to reproduce its support and ideology.25 These different research traditions tend to give priority to just one of the three constitutive aspects of any cleavage: the reference social group and its possible modiﬁcations, the political attitudes and ideological structuring of the group consciousness, and the behavioral element giving rise to the organizational network on which the parties base their strength and of which they are themselves a part. To privilege one dimension or perspective may lead to speculation about the cleavage development that is quite different from that which can be advanced by privileging an alternative dimension. For this reason, in this work I consider them all as constitutive elements of the cleavage structuring. This latter – my dependent variable – is operationalized through different properties, and the cross-country variation of these dimensions is interpreted as resulting from a macroconstellation of systemic factors. The following section outlines the framework as a whole.

DEPENDENT VARIABLES: THE PROPERTY SPACE OF EUROPEAN LEFT VARIATION Summarizing this discussion, Figure 1.1 graphically presents the structuring of a cleavage in terms of the three dimensions of social constituency, organizational network, and cultural distinctiveness. Each of the three dimensions changes empirically from cleavage to cleavage, from country to country, and from historical period to historical period. While the cleavage dimensions of social constituency and organizational network have been largely dealt with by the literature, that of cultural ‘‘distinctiveness’’ is more complex. By this term I do not mean the process of ‘‘objectivization’’ of class whereby a third element is introduced between the class as an objective social condition and the class as an expression of observable behavioral patterns of organizational membership and action; nor do I 25

Kitschelt (1994) stresses this point in his interpretive line. In general, the emphasis on the capacity for adaptive change of the established parties is a dominant theme of the work of P. Mair, whose main contributions in this direction are now collected in Mair (1997).

26

THE POLITICAL MOBILIZATION OF THE LEFT

Figure 1.1. Dimensions of cleavage structuring. mean the ‘‘class interest,’’ as this is deﬁned both independently of what individual workers declare it to be and as the political elites and organizational leadership express it. Cultural distinctiveness instead indicates that a certain number of individuals, without any external constraints, react in a similar way to similar situations and absorb des obligations sociales et des coutumes, a spontaneous kind of moral behavior within a given group. While the social constituency of a cleavage is ascertained in terms of social homogeneity – that is, a similarity of social conditions – cultural distinctiveness pertains more to the dimension of ‘‘community,’’ of a similarity of values and obligations and a sense of belonging. It is difﬁcult to separate this aspect clearly from the communality of the social condition and from the speciﬁc pattern of organizational membership and behavior. It refers to something that is similar to what Aron called the ‘‘objectivization’’ of communities when he referred to features that Weber called ‘‘ethos,’’ and to the equivalent concept of ‘‘habitus’’ in Bourdieu.26 Given that variations in outcomes are used as a key to the comparative search for causes, it is essential to start with an accurate deﬁnition of any such variation. The scheme in Figure 1.1 helps to organize the dependent 26

Aron (1989: 462–471).

CLASS CLEAVAGE: FRAMEWORK

27

variables in a synthetic form. The social constituency dimension points to two aspects of cleavage structuring. (1) The first is the level of electoral mobilization, that is, the electoral strength of the different left political organizations, meaning the extent to which they were capable of mobilizing support and obtaining the electoral loyalty of sections of the population. From this point of view, there are strong and weak lefts in electoral terms. At the same time, the structuring of the class cleavage resulted in (2) different social compositions of its electoral constituency. From this point of view, class cleavages can be ranked in terms of the extent to which they managed to mobilize the putative social target as well as other social groups. Class cleavages will therefore be characterized by a higher or lower level of social homogeneity of support. The second important dimension of variation concerns the organizational network, which can also be seen from two distinct perspectives. The ﬁrst perspective concerns organizational cohesion, that is, to what extent in their organizational development national lefts were capable of maintaining organizational unity or, alternatively, were faced with processes of internal organizational fragmentation. In this perspective, the European experience sees a cohesive and a divided left. The second perspective concerns the degree of organizational density of the class cleavage, that is, the extent to which such a cleavage rested on a densely organized network of corporate, political, and cultural associations or, at the opposite extreme, consisted mainly of ideological and attitudinal opposition deprived of strong organizational vertebration. The issue is thus the extent to which class cleavages were organizationally encapsulated. The dimension of cleavage cultural distinctiveness also includes two subdimensions. At the level of the political elites and participant activists, the most important element concerns the predominant ideological orientation of the movement. The distinction in this case involves different types of class-left ideologies (orthodox Marxism, syndicalism, communism, etc.) and the extent to which they are moderate or radical. Clearly, here I am not interested in the ideological or theoretical schools in themselves, but rather in their spreading and appearing as convincing answers to the problems of the people. The causes and the consequences of the spread of different ideologies are sociological and political questions that have little to do with their exegesis and even less with their success or failure in terms of the accuracy of their historical predictions. At the level of the masses, the important element is the level of cultural solidarity that the cleavage structuring is able to create and reproduce within a social constituency. Different class cleavages rest on different levels of class solidarity, on different types of working-class culture, and on the different intensity of feelings of

28

THE POLITICAL MOBILIZATION OF THE LEFT

belonging. However, despite its importance, this aspect of cleavage cultural distinctiveness will not be included in this study because I have been unable to conceptualize it, let alone measure it, in any way. Although several historically detailed studies exist at the national level, it is extremely difﬁcult to frame this dimension in a way that allows for grounded comparative statements to be made.27 Rather than dealing with this aspect unsystematically, I have preferred to omit it from my research. Only indirect indications will therefore be derived on this topic from other dimensions of cleavage structuring. To sum up, the three dimensions of cleavage structuring are translated into six variables (Figure 1.1), only ﬁve of which will be directly dealt with in this study: electoral strength, social homogeneity, organizational cohesion and density, and ideological orientation The study will attempt to explain to what extent and why the class cleavage is electorally strong, socially homogeneous, organizationally cohesive and dense, and ideologically moderate or radical in different historical experiences. Not only is there cross-country variation, but there is also cross-time variation in each of these properties. That is to say, outcomes are different according to the moment in time at which the study is carried out. On the eve of World War I, for example, the picture would reveal a strong and uniﬁed German socialist movement, while in the United Kingdom the working class was organized through unions that were mostly concerned with the marketplace and voted for the Liberals – what we would call today a ‘‘catchall party’’. Thus, the homeland of industrial capitalism had no socialism, while at the same time, ‘‘backward’’ Finland had a strong radical social democratic party with 37% of the vote.28 Few observers of the time could have guessed accurately where the future of socialist ideology and organization lay. In the early 1920s an evaluation of left radicalism would have pointed to the Norwegian Labor Party as having the most extreme position, given its almost unanimous adhesion to the Third International. Yet, the Norwegian radical takeover was short-lived, and the ensuing history of the party can be more appropriately described as a trend toward consensus.29 In other words, there is a problem with interpreting 27

28

29

Theoretical discussions of this aspect are not as rare as attempts at comparative evaluation of it. See, for instance, Mann (1973). Sombart (1905) pointed to the paradox that the United States, the country he regarded as having the most advanced industrial capitalism and representing the direction of future European development, did not have a socialist working class. The question, then, was whether the United States represented an ‘‘exceptional’’ development or just the future European development, and whether the rise of socialism was a necessary and inevitable corollary of the development of capitalism. See Torgersen (1962).

CLASS CLEAVAGE: FRAMEWORK

29

labor’s responses in the short term as against the more permanent longterm features of national labor movements. This is also true when attributing the relative weight to the ‘‘states of dynamism’’ – characterized by a great ﬂexibility of organizational forms and a capacity for qualitative absorption of new groups and strata previously not forming part of the movement. These states of dynamism are typically marked by attitudes and behaviors of a nonconformist nature, as opposed to those of ‘‘stabilization’’ – which are instead characterized by organizational rigidity and by an absorption that is mainly quantitative, that is, by increases in the organization and representation of strata that were already part of the movement.30 This study is oriented to the interpretation of the more longterm features of the socialist movements. However, these considerations will necessitate a certain amount of contextualization at different points in time if we want to avoid the risk of judging the entire left experience on the basis of its ﬁnal post–World War II outcome. A certain amount of periodization will therefore be needed. Let me mention a ﬁnal insurmountable problem of interpretation of the labor response: Where did the latter originate, and of whom was it the response? Did it emerge from elite groups and social vanguards? From organizational bureaucracies? From the masses and the rank and ﬁle? In a Leninist perspective, the political response of national labor movements originated in vanguards, that is, in politically aware groups leading the rank and ﬁle. This means that labor’s response is the result of a struggle for inﬂuence over the whole movement fought by minority groups. From a Michelsean perspective, socialist leaders and functionaries tended to assume a lifestyle and values close to those of the bourgeoisie. Their identiﬁcation with the labor movement organizations, their defense of the organizations’ priorities, and their natural conviction that organizational interests and imperatives were by deﬁnition the best embodiment of the class interest in the long run dictated the attitude of the socialist movements as a whole. Thus, progressive socialist integration was neither a conspiracy nor a takeover of the interests of certain sectors of the well-off working class; it was, rather, a natural by-product of the oligarchic tendencies typical of all bureaucratic organizations and of the bureaucratic elites’ interest in the status quo.31 Finally, a third line of argument suggests that the long-term response of the labor movement was the result of a progressive adaptation of the elites to the basic inclinations of the rank and ﬁle. The pre–World War I socialist movement was more radical and ideological 30 31

Bauman (1972: 324–325). Cf. Michels (1949).

30

THE POLITICAL MOBILIZATION OF THE LEFT

than the workers themselves. It was directed by ideologues and intellectuals whose sophistication and ideological rigor pushed the ideological stances far beyond the basic claims and interests of the rank-and-ﬁle workers. Thus, the post–World War I developments can be seen as the result of the progressive adaptation of such elites to the more moderate and pragmatic approach of the rank-and-ﬁle workers, as radical appeals proved unsuccessful almost everywhere.32 In this case, the response of labor is, in the long run, attributed to the basic orientation and preference of the movement’s rank and ﬁle, reducing the role of elites and vanguard groups to the early phase of development and leaving little room for organizational imperatives and ends substitution. These three theories are important. According to whichever one is adopted, the role of elites, intellectuals, vanguards, and rank and ﬁle is enhanced or reduced. To a certain extent, the choice of one approach implicitly predisposes the study toward one sort of interpretation of the sources of the long-term response of labor. This, however, does not relieve the scholar of the task of performing a more speciﬁc study of the conditions under which the forces evoked were comprehensive and successful.

FRAMEWORK OF INDEPENDENT VARIABLES AND STRUCTURE OF THE BOOK In the theoretical discussion on cleavage structuring so far, I have discussed a number of theoretical points concerning the process of translation of social oppositions into cleavages through the mobilization efforts of political entrepreneurs. These were framed within the broader processes of state and nation formation and internal democratization that contributed to the speciﬁc national interplay of different conﬂicts and oppositions. It is now necessary to brieﬂy summarize them into a grid of independent explanatory factors; this will also serve as a guide to the structure of the entire book. In Table 1.1, these factors are subdivided into preconditions, sociostructural trends, organizational resources, and politicoinstitutional opportunities. The table starts from the more remote macroconstraining conditions and approaches the conditions of individual, collective, and organizational choices that eventually determine the differentiation of outcomes. We can imagine preconditions and social inputs as being those general conditions that affect the capacity of a given movement to mobilise – ‘‘the basic variables affecting the organizational capacity of a population.’’33 The 32 33

For this position, see the conclusions in Mitchell and Stearns (1971). Stinchome (1965: 145).

CLASS CLEAVAGE: FRAMEWORK

31

Table 1.1. Mobilization of the class left: ordering of influencing variables

difference between the two is that while the preconditions pertaining to the international status of the polity, its degree of boundary consolidation and internal centralization, and its internal cultural homogeneity are constant for any given polity, sociostructural trends in industrialization, urbanization, working-class constituency formation, education, and community formation have a distinct temporal variation within each polity. Organizational features and political and institutional opportunities are given a causal role that is closer to the outcomes within a context of given preconditions and sociostructural inputs. They are not exogenous variables but can also be related to preconditions and social inputs. They are also associated with each other, as, for instance, when organizational resources depend considerably on the institutional opportunity structure of the polity. Similarly, the opportunity for political alliance depends on organizational consolidation, as well as on the openness or closure of the institutional structures. The variables listed under the label of ‘‘political opportunity structure’’ include features that clearly lie outside the sphere of choice of the socialist movements – such as the nature of the incumbent dominant alliance or the preexisting traditions of popular movements – as well as features more open to choice opportunity – such as the possibilities of social and political alliances. The last column of Table 1-1 lists the differentiation of outcomes that were presented as independent variables in the previous section. At this stage, it is premature to discuss details about causal linkages.

32

THE POLITICAL MOBILIZATION OF THE LEFT

In the various sections devoted to each set of features, I introduce and discuss the theories that link them to the mobilization process, and I try to test them as accurately as possible. It is evident, however, that links established between, say, social inputs and the dependent variable also affect intermediate-level variables. In other words, the inﬂuence of industrialization levels and working-class constituency size on electoral mobilization can be seen as direct and also indirect, expressing itself through, for instance, trade union organization and recruitment. Moreover, complex interactions between the dependent variable of electoral mobilization and the independent variables of institutional and political opportunity structures are evident. The overcoming of an institutional barrier – for instance, regime liberalization or ﬁnal enfranchisement – can be seen as something that both affects and is affected by mobilization of the left. These special interactions between potential causes and effects will be dealt with carefully in the speciﬁc sections. The concluding part of this work links these processes together, forming a dynamic model that tries to assess the relative weight of each process. The guidelines for this work can be described here: The more general question concerning class cleavage mobilization is whether this should be regarded primarily as a political response rooted in forms of particularly acute social dislocation and economic alienation or if, by contrast, the main interpretive element should be linked to the idea of ‘‘political alienation.’’ Reformulating Huntington’s conceptualization of the linkage between social mobilization processes and levels and forms of political stability,34 we come to the following set of relationships: social mobilization economic development

⫽ social frustration

social frustration ⫻ organizational development

⫽ political mobilization

political mobilization institutional integration

⫽ political alienation/integration

This grid starts from social mobilization (industrialization, urbanization, education and literacy development, and socioprofessional transfor34

Huntington’s (1968: 54–55) original formulation was mainly concerned with the link between the social mobilization process and the levels and forms of political stability in modernizing countries. His original formulation of the relationship between the two was synthesized in the following formulas: social mobilization / economic development ⫽ social frustration; social frustration / mobility opportunities ⫽ political participation; political participation / political institutionalization ⫽ political instability.

CLASS CLEAVAGE: FRAMEWORK

33

mations) as processes generating rising expectations and demands. The level of satisfaction of these demands – according to certain theories – or the structural conditions of work – according to others – determined by given rates of socioeconomic development foster different levels of social frustration. In this perspective, the historical conditions of socioeconomic alienation were favorable to the birth of radical political protest and organized political action. This ﬁrst part of the scheme, however, says little about how socioeconomic alienation transforms itself into organized political action. The intervening variable inserted by Huntington to arrive at political participation was ‘‘mobility opportunity.’’ An individual’s willingness to participate is the result of a ratio between his or her social frustration and the opportunity for mobility. This replicates an individualistic and socially driven process of political mobilization, thereby underestimating the role of political organization in fostering participation (particularly the participation of citizens with few individual sociocultural resources).35 Moreover, in contrast to the United States, in Europe during the industrialization phase, the mobility opportunity for the lower classes does not seem to have been an important aspect militating against the deﬁnition and defense of collective interests. Therefore, political participation is seen in my scheme as the result of the combined effect of social frustration and organizational development, which – in the context of socialist history – often took the form of a rapid imitation of political forms of earlycomer by latecomer countries. In the continental context, the spread of organizational forms and techniques of action was relatively fast, faster than the spread of other underlying social and economic processes. Organizational and political leads and lags, therefore, do not correspond to socioeconomic leads and lags. Finally, Huntington’s third equation maintains its importance in this context. The relationship between the levels of participation and the establishment and institutional consolidation of the channels for its expression (institutional integration) gives rise to different levels of political alienation/integration as a result of lack of citizenship and feelings of political exclusion. Political institutionalization must be conceived here mainly in terms of the adaptability of traditional political institutions to the pressures from growing political participation. There are three positive aspects in this revision of Huntington’s scheme. First, it allows dynamic relationships to be established between the macroprocesses of modernization that have so far been considered only 35

See Verba, Nie, and Kim (1978) for their discussion of this role of political institutions.

34

THE POLITICAL MOBILIZATION OF THE LEFT

from a static point of view. Second, it reintroduces the crucial variable of organizational development. Third, it neatly distinguishes the sources of economic alienation from those of political alienation. The last point is particularly important. We can in fact oppose a socioeconomic to a political alienation theory as a source of working-class politics in the period between 1848 and 1920. If one maintains that the sources of class politicization and mobilization were mainly those of socioeconomic alienation, rooted in material conditions and the organization of production, then a set of hypotheses that requires a special emphasis to be given to the socioeconomic conditions of working-class formation and lower-class makeup is in order. If, by contrast, we regard political alienation as an essential feature of the emergence of a class conﬂict and cleavage, then we need to emphasize the response given to requests for full citizenship, political integration, and representation, that is, to the process of the civic (individual emancipation, freedom of association, etc.) and institutional integration of the lower classes (franchise, liberalization, policy inﬂuence, etc.).36 By emphasising the opposition between the socioeconomic and politicoinstitutional roots of alienation, one overcomes the implicit equivalence between the two that is not infrequent in the literature on the development of working-class politics. Moreover, this allows us to consider a trade-off between the two spheres; that is, there is the possibility that mutual compensations might occur between political and economic alienation, that higher levels of political alienation might be bearable in exchange for lower levels of social frustration (a Bismarckian strategy), or, alternatively, that higher levels of economic frustration and alienation might be accompanied by, and in part compensated for, higher levels of political representation (a liberal strategy).

36

That working-class politics is mainly a response to political alienation is a thesis that has been discussed by R. Bendix since the time of his early writings. The problem was formulated as follows: Nineteenth-century Europe’s problem concerned the way in which societies undergoing industrialization would ﬁnd a solution to the dilemma of incorporating the newly recruited industrial workforce into the economic and political community of the nation. Bendix’s early studies concerned ideologies of management that he felt to be important because they indicated the way each society shaped the answer to this problem of integration. The main reference is Bendix (1974: 436–450). A revision of this piece is found in Bendix (1984b: 70–90). Bendix’s thesis that the political movements of the nineteenth century aimed more at citizenship recognition than at economic revolution, and his criticism of Marx’s view on the point are in (1984a: 91– 107). The problem was also discussed in (1964: 89, 112–122).

CLASS CLEAVAGE: FRAMEWORK

35

METHODOLOGICAL CHOICES Four methodological issues have been of particular interest to me during the preparation of the research outlined earlier. The ﬁrst concerns the deﬁnition of what needs to be explained about the electoral history of the left. The second concerns the level of spatial variance dealt with in this study and, more precisely, whether a cross-country comparison is too aggregate. The third issue concerns how to deal with the temporal dimension of variation of almost one century and how to relate a comparative synchronic analysis to a comparative historical analysis, that is, how to combine spatial variance with temporal variance to improve control. The fourth and ﬁnal issue relates to the debate on whether an adequate explanatory attempt necessarily requires a microfoundation anchored to individual motivations, choices, and beliefs. This section delineates the methodological boundaries and ambitions of the research, as well as the methodological biases of the author. WHICH CASES AND WHAT PROBLEM? In the history of class cleavage structuring, what needs to be explained and which cases is it worthwhile to consider? These two questions are interlocked because the choice of cases inﬂuences what we can study, and what we want to study inﬂuences what cases we should consider. The cases I investigate are thirteen Western European countries for which a sufﬁciently long period of regular competitive politics and sufﬁcient historical data and studies are available. Spain, Portugal, Greece, Iceland, and Luxembourg are excluded, only because of the difﬁculty of obtaining reliable and systematic secondary-source historical data on many of the aspects included in this research. For Portugal and Spain, the very long interruption in liberal institutions and democratic elections after World War I is also an important reason for exclusion. However, no country has been excluded for considerations pertaining to size, importance, or peculiarity. The thirteen countries considered are very different in population size, economic strength, and international status. Each of them enters my analysis as a case with the same weight as all the others. This is important, particularly when I base comparative judgments on the yardstick of all-cases average or typical development. If the cases were different, such a comparative yardstick might have to be different. It is important to bear this in mind because, according to some scholars, the comparative study of domestic internal development should be conducted primarily across countries that are large enough and resource-

36

THE POLITICAL MOBILIZATION OF THE LEFT

ful enough to enjoy relatively autonomous domestic developments. Barrington Moore has offered this as a reason for his concentrating on major international powers in his study of the social roots of democracy and authoritarianism.37 I ﬁnd the reasons for concentrating on large countries unconvincing. Large countries and powers, thanks to their economic and military resources and cultural autonomy, may well show more ‘‘autonomous’’ domestic developments and successfully shelter internal developments from international pressures. However, their high international status may also be a burden that weighs heavily on domestic developments. It may thus entail international interests and/or responsibilities the perceptions of which alter internal political developments. Smaller countries, on the other hand, enjoying less international status, may be less able to avoid external pressures and inﬂuences and thus be more prone to institutional imitation. These aspects necessarily have an impact on their internal developments, making them generally less willing to engage in international or military initiatives. In other words, strains on internal domestic politics are not so clearly associated with size. Moreover, international status may change quite rapidly (e.g., the Hapsburg empire into Austria). In reality, in selecting only large powers, it is not the most important, signiﬁcant, or autonomous developments that are selected. More simply, these are cases for which one important dimension – that of international status and the strains and responsibilities associated with this status – is turned into a constant. This prevents variance occurring along this dimension, as well as the comparison of cases of high international power and responsibilities with those of a lesser international role. In other words, the international status of a country should not be a criterion for selecting cases, but instead a variable whose different value/status may have an impact. This point is of particular importance for a study of the class cleavage. Class divisions, class ideologies, and class movements were not only forms of internal voice whose nature was particularly divisive for the polity; they were also associated with international movements challenging consolidated state boundaries. It is therefore very likely that the international status and involvement of each country had a strong impact on how the political structuring of the class cleavage was faced and treated. What needs to be explained is never a neutral question. This is even more true when the issue concerns class politics and the left, on which so many different evaluations exist of what is a success or a failure, and so many analyses have accumulated over time, taking the lead from theories of how objective class positions should translate into politics. For this 37

See Moore (1966: xii–xiii).

CLASS CLEAVAGE: FRAMEWORK

37

reason, it is worthwhile to devote some time to what needs to be explained in the electoral history of socialism and to the implications of the choice of one explanandum rather than another. The recent inﬂuential works of Przeworski and Przeworski and Sprague38 about the electoral history of socialism and its predicaments is an excellent and stimulating starting point, as their goal is the exact opposite of the one I have set for myself. Przeworski and Sprague’s central question is the following: why has the left (for them the socialist parties) never or only occasionally reached the 50% threshold? Why has the left never obtained the vote of all the workers? Przeworski argues that ‘‘by invoking relative standard (i.e., wondering why the Swedish left is stronger than the Dutch, etc.) one often forgets that absolute standards exist as well: One proceeds as if the common fate did not require an explanation. And yet it is that which different histories have in common that illuminates the limits and historically inherited possibilities. Before asking why this party has been more successful than that one, we can not ignore the fact that none has been successful in terms of its own dreams and designs, that no one has brought to realization the very purpose of its foundation.’’39 The common feature, the general development that assimilates all such political movements, is said to be a common electoral failure with respect to the left’s original purposes and expectations. I disagree with this point of departure on both methodological and substantive grounds. Elsewhere, I have argued40 that if there is a development common to all units, it is difﬁcult to explain. In any case, the key to the interpretation of a general development or common fate is the variance across cases in the timing, intensity, and so on of this ultimate outcome. Looking at the European experience in its entirety – without excluding embarrassing cases – what strikes me is not that socialism has never reached 50% but rather the striking variance in its strength, between the average 10% of the Irish left and the 45% of the Austrian left. If the focus is on the notion of common failure, both the Irish and the Austrian left have failed, although to different degrees. If instead these two cases are considered in the context of the Europeanwide empirical variance, the ﬁrst is a dramatic historical failure, ranking on average 25% votes below the Europeanwide mean, while the second is an outstanding success, having around 10% more votes than such a European mean. If the within-group variance can be explained, then this is the key to interpreting the group mean. In other words, the 38 39 40

Przeworski (1985) and Przeworski and Sprague (1986). Przeworski (1985: 103–104); the emphasis is mine. Bartolini (1993: 158–165).

38

THE POLITICAL MOBILIZATION OF THE LEFT

explanation of the 50% threshold follows as a variant of the former explanation (concerning the appropriate comparative yardstick, see the later section ‘‘Temporal Variation’’). The consequences of the choice of the explanandum often extend to the kind of analysis that is carried out. To justify the relevance of the ‘‘why never above 50%’’ or ‘‘why never all the workers’’ questions, reference must be made to the aspirations, desires, and beliefs of the early socialist thinkers and leaders and therefore to an implicit theory about the relationship between class condition, size, and class action. It is because the people in a working-class condition were the majority in a given historical period (and can this really be said to be true?) or were expected to become so over time that the problem arises as to why this was not reﬂected in a similar amount of class consciousness and class action. The 50% question presupposes a model that takes as normal the translation of a class condition into class consciousness and action and investigates the intervening variables that make this normal translation more or less successful.41 The aspirations and beliefs of the protagonists may be a poor yardstick for the deﬁnition of historical problems. Why have left parties been unable to realize such aspirations and desires? One could pose the same question for any other political formation. Why did the liberals never gain the total support of the rising bourgeoisie, or the religious movements the total support of religious people, or nationalist movements that of the whole nation? Catholics, Protestants, and ethnic, liberal, conservative, and agrarian movements all shared the same hopes and aspirations to fully mobilize their potential electoral constituency, and there are good reasons for claiming that they failed. If it is argued that the socialists were prevented from reaching their goals by intervening forces, could not the same be said for all the other political forces? To treat the history of socialism from the point of view of its failure to mobilize the working-class electorate implies the assumption that mobilization along functional socioeconomic interests and identities is normal and that the others are not normal or are exceptional. That class is the dominant factor of political alignment implies the residual nature of other factors.42 If and when class is politically articulated 41

42

‘‘It is not necessary to explain at length workers’ vote in favour of communist and socialist parties. Such vote is normal and has its logic, good or bad, which it is not our job to evaluate’’; Dogan (1960: 25). Przeworsky and Sprague (1986: 28, 59, and 78) make this postulate very clear in their analyses: ‘‘The lines of conﬂict and the modes of individual behaviour that emerge if and when class is not the force guiding the behaviour of workers are beyond the scope of this analysis’’; it ‘‘decreases the salience of class as the basis of collective identiﬁcation. It leads, therefore, to the resurgence of other bases of collective identiﬁcation, whether they

CLASS CLEAVAGE: FRAMEWORK

39

and appealed to, then it is the force that guides electoral behavior; when it is not, then other forces play a role. Other collective identities, such as religion or language, (re-)emerge only when the class appeal is attenuated or muted. This way of approaching the problem means prejudging the crucial empirical question of whether class is mobilized before or after other lines of cleavage and which of these other lines are mobilized. It also means simplifying the whole process of competitive mobilization along different fronts and excluding those processes that actually produced the interaction between class and other lines of conﬂict and that explain when class becomes a strong source of political behavior. The problem cannot be solved from the perspective of the aspiration of one actor; it is the whole pattern of interaction of mobilization efforts by politicized groups and elites that needs to be considered. Implicit adherence to a model according to which the true or primary identity is class has a further implication. If class is a normal or natural basis of mobilization, then it will mobilize if somebody cares to mobilize it qua class. If people vote according to their interests if somebody cares to articulate those interests and appeal to them, then the question becomes one of when and why the elites did or did not articulate and appeal to a speciﬁc class interest that exists as a reality.43 Necessarily, the variance in the extent of class mobilization ends up being mainly due to the subjective efforts of the mobilizers. Little room is left for factors that are external to the strategy of the left parties. The explanation of a fate that is deﬁned as common (failure to become majority forces) cannot be found in the external nonhomogeneous environments (institutions, competing forces, etc.). It must be identiﬁed within the movement itself, in its common and universal strategic predicament. Przeworski and Sprague’s argument is a modern version of a classic tradition that, from a different – historicist – perspec-

43

were based on the size of revenues, character of work, religion, language, or race’’; ‘‘workers becomes Catholics.’’ All emphases are mine. The keystone of every construction based on the naturality of class interests is the belief in the existence of such class interest that is neither the sum of the individual preferences of the workers nor the preferences expressed by their political organizations. If the class interest is identiﬁed with the positions articulated by working-class political organizations, then it would be impossible to judge the extent to which such organizations were actually articulating such class interest. By deﬁnition, they would do so. If class interest is identiﬁed with the empirical preferences of the people in the objective class position, then it would be equally impossible to argue that political organizations do not appeal and articulate such sets of diversiﬁed preferences. It is necessary to build the concept of an abstract class interest if one wants to argue about the extent to which political organizations actually appeal to the class qua class. The discussion of this Hegelian class interest is not pursued in this context.

40

THE POLITICAL MOBILIZATION OF THE LEFT

tive, attributes fundamental importance to the role and the choices of elites, that is, to the ‘‘historical adequacy’’ of such choices. In this study, I do not place any particular importance on the 50% threshold; I do not assume class to be a natural or normal or predominant basis for political mobilization; and I am unable to establish what electoral failure was other than in relative terms. Concentrating on cross-country differences redeﬁnes what comparative failure and success is. If one allows that the socialists were not alone in the electoral arena and that competing actors and forces existed, then the 50% threshold loses any signiﬁcance. Class can be seen as a residual function of other, preceding or parallel, political identities, rather than as a normal source of mobilization, following Finer’s acute observation that ‘‘class was important in Britain because nothing else was.’’44 Without a priori acceptance of any of these positions, I see this work as an investigation of how difﬁcult, complex, and in many ways nonnormal it was to translate objective class positions into a basis for political mobilization against traditional forms of territorial representation, established cultural commitments, the authorities of traditional elites, individual and organizational costs, and other competing mobilizing agencies and actors. The extent to which this translation was successful and which groups were more receptive to socialist appeals is a matter of empirical investigation. I ﬁrmly believe that the mobilization outcomes resulted mainly, but by no means only, from the efforts of class parties. Other actors were involved, and their choices were equally if not more important for the ﬁnal outcome. I also believe that the history of political and organizational struggles determines what people’s interests are, but it can by no means be claimed that such political and organizational struggles were only class struggles; other struggles could also precede, interfere with, or redeﬁne the political signiﬁcance of class. If one argues that people in the same class condition should in principle have voted left, not because their interests pushed them to do so but only because somebody had convinced them that it would be in their interest to do so, then the possibility that the speciﬁc history of political and organizational struggles convinced them that this was not in their interest must also be introduced. Any dilemma or electoral predicament disappears in this case. What led a Belgian, Catholic, Flemish worker to decide which of these three social roles and circles should be predominant in the deﬁnition of his political identity is a most intriguing historical puzzle. It cannot, however, be solved via assumptions about the 44

Finer (1970: 142).

CLASS CLEAVAGE: FRAMEWORK

41

normal political translation and about when nonnormal translations come into play. To consider actors’ choices in the reconstruction of historical outcomes becomes extremely difﬁcult indeed when one realizes that the actors involved are many and varied. To reconstruct the information and preferences of the actors, the viable alternatives, the causal structure of the situation of choice when considering socialist leaders, liberal elites, Catholic hierarchies, dynastic circles, and so on (and for each of them, who counts: certain leaders? some hundreds of active militants? some thousands of electors?) yields a very complex matrix. It is understandable that one actor only is chosen and is separated from the others. This has, however, implications and costs. As indicated in the ﬁrst part of this chapter, the emphasis in this research is placed mainly on the attribution of macroelectoral outcomes to equally macroeconomic, cultural, political, and organizational factors. This tends to subsume the factors of choice, and the consequences of actors’ choices, within the framework of constraints posed by the macrofactors. This also has implications and costs. In this strategy, the inﬂuences of the environment are as dominant as those of the strategic choices in the actorcentered approach. I believe, however, that for a historical understanding of left electoral mobilization variance, such a macrostructural constellation can offer a richer comparative framework. SPATIAL VARIATION: ARE CROSS-COUNTRY COMPARISONS TOO AGGREGATE? Most of my data concern the electoral mobilization of the left at the national level. As a result, other independent variables are necessarily measured at that same level for an election, a period, or a general country mean. The skepticism in regard to widespread geographical research is sometimes argued from the methodological point of view, referring to the ‘‘too different experiences’’ argument and to the ‘‘too much is lost in national aggregation’’ argument. In my opinion, such theses are not justiﬁed in terms of variance in independent and dependent variables. The level of spatial aggregation at which research is conducted is crucial only for the deﬁnition of those aspects that need to be regarded as constant (those that belong to higher spatial levels than that chosen), those that can be considered as operative variables and are therefore subject to the parameterization controls (those that belong to the same spatial level) and those aspects that cannot be considered in – and are therefore excluded from – the analysis (properties belonging to lower spatial units).

42

THE POLITICAL MOBILIZATION OF THE LEFT

The variance in the dependent variable – for example, the variance in communist electoral strength – is clearly not related to the level of the territorial units of analysis. Whether the territorial units are ﬁfteen countries or ﬁfteen regions of a given country or ﬁfteen cities of a given region or ﬁfteen districts of a given city or even ﬁfteen polling ofﬁces of a given district, the variance varies randomly. This is to say that there may be more variance among the Communist electoral strength of ﬁfteen different Florentine districts than among ﬁfteen European countries. What about variance in independent variables? Moving down the ladder of territorial aggregation, some of the variables that did indeed vary at a higher level of territorial aggregation (e.g., the electoral system, the timing of enfranchisement) become constant and cannot, therefore, be invoked in explaining within-country variations in the dependent variable. At the same time, other independent variables that could not be considered at a higher territorial aggregation level now become potential operational causes of variation. Thus, descending in spatial level, (1) those variables that vary at a higher level of territorial aggregation are transformed into constants; (2) the inﬂuence of the speciﬁc level variables, which had to be considered as randomly distributed and therefore noninﬂuential, at the higher level is ampliﬁed; and (3) all those potential independent variables that belong to lower territorial levels continue to be excluded. The same logic applies at every level. If the study were to be conducted across the electoral districts in the city of Florence, for example, we would be forced to consider as potential causal conditions district-speciﬁc effects such as the 1966 ﬂood having radically changed the social composition of some districts while leaving that of others substantially unchanged. At the same time, every property that can be studied and attributed to the city, the region, or the country necessarily becomes a constant feature across the city districts and cannot be invoked to explain variance among them. In a comparative study, any changing of the level of territorial analysis produces changes in the variables that can reasonably be considered as potential causal conditions of that level variance. However, while some variables acquire such status at a lower level, others lose this status at the same level, because for each level of spatial unit chosen, we must assume the random distribution of lower-level features and the constant distribution of higher-level features. Once we have listed all the microcontext variables to explain the Communist success in Siena versus its failure in Lucca, how can its success in Tuscany versus its failure in Veneto be explained? Or its success in Italy versus its failure in the Netherlands? There is obviously no way to maintain the status of potential causal conditions for all the variables while descending or climbing the ladder of

CLASS CLEAVAGE: FRAMEWORK

43

territorial aggregation. So, for variance-explaining research, the argument that one becomes more speciﬁc, context sensitive, accurate, and so on by studying phenomena at a low level of territorial aggregation is logically inconsistent. Therefore, the issue of the level at which we decide to study a given phenomenon has nothing to do with context sensitivity, historical speciﬁcity, or level of abstraction. It simply points to different sets of independent variables that we want to consider as potential causal conditions in a situation in which the choice of one set precludes the consideration of other sets. The choice of a spatial level of aggregation is therefore a choice about which sorts of potential sources of variation we can control and which ones we must leave uncontrolled. The reason some local studies appear to be more appealing and context sensitive is simply that they combine a great deal of local speciﬁcation with more-general knowledge concerning a higher level of territorial unit: the region, the country, or even the continental context. The latter aspects can, however, only be used as other knowledge and cannot be incorporated into the comparative research design because they do not vary. This is the same case as when, in studying a phenomenon at a given level – for instance, the cross-country level – one adds several subnational aspects, regional or local, to the picture as explanatory elements. It can be done, and it is done, but these cannot be incorporated into the comparative research design properly, and they remain ad hoc interpretations that cannot be submitted to a comparative test. The manipulation of the variance under examination is achieved only through the selection of cases – either more homogeneous or more different cases – not through the selection of the spatial level of analysis. Selecting cases with the goal of manipulating the sample’s variance obviously changes the research design. If one selects, for instance, cases of strong Communist inroads (minimizing the variance in the dependent variable), one is forced to search for the common causes of a common result. In making constant the effect (Communist electoral success), it is necessary to look for features common to all cases that are likely to explain it. If, on the other hand, the aim is to parametrize the causes (for instance, a common state-bureaucratic tradition) through the selection of cases, the search must focus on the differences between the cases that may explain the different outcomes. This is not the place to engage in a discussion of the methodological implications of the different strategies of parameterization of potential causes or effects in comparative research. I am making the point only to underline that the methodological implications of strategies of maximum variance and minimum variance in both the indepen-

44

THE POLITICAL MOBILIZATION OF THE LEFT

dent and dependent variables have nothing to do with the spatial level at which the research is conducted, as they can be followed at each of these possible levels. My conclusion is, therefore, that the objection that in considering countries as units one cancels internal variation is immaterial from a logical point of view (while it remains important in terms of the selection of independent variables). The question of how much is lost and how much is gained in moving up and down the ladder of spatial unit aggregation is a general methodological question, not a commonsense argument about the presumably more concrete, speciﬁc, or contextsensitive quality of local or country studies versus wider comparative research. This study takes the Western European countries as its main spatial units of analysis. Most of the time, these units are national elections taken as a single piece of data or as a country average. It is clear, therefore, that cross-country variations will be implicated and within-country differentiation will have to be placed to one side. The implication of this choice in terms of which aspects can be submitted to empirical control and which are deﬁnitely outside it should by now be clear. It should be added that within the range of thirteen countries and about 370 elections, subsets of cases will often be selected to manipulate variance and to perform more limited comparisons among sets characterized by the maximum and minimum variance in the dependent or independent variables. This, however, will not modify the ‘‘systemic properties’’ framework of the research. TEMPORAL VARIATION This study is not a synchronic cross-sectional comparison performed at a given chronological point or functionally equivalent time. It covers almost a century of European politics. This involves temporal variations within countries that are as large as, if not larger than, the cross-country spatial variations. Elsewhere,45 I have developed my methodological ideas about the difﬁcult relationship between spatial and temporal variance in comparative research, and there is no need to restate them here. The main conclusion is that, for the purpose of advancing causal sequential generalizations, resorting to ‘‘history’’ as the analysis of temporal variance, without the simultaneous analysis of synchronic cross-sectional variance, is far from an ideal solution; rather, it has its own methodological pitfalls. It is pointless to discuss whether the pitfalls of a synchronic analysis deprived of a temporal dimension are more or less intense than those of a longitudinal analysis deprived of synchronic contrasts among different units. For control 45

Bartolini (1993).

CLASS CLEAVAGE: FRAMEWORK

45

purposes, the safest method is to contrast cross-sectional and cross-temporal results; therefore, research designs should try to observe both types of variance. This research on the political mobilization of the left also has, therefore, methodological objectives. It is an attempt to put into practice these methodological guidelines. I apply a comparative research design that combines variation in both the spatial and temporal dimensions, and I hope that the substantive results support the conviction that systematic cross-spatial and cross-temporal research designs are feasible and preferable to other strategies of comparative inquiry. My main methodological tool to check sequential generalizations is what I have called the ‘‘slides of synchronic comparisons through time,’’ that is, synchronic cross-sectional comparisons made at different points in time and contrasted with the results of general analytical and developmental relationships among the aspects under examination. The relatively limited temporal dimension of this developmental comparison should make it less dependent on general theories and should also make more manageable the crucial problem of deﬁning with sufﬁcient precision the temporal units of the analysis. The absence of any ambition to reach spatial universality allows for a more culturally bounded and less abstract conceptualization, the operationalization and empirical referents of which are easier to identify. The properties that can be considered in this temporally and spatially restricted set of units can thus be made both more numerous and empirically more speciﬁc than those usually considered within general development theory, although substantially less numerous than those in the case studies. Meeting these conditions in the study of the patterns of electoral mobilization of the class left allows for four distinct modes of analysis that guarantee the maximum control over generalizations. These four modes are represented in a simple exemplary form in Figure 1.2, applying them to a set of units for each of which variation over time in two variables (A and B) is considered. I will examine four different types of variance. First, by combining all the spatial units together (countries and elections), relationships among the variables can be established at the general analytical level as if they were time nondependent (ﬁrst graph in the ﬁgure). Units enter into the picture according to the value/state of the variables at each different time, and as such, the different temporal units are not logically distinct from the spatial cases. Time is simply a unit-deﬁning parameter. Second, combining all the spatial units along the dimension of time – as in the second graph – allows for the study of the overall temporal variations and the relationships among the historical trends of the variables, which are studied in terms of their development over time. In this case, the develop-

46

THE POLITICAL MOBILIZATION OF THE LEFT

Figure 1.2. Four modes of analysis combining cross-time and cross-space variation. ment trend is analyzed in terms of its general characteristics or average values (phases of statemaking, party system consolidation, suffrage expansion, etc.). This picture of how and when the relationship between the variables changes over time, or in different periods of time, and the possibility of identifying thresholds of structural change reinforces the understanding of the temporal association and causal priority among the variables. These two types of variance – analytical and time dependent – can be checked in the history of each unit, so that we can compare whether the general analytical and temporal relationship resulting at the overall level is reproduced in the case of each single unit or whether it results from different patterns. In the third and fourth graphs, cross-unit variance is therefore added. The ﬁrst two strategies are reproduced in each unit, thus

CLASS CLEAVAGE: FRAMEWORK

47

obtaining cross-sectional variance and the possibility of comparison in terms of both the analytical and the temporal relationship between the variables. The analytical and temporal relations studied in the ﬁrst two graphs constitute the general reference point with respect to which the crosscountry variance can be identiﬁed: both in terms of the differences between units and, what is more important, in terms of the differences between the country-specific analytical and temporal relationship, on the one hand, and the general analytical and temporal relationship, on the other. The latter provide the yardstick against which statements of timing, tempo, and mode can be expressed and against which deviant cases can be identiﬁed. In other words, the overall analytical relationship and the overall temporal relationship are considered as the yardsticks against which each unit’s analytical or temporal relationship can be judged as low or high, strong or weak, early or late. In sum, they provide the reference points for arriving at these comparative statements. Thus the variance on which I will concentrate is not the difference in any one aspect between unit A and unit B, but the difference of each unit from the overall trend or analytical relation. This point is far from trivial. Assumptions about earliness/lateness, presence/absence, intensity, and so on frequently form part of comparative assessments, but more often than not, the reference point is far from clear. The extension of the suffrage in country X is early – but with respect to what? In a two-country comparison, the answer is simple; but in a twenty-country comparison, it becomes more difﬁcult, and here the yardstick should probably be represented by a general temporal study of suffrage expansion over the whole set of countries. Depending on the nature of the measurement of the variables, such a general study could identify a general trend, or different phases, or different historical thresholds with respect to which national trends, phases, or historical thresholds can be judged. In other words, a matrix of temporal and spatial variance cannot be identified empirically and systematically solely on the basis of comparisons among individual units, but only by reference to the temporal and analytical relationships that exist more generally. This transformation of the variance to be explained is, in my opinion, of paramount importance. Let me clarify further the process involved with an example. Let us assume that the analytical relation between workingclass constituency size and left vote is 0.9 in country A and 0.2 in country B. If we concentrate on a direct comparison, then what has to be explained is the difference of 0.7 between the two cases, which may be attributed to any third intervening variable. If, however, we decide to take the overall association as the relative comparative yardstick, the variance to be explained changes drastically. Let us assume that in the entire universe of

48

THE POLITICAL MOBILIZATION OF THE LEFT

400 elections, such an association is 0.55. Thus, both countries are deviating by the same amount (0.35) from the overall association, although in different directions. If, however, the overall association is 0.85, the picture is quite different. Country A is basically in line with the overall association, while country B is an exceptionally deviating case. In this latter example, the variance that interests me is the 0.65 that sets country B apart from the general association. It is thus possible to see that the general analytical and temporal yardsticks not only make comparative statements across many cases more accurate, reliable, and valid intersubjectively, but also redeﬁne the general variance to be explained. In the course of this study, there are many instances in which this redeﬁnition of the variance is surprising and raises unexpected questions. MICRO VERSUS MACRO: WHAT KIND OF STUDY Nothwithstanding recent attempts to underline the advantages of ‘‘holistic’’ approaches to the comparative study of political systems,46 the dominant emphasis in political science in the 1980s and 1990s was different. A ‘‘decompositional’’ approach taken to its extreme consequences – to the fundamental actor, the basic particle of political phenomena: the individual – has recently and aggressively claimed its epistemological superiority and its objective of standardizing social science into a dominant paradigm. The (re)discovery of a few basic propositions has relaunched various versions of the ‘‘actor-centred’’ study of politics, including the idea that individuals have much more control over their own fate than was previously conceded; that therefore attention should be focused on them or, more generally, on the ‘‘actor’’ (collective in this case); that their motivations should be taken seriously, as seriously as their calculation concerning the best way of achieving what they want; and that the outcomes of political processes should hence be seen primarily as the intended or unintended consequence of subjective intentions and actions, interpreted within the framework of rational calculations of alternative costs and advantages. Differences exist in these rational-choice approaches according to the point of reference for the ‘‘rationality’’ of the actor and the different role attributed to the effect of cognition through and in the process of choice. There is little doubt, however, that these approaches are bound together by the basic idea of assuming individual and collective rationality in pursuing his/her/its interests as the basis of the choice and of regarding as 46

See Easton (1990).

CLASS CLEAVAGE: FRAMEWORK

49

external constraints of such a choice context the ‘‘structures’’ (in Easton’s sense), the ‘‘institutions,’’ and that broad and vaguely systematized set of factors usually referred to as ‘‘history.’’ Recently, this approach has spread to the ﬁeld of comparative politics and even that of comparative historical research. In so doing, it has somehow relaxed the set of postulates underlying the more theoretical endeavors, producing works that are both substantially and methodologically challenging.47 This book belongs to the opposite tradition. It is mainly concerned with the environmental constraints and macrostructural features of the political system and leaves little room, if any, to the study of the individual or collective actors’ motivations, choices, and strategies. The choice of the explicandum – the variation in the left mobilization patterns, the choice of the cross-country spatial dimension, and the choice of a retrospective strategy of identiﬁcation of crucial historical junctures – is in part responsible for such a design, which overemphasizes constraints on action imposed by structures over the explanation as identiﬁcation of actors’ motivations and choices. However, it is responsible only in part. The other part is simply due to my own preferences. Those who have a fairly eclectic methodological approach orientation and are convinced that, ﬁrst, the social sciences will never agree on a superior approach; second, the social sciences will continue to be characterized by cycles in which one approach gains predominance, without dislodging others, until it declines physiologically to the advantage of something else, when the epigones have completely taken over the early original innovators; and, third, that good as well as mediocre studies exist and will exist within each different tradition and approach should not involve themselves to any great extent in an approach debate. However, something should be said against the social science paradigmatic standardization sometimes pursued by rational-choice supporters and the claim made by actor-centered explanations of being the only legitimate approach, together with its correlate claim that macroassociations between collective behavior and historical and structural factors are meaningless unless they specify micromotivations. This is undoubtedly refreshing when it is presented as a new way of restating and reviewing classic problems, but it becomes fatuous when it is presented as the ‘‘new paradigm,’’ outside of which oldfashioned, prescientiﬁc research persists in the fossilized belief that political action can be interpreted as a response to some sort of external stimulus that can be generalized. 47

In the comparative party history ﬁeld, one should mention Przeworski and Sprague (1986) and Kalivas (1996).

50

THE POLITICAL MOBILIZATION OF THE LEFT

The emphasis on individual preferences and rational calculations versus that of the study on constraining conditions – or on ‘‘conﬁning conditions,’’ as Kirchheimer once put it48 – should not be seen in epistemological terms, but more as a question of alternative research strategy. First, to consider collectivities and to study the constraining factors external to them and their collective exposure to similar external stimuli is not to forget factors of choice and individual evaluation in making action decisions. Even in the period of dominant behaviorism, there was a keen awareness of the key role of individual motivations, beliefs, and interests.49 The issue was not, and is not, whether the action of the actor is or is not a rational adaptation (of some sort) to his or her personal situation in a given system as the actor perceives it subjectively. From the individual’s point of view, this is unquestionably what happens. Individual action will always be the action of subjective rational adaptation to personal situations. The question is rather whether taking the lead from this subjective perception is the best research strategy from the external point of view of the researcher. This is clearly quite a different question. The obvious starting point for every approach is the realization and attempt to explain some regularities and conformities in the behavior (or actions, if one prefers) of individuals. In an approach that relates macrophenomena to the aggregate behavior of collectivities, the subjective perceptions of individuals are implicitly thought to be shaped by social characteristics and phenomena that are external to the mind of the actor and that produce the behavioral group conformity and regularity that is observable. It might be methodologically correct and wise to specify the personal motivations that are thought to be produced by a constellation of external stimuli, but it remains obvious that such individual motivations are not and cannot be studied per se and are only induced from the inﬂuences of the macrofactor. If, by contrast, the search is initiated at the point of the personal motivations of the individual (or of the interest of the collectivity), it 48 49

Kirchheimer (1965: 964–974). For instance, in works like Lazarsfeld, Berelson, and Gaudet (1948) and Berelson, Lazarsfeld, and MacPhee (1954). This quotation is a perfect example of this awareness: ‘‘Social-position variables, such as class or place of residence, do not ‘cause’ any speciﬁc behaviour in the sense that they are requisite for, or the immediate antecedent of, given acts. Social conditions. however, do form personalities, beliefs and attitudes which, in turn, do ‘cause’ (are requisite to) speciﬁc acts such as participation in politics. . . . It is a simple matter, then, to ﬁnd correlational relationship between social position variables and political participation, but the reader should keep in mind that the effect of social position variables must be mediated through personality, beliefs, and opinions.’’ Milbrath (1965: 110).

CLASS CLEAVAGE: FRAMEWORK

51

cannot be said that we know them or study them better. Very simply, we postulate them or reconstruct them deductively from some general principle such as ‘‘self-interest.’’ It seems highly unlikely that either the most radical-rational choice approach or the most deterministic one can know or understand the motivations of individuals, particularly when these number thousands or millions. To the extent that actions show regularities in groups of individuals, these can be explained by making references to common perceptions, preferences, and situations of choice, but these common personal situations and perceptions can also imply the presence of external inﬂuences determining them and external conﬁning conditions shaping and directing them. In a nutshell, if we start from the observation of group regularity in behavior, we may well conclude that this exists because all members of the group have made the same rational adaptation to the situation as they perceive it, but we may also legitimately inquire whether these common situations/perceptions do not come from some external source. If the formation of individual motivations (or preferences) were not somehow shaped by external inﬂuences and constraining structures, it would be hard to interpret group behavioral regularities and to escape a total randomness of individual ends. In this perspective, the issue of individual motivations versus structural constraints is not that of a choice between the ‘‘good’’ and ‘‘adequate’’ approach against the ‘‘wrong’’ one, but more simply the choice between alternative approaches, both of which may have advantages and disadvantages. Basically, by linking macrostructural determinants to aggregate individual behavior, we bypass individual motivations by assuming that aggregate behavioral regularity is an adequate although indirect indicator of regularities in individual motivation formation. By linking micromotivations to aggregate individual behavior, we bypass those forces that can account for the similarity in individual motivations that leads to group regularity and conformity in individual choices. The question is basically where we put the ‘‘black box’’ that seems to be inescapably associated with every human science research approach.50 If general processes, outcomes, and institutions are reconstructed as aggregate results of individuals’ purposeful rational actions in a given context, it has to be admitted that while the motivations, perceptions, and values of the individuals and groups can be postulated or reconstructed, 50

This is why I ﬁnd greatly exaggerated the ‘‘holistic’’ paradigm label attributed by Boudon (1979: 187–252) to all approaches that do not take the lead from the motivation of the actor.

52

THE POLITICAL MOBILIZATION OF THE LEFT

their process of formation and actual content is unknown. This is the black box in this case. In contrast, by directly linking the study of macroprocesses and institutions to observable behavior, we assume that somehow these conditions determine a regularity in the productions of goals, perceptions, motivations, and so on that, in turn, leads to behavioral choices. Obviously, in this case, these motivations are not studied directly, nor are they postulated a priori; they disappear into the link that is established between conditions and behaviors. From this point of view, the accusation in the ﬁeld of comparative research that macrostructural approaches do not specify the micromotivational foundations of their work is equivalent to the accusation that rational-choice perspectives do not specify the historical, cultural, and institutional constraints that produce and shape individual motivations. Is it better to examine behavior starting from proximate motivations and leaving aside the processes that structured their formation, or is it better to examine it starting from the more distal point of constraining conditions and postulating their effect on motivations? Is the arbitrariness of postulating motivations and deducing choices within a set of constraining conditions higher or lower than the arbitrariness of inducing motivations and choices from the association between macrofactors and collective behavior? It is, quite honestly, difﬁcult to give an answer to these questions. I am ready to be convinced that the actor-centered perspective is more productive, but this view must derive from the substantive results of comparative research conducted with this paradigm, not from the inappropriate conclusion that, as human history is obviously made up of individuals’ choices and actions, scientiﬁc research must start with them. As I have been arguing here, the issue of what it is that ultimately constitutes human history, and the question of which is the best paradigm by which to interpret it, are separate questions. In conclusion, this issue of paradigms – provided that it remains with us long enough before being overshadowed by some new, fashionable rediscovery or comeback – should be treated pragmatically, looking at the quality of the results produced by the different orientations. For this, a comparison of the results of different strategies is necessary. Mutual checking is better than paradigmatic uniformity. As a result of my intellectual training, I have come to believe that collective behaviors and environmental constraints are the visible part of the moon and that individual motivations, beliefs, preferences, perceptions, and eventual choices are its hidden one. If the intellectual climate has changed and it is now viewed as easier and more appropriate to start from the latter, I am ready to accept a

CLASS CLEAVAGE: FRAMEWORK

53

deﬁnition of the approach followed in this book as traditional. After all, even supposedly disappearing minorities have rights to vindicate; tradition sometimes defends the pluralism facing modernity and checks against the oversized victory of the moderns.

2

THE EUROPEAN LEFT: SIZE, IDEOLOGICAL ORIENTATION, AND ORGANIZATIONAL COHESION his chapter is mainly descriptive. Relying on secondary literature and electoral and organizational data, it describes the variation in my dependent variable: the national experiences of the class left in the three main dimensions of size and electoral development, early ideological orientation, and organizational cohesion.

T

SIZE AND ELECTORAL DEVELOPMENT In the 361 national elections held in the thirteen countries between the beginning of the 1880s and the end of the 1980s – almost 110 years – at least one class-left party was present and competed in 347. The fourteen missing cases are concentrated in three countries where free elections were held in the 1880s and 1890s without a socialist-type party taking part: in Sweden between 1887 and 1899 (six elections); in the United Kingdom between 1885 and 1895 (four elections); and in Norway between 1882 and 1891 (four elections). These three countries have, therefore, a late entry of the left into electoral politics with respect to the beginning of competitive elections. In the twentieth century, socialist, communist, and other splinter parties of the same family collected one-third of the European vote averaged by country (33.2%; see Table 2.1), 30.3% of the vote of the total European electorate (see Table 2.2), and 25.1% of the European valid votes. Table 2.1 reports the country average strength over the whole century and the standard deviation of their results, as well as the number 54

THE EUROPEAN LEFT

55

Table 2.1. European left average vote by country (% of valid votes, 1880–1989)

Table 2.2. Average European left vote development (five-year means in thousands)

56

THE POLITICAL MOBILIZATION OF THE LEFT

of elections in each country. In a century-long perspective, only three left parties overcome the 40% average barrier: Austria, Finland, and Sweden. The largest group of seven countries averages between 32% and 35% (Belgium, Denmark, France, Germany, Italy, Norway, and the United Kingdom), which is approximately the European mean. Three countries have a very small left – the Netherlands, Switzerland, and Ireland – which are by far the smallest left in Europe. Six countries have a standard deviation (SD) that is much lower than the mean – Ireland and Switzerland, Germany and Belgium, Finland and Austria. In these cases, the electoral scores of the left have remained within a relatively close range, with over-time variations much lower than those in the other countries. If they ﬁnally resulted in a strong (Austria and Finland), medium (Germany and Belgium), or small (Switzerland and Ireland) left, they were quite strong, medium, or small from the beginning. The differences in the way in which the electoral performance of the national left is spread over time can be best appreciated in Figure 2.1, where a boxplot of their electoral result is produced. The boxes in Figure 2.1 include 50% of the values (their top side represents the 75th percentile and their bottom side the 25th percentile), and the line within the boxes is the within-box median value of each distribution. The top and bottom values that are not outliers are shown with lines drawn from the top and the bottom of the box to them. Finally, outlyers are identiﬁed with two symbols: o if they are values within a range of 1.5 to 3 box lengths and * if they are values beyond three box lengths (extreme values). The length of the box – the interquartile range – gives a good idea of the variability of the observations. Outliers are located in the lower part of the boxes, indicating in most cases the early elections that the left contested in these countries. However, these outliers are few and are concentrated in those countries with little variation in left electoral results, like Austria, Belgium, Germany, and Ireland. Finland and Austria are impressive for the solid homogeneity of their electoral result over almost a century. Finland does not even have any outlier, and the low position of its median within the box indicates that its electoral history is made up of a very high and historically consistent level of around 40%, with half of the elections ranging between 36% and 42% and the other half spreading from 42% up to 52%. Austria is very similar. The ﬁrst two elections fought under the constitutional monarchy are deviating; after them, all the Austrian electoral results spread within the narrow range of fourteen points, between 38% and 52%. With a smaller size, Ireland has a variance that is even lower. In Germany, Belgium, Sweden, and the United Kingdom, the variabil-

THE EUROPEAN LEFT

57

Figure 2.1 Boxplot of total left vote by country (1880–1989). ity of the electoral results is intermediate. These countries all have outliers in the ﬁrst two or three elections, but since then their range has stayed fairly contained. At the other end of the spectrum, Denmark, France, Italy, Norway, and the Netherlands evidence the largest spread. The French distribution is the most striking, standing at about 50 electoral percentage points. Italy is not far from this extreme, and even in Denmark and Norway the general spread of values is high, although the bulk of the 50% values in the box is greatly reduced compared with those of France and Italy. The latter countries are the best examples of stormy electoral histories, starting with very low levels that, however, do not appear as outliers, and moving slowly to higher electoral levels, with frequent ups and downs that cause great variation in their central 50% values. The median value is more or less the same, but whereas in Italy the top 25% values are quite compressed in a reduced range, in France the top 25% of the cases ranges between 40% and 55%. France’s left is characterized by great electoral victories as well as great electoral defeats and probably by a persistent instability and incapacity to stabilize a solid rock of identiﬁed voters. In contrast, the British case is paradigmatic of a normal electoral development. The few outliers simply indicate the understandably

58

THE POLITICAL MOBILIZATION OF THE LEFT

poor result of the ﬁrst entry, while the median value is very close to the center of the box and the two top and bottom 25% sections are not too different. Country median values do not take into account either the different sizes of country electorates or the different number of elections in each country. A better picture of the left mobilization over time is obtained by referring the votes collected by class left parties to the global European electorate. In order to equalize the number of elections, we can take for each country a ﬁve-year average vote (Table 2.2). Not all countries are present in every period, so the last column of the table indicates how many countries contribute to the average. Even with the limitation of the different number of countries per period, these ﬁgures give a more accurate idea of the overall pattern of European left historical development and rate of growth. In the 1906–1910 period, the decline in the left vote is due to the large increase in the number of available country results, a step that adds weaker lefts to the picture. With this exception, the left keeps an average positive rate of growth from the end of the nineteenth century to the 1930s. Then the rate of growth begins to diminish. It is again high in the immediate post–World War II ﬁve-year period. Following this, for most of the post–World War II period, the European left keeps a stable percentage of the electorate. Throughout the post–World War II period, all countries are present, and although the European electorate passes from 147 million in 1946–1950 to 211 million in 1981–1985, the share of the class left vote tends to remain the same. However, in 1981–1985 for the ﬁrst time, we observe a negative variation that is quantitatively considerable. Other aspects of this historical development can be appreciated if, instead of considering ﬁve-year averages, we look at the trend of all elections in all countries. Figure 2.2 shows the 347 elections contested by some of the left parties. The set is divided into four periods: the pre– World War I period from 1880 to 1917, the interwars period from 1918 to 1944, the ﬁrst post–World War II period from 1945 to 1965, and the second postwar period from 1965 to 1985. For each of these periods, a lowess ﬁt line is used that produces a locally weighted regression scatterplot smoothing method to ﬁt a line to a set of points. This method ﬁts a speciﬁed number of data points carrying out a speciﬁed number of iterations. In Figure 2.2, 70% of the points have been ﬁtted through three iterations.1 Figure 2.2 identiﬁes four clearly differentiated patterns of over-time 1

Up to 99% of the points can be ﬁtted. In doing so, the basic trend of the lines does not change, but the lines themselves increasingly approach a pure linear regression.

THE EUROPEAN LEFT

59

Figure 2.2. Electoral development: lowest fit lines by period. development, each period characterized by a speciﬁc pattern of absolute and relative increases. The data sets that appear in the low part of the ﬁgure, starting from about the middle of the 1920s, are the results of the Irish left, deviating from all the other countries and generally very stable over time. The global ﬁgure seems to indicate a typical Gompertz function. In the ﬁrst period before World War I, the European left passes from a genetic to a growth phase. In the genetic phase, up to the ﬁrst years of the twentieth century, the developments and rates of growth are minor. This phase was characterized by the ﬁrst appearances of the class left, generally in contexts of suffrage that were still restricted. By the ﬁrst decade of the twentieth century, the slope transforms itself into an exponential-growth curve characterized by strong rates of growth and increase. In the period between the two world wars, the left passes to its maturity phase, characterized by declining increments. The trend is still growing, but it has already changed its shape. It becomes linear, with homogeneous increases over time. It is possible that the steepness of this line is reduced by the seven Irish elections between 1927 and 1945, and indeed their impact is visible in the smoothed line in 1927. However, this impact does not modify the radical difference between the lines of the ﬁrst two decades of the century and that of the second two decades – a difference that highlights the passing from the takeoff phase to the steady-growth phase. The ﬁrst post–World War II period corresponds to a phase of stability.

60

THE POLITICAL MOBILIZATION OF THE LEFT

In the aftermath of the war, the European left reaches its peak, but immediately afterward, in about 1948, it acquires its period level, which remains more or less stable to the end of the 1960s. The left in this period faced the saturation of its potential electoral market. In the last phase, between 1965 and 1985, there are, in fact, signs of electoral decline. These are not dramatic, but they are clear. What is more important is that for the entire post–World War II period the European left does not evidence any clear secular trend and thus, historically, it has reached its ceiling. The recent literature that has so strongly emphasized the crisis, and sometimes even the electoral decline of the traditional class left, should be evaluated in the face of this secular trend. At the overall European level, the left had already stopped growing by the late 1940s, and the recent mild signs of decline do not justify the claim of a collapse. Let us now combine historical development with cross-country differences. Table 2.3 offers a comparative appreciation of national left size by period and its change over time (for the temporal variation by country, see the later discussion). Before the Great War, only the Finnish left had reached the size of a strong, governmental party. It was followed – though at a considerable distance – by the Austrian, Swedish, and German lefts. The German Social Democratic Party could claim to be the oldest in Europe, but even in the pre–World War I period it was not the strongest electorally. At the opposite end, the United Kingdom and the Netherlands had the weakest left parties in Europe at the time (excluding the nonexistent Irish left). In the second period, between the world wars, the generalized pattern of growth reduces the cross-country variation in left size, and weak lefts tend to gain more than strong lefts in a general process of homogenization. The exception is Sweden, whose left – although it was already strong before the end of the Great War – grew more than any other left except the British and became the strongest in Europe in the interwar years. The Dutch and Swiss lefts became the weakest, followed closely by the French, while the British left caught up with all the other countries with an accelerated rate of growth. In passing from the interwar period to the ﬁrst post–World War II period, the difference is still positive but greatly reduced. The size of the left grows in all countries except Germany (⫺3.0%). In two countries – Belgium and Switzerland – the left has clearly stopped growing, although the prewar level of these lefts was not exceptionally high. The same happens to Sweden, but with the difference that it had already become the strongest European left before World War II. It is interesting to note that in 1945–1965 the German left is one of the weakest in Europe, stronger only than the Swiss and, obviously, the Irish. Norway and the United

THE EUROPEAN LEFT

61

Table 2.3. Average electoral strength of the national left by period and differences between periods

Kingdom experience the highest growth difference between these two periods. The countries with medium growth are those that already had a strong left in the interwar period (Sweden, Finland, Austria, and Denmark). If we exclude the stable lefts and the declining German left, in all the other countries there is a certain inverse proportionality between the prewar size of the left and the electoral increases in the postwar period. On the whole, in the 1945–1965 period, we have two lefts that are on average majoritarian (Sweden and Norway) and four that approach that level (Austria, Denmark, Finland, and the United Kingdom). Three countries have what can be called ‘‘intermediate’’ strength at around 40% of the vote: France, Italy, and Belgium; three others are clearly below the European average, between 30% and 35%: the Netherlands, Germany, and Switzerland. Ireland continues to stand alone as an exceptional case. In the last period, 1965–1985, few major ranking changes occur. The general level of growth has stopped, as we have seen, and its value is negative. Six countries have positive interperiod differences, and seven have negative ones. The most pronounced change concerns the breakdown of the Belgian left, dramatically affected by the crisis of national politics after the mid-1960s (⫺8.9%). Other noticeable declines from a previous high occur in the United Kingdom, Norway, and Finland. A less pronounced decline occurs in Sweden. There is no such decline in Austria and Denmark, while France, Italy, and Germany show the most signiﬁcant increases

62

THE POLITICAL MOBILIZATION OF THE LEFT

in electoral support. In general, there are signs that the stronger lefts of the 1945–1965 period tend to have an earlier and more pronounced decline than the weaker lefts, although the Austrian and Danish lefts manage to avoid it. In contrast, some of the medium-sized lefts of the 1945–1965 period still seem to have some margins of growth. Finally, it must be noted that the Dutch and Swiss lefts decline in the last period, although they had never been particularly strong before. In general, therefore, one cannot speak of a uniform and generalized decline since the 1960s. At the end of the twentieth century, therefore, the picture changes in the relative ranking of countries – apart from the obvious historical failure of the Irish left; also, the Dutch, Swiss, and Belgian lefts appear to be electorally undermobilized. Sweden, Austria, Norway, and Denmark maintain their historical position as leaders, while Italy and France reach the level of electoral mobilization of the latter two countries in this last phase. Finally, Finland and the United Kingdom fall to around the all-country European average together with Germany. With respect to the beginning of the electoral history of European socialism, this last phase has therefore seen some radical changes in size ranking, the clearest of which are (1) the drastic retrenching of the electorally promising Belgian left; (2) the lack in breakthrough of the early powerful German left; (3) the rising rank position of the relatively late French and Italian lefts; and (4) the parabola of British labor, which in the last phase falls back more or less to the electoral size of the interwar years. The continuous governmental potential of the British left in the context of its party system and electoral rules should not overshadow the fact that in the 1965–1989 period this left is only 4 percentage points stronger than the Dutch left. Period means by country allow a synthetic appreciation but also conceal a great deal of over-time variation. To examine the temporal development patterns of the thirteen countries, in Figure 2.3 they have been classiﬁed into three groups according to the predominant shape of their electoral history. In Figure 2.3a, the United Kingdom, the three Scandinavian countries, and Finland are charted. These cases demonstrate a comparatively late entry of the class left into the electoral market. In Norway, the United Kingdom, and Sweden, the left appears only at the beginning of the twentieth century, and, as a consequence, its growth phase is extraordinarily rapid. Denmark has an earlier start and a less rapid growth phase. The Finnish left’s appearance is one of the latest in Europe due to the delay of political rights and partisan politics in Finland; but once it appears, its level is extremely high. Finland is the only country in which the left ﬁelded about 40% of the votes before World War I. Consequently, it does not ﬁt the four phases of the overall European development. To

THE EUROPEAN LEFT

63

begin with, it was an oversized left, and its successive development was therefore less trendy. All cases peak in the middle of the 1940s, all had about 50% of the votes in the 1945–1965 period, and all have tended toward a negative and declining trend since that date, even if such a trend has been more accentuated in the United Kingdom and Finland than in the Scandinavian countries. Note the exceptionally continuous and rapid growth phase of the Swedish left up to the 1940s. The French, Italian, and Austrian lefts are charted in Figure 2.3b. The growth phase in these countries is less rapid, also because their entry level is higher than that of the previous group: when they ﬁrst entered the electoral races, they immediately obtained between 10% and 20% of the votes. They therefore tend to grow less rapidly than those in the ﬁrst group. Moreover, no signs of a ceiling emerge in the middle of the 1940s. In addition, none of these lefts is characterized by a negative electoral trend after 1965. Quite the contrary: They continue to grow even after World War II. Even if the immediate post–World War II election years are peak years, all of them tend to show better electoral performances in the 1970s than in the previous decades. This is particularly evident in Italy, whose trend up to the middle of the 1980s is one of continuous albeit moderate growth. France is interesting because the peak reached in 1945–1946 is followed by a lasting and critical period of decline throughout the Fourth Republic (1945–1958) and by a slow but continuous recovery during the Fifth Republic (1962–1986), which brings it to its highest-ever electoral level. The striking decline of the left during the 1950s is an exclusively French feature. The Austrian left overcomes the 50% voter threshold in the 1970s. Although it was an early strong left, reaching 40% of the votes after World War I, it took another half-century to reach its highest point. Finally, it should be pointed out that in the post–World War II period, it was only in Italy that the electoral growth was constant. Figure 2.3c regroups the remaining countries. I have included Ireland, which, as I said earlier, is a unique and totally deviating case. The other cases – Germany, Belgium, the Netherlands, and Switzerland – do have something in common, however. First of all, none of these lefts ever approaches the 50% threshold. Second, they tend to be early mobilizers, particularly in Germany and Belgium, but they also reach their maturity level before the other cases. For Germany, Belgium, and Switzerland, the immediate post–World War I years deﬁne their maturity level, above which they will never rise. The Netherlands is slightly different because it grows in the interwar period, too, and peaks and reaches maturity in 1945, like the countries of the ﬁrst group. Its growth is, however, much less marked and rapid, and the overall maturity level reached is much lower.

64

THE POLITICAL MOBILIZATION OF THE LEFT

Figure 2.3. National electoral developments (1880–1989).

THE EUROPEAN LEFT

65

With the partial exception of the Netherlands, then, it can be said that in these countries electoral growth stops between the two world wars at intermediate levels. Switzerland and Ireland are striking for the marked stability of their results; they seem to have been untouched by the huge sociopolitical changes since the end of the First World War. To a certain extent, the Belgian left may also be included in this group, because it remained remarkably stable throughout the long period from 1919 to 1965, changing only when the politicization of ethnic and linguistic cleavages in the latter decade weakened it. These continental cases differ from the Nordic ones in the absence of growth in the interwar period. Germany is noticeable for its unique electoral shape. It starts early – in fact, it was the earliest – and it grows rapidly. By the end of World War I, it has reached an electoral size of about 45%, which is similar to that of the Nordic countries of the ﬁrst group. However, the German left then enters a long period of decline and stagnation that lasts for almost four decades. It starts to grow again only at the end of the 1950s. The peak reached at the beginning of the 1970s is practically the same level as that reached in 1918. The four phases of European development discussed in connection with Figure 2.2 were therefore the result of quite different national develop-

66

THE POLITICAL MOBILIZATION OF THE LEFT

ments. In terms of general shapes of national development, and irrespective of the general size of the left and the precise dates at which variations occurred in different countries, we can distinguish four patterns. The ﬁrst is a pattern of rapid growth, a high peak, and an equally accentuated decline, which is epitomized by the United Kingdom but which also characterizes Finland, Sweden, and Belgium quite well. The second is a pattern of less rapid but continuous growth, typical of Italy and also applying to Austria. The third is a more complex pattern, characterizing those countries that experience a period of decline located between two peaks: the typical sine curve of France, but also of Germany and Finland. The fourth and ﬁnal pattern is that of a growth phase followed by a long stagnation period, a pattern epitomized by Switzerland but also applying to the Netherlands, Norway, and Denmark. After World War II, these lefts have an electoral ceiling around which they oscillate, with no consistent pattern of growth or decline. On the whole, then, during the postwar period, the European left was generally incapable of improving its level of support among the various national electorates. The increase in support between the pre- and postwar averages was achieved only in the immediate aftermath of the war. Later on, with the exception of Germany and Italy, these increases become marginal. Therefore, in the second half of the 1940s, the European left reached national peaks that later proved difﬁcult to exceed, and it crossed the 50% threshold in only six countries during the entire period: Norway, Sweden, Denmark, Finland, Austria, and France. Even among these countries, it is only in the ﬁrst two that this phenomenon appears to be a lasting achievement. In the others, it is a more occasional result. In Norway, the left won more than 50% of the votes between l945 and l973, while in Sweden this occurred between l936 and l956 and again during the 1960s. In the United Kingdom, the left approached this crucial threshold on several occasions. In France and Italy, this has occurred only in recent years.

IDEOLOGICAL ORIENTATION AND RADICALISM The European variance in the labor response also includes the predominant ideological orientation of the socialist movements. Slightly more complicated is the possible relationship between radicalism and organizational cohesion. Therefore, some prior clariﬁcation of the relationship between ideological orientation, radicalism, and organizational cohesion is required, as

THE EUROPEAN LEFT

67

the three phenomena are sometimes inextricably entwined. To approach the question via a general deﬁnition of ‘‘radicalism’’ is inadequate because such deﬁnitions are in general too broad and rarely useful enough for a historical comparative analysis of several cases.2 For this reason, the usual strategy used to identify radicalism refers to three phenomena, that are taken as indicators: 1. The spread or strengthening of a particular ideological version of socialist thought identiﬁed as ‘‘radical’’ or more radical than others. 2. The presence, intensity, and extension of specific behavioral patterns regarded as radical – for instance, the frequency, intensity, and adversarial nature of strike activities; the frequency of speciﬁc working-class protest activities, confrontations with the forces of order; and revolts and attempted revolutionary actions. In short, the extent to which working-class collective behavior was militant and/or aggressive. 3. The organizational fragmentation of the socialist movement and, in particular, the strength of specific organizational forms of the socialist movement. The inability of different groups to live together within broad, organizationally uniﬁed movements can be regarded as a sign of internal tension and of radical ideological and organizational strife. In particular for the post–World War II period, the strength of the communist organizations is considered to be the main indicator of a radical orientation. The degree to which socialism and working-class politics are moderate/ radical is usually identiﬁed with one of these three aspects, which are not, however, necessarily associated with one another. Finally, they do not coincide historically, as each aspect seems to be particularly important or prominent in a different period. Let me clarify these three points. 2

R. C. Tucker (1967), for instance, deﬁnes radicalism by making reference to ‘‘attitudes or demands which do imply, request or expect basic and fundamental changes in societal structure.’’ ‘‘Attitudes’’ and ‘‘demands’’ are not the same thing. Radical attitudes in the bulk of the rank and ﬁle may be matched by more moderate demands coming from the organizations and the political leadership; alternatively, radical leadership may be nonrepresentative of rank-and-ﬁle basic attitude systems. Second, why should radicalism be expressed by or through demands of ‘‘societal’’ change? Radical demands can also concern the state and, more generally, other political aspects. Even if we admit that ‘‘attitudes’’ and ‘‘demands’’ match and that ‘‘societal’’ includes both socioeconomic and political institutional demands, what is ‘‘basic and fundamental change’’? How can one state that there was more of such demand here than there, now than before? A general deﬁnition, even if good, does not serve my purpose because it is not sufﬁciently denotative to help identify empirical historical variance among cases.

68

THE POLITICAL MOBILIZATION OF THE LEFT

To concentrate attention on the spread and importance of the ideological system guiding the action of the organized part of the socialist movement may focus too heavily on elite conceptualizations and doctrinal debates; it may give too much weight to minor intellectual groups or party bureaucracy; and it may privilege a coherence of thinking and intellectual elaboration on actual class action. Alternatively, to concentrate mainly on mass collective action poses the problem that the radicalization of forms of action may be unrelated to radical goals and strategies. In other words, radical forms of class action do not necessarily indicate radical political socialism. To identify radicalism with the organizational and electoral success of communism is to gainsay too many of the early preconditions for such success.3 The three aspects do not necessarily covary. Thus, orthodox revolutionary Marxism may combine with the quiescent collective behavior of the socialist movement, just as strong communist organizations may resort to radical collective action regarding events and concerns that are outside the internal dynamics of the national labor movement, and strong anarchist or syndicalist wings may shatter the labor movement without being able to consolidate important political or corporate organisations. The three aspects are historically separate. If the question of socialist radicalism is raised for the early phase of the consolidation of socialist movements before World War I, the answer to it is likely to emphasize the dominant ideological orientation of the movement. Radicalism was probably associated with the spread of anarchism or revolutionary syndicalism, or even of orthodox Marxism, as opposed to other, more moderate forms or socialist traditions. During the revolutionary post–World War I European crisis, radicalism was mainly identiﬁed with the shattering of the traditional socialist political and union organizations by new forms of collective action, by the intensity of the social conﬂict, and by the existence of the threat of social revolution; that is, it was mainly identiﬁed by a pattern of radical class action. The 1920s mark the beginning of the process of organizational fragmentation of the socialist movement after many decades and efforts at political centralization and uniﬁcation. From the 1930s on and after World War II, the level of ‘‘domestication’’ was judged in reference to the strength of the communist component in each movement and, more generally, to the extent to which the socialist move3

The speciﬁc identiﬁcation of communism with radicalism could be subsumed under point (1), that is, communism could be regarded as the spread of a speciﬁc ideological current (Leninist doctrines). However, it is better to regard this as a separate case. This is so because the essence of communist radicalism is seen – rightly or wrongly – in its organizational expression, as an organizational weapon.

THE EUROPEAN LEFT

69

ment was organizationally uniﬁed or divided. Finally, it is important not to characterize radicalism as a secular process according to which the situation crystallizing in the 1950s and 1960s was the ﬁnal result of a long history whose roots had existed from the beginning. In fact, the countries that in the long run are considered the homes of moderate reformist socialism may have had radical labor movements before World War I or in the interwar period. For this reason, I have decided to organize my description of the internal dynamics of the European left by separating these three aspects into the temporal phases in which they predominated or were the most important feature. I will therefore deal with 1. The early predominant ideological orientation of the socialist movement, before World War I (when the movement was organizationally uniﬁed or tended toward national political uniﬁcation); 2. The phase of collective class action radicalization between the 1910s and the 1920s – the result of the war and revolutionary crisis; 3. The subsequent process of organizational fragmentation of the socialist movement, whose most conspicuous element is the communist split and eventual consolidation. Early ideological orientation, intermediate militant collective action, and ﬁnal organizational (communist) fragmentation form speciﬁc combinations in each country. Therefore, the poles of the empirical variation can be identiﬁed as, on the one hand, an early moderate orientation, lack of revolutionary crisis and mobilization, and absence of organizational fragmentation and communist inroads, and, on the other hand, early radical ideological traits, profound revolutionary crisis, and deep organizational splits. Using this framework, it is possible to remold the inadequate unidimensional opposition between ‘‘radical’’ and ‘‘moderate’’ while maintaining a focus on the ideological, behavioral, and organizational faces of radicalism and, at the same time, detecting historical empirical variance among the cases. The following two sections are devoted to the ﬁrst two aspects; organizational fragmentation and communist strength will be dealt with separately in the ﬁnal part of the chapter. EARLY IDEOLOGICAL ORIENTATION The starting point for the spread of the ideological tradition that will be taken to represent more or less radical idea systems is the contrast among the United Kingdom, Germany, and France. These three countries are

70

THE POLITICAL MOBILIZATION OF THE LEFT

most representative of certain ideological traditions: moderate trade unionism in the United Kingdom, orthodox revolutionary Marxism in Germany, and a combination of the latter with important anarcho-syndicalist components in France. These variants have also been examined from a normative point of view: While the moderate reformer underlines the realism, pragmatism, and humanitarianism of British Fabianism against the radical, rationalistic continental schemes, the Marxist thinker contrasts the theoretical quality of German socialism with the ‘‘adventurousness’’ and lack of coherence of the French movement and the vulgar pragmatism of the British one. At the same time, the eclectic French socialist criticizes both trade unionism and ‘‘bureaucratic state socialism.’’ The experiences of other countries in the historical phase up to the time of the Russian Revolution are interpreted within this triangle of ideological genealogy. Trade Unionism The strongest opposition is often thought to be that of the United Kingdom versus the Continent4 and the connected question of the absence of any Marxist penetration in the British Isles. Without question, of all the theories and ideas that forged British socialism, Marxism played the smallest role. In fact, the initial cultural characterization of the British labor movement was permeated by the liberal culture of its early leaders and organizers. Liberal radicalism was traditionally associated with religious nonconformism, and in the nineteenth century such denominational sects constituted the cultural basis of the labor movement, just as Methodism and its theology of consolation were widespread among the working class. Equally important was the romantic tradition of social criticism, with its strong moral and ethical bases for social protest and change, and the consequent idea that moral reforms were as important as, if not more important than, social ones, and that socialism was fundamentally a process of enlightenment for both workers and employers. At the same time, the origins of the labor tradition of the early trade unions were Owenite and utilitarian, concentrating on reforms within the existing system and redistribution through economic action, that is, by distinguishing between 4

The recent work by S. Berger (1994: 248) denies this opposition and argues that the differences between the British and German parties were more a matter of degree than of substance. He concludes that ‘‘the hypothesis that ideological and organisational differences made the Labour party an altogether different type of working-class party from the one to which the SPD belonged can not easily be maintained.’’ This two-country comparison is conducted, however, without an overall comparative framework of reference, and it is based on examples that are selectively drawn from various sources and deliberately emphasize similarities and minimize differences.

THE EUROPEAN LEFT

71

fair and unfair competition and good and bad capitalists. These old cultural traditions were unfavorable to the basic tenets of Marxist theory, in particular to the idea of class warfare and the recognition that political and economic power were profoundly related to the question of authority in industry and the state.5 The fundamental belief in the neutrality of the state machinery caused mainstream laborism to sidestep this feature of continental socialism and Marxism in particular. Attempts to introduce more politically radical versions of socialist thinking occurred only outside the world of the trade unions. The large middle-class, intellectual composition of the early socialist societies and socialists alike was regarded with suspicion by the unions. Thus, the 1880 Social Democratic Federation of Marxist inspiration was formed by a few intellectuals, whose program of integrating and assimilating the various streams of laborite political thought and movements was a total failure. Any deliberate socialist ideology/program was systematically rebuffed by the unions, and even the formation of a more moderate Independent Labour Party outside a trade union context was unsuccessful (as evidenced by its poor 1895 electoral results). Radicalization phases were not, however, unknown in the British labor movement. In the 1880s, the original concentration of British trade unionism within the more prosperous and skilled segments of the working class, and its main concern with salary and working hours, was challenged by the rise of ‘‘new unionism’’ and its decisive push toward the unionization of unskilled workers. The conservative character of British trade unionism was shattered but ultimately unaffected by this crucial step (which in other continental countries occurred much later, some time after the Russian Revolution). Rather than causing radicalization along Marxist lines, it produced the two basic streams of British pre–World War I socialism: Fabianism and guild socialism. Fabianism, notwithstanding its intellectual attitudes and origins, became the dominant theoretical reference for British laborism; in fact, it was more important than the actual number of Fabians may suggest. This was because it concentrated on a theory of political action and programmatic strategy that was compatible with the cultural and organizational traditions of British labor. The Fabian essays of 18896 – the same year of the foundation on the Continent of the Second International, dominated by Marxist-oriented parties – were written by various thinkers and incorporated various streams of thought, but virtually none of the themes so 5 6

See Willis (1977: 417–419) and McKibbin (1984: 297–331). Webb, Shaw, Wallace, Oliver, Clarke, Besant, and Bland (1962).

72

THE POLITICAL MOBILIZATION OF THE LEFT

typical of Marxism and anarchism were included. The essential tenet was that socialism was simply the fulﬁllment of democracy in an industrial society; democratic procedures were taken for granted; the transformation was aimed at modifying the attitudes of the large majority of people and therefore had to be gradual; and it was through the state that this would be achieved. In a nutshell, the onset of socialism was constitutional and peaceful. Guild socialism represented the radicalization of some sectors of the working-class movement before World War I. Its very name linked it to the original anti-industrialism protest, and its anticapitalism was coupled with an antistate attitude, aimed at transforming the state into a coalition of guilds, unions, and professions, each running its own affairs but being held together by some sort of nonstate, which some wanted to call ‘‘The Commune’’ and others ‘‘The Industrial Guild Congress.’’ It has been observed that guild socialism echoed elements of the revolutionary radicalization on the Continent, and it is certain that the movement had several points of contact and correspondence with the continental syndicalism of that time, particularly in France, Spain, and Italy.7 However, while on the continent the end of World War I brought about a further radicalization of the labor movement, in the United Kingdom it saw Fabianism predominate over guild socialism. Unquestionably, British labor developed a local ideological tradition, which Marxism and anarcho-syndicalism inﬂuenced only minimally. Orthodox Marxism When discussion focuses on British versus continental opposition, it is frequently the British versus German opposition that is meant. Germany is regarded as the homeland of orthodox Marxism and as the ‘‘quality side’’ of socialist theory. Thus, the socialist ideological orientation is mirrored by its programmatic development. The 1869 Eisenach Program of the Social Democratic Worker Party stressed the democratic tradition of the then dominant Saxonian progressives as well as Marxist internationalism. The 1875 Gotha Program of Uniﬁcation was criticized by Marx and Engels for its Lassallian emphasis on welfare state measures and its progressive stress on the democratization of state and society. Due to the still profound inﬂuence of Lassalle in the trade union movement, these early socialist programs favored legalism and a positive attitude to social legislation, although they were more sophisticated than British trade unionism. This tradition did not, however, prove strong enough to guarantee the predom7

See Holton (1985).

THE EUROPEAN LEFT

73

inance of ministerial socialism, with its corollary of the acceptance of the state as a positive force and actor. The 1891 Erfurt Program mirrors the extent to which Marxism had by then become the exclusive theoretical basis and inspiration of the newly named German Socialist Party (SPD). Virtually all the previously mentioned Lassallian elements were eliminated by the new party program. The ideological penetration of Marxism, against the earlier Lassallian and progressive orientation of cooperation with the state while trying to reform it, strengthened during the underground period of repression and suppression. Marxism took over the party and neutralized the unions’ positions in the twelve years of repression between 1878 and 1890. Kautsky and Bernstein were the chief popularizers of Marxism in the following period. Their efforts were built on and backed by a socialist press that was impressive for its quantity, diversity, and circulation – a situation unmatched in any other country with the possible exception of Austria (and, to a much lesser extent, Denmark).8 If the German labor movement was the homeland of Marxist orthodoxy, it should not be forgotten that its internal intellectual and doctrinaire debate was also so rich that, in addition, it became the homeland of the most coherent and consequent reformist version of Marxism itself. It was Bernstein who represented the alternative political strategy, advocating, ﬁrst, the need to attract middle-class support in the political struggle against the authoritarian imperial government; second, a more realistic (and not antagonistic) policy on the agrarian question and the peasant problem; and third, the possibility of playing a more effective political role in German policy making and legislation even before the deﬁnite political victory of the labor movement.9 Bernstein’s analysis of capitalist development was rejected by the center and the left of the SPD. But while the revisionist theory was rebuffed and the orthodox theory was kept pure, the practical political guidelines of the party met a different fate. In Germany and in many other European countries inﬂuenced by the German debate, the Kautsky–Bernstein dispute opened up two possibilities. The theory could be kept pure and orthodox, in which case it would be necessary to ﬁnd new and alternative explanations for the socioeconomic 8

9

On the richness of the Social Democratic press and the immense educational and cultural policy carried over by it, see Hall (1977), whose book provides useful information about the diffusion of different journals and the amount of internal theoretical debate. The author underlines more than once the exceptional number and diffusion of this press in comparison with those of other neighboring countries. Bernstein (1899). For the analysis of the situation that led to his political conclusions, see Colletti (1972: 61–147).

74

THE POLITICAL MOBILIZATION OF THE LEFT

development identiﬁed by Bernstein or to live with a progressively schizophrenic split between pure theory and inevitably corrupted and compromised practice. Alternatively, the theory could be diluted, softened, and made less important, so that the compromise political tactics would not seem a corrupted and betraying behavior. In this book, I argue that the choice effectively made by the socialist movements in continental countries was largely determined by the response of the political environment of the socialist movement. In Germany, one could interpret the ideological debate within the SPD before World War I as a series of theoretical equilibriums developed in order to save orthodoxy. To offer alternative explanations to trends that Bernstein had interpreted in revisionist terms, Kautsky introduced the ‘‘relative’’ element in the theory of working-class impoverishment and argued that temporary checks compensated for the general tendency toward a worsening of the proletariat’s conditions.10 Luxemburg and the left wing of the party came up with the idea that the capitalist breakdown was not materializing because of the exploitation of colonies, that is, of noncapitalist parts of the world.11 These new interpretations were compatible with the orthodox theory, but they maximized contradiction and tensions between theory and practice. In conclusion, at least until World War I, in the German socialist movement of Marxist orthodoxy was defended more successfully against moderate revisionism in theory than it was in practice. Socialist Eclecticism In France, neither a trade unionist tradition nor orthodox Marxism seems to represent adequately the ideological temperament of the labor movement. In fact, French socialism presents a number of paradoxes. Its political situation and tradition in the second half of the nineteenth century were the most favorable to socialism development in Western Europe. France had a long tradition of working-class militancy stemming from its long revolutionary tradition. Some of the key reference points for a socialist ideology were identiﬁed with French history of the late eighteenth and nineteenth centuries – a history that offered to the birth of socialism a cultural and ideological legitimacy that was absent in most of the other Western European countries. The philosophy of the Enlightenment and Cartesian rationalism were elements of the national culture claimed by the left as its heritage and were the precursor of the founding values of the 10 11

Kautsky (1899). Rosa Luxemburg replied immediately to Bernstein’s analysis in Luxemburg (1899). For the development of her thinking later on, see the collection of her writings in Luxemburg (1963).

THE EUROPEAN LEFT

75

socialist anthropological vision. The 1789 Revolution12 was interpreted in the nineteenth century as a popular movement, and the inﬂuence over French political culture of works like Lamartine’s Histoire des Girondins13 and Michelet’s Histoire de la re´volution14 contributed to the setting up of a republican tradition as a set of values and political principles (antiauthoritarianism, anticlericalism) that came close to those of socialism. At the moment of consolidation of the socialist movement, the country had already gone through three republics. The brief workers’ governments in Paris in 1848, 1870, and 1871 were republican founding experiences that were largely assimilated as positive. Ideas on popular rule were therefore not only theoretical, but had been tried and tested on several occasions. Louis Blanc, Auguste Blanqui, Etienne Cabet, and Pierre-Joseph Proudhon were legendary ﬁgures who had spent their lives in the labor movement. Thus, contrary to the cultural isolation of its German counterpart, French socialism was both culturally and politically well integrated into a legitimate and powerful republican and democratic tradition. However, these conditions created a very complicated ideological scene; a class movement was created that was as rich in cultural traditions and political experience as it was organizationally weak and ideologically eclectic, in which none of the main streams of socialist ideology ever managed to integrate the entire movement. The early socialist thinking of the schools of Saint Simon (1760–1825), Fourier (1772–1837), Blanqui (1805–1881), and Proudhon (1809–1865) shared an incapacity to envisage a reform stemming from mass-organized collective action. In this sense, these thinkers and agitators were all deeply rooted in the mood of the Enlightenment.15 In reacting against the bureaucratization of modern industry and the modern state, none of them saw, appreciated, or anticipated modern collective action, and thus they provided neither a rationale nor a guide for it. An organized collective movement for the realization of ideas was alien to them and – to a large extent – would remain alien to the entire French political culture. Any success that Marxism in France had16 was due largely to Jules Guesde and his review, L’Egalite´ (1877), and was originally conﬁned to 12

13 14 15

16

Of which the Russian Revolution was seen as a derivative. In his speech at the Congress of Tours in 1920, Cachin considered Lenin a true disciple of the lumie`res. De La Martine (1895). Michelet (1847–1853). That their hope for the reforming philosopher, their search for philanthropists or technocratic state engineering, their belief in insurrection, conspiracy, or a coup d’e´tat had in common the idea that the solution had to be found in favor of or in behalf of the oppressed, more than by the oppressed themselves, is underlined by Ulam (1979: 98– 99). See Dommanget (1969) and Pre´lot (1939).

76

THE POLITICAL MOBILIZATION OF THE LEFT

this section of the socialist movement. The Guesdesists,17 although making reference to Marx, maintained a strong anarchist and Blanquist imprint, and the doctrinal formation of the leaders of the Parti Ouvrier Franc¸ais was the object of frequent complaints by Engels and Marx.18 But Marxism in its Guesdesist French variant met formidable obstacles along its path to development. At the political level, the Marxist wing was unable to coordinate, and even less to hegemonize, the extremely ideologically, geographically, and organizationally fragmented socialist tradition.19 The attempts to differentiate French socialism from the German experience – consciously pursued by Lucien Herr, by the municipal socialism of Paul Brousse, and by the intellectual humanitarian and in some ways bourgeois political sensitivity of Jaure`s and later by Blum’s socialism – represented powerful barriers for a group with very weak inroads into the trade unions. Marxism was perceived as a German doctrine in a country where the memories of the 1870 war had left a persistently anti-German attitude, which was reinforced by the 1914 SPD vote for the war credits.20 One does not need to accept the radical thesis of Robert Aron, who sees Marxism as a foreign inﬂuence with no national roots in France and whose only effect was to ruin the autonomous local traditions,21 to recognize the deep anti-German feeling of the nationalist tradition of the republican left. In 1882, Brousse criticized Guesde, arguing that ‘‘Les ultramontains ne peuvent pas obeir a` la loi de leur pays parce que leur chef est a` Rome. Les Marxistes peuvent pas obeir aux de´cisions du Parti et a` ses Congre`s parce que leur veritable chef est a` Londres. . . .’’22 Blum also saw Guesde and his ideas as the French representative of ‘‘German collectivism.’’ Many other examples of mistrust in the German origins of Marxism can be found in the early 17

18

19

20

21

22

For the role of the Guesdesists in the formation of the Socialist Party and in the introduction of Marxism, see Perrot (1967: 701–710). Marx’s Das Kapital was originally translated into French by Jules Roy and was published in 1875. It was not very successful. Far more inﬂuential was an abridged version published in 1876 by Gabriel Deville. The correspondence between Engels and Lafargue, which is a crucial source for the reception of Marxism in France, is full of such complaints; Correspondance (1956). Beyond tactical and short-term issues, the fundamental criticism was that Guesde’s materialism resembled that of the eighteenth century and had no sense of dialectics. The piece devoted to socialism by Zeldin (1979: 361–423) clariﬁes that this fragmentation was geographical and social before being ideological and organizational. J. Droz’s (1973) interpretation attributes to the difﬁdence originating in 1870 a primordial role in determining the difﬁcult reception of Marxism and, more broadly, the difﬁcult relationship between the German SPD and French socialism.. Marxism is regarded as a foreign graft by Aron (1971). Kriegel (1969) interprets communism in the same vein. Speech of P. Brousse at the Congress of St. Etienne, where the split between Guesdesists and Broussists occurred. Quoted in Lefranc (1963: 47).

THE EUROPEAN LEFT

77

history of French socialism.23 In the end, the adoption at the political level of more systematic Marxist stands as ofﬁcial party doctrine was always tempered by the inﬂuence of many other autonomous local traditions and leadership and remained therefore a surface phenomenon with respect to the broader assimilation of the socialist intelligentsia with radicaldemocratic values and beliefs. A further element that prevented Marxism from making signiﬁcant inroads into France was more overtly political. Initially, Jaure`s and then Blum both perceptively pointed out that Marxism, and in general German socialist doctrine, was the result of the forced political impuissance of the German SPD in imperial Germany and that the political conditions of France and therefore the political problems of French socialism were fundamentally different. At the trade union level, the situation was even more difﬁcult. The Guedesists did not manage to penetrate the trade unions, which, indeed, until 1914 constantly afﬁrmed their independence from them. The French trade unions were weaker in organizational terms than their German and British counterparts; they remained non-Marxist, like the British; but unlike the British and the Germans, they placed themselves to the left of the political movement, in the direction of syndicalism. The self-reliance of workers on their direct action qua associations and the mistrust of the professionalized political branch became important features of the movement. The syndicalist tradition, with its emphasis on direct class struggle, 23

For instance, the explicit accusation in the writings and political positions of Lucian Herr that the ‘‘solid’’ and ‘‘hard-working’’ Germans received the ideas from the main French socialist thinkers and had only to systematize them; or the explicit claim of French leadership in the political ﬁeld mentioned by Blum: ‘‘Ce n’est pas le prole´tariat allemand qui a conquis le suffrage universel. Il l’a rec¸u d’en haut’’ (Conference of Amsterdam of the Second International, 1904, cited in Lefranc [1963 :119]; or the comment of Charles Andler on the 1848 Communist Manifesto: ‘‘La the`se du petit livre e´dite´ par la Librairie, c’est que, dans Marx, ce qu’il avait de grand c’est la tradition socialiste franc¸aise et anglaise et qu’aupre`s d’elle l’originalite´ de Marx apparaissait moins e´crasante’’ (Andler (1932: 139). The review Bibliothe`que socialiste was launched in those years as an instrument of research of an original socialist and, above all, French elaboration. Blum, Andler, and other intellectuals of the group close to Herr and Jaure`s published in it several articles and studies whose spirit and content was decidedly antiGerman and anti-Marxist. This attitude had been fully perceived by Marx himself, as is clear from a letter to Sorge of November 5, 1880; after that, in spring 1881, he elaborated, with Engels and Guesde (with whom he had entered into epistolary contact in 1878), a program in the form of a manifesto published in the Egalite´ that year: ‘‘Je n’ai pas besoin de te dire (car tu connais le chauvinisme des franc¸ais) que les ﬁcelles secre`tes a` l’aide desquelles leur leaders (Guesde and Malon) ont e´te´ mis en mouvement doivent rester entre nous. Il n’en faut rien parler. Quant on veut agir pour MM les franc¸ais il faut le faire anonymement pour ne pas froisser leur sentiment national’’; letter of Marx to Sorge, November 5, 1880, reprinted in Lefranc (1963: 43).

78

THE POLITICAL MOBILIZATION OF THE LEFT

on trade unions as a direct political force, on direct action and the general strike, its polemical attitude toward middle-class and intellectual socialism and socialists, and its antipartitism and antiparliamentarism24 contributed heavily to preventing the integration of the socialist corporate and political movements. This combination of factors shaped the French ideological tradition. It was characterized primarily by the unique mixture of elements of Jacobinism, utopianism, anarcho-syndicalism, republicanism, state-technocratic socialism, and Marxism. This unusual mix was reﬂected in the ﬁgure of Jaure`s, who reconciled the bourgeois and working-class traditions and invented ‘‘that sort of sensibility which is typical of the French left,’’25 nurtured of the memories of the Great Revolution, the ideas of utopian thinkers, and those of republican humanists. French socialism was a broad political movement with great cultural legitimacy, but it was organizationally weak and ideologically divided. A European Map of Early Ideological Orientation These three paradigmatic cases can serve as a point of reference for the characterization of the other cases. Sometimes they directly inﬂuenced neighboring countries. For instance, the United Kingdom successfully screened Ireland from the ideological turmoil of the Continent in the period in which Irish trade unions were effectively part of the British trade union movement until 1894. Similarly, the experiences of Italy and Spain are associated with that of France for the mix of Marxism among political and intellectual elites and of revolutionary syndicalism in the unions and the working-class movement. Indeed, the origins of the socialist movement in Italy and its peculiarities cannot be understood without considering the role played by Michele Bakunin, who arrived in Italy in 1864. Bakunin’s inﬂuence was decisive in the early orientation toward radical conceptions of the revolutionary action of popular and workers’ circles in the peninsula. Much of its prestige, particularly in southern Italy, was also due to the fact that it appeared as the symbol of that glorious, and at the same time mysterious, International under whose colors the Parisian communards had fought.26 At the beginning of the 1870s, ‘‘anarchism, socialism, and inter24

25

26

For a detailed presentation of the position of French syndicalists, see Lorwin (1954: 30– 36). See also Bergounioux and Manin (1979: 66–71). Bergounioux and Manin (1979: 66). It was customary for historians to criticize the theoretical weakness of French socialism. See in particular Ligou (1962) and Judt (1976: 71–97). Particularly when Mazzini took an explicitly negative attitude vis-a`-vis the Commune,

THE EUROPEAN LEFT

79

nationalism were in Italy, if not synonymous, at least equivalent, and the name of Bakunin was far more famous than that of Marx. It was under this inﬂuence that the Italian sections of the International multiplied.’’27 Malefakis goes far as to argue that the strength of revolutionary syndicalism in Italy (and even more so in Spain) was greater than in France, and that this has been overshadowed by the longevity of French syndicalism, its spread from there to the rest of Europe, its identiﬁcation with the writing of George Sorel, and its long-standing inﬂuence on the largest French trade union organization, the General Confederation of Workers (CGT).28 Not all scholars share these conclusions, but all agree on the fundamental similarity of these socialist experiences.29 The assimilation of the French and Italian experiences is not limited to the role played by revolutionary syndicalism until after World War I, but extends to the unusual – in comparative terms – internal ideological differentiation of the socialist movement. While in Romagna socialism was inserted over a strong republicanism with its patriotic heritage, in Lombardy a worker was ashamed to call himself a republican, and the opposition between Marxist socialism and the Mazzinian democratic vision and ideals was intense.30 A considerable amount of patriotism was a typical feature that originated from the Risorgimento tradition, and it explains in part the yielding of certain socialist groups to interventionist stands that were taken in 1914. A ﬁnal afﬁnity with the early French socialist tradition was that its character was more popular than working class.31 The diffusion of Marxist ideas, programmatic lines, and organizational principles was difﬁcult within the socialist party. Marxism in Italy was originally strongly mixed with positivism, and its diffusion and revision in relation to the Italian situation were largely the result of the intellectuals who were external to the party itself, such as Labriola and Mondolfo.32

27 28 29 30 31 32

which increased the prestige of Bakunin. Little was known of the violent polemics that were taking place in those years between Bakunin’s and Marx’s followers within the First International or of the ambivalent attitude that the General Council of the International had initially expressed toward the Commune. See Cole (1954: 176–236). Procacci (1970: 395) Malefakis (1974: 7–8). Lorwin (1954: 36), Judt (1979: 285), Andreasi (1981), Scho¨ttler (1986: 419–476). Cf. Morandi (1978 (1st ed. 1945): 40). Morandi (1978: 39–43). On Italian intellectual Marxism and, in particular, on the role played by Labriola in revising Marxism toward a less fatalistic and more voluntaristic vision of the development of the revolutionary conditions, see Santarelli (1964). For more in general on the Italian Marxist intellectual debate, see Croce (1938). On the poor circulation of the Marxists’ basic writings, see Bosio (1951: 268–284, 444–477).

80

THE POLITICAL MOBILIZATION OF THE LEFT

Germany, the quality side of orthodox Marxism, is regarded as the source of the spread of this ideological version of socialism to the Netherlands and Denmark in the north (and through Denmark to the other Scandinavian countries) and Austria and Switzerland in the south. Doubtless, what is called the introduction of Marxism was actually the introduction of Engels’s version – or vision – of Marxism after 1883.33 The correspondence of Engels, and later that of other German leaders, was more politically inﬂuential than were the original writings of Marx, most of which were inaccessible outside Germany. Even Marx’s famous critique of the Gotha’s 1875 SPD Program was actually unknown to German social democrats for almost sixteen years, until it was published in 1891. In the previously mentioned countries, Engels’s Marxism penetrated profoundly at the ideological level, but – with the possible exception of Belgium and the Netherlands, where until the end of the century some syndicalist traditions remained important in the unions and workers’ societies – in none of them did Marxism have to contend with strong national variants of socialism. The question in these cases is not how Marxism interacted with other ideological streams – amalgamating with them or being rebuffed by them – but, rather, how intensely Marxism orthodoxy permeated these movements and persisted over the forces of ideological or practical revisionism. In a recent systematic comparison of the reception of Marxism in several countries, Steenson has analyzed in detail Engels’s correspondence with foreign party leaders after Marx’s death. Instead of relying on theoretical works (as do most other national or international studies of socialism), he has concentrated on party press and party programs in order to judge the Marxist orthodoxy of several socialist movements, in particular the German, Austrian, French, and Italian ones.34 The conclusions of his work show a clear division between the more eclectic and less orthodox stands of the French and Italian movements versus the orthodox ones of the German and Austrian movements. The Austrian movement was not only Marxist, radical, and orthodox, but also theoretically rich and fertile. The full range of the ideological character and original elements of what is called ‘‘Austro-Marxism,’’ represented by the writings and positions of Otto Bauer, were displayed only after World War I, in the conﬂict between Leninism and Second International socialism. The theoretical work of Otto Bauer, and in general of the Austrian Marxists, interpreted the Russian Revolution and the Bolshevik ideology as a phase of the bourgeois revo33 34

Steenson (1991: 3–46) amply argues and demonstrates this. Ibid.

THE EUROPEAN LEFT

81

lution in Russia and, in particular, as a radical and utopian ideology of the plebeian phases of bourgeois revolutions. Their strategy for the revolutionary forces of the advanced countries centered on the maintenance of democratic political forms, the rejection of the nationalization of the means of production35 as a strategy that would be dangerous for democracy,36 and the idea of an economic democracy at the level of the local administration of each enterprise. It should be added that the Austrian socialists also possessed an unusual ideological-organizational encapsulation and self-reliance. Adler gave a digniﬁed cultural deﬁnition of the process of powerlessness and isolation of the movement with the idea of the poetische Politik, that is, a policy that concentrated fundamentally on infusing the movement with a strong symbolic content, introducing a sacred aura around its organization and other activities, and emotionally charging the social militancy and participation in order to foster the members’ identiﬁcation with the organization and with each other.37 No other European socialist movement seems to have succeeded in building such a strong emotional solidarity as the Austrian one. The practical components of this vision were the festivals, celebrations, and political rallies that were staged and the immense effort that was devoted to the development of working-class cultural societies and all sorts of other ancillary organizations. The basic goal was to create a mass psychology of the movement, a strongly distinctive organizational identity of the working class and ‘‘the ghetto mentality.’’ The Austrian movement, therefore, is a case of an orthodox radical Marxist movement, which was doctrinaire and theoretically fertile and in which ideology mattered both for the organization and for political action. In reality, in no other country was Marxist orthodoxy so accentuated or important for organization and action as in Germany and Austria. In the Netherlands, Belgium, Switzerland, and the three Scandinavian coun35

36

37

‘‘Nationalization’’ may not be the best word in this case. What was meant was direct state – i.e., state bureaucracy – ownership of the means of production. On the various meanings of the terms relating to the ‘‘public’’ ownership of the means of production in the socialist ideology and jargon, see Waldbrunner (1952: 13–17), who reviews the utilization and meaning of such terms as ‘‘socialization,’’ ‘‘common ownership,’’ ‘‘public ownership,’’ ‘‘collectivization,’’ ‘‘nationalization,’’ and ‘‘state ownership.’’ The Austrian Marxists and Bauer in particular proposed the socialization of the public economy and, in particular, of big industry. The latter was to be based on expropriation, to be indemniﬁed through a progressive tax hitting the capitalists whose property had not been expropriated. On the programmatic points of Austro-Marxism, see Agnelli (1969). This cultural and organizational policy of the Austrian movement is discussed at length in McGrath (1974).

82

THE POLITICAL MOBILIZATION OF THE LEFT

tries, Marxism, although widely accepted as the dominant intellectual doctrine, never acquired such an important role for political organization or, above all, for political action. Alongside the Marxist formal orthodoxy a revisionist practice developed early on, but the potential contradiction between the two never exploded into a major political crisis. The experiences of these countries, even if they can be regrouped under this broad characterization, should not, however, be viewed as absolutely similar. Three cases seem to be more ideologically radicalized than the others: those of the Netherlands, Denmark, and Sweden. Initially, the Dutch socialist movement was revolutionary in nature, although this aspect was progressively curtailed. Until the beginning of the twentieth century, there persisted within the trade union movement a relatively strong syndicalist orientation that was often in conﬂict with the political wing of the socialist movement. The Social Democratic League, forerunner of the Socialist Party, had basically borrowed its program from Gotha’s SPD program. Later, Erfurt’s program became the basic inspiration for the Social Democratic Workers Party (SDAP), and at the ofﬁcial ideological level the party followed Kautskyan Marxism quite closely. Even as late as 1913, on the issue of socialist participation in bourgeois governments, Kautsky was consulted.38 At the ideological and programmatic level, therefore, Dutch socialism was very internationally oriented and particularly focused on the German SPD experience. At the same time, historians have argued that a patriarchal family ideology was accepted by substantial parts of the socialist movement, while it was, of course, the basis for mobilization on religious-ideological grounds of Catholic and Protestant workers against liberalism and socialism. The socialist attempt to counterbalance or neutralize the conservative hegemony failed when it was confronted with the control exerted by bourgeois and religious organizations (state or private) over virtually all aspects of the socialization process: the press, schools, and so on. The socialist organizational pillarization, in this perspective, was defensive, not offensive. It reinforced acquired loyalties but was unable to expand beyond them. Therefore, in its pre–World War I phase, Dutch socialism can be seen as one of the best examples of the maintenance of an ofﬁcial orthodox Marxist doctrine, only slightly adapted to Dutch conditions, and, at the same time, of an increasingly moderate and pragmatic policy pursued by the party majority. It was not that Kautskism was rejected but, quite simply, that practice was separated from theory within the movement. In the Netherlands, this growing internal tension ﬁnally led to organi38

Ruitenbeek (1955: 25–26), quoted in Irwin (1989: 40).

THE EUROPEAN LEFT

83

zational consequences. At the beginning of the twentieth century, the Marxist stream started to make its voice heard, criticizing the party line for not being Marxist enough. The key issue was indeed party strategy: Predicting a crisis of industrial capitalism, the Marxists argued that until then, the party should limit its support to the relatively small groups of class-conscious workers who were free from conservative religious ideological inﬂuences. Theoretical clarity and education of the proletariat would thus be the short-term goals. This debate was transformed into a power struggle within the party that culminated at the 1909 congress, when the parliamentary party expelled the radical left. The latter formed a Social Democratic Party that was later to become the Dutch Communist Party. This turn of events is quite similar to developments in Italy and, to a lesser extent, in Sweden in the same period. Only in these three countries did the socialist movement split into radicals and moderates before World War I. In Sweden (and Norway), Marxism was introduced via Denmark, and cooperation among socialist movements was institutionalized in the Nordic Labor Congress rather than via the Second International. The inﬂuence of the German socialist debate was expressed mainly in party programmatic statements but was not accompanied by any profound ideological debate or doctrinaire factionalism. However, a radical minority of the Swedish Social Democratic Party (SAP), which had rejected the moderate and reformist strategy of the majority since the 1890s, acquired greater inﬂuence after 1908. This radical wing had identiﬁed itself at the beginning of the century with the Young Socialist League (SUF), which advocated moredirect action, general strikes, and no parliamentarism. In 1908, its two most prominent leaders were expelled from the SAP for these antiparliamentary and pro–general strike activities. As a result, they founded the Swedish Young Socialists Party (SUP), which, being antiparliamentary in attitude, did not contest elections. After the bitter confrontation of 1908– 1909 and the 1909 general strike (the consequence of a series of strikes in response to lockouts and employers’ attempts to lower wages) ended in defeat for both the SAP and the trade union confederation (LO), the Young Socialists managed to push through the foundation of a Swedish Worker’s Confederation (SAC) with a syndicalist orientation. Even in Sweden, therefore, the pre–World War I confrontation between radicals and moderates produced organizational splits, even if they had no parliamentary consequences for the main socialist party. This was the beginning of a long history of minor splits in both the communists and socialist left in Sweden. In Denmark and Norway, the internal ideological strife seems to have

84

THE POLITICAL MOBILIZATION OF THE LEFT

been comparatively less and did not produce important organizational splits before the war. Although Danish party programs were directly inﬂuenced by the German ones, from the beginning they omitted any references to ideological formulas such as the ‘‘iron law of wages,’’ and they included special paragraphs devoted to the agrarian problem. Notwithstanding the early marked internationalism and the intense effort at party publication development (a literature of primarily Marxist orientation),39 theoretical issues and Marxism did not become a major source of internal conﬂict. This was even more true of Norway, where such issues were hardly even a source of debate. The approach of the Norwegian socialists to doctrinal issues was instead inherently pragmatic. If some of the party leaders were familiar with the socialist classics, they were never interested in the ﬁner theoretical differences. Different theoretical traditions were seen as equally important (Lassalle as much as Marx). Although the party was represented at the 1889 Paris Labor Congress and although from 1890 onward the First of May was ﬁrmly established as a day of international socialist demonstration, the Second International was viewed as fairly unimportant as regards other matters, overshadowed as it was by the role of the Nordic Labor Congress, where the Scandinavian socialist parties cooperated. It is interesting to note that these early differences in ideological characterization have tended to persist at the programmatic level. After an intense post–World War I radicalization (see later), Norway was the ﬁrst Scandinavian country to overtly deradicalize its program in 1930. Denmark was slower in programmatic adaptation. Its 1923 program was still Marxist in overtone and remained basically unmodiﬁed until 1961, when a new program of total revision and radical deideologization was adopted. Sweden has had the most continuous programmatic development. Successive programs have only slightly adapted ofﬁcial party positions. Indeed, the 1960 and 1975 programs still preserved elements of the original Marxist ideology (e.g., the ‘‘third stage of democratic development’’) and a strong collectivist conception of economic and political democracy. Finland is a difﬁcult case to classify in a European perspective. In 1903, the Finnish Socialist Party, like so many others, adopted a program that was directly borrowed from Erfurt’s German program. However, the impact of Marxist ideas on the development of the party seems to have been minimal before 1905. Socialist literature, particularly Marxist literature, was unavailable in Finnish. It seems that Finnish socialist theory was 39

See Callesen (1990: 142–143) for documentation of this effort to educate the working classes through party publication activities, in many ways similar to the efforts in this ﬁeld typical of the German and Austrian parties.

THE EUROPEAN LEFT

85

relatively rudimentary, and not only did external ideological inﬂuences have little impact on it, but few internal ideological debates were launched. However, the party stand on orthodox ground was helped by the fact that in the illiberal climate of Russian domination, the party did not have to face the complex strategic choices and practical decisions that a more liberal political context posed. From this point of view, its position until the end of World War I and independence was similar to that of the German and Austrian socialist movements. Finally, Switzerland and Belgium are the two clearest cases of socialist movements that formally adopted Marxist elements in their programmatic proﬁle and yet demonstrated a total disjunction between these proﬁles and moderate and reformist political practice. In Switzerland, prior to the sudden and short period of radicalization at the end of World War I, it is hard to identify any signiﬁcant radical tendencies, in part because the extreme territorial and cultural fragmentation of the country made it difﬁcult for these groups to forge any solid organizational links. In Belgium, the early ideological matrix of the socialist movement was very complex. The fact is that from the beginning the Belgian Socialist Party was a conglomerate of associations of every kind: professional, economic (cooperatives), social (mutualistic and insurance organizations), cultural (schools, newspapers, libraries, conference centers, local clubs for workers education, e´coles ouvrie`res superieures), medical and pharmaceutical (chemists, clinics, nursing homes, kindergartens, assistance to women), political (workers’ leagues), sports, travel, dramatic art, ‘‘Friends of Nature,’’ and social groups (women, retired workers, young-socialists guards, ex-soldiers, peasants, etc.). Moreover, the need to resist the interference of the state, then in the hands of the Catholics, obliged the early socialists to build a dense network of strongholds at the communal level, where they were in the majority. Communes were traditionally important, numerous, and autonomous entities, and socialists used them to insulate themselves from state inﬂuences. It was hard to give ideological coherence to such an organizationally complex world. Its ideological cement seemed to be that of anticlericalism, free-thinking, and Freemasonry, given the dominant role of the church and Catholics during the nineteenth century. The party concentrated most on winning political reforms – not only universal and equal suffrage, but also the battle for cremation and civic burial, for free and Catholic-emancipated schools – battles that were eventually hegemonized by the rather ‘‘Voltaireanian’’ Belgian bourgeoisie.40 The party, in a nutshell, was almost en40

The trust in the general strike that characterized Belgian socialists concerned the politi-

86

THE POLITICAL MOBILIZATION OF THE LEFT

tirely revisionist and nondoctrinaire, so much so as to elicit the strong criticism of Karl Kautsky: ‘‘They have nothing to revise, for they have no theory.’’41 However, the moderate and petit bourgeois nature of the Belgian socialists was not only criticized by the father of orthodox Second International Marxism; Sorel also considered Belgian socialism to be one ‘‘where the drugstore is erected to sacerdocy,’’42 and one historian has deﬁned it as an ‘‘a-political socialism.’’43 A broad classiﬁcation of the ideologically predominant characterization of pre–World War I socialist movements results, therefore, in a map such as that in Figure 2.4. Opposed to a predominantly trade unionist variant (the United Kingdom and Ireland) is a mixture of a strong anarchosyndicalist tradition with intellectual Marxism (France and Italy – and Spain – and only to a certain extent Belgium); then there are those countries where theory, organization, and political strategy were based on orthodox Marxism (Germany and Austria), and ﬁnally, those where the inﬂuence of Marxism was programmatic but accompanied by early revisionist political practices and strategies. It would seem that internal tensions between doctrine and praxis were greatest in the Dutch and Swedish cases, intermediate in the Danish and Norwegian cases, and minor in the Belgian and Swiss cases. It is hard to collapse these different types into a single dimension of radicalism versus revisionism or moderation. If one were forced to do so, the only possibility would be to oppose the basic reformist stand of the British, Irish, Belgian, and Swiss socialist movements to the more radical German, Austrian, French, and Italian movements, leaving in the middle the cases of the Netherlands and Sweden, Denmark, Finland, and Norway.

WAR AND REVOLUTION: THE RADICALIZATION OF WORKING-CLASS ACTION The pre–World War I characterization of the ideological orientation of different socialist movements is in many cases utterly transformed by the general radicalization of working-class politics on the eve of, during, and

41 42 43

cal strike. Before 1914, the Belgian Socialist Party organized four general strikes, all of which were for political objectives linked to democratization goals. Letter to Adler quoted by Lorwin (1966: 156). Quoted in Moulin (1981: 196). See Liebman (1979: 185, 187) for other colorful descriptions of the same feature: ‘‘politics, the party’s poor relative’’; ‘‘theory condemned to ostracism’’; ‘‘the co-operatives: the small village and the inn.’’

THE EUROPEAN LEFT

87

Figure 2.4. A map of early socialist ideological orientation. immediately after World War I. This radicalization is a Europewide phenomenon and is expressed mainly through growing membership, militantism, and internal ideological strife within labor movements. Its roots can be traced back to four main factors: (1) the broad increase in industrialization and the working classes themselves in many countries in the ﬁrst decade of the century; (2) the growing problem of integration within the socialist unions and political organizations of new waves of mainly unskilled workers; (3) the Great War mobilization and its consequences; and (4) the impact of the Russian Revolution. The war and revolution watershed not only fueled and renewed internal ideological and political strife, it also marked the beginning of the organizational divisions of the socialist movement. The results of almost half a century of efforts to organize, unify, and centralize socialist movements suddenly melted away almost everywhere between 1910 and 1920. After World War I, the issue of radicalism must be reinterpreted as the issue of the organizational fragmentation of the socialist movement and, in the ﬁnal analysis, as the issue of the birth and consolidation of communist parties. A distinction should, however, be maintained between (1) the conditions of generalized radicalization in most European countries; (2) the

88

THE POLITICAL MOBILIZATION OF THE LEFT

level of social unrest and turmoil that followed the war; and (3) the ideological strife and organizational fragmentation during the war and its aftermath, with the formation of Third International communist splits. These aspects will be discussed as potential causal conditions in the following chapters, and particularly in the chapter devoted to the political sociology of the communist split (Chapter 9). In this introduction, my aim is limited to a description of their impact and, in particular, of the social turmoil and organizational fragmentation that resulted from them. The most important aspect to highlight remains, as usual, not the general phenomenon but the differences among the cases. The point is that the accumulation of radicalism and revolutionary potential was unequally distributed across countries as a result of various pre-, during- and postwar national situations. The ideological and organizational strife within national socialist movements after the war was linked to (1) their record of unity and division before the war; (2) the geopolitical war position of each country, that is, its being belligerent or neutral and, if belligerent, its being defeated or victorious, and, if neutral, its being invaded or unaffected by the war; (3) the positions adopted by the national socialist parties during the war with respect to the national war effort; and (4) the national impact of the Russian Revolution and Leninism. Radicalization and internal ideological strife were already well developed in several, albeit not all, of the socialist movements before the war. Recent historical research on Italy, France, and Germany44 has shown that after the war, the divisions among socialists took place on the basis of prewar factions of the International, social divisions in the working class, generational gaps, and divisions deriving from previous experiences of the organized socialist movements and other organizations of the working class. Splits would have occurred even without the interference of the Third International, probably in a less mechanical way, with different timings and alliances among the factions.45 Generalizing this point, it has been argued that ‘‘in the political democracies where a left opposition failed to win a strong position in the socialist movement before the rise of international communism, the Communist Party never became a major force of its own.’’46 This point attributes the extent of the postwar ideological and organizational fragmentation to the situation of the prewar socialist parties. We have noted that in three countries – Italy, the Netherlands, and 44 45 46

Cortesi (1973: 451) and Lindemann (1977: 451). In support of this point, see also Schumpter (1976: 358). Lorwin (1966: 164).

THE EUROPEAN LEFT

89

Sweden – important organizational splits along the radical versus moderate dimension had already occurred before the war, even if in the last two countries they had no electoral and parliamentary consequences because the splinter group decided not to stand for elections. In Germany, no organizational split occurred before the war, but the revisionist debate triggered by Bernstein at the end of the 1890s had already split the party profoundly on ideological issues into a revisionist wing, the orthodox Marxist center of Kautsky, and the Luxemburg left wing.47 However, it is hazardous to generalize about the relationship between prewar socialist internal divisions and postwar radicalization and socialist splits. All socialist movements experienced some internal ideological fractionalization in this period. Even the moderate British trade unionism in the period 1908–1913 saw an increase in industrial militancy and a growing willingness to strike in sympathy with other workers, indicating a growing class consciousness. But for the declaration of war, it is likely that the spreading anger and the many strikes between 1910 and 1914 – resulting from dissatisfaction with real wages and the deﬁciences of the collective-bargaining system – may have culminated in a general strike in September 1914. Syndicalism, too, had made inroads in this period. Introduced into the United Kingdom by Tom Mann in the early 1900s, it was present by 1914 in all the major unions, although in a minority position.48 The radicalization of forms of action was therefore quite generalized, and it was no accident that these actions corresponded to a broader crisis, challenging the intellectual and political base of liberalism, whose rationalistic foundations had come under attack from the philosophies of Bergson and Freud, and whose political theory had been criticized by the theories of Mosca, Ostrogorski, Michels, and Pareto. The outbreak of war and, in particular, the unanimous vote of the SPD on August 4, 1914, in favor of war credits made it clear that workingclass internationalism could not prevent European governments from going to war with each other and that internationalism was a myth. The reaction to this was particularly strong among French and Belgian socialists, who were both directly and immediately involved. However, even socialists of the more distant and neutral areas, such as the Netherlands, Switzerland, and the Scandinavian countries, felt let down, to say the least. The inevitable consequence was that almost all the existing social democratic parties were driven to transform themselves into patriotic defenders of their re47

48

On the characterization of these three ideological wings and on their ﬁnal split with the coming of the war, see the classic Schorske (1955). See Holton (1985: 266–282).

90

THE POLITICAL MOBILIZATION OF THE LEFT

spective societies. At the same time, minority groups from the intellectual middle class and the working class were even more radicalized by this ‘‘patriotic slaughter’’ and became completely alienated from and opposed to the liberal tradition and order. The extent of this alienation depended to a large extent on the war position and involvement of each country, that is, whether it was belligerent or neutral, defeated or victorious. In occupied Belgium, ‘‘in the long-neutral little nation which, as King Albert said, had been ‘driven to heroism,’ revolutionary defeatism and post-war recrimination against social patriotism were absurd.’’49 The ‘‘revolutionary accumulation’’ was likely to be stronger among countries at war than among neutral countries and stronger among defeated than among victorious ones.50 Revolutionary potential was clearly associated with the collapse of public order and political control in both belligerant and defeated countries such as Austria, Germany, and Finland,51 where, indeed, revolutionary defeatism and postwar recrimination found favorable soil. The third important factor mentioned at the beginning of this section was the extent of the involvement of the socialist movement and leadership in the national war efforts. Internal radical opposition was fed largely by the reaction against socialist and trade union support of the war effort in belligerent countries, particularly in France, Germany, Austria, and Italy. Recriminations against social patriotism and participation in ‘‘bourgeois’’ wartime governments were of course, more or less intense according to the extent of socialist involvement in the ‘‘imperialist war.’’ Such involvement was far more open and intense in France and Germany than in Italy and Austria. As a reaction against the early SPD support of what the French regarded as a military aggression, the French Section of the Workers International (SFIO) supported the national coalition. The Austrians were spared the difﬁcult choice of voting for or against war credits thanks to the closing of the parliament by imperial dictum and therefore managed to maintain a more ambivalent position. Thanks to the delay of Italian entry into the conﬂict, the Italian socialists had more time to oppose the conﬂict, 49 50 51

Lorwin (1966: 164). Kriegel (1974: 20). This collapse of public order was probably more acute in Finland than in any other country, even if formally Finland could not be deﬁned as a defeated country. The 1918 civil war was largely the result of the fact that, after the Russian Revolution, the dominant Finnish groups lacked any organized military force. This explains the ease of formation of the Red and White guards in the absence of military and political forces upholding authority. Even in the 1920s, the White-inspired civil guard was far larger than the army (100,000 armed men compared with 25,000 soldiers). Cf. Alapuro and Allardt (1978: 125).

THE EUROPEAN LEFT

91

and ﬁnally, although with great hesitation, they followed the line of ‘‘ne´ aderire ne´ sabotare.’’52 This was later very important in explaining why the Livorno communist split in the PSI affected only a minority of the members, while the Tours communist split took with it the majority of the old SFIO (see Chapter 9). An uncompromising ‘‘patriotic’’ socialist alignment in a belligerent and ultimately defeated country was most likely to trigger the highest level of internal criticism, recrimination, and opposition within the socialist leadership and the rank and ﬁle. When this was accompanied by the ideological crystallization of different organized ideological factions, the situation became even more explosive. Finally, the Russian Revolution and Leninism contributed to the fueling of this internal debate. The political and ideological meaning of the Russian Bolshevik Revolution lay not only in the emotional attraction represented by the long-awaited establishment of the ﬁrst socialist state for the entire socialist movement, but also in the great upheaval of Marxist theory and socialist ideology it produced. Lenin’s revolution destroyed the model of historical development based on the Marxist analysis of capitalist society and on the deterministic deﬁnition of the conditions of development and revolution as the basis of the socialist understanding of reality and socialist strategy. It proved the bankruptcy of any determinist theory and highlighted the role of organizational and ideological factors with regard to political practice. The fact that a revolution occurred in a country where the socioeconomic conditions were not ripe seemed to prove that revolution was possible under any circumstances. This universal possibility of revolution thereby implied a rejection of, or at least a reconsideration of, any theory of capitalism and, at the same time, a reconsideration of the theory of the development of socialist movements under capitalism. The study of the conditions for the arrival of the revolution now concentrated on the tactical and organizational conditions. A theory was still necessary, but this theory was largely transformed from a theory of capitalist development to a theory of revolution. The universal possibility of revolution, exempliﬁed by the Russian events, threw the blame for the failure of the revolution in the more advanced European countries on the prewar and wartime socialist leadership, particularly in the larger Western European countries. The idea of 52

Actually, the position of the Italian socialists with regard to the war was less coherently internationalist than the formula implied. The socialists put pressure on the government to avoid intervention in alliance with the central empires, stressing the difference between the two belligerent parties. Cf. Cortesi (1973: 86–87).

92

THE POLITICAL MOBILIZATION OF THE LEFT

party that Lenin progressively developed between the writing of What Is to Be Done (1902–1903) and the later revolutionary experience53 was an effective weapon in an atmosphere of clandestine activity and repression that occurred during political upheavals but was far less so under any other conditions. Similarly, the three propaganda proposals – ‘‘land to peasants,’’ ‘‘all power to the Soviets,’’ and ‘‘stop the war now’’ – were ideal tactical weapons in a short-term revolutionary situation but hardly an exploitable program under any other condition. The question of the short-versus longterm revolutionary solution has in fact been a vital problem in the development of the international communist movement. The correct evaluation of the nature and possible outcome of the crisis was closely related to the choice of the kind of party and strategy the Third International forces had to organize and follow. If the revolutionary crisis brought about by the war in several countries was to produce a short-term revolutionary outcome, then a leading Bolshevik party was an immediate necessity. If the crisis was leading to restoration rather than to revolution, the new parties had to be able to survive and wait for a new crisis and for a new and favorable conjuncture. In view of the uprising they foresaw, the Bolsheviks considered it essential to drastically change the unitary social democratic parties. It remained to be decided how to organize such a change: the moment of the internal confrontation, the choice of the confrontation line (with whom and against whom), and, if it was the case, the choice of the breaking modality: by minority split or by internal purge. In conclusion, the prewar socialist divisions, the country’s war role, and the socialist attitude to wartime cabinets were the powerful contextual conditions that helped to shape the receptiveness of the Leninist interpretation of the postwar revolutionary potentials and opportunities within the old socialist parties. When by the end of 1921 it became evident that the revolutionary crisis was over, the problem of revolutionary parties working in objectively nonrevolutionary situations became that of settling and organizing the strong extremist nucleus of the early postwar phase. Lenin summed up the problem with this pointed remark addressed to Costantino Lazzari in July 1921: ‘‘il faut reculer pour mieux sauter.’’54 However, the overestimation of the revolutionary potential in the West was to have farreaching implications. The main consequence of the political strategy chosen to deal with European socialism was that no communist party would succeed in retaining the majority of the politicized working class. 53

54

The best account of Lenin’s ideas in general, and of his organizational theory in particular, is Lundquist (1982: 297–320). Sechi (1977: 39).

THE EUROPEAN LEFT

93

To what extent was this different accumulation of revolutionary potential in different countries actually mirrored in the radicalization and intensity of class action? The question may be answered by comparing the intensity of industrial conﬂict across countries during the war and revolution years and the extent of political turmoil and instability brought about by them. It is necessary to differentiate between these aspects because industrial conﬂict is not by itself a sign of political crisis or revolutionary potential; that is, it does not necessarily indicate a political radicalization of the class conﬂict. Social and industrial unrest and workers’ militantism increased everywhere in this crucial period. To evaluate where such radicalization in collective behavior was stronger, one must rely on information about protest events and ‘‘collective action pursuing an explicit goal by the use of adversarial, disruptive or even violent means.’’55 However, systematic comparative and historical data are difﬁcult to ﬁnd in this ﬁeld.56 For a global comparison, therefore, the only systematic sources are those on strikes. Strike actions are composed of three elements that do not necessarily coincide or covary: The number of strikes launched, the average number of strikers, and the average length of strikes and country traditions are very different in this respect.57 For my purposes, I have combined the number of strikers and the number of strikes into a multiplicative single index (see Table 2.4).58 The ﬁgures in Table 2.4, which cover the period 1901–1940 in ﬁveyears blocks, show how the intensity of the industrial conﬂict grew steadily before the war, peaked in the years 1918–1925, and declined rapidly thereafter up to the Second World War. The high level indicated for the 1936–1940 period is due exclusively to the absolutely exceptional levels of conﬂict in France, which is the only European case of industrial conﬂict mobilization in the second half of the 1930s. However, even leaving aside this case, the data show great variation among countries. In ﬁve countries 55 56

57 58

Rucht and Ohlemacher (1992: 77). The pioneering work of Charles Tilly on strikes, collective violence, and ‘‘contentious gatherings’’ does not allow a synthetic appreciation of differences among our thirteen countries. See Tilly (1978). Other studies of protest events concentrate on the post– World War II situation and are less useful for me in this context. See the World Handbook of Political and Social Indicators (1964, 1972, 1983). See Schorter and Tilly (1974: 306–334). These data are available in Flora, Alber, Eichenberg, Kohl, Kraus, Pfenning, and Seebohm (1987: 679–753). On the length of strikes, the data are far less complete and systematic. The data contrast the cases of Norway, Sweden, and the United Kingdom to all the others. In the former, the relative number of working days lost is far higher, and systematically higher, than in all other countries, both throughout the forty-year period from 1901 to 1940 and in the ﬁve-year periods immediately preceding and following the war.

94

THE POLITICAL MOBILIZATION OF THE LEFT

Table 2.4. Industrial conflict levels (N strikers * N strikes) before and after World War I

Note: N of strikes and N of strikers are computed per 100,000 persons of the economically active population.

industrial conﬂict remained comparatively low, with ﬁgures consistently below the European mean levels both before and after the war: Switzerland ﬁrst of all, but also Denmark, France, Ireland,59 and the United Kingdom, where strikes tended to be numerous and long but involved few strikers. At the other extreme, the countries that show the highest level of industrial conﬂict throughout the entire period are Sweden and Norway, both of which remained neutral during the war. The Netherlands and Italy, which we have already singled out as the countries where there was intense ideological strike action and organizational splits in the socialist movement before the war, show levels of industrial conﬂict that were higher than the mean before the war rather than after it. By contrast, the level of conﬂict was higher after the war in Austria,60 Finland, Germany, and Belgium. The interpretation of these data is not straightforward. The level of industrial disputes does not necessarily mean a corresponding level of social or political tension. According to the ideological tradition and the position 59

60

For Ireland, systematic data exist only after 1920; however, there is no indication of particularly intense strike activity before that date in the literature. For Austria, prewar data are, unfortunately, not available. Between 1921 and 1926, Austria had the highest level of industrial conﬂict registered (although strikes probably had a short duration).

THE EUROPEAN LEFT

95

of trade unions with respect to the political party, each level may acquire a different meaning. The industrial militantism of the British working class after the war years developed within a comparatively liberal political atmosphere and a tradition of relative autonomy between political and industrial action. Although the Irish data are missing for the period, class polarization was not the core issue at that time. Conversely, cases of revolutionary syndicalist trade unions and of orthodox Marxist party– linked trade unions gave a more important political meaning to the industrial disputes. In reality, in only four of our thirteen cases was the postwar high industrial conﬂict connected to revolutionary threats and political instability (Table 2.5). In Germany, the military and state collapse of November 9, 1918, triggered a revolutionary phase that carried through the election of the ﬁrst National Assembly in February 1919; the summer 1919 repression of the Bavarian Council Republic and the workers’ agitation in several other parts of the territory; the spring 1920 attempted Kapp Putsch and the following Ruhr workers’ rebellion; and, ﬁnally, the failed communist upheaval in October 1923. German historians do not agree on the evaluation of this period, namely, whether to consider it as an uninterrupted revolutionary cycle from 1918 to 1923. The majority seems to believe that a revolutionary upheaval would have been possible only in the ﬁrst phase, before the election of the National Assembly. The developments that followed, with the explosion of the workers’ protests in the face of the result of the ﬁrst phase, are viewed as having been unable to challenge the political equilibrium and the legal and economic choices that were already consolidated.61 The Bavarian and Spartakist attempts could still be considered as events linked to the collapse of the empire and the war defeat;62 the Ruhr rebellion has been interpreted as a defensive measure in response to the threat to the Weimar Republic;63 and the attempted communist revolutions of 1921 and 1923 are judged as less signiﬁcant challenges and easier for the government to repress than the Ruhr revolt, as they were more clearly minoritarian, insurrectional, and disloyal attempts.64 Two other countries witnessed unstable and politically strained postwar development. In Finland, as in Russia, the war slid into a civil war between Reds and Whites, with the eventual military victory of the latter 61 62 63 64

See Ribeni (1989: 139). For the Bavarian revolutionary attempt, see A. Mitchell (1965). Moore (1978: 328, 340). This is the interpretive conclusion of Angress (1972: 459–461). Actually, the repression that followed was mild, and the Communist Party was made illegal for only a few months.

96

THE POLITICAL MOBILIZATION OF THE LEFT

Table 2.5. Configurational aspects of the ‘‘war and revolution’’ crisis of the socialist movement

setting the terms for a bitter confrontation with the communists, who were forced into a clandestine position and banned in 1923 (the communist-controlled Trade Unions Confederation was banned in 1930). In Austria, the First Republic was proclaimed by the socialists. The communist attempt to penetrate the workers’ councils failed completely, and a coalition government of major political forces ran the country from 1918 to 1920. Therefore, the immediate political transition seemed to be under control. However, class warfare was radicalized by the electoral victory and the governmental coalition of the Christian Social Nationals and the German Nationals in autumn 1920. The emergence of armed organizations, such as the Frontka¨mpfer, the SA, and the Heimwehren on the extreme right led to a corresponding development in the Social Democratic ﬁeld, with the armed Republikanischer Schutzbund, and to a situation of

THE EUROPEAN LEFT

97

radical and even violent confrontation. In these three cases, the postwar political instability resulted largely from the collapse of the old prewar authoritarian regime as a consequence of the military defeat it had undergone. Italy is the fourth European case of high postwar social turmoil and political instability. Although the country fought in the war and was eventually victorious, high social tension permeated the entire post–World War I period until Mussolini came to power. However, to deﬁne the Italian situation in 1920 as potentially revolutionary is probably incorrect. The occupation of the factories that took place in September in Turin and in other parts of Italy was in no way a planned revolutionary attempt. It was rather the result of industrial disputes in a period of a rapidly growing political mobilization of the Italian working class and was accompanied by the growth of new working-class institutions: the internal commissions and the factory councils. It was initially a defensive move, that is, a way of protecting workers from the threat of employers’ lockout. The interpretation of this period as a failed revolution is usually advanced in reference to the thinking of Antonio Gramsci, who aimed to direct events toward such an outcome.65 In reality, it was the political skill of Giolitti and the permanent reformist orientation of the General Confederation of Labour (CGL) that limited any revolutionary potential and led the confrontation to a negotiated settlement. The subsequent political threat, which came mainly from the nationalistic groups and the new fascist movement, forced the socialist movement to be permanently on the defensive. Beyond these four cases, there was no challenge to the political order during the postwar phase in Europe. In an attempt to systematize the experiences discussed so far, we can regroup the factors mentioned in Table 2.5. It was within the resulting conﬁguration that socialist organizations fragmented further that communist splits developed.

ORGANIZATIONAL FRAGMENTATION AND COMMUNIST STRENGTH The Great War and the crisis that ensued triggered a process of progressive and increasing political fragmentation of the European left. The enormous efforts in almost all countries between 1860 and 1900 to foster the pro65

A typical example of the political situation being interpreted through the lens of Gramsci’s thinking is the work by Clark (1977). On the occupation of the factories, see the more sober judgment of Spriano (1964).

98

THE POLITICAL MOBILIZATION OF THE LEFT

Figure 2.5. Left electoral fragmentation by decade. gressive uniﬁcation and centralization of working-class organizations, in both the corporate and electoral channels, evaporated. Under the pressure of powerful internal ideological dissent and international pressures, in most countries the left fell victim to internal ideological and organizational differentiation. In Figure 2.5 this is shown clearly by reporting for each decade Rae’s index of fragmentation applied to the total left alone rather than, as usual, to the entire party system.66 Some fragmentation was already evident before World War I, but the process accelerated from the 1920s to the 1940s. With the electoral stability of the 1950s, fragmentation was again reduced, but afterward it increased progressively in each decade. Comparing the countries in the four electoral periods that we have discussed, we can observe quite different levels and patterns of electoral fragmentation. In the ﬁrst period before World War I, most of the left was represented by a single uniﬁed socialist party, and other very minor social66

That is, having made equal to 100% the size of the total left, the index measures the degree of its internal divisions in electoral terms.

THE EUROPEAN LEFT

99

ist formations existed only in France and Italy.67 After World War I, the general tendency toward growing fragmentation is visible in Belgium, Denmark, Finland, France, Ireland, Italy, and Norway, while the situation is stable in Austria, the Netherlands, Sweden, Switzerland, and the United Kingdom; a clear decline can be seen only in Germany. In Italy and Finland, the interwar period indicates a relatively low index of fragmentation simply because the communist party was outlawed in Finland, and in Italy the only available electoral data refer to the elections of the early 1920s. Germany has the highest electoral fragmentation of the interwar period, but it also shows the sharpest concentration process after World War II and the division of the country. It should perhaps be pointed out that the outlawing of the Communist Party in Germany after World War II was not responsible for this trend; it had been reduced to a tiny proportion of the electoral whole even before it was outlawed. At the opposite end, four countries are almost immune from politicoorganizational fragmentation – Austria, Switzerland, the United Kingdom, and Ireland. Considering the characterization of the early ideological orientation of the left and the ‘‘war and revolution’’ mobilization phase, the capacity of Austrian socialism to maintain its organizational cohesion is somewhat surprising: preconditions seemed favorable to an early and deep ideological-organizational split of the party. All the other cases fall into a middle position between these extremes, with some change from period to period. In Denmark and Belgium, for instance, the tendency is toward a continuous increase in fragmentation. In the 1966–1985 period, these countries exceed the European mean level and come close to that of Italy, France, and Finland. The Netherlands, Norway, and Sweden instead keep a medium and fairly stable level for the entire period from the end of World War I to the 1980s. THE MODALITIES OF ORGANIZATIONAL SPLITS There are three countries in Europe where no serious organizational split occurred after World War I, Ireland, Austria, and the United Kingdom. In Austria, Victor Adler managed to balance the center position of an orthodox Marxist movement between the patriotic tendencies on one side 67

As mentioned before, a split took place in Sweden in 1908 (SVP), but the splinter party did not participate in elections until 1917, when it obtained considerable success with 8% of the votes. In the Netherlands a similar split occurred in 1909, but the radical splinter group (SDP) did not stand for election before or during the war period.

100

THE POLITICAL MOBILIZATION OF THE LEFT

and the antimilitary tendencies on the other (led by Bauer and Renner), and avoided the party split a` la Germany. This difﬁcult position was helped by a number of circumstances. First, the restiveness of the various national wings within the socialist movements – in particular, that of the Czechs – helped the center to keep the party united. The dangers of national splits and the centrifugal drives associated with them probably constrained the ideological strife between the moderates and radicals of each national section, and national idiosyncracies prevented them from joining forces cross-nationally. At the same time, to deal with the war, the imperial regime suspended a number of constitutional rights, and by resorting to Article 14 of the constitution, which granted exceptional power to the emperor, the parliament was dissolved. The party fought against this limitation of rights together, while at the same time avoiding the internal debate on the war credits vote (which had triggered the internal conﬂict in the SPD) and on many other lacerating issues. The split in the party, created by the extreme antiwar left, occurred as early as November 2, 1918, but it was small and very marginal. This splinter group became involved in an unsuccessful revolutionary coup, initiated in conjunction with the Bavarian one, and from then on remained an ultra-left sectarian group, divided internally by dissidence and factional strife, boycotting the elections and unable to penetrate the social democratic working-class organizations. The Austrian Socialist Party rejected Lenin’s twenty-one points, proposed by Zinoviev; but at the same time, they did not return to the Second International, which they accused of having been affected by ‘‘social patriotism.’’ Instead, they formed the ‘‘Work Community of the Socialist Parties’’ in 1921, which was immediately deﬁned disparagingly as the ‘‘Two-and-a-Half International.’’ They returned to the Second International only in 1923, when the most intense revolutionary crisis was over. In the British case, one should not speak of a split, as the Communist Party of Great Britain (CPGB) founded in London on August 7, 1920, was rather a fusion of a number of Marxist groups, including the British Socialist Party, the Labour Socialist Party of Glasgow, and other nuclei, none of which can be regarded as a split from the Labour Party. At its second congress, held in Leeds in 1921, other political groups joined, such as the Socialist Federation of Workers and the left of the Independent Labour Party, and the party formally adhered to the Third International.68 The party tried to afﬁliate itself with the Labour Party, but it was rejected from the beginning. Following the election of a Communist on the ofﬁcial 68

Wood (1959).

THE EUROPEAN LEFT

101

Labour Ticket in the general election of 1922, the Labour Party decided in 1924 to formally preclude the nomination of CPGB members as Labour candidates in future elections. A year later, in 1925, it ﬁnally decided to exclude CPGB members from its constituency branches.69 As a result of its electoral weakness – with the few exceptions of small zones like Rhondda East, Motherwell, and North Battersea – British communism has always been considered insigniﬁcant, although in the 1920s its inﬂuence was considerable in the mining and engineering unions. In some trade councils, such as that of London, relatively strong CPGB cells existed. The CPGB also formed some important organizations like the National Unemployed Workers’ Movement (NUWM), whose membership ﬁgures are still highly debated by historians. After 1927, its inﬂuence in the trade unions was reduced substantially due to the strong attempt to achieve this by the Trade Union Confederation (TUC) reformist leadership and to the CPGB’s suicidal policy of attacking moderate trade unions as ‘‘social fascists.’’ The radical split in the Irish Labour Party had few effects. An Irish Communist Party was founded in 1921, but it had already dissolved by 1923 to merge into the League of Irish Workers, its activities limited to the twenty-six Southern counties. Communists continued to work through autonomous structures in the six counties of Northern Ireland until June 1933, when a national congress reunited the Irish Communists across the whole country. This small and politically insigniﬁcant group continued to split and reunite until its formal dissolution in 1941. In 1948, the Communists in the South reorganized themselves into the League of Irish Workers. In 1970, the North and South once again united under the name of the Communist Party of Ireland. In Austria, Ireland, and the United Kingdom, therefore, war time and postwar radicalization resolved itself either with no formal split or with the formation of totally insigniﬁcant sects. In the Netherlands and Sweden, the postwar formation of communist parties should be seen in the light of the prewar radical splits in the socialist parties. In both countries, the radical wings had been expelled and had formed new parties of the left: the Social Democratic Party in the Netherlands and the Swedish Young Socialist Party (SUP) in Sweden in 1908 and 1910, respectively. At a congress held in 1918, the Dutch Social Democratic Party changed its name to the Communist Party of the Netherlands. The Swedish development was more complex. The new radical party pushed through the foundation of an anarcho-syndicalist movement 69

For the contact at election times, see Salles (1977: 407–427). On the formation of the Communist Party, see also Mair (1979: 157–182).

102

THE POLITICAL MOBILIZATION OF THE LEFT

in Sweden (SAC). When the Swedish Social Democrats participated in a liberal government in 1917, a second split occurred, with the left wing of the party forming a new party: Sweden’s Social Democratic Left Party (SSDLP). This party was the embryo of the various communist parties that would emerge over the following decades. It adhered to the Third International in 1919 and changed its name to the Swedish Communist Party after having reunited various political and unions left-wing groups. What is striking about the fragmentation of the Swedish organizations after World War I is that the formation of a Communist Party was not its end but its beginning. In the following years, continuous confrontation with the Third International led to two right-wing splits: in 1926 the so-called Ho¨glung Communists, who soon returned to the Social Democrats, and in 1929 the Kilborn Communists, who returned to the Social Democrats in 1937. Although there are differences in the process of Communist Party constitution, the Dutch and Swedish cases are similar because the post– World War I birth of a Communist Party resulted from the adherence to the Third International of previously established radical left formations; the original motivations and sources of these groups were independent of the speciﬁc push determined by the Russian Revolution and by the initial strategy of the Third International. Unlike the Netherlands and Sweden, in Belgium, Denmark, and Switzerland the split originated in the last phase of the war and had no speciﬁc organizational links with prewar parties. The experiences of the three countries present some variations, but in all cases the split resulted in a minor and nontraumatic separation with a decidedly tiny left-wing minority. In Denmark, where the impact of World War I was minimal thanks to its neutrality, there was little right- or left-wing factionalism. When Thorvald Stauning entered the cabinet as the ﬁrst Social Democratic minister in 1916, the extraordinary party congress authorized this move by a vote of 291 to only 32,70 even though the continued governmental cooperation with the Radical Liberals was approved by a much more restricted majority in 1918 (220 to 156).71 Very minor socialist parties started to organize during 1918.72 More important was the radical attitude of the Social Democratic Youth organization (the SUF), which adhered to the Zimmerwald left. Having failed to obtain the termination of the socialist participation in government and ﬁrmer antimilitarist stances, the SUF 70 71 72

Logue (1982: 66). Ibid. 68. Namely, a Socialist Workers’ Party (SAP) organized by some leaders of the syndicalist opposition unions movement and a second small socialist splinter party. However, they obtained only 2,500 votes of the 920,000 cast in the 1918 election.

THE EUROPEAN LEFT

103

broke away from the party and, in November 1919, joined other minor socialist groups to form Denmark’s Left Socialist Party, which one year later accepted the Comintern’s twenty-one conditions and renamed itself as a Communist Party. Subsequently, in 1921, it merged with the syndicalist trade union opposition (FS).73 As in Sweden, the Danish Communist Party suffered a major split in 1958, when the wing led by one of its major leaders, Aksel, split to found the People’s Socialist Party (SF). In Belgium, the Communist Party was formed in 1921 by the fusion of two left-wing groups that had detached themselves from the Belgian Socialist Party at two different moments. The party was the fusion of two newspapers and of the political groups around them, L’ouvrier communiste and L’exploite´, directed by Joseph Jacquemotte, who was to become the major leader of the new formation. There was a certain amount of continuity between these groups and their internal opposition over the question of the socialist participation in bourgeois government since 1910.74 The ﬁnal outcome of the Swiss radicalization and organizational fragmentation was similar, but the modalities of the split are particularly interesting and merit review. The Swiss Socialist Party is generally portrayed in comparative terms as a moderate movement. However, radicalization had also occurred in Switzerland during the second half of the war, when restricted food and supplies and income reductions accentuated social tensions. This resulted in the general strike of 1918. In the ﬁrst days, the level of participation was impressive even by non-Swiss standards. However, the strike was unexpectedly stopped by a decision of the executive body of the strike itself and by the Socialist Party. The nine-point set of goals formulated by the strike committee (the Committee of Olten) was abandoned, even though some of the reforms advocated (in particular, proportional representatation and a forty-eight-hour work week) were implemented. This sudden halt produced bitter resentment and opposition among many of the participants and gave dynamic momentum to the radical opposition within the Socialist Party.75 As a result, a motion in favor of leaving the Second International was approved by 459 delegates and rejected by only 1 in the 1919 Basel Congress. The successive motion 73 74

75

On the origins of the Danish Communist Party, see Rohde (1973: 3–33). The split has been related by some historians to the growing manifestation of a revolutionary syndicalist and radical Marxist left within the Belgian Socialist Party from 1908 on. Steinberg (1971: 3–34), in particular, has underlined these links. Jacquemotte was already the leader of this pre–World War I radical wing, but it was a restricted group of people with only personal links. On the crucial importance of the 1918 general strike and its interpretation as the high point of working-class protest in Swiss socialist history as a whole, see Stettler (1980: 5– 43).

104

THE POLITICAL MOBILIZATION OF THE LEFT

for adherence to the Third International was passed by a majority of 328 to 147. Like the Norwegian and French socialist parties, the Swiss party adhered by majority vote to the new International. However, the minority asked for the decision to be conﬁrmed by a referendum among the members of the party. The surprising result was that the congress’s decision to adhere was rejected by 14,612 votes to 8,722. Yet, notwithstanding this burning popular defeat, the supporters of the Third International still numbered more than one-third of the party members. In the extraordinary Congress of Berne in December 1920, the motion to adhere to the Third International was reproposed and this time was defeated by 350 votes to 213. In spite of this result, the left of the party thought that the supporters of the Third International were more numerous than was shown by the vote because the majority of the delegates approved action to obtain modiﬁcation in the twenty-one conditions of adhesion. It was now their turn to ask for a referendum among the members, but they were again disappointed by the results: The congress decision was conﬁrmed by 25,475 votes to 8,777. Finally, after the sudden and intense resentment had passed, the communist left split in March 1921, when it fused with other older organizations of anarcho-syndicalist tradition.76 From the mid-1920s on, the Swiss Communist Party experienced accentuated Bolshevization. The old guard was dismissed and the direction modiﬁed a` volonte´ until strict Stalinist orthodoxy was granted.77 In the interwar period, the party was outlawed in some cantons (Geneva, 1932) and, during the war, at the federal level. The bulk of the party’s members were in German-speaking Switzerland, and it never succeeded in making inroads into Romandy (although it was moderately successful in Geneva). Federalism and the deeply rooted Swiss traditions of local autonomy were regarded as vestige bourgeois; but throughout the 1920s and early 1930s, conﬂicts between regional committees (particularly Romance ones) and the central committee were frequent. Relationships with the trade unions had been difﬁcult from the beginning and remained so. The Socialist Party took a very hard line against the opposition, reacting strongly against the creation of communist cells. Membership ﬁgures show how politically 76

77

On this fusion, see Studer (1993: 5–38.) On the historical origins of the Swiss Communist Party, the best source is Stettler (1980). On the Bolshevization process, see Stettler (1980: 61–92). The party was dominated for a long time by Jules Humbert Droz, who reached the highest echelons of the International. As secretary responsible for the Latin countries and as a member of the executive committee, he spent more than ten years in the service of the Third International. During this period, the Swiss section of the International became an essential support for Droz in its internal dissenting position vis-a`-vis Stalin. So, the party probably also suffered from this conﬂict as well as from Droz’s ﬁnal expulsion.

THE EUROPEAN LEFT

105

insigniﬁcant Swiss communism was and give some idea of its steadily continuing decline: the 6,356 members were reduced to 3,500 by 1927, to 1,700 by 1933, and to a few hundred in 1938–1939.78 In the last group – France, Italy, Germany, Finland, and Norway – the splits were more dramatic, the forces more evenly distributed, and the depth of the divisions more profound. In some cases, such as Norway, Italy, and Germany, the socialist movement split into three major groups. The Norwegian Workers’ Party was one of the founders of the Third International, and its original adherence was almost unanimous. However, this was more an act of international solidarity, despite its reﬂection of the postwar radicalization of the movement. The growing divergence with the Comintern that soon emerged produced the right-wing split of a large group of reformists who founded a Social Democratic Party two years later. An idea of the respective strength of the two groups was given by the election of the same year, when the reformists obtained about 9.2% of the votes compared with 21.3% obtained by the Workers’ Party. In other words, a third of the party had split. The resulting growing resistance of the Workers’ Party to the Comintern’s intrusions into internal party affairs and the skepticism of the trade unions – which had not adhered to the Red Trade Union International – precipitated a withdrawal of the party from the International in November 1923. The pro-Moscow minority immediately founded the Norwegian Communist Party (NKP). As a result, the Norwegian left was split into three separate parties at the 1924 elections, each of them representing a considerable proportion of the vote: 8.8% for the reformist Social Democrats, 18.4% for the ‘‘centrist’’ Workers’ Party, and 6.1% for the Communist Party.79 The German and Italian experiences produced similar results, even though they took place in a context of more pronounced social turmoil and political instability. The German division into revisionist, orthodoxcentrist, and revolutionary groups was an old and well-established internal articulation within the SPD from the beginning of the century, but the war and the war credit vote now precipitated organizational fragmentation. Here the three-party format was the result of successive left-wing splits. In April 1917, the opponents of the patriotic stance of the SPD left the party (after having already constituted an autonomous parliamentary group) to form the Independent Social Democratic Party (USPD), which included both Kautsky and Haase and the radical left Spartakus group led by Karl 78 79

Studer (1993: 19). On the stormy 1919–1923 history of the Norwegian labor movement, see Sparre (1981a: 691–704) and (1981b: 393–413).

106

THE POLITICAL MOBILIZATION OF THE LEFT

Liebknecht and Rosa Luxemburg. Between 1918 and 1919, the Spartakus group left the USPD to found the Communist Party of Germany (KPD). The KPD refused to participate in the 1919 election, but in 1920 the socialist movement was electorally divided into a three-party, left-skewed situation: 21.6% of the votes to the ‘‘majoritarian’’ SPD, 17.9% to the Independent Social Democrats, and 2.1% to the Communists. In later years, the Independents were split by the centrifugal attraction of the SPD on the right and the Communists on the left, but they continued to participate in elections up to 1928. In Italy, the socialist movement had split before the war, when the reformist wing led by Bonomi and Bissolati – which was sensitive to the Bernstein revisionism, as well as to the British model of trade unionism – was expelled from the Socialist Party.80 During the war, the Italian Socialist Party (PSI) had been an active member of the Zimmerwald (September 1915) and Kienthal (April 1916) Socialist International Conferences and had maintained a paciﬁst, antimilitarist, and antiwar position, while at the same time trying not to sabotage the military stand taken by the country. The Russian Revolution of 1917 and the end of the war exacerbated the internal divisions of the party. However, under the leadership of the left led by Serrati – within which the intransigent group led by Bordiga and Gramsci started to be noticed – the party seemed capable of maintaining its organizational unity and negotiating its global adherence to the Third International. It was thus only at the beginning of 1921 that the Leninist wing split to found the Communist Party of Italy. The effort to keep the party united while negotiating its entrance into the new International had failed, but the Communist split was not sufﬁcient to safeguard the organizational unity between maximalists and reformists, and the issue of the relationship between socialism and Russian-inspired communism, as well as the interpretation of the Italian political situation and that of the incoming fascist movement, led to a further split of the reformist wing led by Turati and Treves, which founded the Unitary Socialist party (PSU).81 By 1922, the Socialist Party of 1910 had split into four groups: the majoritarian Socialist Party; the reformists of 1912, who competed autonomously in the elections of 1913 and 1919 (gaining 3.9% and 1.4% of the vote) but not in 1921; the Communist Party, which competed for the 80

81

It is interesting to note that the left led by Lazzari and Mussolini managed to expel the reformist wing, using as a formal argument the fact that its leaders had violated the socialist tradition, having agreed to be consulted by the king during the previous ministerial crisis. See Arfe´ (1965: 137–148). Even in this case, the formal occasion was the fact that Turati had participated in political consultation following the crisis of the Facta government.

THE EUROPEAN LEFT

107

ﬁrst time in 1921, gaining 4.6% of the vote; and the Turati–Treves reformists, who never contested elections. So, from the strictly electoral point of view, the fragmentation was never as sharp as in Germany or Norway, although politically it was similar. The French split resembled that of Norway: The right wing of the SFIO left the party, while the majority voted for unconditional adherence to the Third International. Although the entry of Guesde into the national defense government met with little internal opposition at the beginning of the war,82 criticism of the unconditional support for the war and the request for a policy of peaceful conclusion of the conﬂict started to grow from 1915. As a result, the left fraction established contacts with other European revolutionary socialists, as well as with Lenin at the Kienthal and Zimmerwald congresses. The internal radicalization and factional strife increased to high levels only at the end of the war regarding the issue of adherence to the Third International. The right wing rejected the organizational and political principles of Leninism, the center favored adherence but wanted to negotiate conditions, and the left requested unconditional adherence. In a series of successive votes in the 1919 and 1920 congresses, the unconditional pro-adhesion group gained a strong majority of 67.4% and ﬁnally the right and the center abandoned the Tours congress as a minority.83 A second element that makes the French split very striking is that the communists merged with the syndicalists and even the anarchists. Moreover, the political split was accompanied by a split along the same lines in the trade union movement, leading eventually to the formation of the General Confederation of the United Workers (CGTU).84 This is virtually a unique phenomenon, as in most cases trade unions resisted 82

83

84

The issue of the socialist participation in bourgeois governments had been highly divisive from 1901, when the participation of Millerand in the government of Waldeck-Rousseau convinced Guesdes to distance himself from the slowly organizing French Socialist Party. This may be regarded as a very early prewar internal split of the socialist movement, but it is true that the Socialist Party was still in a formative period. However, the new condition of foreign attack in 1914 did not create the same tensions. For an account of the episode of the Guesde’s sharp opposition to Millerand, see Lefranc (1963: 111–112). In 1919, the different positions were distributed as follows: right (to stay within the Second International): 757 mandates; center (to stay provisionally within the Second International but to break relations with the German Social Democrats and maintain good relations with Russia: 894 mandates; left (to abandon the Second International for the Third): 270 mandates. In 1920, at Tours, adherence to the Third International was approved with 3,208 votes, while the centrist position obtained 1,002 votes and the right obtained 397 abstentions. A second vote – required by the majority to give the center a second chance to adhere – failed: 3,247 votes to 1,398. See Kriegel (1969: 386). On the modalities of the trade union movement split and on the extremely heterogeneous nature of the newly founded CGTU, see Robert (1980: 180–182) and Lorwin (1954: 57–58).

108

THE POLITICAL MOBILIZATION OF THE LEFT

internal strife on moderate positions much more successfully. Unquestionably, the formation of the French Communist Party (PCF) – which absorbed the majority of the old SFIO – was the result of political issues concerning the organizational nature of the new political party and its strategy, which were not clearly delineated. This lack of clarity created a larger majority for the Third International, but the extremely heterogeneous nature of the new party led to a more profound and shocking intervention on the part of the Third International later. In Italy, the minority groups of Bordiga and Gramsci had recognized the political and organizational implication of Leninism more clearly, and had proved less inclined toward groups that did not share such implications for the life and organization of the new party.85 The Finnish Communist split does not resemble any of these others because of the close link between Finnish and Russian history. Strictly speaking, there was no split, as the Finnish Communist Party (SKP) was founded outside Finland, in Moscow, at the end of August 1918 by the survivors of the defeated Red Guard in the civil war of January–May 1918. Its history was, therefore, from its beginning, intimately linked to that of the Russian Bolsheviks; its culture was that of a clandestine armed group, and it developed outside Finland until its fourth congress in 1920, which was the ﬁrst held at home. This did not, however, prevent the party from gaining a remarkable 14.8% of the vote at its ﬁrst appearance in the national electoral competition. The division with the moderate socialists was particularly deep, as the physical proximity of the Russian Revolution and the national opposition between Finland and the Soviet Union deepened the cleavage between the communists and socialists. One example of this can be found in the contrast that was accentuated by the foundation of the autonomous Soviet Karelia on the Russian side of the Finnish border, which was headed by a number of the central ﬁgures of the defeated revolutionary group in 1918. Communists were considered an enduring threat, and the socialists moved to the center even in economic policy. Thus, unlike other Scandinavian socialists, Finnish socialists accepted and went along with classical liberal economic policies that were costly and painful to the working class. In 1923, three years after its return ‘‘home,’’ the Communist Party was dissolved by the Finnish Diet when it held twenty-seven seats. In 1930, the Trade Union Confederation (SAAJ) was also dissolved because it was controlled by the communists.86 In conclusion, the communist–social85 86

This is the opinion of Galli (1976b: 39–40). For a general discussion of Finnish communism see Hodgson (1967, 1973), Laulajainen

THE EUROPEAN LEFT

109

ist split was more bitter in Finland than in any other country. In fact, even during World War II, which was generally a period of rapprochement for anti-Nazi national coalitions, the two parties remained in opposite camps. This happened as a result of the Finnish–Soviet War that began on November 30, 1939, following the Finnish refusal to yield to the Soviet request for territorial concessions aiming to guarantee the defense of Leningrad. The SKP boycotted the war and established a Soviet government in Terijoki under the direction of the leader Kuusinen. This move was interpreted by the other political forces as a step toward the annexation of Finland to the Soviet Union and precipitated not only the persecution of communists as national traitors, but also the Finno-German alliance of June 1941 and the unique European case of a socialist party cooperating with the Nazis: Vaino Tanner, leader of the Social Democrats, was indeed to participate in all the war governments.

THE ELECTORAL DEVELOPMENT OF THE LEFT’S COMPONENTS The three main components of the left are represented in this study by, ﬁrst, the ofﬁcial socialist, social democratic, or labor parties, all of which maintained their original afﬁliation with the socialist International (only in Italy were two socialist parties members of the International at the same time as a result of the 1947 split between socialists and social democrats); second, the communist parties; and a third component represented by all those ‘‘third’’ parties that originated as right-wing or left-wing splits from either the ofﬁcial socialists or the ofﬁcial communists. This is a far more heterogeneous group in terms of ideology and organizational orientation. Any further internal differentiation of this group has proved very difﬁcult and I have decided to regroup all these parties into a general category called ‘‘other left parties,’’ that is, the left that was neither orthodox communist nor ofﬁcial by socialist. These three families have varying weights within the overall European left. Socialist parties took part in all the 347 elections that constitute this sample. The average size of their vote was 27.7% with a 12.7% standard deviation. In the frequency distribution of the electoral results of all socialist parties, there is a discontinuity for the values between 30% and 38%. This indicates that this was a point of instability for the socialist performance. Thus, socialist parties reaching this level were either rapidly able (1979), and Haapakoski (1974). For a historical recollection by a former general secretary of the Finnish Communist Party, see Tuominen (1955).

110

THE POLITICAL MOBILIZATION OF THE LEFT

to overcome it, becoming large parties with over 40% of the vote, or rapidly regressed below that level, stabilizing at between 25% and 30%. Few socialist parties managed to stay at that level for a long period of time. Communist parties participated in only 217 of the 347 elections. A communist party is absent in the Republic of Ireland. In the United Kingdom, the Communist Party (CPGB) consistently won between 0.1% and 0.4% of the vote after 1918, but it has only won three seats in the House of Commons in its history (one in 1922 and two in 1945), and since 1964 most of its candidates have lost their deposits. The distribution of the electoral fortunes of the communist parties is totally skewed. Their mean result was 6.6% but their standard deviation is higher (7.97%), indicating the remarkable cross-country variation and cross-time instability. In more than three-quarters of these elections, Communists scored less than 7% of the vote. In fact, in all elections they have obtained more than 11% of the vote in only four countries: France, Finland, Italy, and interwar Germany. Therefore, Communist parties were either strong or marginal.87 Finally, the role of ‘‘other left’’ parties is even less pronounced. Such parties took part in less than one-third of all European elections: 114 of 347. The overall mean across the century is only 4% of the vote, and the standard deviation is as high as that of the Communist parties. Their electoral performance is on the whole modest: In only eleven elections did such parties manage to collect more than 10% of the vote, while 50% of the cases fell below 3.5% of the vote threshold. The electoral development over time of the three components for all European elections and by country is reported in Figure 2.6. The overall development of the left, discussed in the ﬁrst section of this chapter, is actually the result of the different trends of the three components. The socialist party electoral record is one of intense growth up to the 1910s, followed by a period of modest increases, which continued up to the 1950s. After this time, the socialist parties tended to stagnate, and by the 1980s a decline is evident. Communist parties, in contrast, grew moderately during the 1920s and 1930s and peaked in the 1940s, particularly in the 1945–1948 period, reaching their maximum of about 10% of the European vote. Since then, they have declined consistently, losing more than half of their percentage points. Finally, ‘‘other left’’ parties showed a third 87

Unfortunately, the exclusion of Spain, Portugal, and Greece for the reasons mentioned in Chapter 1 deprives us not only of three cases in which the communists made considerable electoral inroads, but also of the cases in which such inroads actually represented an intermediate level between the strong and marginal cases. In other words, their inclusion would give a more continuous character to the variance of communist electoral strength.

THE EUROPEAN LEFT

111

pattern of historical development. They had their maximum impact in the radicalization phase before and during the First World War, but then they tended to decline considerably and consistently from the 1920s to the 1950s. However, since the 1960s, their electoral performance has been growing again. This growth however, is not sufﬁcient to compensate for the decline of the communists and the stagnation/decline of the socialists. However, it is the only component of the left that has managed to increase its share of the electoral market in the last three decades. A close look at the graphs in Figure 2.6 is fascinating because it shows the great variety of over-time developments of each component that is concealed behind the overall development of the left in each country (speciﬁed in Figure 2.3). Moreover, these ﬁgures show how the growth– decline of one component relates to the growth–decline of the others in each country. A detailed descriptive discussion of these ﬁgures and data is not necessary and can be left up to the reader. I will limit my comments to a few important and general points. Given that there is an ofﬁcial Socialist Party in all the countries and an equally ofﬁcial communist party in all of them, with the exception of Ireland, then, in terms of composition, the electoral presence of other left parties is the main distinguishing feature between one left and another.88 No other left party has ever challenged the monopoly of the socialists and communists in Austria and the United Kingdom. Although other left candidates and groups have arisen from time to time, they have been relatively unimportant and can be easily overlooked (although they will be included in the electoral computation). In the United Kingdom, for instance, the British Labour Party never organized in Northern Ireland. Such groups as the Northern Ireland Labor Party, Republican Labour, and the Social Democratic and Labour Party were represented in the Stormont (i.e., the Northern Ireland Parliament) when it functioned and occasionally in Westminster. In the United Kingdom, there is the problem of how to interpret the 1981 split of the Social Democrats. This party was originally considered a truly other left party, which would make the United Kingdom the country with the most recent and late occurrence of left fragmentation. However, the group – consisting of thirteen Labour members and members of Parliament – rapidly merged with the Liberal Party, making their individual identiﬁcation impossible after the 1983 election. For the 88

I have not taken into consideration the development after 1985. Insisting on electoral presence is crucial. In the history of the socialist movements, there have always been numerous groups, associations, small parties, etc. in each country and in each period, which, however, have not contested the elections autonomously or, if they have, have not been able to gain any signiﬁcant or measurable support.

112

THE POLITICAL MOBILIZATION OF THE LEFT

Figure 2.6. Electoral development of the internal components of the left.

THE EUROPEAN LEFT

Figure 2.6. (cont.)

113

114

THE POLITICAL MOBILIZATION OF THE LEFT

purposes of this book, I have decided not to consider the Social Democratic split as a third left party.89 In Austria, notwithstanding its long experience of a proportional electoral system, a socialist splinter group emerged only once: the Democratic Progressive Party, which was formed from exmembers of the Socialist Party and led by a former minister of the interior. In 1966 it won slightly more than 3% of the vote, but afterward it dissolved immediately. In all the other European countries, however, the socialist/communist monopoly has always been challenged at one time or another. France is unique in that the class left panorama has always been characterized by the presence of more than one socialist party. However, before the formation of the Uniﬁed Socialist Party (PSU) in the 1960s by dissident socialists and Communists, the phenomenon of other left formations can be mainly regarded as an expression of the loosely structured French party system, characterized by the almost continuous presence in parliament of independent personalities, whose election has been facilitated by the emphasis on uninominal constituencies and the double ballot. Such personalities have regrouped under labels such as Divers Gauche or Inde´pendents de Gauche. In Italy, too, a third party can be traced back to the pre–World War I period, with a right-wing splinter party gaining 5.2% of the vote in the 1913 elections. In the two elections before the rise of fascism there, loosely organized groups of reformist Socialists were excluded from the ofﬁcial Socialist Party, which was sharply divided by the awkward dilemma of participating in the new International while maintaining the maximum unity compatible with this decision. Since 1945, the presence of third parties in Italy has been continuous as a result of the split of the Social Democratic Party from the PSI in 1947 following the latter’s orientation toward the Popular Front. The picture has since been complicated by additional mergers and new splits in the PSI and in the 1970s by the participation in national elections of groups on the extreme left, which combined some of the ideological features of the new left with a somewhat Leninist or pseudo-Leninist vocabulary and ideology. In the Netherlands, the presence of third parties in the left panorama has been a permanent feature since World War I. For almost the entire interwar period, a Socialist Party (SP, 1918–1925) followed by a Revolutionary Socialist Party (RSP, l929–1933) was present, but only in the 1933 election was the latter able to win more than 1% of the total vote. In the post–World War II period, a left-wing split from the Labor Party, result89

The best account of the British SDP is Crewe and King (1995). They ﬁnally regard the SDP as a political failure.

THE EUROPEAN LEFT

115

ing from disagreement on North Atlantic Treaty Organization policy, gave rise to the Paciﬁst Socialist Party (PSP, 1959) and provided the opening skirmish in a much larger battle that culminated in what may be considered the highest fragmentation of the left in Europe in terms of the number of parties. In the 1971 election, two more parties were added to the Dutch left. One, the Democratic Socialists 70 (DS, 1970), was a right-wing split from the Labor Party; the other, the Radical Political Party (PPR), which had originally been a left-wing split from the Catholic Party, evolved into a distinct left party and since 1971 has been considered to be a full part of the other left. Even though these parties lost considerably in successive elections, some of them have remained among the largest and most consistent other left forces in the European party systems. In Sweden and Germany, a three-part division within the class left was limited to the interwar period. In Sweden, as in the Netherlands, the 1908 radical split from the Social Democratic Party generated a Left Socialist Party, which refused to participate in elections until 1917. In 1921, the party joined the Communist International and founded the Swedish Communist Party. However, a minority decided to remain independent and stand for the 1921 election under the earlier label. In 1924, 1932, and 1936, the Swedish other left was composed only of splinter groups from the ofﬁcial Communist Party, sometimes joined by a few dissident leftwing Social Democrats. In Germany, too, the ﬁrst split occurred during the war years. The split of the SPD into majoritarians and independents was the most profound division of the socialist movement in Europe before the communist splits. However, it lasted for only a decade, and most of the Independent Socialists then provided the electoral basis for the growing German Communist Party. In neither Germany nor Sweden was there another split or a new party after 1945. It is interesting to note that there is no instance of other left party fragmentation emerging for the first time in the ﬁrst post–World War II period. Either such fragmentation was already a feature of the left or it emerged only beginning in the mid-1960s. In almost half of these countries, the presence of other left parties is mainly or exclusively a feature of the mid-1960s (Denmark, Finland, Norway, and Belgium) or even the 1970s (Ireland and Switzerland). In Denmark, this process was initiated by a right-wing split from the Communist Party, and it later escalated when the Danish Communist Party’s leader, Aksel Larsen, who returned from the Twentieth Congress of the CPSU looking for a national road to socialism, was expelled from the party. In 1958, he formed the Socialist People’s Party (SF), which was clearly located between the ofﬁcial pro-Soviet Communist Party and the Social Democrats. The left wing of the Danish

116

THE POLITICAL MOBILIZATION OF THE LEFT

political spectrum subsequently became crowded with other parties – one stemming from a left-wing split from the Socialist People’s Party (the Left Socialists, VS, 1968), the other from a right-wing split from the Social Democratic Party (the Centre Democrats, CD, 1973). In Finland, a leftwing faction of the Social Democrats began to organize itself in 1957, presented some candidates autonomously in the elections of 1958, and ﬁnally set up the Social Democratic League (TPSL) in 1959.90 The Norwegian picture is more complicated. Strictly speaking, an other left party had been formed in the 1920s. In 1921, a Socialist Democratic Workers’ Party (NSA) was formed to the right of the Labor Party when the latter decided to join the Comintern. The situation was unique in that there were thus two ofﬁcial socialist parties, one of which retained its name but was afﬁliated with the Third International. When this afﬁliation was terminated two years later, the Norwegian left was for a while divided into three parties – like the German, Italian, and Swedish left in the same period – because a minority of the Labor Party remained in the Comintern. This division lasted for only a short period, however, for in 1927 the Labor Party and the Social Democrats reunited. It was therefore only in the 1924 election that three left parties competed for the Norwegian electorate. A signiﬁcant and stable presence of third left parties in Norway can thus be dated with the recent rebirth in 1961 of radical socialism with the Socialist People’s Party (SF). Its success in the 1973 election, in alliance with the Communist Party, was mainly due to the difﬁculties experienced by the Labor Party in dealing with such foreign policy issues as membership in the European Economic Community (EEC) and NATO. This has coincided with the particularly strong impact of some typical new left concerns such as Third World ideology, paciﬁsm, and neutralism in a country where proximity to Sweden and the widespread awareness of an ex-colonial periphery constantly recall the advantages of neutrality. In Belgium, three other left parties emerged when the formidable shattering of the party system occurred due to the linguistic remobilization of the electorate. The ﬁrst of these parties, the Walloon Workers’ Party, competed only in the 1965 and 1968 elections and was regional; the other two emerged in the middle of the 1970s (All Power to the Workers– Labour Party in 1974; Revolutionary Workers–Socialist Workers in 1977) as more permanent and radical left groupings. However, all of these were 90

In 1973, the Social Democratic League of Workers and Small Peasants reunited with the Social Democratic Party. However, a fraction within it repudiated this alliance and founded the Socialist Workers’ Party (STP).

THE EUROPEAN LEFT

117

electorally insigniﬁcant. None ever managed to reach 1% of the vote, and only in 1965 did the Walloon Workers’ Party win a parliamentary seat. Finally, in Ireland and Switzerland, the birth of a new post-1960 left was delayed until the middle of the 1970s. In Ireland, mention should be made of a number of Independent Left representatives elected in the period 1923–1937 and of two others elected in 1977. Moreover, a splinter group of the Labour Party broke away in 1944 as the National Labour Party but reunited with the former in 1950. Also, the National Progressive Democrats, which contested the 1961 election and won two seats with 1% of the vote, is generally regarded by Irish commentators as a socialist formation. It later merged with the Labour Party. Even in this case, however, its participation was intermittent and of very little electoral importance. A continuous and consistent presence of other left parties in Ireland can be identiﬁed only in the mid-1970s and early 1980s. In 1973, a Workers’ Party began to compete with some success (between 1% and 3% of the vote up to 1985 and between one and three seats). There may, however, be some problem in considering it as a class left party, as it was the ofﬁcial wing of the Sinn Fein. A Socialist Labour Party competed in the 1981 elections (0.4% and one seat). Since 1982, a Democratic Socialist Party has entered the elections, winning about 0.5% of the vote. Both of the latter parties were founded by ex-Labour candidates or independent Labor candidates. In Switzerland, two small other left parties have existed since 1971, when a split from the Communist Party (Progressive Organizations) and a left-wing breakaway from the Ticino Social Democrats (Autonomous Socialist Party) stood for elections. Table 2.6 sums up the situation and indicates the periods in which other left parties have been present in the party systems. The presence of third parties has generally been intermittent and, in most cases, of little electoral signiﬁcance over the entire period. Only in Norway, the Netherlands, and Denmark does the total electoral strength of these parties exceed that of the respective communist party in a large number of elections since the 1960s. The only country in which the other left has reached a considerable electoral size is Denmark, where it ranges between 10% and 15%. In the same period, the second strongest other left parties are those of Norway, Sweden, and Italy, all ranging around the 5% threshold. Therefore, in the period since World War I and in the context of a given national size of the left, the variance in the size of the Communist parties is in large part the result of variance in the size of the socialist parties from which they originated and has been little inﬂuenced, so far, by the force of other left parties. This inverse relationship between the size of the Communist and socialist parties, however, presents, a clear discontinuity (see Table 2.7). For a ﬁrst group of countries, the size of the

118

THE POLITICAL MOBILIZATION OF THE LEFT

Table 2.6. Electoral presence of left parties other than communist and socialist parties

communist parties bears no relationship to that of the socialists. Equally small communist parties exist in the Netherlands and Switzerland, as well as in Norway, Sweden, Austria, and the United Kingdom, so that no inﬂuence over the size of the socialist support can reasonably be attributed to the Communist split. Moreover, the relatively small changes that took place after l945 in this group of countries indicate that the problem of the respective size of the parties – whether the communist party was to be large or small – was decided during the 1917–1944 period. On the other hand, for the remaining countries, it is evident that the greatest variations occurred after 1945. In three cases – France, Finland, and Italy – this change has led to the strongest European Communist parties. In one, Germany, it sharply reduced the size that the Communist Party had reached between the world wars. In only France and Italy did the Communist Party eventually manage to surpass the strength of the socialists. However, the group of countries that developed large Communist parties after World War II was already clearly distinguishable from the other group in the 1917–1944 period. In short, it might seem that the large postwar communist parties were already in nuce before World War II. The uncertain case of Italy and that of Germany, however, weaken this conclusion. Because the Italian Communist Party was not as strong as the Finnish or French parties between the world wars, it is difﬁcult to know whether this is due to the fact that Italian Communists participated in only one election against the socialists (1922) before the rise of fascism and the demise of democracy, or wheather it was due to its own weakness. In the second hypothesis, the post–World War II strength of Italian communism can be traced back to the fall of fascism, military defeat, and the resistance movement – that is, to the 1943–1945 period. The German case is quite the opposite. Communism here was from the beginning a strong political movement, as strong as those in France and Finland; but after the Nazi regime, it fell to an insigniﬁcant size. This

THE EUROPEAN LEFT

119

Table 2.7. Electoral size of socialist and communist parties and of the total left (period averages)

Note: The number of elections in which the party competed is given in parentheses. The party average electoral size was computed taking into consideration only the elections in which the party actually competed. Corresponding ﬁgures for the total left are given in Table 3.3.

major change in the position of the German Communist Party did not produce a corresponding social democratic growth, since after 1945 the SPD simply recovered its Weimar electorate. Even in this case, it is difﬁcult to say whether this is a case of Communist Party electoral failure, notwithstanding its strong electoral inroads in the interwar period, or, alternatively, whether it is the result of the powerful inﬂuence of the division of the country and the border with communist Eastern Europe. As a reaction to its uneasy international position, the West Germans played down any potential internal divisions and tended to look ‘‘with suspicion at any type of criticism parallel with, even if organizationally independent of, that coming from East Germany’’ and the Communist world.91 The negative statistical correlation between the electoral support of communist and socialist parties increases strongly after World War II (1918–1944 ⫽ ⫺.466; 1945–1965 ⫽ ⫺.776), thus indicating a closer link between the respective sizes. At ﬁrst glance, this could lead us to conclude that the postwar stabilization of the total left vote in Europe and the corresponding saturation of the left electoral market necessarily brought about more direct competition between Communists and socialists for this total left constituency. After World War II, these parties were less 91

Kirchheimer (1966a: 249).

120

THE POLITICAL MOBILIZATION OF THE LEFT

likely to grow together electorally and, in this sense, less likely to develop independently.

A SYNTHETIC MAP To conclude this descriptive chapter, I will summarize the discussion, offering a general classiﬁcation of the countries along the three main dimensions discussed so far – size, cohesion, and ideological orientation, that is, the main matrix of European left variance. Table 2.8 derives from the discussion in this chapter and more precisely from Table 2.3 (size), Figure 2.4 (early ideological orientation), Table 2.5 (war and revolutionary radicalization), and Tables 2.6 and 2.7 (Communist size and other left). It goes without saying that this is a very risky exercise. Dichotomizations force intermediate cases. Moreover, given the long time span, countries may belong to different classes in different periods. Nevertheless, it is hoped that a simpliﬁed ‘‘map’’’ will be useful to the reader as a guide in the following chapters. Two points should be underlined. First, I have resorted to a double classiﬁcation of national experiences only for Germany. Its comparatively different total size before and after World War II and the discontinuity in Communist Party size before and after World War II allow for no other Table 2.8. Synthetic map of the European left experiences

Note: Level of ‘‘war and revolution’’ radicalization given in parentheses.

THE EUROPEAN LEFT

121

solution. Second, on the ideological orientation dimension it is well known that there is a common historical trend toward a generalized deradicalization of socialist movements and that, therefore, individual national experiences need to be classiﬁed differently for different periods. I have tried to solve this problem through time by considering the early ideological orientation of the working-class movement in its genetic phase, from the 1880s to the World War I, by characterizing the war and revolution radicalization between the 1910s and the 1920s, and, ﬁnally, by considering communist electoral inroads for the period following this one. After World War II, left radicalism was to a very large extent linked to communist strength. This division means that one can try to identify more-radical workingclass political conditions in each of these three phases while also attempting to characterize the experience of the entire century. In this case, the most radical lefts will be those that fall into the polar type of orthodox Marxist or Marxist-syndicalist early ideological orientation followed by intense radicalization around the time of World War I and by the development of strong communist parties. The polar opposite is represented by lefts with an early trade unionist or pragmatic Marxist orientation, followed by low radicalization after World War I and by the absence of important communist splits. The main goal of this chapter has been to chart differences. The goal of the following chapters will be to interpret them. The questions that need to be answered are as follows: Why was the overall left mobilization higher in certain countries and lower in others? Why was its electoral development earlier in certain cases than in others? What macrofactors can account for the different patterns of organizational cohesion and division? Why did communism succeed electorally in certain countries and completely fail in others? Is it possible to make a general interpretation of the patterns of genetic ideological orientation and of the radicalization and deradicalization trends?

3

INDUSTRIALIZATION, URBANIZATION, AND LABOR’S RESPONSE ndustrialization and urbanization are regarded as preconditions of working-class political mobilization. These processes create and intensify the social problems and grievances of the working classes and lower classes in general; at the same time, they constitute the structural preconditions for these problems to become sources of organization and mobilization efforts. The resulting social mobilization gives rise to new social groups; it increases the self-awareness of those already existing; and it intensiﬁes existing conﬂicts and provokes the explosion of latent ones. Linking social mobilization processes with the left’s electoral mobilization postulates the following underlying causal chain: that the formation of the working class – in the sense of the creation and spread of given class conditions – impinges directly on the development of class consciousness, which in turn leads to the structuring of the class cleavage. According to this hypothesis, cross-country variance is to be explained by differences in the available proportion of potentially mobilizing voters, the latter in turn resulting from the quantitative spread and qualitative consequences of the formation of working-class conditions. This general hypothesis is the subject of this chapter.

I

CONCEPTUAL FRAMEWORK The belief in a relationship between industrialization, working-class formation, and the political response of labor is common to both historical socialist thinking and more recent post–World War II literature. However, 122

INDUSTRIALIZATION AND URBANIZATION

123

the emphasis on what this means is different. The relationship dominated the nineteenth-century debate on the ‘‘social question’’ and was at the heart of the theorizing of socialist thinkers. Marx’s analysis of the social consequences of industrialization implied the formation of a new type of worker in large, machine-intensive manufacturing, which he characterized as socially homogeneous, functionally interchangeable, and economically alienated, that is, ‘‘abstract labor’’ potential. For the socialists of the Second International, this discussion took the form of the debate over the ‘‘maturity of the socioeconomic conditions’’ that would allow for the passage to a phase of socialist society building. This was because Second International socialism believed in a long-term ‘‘constituency effect’’ of industrialization. Industrialization and economic development were perceived as benign forces, creating favorable conditions for the development of the labor movement in all its possible conceptions – in terms of quantitative development and the constitution of the critical mass, of ideological education and political consciousness, and of political action. However, around the turn of the twentieth century, doubts began to be felt about the linear relationship between industrialization, the growth of the working class, and the strengthening of political socialism. More clearly than anybody else, Bernstein suggested that this process was not linear. He envisaged the potential long-term integrative effect of economic growth as having an impact on the quantitative development of the working class, on its internal composition, and the consequent tendency toward political integration. However, opinions of this nature remained in the shadows for a long time. The labor movement’s development toward a reformist integration within the existing system was then interpreted in the light of the ‘‘aristocracy of labor’’ phenomenon,1 which some saw as a sheer and almost deliberate betrayal of ‘‘whole’’ class interest by better-off workers and their leaders and others considered more as a sociological tendency.2 In this context, the theory explained the ‘‘reformist’’ or integration deviation from a general path that was aimed at linking growing industrialization with the growing strength of radical (if not revolutionary) socialism. What was conceptualized as problematic and in need of an 1

2

This concept has its origins in Lenin’s work on imperialism (1939), although historians have traced it back to an article by Engels in 1885. For a detailed discussion, see Gray (1981). The thesis had various nuances. Moreover, even in a country like the United Kingdom, where the thesis was meant to explain most, there were convinced supporters and skeptical critics. Although Hobsbown endorsed it, Pelling and Joyce criticized it, arguing that, rather than being moderates, the labor aristocracy (craft and skilled sectors) were exactly the opposite: the most radical and militant sector of the working class. Hobsbawm (1964a), Pelling (1968a), Joyce (1980).

124

THE POLITICAL MOBILIZATION OF THE LEFT

explanatory theory was the failure of radical socialist mobilization of sectors of the working class.3 The basic theme of the renewed debate of the 1950s and 1960s can be introduced with a quote: ‘‘The growth, the structure and ideology of the labor movement of any country are conditioned by the nature of industrialization process, that is by the character of the tempo and direction of industrial development. This is not to say that political and cultural factors are not important. . . . But the dynamic element . . . appears to be the complex of events subsumed under the concept of Industrialisation.’’4 This thesis shows the sophistication of the original link between industrialization and working-class politics. It includes a complex dependent variable: the growth, that is, the pattern of quantitative development over time; the structure, that is, the internal composition and/or organizational setup; and the ideology, that is, the predominant Weltanschauung that deﬁnes the elements of consciousness, identity, and sense of destiny. There is also an equally complex independent variable: the character, tempo and direction of industrialization. The possible linkages implied by such a paradigmatic statement have, since the 1950s, been explored by a large number of writers. Most of this debate has been conducted by nonsocialist scholars. It originated in or was largely inﬂuenced by the totally different experiences of the labor movement in the United States; it was more systematically comparative and less linked to the experiences of the paradigmatic bigcountry cases; and ﬁnally, it witnessed a less linear conception of economic development and industrialization. While the latter two remained the essential independent variables explaining the development of labor movements,5 the post–World War II climate changed the conceptualization of the problem or, more precisely, the deﬁnition of what is in need of explanation. ‘‘Radicalism’’ became the pathological phenomenon to be explained. The 1960s witnessed an important debate on the issue of whether economic development and industrialization had a moderating integration effect on working-class politics or the contrary: a radicalizing and nonintegration effect. The question of the response of working-class politics to 3

4 5

The inﬂuential reconstruction of the paradigmatic case of working-class formation by Thompson (1963) is indicative of the resilience of this conception. Galenson (1952a: 105). To the best of my knowledge, Blumer (1960) is the only scholar who has not attributed a speciﬁc impact on working-class politics to industrialization. He explicitly considers industrialization a ‘‘neutral’’ process vis-a`-vis the working-class response. The latter, he claims, depends on a rather complex interaction of other factors. The list he proposes is rather nonparsimonious and ends with a conﬁgurative description of each national experience.

INDUSTRIALIZATION AND URBANIZATION

125

economic development was now posed in more problematic terms, namely, in reference to a typology of different forms of industrialization. The major problem of the debate was now the timing and the tempo of industrialization, that is, its earliness versus lateness and its gradual versus rapid growth and spread. According to Lipset, ‘‘economic development producing increased income, greater economic security, and widespread higher education largely determines the form of ‘class struggle’ by permitting those in lower strata to develop [a] longer time perspective and [a] more complex and gradualist view of politics. A belief in reformist gradualism can be the ideology of only a relatively well-to-do lower class.’’ Successful industrialization had the long-term effect of integration and domestication of the labor movement. Less successful industrialization had different political response outcomes.6 This general idea was taken up and developed later by several other authors. Lipset, in particular, introduced the concept of ‘‘half-way’’ industrialization. This industrialization was not ‘‘late’’ in temporal terms, but was instead incomplete and unbalanced in structural terms. In this case, the permanence of traditional family-based small businesses was disadvantageous to the development of strong, stable trade unions. The nonhomogeneity in the class condition prevented full-scale mobilization and integration in reformist movements and, at the same time, created the conditions for political division, with a commitment to radical anarchism, syndicalism, and so on of some sectors of the labor movement.7 Similar conclusions were advanced by Lorwin about the French labor movement and, in general, about the relationship between forms of economic development and working-class conditions; he attributed similar consequences to sluggish economic growth, small workshops, and lack of entrepreneurial risk taking.8 In the 1920s, Edward Bull, contrasting the rapid Norwegian industri6

7 8

Lipset (1960: 60). Actually, even before World War I, Commons had made a clear distinction based on a then typical comparison between the United Kingdom and the Continent. He distinguished between forms of capitalism characterized by gradual development of the factory system (in the United Kingdom) that were conducive to a reformist, integrated, and politically nonmilitant working class and a form of merchant capitalism, characterized by the deﬁnitely minor role of manufacturing, mainly performed in small-scale, scattered workshops (combining masters, journeymen, and apprentices), which tended to foster radical responses of anarchist, syndicalist, radical Marxist militant working classes and labor movements (the French case). He therefore emphasized the type of industrialization and economic development, rather than the quantity. See Commons (1962: 681–685). See in particular Lipset (1969). Lorwin (1954) and (1967b).

126

THE POLITICAL MOBILIZATION OF THE LEFT

alization through electriﬁcation with the slower and less dislocative industrial development in Denmark (Sweden being an intermediate case), pointed out the extreme radicalizing effect of late and rapid industrialization.9 The sudden displacement of workers from rural backgrounds and traditional surroundings to the turbulence of the new industrial surroundings created a working class that was less well versed in the socialist ideas, discipline, and traditions established in the consolidated sectors of the trade union movement. Late and rapid industrialization was thus a vehicle of social disruption and anomie, leading to radical responses and organizational fragmentation of class action.10 The thesis of the destabilizing force of rapid economic growth was later reasserted and made known internationally by Galenson,11 in line with the tenets of the literature on mass society and its anomic uprooting.12 It was later taken up and supported with varying qualiﬁcations by other studies, such as those of Halebsky, Mancur Olson, and Dubofsky.13 A large group of historical microstudies also tended to support the idea that the enlargement of the socialist movement beyond its original highly conscious craft and specialized workers took place through the recruitment of new workers from relatively homogeneous rural surroundings. Their completely new experience, the sudden breakdown of traditional values, and their great difﬁculties in adapting to and being integrated into the working-class movement and its organizations, dominated by skilled-craftsman membership and leadership characterized by different status and self-esteem, were signiﬁcantly correlated with radical feelings and new organizational forms.14 As usual, dissenting voices can be heard. William Lafferty’s very thorough control of Bull’s thesis for the Scandinavian countries has offered partial support for this – qualiﬁed, however, by a number of intervening variables that refer to the political integration of the working-class movement.15 R. H. Tilly, on the other hand, directly criticized Bull’s thesis in a study of German industrialization.16 9 10 11 12 13 14

15

16

Bull (1922: 329–361). For a worldwide application of this scheme, see Lipset (1958: 173–192). See in particular Galenson (1949) and (1952a). See in particular Kornhauser (1959). Halebsky (1976), Olson (1963), and Dubofsky (1966). See, for instance, Stra˚th (1983: 261–291), who compared the attitudes of the early working class in two Swedish cities (Malmo¨ and Gotheberg) and a German city (Bremen); and Lucas-Busemann (1976), who studied workers’ radicalism in two German towns of the Tuhr district. See the two fundamental works by Lafferty on Norway (1972) and (1974) and on the Scandinavian countries (1971). Tilly (1980).

INDUSTRIALIZATION AND URBANIZATION

127

The moderation/integration versus dislocation/radicalism interpretations of rapid social change can be reconciled, to some extent, by assuming a nonlinear relationship over time. In their early phase, industrialization and urbanization produce disruptive social effects, raise expectations beyond possible satisfaction, and result in radical responses; in their later phase, adaptation and learning, organizational responses, and the distribution of economic beneﬁts lead to integration and moderate responses. Such a thesis17 corresponds to an application to working-class politics of a transitional model of political development according to which modernization processes imply the passage from a premodern, relatively steadystate situation to a modern ﬁnal and again steady-state one after passing through a more or less prolonged transition phase during which social dislocation is at its peak, learning and accommodation processes are not yet effective, and radicalism, anomie, and instability are the likely results.18 The previous review points out the multifarious possible ways of operationalizing the process of industrialization. Industrialization sets the structural transformation that creates the phenomenon and the magnitude of the ‘‘class condition,’’ which can be seen in both quantitative and qualitative terms. In quantitative terms, industrialization determines the sheer size of the potential constituency for socialist mobilization in terms of sector (industry versus agriculture) and status (dependent versus independent labor) transformations. In qualitative terms, one can emphasize the character or the nature of the class condition, underlining the importance of the size of the workshop, the homogeneity of the composition of the working class, and the disparity between sectors and/or geographical areas. In addition, industrialization can be judged as early versus late in terms of timing and gradual versus rapid in terms of the tempo of the transformation. Both of these dimensions have been considered important factors in explaining labor’s response independent of the quantitative and qualitative aspects of the class condition formation. So, at each given point in time, the same level of industrialization in terms of structural sector or status transformation can be conceived as the result of a more or less early or late development; of a more or less gradual or rapid growth; and of the predominance of the factory system or of a merchant capitalism with a small amount of manufacturing centered in small, scattered workshops, of half-way, imbalanced, or incomplete industrialization. Table 3.1 summa17 18

See Soares (1966: 190–199). Examples of transitional models of political, social, and economic modernization, respectively, are those of Huntington (1968), Lerner (1958), and Fourastie´ (1963). For a discussion of the characteristics of transitional models of modernization, see Flora (1974). A similar transitional perspective is adopted in the critical analysis by Ulam (1979).

128

THE POLITICAL MOBILIZATION OF THE LEFT

Table 3.1. Dimensions and indicators of industrialization

rizes the possible dimensions of industrialization from the viewpoint of its potential impact on working-class formation and the possible different indicators that each of them implies and requires. At the conceptual level, the role of urbanization must be kept separate from that of industrialization. The theoretical dimensions of the process of urbanization are, ﬁrst, the spatial concentration of the population and, second, the growing division and specialization of labor. These two aspects apply to all historical forms of urbanization: ancient as well as medieval urbanism industrial urbanization, and the urbanization of developing countries. From an economic point of view, the lessening of the distance among individuals implied by spatial concentration minimizes the spatial ‘‘friction’’ of the market, which was one of the conditions most favorable to economic development (qualiﬁcation and mobility of labor, decreasing costs of infrastructure, economies of scale). From a sociological perspective, urbanization is mainly seen in light of the growing specialization and division of labor and of the formation of new social relationships, groups, and identities. The concentration of individuals in urban centers multiplies their contacts. It also changes the nature of their contacts, making them briefer and more motivated by speciﬁc goals. Urban contacts tend, therefore, to leave aside personal characteristics, and relationships with other people are based on role and function differentiation.19 Spatial concentration creates social competition, 19

On this point, see Simmel (1903: 185–206).

INDUSTRIALIZATION AND URBANIZATION

129

which in turn fosters the division of labor and a new type of social solidarity realized through differentiation, that is, through separateness and complementarity. The social consequences of urbanization are also crucial in the formation of new social conditions and the consciousness of them. In the passage from rural communities to urban quarters, loss of contact with the extended family and the soil, weakening of personal social relationships, detachment of place of work from residence, and loss of social and economic security all contribute to the formation of new groups and identities. The spatial concentration of the industrial proletariat tended to foster new forms of group solidarity. Class consciousness or, more generally, socialist ideology has been interpreted as a response to the ‘‘casual nature’’ of social relationships that result from the spatial concentration (workplace as well as residential) of people from very different backgrounds.20 Ideology and increased group consciousness can be regarded as the natural identity-building mechanism for people who share nothing or have very little in common, as a process that rebuilds a community of shared symbols facing the social dislocation that has destroyed the old community of provenance. From the political point of view, these economic and social characteristics and consequences of urbanization result in three related processes of political change. First, urbanization and urban concentration are the foci of new problem pressures. The articulation of new political demands is linked to the intensity of the socioeconomic problems generated by urban concentration. Second, urbanization indicates higher levels of communication and more sophisticated communication structures. Communication narrows the geographical and social distance between people and facilitates the spread of conformity of behavioral patterns and norms, as well as facilitating propaganda, recruitment, expression of interest, and imitation of organizational forms and experiences. Third, the new socioeconomic conditions of urban life can be conceived as particularly favorable to the ‘‘depersonalization’’ of politics and the passage from predominantly territorial to predominantly functional forms of political representation. Thus, in the nonurban setting, the transition from localism and personalism to modern national politics was slower. For individuals and groups to shift from indifference to political participation, a perception of their involvement in the nation and of the impact of national affairs on persons and localities was required. The latter, in turn, required that they be reached by a ﬂow of information, messages, agitators, and so on.21 20 21

Pizzorno (1962: 41). But even when this ﬁrst step was achieved, nonurban politics persisted, sometimes for a long time, in archaic localist and personalized forms. A masterly description of the

130

THE POLITICAL MOBILIZATION OF THE LEFT

Urbanization and the context of urban life are therefore particularly favorable to, and speciﬁcally more favorable than, rural or small village settings to the forms of electoral corporate and partisan mobilization that constitute the object of this study. Urbanization can therefore be imagined as an indirect indicator of the formation and concentration of the industrial proletariat that can be considered as complementary to, if not independent of, social stratiﬁcation data that indicate mainly its quantitative dimension and formation. The extent to which industrialization and urbanization are correlated at different points in time and in different countries is an empirical question to be addressed in the rest of this chapter. Moreover, urbanization points to preconditions facilitating the development of socialist organizational mobilization, recruitment, and propaganda: The more urbanized the context, the easier it is to recruit organizational membership; the more developed the latter, the stronger the socialist electoral support.22 Finally, urbanization indicates a set of social relationships and conditions that reduce the weight of traditional forms of social control by established elites and institutions and that build areas of status equality facilitating participation and collective action.23 This short review clariﬁes the wealth of linkages that can be established between the processes of industrialization and urbanization, on the one hand, and the political responses of labor discussed in Chapter 2, on the other. In this chapter, I concentrate on the relationship between the quantity, character, timing, and tempo of industrialization and urbanization and the overall electoral mobilization of the class left. The relations between industrialization/urbanization and radicalism/organizational cohesion of the labor movement will be discussed separately in Chapter 9.

WORKING-CLASS CONSTITUENCY FORMATION AND SOCIALIST MOBILIZATION The great transformation that took place in the European economy after the end of the Napoleonic wars and brought about the mature industrial

22 23

archaism of local politics in rural France can be found in Weber (1979: 241–277). There the gap in political mobilization between countryside and urban centers was particularly wide. Similar accounts can be found in almost all countries. For Italy, see De Sanctis (1876); and for Norway, see Rokkan (1970c). More generally, for the special conditions of national politicization in nonurban settings, see Urwin (1980: 87–110). See Linz (1959: 346–350). Hintze (1980: 15) has argued that these elements tended to favor the birth of parties rather than estates. It is therefore worthwhile to investigate the potential autonomous impact of urbanization on political mobilization.

INDUSTRIALIZATION AND URBANIZATION

131

society will be considered here only from the labor allocation point of view.24 In a discussion of industrialization in reference to the potential mobilization capacity of socialist movements, it is logical to emphasize the changes in social structures that resulted from the rapid growth of industry and to concentrate on socio-occupational consequences. These were expressed in two major processes of labor force displacement: One concerned sector shifts; the other, status shifts. In the ﬁrst case, this structural change was characterized by an accelerating shift of the labor force from agricultural to industrial activity and, later, to services.25 With its mass production dominated by an increasing use of technology and inanimate energy, industrialization entailed a progressive and continuous shift of the labor force to industry, to the point where mature industrial society was largely dominated by industrial occupations, while both agricultural and traditional service sectors declined. In terms of occupational status – rather than sector – the same long-term processes determined a shift of labor from independent to dependent wageor salary-remunerated positions. Next to the progressive reduction of the employees, particularly the self-employed, status shifts within the dependent labor force strengthened a homogeneous category of workers at the expense of less well deﬁned and less homogeneous groups such as family workers and apprentices and, in a second phase, determined a progressive increase in the number of salaried employees.

THE MATURITY OF INDUSTRIAL SOCIETY The transformation associated with sector shifts best suits the aim of deﬁning structural conditions for a mature industrial society. It is characterized by three universal trends: the decline of the active population in agriculture; the growth of an active population in industry; and – at a later stage, according to the theory – the growth of the active population in the services. Each of these trends can be looked at comparatively across coun24

25

Other measures refer to productivity, investments, production outputs, or similar economic indicators. The GNP measures are not necessarily a direct indicator of the level of industrialization. On the other hand, indices of sector production are more varied and difﬁcult to compare. Of the various attempts to measure the level of industrialization in terms of output per capita, the most systematic and reliable one concentrates on manufacturing output, excluding not only construction, transport, and utilities, but even mining. Bairoch (1982) reviews all attempts to measure industrial production directly and provides a rich appendix on the method followed and the problems involved in arriving at his estimates. At the end of the 1950s, Fourastie´ (1963) and others predicted a further and different long-term sector transformation destined to lead to the ﬁnal predominance of a new modern service sector.

132

THE POLITICAL MOBILIZATION OF THE LEFT

tries in Table 3.2, where the percentage of the active labor force in each sector is reported by country in the decades between 1880 and 1970 on the basis of census data.26 The ﬁgures in Table 3.2 help to show comparatively the transformation in each sector. The average level of agricultural occupation in Europe declined from a mean of about 45% of the active population in the decade 1880–1890 to about 12% in the 1970s, a drop of 33 percentage points.27 In industry, the same time span saw the Western European mean increase from about 25% to about 40% of the active population. In service occupations, the corresponding increase is from about 24% to about 45%. Therefore, of the roughly 35 percentage points lost by agricultural occupations during the century, about 15 have been gained by industry and about 20 by services. So, on average, service occupations have grown more than industrial ones. The huge disparities in sector occupation before World War I lessened over time, and in the 1970s they were minimal with respect to the starting point. This implies that the rate of change was particularly accelerated or particularly slow for those countries that were further from the European mean value in the early phase. The difference between the highest and lowest percentages of agricultural occupation was about 60% in the 1880s (10% in the United Kingdom; 70% in Finland), and only 20% in the 1970s. In industrial occupation, it was about 30–35% in the 1880s (40% in the United Kingdom and Switzerland; less than 10% in Sweden)28 and 20% in the 1970s (about 50% in Germany and Switzerland; about 30% in Ireland).29 In the service sector, the early difference was about 30% in 26

27

28

29

This time span is imposed by the availability of comparable census ﬁgures throughout Europe. However, before that date, industrial development had progressed extremely slowly, not only worldwide, but also in Europe. The changes that had begun in English industrial production in the mid-eighteenth century took over half a century to be imitated and followed elsewhere. Considering also that, with the exception of the United Kingdom, Switzerland, and Belgium, the rate of population growth exceeded the rate of growth of industrial output, it is likely that the nature of socio-occupational changes brought about by industrialization before the 1830s was limited. The rate of growth in industrial output in Europe began to speed up after the 1860s and accelerated at an even faster rate after the depression of the 1870s to 1890s. See Bairoch (1982: 272, 274, 290) and Rostow (1978: 111–162). The decline of the agricultural population in this table, as well as in the following tables of this chapter, is particularly accentuated by the fact that it is expressed as a percentage of the active population or of other groups. The absolute number of people involved did not decline so rapidly, and actually, according to Dovring (1964: 78–98), it may have increased in the early stages of economic development. No data are available for Ireland and Finland in this period. They probably had an even lower percentage of active population in industry. This process of leveling off is also underlined by measures of industrial output. Indeed,

Table 3.2. The European sector transformation (1880–1970)

134

THE POLITICAL MOBILIZATION OF THE LEFT

the 1880s and about 15% in the 1970s. Of the three trends, that of the service sector is characterized by a more pronounced process of homogenization across the countries and also by a less pronounced spread of all countries around the European mean. A further aspect worth underlining is the considerable acceleration in all three trends from the 1950s on. Up to then, cross-country differences were still quite marked and tended to persist. From the 1950s on, all countries rapidly concentrate around European means and the standardization of the sector composition of European labor forces progresses rapidly. Finally, the homogenization process is general and without exceptions or reversals for agriculture and service, although this is not the case for industry. Indeed, the percentage of the population active in industry starts to decline at the European average level only in the 1970s, when early signs of deindustrialization begin to appear. At this stage, deindustrialization reintroduces elements of differentiation among Western European countries, whereby those above the European average anticipate the future trends of all the others. The United Kingdom, Belgium, Switzerland, the Netherlands, and, to a lesser extent, Germany are clear earlycomers, anticipating the temporal trends with a greater than average industrial sector and a weaker than average agricultural one. Finland, Ireland, and, to a lesser extent, Italy and Austria, are clear latecomers, with a weaker than average industrial sector and a greater than average agricultural one. France locates along the allcountry average in all three sectors. The three Scandinavian countries started as latecomers but rapidly reached average European levels in all sectors. In Sweden, the growth of the industrial sector and the decline of the agricultural sector were particularly accelerated. Finally, two of the earlycomers – the United Kingdom and the Netherlands – are also countries with very large service sectors, while among the latecomers, Finland and Italy (Austria too, but only until the 1930s) are not only predominantly rural but also underserviced societies. It is interesting to compare this characterization of Western European countries in terms of sector transformation with their industrialization level, as indicated by manufacturing output per capita. Table 3.3 reports these levels for selected years in relation to the British level in 1900, which is made equal to 100. With the exclusion of the United Kingdom, Belgium, and Switzerland, between 1750 and 1860 industrialization had it is even more marked considering the manufacturing output per capita due to the early deindustrialization of this sector in the very earlycomer regions, such as England, the Belgian Walloon, and the French northeast. See Bairoch (1982: 300–306).

INDUSTRIALIZATION AND URBANIZATION

135

Table 3.3. Levels of manufacturing output per capita (UK in 1900 ⫽ 100; triennial averages except for 1913)

Note: Austria-Hungary up to 1913; after 1938 East and West Germany. Source: Bairoch (1982: 294, 302, 330–1).

proceeded very slowly indeed. These three countries remain the forerunners, with Italy, Finland, and Ireland the latecomers. However, there is great variation in the ranks of the other countries. Germany, having started somewhat later, had by 1900 already reached the Belgian and Swiss level.30 France, by contrast, industrialized relatively early, by the 1880s, but after this period its development decelerated and stagnated. Indeed, by 1938, the per capita industrial output of Sweden, Norway, the Netherlands, and Denmark had passed that of France, and the output of Austria had reached it. Another case worth mentioning is that of the Netherlands. The sector composition of the labor force in this country appeared extremely advanced, but its manufacturing output did not correspond to this image. The key probably lies in its large service sector. For Norway, the lower ranking in output rather than in sector structure is probably attributable to the exclusion of electric-energy output from the industrial output. Italy is a clearer example of a latecomer more in industrial output than in occupational structure. Sweden, by contrast, shows an impressive level and rate of industrialization: By 1963, it had become the Western European country with the highest per capita manufacturing output. On the whole, however, all the differences are not so pronounced, and the relative ranking 30

On the speciﬁc relation between industrialization and labor force growth and change in Germany in reference to other earlycomer countries see, Zwahr (1987).

136

THE POLITICAL MOBILIZATION OF THE LEFT

of the countries is not modiﬁed in any substantial way in considering industrial output instead of sector occupation. It is reasonable to expect that the most favorable conditions for socialist mobilization would emerge in those societies that experienced early attainment and prolonged duration of the occupational structure typical of a mature industrial society. To describe a mature industrial society, we need to use thresholds for deﬁning a prevalently agricultural society as opposed to those that are industrial or postindustrial. Industrial society enters its mature phase when the proportion of the active labor force in industry overtakes the proportion of the labor force in agriculture; postindustrial society begins when the population active in industry is bypassed by the proportion active in providing services. According to the ﬁgures on European occupational sectors elaborated by Kaeberle (and reported in Figure 3.1), in Western Europe a mature industrial society, as just deﬁned, appeared between the 1920s and the 1970s. The industrial sector predominated over the service sector from the beginning of industrialization to the 1970s, but only in the 1920s did industrial occupation bypass agricultural occupation.31 With respect to this Europewide picture, national differences concerned not only timing but also patterns of structural differences. A mature industrial society was not only reached earlier or later, but could be shorter or longer, not exist at all, or develop with completely different features in some countries. Using the threshold mentioned earlier, in Table 3.4 I have cross-tabulated the countries according to when the industrial societies were reached and according to their duration. The countries that best indicate the European model of sector transformation are undoubtedly the United Kingdom, Belgium, Switzerland, and Germany,32 because an early decline of the agricultural sector and a very long period of predominance of the industrial sector over the service sector is evident in all of them. In the United Kingdom, industrial society dates back to the 1820s. In this 31

32

Kaelble (unpublished manuscript) has argued that the idea of a sequence from agricultural to industrial and ﬁnally to postindustrial society, and the notion of industrial and postindustrial society make sense only in the European context. For other parts of the world, both developed and underdeveloped, such categories do not have the same signiﬁcance. Kaelble, in the previously cited manuscript, produces a similar set of ﬁgures and tables for a number of European and non-European countries. My tables are based on different data taken from the cited source of Peter Flora. This produces slight and insigniﬁcant differences in the graphs. What is more important is that the conclusions and the classiﬁcation of individual country experiences are different, as Kaelble concentrates exclusively on the relationship between the service and industrial sectors, leaving aside data about the agricultural one, which from my point of view are essential for the identiﬁcation of the maturity of industrial society.

INDUSTRIALIZATION AND URBANIZATION

137

Figure 3.1. Thresholds of mature industrial society in Europe. Note: Data in this ﬁgure are not means of European countries. They represent actual occupation ﬁgures summed up for all European countries and computed as a percentage of the Europewide active population. Source: Kaelbe (unpublished paper: 43). case, however, by the 1930s the service sector was almost as large as the industrial one. Here, the predominance of industrial occupation over agricultural and service occupations lasted for almost a century. In a second group of northern countries, the pattern is strikingly different. The beginning of mature industrial society can be identiﬁed at different points in time. In the Netherlands, this point was reached very early, in the 1890s33 (that is, more or less at the same time as the four countries discussed earlier); in Denmark, in 1910; in Sweden, in 1930; in 33

The Netherlands is slightly different from the other earlycomers. Notwithstanding its advanced sector composition of the labor force, industrialization was not very early nor was it characterized by big factories and industrial concentration: ‘‘The Netherlands industrialized relatively late, and until 1914 was typiﬁed by extensive agrarian and merchant capitalist sectors. The small ﬁrms remained dominant for a long time, while industrial concentration developed only sporadically. These factors explain the birth of a

138

THE POLITICAL MOBILIZATION OF THE LEFT

Table 3.4. Timing and length of industrial society and left electoral mobilization

Note: Numbers in boldface represent all-countries averages

Norway, in the 1940s; and, ﬁnally, in Ireland, in the 1960s. What makes this set of countries an interesting separate group is not the timing of the beginning of mature industrial society, but rather the fact that industrial predominance never really materialized. In all these cases, the service sector always remained quantitatively more important than the industrial one. So, while industry eventually predominated over agriculture, both remained quantitatively less important than services. Therefore, these countries therefore do not adhere to the model sector transition since something made them service economies from the beginning. Finally, in France, Austria, Italy, and Finland, the common feature is the late arrival of industrial maturity: in the 1940s, 1950s, and 1960s, respectively.34 At the same time, all these cases are characterized by a

34

movement of artisans from the 1860s onwards when mechanisation began to take place, and also the limited extent of this craftsmen’s movement, the relatively late growth of national trade-union organizations, and the weak organizational structure which often did not allow the permanent entrenchment of these organizations. Only at a much later stage, when accelerated industrialisation began and a real industrial working-class started to form, did the artisans’ movement begin to change into a workers’ movement – if limited in the ﬁrst instance to a workers’ elite – and this was accompanied by changing organizational structures’’ (Buiting, 1990: 81). Ireland should be added from this point of view. But, as mentioned, its most striking feature is the dominant role of the services. This aspect differentiates it from the third group.

INDUSTRIALIZATION AND URBANIZATION

139

marked balance between occupation in industry and services, with the result that industrial occupation remained relatively low. In Italy and Austria, the predominance of industry over services is marginal; in Finland and France, it never materialized.35 As can be expected, the earlycomers to industrial predominance maintained their predominance for a long period, while the latecomers were necessarily characterized by a shorter phase of pure mature industrial society. The most interesting cases are those in the ﬁrst two rows of Table 3.4: societies that went from an agricultural to a service economy without passing through a clear industrial phase. Three types can be distinguished: early industrializing countries, where industrial predominance never existed (the Netherlands); late industrializing countries, distinguishing in terms of the length of industrial predominance between the group of northern-periphery countries, where such predominance never existed, and a southern-continental group, where such predominance was short (Austria, France, and Italy). In Table 3.4, I have added to each country the mean level of electoral mobilization for the whole 1880–1965 period and for the early pre–World War I period (to check if it is the earliness of electoral mobilization, rather than its size, that is associated). The data do not correspond to expectations. Each category is nonhomogeneous, and in every period the level of left electoral mobilization is higher in the latecomers to industrial society than in the earlycomers (31.9% versus 27.8% for 1880–1965; 22.9% versus 17.5% in the early period; 38.4% versus 34.6% in the post–World War I period 1918–1965). The mean level of electoral mobilization of the left is not enhanced by the maturity of the industrial society; if anything, it is the opposite. The same can be said for the amount of time that industrial society existed. Whether this period was long, short, or nonexistent does not produce any signiﬁcant differences in the electoral mobilization potential of the socialist left. In each period, the difference between the two most opposite groups of countries (early and protracted versus late or never-achived industrial society) is always favorable to the latecomers and the service-dominated societies: 27.8% versus 31.4% for the whole period; 18.5% versus 25.9% for the 1900–1915 period). Let us now look at the sector transformation data from a different perspective. Rather than considering mean levels of mobilization by type of industrial society, let us consider the association between populations active in different sectors and the levels of electoral socialist mobilization 35

It is interesting to note that this pattern looks rather ‘‘southern.’’ In fact, the cases of Spain and Portugal are rather similar to those of France and Italy. Greece is more a case of continuous predominance of the service sector. See the data in Kaelble (unpublished manuscript: 41–42).

140

THE POLITICAL MOBILIZATION OF THE LEFT

for the entire set of elections held in the countries, pooling all the data together, without any reference to particular countries. In Table 3.5 the correlation coefﬁcient is reported by economic sector, by period, and by decade. Over the whole period (1880–1975), the levels of electoral mobilization are associated with the industrial transition as expected: negative association with agricultural occupation and positive association with both industry and service occupations. However, this association concerns data collected over time and with clear trends. The left tends toward secular growth, as well as industrial and service occupation, as we have seen, while agriculture declines sharply. These covariations, therefore, may point to historical multicolinearity rather than to causal linkage. To control for covariation in time, we can look at the associations over shorter periods or even single decades. As the census data are collected on a ﬁveto ten-year basis, within each decade occupational data do not vary much. For each point in time, the association is no longer blurred by potential linear temporal covariation but indicates only the cross-space association. This means that a temporal hypothesis – whether the growth (or decline) of A is associated with the growth (or decline) of B – is transformed into its cross-sectional synchronic equivalent, that is, whether higher (or lower) levels of A are associated with higher (or lower) levels of B at each given point in time. Looking at the data by period produces a signiﬁcant change in the picture. In the period from 1880 to 1917 (Ireland is missing), there is a striking positive association between agricultural-sector size and total left vote and a negative association between industrial and service sector size and total left vote. In the following two periods, the situation is reversed and agricultural sector size is not associated with electoral mobilization. However, the association levels are weak or nonexistent. Considering decades, the associations are important up to the 1930s, becoming smaller thereafter. Moreover, in the early phase, higher levels of socialist electoral mobilization are positively associated with high agricultural-sector occupation and negatively associated with the size of industry and service sectors. Only in the period 1930–1960 is a positive association between industrial occupation and left mobilization evident. In conclusion, whether one considers the timing of industrialization, the length of the industrial-sector predominance, or levels of sector occupation, it does not seem that socialist mobilization was earlier or stronger in the more industrially advanced economies. Only in a global developmental perspective – that is, over the whole period of time covering the occupational transition – does the sector transformation of the labor force seem to inﬂuence levels of socialist electoral support. Political mobilization is asso-

INDUSTRIALIZATION AND URBANIZATION

141

Table 3.5. Correlation between active population in the economic sector and electoral socialist mobilization

Note: Number of elections given in parentheses.

ciated over time with the transition from predominantly agricultural economies to industrial and service ones; however, any synchronic control concerning the signiﬁcance of the early attainment of such a transition, or its more or less protracted nature, fails to offer a clear base of discrimination among national patterns of socialist political mobilization.

STATUS TRANSFORMATION Status transformation is historically a process of shifting from independent and self-employed status position to dependent ones and, within the dependent labor force, trendy processes of change concerning employees, workers, apprentices, and family workers. Census data concerning the status distribution of the active populations of European countries present problems of comparability represented by the crucial category of ‘‘family workers’’ and their uncertain status.36 I have chosen to add wives, sons, or 36

Most censuses distinguish family workers from both dependent and independent groups. Some include family workers in the category of independents (as, for instance, the early

142

THE POLITICAL MOBILIZATION OF THE LEFT

close relatives of the self-employed to the ‘‘independents’’ category because this minimizes comparability problems for all those cases in which the two categories are not clearly differentiated in census data; it also makes sense with respect to the speciﬁc goal of this analysis.37 However, to control for differences between the two ways of deﬁning the independent group, in the following analysis I have also reported the data concerning the independents taken in the more restricted sense, that is, excluding family workers. As these data are less interesting in the context of this analysis, I have decided not to present them in table form. The global picture is the tendency toward the progressive decline of independent labor to the advantage of dependent labor. At the European average level, this trend has progressed at a relatively slow pace and has accelerated only since the 1950s. In the 1940s, more than a third of the average active population in Western European countries was composed of employers, the self-employed, and family workers; from 1880, the decline was only 5–8% in about seventy years. Countries can be classiﬁed into four different groups. In the ﬁrst group, dependent labor prevailed early over independent labor, and the bulk of the historical transformation was accomplished in the 1880s, before electoral enfranchisement and party mobilization. In the United Kingdom, Belgium, and the Netherlands (and, to a lesser extent, in Germany), the over-time trend is almost imperceptible because, since the beginning, dependent labor accounted for 70% or more of the active labor force. Next are Denmark, Norway, and Switzerland, most of whose labor was dependent from the beginning of the period and tended to follow the general European average path of transformation. France and Italy must be considered a third group in which dependent labor was more numerous than independent labor from the 1880s, but the difference was minor and remained so for a long period of time. In these relative latecomers to industrial society, the proportion of independent

37

censuses of Austria and Italy). In a few cases, some doubt remains about their attribution: in particular in the 1881–1891 French censuses and in the Dutch ones up to the 1930s. Finnish data present different problems. A separate category exists for family workers, but up to the 1940s employees were counted together with the independents. In this case, I have resorted to linear extrapolation from later censuses to deduce the percentage of employees from the independent group. Whether to exclude or include family workers among the independents is an important decision, given that they are often a substantial proportion of the active population, particularly in agriculture. In the censuses of this century, they represent, on average, roughly 26% of the active population in Finland and about 16% in Italy and Austria (at the other extreme, only 0.4% in the United Kingdom). See Flora, Alber, Eichenberg, Kohl, Kraus, Pfenning, and Seebohm (1987: 445) on the problem of family members in European censuses.

INDUSTRIALIZATION AND URBANIZATION

143

labor remained virtually unchanged until the 1940s–1950s at about 40– 45% of the active population. After 1950, these countries caught up rapidly with the others, with higher rates of change. Finally, there is a fourth group of countries in which independents labor was more numerous than dependent labor until well into the twentieth century. In Finland, this was the case up to the 1950s; in Ireland, up to the 1930s; in Austria and Sweden, up to the 1910s. In Table 3.6 the levels of left political mobilization are related to (a) the four different patterns of status transformation and (b) the levels of dependent/independent occupational status. The results are surprising. From earlycomers to latecomers, passing through two intermediate categories of countries that closely follow the European average pattern, and of those that have minoritarian but stable independent sectors, the left electoral mobilization increases. This overall situation is determined by the earlier period, where the level of left mobilization of latecomers is twice that of all other groups. In the interwar and post–World War II periods, the differences tend to disappear and the four groupings become insigniﬁcant. Considering the correlations in part (b) of Table 3.6, signiﬁcant associations are reported in the expected direction over the entire set of elections in the century. The growth of the left electoral force is associated with the transformation of the status condition of the labor force. Considering that the status transformation trend is less marked than the sector transformation trend, and that rates of change are more muted, these associations appear more signiﬁcant than those previously reported concerning the sector shift in labor. However, even in this case, such overtime association points to a possible developmental cause that should also be checked by cross-sectional measures. In fact, the period and decade data immediately modify the overall developmental picture. The problem is the ﬁrst period and the early decades, when the total left vote is associated positively with independent labor and negatively with dependent labor, contrary to the whole developmental association and to what happened in the other two periods. This is because, between the end of the nineteenth century and the ﬁrst two decades of the twentieth century, the left electoral mobilization was not higher where the percentage of dependent labor was higher – quite the contrary. The contrast between cross-time and cross-space associations leads to conclusions similar to those reached with sector data. The overall crosstime association is in the expected direction, but the slices of synchronic comparison do not allow us to transform this developmental association into the idea that the early timing or the levels of status transformation were positive preconditions for left mobilization. We can read these results

144

THE POLITICAL MOBILIZATION OF THE LEFT

Table 3.6. Status transformation and electoral mobilization

Note: Number of elections given in parentheses.

in two different ways. First, this analysis can be read as conﬁrmation of the fact that early left development took place more easily and strongly in countries that were latecomers to the industrial transformation, taking the high proportion of independent labor as a further indicator of such late transformation and leaving aside at this stage the question of why this was the case. Alternatively, the analysis could be read as indicating a more direct causal link between the size of the social group of independent labor and the vote for the left, in the sense that parts of this group, probably to be identiﬁed among the agricultural independents and particularly the small farmers, offered direct electoral support to left parties in lateindustrializing countries. Some support for this idea comes from considering the category of independents that excludes family workers. Family

INDUSTRIALIZATION AND URBANIZATION

145

workers are heavily concentrated in agriculture, and their inclusion in the category of independents makes the latter largely dominated by the agricultural sector. In addition, a large group of family workers points to an agrarian structure characterized by a large number of small independent farmers. When we exclude them, the negative association between independents and left vote is strengthened, doubling over the whole period and becoming more negative with each period and decade (Table 3.6).38

WORKING-CLASS CONSTITUENCY The previous discussion centered on the relative position of individual countries in the industrialization process. In this section, I continue the analysis, considering the size of the working-class constituency, which identiﬁes the putative support base of socialist parties. The size of the constituency is not only a presumed quantitative asset; it also produces social and behavioral effects under the thesis that the homogeneity of a social environment – that is, the numerical predominance of a given social group in a community – tends to increase the social pressure of the group on its members toward attitudinal and behavioral conformity. The larger the proportion of a social group in an area, the higher the frequency of social contacts among members of the group; and the higher the frequency of these social interactions, the greater the social pressure that forms and reinforces group identity and organizational behavior, and the greater the sanctions for group nonconformity. Therefore, the more dominant a social group is in the local context, the more uniform its political behaviors are likely to be.39 Tingsten’s argument was developed to explain cross-area differences, and has been mainly used and tested at the level of withincountry variance,40 but there is no reason why it should not be used at the 38

39

40

The difference introduced by the exclusion of family workers is enhanced if one considers that the category of independents including family workers is more negatively trendy over time than the category of independents excluding family workers. So, one would expect independents with family workers to correlate more negatively with the growing trend of left electoral development. The reverse is actually the case. This thesis was formulated and tested explicitly for the ﬁrst time in Tingsten (1937: 177–180). He based this conclusion on an analysis of district voting in the city of Stockholm in the 1932 elections. Mixed social areas are contrasted with homogeneous ones in terms of the corresponding homogeneity of political behavior, in particular electoral participation and class voting. For instance, Butler and Stokes (1970: 144–48) ﬁnd that the support for labor grow in a curvilinear way with the increase in the percentage of workers in the district. Worlunbg (1990: 49–57) has studied the variance in support for the Swedish Social Democratic Party between 1921 and 1940, ﬁnding a straight-line relationship rather than a curvilinear one.

146

THE POLITICAL MOBILIZATION OF THE LEFT

cross-country level as well. Working-class size should therefore have both sheer quantitative potential constituency effects as well as qualitative social pressure effects, both of which, in principle, foster socialist political recruitment and mobilization. Workers can be deﬁned as lower-status dependent laborers or bluecollar personnel, including home workers.41 The working-class constituency is the result of the combination of status and sector dimensions, and the type of working-class constituency changes according to how many sectors and sub-sectors there are to consider. So the ‘‘working class’’ can be deﬁned as the proportion of lower-status laborers or blue-collar personnel in all sectors of economic activity – industry as well as agriculture and service. Alternatively, the number of sectors can be limited to the industrial ones, thus arriving at a more restricted deﬁnition of ‘‘industrial working-class.’’ Within industrial subsectors, the key ones, such as manufacturing and mining could also be selected for an even more restricted concept of ‘‘core industrial working class,’’42 mainly in reference to largescale productive units in core industrial sectors. Notwithstanding the early socialist thinkers’ conviction that the working class was already or was soon to become the majoritarian group in society, this was hardly the case in Western Europe, even adopting the broadest possible deﬁnition of ‘‘worker’’ so as to include all dependent lower-status and blue-collar workers in all possible sectors of economic activity. Table 3.7 reports the percentage of this social group in the active population in Western European countries between the 1880s and the 1970s. The United Kingdom and Ireland are not included, for up to the 41

42

The distinction between workers and employees is problematic: Over time and across countries, the census distinction between the two groups may be different, sometimes emphasizing the type of work (in particular, manual versus nonmanual), the place of work (in particular, factory versus ofﬁce), and ﬁnally the legal status (in labor and social security laws). However, for a macrocharacterization of cross-country differences, these problems can be considered minor, and I have therefore relied on the standardization made by Flora and his associates. The measures of industrialization based on industrial output generally exclude not only transport, building, and services, but also mining (and electrical power). Manufacturing output is regarded as the best indicator. This is based on two arguments. The ﬁrst is that if one wants to differentiate advanced industrial countries from less developed ones, this is necessary because expansion of the mining industry has been a feature of underdevelopment in the Third World. The second argument is conceptual: If one includes mining in industry, one must also include agriculture, which for a long time was one of the main sources of raw materials for industry. The coal miner is therefore regarded as outside manufacturing even if his work is destined to serve the latter. See the discussion in Bairoch (1982: 321–322). In a homogeneous Western European context and in reference to electoral mobilization problems, it is natural to speak of an ‘‘industrial working class’’ that also includes mining, building, etc.

INDUSTRIALIZATION AND URBANIZATION

147

1960s there prevailed in these countries the tradition of counting ‘‘workers’’ and ‘‘employees’’ together in national censuses (see Appendix); the Dutch ﬁgures up to the 1930s are possibly overestimated due to the inclusion among ‘‘workers’’ of other unknown categories, including some family workers (but not assisting spouses, who are kept separate). Nevertheless, it is clear that lower-status dependent labor groups were never as overwhelmingly majoritarian in Europe as has often been implied. In addition their variation over time during the century is modest and, if anything, has a negative trend. The European mean ranges from a starting point of 55% in the 1880s to about 47% in the 1970s, the other half consisting of the independent, the self-employed, family workers, and employees. Even in the group of forerunners of industrialization, the numerical predominance of workers tended to decline over time, and by approximately 1940, workers represented 50% or less of the active population. All the other countries show levels that were clearly below the majority point, and only Sweden (and possibly Ireland) shows any growth. The ﬁgures have no historical trend and remain fairly stable, with percentages between 40% and 50%. It is not surprising that the association between this broad conception of the working class and the level of organizational mobilization of the socialist parties is of little signiﬁcance and actually is systematically negative over time, in each historical period, and in each decade.43 The main change over time in the working-class constituency was its internal transformation rather than its increase – that is, the long-term growth of the industrial working class (lower-status laborers and blue-collar workers employed in the sectors of mining, manufacturing, utilities [electricity, gas, and steam works and supply], construction and transport) at the expense primarily of the agricultural working class (lower-status laborers in agriculture, forestry, and ﬁshing) and, to a lesser extent, of the service working class (all other residual categories, ranging from public administration to commerce, banking, and services [restoration, cleaning, domestic services, etc.]). These trends are fairly uniform throughout the set of countries, although at different levels. The only exceptions are Italy and Sweden, where the proportion of the agricultural working class grew remarkably from 1880 to 1910 and from 1880 to 1920, respectively. For all these reasons, only the industrial working-class constituency is 43

Exact ﬁgures are ⫺.412 for the whole 1880–1975 period; ⫺.382, ⫺.210, and ⫺.330 for the three pre–World War I interwar and post–World War II periods; and .157, ⫺.051, ⫺.270, ⫺.416, ⫺.054, ⫺.234, ⫺.582, ⫺.212, and ⫺.317 for the decades from 1890 to 1970, respectively.

148

THE POLITICAL MOBILIZATION OF THE LEFT

Table 3.7. Global working-class constituency (workers in all sectors as % of the active population)

taken as the reference point for the mobilization capacities of socialist movements. With this restricted deﬁnition of the working-class constituency, any perspective on majoritarian size is out of the question. The industrial working class never approached the mythical 50% threshold, and only in the United Kingdom did it come close. Although there was a consistent increase from 1880, industrial workers could never hope by themselves to represent a majority of the active population, and therefore of the population as such. From 1880 to 1960, the industrial working class at the European mean level grew from about one-ﬁfth of the active population to about-one third of it. Moreover, before they reached this level, the process of deindustrialization started to make its effects felt. That is to say, even if the entire social group of the industrial workers had massively supported the socialist movements, it could not have hoped for an electoral takeover. I will therefore relate the cross-time and cross-country variation in the size of the industrial working class to interpret the variance in electoral support for the European class left. These associations are reported in Table 3.8 for the entire set of elections, by individual country, by broad electoral period, and by decade. Over the whole set of elections, the association between the industrial working class and the total left vote is positive (.257), albeit weaker than might have been expected. Within each country,

INDUSTRIALIZATION AND URBANIZATION

149

Table 3.8. Industrial working-class size and left political mobilization

Note: Number of elections given in parentheses.

however, the association is much higher. This is due in part to the smaller number of cases involved in each country, although the associations are fairly homogeneous. In the three Scandinavian countries and Italy, the levels of association are around .900. Earlycomers to industrialization (the United Kingdom, the Netherlands, Switzerland, and Belgium) have lower associations than the others, because early industrializing countries had high levels of industrial workers relatively early, and such levels tended to vary little over time. The overall correlation between the industrial working class and the total left (.257 over 292 elections) is the result not only of high individual country associations but also of quite different associations in the three periods. While in the interwar and post–World War II periods the correlation is positive (.171 and .183), in the pre–World War I period it is strongly negative (–.326). This further conﬁrms that it was not in the early industrialized countries that the left grew stronger between 1880 and 1920, but the opposite. Over time, the initial gap in favor of late-industrializing countries is reduced, but initial differences are never completely overcome. A close examination of the plot of elections in the pre–World War I period reveals that the quite marked negative regression line is largely determined by two sets of elections: the Finnish elections between

150

THE POLITICAL MOBILIZATION OF THE LEFT

1907 and 1917 (eight elections) and the United Kingdom elections between 1904 and 1910 (four elections). These two extreme cases in terms of the size of the industrial working class are also extreme cases of, respectively, very strong and very weak left electoral mobilization. This weighs heavily on the overall negative association for the period; if these elections were excluded, the association would become slightly positive, (.166). In the interwar and post–World War II periods, the trends are modestly positive, but the Irish elections are so outlying as to be almost totally responsible for a positive association of this kind. Without the Irish elections, in both periods the association becomes negative (r ⫽ ⫺.123; r ⫽ ⫺.084). Finally, in the decade correlations, the same different phases emphasized by the period analysis emerge. In the 1900s, 1910s, and 1920s, the left is stronger where the size of the working class is smaller. In the 1900 and 1910 decades, this result may be due to the disproportionate effect of Finland and the United Kingdom, but this is not the case for the 1920s. From the 1930s to the 1960s the association becomes positive: left mobilization tends to be higher where the size of the working class is also larger.44 Figure 3.2 shows the linear regression of each country for all the elections over the whole century. I prefer this graphic presentation to a discussion of regression coefﬁcients, as it clariﬁes more effectively how the high within-country associations turn into a relatively weak global association. The different starting points of industrial development explain the overall effect of the poor correlation between industrial working class and left vote. The countries can be classiﬁed into three clearly distinguishable groups. In the ﬁrst group, the electoral development of the left started when the level of the industrial working class was around 10–15% of the active population, and this development shows very steep regression lines: Austria, Denmark, France, Italy, Norway, and Sweden. Next is the group of the ﬁve early industrialized countries – the United Kingdom, Germany, the Netherlands, Belgium, and Switzerland – which all have a higher starting level of industrial working-class size, a consequent lower internal

44

It may be thought that these associations are overly inﬂuenced by the different number of elections in each country, an argument that is particularly true in the ﬁrst phase, when some countries deﬁnitely have more elections than others. Finland, for instance, has eight elections between 1907 and 1917. In the same period, Austria and Belgium have only two. To control for the number of elections, one can conduct the same analysis on a decade base, entering decade averages for each country decade in both left electoral mobilization and working-class size. The correlation coefﬁcient changes, but not the general picture. In the ﬁrst period, the association remains negative (⫺.119); and in the following two, it is positive (.236 for 1918–1944 and .187 for 1945–1975).

INDUSTRIALIZATION AND URBANIZATION

151

variation, and therefore far less steep regression lines. Finally, the two totally deviant cases of Finland and Ireland should be mentioned. These two countries start with the lowest industrial working-class levels and should therefore resemble more closely the intermediate group of relative latecomers rather than the advanced industrial societies. However, they follow either of the two group patterns: Their regression lines are ﬂatter not because they have little over-time variation in industrialization (quite the contrary) but because they have little over-time variation in the total left vote. In relation to their respective industrial working-class levels, Finland has always had a higher left vote; Ireland, by contrast, has had a disproportionately low left vote. These data justify a provisional hypothesis: the political left started to organize electorally in European countries at roughly the same time (see the relatively minor difference in early organizational consolidation in Chapter 6) as a result of political imitation and diffusion. When such organization took place in the context of already very large industrial working classes, the political mobilization effect was minor. By contrast, when the left organized politically where there were low levels of industrial working-class development, it was able to fully exploit the subsequent rapid growth of the industrial constituency. In sum, the question seems to be: Was a speciﬁc political organization for mobilization already in place when the industrial working class underwent its major growth? In conclusion, the developmental association between the size of the industrial working-class constituency and the levels of left electoral mobilization is strong within each country; it is clear – although blurred by different starting points – over the whole set of elections; and it is weak or negative at the cross-country level in the ﬁrst decades and also in successive periods. Finland and the United Kingdom counterbalance each other as deviant cases in the ﬁrst period; Ireland’s deviance in the second and third periods contributes to the slightly positive association. Without these extreme cases, the association would be more homogeneous in the three phases and would be very weak. WORKING-CLASS HOMOGENEITY Forms and levels of political mobilization can be linked causally to the more or less homogeneous nature of each given working class. That the homogeneity of the working class might be as important as, if not more important than, its sheer size to determine favorable conditions for socialist electoral recruitment is suggested by a number of considerations. On the one hand, the social homogeneity of the environment is essential for

152

THE POLITICAL MOBILIZATION OF THE LEFT

Figure 3.2. Industrial working-class and left vote: regression line by country and period. building an ‘‘equality area’’ within which political consciousness, mobilization and participation are fostered. On the other hand, a vast literature underlines the job instability of the agrarian proletariat, its inability to exploit certain forms of collective action that are too disruptive and ineffective in the agrarian sector, and its extreme organizational and membership volatility depending on the economic cycle.45 These points suggest the greater potential for the organized political mobilization of a sectorhomogeneous and large-workshop-concentrated working class. Workingclass homogeneity can be evaluated through its internal sector composition, that is, whether the working class, independent of its size, is composed of workers concentrated in core industrial-production sectors. A nonhomogeneous or less homogeneous working class, by contrast, is one that is spread more widely across many production sectors, most of which are not core 45

Malefakis (1974).

INDUSTRIALIZATION AND URBANIZATION

153

industry (agricultural or service workers or workers in non-core industry sectors). This dimension can be approximated in operational terms by considering the proportion of workers employed in industry (mining, manufacturing, building, transport, utilities) with respect to those in agriculture and other residual sectors. Table 3.9 shows the correlation coefﬁcients between the three main internal components of the global working class and the left electoral mobilization level according to the usual scheme: over the whole century, by periods, by decade, and ﬁnally by country. The data show in a much clearer form a pattern that has already emerged. Over the whole set of elections, the correlation ﬁgures are, as expected, very positive for the industrial component (.455) and clearly negative for both the agricultural and residual other components (⫺.386 and ⫺.231, respectively). These associations are much stronger than those concerning the size of industrial and agricultural workers’ constituencies. When the same data are plotted by country, indicating the regression line of each country as well as of each period, it becomes evident that a small number of outliers sharply reduces the overall temporal association. These outliers are the early Finnish elections (1907–1917) (mentioned earlier) and the early Swiss elections (1899– 1914). Excluding these thirteen elections, the overall correlation over time rises to .622. Moreover, the correlation for each country is higher, ranging from .600 to .900, with the single exception of Switzerland (.203), where the proportion of industrial workers in the whole working class remains stable in the twentieth century, having already reached a high level at the turn of the century.46 Once again, the picture is different when one looks at analytical and non-time-dependent associations, as the global and within-country pattern is not repeated at the period or decade levels. In each period and decade (with the exception of the ﬁrst and the last), the association is systematically negative. This indicates that at no point in time was the left electorally stronger where the sector homogeneity of the working class was 46

At least a part of the small variation in the industrial-working-class proportion of the whole working class in Switzerland is due to the rather ambiguous case of the 1941 census ﬁgures for the industrial working class and for the working class altogether. The proportion of the ﬁrst over the latter is 65% in the previous census and 72% in the following one. In 1941, however, the census ﬁgures give a proportion of 62%, that is, declining with respect to the 1931 census. This is a doubtful ﬁgure. In no other case is such a remarkable drop in working-class constituency found. This is likely to be one of those cases in which some change in accountancy or attribution rules accounts for the change. However, I have accepted the data, avoiding any interpolation. The result is that for almost twenty years the variable remains substantially unchanged, with obvious consequences for its association with the size of the left.

154

THE POLITICAL MOBILIZATION OF THE LEFT

Table 3.9. Internal composition of the working class and left electoral mobilization

Note: Number of elections given in parentheses.

INDUSTRIALIZATION AND URGANIZATION

155

Table 3.10. Mean left electoral mobilization by type of working class

Note: Number of elections given in parentheses.

higher. If anything, quite the contrary: It appeared to be stronger where the latter was weaker. The cases that are missing with respect to the analysis conducted for the size of the working class would have no effect on the general picture. The United Kingdom had an early homogeneous working-class and weak electoral mobilization until the 1930s, while Ireland had a small left and presumably a nonhomogeneous working class. It may be wondered whether the respective effects of industrial working-class size and working-class homogeneity can be distinguished from one other and which effect is more important. To this end, Table 3.10 presents the mean values of the left vote for the type of working class resulting from the cross-tabulation of size and homogeneity. The axes represent the mean value of the distribution of the two variables (28.5% for the industrial working class as a percentage of the active population and 55.2% for the industrial working class as a percentage of the total working class) and identify four groups of cases: those in which the size of the working class is lower than the mean and where the homogeneity is also lower; those in which, by contrast, both size and homogeneity are above average; and the two mixed groups of high size and low homogeneity and low size and high homogeneity, respectively. In Table 3.10, the difference in support for the left between the quadrant of small size and low homogeneity (26.3%) and that of large size and high homogeneity (38.7%) reﬂects the growing temporal trend in both variables. This is not the case for the difference between the quadrant of the small, homogeneous working class (39.4%) versus that of the large heterogeneous one (19.6%). As shown in the discussion of the sector and status transformation of Western European societies, there is no temporal trend moving from a large, heterogeneous working class toward a small,

156

THE POLITICAL MOBILIZATION OF THE LEFT

homogeneous one or vice versa. In this case, the remarkable difference in the mobilization of the left (39.4% versus 19.6%) cannot be attributed to over-time temporal association, but instead reﬂects the weight of homogeneity versus size independent of time. Moreover, it should be noted that in both categories of working-class size, the cases of higher homogeneity show a higher level of electoral mobilization (39.4% versus 26.3%; 38.7% versus 19.6%). Finally, the general effect of size and homogeneity in the entire set is also clearly favorable to the latter: a difference of 7.3 points between small and large size versus a difference of 13.3 points between low and high sector homogeneity. These data have, however, the problem of large frequency differences among boxes. A second way to search for an impact of homogeneity independent of its over-time association with size is to separate the cases located above the regression line between the two variables from those located below the line. If one imagines the regression line between the two variables as representing the common over-time development, then the cases above the line are those that at each given moment (and independent of overall averages) are characterized by a working-class homogeneity that is higher than one would have expected on the basis of its size. Conversely, the cases below the regression line are those in which, again irrespective of time (represented here by the regression line), the working class is less homogeneous than one would have expected given its size. This analysis is carried out in Figure 3.3. Here, 122 cases fall above the regression line and 127 cases fall below it. The difference between the two groups in terms of average left electoral mobilization is considerable: 38.7% for more homogeneous versus 27.7% for less homogeneous. Considering that in this case no weight can be attributed to over-time variation or to the corresponding concentration of certain cases in certain periods, the conclusion is that homogeneity, in the sense of the predominance of the core industrial working class in the composition of the whole working class, has an effect of its own that appears to be more important than that of the size of the industrial working class. The data concerning the proportion of agricultural workers within the total working class do not require extended comments, as they point to a pattern that is specular to that discussed for industrial workers: a strong negative over-time association accompanied by a positive association in each of the three periods and in individual decades between 1880 and 1970. Why, then, do the industrial and agricultural components of the working class evidence symmetric and unexpected patterns of inﬂuence on the left vote? For the industrial working class, the over-time association is positive, although synchronic associations indicate a stronger left vote in

INDUSTRIALIZATION AND URBANIZATION

157

Figure 3.3. Types of working-class structures and left vote. the context of a weaker industrial component; in contrast, for the agricultural working class, the over-time association is negative, but synchronic associations indicate that the left vote was stronger where this agricultural component was higher. This is a clear example of how misleading it can be to rely exclusively on over-time association among variables. The dyarchic conclusion that the growth of the industrial working class is associated with the growth of the left cannot be transformed into the synchronic statement that the left will be stronger where the industrial working class is higher in number or internal proportion. Two processes coexist, the over-time and the synchronic association, each of which points in a different direction. The fact that positive initial associations for agricultural laborers and negative synchronic associations for industrial workers persist in each period of time, while over time the two processes show a different trend, indicates that these associations translate over time without changing. Whatever initial factors were responsible for the left growing stronger where the working class was less industrial and more agricultural, these initial differences have been maintained through time. The key problem becomes that of identifying those factors responsible for the initial pattern.

158

THE POLITICAL MOBILIZATION OF THE LEFT

WORKING-CLASS CONCENTRATION After analyzing sectoral and status transformations, working-class size, and working-class homogeneity, a further approach is to consider working-class workshop concentration. It can be argued that the more shop-concentrated (rather than dispersed in a larger number of small workshops) the industrial working class is, the more it leads to social homogeneity and an equality area, which in turn are favorable to ideological and organizational mobilization. This is because, for instance, in large workshops more impersonal relationships between workers and employers predominate, against the more direct face-to-face relationship in small workshops. The latter tend to increase social control, deference, and paternalistic attitudes among the workers and generally strengthen the perception of a common interest of both owners and workers in the fate of the small workshop. In the absence of direct comparable over-time information on workshop size, an indirect indicator can be devised considering the ratio of employers and the self-employed to those occupied in the core manufacturing industry. Assuming that the number of employers reﬂects the number of workshop units, this ratio indicates the rate of fragmentation to concentration of the working class in manufacturing. Low rates indicate a higher number of workers per unit; high rates indicate the opposite. Table 3.11 describes the distribution of manufacturing working-class concentration in Western European countries from 1880 to 1970. The ﬁgures should be read as indicating the number of employers in manufacturing for every 100 workers occupied in the same sector. As the means of decades indicate, there is a long-term growth trend in the concentration of the manufacturing working class in larger and fewer workshops. At the beginning of the twentieth century there were approximately four workers for one employer/ self-employed person in manufacturing; by the 1930s the ratio had risen to seven to one, and in 1970 it was twenty to one. It is not surprising to ﬁnd that the British manufacturing working class was the most concentrated, followed at a considerable distance by the Belgian one. In the 1880s, the ratio of British employers to workers was lower or similar to that of all other countries in the 1930s. Yet the differences among the countries do not systematically reﬂect the differences between earlycomers and latecomers. Notice, in particular, the high manufacturing concentration that always characterized the small Finnish working class. At the opposite end, the most fragmented working classes are those of Italy, followed by France, Denmark, and Norway. In Italy, the process of over-time concentration is less evident than in the other countries. In the 1970s, its level of manufacturing concentration was three

INDUSTRIALIZATION AND URBANIZATION

159

Table 3.11. Employers/occupied ratio in Western European countries

Note: Figures are multiplied by 100. They indicate the number of employers for every 100 occupied.

times lower than the average (and eight times lower than in the United Kingdom). In Table 3.12 the usual correlation analysis is performed on the whole set by periods and decades. Here, not only is the employers/workers ratio in manufacturing strongly associated, as expected, with the left vote over the whole period – .604 (i.e., the smaller the concentration the lower the electoral support) – but this relationship is conﬁrmed across countries as well as within countries. Even the ﬁrst pre–World War I period shows a strong negative association, while for most other variables discussed so far, it turns out to be quite deviant. In the early mobilization phase, unlike the other aspects discussed so far, the concentration of the working class in large workshops proves advantageous for left mobilization. The same remains true in the interwar period, and the association fades only after World War II. The analysis by decade conﬁrms the consistency of this association and its slow fading away over time. Of all the indicators discussed so far, this is the only one that shows a strong (the strongest) developmental association, as well as a corresponding synchronic association with the left vote,.

RATES OF GROWTH The last aspect to be considered is the rapidity or tempo of the transformations analyzed so far, as opposed to their earliness or quantity. I use the indicators discussed in the previous sections, and in particular the interde-

160

THE POLITICAL MOBILIZATION OF THE LEFT

Table 3.12. Working-class concentration and left electoral mobilization

Note: Number of elections given in parentheses.

cade rate47 of change, in regard to the following four dimensions: (1) population occupied in industry; (2) dependent labor over the active population; (3) the proportion of manual and lower-status workers in key industrial sectors (mining, manufacturing, building utilities, and transports) over the active population; and (4) the proportion of the global working class constituted by the industrial core workers (deﬁned as belonging to the sectors mentioned in dimension 3). The ﬁgures are derived from census data already presented and can be found in the Data Appendix. The data can lead to several observations on comparative industrialization, but I will limit myself to highlighting some distribution features that appear to be important in this context. Industrial latecomers (Ireland, Finland, Sweden, Austria) experienced high rates of change, while industrial earlycomers (Belgium, the United Kingdom, the Netherlands, Switzerland, and Germany) had comparatively lower rates of change. Looking at the means of each decade, there are some differences in the four dimensions. For the sector transformation, the decades showing a high degree of change were 1890–1900, 1940–1950, and 1950–1960. The period between the end of World War II and the 1960s is the phase of highest growth for all indicators in all countries, Belgium and the United Kingdom exluded. In the interwar decades, the growth of the active population in industry and that of working-class homogenization slow down, and 47

Examining changes between elections produces too much electoral volatility and too little social change (decade census data).

INDUSTRIALIZATION AND URBANIZATION

161

the working-class constituency actually declines (in 1920–1930 by 1.0%). This happens again in the 1970s as deindustrialization occurs in all countries except Ireland and Finland (and, to a lesser extent, Austria). In the same period the working-class constituency declines sharply by 2.3%. The French industrial development, despite being one of the most advanced in Europe by 1880 in terms of scientiﬁc tradition and technical innovation, particularly in the banking and ﬁnancial sectors, slowed to a virtual standstill thereafter. Between 1890 and 1940 all indicators point to a sluggish development rate, which is unique in a European comparative perspective.48 The elements that signiﬁcantly limited the potential for development of the industrial proletariat and the French failure to enter the twentieth century as a mature industrial state have been the subject of considerable debate. The causal factor that appears more consistently in comparative analyses as responsible for this stagnation – and for its correlated continuous predominance of handicraft production,49 slow growth of factory industry, weak industrial concentration, and little large-scale expansion – is demographic. France had a slow rate of population growth, and no demographic revolution took place. This implies a persistence of regional specialization, of only partially integrated provincial markers, and of the use of local materials.50 In France, major rates of change in the four dimensions appear only after World War II. Sweden and Finland experienced the fastest rate of industrial transformation in all its dimensions. From 1870 to 1914, no European country exceeded Sweden in terms of economic growth per capita. The crucial role in creating this situation was played by foreign trade: Swedish raw materials, especially timber and iron ore, were in demand by the growing European economies, and a large share of foreign capital helped to ﬁnance industrialization. Large enterprise played a more important role in Sweden than it did in France, for example. Small businesses, by contrast, were less important than in other continental countries.51 Therefore, the industrial 48

49 50

51

The stagnant nature of the French economy between 1880 and 1930 is underlined by French economic historians. See Dupeux (1964: 170–173); Le´vy-Leboyer (1968: 281– 298) provides a British–French comparison of the early stages of industrialization; Sauvy (1967: 536) provides a comparison of French industrial production levels with American, British, German, and world levels; see also the classic French–German comparison of Clapham (1936: 53–56 and 402–409). For the artisan, rather than factory worker, see in particular Sewell (1986: 49). See the titles in note 48 and Milward and Saul (1977: 125–139); Trebilcock (1981: 140–150, 194); Landes (1969: 245–246). See Tilton (1974: 561–571). In 1872, the enterprises employing more than 100 people employed 65% of the Swedish industrial working class; Simonson (1990: 86).

162

THE POLITICAL MOBILIZATION OF THE LEFT

working class in Sweden was highly concentrated in terms of industrial units. At the same time, industries were located not in urban centers, but in rural ones (as in Norway). Swedish industrial development was thus not characterized by the rapid growth of urban industry, attracting labor from rural districts. With the exception of the textile industry, industrial workers tended to live outside the large towns. As a consequence, urbanization was scarcely a signiﬁcant phenomenon (as it was, for example, in Norway).52 Although the studies by Bull, Galenson, and Lufferty have singled out Norway as the country with the most accelerated changes, with Sweden coming second and Denmark third, nothing in my data supports this picture: Norway is always clearly characterized by lower rates of change than Sweden (and also Finland). This may, of course, be the result of different types of data.53 Norwegian industrialization was hampered by a lack of coal until the end of the nineteenth century, but subsequent electriﬁcation rapidly overcame this disadvantage. Electrical-energy industrialization was very rapid in terms of output and national product, but some of its features dispersed rather than concentrated industry and the working population. This may explain for Norway (as well as for Switzerland, which also underwent rapid electriﬁcation) why rapid rates of change in energy use and output may not be fully reﬂected in high rates of occupational structural change and (see later) urbanization.54 The analysis of the relationship between the tempo of industrialization and left electoral development can be organized on the basis of two hypotheses. The ﬁrst suggests that cases and periods of higher rates of socioeconomic change are associated with cases and periods of higher rates of growth in left electoral support. The second hypothesis argues that cases and periods of higher rates of socioeconomic change are associated with higher levels of total support, not with higher levels of change in such support. The two hypotheses have different theoretical meanings. To assume an association between rates of economic change and rates of electoral change postulates a straight and immediate correspondence or translation of economic change into political change. When one relates rates of economic change to levels of left electoral support, the logic of the argument is different. 52 53

54

Simonson (1990: 87). The data used by Lafferty are indeed different from census data and refer to output indicators of economic development. See Lafferty (1971: 37). On pages 48–51, Lafferty discusses the implications of using different kinds of data to compare the three Scandinavian countries. A waterfall sufﬁced to provide cheap energy. Peripheral and rural communities could attract employers, thus favoring high dispersion of industry. On the special features of electriﬁcation-led industrialisation see Trede (1992: 204), who sets the Swiss, Norwegian, and Swedish models aside from the rest of the European experiences.

INDUSTRIALIZATION AND URBANIZATION

163

The question is whether left electoral support was higher in countries and periods of rapid economic transformation vis-a`-vis countries and periods of less rapid or even sluggish economic transformation. It seems to me that the general theses concerning the rhythm of change are better expressed and tested in this second form. To perform this test in Table 3.13, I report the correlation between the mean electoral size of the left and the mean rates of change in industrialization measures in different periods. The associations result from a small number of cases, as each country enters with only one average value, and are remarkably high, particularly in the ﬁrst two periods, 1890–1910 and 1920–1940. Considering the periods 1890–1940 or 1890–1970 still yields systematically positive associations: The working-class constituency effect remains the weakest of all and the working-class homogeneity effect the strongest.55 These data conﬁrm the hypothesis advanced earlier that the coincidence between the organizational development of socialism and the rapid growth of the industrial constituency is a far better predictor of electoral mobilization capacity in the industrial society than is the level of development. Industrialization earlycomers had a much larger industrial constituency, which had, however, already experienced the phase of intense growth. Latecomers had a smaller industrial working-class constituency, but in these cases socialism organized precisely during the phase of major industrial society growth rates.

URBANIZATION AND LEFT MOBILIZATION If one identiﬁes urbanization with concentration in large cities, a threshold of 100,000 inhabitants is adequate. If the intention is to differentiate dispersed settings and microvillages from other settings, a 5,000 or even 2,000 cutoff point might be ideal. However, as in census statistics, the units identiﬁed by the threshold do not necessarily correspond to an ‘‘urban settlement.’’ The threshold of 20,000 is generally retained as a good approximation between the geographical unit of measurement and the urban settlement. This is because the most important ingredients of urban social life are best guaranteed once the unit size exceeds 20,000 inhabitants. Table 3.14 shows the levels of urbanization by country, taking as a reference point the percentage of the population living in cities larger than 20,000 and 100,000 inhabitants. The lines indicate the point at which a 55

Remember, however, that the United Kingdom and Ireland have no data for this indicator. Ireland, in particular, works as a deviant case, reducing the level of association.

164

THE POLITICAL MOBILIZATION OF THE LEFT

Table 3.13. Correlation between mean left electoral size and mean rates of change in industrialization measures by periods

Note: Number of elections given in parentheses.

certain level of urbanization was reached in each country, and they facilitate cross-country comparisons. For the 20,000-inhabitant threshold, the line represents 33% of the population; for the 100,000-inhabitant threshold, the line represents 20% of the population. However, in describing the levels of urbanization of European countries throughout the last century, it is also useful to make some reference to the 2,000- and 5,000-inhabitant thresholds. Taking the 20,000-inhabitant threshold, the United Kingdom and the Netherlands lead in urbanization, followed by Germany, Belgium, and Italy (the most highly urbanized country following the 2,000 or 5,000 criteria), which show average cross-European levels. Northern countries, including Ireland, remain the least urbanized part of the Continent together with Switzerland (which is one of the most urbanized countries of Western Europe by the 2,000-inhabitant threshold). The 100,000inhabitant threshold does not substantially modify this picture except for the cases of Austria and Ireland, which both fall into the group of early urbanized countries, although they still are behind the overurbanized cases of the United Kingdom and the Netherlands. After World War I, Austria, Ireland, and, to a certain extent, Denmark are ‘‘large-capital’’ countries, whose urbanization is due to the disproportionate impact of the capital urban center on the whole population. Comparing the different thresholds yields the following characterization: The United Kingdom, the Netherlands, and Germany (to a lesser extent) are homogeneously both early and highly urbanized countries. At the other extreme, the Scandinavian countries are the least urbanized in Europe. France should also be included among the countries with relatively weak and late urbanization irrespective of the threshold chosen. Ireland and Austria would also be in this group,

INDUSTRIALIZATION AND URBANIZATION

165

Table 3.14. Urbanization rates in Western Europe

but for the weight of their large capitals. Switzerland is one of the least urbanized countries in terms of medium or large cities, but it had a relatively early level of small-city urbanization. Similarly, Italy is the most and earliest urbanized country in terms of small cities (2,000 to 5,000) but is average or even below average for medium- and large-city levels of urbanization. Consequently, the correlation among these various measures of urbanization is lower than might be expected.56 The relationship between urbanisation and levels of left electoral mo56

They are the following: U2,000 U5,000 U20,000

U5,000 .9185 (104)

U20,000 .7232 (104) .8942 (104)

U100,000 .4109 (103) .5544 (103) .8749 (111)

166

THE POLITICAL MOBILIZATION OF THE LEFT

bilization is summarized in Table 3.15 following the usual distinction between overall cross-country and cross-time association followed by periods, decades, and country associations. Given the extensive discussion of this type of table in the previous sections, it is no longer necessary to specify the meaning I attribute to the different types of associations. At the overall cross-country and cross-time level, the association between urbanization and left mobilization is positive but weak; it tends to increase systematically with the level of the urban threshold, from .051 for the 2,000-inhabitant threshold to .216 for the 100,000-inhabitant threshold. This suggests that the element of urban settlement that most fosters socialist electoral mobilization efforts is large urbanization. However, even in this case, the picture of developmental association is reversed by crosssectional synchronic analysis even more clearly than was the case for industrialization. Urbanization levels are negatively associated with left electoral development in almost all temporal units, particularly in the 1890–1917 period and in the ﬁrst decades in general. The different indicators all behave roughly the same way: For medium- and large-city urbanization the association disappears altogether from the 1930s on, while for smallcity urbanization it continues to be consistently negative (if very weak) throughout the entire century. In this case, a more detailed look at scattergrams by periods is not particularly revealing, as no spatially deviant cases can be clearly identiﬁed. The conclusion is that before the 1930s low urbanization contexts proved a more favorable setting for left mobilization. To put it differently, such contexts did not prevent or hamper signiﬁcantly such mobilization. This conclusion may seem to contradict the widespread assumption that within each national context socialist electoral mobilization found more favorable conditions in urban settings. It does not, however: There is no contradiction between this and the ﬁnding that at the cross-national level, socialism developed earlier and/or faster in less urbanized countries. A ﬁnal check concerns the rates of growth in urbanization as opposed to its actual levels, as was done previously for the indicators of industrialization. If the rate of change from one census decade to the next is computed relating the four indicators of urbanization to either the level of the left vote or the rate of change of the vote, no signiﬁcant associations emerge.57 When the data are aggregated at the period level (three periods per country), the associations remain weak (around .100), although they become positive. Only when the data are aggregated at the country level does a signiﬁcant correlation emerge. The average per country rate of 57

The correlation coefﬁcient range between ⫺.065 and .080 over a set of eighty-nine to ninety-eight decades.

INDUSTRIALIZATION AND URBANIZATION

167

Table 3.15. Electoral mobilization and urbanization levels

Note: Number of elections given in parentheses.

urbanization growth (using the 20,000-inhabitant threshold) is linked to the average per country left vote with r ⫽ .495 and only one deviating case emerges: Norway, with a total left vote that is far higher than expected on the basis of its average urbanization growth rate (without Norway, the association is r ⫽ .625). In sum, contrary to the case of industrialization measures, there is not enough evidence to conﬁrm that the rates of change in urbanization levels are associated with left electoral mobilization levels.

SOCIAL MOBILIZATION MODEL Urbanization and industrialization measures are not highly associated with each other. In the global set of census decades considered in this study, the association between different measures of urbanization and the two key measures of industrial society development (the percentage of population employed in industry and the percentage of industrial working class over the active population) ranges between 0.5 and 0.7. Moreover, these associations are stronger in the pre–World War I period and tend to be reduced in the two following periods. This is not surprising considering that in describing patterns of industrialization and urbanization, several differences have emerged. Some of the early and most industrialized countries, like the United Kingdom, are also the most urbanized. Some of the less urbanized countries, like the Scandinavian countries, Finland, and Ireland, are also the latest and least industrially developed. There are, however,

168

THE POLITICAL MOBILIZATION OF THE LEFT

some important deviations. Switzerland is an early and strongly industrialized country and is also one of the least urbanized societies of Western Europe; the weak urbanization explains to a certain extent the Swiss peculiarity of early and thorough industrialization occurring without the creation of a proletariat.58 France is also comparatively more industrialized than urbanized. The exceptional urbanization of the Netherlands does not correspond to similar levels of industrial development. It might be useful to combine these two processes together into a single social mobilization model to see whether their combined effect is a better predictor of left electoral mobilization than its separate elements. Given that (1) the process of growth seems more important than the level of each variable; (2) urbanization and industrialization magnify each other rather than compensate for each other, and (3) the relationship between the two is very different in different countries, the combination of levels of industrialization and urbanization into a common index does not increase the association levels with the overall electoral strength of the left.59 Rather, we need to consider to what extent the two processes overlap and 58

59

For this aspect of the Swiss industrialization see, Biucchi (1976–1977). In 1860, twothirds of Swiss workers lived in their own houses, and the vast majority were part-time farmers, whose land was often made available by the commune. Cf. Gruner (1968: 94). The product between the two factors is better than the mean (or the sum) if one assumes that the impact of medium levels of industrialization and urbanization have to be regarded as more signiﬁcant for a social mobilization syndrome than the combination of high industrialization and low urbanization or the reverse. The major impact of parallel growth processes has been argued, and this would speak for a multiplicative index. However, when dealing with levels rather than rates of change, this argument is less applicable. Actually, the test shows that the way the index is computed does not make much of a difference in terms of association with the left vote. The following table shows both combined indexes of industrialization/urbanization: the additive standardized one and the multiplicative one, taking the natural logarithm of such multiplication. Their association with the left vote is compared with the association of the single processes of urbanization and industrialization.

Social mobilisation and electoral development social mobilization N 1880–1975

278

1880–1917

industrialization

urbanization

ln(I*U)

(ZsI + ZsU)

.234

.189

.218

.233

82

⫺.326

⫺.376

⫺.447

⫺.375

1918–1944

91

.174

⫺.043

.061

.075

1945–1965

76

.179

⫺.035

.050

.076

1965–1975

29

.031

.152

.231

.172

Note: Based on the same number of cases for all variables.

INDUSTRIALIZATION AND URBANIZATION

169

reinforce each other or, alternatively, are historically separated. To this end, in Figure 3.4 the countries are represented by their decade values in both industrialization (percentage of industrial working class over the active population) and urbanization (percentage of population living in cities with more than 20,000 inhabitants). They are regrouped into four groups to facilitate the reading of the graphs, but also because different patterns are roughly indicated by each group slope. The relationship between the quantitative scope of the processes of industrialization and urbanization in each country is indicated by the ratio between the increase on the x axis (industrialization) and the y axis (urbanization). In the top right-hand part of the ﬁgure, the United Kingdom, the Netherlands, Switzerland, and Germany show similar ranges in industrialization as earlycomers and highly industrialized countries. At the same time, they have different urbanization starting points, ranging from the highest values for the United Kingdom and the Netherlands to the lowest for Switzerland. They have in common that their twentieth-century transformation is more pronounced in the urbanization level than in the industrialization level. In the second graph (bottom right), Finland, Sweden, and Italy are industrial latecomers that undergo a process of accelerated transformation in both dimensions, but industrialization exceeds urbanization.60 The transformation pattern of Denmark, Austria, and France is one of relatively balanced but modest growth in both processes. The starting points are neither very low nor very high; the ﬁnal points have medium values; and, as a result, the paths of transformation are slow and modest in unitary terms, showing no major steps or jumps. The graph in the lower right-hand part of Figure 3.4 groups the three remaining countries: Norway, Belgium, and Ireland. The modest correspondence in the trends of the two processes is determined here by industrialization developments that are far wider in scope than the urbanization changes. Norway is the clearest case of industrialization without urbanization, but the other two countries are also relatively and comparatively ﬂat as far as urbanization growth is concerned. If one considers (1) that the combination and cumulation of both processes of change can be regarded as particularly propitious to early and strong socialist mobilization; (2) that as far as industrialization indicators are concerned, the rates of change were more important than the actual 60

Italy represents a more complicated pattern: After the ﬁrst three decades (1890–1920) of modest growth in urbanization without important growth in industrial population, the pattern becomes similar to that of Finland and Sweden, with accelerated growth in both dimensions but also with a predominance of industrialization over urbanization.

170

THE POLITICAL MOBILIZATION OF THE LEFT

Figure 3.4. Patterns of industrialization and urbanization. levels; and (3) that for urbanization the same was not true, and the levels seemed more important than rates of change, it is possible to draw a number of conclusions about the development of the left in different cases. It is likely that conditions for left political mobilization are best in cases of parallel and combined signiﬁcant rates of change in both industrialization and urbanization, that is, in the countries of the graphs in the top righthand and lower left-hand part of the ﬁgure. By contrast, I would expect lower levels of left mobilization in those countries characterized by the predominance of one aspect over the other: urbanization over industrialization or industrialization over urbanization. The extent to which these hypotheses are supported by left vote levels is reported in Table 3.16, where, for each group of countries characterized by a speciﬁc historical pattern of social mobilization, the corresponding mean left vote is reported. In fact, the countries characterized in the 1890–1960 period by a

INDUSTRIALIZATION AND URBANIZATION

171

Table 3.16. Patterns of historical social mobilization and left vote

coincidence and a parallel development of industrialization and urbanization show remarkably higher levels of left support than those in which one of the two processes prevails over the other. The idea that modern mass parties, and in particular class parties, are not necessarily the product of higher levels of social mobilization (industrialization and urbanization) – as Duverger wrote in the 1950s61 – but rather tend to be stronger when both processes are coincident and achieve accelerated parallel growth62 is supported by these data. Those countries characterized by the decoupling of these processes show much lower left support. Differences with respect to the left vote mean for the entire set of European elections (32.6) are not marginal. At this stage, one can test the predictive capacity of the social mobilization processes by devising a combined index including all variables that have proved signiﬁcant in the analysis carried out so far. For the constituency-size effect, I consider the weight of the industrial workers on the active population (positive long-term effect) and the weight of the independents without family workers (negative long-term effect). For the working-class homogeneity effect, I examine the proportion of industrial workers over the entire working class.63 For working-class concentration, I consider the employ61 62 63

Duverger (1967). Pizzorno (1966b) and (1969). To prevent the indicators of homogeneity from producing the loss of the British and Irish cases and of the French elections after 1945, I have standardized the variables and assigned the value of 0 (i.e., the overall mean) to these cases. This means that in summing up all the indicators, for these cases the missing one is absent and only the remaining three are counted.

172

THE POLITICAL MOBILIZATION OF THE LEFT

ers/occupied ratio. Finally, to consider the variable of the combined coincident rate of growth of industrialization and urbanization, I rank the groups of countries identiﬁed in Table 3.16 as those where the two processes do not combine (small growth in one and high or medium growth in the other [value ⫽ 1]); those where the processes combine with moderate rates of growth (value ⫽ 2); and those with high rates of growth for both industrialization and urbanization (value ⫽ 3). The formula of the combined social mobilization index is therefore the following sum of standardized values: Process constituency size

Variables

percentage of industrial workers in active population percentage of independents in active population constituency homogeneity percentage of industrial workers in total workers constituency concentration employers/occupied ratio rate of transformation industrialization/urbanization classiﬁcation

Sign

Index

⫹ ⫺ ⫹

social mobilization

⫹ ⫹

Figure 3.5 reports the scattergram of this combined social mobilization index and left vote for the whole set of elections from 1880 to 1975. The only group of clearly outlying elections is now represented by the early British elections (the ﬁrst four elections, 1900–1913). This social mobilization model results in an overall developmental association of .644 over 292 elections, corresponding to an explained variance of 41.4%. The national regression lines in the ﬁgure are now much more homogeneous in their slope, with the cited exceptions of Finland, the United Kingdom, and Ireland. The cases of Ireland and Finland are the main novelty of the social mobilization index compared with the various independent indicators of industrialization and urbanization discussed before: Ireland and the Finnish elections no longer appear deviant when set in the overall distribution of elections of Figure 3.5, although it remains true that their overtime variance in left electoral performance is not very closely associated with the over-time growth of social mobilization. In addition, their regression lines are ﬂat compared with those of all the other countries. Nevertheless, their range of values is consistent with the general developmental model. Combining the various indicators of industrialization and urbanization does not entirely dissipate the originality of the Irish and Finnish development, but it does reduce it considerably.

INDUSTRIALIZATION AND URBANIZATION

173

Figure 3.5. Social mobilization and the left vote. A cumulative index allows for charting of elections and visual control of the structure of the data, but it does not correctly measure the multivariate association. Running a regression between the left vote and the ﬁve previously mentioned variables yields an r of .755 and an Rsq of .570. The variables have the following beta coefﬁcient: industrial workers as a percentage of the active population (working-class constituency) independents as a percentage of the active population (‘‘bourgeois’’ constituency) industrial workers as a percentage of all workers (working-class homogeneity) employers/occupied ratio in manufacturing (workingclass concentration) urbanization/industrialization coincidence and rates of growth

.004 ⫺.070 .259 ⫺.429 .490

174

THE POLITICAL MOBILIZATION OF THE LEFT

In these data we ﬁnd an additional conﬁrmation that the weight of the constituency-size variables is less with respect to the other variables (actually, in a stepwise regression, the two constituency variables do not enter the equation). Let me ﬁnally underline the key difference between this analysis and other similar attempts. The test of the model is not performed within each country. The latter strategy normally results in much higher associations (it yields an Rsq higher than .800 and in three cases higher than .900), because the number of cases is lower (sometimes simply too low) and the historical multicolinearity of social and political processes is maximum. Comparing within-country model performance allows an evaluation whereby the model works better, but it does not allow any identiﬁcation of deviance from a general European model. Within-country data can be ﬁtted very well by regression lines even if they are totally deviant with respect to the overall model association, or the reverse. In addition, withincountry regression cannot incorporate comparative ranking variables, which are obviously constant for each individual country. The purpose of my attempt is to identity a European model and then apply it to individual countries to identify their deviation from such a model. For this reason, data are standardized across all the European elections. The consequent variance refers to a ﬁt between social mobilization processes and the left electoral vote, which is established as if country boundaries did not exist. It is thus a truly comparative cross-country variance, according to the logic speciﬁed in the introductory methodological notes of Chapter 1.

CONCLUSION: THE LIMITS OF THE SOCIAL MOBILIZATION MODEL The analysis carried out in this chapter is lengthy and detailed for two reasons. The ﬁrst is that the relationship between social mobilization processes and left development is often taken implicitly as a matter of fact, and it therefore required thorough investigation. The second is that this relationship takes a variety of forms and versions according to different conceptions of the independent variable. I have tried to explore as many of these varieties as possible. The discussion has shown that all factors related to industrialization and urbanization are characterized by well-deﬁned temporal trends that are similar across the countries. The same is true of the electoral development of the left, as discussed in Chapter 2. Under these conditions, any analysis of the association among these historical processes is subject to the

INDUSTRIALIZATION AND URBANIZATION

175

risk of interpreting a merely historical multicolinearity pattern as a causal link. I have tried to clarify the difference between the two in three main ways. The ﬁrst way was to apply sequential logic, by reformulating the association between social transformations and left development in terms of relative timing – of earliness versus lateness. The second was through a reformulation of developmental associations to synchronic ones tested at different points in time. The third was a comparison of country levels and trends in socio-occupational transformations with European average trends and levels, with the goal of creating typologies of different paths to be then related to the global left size. Some of the results are very clear. None of the hypotheses concerning the comparative timing of social mobilization can be sustained. The earliness of all possible dimensions of social mobilization is not related either to the earliness of left electoral mobilization or to its average strength over longer periods of time. By contrast, the tempo of social mobilization, that is, its rate of growth and its rapidity, have proved to be more signiﬁcantly related to the patterns of left growth and to left levels. The interplay between the role of timing and tempo explains the more ambiguous results of the data concerning the social constituencies and their homogeneity and concentration. For these variables, a systematic discrepancy between overtime developmental associations and cross-country synchronic ones was reported. The former generally went in the expected direction: The left levels of electoral mobilization were positively associated with all indicators of industrial society, industrial working-class constituency and homogeneity, and urbanization level and were negatively associated with agricultural occupation, independent status groups, agricultural workers, and so on. Some of these variables proved to be more strongly associated than others (the internal homogeneity and concentration of the industrial working class more, for example, than its size). However, the variables pertaining to urbanization, social constituency size, and homogeneity do not reproduce at the synchronic level their expected over-time association with electoral mobilization. The left tends to be weaker where industrialization levels, industrial working-class size and homogeneity, and urbanization levels are higher. Variables that are associated negatively with left mobilization at the over-time developmental level are associated positively at each point in time, whether a period or a decade. Thus, the agricultural component of the active population and the agricultural component of the working class, both negatively associated over time, tend to be positively associated by period and decade. The structure of the data that produce this result is shown in Figure 3.6. Positive regression lines exist for each country, and they result in an

176

THE POLITICAL MOBILIZATION OF THE LEFT

Figure 3.6. Structure of the social mobilization model data (for data that are associated positively over time with the left vote). overall association that is, however, weakened at this level by the fact that each country enters the period under study with widely diverging starting points for social mobilization levels. However, this ‘‘sheaf’’ of parallel overtime associations for each country is downgraded in such a way as to produce, for each time slice, a negative association. These early differences tend to remain the same over time; they are not reduced. It is as if the early imprint was maintained later within the context of generally growing left movements. A contrasting but symmetrical argument holds for those factors that are negatively associated across time but positively associated across space with the electoral development of the left. To a certain extent, the discrepancy between over-time and synchronic associations can be explained by the identiﬁcation of a few country elections that clearly deviate. At different stages, Ireland, Finland, and the United Kingdom deviate drastically from the implicit hypotheses of the social mobilization model, although in different ways. Ireland is one of the least and latest developed industrial economies and has the weakest left mobilization in Europe, which is perfectly consistent with the hypotheses of the model. However, the undermobilization of the Irish left goes well beyond its socioeconomic comparative backwardness. Finland belongs to the same genus in socioeconomic terms but has levels of socialist mobilization that are the highest in Europe in the 1910s and 1920s. The United

INDUSTRIALIZATION AND URBANIZATION

177

Kingdom is the prototype of the early mature society, but its left lags behind that of almost all the other countries before the 1930s. These cases will therefore need special attention to see whether other factors may help to account for their abnormal position. However, their deviance alone is not enough to reconcile the developmental and synchronic associations. The data structure in Figure 3.6 also helps to show that the different cross-time, cross-country, and within-country associations between social mobilization and the left vote are not contradictory. Within-country over-time associations are generally very strong, and yet the within-country studies of synchronic associations have produced quite divergent results. While the historical literature of individual countries generally shows that earlier and more intense socialist mobilization took place in the most industrialized and urbanized areas, ecological studies do not systematically conﬁrm this thesis.64 The coexistence of these different generalizations is possible: (1) at the overall level, left growth is associated with social mobilization growth, while at the synchronic cross-country level, the left is stronger in less industrialized countries; (2) within each country, left growth and social mobilization are associated, while the ecological association is absent or weak. Therefore, the degree of satisfaction that can be attained with a social mobilization model depends on what one is attempting to explain. If the aim is to explain the general phenomenon of the growth of the left in Europe and within each country, the model performs satisfactorily. Patterns in occupational transformations typically associated with industrial society growth are unable to account for cross-country differences in left voting. The key question then becomes why these early differences in left electoral development existed at time 1 and why they remained stable in the following periods. Additional aspects of a different nature need to be included in the analysis to explain these early differences. General noncountry-speciﬁc factors need to be found that explain early synchronic differences that later remained as permanent historical patterns. That social mobilization models explain so few of the differences among the countries at each historical moment is astonishing.65 If we 64

65

For instance, Korpi’s work about Sweden (1983: 120) shows that support for the left was not only stronger among industrial workers than among agricultural workers, but also that it was stronger in industrial and urban areas than in rural ones. Similarly, Urwin and Aarebrot (1981: 241–273) show that within Germany the left vote correlates positively with industrial employment and negatively with agricultural employment. In France and Italy, however, this was not the case, as shown by Dogan (1967: 135–136) and by Galli (1968: 240–242). A counterargument, brought to my attention by an early reviewer of this chapter, calls

178

THE POLITICAL MOBILIZATION OF THE LEFT

consider the number of historical national accounts that explain late or early, stronger or weaker left electoral development by levels of industrialization and urbanization, or more generally by the more or less ‘‘modern’’ or ‘‘advanced’’ nature of the social context of one country with respect to others, we must conclude that too much weight is given to these processes. Industrialization and urbanization unquestionably constitute the global context of the formation of working-class constituencies. If, however, they fail to explain differences between countries, phases, or periods in the levels of political mobilization of the class left, it is exactly what intervenes between working-class formation and political mobilization that needs to be investigated. Marx’s ‘‘de te fabula narratur’’ in reference to the British social transformation in the eighteenth and nineteenth centuries66 was an adequate description of the future developments of latecomers as far as the trends in occupational status and working-class constituencies were concerned. Their rapid assimilation into the occupational structure of early industrial nations and the global progressive homogenization of such structures was indeed the main structural trend in the century following its forecast.67 However, the

66

67

into question the core hypothesis of the mobilization approach: that the growth of the left should be linked to the social transformation brought about by industrialization and urbanization. According to this argument, it would be wrong in principle to assume an association between left political mobilization and, for instance, the size of the industrial working-class constituency. Actually, left mobilization could have proceeded by progressively acquiring growing proportions of a given industrial working-class constituency without being linked to its size. It would therefore be unnecessary to expect a close association between the growth of the constituency and the growth of its mobilized section. The argument is that a close correspondence between the working class and the growth of the left does not need necessarily to manifest itself in associations between levels of the two processes. On close inspection, the argument seems to me self-defeating; it ends up denying what it seeks to afﬁrm. If socialist electoral mobilization occurred by capturing growing sectors of the national working-class constituency in each country, and if one assumes that such a capturing capacity was independent of the size of the constituency, one must also conclude that this capacity depended on other, non-classconstituency-related factors. The argument implicitly states that industrialization and urbanization, and the class positions they created, were not the essential factors in explaining the increase in class consciousness and political action. This is, in essence, what I conclude, starting from the different premise that, ceteris paribus, the mobilization capacity of the socialists should depend on favorable social constituency conditions. ‘‘. . . [T]he industrially more developed country presents to the less developed country a picture of the latter’s future’’; Marx (1973: 16). Gerschenkron (1962) has criticized this conclusion in several essays, contending that it is only half-true and that it conceals differences between earlycomers and latecomers. According to his famous analysis, the latter will probably not follow the path of advanced countries, but rather will skip certain stages and substitute for missing prerequisites. His criticism, however, was mainly directed against the uniformity in the stages of development postulated by Rostow (1960): ‘‘traditional,’’ ‘‘preconditions,’’ takeoff,

INDUSTRIALIZATION AND URBANIZATION

179

earlycomers’ fabula was not an adequate description of political developments for latecomers – quite the contrary. While the United Kingdom was the model for industrial transformation, it was not the model for political mobilization. Instead, latecomers proved more favorable contexts for left electoral development well before they even approached earlycomers’ levels of social mobilization. The accelerated rhythm of their industrial development may well be a key to this divergent pattern. When the major push of social and occupational dislocation in the labor force determined by industrialization and urbanization took place in parallel with the organizational development of socialist parties and the granting of the vote, socialist parties were able to capitalize on such transformations. By contrast, when these major social and occupational dislocation effects took place before the socialist organization and before suffrage extension, the mobilization capacity of the socialist parties seems to have been less. The rapidity of the economic transformation is causally associated with the level of left electoral development. This means that socialists were best able to capitalize on political support where and when the bulk of the industrial and urban transformation was taking place in parallel with the extension of the suffrage and the mobilization efforts it triggered. In this sense, the dynamism of social change was far more important than the level it reached. By contrast, in the case of the earlycomers, the organizational development of socialism and the process of suffrage extension occurred after the bulk of industrial transformation had already taken place. Here, socialism organized politically after the fabula had already been told. This possible line of interpretation is checked in Chapter 5. Before reaching this stage, however, I discuss the potential impact on a second general model relating to socialist mobilization capacity: the cultural homogeneity of the class environment in which this mobilization had to take place. ‘‘drive to maturity,’’ and ‘‘high mass consumption’’ and in a world perspective, where the latecomers are Third World countries. In a within-Europe perspective, the leveling off in production outputs per capita and in socio-occupational structure is undeniable.

4

CULTURAL HETEROGENEITY

DIMENSIONS OF CULTURAL HETEROGENEITY ation formation is the second macroprocess that sets the context for the development of working-class movements. In particular, the point of interest here is national cultural standardization. Although the nineteenth-century socialist movements were decisively characterized by their international character, that is, they were ‘‘anational’’ when not deliberately ‘‘antinational,’’ their successive history showed plenty of evidence of their actually being national movements that were part of and an expression of the formation of a national culture and identity. Where the formation of a relatively homogeneous cultural national context was lacking or weak, working-class movements experienced profound problems of organizational consolidation and spread of appeal. The heterogeneity of the class cultural environment will be regarded in this chapter as a crucial element for the successful establishment and consolidation of a workingclass movement and of a class left electoral mobilization. A class and social group analysis of the political mobilization process rests on a model that links the formation of social position, the development of group solidarity, and uniformity in group political action. Socialist thinkers of the second half of the nineteenth century clearly regarded any social (and political) identities that were not rooted in the social position related to the productive process as supra structural factors amenable to false consciousness. They therefore assumed that the role of such identities was deemed to disappear with time. This position is echoed in recent

N

180

CULTURAL HETEROGENEITY

181

works that implicitly or explicitly assume that the pro-left party and organizational behavior of workers is normal and that only deviation from this pattern needs to be accounted for. In this sense, other political identities are the result of the failure of the class identity. This interpretive picture inverts the order of the factors. Preindustrial cleavages created solidarity feelings and identities along several cultural as well as functional lines of conﬂict well before the class cleavage could acquire sufﬁcient momentum, thanks to the processes of industrialization and concentration in urban environments. Thus, the class cleavage was the outsider striving to establish its domain within a space that was already occupied by strong and consolidated social identities. The distribution and relationships between such preindustrial group identities limited the space for maneuver and the mobilization potential of the class appeal rather than doing the opposite. Obviously, when the social dislocation associated with accelerated processes of social and geographical mobility related to urbanization and industrialization occurred, preindustrial social identities were broken down in certain social milieus, but in no way were they simply erased. The class cleavage did not, therefore, start to recruit and mobilize its potential constituency in a vacuum of solidarity and identity. Therefore, the links postulated by the social-mobilization model between workingclass formation and working-class mobilization (via collective identity) should be regarded as more or less valid according to the conditions of contextual cultural homogeneity. ‘‘Class in a Culturally Homogeneous Environment’’1 was not the same thing as class in a culturally heterogeneous context. In this chapter, I analyze those characteristics that pertain to the cultural homogeneity of any given society. Cultural homogeneity is in many ways an essential ingredient for the spread of a nationwide appeal to economic and functional cross-local conﬂict. My aim is to see to what extent the consideration of these essential preindustrial bases for class mobilization improves the working of the social mobilization model, specifying the best conditions of its application and explaining deviance from it. Cultural heterogeneity (or homogeneity) can be conceptualized along horizontal and vertical dimensions. In the horizontal dimension, we can include those elements of cultural nonhomogeneity that separate segments of the society characterized by clear religious, ethnic, linguistic, or other types of cultural/identity standards and differences. Although to different degrees, each segment is composed of both masses and elites groups. 1

This is the speciﬁcation made by Schumpeter (1951) in the title of his essay about social classes: Social Classes in an Ethnically Homogeneous Environment.

182

THE POLITICAL MOBILIZATION OF THE LEFT

On the vertical dimension, cultural homogeneity refers to the possibility for community development in Toennies’s original meaning of the term. The development within a given territory, national in this case, of the possibility for cross-area communication and organization requires the development of links between masses and elites that can be available only under certain conditions made possible by the spread of education, linguistic standardization, and mass media communication development. The spread of ideas, ideologies, organizational forms, and collective crosscultural identities requires, even in this case, that we identify thresholds above which conditions are equalized and below which basic preconditions are lacking. The formation of the nation, of a national community able to exchange not only goods but also symbols with a shared meaning, seems to be just such a necessary precondition for the development of a class type of cleavage. The spread of both formal education and mediated communication should be considered here (Figure 4.1). The process of homogenization of religious practices obviously goes back to the dynastic confrontations and religious wars of the sixteenth and early seventeenth centuries. Language as a symbol of national identity became more important, if not dominant, during the period of nationalism following the Napoleonic wars and the expectations they left behind. Dominant versus subject linguistic standards have always existed, but they became dominant elements of identity and of political mobilization only in this relatively recent phase. Finally, the leveling of cultural stratiﬁcation in terms of cognitive capacities is a third type of process that was accelerated in the nineteenth century in most countries but that can also be traced back to the sixteenth century in the Reformation territories. My goal in this context is not to recall these processes and the differentiated cultural map of Europe they produced.2 It is rather to consider these processes in terms of their end result – as they were at the moment of the early political mobilization of the class cleavage and in its ﬁrst decades of development. They will be considered as historical givens that are likely to determine the extent to which the social transformations considered in the previous chapters were more or less likely to result in straight patterns of the political mobilization of the lower classes. The extent to which cultural standardization within a national territory was extensively achieved or, rather, considerable levels of cultural segmentation and/or cognitive inequalities were left within this territory, inﬂuenced the chances of the political mobilization of the left. Why should 2

On this point, see Rokkan and Urwin (1983) and Rokkan, Urwin, Aarobrot, Malaba, and Sande (1987).

CULTURAL HETEROGENEITY

183

Figure 4.1. Cultural heterogeneity dimensions. a less homogeneous environment have a negative effect on class mobilization? What are the intermediate steps that link these two macroprocesses? For linguistic and religious segmentation, the hypothesis is simply that of the social distance created by group membership. The stronger the boundaries of preexisting group membership, the more difﬁcult the superimposition of new group loyalties. It is reasonable to expect that in contexts characterized historically by differentiated and often opposing (when not directly conﬂicting) religious or linguistic/ethnic group membership, the social distance (boundary) between these groups was higher than in countries that were largely homogeneous from the same points of view. Therefore, the superimposition of a new cross-religious or cross-linguistic group identity could prove more difﬁcult, even in the presence of a rapid sociostructural transformation of the individuals concerned. With regard to cognitive stratiﬁcation, it seems plausible to argue that both organizational recruitment and electoral support in underprivileged groups was enhanced by the possession or acquisition of basic cognitive instruments. Without them, not only it is likely that a break with traditional social ties and social elites was more difﬁcult to achieve, but also that propaganda and recruitment efforts had to be based almost exclusively on enormous human resource–intensive organizational efforts. It is not necessary to postulate that the greater the cognitive competence, the easier the understanding by the people of their ‘‘true’’ interest – even assuming that the true interest of the lower classes was to vote for the socialists. It is sufﬁcient to assume that cognitive capacities make it easier for a new and outside

184

THE POLITICAL MOBILIZATION OF THE LEFT

political actor to spread its political message without being obliged to resort to direct, face-to-face intense organizational efforts. I therefore consider cultural homogeneity as indicated by three variables: (1) the extent of the religious homogeneity of the population, in the sense of adherence to a single religious standard; (2) the extent of linguistic homogeneity of the population, to be understood as adherence to a single common linguistic standard; and (3) the extent of equalization of basic cognitive capacities, such as those of writing and reading. Together, these factors should offer a general picture of the extent of the homogeneity of the social environment in which working-class parties organized their political and electoral activities. I would like to emphasize that the variable taken into consideration here is not religious or linguistic mobilization, that is, the existence, nature, and pervasiveness of political movements built around linguistic or religious identities. I consider cultural heterogeneity in strictly demographic terms, making no reference to political mobilization. In a later chapter, I come back to the question of the extent and the character of political mobilization along different lines, including linguistic and denominational ones. The relevance of this distinction will become clear when we discuss the extent of cultural heterogeneity of different European countries in the next section. In this chapter, I proceed as in the previous one. First I deal with each individual factor, trying to identify whether it has an impact of its own. I then combine several factors to evaluate the extent of their combined effect, as well as to see if they are independent or autocorrelated. Finally, I discuss a general model of left development as it emerges on the basis of the cultural heterogeneity variable and then relate this model to the social mobilization model discussed previously.

CULTURAL SEGMENTATION The data on the linguistic and religious composition of the populations of Western European countries are derived from the census.3 For both religious and linguistic fragmentation, I have decided to consider only the major religious and ethnolinguistic groups in order to avoid inﬂating fragmentation when many very small cultural minorities are present. This is particularly important because, in order to obtain a single value for linguistic and religious heterogeneity in each decade since the 1880s, I 3

As reported by Flora, Alber, Eichenberg, Kohl, Kraus, Pfenning, and Seebohm (1987: 55–85).

CULTURAL HETEROGENEITY

185

have resorted to the index of fractionalization, which is very sensitive to the number of groups as well as to their respective size.4 The inclusion of many very small minorities would increase the index considerably, while in reality, the point of interest here is the major cultural divisions. In Table 4.1 the score of the religious and linguistic fractionalization by country and decade is reported. In a few cases, census data were missing for a decade, and I have reported the interpolated value. Countries tend to divide into two distinct groups. First, there are the clear-cut cases of cultural heterogeneity of Germany, the Netherlands, and Switzerland for religion and Belgium and Switzerland for language; second, there are all the others. Switzerland is the only case of pronounced cultural heterogeneity on both fronts. A few intermediate-level cases also merit further comment. The Finnish case is in an intermediate position, thanks to the persistent presence of a Swedish-speaking minority. Relatively high levels of linguistic fragmentation are also typical of Ireland and the United Kingdom. In these two cases, a common and strong linguistic standard has been observed with regard to the English language for a long time. The data on linguistic heterogeneity, resulting from questions concerning the language that is regarded as the participant’s native tongue, may reﬂect an ideological orientation toward stressing personal identity more than the existence of clearly differentiated linguistic subcommunities. This is evident when we look at the Irish data, where the linguistic heterogeneity that results from the deﬁnition of Gaelic as the ﬁrst language increases slightly over the period from the 1950s to the 1970s. This is less likely to be the result of a renaissance of Gaelic than of a new desire to identify culturally via such a choice. Another case worth commenting on is the great linguistic homogeneity attributed to Italy and France. Italy is characterized by a large number of minority linguistic groups of small and sometimes minute size. Their inclusion would make for a disproportionately high index of fragmenta4

This index is well known. Its formula is the following: n

F ⫽ 1 ⫺ (∑ Ti2) i ⫽ 1

where Ti is each religious or linguistic group’s decimal share of the total population. See Rae (1971: 55–58) for a presentation of the index and Pedersen (1980: 387–403) for a critical analysis of its performance. Pedersen rightly underlines the inadequacy of the index to capture the dynamic of over-time changes as it responds to changes in both the number and strength of groups without being able to discriminate between them. This weakness of the index does not concern us in this case. Cultural heterogeneity is fairly stable over time, and I therefore do not want to concentrate attention on over-time variation.

186

THE POLITICAL MOBILIZATION OF THE LEFT

Table 4.1. Religious and linguistic heterogeneity in Western European countries

tion. A similar picture can be drawn for the French linguistic border minorities, where the Corsican case clearly throws the values of high linguistic homogeneity into some doubt. In this case, however, not much can be done to improve the data, given the fact that Corsica was not given the status of a language or identiﬁed separately in census data. On the whole, it is likely that France and Italy5 should be regarded as far less homogeneous from the linguistic point of view than they actually are according to the ﬁgures in Table 4.1. Religious heterogeneity ﬁgures raise fewer problems of reliability and comparability. They do, however, present a problem that is common to all 5

For all countries, but with particular reference to France and Italy, I should also mention the key question of how language is deﬁned in contrast to dialect. I am unable, however, to take into consideration an alternative distribution of linguistic groups according to any different deﬁnition of languages.

CULTURAL HETEROGENEITY

187

countries: in historical statistics, no speciﬁc place is left for those who profess to have no creed.6 In this context, however, this is not a problem that needs to be taken into account. As speciﬁed earlier, I am not dealing with the extent or strength of a religious cleavage, which would necessarily require reference to true believers versus nonbelievers. Given that my intention is simply to check whether religious homogeneity versus heterogeneity plays a role in left mobilization, the intensity of the engagement is unimportant, and attention must be centered on the existence of multiple religious choices and groups versus the lack of this pluralism. On the whole, while linguistic heterogeneity – in part as a result of the inadequacy of the data – tends to dichotomize the group of countries between those that are heterogeneous and those that are not, religious heterogeneity is more evenly distributed. It should also be noted, ﬁnally, that none of the dimensions of heterogeneity show any temporal trend: They are features that characterize a polity throughout the whole period under examination. Consequently, they will be unable to explain any within-country variation over time, and they should be most important in discriminating crosscountry variation. Moreover, no problem of historical multicolinearity arises with left electoral development. The ﬁrst simple way of approaching the problem of the inﬂuence of religious and linguistic heterogeneity on left political mobilization is to look at the mean level of left support through levels of cultural heterogeneity. To this end, in Table 4.2 I have divided the variance in religious and linguistic fragmentation into low, medium, and high, and I have reported the corresponding mean levels of electoral support for the left in these groups. In order to avoid too skewed a distribution of cases, given the very different distribution of values for the two variables, I have chosen slightly different thresholds, deﬁning as low religious fragmentation those cases that show levels between 0 and .10, medium for those between .10 and .50, and high for those over .50. For linguistic heterogeneity, the corresponding thresholds are 0–0.5, 0.5–0.40, and greater than 0.40 (see the values in Table 4.1). The ﬁgures in Table 4.2 point to a very clear impact of both dimensions of cultural heterogeneity on the mean electoral strength of the left. On average, in both cases there is a difference of about 7 to 8 percentage points between the electoral strength of the left in religious or linguistic homogeneous contexts and those in highly heterogeneous ones. Such a difference represents one quarter of the overall mean of the left vote over 6

Figures on religious denomination, including ‘‘no creed,’’ are reported in Taylor and Hudson (1972: 271–274). However, these data have no historical depth.

188

THE POLITICAL MOBILIZATION OF THE LEFT

Table 4.2. Mean left electoral support by levels of religious and linguistic fragmentation

Note: Number of elections given in parentheses.

the whole set of about 300 elections between roughly 1890 and 1970. In both cases, the medium levels of religious and linguistic heterogeneity produce levels of left electoral strength that fall into the middle of cases of high or low heterogeneity, pointing to the existence of a rather linear effect. The level of cultural heterogeneity of the social environment unquestionably reduces the potential for electoral mobilization of the left. The next step is to consider the possible interaction effect between these two variables. In reality, both variables could tap the same phenomenon, to the extent that religious and ethno/linguistic heterogeneity coincide. Alternatively, they could be independent factors that cancel each other. Finally, they could both be independent of each other and have a reinforcing effect. The interaction effect can be appreciated by simply dividing our set of elections into four types: (1) those that occur in environments that are both religiously and linguistically heterogeneous and should constitute the most difﬁcult cultural environment for the left; (2) those that occur in both religiously and linguistically homogeneous environments, setting the best conditions for left development; and (3) and (4) those that take place in a split environment, homogeneous with regard to language and religion or the reverse, and that should evidence intermediate levels of left electoral support. In Table 4.3 the mean left vote in these four contexts is reported. The interaction between the two forms of heterogeneity seems to be virtually uniform and cumulative. Within each category of linguistic cultural heterogeneity, the left vote increases with the homogeneity of the religious context. Similarly, for each category of religious heterogeneity,

CULTURAL HETEROGENEITY

189

Table 4.3. Mean left vote by types of cultural fragmentation

Note: Number of elections given in parentheses.

the left vote increases with the homogeneity of the linguistic context. Thirteen percentage points separate the set of elections taking place in an environment that is both religiously and linguistically heterogeneous (basically, this cell corresponds to the only case of cumulative heterogeneity, that of Switzerland) from those taking place in environments that are homogeneous on both dimensions. This evidence stresses the fact that religious and linguistic heterogeneity has an individual independent effect on left mobilization and that these independent effects can be treated as cumulative with respect to the left vote.7 Let us now look at the relationship between cultural heterogeneity and the left vote over time as well as at different points in time. In Table 4.4 Pearson’s correlations between the left vote and the index of ethnolinguistic 7

This analysis by type of social environment has also been performed using the threecategory operationalisation of Table 4.2: low, mid and high. Due to the already discussed skewed distribution of cases, this yields a nine-cells typology where, unfortunately, some cells have too low or even null frequencies. However, the results are the same as those discussed with a four-cell scheme if one aggregates the nine cells into three groups: those in the upper right corner with values of high-high, high-mid, and mid-high religious and linguistic heterogeneity; those in the bottom left corner with low-low, low-mid, and mid-low values; and those on the diagonal corresponding to high-low, low-high, or midmid values for religious and linguistic heterogeneity. The results are the following: homogeneous mixed (diagonal) heterogeneous

36.5% (165) 28.1% (101) 23.2% (26)

190

THE POLITICAL MOBILIZATION OF THE LEFT

and religious heterogeneity are reported for the whole set of elections, as well as for the usual temporally deﬁned subsets. In addition, the same correlations are computed for a combined index that is the result of the simple sum of the two. The ﬁgures in the table add a few other elements. First, all correlations are homogeneously negative throughout the whole period, in each subperiod and for each decade. The only exceptions can be found in the very ﬁrst phase, and particularly in the 1890s, when positive associations are observable. There can be little doubt, therefore, about the negative impact and very little need to search for period effects. The positive associations of the early decades are understandable if one remembers that cultural heterogeneity does not change over time but remains stable. In the early phases, socialist mobilization was of necessity extremely weak, irrespective of the levels of cultural heterogeneity that were already established. From this derives the lack of association or even the possibility of positive associations. The second point to be underlined, and that follows from the ﬁrst one, is that the associations between linguistic and religious heterogeneity indices and the left vote are stronger in each period and decade than generally over time: ⫺.175 and ⫺.187, respectively, for linguistic and religious heterogeneity are not impressive negative associations, and they are, on the whole, lower than those that can be obtained in each phase or decade. This shows that cultural-heterogeneity variables discriminate across countries when the temporal variance of the left is somehow controlled, but they are far less important when this temporal variance exists, that is, in the overall bulk of a century’s elections. The third point worth mentioning is that the negative association of ethnolinguistic fragmentation increases remarkably over time, from the decades around the end of the nineteenth century to those after World War II. In contrast, the negative association of religious heterogeneity remains relatively stable throughout the twentieth century. Cultural heterogeneity is not so negatively associated with left size at the beginning, but as time passes it increasingly becomes a structural barrier to the development of the left. It is thus a constraint that grows in importance after the ﬁrst beginning phase and that sets deﬁnite boundaries on left electoral development. The negative association of linguistic heterogeneity increases more than that of religious heterogeneity and appears, in the ﬁnal analysis, to be far more important. If we look at the associations concerning the left vote and the combined index of religious plus linguistic heterogeneity, it is clear that up to the 1920s it is religion that produces negative associations, while from the 1930s on, it is language that plays the more important

CULTURAL HETEROGENEITY

191

Table 4.4. Religious and linguistic heterogeneity and left vote (Pearson’s correlation)

Note: Number of elections given in parentheses.

role. We may have a hint here of the historical sequence of identity mobilization, with religious identity becoming politically important from the beginning of the century and ethnolinguistic (and regional) identity becoming more so after World War II. However, this refers to a feature that belongs to the political mobilization of persons with different identities, and this I have decided to leave aside in this context, where only basic social homogeneity is evaluated. The combined index of the third column of Table 4.4 conﬁrms the additive nature of the interaction that has already been argued: In most cases, it yields higher negative associations than either of the two. On the whole, these data highlight a pattern this is the opposite of that resulting from the social mobilization analysis. While social mobilization pointed to a long-term positive impact on the left vote over time, but tended to disappear as a cross-country discriminating factor when time was parametrized in different periods and decades, cultural heterogeneity

192

THE POLITICAL MOBILIZATION OF THE LEFT

factors worked the opposite way, adding less to the explanation of the temporal development of the left than to the cross-country differences at each given moment in time. This suggests that the combination of these two models may prove a valid way of reducing unexplained variance on both the temporal and spatial dimensions. From a developmental point of view, while social mobilization is a factor of over-time growth, cultural heterogeneity is a factor of over-time containment.

COGNITIVE STRATIFICATION The third factor to be considered as an indicator of cultural homogeneity/ heterogeneity is the cognitive stratiﬁcation of a country’s population. While cultural segmentation tends to split a society into communities, within which both elites and masses are present, cognitive capacities deﬁne strata endowed with different capacities for communication and tend to separate groups along an elite-versus-masses dimension. Cognitive mobilization is the long-term process by which such capacities are equalized through the spread of individual skills, as well as the spread of new means of communication. It involves several features. In political development schemes and theories, it is associated with the idea of rising expectations and therefore of growing demands on the political system. However, cognitive mobilization is also a resource for the system, a ‘‘capability’’ at its disposal.8 From the point of view of this research, the key aspect is the relationship between participation and the ﬂows of political messages that reach the individual citizen. Generally speaking, the spread of cognitive capacities results in a growing number of individuals who can be reached by a wider variety of sources of political messages, through a larger number of channels of communication, and with potentially longer exposure to them. These changes in sources, channels, and exposition to political messages have progressively transformed the nature of political propaganda. In a context in which the mass media are absent or where their impact is profoundly reduced by the lack of decoding capacities in the public, political propaganda efforts and incentives to activate participation are limited to nonmediated exchanges, that is, direct contact and experience. This requires a form of political organization that is, by necessity, characterized by human capital-intensive techniques. The spread of decoding capacities has marked the beginning of a longterm process of ‘‘privatization’’ of political information and propaganda, 8

Deutsch (1961) and Coleman (1965b).

CULTURAL HETEROGENEITY

193

leading to a progressive reduction of human capital in the face of a growing capital-intensive technique based on mass media messages. This process reached such an advanced stage in the last quarter of the twentieth century that it has profoundly modiﬁed the structural characteristics of political parties. In this context, I am interested in the initial steps of this process. For the early mobilization of the European left, the successive waves of cognitive mobilization of European citizens are probably less important than the ﬁrst ones, identiﬁed with the spread of reading and writing. So, the focus of this section is not the general educational level of a population, but rather the basic preconditions for becoming potential objects of mediated written political messages, as well as the minimum threshold capacity for their interpretation. General education can, and has been considered, as a predominantly conservative institution, transmitting and fostering traditional values and attitudes. At the same time, however, at least minimum levels of education have been regarded as important in fostering socialist propaganda and mobilization capacities, promoting the development of the collective consciousness that is essential to class action and contributing to the creation of individual cultural resources capable of counteracting traditional forces and values in the social environment. It is difﬁcult to ascertain literacy levels understood as the capacity to read and write. This is in part a result of the fact that the means of ascertaining such skills have been complex and nonuniform through time and space. At the level of census statistics, the most reliable data concern school enrollment in different levels of educational institutions. However, they present some major problems. First of all, they indicate levels of presumed literacy among age groups that are still far from the age of potential political participation, and they say very little about the literacy capacities of their families. Moreover, in addition to representing the cognitive skills of the future adult population, enrollment statistics present a second problem – this time technical. Computing the percentage of people of a given educational level implies the choice of a reference age group that is not necessarily homogeneous across countries. Changes in the length of primary education, for instance, make this type of percentage statistic unreliable for cross-country comparisons, even global ones. There have been attempts to synthesize and standardize as much as possible the statistics concerning the enrollment of pupils in primary schools as percentages of the 5- to 14-year age group.9 For the late nineteenth and early twentieth centuries, these statistics may still be able to give a comparative estimate of the extent of primary education in 9

See Flora, Alber, Eichenberg, Kohl, Kraus, Pfenning, and Seebohm (1987: 553–633).

194

THE POLITICAL MOBILIZATION OF THE LEFT

different countries, but as soon as postprimary education in the 11–14 age group develops across European countries, higher levels of education are transformed into declining levels of the index. As many pupils between 11 and 14 go to postprimary school, the percentage of pupils in primary school over the 6–14 age group may decline.10 Moreover, literacy abilities are by no means linked to school attendance, particularly in Protestant countries. Alternative indirect indicators of literacy are provided by special groups statistics, such as those reported by Mitchell and taken from marriage registers (which refer to the ability to sign marriage contacts) or statistics based on military recruitment tests of literacy.11 However, these statistics also refer to special groups rather than to the total population, and, in addition, they are also not systematically available across the thirteen European countries. Therefore, I have decided to rely on the collection of estimates made for literacy levels among global populations by Flora in his work devoted to literacy and urbanization.12 Such estimates are reported in Table 4.5. These estimates offer a number of important points that directly concern the question of potential class mobilization. First, with the exception of the high and persisting levels of illiteracy in Italy, which clearly set it apart from the northern and continental model in this ﬁeld, the bulk of the literacy process was already accomplished in the European countries at

10

11 12

Following is the percentage of pupils enrolled in primary education among the 6–14 age group drawn from the previously mentioned source: Au

Be

De

Fi

Fr

Ge

Ir

It

Ne

No

Sw

Sz

Uk

1880

56

.

.

.

82

75

.

35

.

.

72

.

.

1890

63

28

60

.

83

75

.

36

66

65

75

76

56

1900

67

59

60

.

86

73

.

37

70

67

69

72

75

1910

70

62

68

26

76

70

.

43

71

69

68

70

79

1920

74

70

65

36

68

67

83

50

71

69

60

72

79

1930

77

72

67

55

80

75

89

58

74

72

62

70

81

1940

79

70

65

61

79

76

88

43

68

68

63

70

74

1950

76

68

61

62

72

80

87

59

65

67

59

66

67

1960

73

66

71

59

64

69

88

53

62

69

62

69

61

1970

74

65

67

50

54

68

82

56

61

62

59

65

66

See Mitchell (1975). Flora (1973: 245).

CULTURAL HETEROGENEITY

195

Table 4.5. Illiteracy rate (% unable to write and to read)

* Austria: territory of the Republic ** 1950–1960 GDR and GFR rates are the same ***Ireland to 1910; Eire since 1920

the beginning of the period considered here, that is, in the 1890s. This means that in Europe the long-term development of literacy preceded the processes of urbanization and industrialization.13 In this sense, the basic cognitive mobilization of European citizens can be said to be part of the preceding process of nation building rather than that which followed the Industrial Revolution. This is due to the fact that, in contrast to the processes of urbanization and industrialization, literacy development could be fostered more easily by the nation-state if it was deliberately pursued through cultural agencies and/or by state bureaucracy. The second important element in the distribution of Table 4.5 is that, given its older origins, the process of basic literacy development tended to be completed by the 1920s. In 1930, with the exception of Italy, almost all the other cases present estimates that are fairly similar. By that date, no trend is present in the data and no signiﬁcant cross-country differences can be seen. This means that literacy and illiteracy is a dimension that is 13

The key point in Flora’s article (1973: 227) is that literacy development preceded urbanization in Europe, contrary to what has been postulated by modernization theory sequences. The experience of other parts of the world is reversed: Urbanization predates literacy development.

196

THE POLITICAL MOBILIZATION OF THE LEFT

important, if it is important, only for the ﬁrst part of the period considered here, that is, for the very early mobilization phase up to roughly World War I. The most obvious characteristic of these data is the sharp difference in literacy levels between Protestant and Catholic countries. The ﬁt is not perfect, but the countries that show earlier and higher levels of literacy are those of Scandinavia, followed by Switzerland and Germany and, very close behind them, Finland. At the other extreme – apart from the alreadymentioned Italian case – France, Austria, and Belgium show the highest levels of illiteracy. There are, however, two major exceptions. The levels of illiteracy in England, Wales, and Scotland are far higher than those of the other Protestant countries between 1870 and 1900 and are not much different from those of the Catholic ones. Ireland, although still characterized by relatively high levels of illiteracy in 1870 (25.8%), rapidly recuperates: By 1910, its estimates of the illiterate population are among the lowest in Europe. All in all, however, the difference between predominantly Protestant or religiously mixed countries, on the one hand, and homogeneously Catholic ones, on the other, is quite impressive. The standard interpretation takes the lead from the classic theses of Max Weber about Protestantism as the Christian religion with the strongest national language written tradition and, therefore, with the strongest incentive to literacy development among believers and the national population at large. This is conﬁrmed both by the early introduction in Protestant countries of compulsory national primary education and by the more informal role played by Protestant churches in giving incentives to the spread of reading and writing capacities, a feature that is absent in Catholic countries. Thus, for instance, compulsory-education schemes were introduced in Prussia as early as 1763; in Denmark, as early as 1814 (for children up to seven years of age but for only three days each week; in 1849, this was increased to six days each week); in Norway, between 1848 and 1889. In Sweden, from 1842 to 1848, the parliament introduced an elementary school system lasting for six years for all children. Parishes were responsible for starting and ﬁnancing elementary schools to teach writing, reading, and arithmetic. However, even before this time, the exigencies of teaching Luther’s catechism pushed parishes to enforce educational requirements for marriage and other social occasions. In Finland, institutional thresholds of this type cannot be identiﬁed, because the country was not independent until the twentieth century, but it has been reported that in the ﬁrst 100 years after the Reformation, the Lutheran Church made the ability to read a prerequisite for enjoying certain civil rights, as well as for marriage and communion. Similar forms of incentives are to be found in the history of

CULTURAL HETEROGENEITY

197

Sweden, Norway, Denmark, and even Scotland. Teachers were often recruited from the local community; they were soon to become important local cultural and political ﬁgures, often acting as counterweights to the Lutheran national state church, as well as guides and leaders of the popular movements that spread throughout the Scandinavian countries, particularly Norway and Sweden. By 1870, most young workers in these countries were able to write and read.14 This was a crucial factor for development of the socialist press, as well as for the propaganda published and disseminated by the socialists. To this religious factor, Flora has added two other hypotheses in order to interpret basic differences in literacy rates: the existence of communal self-government as a further incentive to literacy development and the international status and geopolitical position of the country. In the latter point, he stresses the need for countries at the center of the European system of interstate relationships, concerned with self-defense and selfassertion, to develop efﬁcient bureaucratic machines (both military and civil). This helps to explain the differences between southern peripheral Catholic countries – such as Italy before national unity, Spain and Portugal – and the northern Catholic countries: France, Austria, Belgium, and Ireland. For Ireland, in particular, the accelerated literacy development from the end of the nineteenth century can be interpreted as part of the struggle for national independence and cultural self-expression: ‘‘In the nineteenth century, compulsory education became a status symbol of national independence.’’15 The interpretations of the different cross-country levels of literacy need not concern us here except for one reason: They show that literacy development is the product of cultural/bureaucratic incentives and needs that are independent of the development of urbanization and industrialization. Precisely for this reason, around 1910 Finland and Ireland were the most literate countries of Europe while remaining among the least urbanized and certainly the least industrially developed. With the exception of Switzerland, none of the other earlycomers to industrialization and urbanization were characterized by particularly high rates of literacy either. Therefore, even in this case there are no problems of a correlation with social mobilization variables. To investigate the link between left electoral levels and literacy development it is necessary to interpolate the estimates reported in Table 4.5 for the decades 1880, 1900, 1920,1940, and 1970 to ensure a complete 14 15

For this information see Cipolla (1969) and Harvey (1989). Flora (1973: 230), drawing from MacElligot (1966).

198

THE POLITICAL MOBILIZATION OF THE LEFT

Table 4.6. Illiteracy and left electoral mobilization

Note: Number of elections given in parentheses.

series of decade values. This is done by taking the mean value between the two. Whenever an estimate was indicated by a range, I have taken the mean value of the range. British decade data have been computed, considering English and Welsh as well as Scottish percentages and computing them in a single percentage for Great Britain that takes into account the relative proportion of each population. With these data modiﬁcations, it was possible to calculate the correlation levels between illiteracy and left vote levels in the set of our elections (or decades); the results are reported in Table 4.6. Over the 293 elections for which data are available, the association reaches ⫺.399, the highest level encountered so far for the whole period. The signiﬁcance of this association should be underlined because the illiteracy data – like those on cultural segmentation – are not characterized by a strong or clear long-term trend. As stated earlier, any trend disappears around the 1930s. This element became evident when we look at the data by period. The negative association diminishes progressively, disappearing after World War II, when rates of literacy remain virtually unchanged in the European countries. At the same time, the high negative correlation in the 1880–1917 period is of great importance in the context of this discus-

CULTURAL HETEROGENEITY

199

sion. Such a correlation goes in the same direction as the overall developmental one; that is, in each historical period, particularly the ﬁrst one, the pattern is the same: Higher levels of left electoral support are associated with higher levels of literacy. This is clearly conﬁrmed by the correlation measures at the decade level. In each decade, higher literacy tends to manifest a stronger left. Illiteracy levels – in contrast to the other two cultural heterogeneity variables, which tended to grow in importance over time – are important factors that help to discriminate among countries in the ﬁrst early phase of left development. Growing levels of literacy accompanied a growing level of left support, and – more important – the early left movements were stronger where literacy levels were higher.

CULTURAL HETEROGENEITY AND LEFT ELECTORAL DEVELOPMENT Given that all the aspects of cultural heterogeneity relate as negatively as expected to the left vote, I have summed the standardized values of the two variables of ethnolinguistic and religious fragmentation and those of illiteracy levels. The combined index consistently yields strong negative associations both over time and in each period and decade. As usual, the model works least well in the very early phase of development, and its explanatory power tends to increase with time. Unquestionably, cultural heterogeneity sets limits on the development of the left, and these limits become increasingly clear and strong after the ﬁrst initial phase of establishment (see Table 4.7). In a scattergram of cultural heterogeneity and the left vote (Figure 4.2) in the context of a 20% of variance explained over the whole period, several important outliers can be noted. Most of the early Swedish, Danish, and Norwegian elections appear in the bottom left part of the scattergram; given the high level of religious, linguistic, and literacy homogeneity of these countries, the model predicts a far stronger left than was actually observable at the end of the nineteenth and the beginning of the twentieth centuries, when the Scandinavian left was even weaker than the others. Ireland is also a rather deviant case, given its high religious homogeneity and high literacy level, attenuated only in part by linguistic differentiation. In the seventy-nine elections between 1880 and 1917, the model performs poorly, explaining roughly 10% of the variance. This is mainly due to the low left vote of the three Scandinavian countries, compared with their high cultural homogeneity, and to the high Finnish vote for a lower level of cultural heterogeneity. In the periods 1918–1944, 1945–1965, and 1965–

200

THE POLITICAL MOBILIZATION OF THE LEFT

Table 4.7. Cultural heterogeneity and left vote

Note: Number of elections given in parentheses.

1975, the outlying cases are progressively reduced to the Irish elections alone, and the variance explained increases from 10% to 18% to 30% and ﬁnally to 50% in the last period.

CONCLUSION: SOCIAL MOBILIZATION AND CULTURAL HETEROGENEITY Considering the evidence discussed in the previous two chapters, it is possible to recognize three basic patterns at work. The variables pertaining to industrialization and urbanization, which I have called the social mobilization model, have a developmental potential for explanation, that is, they point to a left that increases with the growing level of social mobilization, both within the whole set of elections and within each country. However, they fail to account for differences across countries, particularly in the ﬁrst formative periods but also in general. In contrast, religious and linguistic fragmentation are deprived of signiﬁcant power of over-time explanations, both for the whole set and in any country, but they are particularly powerful in explaining cross-country differences in each period; this they do particularly from the 1930s on, although they also are unable

CULTURAL HETEROGENEITY

201

Figure 4.2. Cultural heterogeneity and the left vote (scattergram). to account for early differences in left mobilization. Finally, levels of illiteracy are signiﬁcant as a developmental explication in the sense that left votes increased over time as literacy levels increased, and at the same time they discriminate across countries, but only in the ﬁrst formative periods. Illiteracy is the only variable encountered so far that has some capacity to account for the initial differences in the level of left development. These different patterns can be combined, as they explain different aspects and different phases of the general phenomenon. We can therefore study the two models of social mobilization and cultural heterogeneity together to see whether their interaction compensates for their respective weaknesses. This interaction can be analyzed, ﬁrst, by looking at the mean left vote in different types of settings; second, by devising a combined index that subtracts the cultural heterogeneity from the social mobilization levels (social mobilization⫺cultural heterogeneity); and, ﬁnally, by studying their mutual inﬂuences, as well as their independent roles, in explaining electoral development via a regression analysis.

202

THE POLITICAL MOBILIZATION OF THE LEFT

In Table 4.8 the interaction between social mobilization and cultural heterogeneity is presented by reporting the mean levels of left support in different sociocultural contexts. I have identiﬁed levels of high, medium and low social mobilization and cultural heterogeneity by taking the tertile threshold of the respective frequency distribution of each index. The combination of the two models considerably improves their respective performance. Within each level of cultural heterogeneity the left vote increases with the growing level of social mobilization and within each level of social mobilization and the left vote declines with the growing level of cultural heterogeneity. While the difference between cases of low and high mobilization is about 18% and the difference between cases of high and low heterogeneity is about 12%, when the two are crossed, the difference between the context of low social mobilization and high cultural heterogeneity (the worst left mobilization context) and that of high social mobilization and low cultural heterogeneity reaches 35%. One could hardly have expected a clearer interaction between the two macrofeatures. If we then relate a combined social mobilization and cultural heterogeneity additive index16 to the size of the left, it shows a strong and positive association, homogeneous through time and space. Over almost a century of elections, the remarkable correlation of .778 is reached. Within each period (before 1918: .419 [79]; interwar period: .706 [91]; 1945– 1965: .712 [76]; 1966–1975: .844 [29]) and decade, the association remains positive and signiﬁcant, although it clearly tends to increase with time. For the ﬁrst time, through the interaction of social mobilization and cultural heterogeneity, the period before World War I does not appear to be so deviant a phase. Finally, if we regard the same data in a scattergram form (Figure 4.3), we can see that the number of outliers is deﬁnitely reduced and that no country now appears as generally and systematically deviant. At this stage, the most deviant cases are the early British, Norwegian, and Swedish elections for their underdeveloped socialist vote and the late Irish elections (i.e., by the 1940s – but no longer the elections of the early 1920s and 1930s) for the same reason; and again the early Finnish elections for a left vote that is still, although to a much smaller extent than before, higher than that predicted by the model. The model performs less well in the formation period than in any other and is still unable to account for the early original differences in the 16

Given that social mobilization variables are expected to enhance left support and culturalheterogeneity variables are expected to depress it, the index must be built as social mobilization⫺cultural heterogeneity. Values are standardized given the different magnitude and range.

CULTURAL HETEROGENEITY

Table 4.8. Mean left vote in different sociocultural contexts

Note: Number of elections given in parentheses.

Figure 4.3. Scattergram of the left vote and the sociocultural context.

203

204

THE POLITICAL MOBILIZATION OF THE LEFT

Table 4.9. Regression analysis of the sociocultural model

socialist vote between the end of the nineteenth century and the beginning of the twentieth. While the total variance accounted for over the whole period is about 60 % – a remarkable result indeed for a model applied not within each country, but to the universe of hundreds of elections scattered across thirteen different countries and almost a century – its capacity to explain the pre–World War I variance does not exceed 20%. However, the ﬁgures are certainly less puzzling than those discussed for social mobilization alone (Table 4.9). In the regression, cultural heterogeneity emerges as the more important variable of the two in each subperiod, although not for the whole period. On the whole, it adds the same amount of explained variance as social mobilization alone had over the whole period. Many other factors discussed earlier help us to conclude that this ordering is not a simple computing artifact. Of the two aspects, it is cultural heterogeneity that

CULTURAL HETEROGENEITY

205

carries the most weight in determining the cross-country limits of left development and explains most of the early differences among the countries. To some extent, it can be said that cultural heterogeneity sets the conditions of validity of the hypothesis linking left development to the patterns and levels of industrialization and urbanization more than the latter do for the validity of the cultural heterogeneity hypotheses. This is evident when we compare the different orders of correlation. The ﬁrst order of correlation between social mobilization and the left vote is modest in the ﬁrst period (.180), but it is almost doubled once the impact of cultural heterogeneity is controlled (.331). We are still missing a genetic model of left development to account for the wide differences in the starting conditions of political mobilization. Up to now, variables have been limited to the social, prepolitical conﬁning conditions of electoral mobilization. In the following chapters, I add to the scheme other variables that are more political, that is, concerning the institutional conditions and the organizational environment in which electoral mobilization was taking place. The aim is the same as before: to search for general cross-country and cross-time factors capable of deﬁning the macro constellation of left mobilization and restricting the scope and role of more-idiosyncratic factors related to the different structure of opportunity offered in each period and country. The institutional context of electoral participation, that is, the process of political democratization, is the subject of the following chapter.

5

ENFRANCHISEMENT

THE ROLE OF THE FRANCHISE he social mobilization and cultural heterogeneity factors are related to the speciﬁc historical context and its traditions. It is unlikely that any generalized notion of ‘‘industrial society,’’ conceived as a syndrome of structural/cultural features, will help to delineate the contextual political responses of the early phase. Similarly, the inhibiting capacity of cultural heterogeneity on the development of the left is unlikely to make its impact felt clearly in the very early phase of electoral development. It performs better as a limiting condition over the long term or as a potential boundary for mature socialist movements. While the long-term forces of assimilation and standardization may ultimately create similarities and bring about a leveling of both the social structure and cultural attitudes of industrial societies, in the very early phase of mass politics the opposite is more likely: The contextual features of presocial mobilization are more important than long-term developmental forces. Therefore, in the course of this book, I will consider more context-related factors. The development of political rights, in particular the right to vote, was the end result of a long historical process going back to the eighteenth century and was rooted in the development of civic rights. Civic rights developed primarily in relation to the market as rights of property, contract, unrestricted choice of residence and workplace, and so on. Civic rights also refer to the potential for associability in a society when they touch on freedom of faith, thought, speech, assembly, and association. The combination of these civic rights constituted the point of departure for the

T

206

ENFRANCHISEMENT

207

opening up of political public space and opinion. The successive development of strictly political rights constituted the decisive push for political mobilization, granting the legal basis for the development of interest groups and political parties.1 The development of voting rights should not, however, be seen as a linear development of previous and prerequisite rights of expression, association, and opposition. Even in the second half of the nineteenth century, the combination of all these factors was rather complex. We ﬁnd that cases of extended voting rights were not accompanied by ﬁrmly established rights of association and expression, as well as the reverse. This is due to the ambiguous role that the process of enfranchisement played in the eyes of the ruling elite. From 1848 to the First World War, the level of suffrage granted indicated two quite different situations: suffrage as a device of national integration and suffrage as a device of political representation. In the ﬁrst case, high levels of suffrage were granted from above as an instrument for integrating social groups nationally, but they were blocked as an instrument of representation by a vast set of inequality devices: from the curia/estate system to plural voting or institutional barriers, from irresponsible government to second-chamber predominance; and so on. Such institutional devices actually prevented suffrage from properly representing the electors and granting them a share of parliamentary and executive power. These inequalities should be incorporated in the analysis of the franchise to avoid drawing false inferences from sheer numbers of the enfranchised population. On the other hand, restricted or enlarged representational suffrage was based on and impinged on already established opposition and association rights and constituted their representational expression. The nature of the suffrage is therefore characterized by other factors that differ from the suffrage itself and that clarify its role and function within the political system. A discussion of this aspect requires a typological cross-tabulation of both democratization and liberalization, which will be dealt with in Chapter 7. In this chapter, therefore, I concentrate on classifying European experiences in terms of the levels of electorate enfranchisement and turnout. It is difﬁcult to ﬁnd a common dimension along which to rank-order national cases.2 In this chapter, I ﬁrst distinguish cases along an early versus 1

2

For the waves of developments of rights, see Marshall (1965: 71–134) and Bendix (1964: 89–126). A separate issue concerns the database. In historical sources, there is little systematic standardization of estimates of enfranchised populations. The electorate is sometimes calculated as a percentage of the total or male population and sometimes as a percentage of the adult (male or total) population, where the latter is deﬁned by the legal standard

208

THE POLITICAL MOBILIZATION OF THE LEFT

late dimension, a sudden versus gradual dimension, and a continuity dimension (with or without important reversals). Subsequently, I deal with the comparative levels of participation (turnout) to ascertain the extent to which the formal right of voting was actually exercised. After describing the cross-country cross-time variance, my analysis concentrates on the relationship with the class left vote, treating the following questions: Were levels of franchise, turnout, and their combination important in determining levels of the left vote? Was the pattern of enfranchisement (tempo and timing) responsible for the earliness or level of left development? Did big jumps count the most, or were step-by-step enlargements more favorable to the left? Was the left deﬁnitely launched by universal male suffrage? What was the impact of female enfranchisement? Finally, at the end of the chapter, I discuss the causal role of the enfranchisement process in the left mobilization model. So far, electoral support for the left has been analyzed as if enfranchisement levels were equal across the countries, which was not the case in the period between 1880 and 1920. In cases of highly restricted and/or particularly unequal franchise, left mobilization is unlikely to be high, whatever the levels of other independent variables. Given the link between suffrage development and left vote development, it may seem surprising that the analysis did not start immediately with this factor. There is, however, a reason for this choice. Social mobilization and cultural heterogeneity are exogenous variables in this model; they represent givens whose origins lie outside the scope and interest of this research. The status of enfranchisement is different and more complex. It is different because it cannot be considered a of the country for adulthood or by the age indicated in the different laws: thirty, twentyseven, twenty-four, and, later, twenty-one and eighteen years (sometimes independent of and generally higher than the legal age of adulthood for civic rights, as voting was regarded as an especially demanding activity). Sometimes, instead, ﬁgures are offered as a percentage of the enfranchised age group, i.e., of the population above the legal age indicated by the electoral law. This makes a rigorous comparison across time and space very difﬁcult. A single measure is necessary that takes the number of people with the legal right to vote as a percentage of a reference group that is relatively stable and homogeneous over time and across countries. Take a simple example to illustrate the problem at its worst: Certain reform acts of the second half of the nineteenth century actually lowered the property, income, capacity, or census requirement for voting, increasing the number of enfranchised people. However, at the same time, the reform modiﬁed the age limit for voting by lowering it – for instance, from thirty to twenty-four years. The ﬁnal result of these sorts of changes is that when the percentage of enfranchised people is computed in reference to the adult population as deﬁned by the electoral law, these reforms can actually result in a decline of the percentage of the enfranchised. I have used a single international source that offers the ﬁgure of the electorate as a percentage of the total population (male and female) twenty years of age or older (Flora et al., 1983). This solution has clear advantages.

ENFRANCHISEMENT

209

potential explanation of the vote but rather a necessary precondition: Voters did not choose socialist parties because they were allowed to vote but rather when they were allowed to vote. It is more complex because enfranchisement cannot be regarded as totally independent of the left vote, nor can it be considered an exogenous variable with regard to social mobilization/cultural heterogeneity. The level of franchise can be seen in three ways: 1. As an endogenous variable in the sense that its level inﬂuences left electoral mobilization but is also inﬂuenced by social mobilization and cultural heterogeneity. In this case, the relationship between the variables is as indicated in Figure 5.1a. The effect of the franchise on left development includes both a direct component and an indirect one, represented by the effect of social mobilization and cultural heterogeneity on the level of the franchise. 2. As a conditional variable vis-a`-vis social mobilization and cultural heterogeneity: The effect of the latter depends on the value of the former, and the possibility of social mobilization (or cultural heterogeneity) having an effect on the left vote is conditional on a certain level of enfranchisement. The conditional variable determines the size of the causal effect of other exogenous variables, as presented in Figure 5.1b. 3. As a reciprocal causation variable: The extension of the franchise may foster left electoral mobilization, but the level of left electoral mobilization may also be responsible for further enlargement of the franchise (as indicated in Figure 5.1c).

EARLINESS How many individuals were allowed to vote during the nineteenth century? The ﬁgures reported in the Appendix are divided into four periods: 1830–1880, 1881–1917, 1918–1944, and 1945–1975. The same data are charted in Figure 5.2 (see later). In the 1848–1880 period, three countries stand out as earlycomers to relatively large suffrage. France, after experimenting with several formulas in the post-Napoleonic period from 1815 to 1846,3 suddenly introduced universal male suffrage (the suffrage had 3

In the period immediately after the Restoration (1815–1816), suffrage was kept almost universal and equal for males (dependants were excluded), although this was only for primary elections to elect lifetime members of electoral colleges. It was the latter who elected members of parliament. Between 1824 and 1830, elections became direct, although suffrage was restricted and unequal (the upper 25% of the electorate, paying the

210

THE POLITICAL MOBILIZATION OF THE LEFT

Figure 5.1. Role of the franchise. already been equal since 1831) for citizens over twenty-one years in 1848, with an increase in the electorate from about 1–2% to 36%. In Switzerland, universal male suffrage for citizens twenty years of age or more was introduced with the constitutional reform that followed the Sonderbund of highest direct taxes, was made into an additional electoral body) for men over thirty years of age. Between 1831 and 1846, suffrage was equalized for men over twenty-ﬁve, but it still remained highly restricted. For detailed information, see Campbell (1958) and Huard (1991).

ENFRANCHISEMENT

211

autumn 1847, bringing the enfranchised electorate to about 30%. Swiss citizens, however, had never really been subjected to any re´gime censitaire and had a long tradition of general voting in the mountain cantons.4 So, in the Swiss case, there was less of a break with the past. Finally, Denmark introduced equal male suffrage in the wake of the 1848 revolution, and this produced a radical change from autocracy to proto-democracy in the kingdom. The principle, however, was tempered by many more restrictions than in the other two cases5 and was applied to men over age thirty. This resulted in an smaller enfranchised electorate, standing at about 25%.6 The fourth earlycomer to high levels of male suffrage was Germany. The short-lived Frankfurt Assembly of 1848 was elected by universal male suffrage, and after that date many German states could count on a fairly large male electorate. In particular, in the primary elections in the Kingdom of Prussia, every male citizen twenty-four years of age or more was entitled to cast a vote,7 although the suffrage was not only indirect but also highly unequal.8 The two elections held in 1867 for the Reichstag of the North Confederation involved universal male suffrage (for citizens twenty-ﬁve years of age or more), and the electorate was estimated to be about 35% of the adult population. After 1871, all elections were direct, equal, and male-universal. In the middle of the nineteenth century, all the other countries had very restrictive suffrage requirements, with electorates ranging between 3% and 8% of the population twenty years old or more; it makes little 4

5

6 7

8

Electoral inequalities had been intrinsic to the system, however, through the electoral privileges of the cities (of the plateau), against which the 1830 and 1833 ‘‘revolutions’’ had been directed and which had led to the enlargement and equalization of the suffrage. Martin (1980: 252). Electors had to reside in the electoral district for at least one year; dependent people without a family were excluded, as well as those who received poor relief and those whose patrimony was under bankruptcy proceedings. For details about these limitations, see De Kiriaki (1885: 82–88), which is an invaluable source for the detailed legal aspects of most early-nineteenth-century electoral systems. See also Castairs (1980: 75–77). On Danish electoral reforms see Elklit (1980: 366–396). For details about the electoral requirements of Prussia and several other German states, see Nohlen and Schultze (1971). The primary elections selected grand electors. Primary electors were not subject to any restrictions but were divided into three classes on the basis of the taxes they paid. The total sum of the taxes paid in the constituency was divided into three equal parts. The ﬁrst class included the biggest taxpayers up to a ceiling of the ﬁrst one-third of the total tax income. The second and the third classes were determined by the same principle, with obvious inequalities as a result. According to Casertano (1911: 204), in 1893 the proportion of the electorate in the three classes was still 3% in the ﬁrst, 11% in the second, and 86% in the third. He cites the curious example of the cases of Essen and Frankfurt, where the Krupps and Baron Rothschild respectively, were the only members of the ﬁrst class.

212

THE POLITICAL MOBILIZATION OF THE LEFT

sense to distinguish between them. One case that can possibly be singled out is that of Norway. The 1814 Norwegian constitution introduced the most liberal voting qualiﬁcations of the time, enfranchising about 25% of all men.9 This corresponded to roughly 10% of the population aged twenty years or more, which, up to 1848, was the highest franchise level in Europe. However, 1848 passed unnoticed in Norway and the electorate remained stable at around 9%, even declining to 8% in the 1870s. By that time, the Norwegian franchise could no longer be regarded as comparatively high. By the late 1880s in Austria, Sweden, and the United Kingdom, the electorate had passed the 10% threshold. In Austria, the ﬁrst direct elections to the lower house were held in 1873 (in previous elections, the deputies had been indirectly elected by provincial diets), with an enfranchised electorate of about 10%.10 Similarly, in Sweden, the establishment of a second chamber and of centrally recorded elections came after 1866, and throughout the 1870s the electorate was about 10%. Finally, in Britain, the electoral reforms of the 1867–1872 period brought the electorate to about 15% of the male population twenty years of age or older. The 1880–1920 period is the crucial phase of suffrage extension. The ﬁrst countries to extend the right to vote substantially were Ireland and Great Britain through the reforms of the mid-1880s, which introduced a uniform household franchise, a uniform lodger franchise, and a uniform £10 occupation franchise in every borough and county throughout the country while leaving ownership franchise differentiated. The electorate was increased by 80% by these measures and reached about 30% of the adult population.11 Three other countries enlarged their electorate to reach a third of the adult population, corresponding roughly to universal male suffrage, before the turn of the century: Belgium (1894 ﬁrst election), 9 10

11

Rokkan (1966a: 247). The information in the text refers only to the seven provinces carved out of the Germanspeaking westernmost part of the pre–World War I empire – Cisleithanian Austria – which subsequently became the Republican territory (Treaty of Saint-Germain-en-Laye of September 10, 1919). Burgerland became an Austrian province later in 1921; earlier, it had consisted of the westernmost counties of the three Hungarian provinces of Moson, Sopron, and Vasvar. Obviously, the series of reforms introduced in this period in the United Kingdom, in particular the Corrupt and Illegal Practices Act (1883), the Representation of the People Act (1884), and the Redistribution of Seats Act (1885), had a number of other political implications. Among the most important ones were an alteration in the balance of representation in favor of urban and industrial areas, the single constituency as a rule (although a few exceptions remained), and a sharp increase in the competitiveness of contests, with a decisive drop in ‘‘unopposed’’ constituencies. See Butler and Cornford (1969: 334). Here we are concerned only with the extension of the electorate.

ENFRANCHISEMENT

213

Austria (1896), and Norway (1900). The Belgian 1893 reform that suddenly increased universal male suffrage was introduced for the national assembly for male citizens age twenty-ﬁve or more, increasing the electorate from 3.9% to 37.3%, even though huge inequalities were maintained.12 In Austria, elections to the lower house (the Abgeordnetenhaus) continued to be held according to a curia system that divided the house into four classes.13 The total enfranchised electorate was estimated at about 12–13% in the 1890s. The 1896 reform added a ﬁfth curia (72 seats compared to the 353 deputies of the other four), which had a general character and male universal suffrage. This brought the enfranchised population to about 36% of the adult population (see the legend in the Appendix table). Finally, in Norway, the suffrage increased slowly from 10% to 16–17% by the end of the nineteenth century through gradual reforms that extended suffrage from property and occupational requirements to those of citizens paying a minimum tax on income (1885). In 1898, universal suffrage for men aged twenty-ﬁve years or more was achieved, bringing the electorate to about 35%.14 Later, the Norwegian electorate continued to grow by marginal increases and progressive enlargements; thus, in 1907, a proportion of the female electorate (about 48%) was enfranchised; in 1913, universal suffrage for adult women was introduced (it had been preceded by women’s universal suffrage at the local level in 1911), bringing the enfranchised adult population to 77%, the second highest level of enfranchisement in Europe at that time after Finland. Overall, the Norwegian pattern of extension was gradual and consistent. Finland represents the unique case of a relatively late and extremely sudden male universal suffrage (but early female suffrage). Between 1809 and 1867, a four-diet system represented the heads of noble families, the clergy, the city dwellers (one or two representatives for each town or group of towns), and the peasants (one representative by jurisdictional district). From 1872 to 1904, the procedures and qualiﬁcations for voting did not change fundamentally, except for allowing school and university teachers and civil servants to vote in the clergy curia. The franchise was probably very restricted in this period. In 1904, an increase of the electorate brought the enfranchised adult population to about 9%. The 1906 reform, follow12 13

14

See Gilissen (1980: 338–365). The ﬁrst (85 seats) consisted of male landowners who paid at least ﬁfty ﬂorins in taxes per year; the second (21 seats), of members of the Chambers of Commerce and Trade; the third (118 seats), of all male urban dwellers twenty-four years of age or more who paid ten ﬂorins or more; and the fourth (129 indirectly elected seats), of male ruralcommune residents who paid at least ten ﬂorins. For more details, see Rokkan (1970c).

214

THE POLITICAL MOBILIZATION OF THE LEFT

ing the temporary loosening of the Russian hold on Finnish political affairs, suddenly introduced universal male and female suffarage (over twenty-four years of age) direct and secret elections, and even proportional representation. In a single reform, the new Finnish unicameral Parliament (Eduskunta) was elected by 76% of the adult electorate.15 The Netherlands, Italy,16 and Sweden adopted universal male suffrage last. In Italy, the Zanardelli Reform Act of 1882 had signiﬁcantly increased the electorate to about 13% by lowering the male voting age from twentyﬁve to twenty-one and by reducing the tax, wealth, and educational requirements. Before this reform, 80% of the electorate was given the right to vote on the basis of tax and property qualiﬁcations; after the reform, this percentage dropped to 34.7%, while 63.5% were included due to their intellectual and educational capacities.17 However, in the period of antisocialist legislation starting in 1894, electoral registers were revised and educational tests were made more stringent, with the result of actual disenfranchisement of almost 5% of the adult population. The electorate fell back to pre-1880 levels until the 1912 electoral reform, which introduced almost universal male suffrage (for men over thirty) and brought the electorate to 42% of the adult population.18 The Swedish pattern of franchise development resembles the Italian one closely. Although no marked disenfranchisement occurred in Sweden, the electorate remained fairly stable throughout the 1880s and 1890s at around 10% of the adult population and rose to about 15% in the ﬁrst decade of the twentieth century. In 1909, only a couple of years before Italy, almost universal male suffrage for citizens twenty-four years of age and over was introduced, doubling the electorate from 15.8% to 32.8%. The pattern of the Netherlands differs from those of Italy and Sweden, as the ﬁrst steps of suffrage enlargement took place earlier and the whole process was more gradual. The reform of 1887 doubled the electorate from 5.7% to almost 12% of the adult population by lowering economic requirements. A second doubling of the electorate from 11% to 20% took place with the reform of 1896, 15 16

17 18

For Finnish early electoral developments, see Mylly (1984: 9–27). The referenda (plebiscites) in central and southern regions from 1860 to 1870 for ratifying their annexation to the Italian kingdom were held under almost universal male suffrage. However, this enlarged franchise was an instrument of international legitimation of the new state. By contrast, the ﬁrst elections of the Italian parliament in 1861 were held under very restricted suffrage, and only about 240,000 people voted (57.2% of the electorate). See AA. VV. (1955), Shepis (1958), and Ballini (1985) and (1988). The suffrage was universal for males over age thirty, but certain citizens between twentyone and thirty had the right to vote if they paid a minimum level of taxes or had completed military service, ﬁnished primary school, or exercised ofﬁcial functions.

ENFRANCHISEMENT

215

which enfranchised large parts of the lower middle classes, many highly qualiﬁed workers and sections of the rural proletariat, and small farmers and tenants. The electorate grew to almost 28% of the adult population in the last pre–World War I election (1913), and universal male suffrage was introduced after the war in 1918. The ﬁnal stage of enfranchisement concerned the female electorate, and in most cases it was a sudden decision. Only two countries had enfranchised women before World War I: Finland in 1907, together with men, and Norway between 1909 (for women whose own or husband’s income exceeded a minimum) and 1915. In Austria, Denmark, and Germany, female enfranchisement took place in a single step immediately after the war, between 1918 and 1919. The United Kingdom and Ireland enfranchised women thirty years of age or more (with certain minimal limitations) in 1918, completing the process in Ireland in 1923 and in the United Kingdom in 1928. In both cases, the age limit was brought down to twenty-one years, as for men. The Netherlands and Sweden extended suffrage to women in 1921 and 1922, respectively. Finally, Italy, France, and Belgium granted female suffrage only after Word War II, between 1945 and 1948. Well in the rear came Switzerland, which gave suffrage to women at the national level only in 1971, 123 years after the same right was granted to men. This brief description shows that a simple and straightforward classiﬁcation of national experiences is difﬁcult even along a single dimension such as earliness/lateness. Beyond the four clear-cut cases of France, Germany, Switzerland, and Denmark, the relative positions of the other countries has changed from decade to decade. Thus, Switzerland and France, which were in the lead up to the 1860s, formed part of the group at the lowest level by the 1920s; Denmark’s early start was followed by a stagnation that allowed several countries to catch up with it by the 1870s and 1880s. Germany is probably the only country that consistently maintained its position in the ﬁrst ranks.

TEMPO Considering the same historical data in terms of the rapidity or lengthiness with which suffrage was extended clariﬁes whether similar levels at a given time were the result of gradual growth or sudden expansions of the electorate. To help the reader follow the discussion, I have reproduced the development of the electorate over time in graphic form for each country (Figure 5.2). Next to the line showing the development of the electorate,

216

THE POLITICAL MOBILIZATION OF THE LEFT

Figure 5.2. Franchise, turnout, and left vote: national developments.

ENFRANCHISEMENT

Figure 5.2. (cont.)

217

218

THE POLITICAL MOBILIZATION OF THE LEFT

there is also the dotted line of the development of the turnout over time (the percentages of valid votes, although occasionally only total votes were available) by the electorate, as well as the percentage of valid votes obtained by the left. These two latter lines will help this discussion later. For the moment, let us concentrate on the sudden versus gradual nature of the electorates’ development. In the vast majority of cases, enlargement of the suffrage proceeded in relatively large steps. Sudden and large changes doubling the electorate occurred where there was female enfranchisement in all but three cases. Only in Norway, Ireland and the United Kingdom did female enfranchisement proceed in two main consecutive steps: in 1909–1915, 1918–1923, and 1918–1929 (I refer, as usual, to the ﬁrst elections under new rule). Changes in male suffrage were more highly differentiated, but in this case too, jumps predominate over gradual evolution. If a jump is deﬁned as an increase of more than 10%, a change of at least this size is present everywhere except in the four earlycomers (France, Germany, Switzerland, and Denmark), which, after their early sudden enlargement, remained at the same level for almost seventy years, until after World War I. Earlycomers could not have gradual development, so cases of early and gradual enlargement are as impossible as those of late and gradual enlargement. All the other cases were characterized by larger or smaller jumps. Going from the earlier to the later jumps, the cases are the following: The United Kingdom produced the ﬁrst important jump of about 13% in 1885 and again of about 13% in 1918; Ireland jumped by 18–19% at the same time; Belgium, by 33.4% in 1894; Austria, by 22% in 1896; Norway, by 18.2% in 1900; Finland, by almost 30% in 1907; Sweden, by 16.7% in 1911; Italy, by 27.2% in 1913; and the Netherlands, by almost 12% in 1918. The size of these changes varied. The most sudden increases, affecting a third of the adult population, were no doubt experienced by Belgium at a very early stage and by Finland and Italy later. Austria’s sudden increase affected about a ﬁfth of the adult population in 1896. Ireland (1886), Norway (1900), and Sweden (1911) experienced smaller jumps, to around 18% of the adult population. Finally, the Netherlands had only one jump, which came very late, after World War I, and was also the smallest – just above the 10% limit, like the two British increases. Any other changes not mentioned here can safely be considered gradual adaptations due to small modiﬁcations in economic and/or capacity requirements and in revisions of the electoral lists. In Table 5.1, the average yearly increase in (only) the male electorate from the 1860s to the 1920s is reported. Countries are regrouped according to the size of major increases, and the table offers information about

ENFRANCHISEMENT

219

Table 5.1. Rates of growth per decade in the male electorate

a

Stable: although no precise data are available it is known that no substantial modiﬁcation of the electoral legislation had taken place in the period. b Data drawn from estimates in secondary literature. n.a.: neither data nor estimations available.

the location and size of the major increases. Following the table from top to bottom, there is a decline in the size of the per annum increases in the case of major enlargements and a growing level of per annum increases in the decades that are not characterized by any major redeﬁnitions of the franchise. The Netherlands shows the most clear-cut case of gradual enlargement. It presents a jump exceeding 10% only in the ﬁnal phase after World War I. In all the other decades (but one), the average increase is considerably higher than in the other cases, indicating a process of truly progressive enlargement of the electorate. In the three decades preceding the ﬁnal granting of universal male suffrage after World War I, the electorate was increased by about 6%, 10%, and again 6%, for a total of 22% (see the table in the Appendix and Figure 5.2i). Britain, despite its reputation for very gradual development, presents two peaks, while in other decades the rates of growth are close to zero. Ireland shows a rela-

220

THE POLITICAL MOBILIZATION OF THE LEFT

tively important increase in the 1880s – bigger than that in Great Britain (2.07% versus 1.29%) – but, unfortunately, there are no data for the 1890s and 1900s. Norway is also usually associated with gradual development, but the increase in universal male suffrage in the 1890s was big. Gradual development characterized the pre-1890 and post-1900 periods, but in the 1890s the Norwegian electorate increased by about 20% of the adult population. Without forcing national cases too much into the comparative framework, I can now provide in Table 5.2 a classiﬁcation of the Western European enfranchisement process along the two dimensions of its timing and tempo. The most difﬁcult case to classify is that of the Netherlands, whose development is unquestionably gradual, but which in terms of timing is a relative latecomer until the end of the ﬁrst decade of the twentieth century. We could thus classify it as gradual and intermediate as well as gradual and late. I have, after much thought, placed the Dutch case in the intermediate timing group because the levels of enfranchisement in the 1890s and 1900s were considerably higher than those of Italy and Sweden (classiﬁed as latecomers).

REVERSALS A third dimension of suffrage enlargement is the existence or absence of reversals, that is, the more or less linear nature of the enlargement itself. France is the classic case for which the ‘‘early, sudden and followed by reversal’’ category was originally developed. It is not correct, however, to concentrate on the Restoration reversal vis-a`-vis revolutionary times, when the outstanding characteristic of the French pattern remains that of very early universal male suffrage. After 1848, a revision of the electoral lists (the law of May 31, 1850) attempted to restrict the franchise by demanding, as a prerequisite, three years of residence in the voting place. However, this reform had only minor effects: The electorate changed from 9,837,000 in 1849 to 9,836,000 in 1852 to 9,490,00 in 1857, rising again to over 10,000,000 in 1863, which can hardly be regarded as mass disenfranchisement. In Denmark, the original democratic promises of the 1848 revolution were to some extent muted in the following decade. Confrontation between the king and the conservatives, with their strongholds in the ﬁrst chamber, and the rurally supported liberals in the second chamber resulted in a minor dedemocratization of the constitution itself. In 1866, suffrage was also restricted, but the impact of such changes was felt particularly in the

ENFRANCHISEMENT

221

Table 5.2. Comparative enfranchisement: timing and tempo

Landsthing (the ﬁrst chamber), where higher property qualiﬁcations were introduced. However, strictly from the point of view of the suffrage, these conﬂicts were manifested more clearly in the stagnation of a relatively high but not yet universal male suffrage throughout the 1850s and 1860s. No real signs of signiﬁcant disenfranchisement are evident in the post-1849 ﬁgures.19 The only case of a franchise reversal that has signiﬁcant implication for our analysis is that of Italy in the 1890s. The revision of the electoral registers carried out in 1894 reduced the electorate from 2,934,000 in 1892 to 2,121,000 and constituted a clear rupture in the process of growth initiated by the reform law of 1882. Almost a third of the electorate lost its right to vote in 1895. Moreover, this disenfranchisement was without doubt one of a set of measures set up by the government in a clearly antisocialist operation.20 The effects of the disenfranchisement lasted for a long time; it took seventeen years and ﬁve elections before, in 1909, the electorate reached the same numerical level at which it had been in 1892. This revision was of great political importance because, without it, Italy would show only a very gradual pattern rather than a sudden and late development. So, only in Italy do we ﬁnd a disenfranchisement that combines size, lasting inﬂuence, and a clear antisocialist political orientation.

TURNOUT During the process of enlargement of the right to vote, the propensity to participate varied remarkably over time and across countries. Here, the interest is not in short-term ﬂuctuations from one election to the next, but rather in systematic cross-time and cross-space differences in levels of 19 20

See Sva˚sand (1980: 398–411). See Mastropaolo (1980: 97–124).

222

THE POLITICAL MOBILIZATION OF THE LEFT

electoral participation and their relationship to the process of enfranchisement.21 The correlation between the electorate and turnout tends to be weak. It is .583 for the 390 elections between 1831 and 1975, but it is slightly negative in the period before 1880 (⫺.067) and in the period between the two world wars (⫺.085), while it is positive in the period 1881–1917 (.138) and in the period following World War I (.439). These data raise some doubts about the validity of the thesis that census- or capacity-restricted electorates have a higher turnout than enlarged electorates as the result of the strong propensity of the richest and most highly educated strata of the population to vote. According to this thesis, successive enlargements of the franchise, resulting in granting of the right to vote to citizens of a lower socioeconomic and cultural status, should result in a lowering of the turnout. This would be particularly true with large, sudden increases of the electorate and, I should add, with enfranchisement of women, whose sudden involvement in a political role traditionally associated with and reserved for men should produce a pronounced fall in turnout levels. However, when we consider the mean level of turnout by decade between 1831 and 1980,22 the ﬁgures offer only modest proof of a curvilinear historical effect of franchise development over turnout levels. These levels are no higher during the phase of very restricted suffrage than during the period of suffrage enlargements. This conclusion is strengthened when we examine the mean level of turnout by levels of enfranchised electorate in Table 5-3. The only fall in turnout (from 75.4% to 64.8%) comes in the category of the electorates ranging from 60% to 80% of the adult 21

22

There are no systematic historical studies of turnout levels and very few solid generalizations about their variation. One of the few generalizations is that turnout tends to correlate historically and cross-sectionally with levels of urbanization – ﬁrst stated by Thompson (1927). Urwin (1980: 112–123) has argued that the data presented in this research did not fully support the author’s conclusion, and he concludes that no clear, systematic ‘‘rural lag’’ in electoral participation can be identiﬁed. My data do not concern within-country variance between urban and rural areas but only cross-country differences in levels of urbanization across time. Turnout ﬁgures are associated positively with the four different urbanization levels over the entire period and in each decade. The associations are systematically stronger for the 2,000 and 5,000 threshold than for the 20,000 and 100,000 ones. The associations are stronger in the period from the 1910s to the 1950s than in the others. This suggests that turnout levels tend to increase with the increase in urbanization and also that more-urbanized countries tend to have higher levels of turnout than more-rural ones. In short, my data support Thompson’s conclusion more than Urwin’s. I repeat, however, that they both refer to within-country and not to cross-country variation. Note that before 1870–1880 the data are far less representative because some countries are missing and others held very few elections (see the table in Appendix for a list of the elections considered).

ENFRANCHISEMENT

223

Table 5.3. Mean turnout by levels of the enfranchised electorate

population and can be interpreted as the result of female enfranchisement.23 In conclusion, at the average European level there is no evidence of a long-term curvilinear development in turnout. If anything, it seems that the turnout has experienced a long-term structural growth parallel to that of the franchise. However, national patterns could be more supportive of the curvilinear thesis. Checking the plot of the electorate and turnout ﬁgures for each country reveals that this is the case only in Norway. The fact that Rokkan gave a great deal of attention and publicity to this Norwegian pattern has probably turned it into something more than a speciﬁc national pattern. However, there is no evidence of its actually being so. Major enlargements of the suffrage can be accompanied by a fall in turnout even if they do not result in an overall curvilinear historical development. I have checked for each major increase (greater than 10%) in the electorate, as well as the corresponding variation in the turnout level for the same election and the following ones when useful (Table 5.4), separating the major increases resulting from exclusively male, male and 23

Since I consider the enfranchised population as a percentage of the total population aged twenty years or more, two countries that have both universal male or male/female suffrage but apply different voting ages – for instance, twenty-one and twenty-nine years – result in two different percentages, of which the ﬁrst is much higher. To avoid this result, one should compute the enfranchised population as a percentage of the enfranchised age group, i.e., the population above the legal voting age. However, to do this would involve even bigger distortions. Given that the voting age in the past was very different from one country to another, I would lose much of the common bases for comparative purposes. How, for instance, can an 80% enfranchised population over thirty years of age vis-a`-vis a 70% enfranchised population over twenty years of age be evaluated in comparative terms?

224

THE POLITICAL MOBILIZATION OF THE LEFT

Table 5.4. Major increases in the electorate and corresponding variations in turnout

* The turnout differential is missing, as no pre-1907 data are available. However, in the elections following 1907, the turnout dropped drastically.

female, and exclusively female enfranchisement. No general tendency toward a fall in turnout can be ascertained in the case of major enlargements of the franchise. Paradoxically, the most supportive instance of the three listed in the table is that of male enfranchisement. If we consider that the Belgian 1894 and the Dutch 1918 enfranchisements were associated with important increases in turnout because of the introduction of compulsory voting, we are left with three cases of turnout decline at the moment of electorate enlargement (Sweden in 1911, Norway in 1900, and Italy in 1913) and two of turnout stability and even growth (the Netherlands in 1897, ⫹16.7%; France in 1848, stable). In the enlargements involving part of the male and part of the female electorate (middle three columns), the cases of turnout fall still prevail slightly (they all occurred between 1907 and 1918). The most signiﬁcant, however, is the British drop of 14.6%, while the others are minor changes. Finally, and most surprisingly, no general support for the thesis advanced earlier is offered by the female enlargements, which produce an equal distribution of increases and decreases in turnout. In a nutshell, it is hard to interpret these data in view of a general or predominant tendency; indeed, national contexts seem to dominate and be responsible for individual outcomes. The only two countries where increases in the electorate invariably result in falls in turnout are Sweden and, in particular, Norway. The last point worth discussing is the comparative cross-country vari-

ENFRANCHISEMENT

225

ation in average levels of turnout before the 1920s. The different levels of turnout in the European countries can be appreciated by looking at Figure 5.2. A close examination of these graphs shows a great variety of national patterns. Belgium, Austria, and, to a lesser extent, France and the Netherlands are the prototype of the high-turnout country irrespective of the level of the electorate. In these cases, the level of the electorate does not seem to inﬂuence signiﬁcantly the level of turnout, which is always comparatively high. The United Kingdom and Finland have a medium level of turnout in this period, between 60% and 70%, and it oscillates around these ﬁgures irrespective of electorate development. In Denmark and Germany, turnout grows steadily over time irrespective of the trend in the electorate. In these two cases, but more clearly in the Danish one, it seems that electoral mobilization takes place mostly through the progressive involvement of enfranchised people and that this increase tends to be gradual over time without being profoundly affected by changes or stability in the electorate. Finally, in Sweden, Italy, and Switzerland, turnout is comparatively low throughout the entire period. In Italy, it remains stable below 50–60% until after World War I. In Sweden, it grows over time, but (see Figure 5.2k) its starting level was so low (only 20% in 1872) that at the beginning of the twentieth century less than half of the restricted electorate actually voted. Switzerland has a stable, secular low level of turnout, with the exception of the jump that occurred in the twenty years between 1919 and 1939.

ENFRANCHISEMENT AND THE DEVELOPMENT OF THE EUROPEAN LEFT Each feature of suffrage development discussed so far – earliness, graduality, and levels of legal and real participation – may have important consequences for the electoral and/or organizational and/or ideological response of labor. The effects of these features on organizational and ideological development will be taken up in later chapters. Here, I will continue to deal mainly with their effects on electoral developments. PATTERN OF DEVELOPMENT: TIMING AND TEMPO Four fundamental patterns of development can be derived from Table 5.2, where the countries were classiﬁed according to the timing and tempo of their enfranchisement. At one extreme are those countries that show early

226

THE POLITICAL MOBILIZATION OF THE LEFT

or intermediate timing and a necessarily gradual development (the United Kingdom, Ireland, and the Netherlands); at the other are those that show late or intermediate timing and a rapid tempo (Finland, Italy, and Sweden). In the intermediate cases, one can distinguish between the earlycomers with sudden enlargements (France, Denmark, Germany, Switzerland), on the one hand, and the intermediate cases on both dimensions on the other (Belgium, Austria, and Norway). In Table 5.5 the average left vote for each period and for the whole century for each of these four groups of countries is reported. The order is quite consistent for each of the four patterns and for the whole period. In all cases, the left appears to be electorally stronger moving from the cases of early timing to those of late timing and, within each of these groups, moving from gradual to sudden enlargement. The contrast between the early/ gradual and the late/sudden groups is remarkable, and the two intermediate groups always remain within this range of extreme values. The most striking difference is in the ﬁrst period, when the early/gradual group has an average left vote almost ﬁve times smaller than that of the late/sudden group. However, these early differences seem to remain steady through time. In terms of patterns of development, the lateness, on the one hand, and the sudden change, on the other – far from being negative aspects for the electoral development of the left – seem to have set conditions that are more favorable, ceteris paribus, than very early enfranchisement and more intermediate and gradual developments. In these data, we can see supporting evidence for an argument that has already been advanced for social mobilization processes: It is not the earliness per se or the levels reached that create more or less favorable conditions for political mobilization, but rather the coincidence or lack of coincidence between such processes and the organizational developments of the socialist movement. The ﬁrst two groups of countries, those with early timing, had enlarged their suffrage completely or to a great extent before the organizational consolidation of a socialist movement, while in the other two groups the processes tended to be parallel. LEVELS OF THE ELECTORATE AND OF TURNOUT In Table 5.6 the relationship between the level of electorate and the level of support for the left is presented with ﬁrst-order correlation coefﬁcients. At the overall level, the association is positive and high at .678. However, the overall association throughout the century results from two distinct

ENFRANCHISEMENT

227

Table 5.5. Pattern of enfranchisement and left vote

Note: Number of elections given in parentheses.

Table 5.6. Electorate and turnout levels and size of the left vote

Note: Number of elections given in parentheses.

groups. One group can be roughly identiﬁed by a level of the electorate that is below 65% and that corresponds to the male or male and partfemale enfranchised population. In this group, the regression line is much steeper than in the whole set and the correlation is higher. The second group consists of the fully enfranchised electorates, where, not surprisingly, the association between the electorate and the left vote is weak or nonex-

228

THE POLITICAL MOBILIZATION OF THE LEFT

istent.24 The association is maximized in the period before World War I and drops considerably after that date. The minor increase between 1918– 1944 and the post–World War II period should not be regarded as significant, because a closer examination of the graphic form of the data shows that it is due to the outlying position of a few 1946–1967 Swiss elections and to the 1946 Belgian election held with an all-male franchise. These six elections deviate enormously from all the other 109 and therefore contribute to a slightly positive regression line. The link between turnout levels and left development is less strong (.419). More surprisingly, the association seems to go in the opposite direction than that with the electorate. It is absent in the ﬁrst phases of electoral mobilization and reaches a signiﬁcant level only after World War II. In conclusion, the level of the enfranchised electorate is a discriminating variable in the phase of mobilization but ceases to have any important effect once the electorate approaches the saturation level of full mobilization. In contrast, turnout levels appear to be irrelevant in non-fullymobilized electorates, while they gain some importance when they reach the state of full mobilization. By the 1920s, the level of the enfranchised electorate loses the strong predictive power of the 1880–1919 period, and by the 1940s the importance of turnout levels starts to grow. To sum up, in the ﬁrst phase, left mobilization rests mainly on enfranchisement; in the later phase, it depends increasingly on turnout levels. In these data, over-time developmental associations and cross-space synchronic associations are compatible. We can safely transform the development generalization according to which increases in the electorate bring about increases in the left vote (in the ﬁrst period) into the synchronic generalization that countries with a higher electorate tended to have a higher left vote. Indeed, the decade associations show correlation levels as high as those over time, particularly in the four ﬁrst decades between 1880 and 1920. The relationships do not result, therefore, from pure temporal covariation. To conﬁrm that this is indeed the case in each period, we can 24

The small number of Irish elections falls into this second group. The sharp drop in the correlation since the 1920s could be due to these elections for their weak left and large electorate. The decade associations computed without Irish elections are substantially higher (1920: .375 [40 elections]; 1930: .296 [29]; 1940: .419 [25]), but are still much smaller than those of the 1890s, 1900s, and 1910s. Moreover, if we want to exclude outliers, the Swiss elections between the 1930s and 1967 should also be left out. Given the nonenfranchisement of women, the Swiss electorate was roughly half that of all the other countries. The presence of these elections increases the positive association; without them, the association between the electorate and the total left vote even becomes negative from the 1940s on: ⫺.184, ⫺.025, and .087 (24). Therefore, the outlying cases of the Irish and Swiss elections compensate for each other to a large extent.

ENFRANCHISEMENT

229

regress each variable – the electoral and the left vote – independently against time and study the association among the residuals of such a regression. A signiﬁcant association remains among the residuals (r ⫽ .353 over 293 elections), and this supports the conclusion. Finally, the strong positive association in the ﬁrst four decades is important because, in the previous analysis, the ‘‘genetic’’ differences in left mobilization were unexplained by variables concerning the levels of social mobilization and cultural heterogeneity. Here, it is possible to ﬁnd the ﬁrst element of the genetic explanation. Early differences in the left vote were due in part to different patterns of enfranchisement. The latter, therefore, should be particularly useful in explaining early genetic crosscountry variations. INCREASES IN THE ELECTORATE AND TURNOUT Here, the question is whether the major jumps in the electorate corresponding to waves of enfranchisement or major increases in turnout levels were particularly important in determining the electoral growth of the left. A distinction needs to be made between male and female enfranchisement. In Table 5.7, I have synthesized the analysis, showing the level of the electorate, the turnout, and the left vote in the years before and after the main periods of male and female enfranchisement. In the last three columns of the table, I show the increase in the electorate, the turnout, and the left vote between these two elections.25 The data show that male enfranchisements were indeed crucial to the initial development of the left, as they are all associated with major jumps in its force in Belgium, Sweden, Finland, and the United Kingdom (and probably in Austria if data were available; in 1907 the Socialist Party recorded a level of support of 21%, while before this date it was judged as electorally unsuccessful by national historians). By contrast, the Netherlands and Norway show smaller increases in left electoral strength. In the Netherlands, this can be explained by the already mentioned fact that the 25

Unfortunately, electoral information is missing for Austria prior to 1907 and the introduction of universal male suffrage. For Ireland, electoral data are available, but the Irish Labor Party deliberately abstained in the ﬁrst elections during the transition to the Republic. I should also point out that in Finland and the United Kingdom, in 1907 and 1918 respectively, it is impossible to distinguish between male and female enfranchisement because the two were granted together. I have included them in the male table. Finally, the Norwegian case is also complex, as in 1912 only part of the female electorate was enfranchised (the process was completed in 1915) and the enlargement also concerned a section of the male adults.

230

THE POLITICAL MOBILIZATION OF THE LEFT

Table 5.7. Differentials in electorate, turnout, and left vote before and after male and female enfranchising elections

a) n.a.: electoral data not available; *: male suffrage reached before socialist party was founded; **: party already founded but not running at election; ***: male and female enlargement b) *: impossible to distinguish between male and female enfranchisement; **: party already founded but not running at election; ***: In Norway in 1909, only some women were enfranchised.

ﬁnal step to universal male suffrage was gradual. In Norway, the ﬁnal step to male franchise concerned an increase of 18% of the electorate. However, as mentioned before, this increase of the electorate was accompanied by a sharp drop in the levels of electoral participation (turnout dropped from

ENFRANCHISEMENT

231

85.3% to 55.9%), and this may explain why the Socialist Party gained so little (only 2.3%) from this enlargement.26 The most surprising case is therefore Italy, where male enfranchisement resulted in an increase of 17% of the electorate and only a modest 3.8% strengthening of the socialist forces, while the turnout remained at similar levels. The cases of female enfranchisement present a different picture. In four cases, they coincided with an electoral decline of the left (Belgium, 1946– 1949; the Netherlands, 1918–1922; Switzerland, 1967–1971; Denmark, 1913–1918), while in six there was an increase in the left vote. However, one should consider that in four of the six cases of growth, female enfranchisement coincided with a postwar election and party system restructuring (Austria, 1911–1919; Germany, 1912–1919; France, 1936–1945; Italy, 1921–1946). In these cases, the increases in the left vote cannot be linked to the new female vote. It is far more likely that they instead represent a generally favorable trend arising from the collapse of previous authoritarian or fascist regimes. In the two cases where female enfranchisement occurred under normal conditions, that is, between two consecutive elections that were not separated by long periods, wars, or resistance movements, the left grew (Sweden, 1920–1921; Norway 1906–1909). The conclusion is that female enfranchisement tended to have no impact or a negative impact on left size, because the cases of left growth are more attributable to factors that are independent of new female votes. This conclusion is in line with large survey data evidence that the ﬁrst female voting practices are more conservative than those of males. Note also that female enfranchisement was more often associated with drops in the level of electoral participation. If we generalize the argument for the major enfranchising changes in the electorate, we can check whether and when the increments in the electorate were associated with the increments of the left vote (not with its level). In general, this association is quite strong in the ﬁrst pre–World War I period (.408), declines in the interwar period (.370), and becomes null in the post–World War II period (.085). This association between increments is independent of the level of the electorate, as the partial correlation controlling for the electorate level shows (.414, .380, and .095, respectively). Conversely, the increases and decreases in the levels of participation are not associated with those of the left in any of the three crucial periods (and controlling for levels of turnout does not alter the picture signiﬁcantly). This suggests that the level of electoral participation should 26

It should also be remembered that Norwegian elections continued to be indirect in this period.

232

THE POLITICAL MOBILIZATION OF THE LEFT

be considered important in only some cases, but that this does not point to a general inﬂuence over the levels of left electoral mobilisation.

ROLE OF THE FRANCHISE IN THE MODEL Before turning to a discussion of the franchise in combination with the sociocultural model, let us comment on the theoretical role of this variable. As mentioned at the beginning of the chapter, the extension of the franchise plays a relatively ambiguous role in any study of the development of mass politics. In Figure 5.1 three models of interaction were presented in which the franchise was ﬁrst an intermediate, independent, nonexogenous variable; then an intervening conditional variable; and, ﬁnally, a dependent/independent variable with respect to the left vote. The association between enfranchisement and the left vote does not resolve the problem of which was the primum mobile and what type of relationship it was. In this section, I review the evidence supporting the various positions in regard to these problems. Let us start with the major issue: reciprocal causation. Was it the force of the left that captured the wider suffrage, or was it the wider suffrage that offered new chances to convert to the left? To test which variable comes ﬁrst, one can lag both of the two variables and control their level of association with each other. If the level of the electorate has a causal status with respect to left electoral size, it can be expected that lagging it will not modify the association. Conversely, if the strength of the left is the independent variable that explains increases in the electorate, lagging it should increase the level of association, as the association between the left vote and electorate at time t cannot be attributed to the left vote, given that the former is prior to the latter. In Table 5.8, the data concerning the ﬁrst and crucial period (1880–1917) show that while the lagged values of the level of the electorate maintain roughly the same level of association with the total left vote, the lagged values of the left make the association decline sharply from .703 to .428 to .382. In other words, if we want to explain the levels of the electorate at each election from the strength reached by the left in previous elections, we fare worse than if we do the opposite. The test is also revealing because resorting to an association measure requires exclusion of all those elections in which left parties had not yet competed, that is, all those numerous cases in which the electorate was extended before the left entered political competition; therefore, those

ENFRANCHISEMENT

233

Table 5.8. Electorate levels and left vote levels: association with lagged variables (1880–1917)

Note: Number of elections given in parentheses.

extensions deﬁnitely cannot be attributed to the strength of the left.27 Of the two possible directions of the association, that going from the level of the electorate to the strength of the left is much more powerful than that going from the level of the left support to the extension of the franchise. Having clariﬁed the direction of the association, we can now consider the role of the franchise as a conditional intervening variable that is able to boost the effect of social-mobilization variables. If this hypothesis is true, once we control for the level reached by the electorate, the association between social-mobilization indicators and left development should grow. One way of doing this is by removing the linear effect of electorate levels from both social-mobilization variables and the left vote. However, the ﬁgures in Table 5.9 show that partial correlations are lower than or equal to the zero-order correlation for both the whole period and the ﬁrst period. Once we control for the effect of the franchise levels, the social-mobilization and the cultural-heterogeneity models work worse rather than better. This enables us to reject the idea of considering the franchise as an external conditional variable deﬁning the size of the direct causal effect of the others. 27

When the test is repeated for each country, the results are consistent. Where universal male suffrage was achieved early and before political organization of the class left (France, Germany, Switzerland, Denmark), no doubt can exist about the direction of the relationship. For the remaining nine cases, in six the pattern is exactly the same as in the bulk of all the elections; in only three cases, a considerable drop in the association between the level of the electorate and lagged values of left strength is absent.

234

THE POLITICAL MOBILIZATION OF THE LEFT

Table 5.9. Association between social mobilization and left vote, controlling for electorate level

Note: Number of elections given in parentheses.

We are therefore left with the model that assumes that the franchise is an intermediate independent variable standing between social mobilization and cultural heterogeneity on the one hand, and the left vote, on the other. This model assumes that social mobilization and cultural heterogeneity inﬂuence the total vote of the left via the extension of the franchise. This view is consistent with the fact that by removing the impact of the franchise on the left vote – as was done in the previous test – we also remove a part of the impact of social mobilization and cultural heterogeneity on the left vote. For this type of interaction to be supported, we would expect a considerably positive impact of social mobilization on the franchise and a considerably negative impact of cultural heterogeneity on the franchise. In reality, controlling for cultural-heterogeneity levels, the association between social mobilization and the franchise increases to a zero-order .568 correlation and a .647 partial correlation. The same is true when we control the association between cultural heterogeneity and franchise for social mobilization levels: a ⫺.272 zero-order correlation and a ⫺.438 partial correlation. In conclusion, the model in Figure 5.1a seems to be the most suitable. The conditional-effect model did not work according to theoretical expectations, and the reciprocal-causation model is not supported by the general data set, although it remains plausible for some individual cases.

ENFRANCHISEMENT

235

ENFRANCHISEMENT PATTERN, SOCIAL MOBILIZATION, AND CULTURAL HETEROGENEITY In this chapter, we have seen that the aspects of the enfranchisement process that are most signiﬁcantly associated with the development of left electoral mobilization are the timing/tempo pattern and the level of the enfranchised electorate, while the increases in the electorate are signiﬁcant only for the increases in the left vote. Both the levels and the increases in turnout are signiﬁcant in some cases, but they do not show a systematic or consistent impact. A speciﬁcation concerning the role of turnout is, however, in order. A general additive or multiplicative model combining the electorate and the turnout in a single index of electoral mobilization is inadequate to support a general interpretation of left strength. There were cases in which the turnout remained constant at the same high or low levels (Belgium, the Netherlands, Italy, France, and Switzerland), cases in which it tended to grow parallel to the growth of the electorate (Austria, the United Kingdom), cases in which it grew within a constant electorate (Germany and Denmark), and at least two cases (Norway and in part Sweden) in which turnout declined while the electorate increased. It is therefore difﬁcult to identify a systematic effect of turnout. Electoral mobilization through enfranchisement and electoral mobilization through turnout increase have a mixed and varying inﬂuence in different cases. The only thing that can be done is to describe the situations in which the two play different roles through a close inspection of the early development of these variables (see Figure 5.2 ). If we were to maximize the impact of the general process of electoral participation development in each country, we would mix the two dimensions in different ways for each group of countries. We could leave aside turnout in certain countries, consider it as the most important variable in others, and, ﬁnally, combine electorate and turnout levels in others. For these reasons, I have decided not to include turnout in a general index of the enfranchisement process. The latter will therefore be composed only of the levels of the enfranchised electorate and the modality of franchise enlargement. To obtain such an index, I have added the standardized values of the enfranchised electorate to the standardized values of the timing-tempo patterns reported in Table 5.5. Each country has been attributed a rank-order from 1 (early/gradual) to 4 (late/sudden) (see Table 5.5). The index is the following:

236

THE POLITICAL MOBILIZATION OF THE LEFT

franchise levels ⫹ timing/tempo of enfranchisement ⫽ enfranchisement pattern combined index

Note that the levels of legal franchise and the timing/tempo classiﬁcation are not associated with each other (⫺.021 over 307 elections). While the franchise level represents an over-time variation across and within countries, the timing-tempo classiﬁcation represents cross-country differences in sequential patterns. The improvement of the combined index with respect to the individual variables is shown in the correlation of Table 5.10. While the role of franchise levels is important mainly in the ﬁrst period and tends to decline later on, the differences in the left vote associated with different timing/ tempo models of enlargement tend to persist over time. Combining the two indices into the enfranchisement pattern index, we obtain a more homogeneous association through time, and the total over-time association is also improved. When these data are plotted, no clear subset of outliers emerges. At this point, it is possible to incorporate the enfranchisement pattern within the general model together with social mobilization and cultural heterogeneity. The result of a stepwise regression of the total left vote against the three indices of social mobilization, cultural heterogeneity, and enfranchisement pattern (see Table 5.11) shows the considerable improvement achieved. Over the century as a whole, this model yields an explained variance of about 70%, each of the three variables maintains a strong impact, and in each of the four periods the model is consistent. The variance explained is now considerable in the ﬁrst period as well: 56% before World War I, 62% between the wars, and 54% and 71% in the two post–World War II phases. The relative weights of the different variables change in each period. The role of the enfranchisement pattern is particularly strong before World War I and decreases thereafter, to the advantage of the other two (so much so that it does not enter a stepwise regression equation with standard entry and removal). The weight of cultural heterogeneity is increased constantly from the ﬁrst period on, while the impact of social mobilization is at its maximum starting from the interwar period. On the whole, the interaction among the three variables guarantees a relatively constant over-time level of explained variance, thanks to their differential impact in each period. Moreover, with this model, one no longer faces the contradiction between over-time and across-space associations. Once the interaction between cultural heterogeneity, social mobilization, and the enfranchisement pattern

ENFRANCHISEMENT

237

Table 5.10. Enfranchisement pattern: individual and combined index association with the left vote

Table 5.11. General determinants of the left vote

is taken into consideration, each of the three variables maintains the same sign of inﬂuence and a signiﬁcant impact in each period as well as across all of them. Given the type of interaction that emerges among the three factors, we can chart cases using a simple additive index. In Figure 5.3, such an index has been computed and plotted against the left vote. Its predictive capacity is almost the same as that of the regression model (.694 versus .709), which conﬁrms how well the three processes combine in a simple additive manner. In addition, the scatterplot of Figure 5.3 also shows the extent to which the combined index reduces the presence of clear-cut outliers. No country or country–period stands out as a clearly exceptional case; even the Irish elections no longer stand completely alone. At this stage, the bulk of outlying elections in the lower-left part of the plot and below the regression line represent not only Irish elections, but also the early elections in the northern countries of Sweden, Norway, and the

238

THE POLITICAL MOBILIZATION OF THE LEFT

United Kingdom, whose lefts are far below the levels that the model would predict. What is deviating in Ireland is not its early low levels, in line with the general model predictions, but its lack of development over time. This model can be used to evaluate each country ﬁt, comparing the Europewide predicted left vote with the real national left vote. Applying to each country the general Europewide model, it is possible to evaluate the extent to which each deviates from it. The higher the residuals, the less well the general model performs in ﬁtting the actual national electoral results. It should be emphasized once more that the model is drawn from an all-country analysis, and the country residuals are not the unexplained variance in left vote as predicted by national ﬁgures of urbanization, industrialization, and enfranchisement, but rather the unexplained variance with respect to European-based predictions. The analysis of the residuals of national regressions identiﬁes a few experiences as particularly deviating from the general European pattern. The Irish case is no surprise. On the contrary, it is now surprising to see that in the ﬁrst period, the countries that represent a wider divergence from the model are the Northern European countries of Great Britain, Sweden, and Norway. In the interwar and immediate post–World War II periods, there are fewer marked deviations from the predicted values. In the ﬁnal period, the cases that stand out for their unexplained variance are Italy, Finland, and Belgium. Note that at this point, with the enfranchisement pattern inserted within the general model, the Finnish left vote no longer appears as deviating. The countries that tend to be best explained by the Europewide model are Austria, the Netherlands, and Switzerland, and to these one should add Belgium except for its last post-1965 phase. The signs of these residuals point to a left that is weaker or stronger than one would have expected on the basis of the general European pattern. What is more important, however, is that in the following chapters, when discussing additional aspects, we should keep in mind that as far as the electoral size of the left is concerned, some countries will require the identiﬁcation of additional potential sources of variation more than others. In conclusion, the interaction between the processes of social mobilization associated with the patterns of industrialization and urbanization, the early levels and models of enfranchisement, and, ﬁnally, the crosscountry variation in the extent of cultural homogeneity now constitute a satisfactory general model for the interpretation of the basic differences in the electoral development of the class left across Europe, both at the overall developmental level and at the synchronic cross-country level. With respect to this general Europewide model, more or less important residuals that

ENFRANCHISEMENT

239

Figure 5.3. Left vote and the combined index of social mobilization, cultural heterogeneity, and enfranchisement pattern. need to be explained by resorting to aspects other than those analyzed so far exist in some country–periods. In the following chapters, more directly political aspects will be invoked to account for such residuals. Moreover, the general model described so far will be recast and applied to the other aspects of the dimensions of the left that pertain to its organizational development and cohesion and to its more or less radical orientation.

6

ORGANIZATIONAL STRUCTURING AND MEMBERSHIP MOBILIZATION n this chapter, attention shifts from the structural conditions of cleavage structuring to the organizational and behavioral dimension. The central questions concern the relationship between organizational development and electoral mobilization; between corporate and political socialism; and between forms of interest representation and political representation. These themes are discussed along the two dimensions of organizational consolidation and membership mobilization. The former refers to the establishment and consolidation of speciﬁc political organizations into the corporate-group and the electoral-party channels, and to the linkage established between them; the latter refers to the capacity of the same organizations to mobilize individuals into such groups as trade union members, party members, and voters. This perspective is important for three reasons. First, the levels and patterns of early organizational consolidation and structuring of the cleavage are instrumental in explaining levels and patterns of political mobilization. That is, organizational strength can be translated into both capacity for electoral mobilization and stability of electoral mobilization. Second, once a political organization achieves a certain amount of consolidation, it acquires relative autonomy from the environment, becomes an agent of mobilization of its own, and shapes its environment through ideological and organizational encapsulation and mobilization. In this sense, organization is a way of making the cleavage relatively independent of its sociostructural bases and of strengthening its own ideological and cultural distinctiveness at the same time. Finally, early organizational patterns and the relationship they establish between the party and the environment of

I

240

STRUCTURING AND MOBILIZATION

241

corporate interests determine the latitude of action of the party’s electoral strategies. The chapter has three main sections. First, I provide a general framework for the study of the relationship between organizational and electoral mobilization and review the comparative experience of early organizational structuring of working-class movements. Second, I describe and discuss national patterns of membership mobilization in the party and trade union organization domain. Finally, I investigate how these factors relate to different histories of the electoral left, and I determine different types of class cleavages from the organizational/electoral linkage point of view.

EARLY ORGANIZATIONAL CONSOLIDATION PARTIES AND GROUPS IN THE FIRST POLITICAL MOBILIZATION A crucial aspect of the early organizational consolidation of every political movement is the linkage between the various arenas and types of mobilization bases. Such linkage varies considerably according to the type of interests and/or identities being mobilized, but also according to the general character of the political system environment of the claimant. Table 6.1 systematizes some possible forms of this linkage,1 separating the two channels – electoral and corporate – of mobilization. Widespread networks may be the only organizational base present in both of them – professional, craft, and other types of organizations, on the one hand, and candidates, caucus, and club-type organizations, on the other. At the other extreme, one can ﬁnd formal membership organizations in both channels. Electoral and corporate mobilization can, of course, take place without necessarily being dependent on central forms of coordination. Only party membership mobilization is necessarily dependent on the consolidation of a central political organization. The historical experience of Western European parties reveals various modalities.2 Voters can be mobilized without building on distinctive membership organizations in either the corporate or the political channel, relying instead on widespread networks of professional or other types of 1

2

The scheme is inspired by a short article in which Rokkan (1977) responded to criticism of his overlooking of the interest group formation process. I have modiﬁed it to suit my purpose here. The critique was from Alford and Friedland (1974). Important analytical considerations concerning the process of party formation are found in Nedelmann (1975), Sva˚sand (1978), and Panebianco (1982).

242

THE POLITICAL MOBILIZATION OF THE LEFT

Table 6.1. Type of interlinkage between electoral and corporate mobilization

organizations. This is usually the case in countries where enfranchisement was early and extended before any decisive waves of urbanization and industrialization, thus predating organizational developments. France during the Second Empire and Third Republic and the United States are examples of this pattern. Where this was the case, the forms of representation tended to be territorial in nature and not easily converted into a cross-local organizational structure. Indeed, the formation of strong corporate and political mass organizations proved very difﬁcult later on. A second model is the early foundation of speciﬁc organizations in the electoral channel not matched by and not relying on the support of speciﬁc organizations in the corporate channels. This classic model of British liberalism and conservatism proceeds from the electoral caucus to local branches, from registration activities to electioneering to membership and organizational development. This model does not tend to organize support along a single and clearly identiﬁable cleavage line, but rather across many lines, through a variety of alliances, although none of these are given visible expression in cleavage-speciﬁc organizations allied to the party. This modality can be associated with parties that developed in liberalized, but not electorally inclusive, contexts. ‘‘Indirect’’ parties3 are the end result of the political mobilization originally taking place through diffuse electoral networks and are the manifestation of speciﬁc corporate organizations that preceded the formation of the political organization. The support provided by the corporate organization to the electoral network of candidates and political associations is contingent at ﬁrst and later progressively interlocking, and it was typical of most of the denominational and agrarian parties, as well as of 3

This terminology was introduced by Duverger (1967).

STRUCTURING AND MOBILIZATION

243

some socialist ones. In this case, mobilization tended to be along a single and easily recognizable cleavage line, deﬁned by the preexisting corporate organization. A further model is that of the relatively parallel development of speciﬁc organizations in both domains, electoral and corporate. The relationship is one of mutual support and reinforcement; representation is more functional than territorial and takes place along well-established cleavage lines. The relationship between the forms of mobilization in the two channels is represented on the right side of Table 6.1. As can be seen, the crosslinkage between corporate and electoral organizations may be absent or very weak. Alternatively, it may be of a demand group type when organizations in both channels retain full freedom to bargain for support. In this type, they relate to each other in the classic terms of the pluralist model of democracy – each being a ‘‘client’’ of the other in its respective ﬁeld of action, exchanging electoral for policy support. The relationship may be contingent and thus expressed through alliances and coalitions based on interest proximity and goal similarity, but deprived of solid organizational, ideological, and personnel interpenetration. Although closely allied, the two organizations preserve distinct identities and independent latitude of action. The cross-linkage may be interlocking, that is, it may be characterized by a profound interpenetrating of corporate and electoral organizations that reinforce each other on the basis of leadership and membership overlap and interchange, support-base coincidence, and a range of common collective activities. Finally, the relationship may be one of dependency, in which one of the two forms of central organization tends to direct the second both politically and organizationally. The mode of representation does not depend on the organizational base or the cross-linkage, but instead on the extent to which a one-to-one monopolistic relationship of representation is present in both channels. The fragmented mode points to a situation of both corporate and electoral fragmentation. In neither the corporate channel nor the electoral channel does a single or dominant organization exist that can claim a monopoly of the representation of the interests to which it refers. Instead, the link between the interest sector and electoral politics results in a plurality of actors all claiming a representational right. The fragmentation of both channels is often the consequence of mutually reinforcing drives and can be of different types: (1) segmental pluralism, that is, the organization of competing ‘‘pillars,’’ with a close interlocking of social movements, interest groups, educational and cultural agencies and activities, and political parties divided along culturally distinctive identities (linguistic, religious, or

244

THE POLITICAL MOBILIZATION OF THE LEFT

cultural-peripheral);4 (2) ideological fragmentation along the more classic left – right dimension of competition;5 and (3) territorial fragmentation. The representation mode labeled ‘‘corporate’’ requires, by deﬁnition, a form of interlocking linkage between the two channels. The corporate organization is the only, or the dominant, representative of a well-deﬁned interest population in the interest channel, and the electoral organization with which it is interlocked shares the same prerogatives in the electoralparliamentary channel. A monopoly of representation exists, therefore, in both the corporate and electoral channels. I will now use these analytical categories to characterize the relationship between corporate and electoral organizations within the class cleavage.

EARLY ORGANIZATIONAL STRUCTURING OF CORPORATE AND ELECTORAL SOCIALISM Similar to agrarian parties, socialist parties were, in general, the extension of or parallel to organizations that were already active in the corporate channel. The tendency was to start from an economy-based response and to move toward the multiplication of speciﬁc cultural and political organizations built around economy-oriented identities in both the corporate and political channels. If one takes as a point of reference the moment of the formation of a nationally centralized organization in both channels – that is, the founding of national parties and central national confederations of trade unions – important differences emerge between national contexts concerning opportunities, difﬁculties, and choices. The section that follows is a review of the origins of socialist organizations focusing exclusively on the relationship between consolidation in the economic and political spheres of action in the pre–World War I period. The historical information is framed within the conceptual scheme summarized in Table 6.1. A comparative appreciation should pay attention to (1) the organizational sequence of the process, that is, which of the two organizations centralized ﬁrst, whether it be in the political or the economic sector; (2) the timing of the process, that is, whether, in a comparative European perspective, centralization in the political and economic spheres was early or late; and (3) the length of the process, that is, whether the efforts at national centralization were long and difﬁcult or more rapid. This description, based on threshold points, needs to be complemented by 4 5

Lorwin (1971: 147–187). See Sani and Sartori (1983: 307–340).

STRUCTURING AND MOBILIZATION

245

a discussion of the nature of the relationship between organizations active in corporate and electoral/parliamentary channels, that is, (4) the corporate–electoral cross-linkage and (5) the mode of representation (see Table 6.1). The combination of these ﬁve dimensions offers a comparatively perceptive picture of the modalities of organizational consolidation of early European socialist movements. Organizational Sequence Raw information concerning organizational consolidation is synthesised in Table 6.2. For each country, a date is suggested that refers to the ﬁrst push toward the national centralization in trade union and party/ political movements. The dates corresponding to the successful formation of a national trade union confederation and socialist party are reported in different columns. When possible, the table indicates in parentheses the length of the process of organizational central consolidation in both channels. In the majority of cases, the centralization of political organizations took place before a similar process was accomplished in the corporate channel: Denmark6 and Belgium7 are the clearest cases of this, followed by less important delays in the Netherlands,8 Italy,9 Norway,10 and Sweden.11 In Germany,12 Austria,13 Finland,14 and France,15 the processes were paral6

7

8

9

10 11 12

13

14

15

For Denmark, see Galenson (1952b: 104–172), and Foverskov and Johansen (1978). Historians have underlined the importance and the persistent role of craft unionism as a distinctive feature of the Danish labor movement. Its strength is probably due to the slow industrialization of the country. This helped Danish guilds – among the strongest in Europe, and whose privileges were not eroded until 1862 – to transform slowly into unions. By 1895 almost all the former guilds had adopted the trade union organizational format. See Galenson (1952a: 10–32). My sources for the early organizational history of the Belgian socialist movement are Spitaels (1974: 9–21) and Delsinne (1936). On the Netherlands see Daalder (1966b), Kossmann (1978: 344–349) and Buiting (1990: 73). On the party–unions relationship in Italy see Horowitz (1976), Pepe (1971), and Pappalardo (1989). The last author reviews the historians’ debate on the issue. See Galenson (1949: 7–77), Korpi (1983: 107), and Millen (1963). For Sweden see Galenson (1952b: 104–172) and Carlson (1969: 14–30, 46–50). For the German party–union early relationships, I have relied on Schorske (1955) and the contributions in Mommsen and Husung (1985). On the close contacts with the political party, see Moses (1982: 67–86). Sources for the early history of Austrian socialism are Steiner (1964), Knopp (1980), and Klenner (1951). For the early Finnish labor movement, see Suviranta (1987: para. 275) and Berglund and Lindstro¨m (1978: 35–36). On the French early organizational development, see Zeldin (1979: 210–13, 243), Duveau (1946: 212, 217–231), Lorwin (1954), Goguel (1950: 271–359), and Huard (1996: 268–289).

246

THE POLITICAL MOBILIZATION OF THE LEFT

Table 6.2. Socialist parties and trade union formation

lel, while in Switzerland16 and particularly in Great Britain17 and Ireland,18 centralization in the trade unions sector preceded that in the political sector. This is paradoxical: There can be no doubt that in most Western European countries, worker mobilization in the corporate channel largely precedes political mobilization in the electoral one. Nevertheless, in almost all cases, the effective organizational centralization of a national trade union confederation occurred after a political party was set up at the national level. According to one line of argument, when industrial revolution preceded the extension of the parliamentary franchise to the working class, the unions organized themselves in the labor market before seeking political representation through political organizations (the British case). By contrast, on the Continent, where unionization took place after the extension of the franchise to the working class, unions tended to be the result of party activity and were more dependent on and subordinated to the political organizations.19 This interpretation dramatizes the Britain–Continent opposition and reduces excessively the variation among continental 16

17 18

19

My sources for the Swiss early socialist movement are Balthasar, Gruner, and Hirtner (1988), Garbani and Schmid (1980), Reymond-Sauvain (1966), Gass (1988), and Vuilleumier (1988). Pelling (1965) and (1976a). For Ireland, I have used Roche (1990), Berresford Ellis (1972: 167–183), Chubb (1974), Mitchell (1974), and Gallagher (1982). For this typical Britain-versus-the-Continent interpretation, see Kahn-Freund (1972), in particular the Introduction. On a similar line, broadening the opposition to an AngloSaxon (including the United States and the former British colonies) versus continental European interpretation, see also Commons (1962: 681–685).

STRUCTURING AND MOBILIZATION

247

countries. As we have seen, the British pattern of franchise extension is not very different from some of the continental patterns. A ‘‘corporate-ﬁrst model’’ is also evident in Ireland, where the precedence of the Industrial Revolution over extension of the franchise is difﬁcult to argue, although in this case, the organizational imitation of neighboring Great Britain could be the explanation. However, in Switzerland, corporate centralization also preceded party centralization, despite the fact that extension of the suffrage to the working class occurred very early. The following are also probably important factors: (1) the need for centralization was felt more clearly in the political sphere; (2) the crosscountry transmission and spread of organizational-political forms was easier in the political sphere; and (3) there were important intrinsic differences in the opportunities for mobilization of the same constituency in different channels. In their original efforts to politically mobilize previously inactive citizens, socialist ‘‘agitators’’ and parties had to motivate individuals to commit themselves emotionally to abstract ideas, abstract categories of individuals, and interpersonal relationships. A feeling of solidarity had to be created among individuals who were physically and socially distant. Individuals’ traditional emotional ties to other, more direct primary or territorial groups thus had to be loosened. In every case, socialists had to induce their potential constituency to act in a way that contradicted the norms and roles dominant in their geographical-social milieus and, in general, against the authority of the social hierarchy at the local level. Compared with these difﬁculties of electoral and partisan mobilization, corporate mobilization enjoyed the advantage of focusing more on problems of the constituency in the workplace than on abstract ideas. The direct functional-economic interests and direct experiences of common destiny were more easily perceived than broader national political interests; the local factory ‘‘enemy’’ was closer to home than the ‘‘capitalist class’’ and the state. The perception that individual and collective fates could be inﬂuenced by remote political decision making demanded higher intellectual sophistication. Corporate recruitment and mobilization took place in a far more socially homogeneous environment than partisan and electoral activities.20 Therefore, it is not surprising that the ﬁrst mobilization of labor took place in workers’ societies, cooperatives, mutualities, and trade unions rather than in political associations and parties. However, when one considers the problems of national organizational coordination and consolidation – as opposed to that of early mobilization – 20

For the signiﬁcance of the homogeneity of the social context of participation, see the discussion in Chapter 3 and Rokkan (1970d).

248

THE POLITICAL MOBILIZATION OF THE LEFT

some of these advantages fade away and even become liabilities. There were additional difﬁculties for trade unions. In order to consolidate, national confederations – just like political parties – needed to establish a growing authority over and coordination of local organizations, that is, to overcome territorial fragmentation (chambers of commerce and similar early territorial forms of organizations) that, in several of the cases mentioned, delayed national organizational development. While for political parties this process was facilitated by the obvious centralization of power and decision making in national executives and parliaments, there was nothing similar to foster the central organization of the trade unions. Policy was, of course, decided at the center, but the immediate interests of members were more clearly dealt with at the local factory or branch level. As in the case of the party, it was necessary to add an element of higher solidarity to foster central federal and confederate development. It is not surprising that this stimulus was often produced by the political organization, which by necessity was more center-oriented in its actions. Moreover, early unionism was craft-based, uniting workers with higher qualiﬁcations, higher incomes, and generally greater resources for organizational participation. Such workers were part of a restricted and not very competitive labor market, and their self-identiﬁcation and status made them a distinct group with a higher tendency toward behavioral conformity. The process of sector integration of different craft traditions, qualiﬁcations, and professional ﬁgures and the organization of all employees of a given industry, leaving to one side the skills and tasks of the individual employee, was difﬁcult. Similarly difﬁcult was the integration of several sector unions into a single national confederation. In other words, status differences within the working class were less important and less damaging to the party than to trade unions, which were highly stratiﬁed along craft and professional lines. Within the political sphere, there was an ideological area of equality among members that was far more difﬁcult to achieve in a still highly segmented labor market.21 Finally, it is likely that the organizational imitation of earlycomers was easier in the political sphere than in the corporate one. The early formation of the party in Germany, for example, exerted an important inﬂuence. The strong inﬂuence of the German model and German-born socialist agitators in Denmark (and via Denmark to Norway and Sweden), as well as in Austria, the Netherlands, and Switzerland, is well documented. In addition, the role of the Second International was instrumental 21

On the crucial importance of ‘‘areas of equality’’ in fostering political participation, see Pizzorno (1966b).

STRUCTURING AND MOBILIZATION

249

almost everywhere on the Continent in fostering the formation of national parties. The most telling cases are probably the Scandinavian countries, where the development of socialist organizations was extremely early and rapid, considering the social background and conditions of these countries. In other countries leading the industrial transformation (Belgium and Switzerland), the tradition of working-class self-organization, mutual societies, cooperatives, unions, and so on was already quite strong, and the external political inﬂuence mixed with internal traditions to offer a picture of ideological syncretism and nonhomogeneity. In France, and to a lesser extent in Italy, German and Second International inﬂuences were met with considerable opposition from national organizational and ideological streams, which possibly delayed the political uniﬁcation of the socialist movement. The combination of the three factors discussed earlier may explain why, historically, the mobilization of unions predates forms of political centralization and, with few exceptions, ends up with their national coordination and organization after the parties and often under their auspices. Timing There is a small correspondence between the earliness or lateness of central organizational consolidation in the two channels. In the corporate channel, early national coordination characterizes the experiences of the United Kingdom, Germany, and Switzerland, in all of which such a process had already begun before the 1880s. At the other extreme, in Italy, the Netherlands, and France, national central organization was achieved in the ﬁrst decade of the twentieth century. All other cases fall in between, with the establishment of central national confederations taking place in the 1890s. The pattern is quite different for political organizations. Germany and Denmark had set up central national party headquarters by the 1870s; Austria, Belgium, Norway, Sweden, and Switzerland did so in the 1880s; Finland, Italy, and the Netherlands established theirs in the 1890s; and the latecomers, France, the United Kingdom, and Ireland, did so only in the ﬁrst quarter of the twentieth century. Germany and France are therefore homogeneous cases of very early and very late organizational consolidation, respectively, in both channels. The Irish experience approaches that of France; the Swiss experience, that of Germany. All the other countries fall between these extreme cases. While Italy and the Netherlands tend to approximate the pattern of France as latecomers in both channels, the United Kingdom stands out as the clearest case of total bifurcation in the development of the two channels: Organizational consolidation in the corporate channel was as early as the corresponding political

250

THE POLITICAL MOBILIZATION OF THE LEFT

development was late. Denmark is the clearest case of the reverse pattern: National party formation is astonishingly early compared with its relatively late trade union central organization. The earliness of the organizational consolidation of the socialist parties is related to the extension of the franchise or to the level of socioeconomic modernization of the country. The hypothesis is that the higher the levels of enfranchisement and social mobilization, the more differentiated the interest representation is in the political system, the greater the need for organization, the greater the potential interparty competition, and therefore, the earlier the development of mass parties. In a recent work, Sva˚sand has tried to test this thesis with qualitative information concerning the Scandinavian countries.22 He argues that in order to be supported, the organization of major mass parties should soon follow major extensions of suffrage. No clear relationship is evident in Table 6.3 among the earlycomers to large franchise (France, Denmark, Switzerland, Germany) and party formation, France being a clearly contradictory case. Among the latecomers to large suffrage (Finland, Italy, Sweden, the Netherlands) party formation was in general late (the 1890s: Italy, Finland, the Netherlands). Sweden had late enfranchisement and relatively early party formation (the 1880s). In the countries where socialist central consolidation was very late – France, the United Kingdom, and Ireland – enfranchisement was early or intermediate. On the whole, therefore, the relationship between level of enfranchisement and socialist party formation is weak. The hypothesis of the short time lag between male suffrage and party formation seems to ﬁt only the British and German cases; in Denmark, Switzerland, and especially France, the time lag is too large to be considered a valid case of conﬁrmation. In almost all other cases, like Belgium, the Netherlands, Finland, Italy, Ireland, and Austria, the parties appear to have formed before major suffrage enlargements occurred. Considering the level of the enfranchised electorate at the moment of the ﬁnal consolidation of nationally centralized socialist movements (the third column of Table 6.3), one can see that in seven of the thirteen cases socialist parties were founded with electorates that were enfranchised below the 15% threshold, which can hardly be viewed as a level capable of triggering the organizational consequences of the competitive drives supposed by the thesis. For example, considering the time lag between party formation and male suffrage (the ﬁfth column of Table 6.3) in Germany, the development of the parties and the franchise was almost parallel. In 22

Sva˚sand (1980). His conclusion is that the thesis does not hold for Sweden and Norway.

STRUCTURING AND MOBILIZATION

251

Table 6.3. Socialist party foundation, franchise, and social mobilization levels (cases rank-ordered according to the earliness of national party foundation)

Note: Cases of very early male suffrage extension are in bold. Cases of late extension are in italics. *1920s mean value.

Denmark, France, and Switzerland suffrage predated socialist parties by twenty-ﬁve to ﬁfty years. In all the other cases, socialist parties organized ﬁfteen to twenty years before universal male suffrage occurred. It is therefore impossible to argue that the formation of socialist parties was closely related to major enfranchisement. Finally, if one considers the combined index of social mobilization (sector and status transformations, working-class size, homogeneity, and concentration; cf. Chapter 3), it is clear that no relation whatsoever can be identiﬁed between such a ranking and the timing of party formation. Length The length of the process of organizational consolidation is the most difﬁcult factor to evaluate. Any answer depends on the deﬁnition of the starting point of political and corporate national consolidation. Three groups of countries can be safely identiﬁed as representing very distinctive experiences. The bulk of the northeastern Protestant countries (Denmark, Sweden, Norway, Finland) stand out for their short and apparently nonproblematic process of political centralization. Only six years elapsed in Denmark between the ﬁrst signiﬁcant political predecessor – the Danish Federation of the First International (1871) – and the founding of the national political party (1876). Norwegian socialism, which had stayed at the margins of the First

252

THE POLITICAL MOBILIZATION OF THE LEFT

International, founded a Social Democratic Association in 1883, which transformed itself into the Norwegian Workers’ Party (DNA, 1885) two years later. The Swedish Social Democratic Workers Party was founded in 1889, and it is difﬁcult to identify a predecessor.23 In Finland, the ﬁrst attempt to federate the workers associations, led by the philanthropist industrialist Victor Julius von Wright (1893), preceded by six years the foundation of the Finnish Workers Party (Turku Program, 1899). By contrast, political centralization proved painfully long and difﬁcult in continental Europe; particularly in Catholic countries like France, Belgium, and Austria but also, to a lesser extent, in religiously mixed countries such as the Netherlands, Switzerland, and Germany. The problems of political coordination were not the same. In Belgium and France, the almost twenty-ﬁve year period that elapsed between the ﬁrst forerunner political organizations and the establishment of identiﬁable national political parties was characterized by ideological and organizational divisions due to the differentiated initial ideological environment. The difﬁculties of political coordination of the Belgian socialist movement, notwithstanding early industrial development, have been attributed to several factors: (1) the early diffusion of socialist ideology came from political emigrants – Buonarroti, Proudhom, Rochefort, Marx, and the refugees of the Paris Commune, who gave Belgian socialism an anticlerical spirit, burning rationalism, and several ideological divisions; (2) in its formative period, the movement was divided between a Flemish part, where socialist cooperative associations predominated, and a Walloon part, where anarchic tendencies were strong and opposed to any spread in inﬂuence of the political party. In 1885–1890, political initiatives and strikes in Wallony were not controlled by the party but instead by anarchist militants. In France, the SFIO, founded in 1905, was a merger of at least ﬁve distinct formations;24 its formation had been delayed by the strong antagonism and competition

23

24

One might mention the Alma¨nna Svenska Arbetarefo¨reningen (ASA) founded earlier (1885) at Malmo¨ by August Palm, an e´migre´ tailor who, in Germany and Denmark, had come into contact with Lassallian socialism. Although the ASA was certainly the ﬁrst socialist association in Sweden, it is improper to regard it as the forerunner of the SAP; the following years did not see the slow process of political uniﬁcation of different organizations. The Workers French Party of Jules Guesde (Marxist); the Broussistes, from Paul Brousse, gradualist, reformist, vaguely syndicalist, and open to bourgeois collaboration; the Workers Socialist Party of Jean Allemane, a fundamentally ‘‘workerist’’ party devoted to the revolutionary general strike and to extreme forms of antimilitarism and anticlericalism; the Socialist Revolutionary Party founded by Louis Blanqui and Edouard Vaillant; the Independent Socialists of Millerand and Jaure`s.

STRUCTURING AND MOBILIZATION

253

among these different parties and political organizations25 and by the longstanding syndicalist hegemony over the workers’ vanguard.26 In Austria, the difﬁculties that hampered national political centralization between 1874 – when an Arbeiter-Bildungsverein was founded in Neudo¨rﬂ – and 1888–1889 – when the process was concluded, under the leadership of Victor Adler, with the formation of the Sozialdemokratische Arbeiterpartei in Oesterreich (SDAPO), were of two types. First of all, there was a strong repressive policy that made the party almost disappear between the middle of the 1870s and the end of the 1880s. Second, there was the serious internal organizational problem of appealing to several nationalities. Not only in the empire, but also in Vienna and all the regions where ethnic Germans were predominant, large sections of the working class were composed of Czech, Polish, and other Slavic groups, often with their own organizations. In the Netherlands between 1881 and 1894, the process leading to the formation of the Dutch Social Democratic Party (SDAP) was characterized by the conﬂict among socialist leagues. Those leagues that favorably viewed political-parliamentary action were defeated in the League Congress of 1893 and split when founding the SDAP the following year. Thus, Dutch political socialism emerged from a split in the socialist movement, and its dominant position within this movement did not occur until the beginning of the twentieth century, when other forms of socialism – in particular the syndicalist one – suffered important strategic setbacks. In Switzerland, the major problem was probably territorial fragmentation. The party formed in 1888, eight years after the central Trade Union Confederation and after two failed attempts to do so in 1870 and 1880. Swiss socialism was preoccupied with the issue of centralization against cantonalization, and it remained so throughout its entire history. Linguistic barriers, and its own early almost exclusively Swiss-German roots,27 25

26

27

‘‘Control by a socialist party might have been endurable. What was unendurable was the competition of rival socialist parties for control of the struggling young unions. The antisocialist current was strengthened when in 1890s a considerable number of anarchists went into the union’’; Lorwin (1954: 20–21). Malefakis (1974) has argued that the internal divisions and the syndicalist role in France should not be exaggerated. France should be assimilated more to the continental experience than to the Southern cases of Italy and Spain. He may be right in stressing more differences than similarities in the role of anarcho-syndicalism in France versus Italy and Spain. However, when the comparison is made between France versus Belgium and the Netherlands – the other two countries in which syndicalism played an important role in the early phase of the labor movement – the picture is reversed. Classiﬁcation of countries in paired comparisons may differ from classiﬁcation in a Europewide set. The German-speaking part of the country offered more favorable conditions for socialist mobilization: higher industrialization, political inﬂuence of e´migre´ German-socialist

254

THE POLITICAL MOBILIZATION OF THE LEFT

made it difﬁcult to centralize an organization whose cantonal sections were facing different party systems, alliance opportunities, and power structures. If, in comparison with other Swiss parties, its federal organizations were more important than its cantonal ones, their coordinating capacity and central authority were unimpressive compared with those of other European socialist parties. Notwithstanding all its efforts, the party was best described as a federation of local sections resting on the support of some large unions (in particular the iron metallurgy and building sectors). The German and Italian cases are more difﬁcult to characterize, as the length of the process of political centralization is less indicative of its difﬁculty. In Germany, a speciﬁc political organization (Sozialdemokratische Arbeiterpartei, SAP) was set up very early in 1875 as a result of the fusion of earlier organizations (in particular, Lassalle’s General German Workers’ Association [1863] and Bebel’s Union of German Workers’ Societies). The length of the process should be seen in the context of its remarkable earliness (see the earlier discussion of timing). No imitation advantage – like those that certainly played an important role in the Scandinavian countries – was available in this case. In contrast, the Italian experience resembles the French one in many ways in that the political organization founded in 1892–1893 was the result of difﬁculties in amalgamating several tendencies and excluding others. Anarchist, ‘‘workerist,’’ and radical-republican (in central Italy) tendencies were ﬁnally marginalized but did not disappear. The awareness of the need for a speciﬁc political organization won over strong resistance. At the beginning, the party (from 1895 on, the Italian Socialist Party) had four presidents to guarantee the autonomy and representation of its various ideological components. Despite this, political centralization was achieved faster in Italy than in France for reasons that are difﬁcult to determine. In all likelihood, the less distinctive ideological and organizational proﬁle of the Italian forerunner organizations, as compared with the French ones, made their amalgamation easier. It is difﬁcult even to clearly identify the beginning of a phase of national political centralization before 1892, although the Workers Italian Party that emerged in the 1880s should be mentioned.28 Finally, the experiences of Ireland and the United Kingdom represent a modality differing from those of both continental and Northern Europe. In these two cases, the period between the ﬁrst predecessor and the ﬁnal

28

militants and leaders, and absence of the inﬂuence of the historical Radical Party. See Vuilleumier (1988). The Socialist Revolutionary Party of Andrea Costa was strongly localized in Romagna, and, in addition to having its origins mainly in the peasantry and agricultural laborers, it incorporated a predominantly anarchist component.

STRUCTURING AND MOBILIZATION

255

party formation is extremely long. The ﬁrst Irish political organization with a socialist program was the Republican Socialist Party (ISRP), founded in 1896 by the legendary James Connolly, but no organizational continuity can be claimed between this group29 and the Irish Labour Party (ILP), whose foundation date is still a matter of debate among Irish historians. Some indicate the date of 1912 for an embryonic political organization; in 1914, however, the Irish Trade Union Confederation (ITUC) added ‘‘and Labour Party’’ to its name, indicating at least minimal differentiation between the two. Formally, the party obtained independent existence in 1922, but it was only in 1930 that the true separation between ITUC and ILP occurred. It is also hard to identify clear predecessors for the British Labour Party, founded ofﬁcially in 1906 but ﬁrst formed in 1900. A Marxist Social Democratic Federation had been founded in 1884, but it had remained extremely weak. The Independent Labour Party was founded in 1893; this was less ideologically orthodox than the latter, and it attracted larger numbers of members. Fabian Societies had existed, of course, for many years. All these forces ﬁnally joined together, but it is clear that it was the support of the Trade Union Confederation (TUC) that created the crucial momentum for founding the party.30 Indeed, strictly speaking, the Labour Representation Committee created in 1900, mentioned earlier, was a coalition of organizations and not an institution in its own right. Real organizational separation can be dated only to 1919, when the party began to accept individual members. In both the Irish and British cases, therefore, the process of party formation spread out over time, but the leading position of the trade unions makes this delayed formation different from that of the continental countries. The four groups of countries discussed in relation to political centralization (northeastern, northwestern, continental Catholic, and mixed) are not so neat when one considers the length of the process of corporate centralization. Certainly in most cases, there is a correspondence between the difﬁculties of political and corporate centralization; and as the reasons are frequently the same in both cases, no separate discussion is necessary. However, it is interesting to emphasize that the processes of corporate centralization of the three Scandinavian countries are quite long and sharply contrast with their rapid political centralization. There might be special national reasons for this. In the Danish case, for instance, historians 29

30

Which remained extremely weak and was shattered by the obscure story of its leader’s emigration to the United States in 1903. Pelling (1965: 205).

256

THE POLITICAL MOBILIZATION OF THE LEFT

point to the ongoing presence of craft unionism as a particularly strong impediment to national confederation development. In Sweden and Norway, however, this was not the problem. Rather, the relationship between the party and the trade unions was so close that the party simply operated for a long time as the coordinating body of local and sectoral unions. Between 1885 and 1899, for instance, the Executive Committee of the Norwegian Labor Party functioned as a confederal chamber of the local unions’ branches (samorganisasjon), corresponding to the continental chambers of labor. In these two cases at least, the length of the centralization process in the corporate channel simply indicates that this central coordinating function was taken over by and performed within the political party without provoking any signiﬁcant tension in the socialist movement. Finally, in only three cases is the time lag between the ﬁrst attempt at national coordination and the setting up of a central confederation very short: Germany, Switzerland, and the United Kingdom. Cross-linkage It comes as no surprise that the prevalent cross-linkage between socialist organizations in different channels was interlocking, with overlapping ancillary organizations, leadership, and activities. Even if this was not the case initially, it tended to become so over time. One already familiar example of this is the early organizational, leadership, and activity interpenetration of the Scandinavian cases. Although Austria is similar in its end results, the circumstances in which this took place are not. Even though Austro-Hungarian socialism was ultimately divided by the disintegrating forces of nationalism, the hostile environment and the extensive and protracted governmental repression resulted in a relationship between the party and the unions that was probably closer than anywhere else in continental Europe. Trade unions offered virtually unqualiﬁed support to the party; they recognized from the beginning the importance of political action and the role of the party in this ﬁeld. Unlike the German socialist movement, where unions sympathetic to socialism ﬁnally gained the upper hand in matters of tactics (strike decisions in particular) regardless of their initial dependence on the party, Austrian trade unions always developed their strategic decisions in close consultation with the political party.31 In Germany, Finland, Great Britain, and Ireland, the relationship between parties and unions was characterized by an initial phase of dependency – trade union dependency on party personnel, political, and ideological guidance in the ﬁrst two cases and party dependency in the others. 31

For the German–Austrian comparison on this point, see Steenson (1991: 196–213).

STRUCTURING AND MOBILIZATION

257

This situation evolved over time in the classic interlocking pattern. However, the idea of ‘‘dependency’’ needs to be qualiﬁed here. Germany is regarded as a clear-cut case of a party-led socialist movement. The trade union movement was initially fostered by socialist political agitators who saw them ‘‘primarily as recruiting agents for the political labour movement.’’32 Once the early unions’ national center had centralized to some extent, the movement kept it in a relatively dependant position; this was also due to Bismarck’s antisocialist laws (Sozialistengesetz), which severely circumscribed the unions’ activity and pushed them to seek political allies. Such dependency did not last, however. The initial silence of the trade unions with regard to the party theoretical debate between reformists and revolutionaries disappeared, and increasingly they cast the decisive vote. Thus, until 1918, the relationship between the two was not one of subordination or equality in the classic sense of a division of labor and independent spheres of competence. Rather, it was a form of convergence dictated by common goals and forms of struggle imposed by the authoritarian system, by the extremely compact working-class culture, and by a basic ideological orientation. All these worked to render the goals, tactics, and strategies of both kinds of organizations inseparable. In this situation, however, it is easy to see the early imprint of political over corporate socialism. Germany is not the best example of party predominance over the unions; Finland is a more clear-cut case. Here, the dependence of the unions on the party was difﬁcult to overcome because not only had the local as well as central unions been fostered by party political activism, but also because they had remained extremely weak in organizational and membership term for many years. Sufﬁce it to say that still in 1914, while the party enjoyed massive electoral support and claimed 70,000 members, the trade unions claimed only around 30,000. In Ireland and the United Kingdom dependence of the party on the trade unions is usually found. The Irish case, however, should be differentiated from the British one, not only because this dependence lasted for a longer period of time and was not overcome until World War II, but also because it was different in nature. Moves to broaden the membership of the Irish Labor Party beyond the collectively afﬁliated unions were regularly defeated throughout the 1920s.33 Until at least the 1930s, the party 32

33

Schorske (1955: 9). Lassallistes were particularly active in this job since their founding of the ﬁrst German trade union in the cigar-makers sector (1865). Mitchell (1974: 85). It is interesting to recall that the 1914 congress was worried and concerned ‘‘to prevent politicians from worming their way into the new party’’ and ordained that any Labor candidate for public body must be a trade union member.

258

THE POLITICAL MOBILIZATION OF THE LEFT

remained trade unionist in personnel and mentality; after 1930, it remained so in spirit.34 In Britain, the early relationship between the party and the trade unions is not one of pure party dependency. The British Labour Party soon developed into an unusual coalition of forces similar in some ways to the early Belgian Socialist Party. Trade unions played a predominant role through collective afﬁliation, but they never ‘‘colonized’’ the party in an Irish type of relationship. Various circumstances and features of British trade unionism may explain this: its sense of force, autonomy, and self-awareness; its strong concern with organizational integrity; its attention to the priority of industrial issues; and, ﬁnally, a certain sensitivity to party viability and a sense of constitutionality. All this always left a wide range of maneuver to the Labour Party. In many ways, although types of relationships were the same, the difference between this and the Norwegian and Swedish experiences lies in the greater division of labor and in weaker levels of interpenetration. In contrast to these four cases, even the Belgian socialist movement has sometimes been described as being characterized by early trade union dependence on the political movement. The foundation of the General Confederation of Belgian Workers (CGTB) resulted from party activities. In 1898, at the bidding of the party’s twenty original founding local unions, the party had studied the opportunity of forming a Central Trade Union Committee (Commission Syndicale), in fact, the embryo of the CGTB. The committee was appointed by the party executive; it had no independent policy line even if, in principle, it was open to unions of any political tendency. Only in 1907 did the committee obtain its own statutes and congresses. Some experts have claimed that the party brought the unions under its direct control from the beginning. This is correct if one looks at the formal relationship; however, the party continued to draw a large part of its strength from the extraordinary (and absolutely exceptional in European comparative terms) development of the cooperative and mutuality movement, whose organizations were among the elements constituting the party itself. While the party was fostering union centralization, other forms of corporate organization were feeding the party. This original pattern certainly underlines the difﬁculty of the Belgian unions in organizing autonomously at the center without external coordinating help. Even so, one hesitates to describe the pattern as being party dominated.35 34

35

Gallagher (1982: 4). Until World War II, candidates and party ﬁgures were almost exclusively trade unionists. Their explicitly indicated function was to defend trade unions’ interests in the Dail. Contrary to most other European experiences, the ILP suffered trade union conﬂicts and internal strife more than the reverse. Belgian historians argue that at the national level the development of unionism and

STRUCTURING AND MOBILIZATION

259

In the French, Italian, and Swiss cases, I regard the cross-linkage between corporate and electoral organizations as predominantly contingent. In these countries, such cross-linkage was organizationally weaker, less stable, and more, conﬂictual, and had a less clear division of labor. The French situation is the purest from this point of view, and it can be seen in the number of occasions on which political and union socialism have followed conﬂicting lines on which the political party was obliged to deal with autonomous and previously uncoordinated actions, together with the continuously asserted independence of unions vis-a`-vis politics. In this period the French trade union movement was sometimes tempted to become a pure ‘‘demand-group’’ vis-a`-vis the party. In Italy, the Socialist Party remained extremely weak throughout the 1890s and was unable to exercise a dominant role. The relationship with the unions is more akin to the French type than to the German one – close but problematic and without any clear predominance of one over the other. Confusion about the roles of the two organizations persisted for many years, and while the division of labor never prevented the reformist leadership of the trade unions from taking speciﬁc political action, sectors of the union movement debated at length the suitability of founding a political organization to be the direct expression of organized labor.36 As in France, the politicization of the trade union movement was probably the consequence of the internal ideological and organizational fragmentation of the entire socialist movement. The Swiss case is very different. Here, the organizational weakness and the poor centralization of both organizations always meant that differences between party and trade unions were maintained over the national territory while coordination between the two remained very poor, making impossible and nonviable any dependence or strictly interlocking relationships. Mode of Representation The nature of the cross-linkage between political and union central organizations needs to be placed within the context of the broader system of interest representation in the polity. The corporate and fragmented modes of representation point to the presence or absence of a one-to-one relationship in the representation of both channels (see Table 6.1 and its discussion). A socialist-inspired trade union organization can be strongly interlocked with the socialist party, but be unable to claim full representation of the organized labor movement. On the political side, ideological

36

union activities was hindered by an extremely widespread cooperative movement; Spitaels (1974: 9–21). The Labor Party dreamed of by Rigola; see Cartiglia (1976).

260

THE POLITICAL MOBILIZATION OF THE LEFT

as well as segmental or territorial fragmentation can give rise to organizations competing for political representation of labor. In the corporate channel, this issue can be operationalized in terms of how much of tradeunionized dependent labor is actually represented by the main socialist confederation and, second and less important, whether such a confederation is ideologically divided internally. In the political channel, the issue is whether the political representation of labor is claimed by a single united political party or by a variety of political formations. Clear examples of a single-organization monopoly of representation in corporate and political channels are those of Sweden, Norway, Denmark, and the United Kingdom. A mode-disputable case is Austria, where both a Catholic-inspired party and a Union Center were founded early on, in direct competition with the socialists, while no important organizational fragmentation within the left occurred. However, superposition of the class and religious cleavage, rather than their cross-cutting, prevailed, with the result that the class alignment tended to coincide with the denominational one to a far greater extent than in other countries with Catholic political and interest organizations. It should also be emphasized that in their formative phase the trade unions had no ideological competition in organization (as mentioned, they faced national splits); thus, nearly all organized workers belonged to unions with close ties to the Socialist Party until the end of World War II. The Austrian experience can, as a result, be assimilated more to the corporate mode of representation than to the fragmented one. In Italy and the Netherlands, socialist political fragmentation was endemic even before World War I. Eventually, the ﬁrmly established link between political and corporate socialism ended a long process of consolidation during which both organizations had ﬁnally defeated their left and/ or right wings. At the same time, however, they had seen the parallel development of competing Catholic and Protestant labor organizations,37 which hindered the development of socialist trade unions and fragmented the representation of dependent labor. In Germany, socialist political fragmentation started during World War I, and political and corporate fragmentation was increased by signiﬁcant communist splits in France, Italy, Finland, and Germany after World War II.38 Religious union mobilization 37

38

The Christian National Federation of Crafts (Protestant) and the Organizational Ofﬁce of Crafts (Catholic) were founded at about the same time as the Dutch Federation of Trade Unions. Also, the Norwegian socialist movement was highly radicalized during the First World War and was strongly affected by the Third International split. However, this resulted in minor short-term fragmentation. The Socialist Party moved in a majoritarian and

STRUCTURING AND MOBILIZATION

261

produced the lack of corporate representation monopoly in Belgium and Germany very early in the nineteenth century and in Switzerland at the beginning of the twentieth century.39 In most of these cases, fragmentation of representation in one channel was paralleled by fragmentation in the other. Finland provides the only case of a uniﬁed socialist union movement with two political parties espousing socialist and communist principles as a political reference point. The 1918 civil war left the unions’ confederation (SAJ) largely dominated by the communists until its ﬁnal dissolution by court order for subversive activities in 1930. The new Central Organization of Finnish Trade Unions (SAK), founded the same year, was under socialist control; but although it managed to maintain organizational unity until after World War II,40 internal strife was endemic and profound. To conclude, political representation was fragmented and highly polarized; corporate socialism, although formally uniﬁed organizationally, was marked by intense ideological strife. Table 6.4 summarizes the statuses of each country’s experience in the set of ﬁve properties of organizational consolidation selected here, distinguishing between corporate and electoral channels. On the basis of this comparative characterization, we can now turn to the central question of the relationship between organizational consolidation and membership mobilization.

MEMBERSHIP MOBILIZATION In this section, attention is turned from patterns of organizational consolidation to levels of membership mobilization. Data concerning the membership of parties, general trade unions, and socialist trade unions have been collected from a variety of sources documented in the Data Appendix. A comparative analysis based on row ﬁgures is obviously misleading for countries with very different total populations. Therefore, it is necessary to

39

40

orderly way to the Third International, and in an equally majoritarian and orderly way it came back to the Second three years later. Organizational and ideological fragmentation were in no way comparable to the other cases listed in the text. In 1906 (Basel Conference), when the socialist-sympathetic Trade Union Confederation (SGB-USS) decided to abandon its principle of political and religious neutrality (codiﬁed in 1902) in favor of a class-struggle platform. When in 1960 – like in France and Italy earlier – moderate Social Democrats split from the SAK and founded their own confederation, which resumed the old name of SAJ. They accused the communists and left-socialist members of the SAK of using the trade union movement for their political purposes. Only in 1969 did the new internal and international situation permit a reuniﬁcation of the labor movement.

262

THE POLITICAL MOBILIZATION OF THE LEFT

Table 6.4. Trade unions and parties: organizational consolidation (1880–1940)

use indices that standardize for the size factor, taking as the reference the total dependent labor force, the total vote for the party, and the total electorate. The indices used in the analysis are therefore the following:41 1. Partisan mobilization a. Membership of socialist parties as a percentage of the national electorate (socialist partisan density [SPD]) b. Membership of socialist parties as a percentage of the socialist vote (socialist voters’ encapsulation [SVE]) c. Membership of communist parties as a percentage of the national electorate: (communist partisan density [CPD]) 41

These indices have been computed on a year-to-year basis. In this chapter, they will be analyzed by decade averages and by election years. When corporate or partisan membership ﬁgures are referred to the election year, the actual election year ﬁgure is an average of the ﬁve years around the election year (two beforehand, the election year, and two afterward).

STRUCTURING AND MOBILIZATION

263

d. Membership of Communist parties as a percentage of the Communist vote (communist voters’ encapsulation [CVE]) e. Total cumulative membership of socialist and communist parties as a percentage of the electorate (left partisan density [LPD]) f. Total cumulative membership of socialist and communist parties as a percentage of the total left number of votes (left voters’ encapsulation [LVE]) 2. Corporate mobilization a. All trade union members as a percentage of the total dependent labor force (trade union density [TUD]) b. Left-oriented trade union members as a percentage of the total dependent labor force (left trade union density [LTUD]) c. Left-oriented trade union members as a percentage of the electorate (trade union electoral density [TUED]) The distinction between party organizational density and encapsulation points to different theoretical meanings of the different indices. When membership ﬁgures are computed as percentages of the general electorate, the organizational presence of the party partisan machine is measured within the global electorate. This measure can therefore be viewed as a reliable comparative indirect indicator of the organizational presence within the society. When membership ﬁgures are related to the party or party-family actual vote, the reference point varies widely from one party to another. What this index can actually indicate is not the absolute partisan density of the given party within the society, but the level of closeness between members and voters. This index, used widely by Duverger in his classic work on parties, is totally inadequate for cross-country or crossparty comparisons, although it may be of some use for cross-temporal comparison.42 PARTISAN MOBILIZATION43 The entire historical process by which personal experience and group identity were transformed into political choices ‘‘could not have got under 42

43

For this reason, the cross-country voters’ encapsulation is not discussed here. A graph with these data can be found in Bartolini (1983: 187). The levels and rates of corporate and partisan mobilization may depend on the more or less closed nature of the channel in which they operate. Political and corporate mobilization requires association and collective-action rights to be granted; freedom of association is therefore a decisive intervening variable between organizational consolidation and political mobilization. These variables are discussed in the next chapters.

264

THE POLITICAL MOBILIZATION OF THE LEFT

way without the organizational efforts of the parties and their militants, without their educational and promotional activities, and without their tireless drive to mobilize more and more of their clientele.’’44 Party members and activists in the pre–mass media era were the protagonists of the scramble for additional support throughout decades of hidden and obscure work on each side of the main cleavage lines. The fact that some socialist parties were preceded by other parties with a mass organization45 might have facilitated their development in that the legitimacy of this kind of organization had already been recognized. However, in most cases, socialist parties developed as outsiders in hostile environments, needing to mobilize resources in order to overcome the ostracism of the established elites. As such, the organization of a large number of followers was likely to be their answer to the problem of entering into the respective political systems. This led the organizational network to enforce a strong degree of formalization and role differentiation.46 Members were the most important, if not unique, ﬁnancial source for the parties and their press; they carried out the work of educating (not only politically) a working population with a low level of class awareness; they provided the party with inexpensive personnel for all kinds of necessary work; in many cases, they also provided medical care, legal assistance, and protection to workers and to groups of potential supporters of the party. Membership entailed militancy. It could be hypothesized that the electoral results of the European socialist parties during the period of their formation and consolidation depended on their increase in membership and their organizational growth. If one contrasts this early picture with the current situation, it seems that the traditional role of membership in social democratic parties has been severely damaged by the major changes in the post–World War II economy, social structure, and communications. The goals of a high level of involvement and of rank-and-ﬁle agitation are incompatible with the socialist parties’ massive participation in government. The growing importance of nationwide mass media as channels for political messages has had 44 45

46

Rokkan (1970e: 422). As in the Netherlands and Belgium, in Great Britain an organization that was truly a mass organization can be dated from the Chamberlain reorganization of the Liberal Party’s ‘‘Birmingham Caucus’’; see Ostrogorski (1970, Vol. 1: 161–203). Even the Conservatives, after much criticism of Chamberlain’s system, developed some of the features of an early mass popular party; Cornford (1964). Neumann (1956a) argues that the ‘‘total integration’’ at which communist parties were aiming was both a function of and a prerequisite for their nondemocratic nature. However, even if, in some of the Western communist parties (although not in all), quantity of membership was sacriﬁced to quality, the role that members played within the party was not that different.

STRUCTURING AND MOBILIZATION

265

far-reaching consequences for the communication processes of the parties and their organizations. The development of the welfare state has inevitably taken over most of the previously mentioned functions of education, legal aid, medical care, and so on. Finally, even the ﬁnancial function of the membership has been challenged by the parties’ involvement in business and commercial enterprises and, last but not least, by the introduction of ﬁnancial aid from public sources since the 1960s. Paradoxically, such postwar developments have led some experts to speak about a ‘‘contagion from the right’’ of the social democratic parties47 and to emphasize, regretfully, the process of growing uniformity among European parties toward a model implying the ‘‘downgrading of the role of individual members.’’48 We can now look at the historical data in the light of these opposing perceptions. Three European countries – the United Kingdom, Sweden, and Norway – have a socialist membership that includes individual cardcarrying members as well as collective afﬁliation through trade unions, and this creates problems of comparability that need to be kept in mind. In the British case, separate ﬁgures for individual members are available for 1928 on, and these data are used in the ﬁgures and computations. Therefore, British Labour partisan density and encapsulation are systematically underestimated to an extent that is difﬁcult to guess, corresponding to the proportion of collective members who would have been individual members were this the only way of joining the party. No separate ﬁgures for individual members of the Swedish and Norwegian Labour Parties are available, but in Sweden the collectively afﬁliated members are estimated to be about 70% of the total postwar membership and in Norway, about 40%.49 These partisan densities are therefore systematically overestimated by the unknown proportion of collective members who would not have joined the party on an individual basis. Nevertheless, the problem of the Scandinavian countries is less acute because, particularly in Norway, indirect membership is not channeled through nationwide federations. Rather, each local union makes a decision on afﬁliation by majority vote, and individual members can thus ‘‘contract out’’ fairly easily. For counting purposes, this kind of afﬁliation indicates a more serious political involvement in the party than is the case for the United Kingdom.50 In Figure 6.1, the levels of socialist party membership as a percentage 47 48 49 50

Epstein (1967: 257–261). Kirchheimer (1966b: 190). Elvander (1977). In ﬁnancial terms, this often does not mean very much, as trade unions support the party ﬁnancially from a common fund without specifying the names of contributors. See Rokkan and Valen (1962).

266

THE POLITICAL MOBILIZATION OF THE LEFT

Figure 6.1. Socialist party membership as a percentage of the electorate. of the total electorate (SPD) from 1890 to 1980 are charted. Between 1890 and 1919, in the context of the still very restricted suffrage, the countries show high levels of partisan density, which for many of them will never be matched later on. When the right to vote was extended to men between 1917 and 1920, the index shows a considerable drop for most countries. In the period between the two world wars, socialist partisan density begin to differentiate across countries. Most of them remain fairly stable at the level reached after World War I, while in the three Scandinavian countries and Austria, socialist partisan density continues to grow during the entire interwar period and reaches a level ranging from 9% to 13% of the electorate. This level is on average three times larger than that of the other countries. Between 1945 and 1965, two of the four high partisan density countries – Norway and Denmark – start a decline that is remarkable for its sharpness and consistency. Their index shows an impressive parabolic distribution over the twentieth century, peaking in the 1950s. This means

STRUCTURING AND MOBILIZATION

267

that the decline of socialist members in the electorate at large precedes by ten to twenty years the electoral crisis of these parties in the second half of the 1960s. By contrast, in Austria and Sweden, the organizational presence of the socialists continues to grow, both countries reaching an average level of 15 socialist members for every 100 electors. Austria recovers the exceptionally high level reached at the beginning of the 1930s, while Sweden continues to increase steadily. The particularly strong corporate model of Swedish society has allowed the Social Democratic Party to make up through organizational afﬁliation – which, incidentally, goes well beyond the traditional trade union sectors – what it was losing in individual membership.51 In Austria, however, membership was, and remained, strictly individual, and the party’s attempt to transform every voter into a full party member was linked to the lager model of societal development and the necessity within the consociational model to organize the respective subcultures fully.52 Among the remaining countries, only three are worth commenting on. In the 1945–1965 period, Switzerland maintained a higher level of partisan density than the other countries simply because its electorate was smaller than that of the others (no female vote). France, by contrast, underwent the organizational demobilization of its socialist party, which became the least densely organized party in Europe. Belgium shows a small but consistently growing trend in the 1945–1965 period,53 which is surprising, as it occurs in a particularly unfavorable situation of sharp electoral decline and in a political milieu characterized by the relative apathy of the citizens, by elitist values and culture, and by a minor role attributed to militants and members.54 In the ﬁnal period, following 1965, Austria and Sweden show signs of a declining partisan presence. The Norwegian and Danish socialist parties continue to decline at a rapid pace; indeed, the Danish decline amounts to an organizational demobilization, losing two-thirds of its organizational presence in the electorate. The Norwegian decline is similar, although less rapid. Most of the other countries show a declining trend, which is less dramatic given their low starting levels. The only exceptions are Germany 51

52 53

54

Experts have emphasized that if the effects of collective membership are disregarded, there has been a long-term decrease in the total number of members since the 1950s; Sjo¨blom (1978: 66). Gerlich (1987: 81–83). The Belgian data are available only for the postwar period; before that time, the party was structured indirectly and federally; see Dewachter (1976: 309). Moreover, the data have been fully reliable only since the increasing efforts to administer the party centrally in the 1960s; see Rowies (1977: 14). Dewachter (1987: 311–330).

268

THE POLITICAL MOBILIZATION OF THE LEFT

and Finland, which show signs of an increase in the late 1960s and early 1970s and which have since then stagnated. The case of the SPD is worth noting insofar as it challenges conventional wisdom about the development of membership. In clear contrast to all the logical speculation and hypotheses summarized at the beginning of this section, the German SPD experienced a major decline in membership until l953 under the leadership of Schumacher, that is, during the period in which the party adhered very closely to the ‘‘party of mass integration’’ model and during which the ideological emphasis and the functional uses of the members were still of a classical type.55 Conversely, after the party began its transformation into a catchall organization, its membership began to increase rapidly. The German SPD kept abreast of the German electorate as it increased – this was no mean achievement – and it even managed to increase its presence after l965. Finally, there is a striking decrease in the individual membership of the British Labour Party, which is even more pronounced if we consider that in the 1960s ‘‘constitutional’’ amendments were introduced to increase the differentiation between trade union and party members and to encourage people to afﬁliate individually.56 In Figure 6.2 the data concerning the communist parties are reported. In all cases, the over-time development of communist party membership is a normal curve, with a growth phase from 1920 to 1945–1950, with a peak immediately after World War II, and a subsequent generalized tendency to decline. The only exception is that of Finland, where the peak moves to the early 1960s. In eight countries, the partisan density of the communist movement never went beyond 1%, that is, 1 party member for every 100 electors. Only in Italy, France, and Finland has it acquired a higher level; in Belgium and Austria, communist partisan density peaked at a high level for only one or two elections after World War II, and then it rapidly declined to below the 1% level. A few other points are worth mentioning, as they question widely held assumptions. The French Communist Party is depicted as a large partisan machine in comparison with the other French parties; it looks less impressive, however, in a cross-country perspective. The French communists 55 56

Hartenstein and Liepelt (1962). In 1962, a constitutional amendment stated that only trade unionists who were individual members could be trade union delegates to the party conference. In 1965, the same thing was suggested in relation to attendance at constituency meetings. Moreover, in the same period, the Labour Party adopted the rule that local units should afﬁliate at a minimal membership of 1,000. As it seems that the membership of an average local unit was no more than 500, it follows that the real level of individual membership in the 1970s was much lower than the ofﬁcial ﬁgures suggest. See Leonard (1975), Martin and Martin (1977), and Committee on Financing of Political Parties (1976).

STRUCTURING AND MOBILIZATION

269

Figure 6.2. Communist party membership as a percentage of the electorate. strengthened their partisan network only in the 1930s and peaked at about 3% of the electorate after World War II. Immediately after this, their support started to collapse so rapidly that by the 1970s it was only slightly higher than that of the tiny communist parties of other countries. If one compares the partisan mobilization with the electoral support in that period, the conclusion is that the party can hardly be regarded as a powerful partisan network and that its electoral support went far beyond its organizational presence. Again, by social democratic standards, even the Finnish communist movement cannot be characterized as densely partisan, as its social presence never went beyond the 2% threshold. The same argument applies to the German Communist Party at the peak of its electoral fortunes. Only the Italian Communist Party shows a signiﬁcant level of partisan mobilization after World War II, when there were roughly 8 members for every 100 voters in 1945; this was roughly the same level as the Norwegian Socialist Party in the same period and was well above all the other socialist parties except for the Danish, Austrian, and Swedish ones. Even at the end of the 1980s, shortly before its deﬁnitive abandonment of every reference to the communist tradition, the party’s partisan

270

THE POLITICAL MOBILIZATION OF THE LEFT

density was about 3.5%, still above that of the majority of European socialist parties.57 In principle, one might have expected that, within the limits of the partisan mobilization propensity of a country, communists and socialists would share the partisan market in quotas that are somehow related to their electoral size. However, no association exists between the partisan density of communist and socialist parties. Only in Finland does there seem to be an equilibrium among the partisan density of the two parties. A tentative interpretation of the ﬁnding that the partisan market is more ‘‘monopolistic’’ than the electoral market is that in the far more ideologically tense world of membership and militant confrontation, one of these two parties necessarily takes a clear lead over the other. In conclusion, it is worth emphasizing that the overall trend of Communist membership developments since 1945 is similar to that of most socialist parties. If, as is sometimes argued, the Communist parties tried to resist such a tendency and to maintain a mass organization, they failed. Because they are both expressed as proportions of the electorate, the partisan density of socialist and Communist parties can be summed into a partisan density of the whole left. Considering only the post–World War I data, so as to reduce the inﬂuences of changing electorate size, the set of countries can be ranked in terms of average partisan density (Table 6.5). Such a ranking is fairly stable throughout the period and adequately illustrates their differences. I relate these patterns to electoral mobilization later in the chapter. However, a ﬁrst examination of the main tendencies in Figure 6.1, Figure 6.2, and the averages in Table 6.5 suggests that there are different national propensities toward a card-carrying involvement that should be linked to systemic features and that tend to stay unmodiﬁed. A second, and linked, consideration is that, notwithstanding the sharp decline in the memberships of the Danish and Norwegian labor parties, the small countries retain their systematic lead over those with larger populations. In the 1970s, the six countries with the highest ratio between socialist party members and voting population are Sweden, Austria, Norway, Belgium, Denmark, and Finland. Only the Netherlands and Ireland are exceptions. Smallness thus seems to facilitate socialist recruitment and 57

It is possible to read this information as proof that the Italian communist movement was, at least from the organizational point of view, qualitatively different from the traditional Communist Party from 1945 on. This thesis has always been defended by the party leadership, but it has also been supported by much of the scholarly research on the party. See in particular the collection of essays in Blackmer and Tarrow (1975). See also Tarrow and Ascher (1975).

STRUCTURING AND MOBILIZATION

271

Table 6.5. Average partisan left density (1918–1989)

a

Limited comparability due to high proportion of collective membership. Limited comparability due to female enfranchisement in the 1970s. c Only individual members. b

penetration, strengthening the old generalization about the negative relationship between size and participation. The analysis of the data suggests a few concluding considerations and speculations about party membership. The declining fortunes of socialist membership were not compensated for either by the growing membership of other left parties58 or by the growing membership of the non-left parties.59 The overall partisan density of the European party system has tended to stagnate or decline since World War II, because it has been 58

59

A discussion of radical socialist parties in Scandinavia can be found in Logue (1982: 249–290); membership ﬁgures for these parties are available in Katz and Mair (1992: 215, 745, 792). In some countries, various religious and conservative parties show increases in membership in the 1970s: the Finnish Christian League, the Irish Fianna Fail and Fine Gael, the Parti Chre´tien of Luxembourg, the Norwegian Conservative Party and Christian Democratic Party, the Swedish Center Party and Christian Democratic Party, and the German CDU, CSU, and Liberal Party. Katz and Mair document a few other similar cases in the 1980s. See Mair and Katz (1992). However, in no way do these increases compensate for the left secular losses.

272

THE POLITICAL MOBILIZATION OF THE LEFT

unable to keep abreast of the growth of the electorate in these countries. The basic problem thus becomes whether, from an organizational-resource point of view, this decline implies a decline in the organizational interest of parties for this type of resource, which in turn determines a corresponding decline in the investment made to promote it, or, from the individual motivation point of view, the data point to a decline in the political interest of citizens in this kind of participation and/or a shift toward other channels of participation. From the organizational perspective, party membership should be viewed mainly as a resource and as a consequence of the organizational incentives offered by the party leadership and ofﬁcers.60 To maintain or increase the levels of membership and activism requires an organizational effort that may or may not be rewarded in terms of money, work, and time. This will thus depend on how party leaders perceive and value the resource of membership and its ‘‘by-product,’’ that is, activism. Such an evaluation implies a cost–beneﬁt analysis conducted in terms of politics, ideology, and economics in comparison to the other resources available to the party, which can be strongly dependent on certain institutional characteristics of the political system, as well as on the individual-level phenomena discussed later.61 There are three main sources of party interest for partisan density as a resource. The ﬁrst is party ﬁnance. During the 1960s and 1970s, most European democracies introduced some form of public ﬁnancing for the political parties. Of the many changes that have touched the political parties since their development, this may have had some of the most farreaching consequences. Public ﬁnancing tends to decrease the organizational incentive to increase membership, even if much depends on whether alternative ﬁnancial sources are open to the leadership.62 This is because 60 61

62

Clarke and Wilson (1961), Conway and Fiegert (1968), and Lange (1977). I have tested a number of hypotheses relating the ﬂuctuation in socialist party membership to conditions of party environmental disturbances: the government/opposition role of the parties, the electoral and postelectoral years as compared with nonelectoral years, and the cases of electoral losses. The results of the test were mixed. More important, however, was the fact that the instances conﬁrming the role of membership as the organizational regulator of the party’s relationship with the political environment tended to decline over time, suggesting that parties either renounced efforts at organizational encouragement or lost the capacity to make them effective and rewarding. Cf. Bartolini (1983). A comparison of the German SPD and the British Labour Party is particularly interesting in this regard. The former has always raised substantial sums from its members. In a sense, it was obliged to do so (despite the fact that some public funds were already given to parties in 1959) by the fact that after World War I the trade union movement was declared to be nonpolitical and, in contrast to the United Kingdom and many other

STRUCTURING AND MOBILIZATION

273

most of the funds provided by the state are directed toward campaign and research expenses, a factor that could well make the candidates and the party machine more independent of the members. The second major challenge to the relevance of membership as an organizational resource derives from the postwar changes in the channels of political communication. In literate and urbanized societies, the mass media tend to replace party membership as the key means of conveying the political messages of the parties to the largest possible constituency. The role of traditional campaign and propaganda activities, such as local meetings and speeches by candidates and party representatives and doorto-door propaganda, declines in the face of more-effective techniques of mass propaganda. This thesis was challenged some years ago. It was argued that party leaders may grossly overestimate the capacity of the mass media to convey political messages, as the latter rarely have widespread impact unless they are relayed and reinforced within the innumerable face-to-face environments in each community.63 However, this criticism preceded the membership decline of the left parties and the further explosion of mediated communication. The environment of political persuasion has been increasingly privatized in recent years. This has reduced the role of face-toface communication and, consequently, the marginal utility of membership and militant activism. The contradiction between membership activism and other forms of party-efforts need not be absolute, but it is difﬁcult to deny that traditional local-party activity today merits less attention from the party leadership. Leadership attitudes toward membership are also inﬂuenced by its role in the formation of party policy. After participating several times in government, leadership groups have learned that their fate is much more closely linked to their responsibility to the electorate at large than to the possibly unrepresentative sample of party members.64 By increasing its

63

64

European countries, was not allowed to give substantial ﬁnancial support to the Socialist Party. Hence, the SPD emphasis on a dense network of local branches and membership mobilization (about 60% of party income comes from membership fees; cf. Morgan [1969: 38]). Relying on the pioneering work of Paul Lazarsfeld on the impact of the new mass media on the opinions and tastes of the U.S. public and on his two-step model of communication ﬂow, Rokkan argued that the mass media could reach and inﬂuence only active and interested citizen; passive and uncommitted individuals pay little attention to messages that do not ﬁt into their previously established orientation. See Rokkan (1970e: 425). See Seyd and Minkin (1979), who base their analysis on this argument. Both Pizzorno (1967: 28l–283) and Kirchheimer (1966b: 177–200) have stressed this point. They believe that the growing autonomy of political leadership in contemporary political organizations results from the need for all success-oriented organizations to accept conformity to the values held by the society as a whole in selecting their ruling

274

THE POLITICAL MOBILIZATION OF THE LEFT

autonomy and strengthening its position, party leadership becomes a resource in its own right, in the sense of the mass public image of leaders. Leadership needs visibility and autonomy of action to be exploited as a resource, whereas increases in membership and militancy for electoral purposes can easily occur at the cost of reducing leadership autonomy both inside and outside the party.65 From the individualistic perspective, membership decline results from a decline in the positive orientation of individuals toward the speciﬁc form of card-carrying political involvement. This perspective gains importance whenever collective enthusiasms and mobilizations progressively give way to phases of ‘‘individualistic mobilisation,’’66 when the individual motivations leading to party membership become less stable than those leading to the party vote.67 Individual-level arguments relating party membership ﬂuctuation to economic cycles or to material return distribution are of little use for interpreting the data described earlier. Cycle hypotheses cannot explain secular trends. Material-rewards hypotheses imply that membership increases when parties have more power and greater access to government,

65

66 67

groups. Consequently, the success of a leadership rests more on its identiﬁcation with the needs and values of the whole society than on the institutional goals and programs of its own organization. Associated with this ‘‘leadership autonomy’’ need is the parties’ interest in establishing links with new types of organizations designed to express the interests of new sectors of society that are not traditionally part of the party network. Parties may reduce their efforts to increase and strengthen membership in the traditional social units and instead devote their energy to recruiting from different functional groups in society through the direct incorporation of their organizational leadership. However, the chance and opportunity for corporate links with various organizations is dependent on the nature of societal articulation. Moreover, the recognition that other kinds of organizations are becoming increasingly attractive to politically active people does not necessarily lead to a reallocation of parties’ organizational resources and efforts to establish direct links with such organizations. The challenge might be regarded as a stimulus to devote more energy to increasing their own levels of membership and activism. Pizzorno (1967) and Hirschmann (1982). Actually, members are more sensitive to internal party life (splits, uniﬁcations, leadership crises) and are more unstable than party voters. An aggregate comparison of the stability of the socialist electorate with that of the socialist-party membership can be made through an index – albeit a rather rough one – computed in the following way: (the difference between the highest and lowest votes of each party divided by the average vote of the period) (the difference between the highest and lowest membership ﬁgure divided by the average membership of the period). If the index has a negative value, this means that the electorate of a given party is more stable over time than its membership. With the marginal exception of a few cases during the pre-1914 period, the clear result is that party electorates are everywhere more stable than party membership. All the socialist parties considered can be ranked according to this index, starting with the Dutch Labor Party, which shows the lowest difference in favor of electoral stability.

STRUCTURING AND MOBILIZATION

275

and, in the long term, membership is higher in countries with a longstanding socialist participation in government. Neither of these hypotheses, however, is borne out by the data.68 It is more promising to focus on the role of ideological motivations in stimulating membership, considering the level, intensity, and spread of ideological beliefs in different historical periods as the single most important determinant of ﬂuctuations. In the postwar period, we experienced a decline in the appeal of the traditionally integrative and all-encompassing ideologies that dominated the ﬁrst half of the twentieth century, even if we have not necessarily witnessed the end of ideology as such. However, the association between the clear decline of ideological tension of citizens’ political attitudes, on the one hand, and the long-term decline of partisan density, on the other, retains a distinctively qualitative ﬂavor. From another point of view, the growth and diversiﬁcation of leisure, as well as cultural and communal activities, has made membership in political parties, as well as other forms of direct political participation, increasingly marginal and unattractive for those whose membership choice was based on motivations of solidarity and identity. A larger and more differentiated number of voluntary associations are currently available, and these are often less demanding in terms of real involvement. Yet another perspective suggests that membership stagnation and decline, rather than being a sign of depoliticization, may be attributed to the shift toward alternative and competing channels of political participation. Even if it seems that, up to the 1960s, different forms of political participation were not mutually conﬂicting or exclusive, but rather produced spillover,69 the question remains of whether this has changed with new and unconventional forms of political participation; there is still limited comparative evidence on this problem.70 However, even individuals’ instrumental moti68

69

70

Bartolini (1983: 198–199). This ‘‘political economy’’ argument is certainly not supported by the Norwegian and Danish lefts, whose memberships’ secular decline started and was accentuated exactly in the period in which their lefts were enjoying a hegemonic governmental role. Considerable evidence on the positive relationship between party membership and other forms of political participation (and even of nonpolitical participation) was accumulated in the 1960s: Allardt and Pesonen (1960: 27–39), Barnes (1967: 122), Lane (1959: 166), and Berry (1969: 204). Two important works on political participation – Verba, Nie, and Kim (1978) and Barnes and Kaase (1979) – do not deal with this problem at all. The former deals only with conventional forms of political participation and, largely inﬂuenced by the American conception of and experience with party and party activism, considers membership as a mere indicator (among others) of party identiﬁcation. The latter, by contrast, studies unconventional participation and protest, but, developing its analysis exclusively at the level of individual attitudes, it rules out the impact of political institutions. These

276

THE POLITICAL MOBILIZATION OF THE LEFT

vations toward party activism have suffered considerably from the changes in the role of party politics. The struggle for collective goals that characterized the period of the structuring of mass politics was based on behaviors largely determined by moral commitments, collective loyalties, and nonspeciﬁc participation. The changes just mentioned have not simply brought about a new politics of ‘‘individual possessivism’’; they imply a more far-reaching change in the conditions in which various interests can be articulated and represented in the political arenas where parties are conﬁned to exercising an institutional role of mediation. These changes are seen in the growing importance of private, organized interest groups brought about by a type of political behavior that is increasingly inﬂuenced by instrumental and selﬁsh calculations coupled with speciﬁc forms of participation.71 The resulting trend is toward the collectively organized representation of speciﬁc and limited interests in order to satisfy sectional demands that are articulated autonomously. This behavior is based on the growing awareness among large sectors of the public that the living conditions of individuals are increasingly dependent on speciﬁc governmental policies in the welfare state and that they are no longer the result of enduring collective struggles under the leadership of political parties and their ancillary organizations. In facing these new forms of political behavior, parties – and particularly traditional mass parties – are further handicapped by their tendency to be in some way the ‘‘captive’’ of those interests and organizations that led to their initial emergence.72 More generally, individual ideological, solidarity, and instrumental sources of party commitment have been challenged by new modalities of political participation such as action groups, local movements, single-issue parties, and so on. Precisely because they concentrate on a narrow range of problems and are not responsible to large electoral constituencies, they can

71

72

choices are legitimate. However, they weaken the ﬁnal extrapolations concerning the problems and strains that contemporary institutions of political involvement – namely, parties, interest groups, and mechanisms of representation and election – may undergo as a result of the emergence of the new issues, attitudes, and cleavages documented in the book. These new phenomena have been very skillfully described by Schmitter (1977). However, this analysis is based to a large extent on the experiences of central and northern European countries, and these phenomena may be less visible and important in Southern European parties. On this point, see Linz (1979: 180). Offe (1972) elaborates on this point, identifying the ‘‘repressive’’ nature of the modern liberal democratic political systems in this intrinsic selectivity of interest representation. From a different perspective, recent studies on postmaterialist protest come to the conclusion that unconventional forms of behavior emerge as a consequence of inadequate representation; cf. Inglehart (1977) and Barnes and Kaase (1979).

STRUCTURING AND MOBILIZATION

277

maintain the commitment of their membership, something that is simply not feasible in a larger mass party. In this way, they satisfy the demands that party activism is less suited to meet. It should also be added that the ever-increasing interest in local political issues would seem to favor new kinds of militancy other than that of party membership. Traditional parties, and particularly socialist parties, have tended to nationalize political issues and to place them in a broad perspective, thus ﬁnding it difﬁcult to deal with them satisfactorily in purely local terms. In conclusion, the interpretation of the historical decline of left party membership needs to be framed within three intertwined processes. First, there are cross-national differences in partisan mobilization resulting from national systemic properties (historical experiences, the closeness of societal groups to political parties, the role of parties as opposed to other agencies in early mobilization or regime founding, etc.). These elements determine different national propensities toward this form of political participation and explain the main cross-country differences irrespective of time and historical periods of collective mobilization exuberance. Second, the shortterm variation in partisan mobilization can be referred to as an organizational perspective that responds to the party’s strategic decisions to foster or dampen this type of mobilization. Finally, long-term trends common across countries cannot be interpreted as the result of national systemic features or the strategic choices of the individual parties. Instead, they can be best explained by looking at membership from an individual perspective, as a factor external to party organization that inﬂuences the propensity to adopt this type of political involvement. Basic changes in individual motivation toward partisan activism, and declining organizational investments in it have combined to determine the post–World War II general decline in or stagnation of membership ﬁgures and a corresponding decline in party activism. UNION MOBILIZATION Whereas party membership is the result of the creation of a nationalized and centralized political organization, union membership developed before any form of national and centralized union movement occurred. The precondition for party membership was freedom of political association; the precondition for electoral mobilization was the right to vote; and the precondition for corporate mobilization was freedom of economic association. Beyond this, however, the possibility for unions to mobilize potential support rested on their capacity to deliver collective goods for their potential audience. This, in turn, depended mainly on the unions’ capacity to

278

THE POLITICAL MOBILIZATION OF THE LEFT

provide collective contracts that could beneﬁt an entire group of persons with a single agreement on wages, work conditions, and the like. In this section, I examine the European variation in levels of corporate mobilization by looking at the levels of union membership. I then relate their variation to the modalities of collective-agreement developments. Data on union membership are more reliable than those on party membership because more research has been devoted to them, and excellent data collections have been produced. However, not all unions were socialist or ideologically linked to left parties, so the latter have to be separated from those of different political orientations. The attribution of unions to one camp or the other has been based on secondary sources and has not posed any major problems (see the Data Appendix for the sources).73 Figures 6.3 and 6.4 chart the historical trend in overall and left trade union density across the European countries. The difference between Figure 6.3 and Figure 6.1 concerning left party membership is striking. First, there is a secular tendency to growth and no sign of decline in the most recent decades. Second, the over-time development suggests that levels of trade unionization of the dependent labor force were relatively uniform in the pre–1918 period and tended to differentiate later on. Before the turn of the twentieth century, the highest levels of trade unionization (in the United Kingdom and Denmark) stood at about 16–18% of the dependent labor force, while the lowest ranged from 2% to 5%. By the 1980s, the differences between the extremes (even excluding France) are on the order of 50–60%. The period in which cross-country differentiation occurs is therefore the one after World War II. In 1945, the range was still limited to about 30%, and it has doubled since then. For almost all countries, there are two clear peak periods – 1917–1922 and 1945–1947 – followed by a clear decline in both cases. There was a period of general growth until World War I and a peak after it; then a decline in the 1920s, followed by new growth in the 1930s and a peak immediately after World War II. After this, some countries continued an upward trend: Sweden, Austria, Denmark, Norway, Belgium, and Ireland. Others underwent a decline in the 1950s and then a new growth phase later on: Italy, Finland, and the United Kingdom. Still others remained fairly stable throughout the post–World War II period: the Netherlands and Germany. Finally, there are only two cases of clear and protracted 73

In Austria, it is more difﬁcult than elsewhere to estimate the percentage of left trade union members over all members. I have based this estimate on the percentage of seats in the Trade Union Confederation governing body, which has been stable at about 77% for the SPO, 20% for the OVP, and 3% for the KPO. A discussion of these ﬁgures is available in Gerlich and Mu¨ller (1983: 343).

STRUCTURING AND MOBILIZATION

279

Figure 6.3. All trade union members as a percentage of the dependent labor force.

postwar decline: Switzerland and France. For the latter, it is possible to view this trend in terms of a postwar collapse from which the unions never recovered. In the 1980s, there were some signs of decline in Ireland, the United Kingdom, the Netherlands, Germany, and Austria. However, this trend was not generalized, as it was in the case of party membership. The picture of left union membership (Figure 6.4) is not substantially different, apart from two features. The ﬁrst concerns the post–World War II period, during which the density of the left unions give an impression of stability rather than growth. Only in Finland, Sweden, and Denmark did left trade unionization grow sharply. This implies that post–World War II trade unionization growth has been greater for non-left unions than for left ones. The second point is that the difference between the Protestant north (Denmark, Sweden, Finland, the United Kingdom, Norway) and the Catholic or religiously mixed center and south of the Continent (Austria, Belgium, Germany, Italy, Switzerland, the Netherlands, France) is far more evident for left trade unionization than it is for overall trade unionization. The decline in left party membership after 1945 was paralleled by

280

THE POLITICAL MOBILIZATION OF THE LEFT

Figure 6.4. Left trade union members as a percentage of the dependent labor force. increasing levels of trade unionization in European society. Although traditional working-class unionization was generally regarded as supporting and reinforcing the electoral and membership development of left parties, the situation became less clear with the new middle-class and publicemployee unionization. Trade unions might well have been seen by many people as a more effective channel of interest representation than the parties insofar as they were able to preserve their ideological standpoints and the defense of their sectional interest more effectively than the parties. The ratio of left trade union density to total trade union density points to a varying degree of corporate or fragmented relationship in the interest channel. This was discussed before and can now be looked at more precisely for the unions. In Figure 6.5, the lines of total trade union density and left union density are reported for each country. The difference between the two is indicated by the gray area, which represents the level of trade unionization outside the socialist world. In one group of countries – Austria, Belgium, the Netherlands, Switzerland, and Italy – the union

STRUCTURING AND MOBILIZATION

281

movement has been historically characterized by the presence of nonsocialist unions; therefore, the mode of representation in the interest channel is historically fragmented.74 This historical fragmentation has grown over time to the advantage of the nonsocialist unions, which have grown more since World War II. In another group – France, Germany, and Ireland – there is also the historical tradition of Catholic and/or religious unions, but there is no historical increase in the share of the unionized labor force of this nonsocialist component. Finally, there is the group of the homogeneously Protestant northern countries. In Denmark, Sweden, and Finland, nonsocialist unions were not the result of religious countermobilization, but rather of the unionization of social groups that did not identify politically with social democracy. In these three countries, these unions were weak up to the 1950s, but they exploded from the 1960s on, representing the most dynamic sector of new trade unionization. The same applies to Norway, the only difference being that until the 1950s there were almost no nonsocialist unions there. Britain also belongs to this group but does not show the growth of nonsocialist unionism, as in the other cases. With the exception of Finland, these northern countries provide the clearest historical cases of corporate representation between the political and interest channels. However, more recent developments have broken this pattern: The monopoly of representation of socialist unions disappeared by the end of the 1980s, even if the new nonsocialist unions did not necessarily move closer to other, nonsocialist parties. Table 6.6 reports a rank ordering of European countries for the period 1918–1986 in terms of the level of general and left trade unionization. Then the average ﬁgures for the post-1945 period are reported in order to help judge the extent to which the general ranking for the whole post– World War I period reﬂects later developments. Sweden and Denmark lead the trade unionization rank ordering, both in general and for the left, with Austria and Norway closely following. The difference between Norway and Austria is that left trade unionization grew more strongly in Norway after World War II, while it was muted in Austria in the same period, also as a result of the success of other nonsocialist unions. Belgium and the United Kingdom follow with slightly lower levels, but in Belgium left trade unionization reaches barely half of the overall trade unionization. On the whole, these are the countries for which it can be safely said that 74

The ﬁgure shows nonsocialist unions appearing only after World War II in Italy. However, even before the rise of fascism, Catholic unions existed, although it is difﬁcult to estimated their membership.

282

THE POLITICAL MOBILIZATION OF THE LEFT

Figure 6.5. Non-left trade unionization by country.

STRUCTURING AND MOBILIZATION

Figure 6.5. (cont.)

283

284

THE POLITICAL MOBILIZATION OF THE LEFT

Table 6.6. Trade union and left trade union membership as a percentage of the dependent labor force: country rank ordering

there was a comparatively high European level of trade unionization throughout the whole period under consideration. In Ireland and Finland, by contrast, there is a marked difference between the average for the 1918–1986 period and that for the post– World War II period; in both cases, the period of major growth followed World War II. Switzerland and France stand at opposite ends of the continuum: Their averages were below the European average in both general and left trade unionization. By the 1920s, a general ranking can be said to have been established; and since then, countries have kept their relative levels of overmobilized, medium-mobilized, or decidedly undermobilized corporate channels. This was less the case in the development of party membership. In other words, if the overall temporal trend is put to one side, in each main period most of the variation is among countries rather than within them.75 75

An analysis of variance for the 1945–1986 period shows that about 70% of the variance is between country groups compared to less than 30% within country groups. For some of the systematic hypotheses concerning these national differences, see Stephens (1979) –

STRUCTURING AND MOBILIZATION

285

To conclude this section, I shall now deal with the question of the relationship between cross-country and cross-time variation in the levels of mobilization and the unions’ incorporation into the industrial-relation system by establishing collective bargaining and agreements. In this area, a cross-country comparison is too crude, as union growth in different sectors should be related to the spread of collective agreements in those sectors, but it does offer some insight into the role that collective agreements played in the development of union strength. To examine the onset of collective bargaining in European countries and to deﬁne the point at which it became an established practice is complicated because (1) in some cases, it is difﬁcult to indicate a clear point at which collective bargaining started; (2) the interval between the emergence of a ﬁrst agreement in a skilled trade and the time when collective bargaining covered a signiﬁcant section of the working population can be considerable; (3) a general agreement between national employers and trade union confederations did not always lead to formalization of the system; and (4) the date of legal recognition may be misleading, either because it came many years after a system had actually been established (for example, in the Netherlands) or because formal recognition was not matched by practice. Therefore, for a comparative assessment, I have collected information concerning (1) the date at which collective bargaining showed signs of emerging; (2) the date at which laws governing bargaining were enacted; and (3) the point at which collective bargaining became an established practice, that is, the ‘‘critical mass’’ point at which bargaining was widespread and accepted.76 A comparative overview of these ﬁndings is reported in Table 6.7. The table summarizes the development of collective agreements, reporting the date of the ﬁrst agreement when known, the date of legal recognition and the regulation of such agreements, and the period in which collective agreements spread to become common practice. Together with these three dates, comments are made that add qualitative information.

76

who emphasizes the degree of concentration of capital as a key explanatory variable – and Rothstein (1992) – who emphasizes the differences in government labor-market institutions, in particular the policies and institutions of unemployment insurance (the Ghent system versus other solutions). In this section, I have relied on the research assistantship of Simon Dubbins. Only the ﬁnal results of this comparative analyses are reported here. I have used the following sources: Crouch (1993), Traxler (1992), Vilrokx and Leemput (1992), Scheuer (1992: 172), Skogh (1984), Brunn (1991), Lilja (1992), Goetschy and Rozenblatt (1992: 404– 408), Stearns (1971), Neufeld (1974), Ferner and Hyman (1992), Cella and Treu (1989), Windmuller (1969), Galenson (1949), Prodzyznski (1992), Fulcher (1991: 141), Parri (1987), Katzenstein (1984), and Musson (1972).

286

THE POLITICAL MOBILIZATION OF THE LEFT

Table 6.7. Development of collective bargaining in European countries

STRUCTURING AND MOBILIZATION

287

The various national experiences can be organized along two dimensions, which will now be discussed. Timing of Development Three groups can be identiﬁed – early, medium, and late. In the early category – 1890–1914 – have been placed Britain, Switzerland, Austria, Denmark, and Belgium, where the critical-mass point was reached either at the end of the nineteenth century or in the ﬁrst decade of the twentieth century. In the middle category – 1914–1945 – are Germany (excluding the period of Nazi rule), Sweden, Norway, the Netherlands, Ireland, and, to some extent, France. In the late category – after 1945 – are Finland, Italy, and again France. France appears in both categories because bargaining develops between the wars, although it becomes established only much later. Germany is also problematic: While bargaining did develop after World War I, it was accepted only by some employers and was destroyed by the Nazis. Therefore, it may be better to place Germany in the third category. Types of Development There seem to be three types of bargaining developments: 1. Steady evolutionary developments in which a steadily increasing proportion of the workforce have their terms and conditions regulated by collective agreements. This group includes the United Kingdom, Austria, Switzerland, Germany, Belgium, and Ireland. 2. A more explosive style in which bargaining is resisted and then, after intense struggles, gains acceptance spreading rapidly to include large sections of the workforce. This group includes Finland, Norway, Sweden, the Netherlands, and Denmark. 3. An unbalanced growth in which bargaining develops for small sections of the population, but with constant reluctance to bargain until very recently. This group includes France and Italy. These three types of bargaining result in a typology of different national experiences, which is reported in Table 6.8. From a strictly individualistic perspective, it can be argued that the capacity of an organization to deliver collective goods should lead to stagnation or even decline in its membership.77 Unions that are unable to force collective agreements may still attract members who desire those 77

Olson (1971).

288

THE POLITICAL MOBILIZATION OF THE LEFT

Table 6.8. Timing and type of development of collective bargaining

agreements and perceive their additional contribution to the organization as instrumental to such a goal. When the capacity to deliver the collective good is ﬁnally achieved, this drastically alters the cost–beneﬁt calculations of potential members and leads to membership stagnation or decline. From a contrasting perspective, collective agreements may be viewed as the sign of the unions’ capacity to deliver what their claims promise, that is, as the evidence of unions’ claim that organizational mobilization pays. As such, collective agreements may act as a powerful incentive to membership recruitment. In this collective rather than individual perspective, the unions’ capacity to win collective concessions actually fosters unionization; therefore, collective agreements mark the turning point in union growth rather than their decline. We can look at the historical development of union membership discussed before in the light of these two opposite hypotheses. A relationship exists between the earliness of collective agreement institutionalization and the overall unionization rate of the labor force (Table 6.9). Countries where collective agreements developed late tended to have a lower density overall and, in particular, a lower left union density, while the highest density ﬁgures are shown by countries with early collectivebargaining practices. However, of the ﬁve countries that introduced widespread collective bargaining either early or at an intermediate time, at least two never reached high trade unionization levels: the Netherlands and Switzerland. In contrast, among the latecomers, Ireland, Finland, and Italy have trade unionization levels that are higher than those of Switzerland or the Netherlands. A more decisive test of the role played by collective bargaining and agreements in relation to trade union growth is performed by comparing the general union density in the period of the spread of collective agreements

STRUCTURING AND MOBILIZATION

289

Table 6.9. Levels of trade union density by timing of institutionalization of the collective bargaining

(as deﬁned in Table 6.7) with that of the two decades before and the two decades after this phase. The choice of the two decades rather than a different number of years is arbitrary, but I feel that to evaluate the impact adequately, long periods are necessary. For the same reason, I identify the phase of wide spreading with a long period to avoid identifying the date of collective-agreement institutionalization with that of legal acceptance. In Table 6.10, countries are ranked according to how early collective agreements were institutionalized.78 The institutionalization of collective agreements did not start earlier in those countries where the trade unionization level was comparatively higher. This suggests that earlycomers introduced collective agreements not as the result of union growth pressure but because of other institutional or political features of the political system. In these countries, union growth occurred primarily after the introduction of collective agreements. This is true even for the United Kingdom, where union density doubled in the two decades after their institutionalization. Even for latecomers, the introduction of collective agreements led to union growth both during and after the introduction, even if, obviously, this introduction occurred when unions were stronger than was the case for the earlycomers. There are only three cases that lend some support to the idea that collective agreements were the result rather than the cause of union growth and that they were later responsible for union stagnation or decline. Italy 78

In the ﬁnal column of Table 6.10, the average trade union density for the entire period after collective-agreement institutionalization is reported. This allows a control, particularly for those countries where collective bargaining agreements were established very early in the century, on whether the change in the two decades immediately afterward was major or minor with respect to the overall change in the whole period after the critical-mass phase.

290

THE POLITICAL MOBILIZATION OF THE LEFT

Table 6.10. Trade union density and development of collective bargaining

is a borderline case. Union density had already reached about 40% before collective agreement spread, and there was only a minor increase during the spread itself. Still, the decade following this spread saw a remarkable increase of 10%. In Germany, union density was low before the introduction of collective agreements and increased mainly during the period of their proliferation between the two world wars. In the following period, growth stopped and remained more or less stable. France is the only case in which union density declined both during and after the spread of collective agreements. On the whole, however, it cannot be maintained that in Western Europe the unions’ capacity to deliver collective goods to their potential constituency reduced the willingness of employees to join them. Collective agreements appear almost everywhere as a resource for union growth and as a sign of the success and credibility of their representational claims that did nothing to discourage participation.

ORGANIZATIONAL STRUCTURING, MEMBERSHIP MOBILIZATION, AND ELECTORAL DEVELOPMENT So far, I have discussed (1) the modalities of the early organizational consolidation of the socialist movement in both its political and corporate aspects; (2) the successive levels of partisan mobilization that ensued in the political channel; and (3) the union density levels and growth. In this section, these elements will be related to the electoral developments of

STRUCTURING AND MOBILIZATION

291

socialism. At later stages, they will also be used in reference to other features of the class cleavage. Three issues are discussed in the following sections: (1) to what extent differences in early central organizational consolidation (sequence, timing, length, cross-linkage, representation mode) later inﬂuenced membership mobilization capacity; (2) what kind of relationship can be said to exist between corporate, partisan, and electoral mobilization; and (3) whether the combination of the three types of mobilization actually identiﬁes different styles or types of socialist movements and class cleavage mobilization.

ORGANIZATIONAL CONSOLIDATION AND MEMBERSHIP MOBILIZATION To model the organizational consolidation pattern, I have arranged its dimensions as shown in Figure 6.6. The earliness or lateness of the central consolidation of unions and parties combine to determine the party/union sequence (party ﬁrst, union ﬁrst, or parallel). The correlation coefﬁcient of the length-timing classiﬁcations indicates the extent to which, for instance, early timing of union development is associated with a resulting pattern of union precedence over the party; similarly, the association between party timing and sequence indicates the extent to which early party formation accompanies a party-led sequence.79 Party and union length and timing are ordered by giving the highest value to the respectively most favorable situation (early development and short consolidation). This makes it necessary to have two ‘‘best sequence’’ ordinal scales: one for the unions – attributing the highest value to the unions leading over the party – and one for the parties – attributing the highest value to the party leading over unions. This explains why the correlations have double signs but equal value.80 79

80

The correlation in the table results from the following ordering of the organizationalconsolidation dimensions. Unions’ Sequence: 3 ⫽ Unions ﬁrst; 2 ⫽ Parallel; 1 ⫽ Party ﬁrst; Party sequence: 3 ⫽ Party ﬁrst; 2 ⫽ Parallel; 1 ⫽ Unions ﬁrst; Unions’ timing: 3 ⫽ early; 2 ⫽ mid; 1 ⫽ late; Party timing: 3 ⫽ early; 2 ⫽ mid; 1 ⫽ late; Unions’ length: 3 ⫽ short; 2 ⫽ mid; 1 ⫽ long; Party length: 3 ⫽ short; 2 ⫽ mid; 1 ⫽ long; Cross-linkage: 3 ⫽ interlocking; 1 ⫽ contingent; Mode: 3 ⫽ corporate; 1 ⫽ fragmented. If unions’ length ordering is correlated with the sequence for unions (with unions’ leading being assigned the highest value), the association is positive. If the same variable

292

THE POLITICAL MOBILIZATION OF THE LEFT

Figure 6.6. Modeling the organizational consolidation process. If party and union length and timing determine a speciﬁc sequence, the latter impacts on the relationship between parties and the unions. Party-led processes of central consolidation should thus more logically result in union dependency or interlocking patterns. At the same time, union-led processes should more often lead to contingent or interlocking relationships. As far as the mode of representation is concerned, union centralization preceding party development is associated more with ﬁnal fragmented representation in both channels, while the opposite is true for sequential development in which party development clearly precedes union centralization and presumably plays an important role in this process. With this kind of framework, we can investigate the relation between each pattern and the actual mobilization capacity in the different channels of the socialist movement. To this end, Table 6.11 reports the mean levels of the three types of mobilization – left trade union density, left parties’ partisan density, and total left vote – by the modalities of organizational consolidation in the period between 1900 and 1940, when all countries had elections and the franchise tended to converge toward universality. Corporate mobilization is more than twice as high with early or intermediate consolidation than with late consolidation. It is also inﬂuenced by the length of the consolidation process in an unexpected direction: higher levels of corporate mobilization with longer centralization processes. Note that factors relating to political centralization and representation have a strong impact on corporate mobilization, which is roughly double (1) in is associated with the sequence with the party leading the unions, the correlation has the same value but the opposite sign.

STRUCTURING AND MOBILIZATION

293

Table 6.11. Pattern of organizational consolidation (1880–1920) and levels of political mobilization (1900–1940)

Note: Number of elections given in parentheses

cases of early party centralization and (2) in cases of interlocking party– union cross-linkage. A corporate mode of representation is associated with higher levels of corporate mobilization in only a limited way. Finally, sequence – whether party consolidation comes before or after union consolidation – does not inﬂuence the membership development of the unions. Although the role of political parties and activists in fostering trade union development is known, it is somehow striking that political and representational factors (the linkage and the mode of representation) have such a strong effect on the levels of union membership mobilization. A similarly notable surprise is that the variables belonging to the process of organizational consolidation do not have a signiﬁcant impact on the levels of electoral mobilization. The differences in the means are far lower than for the other forms of mobilization. The only relevant factors are party-related factors:

294

THE POLITICAL MOBILIZATION OF THE LEFT

the early formation, the rapidity of its centralization, and interlocking party–trade union linkage. In contrast, none of the factors relating to trade union consolidation are signiﬁcantly associated with electoral mobilization. Note that while a fragmented mode of representation is associated with somewhat lower levels of corporate mobilization, it shows little impact on electoral mobilization. The electoral success of the left parties appear to be quite independent of the modalities of organizational development in the corporate channel. Party organizational and representational factors influence corporate mobilization much more than corporate organizational factors affect electoral mobilization.81 Partisan mobilization is indeed more highly inﬂuenced than the other forms by organizational consolidation aspects. Almost all organizational characteristics do have a signiﬁcant impact on partisan mobilization (the weakest one being the length of party consolidation). Not surprisingly, the factors that play the most important role are the sequence (the party-led sequence is favorable), the timing of party consolidation, and, in particular, the linkage and the mode of representation. Interlocking cross-linkages are associated with two-and-a-half times higher levels of partisan mobilization; corporate monopolistic trade unions and political representation fosters a partisan mobilization that is three times higher than does fragmented representation. Parties that grew in close organizational interpenetration with uniﬁed trade union movements gained advantages from the dense network of organizational cross-linkages. The model in Figure 6.6 can be arranged to create syndromes of the most favorable and most unfavorable sets of conditions for each type of mobilization. For trade union mobilization, one can hypothesize that the most favorable conditions occur when trade unions organize early, within a short period, before the party, with an interlocking or party dependency cross-link with the party and with a corporate mode of representation. The most favorable organizational combination for electoral and partisan mobilization consists of an early and short process of party centralization, a party-led sequence, and again, interlocking linkage with unions and a corporate mode of representation. The opposite combinations should be regarded as the least favorable for each form of participation. To test this hypothesis, we can run a regression of each type of mobilization against the set of factors that identiﬁes its supposed best conditions and then compare this with a similar analysis in which all the 81

This thesis seems to be valid even for the British case. With a geographical analysis of electoral changes and of unions’ density, the important research of Tanner (1990) has shown that the development of the electoral support for the Labour Party does not reﬂect the areas of union expansion in the 1920s.

STRUCTURING AND MOBILIZATION

295

organizational variables are included. For each speciﬁc form of mobilization, this can identify the most discriminating factor among those discussed here (controlling for the combined effect of all) and the global weight of the whole set. This is reported in Table 6.12. For each type of mobilization, the ﬁrst column shows the result of a regression performed with only the organizational variables that concern that speciﬁc form of mobilization and set the best conditions for it. All the organizational variables are used for the same exercise in the second column. First, this test conﬁrms that the process of organizational consolidation is far more important for partisan and corporate mobilization than for electoral mobilization, for which it explains little variance (standing at .177 compared with .477 and .311). Second, entering the party consolidation variables into the regression concerning corporate mobilization increases the association considerably, adding about 11 percentage points in explained variance (from .311 to .421). The timing of party development, in particular, takes on an important weight. When the variables concerning union consolidation are added to the regression of electoral mobilization, nothing changes in the overall association (.182 versus .177) and very little in the order and relative weight of the variables. This confirms that left electoral mobilization does not depend on the forms of corporate-channel organizational consolidation, while corporate mobilization does depend to a considerable extent on the pattern of party consolidation.82 Partisan mobilization lies in an intermediate position, the bulk of the variance being explained by party-related variables. However, the inclusion of the union consolidation variables causes a signiﬁcant increase in explained variance, as both trade union timing and length appear important. The fragmented or corporate representation model goes a fairly long way toward determining levels of partisan membership. The data indicate that political and organizational divisions impact most strongly on partisan membership participation. By its nature, the latter is an intermediate form of participation that combines the character of voluntary and active adhesion in a way that is appreciably more ‘‘expensive’’ than voting (like corporate mobilization, in this sense) and the character of remoteness from 82

To check this conclusion further, I have performed a regression of each type of mobilization with the variable that, in principle, should have a lower impact on it: that is, corporate mobilization with party consolidation variables and partisan and electoral mobilization with union consolidation variables. While party consolidation variables alone explain more variance in corporate mobilization than do corporate consolidation variables (about 38% compared to 31%), in the other two cases union consolidation variables explain less variance than the party variables (44% versus 48% in partisan consolidation and 14% versus 18% in electoral mobilization).

296

Table 6.12. Relative weight of organizational consolidation factors (1900–1940)

Note: method ⫽ enter; entry .05; removal .10.

STRUCTURING AND MOBILIZATION

297

functional speciﬁc interests and actions, in a way that is appreciably more ‘‘abstract’’ in motivation and incentives than union membership (like electoral mobilization, in this sense). ‘‘Costing’’ more than voting does and being more ‘‘remote’’ from interests than union membership is, partisan mobilization is more strongly affected by organizational and ideological incentives than are corporate and electoral mobilization. RELATIONSHIP BETWEEN FORMS OF MOBILIZATION As far as I know, the association between corporate, partisan, and electoral forms of mobilization has never been examined in a comparative perspective. For the period under consideration and for the set of approximately 320 elections, the associations are shown in Figure 6.7, including both left and general trade union density. There are several striking features in these ﬁgures. First, they are all very similar. Whether we consider trade union density or left trade union density, the association between partisan and corporate mobilization is basically the same and has a level that comes close to that between partisan and electoral mobilization. Second, the correlation between corporate and electoral mobilization is slightly higher than the others for both total trade union density and left trade union density (.563 versus .491). Moreover, electoral mobilization is associated more strongly with trade union density than with left trade union density. Finally, and most important, these associations appear to be surprisingly low, considering the close historical relationship between the left parties and trade unions and considering the fact that electoral and union developments both show a historically growing trend that should foster their association over time. Two theses can be used to justify the overall weak association between corporate and electoral socialism in the nineteenth century. The ﬁrst can be labeled the ‘‘learning process thesis.’’ It states that, although both processes show a secular growth trend, in the early days of socialist mobilization workers who were already mobilized in the corporate channel were not yet translating this membership into voting practices for class parties due to lack of experience and/or class consciousness, to the habit of abstaining, or to anarcho-syndicalist political or antipolitical stands. It was only over time, and after a long process of apprenticeship and political education, that they came to develop the habit and realize the importance of electoral politics. If this were true, one would expect an increase over time in the association between these forms of political engagement. Thus, a generally weak association should result from weak levels of association in

298

THE POLITICAL MOBILIZATION OF THE LEFT

Figure 6.7. Association among partisan, corporate, and electoral mobilization. the early days of the socialist movement’s development, with a growing level in more recent decades, say from the interwar period or from World War II on. The second thesis can be labeled the ‘‘golden age’’ thesis. Here, the idea is that in the early phase of class mobilization, clearer class alignments and greater social homogeneity of socialist support existed. Later on, goes the argument, the pattern was complicated by several factors, ranging from the progressive abandonment of the class mobilization appeal of socialist parties in favor of broader nonclass appeals, to the social and cultural blurring of class boundaries, to the inﬂuence of the countermobilization of religious or nationalist movements, and so on – a long list of explanations of why class is less closely linked with political behavior today than it used to be. In this second case, the overall weak association should result from a temporal pattern that runs in the direction opposite the one described previously: A strong association in the early ‘‘golden age’’ of class alignments should decline over time as other factors impact on the relationship between corporate and electoral socialism. The learning process and golden age theses can be tested by analyzing the temporal development of the association between the different measures of mobilization. In the ﬁrst case, we would expect that low association levels in the early formative period become increasingly strong with time. In the second case, the opposite should result: A strong association in the early golden age should become weaker over time. Figure 6.8 reports the values of the correlation between corporate and electoral mobilization, decade by decade, over almost a century of electoral politics. The decade correlations tend to be very low in the period up to the 1920s, peaking in the decade 1930–1940 and tending to decline later to levels roughly similar to those of the ﬁrst period. This evidence does not ﬁt either the learning process or the golden age thesis. If instead we use the logic of both arguments to interpret the curves, we can argue that the early phase

STRUCTURING AND MOBILIZATION

299

Figure 6.8. Correlation by decade between left votes and trade union density. was not characterized by strong association because of the delayed translation of corporate membership into political behavior; that the golden age of association between class and electoral behavior is to be located in the decades around the middle of the century; and that the class dealignment factors mentioned previously have tended to make union membership and the left vote increasingly less associated since the 1950s. If there was a period in which left parties became the captive of their own high workingclass support, this was in the 1930s to the 1950s. Who voted for the socialists in the early days, then? As far as the class composition of the socialist vote is concerned, the ﬁrst formative phase should be regarded as being at least as problematic, in terms of class alignment, as the last and most recent phase from the mid-1960s on. General association measures hide cross-time and cross-country differences. To facilitate the direct cross-country, cross-time, and cross-channel comparisons of political mobilization processes within the socialist movement, data have been averaged in Table 6.13 by decade, computing left union membership and party membership in percentages of the national electorate. Taking the values of the decade means at the bottom of the table, it is easy to see that in the 1930s there were, on average, 12.6 members of left-oriented trade unions, 5 members of the left parties for every 100 electors, and 34 left voters for every 100 valid votes. Compared with the same ﬁgures in the decade of the 1900s, partisan density remained stable, corporate density increased by about 20%, and electoral support increased by 100%. Therefore, in the global experience of European socialism, electoral mobilization lags behind corporate mobilization until approximately the 1890s, but from then on there are far more left voters than left union members. Note also that partisan mobilization is the most stable of all.

300

THE POLITICAL MOBILIZATION OF THE LEFT

Table 6.13. Levels of corporate, partisan, and electoral mobilization by decade

n.a. ⫽ data are not available

Indeed, between the decade of the 1910s and the 1930s it did not change much, despite the fact that both corporate and electoral mobilization were on the increase. Moving to national differences, Finland stands out as the only country where early electoral socialism was totally unmatched by corporate mobilization: The Finnish trade unions never had the enormous early success of the Finnish Socialist (and later Communist) Party. France and Italy also stand out as cases of predominantly electoral socialism. The weakness here of the socialist movement in both corporate and partisan channels is aston-

STRUCTURING AND MOBILIZATION

301

ishing when compared with their roughly average electoral development. It is highly unlikely that in this period the considerable electoral strength of these socialist movements lay in their organizational infrastructures.83 Five countries show a relatively parallel development in the electoral and corporate channels: Austria, Belgium, Switzerland, Denmark, and Sweden. We can therefore come to a second conclusion: Independent of the level of association between corporate and electoral mobilization, the former tended to be higher than the latter in the early decades. After this, the growth in votes began to reﬂect a higher and more accelerated increase than that of union membership. Over time the experiences of the different countries tended to become more similar in electoral terms; although converging, they remained more differentiated in corporate mobilization; the profound differences between them in the partisan mobilization channel continued. There is no single instance of a reversal or a recovery in this ﬁeld. Socialist movements that were weak in partisan mobilization remained in the same class throughout the period. In this context, Austrian socialism provides an impressive image of partisan mobilization: There was one socialist party member for every seven or eight voters. The ﬁgure was one for every twelve for Denmark and Sweden and one for every eighteen to twenty for Norway. No other socialist movement even approached these levels of partisan mobilization within the electorate. Intermediate cases, such as Germany and Switzerland, already had one party member for every thirty to forty, voters. To see the contrast, it is interesting to compare these ﬁgures with the protracted structural weakness of the French, Italian, Dutch, Belgian, British, and Irish socialist movements; by and large, in these cases, there was one left party member for every 60 to 100 voters. In these cases, therefore, the partisan activities of recruitment, propaganda, social work, and the like were probably concentrated in a few regions or urban areas. If one directly compares left union members and left voters, considering the ratio between the two (how many left trade union members existed at each moment for 100 left party voters), three countries stand out as clear-cut cases in which socialist mobilization in corporate channels tem83

For France, this may reﬂect the general poverty of association life, which was lamented by Tocqueville and later mentioned by Durkheim in The Suicide (1958: 106), Laski (1919: 321), Lerner (1957: 27–32), Kornhauser (1959: 84–87), and many others. However, MacRae (1967: 28–32) and Tilly, Tilly, and Tilly (1975: 40–44) have rejected the thesis of the ‘‘underorganization’’ of French society. Even for Italy, and particularly for southern Italy, it has often been argued that the people are not inclined to create voluntary associations. See Banﬁeld (1958) and the criticism by Pizzorno (1966a: 55– 66). For Finland, I have not been able to ﬁnd any similar complaints about the quantity and quality of association life in the literature.

302

THE POLITICAL MOBILIZATION OF THE LEFT

porally precedes electoral mobilization: Denmark, Sweden, and the United Kingdom. Switzerland, on the other hand, shows an almost one-to-one relation between union members and socialist voters. In all the other cases, electoral socialism at the turn of the twentieth century was already supported by large sections of previously unmobilized voters.84 In Italy, Austria, Belgium, and Germany, there was one member of a left trade union for roughly every two left voters; in Norway and the Netherlands, this number decreases to one for every two and a half; and in France, to one for every three to four. Differences are more pronounced in terms of the ratio between voters and party members. Sweden, Norway, and Denmark are cases where partisan mobilization preceded electoral mobilization (unfortunately, Austrian Socialist Party membership ﬁgures before 1913 are not available). In the ﬁrst ten years of the century there, were three-and-a-half party members for each voter in Sweden, two-thirds of a party member for each voter in Norway, and one party member for every two voters in Denmark. These early differences tended to decrease over time, but they never disappeared. During the 1900–1940 period, Norway, Denmark, Austria, and Sweden had one socialist member for every two voters. In contrast, Italy, France, Germany, the Netherlands, and the United Kingdom (Belgium and Ireland could be added on the basis of period estimates) had only one member for every six to ten voters. These data can be seen from two different perspectives. In those cases in which the ratio of members to voters is low, it could be argued that electoral mobilization did not necessitate a strong organizational infrastructure; that electoral socialism had different sources of appeal; and that votes were indeed, from this perspective, organizationally ‘‘cheap.’’ Alternatively, one could read the data as indicating the varying strength of the electoral rooting of socialism as deriving from the strong organizational bonds of its electoral constituency. In fact, in the long run, the kind of socialism that was solidly encapsulated proved more electorally stable and more socially hegemonic than weakly encapsulated socialism. This is not to say, however, that it was necessarily stronger in electoral terms. 84

Shalev and Korpi (1980: 54–55) come to the conclusion that although unions and working-class parties are different things, ‘‘unionization may in fact precede and support political organization and mobilization’’ (p. 55). They argue that this is one of the key explanations of American exceptionalism, i.e., that the weakness of the working-class party in the United States is linked mainly to the weakness of trade unionization. They cite as a counterexample the British case, where the party resulted from the unionized workforce. However, as a general thesis, this is wrong. The authors may have been too strongly inﬂuenced by a British-versus-American example, where their thesis works, and also by the experience of the Scandinavian countries, where, as we have seen, trade unionization tended to precede working-class party strength.

STRUCTURING AND MOBILIZATION

303

TYPES OF SOCIALIST MOVEMENTS In Figure 6.9, I have tried to summarize the picture discussed so far, charting countries by their mean level of partisan and corporate membership density and indicating their mean left vote for the 1918–1985 period. Those countries that experienced a signiﬁcant Communist split of the working-class movement are indicated by a large circle; those with the latest collective bargaining development by a smaller circle; and all the others by a square. In Austria, Denmark, and Sweden, mobilization was high in all three channels, the strong electoral support for the socialist movement mirroring strong organizational penetration of the electorate. The Norwegian and German means in each channel are close to the overall means. Germany shows average mobilization in the corporate channel and weak mobilization in the partisan channel. This is interesting, for it is at odds with the usual image of German social democracy as being the most strongly organized and solid socialist example.85 This was true in the very early phase of socialist development, between the 1860s and 1890s, but it faded at the turn of the century. Perhaps it was based more on strong centralization than on actual levels of partisan mobilization. The two European early industrialized countries – Belgium, the United Kingdom – show mean corporate mobilization but lower than mean partisan and electoral mobilization. Finland is the opposite, as already noted. At the other extreme are those countries where socialist mobilization remained below the overall European mean level in corporate and partisan as well as electoral mobilization; such as the Netherlands, and Switzerland. What characterizes these countries comparatively is simply their class cleavage undermobilization in all dimensions. The correspondence between organizational density and electoral strength is weakened by the three cases of France, Finland, and Italy, where the left is on average above 40%, although it also shows comparatively low levels of membership density. These are three of the four countries that experienced profound and lasting communist splits. In none of these countries where the communist split was relatively successful was membership density high or collective bargaining an early development. Several times in this chapter, the different ‘‘logics’’ of corporate versus political socialism, the lack of any strong association between the two, and

85

To quote only one source: ‘‘with a membership of more than one million in 1914, the SPD was the largest and best organized Socialist party in Europe.’’ Breitmen (1981: 6).

304

THE POLITICAL MOBILIZATION OF THE LEFT

Figure 6.9. Organizational density of the class cleavage (1918–1985). the preeminence of the latter over the former have been underlined. In reality, these ﬁndings point to qualitatively different types of early socialist movements characterized by different mixes of organizational mobilization. I have divided these into four basic types, reported in Table 6.14. Union lefts were more highly mobilized in the corporate channel than in the electoral one. Encapsulated lefts were densely organized and strongly interlinked between all the channels. Ideological lefts were far more successful electorally than they were organizationally strong. Finally, undermobilized lefts simply lagged behind in all the mobilization spheres. Such typological characterizations suggest a number of considerations and point to questions that need to be answered. First, this research underlines the need to separate interpretations of corporate versus partisan versus electoral mobilization. The form in which the early consolidation of speciﬁc organization in both channels of interest representation, and in which voting and direct political action took place,

STRUCTURING AND MOBILIZATION

305

Table 6.14. Types of left movements by organizational and electoral mobilization

was important for later waves of mass political mobilization. The initial organizational pattern was more inﬂuential on corporate and partisan mobilization than it was on electoral mobilization. In most cases, from the beginning, electoral socialism extended far beyond the politically mobilized sectors of the labor movement. The few exceptions to this general picture are, however, important; it is exactly where early corporate and partisan mobilization preceded or paralleled electoral mobilization that socialism proved in the long run to be more successful, more electorally stable, and, in a word, more ‘‘hegemonic’’ than in the other cases. Yet this should not conceal the fact that an electorally large socialist movement was able to develop without building on other forms of preelectoral mobilization; that is to say, it managed to maintain its strength even though it rested on narrow organizational bonds and bases. Particularly in these cases, however, electoral socialism was, from the beginning, ‘‘trapped’’ with large electoral allies – wanted or unwanted – that were linked to socialism for politico-ideological reasons but were deprived of any solid encapsulation in its organizational network. In a nutshell, the explanation of corporate and electoral socialism need not be the same, and a clear distinction between these two channels helps to reconcile the often widely diverging general explanations of socialist development sub specie of ‘‘economic response’’ and ‘‘political response.’’ Second, as far as electoral overmobilization is concerned – and, in general, for electoral mobilization greater than corporate mobilization –

306

THE POLITICAL MOBILIZATION OF THE LEFT

this seems to have been particularly true (1) where protracted authoritarian rule forced the socialist movement to give priority to the political goals of democratization and liberalization; and, to a lesser extent, (2) where religious and political mobilization were oriented to socialist sections of the electorate that were motivated by rationalistic and anticlerical attitudes. In both instances, electoral socialism was able to attract sections of the middle classes more easily, while, at the same time, it was obliged to relinquish its monopoly over the representation of organized labor. It is important to emphasize that these aspects have traditionally been considered liabilities for socialism, as they have hampered its chances of full mobilization of the working class. However, while these were liabilities for corporate socialism, they were not necessarily so for electoral socialism. This is because what they lost in terms of organized labor support was compensated for in terms of middle-class ideological support. Of course, this resulted in different kinds of socialist movements: These were different precisely because and according to the relationship between electoral, partisan, and corporate support. These considerations conﬁrm the need to distinguish clearly the statebureaucratic environment response from the economy-class environment response in the history of socialism.86 Patterns of economic and industrial developments – the ‘‘maturity of socioeconomic conditions’’ – were more closely associated with levels of corporate class mobilization than with levels of political class mobilization. For the latter, the relationship appears to be inverted. Electoral socialism proﬁted from the strength of the state/ political opposition that it encountered, and grew both earlier and easier in the political sphere where it was forced to ﬁght for goals unattained by inﬁrm or ineffectual liberal-democratic political movements (suffrage extension, parliamentarization, state–church separation, bureaucratic power restriction, antimilitarism, abolition of late feudal remnants, etc.). In contrast, the earlier and easier attainment of these goals, by making mobilization in the interest channel both easier and faster, had the effect, at the same time, of making partisan and electoral mobilization lag behind. When this occurred, the corporate movement did not need to develop a speciﬁc political tool rapidly – because it could lend its growing organizational support to other sympathetic candidates or political organizations – and electoral socialism found it difﬁcult to raise support beyond the groups already mobilized in the corporate channel. Third, reading Table 6.14 along the row of organizational undermobilization, the political fragmentation of the labor movement is seen to 86

On this theme, see the stimulating reﬂections of Szabo (1982).

STRUCTURING AND MOBILIZATION

307

associates clearly with weak corporate, but not electoral, mobilization. More speciﬁcally, in all those cases in which the Communist split proved to be successful in organizational terms – in particular, in France, Germany, Finland, and Italy – the socialist movement rested on a relatively narrow organizational encapsulation along both partisan and corporate dimensions, which was greatly inferior to its electoral mobilization. Does this support the thesis that the prospects for the success of the ‘‘international revolution’’ cleavage were boosted by the weakness of the socialist organizational encapsulation of the working class in the aftermath of World War I? In these cases, the socialist electorate was composed of relatively large sectors that were not organizationally linked to the socialist party and unions, and a massive delayed trade unionization took place during the period of maximum appeal of the Russian Revolution. Organizational competition between the socialists and communists did not take place within large, established corporate movements, but rather within weak organizations subject to the accelerated recruitment of previously nonmobilized (and mostly unskilled) sectors of the working class. Finally, a number of more speciﬁc questions are left open: What were the nonorganizational sources of the exceptional electoral success of Finnish socialism? Why were the Dutch, Italian, and French movements so undermobilized organizationally? Were there non- or prepolitical networks that made it easier for Scandinavian and Austrian socialism to build extraordinarily dense organizational networks? The national literatures do not lack speciﬁc answers; the problem is to make them compatible with a comparative framework. What all these concluding hypotheses call for is a more detailed analysis of institutional developments and political alignments. This will be carried out in the next three chapters.

ORGANIZATIONAL CONSOLIDATION, MEMBERSHIP DEVELOPMENT, AND THE GENERAL MODEL The three organizational factors that I have found to be most closely associated with left electoral development are (1) the modalities of early party consolidation (those of central union consolidation were not signiﬁcant); (2) the relationship between party and unions; and, ﬁnally, (3) the level of organizational-membership density. These three variables are operationalized as follows. Party organizational consolidation is measured on a seven-point scale ranging from the lowest value of 3 (late founding [timing], long centralization process [length], and preceded by unions [se-

308

THE POLITICAL MOBILIZATION OF THE LEFT

quence]) to the highest value of 9 (early party founding; short centralization; before the unions). The party–unions relationship is a three-point scale ranging from the lowest value (2) when the mode of representation is fragmented and there is no interlocking relation between party and unions, to the intermediate value (4) when the party–union relationship is interlocking but the mode of representation is fragmented, and, ﬁnally, to the highest value (6) when the relationship is based on an interlocking linkage and a corporate mode of representation. The values result from the summing of the individual case ranks on the linkage and mode of representation variables. Finally, organizational density is measured by the sum of the ratio of left trade union members and party members over the electorate.87 When entered into a regression with the left vote, these variables yield a multiple R of .459 corresponding to an Rsq of about 21% (see Table 6.15). The weight of the type of relationship between unions and the parties changes drastically over time. It is negative in the early phase, when corporate representation and an interlocking link were not favorable to electoral development, and becomes equally highly positive after World War II. It is as if parties with this kind of relationship with the unions were late in electoral mobilization. The model of early party consolidation is positively correlated with the left vote in every period, and it is the strongest positive factor in the phase before World War II. Finally, organizational density proves a much weaker predictor for each period than it is for the overall series of elections. Before entering the organizational variable into the general model, which includes the determinants discussed in previous chapters, let me ﬁrst brieﬂy report how the organizational aspects relate to social mobilization and cultural heterogeneity. As expected, the relationship between 87

To avoid losing several cases where membership ﬁgures are missing, I have interpolated the missing values. There are only three countries where this is problematic because it concerns not one single missing ﬁgure within a series of reliable data but rather an entire period. In Belgium, socialist (but not communist) party membership ﬁgures are not available from 1900 to 1939. In the following decades, partisan density ranges from the minimum of 2.2 in 1950 to the maximum of 4.3 in 1978. I have arbitrarily set the level for the 1900–1939 elections at 2%. In Ireland, the same problem applies to the 1927–1961 period. Here, however, partisan density is so low that errors can hardly be made in the estimation. Membership ranges between 0.2% and 0.4% between 1965 and 1989; I have set the level for the 1927–1961 period at 0.2%. Finally, we do not have individual membership ﬁgures for the British Labour Party up to 1924. In 1929, partisan density was 0.8%, and it grew to its highest value of 2.7% in 1950. For the few elections between the founding of the party and 1924, I have set the level of partisan density at 0.5%. As is evident, these three are all cases of extremely low and belowEuropean-mean partisan density, and for none of them does the literature offer any reason to believe that this was different in earlier years.

STRUCTURING AND MOBILIZATION

309

Table 6.15. Regression of organizational variables on left vote (beta coefficients)

Note: Number of elections given in parentheses.

social mobilization and corporate mobilization is always positive.88 Social mobilization is a better and more direct predictor of corporate socialism than it is of electoral socialism, while the relationship of social mobilization with partisan density is neither strong nor systematic in sign. Cultural heterogeneity negatively inﬂuences all aspects of class cleavage mobilization. As expected, the negative association is higher for left trade union membership than for overall union membership. It is also always extremely high for partisan membership. This supports the idea that in culturally segmented societies partisan mobilization in speciﬁc political organizations is either less necessary or more difﬁcult due to the already high and functionally equivalent membership mobilization of all the various subcultural associations and organizations that constitute the bricks and mortar of the cultural pillars of the society. To enter the signiﬁcant aspects of the organizational development into this general model, I combined the three factors of membership density, union–party relations, and party organizational consolidation into a single standardized index. I then ran a general regression in which the left vote was predicted on the basis of the four macrofactors discussed so far: the indices of social mobilization, cultural heterogeneity, enfranchisement pattern, and organizational development The results are reported in Table 6.16, which should be compared with Table 5.11 in the previous chapter, where the same analysis was performed without the organizational dimension. This new variable does not add to the overall explained variance, which remains at around 70%. 88

In contrast to this evidence, Wallerstein has found a negative association between the size of the labor force and its degree of unionization, arguing that it is not surprising that the larger the potential constituency of wage earners, the more incomplete is the process of their recruitment. See Wallerstein (1989). However, he does not use, historical data to prove his point.

310

THE POLITICAL MOBILIZATION OF THE LEFT

Table 6.16. Social mobilization, cultural heterogeneity, enfranchisement, organizational model, and left vote (beta coefficients)

It does, however, inﬂuence the weight of the other factors and the variance explained in each period. The organizational aspects are quite unimportant in both the ﬁrst pre–World War I period and the ﬁnal post-1965 period. However, they do play a signiﬁcant role in the interwar period and in the ﬁrst post–World War II period, between 1918 and 1965. In this central phase, they increase the explained variance considerably, from 62.6 to 71.1 and from 55.6 to 60.2. Thus, while the sociostructural variables maintain a constant important role, the enfranchisement pattern is crucial in the ﬁrst period and declines progressively in the others, the impact of cultural heterogeneity grows over time, and the organizational dimension impacts mainly in the second and third historical periods. In this chapter, much descriptive space has been devoted to left organizational development, as I believe this to be the least developed ﬁeld in the comparative history of socialism. However, these conclusions suggest that for further investigation and more accurate control, organizational data need to be complemented by data concerning the timing, easiness, and modalities of entry of the socialist movement into the political system. To test macrohypotheses about the historical development of the European class left requires a model that links sociocultural conditions, organizational mobilization instrumentality, and structure of political opportunities. It is to the latter that I turn in the following three chapters, where the argument will acquire a more qualitative ﬂavor. It is likely that the speciﬁc pattern of institutional incorporation in the two channels decisively inﬂuenced the extent to which the two paths, corporate and electoral, were separated or remained closely linked. For instance, the earlier collective bargaining, trade union recognition, and so on occurred, the less need there was for unions to rely on politics and the more independent the paths could become in corporate and electoral channels and organizations. At the same time, where the political channel remained closed longer, one would expect a politicization of the organiza-

STRUCTURING AND MOBILIZATION

311

tions in the corporate channel. Generally speaking, the earlier and the more thoroughly the institutional barriers were separated in the two channels, the more we would expect the corporate and electoral paths to be differentiated and less dependent on each other.89 The respective timing of institutional incorporation may also have had a signiﬁcant impact on the prevailing ideological orientation of the early socialist movements. The stronger the trade union movement before party development and the smoother its institutional role and development, the less likely it was that an orthodox Marxist orientation would prevail within the party and in the party–union relationship and the more likely a party dependence on trade unions or a strong corporate cross-linkage with a moderate Marxist platform would be. Thus, weak and unrecognized unions may become instrumental in the fragmentation of the political representation of the class left. 89

See Ebbinghaus (1992: 5–6) for a discussion of this problem.

7

POLITICAL INTEGRATION

he socialist movement was an outsider for the conservative or liberal groups that controlled the political system in the second half of the nineteenth century. To establish itself, it had to penetrate the system and it faced varying degrees of resistance in this enterprise. The ease with which it came to be recognized as a legitimate political actor and the barriers it had to overcome to achieve this status – in a nutshell, the pattern of political integration – inﬂuenced its electoral strength, organizational cohesion, and radicalness of stance. Thus, the extent to which working-class political movements felt they could pursue and achieve their goals and values within the existing framework of state and political institutions was shaped by the learning experience deriving from the responses of the dominant groups and established elite when facing their demands. The goal of this chapter is to assess the European variation in institutional ‘‘openness’’ to the new claimants’ demands and organizational efforts, assuming that political alienation was as important as, if not more so than, social dislocation and economic deprivation in setting the context of the political response of new lower-class claimants. I compare the earliness and the level of left institutional integration, concentrating on four institutional obstacles to entry into the dominant pattern of political competition:

T

1. The extent to which demands made by the political movement of the working class were met by the established elite and institutions’ use of nonpolitical means of confrontation, that is, direct legal and administrative repression and harassment. In other words, I consider to what 312

POLITICAL INTEGRATION

313

extent and with what efﬁcacy the state, as an administrative and policing machine, was used against the rising working-class movement. 2. The extent to which the institutional regime was liberalized, that is, the extent to which the resources of power and inﬂuence in the administrative and bureaucratic machinery of the state, or in the social and market relationships of the society, were balanced by the parliamentary inﬂuence and negotiation based on electoral weight. The parliamentarization of the executive opened a new structure of opportunity, allowing the socialist movement to concentrate part of its resources and efforts on directly inﬂuencing the executive, rather than on nonparliamentary confrontation with employers, as well as with administrative, bureaucratic, and legal obstacles. 3. The types of obstacles set up to impede the fair parliamentary representation of the electoral strength of the movement. Even liberalized regimes resisted, with differing styles and vigor, the full transformation of the left movement’s electoral force into parliamentary inﬂuence by various mechanisms of electoral misrepresentation. 4. The kinds of obstacles that were used to block socialist parties’ access to executive responsibilities. Stateness and repression, responsible government, fair representation, and executive power access are linked into a syndrome pattern. In the following pages, I deal with each of them separately, leaving a global description of the pattern of each country for the ﬁnal section. The losses in case-conﬁguration power resulting from the analytical sectioning of the global process are compensated for by the increased accuracy of the comparative analysis. The qualitative nature of this in the following chapters is due to the type of information and material discussed, as well as to the logic of the analysis. Multivariate techniques isolating the role of individual variables are less appropriate when a potential syndrome effect exists: that is, when it is the speciﬁc combination that establishes a more or less favorable condition.

STATE RESPONSE STATENESS The legal and administrative repression and harassment of the workingclass political movement in the nineteenth and early twentieth centuries highlight the role and the resources of the state. Recent literature has

314

THE POLITICAL MOBILIZATION OF THE LEFT

emphasized the importance of the state’s tradition and role in the process of political modernization, linking it to the alliances between social groups that have determined the liberal, authoritarian, or totalitarian outcomes of the modernization processes,1 to the sources of revolutions and their outcome,2 to the process of social group formation,3 and to collective action.4 Here, I focus on the state as a set of resources to which established and dominant groups could have recourse in their attempt to obstruct the emerging socialist movement. More precisely, the calculation of the cost of internal repression or tolerance made by dominant groups depended, ﬁrst of all, on the availability of the means of repression, that is, on the availability of a bureaucratic tradition and bureaucratic machinery that could be exploited to oppose new movements successfully. We can put the question in the following terms: How did the nature of state formation (external challenges, boundary consolidation, centralization, development of apparatuses, army, etc.) inﬂuence the relative costs of tolerance and repression of the socialist movement? Levels of administrative and legal harassment are linked to stateness but are not deﬁnitely determined by it. Where stateness was weak, a repressive strategy was not available or was likely to be very costly; where it was strong, such a strategy was possible, even if other variables may have inﬂuenced the willingness of the established elite to choose it. The ﬁrst step is to assess the level of stateness independently of the nature of the institutional and political regime.5 The second is to evaluate the extent to which existing state resources were actually used against the rising socialist movement. In dealing with the ﬁrst problem, I will not draw on the ideal type of experience of a few well-known countries (generally France versus Britain versus Prussia). Such ideal or polar types are not effective when dealing with a large comparative design classifying all available cases. It is more fruitful to break the concept into a number of constitutive dimensions and to compare individual cases along each of them, and to reconstruct their overall relative positions later.6 Stateness indicates the extractive, regulative, and repressive resources that each state holds vis-a`-vis society. This accumulation of resources is 1 2 3 4

5 6

Moore (1966). Skocpol (1979). Katznelson (1985). Birnbaum (1988). See also Birnbaum and Badie (1982: 189–190). The third part of the book is devoted to an ideal-type analysis of types of states. See Nettl (1968). A decomposition strategy overcomes the criticism that the characterization of stateness at the macrolevel does not take into account possible variations among sectors within a given state; see Atkinson and Coleman (1989: 49).

POLITICAL INTEGRATION

315

primarily the result of the pattern of state formation. The latter identiﬁes the process of political, economic, and cultural uniﬁcation of the central dominant elite. On state formation, the comparative-politics literature generally agrees on the following propositions: 1. If a state-based legal order and a consolidated boundary clearly predate the formation of a strong central bureaucracy, then the repressive powers of the state are more likely to remain fairly independent of the requests and the direct control of the executive. 2. If the consolidation of the state’s basic institutions is uncontested or only relatively so, then the pressure for a centralized administration of law-and-order apparatuses will be weaker and the legal tradition will be less likely to be state-centered. 3. If the creation and consolidation of the state structures is early enough and does not involve processes of large-scale political mobilization or violent popular resistance, then the status apparatuses are likely to be less politicized and their repressive means are likely to be more responsible to local judicial bodies. 4. If the central bureaucracy never consolidates, or if it consolidates after representative institutions have been introduced, then the state apparatuses are not a resource for the repression of newcomers.7 Taking these points together, European countries can be arranged from a minimum to a maximum of stateness from left to right, as in Table 7.1. The end result of the state formation process consists technically of the development of the bureaucratic, ﬁscal, and military apparatuses. Therefore, the ﬁnal level of stateness of the political processes can be identiﬁed along four dimensions: 1. Creation of the organization for the mobilization of resources: bureaucracy and tax burden. 2. External consolidation of the territory: army. 3. Maintenance of internal order: police and judiciary. 4. State activism in regulatory activities and in economic and social intervention. It is, unfortunately, difﬁcult to assess directly the last of these dimensions (the extent of state intervention in the economy and society). Whereas 7

I have reformulated theses that have been widely discussed in the literature of state formation and political development. In particular, see the articles in Tilly (1975) and Dahl (1971: 48–49).

316

THE POLITICAL MOBILIZATION OF THE LEFT

Table 7.1. Modalities of state formation in Europe

it would be relatively easy to build an ideal type placing ‘‘Colbertism’’ and ‘‘cameralism’’ in opposition to ‘‘laissez-faire’’ and ‘‘decentralized, amateurish public administration,’’ it is impossible to set up a systematic charting of cross-country differences along this dimension.8 As a result, for the comparative appreciation of each country’s extractive resources, I propose to look at the following aspects: 1. The tax ratio, that is, the general governmental taxes as a proportion of the gross domestic product. 2. The tax centralization, that is, central governmental taxes as a proportion of general governmental taxes. 3. The general central administration personnel as a percentage of the population. For an evaluation of the means of control over external and internal order, I will examine: 1. The military personnel as a percentage of the male population between twenty-two and forty-four years old (excluding war period ﬁgures and post-war demobilization ﬁgures in order to account for the belligerent or non-belligerent status of countries during World War I). 2. The police and judiciary personnel as a percentage of the total population (the judiciary calculated as personnel employed by the ministry of justice).

8

Mann (1993: 358–395) provides several tables of data about what he calls the ‘‘size’’ of the state and discusses this dimension to some extent. However this is done for European major powers only.

POLITICAL INTEGRATION

317

I have computed the ﬁgures for the ﬁve indicators, averaging the census decade data of the 1880–1920 period.9 In Table 7.2, the average value of each dimension is reported, together with the mean of the ﬁve indicators as an aggregate index of stateness. Note that ﬁgures are expressed in zeta scores of the distribution.10 The ﬁgures show that the ﬁve dimensions are not always homogeneous. Switzerland, Norway, Denmark, and Sweden have values below the Western European mean in all the ﬁve dimensions analyzed. Austria and Belgium tend to be homogeneously above the European mean, without, however, reaching very high levels in any of the ﬁve dimensions. The United Kingdom oscillates in each dimension close to the European mean, while the Netherlands has medium values in all dimensions with the exception of tax centralization, where it appears to be the most tax-centralized country of Europe at the turn of the twentieth century. Italian stateness, by contrast, looks quite unbalanced. While Italy is more highly bureaucratized than any other country, with an extensive central bureaucratic personnel and high tax centralization, it remains below the mean in other indicators (unfortunately, police and judiciary personnel data are not available). Germany’s stateness levels are very high in terms of repressive resources and, in particular, in terms of internal law-and-order means, while they are lower in terms of central resource extraction. French stateness is strong and is above the mean homogeneously along four of the indicators, falling below it only in terms of internal law and order. In the cumulative stateness index, ﬁve countries stand out as characterized by a high level of stateness of their political processes: Italy, France, Austria, Germany, and Belgium. Three countries present intermediate values that come very close to the European mean – the Netherlands, Ireland, and the United Kingdom – and ﬁve fall clearly below the European mean level, characterized by weak levels of stateness in their polity: Finland, Denmark, Sweden, Norway, and Switzerland. If we compare these ﬁndings with the usual descriptions of the strength of the state, a few surprising points emerge. For example, the ideal type of the weak state, the United Kingdom, does not look so weak when it is merged into a Europewide comparison. It is far weaker than France and Germany, but it is in no way the model of the weak state. Presumably, the impressive 9

10

Data are elaborated from Flora, Alber, Eichenberg, Kohl, Knaus, Pfenning, and Seebohm (1983). Note also that not all ﬁgures are available for all indicators. In computing the combined index, I have set the missing data at the 0.0 level, that is, to the overall distribution mean.

318

THE POLITICAL MOBILIZATION OF THE LEFT

Table 7.2. Dimensions of stateness (1880s–1920s) (average standardized scores)

⫺ ⫽ data not available. ( ) ⫽ data refer to the 1920–1940 period.

correspondence of British ranking in all dimensions contributed to forge the image of a weakly state-driven polity. Beyond the unsurprising case of Switzerland – the real prototype of the low-stateness polity – the low levels of state capacities in the Scandinavian countries are to a certain extent surprising. Sweden and Denmark represent cases of many centuries of dynastic rule and of dynastically driven state and nation building;11 Denmark, in particular, has often been considered the most ‘‘Prussian’’ of the Scandinavian countries. The countries share deep-rooted traditions of centralized bureaucratic rule that have been inherited from their long-standing absolutist regimes.12 However, even in Denmark and Sweden, the resources controlled by the state at the end of the nineteenth century were far lower than those of any other European nation.13 11

12 13

If the strength of a state is to be seen as resulting from both its capacity to enforce authority and the existence of sentiments of national solidarity – as suggested by Linz (1973: 34–35) – in Scandinavia the identiﬁcation of the population with the nationstate is high and unchallenged. Even in the peripheral countries that gained independence only in the twentieth century – Finland, Norway, and Iceland – the sense of national identity was fairly consolidated long before. Rokkan (1966b: 74). See also the various research project reports in Stra˚th (1988b). The strong persistence of a relatively independent nobility, the maintenance far longer than anywhere else of older estate principles, the persistence of strong conciliator or collegial forms of rule, and the peculiar bureaucratic separation between policy-making

POLITICAL INTEGRATION

319

The intermediate position of the Netherlands comes as no surprise, as state formation in the real sense occurred after 1785; it is only from this time on that one can speak of a central government and direct control over citizens. At the same time, the extent of bureaucratization and centralization brought about by the Napoleonic conquest and reform should probably not be exaggerated in a comparative perspective. In this particular case, the Netherlands may well point to a problem typical of historical systematic comparisons. A process that is unquestionably clear in historical terms with respect to the previous situation in the country – and that is strongly emphasized as such in a developmental perspective by historians – may well remain relatively modest in a cross-national comparison.14 Comparing the levels of stateness indicated in Table 7.2 with the modalities of state formation reported in Table 7.1, it is clear that the countries that never consolidated as a unitary state (Switzerland), those with early and unchallenged consolidation (the United Kingdom, Sweden, and Denmark), and those with late and gradual consolidation (the Netherlands and Norway) are all low- or medium-stateness countries. The only exception is Belgium, whose level of stateness is, however, only moderately higher than average. The countries with early consolidation, but with considerable resistance (France and Austria), and those with late and violent consolidation (Italy, Germany, and Austria)15 all tend to be highstateness countries also because they were deeply involved in the nineteenth-century continental warfare. In this group, Finland and Ireland are the exceptions, both having low stateness in the context of a late and violent consolidation. This is not surprising considering that we are dealing with peripheries of large empires (the British and the Russian) whose state structures and resources were largely imported externally (and later withdrawn) rather than generated internally. The process of state formation and its end result in terms of stateness made available potential resources that were not necessarily used or directed against the socialist movement. In a situation of low stateness, the calculus of the established elite regarding the potential costs of repression versus toleration was likely to be negative. That is, dominant groups could hardly embark on a repressive strategy based on the state. Strong stateness

14

15

boards and executive agencies have been invoked as an explanation of this end result. See Daalder (1995: 119–120) and Anderson (1974: 173–191). On the debate about the levels of bureaucratization and centralization in the Netherlands after the Napoleonic war, see Daalder (1990: 19–20). The classiﬁcation of Austria depends on whether one considers the Hapsburg Empire or the republic.

320

THE POLITICAL MOBILIZATION OF THE LEFT

made a state-repressive strategy available, and this availability may have inﬂuenced the calculations of dominant circles. However, in no way was such a strategy either mandatory or without other alternatives. It is therefore necessary to consider the extent to which the state apparatus was actually used against the socialist movement. REPRESSION AND TOLERATION Some antisocialist hostility and repression existed in all European countries. National accounts, although rich in information about this hostility, rarely make comparative assessments of its relative level.16 My attempt here is to systematize in relative terms the various kinds of information available in this literature about the fundamental freedoms of the press, association, and the strike. The achievement of freedom of the press across European countries can be divided into three fundamental steps: the abolition of censorship, the later progressive softening of postpublication prosecution, and, ﬁnally, the abandoning of any sort of legal, and possibly insidious, mechanisms by which the socialist press was made less viable and more difﬁcult to circulate. Table 7.3 reports the approximate dates when these barriers were overcome in different countries and indicates if any important setbacks occurred after World War I. By the end of the Napoleonic wars, Norway was probably the most liberal country in this respect. By the beginning of the 1850s in the United Kingdom, Sweden, the Netherlands, Switzerland, and Belgium, censorship had technically been abolished, although a wide variety of administrative techniques of control and postpublication prosecution were still in existence. At the other end of the continuum, there were still major restrictions on press freedom in Germany, Austria (to a lesser extent), and Finland (to an even lesser extent) on the eve of World War I. In Germany, censorship had ﬁrst been banned in 1849 in most of its states and again by a press law of 1877 in the Reich, but even so, extensive, severe prosecutions and sanctions continued until the end of the empire. 16

For this reason, in this section I have relied more on comparative sources than on standard national accounts. The only comparative work I am aware of that is devoted to the analysis of political repression in nineteenth-century Europe is Goldstein (1983), which I have used extensively even though Goldstein stresses the commonalities among cases rather than the relative differences. Other works on the history of labor have been very useful: Lafferty (1971), Geary (1989), Abendroth (1965), Kassalow (1963), Castles (1978), Droz (1972), Galli (1976a), Greene (1971), Kendall (1975), Kriegel (1973), Landauer (1959), Lindemann (1974), E.N. Mitchell and Stearns (1971), Piro and Pombeni (1981), and Steenson (1991).

321

Table 7.3. Press, association, and strike freedoms: repression and harassment

322

THE POLITICAL MOBILIZATION OF THE LEFT

Similarly, Austria abolished press censorship in 1867 but continued to prosecute opposition newspapers with a socialist or nationalist orientation, and the sentences were particularly severe in the 1877–1880 period. Finland remained under the strong censorship regime of Tsarist Russia, softened to some degree by the relative level of its autonomy, but constantly subject to further restrictions whenever the Russian autocracy felt politically menaced or needed to gain a stronger hold on Finnish affairs. France and Italy are located in an intermediate position with respect to the other countries. They formally abandoned censorship in 1830 and 1848, respectively, but in both cases administrative and press restrictions continued to be common practice until near the end of the century. Obviously, in Germany, Austria, and Italy, the collapse of democracy was a sharp setback to this process, bringing about unprecedented levels of censorship. Even in Finland, although to a lesser extent, the political upheavals from the civil war of 1918 to the crisis at the end of World War II led to considerable limitations on the press. Censorship and postpublication prosecution were not the only forms of harassment for the opposition press and, in particular, for the socialist press. There were several indirect, less insidious, but still effective forms of ﬁnancial control. For instance, publishers could be asked to pay a security bond and caution money, and newspapers to pay special taxes. The use of these measures, sometimes left to administrative discretion, often damaged the socialist press since they led to higher costs and higher prices. Germany and Austria, once again, used ﬁscally discriminating measures most frequently and abolished them only as late as 1874 and 1899, respectively. In France, similar provisions also continued until 1881, being imposed and relaxed according to the liberality of the regime. They were not totally abolished in the United Kingdom and the Netherlands until 1861 and 1869, respectively.17 In conclusion, the general picture suggests that – apart from the exceptional case of Norway – Sweden, the Netherlands, Belgium, Switzerland, the United Kingdom, and Denmark had abolished strong administrative control of the press by the middle of the nineteenth century and that there were no other major setbacks from then on. The liberalization of the press occurred before the development of early socialist movements, and the latter were therefore free to develop in a fairly liberal press environment. Italy and France were more resistant in giving up forms of press control, particularly those affecting the socialist and anarchist press 17

Goldstein (1983: 41).

POLITICAL INTEGRATION

323

and publications. Finally, in Germany, Austria, and Finland, the early socialist movement had, in contrast, to combat the strong control over the press by the state bureaucracy and to continuously invent methods by which to avoid these restrictions and prosecutions. This is clear from the behavior of the socialist press and publications in these countries: Frequent changes in titles, places of publication, and so on were necessary to enjoy the initial advantages of the formally liberal press law and to delay the hindrance that inevitably followed a certain period of activity. Almost all Western European states severely limited the freedom of political assembly and association before 1848. The only exceptions were Belgium, under the liberal constitution of 1830; Switzerland, after the canton revolution of 1830 (and again in Switzerland in 1848); and the United Kingdom, where only some kinds of obstruction were practiced up to 1848. After the 1848 revolutions, these freedoms were granted in Piedmont (later Italy, in 1859), the Netherlands, and Denmark. The subsequent constitutional guarantees of these rights were granted in Austria (1867) and Germany (1869). However, constitutional and statutory guarantees of freedom of assembly and association were often suspended or discontinued (Germany in 1878–1890, Austria in the 1880s, and Italy in 1893–1894 and 1898) by the introduction of exceptional legislation in periods of high political tension. Moreover, even after ofﬁcial recognition, actual restrictions of a more administrative and police nature were continued in many cases. Rather than follow the evolution of constitutional and administrative practices, I concentrate on the actual record of what happened to the basic rights and claims of the socialist movements, in both their political and union components, by investigating the following issues: (1) Were unions, union activities, or political parties ever banned? (2) When, approximately, was this ban repealed in constitutional or legal terms? (3) When did the severe administrative harassment of unions and party activities actually end? (4) When were unions recognized by both the state and the employers as legitimate representatives of employees? (5) Finally, were there any setbacks in these processes? In Table 7.3, all the relevant information has been concisely synthesized. However, some explanation is needed, and the following sections describe the reasons for my choices. In Norway, a law banning or otherwise making illegal the association activity of the unions never existed. The most serious episode of repression was the Thrane movement in 1851. Once the earlier strong dependence on Denmark was replaced by the milder one on Sweden (in 1814), Norway developed into a fairly liberal, nonviolent state. A few minor confrontations occurred after the 1850s, and historians remind us that during a labor

324

THE POLITICAL MOBILIZATION OF THE LEFT

conﬂict in 1881 one worker was killed,18 but in general the police and the army remained in the background of social confrontations. During the radicalization phase of the labor movement in the years before and after World War I, renewed nervousness and alertness on the part of the state led to soldiers presiding over various labor conﬂicts (especially in the northern areas of the country), but this did not lead to any violence. Similarly, the life of the socialist political movement was by any comparative standards relatively easy. There was little obstruction, and no signiﬁcant cases of repression, conviction of leaders, banning, or other administrative restrictions have been reported by historians. Switzerland presents more than one similarity with Norway. The canton constitutions of 1830–1831 and the federal constitution of 1848 formally recognized the freedoms of press, association, and political action, and unions were never formally outlawed. In Switzerland, too, the socialist movement had precursors in the so-called democratic movements of the 1860s.19 At the same time, the early unions and socialist organizations had to deal with a ‘‘minimal’’ state in terms of centralization and resources even after the 1848 and 1874 constitutional revisions. In fact, the army was of a militia type created by conscription, and it was rarely called on to maintain law and order. The executive branch of government was weak, not only at the federal level but also at the canton level as a result of the division of powers between cantons and municipalities. Power was exercised mainly through associations and, above all, through groups of prominent citizens:20 an oligarchic rule that could not, and probably did not need to, rely on state regulations or state apparatus to defend its socially prominent position. Between 1870 and World War I, unions and socialist parties were harassed rarely, and the episodes of state (canton) direct intervention and repression were unusual and hardly need to be mentioned; casualties resulting from clashes with the police or soldiers were truly exceptional. Most of the repressive measures that did take place focused on anarchists and radical socialist refugees and their activities and resulted from pressure from neighboring major powers (France and Germany in particular) that were eager to persecute internal opponents even outside their national boundaries. The frequency with which such requests were made and the infrequency with which they were complied with by the Swiss authorities indicate the liberal internal climate. 18 19

20

Terjesen (1990: 12). See Schaffner (1982), in particular pp. 144–197, where the author moves from the analysis of the Zurich canton case to that of other cantons. Jost (1967: 203–219).

POLITICAL INTEGRATION

325

The image of a nonrepressive or rarely repressive history of British labor may be misleading, considering the many episodes of sharp clashes and repression in the ﬁrst half of the nineteenth century. The Combination Acts of 1799 and 1800 forced labor organizations underground. In 1794 and 1817, habeas corpus was twice suspended during the ﬁercest phases of political confrontation. Even after the repeal of the Combination Acts in 1825, for many years workers remained liable for combination under common law, which prohibited strikes, and unions continued to be harassed. This situation ended only in about 1875, when the Conspiracy and Protection of Property Act removed the taint of conspiracy from trade disputes and ensured that violence or intimidation would be dealt with under the normal course of law. In addition, the Employers and Workmen Act (1875) established equality of contract in the employment relationship, replacing the paternal control of the Master and Servant Law; the Trades Disputes Act (1906), which became known as the ‘‘charter of trade union liberties,’’ ﬁrmly established the legal immunity of the unions and the right of peaceful picketing; ﬁnally, the Trade Union Act (1913) reversed the Osborne judgment of 1909, which had prohibited the use of union funds for political purposes, and thus legalized the ﬁnancial support of the Labour Party by the unions.21 Many other legal issues were at the center of union activities and battles in the period 1870–1920. The ﬁnal decisions concerning trade union support for autonomous and independent labor representation in Parliament was precipitated by a court decision in 1901 that declared the Railwaymen’s Union liable for the actions of its members and ﬁned it £40,000. In Britain, therefore, the fate of labor was regulated by a series of legislative enactments, and the law was historically the central arena of struggle. British labor did dispute the laws that limited, constrained, or repressed its actions, but this was done in most cases by using or addressing the law itself; for this reason, legal issues were often the cause and the occasion of political mobilization.22 It should be added that the state was mainly identiﬁed with Parliament and the law also because bureaucratic centralization was very weak and, until the late nineteenth century, the 21 22

Price (1990: 7). The role of the courts needs to be qualiﬁed, however. Hattam argues that the difference in U.S. and United Kingdom working-class formation and political strategy has much to do with the role of the courts and the difference in state structure. While the legislative victories of Labour produced signiﬁcant changes because the courts were less powerful and allowed the government to regulate the labor market, in the United States, where more powerful courts could not be challenged by political power, legislative victories resulted in little change. Hattam (1992: 153–187).

326

THE POLITICAL MOBILIZATION OF THE LEFT

administrative machinery was relatively minor; legislation was administered and enforced by local bodies, and even the police were controlled locally. The greater bureaucratization of administrative structures, the professionalization of the municipal governments (the ousting of unpaid gentry), the diminishing role of voluntary associations in providing social needs and their progressive substitution by large insurance companies, and the state responsibility for education (acts during 1870–1880), and so on – to sum up, a more interventionist role of the state – developed in parallel with the labor movement and, not infrequently at labor’s own request. In conclusion, direct repression of the labor movement occurred early, preceding mass political mobilization; it was directed mainly at the market arena activities (strikes, coalitions); and it was dominated by a liberal ideology and not by the repressive political fears of the state. The state as an autonomous apparatus was as invisible in practice as it was underconceptualized in theory. The early socialist movement came to consider existing institutions mainly in terms of viable channels for its grievances. At the same time, the absence among the dominant elite of a speciﬁc conception of the state, and of its role and mission, made them perceive such grievances less as a global challenge than as speciﬁc requests. In Sweden, the eighteenth-century laws banning union organization were dismantled between 1846 and 1864 (together with the restriction on employers to increase wages). However, no extensive form of labor organization came into being in Sweden before 1870. Legislation affecting workers’ organizations and the attitudes of the employers and authorities were certainly less hostile than in many other European countries. The governing authorities sought to avoid conﬂict with the socialist movement and also to reject its most radical demands by a policy of moderate concessions.23 Unions and socialists were not entirely satisﬁed with these concessions, but they found it difﬁcult to launch a general mobilization and attack. The only signiﬁcant dispute that I have been able to identify took place in 1876, when a workers’ movement at the Sundswall lumberworks, which had struck for higher pay and better working conditions, was repressed by military troops. By 1890, the rights of organization and action were generally recognized, and up to 1931 no worker was ever killed in a street or workplace dispute. The last and most notable piece of antiunion legislation was probably the Akarp Law (passed in1899), which prohibited striking workers from threatening or physically hindering strikebreakers. Legislation against workers’ organizations was passed with difﬁculty 23

Examples of these moderate concessions in response to radical requests are presented in Simonson (1990: 101–102).

POLITICAL INTEGRATION

327

because the farmer-dominated lower chamber normally opposed it. At the same time, the apparatus for national control and repression (the police, for instance) was not centralized. Instead, forces were set up through a combination of state police and local constables who were appointed by the local farmers. Courts, too, combined central and local features, with judges appointed by the state and jurymen appointed by the free farmers from among their own ranks. In evaluating the attitude toward the state in Sweden, one should also remember the existence – before the working class started to organize itself – of local institutions allowing for some level of democratic participation, institutions that traditionally resisted any central authority with which it did not seem possible to integrate or be represented. On the other hand, however, it should be pointed out that in Sweden the right to engage in political propaganda and agitation was fought far more ﬁercely than the right to use corporate activity. Throughout the 1880s, socialist political recruitment and propaganda met with considerable police harassment, and at the beginning of the 1890s there was still substantial ofﬁcial opposition to socialist meetings. In reality, Sweden – like other nonliberalized regimes – tended to recognize, and even to sustain more openly, the corporate representation of workers’ interests (in view of the social peace resulting from this recognition and institutionalization) rather than their political representation. There is some disagreement among scholars of the Danish socialist movement as to the degree of repression it faced.24 Unions and the right to strike were legalized earlier in Denmark than in Sweden, and their legal status remained unchallenged from 1848. Administrative harassment was suspended at more or less the same time, at the beginning of the 1890s, when the police and the courts softened their considerable hostility to the unions. The state opposition to the political wing of the socialist movement was, however, more pronounced in Denmark than in the other two Scandinavian countries. In the 1870s, the Danish Section of the First International, founded by the Lassallian Louis Pio in 1871, was severely repressed. Its leaders were sentenced to several years in jail in 1872, and in 1873 the supreme court actually banned the International Workers’ Association of Denmark (IWAD) and actively persecuted the radical leadership of Pio, Brix, and Geleff. Given this situation, the movement developed in a more moderate direction. Its legitimation was therefore easier, but, 24

Galenson underlines the similarities among the Scandinavian experiences, arguing that early socialism developed as craft socialism, without any signiﬁcant state repression and opposition. Lafferty disagrees with this, claiming that repression of the early socialist movement was greater in Denmark than in the other two Scandinavian states. Galenson (1952a) and (1952c) and Lafferty (1971: 146).

328

THE POLITICAL MOBILIZATION OF THE LEFT

according to Lafferty, ‘‘the terms of entry were clearly those of the authority.’’25 In the Netherlands and Belgium, the unions and early socialist movements developed in a relatively tolerant, if not fully liberal, climate. Early legislation against strikes and unions existed in both countries from 1815 (the Netherlands) and 1831 (Belgium) on. Belgium slowly legalized union activities in 1866–1867, repealing the provisions of the penal code that had made the combination of workers or employers to inﬂuence wage levels illegal. The particularly hostile provision of the compulsory ‘‘workbook’’ (livret) for workers – which made it impossible for them to change employers even if they had complied with all the contractual obligations – fell into disuse in the 1870s and was abolished in 1883. The same happened to a law stating that, in a conﬂict, the employer had to be taken at his word (this was also an article of the penal code). At the same time, it is important to remember that in the 1850s Belgium had instituted a series of Arbitration Councils, the aim of which was to arbitrate and moderate, to judge certain types of conﬂicts, and to perform certain policing tasks. From 1859 on, workers’ representatives gained equal representation on these councils, and the right to vote in these bodies was opened up progressively, ﬁrst to the more prosperous workers and then to all of them (1896). In Belgium, the key issue of collective action in the period from the 1880s was typically that of electoral reform, and particularly the elimination of the insidious plural voting. In 1912, in connection with this issue, there was considerable rioting and violent demonstrations, which resulted in the death of some of the participants. Again in 1913, the socialists scheduled a general strike for electoral reform; this time, it was peaceful, although it was just as unsuccessful as before. In the Netherlands, the most violent confrontation occurred during the general strike of 1903, launched by the anarcho-syndicalist groups within the unions against the conservative government and, in particular, against legislation limiting the right to strike in the public service. There was considerable repression of the participants in the strike, and troops were called in to control the railway stations. However, this relatively violent incident – which led to the demise of Dutch anarcho-syndicalism – was isolated. Dutch agitation for suffrage enlargement was far less strained and came into existence somewhat later (1910 and 1912) than in Belgium. In general, socialism developed in a relatively tolerant political climate, which justiﬁes the statement that ‘‘the Dutch government yielded to lower 25

Lafferty (1971: 148).

POLITICAL INTEGRATION

329

classes’ demands earlier and with considerable more grace.’’26 In neither Belgium nor the Netherlands was the political socialist movement ever outlawed, severely persecuted, or repressed. Ireland is difﬁcult to classify. Formally, it came under British constitutional and legal rule, and its late-developing trade union movement was for many years closely associated with that of the British. While the global environment can be said to have been the same, however, the social problems were different, and by the turn of the century they had begun to be amalgamated with the issues of home rule and independence. Moreover, in sharp contrast to the relative social calm of British politics in the ﬁnal quarter of the century, Irish society was passing through a period of strong social unrest, to which the British replied with a complex mixture of actions, including repression. For example, the 1881 Coercion Act suspended the right of habeas corpus, authorized detention without trial, prohibited illegal meetings and seditious speeches, and so on. In the same year, Charles Parnell, the leader of the Irish National Land League, was jailed and the league outlawed. In 1882, a new Coercion Act presented by Gladstone introduced three years of martial law. In 1886, the Tory government of Salisbury (1886–1892) responded again to a new upsurge of agrarian disorder with sharply repressive measures. These disturbances continued throughout the 1880s; and when the agrarian issue ﬁnally calmed down, new turmoil started to emerge on the issue of home rule, increasingly polarizing both the Protestant and Catholic communities. It was probably only the outbreak of World War I that prevented an escalation in violence. At the same time, this policy of ﬁrm repression was balanced by one of reform. In 1881, the British Parliament passed a major agrarian reform that established dual ownership of land by the landlord and the tenant, ﬁxed the establishment of rent levels by impartial bodies, and limited evictions to cases of nonpayment of rent and other well-deﬁned situations. Each successive wave of repression was accompanied by acts whereby rents would be paid for poor tenants (Arreas Act of 1882) and low interest rates offered so that they could purchase their land (Ashbourne Act of 1885, enlarged and refunded in 1888 and 1891). Thus, the early popular (mostly peasant) movements in Ireland – which were by no means socialist in character – were faced with a stop-and-go policy of repression and appeasement that eventually paciﬁed the agrarian Irish world but left a far from benign image of British rule. 26

Goldstein (1983: 263).

330

THE POLITICAL MOBILIZATION OF THE LEFT

In Italy, France, Austria, and Germany, the early development of the socialist movement faced strong state opposition and resistance. The established elite resorted to nonpolitical means of confrontation, increasingly relying on the pressure of the state administration, the police, and the army. The similarity of the response of these countries is striking, considering that Austria and Germany were the centers of internationally powerful continental autocracies, while France and Italy were more liberal. At the same time, within a common pattern of relatively violent response, the differences among these countries should not be ignored. Early attitudes toward the French working-class movement had been relatively tolerant. The 1791 ban on union activities was repealed in 1848, when the right to unite was formally granted. However, the revolution, its drive for mobilization and nationalization, and the Napoleonic reforms had further centralized the French state, making the persistence and development of localized, restricted pluralist traditions in parts of the country difﬁcult and enhancing the local control of the prefects. Later on, this apparatus was used regularly to mobilize supportive voters and to harass the organizational development of the opposition. Old restrictions on clubs, societies, and meetings were restored, and combinations were made subject to legal prosecution. However, in 1864 the ban on combination ended and strikes became legal, although the right to strike was limited and unions remained illegal. This long-standing prohibition against workers’ trade associations was ﬁnally abolished in 1884 (only a decade after this happened in Britain), and from this date on active state repression perceptively slackened.27 There is, however, a history of administrative harassment in France that is almost totally independent of these legal developments and that marked the birth of the working-class movement indelibly between 1864 and 1884. In the late 1860s, the moderate leadership of the French Section of the First International met with strong state opposition, with troops used to end strikes and with extensive arrests and prosecutions. As a result, the section became increasingly less moderate in its views, and more syndicalist and revolutionary.28 The Commune’s repression was by far the most violent military, police, and judicial repression of working-class movements in Europe in the nineteenth century,29 and it continued to

27 28

29

Zeldin (1979: 205–206). Ibid., 365–369. ‘‘The growth of the moderate Trade Unions’ movement in the 1860s was cut straight, and the opportunity was given for more violent elements to come to the fore’’ (Ibid., 381). According to historical sources, between 20,000 and 25,000 people were summarily

POLITICAL INTEGRATION

331

shape the attitude of the established elite to the working-class and socialist movements for many years. Thiers’s repressive zeal was resumed during the frequent authoritarian crises: from Marshall MacMahon’s attempted ‘‘institutional’’ coup in 1877 to Boulanger’s 1886–1889 attempt to overthrow the republican regime; to the 1892 laws against the anarchists (lois sce´le´rates) to the casualties of Fourmiers and many others; to the 1905– 1908 policies of police provocation and brutal repression that resulted in the deﬁnitive break between radicalism and the working class.30 During this period, and even up to 1936, memories of repression and of violent clashes with the state were repeated by periodic use of the army, which put down strikes, fuelling the workers’ antimilitarism and antistate feelings. The strong resistance of the state to recognizing unions was equaled only by that of the employers. As a result, collective bargaining and bilateral negotiation in the Scandinavian style never really gained a foothold (see Chapter 6).31 In contrast to this strong reaction to the organization of unions, the attitude toward the organization of the political movement was much milder. Its development was comparatively late; but after socialist political organizations were legalized in 1879, no direct attempt to outlaw them was ever made. In France, therefore, workers’ activities and organizations met with strong resistance by the state and with low recognition by the employers. The strong state tradition and its high centralization made resources for repression easily available; at the same time, the politically weak republican liberal and conservative elite was motivated to use such resources. It is difﬁcult to assess to what extent the strong anarcho-syndicalist component in the French working-class movement was a cause, an excuse, or an effect of this intemperate state reaction, even though it is clear that moderate workers’ associations and movements were harassed even before the Commune existed.

30

31

executed and killed; 38,578 were arrested; ten times this number were denounced; and there were large deportations to New Caledonia; (Zeldin 1979: 380). Kergoat reports that in the 1870s ‘‘every year an average of 9% of strikers were brought to justice and sentenced.’’ Kergoat (1990: 165–167). Despite the new style of the Republic of Jules Ferry and Waldeck Rousseau, who instructed the prefects in 1884 not to use the army during strikes, this was largely ignored in the following period. The use of armed forces ‘‘which had dropped to 3% in the total number of disputes in 1882 increased to 10% later.’’ Between 1892 and 1895, historians report prison sentences against anarchist and socialist agitators totaling 249 years, three life sentences and forced labor, and ﬁve death sentences. In the years 1905–1908, the army intervened against workers, with the following result: 20 killed, 667 wounded, and a total of 104 years of prison sentences. In 1933, only 7.5% of wage earners had been affected by such agreements. Cf. Zeldin (1979: 207); see also Schorter and Tilly (1974: 35).

332

THE POLITICAL MOBILIZATION OF THE LEFT

Italy lacked the long-standing tradition of dynastic bureaucracy, the old nation building, and the international status of France. Notwithstanding these fundamental differences, the Italian state-building liberal and conservative elite felt as unsafe as its counterpart, the French republican elite, facing strong opposition from the papal state and Italian Catholics, as well as having an uncertain international status and facing strong internal resistance to central control. Legitimation of the socialist movement risked shrinking the ﬂimsy support base of the dominant elite even further. The repressive orientation of the new Italian state was boosted largely to repress the widespread resistance to its establishment, which took the form, among others, of separatist movements in Sicily, proBourbon sentiment in the south, and Maﬁa and bandits. These created a mood of repressive defense in the state-building elite, in which Bourbonists, clericals, republicans, and internationalists were all considered enemies of the new state. The battle against them was often fought with exceptional legal instruments: preventive arrest, house arrest, limitations on movement, and administrative deportation (confino). The legalization of unions and their activities, including strikes, took place over a relatively long period (1859–1889), and the end of severe harassment (but not of administrative nuisances) occurred only at the beginning of the twentieth century. During this period, intense political repression alternated with moreliberal toleration. The coming to power of the liberal left in 1876 coincided with a more moderate use of repressive means;32 however, repression was renewed after 1885, when socialism started to spread among the agricultural laborers of the Po Valley and the industrial workers of Milan. In 1892–1894, the repression of the Fasci Siciliani rebellion led to martial law in Sicily in 1894–1895; Crispi resorted to intense repressive tactics; and in 1897–1898, civil disorders brought about strong military intervention against demonstrators. Before the time of Giolitti, therefore, the Italian pattern involved frequent use of the army and of administrative and police measures (less of the judiciary), as well as a weak political and legislative response by a largely nonpluralist dominant political elite. Unlike France, however, in Italy the socialist political movement was also repressed strongly. Various socialist political organizations were dissolved during the 1880s, and the Socialist Party, founded in 1893, was made illegal between 1893 and 1895 32

For instance, the use of the ‘‘admonition’’ (ammonizione) – the police instrument that was most open to arbitrary political use – was restrained, and the number of ammoniti declined from 184,000 to 40,000 between 1876 and 1884. However, a more repressive policy in this ﬁeld was soon resumed after 1885. Cf. Carocci (1975: 51).

POLITICAL INTEGRATION

333

and once again in 1898. In addition, the period after World War I did not mark the legitimation of the socialist movement but, instead, its complete eradication. Unions had to wait until 1945 to regain their previous legality, and up to the end of the 1950s major strikes and political demonstrations maintained a confrontational style with the state and police forces (with frequent casualties). Both the French and Italian liberal regimes rested on a restricted popular base and found it difﬁcult to enlarge. This resulted in a strategy of containment, or even rejection, of the political integration of a considerable portion of their national constituencies (see the following chapter on this point). Both Austria and Germany banned workers’ economic associations from the beginning of the eighteenth century; both formally repealed this legislation at the end of the 1860s; both continued to practice severe administrative harassment until the beginning of the 1890s and continued thereafter with milder forms. In Austria, legal recognition of the unions in 1866 was followed six months later by the prosecution for high treason of those who had signed the request for this recognition. The 1884 emergency introduced a seven-year-long police state in which freedom of the press, the right to meet and assemble, and the protection of postal secrecy were suspended, and arrests could be made without any judicial authorization. A signiﬁcant easing of control occurred only during the 1900– 1905 period. The record is similar in Germany, although here there was also the precedent of the anti-Catholic Kulturkampf. In 1871–1876 and in 1878, socialist organizations had been strongly and directly repressed. The ‘‘socialist law,’’ in force from 1878 to 1889, was typical of the domineering attitude of the Prussian-German elite toward socialism and, in general, toward democratization and emancipation movements. Social democrats and socialist unions were stigmatized as enemies of the empire and therefore of the state. From the socialist point of view, the anti socialist legislation made it clear that the state had taken the side of dynastic and property interests against the workers and their organizations. The socialist movement, strongly promoted by rapid industrialization, was forced to accept drastic limitations on both its liberty to act and its will to participate in the legal and constitutional systems of the country.33 In Austria, political rights were regulated by the 1867 law on associations, which prohibited political societies from merging, founding branches, corresponding with each other, or even maintaining contact via a member of the parliament. Strictly speaking, any political activity outside parliament 33

On the repression of the early German socialist movement and party, see Lidtke (1966: 70–88, 242–253).

334

THE POLITICAL MOBILIZATION OF THE LEFT

and the state institutions could be punished by law.34 This explains why the Austrian social democratic societies had a strongly cultural character, and educational associations concealed their political commitment and emancipation goals behind a range of cultural and educational activities. Although unions were never formally recognized by either the state or employers, the autocratic regimes of Germany and Austria tolerated strike and union activities more than political activity.35 Contrary to the British pattern, economic conﬂicts and action in the market were more easily accepted than political conﬂicts and action. This was because the ideal of the autocratic state was one of apolitical corporate representation, which it would probably be willing to recognize, provided that no challenge was made to the political order. The attempt to isolate political socialism from its social base was thus common to both states and was also pursued systematically through the legislation of the early welfare state.36 In Germany more clearly than in Austria, an exchange of recognition for support occurred when World War I began and the socialists attempted to integrate into the state by supporting its war activities. In a single move, these socialist parties wanted to demonstrate their national reliability and respectability, as well as to rid themselves of the accusation of being antipatriotic enemies of the fatherland.37 Following this, trade union ofﬁcials were exempted from military service, state employees were allowed to join the SPD, and the state forced reluctant employers to accept the trade unions as bargaining partners. Finland is difﬁcult to classify because its late-developing socialist movement was forced to take root in the illiberal climate typical of a subordinate periphery of the Russian Empire. The guild system was abolished by the Trade Acts of 1868 and 1879, which repealed the prohibition on journeymen and employees to associate to further their interests. After this date, no legal restrictions were aimed speciﬁcally at the unions, but freedom of association was virtually nonexistent, as governmental approval was necessary to found associations and was granted on a very restrictive basis. Such restrictions were also applied to workingmen’s associations when socialism reached Finland in 1890–1900. In 1906, freedom of association was enacted as a constitutional right. The point is that Finland enjoyed relative independence over its internal affairs, which was, however, 34 35

36

37

Mattl (1990: 295). Several examples of this differentiated attitude of public authorities are documented in Ritter (1963). On the early welfare measures of authoritarian regimes, see Alber (1982: 119–193) and Flora and Alber (1981). See the debate on this interpretation in Tenfelde (1981: 216–247).

POLITICAL INTEGRATION

335

dependent on Russian attitudes. These were largely determined by the empire’s fears of revolutionary movement more than by any speciﬁc intention of fostering Finnish employers’ interests. The historical record, therefore, is not one of intense socialist movement repression, and when this repression was triggered it was more an attempt by the Russians to repress Finnish autonomy than to repress the socialist movement itself. Once Finland became an independent state, however, internal social tensions began to prevail. The 1919 constitution embodied a new Association Act, but despite this, in 1922 an Employment Contract Act became necessary to declare unlawful and punishable actions taken by employers to prevent employees from joining or belonging to a lawful association. The Finnish socialist movement had emerged from the civil war under a largely communist leadership, and the political action in which it engaged awoke the suspicion and mistrust of authorities and employers alike (and often of moderate socialists, too). In 1930, the SAJ Confederation was dissolved for subversive activities by a court decision (a unique case in the twentieth century). In 1931 an act for the protection of labor peace included provisions that established punishments for forcing employees to either join or to abandon associations and/or to take part in a strike. It was, of course, easy to turn this act into an antiunion instrument. After the 1918 civil war, mutual distrust concerning collective agreements meant that they ceased to be used, and it was only after World War II that most employer associations were prepared to accept trade unions on an equal footing with themselves (see Chapter 6). What makes Finland a special case is that the state was weak under Russian rule, with virtually none of its own military or police forces. Again, after independence, there was a situation more of internal social polarization than of state political repression. Thus, Finland between 1914 and 1945 can probably be more appropriately described as a ‘‘praetorian society,’’ where a weak state was easily penetrated by the social and political forces that were stronger at any speciﬁc moment. If the experience of Finnish socialism cannot be said to be one of early and smooth institutional integration, it should at the same time be kept separate from those of the central European empires.

RESPONSIBLE GOVERNMENT The political responsibility of the top executive branch adds a third type of resource in the ﬁght for power and inﬂuence over decision making. Before this principle is fully implemented, resources are limited to the

336

THE POLITICAL MOBILIZATION OF THE LEFT

inﬂuence that individual and political actors have in and over the administrative and bureaucratic machinery of the state, as well as the inﬂuence they have in the market or can mobilize in the society through organizational pressure. Executive political responsibility adds the political resource of parliamentary inﬂuence and negotiation, which are dependent on electioneering and electoral mobilization. When executive elites become dependent on elective bodies, they are forced to take electoral power relationships into consideration. Electoral resources become an element in the transformation of social and/or administrative inﬂuence into executive dominance. Consequently, elective elites have a greater tendency than monarchs, bureaucrats, lords, or militaries to see the conﬂicts and requests coming from new claimants as political opportunities – to beat and weaken the adversary or to defend themselves – rather than as political threats. When this occurs, the political integration of outsiders, that is, their acceptance of the existing institutional framework as one in which their values can be fought for and at least in part achieved, is easier and therefore more likely. I use the term ‘‘responsible government’’ here in a broad sense to refer to the process that extends and ﬁnally grants to elective chambers control over the formation, personnel, duration, and legislative activities of cabinets. In this broad deﬁnition, the process is quite complex. The constitutional history of European countries offers several cases of ambiguous and intermediate situations.38 Cabinets might be dependent on the support of the elective chamber for their continued existence, but their prime ministers (and ministerial personnel) might be freely chosen by other bodies, such as the monarch or the nonelective ﬁrst chamber. Cabinets might depend on chamber support to be installed, but might still be required to resign by forces other than those in such a chamber. Cabinets and prime ministers might be politically responsible to the elective chamber but still face considerable ‘‘legislative’’ conﬂicts with other nonrepresentative bodies, such as the ﬁrst chamber or the dynastic bureaucracy. Therefore, the establishment of parliamentary government was often a long process characterized by periods of uncertainty in the distribution of competencies, mutual veto power among representative and nonrepresentative institutions, steps forward, and setbacks. The analysis of the thirteen countries, therefore, requires considerable contextual speciﬁcation, but I propose the following six yardsticks for comparative evaluation: 38

For a descriptive account of early institutional developments see Anderson and Anderson (1967).

POLITICAL INTEGRATION

337

1. Before the introduction of responsible government, when the elective chamber was unwilling to pass legislation, was parliament dissolved by the crown? 2. At what point in time was the cabinet’s responsibility to the elective chamber of parliament accepted or imposed over the monarch’s resistance? 3. To what extent and for how long did institutional ﬁghts with nonelective or nonrepresentative ﬁrst chambers persist? 4. Were these conﬂicts politico-institutional (concerning the powers of the elective chambers toward the executive), or were they mainly legislative (concerning speciﬁc legislation veto or rebuttal)? 5. To what extent and how long did the monarchy and its immediate advisers and milieus continue to intruded on the life, personnel, and duration of cabinets? 6. After parliamentary responsibility was introduced, when was a cabinet maintained against the will of the elective chamber for the last time? Let me begin with the four uncontroversial cases of the United Kingdom, Switzerland, Belgium, and the Netherlands, where parliamentary governments developed slowly and well before the middle of the nineteenth century. No major politico-institutional conﬂicts concerning the distribution of political power between the two chambers occurred, although conﬂicts over legislation were frequent. No major monarchical interference in the activities of these governments are reported by historians, and those episodes that are mentioned look marginal in comparison with the remaining cases in our set. Finally, in none of these countries did setbacks in the tradition of responsible government ever occur, and no cabinet was ever kept in power against the explicit will of the elective chamber. The Belgian case merits little or no discussion. When in January 1831 Great Britain, Austria, Prussia, and Russia ratiﬁed the independence of the new Belgian state – the result of a rebellion in the southern provinces of the former Austrian Netherlands – and determined that it would be neutral, the constitutional assembly designed a very liberal constitution. The ﬁrst chamber was originally elected directly and remained so until 1893. After this, a reform meant that half of it was directly elected and the other half was elected by the provincial government. Neither this chamber nor the crown ever challenged the primacy of the second chamber and the principle of responsible government. Strictly speaking, in Switzerland the formal principle of government

338

THE POLITICAL MOBILIZATION OF THE LEFT

responsibility at the federal level was introduced by the further centralizing constitutional revision of 1874. However, this practice made little sense within the loose federation before 1848 and even in the 1848–1874 phase, which followed the Sonderbund canton war of 1848 and the approval of the new federal constitution. In any case, the practice of responsible government was already effective in the cantons well before then, even before 1848 in many cantons. The special character of Swiss cabinet responsibility within the directorate structure of the executive is discussed in more detail in the next section. What matters here is that since that time, federal councilors have always been elected by parliament, and it was unthinkable 150 years ago, just as it is today, that they could challenge the latter. After all, in a republic without a president, the Swiss cabinet had no authority beyond the parliament and the citizens on which to rely or to whom to appeal. In the Dutch case, a Whig interpretation of political development emphasizes the continuity from early corporate pluralism in the seven provinces to modern liberalization. According to this perspective, the early break in the development of a modern absolutist state (under a Burgundian–Hapsburg dynasty) resulted in the weakness of the central institutions (military and bureaucracy), an early and prominent role of the bourgeoisie (given the weakness of the aristocracy), and an emphasis on the right to resist any unlawful action of the sovereign, together with a persistent particularism. As a consequence, the presence of many actors with inherent rights fostered the gradual development from early pluralism to modern democracy. The institutional developments in the nineteenth century, and particularly the constitutional revision of 1848, liberalized the regime, strengthening the parliament and transforming the ﬁrst chamber from a royally appointed body to one elected by provincial councils. However, a radical interpretation emphasizes the oligarchic and nondemocratic character of politics in the pre-Napoleonic period and regards the revolutionary break of the radical Patriots movement of the 1770s and the Bathavian–French period between 1794 and 1813 as crucial for liberalization and democratization. According to this thesis, modern democracy resulted more from these breaks than from any kind of continuity with early pluralism.39 Whatever the case, by 1848 the practice of parliamentary 39

In reviewing this historiography debate, Daalder argues that the two interpretations refer to two different conceptions of democracy and can to a certain extent be reconciled. While the Whig interpretation concentrates on a vision of democratic development as basically liberalization and contestation, the radicals’ interpretation focuses on participation as the key dimension. See Daalder (1990: 4–17).

POLITICAL INTEGRATION

339

government was established, and the possible opposition of a ﬁrst chamber appointed by royalty was eliminated by reform. Conﬂicts between the crown and the parliament over power attribution were not infrequent between 1848 and 1868, but their nature, frequency, and intensity did not jeopardize basic parliamentary practice. In the United Kingdom, parliamentary control of the executive occurred comparatively early and gradually, even if there are different opinions about the date when it became fully operational. Some put it as late as 1832, but most argue that the decisive turning point was in the ﬁrst years of the nineteenth century.40 The nineteenth century was marked by constant conﬂicts between the Lords and the Commons over legislative competence and vetoing – which, however, never brought into question the fundamental principle that the prime minister was accountable to the lower elective house. The downgrading of the Lords’ powers had started long before, but the ﬁnal confrontation came in the ﬁrst decade of the twentieth century and was solved by repeated use of electoral decisions, which consistently (in the two 1910 elections) endorsed the liberal position. Ireland shared the institutional developments of the United Kingdom even though, by the end of the nineteenth century, the issues were different and the growing feeling of nationalism gave rise to violent upheavals (and eventually a civil war in the transition to national independence). In this, the Irish case presents similarities to the Finnish case, on which I will dwell further later on. The important point is that 1922 is the year in which both the new Irish state and the arrival of parliamentary government were achieved, and the key issue that mobilized political forces in Ireland before that date was not so much internal liberalization but home rule. France and Norway are intermediate cases where parliamentary government was established between the mid-1870s and the mid-1880s. However, the only similarity between them is the timing. While in France the principle and practice of parliamentary government were uncertain for a fairly long period, in Norway the transition was astonishingly smooth. In France, notwithstanding the constitutional provisions of 1871, the practice of parliamentary government did not consolidate until the deﬁnitive failure of the reactionary President Marshall MacMahon, who, with the support of a small monarchical majority in the indirectly elected Senate, dissolved the lower chamber and called elections in 1877 under conditions of considerable state administrative pressure. The resulting republican majority introduced the principle of parliamentary responsibility and the curtailing of 40

For the developments up to the 1831 reform, see the documented evaluation of Mackintosh (1977: 35–74).

340

THE POLITICAL MOBILIZATION OF THE LEFT

presidential powers. Given this experience, the mistrust of an active presidency was such that no other parliament was allowed to be dissolved during the Third Republic. However, the principle of parliamentary government was unacceptable to large sections of the public and the bureaucracy and to the end of the century – after the attempt by Boulanger to overthrow the parliamentary regime – the republican majorities remained convinced of the need to defend the system against internal enemies. Moreover, the varying political composition and orientation of a predominantly rural and indirectly elected Senate contributed to the considerable legislative conﬂicts with the lower chamber until the end of the 1930s, often with signiﬁcant political consequences. For example, at the end of the 1930s, it was the strong opposition of the Senate to the socioeconomic legislation of the Popular Front governments of Le´on Blum that led to the latter’s downfall twice, in June 1937 and April 1938.41 Norway never faced the problem of curbing the privileges of the upper chamber because the parliament was a unicameral body from the beginning. When it was ﬁrst established in 1814, during the political void created by the end of Denmark’s alliance with Napoleon’s France, ideas regarding the use of the upper chamber to represent and secure aristocratic conservative interests were strong everywhere in Europe, particularly in Denmark and Sweden. However, the aristocracy and the ecclesiastic orders were so weak in Norway that eventually the Storing became a unicameral institution.42 Until 1884, the Norwegian government was not accountable to the Storing, and executive authority was in the hands of nominated governmental ofﬁcials. In 1884, an alliance of peasants and other groups (mainly doctors, teachers, and lawyers) gained the majority in parliament and founded the ﬁrst Norwegian political party – the Venstre. This majority impeached, convicted, and removed from ofﬁce eight members of the 41

42

The French Senate in the Third Republic – indirectly elected at different points in time than the lower chamber, enjoying the right of censure against the cabinet and considerable control over the timing of the dissolution of parliament, mainly representing the local notables of rural France – acted systematically as an important brake on the consolidation of the French party system and on the powers of cabinets and prime ministers. The fall of Blum’s ﬁrst cabinet (June 1937) is revealing: After the Senate’s modiﬁcation of Blum’s ﬁscal program, the socialist prime minister tried to solve the crisis by resorting to the electorate via dissolution and new elections, that is, following a British style of crisis management by the lower house. Radical ministers in the cabinet and the Senate, which needed to agree on an early dissolution, resisted the request. For a detailed analysis of the role of the Senate in the Popular Front governments, see Lefranc (1963: 339–340, 352–353). It was composed of two houses (Lagsting and Odelsting), which met separately in the law-making process but were elected at the same time and had a similar composition.

POLITICAL INTEGRATION

341

royally appointed government, and this ﬁnally forced the Swedish king, Oscar II, to yield to the primacy of parliament, apparently after having carefully considered the advantages and disadvantages of a royal coup. No important internal resistance to this event has been recorded by historians. The absence of a national monarchy, the weakness of the local aristocracy, and the low level of stateness of the political process left the country with few nonparliamentary political resources. Obviously, parliamentarization concerned domestic matters; any control over its own foreign affairs had to wait until 1905, when Norway became formally independent from Sweden.43 Denmark, Sweden, Finland, Austria, and Germany can be said to have established the deﬁnitive parliamentarization of their governments in the twentieth century; but in addition to differences in timing (from 1901 to 1945), there are also differences in process. Denmark and Sweden had a similar pattern of delayed cabinet responsibility and protracted institutional conﬂict between an elective lower chamber, on the one hand, and the crown and the upper chamber, on the other. In both cases, ﬁnal control of the lower house was not the result of military defeat. Finland, Austria, and Germany shared protracted autocratic rule curbed only by military defeat, although in Finland this was the defeat of the colonial power rather than of the local dynastic power. In a broad comparative framework, the Danish and Swedish constitutional developments hardly merit strong differentiation. Both countries had a tradition of dynastic absolutist rule, but whereas in Denmark this tradition was strong and unchallenged until 1834 – when the king was obliged to introduce an ‘‘estate constitution’’ with four different regional assemblies – Sweden is thought to have maintained more continuity in its institutions of representation.44 In Denmark in 1849 and in Sweden in 1866, constitutional reforms abolished the estate diets and introduced bicameral parliaments that, due to a more restricted franchise or to plural voting systems,45 were sharply differentiated in social composition (mainly 43

44

45

Note that Norway had a tradition of sovereign independence. An autonomous Norwegian kingdom existed from 1163 to 1319, and it was only in 1536 that Norway became a province of Denmark. Even later, during the long period of Danish–Swedish wars for predominance in the area (1563 to 1720), Norway often managed to defend itself by its own efforts. See Derry (1968: 44–67, 89–108). Even if Denmark is notable for its rapid and peaceful transition from absolutism and a military state to early constitutional liberalism in the ﬁrst half of the nineteenth century. See Lind (1988: 112–119). To be more precise, the Swedish upper chamber was indirectly elected after 1866 by an electoral college of members of local governmental bodies. However, the system of

342

THE POLITICAL MOBILIZATION OF THE LEFT

farmers versus ofﬁcers, landowners, and capitalists). The conﬂict between the two chambers increasingly came to dominate the politics of the last quarter of the century in Denmark and even longer in Sweden. The lower chamber was probably stronger in Denmark, where it managed to check legislation and even to force the prime minister, Estrup, to resign in 1894. However, there were frequent instances of parliamentary dissolution in this period due to the parliament’s recalcitrant attitude to government legislation. The king continued, meanwhile, to rely on conservative ministries drawn from the restricted-suffrage upper house and on government by decree. Finally, in 1901, the king was forced to accept parliamentary responsibility based on the Folketing and to call the Venstre to power, although the resistance of the conservative Landsting continued by systematically blocking measures to reform itself. The last resistance ended in 1914, when all the Danish moderate and left parties agreed to institute universal suffrage of the Landsting, a move that was boycotted and delayed by the upper house until 1915. The Danish king made a last, and unsuccessful, attempt to force his own will on the lower chamber during a cabinet crisis in 1920. In Sweden, the king resisted the parliamentarization of the executive until 1917. Attempts to introduce parliamentarization had been underway since 1905, however, when the liberals in the second chamber had brieﬂy formed a cabinet. Even before that date, several cabinets had had to resign after losing votes in the parliament, which had already gained power at the expense of the king.46 However, in 1914, King Gustav forced a liberal prime minister to resign over the issue of increasing military expenditures, and, in an explicit act of rejection of parliamentary democracy, he appointed a conservative ministry that governed throughout the war years by relying on the support of the royal house and the upper chamber. When, in March 1917, a new conservative ministry of this type was appointed, this appeared to be a clear repudiation of parliamentary sovereignty; the ministry was ousted by popular riots and by the lack of reliable army support in repressing them.47 Germany, Austria, and Finland introduced the parliamentarization of the executive as a result of the military collapse of their autocratic dynastic rulers at the end of World War I. Until 1809, Finland, as part of Sweden,

46 47

voting in local elections was characterized by substantial property qualiﬁcations linked to plural voting. At the turn of the century, certain people could still cast up to forty votes in this electoral college. Sannerstedt and Sjo¨lin (1992: 100). Stjernquist (1966: 117–121).

POLITICAL INTEGRATION

343

had had no tradition whatsoever of sovereign independence. When, in that year, Finland was handed over to Russia, it became a grand duchy, maintaining the Swedish constitution of 1770 and displaying some of the characteristics of a sovereign state. For example, it had its own central government, central bank, and customs and postal systems; these national institutions were set up in Norway at a later date.48 The Finnish fourestate Diet was convened in 1809 but was not reconvened until 1863. From this date on, the Diet met regularly, and there was also a modest degree of liberalization between 1860 and 1894. So, paradoxically, although deprived of a long tradition of nation-state autonomy and subject to the staunchest autocratic power in Europe, Finland did enjoy a certain degree of autonomy under Russian administration. Finland’s relationship with the metropolitan power deteriorated toward the end of the century, when assimilation tendencies gained ascendancy in Russian imperial circles, which clashed with the by now distinctive national structure and identity of the country. By 1894, a new period of absolutism had manifested itself through attempts to russify Finland: Russian was made the ofﬁcial language of the bureaucracy, and Russian personnel were placed in key governmental positions (1902). There were also attempts to conscript Finns into the Russian army. Freedom of the press, of organization, and of assembly were restricted, and the Finnish national opposition was repressed. The series of measures designed to limit Finland’s autonomy gave rise to a passive-resistance movement headed by Swedish-speaking liberals and Finnish nationalists. In the period between 1907 and 1913, the tsar continuously blocked legislation and dissolved the Diet almost every year, dissatisﬁed with the electoral results, which continuously gave power to a body of nationalists among whom the socialists were the strongest. In 1910, a last attempt was made to achieve direct control of Finland through a Russian Duma law that removed a wide range of matters from the jurisdiction of the Finnish Diet. From 1917 to 1919, Finland declared its independence (December 6, 1917), went through a civil war, and appointed a German king (Prince Froedrick Karl of Hesse in October 1918), who was then dropped after the German defeat and surrender. In July 1919, parliamentarization was ﬁnally established with a governmental act of secession from Russia. Finnish cabinet responsibility after 1919, however, requires some clariﬁcation because of the special executive dualism of the Finnish consti48

Danish rulers allowed the much desired national bank and university to be set up in Norway only in 1813.

344

THE POLITICAL MOBILIZATION OF THE LEFT

tutional model.49 Notwithstanding the formal inclusion of the principle of responsibility in the constitution, Finnish parliamentarism does not present or use any of the three basic mechanisms that usually embody the idea that the cabinet must have the conﬁdence of the parliament: (1) that the president is obliged to ﬁnd out beforehand if the person he intends to appoint as prime minister enjoys the conﬁdence of the chamber; (2) that the chamber votes to express conﬁdence in the cabinet after its appointment; and (3) that the cabinet invites or initiates such a vote of conﬁdence in presenting its manifesto.50 Between the world wars, there were several occasions on which the president appointed cabinets against the will of the parliament. For example, in 1924, Sta˚hlberg dissolved the parliament against the wishes of its majority; and in 1936, a bourgeois minority government was appointed in similar circumstances.51 Similar cases on the borderline of what is usually regarded as the principle of parliamentary conﬁdence can be mentioned. However, within the Finnish constitutional context and its political practice, it would be misleading to consider these cases as a violation of the constitutional rules. We must in this case accept more-contextual criteria of parliamentary control over the executive, because no major institutional crisis has developed between the president and parliament since the 1919 constitution. Therefore, the latter data can be regarded as the starting point of parlimentary control. The situation was different in the two central European empires. The German constitution of 1866 was an intricate ediﬁce devised by Bismarck 49

50 51

Having failed to get a king, the 1919 constitution makers adopted the statements of the 1917 constitution-drafting commission, opting for writing the principle of parliamentary government into the constitution and, at the same time, giving the presidency of the republic a variety of prerogatives and powers so as to make it a check on purely parliamentary government. As a compromise between the sharply diverging constitutional ideologies of the right – favoring a strong and independent presidency – and the left – favoring a purely parliamentary government – the relationship was ﬁnally shaped by practice and conventions, and remained strongly divergent according to the political climate. After the civil war, in the context of continuous minority government, the role of the president in choosing the cabinets and inﬂuencing their lives remained crucial. For a discussion of the complex attribution of powers among the parliament, the cabinet, and the presidency in Finland, see Duverger (1986). For this practice, see Anckar (1992: 154–155). Anckar mentions a telling episode. When in 1926 the Social Democratic Prime Minister Va¨ino¨ Tanner presented his program in the parliament after his appointment, arguing that this was the practice parliamentarism implied, the speaker of the chamber did not allow discussion or voting on the matter, arguing that in the parliamentary rule nothing was speciﬁed concerning this problem. The speaker’s decision was not contested by the parliament’s majority, nor was it referred to the Constitutional Committee (Anckar 1992: 155).

POLITICAL INTEGRATION

345

as a purely personal invention, intended to permit the chancellor to neutralize his domestic enemies and to facilitate German uniﬁcation. Each of the twenty-six states retained its theoretical sovereignty. The federal government was, on paper anyway, very restricted in its powers, and its institutions were correspondingly few. They consisted of the emperor and his only minister and executive agent, the chancellor; the high civil servants; military and naval commands; and a parliament whose lower house was elected by universal manhood suffrage. It was in the German states that the residual authority was vested; their delegates composed the upper house of the federal parliament, and they were in a position to dominate the federal government, enjoying veto power on Reichstag legislation. However, there was an element that distorted the imperial constitution: the role of Prussia, which was so dominant as to profoundly inﬂuence, if not control, both the imperial regime and the other states too. Prussia accounted for about three-ﬁfths of the territory and population of the German federation. The king of Prussia was also the German emperor; the prime minister of Prussia was, as a matter of custom, also the imperial chancellor; and the Prussian army was the overwhelming majority in the German defense establishment. There was no clear limit to the authority of the emperor, who could appoint and dismiss the chancellor and who considered himself as responsible for forming policy. The government of the state of Prussia was dependent on the Prussian parliament. This body was dominated, thanks to its voting system, by a minority of the Prussian electorate, a minority dominated in turn by the dour, parochial, conservative, and virtually bankrupt landowners of eastern Prussia. The Prussian roots of the German Empire were important in another respect: It was in Prussia between 1862 and 1866 that an important constitutional conﬂict took place that defeated the liberal forces and proved crucial for the future of the politico-constitutional equilibrium of the empire. The liberal forces had insisted on having ﬁnancial control over the army; this was strongly resisted by the king and the army itself. The latter together managed to override several parliamentary vetos, resorting directly to the bureaucracy to ﬁnd the money necessary to ﬁnance the army, and dissolved the parliament three times (in 1861, 1862, and 1863) before the king decided to rule without deferring to it at all. The chancellor, as prime minister of dominant Prussia, was dependent on the support of that state’s decaying aristocracy; as chancellor of Germany, he was also dependent to some extent on the democratic Reichstag; and in both capacities, he was dependent on the hereditary monarchy. Even in the imperial constitution of 1871, executive power was vested more in

346

THE POLITICAL MOBILIZATION OF THE LEFT

the Bundesrat, a large, unwieldy body technically presided over by the emperor and accruing to his power.52 Thus imperial Germany not only did not have the principle of parliamentary responsibility, but even the legislative role of the lower house was restricted, and not only on military and ﬁscal matters. Still, in the early 1890s, when the Reichstag refused to extend the expiring antisocialist law, Bismarck considered the idea of a coup to restrict the franchise. The same situation existed in the Hapsburg Empire after Emperor Francis Joseph I recognized the principle of constitutional government through the October Diploma of 1860 and the February Patent of 1861. A similar conservative upper house of aristocratic dignitaries and landowners – chosen and convened by the monarch – existed, and a similar repeated dissolution of the lower chamber occurred between 1861 and 1878, with the emperor requiring different legislation to support his internal and external policies. Finally, parliamentary activities were equally ineffective in controlling the executive and in promoting or controlling legislation. If anything, this latter situation was worse in Austria than in Germany, as the lower house was also paralyzed by its national fragmentation and inability to form cross-national coalitions. Thus, Austria and Germany represent the clearest cases in which not only did executives remain outside the reach of elective chamber control, but legislation also escaped them almost entirely. A growing sense of the powerlessness of elective representatives and of the uselessness of parliamentary activities was inevitably the outcome of such developments. The collapse of democracy in both countries in the 1930s meant that the principle of governmental responsibility and the practice of parliamentary government were not ﬁnally consolidated until after World War II. Finally, Italy has not received in the comparative literature the attention that it deserves as a totally deviant case: Italian liberal democracy collapsed seventy years after the principle and practice of responsible government had been introduced.53 Although historiographic research has convincingly argued that the Statuto Albertino instituted constitutional government rather than parliamentary government, a wider interpretation was immediately given in practice. Historians tend to regard the connubio in the Kingdom of Piedmont – namely, the formation at the beginning of 1852 of an efﬁcient parliamentary majority – as the ﬁrst step toward 52 53

Cervelli (1978: 240–245). In my opinion, Spanish political and constitutional development resembles Italian development more closely than do the German and Austrian cases. A systematic comparison of Italy and Spain in the period 1850–1920/1930 would be particularly revealing in differentiating them from the central empire experience.

POLITICAL INTEGRATION

347

ministerial responsibility. ‘‘From that moment, once the uncertainties of the Calabiana crisis and of the 1857–1858 period had been overcome, it became consolidated practice that government be formed on the basis of the indications of the Chamber of Deputies. The crown, the upper chamber, and the conservative groups surrounding them actually came to be located in a marginal position as far as the political direction of the country was concerned.’’54 The Italian Senate was nonelective but rarely played an important role in limiting the prerogatives of the lower chamber. In reality, the cabinet soon developed a power of senatorial appointment similar to that of the crown, and several prime ministers made wide use of such power in order to ensure that the Senate’s political orientation was compatible with that of the lower chamber.55 The check to the full parliamentarization of the executive was the persistent inﬂuence of the king and the circles around the royal family. Dramatic conﬂicts never developed, but interventions in the life and composition of the cabinets, particularly in war and foreign affairs and ministries, were frequent.56 Prime ministers never completely prescinded from the parliamentary situation, and the king never intervened systematically, disregarding the parliamentary orientation, but the role of the monarchy was important and kept alive discussion about a return to the original design of the Statuto, that is, to a more constitutional monarchy.57 The general situation is best described in the following statement: ‘‘A [king’s] role kept within the limits of a parliamentary regime whenever the political context obliged it to do so; a [king’s] role going beyond such limits (and sometime even beyond the limits of a purely constitutional regime) whenever politically possible; in any case, a role dictated not by the principle and definite acceptance of certain rules of the game, but dictated by the evaluation made in each case of the political convenience to respect them.’’58 Certainly, it should not be forgotten that the role played by the king was instrumental to the nomination of Mussolini as prime minister in October 54 55

56

57

58

Caracciolo (1957: 106) (my translation). The number of new appointments, called infornate, reached 200. The power of appointment of the king and the prime minister was tempered by the autonomous power of the Senate to verify the qualiﬁcations of the newly appointed members. For a list of the interventions of the monarchical circles on the composition of the cabinet and on the resignation of individual ministers disliked by the king, see Fabbri (1991: 103–186). All the conﬂicts between the king and the ministers are listed. The last one was with De Rudini in 1896. During the political and institutional crisis at the end of the century, a prominent political personality, Sonnino, called for a return to a constitutional monarchy with the famous slogan ‘‘Torniamo allo statuto’’ (‘‘Let’s come back to the statute’’). Rotelli (1975: 7–8) (translation and emphasis are mine).

348

THE POLITICAL MOBILIZATION OF THE LEFT

1922 and that he was silent in the face of Mussolini’s dismantling of the liberal regime. Thus, the comparative evaluation of the parliamentarization of the executive in Italy is more complex than in the other cases. This process was never really consolidated, but its causes lie more in the weakness of parliament and of the party system than in the objective strength of the royal powers. It was thanks to the lack of structuring of the Italian partisan alignments in the second half of the nineteenth century and to the lack of clear parliamentary majorities that the king maintained his position of strength. In Table 7.4, I have synthesised most of the information discussed in the previous paragraphs.

FAIR REPRESENTATION DISREPRESENTATIVE DEVICES The greatest obstacle to political representation was the restricted franchise. But even when suffrage was extended, diverting mechanisms were often devised or retained to ensure that it had a limited impact on the political process. In the long run, however, these obstacles were likely to arouse a deep, strongly felt sense of injustice among those who were affected because of the clearly discriminating arguments advanced to support them. In defending the class-weighted voting system in Prussia, one minister argued that classes of people had to vote on the basis ‘‘of their actual importance in the life of the state.’’59 This argument clearly demonstrated the lesser importance of certain individuals and groups and may have created more resentment than exclusion from voting on capacity (cultural and/or economic) grounds. The obstacles were devices that increased the cost of votes and/or the cost of seats for newcomers. These mechanisms persisted throughout the nineteenth century, distorting the free and direct expression of the voter’s choice (indirect voting and open [nonsecret] voting, which violated the one person, one vote principle, and plural voting and curia/estate voting) and hindering the fair transformation of electoral strength into parliamentary strength (disrepresentative nature of certain electoral formulas).60 Oral voting and the showing of hands greatly facilitated pressure and 59 60

Prussian minister quoted in Anderson and Anderson (1967: 307). The information concerning open, indirect, and curia voting have been collected mainly from Goldstein (1983) and Flora, Alber, Eichenberg, Kohl, Kraus, Pfenning, and Seebohm (1983).

POLITICAL INTEGRATION

349

Table 7.4. Timing of responsible government development

manipulation by government ofﬁcials and the local elite, particularly in nonurban settings. During the nineteenth century, these practices were slowly abolished everywhere. This was true in France as early as in 1831, although nonsecrecy remained widespread for a long time and the ‘‘secrecy of the ballot was not well preserved until 1914, when voting ballots and envelopes to put the ballot into were instituted.’’61 In Germany and the Netherlands, provisions were introduced for protecting secrecy at the end of the 1840s. Italy did this in 1861, but ofﬁcial pressures for voting in public in the rural areas of southern Italy were reported up to the beginning of the twentieth century.62 A new wave of secret-ballot provisions were introduced in Sweden, Belgium, Switzerland (at the federal level), 61 62

E. Weber (1979: 271) offers a number of interesting local examples. See De Sanctis (1876).

350

THE POLITICAL MOBILIZATION OF THE LEFT

and the United Kingdom (and Ireland) between the end of the 1860s and the beginning of the 1870s. Norway did the same in 1885. The only latecomers were Denmark, Austria, and Finland, which introduced measures to protect secrecy only at the beginning of the twentieth century. Indirect voting meant that the ballot was cast not for legislative candidates but for ‘‘grand electors,’’ who, in turn, selected representatives to parliament. This two-stage process introduced an additional barrier for new movements and for parties that could not rely on an established and visible social elite, particularly at the local level. In reality, indirect voting was conceived as a ﬁlter against ‘‘dangerous’’ candidates, and the system was more widespread than is usually thought. France, the Netherlands, and Switzerland had used indirect voting in the early nineteenth century, in the periods 1815–1817 (France), 1815–1848 (the Netherlands), and 1815–1848 (Switzerland); once abandoned they never used it again. By contrast, Austria, Finland, Norway, and Sweden retained some form of indirect voting well in to the twentieth century. In Austria, voting was indirect from 1861 until 1901 in the fourth curia and, from 1897 also in the ﬁfth curia, so that the bulk of the electorate was subject to it. In Finland, the vote was indirect until 1906 for the estate of peasants, who obviously constituted the overwhelming majority of the adult population. Norway and Sweden maintained indirect voting up to 1906 and 1908, respectively, even if, in Sweden, the vote was only partly indirect from 1866 to 1908. In Austria, Finland, Sweden, and Norway, indirect voting may be the key to the very late entrance into parliament of the ﬁrst socialist representatives. In Sweden, socialists obtained seats in 1902, thirteen years and several elections after their ﬁnal national centralization. In Norway, seats were obtained only in 1903, after seventeen years of extraparliamentary life. In Finland, the difference was less (eight years) also because the party centralized late (in 1899). In Austria, effective parliamentary entry was also delayed up to the beginning of the nineteenth-century, even though the party had been quite well organized nationally long before this time. However, in the latter two cases, the curia system (about which more will be said later) should also be added to the indirect voting techniques. Indirect voting was not adopted in the imperial elections in Germany, but it was retained in Prussia until 1918. The curia and estate systems assigned a disproportionate number of seats to the upper estates, generally representing aristocratic and wealthy families and the clergy, while formally giving the right to vote to everybody or to a small number of people in each major social group. Such systems were retained throughout most of the nineteenth century in four Western European countries: Sweden until the reform of 1866; Prussia until the

POLITICAL INTEGRATION

351

collapse of the empire; and Austria and Finland until the beginning of the twentieth century, in 1907 and 1906, respectively. The Finnish and Swedish systems were very similar and give a clear idea of how the mechanism worked. The ﬁrst estate allowed direct representation by the male heads of noble families; the clergy estate allowed it by the higher-ranking clergy and included several elected representatives of the lower-ranking clergy; the burgher estate was composed of representatives elected by plural voting and (see later) by burghers meeting professional and income qualiﬁcations; the peasants’ estate was elected with indirect voting (see supra) by independent farmers, with tenants and agricultural laborers excluded. This mechanism yielded four houses, each formally endowed with the same powers by which, in 1900, about 150 noble families had the same weight as the 1,083 enfranchised clericals, the 23,469 eligible burghers, and the 10,184 enfranchised peasants.63 Finally, throughout the nineteenth century, the practice of plural voting – that is, the attribution of extra votes to wealthy and/or well-educated citizens or to representatives of special institutions (churches, universities, etc.) – was continued. This system had the direct effect of separating citizens bureaucratically into voting categories based on class criteria. France had resorted to plural voting only in the 1820s and for a few highly restricted franchise elections. The Swedish eliminated plural voting in the 1866 reform. In the United Kingdom, plural voting continued until well into the twentieth century, when it was drastically reduced in 1914 and eliminated altogether in 1948. However, although in theory it was possible to have up to twenty votes, the weight of extra voting was never overwhelming: In 1911, about 7% of the British electorate cast plural votes as a result of their meeting more than one franchise requirement or of meeting property requirements in more than one constituency. Minor inequalities also persisted in the Irish new state, were progressively removed in 1922, and were completely abolished in 1936. Finland continued to use plural-voting procedures based on professional and income qualiﬁcation for elections in the burgher estate until 1906. In Belgium and Austria, plural voting played a more important role because, rather than functioning as a remnant of a gradually fading tradition, it was either introduced or broadened wider to check democratization tendencies. Thus, it acquired a clear political signiﬁcance that can be described as ‘‘anti-newcomers.’’ In Austria, in fact, plural voting was not widely used up to 1897, after which it was widely expanded in the context of the democratization of the franchise. When an additional ﬁfth curia 63

Goldstein (1983: 13).

352

THE POLITICAL MOBILIZATION OF THE LEFT

with universal and equal suffrage for male citizens over twenty-four years of age was introduced in 1897, the electors of the ﬁrst four curiae all gained a second vote. Thus, between 1896 and 1907, about 40% of Austrian males cast two votes. Even more clear was the political signiﬁcance of the introduction of plural voting in Belgium in 1894. Prior to this time, suffrage had been equal but restricted, and the introduction of the universal manhood suffrage brought with it strong inequalities, adding a vote for house owners and owners of real estate that was worth more than a certain value and two extra votes for persons with higher education and certain abilities. When the two qualiﬁcations were combined, it was possible to reach a maximum of three extra votes. So, while the franchise was democratized, giving the right to vote to 1.3 millions citizens – in contrast to the previous tax-based 136,000 – the wealthier and more highly educated upper classes were compensated through plural voting. In the 1890s, out of the 1.3 million voters, roughly 850,000 had only one vote, 290,000 had two votes, and 220,000 had three, with the result that 500,000 citizens with more than one vote largely outweighed the 850,000 with a single vote. This explains why the early history of the Belgian socialist movement was largely dominated by heated debates and confrontations concerning the equalization of voting rights. In Table 7.5 I have synthesized the major representation inequalities discussed so far. I have also added a column indicating the ﬁnal introduction of proportional (PR) representation, which represents, in all cases but one, the ﬁnal step in the equalization of the voting right (the exception is Belgium, where PR was introduced very early while the plural voting system was retained). It is difﬁcult to reconcile into a single dimension such a wide array of devices and situations. Taking the onset of the socialist movement as a point of reference, we can say that the obstacles to fair representation were many, signiﬁcant, and protracted in Austria and Finland; relatively minor and gradually removed in the United Kingdom and Ireland; numerous but removed early in Sweden (long before the socialist movement appeared); limited to one but protracted in Norway (indirect voting) and in Belgium (plural voting); and absent by the time the socialists started to compete in Denmark, France,64 Germany (although with the important exception of the Prussian state), Italy, the Netherlands, and Switzerland. 64

In France, obstacles to fair representation consisted mainly of widespread governmental gerrymandering in the period 1820–1970s and later of constituency size inequalities that consistently gave extra weight to the countryside and underrepresented the cities and industrial areas; Campbell (1958: 22 and 19–38).

POLITICAL INTEGRATION

353

Table 7.5. Representational inequalities in the lower chambers

MAJORITY AND PROPORTIONAL REPRESENTATION From the point of view of the socialist elite and rank and ﬁle, these devices and obstacles to fair representation had two major impacts. First, there were the psychological effects of unfairness and injustice, a tangible institutionalization of the class structural division of society and a sign of the fact that socialists’ values and interests could not be defended within the framework of the existing system. Where extensively used, such mechanisms shaped the ideological orientation of the socialist movements toward their state and political system in the formative phase before World War I and inﬂuenced the direction of their political action toward this state. The second effect was that of mechanical under representation. These mechanisms delayed the full transformation of the electoral resources of the socialist movement into a direct parliamentary inﬂuence. The signiﬁcance of the ﬁrst effect can be indicated by the effectiveness of the second. Let us therefore look at the underrepresentation of early socialist parties, considering the difference between the percentage of their votes and the percentage of seats won in the lower elective house.

354

THE POLITICAL MOBILIZATION OF THE LEFT

This picture needs be broken down by type of left party and by country. In Table 7.6 misrepresentation ﬁgures are reported by period, type of party, and type of electoral formula.65 With majority systems socialist parties suffered heavy under-representation in the periods before World War I (⫺5.5%) and between the wars (⫺4.35%), enjoying overrepresentation after World War II. In proportional systems they are, on average, always overrepresented in every period. In contrast, communist parties are, on average, always underrepresented, whatever the period or the electoral system. The situation is fairly similar for the other left parties. Their relatively small size reduced their share of the seats even in PR systems. On the whole, PR offered a premium to the socialists that was basically gained at the expense of communist and other left parties. The most signiﬁcant ﬁgure is the 67% underrepresentation of all left parties and of the whole left in the seventy elections that were held under the majority formula between 1880 and 1944. In Table 7.7 average misrepresentations are presented by country, with the usual distinction between type of electoral system and type of party. Data concerning the post–World War II period are not reproduced and do not need to be described or commented on. In the period of party foundation and consolidation before World War I, socialist parties were severely underrepresented in parliaments in Germany, Switzerland, Italy, Norway, Denmark, and Austria. In the ﬁrst three cases, underrepresentation was massive, but the Swiss case is different from the Italian and German ones because at the canton level – which was the most important – proportional formulas were predominant. In Germany and Italy, socialist parties gained, on average, 9% fewer seats than their percentage of votes. Taking into consideration their average size in the period, this means that between one-third and one-half of their electoral support was underrepresented. Even the Norwegian socialists suffered from the late introduction of PR, and presumably also as a result of the persistent practice of indirect voting, even if a modiﬁed proportional system was introduced for local elections as early as 1896. In Denmark, underrepresentation was less pronounced, but the average of ⫺5% refers to a long electoral history of twelve elections. In Austria, there are data only for 1907 and 1911, when underrepresentation was roughly 5%. In all the other countries, underrepresentation was less pronounced. The Dutch, French, Swedish, and British early socialist parties won fewer seats than votes, but the difference ranged only between 1% and 2%. In Sweden and the United Kingdom, this is likely to be due 65

A detailed discussion of the variety of majority systems utilized before World War I is unnecessary to evaluate the macro-outcome of underrepresentation.

POLITICAL INTEGRATION

355

Table 7.6. Disrepresentation by party and by period (% of seats minus % of votes)

Note: Number of elections given in parentheses.

to the fact that the socialists entered the competition late and their early small size could not, in any case, bring about high underrepresentation. A similar argument applies to the weak pre–World War I French Socialist Party. A case worth mentioning is that of the Belgian socialists. In the four general elections held before the war,66 two took place under the majority formula and two under PR. The socialist party was more underrepresented under PR than under the majority formula. This is clearly due to the unusual Belgian reform of 1893, which introduced PR but also heavy plural voting. In the period between the two world wars, PR systems become predominant; of the ninety-two elections held in this period, eighty were held under this formula. The introduction of PR had been easier (that is, less contentious) and earlier where ethnic and/or religious heterogeneity was an issue (e.g., Denmark in the 1850s, with its Schleswig minority; the Swiss cantons at the beginning of the 1890s; Belgium in 1899; Finland, with its Swedish-speaking minority, in 1906). Rokkan has also argued that PR was introduced earlier when the extension of suffrage and the corresponding growth of working-class parties threatened the existence of the traditional regime censitaire parties, for which PR became a way to remain in the ﬁeld (Sweden in 1909 and the Danish lower house in 1915 are the most notable examples).67 Resistance to PR was stronger in larger polities, where the 66

67

More elections took place during this period, but they were partial elections held in different parts of the country to renew half of its parliament. It is misleading to use their results together with those of the general elections. I have therefore considered only the national general elections. Rokkan (1970f: 147–168).

356

THE POLITICAL MOBILIZATION OF THE LEFT

Table 7.7. Mean misrepresentation of the left by country, period, and electoral system

Note: Number of elections given in parentheses.

central government was able to mobilize greater resources against the movement demanding PR (the United Kingdom and France, in particular, but also Germany and Austria, where, without the collapse of the empires, resistance to PR would have prevented its adoption for much longer). After World War I, majority systems were used only in France and the United Kingdom and in the single election of 1918 in Norway. In this period, the interesting comparison is instead between the fate of the socialists and the communists. In the United Kingdom, the Labour Party was, on average, underrepresented by 5%. This, however, was the result of the dispute with the liberals for the position of second party in the British party system and of an alternation of elections with fair representation (1923 and 1924), clear underrepresentation (1918, 1931, and 1935), and equally clear overrepresentation (1929). In France, the majority formula, used three times, did not hit the socialists hard, but instead drastically curtailed the parliamentary representation of the communists (⫺6.3%). In Norway, the last majority elections of 1918 resulted in the catastrophic underrepresentation of the Norwegian Labor Party parliamentary force (⫺17.3%). In PR elections, the socialists were in general slightly overrepresented (the exceptions are Italy, Ireland, and Switzerland), and the communist parties were systematically, even if modestly, underrepresented. France is the only signiﬁcant deviation from this pattern; there, both the communist and the

POLITICAL INTEGRATION

357

socialist parties continued to be heavily underrepresented under the PR formula, by roughly 5.5% of the seats each, which together made for a substantial average underrepresentation of the left of approximately 9%. This is due to the fact that most of the elections I have classiﬁed as involving PR formulas in fact used modiﬁed PR, and the alliance premiums that existed were often deliberately devised to penalize the left parties.68 To the reformist socialists of the SFIO – involved in 1919 in the delicate internal debate with their radical wing on adherence to the Third International – nothing could have been more discouraging than the result of the 1919 elections, the ﬁrst to be fought with the long-desired PR system. Increasing its electoral strength by almost 5% with respect to the 1914 elections, the SFIO lost 7% of the seats it had previously held and was drastically underrepresented: 20% of the votes and 10% of the seats. The disillusionment among the socialists and the discrediting of the parliamentary road to socialism were enormous. After World War II, there were only two signiﬁcant electoral reforms aiming to underrepresent the left: the electoral law of apparentement in France in 1951, which was implemented successfully, and the Italian 1953 reform, which failed. Both reforms gave a seats premium to the party coalition that reached a majority threshold. However, while in France alliances could be concluded at the constituency level, where the threshold was applied, in Italy the majority premium was triggered only if the alliance reached more than 50% at the national level. The majority premium was represented by a pool of national sets in Italy, while it consisted of all seats of any given constituency in France. Finally, there was a crucial political difference: The alliance of centrist parties included the socialists in France and was mainly directed against the two extreme parties, the communists and the Gaullists. In contrast, the centrist alliance in Italy excluded the extreme right and both the communist and the socialist parties. These differences explain why the French reform succeeded and the Italian one failed, the 50.1% threshold was not reached, the majority premium was not awarded, and the law was later repealed. In France, no alliances were made in twelve constituencies; in seven, two blocs faced each other with no strong misrepresentation effects; in seventy-six constituencies, one alliance alone was formed, and in two-thirds of them this was an alliance of centrist parties against the extreme parties. The new law deprived the PCF of about seventy-one seats it would probably have gained under the 1946 law. If we add that the left, this time the entire socialist 68

On the continuously changing electoral laws of this period of the Third Republic, see Campbell (1958).

358

THE POLITICAL MOBILIZATION OF THE LEFT

and communist left, was again largely underrepresented under the new majority double-ballot system reintroduced by De Gaulle in 1958 and continued to be so until the 1970s, it is clear that the issue of representative fairness remained central to the politico-institutional debate in this country throughout the century.

ACCESS TO EXECUTIVE POWER How long did the socialists have to wait before being given cabinet responsibility? For many years, socialist groups debated whether to participate in bourgeois institutions: whether to take part in elections; whether to ﬁll positions of responsibility in the legislative chambers; whether to enter electoral coalitions with bourgeois parties; whether to vote for laws favorable to workers and, in particular, budget or military credit laws; whether to support sympathetic bourgeois governments; whether to formally enter parliamentary majorities; and, ﬁnally, whether to accept ministerial responsibility. The debate began with the split of the anarchists on the matter in the First International; it continued with Rosa Luxemburg’s critique of ‘‘parliamentary cretinism,’’ concerning the real autonomy of the parliament and legislative action; and it was revived after World War I in the early communist movement in Western Europe (for instance, by Bordiga in Italy). However, whatever the internal dissents and debates and the bargainability of the issues they involved, any socialist inclusion in parliamentary majorities and cabinets depended primarily on the numerical relationships among the parties in parliament and on the reaction of the other political forces. Socialist access to these parliamentary majorities should not be understood as separate from the more general problem of the legitimation of the other political forces, such as extreme right parties, communist parties, and substate nationalist movements. In general, therefore, the chance to participate in government was a function of (1) the extent to which socialists were able to ‘‘shrink’’ the governmental space of the nonsocialist parties. This, in turn, depended on their electoral force but also on the governmental space left once the extreme antisystem forces had been taken away; (2) the level of electoral fragmentation and political division of the nonsocialist forces; (3) the ideological bargainability of policy alternatives; and, ﬁnally, (4) the severity of the international environmental pressures that might ease or impede their internal acceptance. In this section, I review the patterns of executive access and the governmental record of the European left, discussing its causes and consequences.

POLITICAL INTEGRATION

359

ENTRY PATTERNS In Table 7.8 I have summarized the basic information about socialist and communist party access to direct cabinet responsibility. Leaving aside the cases of direct socialist support to cabinets in the pre–World War I or war period in neutral Sweden and Denmark and in belligerent France, Finland, and the United Kingdom, the ﬁrst true executive experiences of the socialists took place in Belgium and Sweden during the war years, and in Denmark, Austria, and Germany immediately afterward. The circumstances of these entries were diverse. In Belgium, the foreign aggression fostered an all-party coalition and led to universal suffrage; this is the clearest case of the role of an external military threat favoring internal democratization.69 The German invasion was equally important, as the clandestine contacts between leaders of industry and Catholic, socialist, and liberal unions determined the basis of the pact that would be implemented after the war through important advances in social legislation and collective bargaining.70 In Austria and Germany, the socialists entered cabinet by default, so to speak, as a result of the collapse of the prewar political order and delegitimation of the established political groups and elite. They were thus compelled to inherit the political catastrophe of military defeat associated with the regime’s breakdown, and their presence in the cabinet coincided with the introduction of parliamentary responsibility. Intermediate entries took place during the 1920s in three countries: Finland, Norway, and the United Kingdom. These were normal entries in the sense that no external threat, military event, or regime change fostered them, even if they were all very brief and they were all minority singleparty cabinets. The Finnish Tanner cabinet of December 1926, for example, enjoyed the support of 60 out of 200 members of parliament (MPs) and lasted for only four days, while Britain’s ﬁrst MacDonald cabinet in January 1924 had the support of the 191 Labour deputies in a 615-MPs House of Commons71 and lasted for fourteen days. The January 1928 Hornsrud cabinet in Norway enjoyed the support of 59 MPs out of 150 and lasted for nineteen days. In these cases, strong, rapidly growing labor parties were tested in single-party minority cabinets with which other groups refused to cooperate. They generally presented clear-cut socialist 69 70 71

On the ‘‘democratizing’’ role of world wars, see Therborn (1977: 3–41). Lorwin (1966: 165). The external and informal support of ﬁfty-nine Liberal deputies was not enough to ensure a majority.

360

THE POLITICAL MOBILIZATION OF THE LEFT

Table 7.8. Executive entry of socialists and communists

governmental programs and were immediately disposed of once it was evident that they did not accept the constraining conditions of their minority status. France provides the only example of socialist access to a cabinet in the 1930s.72 The four Popular Front cabinets between June 1936 and April 1938 (two formed by Blum and two by Chautemps) were coalition cabinets involving socialists and radicals, in which the former accepted a juniorpartner role although enjoying a relative majority in the National Assembly. At least three of them were minimum winning coalitions,73 but 72

73

One should not consider as socialist participation in cabinets the entry of independent socialist personalities, such as Alexandre Millerand in the Waldech-Rousseau cabinet. These actions generally embittered the debate within the socialist movement about the usefulness of participating in bourgeois governments rather than opening the way to such participation. Given the high ﬂuidity of the party and parliamentary group labels of French radical and republican forces even in the late phase of the Third Republic, it is difﬁcult to decide exactly whether a coalition is minimum winning or surplus. Moreover, support of deputies was often granted and withdrawn on a strictly individual base. On the whole, however, Blum’s two cabinets and the ﬁrst of the Chautemps cabinets were the result of a relatively close majority in the lower house.

POLITICAL INTEGRATION

361

although they lasted longer than those of other countries in the 1920s (about ﬁve-and-a-half months each), they were unstable and ineffective. The main reason for this was politico-institutional: The legislative discipline of the right wing of the radical and republican groups was unreliable,74 and there was no clear majority in the Senate, which systematically rejected the main legislation of the Popular Front. Among the socialist latecomers to executive responsibility, participation in the Netherlands and Switzerland was fostered by an external threat. The ﬁrst Dutch socialist ministers were nominated in August 1939, when the country was preparing to face Hitler. The Swiss Socialist Party, on the other hand, had tried to win one vacant Federal Council seat in 1929 without gaining the majority support of the Federal Assembly. This instead was achieved in the four-party ‘‘magic formula’’ coalition in 1943, in the most delicate period of Swiss World War II neutrality. The Swiss case is, however, exceptional in that (1) socialists had acquired executive responsibility in some cantons long before 1943; (2) the principle of cabinet responsibility is not applicable to the unusual pattern of the Swiss executive–legislative relationship;75 (3) the executive branch of government was weak as a result of the division of powers between cantons and municipalities; and (4) the forms of popular referendum were powerful counterweights to the role of the cabinet and offered parties and groups not represented in the cabinet (or even in parliament) opportunities to inﬂuence legislative decision making.76 As a consequence, in Switzerland the 74

75

76

The roll-call analysis by Warwich (1977: 154) proves that in legislative behaviors the Marxist–non-Marxist cleavage far outweighed the cabinet–opposition divide. The peculiarities of the Swiss federal executive are many. Each member of the Conseil Federal is elected individually by the chamber for four years, and there is never a total turnover; each member retires and is substituted for independently of the others. The presidency is rotating. A conﬁdence vote is absent, and if a bill fails in parliament or in a popular referendum, the composition of the council is not affected. When a federal councilor once resigned after the failure in parliament of a major proposal originating from his department, his behavior was regarded as inappropriate by most political and social groups. See Fleiner-Gerstern (1987). Popular referendums (since 1874 at the federal level but since 1860 at the canton level) and constitutional popular initiatives (since 1891) have been judged by Neidhart (1970: 96–122 and 313–319) as ineffectual in defending the interests of the lower classes. Yet, among the ﬁrst national laws approved by legislative referendum was an 1877 factory regulation bill setting a maximum eleven hours of work, and banning children’s factory work and a 1879 bill abolishing capital punishment. From that time to the socialist entry into the Federal Council in 1943, the amount of legislation passed by popular referendum with socialist support was considerable. Before World War I, socialists actively participated in four constitutional initiative referendums. Only the last, concerning the introduction of PR, was successful, with the help of the Catholic–conservatives. Between 1891 and 1979, of sixty-seven initiatives seven emanated from the PSS or were strongly supported by it.

362

THE POLITICAL MOBILIZATION OF THE LEFT

cabinet was never viewed as the center of power or of state regulation and legislative production, the role of the central executive was less signiﬁcant, and the outsiders’ inﬂuence was more important compared with the other typical parliamentary democracies of the period. The last socialist parties to gain access to cabinet responsibility were those in Italy77 (1945) and Ireland (1948). Both did so in broad coalitions: the post–World War II antifascist regrouping of liberal, Catholic, and socialist/communist forces in Italy78 and the anti–Fianna Fail all-party coalition of 1948 in Ireland, whose majority depended, however, on the benevolent support of the independent MPs. If one compares the moment of ﬁrst entry with the moment when the socialist parties became a regular governmental party, the biggest time lag is in Austria and Italy. In Austria, the post–World War I socialist cabinet’s experience was short and terminated in 1920. From that time to the collapse of democracy, the socialists were given no other chance to share executive power. In Italy, the large 1945–1947 coalition governments were followed by almost two decades of socialist exclusion before the party became a regular coalition partner from 1963. On the whole, the experience of socialist executive entry falls into three groups: 1. Complete denial: the Austrian case between November 1920 and 1934; in the Netherlands, too, religious parties considered cooperation with the socialists a rare exception; in Italy, of course, fascism ruled it out completely. 2. Introduction to executive responsibility in wide coalitions where socialists played the role of a junior partner or accepted this role even if their electoral strength gave them the right to a more prominent share: Belgium, Germany, Denmark, Switzerland, Netherlands, France, Italy (1963), Ireland. 3. Admission to executive power as minority cabinets during the 1920s: the United Kingdom, Finland, and Norway, and also in Sweden to a certain extent, even if the ﬁrst Swedish cabinet with socialist ministers was a coalition cabinet with the liberals. 77

78

In Italy, the Socialist Party split down the middle during the Cold War, giving birth to the Social Democratic Party (PSDI), which continued to participate in the cabinet between 1947 and 1963, while the main socialist party remained in opposition. As it was impossible to combine the cabinet experiences of two socialist parties, in this section the data refer to those of the main and larger one, the PSI. However, it should be noted that during the liberal period the socialists were offered, and rejected, cabinet ofﬁce in 1903 and 1911.

POLITICAL INTEGRATION

363

Only ﬁve socialist parties have managed to exercise single-party majority government since their ﬁrst cabinet entry: in Norway, Sweden, the United Kingdom (1945), Austria (1971), and France (1984). The Finnish socialists achieved an absolute majority in the Diet in 1916 (51.5% of the seats) but were not able to exercise executive responsibility under the Russian control of the country. Both coalition and minority governments implied controlled access. Thus, socialists were legitimized into an executive responsibility position ‘‘under conditions of enforced moderation.’’79 Communist participation in the cabinet was more reduced. If we exclude the very short periods of executive presence of the Danish, Dutch, and Norwegian communists, these parties shared ministerial responsibility in coalition governments in only ﬁve countries after World War II, and only in one did they become a regular coalition partner. In Austria, Belgium, Finland, France, and Italy, communist ministers entered coalition governments, remaining there for a minimum of eight months to a maximum of twenty-ﬁve. All these experiences terminated with the deterioration of the postwar antifascist coalitions and the beginning of the Cold War. Indeed, by 1948, all the Western European communist parties were in opposition, and only in Finland did they come back as a regular coalition partner after 1966. The brief participation of four communist ministers in the two French Mauroy cabinets for twenty-three months (May 22, 1981 to March 22, 1983), although they should be mentioned, did not mean that the party was a regular coalition partner. From the point of view of the inﬂuence that communist parties had in the cabinet, we can distinguish between (1) Switzerland, Ireland, and the United Kingdom, where the communists exerted no signiﬁcant inﬂuence whatsoever; (2) Belgium and Austria, where the communists formed part of the cabinet for a short time after World War II, to be excluded shortly afterward and to decline in electoral and coalition inﬂuence from then on; (3) Sweden, Norway, and Denmark, where communists exercised some strategic inﬂuence after World War II, when their votes made the difference between majority and minority social democratic cabinets; (4) France and Germany during the interwar years and France and Italy after World War II, where strong opposition communist parties pushed the socialists into government with centrist and moderate forces, reducing the socialist role to junior partners and, at the same time, preventing the moderating effect of such experience. The presence of a strong left competitor that forced these socialist movements to form coalitions with moderate forces in which they were the junior partner contributed to a schizophrenic 79

Daalder (1967: 59).

364

THE POLITICAL MOBILIZATION OF THE LEFT

division between their ideological identity and political practice; and (5) Finland, where communists played an important role in coalitions during the 1944–1948 period and again from 1966 on.

DURATION Once the socialist parties had gained access to cabinet responsibility, did their position as cabinet members last? After 1966, the socialist presence or absence in cabinets was the result of the normal interplay between electoral results and coalition bargaining, and no exclusionary clause or peculiar institutional difﬁculty of entry is visible. Therefore, in Table 7.9, the length of socialist government is reported in percentages of time and of all cabinets only for the 1918–1966 period. In these forty-nine years, there were 378 cabinets (excluding Switzerland).80 Of these, 224 (59.3%) did not include socialist or communist party representatives; 11 were not party cabinets, but caretakers cabinets;81 143 included socialist ministers and 15 included communist ministers, all of which also held socialists. On the whole, the two main components of the European left participated, respectively, in slightly more than a third of all cabinets (37.9%) and in about 4.0% of them. Of course, these proportions were different in the ﬁrst (1918–1944) and second (1945–1966) periods: all communist participation took place after World War II, while socialists participated in only 57 of the 216 cabinets between the wars (i.e., 26.4%) and in 86 of the 162 cabinets in the 1945–1966 period (i.e., 53.1%). In the following period, between 1966 and 1985, the socialist presence in cabinets increased only modestly as a result of the decline in the Scandinavian countries and the increase in Italy, Germany, the United Kingdom, and Finland. With only two exceptions (France and Germany), the length of the 80

81

The main sources for the elaboration of data in this section are Beyme (1973), Flora, Alber, Eichenberg, Kohl, Kraus, Pfenning, and Seebohm (1983), Woldendorp, Keman, and Budge (1993), and Paloheimo (1984). I have reduced cabinet changes to a minimum: Cabinets persisting throughout elections are counted only once, and minor reshufﬂes are not considered. External support and parliamentary toleration are not considered, either for the party composition of the cabinet or for the level of parliamentary support of the cabinet. The level of parliamentary support of the cabinet is reckoned in the ﬁrst chamber at the moment of its installation (even though cabinets may depend on support in both chambers, as in Italy and Sweden up to 1970). Switzerland is excluded because the parliamentary responsibility of the cabinet is not comparable to that of other countries. One in Denmark and three each in Finland and Sweden. The four no-party cabinets in Germany, however, were the presidential cabinets of 1931–1933 from Bru¨ning to Hitler. Another caretaker cabinet was Churchill’s ‘‘election-waiting’’ cabinet of 1945, which was, however, Conservative in its composition and therefore, strictly speaking, not a noparty cabinet.

POLITICAL INTEGRATION

365

Table 7.9. Socialist party presence in cabinets (1918–1966)*

* Considering World War II war cabinets. a Including the four months of the second 1931 MacDonald cabinet, which was supported only by parts of the Labour Party. b Considering the SFIO as participating in the Popular Government cabinets.

socialist presence in the cabinet is higher than the percentage of such cabinets over all cabinets. This difference is particularly notable in Sweden, Norway, and Austria. These ﬁgures suggest that the entry of socialists into cabinets determined an increase in cabinet duration and a generally higher cabinet stability. While the all-cabinets’ average duration was 17.4 months, that of the 162 cabinets that included socialists was 20.6 months and that of the 213 cabinets without socialists was only 14.9 months. In both Sweden and Norway, bourgeois cabinets had, on average, less than half the average tenure of cabinets with socialists. In Norway between 1919 and 1933, fourteen weak minority governments were formed as the result of a situation in which liberals and the Labor Party were unwilling to govern together, as were the liberals, conservatives, and agrarians. Disputes over agrarian and urban conﬂicts of interest added to the strong political division between the urban bourgeois classes and the mainly agrarian middle classes. This situation delayed the entry of Labor into the cabinet in Norway, but also exhausted the nonsocialist camp and paved the way for the ensuing cabinet control of the Norwegian socialists, which was both extended and continuous. Marked stabilization and increased duration were not, however, the

366

THE POLITICAL MOBILIZATION OF THE LEFT

outcome of socialist access to executive responsibility everywhere. The Danish Social Democrats formed part of the cabinet several times, but their presence was not as continuous as in the other Scandinavian countries because nonsocialist cabinets – whose tenure was as long as that of the socialist ones – alternated more frequently with socialist cabinets. Thus, while on the one hand most socialist cabinets were minority cabinets in Denmark, dependent on the toleration or external support of other political forces, on the other hand, in contrast to the other two Scandinavian countries, a liberal–conservative coalition remained a viable alternative to the traditional alliance of social democrats and radicals. In the Netherlands, Belgium, and Italy (since the mid-1960s), the possibilities of an alliance switch remained by which liberals could replace socialists as an alternative partner to the religious center parties. Only in Germany and France was the average duration of cabinets, including those of the socialists, shorter than that of cabinets without socialists. The same was true for the United Kingdom between the world wars, while after 1945 there was no considerable difference between Labour and Conservative cabinets. In the four countries where the left was deeply divided down the middle, cabinets with socialist participation were either more unstable than bourgeois cabinets (France and Germany) or the difference was marginal.

COALITION PATTERNS: CABINET FORMAT, STATUS, AND IDEOLOGICAL COMPOSITION The third relevant question concerns the kinds of cabinets the socialists managed to enter. This question leads to three subquestions. First, how many partners did socialists have in the cabinet, that is, what was the cabinet format? Second, what was the status of the cabinet (i.e., singleparty majority cabinet, oversized coalitions)? Third, what was the dominant ideological composition of the cabinet? Table 7.10 indicates, in percentages the breakdown by country of cabinets with a socialist presence according to the number of parties represented in them. No-party cabinets were formed in only four countries and were numerous and signiﬁcant only in Finland and Germany. In Finland, only the parties have been counted, although most cabinets also included a number of independent personalities and single-party cabinets in particularly often included experts. In Germany, the four no-party cabinets represent the presidential cabinets of the immediate pre-Nazi period, while in Sweden there were three short caretaker cabinets at the beginning of the 1920s and in 1932. The executives in Ireland, the United

POLITICAL INTEGRATION

367

Table 7.10. Percentage of all cabinets including the socialists by number of parties in cabinet (1918–1966)*

* Excluding no-party/caretaker cabinets (De ⫽ 1; Fi ⫽ 5; Ge ⫽ 4; Sw ⫽ 3). a excluding the Delacroix 1918 cabinet. b Only after 1944. c beginning with the Cosgrave cabinet on December 6, 1922.

Kingdom, and Norway were dominated by single-party governments.82 In the Netherlands, Finland, France, Belgium, Germany, and Switzerland, the dominant executive format was a coalition government, but in Belgium a two-party coalition pattern dominated, with alternation between Catholic– socialist and Catholic–liberal alliances, while in the other countries, coalitions were formed by a larger number of parties. Sweden, Denmark, and Italy (whose data refer only to the post–World War II period) had both single-party governments and coalitions for long periods, and Austria switched from one-party rule until 1934 to a grand postwar coalition until 1966. The socialists entered 143 of these 378 cabinets. Excluding the cabinets for which the exact number of parties is impossible to calculate,83 socialists ruled alone in single-party cabinets on only twenty-eight occa82

83

In Ireland, the association of the Farmer Party with Cumann na nGaedheal from 1927 to 1932 is often regarded as a two-party coalition. Actually, this was a single-party cabinet, as the Farmer Party had only one parliamentary secretary in the government. The Italian Parri and De Gasperi cabinets and the French De Gaulle cabinet immediately after World War II.

368

THE POLITICAL MOBILIZATION OF THE LEFT

sions, all of which were concentrated almost entirely in the three Scandinavian countries and in Britain. They held executive power in two-party coalitions thirty-four times (23.8% of all socialist cabinets), this pattern of coalition being predominant in Austria and Belgium and sometimes also in Sweden and Denmark. In the other countries, almost all experiences of socialist participation in the cabinet were in large coalitions of three or more parties. Altogether, this pattern represents 55.2% of all the socialist participation in government. In Table 7.11 all the cabinets of the period and socialist cabinets are classiﬁed according to parliamentary status: (1) single-party majority cabinets; (2) minimum-winning coalition (MWC) cabinets;84 (3) surplus coalition cabinets; (4) single-party minority cabinets; (5) multiparty minority cabinets; and (6) war cabinets. Next to the number of cabinets of each type, the table reports the cumulative duration of this formula for each given country. A comparison of these data with those concerning all cabinets (i.e., those with and without socialists, not reported here) allows an evaluation of the extent to which cabinets with socialists deviated from the general pattern of the country. Almost half of the socialist governmental experiences were in oversized coalition cabinets (46.2%). In one-quarter, the socialists were part of more vulnerable cabinets: single-party minority (13.3%), minority coalition (4.2%), or war cabinets (6.3%). The remaining one-quarter were cabinets in which the socialist party held a stronger position: single-party majorities (6.3%) and MWC cabinets (21%). Ireland was dominated by single-party cabinets, most but not all of which were majority ones. With a single party approaching the majority-of-seats threshold and with a predominant anticoalition political culture, minority cabinets were often more viable than the necessarily complex coalitions of all the others. Indeed, it was only an alliance of all the other parties against Fianna Fail that enabled the Labour Party to be an indispensable ally in this period. This situation made it difﬁcult, if not impossible, for the weak Labour Party to carve itself out a role of importance in governmental coalition. There were, in fact, only two experiences of coalitions, one MWC and one oversized. Labour was a member of both, but its presence was crucial in only one. In Austria and Italy, socialist parties formed part of surplus coalition cabinets only. In Austria, surplus coalitions were the most popular formula of government, but there were also several single-party and multiparty majority cabinets, MWCs, and single-party minority cabinets, accounting 84

I consider as a minimum winning coalition all those coalitions for which the withdrawal of a partner determines a minority parliamentary status. There are other ways to deﬁne MWC. For a complete discussion, see Lijphart (1984: 46–66).

POLITICAL INTEGRATION

369

Table 7.11. Cabinets with socialists, by type and by total months in office

altogether for almost 60% of all the cabinets and 40% over all time. The socialists entered only surplus coalitions, however. Similarly, the Italian cabinets were of all types except for the single-party majority cabinet, but the socialists entered only surplus coalitions. In both cases, the executive visibility of the socialists was quite weak, even though in the Austrian pattern there was a more balanced relationship with the Christian Democrats than in Italy. Even in Finland and the Netherlands, most socialist participation in cabinets took place in oversized coalitions. Finland experimented with a large variety of coalition patterns (excluding single-party majorities). It had either minority or large majority cabinets, and in both cases ministerial stability was very precarious. However, the entire socialist experience is concentrated in sixteen surplus coalitions, with only one MWC and two single-party minority cabinets. In contrast, the Netherlands has a straightforward coalition pattern in which the number of surplus coalitions is on a par with MWCs. Once again, socialists have participated in only the latter, with one single exception of an MWC. In Belgium, as in the Netherlands, majority cabinets and the tendency to form them on a very wide basis were prevalent. Because no party approached the 50% threshold, minority governments were unacceptable and attempts to form them were immediately discontinued (there were only three short-lived single-party minority cabinets in Belgium and none in the Netherlands). In Belgium, however, MWCs were more numerous and of longer duration than the surplus ones. Yet, with respect to the dominant pattern in the country, cabinets with socialists were surplus coalitions more frequently and for longer periods.

370

THE POLITICAL MOBILIZATION OF THE LEFT

France shares with Belgium and the Netherlands a dislike for minority cabinets, either single or coalition. Socialist cabinets were surplus coalitions rather than MWCs. The latter six experiences, including three of the four crucial Popular Front cabinets, lasted for only 25 months out of a total of 114 for this type. Germany, too, is totally dominated by coalition cabinets, most of the time majority but also including ten multiparty minority cabinets. The few German socialist cabinet experiences were concentrated in majority coalitions, with an equal distribution between MWCs and oversized formations. The Scandinavian pattern can be set apart from the continental one because of the absence of surplus coalitions. The predominant formula was either that of single-party majorities (all socialist cabinets in Norway and Sweden), MWCs (the socialists being present in all of them except for one in Norway), or single-party minority cabinets (half of which were socialist). In Denmark, there were a few multiparty minority cabinets, all of which were formed with socialists, while there were none in Norway and Sweden. In a nutshell, then, Scandinavian socialism, although manifesting a clear preference for single-party governments that gave rise to many minority governments, especially but not exclusively in the interwar period, had a more ﬂexible coalition strategy than British socialism. In Sweden, in the 1920–1932 period, none of the parties had a majority, and any coalition between the socialist and other parties was fraught with difﬁculty. At the end of the 1920s, talks with the liberals failed, the nonsocialist parties were also unable to ﬁnd ground on which to cooperate, and all the political parties seemed to prefer to stay out of government.85 Because every election to the lower house turned into a defeat for the governing party, a change of government occurred at least once every year. Although the liberals were in a better position to form a government than the socialists or conservatives, by relying on shifting left or right support, it was not unusual for the conservatives to join forces against the incumbent liberal government. Majority coalition parliamentarism prevailed in the 1933–1941 period as a result of the alliance between social democrats and agrarians. In autumn 1932, the socialists accepted governmental responsibility; they abandoned their previous policy of nationalization and instead emphasized welfare state policies, thereby reducing the antagonism of all the other parties and, in particular, the center-agrarians. In 1941, a period of single-party majority rule began, with the socialists gaining a majority in the lower house (in 1942, they also gained a majority in the upper house). In Denmark, as in Sweden, the issue was primarily one of cooperation 85

Stjernquist (1966: 122).

POLITICAL INTEGRATION

371

with the center parties, but the formulas experimented with were more varied, the alternation in coalition more frequent, and the cabinet tenure of the socialists less safe than in Sweden and Norway. In Norway, after the radical phase of 1921–1923 and the short-lived 1928 minority cabinet, the party remained divided as to how to interpret the results of its brief governmental experience: that is to say, the impossibility of holding power under capitalism or simply the uselessness of being in government without a solid majority. Its 1930 electoral program was even more radical than that of 1927–1928; it was only in 1930–1933 that the economic crisis and the fascist challenge in Europe triggered a clear process of deradicalization that – together with the 1933 electoral victory on a moderate platform and the 1935 alliance with the agrarians – started the long labor season of almost twenty years of single-party majority cabinet (1945– 1963). Throughout the 1919–1935 period, the liberals, agrarians, and conservatives alternated in single-party minority cabinets.86 Thanks to their electoral strength, the lack of internal division, and the divisions of their adversaries, Scandinavian socialists were also able to hold their own in minority governments (this was also the case in Denmark, although to a lesser extent); this allowed them to escape the dilemma of coalition. The British Labour Party could compensate for electoral weakness by means of the bonus of the majority system when the electoral climate was favorable; thus, it was able to avoid being forced into coalitions and it managed to form single-party majority cabinets with clear-cut socialist programs. However, these conditions, combined with the cohesion of the conservatives, imposed a price of limited cabinet tenure. It is paradoxical that, notwithstanding a number of favorable conditions (a balanced relationship between the extraparliamentary and parliamentary parties; unwillingness of the unions to engage directly in politics; federal and indirect structure and weak institutionalization of the party organization;87 internal power structure both congruent with and functional to the parliamentary system), the governmental record of the Labour Party was poor, as can be seen in Table 7.11. After the short-lived MacDonald experience in 1924, a minority Labour government came to power from 1929 to 1931 and from 1976 to 1979. For part of these periods, that is, for almost ten months in 1930–1931 and eighteen in 1978–1979, the Labor government remained in ofﬁce thanks to an informal agreement with the Liberals. The 86

87

To be precise, of the twelve bourgeois cabinets before the 1935 Labor–Agrarian alliance, eight were single-party minority cabinets and four were two-party (Conservatives plus National Liberals) minority cabinets. On this point, see Panebianco (1982: 167–182).

372

THE POLITICAL MOBILIZATION OF THE LEFT

pact was kept secret in 1931–1932; however, in the 1970s, when the same agreement was made public, its importance was diminished by the continuous assertion that no concessions were being made to the liberals. Thus, no real formal coalition was ever formed. Only on two occasions after 1918 did the Labour Party manage to retain a majority large enough to permit the implementation of its policies for an entire legislature: from 1945 to 1950 and from 1966 to 1970, that is, for nine out of almost ninety years. On a number of occasions – from 1930 to 1931, from 1950 to 1951, from 1964 to 1966, and from 1974 to 1979 (eleven years altogether) – there were de facto minority Labour governments or governments whose majorities were so small as to prevent the party from implementing its policies and remaining in charge of the legislature. In other words, for more than half of its entire period of governmental control, the Labour Party did not have a safe majority in the Commons. Leaving aside war governments, it was in power for approximately a third of the period under discussion, and for more than half of this time, it had no majority or one that was unsafe. In conclusion, the governmental record of the Labour Party is modest and has often been overestimated in a comparative perspective. This poor control was due to the compact nature of the conservative opposition and to the Labour Party’s historical unwillingness to enter into formal coalitions with the Liberals,88 despite the fact that from 1920 to 1931 and again from 1960 to 1970, the latter regarded themselves as a center–left party, generally willing to cooperate with the Labour Party in exchange for a reform guaranteeing more equal representation. At the same time, it cannot be argued that its preference for single-party cabinets and its reluctance to form coalitions was the result of its desire to put its manifesto into practice as it stood, since too often, it was unable to do so without calling on external support. Leaving aside cultural explanations, it remains a mystery why the British Labor Party systematically refused any alliance or coalition with other parties, and in particular with the Liberals, giving rise in this way to a Conservative predominance in the executive.89 Finally, let us consider the types of political coalitions in which socialist parties participated. Given the unreliable nature of the pre–World War II cabinet data, I have classiﬁed the 378 cabinets between 1918 and 1966 qualitatively rather than resorting to formalized criteria.90 Cabinets are 88 89

90

On the anticoalition attitude of the party, see Marquand (1981: 166–176). The debate on the payoff of an alliance with the Liberals, based on a few key mutual and tacit withdrawals of candidates, has recently been reopened in a few articles: see Dent (1993: 243–251); and Sharp (1994: 107–131). The political connotation of a cabinet based on the predominance of its internal left– right orientation has recently been operationalized into a ﬁve-point index by scholars

POLITICAL INTEGRATION

373

classiﬁed as either right-wing or left-wing if their parliamentary support was ensured for more than two-thirds of their tenure by right or left parties; center–right and center–left if centrist forces were represented together with right or left parties, respectively; and balanced when centrist forces were predominant or, alternatively, when both left and right parties joined them in the coalition. I have been able to classify only 311 of the 378 cabinets of the period: The 1919–1922 Italian and the 1919–1939 French cabinets proved impossible to classify precisely. Of the 143 cabinets with socialist ministers, 10 either had no precise characterization (war cabinets) or this kind of description was impossible. Of the remaining 133 cabinets with socialist ministers, a large number of them (45.1%) were balanced, that is, cabinets in which either the center political parties dominated or in which there was an equilibrium between center, left, and right parties. In addition, there were four center–right cabinets (2.8%). The conclusion is that the European socialists participated in center–left or left cabinets in only half of their executive experiences: 29.3% of cases (thirty-nine cabinets) were center–left cabinets; 22.6% (thirty) were predominantly left cabinets. Table 7.12 ranks the countries, going from the most adversarial leftversus-right to the more centrist pattern. The thirty left cabinets are almost entirely concentrated in Britain and Scandinavia. In these countries, cabinet participation was unlikely to precipitate internal conﬂicts and divisions, as in all cases it faced an opposition that was almost entirely rightwing. In Belgium, there was only one left cabinet (the minority single-party Spaak cabinet of 1946), but socialists tended to participate in center–left cabinets more often than in balanced ones. This is largely due to the three-party conﬁguration of the party system, in which the socialist– Catholic cabinets should be viewed as center–left (as they excluded the liberals) and the socialist–liberal cabinets as balanced (even if they excluded the Catholics). In all other countries, socialists entered balanced cabinets most of the time. In the countries with denominational parties (Austria, Germany, Italy, and the Netherlands), this was in practice the only possibility.91

91

mostly interested in relating the degree of party control of a cabinet to its policy preference and/or policy performance. For this index, see Castles (1982) and Keman (1988: 192–216). Scores for the ideological makeup of post-1945 cabinets are available in Woldendorp, Keman, and Budge (1993). This analysis has not been done for the interwar period. Note that the criteria for classifying the ideological completion of cabinets make the Austrian socialist–Catholic coalitions balanced, given that during this period no relevant right of the Catholic Party existed. A similar coalition is classiﬁed as center–left in Belgium, given its exclusion of the liberals.

374

THE POLITICAL MOBILIZATION OF THE LEFT

Table 7.12. Ideological composition and duration (in parentheses) of cabinets with socialist parties (ranked by degree of adversarial coalition pattern)

Note: excluding war and caretaker cabinets.

In Finland, and also in Italy and France, the division of the left and the post-1947 opposition of the strong Communist Party increased the impossibility of these socialist parties forming left or even center–left cabinets. All the cases of their participation in the cabinet involved bilateral opposition, that is, cabinets that were opposed by strong forces on the right and the left. In particular, in France, the legitimation cleavage that excluded both the communists and forces of the right from the cabinets of the Fourth Republic forced socialists to enter system-defense cabinets that excluded the extremes and in which the moderate right was predominant. In France, Italy, Germany (interwar), and Finland (up to 1966), the internal division of the left prevented the socialists from even thinking of a left or even center–left government; at the same time, their obligation to share power as a junior partner with centrist and sometime conservative forces helped to reinforce the division with the communists and to expose them to criticism and electoral competition with the latter. Where communist parties managed to establish themselves as stable parties with more than 10% of the vote, their strength was sufﬁcient to rule out relatively homogeneous left or center–left coalitions and to push the socialists into tense centrist coalitions. The vicious circle was complete, and the radical-left

POLITICAL INTEGRATION

375

prophecy of ineffective participation in cabinets with bourgeois forces was therefore self-fulﬁlling.

SOCIALIST GOVERNMENTAL POWER This review of left cabinet experiences shows that an evaluation of the socialist executive tenure needs to consider presence and duration in ofﬁce, but also to supplement this information with that on the parliamentary status of the cabinet (single-party majority, MWC, etc.) to which the socialist party belongs and with the socialist party cabinet status (leading, equal, or junior member). The parliamentary status of a cabinet results from the intertwining of two dimensions: the number of partners (single-party versus coalition) and majority status (minority versus MWC versus oversized). The party cabinet status refers to the weight of the party within the overall coalition, according to its own and its partners’ electoral size. In Table 7.13, I propose a scheme based on these four dimensions of governmental power designed to obtain a comprehensive measure of the overall impact of each party on the government. This results from the rank ordering of the different combinations in terms of descending governmental power (last two columns of the table).92 The extremes of this rank ordering are represented by the single-party majority and party opposition cases. I have chosen to consider the partyleading role in MWC and oversized coalitions as the closest to the top of the ranking and the party junior position in multiparty minority cabinets and in surplus cabinets as closest to the bottom of the ranking. Equalpartner status in MWC and surplus coalitions come logically before a junior status in these cabinets. The greatest difﬁculty lies in ranking minority cabinet experiences and, in particular, the single-party and multiparty minority cabinets where a party has a leading or equal status. Finally, it is equally difﬁcult to ﬁnd the appropriate ranking of necessary external support and necessary abstention.93 For minority cabinets, the systematic external support of other parties may be of crucial importance. Does such an external-support role point to greater or lesser policy inﬂu92

93

This index of governmental power, as well as that of coalition potential discussed later, is more deeply elaborated and justiﬁed in Bartolini (1998: 40–61). These are situations in which, at the conﬁdence vote and at crucial points in the life of a cabinet, the support or the abstention of a party is necessary without its representation by ministers in the cabinet. I underline ‘‘necessary’’ because, if such external support or abstention is unnecessary for the survival of the cabinet, then the party’s inﬂuence on such a cabinet is reduced to nothing.

376

THE POLITICAL MOBILIZATION OF THE LEFT

Table 7.13. Dimension of governmental power

ence compared with the role of a junior partner within the cabinet? I have chosen the second solution. Applying to each cabinet the ranking value of Table 7.13 multiplied by its duration (in months), a governmental power index of the socialist parties is produced (a similar index for communist parties is irrelevant, as they shared little cabinet responsibility; see the earlier discussion) that can be cumulative for the whole 1918–1966 period or an average value for all cabinets (378) or for all legislatures (171).94 In the last column of Table 94

In comparing a cumulative index (for the whole period) with a mean (by cabinet) index of governmental power, only two differences in ranking appear. Finland and France are slightly more powerful with a cumulative than with a mean index. Belgium is a bit less so. However, with both indexes, Belgium remains ahead of both France and Finland. In the end, the only important change is that with a cumulative index the Finnish socialists appear to be more powerful than the Dutch, while the situation is reversed with a cabinets’ mean index.

POLITICAL INTEGRATION

377

7.14, the cumulative (by cabinet) governmental power of the socialist parties is reported. The Scandinavian countries are those where the socialists enjoyed by far the highest governmental potential. Britain, and at lower levels the Netherlands, Belgium, and Austria, are intermediate cases, while the large continental countries (France, Germany, Italy) and Ireland have very low indices. If we compare this multidimensional index with that of simple duration in ofﬁce or the percentage of cabinets (see Table 7.9) some striking differences emerge. For example, Austrian socialists in this period remained in the cabinet about 60% of the time, but in terms of combined governmental power this party was one of the least powerful because it was usually an equal or even junior partner in oversized coalitions with the Christian Democrats. Similarly, in Finland, the socialists were in ofﬁce for more than 45% of the time (and in 40% of the cabinets), but their global governmental power was much lower than these ﬁgures suggest. By contrast, the Dutch and British Labor parties enjoyed more governmental leverage than indicated by their time in ofﬁce. The British Labour Party stayed in power for only a third of the time and in a third of the cabinets, but it emerges as the party with the highest governmental power after the unmatched Scandinavian cases. The Dutch come after the British; although they began to participate in cabinets late and for about 40% of the time, they did so in a key position and role. In this way, the descriptive analysis of the governmental experience of the socialist parties in the crucial period of their access to government is completed, and a general description of the moments when they ﬁrst entered and then became regular governmental partners, as well as the duration of their overall governmental participation in different types of cabinets, is charted (Table 7.14). After World War I, any further institutional integration of the left into the respective national systems depended on the obstacles it found in translating its electoral and parliamentary weight into executive access and function. Given the European variance discussed so far, two issues are crucial: 1. What is the best predictor of the earliness of cabinet entry of the socialist parties? Was their ideological moderation and the bargainability of their positions more important, or were the favorable circumstances that fostered cabinet access and, consequently, policy and ideological domestication more important? 2. What can be said about the forces, independent of party system format

378

THE POLITICAL MOBILIZATION OF THE LEFT

Table 7.14. Socialist parties’ governmental experiences (countries ranked according to governmental power index)

Note: governmental power ⫽ cumulative 1918–1966 governmental power index for all cabinets.

conﬁguration, that fostered or hindered the overall socialist governmental role? To evaluate the resistance to the translation of socialist parliamentary strength into executive roles, we need to identify the part of the socialist governmental role that derived from the party system format and conﬁguration opportunities from that which derived from other factors. It is necessary to ﬁnd a comprehensive indicator of the configuration structure of opportunity offered to the parties by the party system format and then compare it with the actual governmental role played by that party as measured before. I have measured this potential independently of all other ideological, political, and historical circumstances by resorting to the formula of coalition power within decisional bodies originally developed by Shapley for computing an index of decisional power within committee systems.95 This is one of the many algorithms developed for measuring decisional power in voting bodies from a game theory perspective,96 and it can be interpreted as an a priori power index measuring the potential power of an ith member of a certain body (be it a small committee or a

95 96

Details about the algorithm used are presented in Bartolini (1998). Shapley (1953). This index is also known and cited as the ‘‘Shapley–Shubik index’’ because of its ﬁnal formulation by these two authors, with a modiﬁcation for majority games; Shapley and Shubik (1954).

POLITICAL INTEGRATION

379

parliament).97 Given the percentage of seats of a given party, such a measure computes the number of all the possible majorities above the 50% threshold in which that party is a necessary coalition partner, that is, in how many it is a partner whose withdrawal would lead to the coalition’s falling below the 50% threshold.98 To the best of my knowledge, this index has never been applied to the comparative study of election results across countries and over time.99 For each party, its index of coalition potential incorporates all the numerical potentials implicit in a given distribution of strength among actors in a voting body.100 This ‘‘blindness’’ of the index to any other consideration but numerical ones is an asset. It provides a theoretically adequate description of the situation in which the chances of a party being incorporated into a majority are stripped of all political considerations and are dependent exclusively on the numerical conﬁguration of the party system. It is thus a description of the pure and absolute ‘‘coalitionability’’ of all partners, in which each party is deprived of all its political idiosyncrasies stemming from its past history, previous coalition experiences, or policy preferences. Blindness is, however, also a liability for the same reason that considerations of coalition feasibility are ignored. In this analysis, I have introduced a constraint into the algorithm linked to the spatial positioning of the parties, which forces it to take into consideration only those majority coalitions that are stipulated among spatially connected parties. The coalition potential is computed only in reference to these coalitions.101 97

98

99

100

101

An alternative way to read the Shapley value is ‘‘the probability for the i to deliver the necessary votes for a qualiﬁed majority’’ (Bomsdor 1981: 284). A third deﬁnition is the probability that unit i, by casting its votes, establishes a winning majority in any voting sequence or, to put it differently, as the relative frequency with which unit i plays the pivotal role in an inﬁnite number of random majority coalitions among k units (Rattinger 1981: 223). Attention should be paid to the types of majorities considered by the algorithm itself. These are not the MWCs, but rather all the possible majority coalitions in which the party under discussion is crucial or pivotal. This means that the set of majorities from which the value is ﬁnally calculated also includes oversized majorities resulting from the presence of some other party whose withdrawal would still leave them with a majority of seats. The choice of this set of majorities is necessary because the calculation is ﬁnalized to the coalition potential of the party in question. It has frequently been used for the descriptive analysis of the distribution of voting power in speciﬁc legislatures. See Frey (1968), Holler and Kellermann (1977), Miller (1973), Guillermo (1975), and De Fraja (1988). In principle, any of the many theories of coalition formation can be modeled into the algorithm through the introduction of conditions that conﬁne the number and type of coalitions feasible. For the criteria of left–right ordering of the parties, see the Data Appendix. A second

380

THE POLITICAL MOBILIZATION OF THE LEFT

The theoretical framework summarizes resume the relationship among a ‘‘blind’’ coalition potential, a ‘‘spatial’’ coalition potential, and the actual governmental power (as measured earlier), as indicated in Figure 7.1. Configuration and format factors directly inﬂuence the blind coalition potential and only indirectly, through it, governmental power. The blind index should be closely associated with the numerical and distribution conditions of the party system. Traditional ideological factors enter the picture, introducing sympathies or compatibility among the parties as a result of past experiences. The spatial rank ordering of the parties can be considered a valid indicator of these historical compatibilities and can be incorporated into the index of coalition potential when it is computed on the basis of left– right-ordered party units. What intervenes between a spatial coalition potential and governmental power can be conceptualized as the black box of ‘‘other’’ political factors that are not tapped by any potential index. These factors do not pertain either to the numerical structure of opportunities offered by the party system or to the historical ideological sympathies ﬁrmly established among parties. Instead, they include aspects belonging to the cohesion of the adversaries (of each given party), to the issue bargainability of the party stance, and to the presence of legitimacy cleavages concerning that party. Each of these elements may introduce strong deviations in the relations between coalition potential and governmental power by lowering or boosting the governmental role far beyond what would be expected on the basis of the party’s coalition potential. Brieﬂy, in principle, one would expect that, if the coalition potential corresponds to governmental power, then the structure of opportunities of the left was basically neutral. Adversaries’ coalitionability, socialist policy bargainability, and discontinuities in the space of competition can therefore be conceptualized and measured as otherwise unexplainable discrepancies between the potential and actual governmental role. We can now interpret the timing of socialist access to government to check whether the prevailing factor was the ideological bargainability of their positions or the party-system conﬁgurational features. The hypothesis is that in the absence of speciﬁc ideological and/or political factors hampering or favoring socialist access to cabinet responsibility, such access should be entirely a function of the coalition potential resulting from the conﬁguration format of the party system. In Figure 7.2, I have charted the modiﬁcation may relax the stringent criteria of perfect connectivity. We can allow only coalition among connected parties, but permitting a ‘‘spatial jump,’’ that is, allowing one of the connected parties to be absent. This reduces the problems of the ‘‘correct’’ ordering of parties over the space. Other conditions of coalition feasibility can also be introduced by modifying the set of total feasible coalitions.

POLITICAL INTEGRATION

381

Figure 7.1. A theoretical framework for the analysis of governmental power. year of ﬁrst cabinet entry of the socialists against their size and their coalition power at such a moment. For nine out of the thirteen countries (Germany, the United Kingdom, Finland, Norway, France, the Netherlands, Switzerland, Italy, and Ireland), the coalition potential is an almost perfect predictor of the timing of entry. Size represents a far less important factor. In particular, the British, Irish, and Finnish Labor parties entered the cabinet when their coalition potential was much higher than their comparative size. At their entry, the British and Finnish socialists were no stronger in parliamentary terms than their French, Dutch, or Swiss counterparts, which instead entered about ﬁfteen to twenty years later. Four cases deviate from this pattern: Belgium, Denmark, Austria, and Sweden. These earlycomers to cabinet responsibility did not enjoy a high enough level of coalition potential to justify their executive incorporation, that is, their early entry was due to political factors that were not related to their being highly necessary coalition partners. The strong Austrian socialists had a low coalition potential in 1918, and their entry was due exclusively to the peculiar conditions created by the collapse of the imperial regime and, in particular, to the delegitimation and lack of organization of the conservative forces and their inability to form coalitions with the Catholics. It is not surprising, therefore, that this early entry was a short-lived experience. Given their low coalition potential, the Austrian

382

THE POLITICAL MOBILIZATION OF THE LEFT

Figure 7.2. Socialist size and coalition potential at their cabinet entry. socialists were immediately ousted from the cabinet once the Catholic– German nationalist divide was overcome. In Belgium, the early incorporation of the socialists in coalition making, when their coalition potential was not particularly high with respect to the potential of Catholic predominance or Catholic–liberal alliances, was due to the war and the occupation experience. However, the fact that socialists were not excluded from cabinets later on shows that no stable antisocialist bloc ever developed between the Catholics and liberals. Thus, these latter two parties were sufﬁciently antagonistic not to consider their alliance as always preferable to one including the socialists, and neither of them regarded the socialists as the main common enemy. This supports the idea that it was the moderation and bargainability of the Belgian socialist positions in this period that allowed them to be coalition partners with both Catholics and liberals (in 1919–1920, 1926, and 1935– 1938) or with the Catholics alone (in 1925 and 1939).102 102

The total freedom of coalition by which socialists could form cabinets even with the liberals and without the Catholics developed only after World War II.

POLITICAL INTEGRATION

383

The Danish and Swedish early entry is more difﬁcult to interpret. Their coalition power was quite low at the time of their entry, although it rose enormously after it. Even in these cases, the best explanation is that their relative moderation, added to the strong divisions within the bourgeois camp, fostered their capacity to enter cabinets beyond their coalition potential. In Denmark, the split of the liberals and the creation of the radicals offered the socialists an early, stable ally that was unavailable in other Western European countries at the time. Between the world wars, the radicals formed a coalition cabinet only with the socialists, never with the liberals or conservatives.103 For the Swedish Social Democrats, we have seen that the cabinet experiences in the interwar period involved mainly minority cabinets. The key to this formula was, however, that agrarians, liberals, and conservatives could never agree to form a nonsocialist cabinet, and the system only produced either caretaker cabinets or nonsocialist ultraminority cabinets. The Swedish case is the most obvious case of a socialist cabinet experience largely fostered by the division of bourgeois forces. In conclusion, the good ﬁt between coalition power and timing of entry does not support the idea that it was the more moderate socialist parties that entered ﬁrst. In reality, it was those socialist parties with the highest coalition potential in their party system that entered ﬁrst. The relationship between governmental participation and deradicalization seems to go from the former to the latter rather than in the other direction. As I have argued, Austria is no exception to this rule, and it was only in Belgium, Denmark, and Sweden that early socialist ideological moderation and the ideological divisions of the nonsocialist forces played a prevalent role. Let us now consider the overall picture for the 1918–1966 period rather than the situation at the ﬁrst cabinet entry. Using the standardized scores in Figure 7.3, the level of coalition potential and the level of governmental power are charted for each country. This allows a clear appreciation of the extent to which either the sheer numerical opportunities offered by a given structure of the party system were actually exploited or some other powerful politico-ideological forces interacted with this relationship, fostering the downgrading of the socialist chances of a governmental role. The Scandinavian countries, Belgium, France, and the Netherlands pose no interpretive problem, as their period average coalition potential is matched by their governmental power. The more interesting cases are those of Austria, Finland, Germany, Britain, Ireland, and Italy, 103

It was only in 1945 that they offered external support to a liberal cabinet for the ﬁrst time.

384

THE POLITICAL MOBILIZATION OF THE LEFT

Figure 7.3. Coalition potential and governmental power. where the ranking in terms of potential and that in terms of actual power differs considerably. In Finland and Germany, the overall high coalition potential of the left was reduced by internal divisions that drastically curtailed the potential for the connected winning coalitions of the socialists. In addition, the ideological ‘‘summability’’ of the center–right forces proved much higher than that of the left, actually contributing to reduce the governmental role of the social democrats with respect to what they could have had if the divide within the left had been less intense or counterbalanced by a similarly strong divide within the right. In Austria, surplus coalitions after World War II probably offered the Socialist Party more power than it could possibly have enjoyed if an MWC logic had prevailed in the party system dominated by the central pivotal role of the large Catholic Party. In Britain, where indeed a winner-take-all institutional setting and political culture predominated, the Labour Party enjoyed more actual power than coalition potential, probably thanks to its role in the minority cabinets of the transition period in the late 1920s and 1930s. In Italy and Ireland, the distribution of the forces and the presence of a large, centrally located party offered the Socialist Party very little

POLITICAL INTEGRATION

385

coalition power. However, the party’s actual governmental role, although below the overall mean, was higher than expected because in Ireland the dubious left–right location of the parties allowed for all-party coalitions against the Fianna Fail, while in Italy the unique and simultaneous presence of extreme-left and extreme-right forces pushed the Catholics into preferring oversized coalitions, thereby offering a chance to the socialists. It is interesting to compare Italy and France. Before World War II, the only difference between them was that the socialists participated, between 1936 and 1938, in the four Popular Front cabinets in France, while the Italian socialists reached cabinet status only after the war. After this, the main difference concerns the fragmentation of the centrist forces, which was very high in France and very low in Italy and which enhanced the coalition potential of the French socialists. However, given that extreme-right antisystem and illegitimate coalition partner forces were far less important and more intermittent in France than in Italy, the need for socialist partners was reduced, as center–right and right cabinets were a more likely and frequent outcome. In these two cases, the difference in coalition potential and the similarity in governmental power resulted more from the cohesion of the center–right forces and from the presence of a legitimacy cleavage in the party system than from opportunity structures linked to electoral size. This last remark leads me to consider more closely the coalition potential of the communist parties, or more generally of the other left parties, and the issue of the cost of division within the left in terms of cabinet potential. The data for communist parties highlight the enormous difference between the blind coalition potential – depending exclusively on the numerical opportunities created by their size, dominance, and so on – and the spatial coalition potential, which, as a result of their extreme-left position, is curtailed drastically. In Finland, the potential in connected coalitions is, for example, one-tenth of the same potential in unconnected coalitions. The extreme case is Italy: the connected coalition potential of the Communist Party was near zero between 1945 and 1966 because Italian Communists could not attain a majority of seats by allying with neighboring parties on the left or the right. To reach such a majority, their coalition had be extended to the Christian Democrats in a virtually allparty coalition, in which, however, the contribution of the Communists was no longer necessary. The losses incurred by the left as a result of its internal split can be measured in terms of lost cabinet potential by comparing the cabinet potential of the socialists with the cabinet potential that a uniﬁed left would have enjoyed. I have therefore repeated the calculation of the Shapley

386

THE POLITICAL MOBILIZATION OF THE LEFT

index, inserting into the party system a ﬁctitious party whose parliamentary size corresponds to the global cumulated parliamentary size of all class left parties (socialists, communists, and others). In the post–World War II period, the Finnish left would have been a necessary partner in 10% more cabinets if it had been united (the same was true after 1966). Losses of some importance were also evident in the Swedish case, where a united left would have been necessary to 16% more cabinets than was the case for the socialists alone, that is, for almost all the cabinets. The major losses were encountered, however, by the French left in the 1945–1966 period. If it had been less politically divided, it would have been a necessary partner for roughly 25% more cabinets than was the case for the socialists. In the two decades after World War II, the Italian left did not incur high cabinet potential losses as a result of its division: If it had been united, it would have been a necessary partner in 12% more cabinets than the socialists were. However, the costs for the Italian left increased dramatically after 1966 with the declining cohesion of the centrist forces. In the period 1966–1985, a united left in Italy would have been a necessary cabinet partner in 30% more cases than the socialists actually were. This different trend in the costs of divisions in Italy, France, and Finland reﬂects the history of the communist–socialist cleavage in these countries. The deep divide created by the split was soon reconciled by the common antifascist struggle in Italy, and after World War II the Italian left was the only one in Europe in which the main socialist party aligned in opposition with the communists. Only at the beginning of the 1960s was the ancient cleavage revived with the ‘‘autonomist’’ position of the socialists and their cabinet entry. In Finland, by contrast, the cleavage deriving from the split remained ﬁrm until well after World War II and started to decline only in the 1960s, exactly when it was exacerbated in Italy. In France, the rapprochement of the Popular Front was short-lived, and a deep divide remained throughout the whole Fourth Republic and until the beginning of the 1970s. In a broad historical perspective, the within-left division created the highest cost for the left as a whole, as well as for its internal components in Italy. The socialists’ opposition position, and their closeness to the communists in the 1940s and 1950s, deprived them of the possibility of exploiting their coalition opportunities in this period, while their increasing distance from the communists during the 1970s, and even more in the 1980s, boosted their own coalition potential but sharply reduced that of the global left at a moment when the more moderate communist stance made such an alliance conceivable. The Italian socialist–communist relationships went ‘‘against history,’’ with a solid, noncompetitive alliance

POLITICAL INTEGRATION

387

during the Cold War and an increasingly competitive attitude in the 1970s and 1980s. The consequences of the different governmental experiences of the European left were of great importance. Participation in the cabinet offered leverage on policy making and implementation, as well as the potential for electoral stabilization and strengthening. The capacity of socialist governments to inﬂuence macroeconomic and redistributive policies to the advantage of their electoral base has been argued.104 However, outside the Scandinavian world and Britain, the inferences to be drawn from these data are tenuous for three reasons. The ﬁrst is documented in the previous sections of this chapter: Clearly left or even center–left-oriented cabinets were rare occurrences, and most of the time socialists in continental Europe entered the cabinet under conditions of enforced moderation of their claims, in broad oversized coalitions or in barely tolerated minority cabinets. Second, in continental Europe the socialists were neither the ﬁrst nor the only party to foster welfarist, full-employment, and labor-favorable legislation and policies. In the bureaucratic and authoritarian Central European constitutional monarchies, the political elite actively fostered early and advanced social legislation, with the deliberate aim of detaching workers’ corporate organizations from their socialist political representation and of achieving political legitimation without resorting to the liberalization of the regime and relying on the electoral-parliamentary channel.105 Later on, Catholic and denominational parties competed with the socialists, whose cabinet experiences were in most cases carried over in alliances with such forces. However, in Italy, Belgium, the Netherlands, Germany, and Austria, denominational parties governed more frequently and longer than socialists, which makes a case for their claim of having been the main architects of the exceptional development of social policies in the twentieth 104 105

See the articles in Castles (1982). In the 1880s, Austria introduced an impressive array of social legislation deliberately aimed at the integration of workers into the state. This legislation was interventionist in the sense that it anticipated workers’ requests and was often passed over employers’ opposition. For a detailed analysis of Austrian social legislation, see Talos (1981), especially the Introduction and the ﬁrst chapter, which deals with the early history. In Germany, the situation was similar and responded to the same aim of social integration without political representation for the lower classes. Baldwin (1988) has extended the role of the conservatives in early social policy making beyond the Central European constitutional monarchies. Liberal states could not afford such an interventionist policy in opposition with the capitalists and, at the same time, did not need a different source of legitimation from the one they received from the political-electoral open channel. The paucity of social legislation in liberal France before World War I has been amply demonstrated. A broad social legislation program did not come until the Popular Front in 1936: ‘‘a generation later: a generation too late’’; Lorwin (1954: 7, 28).

388

THE POLITICAL MOBILIZATION OF THE LEFT

century.106 A ﬁnal factor weakening the relation between socialist executive participation and actual policy output is that the thirty-year period between 1930 and 1960 saw the development of welfarist and fullemployment policies anywhere and any way,107 independently of the composition of governments. That is to say, the political forces that enjoyed control of the executive during this crucial period gained by default the dividends of such growth. This expansion was probably the key factor in the remarkable electoral stabilization that took place after World War II, differentiating the period from the more marked electoral instability of the interwar years and the post-1973 period.108 The left parties that missed this historical window of opportunity lost the chance of directing such growth to their own and their constituency’s advantage forever. This leads to the consideration that the great opportunities for the left lay not in the actual distribution of the beneﬁts of this growth as such, but rather in the possibility of guiding such growth through regulative policies that would eventually strengthen their electoral and political hegemony in their respective societies. In this lies the exceptionality of the Scandinavian countries. The combination of organizational, institutional, and political factors discussed in the previous chapters made them able to exploit this opportunity fully and allowed them to shape the fabric of their society in a way that was proﬁtable to their continued rule. In the Scandinavian context of high religious homogeneity, the predominance of socioeconomic conﬂicts (see the next chapter), low stateness (with the state, however, accepted as the regulatory agency of the society),109 and stable access to power in the 1930s, the labor movement could shape new employment policies; inﬂuence social, ﬁscal, and distributive processes; and deprive employers of friendly governments and favorable state intervention from the beginning. The focus of industrial conﬂict – which earlier had fallen on the market, where workers confronted capitalists in costly con106

107

108

109

This does not mean that the general welfare projects of denominational and socialist parties were not different in general philosophy and also in the relative satisfaction offered to different groups of claimants, in particular dependent versus independent work. For the different political characterization of welfare programs, see Ferrera (1993: 155–200). See also Esping Andersen (1990: Chap. 3). Growth ﬁgures for the countries of Western Europe can be found in Flora, Alber, Eichenberg, Kohl, Kraus, Pfenning, and Seebohm (1987). Evidence of the striking electoral stability of European party systems in the 1945–1970 period compared with the earlier and later periods is found in Bartolini and Mair (1990: 98–121). Not necessarily in the statist and paternalist version of Germany and Austria. In Sweden, after the initial inﬂuences of Bismarck’s reforms, German statism and paternalism were soon rejected in favor of a more emancipatory, radical-liberal, and humanitarian vision. See Olsson (1990).

POLITICAL INTEGRATION

389

ﬂicts with uncertain outcomes – switched progressively from the market to the more egalitarian arenas of politics.110 This helped to avoid risk taking in industrial conﬂicts and to resort instead to political instruments for redistribution. The early institutionalization of industrial relations and the labor–management pacts outlawing unauthorized strikes and establishing national grievance procedures for resolving disputes over wage contracts changed Sweden from the country with the highest level of labor disputes in Europe to one with a paciﬁed relation pattern. To sum up, governmental control was the key resource for party strategy. Swedish socialists created labor market institutions and policies that strengthened the unions and their recruiting capacity on a continuous basis.111 Although this slowly but consistently established legal and political context was not the only condition for the development and organizational growth of trade union membership, which was among the strongest in Europe before the socialists gained power (see Chapter 6), it did reinforce their authority over members and their centralized decision-making and bargaining power, as well as the extension of their power in social policy implementation. The early access to and stability of executive tenure in Scandinavia also fostered the organizational cohesion of socialism, making left criticism and propaganda difﬁcult under the conditions of a ﬁrmly controlled Social Democratic government. From the 1930s on, radical wage policies and ‘‘outbidding’’ opposition policies were pursued by communist militants at the grass roots level against those of the socialists even in these cases, but the link between these and real-wage developments was tenuous. The communist wage policy was squeezed between the highly centralized and visible socialist governmental wage negotiations (recognition, etc.) and effective local additions to the central agreements.112 The gap between the material realities and interpretive ideological schemes in communist propaganda was too wide. In countries where socialists had no ﬁrm control over the government, their wage policy was in direct competition with that of the communists only in the ideological and propaganda realm, and no comparison between the material achievements attributable to socialists and the language of radicalism was possible. Only two languages were compared with their possible long-term, abstract consequences, and no gap 110 111

112

Korpi (1983: 181). In particular, the Ghent system’s unemployment scheme, by which unions controlled unemployment beneﬁts and workers needed to join unions in order to be eligible for them. Rothstein (1992). The Ghent system is also in use in Denmark, Finland, Iceland, and Belgium (all highly trade-unionized countries), while compulsory state-run systems are in use in all other countries, including Norway. Stra˚th (1983: 277).

390

THE POLITICAL MOBILIZATION OF THE LEFT

between reality and communist propaganda was visible. Instead, where communist strength was great enough to make socialist cabinet participation a rare and not particularly effective exercise, the gap opened between the socialists’ propaganda and ideological identity and their poor policy achievements. It is likely that the difﬁcult conditions of cabinet participation and/or the perception of the poor results of such participation in countries like Germany, Finland, France, and Austria between the wars, and also in Italy and France after World War II, contributed to the strengthening and maintenance of a verbal radical orthodoxy as a ‘‘compensatory myth’’ for poor policy results and uncertain democratic consolidation. This use of governmental regulatory initiatives to foster societal development and to consolidate social and political alliances113 was impractical for the British Labour Party for cultural and political reasons. Not only was the party not in government long enough to enforce such strategies credibly,114 but the cultural and structural conditions of its relationship with the state made this impossible. The tradition of governmental engineering of civic society remained marginal to, even absent from, the basic ideological orientation of Labour’s leaders. Only in 1945, after the exceptional conditions of the war economy and the disruption of the social fabric, was a decisive push in that direction actually achieved. 113

114

This is most evident when governmental reform had direct electioneering goals, as in the early 1960s in Sweden and Norway, when both parties faced the issue of maintaining or gaining new middle-class voters while confronting extreme dissent from the left. In Sweden the socialist strategy of breaking down status divisions between manual wage earners and salaried employees, emphasizing issues that united the dependent labor force, concretized in a proposal for a comprehensive, government-administered retirement pension plan based on average earnings during the peak period of work life. In Norway, something very similar was set up at the beginning of the 1960s, in particular in 1963, when, in its radical program, the party proposed a stateadministered service pension and a comprehensive plan for investments in the economically backward periphery of the country through capital funds to be created by employers’ and employees’ contributions toward the pensions. Lipset (1964: 287) and Rokkan (1966b: 103–104) have argued that the socialist plan was based on electoral considerations. Stijernquist (1966: 124–125) is less convinced. The relevance of long-term government tenure for hegemonic policies is seen in the different outcomes of British and Swedish post–World War II land planning. In Britain, the postwar Labour governments introduced three different systems for taxing the unearned land betterment gains in order to ﬁnance public holdings of land. The competitive strategy of the Conservatives was to promise to scrap the legislation and to restore a free-market logic in land, advising landowners to withhold their properties from the market until they were returned to ofﬁce. Because the landowners expected the Conservatives to come back to ofﬁce soon, these policies failed. By contrast, in Sweden, the success of the socialists’ municipalization of all development land was based largely on the fact that those who controlled the resources in question could hardly hope to repeal such policies in the short term.

POLITICAL INTEGRATION

391

In France, the bureaucratic tradition and modernizing role of the state was as old and important as it was resistant to any partisan use of it.115 Nothing is more telling than the contrast between the conditions of the Scandinavian red–green alliances of the mid-1930s and the French socialist Popular Front of the same years. While, on the one hand, there was a strong, united, and responsible labor movement, on the other hand, the CGT was prey to factional dissension. To the employers’ acceptance of limitations on their authority in Scandinavia there was in France no corresponding adjustment to new relations with the labor movement. To the revival of industrial production and to an enduring governmental policy favoring collective bargaining in Scandinavia, there was in France little corresponding economic recovery under Blum’s governments and no continuity of governmental social policy after him. Finally, Hitler’s March 1936 march into the Rhineland, gambling on the inability of France and the United Kingdom to defend the peace of Versailles and Locarno, deprived the Popular Front and France altogether of the freedom to carry out those social experiments that could not be absorbed without social turmoil. Only the Scandinavian socialists were able to ensure sufﬁcient, consistent, and high-status tenure of ofﬁce to mold their respective societies. Large, united socialist parties in Austria and the United Kingdom were confronted with adversaries too strong and cohesive to permit them to even attempt a similar strategy. Where an adversarial logic prevailed, as in the United Kingdom, socialists were offered high power in single-partymajority cabinets, but with little or no continuity. Where a coalescent grosse Koalition logic prevailed, as in Austria after World War II, continuity of tenure was achieved at the cost of policy control. In Italy, Germany, France, and Finland, deep cleavage lines both within the left and between the left and the other parties reduced enormously the governmental power of parties supported by 40% to 50% of the voters.

STATENESS, INSTITUTIONAL OPENNESS, AND THE CLASS CLEAVAGE STATENESS, REPRESSION, AND LABOR RESPONSE Between 1815 and 1870, most European countries oscillated from phases of internal repression to phases of relative toleration, and these phases were 115

For the modernizing role of the French state, see the work of Legendre (1968) and Brown (1969). The antiparty and antipartisan ideology of the French higher bureaucracy is well documented in Suleiman (1974: 102–108).

392

THE POLITICAL MOBILIZATION OF THE LEFT

uniformly determined by whatever threat was perceived by dynastic rulers. Repression increased almost everywhere in response to the 1820, 1830, and 1848 revolutionary movements, and, in general, revolutionary events in one country were met with preventive repression in the others. Repressive measures were eased in the intermediate periods. The dominant military powers (the Hapsburg Empire, Prussia-Germany, France, Russia) coordinated their efforts to crush national and social movements at home as well as in other countries under the general principle that changes in government and constitutional arrangements in one country actually threatened other European states and legitimized military restorative intervention.116 Only boundary-safe countries (the United Kingdom) and marginal and strategically less signiﬁcant areas (Scandinavia, with the partial exception of Denmark) were left free to follow their own development. After 1870, this pattern of common response stopped, and there was increasing differentiation in the modalities and timing of repression and toleration. Even the response to the French Commune was not uniform. In any case, after 1860, and then after 1890, pure repression was used less and less as a means of social control.117 Repression became more focused on the most dangerous internal groups and movements. Attempts became more sophisticated, aiming to remove the masses from the inﬂuence and leadership of radical groups and adopting ameliorative policies as an alternative form of social control. This change was dependent on the calculation that a purely repressive response was unlikely to solve the problem of lower-class demands and unrest and that the climate of internal turmoil weakened international status and prestige and was increasingly costly and destabilizing. The established European elite thus more frequently admitted and debated in parliaments that reform was necessary to quiet and calm the lower classes and thereby to strengthen the existing regimes. The Austro-Prussian and Franco-German wars had ended any unity of action on the part of major powers and started a new period of large-power continental confrontation during which internal turmoils were seen as possible assets/liabilities to be exploited by these powers, rather than as a challenge to a general and unitary principle of dynastic legitimate law and order. While this was the general trend, some countries were clearly ahead, while others lagged behind. The variations in institutional openness and 116

117

Typical of this logic was the Troppau Protocol of November 19, 1820, in which Prussia, Russia, and the Hapsburg Empire coordinated counterrevolutionary intervention to suppress Naples’s carbonari revolution of the same year. Goldstein (1983: 345).

POLITICAL INTEGRATION

393

the switch in tactics of sociopolitical control were probably determined by the perception of the cost that the upper classes and dynastic rulers would have to pay for delaying reforms and toleration. These costs had an economic as well as a political dimension. Could the dominant economic interests be defended within a political climate of toleration? Politically, could such toleration allow the dominant political elites to retain their security of tenure and their hold on the polity? The other element of the equation was the cost of repression. The issue was, ﬁrst, whether the means of repression were available and, second, if so, whether dominant elites and circles (king, bureaucracy, upper class) were sufﬁciently united to pursue that course of action. Assuming (1) that the resources available for a repressive strategy were mainly represented by the stateness of the political processes, as indicated by the extractive and repressive resources of the dynastic-bureaucraticpolitical establishment, and (2) that the severity of the economic challenge depended on the homogeneity of economic interests, as indicated by the differentiation of dominant economic groups, we can try to interpret the level of the repression-toleration response as a function of these two aspects. To this end, in Table 7.15 I have cross-tabulated the level of early socialist movement repression to the stateness of the polity and to the level of internal differentiation of dominant economic interests. The latter is based on the strength of industrial and commercial interests, on the basis of the earliness of industrialization, and on the numerical presence of independents in commercial and industrial sectors over the total population of the independents.118 The thrust of the argument is that the earlier the industrialization, the higher the differentiation of urban commercial and industrial interests versus the rural ones, and, therefore, the less likely a common and uniform dominant group perception of economic challenge. Finally, in Table 7.15, the cases in boldface refer to the early timing of the parliamentarization of the executive, as discussed in the second section of this chapter and presented in Table 7.4. Bearing in mind the difﬁcult intermediate cases of the United Kingdom, Ireland, and the Netherlands for the level of stateness, and that of Denmark for the introduction of responsible government, we can draw the following conclusions. If we consider, ﬁrst, the cohesion levels of dominant interests, and therefore the likelihood that they will agree on supporting a 118

I have taken 50% as the cutoff point. The average proportion of commercial and industrial independents over all independents in the 1890–1930 censuses is the following: United Kingdom ⫽ 81.5; Switzerland ⫽ 54.2; Norway ⫽ 35.9; Finland ⫽ 22.1; Belgium ⫽ 60.6; Denmark ⫽ 46.2; Switzerland ⫽ 30.3; the Netherlands ⫽ 60.3; Austria ⫽ 44.3; Ireland ⫽ 26.4; Germany ⫽ 54.5; France ⫽ 38.0; Italy ⫽ 25.8.

394

THE POLITICAL MOBILIZATION OF THE LEFT

Table 7.15. Stateness, strength of commercial/industrial interests, and repression

Note: Cases of comparative early parliamentarization of the executive in boldface.

repression strategy, we can note that in early industrialized countries with strong industrial and commercial interests, antisocialist repression tended to be low with one exception, that of Germany. By contrast, in countries with later industrialization, where rural interests still tended to be predominant, we can observe two groups. On the one hand, Finland, Austria, France, and Italy showed a relatively high and frequent use of repressive measures. On the other hand, in the three Scandinavian countries – notwithstanding the overwhelming role of rural interests in the 1890–1930 period – no heavily repressive response was mounted against the socialist movement. The ﬁt is far from satisfactory, although in the Scandinavian case one could mention that, much more than in the other late industrializing countries, rural interests were represented here by a very large class of small, independent farmers. The second point to be noted is that the level of repression before World War I is only partially related to the liberalization of the regime. Of the eight cases where we ﬁnd relative toleration, six are parliamentarized regimes, but the French and Italian parliamentary regimes strongly repressed their internal opposition movements, and in particular the socialist one. In contrast, Denmark and Sweden still had constitutional monarchies that strongly resisted liberalization, but with low levels of internal repression.

POLITICAL INTEGRATION

395

The third consideration is that stateness is a much better predictor of the level of repression of early socialism than either the level of differentiation of economic interests or the liberalization of the regime. To be more precise, the role of stateness helps to reconcile the other two factors in explaining the degree of tolerance or repression of dominant groups. Of the ﬁve countries that had strong state resources at their disposal, four directed them against internal opposition. Unquestionably, there are important differences between France and Italy, as liberal regimes, and between Austria and Germany, as autocratic regimes, but these differences are less marked than the differences between this cell and the others. Of the eight countries I have classiﬁed as having low or modest stateness, only Finland has a pre-1918 history of signiﬁcant socialist movement repression; Belgium is the only country with relatively high state resources whose political elite did not use them systematically to meet the claims of newcomers. Finland, however, is not really a deviant case because the repressive role was played there not by the Finnish state but by Russia. I have emphasized that the repression of Finnish aspirations (and not only of socialists claims) was due fundamentally to the security imperatives of imperial Russia and to the great political insecurity of its autocratic elite. Actually, in exporting repressive resources to Finland, the Russians inhibited the development of an autonomous Finnish state to such an extent that when the center collapsed, Finland was left with practically no state, but rather with a ‘‘praetorian’’119 society that continued for almost thirty more years to see the direct intervention in politics of armed groups and political organizations, with little or no mediating role of state institutions and apparatuses. If we combine the different dimensions, our understanding of the variation improves and stateness emerges more clearly as the crucial necessary condition for a repressive strategy. Whenever high state resources were available to an established political elite, they were used against newcomers. The exception of Belgium can be explained by the combination of its early liberalization and high internal differentiation of interests between rural groups and powerful urban commercial and industrial interests. On the other hand, countries like Denmark and Sweden, even if they had a conservative and autocratic political structure and a predominance of rural interests, could under no circumstances afford a confrontation strategy against newcomers due to their sheer lack of resources. A typical example of this is the long hesitation of the Swedish crown and its militaryaristocratic advisers on whether to yield to the Norwegian request for 119

Huntington (1968: 95).

396

THE POLITICAL MOBILIZATION OF THE LEFT

parliamentarization ﬁrst and independence later. A repressive reaction was considered at length and was probably the preferred option, but considerations of affordable costs predominated, and in these calculations the low level of stateness in Sweden was probably the crucial variable. We can therefore conclude that the level of repression tended to be higher (1) the higher the stateness of the polity, which made available the means; (2) the lower the parliamentarization of the regime, which deﬁned the limits of utilization of these resources and also offered possible alternatives for control; and (3) the lower the internal differentiation of dominant interests, which made more likely their perception of a common threat and their agreement on a common response. The level of repression was therefore minimal in countries with little stateness, high interest differentiation, and early parliamentarization of government (the United Kingdom, Switzerland, and the Netherlands). In Austria, low interest differentiation and dynastic autocratic regimes coalesced to make the perception of the socialist challenge unacceptable, while the high stateness of the polity lowered the costs of repression. In autocratic regimes with low stateness (Denmark, Sweden, Finland, and Norway, which, however, had earlier liberalization), repression could be used systematically only if it rested on external resources (Finland); it was too costly if it involved the scanty internal repressive means of the state. France, Italy, and Germany are more problematic cases. In the ﬁrst two, high stateness combines with early liberalization and with a weak interest differentiation toward a considerable level of repression. In Germany, higher interest differentiation (concentrated, however, outside politically dominant Prussia) combines with high stateness and no liberalization toward a high repression strategy. The difﬁcult prediction of the French and Italian cases, in particular, suggests a speciﬁcation of the model. Whether repression or toleration was the outcome may also depend on whether a political defense of the dominant interests was possible. In other words, a toleration strategy was more likely if political checks could be expected at the electoral level. This, in turn, depended on something I discuss in the next chapter: the extent to which the structuring of the party system made feasible stable politico-electoral alliances capable of contrasting and checking the electoral inﬂuence of the socialists. The hypothesis is that when the lack of structuring of the party system and the persistent threat of illiberal and/or reactionary alternatives made the liberal elite insecure and unsafe, some repressive response was more likely (France and Italy). The same features may also be important in Germany, where industrialization, urbanization, and interest differentiation were considerable. The politico-electoral weakness of industrial and commercial interests

POLITICAL INTEGRATION

397

can be added to the illiberal regime and the high stateness in explaining its repressive strategy (see the next chapter). Political repression was one of the main determinants of early socialist movement behaviors and the instrument through which the state shaped the structure and fundamental forms of labor protest.120 Support for this generalization can be found not only in my general analysis of crosscountry differences, but also in a few cases of within-country differences. For instance, consider the different socialist ideological position of the 1890s and the deﬁant attitudes of the southern states of Bavaria and Wu¨rttemberg – which had universal and equal suffrage, a higher level of respect of civil liberties, and socialist cooperation with the bourgeois parties – and the northern states of Prussia and Saxony – where discriminatory suffrage requirements and repressive controls remained the rule. Other supporting evidence is the sudden transformation of the Austrian Socialist Party in 1890, when the state became more tolerant and the socialists transformed their radical and militant positions almost overnight. Another example is the extreme militant tactics the Belgian socialists used to obtain essentially moderate political reforms, which were therefore soon abandoned once they were achieved. The state response was a decisive inﬂuence on the ideological and protest answer of the early socialist movements along the dimensions of moderation/radicalism, militance/acquiescence, and collective action/parliamentary action. The relationship with the states, however, was inﬂuenced not only by the timing and level of tolerance of the response, but also by the entire model of institutional integration of national socialism. Such a model can be approached in terms of sequences and levels of institutional integration. SEQUENCE OF INSTITUTIONAL DEVELOPMENT The smoothest institutional integration of new political movements is thought to occur when it starts to organize in the context of a liberalized but still undemocratized regime. In this case, the decisive step of suffrage democratization occurs when the party system is already structured and parliamentary control of the executive has been established. The sequence of the crucial steps of suffrage extension, regime liberalization, and socialist party foundation is therefore important. For Germany, Austria, and Italy, we have the usual problem of deciding when to set the date for the establishment of a ﬁnal responsible government, but only in the case of Italy does this 120

Geary (1981: 60–64).

398

THE POLITICAL MOBILIZATION OF THE LEFT

decision actually modify the sequence. Belgium, the Netherlands, Norway, the United Kingdom (including Ireland), and Italy introduced the principle of cabinet responsibility and parliamentary rule before the decisive extensions of suffrage. Austria, Denmark, Finland, France, Sweden, and Germany introduced manhood suffrage before the effective liberalization and institutionalization of parliamentary rule. Socialist parties were founded and centralized nationally before responsible government was introduced in Austria, Denmark, Finland, Germany, and Sweden, while they were founded after responsible government establishment in Belgium, France, Italy, the Netherlands, and the United Kingdom. In Norway, responsible government and socialist political organization developed at the same time. Ireland is a more difﬁcult case to classify. If one regards the preindependence period as one of colonial rule and considers secession as the crucial ﬁnal step in Irish liberalization, then the three processes occur at more or less the same time. Finally, the socialist parties were founded before universal suffrage was granted in Austria, Belgium, France, Italy, the Netherlands, and the United Kingdom; in the other cases, they were founded afterward. Combining these dates (see Tables 6.2 and 6.3, and 7.4) results in ﬁve basic sequences that are reported in Table 7.16. The ﬁrst and ﬁfth sequences represent two polar types in terms of institutional integration patterns. If socialist parties were ﬁnally organized within a liberalized government context and had time to consolidate before decisive steps were taken toward manhood suffrage – having time to channel new waves of lower-class voters – their integration into the existing political order was smoother. This should have resulted in a far less radical posture vis-a`-vis the political order. It could also result, however, in a weaker mobilization capacity in this early phase. This is not surprising: Smoother institutional integration made it more likely that the working class would be represented by other nonsocialist elites for longer periods of time and made socialist parties less able to capitalize on the representation of demands for liberalization, while the slowly enlarged franchise provided for slower growth of their electoral strength. In contrast, the ﬁfth sequence can be regarded as the worst combination of integration steps. Socialist parties had to live for long periods under illiberal regimes, carrying the burden of the battle for both liberalization and democratization. This may have contributed to their ‘‘total opposition’’ temperament; but at the same time, it made them more likely to capitalize electorally on such battles, sometimes even beyond their lower-class social base of support. Because it was unlikely that the lower-class groups would ﬁnd it acceptable to be represented by a liberal and radical elite whose

POLITICAL INTEGRATION

399

Table 7.16. Socialist party foundation and democratization sequences (1880– 1918)

Note: RG ⫽ responsible government; US ⫽ universal suffrage (man); SPF ⫽ socialist party ﬁnal foundation. The Italian case appears (in parentheses) in two sequences, according to whether the practice of RG is considered as achieved before fascism or only at its demise.

commitment to liberalization and democratization was, to say the least, ambiguous (see the next chapter), it is not surprising that these socialist parties were much stronger electorally than the ﬁrst group. The three intermediate sequences are more difﬁcult to interpret. The Irish and Swiss sequence should be assimilated to that of the ﬁrst group of countries for the key common feature of early responsible government. The French and Danish experiences have in common very early male suffrage. France is a unique case in which universal suffrage preceded the ﬁnal (if not complete) liberalization of the regime by about thirty years, and both preceded the late consolidation of a national socialist party. Almost half a century of free and nonfree elections had taken place when the French socialists ﬁnally managed to set up a viable national political organization. On the whole, Table 7.16 shows how signiﬁcant it was for the early electoral development and ideological orientation (see later) of the socialist movement whether the sequence of institutional development started with responsible government, with male universal suffrage, or with the early founding of the socialist party. LEVELS OF INSTITUTIONAL CLOSURE/OPENNESS From these data, it can be conﬁrmed that the institutional integration pattern was important for the early ideological orientation and electoral mobilization capacity of the socialists. In other words, the smoother the institutional integration is, the less radical and electorally mobilized the

400

THE POLITICAL MOBILIZATION OF THE LEFT

Table 7.17. Institutional integration, socialist ideological orientation, and electoral mobilization (1880–1917)

Note: Number of elections given in parentheses.

class cleavage is. I will pursue this line of argument in looking for further proof and reﬁnement of this basic idea. Specifying the crude sequence pattern with information about the actual level of representation and repression improves our understanding of the differences among cases. To do so, the cases can be classiﬁed on the basis of the information discussed so far about the level of parliamentarization of the regime with the level of repression of the socialist movement and the use of representational obstacles against it (this last factor is obtained by combining the level of electoral underrepresentation [Tables 7.6 and 7.7] with the level of unfairness [Table 7.5] of the electoral system).121 In Table 7.17, I present the crosstabulation of the three main institutional integration variables. For each cell, row, and column, the average left electoral size is reported. In addition, cases are identiﬁed with their early ideological orientation, as discussed in Chapter 2. The ﬁgures in the table point to the role of each variable as being 121

The two aspects are separate because unfair mechanisms (for instance, indirect, nonsecret, or curia voting) did not necessarily result in underrepresentation. For instance, if the socialist party is very weak electorally, the level of underrepresentation will by deﬁnition be minor.

POLITICAL INTEGRATION

401

consistent with the theory of institutional integration advanced so far, as well as with a strong, coherent, cumulative interaction effect among the three variables. This justiﬁes the claim that these variables should be considered as additive dimensions of the general problem of institutional integration. The pattern of this integration is strongly related to both the early ideological orientation of the socialist movement and to its early level of electoral mobilization. The cases I have identiﬁed as characterized by the persistence of an orthodox Marxist doctrine up to the 1920s and beyond all fall into the category of high-stateness countries, with high levels of state repression combined with low parliamentarization and high obstacles to representation. The socialist movement faced an adversarial state as well as a closed political system, and under these circumstances, the most coherent version of revolutionary Marxism persisted as a convincing interpretation of the relationship between the working-class movement, the state, and the market class conﬂict. Eclectic socialism, where radical syndicalism survived along with orthodox Marxism and with various forms of reformist socialism (and occasionally even trade unionist visions), was a sort of split answer on the part of the labor movement that was consistent with a contradictory environment. In Italy and France, the repression waged by high-stateness polities against the socialist movement – and in particular against its more radical components – was considerable. At the same time, however, the parliamentarization of the regime and the relatively low representation obstacles (even in Italy) left open within the socialist elite and rank and ﬁle the debate about what the best strategy for labor was. In none of the countries where parliamentarization was achieved early and repression was low did these ideological tendencies prevail. Either trade unionism or mild and pragmatic forms of Marxism existed where theoretical lip service to the orthodoxy was accompanied by the domestication of practices. Where, as in Norway or Switzerland, a radicalization phase emerged after World War I, it was short-lived and did not crystallize into long-term features. At the same time, institutional integration also inﬂuenced the level of electoral mobilization of the left. From the end of the nineteenth century to the post–World War I period, in the eighty-two elections fought in Western Europe, the average electoral strength of the left was 18%; it was 14.5% in the ﬁfty-one elections held in countries with low repression and 23.8% in the thirty-one elections fought in countries with high repression. The left had, on average, 13.3% of the vote in highly parliamentarized countries (ﬁfty-six elections) versus 28.1% (twenty-six elections) in the countries that were still autocratic and 13% (twenty-nine elections) in countries with low obstacles to representation versus 20.7% in countries

402

THE POLITICAL MOBILIZATION OF THE LEFT

with high obstacles. If we contrast the group of countries/elections held under the best conditions of institutional representation (high parliamentarization, low obstacles to representation, and low political repression) to those held under the worst conditions (low parliamentarization, high obstacles, high repression), the difference ranges from 8.7% to 29.9% of the vote. In situations of low institutional integration, the socialist parties obtained four times more votes than in those of high institutional integration. What is more striking is the almost perfect interaction effect within the cells of Table 7.17. Cases of high obstacles to representation show a stronger left than cases of low obstacles for each level of parliamentarization. The consistency of these data is striking. If the institutional sequence discussed before was a distal precondition of institutional integration, the three variables analyzed here point to concrete aspects of institutional integration, and they work much more effectively. Thus, the consistency of the results greatly strengthens the general hypothesis of the early discriminating role played by the pattern of institutional integration. A further provisional test (a more thorough analysis will be performed in Chapter 9) of the importance of institutional integration aspects can be made by controlling their impact on the cohesion of the left and communist vote. According to the logic so far described, the class left developing in closed political systems or in mixed environments where regime liberalization was accompanied by persistent bureaucratic harassment should be more highly mobilized electorally, but also more internally radicalized, more prone to strategic and ideological dissent, and therefore more organizationally fragmented than the left developing in more open political systems. We can therefore control this hypothesis in the period between the world wars, when ideological and organizational fragmentation developed within the European left (see Chapter 3). In the period between the world wars, then, the average communist vote was 8.7% (seventeen elections) in the countries with a history of closed institutional systems, 3.6% (thirty-two elections) in countries in an intermediate position, and 1.9% (eighteen elections) in countries with long periods of institutional openness. In the case of the radical left, and in particular the communists and Third International groups, what had already happened to the socialist family simply occured again: Where integration into the political system was early and good, their development was late and slow, while it was rapid and fast in the opposite cases. Once the socialists had gained a degree of legitimate access after World War I, even in formerly closed systems, memories of past closure and debates about the viability of such integration remained strong, and in many cases

POLITICAL INTEGRATION

403

institutional-exclusion mechanisms worked against the new radical left in much the same way as they had against the socialists a few decades before. Finally, let us consider how the institutional-integration setting inﬂuences the size of the left in various periods by inserting it into the general model developed for each of the macrofactors discussed in the previous chapters. To do so, I insert the index of institutional closure, as measured by the sum of the country ranking in the three dimensions of repression, parliamentarization, and representation obstacles,122 into a linear regression together with social mobilization, cultural heterogeneity, the enfranchisement pattern, and the organizational variables. In Table 7.18, the beta coefﬁcient and the variance explained by this ﬁve-variable model are reported for the whole period and for each phase. These data should be compared with those in similar tables in previous chapters. I have argued that the more closed the institutional setting, the more likely early mobilization and a higher level of support for socialist movements. The difﬁculty of the political environment made socialist propaganda more convincing to voters and promoted early and close contacts between corporate and political socialist organizations. In contrast, an open institutional setting made the early political defense of working-class interests less necessary, the corporate and political organizations less closely 122

The institutional closure rank ordering of countries is as follows: parliamentarization rank ordering

representative obstacles rank ordering

repression rank ordering

Combined rank ordering

Au 13

Fi 13

Fi 13

Fi 37

Ge 12

Au 12

Ge 12

Au 36

Fi 11

Ge 11

Au 11

Ge 33

Sw 8

No 10

It 10

It 25

Dk 7

It 9

Fr 9

De 19

Ir 6

Sz 8

Be 6

Fr 19

Fr 6

De 6

Dk 6

No 18

It 6

Be 6

Sw 6

Sw 16

No 5

Fr 4

Ne 6

Be 15

Ne 4

Ne 3

Ir 4

Sz 14

Be 3

Sw 2

No 3

Ne 13

Sz 2

Ir 2

Sz 2

Ir 12

Uk 1

Uk 1

Uk 1

Uk 3

404

THE POLITICAL MOBILIZATION OF THE LEFT

Table 7.18. Social mobilization, cultural heterogeneity, enfranchisement, organizational model, institutional integration, and left vote (beta coefficients)

connected, and the persistence of nonsocialist representation of the enfranchised working-class segments more likely. The ﬁgures in Table 7.18 offer new support for this interpretation. On the whole, the inclusion of institutional closure/openness in the model changes only marginally the overall explained variance with respect to the four-variable model discussed in reference to Table 6.16 in the previous chapter (70.1% versus 71.9%). However, its impact in different phases is important. For the period before 1918, the inclusion of institutional closure/openness increases the explained variance of 14% (from 54.7% with four variables to 68.1% with ﬁve variables). A considerable impact is also felt during the interwar period (from 71.1% to 78.3%), while this factor loses its weight after World War II, as expected. In the first period, institutional openness/closure is the single best predictor of the earliness and level of left electoral mobilization. It gains this role largely at the expense of the enfranchisement pattern. Thus, early crossnational differences in the mobilization capacities of the socialist movement should be related ﬁrst of all to institutional differences in the liberalization/democratization processes. The overall picture resulting from this ﬁve-macrofactor analysis is the following. The role of social mobilization is enhanced in each period, and this factor now proves to be the most stable systematic inﬂuence on the left vote once the other factors are accounted for. Cultural heterogeneity is similarly enhanced in its weight and remains a systematic factor both overall and for each period. The impact of the enfranchisement pattern is maximum at the early stage and declines over time. The organizationalaspects factor is unimportant in the ﬁrst period, becomes dominant be-

POLITICAL INTEGRATION

405

tween the wars, and declines thereafter. Finally, institutional openness/ closure is dominant in the ﬁrst period and is still important in the second, declining thereafter. In the two early periods, institutional and organizational features determine the conditions of the impact of the underlying structural processes pertaining to social mobilization and cultural heterogeneity. Finally, with the incorporation of the institutional and organizational aspects linked to the enfranchisement pattern, organizational consolidation, and institutional integration, I have developed a model whose performance is roughly the same in each period, ranging from two-thirds to three-quarters of the explained variance at the overall European level.

HOSTILITY TOWARD THE STATE AND IDEOLOGICAL ORIENTATION In discussing institutional responses, the primary emphasis can be placed on the compatibility of the economic demands – in other words, on whether the established interests (landed and/or bourgeois) perceived a fatal threat to their economic position, possibly resulting in an alliance of landed and bourgeois interests against the working-class political movement. This line of argument induces a state/institutional response from the compatibility of class interests. It therefore assumes that the state was the guardian of such compatibility and that, by necessity, anticapitalism and antistate attitudes should combine and overlap. To test this view, one should be able to demonstrate that the variance in the repressive response is a function of economic interest compatibility and that in those countries that yielded more gracefully to pressures from below, the conﬂict of economic interests were more compatible than elsewhere. I have preferred to argue that the definition of the conflict of interests as compatible or not was shaped by extraeconomic power relationships and bases. The resources available for repression and the historical pattern of institutional development could be seen as independent variables in the deﬁnition and delimitation of the conﬂict of interests. That is to say, to rely on state apparatuses, to rely on institutional obstacles and privileges (upper houses, king, etc.), or to rely on political control (see the next chapter) was not always a choice. Birnbaum has related the nature of the state to the development of the ideologies that structured the collective action of the working class: trade unionism, Marxism, and syndicalism.123 His model is based on the inter123

Birnbaum (1988). Katznelson (1985) has followed a similar logic, focusing on the state as a major source of variation in patterns of ‘‘class formation.’’ The U.S. versus Britain

406

THE POLITICAL MOBILIZATION OF THE LEFT

action of four variables: (1) the process of political centralization leading to a ‘‘minimal’’ state; (2) the fusion or differentiation between the state and the ruling class; (3) the industrialization coming from below or above; and (4) the open or closed nature of the bargaining process. Birnbaum relates these variables to the ideological response of labor. In Germany, the consolidation of a strong state together with the ﬁssion of the state with the dominant aristocratic class and industrialization from above124 in a nonliberalized regime, produced a dominant Marxist ideology that was stronger than that of anarchism. In France, the state was differentiated from the ruling class and bargaining was open, leading to a predominance of syndicalism over Marxism. Finally, in the United Kingdom, weak structural differentiation of the state and industrialization from below, together with a liberalized regime, tended to produce a predominance of trade unionism as the ideological response of the working class. Anarchism, in this light, established itself in countries ruled by strongly dominant regimes, such as France, Italy, and Spain. The emphasis on selected wellknown cases gives Birnbaum’s argument an ideal-type ﬂavor. The ideal type of strong state is deﬁned in reference to concepts such as ‘‘differentiation,’’ ‘‘autonomy,’’ and ‘‘institutionalization’’ that are difﬁcult to operationalize; and in the ﬁnal analysis, the classiﬁcation of the British versus the French versus the German experiences is based largely on common knowledge. This makes it difﬁcult to apply the scheme to other European countries, which are not dealt with or are only occasionally mentioned as examples of intermediate or mixed cases. I have tried to generalize the classic argument of a British versus a continental stereotype to explain within-continental variations. To avoid the tyranny of the large countries and of pattern cases, I have developed indicators to locate the whole range of European experiences along dimensions of a model of institutional integration that can account for the fateful overlap between the class cleavage and the antistate cleavage. Stateness and dominant-interest differentiation represented varying conditions for a possibly repressive response by the dominant classes. The openness/closure of the institutional system was an intervening element that made a repressive strategy possible or not, balanced it in some cases, and reinforced it in others. A further element should be added to this picture: the availability of a political control over the socialist movement that permitted its con-

124

contrast, however, describes the British working-class formation in terms that could apply to the whole European experience and that are therefore of little use in differentiating within Europe. Industrialization is seen by this author as an intervening variable. Birnbaum (1988: 79).

POLITICAL INTEGRATION

407

tainment via politico-electoral parliamentary means (rather than via direct repression or institutional privilege). This required liberalization and favorable conditions of structuring for nonsocialist parties and for the party system in general. Repression and institutional exclusion were more likely the less viable a strategy of political containment was. This last component of the labor environment will be analyzed in the next chapter. The factors I have placed in the picture so far are reported in Table 7.19, together with the ideological outcomes with which they are associated. The issue raised in this discussion is the extent to which the class cleavage, whose structural roots were primarily in social antagonisms in the market, came to be ideologically hostile to the state. It is often argued that ‘‘the formation of demands related to the working condition and the organization of various forms of workers’ action inevitably pushed the workers into conﬂict not only against the employers, but also against the state.’’125 Marx was, of course, consciously aiming at this combined opposition because he wanted to transform an emancipation movement into a revolutionary movement. However, there was nothing inevitable in this link; it was not necessarily the case that working-class political mobilization would come into conﬂict with the state. In terms of hostility toward the state, the European experience is far more varied than a British versus continental pattern or a French versus German pattern would indicate. In addition to countries where the state was strong, there were others where it was far less visible or even absent. The use of the state apparatus was also different. The relationship of national movements with their respective national states was therefore different, and the fusion of class feeling with antistate feeling depended on the concrete conditions of state involvement in their emancipation attempt. The opposition strategies of national dominant forces toward the emerging socialist movements in the last twenty years of the nineteenth century and the ﬁrst twenty years of the twentieth gave rise to two possible reactions: (1) to help and favor the success of the Marxist identiﬁcation between the class enemy (the employers) and the state (the class nature of the state) or (2) to recognize that the struggle for socialism was the same as the struggle for a democratic state, that is, for state democratization. The political integration of a social group into an existing political order involves cultural elements and is ultimately expressed through a set of beliefs and attitudes toward the political system. The development of these cultural traits results from organizational activities and the learning experiences concerning the response to these activities. The extent to which the new actors felt they could achieve their aims within the given 125

Valenzuela (1981: 448).

408

THE POLITICAL MOBILIZATION OF THE LEFT

Table 7.19. The institutional integration syndrome and predominant ideological orientation

institutional structure has been indirectly deduced in this chapter from the process of the institutional integration of their political movements. Institutional integration inﬂuences the extent to which political and corporate movements are inclined to concentrate their pressure on the government and on employers through electoral, parliamentary, and judicial means or else to use other resources and conﬂict arenas that are extrainstitutional. The dominant institutional pattern, therefore, deﬁnes a series of opportunities and constraints of the political and social actors and inﬂuences their acceptance of the pattern of political competition prevailing within such a structure.

POLITICAL INTEGRATION

409

Only when the almost universal economic and ideological conﬂict with employers overlapped with an equally intense confrontation with the state apparatus and the closed major political institutions did the working-class movement become more radical and revolutionary, concentrate more heavily on extrainstitutional activities, develop its isolationist and conspiratorist nature, and ﬁnally ﬁnd a convincing ideological guide in radical versions of socialism (Marxism and Leninism). The more a closed political system was, characterized by high stateness of its processes, presenting itself as a moral and teleological entity independent of weakly differentiated and weakly self-represented societal interests, the more radical ideologies appeared to be theoretically persuasive, reﬂecting reality in an appropriate manner and providing convincing strategies to overcome it. Thus, whenever state machinery was used consistently against the socialist movement, opposition feelings were translated directly into deep antistate opposition that combined with the social and economic alienation of the laboring classes to produce the explosive mixture of political and social alienation that is conducive to radical responses and organizational divisions at the same time. The persistence of nonresponsible government and an executive separated from the parliamentary arena and guarding other traditional institutions, meant that opposition political movements cemented their links with social groups and corporate-interest organizations and that the basic conﬂict of all of these groups and organizations focused on those extrainstitutional power bases that prevented the parliamentarization of government. When the resources and the willingness to repress the socialist movement made this repression viable, deep divisions usually appeared within the movement – divisions concerning whether the values sought could be achieved within the existing institutional framework – and profound, long-lasting splits of the working-class organizations along radical versus moderate lines occurred. Thus, moderate calls for a peaceful and piecemeal insertion postulated that institutional channels were open enough to allow for democratic changes, while radical arguments for a revolutionary breakthrough assumed that the institutional regime would never allow them to compete fairly. In accepting the costs of repression, the established elite implicitly assumed that the demands of the new claimants could in no way be accommodated within the existing order. Intermediate cases, where a low differentiation of dominant interests combined with high stateness to foster a repressive strategy and where, at the same time, the openness of the institutional system softened this strategy and seemed to offer alternative future opportunities, tended to result in ideologically divided and eclectic socialist movements, as in Italy

410

THE POLITICAL MOBILIZATION OF THE LEFT

and France. In L’Arme´e Nouvelle (1910), Leon Blum dealt with the concept of the state, afﬁrming – in contrast to Marx and the French Guesdesists – that not only did the proletarians have a patrie and need to defend it, but also that the state was not the instrument of class domination but a possible arbiter among the classes, which ‘‘n’exprime pas une classe: il exprime les rapports de classe, c’est a dire les rapports des leur forces.’’ In so doing, Blum reached a conclusion that was not controversial in many European countries, where verbal agreement with the Marxist programs and orthodoxy had never been fed by profound hostility toward the state. Cases of low or medium interest differentiation and low stateness, unable to afford a cohesive repression response but able to resist the opening up of the institutional system, tended to produce formal Marxist orthodoxy; these cases, however, were accompanied by constant breaks in such orthodoxy in practical everyday political life, as in the Scandinavian countries.

8

CLEAVAGE STRUCTURES

n addition to resting on repression and institutional closure, opportunities for socialist political control depended on the development of electoral-politico-parliamentary blocs capable of enforcing the political domestication of the socialist movement. Many studies stress the importance of early alliances among major economic and social groups, assuming that the nature of the political regime developed from the logic of the subtending socioeconomic coalitions.1 Instead, I place more emphasis on the logic that political coalition results from speciﬁc forms of political representation developed by the main social groups during the phase of the party-system structuring. In this view, large social groups do not make coalitions or form alliances unless they are able to develop forms of political selfrepresentation. This chapter examines the class cleavage in the light of the process of structuring of the national party systems or, more precisely, structuring stable sets of coalitions and alliances among social groups and political organizations. Within each national context, I discuss the following:

I

1. Which other social movement, if any, preceded or accompanied the socialist movement and what its political nature was (antiestablishment, nationalist, popular self-representational, or merely institutional, such as the opposition of the Catholic clergy). 2. What cleavage issues were already politicized when the labor movement entered the electoral race; what kinds of political parties, if any, 1

Moore (1966), Luebbert (1991), and Rueschemeyer, Huber, and Stephens (1992). 411

412

THE POLITICAL MOBILIZATION OF THE LEFT

were already active in electoral and partisan recruitment; what political alignments and national coalitions were already established among the forerunners and to what extent these reduced the range of options open to the socialists. 3. What relationship the incipient socialist organizations established with these forerunner political movements and parties – forms of delegated political representation, formal alliances, or unstable short-term cooperation. This chapter is developed on the basis of two central hypotheses. The ﬁrst is that the earlier and more encompassing the organizational mobilization of most sectors of the community was through non- or pre-socialist political movements, the earlier the socialist entry was but the lower was its capacity to mobilize. Early nonsocialist political mobilization preempted later socialist mobilization by ﬁxing and strengthening political identities, which progressively narrowed the potential market for the support of the latecomers. The second hypothesis is that the alliance choices and alignments made by the early insiders reduced the alliance alternatives and opportunities available to latecomer outsiders and therefore signiﬁcantly shaped their choice of allies and adversaries among both social groups and political organizations. To anticipate the main thesis in brief, class mobilization should be regarded as a residual process whose coordinates were established by the previous mobilization opportunities and the moves of other social groups and sociopolitical cleavages. Instead of providing the determining factor of political alignments in Western European party systems, class cleavage was dependent on such alignments. Only when circumstances prevented, delayed, or made impossible any form of nonclass mobilization did class become the dominant force and take on predominant importance in elections.

OPPORTUNITIES FOR SOCIAL AND POLITICAL ALLIANCES In a model of political mobilization in which social position is easily converted into social and political consciousness and collective action, the main social groups would tend toward self-representation and would refuse to be represented indirectly by other sociopolitical elites. When this occurs, the square of social groups and that of the political movement reported in Figure 8.1 tend to coincide, with the peasant world organizing an agrarian

CLEAVAGE STRUCTURES

413

Figure 8.1. Alignment of social groups and political movements. party; the aristocracy, the landowners, and the dynastic bureaucratic interests organizing a conservative defense of their position; the bourgeoisie expressing its vision of society through a liberal party; and, ﬁnally, the working class endowing itself with a socialist movement. In this imaginary case, the working class enters politics in a political world and party system that are already socially well structured. This perfect coupling is not realistic, however, even in a historical context in which socioeconomic interests tend to predominate and shape the vision of the world. The representation of agrarian interests and society (the upper part of my social group square) against external forces may predominate over the internal division of the agrarian population, and peasants and landowners may ﬁnd their positions very close in the defense of common interests. On the other side of the square, the working class and the bourgeoisie, who are both representative of the industrial society,

414

THE POLITICAL MOBILIZATION OF THE LEFT

may, in an early phase, share interests and forms of representation for quite a while. If we read Figure 8.1 differently, along the axis that opposes the established social and political elite with the lower classes, common interests may exist between parts of the peasantry and the working-class movement, on the one hand, and between the bourgeoisie and the landowners and aristocratic interests, on the other. The picture becomes more complicated if, in the process of political mobilization, forces act with no clear relation to the main agricultural or urban established or emancipationist social groups. Denominational mobilization, as well as nationalist mobilization, had fewer clear social references and were able to drain support from all the most important social groups, as well as to come to represent one of them predominantly. The diagram in Figure 8.1 is useful, as it simpliﬁes the alliances that were important for the incipient working-class movement. Fundamentally, its political representatives had two social groups and political movements with which to deal. Depending on the circumstances and opportunities, they could establish political links with the liberal bourgeoisie or, alternatively, strengthen their links with the peasants and sectors of the agricultural world. In other cases, they had to come to terms with mobilized denominations. They could also join national independence fronts and coalitions when the problem of national-independence was not solved. Finally, we can envisage a case in which the socialist movement could not establish any stable alliances with other social groups, either because they were not self-representational or because the socialist movement was itself divided on such alliances and was unable to make a stable choice. The alliance of the working class with liberal forces against conservatives (1) presupposed a relatively strong, autonomous politically organized urban and/or rural bourgeoisie; (2) had as the main point of its program liberalization of the regime (if this was delayed) and the defense of the rights of the socialist movement and its organizational conquests; and (3) had as its main weakening element the growing industrial conﬂict (if urban interests predominated). In contrast, the alliance with agrarians (1) presupposed a strong and independent peasantry, capable of self-representation and possessing antiestablishment and countercultural traits; (2) the main points of its program were political liberalization/democratization and the exchange of policy support in respective sectors of interest; and (3) the main problem was the compromise on agricultural protectionism and the defense of small-property owners in the countryside.

CLEAVAGE STRUCTURES

415

The alliance with religious dissenting movements (either Catholic or Protestant, where Catholicism was not dominant) (1) required their being sufﬁciently strong and antiestablishment to offer a basis for such an alliance; (2) the main points of its program were social policies and economic interventionism; and (3) the major problems were state secularization and high competition for lower-class representation. The alliance with nationalist movements (1) presupposed that the mobilization of the left was taking place parallel to the struggle for nation building and national independence; (2) was based on the common interest in national independence from foreign control; and (3) the main problems were the political distinctiveness of the socialist movement within such an alliance and its internal differentiation once independence was achieved. The no-alliance option (1) required either that no real structuring of the party system took place, so that no reference point could be offered to the socialists, or that the party-system structuring would create a political bloc that was strong and cohesive enough to do without the socialists; (2) it helped to maintain doctrinal purism and ideological and organizational unity when this was not a default strategy for movements so internally divided as to be unable to choose alliances or to be credible allies; and (3) its main cost was limited policy inﬂuence and, perhaps, delayed executive access. All alliance options had some advantages and disadvantages. Whether these were based on short-term political goals or on policy exchange, they blurred party images, forced ideological compromise, increased gaps between doctrine and practice, and increased internal divisions and their opportunity costs. They always demanded a high investment in order to keep the socialist movement cohesive and united and, in many cases, were crucial in determining its internal political fragmentation. These choices and opportunities were not ﬁxed permanently and changed substantially between the end of the nineteenth century and the 1930s. No signiﬁcant changes occurred after that date until the 1970s. In the following sections, I discuss the European cases and focus attention more on the political movements than on the social groups. Choosing classes or social groups as main actors means overlooking the fact that they are rarely sufﬁciently united and coherent to allow the recognition of their political preferences without making references to the political elite’s interpretation and articulation of them. By the same token, such an approach frequently underestimates the extent to which the organized political elite can contribute to the deﬁnition of the preferences, interests, and demands of a social group.

416

THE POLITICAL MOBILIZATION OF THE LEFT

EARLY POLITICAL MOBILIZATION AND THE ROLE OF LIBERALISM AND THE BOURGEOISIE COOPERATION WITH LIBERALISM In terms of opportunities for alliance with liberal forces, a clear dividing line separates the northern European countries – Scandinavia and the United Kingdom – from other continental nations. In these countries, (1) the entry of socialism and the ﬁnal structuring of the party system were preceded by the existence of signiﬁcant popular, politically conscious movements; (2) although British liberalism was mainly urban, commercial, and industrially based and Scandinavian liberalism was more farmer-based, such middle-class movements were largely self-representational, that is, they did not delegate political representation to other sociopolitical elites; (3) the liberal–labor alliance was a key early feature of party-system formation and lasted for long periods; (4) in all cases, the tensions within liberalism during or immediately after the phase of cooperation with the socialists eventually split the liberals (in 1905 in Denmark; from 1922 to 1923 in Sweden; after World War I in the United Kingdom; with continuous split-ups in the 1930s in Norway); (5) as a consequence, groups with liberal tendencies, even when they opposed socialist-led governments, maintained an internal radical tradition that occasionally surfaced and made their relationships with the respective conservative parties difﬁcult.2 In the United Kingdom, Chartism was the most important presocialist movement. Unions began to organize and develop after the Chartist failure between 1830 and 1848 to obtain universal manhood suffrage, annual parliaments, and the secret ballot. Similarly important was the middleclass Anti-Corn Laws League. This movement was so efﬁcient in organizing middle-class opinion and so successful in achieving its representation that ‘‘its methods were later adopted by the political parties.’’3 In Sweden, the nineteenth century was densely populated with autonomous and selfrepresentational religious, social, and political movements made up of the middle classes and the peasantry. For example, the Free Church Movement wanted to convert individuals to a Christian life and create a Christian society, while the temperance movement fought to save individuals from the excesses of alcohol and to build a sober society. Although the extension 2 3

On Scandinavian liberalism, I have relied on Berglund and Lindstro¨m (1978). On the development of early opposition in Great Britain, the best source is Potter (1966: 26).

CLEAVAGE STRUCTURES

417

of their mobilization is subject to debate among Swedish historians, it was very high by any comparative European standard.4 Norway also has a long history of dissenting popular movements, particularly peripheral ones, whose development and success demonstrate the low level of repression and high tolerance by the central authorities already discussed.5 The earliest of these movements was the successful peasant protest against the head tax and in favor of the return of this money by the state (1765). Again, in the ﬁrst half of the nineteenth century, what was known as ‘‘Hauge’s movement’’ successfully mobilized large sectors of the peasant population. Hauge was a lay preacher of peasant parentage who criticized the ofﬁcial, urban-based, secularized clergy; challenged the traditional and conventional doctrines and preachings of the national church; and organized a movement of traveling lay preachers, primitive cooperative ventures, agricultural communities, and colonies in territories still not exploited. His movement established the ﬁrst elements of a nationwide political organization and raised the political consciousness of the peasantry and rural middle classes, which contributed to the political detachment of the rural groups from the central establishment.6 Although there are similarities to British Methodism and to the theology of acceptence of German Pietism, Haugianism was much stronger and more aggressive toward the national Norwegian Church, which labeled him a ‘‘Jacobin.’’ Equally important for the lower classes was the Thrane movement. This was not an early or protosocialist movement, but it was certainly a class movement; intellectuals were totally absent from its membership, and its social base was composed of a few industrial workers, peasants working for the rich landowners, and small, impoverished landowners from the west coast. It was a forerunner of the trade union movement, active in organizing strikes, advocating revolutionary aspirations, and making general suffrage claims. Its organizational strength was remarkable for the times; between 1848 and 1854, 273 associations were formed with about 20,854 members. It took the Labor Party twenty-three years to reach the same level of membership that the Thrane movement had at its best. Finally, the ‘‘Friends of the Peasants Movement’’ (1865–1879) was a more periphery-against-the-center movement and was the ﬁrst modern political move4

5 6

Stra˚th (1988a: 32–34) estimates that one-third of the population took part in these movements. See Lafferty (1971: 117–119). On Hauge and ‘‘Haugism’’, see Derry (1968: 122–124).

418

THE POLITICAL MOBILIZATION OF THE LEFT

ment in Norway, with a strong press, organization throughout the whole territory, and 21,000 members. It gradually fused with the urban liberals to form the ﬁrst national Norwegian party (Venstre) to push successfully for parliamentary government (1884).7 In Denmark, similar popular movements were less important than in the other two Scandinavian countries because the native landed aristocracy was stronger.8 However, Danish liberals also organized as an uneasy alliance between urban intellectual radicals and rural interest groups. In the history of the liberal movement, there always had been strong tension between such rural interests and the rather radical urban and intellectual wing, and when the agrarian component became dominant in 1905, the party split, giving rise to an unusually radical ‘‘social’’ version of liberalism, the Radical Liberal Party, which remained a loyal ally of the socialists for a long time. These movements in Scandinavia had Puritanism, egalitarianism, and antielitism in common; they regrouped both their urban and rural lowerand middle-class sections into an opposition; and their antiestablishment ideology was, if not democratic, at least popular and egalitarian. They offered an ideal basis for alliance and support to the developing workingclass movement, which, in the 1880s, thereby joined an already existing popular antiestablishment tradition. Thus, socialists, while making a different political offer, found a receptive audience. In Denmark and Sweden, it was belated liberalization and parliamentarization that fostered close political cooperation and common mobilization between the liberals and the socialists. In Norway, after the repression of the Thrane movement, government ofﬁcials and the bourgeoisie tended to form labor associations such as labor clubs, cooperative societies, and permanent-assistance funds, with the goal of integrating the working class into the liberal regime. Thus, about half of the industrial workers in 1885 were members of an insurance association.9 In the 1880s, these associations were rapidly politi-

7

8

9

Rokkan (1967) has depicted these and other movements (the teetotal movement, the religious layman’s movement [mission], and the vernacular [nynorsk] movement) movements as mainly territorial and countercultural. An economic explanation can also be advanced, relating to an important inequality dimension between regions: The southern and western regions had a relatively egalitarian agricultural structure. In areas with a more marked inequality structure (the north), conﬂicts focused on the defense of privilege and inequality (Conservative Party) and the reduction of inequality (Labor Party). In Finland and Norway, industrial poverty and semicolonial dependence almost completely prevented the formation of a native aristocracy. For a discussion of the strength of the native aristocracy in Scandinavia countries see G. Therborn (1977: 3–41). Terjesen (1990: 113).

CLEAVAGE STRUCTURES

419

cized and taken over by the Liberal Party, which, in turn, handed them over to the Social Democrats during the following two decades. In 1893, Venstre still had considerable working-class support with twenty-ﬁve collectively afﬁliated unions in Oslo. In the elections of the 1890s, the Norwegian Socialist Party (DNA) polled fewer votes than its members, as most of the enfranchised workers voted for Venstre, whose more liberalsocial tendencies had begun to dominate the party in 1891. Not only the franchise enlargement but also the majority of social policies up to 1914 were supported and pushed through by Venstre, as the Social Democratic Party had no inﬂuence on policy until that date.10 On the one hand, the national question and opposition to union with Sweden, which dominated the political debate up to 1905, acted as an obstacle to the early mobilization of the labor movement. On the other hand, many of the reforms carried out – including early universal male suffrage in 1898 – can be regarded as concessions to win the labor vote that Venstre needed to achieve parliamentary power and dissolve the union with Sweden. However, there is an important difference between the way in which the peasantry in Sweden and Denmark was mobilized and integrated into the national political community and, as a consequence, the way in which the rural middle classes came to relate to the socialist movement. In Denmark, the long-lasting absolutist rule and the early universal male enfranchisement created a profound conﬂict between the mobilizing peasantry and their landowners (the most powerful in all the Scandinavian countries) and urban-bureaucratic antagonists. In Sweden, instead, the tradition of estate representation eased the entry and domestication of the upper echelons of the peasantry, resulting in a less deep conﬂict than that of Denmark and more varied alliances between early peasant representatives and other forces. Clearly, the political alliance between Venstre and the socialists was widely discussed and criticized within the Danish Socialist Party. The renewal of the alliance occasionally led to dissent and even splits, although, in the end, it remained a central pillar of the party’s strategy. Between 1903 and 1910, the alliance with Venstre became more difﬁcult, as common political goals were limited and growing conﬂicts over economic policy were emerging. The fact that at that time the agrarian party split and the newly born left-wing (and antimilitarist) Liberal Party became a new potential ally was a great asset for the socialists. In Sweden, the more integrated stratum of the peasantry joined with the urban bourgeoisie to form a moderate or even right-wing protectionist

10

See Kuhnle (1978: 9–35).

420

THE POLITICAL MOBILIZATION OF THE LEFT

nationalist party. Only later did the peasant component break away to form a speciﬁcally agrarian party.11 In the United Kingdom, the basis for the alliance was different; it was not based on the request to liberalize the regime, but rather on the defense of the unions.12 The ﬁnal separation between labor and liberalism, a long process, was completed only in the 1920s. The early liberal orientation of the British working class and trade unions rested on a structure of social relations that, at the local level, reﬂected the ‘‘paternal and co-operative basis of class relations.’’13 The employers inﬂuenced the social and political life of their communities and were often voted for and supported by their workers. As the alliance was made among urban social groups, progressive political separation was fostered by economic crises and the harsher employment and wage policies of employers, as well as by the growing strength of trade union strikes. The early liberal political representation of the working class was progressively undermined by the growing number of competing candidates at the local level and in different situations of local compromise or conﬂict. However, at the national level, the close cooperation of Labour and the Liberals continued until World War I. Indeed, the independent identity of the Labour Party was explicitly declared by the Labour Representative Committee only in 1899, with a close vote and little enthusiasm for it. From 1906 on, Labour MPs acted as a wing of the Liberal Party in Parliament. Until 1910, the Liberal Party was the largest and most representative party of the working class, and up to 1914, when the break with the Liberals at the local level had already been completed, at the national level Labour was unable to win a seat if it was not allied with them.14 The ﬁnal break with liberalism occurred in 1918, when the Labour Party organized itself nationwide and adopted a constitutional commitment to collective ownership. Some authors argue that the decline of liberalism and the rise of Labour should not be attributed to a ‘‘class’ dynamic.’’ In other words, the fact that Labour was a working-class party did not necessarily imply that, when the working class was given the right to vote, the class cleavage would become dominant and the Liberal vote would inevitably shrink and eventually be swallowed up.15 Some (liberal) historians have argued that the Liberal Party declined from 1918 on not 11 12

13 14 15

For these notes, see Rokkan (1970g: 67–68). On the British Liberal–Labor relationship in this phase see, Breuilly (1992), which, despite its title, deals exclusively with Britain and Germany. Price (1990: 19). For details, see Luebbert (1991: 24). Matthew, McKibbin, and Kay (1976: 723–752).

CLEAVAGE STRUCTURES

421

because of the inevitable negative social changes and transformations, but due to its internal political divisions and splits during the war period (between Asquith and Lloyd George in 1916), which prevented the party from offering an adequate response to the trade unions’ requests.16 This thesis17 emphasizes crucial political factors in conjunction with the majority electoral system. The Liberals did not come to believe in proportional representation until the 1920s, when they were on a downward slide, and by then it was too late to make electoral reform a condition of their support for a Lib–Lab governmental coalition.18 A further Liberal attempt to achieve electoral reform in exchange for governmental support was made between 1929 and 1931 during another minority Labour government. Once again, there was resentment within the Labour rank and ﬁle about this, and no agreement could be reached between the two parties. If proportional representation (PR) had ﬁnally become law after World War I, as it had in most other European countries, the Labour Party might well have taken about 20%–30% of the vote and would have failed to establish a corporate relationship with the trade unions and the working class. In the intense electoral period between 1918 and 1931 (six elections), notwithstanding the plurality formula, the United Kingdom had a three-party system, with several Liberal splinter groups. Moreover, during the 1920s, nonparty candidates still enjoyed considerable success. On the left, too, the Independent Labour Party and the Communists were not just mere labels. In 1924, the Labour Party became the second party in terms of number of seats, but it surpassed the Liberals in electoral strength only in 1931. Even after World War II, the Liberal resurgence of the 1960s, and later the Social Democratic Party (SDP) and Liberal alliance, and the signiﬁcant advances of the Scottish and Welsh nationalist parties in the 1970s, are witness to the fact that the two-party structure was and still is maintained by the mechanisms of the electoral-plurality law. In the 1920s, in a comparative perspective, Labour’s strength was relatively modest and very much in line with my expectation that early and gradual industrialization, a liberalized context and an open institutional system, and possible political allies for the representation of strong industrial unions are far from favorable conditions for the strong electoral mobilization of the workingclass movement. It is therefore not unlikely that the decline of British liberalism was due to its internal divisions following World War I, which 16

17

18

For this thesis, see Clarke (1971); for a summary of the historiography debate on this issue, see Ball (1987: 82–86). Which is not unchallenged in the literature. The historical analysis of British electoral alignments made by Wald (1983) does not support this line of interpretation. Butler (1990: 457).

422

THE POLITICAL MOBILIZATION OF THE LEFT

prevented it from establishing PR and thus maintaining a longer and larger representation of the working-class world. In conclusion, the strong industrial trade union movement in the United Kingdom was successful in ﬁghting and isolating independent working-class movements that attempted to deﬁne politics in terms of class rather than in terms of union interests, thanks to the political accommodation and representation offered by the Liberals. This helped to shape British socialism in a liberal democratic mood but also made it politically weak. But for the divisions of the Liberal Party in the World War I years and the plurality system, the British Labour Party – deprived of the crucial asset of leading a political battle for the liberalization and democratization of an autocratic regime–would probably have remained a party of medium size and importance. The lack of ideological rigidity in these northern European socialist movements was therefore due either to an early alliance experience with other popular liberal forces in opposition to the established elite or, more typically in the United Kingdom, to delegation of political representation to these forces until well into the twentieth century. This delegation of political representation helped the movement to concentrate on industrial and typical working-class issues without also having to focus on radicalizing and divisive debates about the goals and tactics of direct political action. The enormous differences between the United Kingdom and Scandinavia in industrial development did not prevent a lasting and profound Lib–Lab alliance in both countries, fostered by experiences of common participation in early popular movements. In the United Kingdom, the Liberals succeeded in integrating their own interests with those of workers’ associations until the end of the nineteenth century. In Scandinavia, although the socialist parties organized as an independent political force long before the Labour Party did in the United Kingdom, this alliance lasted even longer. However, the links of the socialists to early liberal forces were broken once the main political goals that held them together (liberalization, dissolution of the union with Sweden, democratization and universal suffrage, and defense of strong union interests) were achieved, and in all these cases this ﬁnal break led to the fragmentation and restructuring of the liberal forces there. THE NONCOOPERATION CASES: BELGIUM, THE NETHERLANDS, AND SWITZERLAND In contrast to the Scandinavian countries, Belgium, the Netherlands, and Switzerland provide the model of an early mismatch between social groups

CLEAVAGE STRUCTURES

423

and political movements. The liberalization of their regimes was early (as in the United Kingdom), but in none of these cases can one ﬁnd any signs of early coalition or delegated representation between the liberals and the socialist movement united against entrenched conservative privileges. Early mobilization soon made the liberals opposed to any forms of denominational mobilization. In Belgium and the Netherlands, the nineteenth century, and in particular the period between 1848 and 1917, was not a liberal period, but rather one in which the liberal–conservative conﬂict was dominated by state–church relations and conservatism had a rather traditionalist outlook as a result of the early denominational antiliberal mobilization. Liberals and Catholics had dominated the politics of Belgium since the founding of the nation, ﬁrst in a grand coalition and later with growing adversarial competition. The 1830 ‘‘revolution’’ against William I, the king of the Dutch–Belgian entity created by the Congress of Vienna, was fought by a pool of Catholics and the liberal opposition; the former were mainly motivated to defend Catholicism from Protestant rule; the latter, although sympathetic to the anticlerical stand of the dominant Dutch circles, were mainly motivated by antiliberal restrictions and by fear of the dominance of the Dutch language over an almost entirely French-speaking Belgian elite. This coalition lasted until 1846, to be followed by the 1846–1884 alternation of liberal and Catholic governments, with liberals generally in a dominant position. The liberal cabinets of this period and the battles on the subject of religion at school between 1878 and 1884 intensiﬁed the need for a structured Roman Catholic political party. From 1884 on, there followed thirty years of Catholic cabinets and Catholic supremacy. Consequently, party machines developed early: The liberal–Catholic split of the Unionist Alliance in 1857 had led both the liberals and the Catholics to start canvassing the support of the lower classes. The Catholics started to build a special party structure in the 1860s (as in Germany) and in 1884 formally created the party, although it had already been organized in a loose network since the 1860s. The Liberal Party also already had a precocious national structure by 1846.19 The Belgian party system was therefore already consolidated in the 1893–1899 period, much earlier than in all the other European countries.20 By that time, the dominant position of the local clerical notables, based on the census-restricted suffrage and the majority system, had ended and the 19 20

Lorwin (1966: 155). On the growth of the role of political parties from the time of the 1831 constitution to the electoral reform of 1893, see Witte (1980).

424

THE POLITICAL MOBILIZATION OF THE LEFT

overthrow of the ruling elite had been completed. It was the socialist partisan mobilization that copied, and improved, the organizational model already functioning among the liberals and the Catholics.21 The country was characterized by an early and particularly strong web of sociopolitical organizations and a corresponding politicization of the social structure.22 Understandably, trade unions, cooperatives, friendly societies, saving banks, and leisure associations acquired a distinctive political character. The competition between Catholics and socialists, and the challenge of social Catholicism to the allegiance of the working class, provided a further stimulus for the transformation of these agencies into elements of social segregation. The supremacy of the Catholics from 1884 on encouraged closer contact between the liberals and the non-Catholic labor movement, but the basis of this trend was mainly ideological anticlericalism. The language issue, on the other hand, was not important in the early development phase of mass politics. Around 1860, a Flemish-language movement developed, but this was soon monopolized by the Catholic movement and was used to bolster the development of the Christian Democratic Party. The ﬁrst election in which parties articulating linguistic demands achieved signiﬁcant support was in 1919 (with 3% of the vote going to a Flemish nationalist). By the late 1920s and the 1930s, the issue had gained more importance but was still only one of the programmatic concerns of traditional parties (administrative decentralization, cultural autonomy, etc.). Later on, nationalism slowly grew in strength in Flanders, and in the mid-1960s the emergence of Francophone parties led to the current situation of high saliency of linguistic issues.23 In conclusion, the ﬁrst wave of Belgian political mobilization was well advanced before labor entered the game. The Belgian working class faced a complex structure of options: It could found independent socialist political organizations, it could organize itself under the leadership of Catholic clericalism, or it could become part of the liberal movement (even before it could identify with an ethnic linguistic identity). The labor movement was anticlerical, which made for a difﬁcult relationship with traditionalist Catholics; it was antimonarchy and anticapitalist, which made for a difﬁcult relationship with the liberals; it was also alone against the two dominating movements of the nineteenth century in asking for full and complete electoral democratization. No period of close alliance with, or 21

22

23

See the review of the debate on early party system consolidation in Belgium in Noiret (1990a). Particularly well documented by Hill (1974: 40–41). See also De Bakker and Claey-Van Haegendoren (1973: 237–247). Hill (1974: 30).

CLEAVAGE STRUCTURES

425

delegated representation to, the liberals of the working-class movement can therefore be identiﬁed. A similar lack of consolidated and persistent alliance between early liberalism and socialism is observable in the Dutch and Swiss cases. These two countries’ political traditions and structures had several features in common in the nineteenth century. They had both achieved international sovereignty with only minor internal consolidation, and although this independent national existence was achieved by force, no dominant center ever developed. The early weakness of central authority in the Holy Roman Empire, the early development of mercantile cities dominating the surrounding countryside, the multicephalus nature of this network of cities, and the absence of large landowners all made for a nineteenth-century polity with no standing army, no strong national bureaucracy, and a weak national educational system. All this reinforced localism and entrenched pluralism. Yet the Netherlands had fewer barriers to social mobility (given its geography); particularism was broken more easily and earlier (it had a common language) and the national identity was therefore more integrated than in the case of the socioculturally heterogeneous and politically deﬁned national identity of the Swiss. Moreover, the Netherlands was unquestionably more centralized and more internationally involved than federal Switzerland.24 As a consequence, in the Netherlands, cleavage politicization was nationwide and was dominated, much as in Belgium, by an early liberal–clerical opposition, to which class was added only later. In Switzerland, truly national cleavage and party systems never developed. The Netherlands was characterized by the very early emergence of an orthodox Protestant movement in reaction to liberalism and religious modernism and in defense of traditional Protestant education. During the nineteenth century, a section of both the elite and the rank and ﬁle grew increasingly dissatisﬁed with the ‘‘latitudinarianism’’ of the Dutch Reformed Church. The leader of this faction, Abraham Kuyper, and his followers founded an Anti-School Law League (1872), the Calvinist newspaper De Standard (1872), and a working-class organization called ‘‘Patrimonium’’ (1976); made massive petitions to the king not to accept secularization bills in schools (1878); and formed the Anti-Revolutionary Party (1879), a separate Calvinist University (1880), and, ﬁnally, a new Calvinist denomination (the Reformed Church, 1886). Not all orthodox persons joined the Reformed Churches; many remained within the Dutch Reformed Church and later joined the Christian Historical Union Party (1908), which never became a strong mass party, unlike the more militant 24

For this comparison, see Daalder (1973).

426

THE POLITICAL MOBILIZATION OF THE LEFT

Anti-Revolutionary Party.25 The founding of the Anti-Revolutionary Party triggered liberal–clerical opposition, stimulated the birth of modern political parties, and determined the upsurge of Catholic political activism. Catholics were the second group to mobilize politically. By supporting the separation between the dominant Dutch Reformed Church and the state, they had by 1853 achieved the reestablishment of the church hierarchy (the ﬁrst archbishop in Utrecht) and had started to rebuild their church. Due to long-term discrimination, Catholic organization and mobilization took place largely under the guidance of the church rather than as an explicitly political organization. Thus, the Catholic Party was formed only at the end of the century and remained a weak actor vis-a`-vis the reestablished Catholic hierarchy. However, in many ways, the Catholic community had already been reached and densely organized since the middle of the century by the reestablished church organization. By the 1880s, Catholic–Protestant relationships had improved so much that in 1888 a Calvinist–Catholic government (1888–1891) started the ﬁrst government subsidies to religious elementary schools.26 In contrast, the socialists and the liberals were the last to mobilize. The liberals, used to being dominant in nineteenth-century politics without any need to resort to much formal organization, developed their party organization late and were never very strong from an organizational point of view. The socialists, too, formed mass organizations late in comparison with other countries. In Switzerland, in contrast to the Netherlands and Belgium, the localregional element predominted even after 1848 and strongly affected the degree of politicization of the various cleavages. While in Switzerland no dominant national church existed, and each canton developed a speciﬁc and highly territorialized Catholic or Protestant proﬁle, in the Netherlands, national uniﬁcation and centralization after 1895 brought about the equality of religious groups. Catholic minorities in the non-Catholic parts of the country developed nationwide organizations and solidarity among Dutch Catholics, even in Brabant and Limburg. Consequently, the resulting regional cleavages were successfully subordinated to the religious one. This produced a nationalization of politics: It split mixed religious communities but established strong organizational links among like-minded believers across the country. Strong Calvinist and Catholic organizations meant the segmentation of the nation, with the secular groups as a third section within which class cleavage could develop, but with a much more restricted audience. In Switzerland, regional factors were never subordi25 26

Daalder (1990: 41). Ibid., 40, note 22.

CLEAVAGE STRUCTURES

427

nated to cross-national political divisions precisely because religious divisions were essentially regional. Class, religion, language, and territory intersected more profoundly, and none acquired a national standardizing role.27 Very often, the potential for the politicization of any one cleavage line was hampered and minimized by rival claims along other potential lines of division. If accommodation characterized the two countries, this was spread among many arenas and dimensions in Switzerland, while in the Netherlands religion was dominant but class came second and the two subordinated all the others.28 Swiss liberalism had two branches. In Protestant cantons, the Liberal Democrats became the majority party and their liberalism was strongly imbued with religious values (as in Scandinavia in the nineteenth century). In Catholic cantons, it instead combined democratic nineteenth-century aspirations with anticlericalism; the nation-building Radical Democratic Party was thus nationalist, populist, and anticlerical and strongly inﬂuenced by Mazzini’s thinking.29 The consequence of this structuring of the party system was a lack of uniformity in the party presence across the country. The radicals and the socialists tended to be present in all cantons; the Christian Democrats were virtually nonexistent in Neuchatel and were very weak in Bern and Vaud; and the liberals were very popular in Neuchatel, Vaud, and Geneva, less so in Bale, and unrepresented in other places. In consequence, a pattern was discernible whereby a multipartitism based on the equidistribution of political forces resulted in the highly urbanized and developed cantons (Bale Ville, Bern, Zurich, Geneva, Vaud, Argovie, Thurgovie); a party system a` l’italienne, with a strong Christian Democratic Party dominating multipartitism, was dominant in Lucerne, St. Gall, Tessin, Soleure, Grisons, and Fribourg; and, ﬁnally, a party system a` la belgique, with three major parties (Catholics, socialists, and radicals) characterized the political situation in the Valais and Zuog cantons. The resulting fragmentation was such that it actually reduced the importance of parties in the federal political process,30 so that this, even if apparently 27 28

29

30

See Kerr (1987). According to Kriesi (1990), federalism and pillarization are two different mechanisms for integration of subcultures into a larger national community. See also the contrast between territorial and sectional integration of Switzerland and the ‘‘lager,’’ i.e., vertical integration of Austria underlined by Lehmbruch (1966). It is interesting to recall that, thanks to the early inﬂuence of a young left-wing Hegelian, Henry Durney, the party adhered to the First International. Nowadays, the Radicals are the most important antisocialist party in the German-speaking part of the country: the parti de l’e´tat cantonale in Romandie and the anticlerical left in the Valais. Swiss liberalism also suffered several splits. Parties have little voting discipline and little inﬂuence over their MPs. Cantonal execu-

428

THE POLITICAL MOBILIZATION OF THE LEFT

related to party politics, retained a low ‘‘partisanship.’’ Thus, it can be conlcuded that in Switzerland, federalism (localism) has undercut party loyalties by giving much more weight to territorial loyalties, which, in turn, were based on and reinforced by religious and linguistic identities.31 In these three cases – Belgium, the Netherlands, and Switzerland – the socialist movement entered a party system in which important waves of early mobilization had already taken place along preindustrial cleavages of a religious, territorial, and linguistic nature. The socialists’ potential electoral market was, therefore, to a large extent already restricted by the competitors’ efforts. The basis of the political identities created by this early mobilization was, of course, cultural, and the alliance opportunities were largely shaped by the coalitions and conﬂicts that the previously mobilized social and political groups had already established. In addition, the early liberalization of the regime made a Lib–Lab anticonservative alliance unnecessary. In none of these cases, although for different reasons, was a lasting and solid alliance between liberalism and early socialism established.

‘‘MALTHUSIAN LIBERALISM’’: FRANCE AND ITALY I have grouped France and Italy as cases in which the party system completely failed to structure itself along partisan lines before World War I, and I have chosen the label of ‘‘Malthusian liberalism’’ because the early predominating liberal forces did not engage in party building or expansive competitive drives. Instead, they pursued a defensive and restrictive competition strategy as a result of the continuing menace for them and the regime they were founding that came from forces refusing to accept its legitimacy. The nature of this kind of early liberalism assimilates the two countries despite the crucial difference in the timing of universal manhood suffrage. Nineteenth-century France had a long tradition of revolutionary outbreaks, mainly of an urban nature, whose level of success was determined by the support they obtained in the countryside. However, there

31

tives are generally directly elected by citizens by double ballot majority voting, which weakens the direct role of the parties in the recruitment of executive personnel. Finally, the tradition of widespread direct democracy was not controlled by parties at the beginning and has become less and less so over time. On this latter point, see Libbey (1970). Kriesi (1990); Scholten (1980) argues that pillarization and federalism are imposed from above and constitute conscious actions by the elite to defend themselves against socialism.

CLEAVAGE STRUCTURES

429

was little growth or consolidation of this early political opposition movement, which could be deﬁned as popular, in the sense of incorporating sections of the urban and rural lower and middle classes, and selfrepresentational, in the sense of relying on self-recruitment within these groups. Italian nineteenth-century politics was characterized by narrow intellectual movements and localized popular revolts, both of which were unconnected and unlinked to broader national mobilization processes. The various ‘‘political sects’’ and the Mazzini movements of the ﬁrst half of the century, the neo-Guelphism of Gioberti, the ‘‘moderatism’’ of Cesare Balbo, the federalist republicanism of Cattaneo, and the democratic radicalism of Pisacane and Ferrari were all, without exception, intellectual and restricted political movements.32 Important popular lower- or middle-class movements, capable of fusing different social groups, both elites and rank and ﬁle, into a large, mobilizing enterprise were completely absent. The distance of the intellectual elite from the masses could not have been greater than in Italy in this period. In France as well as in Italy, there were, at the beginning of the period of competitive politics, two dominant dimensions of conﬂict: On the one hand, the liberal and republican forces were opposed to the confessional and religious-based forces; on the other, conﬂicts of legitimacy concerned the established constitutional order and opposed the same liberal forces to the reactionary and ‘‘legitimist’’ aspirations of monarchical circles in France and to the staunch refusal of the Catholic hierarchy to accept the new state in Italy. The French Radical Party, the strongest defender of the regime and the political values of the Third Republic, had a social base of notables in the countryside33 who mixed social, economic, and political power so that they were indistinguishable, occupied positions of local and national authority, and obtained and maintained their power thanks to a network of local and personal relationships; that is, this was a provincial, mainly professional, bourgeoisie that owned a small amount of land, was economically autonomous, relatively well-off, and very close to the small landowners and the farmers.34 As such, these local notables, unless challenged by functionally based and cross-sectional political movements, were neither vote maximizers nor mobilizers. These forces never consolidated politically along organizational partisan lines and neither did any other force. French conservative and monarchical circles held that political parties were ruinous 32 33

34

On these early political movements in Italy, see Morandi (1978: 10–19). For the classic description of how early manhood suffrage helped the local notables to control rural votes, see Siegfried (1913). These characteristics emerge in the beautiful portrait of the French local notables by Hale´vy (1995).

430

THE POLITICAL MOBILIZATION OF THE LEFT

to the country, disliked the word ‘‘party’’ from the outset, and preferred to call their political organizations ‘‘movements,’’ ‘‘groups,’’ ‘‘rassemblements,’’ ‘‘unions,’’ and so on. The criticism of parties remained one of the key points of conservative platforms and spread well beyond reactionary and conservative circles: parties paralyzed reform and encouraged agitation, sterile quarrels, the feverish unleashing of personal ambitions and ideological passions, and artiﬁcial divisions that prevented them from reaching predeﬁned national goals and ‘‘missions.’’35 Many explanations have been advanced for these persistent antiparty sentiments in the French political and intellectual circles and for the nonconsolidation of partisan politics in France before World War I.36 In my view, the most important factor is the separation of the historical processes that give rise to mass political identities and parties, industrialization, urbanization and enfranchisement. The early introduction of enfranchisement long before the processes of industrialization and urbanization meant that the municipalities, the local notables, and the strong state agencies at local levels all maintained an important role in channeling representation. The fact that the French masses, particularly the rural masses, were politicized early in a phase of preparty politics still dominated by localistic and notable networks had strong negative consequences for the ensuing rootness of the party system. New political movements with a functional and cross-sectional proﬁle such as class movements had to ﬁght these early channels and never managed to replace them completely as the main focus of political identity of the people. Industrialization had started early in France but then progressed very slowly and without the major dislocation effects that rapid industrialization produced in other countries. What is more important, the major pushes of the three processes were widely separated in time. 35

36

Still, in the 1950s, Goguel complained that parties were exaggerating the role of members and militants at the expense of the voters and depriving them of the possibility to choose their representative directly; Goguel (1950: 352). For a recollection of these antiparty criticisms, see Zeldin (1979: 25–26). Individualistic and psychocultural explanations underlining the unwillingness of the French mind to compromise and associate different shades of thought: Buell (1920: 1) and Lowell (1896: 100–108); the antiassociation legal tradition going back to the Chapellier law, combined with a renascent Jacobin-republicanism and the use of integrated territorial administration to discourage the rise of party organizations: Tarrow (1977: 54–55); the emphasis on individual personalities making the system to remain in a charismatic mood and thereby delaying the transition to bureaucratized politics through parties: Hoffman (1974: 63–110); the territorial centralization in Paris of everything of importance and the ‘‘parisiﬁcation’’ of political movements, which rendered the territorial organization of parties unimportant and offered few incentives for the building of national structures: Sva˚sand (1978: 10).

CLEAVAGE STRUCTURES

431

Even after the 1870s, the high level of personalization of local politics, the fragmentation of the political groups, the existence of parliamentary groups with no corresponding organizational force in the country, the wide freedom of parliamentarians to sustain different governmental coalitions, and the very late recognition of the legitimacy of parliamentary groups in the chambers are probably not enough to assimilate French Third Republic politics to the Italian trasformismo of the same period (see later) but instead make the French situation similar to it. Parties started to organize only well after the beginning of the new century, and it was only before World War I that party activities outside parliament began to have some effect on the government and parliamentary life (e.g., party congresses began to have inﬂuence the positions taken by the parliamentary group). In 1936, the number of MPs belonging to parties was still a minority in the chamber, not to mention in the Senate.37 Only at the beginning of the Fourth Republic did parties become key actors in the negotiations for the formation of government and in the deﬁnition of the internal rules of the chambers, the parliamentary committees, and so on. In conclusion, the weak development of French political parties and their failure to establish roots in local and provincial French political life throughout the entire Third Republic are not only the result of the legacy of the Second Empire, with its style of personalistic and populist government and its heavy reliance on state structures and authorities to channel interest representation; equally important was the permanent unwillingness/incapacity of the conservative (and reactionary) right to organize itself into partisan political forms and the aversion of the radical local notables to engagement and compromise in open political and partisan activity. The structuring of the party system presented a similar picture in Italy and ended by offering a similar constellation to the incipient socialist movement, although for different reasons. An overview of Italian political history between the time of national uniﬁcation and World War I shows that once the early mobilizing issues of nationalism and national uniﬁcation were soft-pedaled, the country’s political landscape soon became very varied, with a large variety of political organizations and no stabilization or consolidation along important partisan division lines. Even after 1870, the mobilization level of the country remained extremely low; it is not possible to identify political movements with a popular base of a Catholic, peasant, or bourgeois/middle-class nature. The 1860s had already shown the inconsistency of the much desired bipartitism. A single political movement and organization that could claim to be the heir of the Cavour 37

On the development of parliamentary groups, see Goguel (1950: 348–351).

432

THE POLITICAL MOBILIZATION OF THE LEFT

policies did not exist, and programs and groups instead formed around individual leaders (e.g., Ricasoli, Minghetti, Rattazzi). Parliamentary coalitions were formed and dissolved around interests that were very often of a regional character. In other cases, they were formed around purely personal and clientelistic relationships and promises of political positions. Although a few attempts were made to give a new political meaning to the two wings of the chambers, the left and the right, and also to create new forces around a clear program (for instance, this was attempted by Sella with his ‘‘young left’’), the process of dissolution of the two old preunitary parties continued and the early differences in opinion as to how to solve the problem of national uniﬁcation faded in the following decades and were dispelled altogether in the era of trasformismo.38 Trasformismo was a parliamentary praxis that tended to guarantee the cabinet a considerable majority in parliament, either through preliminary negotiation with its most prominent opponents and with their eventual incorporation into the governmental majority or through the occasional exchange of favors with less inﬂuential MPs of the centrist marais.39 It was viewed as accepted practice by the entire political class, with the exception of the republican extreme left, and it resulted in limited interelite competition within the broad liberal camp. To sum up, a narrow, closed, and socially and politically homogeneous political class lacked internal cleavages over which to aggregate politico-parliamentary formations that could be less ephemeral than these short-lived personal coalitions.40 As things stood, the only nonpersonal political coalitions could be of a regional nature (preunitary regions). However, even in this case, these could not go too far because the political class as a whole, while sharing the ideology of the unitary state, was also so aware of the fragility of this state that it was reluctant to risk organizing political conﬂict along territorial lines. Therefore, the only possibility of organized political conﬂict and competition was along institutional lines (republicanism versus monarchy) and, later, socioeconomic issues (anarchism and socialism). It has been argued that this parliamentary practice, essential in preventing the consolidation of clear partisan alternatives in the liberal era, represented a kind of contract between the bourgeoisie of northern Italy and the dominant social groups of the south. The basis of the compromise was a politics of reform and a liberalization of the new unitary state that respected the interests of the dominant Southern social groups. The latter, 38 39 40

See Morandi (1978: 25–33). Procacci (1970: 407). On this point, see Caracciolo (1957: 112–113).

CLEAVAGE STRUCTURES

433

as a form of additional guarantee, came to be adequately represented within the government.41 It is hard to decide whether this was the cause or the consequence of trasformismo. The crucial point, in my opinion, is that the political positions adopted by the liberal nation-builders rested on extremely ﬂimsy social bases: The papal non expedit had prohibited Catholic intellectuals, organizations, and masses from taking an active part in the new state; the extreme left remained republican and later anarchist and socialist; and the monarchy had not fully abandoned its atempts to intervene directly and needed to be checked continuously.42 The liberal groups responded to Catholic challenges to their legitimacy with prolonged restriction of the suffrage in order to keep both the Catholic masses and later the socialist masses out of the electoral competition. They responded to the ambivalent monarchical attitude toward parliamentary democracy with a political coalescent behavior whereby certain key issues were silenced. Even the institutional cleavage between republicans and monarchists was somehow muted, as the republicans did not pursue the institutional question vigorously after the non expedit of Pius IX. This, however, allowed Catholics and socialists to deﬁne themselves as the pays re´el as against the pays legal.43 The indifference or hostility of all the excluded groups toward the institutions of the new state – in a context in which the means of expressive uniﬁcation of the society and of consensus building around the modernizing elite were weak due to the high level of illiteracy – made it impossible for the liberals to use the extension of the voting right in a defensive way and brought about their political Malthusianism. The lack of a Catholic political role in support of the conservatives and, at the same time, the lack of an autonomous political organization of the Catholics (as in Belgium, Austria, and the Netherlands) deprived conservatism of the potential support of the Catholic traditionalist masses. For similar reasons, liberalism was deprived of the potential support of early working-class organizations and of the radical middle classes. Both coalesced toward a hostile civic society. In the pre–World War I period, an Italian radical liberal party similar to the French, Swiss, or Danish variants never developed.44 The early 41 42

43 44

Procacci (1970: 408–409). In many cases, the crown deliberately avoided designating as premier a political ﬁgure who appeared capable of securing a homogeneous, minimum majority. See Rotelli (1975: 8–9) and Grisalberti (1978). Galli (1968: 33–35). For the relationships between Italian radicals and the early socialist movement, see the opinion of Galante Garrone (1973: 197–198). Socialist intellectuals published in radical journals, but few true political connections existed and the relationship remained largely intellectual. This was due fundamentally to the narrowness of Italian radical liberalism,

434

THE POLITICAL MOBILIZATION OF THE LEFT

Socialist Party instead manifested the hope that an alliance with the democratically advanced sectors of the bourgeoisie would be able to break the protectionism and the conservative bloc formed by Southern landowners and Northern steel producers. However, the absence of strong social groups of small farmers or peasant landowners (as there were in France, for example) made it impossible for the radical liberals to obtain support in the countryside and in provincial Italy. The failure of the small liberal bourgeoisie meant that the democratization process depended essentially on the socialist movement. In these circumstances, the liberal and conservative forces, which were not organized as parties, managed to resist thanks only to the restricted suffrage and, after 1913, thanks only to the possibility of obtaining Catholic votes channeled by the secular organization of the Catholic world, which was by then worried about the growing strength of the socialist movement.45 Subsequently, after World War I, the liberal forces hoped to borrow the support of the growing nationalist movement and, ﬁnally, that of the new fascist movement that emerged from it. The latter attempt proved, of course, to be fatal for the Italian liberal regime. So, in both Italy and France, it can be concluded that parties were organized for parliamentary rather than electoral purposes. Selfrepresentational social movements existed outside parliament, but with few and tenuous linkages to the parliamentary parties. In both systems, religious and institutional issues fragmented representation without achieving expression along clear partisan lines or becoming institutionalized into party formations and alignments. The major difference between Italy and France was that in Italy the delayed extension of the suffrage gave the new mass parties (the socialists, the Catholics, and later the communists) a chance to absorb the masses of new voters into party subcultures in the absence of previous experiences of institutional and nonpartisan representation. In both countries, the ﬁrst elements of socialism emerged as branches of republicanism, and the ﬁrst socialist candidates’ victories were made possible only by alliances with radical liberals. Despite this, a stable and persistent alliance between socialism and radical liberalism never devel-

45

to its cultural more than its political character, and to its almost nonexistent organizational bases. The clerical–moderates alliances made before the 1913 elections are known as the ‘‘Gentiloni Pact’’ after the name of the president of the Unione Elettorale Cattolica. In 1904 and 1909, such alliances resulted from local agreements between liberal candidates and Catholic organizations. The situation in 1913 was new, as the papal non expedit was suspended in 330 districts and maintained only in 178. See Carocci (1975: 198).

CLEAVAGE STRUCTURES

435

oped. The main reason for this is, perhaps, that in this historical phase neither of the two movements consolidated as strong political organizations capable of strategic choices of alliance. This situation is obvious in Italy; while in France the case is less clear-cut, and there have been different interpretations. Luebbert, for example, has argued that France assimilated to Switzerland and the United Kingdom as a case of ‘‘hegemonic role of liberalism’’ and that in France too, there was close Lib–Lab cooperation.46 Much of the difference between his and my interpretation derives from his deﬁnition of ‘‘socialists.’’ While it is true that ‘‘socialists’’ provided essential support for the ministries of Bourgeois and Waldeck-Rousseau and, even more important, for Combes between 1902 and 1905, and while it is true that close cooperation in the defense of republican values was frequent (as, for example, in the Dreyfus affair),47 the fact is that the rather elitist and personalistic parliamentary socialists of Jaure`s represented very little beyond themselves and the local electoral alliances that had brought them to parliament. The formation of the SFIO in 1905 was based on a Guesdesist and Blanquist ideological predominance in opposition to the radical governments. A weak trade union movement was represented politically by a radical Socialist Party in which ideologues predominated over organizational union movements, in contrast to the United Kingdom, where intellectual socialists were defeated and the Labour Party became the mouthpiece of the strong organizational base of the trade unions, with a platform of ministerialism and political integration. If Millerand and his career illustrates the radical foundation on which socialism was built, it does so from the point of view of parliamentary representation only.48 The early radical–socialist cooperation to defend the liberal and republican order from the menace of clerical, monarchical, and reactionary forces was followed by a sudden radical break between radicalism and socialism at the beginning of the century, when this threat started to evaporate. This profound break between radicalism and the working class is represented in the policy of Clemenceau and the Radical Party in 1905–1908, in its campaign against collectivism, its call for the dissolution of the CGT, and its denunciation of the socialists as antipatriotic. In conclusion, political liberalism was a weak, ideologically fragmented, and politically unorganized point of reference for early socialism in both France and Italy. In contrast to Switzerland and the Netherlands, 46 47

48

Luebbert (1991: 27–48) argues this point extensively. For an account of the socialists’ positions on the Dreyfus affair, see Noland (1970: 61– 85). Zeldin (1979: 408).

436

THE POLITICAL MOBILIZATION OF THE LEFT

these two countries lacked a consolidated liberal regime, and their defense against antiliberal movements continued well into the 1920s and until the end of World War II. At the same time, given the unstructured party system and the weak institutionalization of partisan alignments, the socialist movement was the ﬁrst to present modern mass party features and mobilization techniques. As a result, the socialist movement took over a large part of the radical-liberal program and made it part of its complex and composite ideology. These socialist movements remained organizationally weak, and their relationships with the unions were tense, with deep organizational and ideological splits. In this sense, their weakness was simply a reﬂection of the difﬁcult process of party-system structuring. However, at the same time, this made it easier for them to inherit electorally those radical middle classes, both rural and urban, whose anticlericalism, republicanism, and democratic values were only weakly represented by the organized political forces. In this sense, Malthusian liberalism eventually contributed to the large electoral mobilization of the left, as it did to its composite ideology, organizational divisions, and social heterogeneity. NONEXISTENT LIBERAL OPTION Liberalism as an organized political expression of the urban and/or rural middle class and bourgeoisie was nonexistent in Germany and Austria. In Finland and Ireland also, a Lib–Lab alliance option proved to be unavailable in the early phases of national electoral mobilization. This common outcome derives from different factors and contexts, but in all cases there is a common characteristic: the overwhelming role played by early national and nationalistic issues. Austria is probably the most clear-cut case of the lack of a liberal option for the labor movement. In the Hapsburg Empire, with the monarchy relying on groups from the German, Hungarian, and Polish nobilities, the high Catholic clergy, and the high bureaucracy, the main opposition was formed by liberal groups. These groups, however, were particularly intent on promoting German hegemony, which the monarchy, instead, wanted to limit to a certain extent. After 1879, a year marked by growing internal national tensions, liberalism started to disintegrate organizationally and survived only as an attitude rather than as a movement.49 Before World War I, obstructive parliamentary practices and nationalgroup divisions and struggles prevented the establishment of any orderly 49

Engelmann (1966: 260).

CLEAVAGE STRUCTURES

437

or clear government opposition dynamics and fostered (and justiﬁed to a certain extent) the practice of ruling by decree of the king and the bureaucracy. However, by the end of the 1880s, both the Social Democratic Party and the Christian Social Party had been formed, and these, thanks to universal and equal suffrage for men, became the two leading parties by 1907. Austria is the most illuminating example of the difﬁculty of democratizing within nonlegitimate state boundaries. For disaffected national minorities, the aspirations to liberalization and democratization were only partially and instrumentally relevant and were combined with, but more often subordinated to, demands for cultural and political autonomy or total independence. Thus, the socialist movement was both politically isolated and institutionally unintegrated (see Chapter 7). The socialists were given little support by the liberals in their struggle for democratic reforms. This explains why they were never really concerned with the issue of whether or not to cooperate with bourgeois forces. The clearest example of the impossibility of a Lib–Lab alliance was the 1893 electoral-reform bill prepared by the government. This was a radical reform from above that introduced universal male suffrage, even though, in practice, it included barriers for the lower classes and privileges for the upper classes. Details of the bill are not important here; what counts is that the liberal and bourgeois groups (the so-called left of the Imperial Council) opposed and ﬁnally defeated the electoral reform proposal of Taaffe and his government. This was probably due to the fact that their main concern was to keep the national equilibrium that granted German hegemony. If liberalism and liberal political groups in general remained very weak and offered no option to workers interested in political emancipation and social advancement, no other stable political alliances were available either. For years, the largest populist party was the rather reactionary and openly anti-Semitic Christian Social Party, which gained little support among workers. This resulted in a rather unique European case: a type of isolation in which the Socialist Party had no competitors for the support of the working class on either its left or its right. While this was clearly the cause of its political ineffectiveness and isolation, it also provided the structural context in which the development of orthodox Marxism could be maintained,50 with maximum advantages in internal cohesion and minimum external costs. As a result of its involvement in the national question and in the issue of political rights, the Socialist Party was not particularly affected by the 50

On this orthodoxy and its sources, see Steenson (1991: 269).

438

THE POLITICAL MOBILIZATION OF THE LEFT

ideological debate between reformists and radicals. This situation of political isolation did not change substantially after the war. In the 1920s, the two traditional Austrian anti-Marxist groups joined forces to a considerable extent against the socialists, fusing their tickets. Only at the beginning of the 1930s did polarization occur in the non-Marxist camp to produce Austria’s three irreconcilable ‘‘lagers’’: the socialists, the proclerical conservatives, and the German nationalists. The issue was, and remained, the possibility of an elite cartel among parties and closely linked organizations that would represent densely organized political subcultures as well as distinct social groups. In Germany, too, the socialist organizations found it difﬁcult to establish any stable or privileged links with the liberal, or parts of the liberal, forces before World War I. The early general suffrage led to a correspondingly early national political mobilization of the masses. The Catholic population was initially mobilized in opposition to the Kulturkampf by the Center Party.51 A second wave of mobilization was led by the socialists against the antisocialist laws. Finally, political mobilization was enhanced by the government’s attempt to transform elections into a plebiscite for its policy and also by the strong development of interest organizations. With an 85% average turnout in the last elections to the Imperial Diet (1912),52 a strong network of powerful interest organizations, a highly developed central bureaucracy, and a high degree of state intervention in its socioeconomic affairs, Germany was a highly politically mobilized polity before World War I. However, despite this, there was no national party; with the single exception of the Social Democrats, all the others, including the Zentrum and the two conservative parties, were strongly localized and unable to place candidates in every constituency. The pattern of national uniﬁcation under Prussian leadership meant that, to a certain extent, regional loyalties were transformed into party divisions, and regional fragmentation was high from 1870 to 1933. Ethnic minorities and a few important regional parties emerged, among which the Bavarian Farmers’ League, the Wu¨rttemberg Farmers’ and Wine-Growers’ League, and the Guelph Party in the Prussian provinces of Hannover are worth mentioning as the more resilient examples. Again, a wave of regional splinter parties emerged in the 1920s in several states, including Schleswig-Holstein, Braunschweig, Baden, 51

52

Ritter (1990: 53–55) notes that in 1870 voter participation in Catholic majority constituencies was higher than in Protestant ones. Ritter (1985: 25) provides ﬁgures and comments on the turnout of the 1871–1912 elections.

CLEAVAGE STRUCTURES

439

Wu¨rttemberg, Hessen, Thuringia, and Bavaria. These parties were generally prominent in the rural areas. More generally, however, in this early phase, almost all the parties displayed strong regional bases, in the sense of the skewed distribution of their support. Catholics were conﬁned to areas of Catholic concentration, while the conservatives and the DNVP were concentrated in the eastern provinces. Even the Social Democrats and Communists were constrained regionally by the uneven distribution of industrialization, urbanization, and Catholic inﬂuence. Moreover, the constitutional structure of the Reich encouraged the persistence of regional differences, as the state elections were more important than the national ones. In this context of general and protracted regional fragmentation,53 it was difﬁcult for a national party to devise strategies of alliance. Together with the delayed nationalization of political organizations and alignments, the lack of Lib–Lab political cooperation in Germany has been widely discussed in connection with the thesis of the ‘‘weakness of the German bourgeoisie’’; this discussion provides an opening for the examination of possible links between socialists and liberals. The early abdication of German liberalism on both the socioeconomic and political reform fronts54 made it unappealing to working-class associations and left the socialists with a relatively easy and early monopoly over electoral and parliamentary working-class representation. This political abdication has traditionally been interpreted as a German peculiarity (deutscher Sonderweg). In the assimilation of premodern sociopolitical models by the bourgeoisie, a historiographical tendency has pinpointed its political weakness, the feature explaining the entire historical development of Germany up to the rise of Nazism. The thesis of weak liberalism resulting from an economically weak bourgeoisie with upper-class values and aspirations has also been advanced in regard to Austria and Italy. However, the former – in contrast to the latter – underwent this nineteenth-century bourgeois development, in the sense of its material, cultural, and regulative progression, if not its political liberalization, as a ‘‘silent revolution.’’ In the works of Blackbourn and Eley55 and particularly that of Kocka,56 the thesis of the economic weakness of the German bourgeoisie, and of its ‘‘feudalization’’ 53

54

55 56

See Ritter (1990: 57–58) and Rokkan, Urwin, Aarebrot, Malaba, and Sande (1987: 129– 135). For a more detailed analysis of these regional variations, see the essays in Rohe (1990a), and in particular Niehuss (1990: 83–106), Rohe (1990b: 107–144), and Mintzel (1990: 145–178). Breuilly (1985: 3–42). Gagel (1958) provides several examples and proofs of the continuous ambivalent position of German liberals on the political-reform front. Blackbourn and Eley (1984). Kocka (1989a: 51–55).

440

THE POLITICAL MOBILIZATION OF THE LEFT

and deep assimilation to the nobility, is criticized and rejected, and the differences from the bourgeoisie of other countries such as France and the United Kingdom are reduced. The essays by Tilly and Koeble also both conclude that the weakness was political rather than based on socioeconomic factors, resulting from the weak solidarity among different sectors of the bourgeoisie (i.e., on emerging entrepreneurship versus more traditional sectors that were extraneous and/or inimical to modernization) and by the state intervention that limited the cultural autonomy of the bourgeoisie.57 These critiques of the Deutscher Sonderweg are convincing, although the interpretation of the political factors in the weakness of the German bourgeoisie still needs to be speciﬁed. The problem can be reformulated by emphasizing, on the one hand, the role of the strong state and, on the other, the timing of the mobilization sequence that opened up the phase of political autonomy of the national bourgeoisie. As for the role of the state, it can be argued that in strong, resourceful states with autocratic, dynastic traditions of rule, the controllers of the state machine have ample opportunities through state action and legislation to break the unity of social groups, to gain support from speciﬁc sociopolitical groups, and to orchestrate conﬂict between one sociopolitical group and its adversaries.58 This was a typical political tool used by dynastic powers and circles in both Germany and Austria. In Germany, the workers’ support for the empire was sought in exchange for social policies not supported by the employers. In the context of the Kulturkampf, the aspeciﬁc support (negative integration) of liberal forces was instead sought through conﬂict with Catholicism. Before World War I, Social Democratic support and negative integration was attempted by forcing them to support the military commitments of the empire. Rather than view the socioeconomic weakness of a group as resulting in the autonomy of the state, the autonomy of the state can therefore be seen as a resource for curbing and breaking the potential political autonomy of social groups. In addition, there is a problem of timing among different social developments, and this can foster or reduce the political autonomy of the national bourgeoisie. The identity and autonomous political action of such social groups results from (1) a process of differentiation from aristocratic, dynastic, military, bureaucratic, and landed interests, which requires (1a) the maturation of a sense of diversity in interest and style of life with respect to them and (1b) a growing critique of and detachment from the ancient traditional centers of power and social prestige. At the same time, 57 58

Tilly (1989) and Koeble (1989). On the negative integration strategy of the ruling circles, see Lepsius (1973: 72–74).

CLEAVAGE STRUCTURES

441

this historical process interacts with another one that renders the bourgeois and liberal forces (2) antagonistic to the emerging working-class unions and political movements, which also pass through their own (1a) and (1b) phases. From the beginning of the nineteenth century to the beginning of the twentieth century – with differences in timing and tempo – all the national bourgeoisies passed from the predominance of process (1) to the predominance of process (2), accompanied by the progressive fading of the antagonism implicit in process (1). The timing between these two processes determines the window of opportunity for the autonomous political action of bourgeois liberal forces. The longer the temporal gap between the differentiation of the bourgeois liberal forces from dominant aristocratic and dynastic circles and the differentiation of the political movements of the lower classes from the liberal political movement, the more self-conscious and politically autonomous the bourgeois groups and the more profound and protracted the political role of liberalism are likely to be. If such a temporal gap is short or nonexistent, and if the continuing strength of the dominant circles of power overlaps with the early autonomous political mobilization of the lower classes, political liberalism – independent of the socioeconomic strength of the bourgeoisie and middle classes – will be short-lived. In other words, the window of opportunity for liberalism is opened up either by the early curbing of the dynastic–bureaucratic–military–landed aristocracy cluster or by a delay in the effective political mobilization of the socialists. Early socialist mobilization may make Lib–Lab alliances difﬁcult, antagonizing or demobilizing bourgeois liberal forces and eventually making the processes of liberalization and democratization harder. As a result, a delay in the political mobilization of the working class may well be an advantage for the liberalization of the regime and for an extended and autonomous role of liberal forces. In Germany (and also Austria), the protracted power of traditional social and political groups linked to the dynastic power and the early political organization of a socialist challenge greatly reduced the chances of establishing liberalism. Socialism was thus left on its own to represent a potentially large constituency, but it faced the institutional closure of the regime in a situation of political isolation. NATIONALISM OVERSHADOWING LIBERALISM: THE SPECIAL CASES OF IRELAND AND FINLAND The mobilization of the left was also affected by nationhood itself and by the timing of nation building. In some countries, left mobilization took

442

THE POLITICAL MOBILIZATION OF THE LEFT

place within the long tradition of national identity and state building (the United Kingdom, Denmark, Sweden, France). In others, the issue of the national building of cultural, political, and other identities faded just before the decisive issue of electoral mobilization was initiated (the Netherlands, Switzerland, Belgium, Norway). In Italy and Germany, the two questions overlapped for a long time after uniﬁcation and nationalism, and national issues continued to play an important role during socialist electoral mobilization. Formally, Austria, Finland, and Ireland achieved independence in the twentieth century, when the socialist mobilization was already well advanced. For Austria, this ﬁnal change was a redutio from a previously nationally heterogeneous empire, while for Ireland and Finland it was a process of political emancipation from a foreign colonial power with high international status. The Irish and Finnish cases are particularly interesting because from similar geopolitical conditions and a similar predominance of the national independence issue have emerged two completely different party systems and polar types of the left: an early, strong type versus a late, weak one and a persistent ideological trade union orientation versus an orthodox Marxist and a communist orientation. Finland and Ireland both show an absence of liberal forces in the early political mobilization of the masses and a continuous overshadowing of these forces by nationalistic ones in later phases. In both countries, the dominant problem was the relationship with a strong foreign rule exercised by a neighboring imperial power, and – again in both countries – the process of national independence almost ended in civil war. The Irish–Finnish comparison is therefore interesting in view of the relationship that was ﬁnally established between the social issue and the class cleavage, on the one hand, and the national issue, on the other. In Finland, two basic cleavages that were strongly linked to the national question worked as mobilizing forces of the Old Diet (pre-1907) politics:59 (1) the Finnish nationalist movement against the dominant status of the Swedish language and (2) the russiﬁcation policy and the attitude toward the tsar and Russia. As for the ﬁrst issue, a one-seventh minority of the population spoke Swedish, and this language stratiﬁcation mirrored the status stratiﬁcation. Swedish was the ofﬁcial language of the afﬂuent, educated upper class, with the exception of the clergy, and gave privileged access to public ofﬁces. Early mobilization led by Finnish middle-class nationalists often took the form of nationwide collections of funds 59

On the early Finnish political conﬂicts, see Mylly (1984).

CLEAVAGE STRUCTURES

443

and cultural activities favoring the establishment of Finnish-language grammar schools. This was immediately counterbalanced by the formation of a Swedish-speaking political party to defend the privileges of the minority. Thus, the ﬁrst major political cleavage saw Finnish nationalists, predominant in the clerical and peasant estates, opposing Swedish nationalists, predominant in the nobility and burgher estates. The second major dividing line saw ‘‘constitutionalists’’ in opposition to ‘‘compliers’’ (that is, those willing to yield to the tsar on nonvital Finnish interests and issues). Progressively, by the end of the nineteenth century and as a result of repeated attempts at russiﬁcation, the issue of relations with Russia became more important than the linguistic issue and the new focus on political organization.60 Internal language conﬂicts were suspended, and a mass national protest movement swept the country (522,931 signatures were collected throughout the country through the churches).61 Finnish politics, up to then directed internally and focusing on the language issue, suddenly became directed externally against Russian oppression. A new awareness of nationalism was formed, so that in the period around 1894–1906 the political mobilization of a highly literate society combined antiautocratic opposition with that against foreign domination. This perhaps explains the enormous participation of the Finnish electorate in their ﬁrst election with universal suffrage in 1906.62 The attempt to form a liberal party capable of cutting across the crucially salient cleavage between Swedish and Finnish speakers was unsuccessful in the last decades of the nineteenth century, as was the attempt to form a liberal party based on the Finnish-speaking population alone. The issue of russiﬁcation radicalized Finnish nationalism, and only in 1906 was a party of liberal orientation founded, the Young Finns, although the party that is usually regarded as the expression of Finnish liberalism – the People’s Party for Progress – was created only after World War I.63 At the end of the nineteenth century, with the onset of industrialization, the emergence of a new class cleavage tended to follow and reinforce stratiﬁcation lines identiﬁed by language, estate, and education. This new cleavage line was also congruent with the nationalist divide. In conclusion, the socialist movement was not affected by the linguistic issues that weakened 60

61 62

63

On this ﬁrst russiﬁcation process and on the reactions to it, see Wuorinen (1931: 189– 205). Martin and Hopkins (1980: 189). On the early Finnish popular movements and mass organizations, see Alapuro (1988: 101–110). Berglund and Lindstro¨m (1978: 50).

444

THE POLITICAL MOBILIZATION OF THE LEFT

liberalism and, at the same time, it was able to capitalize on the nationalist mobilization that had preceded its foundation and paralleled its extraordinary partisan organizational development (see Chapter 6).64 However, the alliance between the working class and the Finnish constitutionalists did not last. The Russian Revolution of March 1917 opened up the possibility of independence, but Finnish conservatives feared independence with a Social Democratic Party holding the majority in parliament. Moreover, the country faced a crisis in handling political control, as it had no local, autonomous police force or army. Political confrontation began to be organized by the formation of rival armed guards. When the Senate, which was in the hands of the bourgeoisie after the 1917 elections, moved to structure a military and police force at the end of 1917, the Social Democrats interpreted this as a declaration of war against the working class. On January 28, 1918, the party organized a coup to take power. This was successful in Helsinki and southern Finland, while in the north the White Guard began to counterattack. The war between the two sides lasted from January until the ﬁnal victory of the Whites in May 1918, helped by the intervention of the German army.65 The Irish process of nineteenth-century national political mobilization revolves around three issues, which are all interlinked in national resurgence: (1) the land and peasants issue, (2) the Catholic emancipation issue, and (3) the national emancipation issue. The difference with Finland lies in the lower intensity of linguistic issues as a mobilizing factor and the presence of the Catholic emancipation issue. The land and peasants issue and the national emancipation issue are instead fundamentally similar. The crucial difference in explaining the fate of the left thus turns on the way these issues developed sequentially and impinged on each other. The difference in the pattern of sequential solution (typical of Ireland) versus the pattern of sequential cumulation (typical of Finland) is therefore the basis of my comparison. In Ireland, the Protestants made few real attempts to convert the peasant community, and in the ﬁrst part of the nineteenth century various Catholic associations began to agitate for Catholic emancipation, home rule, and parliamentary reform. Catholic emancipation was achieved in 1829, and the issue of the position of the Catholic Church began to evolve over the century, with three signiﬁcant thresholds: in 1869, The Protestant 64

65

For the structuring of the Finnish party system, see Allardt and Pesonen (1967: 326– 329) and Nousiainen (1960: 28–43). A detailed account of this stormy period is presented in Jutikkala and Pirinen (1962: 254–267) and in Alapuro (1988: 150–196).

CLEAVAGE STRUCTURES

445

Church of Ireland was disestablished; in 1873, Trinity College in Dublin ﬁnally abolished its religious entrance test, which usually discriminated against Catholics; and in 1908, a separate Catholic University was established. This ongoing gaining of independence and inﬂuence on the part of the Catholic Church resulted in ‘‘clerical community leadership’’ and ‘‘hierarchical conservatism.’’66 The peasant and land issue evolved in the second half of the nineteenth century as a result of the large number of economically and legally vulnerable tenants, mainly concentrated in the west and south of the country: not only did they have no title of landownership, but their tenancy was unsafe. As I have already pointed out (see Chapter 7), the tenants’ grievances were ultimately resolved before the ﬁnal nationalist phase began, via a series of successive Land Acts that fostered peasant ownership in 1869, 1881, 1885, 1891, 1896, and 1903. The struggle for land was radical, bitter, and often violent, but the ﬁnal result was the consolidation of a class of smallholders: There were 3% of these as opposed to 97% of tenants in the 1870s, and they swelled to 97.4% in 1929 as opposed to 2.6% of tenants.67 This clearly converted the position of the great bulk of the peasantry from radical social protest to social conservatism, thereby neutralizing any potential rural radicalism. The Land League, founded in 1879, was immensely successful and was the ﬁrst mass movement to mobilize the Irish peasantry into a modern political force. However, it was not a self-representative peasant movement of the Scandinavian type. Neither the Catholic emancipation issue not the peasant and land issues ever gave rise to partisan organizations. The national issue was the last of the three to be activated and politicized, initially stimulated by a revival of Gaelic culture and Irish literature. In the second half of the nineteenth and early twentieth centuries, about four-ﬁfths of the Irish members in the British House of Commons were members of the Nationalist Party, which was devoted to obtaining home rule by constitutional means and within the framework of a Union. The other ﬁfth was formed by Unionists, who opposed this view. The lack of any concrete results discredited the Nationalist Party and led to the formation of dissenting groups that aimed at achieving independence from the United Kingdom through a wholly Irish administration that would eventually take over from the British. The 1905 foundation of Sinn Fein provided an alternative to the Irish Parliamentary Party. The following two decades of electoral and parliamentary developments are 66 67

Thornley (1974: 20). Ibid., 23.

446

THE POLITICAL MOBILIZATION OF THE LEFT

crucial in interpreting the failure of class mobilization in Ireland and need to be reviewed brieﬂy.68 The ﬁrst Irish patriots who set up the Dail were unable to hold their own elections. They accepted the result of the United Kingdom’s general election of December 1918 and considered the members of the United Kingdom’s Parliament elected from Irish seats to be members of the new Dail. The ﬁrst Dail, constituted in January 1919, was essentially a revolutionary body; the same was true of the second, constituted on the basis of the acceptance of the May 1921 United Kingdom election, held under the Government of Ireland Act. This resulted in an armed struggle between the Irish nationalists (Sinn Fein) and the British administration, eventually ending in the Anglo-Irish Treaty setting up the Irish Free State as a coequal member of the Commonwealth (1921). Although accepted by a narrow majority of the Sinn Fein deputies sitting in the second Dail (sixtyfour in favor and ﬁfty-seven against), the treaty was bitterly resented by the minority, which refused to cooperate and resigned from the Dail, carrying its position to the point of armed civil war against the majority. This split over the terms of the treaty produced the two anti- and protreaty parties that afterward dominated Irish political life. The Third Dail, elected in 1922, worked as a national convention to frame the constitution of the new state. The fourth Dail, elected in August 1923, was the ﬁrst to be constituted under a purely Irish electoral law. Near independence of the Irish Free State can therefore be dated to 1922, while absolute independence was achieved in various stages between 1933 and 1949, after which the country emerged as the sovereign Republic of Ireland. In 1922–1923, the antitreaty groups continued their policy of absenteeism, ignoring all the institutions of the new state. In 1925, a further split in the antitreaty Sinn Fein was determined by a group of pragmatic members who believed that parliamentary participation was better than total absenteeism. The opposition party, Fianna Fail, was founded in May 1926. In the August 1927 election, antitreaty deputies were voted into parliament, and, in 1932, being the largest parliamentary group, they formed their ﬁrst cabinet.69 How, then, did the working-class and socialist movement enter this series of conﬂicts and alliances? The urban middle classes were economically weak and numerically limited, and small farmers were originally mobilized by Catholic emancipation, radical land requests, and national 68

69

None of the British Electoral Reform Acts between 1882 and 1884 were extended to Ireland without substantial dilution. The enfranchised population was 6.5% of the adult population in 1868 and reached 26.9% in 1885. The secret ballot was introduced in 1872. For a history of the period, see Murphy (1975).

CLEAVAGE STRUCTURES

447

home rule. Under these conditions, the possibility of a delegated representation of the early working-class organizations within a liberal movement was nonexistent, and the labor movement developed with only a weak link to liberal political allies. The rivalry among the three great issues of Catholic emancipation, peasants and land, and home rule sometimes paralyzed Irish development, and their occasional coincidence produced great upheaval. Combined, these issues determined the sporadic and ultimately unsuccessful development of the social-class issues and the proletarian movement. Whenever a choice was to be made about priority among the various issues, socialist and proletarian issues invariably lost out to nationalism, Catholic inﬂuence, and peasant conservatism.70 In both Ireland and Finland, then, the predominance of the national issue subordinated liberalism to nationalism and negated any form of early Lib–Lab cooperation and support. However, this common feature leaves open the question of why this situation resulted in two opposite outcomes in regard to the strength and ideological orientation of the socialist left. I shall start by discussing the Finnish exceptionalism and shall later test the conclusions against the polar exceptionalism of Ireland. One interpretive line centers on the socioeconomic conditions fostering status polarization in the agricultural population and how these established favorable conditions for the electoral appeal of Finnish socialism. In 1910, 85% of the Finnish population lived in the countryside; two-thirds of them were engaged in agriculture, particularly small-scale farming; over three-quarters of the farms were less than ten hectares in size, and the few large estates were to be found in the most fertile parts of the south and southwest. As a result, in 1907, nine-tenths of the socialist votes came from the countryside,71 where the party received most of its support from agricultural workers and the crofters (small leaseholders who paid their rent either by working a number of days for the landowner or by making money payments). Such a high level of agrarian support for the labor movement is unique in Europe. Unlike the peasants in other countries, Finnish peasant communities owned the forests, and as a result of the timber-led industrial development, the effects of capitalism were felt immediately in the countryside; landowners and peasants prospered by entering the market and exploring alternative uses of production factors (e.g., from arable farming to stock raising and dairy farming). This transformation altered the relationship between the landowners and the landless 70 71

Thornley (1974: 37). For English sources on early Finnish socialism, see Alapuro (1981), Knoellinger (1960), and Kirby (1971) and (1990).

448

THE POLITICAL MOBILIZATION OF THE LEFT

peasants and crofters, all of whom were increasingly dominated by the market forces (in a country that, it should be remembered, had a low rate of emigration). The commercialization and rising prosperity of landowners and freeholders widened the gap with the landless and tenants. All these factors led to early radicalization of some sectors of the very large agricultural population and to growing class conﬂicts in the countryside.72 The strength of this rural class conﬂict and the division of the agricultural population into several often antagonistic social groups (the large landowners, the small independent farmers, and a large segment of landless agricultural workers, crofters, and tenants) bring Finland into line with the conditions of large parts of southern Europe, particularly Spain and Italy. On the other hand, the late but rapid and penetrating development of capitalism in the countryside as a result of the forestry economy is a feature almost unique to Finland. The large number of landless or landeager peasants, the high population pressure over the land, and the forms of sharecropping or crofting are characteristics that can be regarded as favorable to the penetration of socialism in the countryside, because they are important for its ideological orientation and organizational cohesion. To understand how the Finnish socialist movement managed to exploit this opportunity to the maximum, four additional political conditions need to be added to clarify how this concrete organizational solidarity between the small working-class population and the huge rural proletarian and tenant population became possible. The first special condition to be recalled (see Chapter 4) is the exceptionally high level of literacy in Finnish rural society: the extensive network of elementary schools in the countryside and the consequent availability of organizational skills among the rural population, as well as local cultural leadership (the school teacher, the church). The second important element is that the Finnish rural communities were exceptionally well organized politically from a comparative point of view.73 The role played by nineteenth-century linguistic and nationalistic movements in fostering this intense political activism and organization should be considered here. The early, strong local political organization of the Finnish socialists is considered to be as difﬁcult to explain as their strong electoral rural inroads, although the former may be the explanation of the latter.74 72 73 74

Alapuro (1981). Quoted by Alapuro (1981: 274) and others. Alapuro (1981: 274) considers this as a peripheral aspect: ‘‘The party was not only strongly supported, but also well organised in the countryside.’’

CLEAVAGE STRUCTURES

449

A third element is the exceptional and unique opportunity for the expansion of the Socialist Party that came about through the sudden mobilization of the 1905 strike and the electoral reform that multiplied by ten the number of those entitled to vote.75 Until 1905, national electoral politics and the Diet were completely closed to Social Democrats, who consequently concentrated on building their own organizational infrastructure, creating a dense network of extraparliamentary local organizations, centers, and branches, soon supplemented by women’s and youth leagues. The party was thus a mobilizing organization for political education and working-class formation and was clearly very efﬁcient organizationally before it had to face any electoral politics issue.76 Indeed, this concentration on organization and penetration was vital in that it enabled electoral success. Given the traditional nature of the political system, practically no other national political organization existed prior to the granting of universal suffrage, and – in particular – the Finnish nationalists (agrarian and conservative) were ill equipped to contest elections. Thus, the Socialist Party owed much of its success at the polls to the fact that it was organizationally better prepared to contest them.77 In addition, the national patriotic character of the 1905–1907 mobilization was crucial in cementing the class movement with the nationalistic issue. This phenomenon was absent in Ireland, where the incipient labor movement left the management of the nationalist issues to other partners (see later). Moreover, the revolution of 1905 and the belief in the efﬁcacy of spontaneous mass action strongly inﬂuenced the subsequent radical militantism of the socialist movement.78 A fourth political feature was the speciﬁcally political character of the socialist movement in Finland that resulted from the weakness of the trade unions. Party membership outweighed the trade union membership (see the data and discussion in Chapter 6). Unions were concentrated in craft sectors and urban centers, offered little or nothing to the rural proletariat, and had little or no inﬂuence on or control over the political organization. As a consequence, the Socialist Party – in sharp contrast with the Irish experience – was not constrained by corporate interlinkages and had great latitude for strategic action. This allowed it to devise appeals to the rural 75 76

77 78

Alapuro (1981: 286–287). On the lead of the Finnish socialist organization over all the other parties and for its concentration mainly on organizational and recruitment issues, see Nousiainen (1960: 28–43). Kirby (1990: 526). On the Red Guard movement mythology of radical protest, see Kirby (1986).

450

THE POLITICAL MOBILIZATION OF THE LEFT

poor and landless or leaseholders without facing internal resistance or constraints.79 These Finnish features can be contrasted with the interpretations of the failure of class politics in Ireland. The explanations focusing on economic backwardness, the predominance of the agricultural population, late industrialization, the small size of the working class, and, consequently, low class consciousness80 do not work comparatively, as these features were present in other countries, too.81 From an Irish–Finnish perspective, the difference concerns the socialist penetration of the countryside and of the urban middle classes rather than the strength of the working class or the level of industrialization. Another explanation of the failure of class politics82 is the intense rural conﬂict throughout the nineteenth century. In reality, this conﬂict did not radicalize rural laborers or the industrial laborers who had been recently recruited from rural society, as was the case in Finland. This was because the Irish ‘‘land war’’ of 1879–1882 was a struggle between landlords and a land movement (Land League) dominated by larger tenant farmer interests; the dominant purpose of the Land League was to obtain landownership, while radical proposals (e.g., Michael Dovitt’s proposal to nationalize land) remained peripheral. Class polarization in the countryside along Finnish lines therefore did not develop in Ireland, and by the beginning of the twentieth-century the land reforms had created a class of small, independent peasants for whom landed property (and deep religious feelings) meant that they adopted a conservative political orientation. On the other hand, Irish employers quickly accepted pluralist rules for industrial relations, and apart from some resistance in the 1907–1914 period, no great militancy by employers was registered. Thus, even in the ﬁeld of industrial conﬂict, no class radicalization was produced. Finally, the early tradition of popular politics dating from the beginning 79

80

81

82

An additional element was the fact that the rise of popular mass organizations coincided with the decline of religious observance and a weakening of church authority. This phenomenon primarily concerned factory workers and the rural population. See Kirby (1990: 532–533). Even in Finland, the temperance movement played an important mobilizing role, although it was not comparable to that in other Scandinavian countries. For instance, Orridge (1975: 484–491) attributes an important role to this factor, combining it with Catholic inﬂuence and nationalism. In rejecting this interpretation, Mair (1992) also provides data about class awareness showing that this is not lower in Ireland than in other countries. The data, however, refer to the end of the 1980s. A second hypothesis attributes the failure of the Labour Party to the clientelist and individualistic mobilization network set up by Fianna Fail in the countryside. The thesis exaggerates the role of clientelism in Irish politics, lacks comparative reference to other cases, and identiﬁes as responsible for the socialist electoral failure exactly what class politics should have broken down.

CLEAVAGE STRUCTURES

451

of the nineteenth century (the Catholic emancipation issue) established, for some historians, a tradition of cross-class political activism, adding another debilitating factor to class appeal. A further explanation is often sought in the Catholic inﬂuence (particularly in the countryside). Secularization was never a source of political division in Ireland,83 and this is an important difference from highly secularized and literate Finland, where not only was Catholicism absent, but where there was also a fairly strong (at least for homogeneously Protestant countries) clerical–anticlerical opposition that eased the entry of the socialists into certain milieus. Combining nationalism with Catholicism, Peter Mair84 has presented a chain of political explanations: (1) cleavage lines dominated by an opposition movement imbued with a nationalistic and Catholic identity; (2) the consequent development of a culture emphasizing solidarity, cohesion, and homogeneity; (3) successful representation by a ‘‘national’’ party that had always insisted on such a role (ﬁrst Sinn Fein and later Fianna Fail); and (4) reinforcement by the incapacity of the Irish Labour Party to advance a class appeal and a policy line consistently. The last factor is nationalism, whose dominant role did not decline after World War I, as it did in Finland. Yet, by the 1920s and 1930s, the failure of Irish class politics had already become apparent. If, therefore, we take nationalistic mobilization to be important, it must explain both why the potential socialist constituency was captured by nationalist politics before the 1920s in Ireland and why this did not occur in Finland in the same period. The failure of class politics in Ireland invites, therefore, a host of reasonable explanations. Of these explanations, the weak structural condition of class antagonism in the market combined with the powerful political attraction of two cross-class reinforcing appeals (Catholicism and nationalism) is convincing. Yet, one doubt remains: Why were labor and socialist organizations, intellectuals, and militants so weak in these developments? Why did they not manage (or even try) to shape and build on anticlericalism, as happened in so many Catholic countries on the Continent? Why did they not try to guide or focus the turmoil in the countryside, as was often the case on the Continent in the same period? And why, ﬁnally, did they not try to ﬁnd a place for themselves within the nationalistic movement, infusing it with socialist values and goals and exploiting its mobilization potential, as happened in Finland? Irish socialism lost a war that it never fought. Its failure to compete is 83 84

Orridge (1975: 484–491). Mair (1992: 383–410).

452

THE POLITICAL MOBILIZATION OF THE LEFT

macroscopic. For example, in 1918, Sein Fein worked actively to ensure that class divisions would not disrupt the national revolution, units of the Irish Republic Army being forbidden to take part in land seizure; indeed, they often worked to end radical factory and land actions. This led to a dilemma for the trade unions and the Labour Party: whether or not to contest elections and thus oppose nationalist Sein Fein candidates. Sein Fein obviously sought to dissuade labor from doing so; they suggested that social issues could be dealt with only after independence. The contrast with Finland is clear here. Not only did the Finnish Social Democrats not renounce their partisan activities until independence was achieved, but the national issue itself was much more closely linked to the social cleavage there, given that part of the Finnish bourgeoisie was aligned with the tsar. In Ireland, the trade unions and the Labour Party ﬁnally decided not to contest the 1918 election, the ﬁrst under universal adult suffrage. The near civil war that ensued between pro- and antitreaty parties radicalized the country on the one issue on which the working-class constituency was possibly divided, but the Labor Party organization did not try to overcome, contain, or direct such a division; indeed, it abstained. Its involvement in the national revolution in a subordinate and muted position contributed to the containment of working-class militancy, and thus it can be concluded that class conﬂict was deliberately contained. It can be argued85 that the decision of Labour to withdraw from the 1918 general election, in which two out of three voters voted for the ﬁrst time as a result of the recent enfranchisement, was fatal and that, by not contesting the election, Labour allowed Sinn Fein, the Unionists, and the Irish Parliamentary Party to monopolize the campaign and orient ﬁrsttime voters, as well as helping to establish the hegemony of nationalist issues and cleavages. However, this damage was not irreparable,86 given that approximately 55% of the newly enfranchised electorate did not actually vote.87 More damaging was probably the subsequent inability of the socialists to compete electorally with the other parties. On the treaty and the Free State’s relationship to the United Kingdom, Labour’s stand was confused and ambiguous in comparison to that of Fianna Fail and Cumann na nGaedheal. It was against the treaty and against the oath of allegiance, but it refused to make these points central parts of its program, 85 86 87

Farrell (1970) and (1971: 487–488). Mair (1978). In its ﬁrst elections in 1922, the Irish Labour Party got 21% of the vote, but this percentage is not reliable, as the already exploded institutional conﬂict between the proand antitreaty factions resulted in a very low turnout, a large number of uncontested seats (30%), and muted party competition.

CLEAVAGE STRUCTURES

453

arguing that the economic issue took precedence over constitutional details.88 On the economic policy dimension, Fianna Fail argued for autarchy and self-sufﬁciency – trying to attract the support of the small farmers and petit bourgeoisie concentrated mainly in the west and southwest – while the government argued for open economic policy and free trade – defended by the large farmers concentrated in the east and southeast: The Irish Labour Party emphasized welfarist policies oriented to the urban proletariat, which had minimal appeal for the majority of the population. The Irish socialists failed to compete electorally over the salient issues that shaped the party system. As a result, they found themselves in a central position by default, advocating either intermediate positions or different themes on both major issues in a highly polarized context.89 In interpreting this situation, it is necessary to stress the importance of Ireland as a periphery of the United Kingdom. The main British contribution to the failure of the socialists was to shield Irish socialism culturally from the continental ideological inﬂuences that appeared to be anachronistic in the British context of strong industrial unions but that might have helped Irish socialism to shape its strategic role more effectively. These inﬂuences were those of Marxism and anarcho-syndicalism, the peasant and land debate faced by many continental socialist parties, but also the development of the anticlerical themes and feelings of continental socialism. The Irish trade union movement was so deeply inﬂuenced by the British model (from which it developed) that it was even incapable of thinking in terms of peasant mobilization and support. Clearly, however, a British craft-style unionism would never manage to radicalize the peasantry. In Finland, political socialism dominated a very weak trade union movement and directed it according to a competitive electoral logic. This organizational resource made it capable of exploiting the sudden democratization within the context of a national liberation movement. The Irish Labour Party was exactly the opposite: an indirect, trade union–dominated, partisan undermobilized party with no resources and little or no capacity for tactical political maneuvering and strategic behavior. It simply represented the organized trade union stand, which was an extension of the 88 89

See Mitchell (1974: 272). A statistical analysis of swings in these elections shows that the Irish Labour Party lost support to both major parties, particularly to the government party, probably because it alienated the laborers on the big farms in the forefront of the export trade to Britain (and favoring an open economy) and also because it had previously enjoyed the support of antitreaty voters who had expressed their dissatisfaction by voting Labour when Fianna Fail abstained. See Mair (1978: 67–68).

454

THE POLITICAL MOBILIZATION OF THE LEFT

more single-minded unions and as such refused to accept the logic of electoral competition. In Finland, late and sudden political and electoral mobilization had been preceded by the strong organizational development of political socialism. In Ireland, the much earlier popular mobilization and politicization of alternative divisive issues took place in the absence of any autonomous organizational consolidation of political and party socialism. The latter was based on and depended on narrow corporate bases and remained, even ideologically, the inexorable prisoner of a small section of the urban working class.

MOBILIZATION OF THE RELIGIOUS CLEAVAGE AND THE FORMATION OF DENOMINATIONAL PARTIES In this section, I concentrate on the intertwining issues of religious and class mobilization and, in particular, on 1. Whether and to what extent religious identities were politicized, that is, politically characterized and crystallized by and through conﬂict over policies that concerned them. 2. Whether religious afﬁliations and allegiances were not only politicized, but also organized by purpose-speciﬁc political organizations in the corporate and electoral/parliamentary channels. 3. The timing of such politicization and organization vis-a`-vis class cleavage politicization and organizational mobilization. Religious issues and religious groups may exist without gaining much political importance, as they are not linked to speciﬁc policy alternatives; or they may have political importance without, however, being linked to speciﬁc electoral and corporate organizations. Depoliticization versus politicization, the presence or absence of speciﬁc organizational mobilization, and the timing of the former with respect to class cleavage development are the three main coordinates of this section. The structuring of a preindustrial cleavage line such as religion impinges on class mobilization and the formation of the class cleavage in various ways. The Protestant–Catholic dividing line was important because the close relationship between state and church and the incorporation of the church into the state bureaucracy in the Protestant world reduced or eliminated the potential for a state–church conﬂict. Moreover, the Reformation produced an early nationalization of the territorial culture.

CLEAVAGE STRUCTURES

455

This favored the processes of mobilization from below, also because the early development of literacy aided and encouraged the mobilization of the lower strata into mass politics. In contrast, in Catholic territories, the supranational nature and attitude of the church tended to favor mobilization from above by Catholic hierarchies; the delay of literacy made the mobilization of the lower classes from below more difﬁcult, and the conﬂict over the control of the educational system and of mass education led to the mobilization of the church against the state. Class unionism appealed to a wide potential base regardless of other identities; it referred to the workers’ interests and conditions in the functional division of labor. However, such an appeal was more successful where other community identities, which could rely on preexisting groups for cultural organizational consolidation, did not exist. So, in general, we may expect the mobilization of religious cleavages to inﬂuence the development of class cleavage in all its dimensions: in its electoral potential, insofar as it may reduce its lower-class constituency; on its social composition, insofar as it may lead to radical anticlerical middle-class socialism; in its organizational nature, insofar as it may give a defensive subcultural style to the class cleavage, making it less expansive and more reliant on closed community groups and traditions; and in its organizational cohesion, insofar as the relationship with a religious political movement that is both a competitor and a potential ally may further exacerbate the internal conﬂicts that lead to organizational divisions. To analyze the variety of political denominational experiences, I use the categories presented in Table 8.1.90 ‘‘National establishment church’’ identiﬁes the experience of the dominant and state hegemonic Protestant churches of the Northern European countries that were religiously homogeneous (with the partial exception of Great Britain) and state hegemonic in the sense that they enjoyed territorial dominance, legal privilege, and privileged access to secular rulers. In these countries, the early incorporation of the national church hierarchies within the national bureaucracy, and the absence of the dualism between state and church hierarchies, meant that a church–state cleavage never developed during the phase of nationalization and political democratization. In the United Kingdom, as well as in Scandinavia, dissident religious movements, in particular Methodism, played an important role in raising workers’ cultural level and self-respect in the very early phase. The lay preachers enjoyed the conﬁdence of their fellow laborers and won their 90

Labels of these types are inspired by the typology of the church’s political attitudes in Latin America by Vallier (1970a) and (1970b).

456

THE POLITICAL MOBILIZATION OF THE LEFT

Table 8.1. Types of churches

trust, and this was important for the formation and rise of independent internal protest leadership among workers.91 In the United Kingdom, however, no independent religiously inspired politico-electoral movement ever emerged, while in Scandinavia small Christian Democratic parties drawing on the Lutheran faith emerged only when the socialist mobilization had been completed.92 Rather than being an early reaction to the political challenge expressed by liberals or to the social challenge expressed by socialists, denominational mobilization appears more as a late reaction to the accelerated secularization and even the de-Christianization of societies.93 The term ‘‘political church’’ can be explained as follows: In continental nations and in Ireland, the churches and denominational movements took clear, active political stands early in the phase of nation formation and mass political mobilization, and religious issues were politicized to a certain extent everywhere, even if not necessarily through the formation of distinct corporate or electoral organizations. The 1864 papal denunciation of liberal doctrines in the Syllabus of Errors and the First Vatican Council dogma of papal infallibility in 1870 were motivational steps for the political and electoral organization of Catholic movements. However, the presence of elected representatives of the Catholic communities (excluding the traditional clerical membership of upper houses of parliament) predates 91

92

93

See Rimlinger (1960: 333). On the impact of Methodism, see Wearmouth (1937: 221– 238). Pelling (1965: 191–192) claims that labor was initially more successful where religious nonconformism had been particularly strong in the past. Only in Norway does a Christian Party date back to the 1930s, but it managed to poll an average of 10% of the votes and gain national importance only in the late 1950s and early 1960s. The circumstances surrounding the late emergence of these religious parties are described and discussed in Madeley (1977). On the religious dimension in Norwegian politics, see Converse and Valen (1971) and Valen and Rokkan (1974). Berglund and Lindstro¨m (1978: 61).

CLEAVAGE STRUCTURES

457

such documents. As early as 1830, in Belgium, 30 of the 200 members of the assembly were priests. The Austrian Reichstag of 1830 included twenty elected clergymen. In the French Constituent Assembly of 1848, three bishops and twelve priests were elected. Twenty-eight Catholic clergy representatives were present in the Prussian national assembly of 1848, and thirteen Catholic clergy representatives were elected in the Frankfurt Vorparlament of 1848 (representing all of Germany). The subtype I have labeled the ‘‘monopoly church’’ was characteristic of homogeneous Catholic countries, where the church enjoyed territorial dominance and legal privilege, as well as continuing control of the educational system. In Austria, the Catholic Church had always been one of the pillars of the imperial central establishment, and the state–church relationship and formal concordats – such as that of 1855 – were extremely favorable to the church’s dominant and monopolistic position. Even in the years of stronger liberal inﬂuence (the 1870s and thereafter), the abolition of the concordat, and the formal separation between state and church, a close link between Catholic interests and state direction was maintained, as were several of the church’s privileges. Catholicism and the Catholic hierarchy were so strongly associated with the authoritarian monarchical state and enjoyed such privileged access to the secular rulers that for many years no need for political self-representation was felt, although the conservative political identity of Catholics and Catholic organizations had been well established for a long time. The Catholic Party was formed by Karl Lueger very late, from 1905 to 1907, and with some reluctance. From the beginning, it was a mass antisocialist political movement, in a country where political liberalism had never developed and where conservative circles rallied around monarchical and bureaucratic institutions, refusing to organize politically for electoral competition. Therefore, after the collapse of the empire, the Catholics inherited the representation of virtually the entire political spectrum from the center to the extreme right. The party was able to combine, if uneasily, dynastic revanchism and state-bureaucratic interests, deeprooted nationalism (Heimwehr), and the peasants’ radical wing (which was the least hostile to the socialists). This coalition of interests also explains why, during the 1920s and 1930s, Austrian Catholicism maintained a residual hostility to the new republican political system and an implicit identiﬁcation with and adherence to traditional state structures and bureaucracies, repeatedly advocating ‘‘corporatist’’ reforms and pressing for constitutional revision in a less liberal direction. The Catholic political movement, having organized itself against the well-rooted socialist mobilization and having inherited establishment in-

458

THE POLITICAL MOBILIZATION OF THE LEFT

terest representation, had strong class-based social support from the beginning – that is to say, it was strong among the urban middle classes and the peasants but very weak among the urban proletariat. In contrast to the other continental cases, Catholic mobilization in Austria did not preempt or cut across class appeals, but rather reinforced the dividing line of the latter.94 This class base and e´tatism made Austrian Catholicism different from the ideological clerical variant that I consider to be typical of the French and Belgian experiences and also different from the opposition or ghetto variants that represent the early experiences of Dutch, Swiss, and German Catholicism. The subtype labeled ‘‘ideological clericalism’’ is typical of homogeneously Catholic countries and is characterized by the following elements: (1) it faced ideological anticlericalism and early attempts to sharply curtail its position and privileges by strong liberal forces; (2) the Catholic Church and community aligned with conservatives; (3) it was politically identiﬁed and drawn into partisan conﬂict at an early stage that preceded the political organization of the socialist movement. In France and Belgium, the Catholic community developed a political identity early in the nineteenth century, but the outcomes of this process were very different: early electoral and political organization in Belgium versus late and failed organization in France. As mentioned before, the political role of the Catholic elite is a constitutive element of the Belgian polity given its role in the secession from the Netherlands. The Catholics had been the system support group and party from the beginning. In the mobilization confrontation with the liberals, they developed an early popular tendency. By 1867, a Federation of the Catholic Workers Societies existed, and by 1886 the Belgian Democratic League was a coalition of Catholic ancillary organizations in all ﬁelds: trade unions, cooperatives, rural banks, mutualities, and so on. Therefore, notwithstanding the ongoing internal strife in the Catholic world between Christian-social and conservative Catholics (the Rexist Movement) and between the Walloon and Flemish wings, Belgian Catholicism mobilized early, more as an antisecular and antiliberal reaction than as an antisocialist one, and was never in a monopolistic position, strongly relying on its own forces as an independent movement. How important the early political inroads of the Belgian Catholics were for the successive development of the socialists is indicated by the strong territorial comple94

On this peculiar class base of Austrian political Catholicism, see Houska (1985: 18–124, 163–166).

CLEAVAGE STRUCTURES

459

mentarity of the two movements in the ﬁrst half of the twentieth century. In Brussels and Walloon, the early electoral impact of socialism was mainly at the expense of the liberals,95 while in the more rural and less secular Flanders, the strong inroads made by the Catholics and the inﬂuence of the church countered the growth of socialist organization from 1910 to 1920.96 Linguistic differences did not play a role at this stage. In France, political Catholicism passed through a number of different political experiences. The shock of the 1789 revolution, the 1848 revolution, and the Commune forged its role as an early antiliberal and antisocialist force. Napoleon III enjoyed Catholic support or neutrality,97 and in this phase the links of the Catholic Church with monarchical sectors and Bonapartists prevented the emergence of autonomous Catholic forces. During the Third Republic, the Catholics supported the moderate, conservative, and monarchical sections. However, after the defeat of the monarchical forces that had formed behind Boulanger, the Catholic conservative forces rallied to the Republic. Even when they confronted widespread anticlericalism and anticlerical legislation in the 1880s and again in 1900– 1905,98 no party was organized. Divided into left and right, as everywhere else, French Catholics were also divided into republicans and monarchists, and attempts to form a Catholic Party failed repeatedly from 1885 on. The Catholic hierarchy could not agree, and the pope often discouraged Catholics’ attempts to organize themselves politically for fear of the development of an overly monarchical party. Other internal divisions of a different nature (the relationship of the pope with a nationalistic hierarchy authoritarian and antidemocratic tendencies, the Dreyfus affair, and the Action Franc¸aise) further complicated the issue, so that no common ground for Catholic alignment was established before World War I or between the world wars.99 In sum, the clerical guidance in politics was weak and unable to move beyond general indications of tendency. 95 96 97

98

99

Hill (1974: 37). Hill (1974: 39, 43). He gained this support through the 1850 loi Falloux – which gave the church extensive rights in secondary education – by sending its troops to crush the Roman Republic and to restore Pius IX to the papal seat, and later by acting as the defender of papal temporal power. With Catholic schools deprived of all state aid and most religious orders expelled, extensive church property taken over by the state, diplomatic relations with the Vatican broken, and state and church separated. There were several attempts to build an autonomous political organization of the French Catholics from the 1890s to 1945: 1896 the Parti de´mocrate chre´tien; le Sillon of Marc Sangnier in 1910; the Action Liberale Populaire between the world wars; and the Parti

460

THE POLITICAL MOBILIZATION OF THE LEFT

Attempts to create a ‘‘social Catholicism’’ remained isolated and unsuccessful, and during the second half of the nineteenth century, no real Catholic political current existed within the working class. From 1894 on, there were several timid attempts to set up mixed trade unions, but only in 1919 was the Confe´de´ration Franc¸aise des Travailleurs Chre´tiens (CFTC), the most successful (mildly) Catholic social movement, founded. In France, therefore, the politicization of religious issues and of Catholic identity was originally antiliberal and antirepublican. The repeated failures in organizational mobilization and encapsulation were due to the fact that these attempts were led by either right-of-center or left-of-center intellectuals and political elites, in contrast with the bulk of the Catholic-motivated electorate, which existed more in the center. Perhaps for these reasons, the Catholic hierarchy never openly supported any of these parties. French Catholicism therefore remained organizationally ‘‘open,’’100 although it was strongly characterized politically, with Catholics supporting mainly conservative and moderate forces or, less frequently, mildly progressive forces. The Italian experience has been characterized in two boxes in Table 8.1. First, it can be likened to the experience of a ‘‘ghetto church’’ (see later) for the period between the 1860s and the 1900s. The Catholic Church’s break with the liberal and conservative elite, which had uniﬁed the country at the direct expense of the papal state territories, was far deeper than in Belgium or France. From the 1900s on, however, the progressive acceptance of the new Italian state transformed this role, so that from 1898 on – ﬁve years after the ofﬁcial founding of the Italian Socialist Party – a Catholic union movement started to develop. This was most successful in the areas of Piedmont, Lombardy, and the Veneto, where the inﬂuence of the church was stronger and relied on a large number of intermediate Catholic associations and institutions (savings banks, mutual societies, etc.). From 1896 on, the Opera dei Congressi started, in its own words, to prepare ‘‘in abstentionism . . . the healthy, educated and organized body of Catholic electors for the moment in which even the conquest of political power will be possible.’’101 By 1910, Catholic unions had established a network of trade associations and a general executive was so de´mocrate populaire in 1924. A similar fate eventually undermined the most successful Christian political experience, the Mouvement Re´publicain Populaire after 1945. On these various experiences, see Vaussard (1956: 86–105). For an analysis of why Catholic party formation failed in France, see Kalivas (1996: 114–165). 100 The distinction between open and closed Catholicism is drawn by White (1981: 7–8, 147). I use it to indicate the level of organizational encapsulation of the Catholic community and electorate. 101 Quoted in Galli (1968: 26).

CLEAVAGE STRUCTURES

461

organized in Bergamo in 1909.102 In 1905, a new papal encyclical (Il fermo proposito) ﬁnally authorized Catholics to vote in the constituencies where the bishop of the diocese deemed it advisable. Sixteen members of parliament were elected in 1905 with crucial Catholic support, and their number increased progressively until World War I. For thirty years (from 1870 to about 1900), church opposition in Italy played a clear antisystem role, depriving the liberal and conservative forces of a potentially broader popular base of support and contributing to the delay of suffrage extension. At the same time, the attitude of the church largely helped shape the closed and self-referential political attitude of the Catholic subculture within the liberal state and society. The intransigence of the church against the political participation of Catholics in the new state meant that what was a potentially important political issue remained dormant. The religious cleavage existed, but it was not given political expression until the ﬁrst decade of the twentieth-century. Thus, socialism developed in this climate for twenty years without any speciﬁc Catholic competition. It should, however, be remembered that throughout this period the suffrage was highly restricted. The level of enfranchisement creates a crucial difference between Italy and France. In France, the attempts by Marc Sangnier in 1909 and by the Parti De´mocrate Populaire in 1924 to form a Catholic party met with ambivalent and uncertain support from the Catholic hierarchy. These were not attempts to represent a Catholic world that had hitherto been politically unrepresented and/or excluded, but rather were attempts to detach this world (or part of it) from a representation delegated to moderate and conservative sectors. This switch from consolidated conservative representation to direct Catholic mobilization was more difﬁcult organizationally and electorally, and also less proﬁtable from the point of view of the Catholic Church hierarchy. In Italy, the territorial dimension of the Catholic opposition and the restricted suffrage (with the exclusion of the rural masses from electoral competition) meant that Catholics could not be represented by conservative forces. The foundation of a Catholic Party, after World War I, paralleled the introduction of universal suffrage and proportional representation, with no popular conservative representation to defeat and substitute. This offered an exceptionally favorable opportunity to the organizers of the Catholic Party. These differences in Catholic political representation were important for the nature of the class cleavage that developed in the two countries. Confronted with the late entry of an organizationally closed, subculturally 102

On early Catholic efforts in this domain, see Rossi (1985: 3–49).

462

THE POLITICAL MOBILIZATION OF THE LEFT

encapsulated, and politically autonomous Catholic opposition, Italian socialism (and later on, communism) developed more-pronounced organizational subcultures of their own to contrast with and ﬁght the Catholic one. In France, strong links between the political parties and civic society never developed in either the Catholic world or, later, in the class left. With a less organizationally and subculturally bound electorate, the French socialists and communists never really felt the need to develop a counterculture or a strong organizational network of party-linked associations. As a result, they remained ideological currents rather than organizational machines; that is to say, they relied more on ideological competition than on organizational mobilization.103 Catholic political mobilization was very early and organizationally dense in Belgium, late but also organizationally dense in Italy, and completely absent in France. However, in all three countries, anticlericalism came to play a crucial role in the ideological identity of radicalism and the left.104 If denominational and clerical mobilization tended to be inherently cross-class, the same is true of its opposite, secularism and anticlericalism. What denominational appeals subtracted from socialist mobilization potential in terms of lower-class constituency, and therefore in terms of social homogeneity of support, could be ‘‘reimbursed’’ by radical anticlerical middle-class support in homogeneous Catholic countries. This contributed to the shaping of a class cleavage that was not necessarily weaker in terms of electoral support than that of countries where denominational mobilization was absent, but it was different in terms of social composition, being less class homogeneous. The subtype I have labeled the ‘‘ghetto church’’ is characterized in its early phase by isolation and defensiveness, either because of minority disadvantages and emancipation demands or because of the will and need to insulate Catholicism from secular forces or competing denominational forces. This is typical of Catholic political representation in the religiously mixed central European countries – Germany, the Netherlands, and Switzerland – where the national and nation-building establishment was predominantly Protestant. In these countries, the Catholic identity was politicized historically in this process of state building and was reinforced by the emancipation ﬁght within predominantly Protestant contexts. The Catholic movement in the German Reich (the Catholic Center 103

104

The tribune role (function tribunitienne) that Lavau perceptively sees as a characteristic of French Communism; Lavau (1968: 445–466). For Belgium, see Kittell (1973) and Strikwerda (1988).

CLEAVAGE STRUCTURES

463

Party) emerged from a broad tradition of Catholic resistance to the secular state and to the Prussian-led German national uniﬁcation that excluded Austria. In the overwhelmingly Protestant monarchy and bureaucracy of Prussia, one-third of the population was Catholic, and already in the 1840s there was considerable mobilization on issues concerning national uniﬁcation (preferred with Austria), civil marriage, and secular education, on which Catholics organizations differentiated their position from that of the liberals. They also diverged from pro-Protestant conservatives. Catholics could, therefore, organize themselves only as a separate group. By 1852, a distinct Catholic Party existed in the Prussian parliament. However, Catholics felt themselves to be a besieged minority even in uniﬁed Germany. The Kulturkampf started in 1871 and culminated with the laws of May 1873 (control over the formation of the clergy, attribution of seats, limitation of the disciplinary power of the church; institution of a national court for ecclesiastical issues), which were weakened from 1878 on due to the conciliatory intervention of the Vatican and were ﬁnally abolished with the 1887 ‘‘peace law.’’ By then, the political mobilization of Catholics was already very advanced. It materialized into a network of Catholic associations and into a strong clerical presence in the cadres of the Center Party. Catholic action in the social ﬁeld occurred as early as that of the socialists. By the end of the 1860s, Bishop Ketteler was promoting the ‘‘social-Christian’’ associations directed at German workers in competition with the Lassallian union organizations. It is estimated that in the general and universal male suffrage election of 1881, 86% of Catholic voters voted for the Center Party.105 The later development of the Catholic Party moved in an increasingly conservative direction. In their search for legalism and recognition, the Catholics ended up identifying themselves with the imperialism and colonialism of Wilhelmine Germany, and, like the Social Democrats, believed that the coming of World War I would provide a good occasion to show their patriotism and obtain ‘‘national rehabilitation.’’ In the Netherlands, the two most important religious minorities – the orthodox Calvinists and the Roman Catholics – had ﬁrst launched a successful emancipation movement some decades before the working class began to respond to the call of socialism. The Roman Catholic minority was heavily concentrated in the southern provinces of Limburg and North Brabant. Strictly speaking, only in 1896 and 1897 did all electoral Catholic associations accept a common political platform (the General Union of 105

Brezzi (1979: 113).

464

THE POLITICAL MOBILIZATION OF THE LEFT

Catholic Electoral Associations), and only in 1904 was the Catholics Political Party ﬁnally formed.106 However, the political action of the Catholics as such went back much further: Since the 1840s, they had fought for the reestablishment of the Catholic hierarchy, which they achieved in 1853. Between 1840 and 1860, Catholics supported liberals and their call for the separation of the state and the (Protestant) church. This alliance was broken in the 1860s and was slowly replaced by an alliance with the Anti-Revolutionary Party, which brought about the electoral defeat of the liberals in 1888. Throughout this period and later on, electoral associations of the Catholics were active, even though a formal party was lacking. It was, however, the struggle for subsidized private religious schools that provided the ﬁrst major catalyst for mass action. The liberal Education Act of 1878 had already triggered massive petition movements, and the religious calls for national action and organization eventually resulted in the formation of the Anti-Revolutionary Party in 1879 (the ﬁrst effective national party, as already mentioned). It was this common Catholic and Calvinist interest in subsidized religious schools that forced the two organizations to cooperate in parliament on the basis of an antiliberal stance.107 Obviously, the uniqueness of the Dutch case lies in the division of the national Protestant establishment, which led to the very early formation of an orthodox Calvinist opposition. Catholics, instead of facing a common strong Protestant hegemonic establishment, had to strive for recognition and emancipation in a complex multicleavage environment that offered ample opportunities for electoral and political alliances long before the socialists started their political and electoral activities. In Switzerland, a strong Catholic identity was reinforced and supported by territorial cantonal concentration and identiﬁcation and by longstanding international issues concerning Swiss alliance politics in international relations. The Sonderbund, defeated in 1847, reinforced this highly territorialized political identity. Confessional rivalries, sometimes masking other conﬂicts of a territorial, economic, and strategic nature, characterized the post-Sonderbund period from 1874 on, when exceptional measures 106

107

Righart’s comparative study of Catholic mobilization in the Netherlands, Switzerland, Belgium, and Austria (1986) lends support to Daalder’s thesis: The development of the Catholic pillar (zuil) was mainly due to the effort of Catholics in the diaspora, i.e., of Catholics in territories not dominated by the Catholics. In the more traditional and less developed centers of Catholic domination (Brabant, Limburg, and the Catholic cantons in the central part of Switzerland), they resisted the formation of a Catholic organizational infrastructure in politics and trade unions. Daalder (1981: 220–222).

CLEAVAGE STRUCTURES

465

were taken by the new federal constitution to limit Jesuit activities, the number of Catholic convents, and so on. Therefore, the Swiss had their own limited Kulturkampf on the issue of the 1874–1876 constitutional reform. However, Catholics had a power base in cantons where they were the majority. They were an isolated but territorially represented opposition. This strong territorial concentration and representation delayed the formation of a federal party because, while parties and movements were present at the cantonal level, it was viewed as difﬁcult, and perhaps unnecessary, to coalesce into a national Catholic movement. Thus, the cantonal basis of Swiss Catholicism actually worked against its centralization, blocking attempts at national political uniﬁcation. When in 1912 a Catholic federal party was ﬁnally founded, its formation was accelerated by the growing strength of the socialists.108 The Dutch, German, and Swiss Catholic political movements are all characterized by the same features: those of minority movements with a relatively high territorial concentration that mobilized early against state/ nation builders on the emancipation and equality issues. The accumulation of the political challenge, with the social challenge coming from the early socialist movement, may explain why party forms developed earlier in Germany than in Switzerland and the Netherlands. The Dutch division on the Protestant front and the active secularism of the liberals offered Dutch Catholics ample room for political maneuvering without the need to build a speciﬁc electoral-parliamentary organization. Similarly, the strong territorial representation of the Swiss Catholics in a federalized state allowed them to defend their autonomy by opposing centralization and national standardization. Neither of these options were available to German Catholics. All these movements took on an early antiestablishment character – which was unknown to ideological clericalism – as the main push toward mobilization was in opposition to the state-dominant Protestant circles; and from the beginning, they tried actively to obtain the support of the lower classes. In all likelihood, the appeal of these subcultural defense movements was more likely to directly affect the lower classes than in the cases of ideological clericalism, and their capacity to inhibit socialist recruitment should be more evident when comparing zones of strong Catholic presence with those of Protestant predominance. The Irish case is difﬁcult to characterize because of the North–South and pre- and postindependence differences. The emancipation of the Catholic Church in Ireland was one of the key politically mobilizing issues of the nineteenth century, and the Catholic Church and the higher as well as 108

See Girod (1964), Gruner (1977), and Seiler (1977).

466

THE POLITICAL MOBILIZATION OF THE LEFT

lower hierarchies took a clear stand in favor of Irish independence. This is probably the reason why no relevant stream of anti-clericalism developed in Ireland. Catholics never felt any need to organize themselves politically under the umbrella of speciﬁc electoral organizations because the entire nationalist movement was imbued with strong Catholic values and orientations. Finally, socialist organization and mobilization was so late and delayed, and so second-rate compared to the national one, that it could hardly challenge the Catholic political defense. After independence and partition, these characteristics were dramatically accentuated in the Protestant-dominated northern part of the country. However, in the Republic, the Catholic Church suddenly took on the role of a monopoly church, with extensive privileges, direct and special access to rulers, and a limited distinction between the state and the church. It is hard to decide whether Ireland lacked a Catholic Party during the interwar period or whether it was, on the contrary, the only country to have two major Catholic parties. Indeed, for both Fianna Fail and Fine Gael, religious appeals, symbolism, and sensitivity were not inferior to those of most continental Christian parties. Since my major concern here is to inquire into the extent to which Catholic identities were politicized and mobilized before the socialists entered the political game, it is safe to regard Ireland as an extreme case of emancipationist ghetto Catholicism, where the combination of the church with the national emancipation movement led to pervasive Catholic political mobilization. Thus, in this case, the absence of the political party should not be regarded as a failure of organizational mobilization, but rather as a sign of its overwhelming victory. We can organize the information discussed so far on the key dimensions of religious mobilization in Western Europe into a synthetic form. This is laid out in Table 8.2, where each case is classiﬁed along the dimension of its religious heterogeneity/homogeneity, religious politicization, religious organization, and timing with respect to the class cleavage organization. This scheme will help the ﬁnal considerations concerning the interaction between class cleavage and non-class-cleavage structuring.

THE PEASANT ISSUE: MOBILIZATION OF THE PEASANTRY The question of agrarian society and its transformation was widely debated by socialist thinkers and intellectuals around the turn of the twentieth century. The position of the agricultural workers was likened to that of the industrial working class, and the key problem was felt to be the

CLEAVAGE STRUCTURES

467

Table 8.2. Patterns of religious mobilization in Western Europe

attitude toward the intermediate agricultural groups: the peasantry and particularly small subsistence farmers and the different types of semiindependent peasantry made up of sharecroppers, tenants, and so on. Marx’s economic model of capitalist development and capitalist concentration led him to forecast the inevitable economic decline of both farmers and the peasantry, while, at the same time, recognizing the high potential for revolt among the latter in a revolutionary situation.109 The ﬁrst question was whether and how fast the small peasantry would disappear, transformed into dependent labor by the mercantilization and capitalist development of agriculture. The second was what policy the socialists would adopt toward it in the transition phase. The importance given to the second question depended on the answer to the ﬁrst question and, more importantly, on the size of this social group. Orthodox Marxism faced a political predicament in terms of short-term versus long-term political action, the choice of which could lead to a pronounced break between economic analysis and political practice. This was because to protect, defend, and represent poor peasants meant to protect and give incentive to holding private property, and this meant acting politically 109

On Marx and the agrarian question, see Mitrany (1951).

468

THE POLITICAL MOBILIZATION OF THE LEFT

against the law of historical development. Thus, to leave poor peasants to their historical destiny was normally the orthodox response. To this Marxist position is usually opposed a ‘‘socialist-revisionist’’ solution to the agrarian question, epitomized in the work of Bernstein. Marx and his immediate successors never considered that agrarian concentration could develop differently from industry, or develop with a slower pace and intermediate steps during which the property of small peasants could survive and even grow as a result of the rationalization of the still semifeudal agrarian structures in many countries. In reality, the same intermediate phases could be obtained if the political elite made deliberate efforts to stabilize their political position by supporting the countryside. With an important series of statistical data, Bernstein demonstrated that small productive units did not decline in number and wealth but tended to prosper and increase, both in industry and in agriculture. The revisionists interpreted this evidence as a direct refutation of the Marxist theory of capitalist concentration, rather than as evidence of a short-term postfeudal rationalization of agrarian structure, which was simply postponing agrarian concentration. The perspective of a long-term survival of small properties led to a global revision of the theory of socioeconomic development and the acceptance by revisionists of small family agrarian properties as something that should be helped to survive. Because the only way to organize the small peasants effectively was to help them defend their economic interests, the typical tool that socialists developed to serve this interest was the cooperative society, and this defense often took place at the expense of the radical pushes of other segments of the socialist movement. Clearly, however, if peasants’ property was destined to last, to ignore the peasantry meant missing a political opportunity and, what was even more dangerous, leaving an additional chance to the class adversaries. The third political solution to the agrarian predicament, different from both of those offered by the orthodox Marxists and the socialistrevisionists, was the Leninist solution. Developed in reference to the still semifeudal countryside in which small peasant holdings were not yet consolidated, this solution attempted to exploit the precapitalist revolt potential of land-hungry peasants by jumping directly to a postcapitalist solution of the full socialization of land. This solved the problem by bypassing small peasants’ property with a single jump. This theoretical solution (the practical political solution that was adopted during and after the Russian Revolution was very different) was dependent on the existence of a potentially revolutionary situation that would radicalize the peasantry. It was hardly applicable to nonrevolutionary situations and was obviously totally ineffective as an electoral strategy. Thus, this line lost

CLEAVAGE STRUCTURES

469

any appeal it ever had for the Western socialist movement, including the Communist parties, once the radicalization period after World War I was over.110 This broad characterization of an orthodox-Marxist, a socialistrevisionist, and a Leninist approach to the peasant question captures the essence of the most important streams of socialist theory and thinking. However, it is less clear that it correctly represents the socialist party situation and attitude toward the national peasant and agricultural groups. When speciﬁc policies and practical solutions became necessary, political divisions emerged and many socialists started to recognize the basic difference between agriculture and factory industry, to have doubts about the economic superiority of large-scale farming, and, in many cases, to start offering different programmatic tickets for agriculture and industry. The agrarian question in fact posed a different problem-pressure in the different European countries according to (1) the weight of the independent peasantry in their respective agrarian structures; (2) the weight of dependent agricultural labor in the same structures; and (3) the prevalent form of political representation that dominated the peasant world. Not only were agrarian structures highly differentiated throughout Europe, but the peasants were also rarely the ‘‘potato sacks’’ that Marx had spoken of. They actively reacted in defense of their interests and their Weltanschauung. Moreover, other political forces had established representational linkages with the peasant world, often before socialists started to face their theoretical dilemmas. In order to understand the variation in the socialist attitude and potential relationship with the agricultural groups, we need to clarify the different types of agrarian structures and the different traditions of preand postsocialist political representation of these groups.

AGRARIAN STRUCTURE PROBLEM-PRESSURE: THE PEASANTRY AND AGRICULTURAL LABOR For a comparative characterization of the numerical weight of the independent peasantry, I have taken the average census values for the decades between the 1900s and the 1930s as a reference (Table 8.3), keeping in mind that the internal distinction within the agricultural employed population is particularly difﬁcult and subject to sudden revision on the basis of different accounting criteria. These ﬁgures give a broad idea of the different national weights, but they conceal important qualitative differ110

On the different socialist attitudes to the peasantry, see Tarrow (1972: 238–241, 265– 267).

470

THE POLITICAL MOBILIZATION OF THE LEFT

Table 8.3. Independent peasantry in Western Europe: basic features (mean values for the 1900 and 1939 censuses; family workers excluded)

Note: The man-land ratio is calculated by Dovring (1964: 60) as the ratio between male workers in agriculture and the agricultural land in hectares. It therefore does not refer to the independent agricultural labor force but to the total labor force. The ﬁgures reported here are the averages of the 1900, 1930, 1950, and 1960 decades. The Gini index of land distribution inequality for the 1960s is from Russett (1964). The index ranges from 0 (maximum equality) to 100 (maximum inequality).

ences in the structure of agricultural society. There are two polar types of agrarian society, labeled by Dahl ‘‘traditional peasant society’’ and ‘‘farmer society.’’ The difference lies in the degree of inequality in the distribution of land, which implies, in turn, an inequality of political resources. While the former has a high propensity for inequality, hierarchy, and political hegemony, the second is more egalitarian.111 The farmer society is characterized by (1) a high weight of independents over the active population; (2) a large number of independents over the active population in agriculture; and (3) patterns of egalitarian distribution of land. Combining these criteria, the best farmer society candidates are Sweden and Norway, where, since the late Middle Ages, primitive yeomanry had succeeded in holding its right; Denmark, where a profound and successful series of land reforms in the nineteenth century brought about the resurgence of a free peasant class; Belgium and Switzerland in continental Europe; and, to a lesser extent, France and Ireland. At the other extreme, Italy, Austria, and Germany appear to be best characterized as 111

Dahl (1971: 53–54).

CLEAVAGE STRUCTURES

471

more traditional peasant societies. Finland is an extremely difﬁcult case to characterize with these categories because the estate system, established as a consequence of the military frontier situation, was reinforced during the union with Russia. At the turn of the twentieth century, Finland had some of the features typical of several Eastern European countries, where there was a landlord class. However, after independence between 1921 and 1922, radical land reforms were carried out that quadrupled the number of independent farmers by 1938.112 However, even this speciﬁcation conceals a great variety of different relationships with the land and with types of agrarian contracts. The point is that the differences within rural societies are less functional than those in urban occupations, and they are based on the different types of relationships to the means of production, the foremost among them clearly being the right to own land. Since the conﬂict of interest is related primarily to the acquisition and distribution of possession, the political expression of such conﬂicts could not be ‘‘oriented to the creation of a new type of society, whatever it may be in modern revolutionary utopias, but to changing the roles in an existing stratiﬁcation system where the beneﬁciaries rather than the principles are challenged.’’113 Thus, each speciﬁc pattern of the right to rural property should be distinguished in a sociological perspective. The independents in agriculture should be clearly differentiated according to whether they are farm owners or tenants, and the latter further differentiated according to whether they pay a rent or share the proﬁts or crops. Both should be distinguished according to the size of the land, the capitalist commercial nature of the enterprise or of the traditional subsistence farms, whether the farmer hired labor or not, and so on, in a long chain of important distinctions that cross-cut each other. In a comprehensive treatment, Sorokin, Zimmerman, and Galpin proposed a stratiﬁed hierarchy of thirteen rural social groups, twelve of which concerned the independents.114 Linz has made a thorough attempt to link this stratiﬁcation scheme to patterns of electoral behavior, showing how these social groups generally tend to be rather conservatively oriented, but also underlining the potential radical orientation of small farmers and sharecroppers under certain conditions.115 From the viewpoint of the integration and mobilization of the socialist movement, Malefakis has proposed a drastic simpliﬁcation of the issue: In 112 113 114 115

Dovring (1964: 244–246). Linz (1976: 366). Sorokin, Zimmerman, and Galpin (1930: 362–370). See Linz (1976). It excludes, unfortunately, Northern Europe and Scandinavia.

472

THE POLITICAL MOBILIZATION OF THE LEFT

Eastern Europe, the peasantry was simply too large not to become the central concern of the socialist movement; in West-Central Europe, the peasantry was too small to be such a central concern; in Southern Europe the peasantry was large enough to pose important strategic problems to the socialist movement but not large enough to become its dominant concern.116 The central point concerning the problem-pressure for socialists, which was determined by size, is crucial, but the geographical characterization leaves out the Scandinavian countries and is probably questionable in some important cases. If it is likely that there is a central continental pattern characterized by a small presence of independents in agriculture, France does not belong to it, but instead resembles cases like those of Italy and Ireland, which had the maximum weight of independent peasantry; Scandinavian countries and Finland come next and Central European countries last (the United Kingdom is well known to stand alone in this respect). The weight of the independents in agriculture over the whole category of independents also presents great variations across Western European countries, and although there is some correspondence with the independents over the active population, this correspondence is not perfect (see Table 8.3). The problem of the role of the peasantry was therefore of minor political relevance in several countries. In the United Kindgom, its weight over the active population was low, the average farming unit was large, the pressure of labor over land consequently was low, and the productivity of the agricultural population was the highest in Europe (see Table 8.4). The agricultural problem did not exist, and the Labour Party was very rarely ever concerned with this issue. The situation was not very different in a number of other countries. In Belgium, Germany, the Netherlands, and Switzerland, the weight of agricultural independents was below 11% of the active population on average, and the gross domestic product (GDP) originating from the agricultural sector was roughly balanced with its numerical weight. In these cases, there were more agricultural independents within the overall group of independents (Table 8.3) than in the United Kingdom, and in Switzerland and the Netherlands there were more agricultural independents in the active agricultural population as such. On the whole, the small bourgeoisie tended to be more rural than was the case in the United Kingdom. These cases show similar levels of weight of the independent peasantry out of the active population, while they differ in the weights of this independent peasantry in the active agricultural population. In Belgium and Switzerland (to a lesser extent), the indepen116

Malefakis (1974).

CLEAVAGE STRUCTURES

473

Table 8.4. Productivity of the agrarian world (1960s data)

Source: Urwin (1980: 83).

dent peasantry dominated the countryside more than in the Netherlands, Germany, and the United Kingdom. From World War I on, the peasantry in these latter countries was not a major political problem. It is therefore not surprising that the theoretically minded and doctrinaire Germans Social Democrats were long reluctant to renounce some of the key tenets of orthodox Marxism and to expose their urban working-class strongholds to the potential contradiction of a small-peasantry policy and alliance. For a strongly organized, mainly urban, ideologically oriented working-class socialist movement, the cost was probably not worth the potential advantages. Only in 1927 did the party ofﬁcially accept the thesis that the law of concentration of capital did not operate in agriculture. However, even then, this had little or no impact on party policies and organizational efforts. The theoretical issue was, however, for outweighed by a pressing practical political problem in the countries where the numerical and economic weight of the peasantry was at least twice that in the previous cases. In the ﬁrst half of the nineteenth century in Ireland, Italy, and France, the independents in agriculture amounted to approximately one-quarter of the active population, while in Finland, Norway, Sweden, Denmark, and Austria, it ranged from 14% to 20% (Table 8.3). France and Italy made the

474

THE POLITICAL MOBILIZATION OF THE LEFT

most strenuous efforts to protect their peasantry and to insulate it from market competition. Particularly in France, protection laws were preserved as a basic creed of the Third Republic. The tariffs on foodstuffs there were the highest in Western Europe, at about 29%, equaled only by those of Austria-Hungary, compared to 24% in Sweden, 22% in Italy and Germany, and 15% in Switzerland.117 France was so successful that the percentage of agricultural independents over the active population remained more or less unchanged until after World War II. Agricultural independents were protected more than in other countries from falls in agricultural prices, by the maintenance of a higher proportion of labor over the land than would have been otherwise possible, and by the postponement of agricultural modernization throughout the ﬁrst half of the twentieth century, so that when these issues were faced in the 1950s and early 1960s, there was a strong reaction by the agricultural population.118 In the other countries, the decline of agricultural independents over the active population is less drastic and, even more important, does not produce a decline in agricultural independents over the active population in agriculture; on the contrary, this proportion tended to increase, not decline, with the development of small property within the population active in agriculture. In these countries, then (see the third column of Table 8.3), the independents in agriculture represented the majority of all independents in all sectors in this period, in sharp contrast with the situation in the United Kingdom, Germany, Belgium, the Netherlands, and Switzerland. To take the extreme cases, in the United Kingdom and Germany, the agricultural independents represented approximately one-quarter of all independents, while in Ireland, Finland, Italy, Norway, and Sweden they represented more than two-thirds. Finally, these countries have the least productive agricultural economies (see Table 8.4), in particular France, Ireland, Italy, and Austria, where there were particularly signiﬁcant negative differences between the agricultural population and the agricultural share of the GDP. In conclusion, the socialist movement in the Scandinavian countries, Finland, Ireland, Italy, and France could not afford to avoid the issues of the independent peasantry. Although the high proportion of the independents in agriculture determined the problem-pressure for an agrarian socialist strategy, the latter did not depend solely on this factor. What the socialists could or could not do with the independent peasantry was also dependent on the size of another agricultural group, whose interests and positions they regarded as 117 118

For these ﬁgures, see Table 1.3 in Tracy (1964: 25). For this characterization of the French agricultural policies, see Zeldin (1979: 173–185).

CLEAVAGE STRUCTURES

475

more compatible with their class appeal: the agricultural laborers. The weight of this category within the global working class of a country was particularly important. In Table 8.5 I report the proportion of agricultural laborers in the national working class for each country and its temporal evolution from the end of the nineteenth century. These ﬁgures are not available for the United Kingdom and Ireland, but they can be comparatively estimated as being well below the European mean for both cases. In the United Kingdom, agricultural workers represent two-thirds of the active population in agriculture (64.9%) in the period from 1920 to 1960, but the proportion of the active population in agriculture is so small that the corresponding weight of agricultural workers over the whole working class cannot be higher than 5%. In Ireland, on the other hand, the proportion of the active population in agriculture is very high, although the proportion of agricultural workers active in agriculture is the lowest in Europe (only 17.7% between 1920 and 1960), so that the proportion of agricultural workers over the whole working class is around 20% at most. For France, no data are available for the post–World War II decades, but they have been estimated at about 25% and 22% in the 1940s and 1950s, respectively.119 If we look at the average ﬁgures for the central period – 1900s to 1930s – we have four quite separate groups. At one extreme, in Belgium, Switzerland, and the United Kingdom, the weight of agricultural workers was insigniﬁcant and the working class can be described as urban and industrially integrated. Next come the cases of Germany, the Netherlands, Norway, Sweden, and probably Ireland, where the weight of the agricultural laborers in the total working class was between 15% and 25%; thus, it was not insigniﬁcant, although it was minor, and therefore it was not able to create important contradictions in the socialist policy toward the peasantry. Austria and Denmark are intermediate cases. The agricultural working class was around one-third of the total working class, making this group fairly crucial for the electoral and political destiny of the Social Democrats. Finally, in three countries – Italy, Finland, and, to a slightly lesser extent, France – agricultural laborers represented roughly one-third to one-half of the working class. Although, obviously, this agricultural working class was concentrated in the areas of large agricultural enterprises (Paris Basin, Herauld, Normandy, Brittany, the Val de Loire in France; the Po Valley and some southern regions, such as Apulia, Sicily, and Calabria in Italy), it inevitably represented a primary target for socialist recruitment and was such a quantitatively important section of the 119

See Bourquelot (1972: 533–556).

476

THE POLITICAL MOBILIZATION OF THE LEFT

Table 8.5. Weight of agricultural workers within the European working class

working class that the socialist strategy in these cases had to start from this stronghold and then move to address other agricultural groups.

THE REPRESENTATION OF THE AGRARIAN WORLD Let us now consider the problem-pressure that different agrarian structures created for the theoretical debate of the socialists, and even more so for their alliance strategies and policy initiatives, with the prevailing forms of political representation in the agricultural world. In Table 8.6, I have tried to synthesize the basic features of the political representation of the agrarian world, indicating the cases of successful or failed formation of an autonomous peasants’ party, its electoral strength and political orientation, and, ﬁnally, in the last column, a comment about how the political representation of the agricultural world was eventually stabilized. In the United Kingdom, the issue was irrelevant. As for Ireland, we have seen that the Irish labor movement, which was dominated by urban working-class unions, adopted what we may call a British perspective toward the agricultural world, although it existed in a country basically made up of relatively small farmers. Whatever one says about the unfavorable conditions that made labor penetration difﬁcult in the countryside, it is a fact that Irish labor history shows no signs of even the slightest ideological or organizational attempt to come to grips with the peasant question. For the other four countries, where both the working class and the bourgeoisie were predominantly urban (Belgium, the Netherlands, Swit-

CLEAVAGE STRUCTURES

477

Table 8.6. Self-representational mobilization of the agrarian world

Source: Adapted from Table 5.2 and Table 6.1 in Urwin (1980: 138–170).

zerland, and Germany), the problem-pressure for the socialists was relatively minor. Short-lived agrarian parties were founded in the Netherlands and in Germany, while, in Switzerland, a similar political experience lasted longer with the foundation of the Farmers’, Traders’, and Citizens’ Party in 1920. However, in all these countries, the predominant political representation of the peasants was gained by their respective Catholic movements, with about 50% of the peasants supporting the Catholic party. In Germany, notwithstanding the early and far-reaching theoretical debate about the agrarian question, the SPD never displayed any strong desire to attempt to mobilize rural areas. The revisionists also used the issue of capitalist concentration in the rural economy more as an example and as a tool to modify the urban and industrial strategy of the Social Democrats than as an argument for a clear policy to obtain support from the countryside. All attempts to form peasant parties from the 1890s on were very conservative in orientation, regional in scope, and unsuccessful or short-lived. About half of the German peasantry had been solidly linked to the Catholic Party since its foundation.120 120

To explain the failure of the SPD to attract more-extensive rural support, Michels argues that the highly status-conscious nature of both rural and urban workers vis-a`-vis each

478

THE POLITICAL MOBILIZATION OF THE LEFT

Among the Central European cases, the question of agrarian support was probably more central in the Netherlands. In the 1897 election, which followed the enlargement of the franchise, the Socialist Party concentrated its campaign in the two northern provinces of Groningen and Friesland, with a program that made far-reaching promises to small peasants and tenants, as well as appealing to agricultural workers. The party’s organizational support was particularly strong in these northern rural provinces, perhaps due to the high proportion of nonreligious people among the population. This initial organizational and electoral dependence on the northern regions considerably inﬂuenced the party programs and positions later on. The lack of social homogeneity of this support reinforced the idea that parts of the petit bourgeoisie, as well as Protestant and Catholic voters, could be won only by taking a cautious electoral course. This electoral compromise and parliamentary strategy was viewed as conforming to the existing politico-cultural conditions. However, it was exactly this that triggered the internal reaction of the Marxist left wing of the party, which eventually split. The attempt to form a Dutch agrarian party did not succeed in the 1920s, and 50% of the farmers remained loyal to the Catholic Party. In Austria, alongside a larger freeholder peasant groups, there were also regions with some dependent agricultural labor. A Landbund party emerged at the collapse of the empire but obtained little support in the 1918 elections. On the whole, Austrian peasants were so deeply and generally supportive of the Catholic Party and so profoundly linked to the Catholic subculture, while the socialist movement was so heavily urban and working class right from the beginning, that there was very little real political or policy debate over attitudes toward the small freeholders. A large peasantry was therefore mainly a problem of the European southern (France and Italy, but also Spain and Portugal, which are not included in our analysis) and northern periphery (Scandinavia and Finland). While Finland and Southern European countries all had a large agricultural labor force together with an urban/rural working class, the former feature was absent in Scandinavian countries. The main difference, however, concerns the pattern of political representation that separates all Northern countries from Southern ones. The Scandinavian peasantry enjoyed both

other made their political alliance impossible; Michels (1906). I lack comparative evidence to test this interpretation. It should be mentioned, however, that for France and Italy the late and sluggish industrialization is thought to have created a smaller cultural gap between the urban working class and the agricultural workers. Large parts of the Italian working class felt very close to their recently abandoned agricultural roots.

CLEAVAGE STRUCTURES

479

economic and political independence in the nineteenth century: Economically, the bases for a landed aristocracy were poor, as a large-scale manorial system was possible in only small parts of southern Sweden and in Denmark. In Norway, two-thirds of the country was owned by the peasants in 1815. The peasants’ socioeconomic position was secured through the allodial privilege, which guaranteed the lineage control of the farm. In addition, in large areas of these countries, the prevalent methods of cultivation made the exploitation of peasant labor almost impossible and unproﬁtable. The political independence of these peasants was established in Scandinavia very early on, with the presence of a fourth estate of autonomous peasant representation. Of these three countries, only in Denmark did the legal and economic subordination of peasants resemble that of Central Europe. However, fundamental land reform at the end of the eighteenth century multiplied the freeholds and gave rise to a large group of selfreliant small- and medium-sized freeholders. Paradoxically, the Danish peasants rapidly became the best organized and most economically and politically powerful peasant class in Scandinavia. As later imitated by Norwegian and Swedish peasants, they mobilized to different degrees against large landowners (natural enemies seen as the embodiment of the old order), the urban bourgeoisie in commerce, banking and industry (which controlled the basic ﬁnancial and banking infrastructure, regarded as exploitive), and the bureaucratic elite (the central culture in Norway, which denied them equal status and political inﬂuence, as in Denmark and Sweden). The impressive element of the Scandinavian farmers’ movements was their astonishing capacity for self-reliance and self-representation, leading to a high degree of independence from the urban ﬁnancial, commercial, and industrial bourgeoisie (establishing their own network of savings banks, cooperatives, etc.). Cultural and political consciousness should also be seen in the light of the very high level of mass education of the Scandinavian countries. In these conditions, the emergence of a workingclass movement was not seen as a challenge. Quite the contrary, in Denmark in particular, the rise of a working-class movement was actually seen as a welcome split in the urban front. For the Scandinavian Social Democrats, the ‘‘agrarian question’’ was therefore mainly a question of political and governmental coalition with a strong, self-representational movement that included the vast majority of farmers. The ‘‘predicament’’ of the agrarian question, if it existed at all, was soon overshadowed by practical considerations concerning the political alliances that could be set up with the freeholder movement. Although the original party policy statements in this policy area ran close to Marxist

480

THE POLITICAL MOBILIZATION OF THE LEFT

positions, by the 1890s in Denmark and by 1902 in Norway, party political action had distanced itself from them. In Denmark, the acceptance of small agricultural properties and their defense was qualiﬁed only by the idea that the land belonging to the church and other large properties should be handed over to the landless laborers or to the smallholders. In Norway, the party instead outlined an agricultural program in 1902 that openly accepted the existence of independent family farming, along with open-ﬁeld farming, and from then on attempted to increase its support among agricultural workers and small farmers on the basis of this platform. Sweden merits a more detailed discussion because of its paradigmatic proﬁle. The Swedish development on the agrarian program issue is indicative of greater theoretical liberty from its beginning. Early Swedish socialists took for granted that large-scale production and land concentration were inevitable. When the issue was rediscussed due to the German debate (around 1894), the party made it equally clear that no expropriation of land was to be envisaged, as socialism did not wish ‘‘to separate labor from the means of production.’’ This skillful formulation reconciled the goal of socialism (to avoid alienation by separating capital, land in this case, and labor) with an obviously revisionist stand (to let the peasants own their land). However, the party leadership was still divided about the necessity and feasibility of winning farmers over to socialism. Some of these (led by Hjalmar Branting) did not want to support small farmers against largescale farming at the cost of delaying or stopping land concentration; others (led by Danielsson Axel) were prepared to make an ideological compromise, arguing that the victory of socialism depended on having the support of the farmers. The dilemma was between having an active reform program in agriculture or supporting small-scale agriculture. From the productive point of view, large-scale agriculture was considered superior, but the idea was advanced that cooperation among small-scale farmers could also work. Around the end of the century, the party suggested that new small farmers should be encouraged, but that the land should remain the property of the state even if given under hereditary rights to the tenants. The idea that farmers and agricultural workers had the same interests in the face of capitalism spread after the franchise reforms of 1907–1909, but at the same time, it was becoming obvious that the process of concentration in agriculture was not going to take place, at least in the short term. In the 1911 program, formal Marxist ideology was maintained, but at the same time, a clear opening toward the agricultural population was made,121 121

A synthesis of the early agricultural debate in the SAP is presented in Simonson (1990: 98–100).

CLEAVAGE STRUCTURES

481

although it seems likely that throughout this period the SAP did not win much small-farmer support anyway (according to survey data from 1950 to 1970,122 only 10% of farmers voted for the SAP, while 50–70% voted for the Center [ex-Agrarian] Party). When peasant parties started to form in the 1910s,123 the agrarian question died out and was transformed into the question of what possible policy could be offered to the Agrarian Party in exchange for governmental coalition. The alliances between urban workers and peasants that gave hegemony to Scandinavian socialism were ﬁnalized in the 1930s: Denmark and Sweden in 1933 and Norway in 1935. This compromise between urban and rural interests took the same basic form, involving policy programs organized around (1) the expansion of public employment, support for public works, social welfare programs, unemployment relief, and prohibition or regulation of strikes and lockouts in exchange for (2) reduction of agricultural property taxes, reduction of agricultural interest rates, state help for more-costly loans, debt relief in general for agriculture, and various forms of state subsidies to farmers. In sum, the alliance was based on a common farmer/worker interest in policies to stimulate the economy, and it resulted in urban workers accepting higher food prices in exchange for peasant support for public intervention and public works.124 It is obvious that the most immediate basis for this alliance was the self-representative character of the Scandinavian peasant movement. If urban workers and peasants had been grouped into different political formations, such an exchange would have been unthinkable, as it would have jeopardized the immediate interest of other groups. Finland resembles the Scandinavian countries in that it developed an early (1906) and strong self-representational agrarian movement second only to that of Denmark. However, from the beginning, its internal divisions were more pronounced because of the more polarized class relationship in the countryside and also because of the stormy and radical political life of the Finnish republic until 1945. At its birth, the Finnish Socialist Party had been an advocate of smallholder interests, and the well-to-do farm owners had remained within the Conservative Party. In time, however, the party became more conservative and attracted the more prosperous farmers, but, as a consequence, it suffered frequent splits throughout 122 123

124

Korpi (1983: 93). I have taken the date of 1922 for the formation of the Agrarian Party. Note, however, that one agrarian party was founded in 1913 and another one in 1915 and the two parties merged in 1922. On the economics of these agreements, particularly in Sweden, see Winch (1966: 168– 176), Lewin (1988), and Pojas (1991: 64–74).

482

THE POLITICAL MOBILIZATION OF THE LEFT

the 1930s, most of which were characterized by a clear smallholder attraction.125 A formal Red–Green political alliance like the one in Scandinavia never materialized in Finland. This was due largely to the split of the Finnish left, but also to the fact that in northeast Finland rural support for the Communist Party was high among smallholders and traditionally the socialist and communist left had enjoyed the support of the agricultural laborers. It is therefore in the countries with a large peasant population, high class polarization of the countryside, and lack of political self-representation that the problem of the attitude to be taken on the agrarian question became crucial and, at the same time, highly divisive for the socialist movement. In Italy and France, the economic independence of the peasantry was weaker; at the same time, their political representation was fragmented and class-polarized.126 The possibility of forming strong, selfrepresentational agrarian movements is, in general, not linked just to the existence of a large population of independent farmers. Secondary but still important conditions concern the intensity of the urban–rural gap over the territory.127 For instance, one key characteristic of the Southern European agricultural area is its high degree of urbanization. Settlement patterns of large villages or agro-towns, with peasants working in faraway ﬁelds, imply greater urban–rural communication, easier cultural hegemony of the urban elite, easier penetration of urban parties (including clientele politics), generally less stratiﬁcation in urban–rural terms, and a lack of rural cultural alienation from the urban world. This generally leads to more urban–rural integration and a higher penetration of urban culture in the countryside. In Italy, in particular, the lack of early and radical land reform (the liberals feared the Catholic peasantry) and the piecemeal and delayed reforms up to the 1950s enabled the survival of a strong group of agricultural workers in the large-scale farming of the north; a strong group of tenants and sharecroppers in the smaller agricultural units of central Italy; and an even more complicated structure in the south. During uniﬁcation, many of the large feudal estates fell into the hands of the Southern urban bourgeoisie, and the provincial middle classes started to take control of large parts of the previous latifondi. Professionals and merchants began to buy small 125 126

127

Sa¨nkiaho (1971: 27–47) and Berglund and Lindstro¨m (1978: 55). For a synthetic summary of the historical evidence criticizing the myth of the French paysannerie as being created by the French Revolution, reinforced and further fragmented by the Napoleonic Code, and representing a solid and largely undifferentiated bloc with common interests and aspirations, see Wright (1964: 2–6). On the conditions for the formation of agrarian movements, see Urwin (1980: 161– 170).

CLEAVAGE STRUCTURES

483

pieces of land, which contributed, on one hand to the low efﬁciency of agriculture and, on the other, to explosive class relationships. In France, the attempt to form an agrarian party failed, as it had failed in forming a Catholic party. In 1927, the Agrarian Party, founded by an eccentric lyceum teacher – Gabriel Fleurant, who preferred to be known as Fleurant Agricola – failed to politically unify the rural world, which was torn apart by internal division and competition with other rural organizations such as the Dorge`re movement, with its conservative and corporativist ﬂavor. All the attempts made during the 1930s to form a political movement of the peasants were successfully checked by the other political forces, the Radicals in primis. Moreover, French farmers never engaged directly in politics. Even in the highest-density areas, their representatives were always village mayors, departmental councillors, deputies, and senators drawn from among the large landowners, the long-established bourgeoisie, or the new elite of the republican period (notaries, doctors, teachers, shopkeepers, civil servants, etc.) rather than working farmers. In France, in the absence of a Catholic point of reference, 50–70% of wealthy and large farmers favored the conservative128 party (curiously, but revealingly, called the Center of the Independents and of the Peasants), but at the same time, 40–50% of small farmers chose the Communist Party. The Italian agricultural world never even tried to form an autonomous political movement or achieve political representation. Instead, it delegated representation to other social groups and political elites. The Catholic Party obtained the support of about 50% of the small farmers, but some of these, including signiﬁcant numbers of tenants and sharecroppers, tended to be socialist and Communist in orientation. In both countries, of course, the large agricultural working class was politically represented predominantly by the left. There was, therefore, substantial rural support for the French and Italian left in general, and in particular for their communist wings. The main difference lay in the fact that, in France, the left took root mainly among small farmers, who were poor but owned their land; in Italy, the left vote grew mainly among tenant farmers (central Italy) and farm laborers.129 The socialist agrarian dilemma was increased by the fact that the socioeconomic conditions of the peasants were similar to those of the workers. At the end of the 1960s, the percentage of peasants describing 128

129

The Radicals, and in part also the socialist tradition of agrarian unionism, regarded the peasantry as a single class from which they excluded only the large landowners. Jollivert (1972: 88–89). Dogan (1967: 143–151).

484

THE POLITICAL MOBILIZATION OF THE LEFT

their families as living in ‘‘somewhat reduced means’’ was higher than the corresponding percentage among workers’ families only in Italy and France.130 Certainly, in terms of objective economic and educational conditions, they were worse off than the urban working class in these countries. The size of these agricultural groups, combined with their marginal economic position and high status and class polarization, clearly made them potential supporters of radical protest movements. At the same time, it made the question of their support, or at least of their political neutralization, a crucial survival issue for the socialist movement. The problem could not be avoided or put aside on account of its quantitative irrelevance, but neither could it be solved by political agreements and concessions, since these groups lacked any solid political self-representation. A closer look at the French socialist attitudes makes the point clear. Universal male suffrage and electoral interests pushed French politicians to take care of rural interests, and the electoral power of the peasants was increased by the electoral laws and by the distribution of the population. At least half of the electoral districts for the second chamber were predominantly rural. No more than one deputy in every four could afford to ignore agricultural interests and problems if he wanted electoral victory. The rural weight was even greater in the Senate, whose members were elected in departmental constituencies.131 All the French socialist programs and policy documents concerning the rural world, from the early Guesdesist Marxist orthodoxy,132 to Jaure`s’s attempt to revise this orientation at the end of the nineteenth century,133 to the position of the CGT,134 were characterized by preposterous ambiguities and internal contradictions, conﬁrming the historical fate of small property and, at the same time, presenting socialism as its defender. After the adoption of a more radical program in 1910, signaling the radicalization of the party in this period and a less prominent role for Jaure`s in shaping ofﬁcial party policy on agriculture, the party soon reverted to its traditional position.135 Again, when the socialists were in government in 1936 and 1945, they wanted to adapt agrarian structures to economic 130

131

132 133

134 135

Data are from the 1970 European Community Study, as elaborated by Lewis-Beck (1977: 453–452). Unfortunately, no comparable Finnish data are available in this form. Wright (1964: 213). Each of the small communes had a grand elector for the Senate, while Paris had only thirty. See Landauer (1961: 212–225). In the 1890s, Jaure`s became the main expert and exponent of agricultural problems. On his programs and positions, which Engels ferociously criticized in an article in the Neue Zeit in November 1894, see Wright (1964: 22–23) and Gratton (1972: 165). Kergoat (1990: 179). Zeldin (1979: 404–405) and Wright (1964: 217–218).

CLEAVAGE STRUCTURES

485

change without calling into question the foundations of the liberal economy and without touching the peasants’ property. They meant to give peasants total control of their economic sector, but policy measures remained marginal, limited to the creation of cooperatives for the collective utilization of machinery.136 The Communist Party’s position developed differently. In addition to debating different forms of value production in the countryside, the basic aim of the party was to establish an intermediate agrarian regime while waiting for the superiority of collective property and collective land management to be recognized. At the same time, considering the small peasantry as a potential ally of the urban working class in a revolutionary situation, the party continued to support all of its demands and claims vehemently, without overcommitting itself to a risky policy of cooperative development. This attitude was correctly deﬁned as ‘‘electoral poujadism deprived of economic viability considerations.’’ It was not signiﬁcantly modiﬁed until the middle of the 1960s. In France, the problem of the peasantry could not be ignored or set aside, given its considerable indirect political inﬂuence. However, this was not a question that could be resolved by bargaining and negotiation with their political representatives. What made the whole issue even more problematic for the class left was the persistence of a large, internally differentiated, and inefﬁcient agricultural sector. Obviously, it was risky to develop comprehensive reform and reorganization policies that might alienate large sections of this world without providing solid political alliance opportunities. However, the presence and weight of this sector still obliged the left to deal with this huge politico-electoral problem as best it could by offering the peasants much more than was envisaged in its doctrine and ideological orientation. This, in turn, weighed enormously on other policy domains. For instance, it continuously prevented the development, ideological use, and policy implementation (when in government) of comprehensive reform policies for the urban working class, all of which required the rationalization and modernization of the agricultural economic system. It can be argued that the problem of a relationship with a large, socially differentiated, and politically divided rural world – typical of the Mediterranean countries, but to which I also assimilate the Finnish situation – is at the root of the reformist policy failure of Southern socialism,137 of its historical role as a movement more of civic and secular reform than 136 137

Tavernier (1972: 127). On the modest economic achievements of the Popular Front in the ﬁeld of agrarian policy, see Wright (1955: 75–86).

486

THE POLITICAL MOBILIZATION OF THE LEFT

of economic reform, and of its deep internal divisions. Debates about alternative strategies and alliances, political as well as socioeconomic, were exacerbated by this unmanageable issue. I will come back to this problem in the next chapter where I discuss the organizational cohesion of the European class left.

CONCLUSION: THE RESULTING CLASS CLEAVAGE In Chapter 7, I divided the response of the established elite to the rising socialist movement into three types. The ﬁrst response was mainly, if not exclusively, repressive; it relied on high stateness resources and implied a high closure of the political system. The most typical examples of this response were the two Central European empires. The second response relied mainly on institutional obstacles; was usually pursued by low-stateness polities; and opposed new claimants by postponing regime liberalization and maintaining the institutional privileges of upper political chambers, monarchical intervention, and representational obstacles, although with some minor directly repressive means. This was typical of the Scandinavian countries. Finally, a response based mainly on political control was predominant in early liberalized countries such as the United Kingdom, Belgium, Switzerland, and the Netherlands. To a certain extent, a mixture of direct harassment, institutional obtrusion, and political control was present everywhere, but in these three groups the predominance of one type of response was clearer, while in the weakly consolidated liberal regimes, such as those of France and Italy, the various types of responses were all used according to the period and circumstances. Although level of stateness – interpreted as resource availability – and social interest diversiﬁcation – interpreted as established elite cohesion – were emphasized as the crucial determinants of the predominance of any one type of response, the latter also depended on the structuring of the party system and on the social and political isolation of the socialists. A purely repressive response required weak political organizations and weak structuring of the party system, just as a political control strategy required the opposite. If, for instance, in Belgium, the Netherlands, or Switzerland, socialists were encountered mainly in the electoral and parliamentary arenas, this was so because the established elite could rely on a variety of cleavages and alliances that granted them considerable control over these arenas. On the other hand, the use of direct repression and institutional intrusion by dominant groups was justiﬁed by their fear of being unable

CLEAVAGE STRUCTURES

487

or their unwillingness to use political means due to the inadequate political organization of liberal and conservative forces. Therefore, the national pattern of cleavage structuring and party system formation was crucial in shaping both the preferred response of the established elite and the character of the class cleavage. It thus determined different openings of the electoral market, different alliance opportunities, and different social bases for the latter. In turn, these elements were important in determining the levels of radicalization and class polarization of the whole socialist movement. I have focused on four main dimensions of cleavage structuring and party system consolidation: early liberalism and denominational, agricultural, and nationalistic mobilization, examining (1) the extent to which these processes reduced working- and lower-class solidarity and mobilization potential and (2) the conditions in which the choice of available political allies did or did not increase the internal ideological factionalism and organizational unity of the socialists. The relationship of early socialist movements with liberalism had two major consequences: It reduced or increased the social and political isolation of the socialist movement, and it anticipated or delayed its direct selfrepresentational efforts. In the countries where cooperation with a strong liberal political movement was early and protracted – the United Kingdom and Scandinavia – socialism was not politically isolated from the beginning. This enabled it to be less concerned with defensive organizational mobilization and, eventually, considerably delayed its autonomous entry into the political arena. The latter is particularly evident in Sweden, Norway, and the United Kingdom, where socialists did not enter politics until the beginning of the twentieth century. The same is not true of Denmark, where the early inﬂuence of German social democracy and the stormy relationship with the Venstre probably determined earlier efforts to create an autonomous electoral mobilization. Where a liberal ally was not available on the Continent, socialists had no choice but to try to organize their own electoral machine as soon as possible. However, liberalism was not a working-class or lower-class unity-dissipating force; instead, it was an alternative presocialist channel of representation. As such, cooperation with the liberals had no direct impact on the later mobilization capacity of the socialist parties. On the contrary, once this cooperation ended in the ﬁrst two decades of the twentieth century, the socialist movement grew much more rapidly than in other countries. In reality, the cooperation experience split the liberals more than it damaged the socialists. At the same time, the link between the early relationships with liberal forces and the main ideological orientation and internal ideological and organizational strife of the socialists should be underlined. The

488

THE POLITICAL MOBILIZATION OF THE LEFT

long experience of social and political alliances with liberal forces made the integration of the socialists into the political systems easier and earlier; it also avoided political isolation during liberalization and democratization struggles. At the opposite end of the continuum, the typical and early abdication of Austrian and German liberalism, for example, left their socialist movements not only in an authoritarian environment, but also in almost total political isolation. The consequence of this was to reinforce the orthodox doctrinaire stance of these parties. Moderate and reformist groups found very little external political support, and their arguments must have seemed extremely unconvincing given the circumstances. Was it reasonable for the party to face the internal ideological and organizational strife inevitably triggered by moderating and collaborative drives, with little hope that these would be rewarded by the payoff of solid, stable cooperation with disorganized and politically ambiguous potential allies? Retreat to the ‘‘centrist’’ defense of Marxist orthodoxy, was a winning response in a context where institutional closure was accompanied by political isolation. This argument is also convincing in regard to the eclectic and ideologically variegated nature of French and Italian socialism. In these cases, political liberalism remained weakly organized, divided into many groups and parliamentary factions, and never represented a solid potential ally for the socialist movement. The liberalized nature of the regime offered opportunities for cooperation, but breaks and conﬂicts were as frequent as rapprochement between the various factions of the liberal and socialist movements. So, relations with liberal forces became a contentious issue within the socialist movement. This ended by increasing the ideological strife within the movement about the pros and cons of cooperation and on several occasions destroyed its ideological and even organizational unity. The reverse is true of denominational mobilization. Religious political mobilization was the single most important factor reducing lower-class unity and solidarity, and it narrowed the potential electoral support of the socialists. It inﬂuenced both the social composition of the class cleavage and its overall mobilization capacity. The greater the politicization of religious identities, the stronger the organizational mobilization of such identities; and the earlier such mobilization occurred with respect to the timing of socialist mobilization, the more it was likely that a religious cleavage would preempt full mobilization along functional class lines, activating competing political alignments and narrowing the potential electoral market of the class cleavage. The extent of religious politicization in the second half of the nineteenth century and the beginning of the twentieth set the Scandinavian homogeneous Protestant and state-hegemonic church experiences – to

CLEAVAGE STRUCTURES

489

which the United Kingdom can be assimilated, notwithstanding its greater heterogeneity – clearly aside from the bulk of continental experiences, where religious identities were politicized everywhere. A second distinction that is relevant here is between the continental experiences characterized by religious heterogeneity, in which Catholic mobilization was of a reactive and emancipationist type, defending equality with other denominations or state denominational neutrality (typically in Germany, the Netherlands, and Switzerland), in contrast to that of homogeneously Catholic countries, where the church was historically associated with one establishment faction at least and where it defended traditional privileges from secular assault (e.g., Belgium, Austria, Italy, and France). The ﬁgures in Table 8.7 clearly show that the absence of politicization of religious issues in Protestant countries was a long-term asset for the electoral development of the left as much as the politicization of competing denominational identities in mixed countries was a liability. The difference between these two groups ranges from 7% to 9% in the periods before World War II and from 15% to 12% in the periods after it, and the size of the difference is systematic through time. In all mixed countries, Catholics started their political organization quite early – before or in parallel with the socialists, from a ghetto church position, and with considerable antiestablishment misgivings. Competition with the socialists for lowerclass support was clearly embedded in this position. In the Netherlands, socialist electoral penetration was already regionally differentiated according to the spread of religious feelings in the 1880s: Where religious inﬂuence was lower, socialist mobilization was stronger.138 The rapidly growing mobilization of religious unions in the period after 1906 (see Chapter 6, Figure 6.5) caused the socialists to feel that their trade union movement could not hope, in the foreseeable future, to make any deep inroads into large parts of the religiously organized working class, particularly in the south of the country, or into other Protestant bastions elsewhere. This resulted in an additional push toward a self-moderating position and political course that the party felt was the only one that would attract the religious workers.139 138

139

In particular, in the province of Friesland and in neighboring Groningen. In the 1890s, the center of gravity of early Dutch socialism shifted from the urban regions of the west to the rural ones of the north. Buiting (1990: 65–66) shows membership ﬁgures by country regions documenting this shift. The support of agricultural workers in northern agrarian regions was due to the agricultural crisis of the period. But the religious element also was particularly important. This is a particularly clear-cut case and example – although not the only one – of the fact that the moderating electoral strategy, rather than being responsible for the loss of working-class sections, was often developed precisely to attract working-class nonsocialist sectors. Even the 1913 refusal to accept portfolios in a Lib–Lab government was due

490

THE POLITICAL MOBILIZATION OF THE LEFT

Table 8.7. Electoral strength by religious composition

Note: Number of elections given in parentheses.

In German electoral history, there is equally strong evidence that Catholicism limited socialist electoral expansion dating back to the 1890s. A systematic gap in left support in Catholic regions compared with Protestant ones has been documented by various electoral analyses.140 In Switzerland, the inhibiting role of religion on class alignments varied markedly as a function of the linguistic community and interacted with the inhibiting role of national versus local identities.141 However, note in Table 8.8 that the nature of class alignments varied as much within linguistic communities as it did between them (for instance, the three Protestant Frenchspeaking cantons versus the Catholic French-speaking cantons). Therefore, linguistic differences may also be the result of the powerful interaction of religion and language. Whatever the case may be, both muted class mobilization.142 In conclusion, in countries with a dominant Protestant estab-

140

141

142

to the socialist fear of alienating potential workers organized by Catholics and Calvinists (Daalder 1981: 225). The ‘‘electoral trade-off’’ imagined by Przeworski – according to which moderating the working-class appeal in order to attract middle-class voters would eventually lead to the decline of working-class support – was not relevant for Dutch (and other) socialist movements in the pre–World War II periods. They wanted to attract Catholic workers, not the middle classes. It is debatable whether they would have been more successful in attracting Catholic workers with a radical class appeal rather than with a moderate, reformist one. For ecological data evidence, see Urwin and Aarebrot (1981: 255) and Rokkan, Urwin, Aarebrot, Malaba, and Sande (1987: 148–149). Ecological research by cantons has shown that the correlation between Christian Democrat support and the percentage of Catholics in the population is incredibly high (.930). Protestantism yields much less signiﬁcant associations, as the Protestant cantons divided their strength mainly among the other three national parties (radicals, liberals, socialists). The socialist vote is positively associated with Protestantism (r ⫽ .517) and urbanization (r ⫽ .745) as a result of the Catholic Party’s strength in Catholic cantons and also in rural areas (Catholicism coinciding to a large extent with rural areas); Rokkan, Urwin, Aarebrot, Malaba, and Sande (1987: 261–264). For further evidence that when language and religion compete with class they preempt the mobilization capacity of the latter see Lijphart (1979: 442–458).

CLEAVAGE STRUCTURES

491

Table 8.8. Proportion of left support by language and religion and by sense of identity in Switzerland

Note: Identity: national ⫽ Swiss; linguistic ⫽ Romand or German; cantonal ⫽ ‘‘genevois,’’ etc. Elaborated from Kerr (1974: 14) and (1987: 153).

lishment, the opposition and the culturally defensive politicization and organization of the Catholic minorities proved a powerful competitor to the socialists for lower-class support. The situation is more complex in homogeneous Catholic countries. The level of electoral mobilization of the left in these cases is intermediate between that of mixed and Protestant countries, but there is higher withingroup variation in mobilization potential than in either of the two other groups. With reference to the timing of organizational mobilization of the Catholic subcultures, a distinction can be made between the cases of Belgium and Ireland and those of Austria, Italy, and France. In the ﬁrst two, Catholic political mobilization was early, pervasive, and clearly anticipated by that of the socialists.143 Even though one could not formally identify in Ireland – as in Belgium – a ‘‘Catholic party,’’ in both countries Catholicism intertwined with nationalistic issues and was an early political mobilizing force because it represented the political identity of the new states as a secession from previous broader and Protestant-dominated polities. In contrast, in Italy and Austria, Catholic political organization was late and its main catalyzing factor was antisocialism. In France, the formation of a distinctive Catholic party failed between the world wars and was short-lived after World War II. In Figure 8.2, I have charted the decade-average level of electoral mobilization of the left continental countries, distinguishing four groups: (1) those that are homogeneously Protestant; (2) those that are religiously mixed; (3) those that are homogeneously Catholic with early Catholic politicization and organizational mobilization; and (4) those that are homogeneously Catholic with late, postsocialist organizational mobilization. 143

Note that Belgium is the only country in which Catholic trade unions were earlier and stronger than the left trade unions.

492

THE POLITICAL MOBILIZATION OF THE LEFT

Figure 8.2. Left electoral mobilization by timing and type of religious mobilization.

These groupings best discriminate differential levels of left electoral mobilization. It is interesting to note that between the 1880s and the 1910s, there are only modest differences among the groups from this point of view (while, of course, there are differences in the levels of electoral mobilization due to other factors not reﬂected by these groupings). It is in fact after this that left electoral histories diverge according to denominational context during the late 1920s and the 1930s. Thus, the left stagnates in mixed countries and in the homogeneously Catholic countries where Catholic political mobilization was early. The opposite is true for both Protestant countries and the Catholic countries where Catholic organizational mobilization was late and was an antisocialist reaction, and this is illustrated in the continued growth of the left in the following two to three decades. Presumably, this is the case for two reasons. On the one hand, this late denominational mobilization in Italy and Austria inherited the political representation of a historically weak political liberalism and also a politically discredited and never well organized conservatism and was thus less well able to challenge the socialist organizational subculture, by then

CLEAVAGE STRUCTURES

493

already well rooted. On the other hand, Catholicism in homogeneous Catholic countries was so historically characterized in reactionary, antiliberal, prodynastic terms that its political mobilization determined a fundamental anticlerical political realignment within the middle classes. The Catholic role in politics added radical, anticlerical, middle-class electoral potential to what it subtracted from the working- and lower-class vote. The stronger the religious hierarchies and their links with the masses, the larger the sections of the middle classes and the bourgeoisie that adopted an anticlerical position and were inclined to vote left.144 In so doing, Catholic political mobilization impinged on the social homogeneity of the class cleavage, and perhaps on the cohesion of the socialist movement, but not on its electoral potential. By the 1980s, the total left of these Catholic countries tended to be even stronger than that of the Protestant countries. Therefore, while denominational mobilization weakened working-class political homogeneity everywhere, reducing the social homogeneity of the class cleavage, it was not always a factor that contained the size of the left. Utilizing both the classvoting index145 for the 1950s–1960s and an index of ‘‘religious voting’’ constructed with the same logic (percent of churchgoers voting for religious parties ⫺ percent of nonchurchgoers voting for religious parties),146 Figure 8.3 shows that the main dividing line between the northern homogeneous Protestant countries and the mixed and homogeneous Catholic continental ones concerns the social homogeneity of the electoral support of the class left. The left parties that appear more able to attract nonworking-class, middle-class votes are those that are also unable to monopolize the representation of working-class constituencies. Finally, denominational political organizations were not easy partners for socialist movements. The relationship between Catholic parties and socialist parties was, with very few exceptions, highly unstable throughout the period we are discussing. Stable, although intermittent patterns of political alliance developed only in Belgium, the Netherlands, and Switzerland (the ‘‘magic formula’’). Such alliances either failed completely or were highly unstable and competitive in Italy between 1919 and 1924 and 144

145

146

Nor should one assume that this phenomenon is typical only of homogeneously Catholic countries. It is also present in the religiously mixed continental countries; cf. Irwin and Holsteyn (1989: 28). I take this as an indicator of the social homogeneity of the vote for the left. For a recent and updated discussion of Alford’s and Thomsen’s indices of class voting and for a detailed analysis of their over-time variation after World War II, see Nieuwbeerta (1995: 12–13, 16–18, 39–56). The data are drawn from Lijphart (1971).

494

THE POLITICAL MOBILIZATION OF THE LEFT

Figure 8.3. Scatterplot of religious and class voting indices (1950s–1960s). after World War II; in Germany and Austria between the world wars, probably with a fatal outcome for democracy; and in France during the Fourth Republic. The speciﬁc cross-class appeal of the Catholic movements created a continuous internal debate and dissension within the socialist camp about the choice between cooperation or adversarial attempts to split this denominational support along class lines.147 Concerning the agricultural population and its political representation, the predicament differed according to which agrarian structure the socialist movement was facing and the different forms of its political representation. Where a class of independent freeholders was present and agrarian mobili147

There were frequent attempts by the socialists to split their Catholic antagonists along socioeconomic lines. The most explicit attempt was made in Belgium in 1925, when the Flemish laborite wing of the Catholic Party joined the socialists in support of the government. The hopes of breaking the Catholic Party were not unrealistic, given its internal structure organized around corporate membership of several ‘‘estates’’ of Catholic labor, Flemish farmers (Boerenbond), and middle classes. However, the attempt failed. Historically, the socialists proved unable to split Catholic parties everywhere.

CLEAVAGE STRUCTURES

495

zation succeeded in representing the largest proportion of this group, the problem was less one of electoral support than of political alliance. A selfrepresentational agrarian movement was likely to accept Green–Red mutual support. Alliances with peasants were therefore political, based on the exchange of policies among self-representational social groups. Nondoctrinaire socialist parties cooperated with farmers’ movements to enact adequate legislation to help them out of the Depression. The socialists found in these movements an early substitute for the previous alliance with more encompassing liberal movements; were less politically isolated; and were not, given the predominantly egalitarian structure, involved in relationships of high class polarization in the rural areas. The social base of the class cleavage tended to be, and to become, even more socially homogeneous. While the agricultural world was unable to form a unitary selfrepresentational political movement, no political alliance with the left was possible, but the latter was still able to penetrate single sectors of the agrarian world. Doctrinaire socialist parties with rigid Marxist programs hostile to the farmer/owner, strongly antireligious, and overwhelmingly concerned with working-class issues and welfare did not attract farmers, particularly in strongly Catholic countries and regions (Germany, Austria). If the representation of the agrarian world was already ensured by denominational parties (Belgium and the Netherlands are the clearest cases; in Switzerland, representation was shared by Catholics and a noninclusive agrarian party), the socialist movement remained a predominantly urban movement in which political alliances were to be made with Catholic organizations, thus mixing the peasant issues with the denominational one. Although socialist penetration was possible in this case, it was minor and limited to small secular sectors and agricultural laborers. In other cases, the representation of the agrarian world was split among various forces. When the agrarian social structure was highly differentiated and relationships of class polarization prevailed among different agrarian social groups, Catholic, conservative, and/or agrarian noncomprehensive movements gained the support of about 50% of the peasantry, leaving the representation of other sectors and sections to other political forces. When the class polarization of rural areas was high – separating large owners and small peasants and the latter two from sharecroppers, tenants, and agricultural laborers – it generally produced the deep penetration of the socialist movement. This not only added to the lack of social homogeneity of the class cleavage, but also contributed to its strong radical position. Subsistence farmers (peasant consumption enterprise), tenants, and sharecroppers in nonreligiously committed areas – and where a relatively small peasant

496

THE POLITICAL MOBILIZATION OF THE LEFT

owner group coexisted with a number of large farmers – tended to support the radical protest movement of the left, joining their support to that of agricultural workers. In France, Italy, and Finland, the socialists’ involvement in the class conﬂict of the rural areas had momentous consequences for their ideological and organizational developments (see the next chapter) and is the feature that sets these cases more clearly apart from both the Scandinavian and the Central European patterns. Combining these dimensions, the European experience presents four major patterns from the point of view of the political environment that socialist movements faced between 1880 and 1930. The ‘‘most favorable’’ conditions were the combination of an early, stable, and protracted Lib– Lab alliance, which was then supplanted by an equally long-standing and stable Red–Green alliance with a self-represented peasant movement within a context of little or no political mobilization of religious groups and little or no class conﬂict in the rural areas. The ﬁrst-phase alliance was mainly oriented to the liberalization and democratization of the political regime.148 The later alliance focused mainly on common economic policy interests and set the stage for a pattern of political competition largely dominated by socioeconomic issues and alignments (nationalization in the 1920s, welfare programs in the 1930s and 1950s, sales taxation, etc.). From an ideological point of view, the concrete payoffs of the Lib–Lab alliance, the perceived need to avoid alienating the large rural population (with demands such as the nationalization of land) and to avoid the political and social isolation of the working class and the ideological weakness of laissez-faire and economic individualistic liberalism149 all favored early and consistent abandonment of an orthodox Marxist revolutionary doctrine and a gradualist position of compromise. At the same time, the absence of strong religious or ethnolinguistic cleavages (and the high level of literacy) made it easier to replicate the class structure in the party structure. This produced the most distinctive and inclusive class cleavage in the Scandinavian countries (see Figure 8.4 for the following discussion), with a political movement that incorporates about 80% of the workingclass vote and whose electorate is composed of about 75% of workers. The United Kingdom’s experience is a variation on this pattern; the main difference lies in the nature of the liberalism and its subtending social groups. In the United Kingdom, liberalism was an urban phenome148

149

Also thanks to the fact that Scandinavian liberalism was not historically linked to the laissez-faire ideologies of the urban bourgeoisie, but rather to the social reformist and democratic ethos of the peasant periphery. On these points, see Castles (1977: 25). It was not only the strong position of the independent peasantry that favored this ideological weakness, but also late industrialization and the strong bureaucratic traditions (in Denmark and Sweden).

CLEAVAGE STRUCTURES

497

Figure 8.4. Inclusiveness and distinctiveness of the class cleavage. Note: The ﬁgures for inclusiveness and distinctiveness are drawn from national surveys conducted in the 1950s and 1960s. Because there are considerable discrepancies among different sources in several cases, when necessary the different ﬁgures have been averaged. Source documentation is reported in the Data Appendix. non and could not give birth to an agrarian ally for labor once the original alliance ended in the ﬁrst two decades of the twentieth century. Therefore, while in Scandinavia the opposition of the central nation-building culture (essentially an urban elite of ofﬁcials and patricians) to the traditions of the periphery was reinforced by the opposition between the interests of the secondary sector and those of the primary sector, in the United Kingdom the two cut across each other, the central culture being upheld and reinforced by landed families. Labor fought alone against a Conservative Party that was able to fuse the landed and urban elites in a country deprived of an inﬂuential small peasantry and agricultural labor.150 The consequence 150

It is important to emphasize that the crucial passage to national politics was more the result of the Conservatives’ efforts. Around 1885, the Conservatives moved into the

498

THE POLITICAL MOBILIZATION OF THE LEFT

was that British labor was unable to achieve ideological and governmental hegemony equal to that of the Scandinavians. While it remains difﬁcult to explain why the British socialists never sought to consolidate a political alliance with the remnants of the liberals (and paid a disproportionate price for this in terms of continuity of governmental tenure), such an alliance would probably not have altered the cleavage structuring to any great extent. Finally, British class cleavage emerged as highly distinctive, similar to that in Scandinavia but considerably less inclusive, involving about 60% of the British working class in the 1950s and 1960s. This lower level of inclusiveness was due to the high working-class deferential vote for the dominant social circles represented by and in the Conservative Party, whose unblemished historical prestige ﬁnds no correspondence on the Continent151 or in Scandinavia. By the same token, the more complex conditions of the structuring of the class cleavage occurred in countries where liberalism and the liberal forces were politically weak and disorganized and offered no shield for the rising socialist movement, where denominational mobilization drastically dissipated working-class solidarity and where large agricultural sectors were characterized by deep class polarization in the countryside. The two best examples of this are Italy and France. In both, Catholicism soon acquired a distinct political identity and added to the social heterogeneity of the class cleavage by successfully acquiring the political support of the lower classes while losing, in exchange, the support of the radical anticlerical middle classes. The fragmented social structure of the agricultural world added further elements of social nonhomogeneity, making any dialogue with it as a whole impossible but permitting considerable inroads into some of its more socially radicalized sectors. The political opportunities opened and closed by the instability and divisions of the liberal forces, combined with the great heterogeneity of the social bases, ensured that almost any political strategy would be challenged by sections of the socialist movement and that almost any ideological variety would be present and appear convincing to different sectors of it: from trade unionist tendencies to revolutionary syndicalism, from ex cathedra reformism to Leninist. This synthesis was far from unsuccessful in overall electoral terms, but it did mean, of course, that ideological and organizational cohesion was

151

boroughs, and this led to a gradual merger of the landed interests with the urban business interests. The Conservatives, therefore, produced the crucial shift from a traditional cross-class territorial representation to a new emphasis on cross-local class representation in Britain. On working-class ‘‘Toryism’’, see McKenzie and Silver (1967) and (1968), Parkin (1967), and Taylor (1978).

CLEAVAGE STRUCTURES

499

impossible to achieve. These class cleavages remained below average in both inclusiveness and distinctiveness. It is, however, the latter – the low distinctiveness – that distinguishes them from the Central European cases. In other words, what is typical is not the inability of the socialist parties to incorporate more than approximately half of the working class, but rather their capacity to attract and recruit outside the working class. This was the key to their historical electoral success and, at the same time, the stigma of their being undifferentiated coalitions from the class point of view. Austria and Germany present a third model. In this model, a liberal option was unavailable in both the early and later stages. Nationalism and the national question continued to agitate and divide the liberal forces, while Catholicism mobilized in both countries and the agrarian world became politically incorporated, to a large extent, within denominational political movements. Yet, compared with Italy and France, the agrarian sectors of Austria and Germany were far less conspicuous and had a less class-polarized internal structure. The two political contexts were similar in terms of both institutional closure and sociopolitical isolation. Under these conditions, any challenge to Marxist orthodoxy presented by cooperation with the bourgeois political forces, or by the need to come to grips with an independent political movement representing the agrarian world, remained unlikely to succeed. By contrast, an orthodox Marxist stand was probably the most effective way of maintaining the organizational and ideological cohesion of the socialist movement, even though, between the world wars, it failed in Germany and was successful in Austria.152 The similarity of the cleavage-structuring context in these two countries ends here, however. In Germany, Catholics mobilized politically very early and structured their political organization as an anti-Protestantestablishment ghetto party in direct competition with the socialists for lower-class support. In Austria, monopoly Catholicism organized much later politically, more or less at the collapse of the imperial order. It became the only force of political defense for Catholics against anticlericals, but also for peasant interests as opposed to urban interests, dynastic and bureaucratic circles against republican groups, and the urban bourgeoisie 152

Note that the parties had different responsibilities in interwar politics. The Austrian socialists were ousted from the cabinet in 1920 and never came back, so during this period their opposition role and tactics were quite uncompromising and they could continue their class-based electoral tactics. German socialists were far more deeply involved in governmental cooperation with Catholics and other forces, which continuously fueled internal debate about tactics and strategy.

500

THE POLITICAL MOBILIZATION OF THE LEFT

and independent middle classes against one of the best-organized and most socially cohesive socialist movements. Under these circumstances, all these cleavage lines came to be superimposed on one another in Austria, while in Germany they continued to cut across each other more profoundly, particularly in Catholic areas and regions.153 Both parties had distinctive working-class proﬁles, but the level of inclusiveness was considerably different. The Austrian socialist movement included the vast majority of the working class at levels not much lower than those in Scandinavia, while the German socialist movement, which had more-forceful competition, remained at average European levels of inclusiveness, representing about 55% of the working class. The fourth model is represented by those cases where an early confrontation between liberal-conservative and denominational political groups shaped the political competition in liberalized regimes. The early pluralism within the established elite of countries such as Switzerland, Belgium, and the Netherlands eased the political entry of the socialist movement but did not offer any solid liberal point of reference for political alliances. The weight of the agricultural world and its interests was far less pronounced than in the Scandinavian and Southern European cases, so that socialism remained a primarily urban movement. At the same time, the politicization of denominational identities was early and pervasive in all cases, although with different degrees of territorialization (maximum in Switzerland and less pronounced in Belgium and the Netherlands). Not surprisingly, the class cleavage structuring was very noninclusive in all cases, with the socialists represting only 40–50% of the working class, but the considerable difference in distinctiveness differentiated Swiss socialism from the stronger Dutch and Belgian socialism (up to the 1960s). The combination of religious, linguistic, and cantonal identity as forces dissipating workingclass unity and solidarity made the Swiss socialist movement not only the weakest in Europe (after the Irish), but also the least distinctive and inclusive, with around 30% of left voters being workers and around 40% of workers voting left. Belgian and Dutch socialism, facing mainly denominational competition, maintained a more distinctive proﬁle (in the former, this declined sharply when the linguistic cleavage was mobilized organizationally in the late 1960s; my data in Figure 8.4 refer to the situation before this development). 153

To exemplify the peculiar Austrian overlapping of class and religious feelings and contrast it with the more common continental cutting across pattern, one may recall that the difference between the church attendance of the middle classes and the working classes is ⫹1 in Germany, ⫹7 in the Netherlands, ⫹9 in Belgium, ⫹14 in Italy, ⫹15 in France, and ⫹59 in Austria. See Lewis-Beck (1977: 456) and Houska (1985).

CLEAVAGE STRUCTURES

501

Finland and Ireland remain the more difﬁcult cases to interpret from this perspective. With other cases, they share the absence of a phase of liberal cooperation, but they mix features of different types. Ireland shows an extremely prolonged and pervasive role of nationalistic issues and movements and, at the same time, an equally persistent and pervasive role of Catholic political identity. Rural conﬂicts were solved with the creation of a large peasantry deprived, however, of any self-representational political organization much before the Socialist Party emerged. Labor was fundamentally unaffected by and uninterested in the agrarian world. The net effect was ludicrously low inclusiveness and extremely high distinctiveness – a working-class party that left the majority of the working class to be represented by other types of political movements, combining a Catholic proﬁle with a clear position on the national issue. Finland combines elements of the Scandinavian and Southern European cleavage-structuring patterns, although the socialist entry was not as easy as in the other Scandinavian countries and the class polarization of the countryside was far higher than in Scandinavia. This meant that the agrarian political movement was less inclusive than its Scandinavian counterpart. At the same time, the socialist movement had, as in Mediterranean France and Italy, considerable early support among sectors of the agrarian world, that is, among the large group of agricultural laborers, as well as tenants and even small farmers. Stable cooperation with the Agrarians Party was impossible in these conditions. Moreover, because the victory in the 1918 civil war was perceived by the Whites as a national liberation war against the Russian inﬂuence and by the Reds as a class war, antagonism was strengthened by the superimposition of national feelings over class ones. As a result, the cohesion and unity of the bourgeois forces in Finland between the world wars was far superior to that of the Scandinavian countries. The Finnish left therefore experienced the ideological and organizational tensions associated with this situation, but, at the same time, the absence of cultural identity-dissipating forces made the class cleavage resemble the Scandinavian model in terms of both inclusiveness and distinctiveness.154 154

For the similarity between Finnish and Scandinavian patterns of party support, see Markku (1980: 84–86).

9

THE COMMUNIST SPLIT: UNITED AND DIVIDED LEFTS n Chapter 2, the distinction between three different processes was argued: (1) working-class radicalization and revolutionary crisis; (2) shortterm socialist organizational and ideological disunity; and (3) long-term communist electoral success. The radicalization of working-class politics during and after World War I, and in some places even before, was a generalized phenomenon, although the organizational and ideological strife and splits affected the national socialist movements differently according to the prewar divisions, the country’s position during the war, and the socialist attitudes toward war efforts and governments. In Chapter 2, Tables 2.6, 2.7, and 2.8 have been compiled for comparative purposes, and a descriptive analysis of the level of fragmentation of the socialist movements has been made. Reference to these sections should be kept in mind while reading this chapter, which considers the overall interpretation of the long-term outcome of these processes. The explanation of the consolidation and the long-term electoral success of communism and the general constellation of forces that can explain the difference between united and divided lefts, require, in my view, a broader and more long-term interpretive framework. This problem is the central theme of this chapter.

I

INTERPRETATION OF COMMUNIST SUCCESS OR FAILURE The modalities of the communist splits documented in Chapter 2 bear little relation to the more long-term electoral and organizational success of 502

THE COMMUNIST SPLIT

503

the communist movements, which instead requires a more general analysis. In three out of the four cases where communism made important electoral inroads, this took place in an unstable liberal–democratic environment.1 Due to the intertwining of the communist question with national integrity and independence, the party was banned in Finland for almost the entire interwar period bracketed by the two civil wars. In Germany, the communist movement was annihilated by Hitler, and after World War II its disappearance in West Germany may well be the result of the partition of the country and the creation of a communist regime in the Eastern part. In Italy, the party remained clandestine during twenty years of fascism, and was resuscitated thanks to the liberation war and the resistance movement. One might say that these are special cases when compared with the less stormy liberal democratic development in other European countries. Consequently, it may be tempting to search for the sources of communist success in these geopolitical situations. However, it should not be forgotten that electoral communism developed in a liberal environment in France, while it failed to do so completely in Austria, in a context as unstable as that of Germany and Finland. Similarly, it has been argued that the Great War, its development, conditions, and outcomes, had a crucial impact on the chances of a radical Communist movement.2 Given that all post–World War I and post– World War II radical left oppositions were united in their condemnation of the support and cooperation shown by their bourgeoisies and socialist and trade union movements in the countries at war, the latter forces were not subject to the same recriminations in the countries that were not at war. Indeed, the low success of the communist movements in these countries owed much to the more patriotic stands of their bourgeoisies during the wars and occupations.3 This contrasts very strongly indeed with the recurring theme of the ‘‘betrayal’’ of the national bourgeoisie that pushed workers and peasants into unnecessary wars without satisfying the requests of those who had sustained the cost of the efforts (typically, France and Italy and, even more so, Germany after World War I) or, worse, leading to defeat and shame (typically again, Italian fascism, France’s Vichy regime, German Nazism, and the Finnish alliance with Germany in 1941– 1944). The low political and social legitimacy of the bourgeois and domi1

2 3

On the implications of their adaptation to electoral politics see Greene (1973); and for the role of leadership in this adaptation, see Tannahill (1976). Kriegel (1969: 184). A detailed analysis of how war affected the electoral development of communist parties throughout Europe is presented in Timoten (1973). I thank P. Hannivaara for his help in reading this article. See also Rice (1973).

504

THE POLITICAL MOBILIZATION OF THE LEFT

nant circles at the end of the two world wars justiﬁed and favored a radical appeal. However, that a warring and victorious country like France experienced radical splits, while a warring and defeated one (with regime change and territorial retrenchment) like Austria did not, requires a more complex set of explaining variables. On the whole, it can be concluded that the role of a country in war is an important intervening factor that added or muted preexisting and more profound socioeconomic, organizational, and politico-institutional preconditions. For these reasons, I will try to place the interpretation within a broader framework, which will not refer to these speciﬁc circumstances but will deal with them as instances of more-encompassing underlying processes. Interpretations of communist strength and success are more numerous than the number of cases on which to test them. The major difference between them lies in those interpretations that refer to individual-level variables and those that refer to systemic or environmental features. Individual and system characteristics can often be translated into each other. Thus, a focus on the distribution of individual motivational, ideological, or social features places the macroprocesses that produced them in the background in the same way that systemic and environmental features need to postulate and induce the individual reactions associated with them that they cannot observe directly. Let me explain why I have disregarded those interpretations of communism that consider it as a typical mass movement. Sociopsychological interpretations of radicalism, and lower-class left radicalism in particular, have often invoked the variable of ‘‘mass society’’ theory.4 The argument is that the decline in social and group links in communities, the associated need for new social links, and the personal malaise and frustration associated with the intense, rapid social dislocation at the peak points of industrialization set favorable conditions for lower-class radicalism, and, in particular, for the creation of mass movements such as fascism and communism. The support for radical communist-inclined working-class movements is a typical example of such ‘‘mass behavior,’’ characterized more by emotional, noncognitive processes or psychosocial grievances – with extremist, antidemocratic, and intemperate correlates – than by rational responses to class needs and political-action choices. Within this debate, it is, ﬁrst, impossible to test these individual-level interpretations with the kind of data utilized here. The test can only be indirect, considering the background variables that identify the condition of a mass society at the structural level. Second, the emphasis on the 4

Kornhauser (1959: 21–113).

THE COMMUNIST SPLIT

505

psychological and attitudinal characteristics of the supporters of radical and communist movements has produced considerable research and much discussion that, in my opinion, have failed to produce convincing supporting evidence. Kornhauser’s speciﬁc interpretation of communism and fascism as ‘‘mass movements,’’ a typical expression of the social and political alienation of a ‘‘mass society,’’5 has been subject to fairly systematic tests in different contexts, and most of the results have been critical of the theory. Studies devoted to communist militants and voters in France, Italy, India, and Chile have come up with results opposite to those expected on the basis of a mass movement interpretation of communism. Studies on Chilean radicalism have also concluded that the most radical left tendencies were not manifested by the socially isolated, least educated, least informed, or low-participating citizens; nor did upwardly or downwardly mobile lower-class individuals or those with status inconsistencies show any greater radical left proclivity. On the contrary, an association between left radicalism and speciﬁc social and political cognitive factors has emerged, linked to the perception that the failure to improve an individual’s conditions derived from external constraints rather than from her or his weakness or other transcendental factors.6 Other research has shown that supporters of radical movements were not necessarily less likely to be involved in local community associations. On the contrary, they took a greater interest in politics and government, had a greater sense of political efﬁcacy, discussed politics more, and were more knowledgeable. Thus, rather than ﬁnding an association between radicalism and the lack of ties, estrangement, a sense of helplessness, and ignorance, the reverse seems to be the case.7 Other criticisms have accumulated in several countries that argue against the ‘‘protest’’ hypotheses that most communist voters in democratic systems are economically deprived, socially isolated, politically vulnerable to extremist parties, and generally alienated from the conventional norms of their society.8 All these contributions conclude that the Communist labor population is usually not a mass population, and they suggest the limited usefulness of an emphasis on the psychological and expressive roots of working-class radicalism, or on its erratic, irrational, or noninstrumental character. A certain amount of supporting evidence however, has, also been 5

6 7

8

Kornhauser (1959: 179–180) recognizes a limit to his theory as far as the different class nature of seemingly similar mass movements is concerned. Portes (1970), (1971a) and (1971b). Halebsky (1976: 139, 146–147). Similar conclusions were reached for France and Italy: Hamilton (1967: 47 and footnote 10) and Tarrow (1967). See Greene (1971).

506

THE POLITICAL MOBILIZATION OF THE LEFT

accumulated in favor of some versions of the mass society theory. Lipset devised the category of ‘‘working-class authoritarianism,’’ stressing that low levels of education, patterns of family life, and high psychological and economic insecurity foster traits of an authoritarian personality, intolerance, prejudice, and temperamental thinking among the lower classes. Working-class conservatism on cultural issues, combined with their radicalism on economic issues, are not unlikely to make them supporters of antidemocratic movements of the left as well as of the right.9 For example, the most extensive ecological study of Swedish communist electoral support has emphasized the long-term stability of their strongholds in regions of low religious practice and no farming in the north of the country. The main interpreting factor was ‘‘social isolation,’’ by which was meant both geographical isolation (on the spatial periphery outside national communication networks), economic underdevelopment and nonintegration into the national system, and occupational isolation – immigrant labor and nonresidential labor populations.10 Again, Allardt, in his classic studies on Finnish communism, distinguishes between (1) traditional communism in the industrial centers of southern and western Finland and (2) ‘‘backwoods’’ communism in rural areas of northern and eastern Finland (after World War II). His interpretation of the latter emphasizes social anomie, alienation from the political system, and legitimacy crises resulting from rapid changes in rural regions. He interprets the communist support more as an emotional protest than as instrumentally oriented political behavior.11 These arguments have been criticized by Korpi in several works. According to him, some crucial variables interpreted by Lipset as supporting his theory are probably spurious and caused by demographic factors. Korpi’s argument also fails to support Allardt’s theory of anomie. In his opinion, there are socioeconomic reasons that justify and explain the communist vote, and it is consequently easier to interpret available data on the communist orientation among workers in Western Europe in terms of a simple model that assumes rational, self-interested voters. The communist vote is neither partially nor wholly irrational. Voting is determined by rational choices on the basis of the relative utility of the policies pursued by parties for the workers.12 9 10 11 12

Lipset (1959: 482–501). Rydenfeld (1954). I rely on the summary available in Lafferty (1974: 43–58). Allardt (1964) and (1970). Korpi (1971).

THE COMMUNIST SPLIT

507

The debate is sufﬁcient to justify leaving aside this line of research. I am rather uneasy about arguments concerning rationality versus irrationality of speciﬁc political behavior and contrasting emotional reactions to instrumental calculations.13 In the context of the approach followed so far, I will be looking mainly for a syndrome of system features that best accounts for major cross-system differences in communist development. I will discuss the way in which such differences may have inﬂuenced the structure of individual and group beliefs, perceptions, and aspirations, but I will not pursue a direct test of the correspondence of the speciﬁc distribution of such beliefs, perceptions, and aspirations and the communist support at the individual level. I will frame the problem by asking the following questions: Under which socioeconomic, organizational, and politico-institutional conditions was an appeal of radical opposition to the dominant economic and political orders most likely to be successful? In other words, when was it most likely to be regarded as a convincing explanation of their condition and as a convincing perspective for changing such a condition by sections of the lower classes? I will concentrate on the differences among cases, trying to identify which syndrome of systemic factors best account for the successful split of the European left. The hypotheses are so many and so strongly interlinked that it is difﬁcult to isolate them from each other. At the same time, it is impossible to combine too many of them without falling into a case-speciﬁc interpretation that is not applicable outside of its own context. I have, therefore, organized the argument into four subsections before extracting the most convincing factors and linking them together: 1. Interpretations relating communist success or failure to the process, pattern, tempo, and type of social mobilization processes and cultural standardization processes. 2. Interpretations relating communist success or failure to organizational features of the socialist movement. 3. Interpretations relating communist success or failure to the process of institutional and political integration of the socialist movement. 4. Interpretations relating communist success or failure to the development of the cleavage system. 13

Explanations that stress aspects of increased class consciousness or greater solidarity among workers have not been discussed because of their implicit circularity. Cf. Tilly, Tilly, and Tilly (1975: 8).

508

THE POLITICAL MOBILIZATION OF THE LEFT

COMMUNISM, SOCIALISM, AND PATTERNS OF SOCIAL MOBILIZATION AND ECONOMIC DEVELOPMENT Possibly the oldest of interpretations of the variation in communist support relates it to different patterns of economic development and industrialization. One thesis is that communist support ﬂourishes not in backward but in late-industrializing societies, where the lateness is not absolute but relative to the earlycomers of the Industrial Revolution. At the end of the 1960s, Benjamin and Kautsky discussed communist support in a worldwide perspective, making the distinction between ﬁve types of society.14 According to them, in the two most advanced types, namely, European societies and the United States, communist support was stronger in the (relative) latecomers, where late and slow industrialization had not penetrated the entire economy and the core working class was a minority group. Again, according to them, in this situation the overall weakness of industrial labor and of trade union organizations obliged them to face a solid and majoritarian antilabor coalition of both industrial and preindustrial property groups. In addition, sections of the peasants who were propertyless or propertied (but with relatively small holdings) and sections of the old middle classes (shopkeepers and artisans), threatened by big industry and big commerce, were tempted by defensive anticapitalist stands. Under these conditions, coalitions of sectors of the intelligentsia frustrated by incomplete modernization, sectors of the working class, and sectors of protest voters may be conducive to widespread support for communist movements.15 Slightly contradicting with this line of argument are the theses that regard intense radicalization (and Communism as the most stable organizational expression of it) as resulting mainly from situations of rapid and accelerated socioeconomic change and from the social dislocation that follows. This idea was forcefully argued in relation to the Scandinavian countries by Edward Bull in his explanation of why the intense radicalization of the Norwegian Socialist Party was absent in Sweden and Denmark.16 This general idea is also implicit in Kornhauser’s theory: Rapid 14 15

16

Benjamin and Kautsky (1968). In contrast with these characters, one can determine the features of the ﬁfth type, the highly industrialized countries, where the industrial working class is strong and integrated. See Benjamin and Kautsky (1968). Bull (1922). His analysis includes other factors: (1) the more open and democratic nature of the Norwegian socialist movement compared to the more centralized Swedish and Danish ones; (2) the weaker parliamentary representation of Norwegian labor; and (3)

THE COMMUNIST SPLIT

509

industrialization and urbanization produce strong support for communist movements because of the difﬁculty of organizing and socially integrating the new workers coming from the countryside and their difﬁculties in adapting to industrial and urban life. Neither Bull’s nor Kornhauser’s generalizations spell out accurately, however, the links between rapid economic development, corresponding intense social dislocation, and political radicalism. An explicit hypothesis can be found in Thorstein Veblen’s work about the sources of German nationalism.17 His basic idea is that overly rapid social displacement forces people into what he calls an ‘‘obsolescent cultural situation.’’ They ﬁnd themselves in a new position while, at the same time, they have not forgotten their previous status and have not yet been exposed to the elements of the new situation long enough to assimilate them. This incapacity to forget their previous status and situation, which is linked to the length of exposure to their new position, produces reactions that Veblen identiﬁed in radical nationalism. However, one could use this argument as a complement to the theory of the radicalizing effects of rapid social dislocation. The picture becomes more complicated if one considers Lorwin’s propositions on the relationship between economic development and radicalism, largely derived from his studies of the Belgian and French labor movements.18 He draws the following conclusions: (1) rapid growth and economic development in the early stages of industrialization generate radical protests resulting from dislocations, sacriﬁces, and difﬁculty of adaptation. At the same time, (2) continued economic growth leads to the quelling of this protest and satisfaction of needs. However, (3) protest attitudes may persist beyond the economic conditions that provoked them. Finally, (4) sluggish economic growth may generate the most profound and lasting protest as a result of the inability to provide economic well-being and social justice adequate enough to match the aspirations created by the beginning of industrialization and the failure of the economic elite to generate conﬁdence. Interrupted, sluggish, or incomplete industrialization and social transformation seem to be the most favorable conditions for a radical class movement here. We can relate this line of thinking to that of other scholars, too. For example, Schumpeter19 advanced the thesis that the radicalization of the working class and the birth of strong communist

17

18 19

the lack of a national leadership in Norway like that exercised by Branting in Sweden and by Staunting in Denmark. Veblen (1994: 21–43, 248–250). This theoretical scheme pervades the entire interpretation of the German imperial experience. Lorwin (1967b: 58–72). In Schumpeter (1951).

510

THE POLITICAL MOBILIZATION OF THE LEFT

parties are most likely to occur in the countries and contexts where the urban and rural elites form a coalition against industrial development. Moore’s argument is that the alliance of strong noncapitalist rural interests with weak urban industrial interests implies incomplete industrialization, as well as permanent pockets of backwardness and social groups that are weak and politically unstable.20 Although they have several interpretive points that can be reconciled, these arguments lead to quite different conclusions. Communism emerged in the 1920s, consolidated in the 1930s, and stabilized – if and where it did – in the 1950s. It makes a great deal of difference whether we argue that it had the best chance of appealing in the most backward European economies and social structures or, alternatively, in those that were latecomers but were, at the same time, experiencing high and accelerated rates of change or, alternatively, in those where sluggish and stagnant economic and social change prevailed. These theories agree about one end of the continuum: Early starting, smoothly progressing, and successfully completed industrialization processes generate unfavorable conditions for a radical working-class response. They do not agree about the other end of the continuum – which conditions are more favorable to this response: sheer backwardness, late and rapid socioeconomic change, or interrupted growth and sluggish economic development? If we relate the characterization of the patterns of industrialization illustrated in Chapter 3 to the strength of communism after World War I (Table 9.1), the cases of strong and persistent communist movements all belong to the group of countries where the industrial society (more active in industry than in agriculture) arrived late and did not permeate the social structure of the society (industrial predominance over the service sector was short or nonexistent). However, two countries where communism never made any inroads (Austria and Ireland) and the Scandinavian countries, where it gained only moderate support, belong to this group too. Moreover, Germany – where communism was important during the Weimar Republic – belongs to the countries with both an early and a prolonged industrial society. We can look at the same problem from a different perspective by comparing the levels of association between communist and socialist support and the main aspects that deﬁne the sociostructural characteristics of the societies. In Table 9.2, these data are reported for the 1918–1980 20

Moore (1966). Moore applies this argument to the chances of democratic consolidation versus authoritarian rule, but it is indeed applicable at a lower level to the radicalization and communist development problems.

THE COMMUNIST SPLIT

511

Table 9.1. Timing and length of industrial society and Communist support

Note: Normal ⫽ insigniﬁcant communist movement Italics ⫽ small communist movement Bold ⫽ strong communist movement Underlined ⫽ Germany: strong in interwar years; nonexistent after World War II

Table 9.2. Social structure and communist/socialist electoral support (1918– 1980 elections)

Note: Number of elections given in parentheses.

512

THE POLITICAL MOBILIZATION OF THE LEFT

elections. Specular associations are systematic. Social democratic strength is associated positively with all the variables that deﬁne an industrial and advanced economic structure and negatively with all those that identify a preindustrial or incomplete industrial society. More than was the case for the total left, social democracy appears to be the typical product of industrial society, while communism appears to be the typical product of unsuccessful or delayed industrialization, as a more agrarian and traditional society phenomenon.21 The association between the social-mobilization model and the left vote discussed in Chapter 3 is the combined result of the different nature of its two components. If applied to the socialist left alone, the social mobilization model performs much better. The latter considerations point to a characterization of electoral Communism as a phenomenon that ﬁnds favorable conditions in the low integration of the working class in two senses: its internal heterogeneity and its low industrial concentration. We can ﬁnd support for this idea by looking at the communist and socialist support (1) in different subsectors of the working class and (2) in relation to the level of concentration of industry (Table 9.3). In the interwar period, the internal composition of the working class does not seem to be associated with the level of voting for either of the two components of the left. This is probably due to the fact that in this period Germany had both strong communism and a highly homogeneous industrial working class. In the subsequent periods, however, it is clear that socialist support is lower in the context of a large agricultural dependent labor force, while communist support is higher in these cases. France, Finland, and Italy have the highest proportion of dependent agricultural laborers in their working class. This is not true for Germany, however, and even Denmark and Austria have a slightly higher than average agricultural working class but are not examples of a deep postwar split. Finally, looking at the factory concentration of the industrial sector of the working class in terms of the ratio of employers to employees in industry (Table 3.11 in Chapter 3), one ﬁnds that France and Italy are the countries with the lowest industrial concentration by far. On average, in the 1920s and 1930s, there was one employer for every four workers in the industrial sector, while in countries such as the United Kingdom, this ratio ranged from one to fourteen to one to twenty. However, Finland and Germany do not ﬁt into this pattern, while in Denmark, Norway, and 21

These associations are not due to historical multicolinearity. Repeating the correlation analysis by each decade from the 1920s to the 1970s yields higher and similar sign correlations.

THE COMMUNIST SPLIT

513

Table 9.3. Working-class composition and support of left parties

Switzerland, industrial plants were on average small and the working class was not highly concentrated in industrial areas. Let us now turn our attention to the rhythm and tempo of the socioeconomic change rather than to its level. The British case is often cited as a case of early radical working-class responses triggered by the social disturbances of industrialization processes. The social and economic changes of early industrialization certainly stimulated vigorous protests and radical responses, but Luddism and Chartism were defeated or collapsed long before suffrage was granted, and by the time Marxism had developed an articulate theory that could have served the radicalism of the early British working class, the impetus for militancy had abated, that is, the destabilizing period of the Industrial Revolution was over.22 The British example has given credence to the idea that radical responses are typical of the early and accelerated stages of industrialization; at the same time, if this phase precedes democratization, radical responses are likely to be absorbed and continued growth is likely to channel the newly formed working-class demands differently. This implicitly means that only when democratization and mass organizational and ideological development coincide with the most intense destabilizing social dislocation of the takeoff phase of the Industrial Revolution will the radical responses be channeled into radical communist movements. Generalizing the argument, it is to be expected that countries experiencing the highest rate of socioeconomic change at the later stage – and 22

Mair (1979: 162).

514

THE POLITICAL MOBILIZATION OF THE LEFT

particularly at the stage of mass political development between the 1900s and the 1930s – will offer the most favorable opportunities for a political radicalization of the protesters. We can associate the level of the communist vote with the rate of change in the main indicator of socioeconomic transformation by decade. The results are so disappointing for the ‘‘rapid socioeconomic change/social dislocation/radical political response’’ hypothesis that they are not worth reporting in a table. All the correlations are negative with the communist vote and positive with the socialist vote. If the socioeconomic variables are lagged by one unit, so that the level of communist support in the 1920s and 1930s is related to the rate of socioeconomic change in the 1910s and 1920s, the situation does not change: Positive associations with the socialist vote grow, while those with the communist vote remain negative or insigniﬁcant. This seems to suggest that rapid change fosters a socialist (moderate) response rather than a communist (radical) response. If we repeat this analysis at the level of individual countries, relating the mean rate of socioeconomic change to the mean electoral success of the communist splits, as is done in Table 9.4 the data continue to offer little support for the thesis. The most successful communist parties in the 1920s are not to be found in contexts of previous rapid and accelerated socioeconomic transformation. Quite the contrary, Italy, France, and, to a lesser extent, Finland are characterized by the lowest rate of growth in this period. In the second group of periods – the 1930s, 1940s, and 1950s – the situation does not change. Sweden and Austria maintain their position of high-growth countries, but Finland bypasses them in every category except for manufacturing output. The Finnish population employed in industry and the industrial working class grow at a rate either twice or three times higher than the European average in these three decades. France, however, remains the country with the lowest rate of growth, and Italy shows only average ﬁgures. If anything, these data – and particularly those for the ﬁrst three decades of the century, which are the most important – show that communist electoral strength was highest in the least dynamic societies of Europe. If the United Kingdom is taken as the example of the continuous growth that weakens early radical protest, France and Italy can easily be regarded as good candidates for the thesis that sluggish and incomplete industrialization favors radical reactions and appeal. They are the best examples of the syndrome that links (1) sluggish economic development and little dynamism in economic growth with (2) predominantly smallscale entrepreneurship lacking initiative, Malthusian-oriented, demanding state protection against labor and competition, niggardly in sharing bene-

THE COMMUNIST SPLIT

515

Table 9.4. Rate of socioeconomic change and communist strength

Note: Cases of strong electoral communism given in italics.

ﬁts, and stubborn to maintain authority in the workplace; with (3) the persistence of large agricultural sectors and high agricultural prices, pockets of backwardness and protectionism, and major inequalities between poorer agricultural areas and industrial regions; and with (4) labor movement pessimism and a persistent sense of injustice, generating little hope for immediate improvements. The result of the French Third Republic’s political alliance of the small rural and industrial sectors against the labor movement and big industry, in a country that had originally led early industrial development, is from this point of view not dissimilar to the Italian latecomer situation of asphyxiated capitalism and marked territorial and social inequalities.23 The indications derived so far from the discussion of the general hypotheses encourage further investigation of the relation between communist support and the rural population. The ambivalence and dualism of communism to the agrarian world was a constitutive feature for the Bolsheviks. They formed an urban movement, consisting essentially of workers and intellectuals who had never worked in the countryside or among the peasant masses, and among them, antipopulist and antipeasant prejudices and hatred for the mugik were widespread. While the Bolsheviks were leading a political revolution of socialist inspiration in Petersburg, with the urban working class as its protagonist, in the countryside a social 23

For this characterization of the Italian economic elite, see Caracciolo (1957: 92).

516

THE POLITICAL MOBILIZATION OF THE LEFT

revolution was occurring in which the forces of the rural petit bourgeoisie prevailed. Notwithstanding the efforts of Lenin, a true alliance between these two groups never materialized. Within one year of the Russian Revolution, the Bolshevik attempts to penetrate the countryside with class politics based on the poor peasants failed, and the preconditions were set for the future fatal conﬂict between them and a peasant movement that saw its raison d’eˆtre in the private possession of the land. In Western Europe, the relationships between the rural world and the communists were less conﬂictual, and given that communists were never in a position to implement their economic plans, they could remain ambiguous. There are some clear-cut cases in which the roots of the communist split were heavily concentrated among the socialist intellectuals and qualiﬁed workers – as, for instance in the Netherlands,24 Belgium, and Germany. However, in many other cases, and particularly those where communism was to prove a lasting feature, the crucial role played by the rural component of society was evident from the beginning. The detailed analysis of party conferences and voting in the Norwegian Socialist Party in the period before the split shows that radical positions and demands were stronger among the representatives of the primary sectors of forestry, ﬁshing, and agriculture.25 Indeed, at the Congress of Tours (December 1920), the representation of rural and peasant elements was predominant in the majority of the French SFIO that decided to join the Third International.26 At that congress, many questions were raised by the moderates to the Third International about how they would then deal with the peasant and rural enthusiasm that had turned toward communism.27 Similarly, it is documented that the Third International faction of the Italian Socialist Party was heavily rural and southern-based and that communist support after the Livorno split was very strong in prevalently rural regions of the country.28 The organizational and electoral development of the Communist Party of Italy between 1921 and 1924 occurred primarily in southern regions and zones – the traditional bastions of the PSI Third Internationalists.29 It is difﬁcult to interpret these common features of Italian, French, 24 25 26

27

28 29

Buiting (1990: 71–72). Lafferty (1971: 257–289). This role is documented by several detailed reconstructions. See in particular Kriegel (1969: 395) and Judt (1976: 46). Sembat explicitly asked Cachin and Frossard: ‘‘Are you sure that the peasants will be with you in the event that a new Commune will occur?’’ Le Congre`s de Tours (1964: 167). See also Lefranc (1963: 233). See the data provided by Detti (1972: 474). Detti (1972: 205–206, 499–500 and 500, footnote 137).

THE COMMUNIST SPLIT

517

and Finnish Communism. The major French historian of the communist Party has extensively argued that the strong propensity of the rural component of the SFIO for Third Internationalism should be related primarily to the strong mobilizing effect of World War I. This war shattered the French countryside, breaking into the social and political isolation of the rural world, placing the peasants in the trenches alongside urban dwellers, making them pay the highest price for the war, and introducing them to politics, orienting them to the more radical and more militant tendencies of the time.30 The rural mobilization of the war was important in France, as well as in Italy and Finland (with its civil war), but obviously it does not automatically explain the failure of socialism to channel this phenomenon. The cases of Italy and Finland are more clear-cut because socialism was above all a rural movement to begin with. From the 1890s, Italian socialism grew among the braccianti (agricultural workers) of the lower Po Valley and of Emilia. In the provinces around the lower Po River, in this region of large land reclamation projects and newly built villages, often deprived of the familiar presence of a church, the agricultural workers represented a rural proletariat with little resemblance to that of other European countries. Unlike the serfs of the German rural areas to the east of the Elba River, they were a class of recent formation, without a past of subjection and submission, and in many ways more similar in mentality to workers and salaried employees than to peasants. Thanks to the early penetration of socialism in these groups, Italian socialism was rural from the beginning and was never characterized by the distrust of the rural masses typical of other socialist movements. Indeed, sometimes, as in Emilia, it was the ‘‘red’’ countryside that surrounded and went on to conquer the ‘‘white’’ urban settlement.31 Third Internationalism and communism simply took over this historical support in a phase of intense radicalization and added to it the traditional left-wing orientation of the widespread groups of sharecroppers (mezzadria) and tenants in central Italy. This situation of early differentiated social support has always prevented Italian ecological studies from ﬁnding any association whatsoever between indicators of industrial development, urbanization, and the communist vote.32 In France, by contrast, the agricultural workers were less numerically signiﬁcant than in Italy and Finland, but also far less concentrated and 30 31

32

Kriegel (1969: 395). For the rural nature of early Italian socialism, see Procacci (1970: 414–415) and Ragionieri (1976: 1774–1785). Cf. Galli (1968: 161).

518

THE POLITICAL MOBILIZATION OF THE LEFT

politically cohesive.33 Communism made electoral inroads among them and among the sharecroppers of the matayage areas,34 but also, more significantly than in Italy, among the small peasantry of relatively backward agricultural areas. Linz has summarized the multifaceted sources of communist support in various peasant areas and rural groups in France: from the small owners of the Mediterranean departments – among whom the less well-to-do used to earn part of their income as farm laborers on the large farms of the area – to the small peasant owners of the more backward region around the Central Massif – in an area of widespread agricultural property but also of precapitalist and low-productivity agriculture.35 The fact that rural support for communism was so socially differentiated36 and unusually widespread among small owners, plus the fact that in the eyes of intellectuals interpreting the phenomenon it could hardly be attributed to the appeal of the communist rural plans and programs – that from the 1920s to the 1960s oscillated continuously and contradictorily between short-term promises to small owners and long-term transformation plans37 – has led to the characterization of the rural communist vote as a backward and protest vote, that is, a vote of resistance to change by declining groups. This interpretation probably underestimates both what the Communist Party and militants were actually offering through their organization to these individual and social groups, and the fact that the defense of the status quo was highly appealing for individuals and areas where feelings of economic insecurity were widespread. The combined appreciation of the comparative role of agricultural independents and laborers helps to characterize the level of class polarization of the countryside. It also clariﬁes where and when the agrarian question assumed an important role in the internal socialist ideological and policy debate and, ﬁnally, how it shaped the proﬁle of the class 33 34 35

36 37

Ehrmann (1952: 39). Linz (1976: 403). Linz (1976: 390–392). He goes so far as to identify speciﬁc historical-tradition factors in given areas, such as the traditional leftism of the Calvinist Protestants that grew out of anti-Catholicism in defense of laicism in departments such as Gard; Ibid., p. 428. This tradition is documented in Schram (1954: 172–218). Relationships between patterns of land tenure and communist support have also been identiﬁed in countries like Indonesia, the Philippines, and, to a much lesser extent, India. See the summary of this research in Zagoria (1973: 9–16). For the documentation of the multifaceted social roots of rural communism in France, see Ehrmann (1952: 25–26), Wright (1952: 361–372), and (1964: 185–208), Micaud (1955: 354–366), and Gaborit (1972: 197–222, 467– 495). For the cartographic documentation of this dualism see Goguel (1969) and (1970: 176). On the French communist rural stands, see Gaborit (1972: 201–208).

THE COMMUNIST SPLIT

519

Figure 9.1. Class polarization in the countryside.

cleavage to a considerable extent. Figure 9.1 is a scatterplot for the 1900– 1940 period of the average proportion of agricultural independents over all the independents and the average proportion of agricultural workers over all the workers. The y axis gives an idea of the extent to which the working class was either integrated into a predominantly urban and industrial society or, alternatively, split into an urban and agricultural component. The x axis indicates the extent to which the bourgeoisie (the independents) were also predominantly urban and industrial or predominantly rural. The four quadrants identiﬁed by the two reference lines of the overall means single out different agrarian structures. In the bottom left quadrant, we ﬁnd cases in which the percentage of agricultural workers out of the whole working class and the percentage of agricultural independents out of all the independents are below the European mean. This means that both the working class and the independent groups were relatively well integrated into predominantly urban groups. The United Kingdom is the extreme case in which the independent rural bourgeoisie was as insigniﬁcant as the group of agricultural workers. In Belgium, Switzerland, Germany, and the Netherlands, this situation is less well clearly delineated because the per-

520

THE POLITICAL MOBILIZATION OF THE LEFT

centage of freeholders in the countryside is much higher than in the United Kingdom. Yet, they are still quite different from all the other European cases precisely because the percentage of agrarian laborers in the working class remains marginal (between 10% and 20%) and the weight of agricultural independents is also limited within the global group of independents. In all these cases, class polarization is mainly a feature of the urban and industrial world. In the bottom right quadrant, the cases of Norway, Sweden, and Ireland (estimated, see earlier) are located. In these cases, the rural component among the independent groups was strongly dominant. Between 1900 and World War II, about 70% of all independents consisted of agricultural freeholders. At the same time, agricultural laborers formed a small component of the working class, which, whatever its size, was composed of four-ﬁfths of the industrial workers. The combination of these two ﬁgures identiﬁes these cases as areas of relatively egalitarian farmers’ communities, with strong status equality predominating in the countryside. Denmark and Austria are intermediate cases. More than half of the independents were rural, which gives them some elements of an egalitarian farmers’ community. At the same time, about one-quarter of the working class was composed of agricultural workers. In Denmark, this intermediate position was clearly reﬂected in the polarization of the agrarian independents, which led to the split of the agrarian party into a more radical and egalitarian small peasant movement, on the one hand, and a more conservative large farmer movement, on the other. In Austria, the entire agricultural world remained deprived of autonomous political representation, and – if one excludes the small Landbund of 1918, which never made an electoral breakthrough – the bulk of the peasantry remained within the Catholic Party. Finally, the most interesting group is represented by the cases in the upper right quadrant. In these countries, the independents were overwhelmingly composed of rural independents, and the working class was, similarly, overwhelmingly composed of agricultural laborers. Particularly in Italy and Finland, and to a lesser extent in France, between one-third and one-half of the working class was represented by rural workers, and this predominantly rural proletariat also faced a predominantly rural group of independents. In these cases, the agrarian question was the crucial question for the socialist movement. This movement was, to a large extent, the natural representative of the rural working class, but, at the same time, it could not ignore the problems and requests of the small peasantry. Thus, the working class was not a predominant and homogeneous urban industrial group, but was split down the middle into its agricultural and urban-

THE COMMUNIST SPLIT

521

industrial components. Similarly, the large world of agricultural independents was characterized by strong internal divisions, and class polarization tended to be high in these cases. Peasants’ political orientation was divided between rural radicalism and rural conservatism, and the many categories of the rural structure – the large owners, the small peasants, the tenants and sharecroppers, and the agricultural laborers – tended to split their political allegiances. The agricultural question presented a true dilemma for the socialist doctrine and practice. The question was not only about the defense of small agricultural property against the supposed law of economic development, but also about how to reconcile such a defense with the large agricultural worker groups and then how to reconcile both of them with the claims and interests of the urban-industrial working class. The radicalism and fragmentation of the left seem to be associated with high social nonhomogeneity of both the bourgeoisie and the working class and with the high class polarization of the rural world, with the single exception of Germany, where, in purely sociostructural terms, one would not expect the communist radical left to be strong in the interwar period.38 The rationale for this association can be identiﬁed in a direct and an indirect effect. Strong post–World War II communist parties all managed to add considerably different and sometimes disparate rural groups to the support of sections of the industrial working class. Where communism failed to make inroads into the countryside, either because early industrialization had already reduced the size of these social groups or because other powerful forces – denominational mobilization or selfrepresentational peasant movements – prevented such penetration, it remained an electorally insigniﬁcant or modest force. Strong communist movements are therefore ‘‘dualistic’’ movements, combining support from the advanced industrial areas as well as from sectors of ‘‘unsafe’’ agricultural groups. However, the direct effect of rural support for communism and the radical left is probably less important than the indirect effect created by the predominance of rural components in both the working class and the bourgeoisie. A large proportion of agricultural labor made the selfintegration of a cohesive urban and industrial working class difﬁcult, making it difﬁcult to build a stable alliance with the continuously incoming unskilled workers. If the skilled workers’ political representatives were unable to ﬁnd a viable alliance to incorporate the agricultural laborers and sharecroppers (and even tenants and small owners in certain areas), the chances for the spread of radical political alternatives increased. At the 38

Lipset (1983).

522

THE POLITICAL MOBILIZATION OF THE LEFT

same time, the ongoing predominance of small agricultural property not only contributed to deep internal conﬂict in the left about strategies, programs, and perspectives toward this social group, it also delayed economic growth and full industrialization and kept agricultural prices high. In a sense, under these conditions socialist parties were obliged to incorporate within their own social base and internal ideological debate the social contradiction of economic development and modernization. In other words, they were driven into the contradiction of simultaneously fostering the social groups that should inherit a fully mature industrial society and those that would be mostly affected by such processes and indeed might have to disappear as a result of it. Therefore, the high percentage of active population in agriculture created continuous problems of internal integration and cohesion for the working class,39 while the predominantly rural nature of the petit bourgeoisie created problems of system integration for the socialist movements and tended to radicalize the internal debates about short-term and long-term plans, policies, and strategies.

COMMUNISM, LEFT RADICALISM, AND THE ORGANIZATIONAL FEATURES OF THE SOCIALIST MOVEMENT No strong communist party developed after World War II if it had not already made considerable inroads between the world wars, and only one failed to keep the strength it had achieved prior to World War II. Therefore, the key to communist endurance must be sought in the organizational developments of the 1920s and 1930s. The strength of the communist success between the world wars has been associated with the organizational consolidation of the socialist movement and, in particular, with the links established between trade unions and the socialist parties in the prewar phase. There are several indications in the national literatures about a possible relationship between communism, left radicalism, and the organizational features of the preceding socialist movement. I summarize these into four theses. Communist penetration and the conquest of important positions within both the political and corporate movements were easier and more signiﬁcant where: 1. A weak and difﬁcult political relationship between the socialist party and the trade unions prevailed. The distance between the party and 39

Cf. Malefakis (1974: 1–7).

THE COMMUNIST SPLIT

523

the trade unions allowed radical agitators to penetrate unions without close partisan afﬁlation more easily. 2. Prewar socialism was a narrow movement, both electorally and, more important, organizationally. This occurred where its links with the socialist electorate were organizationally weak; where the party did not enjoy close segmental linkages with society or a Bund type of party organization; and where its membership structure was more open (as opposed to the collective membership tradition). 3. Trade unionization was low before the war and a large section of the working classes was still outside the socialist movement and was politically mobilized only during and after World War I. This meant that large numbers of industrial and agricultural workers were not yet solidly bound within the socialist organizational network when the new post–World War I wave of radicalization and political mobilization occurred. Thus they joined the left during the phase of intense communist–socialist strife, offering the communists the possibility of penetrating and even overthrowing working-class organizations that were precluded in other cases. 4. Traditional skilled-worker unions reacted to the new wave of postwar mobilization of agricultural workers and new groups of unskilled workers with exclusive and noninclusive strategies, which made it possible for new organizations to recruit heavily among these groups. These ‘‘organizational’’ hypotheses have in common the perception of Communism as a powerful organizational weapon – a perception that pervaded the hostile Western cultural climate of the Cold War in the 1950s and 1960s. Almond, inspired by Duverger and Neumann’s insistence on the total integration nature of the party and the control it was able to exercise over ‘‘the totality of the militant’s existence,’’40 has identiﬁed the nature of the communist parties of the West in their special mass–elite relationship. Mass and elite were separated by different models of communication and different levels of information; the communist leader was a ‘‘tactician of power,’’ and the mass bases of the parties were instrumental to the leader.41 Similarly, Selznick has qualiﬁed communist parties for their capacity (1) to transform members of voluntary associations into disciplined political activists; (2) to deploy their members as controlled agents within chosen groups; (3) to use mobilization and indoctrination techniques that 40 41

Duverger (1967: 141–149) and Neumann (1956b: 395–421). Almond (1957: 6–61).

524

THE POLITICAL MOBILIZATION OF THE LEFT

caused members to shift their loyalty away from other groups; and (4) to penetrate and manipulate institutional objectives.42 These views of communist parties were justiﬁed by the immense tactical ﬂexibility they had shown between 1918 and 1945, a period during which international communism followed at least six different political lines with relatively minor national deviations: the early revolutionary perspective of 1918–1921, the 1921–1923 United Front, the 1923–1934 ‘‘anti social-Fascism,’’ the 1934–1939 Popular Front, the 1939–1941 Nazi–Soviet Pact, and the 1941–1947 National Front. Each of them was characterized by different combinations of social and political allies and by different directions of organizational effort. Recognizing the superior organizational skill of communist militants and organizations, as well as their tactical ﬂexibility in penetrating noncommunist organizations, this line of interpretation assumed that the Communist parties were so effective in maintaining control that they were able to mold the preferences and attitudes of their members more than the party machine could be molded by them. In this sense, once Communist parties were established, their organizational and ideological features granted them a high degree of independence from the sociopolitical environment. Once cleansed of the residuals of the Cold War cultural climate, this perspective implies that the communists’ capacity for success rested fundamentally on the organizational infrastructure of the preexisting socialist tradition and on the capacity to resist and react to the strategies of organizational penetration of precommunist organizations. Communism could thus penetrate more easily where the organizational consolidation of the socialist movement was weaker. Where the syndicalist traditions had prevented an organic link between the socialist parties and the unions – and where the mystique of the e´lan re´volutionnaire was originally seen as a substitute for solidly organized, dues-paying members, and later rationalized into a virtue in itself 43 – unions failed to achieve an early mass base and were more susceptible to communist penetration and organizational take-over. The paradox of French socialism, discussed earlier, was that the wellestablished militant and revolutionary traditions of the workers, and the most complex ideological scene, were combined with weak organization, poor recruitment, and low cohesion (see the analysis in Chapter 6). It was perhaps the typical national acceptance and cultural legitimacy of such traditions that caused the lack of cohesiveness of the left’s organization and ideology. In contrast to France, the British labor movement established a 42 43

Selznick (1960: xiv–xvi). On this point, see Bowditch (1951: 32–34).

THE COMMUNIST SPLIT

525

distinctive organizational identity early on, but this may have prevented the development of an equally distinctive ideological identity, as well as of any widespread cultural hegemony. Thus, an organizational split in the working-class movement was avoided in the United Kingdom before World War I and in the 1920s, not because there were fewer conﬂicts here than elsewhere between radicals and moderates, nationalist and internationalist wings, but because of the organizational features of the Labour Party. The party had only collective membership of afﬁliated organizations, and the position of each organization, thanks to its internal majority rules, was always clear-cut. Some organizations were antiwar, while others were prowar; some were more radical and others less. As such, whatever internal conﬂict existed within the party was a conﬂict among different organizational components. To split the party, a majority had to prevail in a sufﬁcient number of organizations, and very few cross-organizational contacts were available for a minority cross-organizational split. The British–French contrast also casts light on other cases. In none of the European countries where trade union movements were founded and consolidated before the socialist party’s appearance in the last part of the nineteenth century (the United Kingdom, Ireland, and Switzerland; see the discussion in Chapter 6 and Table 6.2) did the radical left, and communism in particular, manage to make any signiﬁcant inroads. At the same time, in countries where closely interlocking links between the party and the trade unions and collective afﬁliation prevailed in the socialist movement (as in Norway and Sweden), this may have helped to maintain strong organizational cohesion in both the unions and the party. Extending this argument further, it may be possible to say that indirect socialist party structures or party structures based on strong social pillarization and relying on intense extraparty organizational networks and Bund-type organizational afﬁliations, were also more difﬁcult to radicalize and to split organizationally. In cases of strong organizational pillarization, the leaders of the party could rigorously and efﬁciently control the internal life of the sections and the activities of their afﬁliated members. Ideological currents and tendencies ﬁnd it difﬁcult to emerge unless they can rely on the support of some strong internal organizational component, able to provide ﬁnancial resources, support, and defenses on which to fall back. In these cases, indirect structures and the indirect control of party resources can counteract the tendency toward ideological internal debate and factionalism. This argument seems appropriate to cases such as the Belgian, Dutch, and Austrian socialist movements. Belgium, in particular, is a crucial test case, as the ideological and cultural inﬂuences of French socialism, the strong anticlerical ideology, and the relatively strong syndicalist traditions

526

THE POLITICAL MOBILIZATION OF THE LEFT

made the Belgian Socialist Party particularly vulnerable to postwar radicalization and splits. However, from the beginning, the party was a mixture of associations of every kind; and this complex but dense, indirect structure and network of ‘‘communal’’ and organizational strongholds – held together by the need to resist the interference of the state, which was then in the hands of Catholic political forces – helped the party to defend itself from internal ideological factionalism.44 The combination of these three features of the relationship between the party and the corporate channel organizations (respective timing of corporate and political consolidation, strong versus weak socioorganizational pillarization, direct versus indirect membership) yields a classiﬁcation of our countries, identifying those where the communist parties became a viable alternative to socialist organizations in the 1920s and 1930s. Germany, France, Italy, and Finland had all experienced a consolidation of socialist trade unions that either followed the consolidation of the political party or was parallel to it. In two cases (Germany and even more clearly Finland), trade union development was directed and fostered by the party’s leadership. In the other two (France and Italy), such relations were difﬁcult and unstable as a result of the party’s attempt to direct ideologically and politically recalcitrant unions. The strong inﬂuences of syndicalism – with its refusal to recognize the reality of politics and its disdain for parliamentary action – prevented effective working relationships between unions and socialist parties (Table 9.5). Let us now consider the other line of argument that relates the narrowness of the pre–World War I socialist movement to its vulnerability to postwar organizational splits. Narrowness can be conceived in either electoral or organizational membership terms. In the ﬁrst sense, a narrow socialist movement is one that is late or below average in its electoral mobilization capacity. Narrow in organizational terms is a socialist movement that organizes only a small proportion of the labor movement in its own voluntary organizations. In my opinion, however, narrowness is best understood in the relationship between the two, that is, a narrow socialist movement is one whose organizational density is low compared to its own electoral mobilization. The link between the fragile establishment of socialism and the communist success can be seen in the fact that between the world wars, narrow socialist movements were confronted with strong waves of politicization and mobilization of sectors of the working class that were beyond the reach 44

Moulin was very critical of the role of organizational pillarization in shaping the moderate petit bourgeois nature of the BSP. See Moulin (1981: 192–202).

THE COMMUNIST SPLIT

527

Table 9.5. Socialist movement organization in the early twentieth century

of the socialist movement.45 These sectors were composed mostly of agricultural laborers, unskilled workers, or even small peasants, and they lacked the tradition of the disciplined organizational action of the more skilled and earlier-organized sections of the working class. This rapid mobilization created problems for consolidated socialist organizations. Quite naturally, the reaction was one of exclusiveness, leading to closure of the social relationship in the organizations that, in defending the advantages of the organized members, tended to exclude or only selectively include new groups. When such nonexpansive and noninclusive logic prevails in traditional organizations, the radicalization of the excluded and their incorporation into alternative organizations is more likely.46 This interpretation has the advantage of linking organizational variables to the sociostructural factors discussed in the ﬁrst section of this chapter and, in particular, to the internal homogeneity of the working class that resulted from different patterns and tempos of industrialization. The greater the 45

46

Kriegel (1969: 431–432) offers a different interpretation of the same association: ‘‘a radical foreign experience that proposes abstract goals whose relationship with the real situation are unclear, easily attracts the interest of a sect.’’ This interpretation is consistent with Kriegel’s overall interpretation of Western communism as a ‘‘grafting,’’ i.e., as the implantation of an external model in an environment different from its own. See Kriegel (1974: 26–27). Cf. Bendix (1964: 108).

528

THE POLITICAL MOBILIZATION OF THE LEFT

internal homogeneity achieved, the more inclusive the working-class organization will tend to be; the more nonhomogeneous the labor world (artisans, craft unions, a large agricultural sector, agricultural laborers, and other groups in the countryside), the more likely is a reaction to exclude some sections of it from the established organizations facing rapid mobilization of new members. At the empirical level, these theses ﬁnd support in several national accounts of trade union history, although few materials are available on the history of socialist mass organization and membership development.47 In cases in which the end of the Great War produced intense radicalization of working-class politics, historians have related this to massive increases in the trade-unionized labor force, accompanied by the difﬁculties of socialist unions in dealing with this rapid increase. For Germany, the detailed and speciﬁc study by Biader48 documents how the December 1916 law on auxiliary service (the forced recruitment of a large number of workers) and the 1917–1918 strikes led to a large increase in trade union membership and a rapid change in the social base of these unions through the inﬂux of unorganized and unqualiﬁed workers. This was linked to the development of an internal trade union opposition. Biader claims that in this period the German unions reinforced tendencies toward leadership immobilization, conservatism, and organizational rigidity that were already present before the war.49 Such tendencies were emphasized by the unions’ support for the war goals in exchange for concessions that were later nulliﬁed by the beginning of the war. In Norway, it has been shown that postwar radicals tended to be more concentrated in the areas of explosive membership growth in the prewar period,50 while the moderate elements of the labor movement were concentrated in the guilds and handicraft unions.51 Sectors with an old-style organization were gradually overwhelmed and defeated in the battle for the control of the movement by new unions, which drew their strength from the rapidly industrializing rural labor force. Similar accounts are available for the same period in France and Italy.52 Generalizing this line of interpretation to all of our cases in Table 9.6, I have reported the measures of bivariate association between the vote for 47 48 49

50 51 52

See Cronin (1980). Bieder (1981: 99–108, 778–781). Tendencies that are also documented by Scho¨nhouver (1980: 333–376). For German trade union development after 1920 see Potthoff (1979). See Lafferty (1971: 257–289). Lafferty (1971: 279). Detti (1972: xxviii–xxix), Cortesi (1973: 161–162), and Lorwin (1954: 52 and 75–76). In France, the inﬂuence of the communists grew after each membership explosion in the trade unions; after World War I, during the Popular Front; and after World War II.

THE COMMUNIST SPLIT

529

Table 9.6. Correlation between the vote for different components of the left and organizational density indicators (1918–1985)

Note: Number of elections given in parentheses.

different components of the left and the main indicators of organizational density of the left and of the social democrats (for a detailed presentation and discussion of these measures, see Chapter 6 and the Data Appendix). The systematic structure of these data is quite impressive. The Social Democratic vote is always positively associated with trade union density and, quite naturally, with a social democratic organizational presence. The Communist vote, however, is systematically associated in a negative way with these indicators. Communist support tends to be weaker where the trade unionization level is higher and the organizational presence of the socialist parties is stronger. Controlling these associations by decade does not change the picture, so that they do not result from temporal trends but are also valid on a comparative synchronic basis.53 The association with the total left vote, which was discussed in Chapter 6, is the result of the combined effect of these two phenomena. As was the case with the socioeconomic structural variables, but to a much stronger extent, communism and socialism here appear as phenomena of two inherently different environments. Let us now look more closely at the cross-country variance. The argument presented earlier links communist strength to the narrowness and organizational weakness of the prewar socialist movement and, consequently, to the high rate of growth of trade unionization in the post– 53

Actually, the negative association with the communist vote is much higher on a decade basis. It is also interesting to note that these organizational indicators are associated positively with the other left, i.e., with the strength of left-socialist but noncommunist parties.

530

THE POLITICAL MOBILIZATION OF THE LEFT

World War I period. In Table 9.7, I report the mean value of the prewar (1900–1920) social democrats’ vote together with the mean value of the trade union density and social democratic organizational strength. In the last column, I report the communist vote in the following interwar period. The ﬁrst clear point is that it was not electoral narrowness as such that exposed socialist movements to postwar fragmentation. On the one hand, Finland and Germany had the strongest prewar electoral socialism; on the other, the weakest electoral prewar socialist movements – like the British, Dutch and Swiss, not to speak of Ireland, excluded from the table due to the absence of data before World War I and the absence of a communist party after it – did not experience wide communist splits. As far as socialist organizational density is concerned, three countries are shown to be consistently weak on both indicators: Italy, France, and the Netherlands. In these cases, the organizational presence of socialist members and activists was exceptionally low. Germany and Finland are either intermediate or weak cases. If we turn our attention to the rate of trade unionization (particularly on the left), Italy, France, Finland, the Netherlands, and Norway reveal organizationally narrow bases. Finally, in terms of left trade union membership over the whole electorate and the socialist vote, the weakest cases are again Finland, France, Italy, Germany, and the Netherlands. It has to be kept in mind that for the prewar period the important differences in the franchise may make it comparatively inappropriate to refer organizational membership to the electorate. To appreciate in a synthetic way the extent to which the prewar socialists encapsulated their electorate organizationally, we can cross-tabulate the level of left trade union density and the level of socialist membership over the socialist votes (partisan density). In Table 9.8, this is done by dividing into ‘‘high’’ and ‘‘low’’ categories the values in Table 9.7. In this arrangement, the organizational encapsulation of the Austrian, Danish, and Swedish socialist movements stands out against the corporate as well as partisan weakness of Finland, Italy, France, and the Netherlands (whether the United Kingdom is high or low on the partisan density dimension depends on how we consider its almost entirely collective membership). For the Netherlands, I have argued that the low level of partisan identity found a functional equivalent in the strong social pillarization of the country and in the host of subcultural organizations surrounding the parties without being formally afﬁliated to it. All cases of strong communist movements therefore fall into this group, with the exception of Germany, whose level of trade unionization was higher. The last analysis to be made concerns the extent to which these

THE COMMUNIST SPLIT

531

Table 9.7. Pre–World War I (1900–1920) socialist electoral and organizational density and post–World War I communist vote

Table 9.8. Pre–World War I organizational density of the socialist movement

countries were characterized by massive trade union membership growth in the post–World War I period. To consider this, I have computed the rate of percentage change in the number of trade union members over the active population in each country by election years. Across the war years, almost all countries experienced signiﬁcant and rapid changes in trade unionization levels. This happened, in descending order, in Germany between 1912 and 1919 (⫹17.6%) and 1919 and 1920 (⫹13.4%); in Belgium between 1912 and 1919 (⫹23.2%) and 1921 (⫹7.9%); in Finland between 1916 and 1917 (⫹22.1%); in Denmark between 1913 and 1918 (⫹20.8%); in the United Kingdom between 1910 and 1918 (⫹18.8%);

532

THE POLITICAL MOBILIZATION OF THE LEFT

in the Netherlands between 1913 and 1918 (⫹13.1%); and in Sweden between 1917 and 1920 (⫹10.5%). Far less signiﬁcant increases were experienced by Austria (1919–1920: ⫹9.7%), Switzerland (1914–1919: ⫹7.6% and 1919–1922: ⫹9.6%), Italy (1913–1919: ⫹7.8%), and France (1914–1919: ⫹6.4%). It is possible that, for instance, in Norway, Finland, or Italy, the communist and radical presence was stronger in the more rapidly growing unions, but it is not true that communism was stronger in those countries with a massive explosion in union membership. In France, union membership grew modestly or stagnated throughout the 1930s (as did the economy), and a major increase occurred only in the years of the Popular Front. Of the countries with massive postwar or cross-war increases, only Finland and Germany saw the growth of a powerful trade union and radical party opposition; this did not happen to any comparable extent in Belgium, Denmark, or the United Kingdom. Germany has the biggest expansion, with an increase of more than 30%. This was immediately followed by an equally massive demobilization, as trade union membership had already declined by 10% in 1924. A similar phenomenon is observable in Finland, where the 22.1% increase in 1917–1919 was followed by a corresponding decline in 1922. On the whole, the evidence of a relationship between rapid growth in union membership and radical left opposition is inconclusive at the comparative level. It can only be mentioned in the cases of Finland and Germany, and even here only in conjunction with the other organizational features discussed previously.

INSTITUTIONAL AND POLITICAL INTEGRATION A further possible interpretation of radicalism emphasizes the politicoinstitutional rather than the socioeconomic or organizational conditions. The emphasis therefore switches from an interpretation of radicalism as mainly sociostructurally or organizationally based to one that stresses politico-institutional alienation. There are several examples of this line of reasoning. In a comparison between Disraeli’s and Bismarck’s governments between the 1860s and 1870s, Schumpeter blames Bismarck for what he regards as a ‘‘fatal mistake’’ that was entirely his own. The attempt to incorporate the proletariat into the state through social legislation, while depriving it of its own organization and leadership in politics and trade unions, was a mistake that produced unforeseen results later on. Schum-

THE COMMUNIST SPLIT

533

peter seems to believe that without such a policy things would have gone differently, and that the antimilitarist, antidynastic, and antibureaucracy attitudes of the socialist leadership in exile, which were obviously unacceptable to the king, would have been avoided.54 In Lipset’s early formulation, ‘‘in nations . . . where access was denied for long periods . . . and where force was used to restrict access, the lower strata were alienated from the system and were led to accept extremist ideologies. . . .’’55 Bendix instead insists that the working class felt politically alienated as incomplete or second-rank citizens, a feeling that led to their radical stands and utter rejection of the existing dominant order in early socialist history.56 Lorwin has made the synthetic but illuminating statement that the labor movements that are most dependent on the state may show the greatest hostility to the state itself.57 It is possible to adapt this line of reasoning to the electoral success of European communism. I have already discussed in detail (see Chapters 5, 7, and 8) the patterns of institutional and political response; here it is sufﬁcient to recall and adapt those results. When I commented on the application of this model to the whole left, I concluded that the level of electoral mobilization of the left was higher in those cases where its institutional integration was initially more difﬁcult. On the overall level, when the variables of institutional integration (stateness, parliamentarization, level of repression, obstacles to representation) are related to the vote for the internal components of the left, the variables of level of stateness and level of repression are the ones that best discriminate between the support for the socialists and the Communists (see Table 9.9). The variable of stateness, which was shown to be unrelated to the overall development of the left (see Chapter 7), is instead very effective in discriminating between Communist and socialist support. The stronger the extractive and repressive resources of the state, the stronger the likelihood of a radical Communist movement. On the basis of this analysis, it can be hypothesized that strong state resources that were directed against the working-class movement enhanced the electoral mobilization of the left, as well as its internal divisions and fragmentation. At the same time, high obstacles to representation and late parliamentarization seem to have damaged the socialist parties more 54 55 56 57

Schumpeter (1976: 341–344). Lipset (1963: 541–569). Bendix (1964: 122–126). Lorwin (1967b: 58–72).

534

THE POLITICAL MOBILIZATION OF THE LEFT

Table 9.9. Institutional integration and vote of components of the left

than the communists.58 This is quite consistent. If communist support is read as a result of poor institutional integration, these aspects are elements of its strength rather than features that hamper its support. If we look at the individual countries rather than at the overall picture (these comments are based on Table 7.19 in Chapter 7), we see that in all the cases where a successful communist party became established after World War I, the level of political repression of the working-class movement had been, and remained for long time, quite high. In none of the countries with a tradition of low repression did communism make any major inroads. Austria is the deviant case in this respect. In the ﬁve countries with high stateness traditions, three experienced profound left splits, while Belgium and Austria did not. On the other hand, Finland experienced a split without having a strong state tradition. The combination of the two dimensions points once more to Austria as a deviant case, because a strong communist split that did not occur might have been expected. On the two other dimensions of parliamentarization and representational obstacles, two of the four cases of successful communist splits (Germany and Finland) both have late parliamentarization and high representational obstacles, while Italy and France have modest or low representational obstacles and early (although not fully consolidated) parliamentarization. Even in this case, Austria associates with Finland and Germany without showing a similar electoral split of the left. In Figure 9.2 the cumulative rank ordering of countries in terms of institutional integration (the higher the rank, the lower the institutional integration; see footnote 122 in Chapter 7) is plotted against the average communist and socialist vote. I have resorted to a longer-term average 58

The very strong negative association between representational obstacles and other left vote is obviously due to the fact that this vote has always been quite small, suffering particularly from representational threshold of various kinds.

THE COMMUNIST SPLIT

535

Figure 9.2. Rank ordering of institutional integration and the socialist and communist vote (1918–1985). (the 1918–1985 period) because the strength of communism is discontinuous in the German and Italian cases. No association exists for socialist parties, whose electoral fortunes are, in the long run, independent of the pattern of institutional closure or openness of the political system. By contrast, communist support appears to be more closely associated with this factor, particularly when the Austrian case, which deviates from the institutional-integration model, is excluded. Finally, let us consider the problem of left division and cohesion in terms of the variables discussed under the label of political integration. Two common features link the countries with a deep left split after World War I. The ﬁrst is the historical lack of a solid and stable political alliance between the working-class movement and at least one part of the liberal forces, whether representing mainly an urban bourgeoisie or a rural one. The second is the fragmented political representation of the agrarian groups when these groups are both large and polarized along class conﬂict lines. The presence of a strong religious cleavage that opposed ﬁrst the

536

THE POLITICAL MOBILIZATION OF THE LEFT

church to the state, then secularism to ideological clericalism, and ﬁnally the church and the religious community to the rising socialist movement is not a common feature. It is present in several countries that never developed divided lefts (Belgium and Austria among the homogeneous Catholics), and it is absent in Finland. A Lib–Lab alliance was totally absent in Germany and Austria, where liberal forces remained weak, incapable of autonomous political organization and largely hegemonized by conservative forces, while the rural bourgeoisie was under Catholic hegemony or, at any rate, strong religious inﬂuences. The only possible alliance strategy was political and needed to be developed toward the Catholic and religious organizational world. However, it was unstable and unsafe for the socialists from the beginning as it was undermined by a strong clerical-anticlerical tradition of opposition and was therefore likely to increase the internal tension in the socialist movement concerning the possibility, opportunity, goals, and perspectives of such an alliance. In Finland, although the cleavage structure had some of the typical Scandinavian features, its late and stormy independence created divisions within all the political forces concerning the status and nature of the political system to be built, and the high class polarization of the variegated rural world prevented both consistent cooperation with a rural liberal movement and, later, consistent cooperation with the agrarians. In France and Italy, I have argued that liberal forces proved to be Malthusian in both their economic and political strategies. Those parts of these forces that were open to the socialist movement remained in a minority, although they were stronger than in Germany and Austria. Above all, such forces, and in general the large and ideologically heterogeneous marais, which in both countries called themselves ‘‘liberalism,’’ never consolidated modern political partisan groups capable of providing sufﬁcient discipline, organizational cohesion, and programmatic consistency to offer a safe political, if not governmental, point of reference for the socialist movement. At the same time, the sheer weight of the rural population obliged these parties to compete in a class-polarized countryside that split its support along class lines among socialists, Catholics, and traditional liberal notable representatives. In all these ﬁve cases, the end result was the protracted political isolation of socialism. The unstable nature of any possible political alliance that resulted from the ambiguous democratic record of most liberal forces and the traditional, if not reactionary, attitudes of the Catholic leadership exposed any attempt to increase the bargainability of the socialists’ policy goals to high costs and low payoffs. The internal divisions within these

THE COMMUNIST SPLIT

537

lefts about what could be actually achieved by cooperation with the bourgeois forces were based on the low organizational and ideological reliability of the opposite camp. Thus, even when socialists accepted executive responsibility under such conditions, coalitions were short and unstable – as in France in the mid-1930s, in Germany in the 1920s, in Italy in the 1960s, and in Finland in the early 1920s – and their policy achievement was generally regarded as minor from the socialist point of view, that is, it was not comparable in scope and stability of policy orientation to the cases of long-standing policy cooperation with liberals and agrarians. Where socialists had limited or no control over government, their wage policy was in direct competition with that of the communists only in the ideological and propagandistic realm. Only abstract, long-term consequences were compared. While this did not create a gap between reality and communist appeals, it did, create a schizophrenic split among socialist doctrine, programs, and political action because of the contradictions between what they represented in the political alignments of their political systems and the image that members and militants actually had of them. Where socialists were sufﬁciently strong and not politically isolated, they could redeﬁne their identity in terms of performance, avoiding this internal factional strife based on a clear contradiction between identity and reality.

CONCLUSION: A SYNDROME MODEL In the preceding discussion, I have deﬁnitely rejected some of the generalizations concerning radical political support and found more or less extended support for others. However, it is clear that no one single factor can account, without exceptions and deviant cases, for the spread of left radicalism from World War I onward. For this reason, a syndrome of various factors has to be considered. The most elaborate attempt to explain the more or less radical nature of socialist movements by a syndrome model has been elaborated by Lipset, who identiﬁes three variables: (1) rigidity of the status system; (2) extension of the franchise; and (3) de jure and de facto recognition of trade unions’ existence and activities.59 Of the three independent variables, the most important is the ﬁrst. According to Lipset, ‘‘the more rigid the status demarcation lines in a country, the more likely the emergence of radical working-class based-parties,’’ and ‘‘rigid status systems are conducive to the emergence of radical working-class move59

Lipset (1983).

538

THE POLITICAL MOBILIZATION OF THE LEFT

ments.’’ Such rigidity of status demarcations (or status differentiation) is maintained by ‘‘continuation of feudal or aristocratic values,’’ ‘‘retention of major post-feudal elements,’’ and ‘‘emphasis on status and aristocracy.’’ My analysis does not support this model. Lipset regards Germany, Austria, Sweden, and Finland as cases of rigid status demarcations and radical working-class movements, with France and Italy as cases of feudal remnants and strong bourgeois concern for status. The United Kingdom, which seems to contradict the association, is dealt with as a conﬁrming case in reference to the Chartists (one century before communism). For Sweden, the supporting evidence for the relationship is that ‘‘the strong support obtained by the Social Democrats in Sweden . . . can be related to the strength of Standenstaat elements in the most status-bound society in northern Europe’’;60 that is, radicalism is deﬁned here as giving strong support to the Social Democrats. However, Norway has a Socialist Party that is just as strong as that of Sweden, but it has no feudalism and aristocratic tradition. Belgium, the Netherlands, Switzerland, and Denmark are grouped together as cases of weak feudal elements that explain moderate working-class movements. My reading of the Danish case is different, much more similar to that of Sweden than to that of the Netherlands. Lipset’s other two independent variables seem less important and receive less attention (particularly the third, de jure and de facto resistance against the labor movement). For the franchise, I have not been able to detect any relation whatsoever supporting the idea that the timing and tempo of the franchise helped to create the most integrated working-class movement. This may be due to the fact that the concept of ‘‘franchise’’ is encompassing in Lipset’s treatment. For instance, it is possible to accept his point that political citizenship was early in the United Kingdom only if liberalization and political rights are mixed with strict voting rights. Similarly, in the discussion of Germany and Austria, franchise development and regime liberalization are mixed. In Table 9.10, I have tried to summarize the factors discussed in connection with the cohesion and division of the European left, classifying cases as precisely as possible. This synthetic table is not only a ﬁnal summary. It should also identify the syndrome, that is, the set of factors that, when combined, were likely to create the conditions under which radical opposition to the political and economic order appeared as a convincing alternative to large sectors of the lower classes, leading to profound ideological and organizational splits within the working-class movements. A syndrome logic is necessary because these factors impinge on one another 60

Ibid.

539

Table 9.10. A syndrome for communist political support

540

THE POLITICAL MOBILIZATION OF THE LEFT

to a large extent. To facilitate this work, the gray areas of the table present all those values that were, according to the theoretical arguments developed in the text, considered to be favorable to the radical outcome. It is clear that no single variable or set of variables neatly separate the set of strong communist split experiences from those of more cohesive lefts. The ﬁrst set of variables pertains to the socioeconomic conditions under which socialist movements developed. Delayed and late industrialization tended to produce internally nonhomogeneous working classes with few exceptions. France, Italy, and Finland are the only countries that present the three features together: delayed and late industrialization, a nonhomogeneous working class and high class polarization of the countryside. Austria and Denmark present similar features but with a lower-class polarization of the agrarian world. Germany is totally deviant in terms of socioeconomic conditions. Thus, it is not possible to link communism to socio-economic circumstances too closely. A thesis that links in a causal way the percentage of the agricultural population (and therefore late, uncompleted industrialization) to the lack of self-integration of the working class, to the lower ‘‘social integration’’ of that same working class, and to more radicalism and less organizational cohesion61 avoids too many intermediate steps and fails to explain why Denmark and Austria did not follow the same pattern as Italy, France, and Finland, and why Germany did in the inter war years. In this set of variables, one ﬁnds, however, the only single variable that discriminates the countries with large post–World War II Communist movements from all the others – the extent of class polarization in the rural world. If we reason according to Mill’s method of concordance – according to which if two or more cases with a common effect have only one circumstance or feature in common, then that feature is the cause of the common effect – we should point to this element as the strongest potential causal condition of communist success. The second set of variables refers to organizational aspects of prewar and interwar socialism. In this case, the situation is more varied, but only Italy, France, and Finland present systematic features that point to generally low organizational encapsulation of the voter–party links, weak trade unionization and poor party–union linkages, and weak socio-organizational pillarization. The experience of Germany is close to that of these countries, although it tended to have much higher corporate and even partisan density than the latter. The contrast between Belgium, Switzerland, and the Netherlands, on the one hand, and the three Scandinavian countries, on the other, is, however, noteworthy. In the small continental countries, 61

Malefakis (1974: 1–7).

THE COMMUNIST SPLIT

541

parties tended to precede unions in national organization (with the exception of Switzerland, where, however, the difference in timing is unimportant), and membership tended to be individual, resulting in low corporate and partisan density. However, in all these cases, societies were strongly pillarized and dense in sociopolitical organizations that, although not necessarily collectively afﬁliated with the party, offered (not only to the socialists, but to all major parties) a dense network of closely linked and supporting organizations. In the Scandinavian countries, the situation tended to be the reverse: Socioorganizational pillarization was weak, but collective afﬁliation and high organizational density played an important role (with the exception of Denmark, where membership was predominantly individual, and of Norway, where corporate density was low in the early period). In the ﬁrst group, the strength of socio-organizational pillarization of the society offered a functional substitute to balance the weak partisan and corporate density, making it less important to establish close voter–party links; in the Scandinavian countries, the strong organizational structuring of both the corporate and partisan channels compensates for the weak societal pillarization.62 In this set of variables, we ﬁnd the elements that set Austria aside from the countries with profound left divisions – the impressive Austrian organizational density in all dimensions and, at the same time, the high level of socio-organizational segmentation resulting from the unique superimposition of the urban–rural, class, and religious cleavages. This, together with the low class polarization in the countryside, seems to provide the main explanation of why the Austrian working class did not eventually split along ideological lines, although it had quite favorable socioeconomic and politico-institutional conditions (see later). The third set of variables relates to our discussion of institutional integration and political isolation. We again ﬁnd that Finland, France, and Italy experienced poor institutional integration and political isolation, features that they share with Germany and Austria. These conditions were already at the core of the strategic contradiction of many socialist parties before the Russian Revolution and the development of international communism. After the Great War, the tension between electoral potential and actual achievement created a powerful ﬁssure within the socialist movements, particularly those that could not rely on strong corporate and 62

For this interpretation of the contrast between these two sets of cases in which the functional equivalence between societal segmentation and corporate/partisan density is discussed, see Bartolini and Mair (1990: 234–235).

542

THE POLITICAL MOBILIZATION OF THE LEFT

partisan organizational networks of their own or on strong social pillarization of the respective societies. The fact that the intensity of internal ideological divisions can be related to the patterns of politico-institutional integration is best epitomized by the German case, where this variable probably had great importance, given the weak sociostructural preconditions of radical opposition. The three alternative strategic lines that agitated the pre- and postwar Social Democratic Party can be read as different responses to different politico-institutional environments. In the German territories in which the repressive politics of the state yielded to moreintegrative strategies, as was primarily the case in the southern and southwestern areas, the reform-oriented wing of the labor movement predominated. By contrast, in Saxony and in the Prussian regions, where more severe repression and greater political isolation were common, more radical politico-theoretical thinking prevailed. This is illustrated not only by the election results and theoretical positions of elected candidates, but also by the pragmatic politics at the regional level and by the many reform proposals of social democrats with regard to the general constitutional system.63 In conclusion, we can summarize our syndrome of the divided left. Communism was socially the expression of the combined support of advanced sectors of the industrial working class of economically late or backward societies, of a developmental middle class intelligentsia, and of considerable sectors of the rural world that resisted and survived its full transformation in a commercial capitalist direction. This potential base was able to support a Communist split only when the socialist movement was marked by weak organization, poor institutional integration, and low political coalition potential. These three elements strongly interacted with each other. A backward agrarian structure tended to be represented by political forces that regarded the isolation of socialism as a fundamental condition of survival. Weak trade unionization weakened the socialist movement in the market, and organizational weaknesses and poor party– union relationships could not check the internal tensions generated by mass mobilization and rank-and-ﬁle radical politicization. Radical political alternatives seemed to provide a convincing way out of the existing structure of social and political control and market weaknesses. The argument that no substantive goals could be achieved without radical breakthroughs was convincing. The alternatives proposed to the combination of high social and cultural control, organizational weakness, and low institutional and political integration could only be of a long63

Tenfelde (1990: 267).

THE COMMUNIST SPLIT

543

term ideological and radical nature, achievable only by destroying and overcoming the structure of control: the state in primis and the church – the traditional social hierarchies of the rural world. The growth and persistence of a radical left opposition became closely related to the availability of sections of the lower classes that felt dissatisﬁed in societies in which not only business and capital, but also the state, government, and the church refused to recognize the legitimacy of their organizations, leaders, and demands and where, at the same time, the organizational weakness of the socialist subculture was fueled by the heterogeneity of social groups derived from late and unsuccessful industrialization and by the weak organizational articulation of the civic society. Ideological radicalism of the communist type did not emerge from the class conﬂict as such, meaning from the opposition in the market with business and employer interests, but only when unsuccessful politico-institutional integration produced the overlapping of such class cleavage (economy) with antistate (politics) (and often antichurch, i.e., cultural) cleavages. The fact that demands directed to employers should have been combined with those directed to the state and other agencies of cultural control into a single coherent ideological alternative – instead of being split and deﬂected by more powerful integration mechanisms separating the market from politics and both of these from cultural and organizational identities – was not natural, but rather the result of a historically speciﬁc syndrome. Within this syndrome framework, the reader may be surprised by the absence of any reference to the role of syndicalism, which several authors have identiﬁed as a crucial factor in protracted radicalism.64 There are three reasons for this exclusion. First, strong syndicalism was important only in a few of our countries – France, Italy, and, with much less vigor, Belgium and the Netherlands. It was virtually absent in Finland and the Scandinavian countries, as well as in Germany and Austria. The second and more important reason is that the supposed linkage between the ideological role of syndicalism and communist splits is spurious because it is mediated by organizational variables. Syndicalism may have played a separate ideological role, but what is more important is that it resulted in poor party–union relationships, weak unions, and predominantly political movements that were fairly strongly detached from the unions. Syndicalism contributed to the absence of organic relationships between the party and the working-class organizations, to the predominant individual membership and nonbund type of socialist political movements, and to the general weak corporate and partisan organizational density of the whole 64

Kornhauser (1959: 156), Lorwin (1954: 38), and Judt (1979: 285–286).

544

THE POLITICAL MOBILIZATION OF THE LEFT

socialist movement. I have decided to give more direct weight to these organizational factors, also because they can be independently assessed in all countries. Finally, interpretations of why syndicalism gained strongholds in countries like Italy and France normally underline (1) their slow rates of economic development, whereby the spread of industrialization over a long period gave the workforce time to develop and organize in a cooperative defense of its independence; (2) the continued prevalence of small workshops and artisan enterprise, which made it possible to argue that their disappearance was not inevitable and that their defense might have served as a basis for a socialist organization of society; (3) the lack of entrepreneurial ﬂare, harsh and arbitrary shop regulations, and narrow-minded employer attitudes, which led to impatience among workers, who saw little chance of improvement thorough either collective negotiations or personal advancement in the industrial and social hierarchy; and, again, (4) poor, loose, and decentralized union structures. These causal interpretations – which tend to characterize syndicalism as the ideology of a declining group of artisans and partially proletarianized workers – are incorporated into my general plan under the socioeconomic variables. Again, rather than using a variable such as syndicalism, which is strongly idiosyncratic and countryspeciﬁc, I felt it preferable to operationalize the underlying causal conditions in a way that is generally applicable to all these cases. In this perspective, syndicalism is seen as an ideological mediating factor between speciﬁc conditions of industrial development and social structure and corresponding levels of organizational cohesion and density of the corporate and partisan channels of the socialist movement. In my syndrome model, the single deviating case is Austria, which presents several features that would have made it a suitable candidate for interwar radical splits that never materialized. Within my framework, the explanation for this is that Austria did not have class polarization in the countryside (instead, urban–rural and religious cleavages overlapped) and that it enjoyed one of the most strongly organized and socially pillarized socialist movements. I could stop here, but an additional element deserves mention. The Austrian socialist movement has a unique feature in the history of twentieth-century working-class movements: It was the only movement to develop for a long time in an extremely heterogeneous ethnic environment. Urbanization in the second half of the nineteenth century qualitatively modiﬁed the ethnic equilibrium of the empire. This happened particularly in Bohemia and Moravia, seats of the large industrial centers of the empire, where the population was prevalently German, while in the countryside the population was prevalently Czech. Prague, formerly a

THE COMMUNIST SPLIT

545

prevalently German city from a cultural point of view, saw the growing predominance of Czechs in only a few decades. In Vienna, too, the Czech population reached 25%. In Trieste, in the phase of economic growth, there was a massive immigration of Slovene workers that considerably altered the relationship among the ethnic groups. The working class of the empire thus found itself not only, as usual, socially subordinated to the national bourgeoisie but also encountering ethnic discrimination in the face of the dominant Austro-German nation. This isolated the German workers, pushing them to join ranks in a panGerman nationalism in defense of their relatively privileged position. This weakness within the imperial working-class movement brought it to a national split, but my argument is that, at the same time, it helped the German part of the movement to maintain its own internal cohesion before and after World War I. Ethnic fragmentation limited ideological factionalism. At the moment of the international crisis of World War I, the party tended to contain its internal debate in order to avoid the splits occurring in Germany, and Adler managed to maintain the center between patriotic pushes, on the one hand, and antimilitary pushes, on the other, thanks also to the skillful play of national tensions.

10

THE MACROCONSTELLATION OF CLASS CLEAVAGE STRUCTURING he power resources of the lower classes, and wage earners in particular, depended primarily on their willingness and capacity to act collectively in the market and in politics, that is, to create and sustain corporate and political organizations for collective action. This problem can be framed in terms of a collective-action calculation. Assuming the existence of a common group interest determined by a common social position, the historical sequences, the institutional setting, and the resulting political opportunities should be regarded as the conditions that shape the individual’s cost–beneﬁt analysis and therefore set the incentives for collectiveaction outcomes. In this case, class consciousness is conceived as the capacity to overcome the free-rider problem in collective action. The value of this perspective rests entirely on its basic assumption: the possibility of deﬁning some basic or minimal common ‘‘interest’’ for the members of a given social group. This is especially hard when dealing with the long-term processes of formation and transformation of social groups. The fact is that interests, and therefore costs and beneﬁts, cannot be deﬁned unless identity is ﬁxed beforehand, since the former are shaped by the latter. What is regarded as an unbearable cost with one given individual identity may well be seen as an inexpensive beneﬁt with another. In Belgium, at the beginning of the twentieth century, the costs and beneﬁts of any given collective-action choice of a Catholic Flemish worker depended on which of these three possible identities he regarded as predominant. In France, the same worker would not have been embarrassed by a possible subnational ethnolinguistic identity, but could still regard himself primarily as a Catholic or as a member of the working class. In contrast,

T

546

CLASS-CLEAVAGE STRUCTURING

547

the Swedish worker of the same period had hardly any chance to identify himself with anything but the political identity of his social reference group. These different opportunities were not chosen by individuals, but given to them by the constellation and combination of macrohistorical processes that preceded them and shaped their choices. Identities and, hence, interests were shaped by long-term macroprocesses. The timing of the struggle for nationhood, state–church relationships, and the ethnic composition of territories set limiting conditions for such choices in much the same way that the long-term development of industrialization, urbanization, and capitalism forced structural preconditions on them. The calculation of gains and losses means little if it was not referred to these options. In this book, I have giver great emphasis to the role of these limiting conditions in the context of individual choices. The test of macrohypotheses about the historical development of the European left required a set of models linking sociocultural conditions, organizational mobilization instrumentality, and the structure of political opportunities within the framework of nation- and state-building processes. Extensive theoretical discussions appear at the beginning and at the end of every chapter. Given that each discussion builds on the previous ones and progressively incorporates their conclusions into a broader model, there is only limited need for a general conclusion. These ﬁnal notes are meant to recast the logic of the exercise and to present the conclusions in a simpliﬁed and synthetic form, linking the development of socialist movements back to the macroconstellation of their integration into national mass politics.

THE MACROCONSTELLATION I will try to organize the range of factors discussed so far into a general analytical layout for the analysis of the variance in class cleavage structuring in terms of electoral mobilization, cohesion and division, social homogeneity, and ideological orientation. This macroconstellation ordering is organized around the effects of the structural restraints on the behaviors of a collectivity.1 This necessitates reasoning backward, reconstructing the general historical constellation from which the structuring of the class cleavage in each country took on its speciﬁc character. Thus, the pattern of left electoral mobilization was part of the global process of competitive 1

See Rokkan (1970d: 19–24) for a discussion of various combinations of micro-macro variables for the explanation of political behaviors.

548

THE POLITICAL MOBILIZATION OF THE LEFT

drives for electoral mobilization along different lines. As such, it was only one of the components in the global process of cleavage system structuring and political party development. The latter was part of an even broader process of mass political nationalization and institutional liberalization and democratization. These modalities of internal-voice articulation and institutionalization were shaped by three preceding or parallel processes: 1. The reduction of external exit options for individuals, groups, and territories, determined by the consolidation of external economic, administrative, and military boundaries of the state. 2. The similar process at the cultural level brought about by the cultural standardization within such boundaries as a result of the development of national loyalties and identities. 3. The powerful social inputs to the deﬁnition and redeﬁnition of social group size and internal composition brought about by industrialization, urbanization, and capitalist development. For the purposes of simpliﬁcation and schematic representation, these processes are charted in Table 10.1 under four main headings: state (formation), economy (transformation), culture (integration), and politics (democratization). For each of them, the macrovariables dealt with in the chapters are listed. The preconditions that are more or less favorable to class cleavage mobilization must be looked for in (1) the consolidation of territorial units (which allowed centralization of conﬂicts, stateness, and international status); (2) the growth of capitalism and industrialism (constituency formation and differentiation of dominant interests); and (3) cultural standardization and secularization (cultural homogeneity of the class environment). The process of cleavage structuring is seen as the result of crossing inﬂuences from the macrodevelopment in the state as Herrschaftverba¨nd, in the economy as Wirtschaft-Gesellschaft, and in culture as KulturGemeinschaft. STATE CONSOLIDATION State formation indicates the political and economic uniﬁcation at the elite level and, technically, the formation of the ﬁscal and military state. It involves the creation of an organization for the mobilization of resources: bureaucracy and the tax burden; the consolidation of the territory; the army; and the maintenance of internal order (i.e., the police and the judiciary). There are four ways in which the modalities of state consolida-

CLASS-CLEAVAGE STRUCTURING

549

Table 10.1. The macroconstellation of class cleavage structuring

tion impinged on those of electoral mobilization in general and of class cleavage structuring in particular. First, economic-functional conﬂicts tended to prevail over other divisions with the formation and consolidation of the economic and administrative center. It was only with the development of the modern state and with the integration of different groups into the central sphere of society that the conﬂicts between these groups became centralized, with the unifying of potentially opposing camps that were not concentrated territorially. Second, the territorial consolidation of the state’s external boundaries can be seen as a facilitating, if not a necessary, condition for the development of functional conﬂict at the center. Both the timing and the modes of territorial consolidation of the state political Verba¨nde inﬂuenced types and modes of the conﬂict articulation ‘‘within the Verba¨nde.’’2 The transition to mass democracy institutionalization was more gradual and less violent in those polities where territorial economic and military exit were less available. Those states that developed in relative isolation from the 2

Generalizing Robert Seeley’s dictum that ‘‘the degree of political freedom within a state must reasonably be inversely proportional to the military and political pressures on its border,’’ quoted by Hintze (1962: 366).

550

THE POLITICAL MOBILIZATION OF THE LEFT

exit opportunities concentrated in the city-belt continental zone and around it3 found it easier for their early representative institutions to survive (England, Sweden, and Norway). Where these exit opportunities were higher – at the very edge of the continental city-belt zone – the building of boundaries was the result of constant struggles against the real possibility of exit, and this led to more-profound bureaucratization, militarization, state control, and so on, leading to a higher level of resistance and a more abrupt transition to mass politics and democracy (France, Austria, Prussia, and Denmark). In the continental city-belt zone, the structuring of internal conﬂicts and voice opportunities depended very much on the inheritance and survival of the traditions of consociational ties within and among cities. Where these consociational ties were strong enough to prevent the development of a centralized state bureaucracy, this structuring was easier (the Netherlands and Switzerland). By contrast, where this tradition was weaker and the city network was only ﬁnally and with difﬁculty incorporated into larger states with an absolutist tradition, the transition was more problematic (Prussia and Piedmont). Third, external boundary consolidation is also essential in that reduced size and insecure international status may increase or decrease the burden of international politics and external inﬂuences. Low international status, on the one hand, calls for more internal cohesion and therefore less conﬂict and, on the other, for more external dependency. In contrast, high international status and involvement as a primary international actor makes internal dissent harder to accept; at the same time, it allows more autonomous internal developments. The geopolitical location of the state in international affairs is therefore of great importance in the sense that internal evolution is usually ‘‘geo-politically assisted.’’4 Finally, there is a fourth way in which external state closure and security impact on internal political voice structuring – by inﬂuencing the stateness of the domestic political processes. Bureaucratic development and the consolidation of the center is a key factor in terms of the resources it makes available to the established dominant elites. The insider’s perceptions of the costs of repression versus toleration depended on the repressive and extractive resources of the state. The amount of resistance and opposition that newcomer movements encountered, and the willingness of established elites to allow their political incorporation, were inﬂuenced by the centralization of state apparatuses, as well as by the extractive and repres3

4

On the concept and characters of the central continental city belt, see Rokkan (1973), (1974), and (1975). I borrow this inspired expression from Mann (1987).

CLASS-CLEAVAGE STRUCTURING

551

sive resources these elites could afford to use against them. Territorial fragmentation and low stateness may be seen as the reverse process. NATION BUILDING Nation formation is the establishment of direct vertical contacts between the elites and ever larger sectors of the peripheral population, as well as the development of a ‘‘national community’’ whose horizontal exchange concerns not only goods but also symbols. Technically, it is achieved by conscript armies, school education, the mass media, religious and linguistic standardization, the spread of myths, and the feeling of national identity. Under this label are regrouped those system characteristics that pertain to the cultural homogeneity of any given society. Such cultural homogeneity is in many ways an essential ingredient for the spread of a nationwide politicization of economic and functional cross-local conﬂicts. Without some form of nation formation, the development of cross-local or crosssectional functional class identities is difﬁcult. In fact, the class conﬂict and cleavage was more a consequence and a feature of nation building than a force superseding or bypassing it. In principle, we can imagine that the greater the cultural homogeneity of the class environment, the more likely it is that class positions and conﬂict will be translated into direct political allegiance and action. Cultural homogeneity has both a vertical and a horizontal dimension. The horizontal dimension includes those elements of cultural nonhomogeneity that separate segments of the society characterized by clear religious, ethnic, linguistic, and other cultural/identity differences. Each segment, although to different degrees, is composed of both mass and elite groups. On the vertical dimension, cultural homogeneity refers to the possibility of community development in Toennies’s sense. Within a given national territory, development of the possibility for cross-area communication requires the development of links between the masses and the elites, which are available only with the spread of education and the development of mass media. Religious and ethnolinguistic homogeneity inﬂuenced the basic conditions for the development of mass democracy to a strong degree. The Protestant–Catholic dividing line was important because, on the whole, in those countries where the Protestant Reform produced early ‘‘nationalization’’ of the territorial culture, this favored the processes of mobilization from below, and also because the early development of literacy aided and encouraged the mobilization of the lower classes into mass politics. Moreover, the close relationship between state and church and the incorporation

552

THE POLITICAL MOBILIZATION OF THE LEFT

of the church into the state bureaucracy reduced or eliminated the potential for a state–church conﬂict, which in turn can be said to have helped the clear deﬁnition and identiﬁcation of the center against which the opposition of the outsiders could be directed. By contrast, in Catholic territories, the supranational nature and attitude of the church tended to favor mobilization from above by the Catholic hierarchies, the late spread of literacy made it more difﬁcult to mobilize the lower classes from below, and the conﬂict over the control of the educational system and of mass education led to mobilization from the church against the state. The second consideration is the position that churches came to adopt in the process of politicization and mobilization during the nineteenth century. This is of crucial importance and depends more on the degree of pluralism and the strength of the opposition of the churches to the nationbuilding elites than on the kind of religion – whether Catholic, Protestant, or Calvinist. Religious pluralism was important because it often offered the basis for political resistance against the elites, imposing a recognition of the limitation of state power in certain spheres and afﬁrming the legitimacy of corporate powers and rights. As such, religious pluralism is a forceful factor independent of the religion in question. In Latin Europe, the dominant Catholic Church (in conﬂict with nation-building liberal elites over the defense of its vested interests) managed both to mobilize vast sectors of the lower and working classes and, in the same process, to fuel the anticlerical feelings of sections of both the middle and working classes. In so doing, it split and politically weakened both of these social groups. The absence or the lower intensity of this split between nation builders and the Catholic Church in Austria and Belgium resulted in a more pronounced superimposition of the religious and class cleavage, with a strong impact on the social homogeneity of the class support of the socialist parties. Minority opposition Catholicism – of the Dutch type – asking for separation between the (Protestant) church and the state and opposing the established elites was more of an antiestablishment movement and maintained considerable control over lower-class groups. Even Lutheranism and Calvinism were more aligned with the vested authorities and interests in some countries than in others. For Lutheranism, this was more the case in Prussia-Germany than in Switzerland, where it was more radical and modernist. In Sweden, Lutheranism was, at the same time, the basis for both an ofﬁcial national state religion and a fundamentalist opposition. Calvinism was much more antiestablishment and nonconformist in some countries than in others, depending on its relationship with the dominant elites and on whether there were stronger links with

CLASS-CLEAVAGE STRUCTURING

553

vested interests and the elites or with the lower classes.5 In conclusion, not only were the type of religion and the extent of the political signiﬁcance of religious identity important, but also the position that churches and religious groups occupied within each national system of conﬂicts and oppositions. Language and ethnic national differences play a similar and equally important role. Whether the class cleavage was structuring before, in parallel with, or after the national issue was settled made an enormous difference to its mobilizing capacity, in its institutional integration, and in its political alliance opportunities. The situation was explosive if mass democracy and general suffrage expansion took place before clear-cut cultural boundaries were deﬁned for the nation. When the channels of mass politics were opened, this easily led to conﬂict over the national deﬁnition of the political unit. The cultural heterogeneity of the class environment proved to be a dissipating force for working-class left support everywhere, but it produced different electoral returns in terms of middle-class support according to the positions that churches adopted toward the liberalconservative nation-building elites. ECONOMIC DEVELOPMENT The idea most frequently referred to in histories and analyses of the working-class movement is that the growth, structure, and ideology of the labor movement of any country were conditioned decisively by the nature of the industrialization process, that is, by the way in which the tempo and direction of industrial development evolved. This general outlook implies that cross-country socioeconomic variation had a direct impact on the variation in the political mobilization process. Under the ‘‘economy’’ label are therefore regrouped those factors pertaining to the development of capitalism – which refer mainly to the spread of the monetary economy and market relationships over the given territory – and those pertaining to industrialization-urbanization – which mainly include its effect on the changing composition and type of the labor force. The social mobilization produced by these processes is, in a general sense, a precondition for working-class political mobilization. Economic growth, industrialization, and urbanization gave rise to new social groups, determining the spread of certain new social conditions. At the same time, they increased the self-awareness of other social groups, intensifying the 5

See Daalder (1966a: 47, footnote 4).

554

THE POLITICAL MOBILIZATION OF THE LEFT

social problems and grievances for which the working classes and lower classes in general then demanded political action. They provoked the intensiﬁcation of existing conﬂicts and the explosion of latent ones, and they constituted the structural preconditions for these problems to become sources of organization and mobilization efforts. Moreover, in contrast with the original antiindustrialism of the working class, the spread of industrialization brought about the ‘‘saturation’’ of this class with industrial values that helped to overcome forms of traditional socialist consciousness based on the role of the philanthropist, the technocratic state engineer, or conversion through examples and model communities, as well as forms of traditional insurrectional and conspiratorial class action. In this sense, these social processes determined the type of ‘‘constituency formation’’ for class cleavage structuring, impacting on the size, the internal homogeneity, the territorial and workplace concentration, and the cultural distinctiveness of the putative social constituency. The economic transformation input is crucial to class cleavage structuring in a second way. The timing, tempo, and nature of industrialization and capitalist development determined the level of urban and rural dominant socioeconomic groups and differentiation of interests at the moment of socialist organization and entry into the political system. The more differentiated these economic interests, the less likely the perception of a common threat by the dominant circles and the less likely their possibility of agreeing on a common response to organized labor. On the economic front, therefore, the huge social transformations triggered by industrialization and urbanization created cross-country variation in the interest differentiation of dominant groups and in lower-class constituency formation. Describing the two polar situations, we can imagine a relatively small, socially nonhomogeneous (rural and urban), relatively dispersed working class facing the relatively low interest differentiation of rural and urban dominant circles versus a large, urban-based, and industryconcentrated working class facing well-differentiated and economically conﬂicting urban and rural dominant circles. POLITICAL DEMOCRATIZATION Finally, under the label of ‘‘politics,’’ I regroup that set of organizational, institutional, and political developments that are more directly linked to the process of left mobilization but that, at the same time, are inﬂuenced by the three previous exogenous macroprocesses. The structuring of a cleavage line is, above all, a translation into politics and through politics. The mechanisms, the path, and the actors of this translation determine the

CLASS-CLEAVAGE STRUCTURING

555

extent to which – within the framework of facilitating or hampering conditions pertaining to state and nation formation and to the social consequences of the economic transformation input – a political movement is built and characterized by a stronger or weaker, organizationally cohesive, socially distinctive, and radical electoral mobilization. Three main elements have been emphasized: the organizational consolidation of the socialist movement; the institutional openness of the political system; and the political alliance opportunities offered by the mobilization of preclass or nonclass cleavages. These aspects of mass political development are strongly linked to one another and, to a large extent, also result from the kind of combination that occurs among the other three macroprocesses of state and nation formation and economic transformation. The pattern of organizational consolidation of early socialism (party organizational consolidation, union–party relations, membership density) is far more closely related to the stateness of the polity, to its cultural homogeneity, to the closure of the institutional system, and to the lack of viable early political alliances than it is to any social-input dimension. The organizational features of socialism are more easily interpretable as responses to the state-bureaucratic environment and to the cultural segmentation environment than to any pattern of socioeconomic development. The institutional closure/openness of the political system can be interpreted as resulting from the combination of stateness levels – making available repressive resources – and the differentiation of dominant interests – creating the political opportunity to use such resources. Combined with the early or delayed partisan structuring of political competition, these variables determine the predominant response the labor movement faced; direct repression, institutional intrusion, or political-electoral control. The level of repression tended to be higher (1) the higher the stateness of the polity, which created the opportunity; (2) the lower the parliamentarization of the regime, which deﬁned how and when these resources would be used and offered possible alternative forms of control; and (3) the lower the internal differentiation of dominant interests, which increased their perception of a common threat and their agreement on a common response. A merely repressive response required a weak structuring of the party system, just as a political-control strategy necessitated the opposite. If, for instance, in Belgium, the Netherlands, or Switzerland socialists were encountered mainly in the electoral market and in the parliamentary arena, this was so because the established elite could rely on a constellation of cleavages and alliances that granted them considerable control of such arenas. By contrast, the use of direct repression and institutional intrusion by dominant groups was justiﬁed by their fear of being unable to handle

556

THE POLITICAL MOBILIZATION OF THE LEFT

the socialists or by their unwillingness to combat the socialists politically, given the inadequate political organization of liberal and conservative forces. Therefore, the national pattern of cleavage structuring and party system formation was crucial in shaping both the preferred response of the established elite and the character of the class cleavage. That is to say, it determined different openings on the electoral market, different alliance opportunities, and different social bases for the latter. In turn, these elements were important in determining the level of radicalization and class polarization of the whole socialist movement. A long tradition of studies of political development stresses the predominant importance of early alliances among major economic and social groups, assuming that the nature of the political regime resulted from the logic of the subtending socioeconomic coalitions. In considering the institutional responses to the labor movement, emphasis can be placed on the compatibility of economic demands – in other words, on whether established interests (the landed gentry and/or the bourgeoisie) perceived the socialist movement as a fatal threat to their economic position, which would lead to an alliance of landed and bourgeois interests against it. This line of analysis deduces the state/institutional response from the compatibility of class interests. It therefore assumes that the state was the guardian of such compatibility and that, of necessity, anticapitalism and antistate attitudes should combine and overlap. It also implies that in those countries that yielded more gracefully to pressures from below, the conﬂict of economic interests was reduced. Dealing with thirteen cases to avoid the tyranny of large countries and pattern cases, I have developed indicators to locate the European experiences along several dimensions of a model of institutional integration. To interpret the European variation in this outcome, I have preferred to argue that the deﬁnition of the conﬂict of interest as being more or less compatible was shaped by extraeconomic power relationships and bases. First, the possibility of using political power and the state to subordinate labor did not depend only on the willingness of dominant economic circles to do so (on the basis of the perceived threat). We can reverse the argument and assume that such willingness depended ultimately on the availability of politically repressive resources; on the historical stateness legacy, as this was shaped by warfare traditions; and on military challenges in continental Europe. These resources did not exist everywhere; and where they did exist, they not only inﬂuenced the response strategy chosen by the dominant economic circles, but they also allowed the survival of more-autonomous dynastic and bureaucratic interests that were able to choose certain economic groups as allies and set up strategies to weaken politically those

CLASS-CLEAVAGE STRUCTURING

557

social groups that they wanted to neutralize. The resources available for repression and the historical pattern of institutional development can be seen as independent variables in the deﬁnition and delimitation of the conﬂict of interests. To rely on state apparatuses, on institutional obstacles and privileges (upper houses, king, etc.), or on political control was not always a choice made by the dominant economic elites and based purely on an evaluation of interest compatibility. It is clear, in my opinion, that this perspective ﬁts the European experiences of political response far more effectively than any other. It provides a better bit to the German case, where any argument about bourgeois weakness and bourgeois–Junker interest compatibility and differentiation is comparatively untenable. It also describes more accurately the experiences of Denmark and Sweden, where the weak state resources available to curb the political power of the labor movement resulted in nonrepressive strategies, even though the dominant economic-interest differentiation was very low. This leads to another closely related point. I have emphasized somewhat the logic of political coalition resulting from the speciﬁc forms of political representation that the main social groups develop during the phase of party system structuring. Large social groups do not ‘‘make coalitions’’ and do not ‘‘have alliances’’ unless they establish political selfrepresentation, and the extent to which they represent themselves politically does not depend exclusively on their economic strength or distinctiveness. The early political organization of nonsocialist movements and the degree of structuring along partisan lines of the party system become crucial for checking the potentially autonomous role of dynasticbureaucratic elites, for using other strategies of political control of socialism, and for easing the institutional integration of the latter. The availability of a political-control strategy of the socialist movement, which permitted its containment by politico-electoral-parliamentary means (rather than by direct repression or institutional privilege), required liberalization and favorable conditions for structuring nonsocialist parties and the party system in general. Repression and institutional exclusion was more likely to be used when strategies of political containment were less viable. The earlier and more encompassing the organizational mobilization of most sectors of the community was through non- or presocialist political movements, the earlier and easier was the socialist political integration but the lower its resulting mobilization capacity. The alliance choices and alignments made by the early insider forces reduced the alliance alternatives and opportunities available to latecomer outsiders and therefore

558

THE POLITICAL MOBILIZATION OF THE LEFT

shaped signiﬁcantly their choice of allies and adversaries among both social groups and political organizations. It is not surprising that the best alliances for socialist movements were political alliances (not social alliances) made along the functional socioeconomic representation axis with politically structured and largely self-representational movements. Neither early liberalism nor agrarian movements were working-class or lower-class unity-dissipating forces. In reality, these cooperation experiences eventually split the liberals and agrarians more than they damaged the socialists. By contrast, denominational mobilization – apart from its forceful role in dissipating lower-class unity and solidarity – weakened political liberalism so as to render it occasionally very weak, occasionally an unviable ally and occasionally an undependable and hostile movement. The impossibility of political alliances with socially characterized political movements and the difﬁculty of cultural alliances (denominational, ethnolinguistic, nationalistic) forced socialism to adopt a set of complex, politically controversial social alliances with sectors of disparate social groups, which proved in most cases to be detrimental to the organizational and ideological cohesiveness of the movement.

THREE MODELS In the previous chapters, I have slowly built a model explaining variations in the structuring of the class cleavage through the three main dimensions of size, organizational cohesion, and ideological orientation. In a ﬁnal attempt to further simplify the complex web of connected factors, I have synthesized them in Table 10.2 in relation to the main historical outcomes. From left to right, the table indicates the ‘‘distal’’ and exogenous macrohistorical processes, their speciﬁc features directly relevant to our problem, and the intermediate-level variables resulting from combinations of the macrofactors. The following columns present the outcomes in a simpliﬁed form. ELECTORAL SIZE Electoral mobilization had both a developmental and a cross-sectional component that were difﬁcult to reconcile. Factors explaining over-time change failed to explain cross-country differences in the early stages. The ﬁnal model, which combined the over-time and cross-space differences into a single explanatory framework, accounting for roughly 70% of the vari-

Table 10.2. Macroprocesses and variation in outcomes

560

THE POLITICAL MOBILIZATION OF THE LEFT

ance along both the time and space dimensions, is summarized in Figure 10.1. While social mobilization points to a long-term positive impact on the left vote over time, which tended to disappear as a cross-country discriminating factor when time was parametrized into different periods and decades, cultural heterogeneity factors work the opposite way: They add less to the explanation of the temporal development of the left than to the cross-country differences at each moment in time. Because industrialization, urbanization, and the cultural homogeneity of the class context – the global context of the formation of working-class constituencies – fail to account for synchronic differences in levels of electoral mobilization, attention is needed to what intervenes between working-class formation and electoral mobilization. Variation in political translation has been attributed to three sets of variables: the pattern of franchise extension, the level of early organizational consolidation, and the closure of the institutional context in which the left developed. The pattern of franchise development inﬂuenced the ability of the left to mobilize in a rather unexpected way. The left grew stronger moving from the cases of early timing to those of late timing and, within each of these groups, moving from gradual to sudden enlargement. Ceteris paribus, the lateness, on the one hand, and the sudden change, on the other, far from hindering the electoral development of the left, seems to have created conditions that were more favorable than very early enfranchisement or more intermediate and gradual developments. The coincidence or lack of coincidence between such processes and the early organizational developments of the socialists adds to an interpretation of differences in mobilization capacity. When the major push of social and occupational dislocation of the labor force due to industrialization and urbanization paralleled the organizational development of socialist parties and the development of the franchise, socialist parties were able to capitalize on such a transformation, particularly if the organizational consolidation was party-led rather than union-led. A similar process is at work with the institutional closure of the political system. Far from being a nuisance to left electoral development, this proved to be a positive factor in general, but particularly in the early period. Before World War I, this is the single most important factor determining the left’s level of electoral mobilization. The more open and liberalized the political context, the lower and slower the mobilization of left voters. The institutional sequences that favored the integration of opposition movements (ﬁrst liberalization, then party organization, and ﬁnally universal suffrage) actually reduced the general left’s electoral

CLASS-CLEAVAGE STRUCTURING

561

Figure 10.1. The electoral mobilization model. strength, although it may have favored its long-term organizational cohesion and fostered ideological moderation. The general picture resulting from the interaction over time of these ﬁve factors is the following. Only after the role of all the other factors is controlled for does social mobilization emerge as the most stable and systematic overall force of left development in each period. If social mobilization is the developmental engine of left mobilization, cultural heterogeneity (and, even more, ethnodenominational forms of politicization) represents the long-term limiting condition of left development. The organizational factors are therefore unimportant in both the ﬁrst pre– World War I period and the post-1965 period. By contrast, they play a signiﬁcant role between 1918 and 1965. The pattern of enfranchisement is a powerful predictor of early strength or weakness, which tends to fade over time. Finally, including institutional closedness/openness into the model changes only marginally the overall variance explained with respect to the other four variables. However, its impact is decisive in explaining cross-country differences in the early period before World War I, and it is still important up to 1944. Early cross-national differences in socialist electoral mobilization, which tended to persist strongly later on, should, ﬁrst of all, be related to institutional differences in the liberalization/ democratization processes, then to the lateness and sudden nature of suffrage extension, and, ﬁnally, particularly between the world wars, to organizational development. This means that if we concentrate on the long-term general forces of left electoral development, we must look at the socioeconomic and cultural

562

THE POLITICAL MOBILIZATION OF THE LEFT

environment of class cleavage structuring. If, however, we focus on the huge cross-country differences in the timing and extent of this mobilization, we should give priority to the institutional and organizational environment of the socialist movement. Given that cross-country differences proved enduring within the context of general historical growth, it could be concluded that the development of national socialist movements in terms of the earliness and scope of electoral mobilization depended more on the state-bureaucratic-political environment than on underlying sociocultural processes. It is difﬁcult to describe ‘‘what socialism was,’’ but if we deduce its nature from the forces underlying its strength, we can conclude that it was more a movement for the national political emancipation of conscious segments of the lower classes than a movement of socioeconomic protest and revolution. Therefore, not only was class not extraneous to the question of citizenship, but the extent (and the nature; see the next section) of class appeal in electoral terms was also dependent on the conditions of citizenship. Socioeconomic positions and their associated grievances created a structural background, but the speciﬁc strength and weakness of the political articulation of these grievances was set by the type and number of obstacles blocking their expression through normal (and not even necessarily socialist) corporate, party, and parliamentary channels. IDEOLOGICAL ORIENTATION The simple dichotomy between radicalism and reformism that subtends the various ideological orientations of the early socialist movements in Europe was reshaped into four main types in Chapter 2: Orthodox revolutionary Marxism, eclectic socialism, unionism, and moderate reformist ‘‘fac¸ade’’ Marxism. In this book, the differential success of these ideological variants has not been analyzed from the viewpoint of the theoretical value of these analyses or from the viewpoint of their success or failure in terms of the accuracy of historical predictions. Instead, the issue is the conditions under which the questions and answers that these theories raise and offer were actually considered important and convincing by sections of the working population and by the socialist rank-and-ﬁle and leadership. The development of the class conﬂicts in the middle of the nineteenth century was natural. At that time, the sharp contrast was evident between the old agrarian order – maintaining traditions and providing some economic and social security, although little economic well-being – and the new order, with its dynamic, ever changing productive forces, the instability of employment, and the social and economic security it provided, and

CLASS-CLEAVAGE STRUCTURING

563

the resulting necessity for constant adaptation to new habits, skills, and concepts. For newly uprooted peasants and artisans, the clash between these two worlds was very strong, and it is no wonder that it was exactly in this period of transition – in which the two worlds were still potentially viable alternative ways of life for individuals and groups – that the opposition to capitalism/industrialism grew stronger. However, this opposition focused on capitalism/industrialism rather than on simple capitalism as such, which, after all, had existed for some time. These radical emotions rested on original anti-industrialism feelings. Marxist thinking and political action, adding to this initial emotion the logic of the progressive worship of science and technology and the faith in their limitless possibilities for industrial progress, offered working-class groups a reason and some hope for action within the industrial society (inevitable) and against capitalism (avoidable). The success of this message rested on the fact that its goal of destroying capitalism coincided with the proletariat’s instinctive reaction against industrialism. Without the antiindustrial emotion of a hitherto mainly agrarian and traditional society, early socialism would not have been a revolutionary movement. Equally, if it had had no faith in the potential of industrialism, it would have spent its energy ﬁghting the unavoidable forces of modern life.6 However, the progressive saturation of the working class with industrial values tended to deprive radicalism of its implicit revolutionary character and aspirations. Once socialism had helped the working class to become pro-industrialist, it risked losing the emotional basis for its revolutionary goals; that is, anticapitalism feelings, if deprived of an antiindustrialist underpinning, could easily turn into a demand to reform the industrial society without a revolutionary potential. Marxism provided a second source for a radical interpretation of the working-class role, linking anticapitalism and antistate attitudes and action into a coherent theoretical argument. The class cleavage rested on a divide between suppliers and buyers of labor and necessarily implied an opposition between workingclass organizations and capitalist or bourgeois interests. However, to assert that the working classes were hoodwinked by the capitalists is one thing; to assert that the state is the ‘‘executive committee’’ of this exploiting class is another. Marxism linked the class cleavage to revolution via the antistate emotion; this provided the basis for a theory that explained why the most conscious sections of the working class could not improve their conditions other than by a political revolution that would destroy the state. However, for this synthesis to become the major source and inspiration of socialist 6

On this combination of emotion and anti-industrialism, see Ulam (1979: 55).

564

THE POLITICAL MOBILIZATION OF THE LEFT

action and propaganda and a convincing answer to the problems perceived by each national working-class movement, a theory linking the state to the defense of capitalism was not enough. A reality – that is, an actual situation in which the state defended vested interests, directing its resources to the goal of curbing labor – needed to be added to the everyday experience of the lower classes. In interpreting the variations of the early ideological orientation of socialist movements, attention must therefore be paid to the extent to which the class cleavage, whose structural roots were primarily in social antagonisms in the market, came to be ideologically tinted with hostility against the state. There was nothing compulsory about this link; nothing necessarily implied that working-class political mobilization should come into conﬂict with the state. In the long run, however, it depended on three fundamental factors: the historical stateness of the polity, the degree of interest differentiation of the dominant urban and rural groups, and the partisan structuring of the party system. The ﬁrst two provided the terms for the calculations of the costs and beneﬁts of strategies of toleration or repression. The third set, at the same time, the conditions for the presence or absence of political allies for the socialist movements, and the possibility of its political control via partisan competition, available only when non socialists forces were able to organize their electoral defense in the political arenas early on (Figure 10.2). In terms of hostility toward the state, the European experience is far more varied than a British-versus-continental pattern or a French-versusGerman pattern would indicate. In addition to countries where the state was strong, there were others where it was far less visible or even deﬁnitely weak. The use of the state apparatus was also different. The relationship of national movements with their respective national states was therefore different, and the fusion of class feeling with antistate feeling depended on the speciﬁc nature of state involvement in their emancipation attempt. The opposition strategies of dominant national forces to the emerging socialist movements in the last two decades of the nineteenth century and the ﬁrst two decades of the twentieth century led to two possible reactions: (1) to help and favor the success of the Marxist identiﬁcation between the class enemy (the employers) and the state (the class nature of the state) or (2) to the recognition that the struggle for socialism was the same as the struggle for a democratic state, that is, for state democratization. The political integration of a social group into the existing political order involves cultural elements, and it is ultimately expressed through a set of beliefs and attitudes toward the political system. The development of these cultural traits results from organizational activities and from the learning experiences linked to the response to these activities. The extent

CLASS-CLEAVAGE STRUCTURING

565

Figure 10.2. The ideological orientation model. to which new actors feel they can achieve their goals within the institutional structure inﬂuences the amount of pressure exerted by the political and corporate movements on the government and the employers through electoral, parliamentary, and judicial methods or, rather, the extent to which they are prepared to use other resources and conﬂict arenas that are extrainstitutional. Thus, the more a closed political system presented itself as a moral entity independent of sectional societal interests, the more theoretically persuasive the radical ideologies appeared to be. There were deep divisions within the movement as to whether their aims would ever be achieved, and profound ideological strife in the working-class organizations occurred along radical versus moderate lines. Moderate calls for a peaceful and piecemeal insertion believed that institutional channels were open enough to allow for democratic change. Radicals, in contrast, arguing for a revolutionary breakthrough, assumed that the regime would never allow them to compete fairly. In accepting the costs of repression, the established elite assumed that the demands of the new claimants could in no way be accommodated within the existing order. For these reasons, in interpreting variations in early ideological orientations, I have placed greater emphasis on patterns of institutional and political integration, ﬁnding little evidence of direct links with socioeconomic, cultural, and organizational features. The countries characterized by the persistence of an orthodox Marxist doctrine up to the 1920s and afterward are all high-stateness countries with a high level of state repres-

566

THE POLITICAL MOBILIZATION OF THE LEFT

sion combined with low parliamentarization, high representational obstacles, and political isolation due to a lack or weakness of politically reliable allies. Moderates and reformist groups had very little external political point de repe`re, and their arguments must have appeared quite unconvincing given the circumstances. Eclectic socialism, in which radical syndicalism survived along with orthodox Marxism and various forms of reformist socialism (and occasionally even trade unionist visions), was the labor force’s divided answer consistent with a contradictory environment. In Italy and France, the repression used by high-stateness polities against the socialist movement – and, in particular, against its more radical components – was considerable, but at the same time, the parliamentarization of the regime and the relatively low level of representation obstacles triggered an ongoing debate among the socialist elite and rank-and-ﬁle about which was the best strategy for labor. These are cases of weak and late consolidation of the party system, largely because of the weakening of the conservative forces due to the ambivalent attitude of the Catholic masses to the regime, as well as to their unwillingness (or incapacity) to organize politically within a liberal regime framework. This combination of political and institutional circumstances made it likely that almost any political strategy would be challenged by sections of the socialist movement, and for almost any ideological variety to seem available and convincing to different sectors of the movement – from trade unionists to revolutionary syndicalists, from ex cathedra reformists to Leninists. This synthesis was far from unsuccessful in overall electoral terms; but clearly, ideological and organizational cohesion was impossible to achieve. In none of the countries where parliamentarization was achieved early and repression was low did these ideological tendencies exist. Either trade unionism or mild and pragmatic forms of Marxism – where theoretical lip service paid to orthodoxy was often accompanied by a moderation of practices – prevailed. ORGANIZATIONAL COHESION The model that best accounts for the cases in which the socialist movement split down the middle permanently requires further speciﬁcation and additional variables. Sluggish, incomplete, and relatively late industrialization, on the one hand, and low institutional integration and few opportunities for stable social and political alliances, on the other, represent the general context of a deeply divided left. The general background, therefore, shares most of the features of the early orthodox Marxist or eclectic socialist

CLASS-CLEAVAGE STRUCTURING

567

ideological orientation. In reality, communism made no lasting inroads into any country that had no such ideological tradition. However, we should not link communism too closely to these circumstances alone. A thesis that links late and incomplete industrialization in a causal way to the lack of self-integration of the working class, to the lower social integration of the same working class, and to more radicalism and less organizational cohesion leaves out too many intermediate steps and fails to explain why Denmark and Austria did not follow the same pattern as Italy, France, and Finland, and why Germany did so in the interwar years. Similarly, an emphasis on poor institutional integration and political isolation alone is not sufﬁcient. While it may help to include Germany between the world wars – where socioeconomic conditions seemed far from favorable to the organizational division of the socialist movement – it leaves out another deviant case: Austria, where socioeconomic, politicoinstitutional, and ideological conditions seemed the perfect preconditions for a long-lasting and profound organizational split. Two further factors need to be added to this background: the weakness of the organizational web of the socialist movement in the crucial post– World War I and post–World War II periods, and the intense class polarization of the countryside resulting from the conﬂicts between a large agricultural landless or land-insecure working population and an equally large agricultural bourgeoisie. In countries with successful communist splits, the socialist movement rested on a strong imbalance between the organizational encapsulation of both partisan and corporate dimensions, on the one hand, and far stronger electoral mobilization, on the other. In these countries, the socialist electorate was composed of relatively large sectors that were not organizationally linked to the socialist party and unions, and a delayed and massive trade union movement occurred precisely during the period of maximum appeal of the Russian Revolution. Organizational competition between the socialists and communists took place not within large, established corporate movements, but within weak organizations subject to the rapidly growing recruitment of previously nonmobilized (and mostly nonspecialized) sectors of the working class. Italy, France, and Finland all had comparatively low organizational encapsulation of the voter–party links, weak trade unionization, poor party–union linkages, and weak socio-organizational pillarization. The experience of Germany is close to that of these countries, although it tended to have higher corporate and partisan density. The three small continental countries – Belgium, Switzerland, and the Netherlands – had fairly similar organizational features, but they were also strongly pillarized and dense in sociopolitical organizations that, although not necessarily collectively afﬁl-

568

THE POLITICAL MOBILIZATION OF THE LEFT

iated with the party, offered it (and not only to the socialists, but to all major parties) a dense network of closely linked, supportive organizations. By contrast, in the culturally more homogeneous Scandinavian countries, the socio-organizational pillarization was not high, but the important role of collective afﬁliation and high organizational density in general placed them in the group of the earliest and more densely organized pre– and post–World War I movements. In the ﬁrst group, the strength of the socio-organizational pillarization of the society offered a functional counterbalance to the weak partisan and corporate density, making it less important to establish close voter–party links; in the Scandinavian countries, the strong organizational structuring of both the corporate and partisan channels compensated for the weak societal pillarization. These factors set Austria apart from the countries with profound left divisions – the impressive Austrian organizational density in all dimensions and, at the same time, the high level of socioorganizational segmentation resulting from the unique superimposition of the urban–rural, class, and religious cleavages. Together with the low class polarization in rural areas, this provides the main explanation for why the Austrian working class did not ultimately split along ideological lines, although it had favorable socioeconomic and politico-institutional conditions for doing so. In conclusion, the syndrome of the divided left is best summarized in Figure 10.3, where background conditions and speciﬁc differences are listed. Communism was the social expression of the combined support of advanced sectors of the industrial working class of economically late or backward societies, a developmental middle-class intelligentsia, and considerable sectors of the rural world that resisted their full transformation in a commercial capitalist direction. This potential base became the support for a communist split only when the socialist movement had weak organization, poor institutional integration, and little political coalition potential. The three elements interacted strongly with one other. Thus, a backward agrarian structure tended to be represented by political forces that regarded the isolation of socialism as a fundamental condition of survival; weak trade unionization also weakened the socialist movement on the market; and organizational weaknesses and poor party–union relationships could not check the internal tensions generated by phases of mass mobilization and radical politicization of the rank-and-ﬁle. Under these conditions, the alternatives to the combination of high social and cultural control, organizational weakness, and low institutional and political integration required a long-term ideological and radical nature. This was achievable only by destroying structures of control – the

CLASS-CLEAVAGE STRUCTURING

569

Figure 10.3. The organizational division model. state in primis, but also the church and the traditional social hierarchies of the rural world. The growth and persistence of a radical left opposition was linked to the existence of sections of the lower classes dissatisﬁed by the societies in which they lived. In these societies, not only business and capital, but also the state, government, and church, refused to recognize the legitimacy of their organizations, leaders, and demands, and, at the same time, the organizational weakness of the socialist subculture was fed by the heterogeneity of social groups deriving from late and unsuccessful industrialization and by the weak organizational development of the civic society. Demands against employers and demands against the state and other agencies of cultural control were combined into a single coherent ideological alternative, rather than being split and deﬂected by more powerful integration mechanisms separating the market from politics and both of these from cultural and organizational identities. This was the result of this historically speciﬁc syndrome.

CONCLUDING CONSIDERATION The literature has made several attempts to classify the different experiences of national socialism. In the 1920s, Sombart7 proposed a classiﬁcation 7

Sombart (1924: 357–360).

570

THE POLITICAL MOBILIZATION OF THE LEFT

of socialist experiences that identiﬁed three basic types of labor movements: the German, the Anglo-Saxon, and the Latin. The German type appeared to be the socialism of the professor, who builds a theory and applies it to reality; this socialism has a strong philosophical and theoretical impact on the elaboration of social and political-action strategies. The Anglo-Saxon variant, by contrast, was the empiricism of the practitioner. The Latin variant was the socialism of the artist, inﬂuenced by the role of great personalities and by the idiosyncratic character of the individual. This typology was clearly inﬂuenced by reference to the ‘‘national character’’ and to the national collective mentality. Mommsen8 has, instead, contrasted a stream of Western continental socialism (in which the syndicalist orientation was the dominant element), an Anglo-Saxon labor movement (characterized as a speciﬁcally working-class organization and a corporate movement), and a Central, Eastern, and Scandinavian variant (in which the socialist movement was dominated by the Social Democratic parties). His typology refers to both ideological and organizational features, in particular to the relationship with trade unions. M. van der Linden9 contrasts Western socialist movements (including the German and British variants) as integrated movements versus the Eastern and, in particular, the Russian working-class movement, which are viewed as the least integrated. Valenzuela10 has developed a ﬁve-factor typology of labor movement insertion into the national political process. Three types apply to the Western context: the ‘‘Social Democratic’’ type, characterized by a single union connected with a strong single party (Scandinavia, the United Kingdom, Austria, West Germany, Belgium, the Netherlands); the ‘‘contestatory’ type,’’ characterized by the communist split of both unions and the party (Italy, France, Finland); and the ‘‘pressure group’’ type, characterized by union links with a preexisting party or fragments of party (as in the United States and also the United Kingdom in the nineteenth century). None of the efforts to neatly classify the socialist movements in Europe ﬁts well with the schemes or the analysis presented in this book. The reason is probably that each of them selects and privileges one or two main dimensions of differentiation in the Western experience, while I have resorted to a multidimensional characterization of the process of class cleavage structuring, combining electoral mobilization with the social homogeneity of this support, organizational cohesion with the ideological 8 9 10

Mommsen (1979: 31). Linden (1988: 286). Valenzuela (1992). The ‘‘state-sponsored’’ and ‘‘confrontational’’ integration models refer to authoritarian regimes and are less relevant in this context, although they could be related to the early experiences of some socialist movements.

CLASS-CLEAVAGE STRUCTURING

571

orientation, and all of them with the pattern of organizational encapsulation. The classiﬁcations based on electoral size do not correspond to those based on ideological orientation; those based on organizational cohesion do not correspond to those based on social-support distinctiveness; and so on. The resulting typologies combining two or more of these features are therefore ‘‘unstable’’ in that they depend entirely on which factors are chosen as the primary classiﬁcation dimension. The frequent attempts to classify in this book were designed mainly to parametrize qualitative variables and cannot be used to obtain a ﬁnal general typology. My objective was, rather, to see how the processes of state consolidation, national cultural formation, social transformation, and political democratization intertwined with each other, producing different sets of characteristics – strong and socially distinctive movements versus equally strong but socially nondistinctive movements, conditions under which early radicalism preserved organizational cohesion or undermined it, and so on. Class mobilization and class cleavage structuring was a residual process whose coordinates and characteristics were set by the conditions of consolidation of the state-bureaucratic structures and the standardization of the national culture, as well as by the consequent capacity for organization and mobilization of preindustrial or nonindustrial groups, conﬂicts, and cleavages. If we see socialism as a ﬁnal step in the broader process of mass nationalization and integration of the lower classes into the national political order, it is natural that the setting for the direct expression of social and economic interests and grievances was fundamentally shaped by the scope of political representation, equality, and citizenship. For this reason, the latter processes proved a far more powerful determinant of the nature of political socialism than the mighty social inputs of economic development and industrialization. This is not to deny the importance of the class cleavage within the alignments of European politics. It simply suggests that the quality and nature of the class cleavage that resulted in each country in terms of mobilization size, internal organizational cohesion, social-support homogeneity, external closure, and ideological orientation can be interpreted as a function of opportunities and constraints set up by more long-standing, and in many ways premodern, forces and structures. For at least twenty years, empirical evidence and theoretical interpretations have accumulated, pointing to a transformation and/or decline of the class cleavage. Symptoms and signs of this decline have been identiﬁed in the quantitative and qualitative changes of the working-class constituency due to deindustrialization and the new division of labor; in the organizational weakening and declining linkages between political organizations, corporate groups, and citizens; in the declining social distinctive-

572

THE POLITICAL MOBILIZATION OF THE LEFT

ness of social support; in trends toward ideological undifferentiation and the decline of ideological incentives; and in the emergence of competing cleavages and party competition lines. In this work devoted to the ‘‘golden age’’ of class politics, none of these symptoms have been discussed or need to be discussed. There is a lesson, however, that can be drawn from this work: Such symptoms all point to a new macroconstellation in which the role of the state and the loosening of its economic, administrative, and cultural boundaries, the rebirth and redeﬁnition of cultural identities, the new international military and high diplomacy context, and the new and powerful processes of economic cross-national competition open up new conditions for the deﬁnition of group interests and for the direction of their claims toward new decision-making centers. These new conditions modify both the capacities and the modalities of collective organized action, leading to a new macroconstellation of political opposition structuring in loosely bounded territories, in culturally growing heterogeneous populations, and in multiple and nationally displaced decision-making processes. The interpretation of the fate of the class cleavage within this new framework is another story – and, of course, a different book.

DATA APPENDIX

PARTY COMPOSITION OF THE CLASS LEFT BY COUNTRY Country

Left Parties

Austria

Social Democrats, Communist Party, Democratic Progressives

Belgium

Workers Party/Socialist Party (later Flemish Socialist Party, Francophone Socialist Party), Communist Party, Partij V. de Arbeid, Walloon Workers Party, All Power to Workers/Labor Party, Revolutionary Workers/Socialist Workers

Denmark

Social Democrats, Communist Party, Common Course, Socialist Peoples Party, Left Socialists

Finland

Social Democrats, Socialist Workers Party, Social Democratic League, Democratic Alternative

France

Socialists, Independent Socialists, Communist Party, SFIO, EFD-PSU/Extreˆme Gauche, Divers Gauche

Germany

Social Democrats, Communist Party, Independent Social Democrats, Action for Democratic Progress

Ireland

Labour, National Labour, Independent Labour, Pro-Labour Independents, National Progressive Democrats, Socialist Labour Party, Workers Party, Democratic Socialist Party

Italy

Independent Socialists, Reformist Socialists, Socialists, Communist Party, PSI, PSIUP, Extreme Left, USI, PSDI 573

574

DATA APPENDIX (cont.)

Country

Left Parties

Netherlands

Social Democratic Workers/PVDA, Social Democratic League, Socialist Party, Revolutionary Socialist Party, Communist Party, Paciﬁst Socialist Party

Norway

Labor Party, Social Democratic Workers Party, Communist Party, Marxist-Leninists, Socialist Peoples Party, Socialist Left Party

Sweden

Social Democrats, Left Socialists, Communist Party, Hoglund Communists, Socialists, Kilbom Communists

Switzerland

Social Democrats, Communist Party, Autonomous Socialist Party

United Kingdom Labour Party, Independent Labour Party, Communist Party, National Labour, Social Democrats, Social Democratic and Labour Party.

Unless otherwise indicated, my source for electoral and parliamentary data is Mackie and Rose (1974 and following editions). For the early national elections in the nineteenth century, however, more-detailed sources have been used for a number of countries that are documented in Bartolini and Mair (1990: 314–322).

DATA FILES Census data are normally collected by decade or ﬁve-year periods; urbanization data are normally reported by decade; many other data, including those on school enrolment, labor disputes, trade union and party membership, state budgets, and so on, are available on a year-by-year basis. Finally, electoral and cabinet data are available on an irregular basis. To deal with these different time units, I have created four data ﬁles: 1. An election year database. Nonelectoral variables were included, taking the election year as reference. The rule has been to take the nearest census ﬁgures, provided that they were within five years of the election year. Otherwise, I have interpolated the data. Year-based data have been included in the election ﬁle, taking their average value in the four to ﬁve preceding years. In a few cases (when data were missing for the ﬁrst or last election year), I resorted to extrapolations (the Netherlands in 1888; Norway in 1969–1973; Sweden in 1965–1979 for the data concerning WO2AP, WP3AP, WO2WO, and WO3WO; and Sweden

DATA APPENDIX

575

in 1969 for all WO data [see the list of variable names at the end of the Data Appendix]). 2. A census year database. In this case, the other types of data have been considered for periods around the year of the census. Election data, for instance, refer to the two nearest elections, one before and one after the census date. When the census year is also the election year, the number of elections considered becomes three. In a few cases, the single nearest election has been used because it was the only one available. 3. A decade database. In this case, all variables not measured on a decade basis have been averaged for the period concerned (note: the ‘‘1910’’ decade indicates the 1910–1919 period). 4. A cabinet data ﬁle, which includes the 378 cabinets formed in the 1918–1966 period. At the end of this Data Appendix, only the election ﬁle codebook is documented.

NOTES CONCERNING THE SOCIOECONOMIC DATA When not otherwise speciﬁed, my main source for the socioeconomic data is Flora, Alber, Eichenberg, Kohl, Knaus, Pfenning, and Seebohn (1983) and (1987). Sometimes the data have been adjusted to my speciﬁc needs. GENERAL 1. ‘‘Dependent’’ refers to employees plus workers. 2. ‘‘Workers’’ are deﬁned as low-status, dependent laborers or blue-collar personnel, including home workers. 3. Unless otherwise stated, ‘‘workers’’ always include the ‘‘apprentices,’’ deﬁned as ‘‘persons in vocational training in ﬁrms.’’ 4. Unless otherwise indicated, ‘‘workers in agriculture’’ exclude ‘‘family workers,’’ that is, assisting spouses, children, and relatives. • Austria: Data before World War I refer to the Austrian part of the Hapsburg Empire as deﬁned by the 1867 constitution. 1880–1990: these censuses combine workers and family workers. I

576

DATA APPENDIX

have estimated the ﬁgures for the workers on the basis of the 1900 census. • Belgium: 1880–1890: family workers are not included in the labor force 1880: workers and employees are grouped together. I have estimated the respective weight on the basis of the 1890 census. Note, however, that for the core sectors the proportion of employees is very small. • Denmark: 1880–1890: workers and employees are grouped together. I have estimated the relative proportion as indicated for Belgium in 1880. In these two censuses, the category ‘‘workers’’ may be underestimated due to the existence of the category ‘‘servants,’’ including people ‘‘engaged in domestic work.’’ Moreover, the two categories of ‘‘workers’’ and ‘‘servants’’ must also include ‘‘family workers,’’ a category normally kept separate. In fact, family workers are included in the labor force, but with a classiﬁcation that is unknown. In this case, I was obliged to combine ‘‘workers’’ and ‘‘servants,’’ preferring the likely overestimation resulting from the inclusion of the family workers to the underestimation resulting from the global exclusion of servants. 1930: the distinction between ‘‘employees,’’ ‘‘workers,’’ and ‘‘family workers’’ is available only for some sectors. The ﬁgures are reconstructed using interpolation from the previous (1920) and following (1940) censuses. • Finland: 1910–1940: ‘‘independent in agriculture’’ includes ‘‘employees.’’ Given the low percentage of employees in agriculture anyway, and in particular in this period (in all countries less than 1%), I have felt no need to modify the data. 1910–1930: ‘‘independents as a percentage of the active population’’ (INDI) and ‘‘independents in agriculture as a percentage of total independents’’ (INAGIN) also include ‘‘employees.’’ I have estimated these ﬁgures by extrapolation, and I have checked for estimates in alternative source. • France: French data present the greatest problems of comparability, and in many cases I was obliged to resort to missing values. 1881–1891: no category for ‘‘family workers’’ is available. They are probably included in ‘‘workers’’ and in ‘‘employers.’’ The category ‘‘workers’’ is therefore inﬂated compared to the other countries, and I left out all data when it was impossible to exclude the family workers. By contrast, ‘‘workers in mining, manufacturing, utilities, construc-

DATA APPENDIX

577

tion, and transport’’ are considered as a percentage of the active population (but not of the overall ‘‘workers’’), given that the percentage of ‘‘family workers’’ is very small in these sectors. 1881: ‘‘workers in construction’’ are included in the category ‘‘manufacturing.’’ The percentage of ‘‘workers in mining and manufacturing’’ on the active population is therefore missed. 1896–1901: employers and workers are classiﬁed together. The percentages have been estimated by interpolation of the previous and following census. 1896–1936: a total ﬁgure for ‘‘family workers,’’ but no breakdown by sector, is available. No agricultural occupation ﬁgures can therefore be properly computed. I computed only the percentage statistics for industrial workers, where the weight of family workers could be overlooked. 1901–1936: ‘‘casual workers’’ and ‘‘workers’’ have been summed together. 1946–1980: ‘‘employees’’ and ‘‘workers’’ are included in the same category. The percentages concerning ‘‘workers’’ can be accurately computed only for those cases in which the percentage of employees is marginal with respect to the active population, that is, ‘‘workers in agriculture’’ and ‘‘workers in mining and manufacturing.’’ For ‘‘workers in construction, utilities, and transport,’’ the proportion of employees to workers is higher, and therefore, statistics are overestimated in this sector. • Ireland: ‘‘workers’’ and ‘‘employees’’ are grouped together. Corresponding statistics concerning workers have not been computed in those cases in which the percentage of employees can be considered irrelevant with respect to active population, that is, ‘‘workers in agriculture’’ and ‘‘workers in mining and manufacturing.’’ For ‘‘workers in construction, utilities, and transport,’’ see the text for France. • Italy: 1881: separate ﬁgures for ‘‘family workers’’ are not available, but they were probably not included in the labor force. Workers and employees are grouped together; the corresponding statistics have been computed only for the cases listed previously under Ireland. 1901–1921: ‘‘family workers’’ are included in the ‘‘independents,’’ which, therefore, are comparatively overestimated. 1936: ‘‘workers’’ and ‘‘servants’’ are grouped together.

578

DATA APPENDIX

• The Netherlands: 1899–1930: ‘‘family workers’’ are included in part (assisting spouses) in the self-employed status group and in part in other unknown status groups. In this case, statistics concerning ‘‘workers’’ have been retained, even if, particularly in the case of ‘‘workers in agriculture,’’ they result in some overestimation compared with other countries. • Norway: 1930–1946: ﬁgures concerning ‘‘workers in agriculture’’ are doubtful when compared to those in previous and following censuses. Corresponding statistics have been interpolated from the 1920 and 1950 censuses. 1930–1950: separate ﬁgures for ‘‘family workers’’ in agriculture and services have been derived from the respective status and sector categories. 1960: ‘‘family workers’’ in agriculture are included within ‘‘workers.’’ Their proportion has been estimated by interpolation from the 1950 and 1970 censuses. 1960: ‘‘employees’’ and ‘‘workers’’ are grouped together. The relative proportion have been derived by extrapolation from the 1946 and 1959 censuses. 1970: ‘‘workers’’ and ‘‘employees’’ are grouped together; no reliable estimation was possible. Only the proportion of total workers was estimated by extrapolation from the previous census. • Sweden: 1880–1920: the ﬁgures for ‘‘family workers in agriculture’’ have been subtracted from the ﬁgures for ‘‘workers.’’ 1965–1975: ‘‘workers’’ and ‘‘employees’’ are grouped together. No statistics concerning ‘‘workers in agriculture’’ or ‘‘workers in industry’’ can be accurately calculated. I have extrapolated these data from the 1940 and 1960 censuses. In this case, therefore, the same argument developed for France and Ireland applies to Sweden. • Switzerland: 1888: ‘‘employees’’ and ‘‘workers’’ are grouped together. The percentage of workers has been estimated on the basis of the 1890 census. • United Kingdom: The separate censuses of (1) England and Wales and (2) Scotland have been summed together. 1891–1901: ‘‘family workers’’ are excluded from the labor force. 1891–1951: ‘‘employees’’ and ‘‘workers’’ are grouped together. Only those sectors in which the number of employees is small have been

DATA APPENDIX

579

considered: agriculture, mining and manufacturing, construction, and transport. As already mentioned, the last sector is the most problematic, and its inclusion in the core working class determines its slight overestimation. No ﬁgures are available concerning the whole category of ‘‘workers.’’ 1931: an ‘‘out of work’’ category has been deducted from the labor force in order to maintain comparability with other censuses. 1931: ‘‘family workers’’ are included in the labor force, but with an unknown classiﬁcation. 1931: ‘‘managers’’ are included in ‘‘employees and workers,’’ as in previous censuses. 1961–1971: The category ‘‘unknown,’’ including ‘‘out of work,’’ has been excluded from the labor force. THE TEMPO OF INDUSTRIALIZATION (DATA DISCUSSED IN CHAPTER 3) Rates of sector transformation (active in industry as a percentage of the active population) 1880– 1890

1890– 1900

1900– 1910

1910– 1920

1920– 1930

1930– 1940

1940– 1950

1950– 1960

1960– 1970

country mean

⫺0.1

0.9

2.2

4.9

4.9

⫺1.6

5.4

3.5

1.0

2.3

Belgium

2.3

4.1

3.8

0.8

1.9

0.5

⫺1.3

⫺1.3

⫺2.0

1.0

Denmark

1.8

⫺0.5

0.9

1.0

⫺0.1

5.4

1.9

3.4

0.2

1.6

2.0

1.7

3.7

9.2

3.8

2.8

3.9

Austria

Finland France

2.1

1.3

⫺0.2

1.7

0.9

⫺3.0

6.5

3.0

⫺4.0

0.9

Germany

3.6

2.5

1.1

1.1

⫺1.4

2.3

0.2

5.9

0.7

1.8

3.4

0.4

6.1

2.5

4.6

3.4

4.5

0.5

1.5

9.0

2.8

1.7

Ireland 3.1

⫺2.9

1.2

1.4

1.4

0.6

⫺1.2

3.8

3.8

⫺6.4

0.6

4.0

⫺0.9

1.0

⫺0.4

5.7

4.2

0.0

0.8

1.8

5.3

5.0

5.7

0.5

6.1

3.6

3.5

⫺5.0

3.2

Switzerland

3.5

0.5

0.9

⫺1.3

⫺0.2

2.6

4.2

⫺2.2

1.0

Un. Kingdom

1.1

⫺1.0

⫺2.6

4.1

1.6

1.5

⫺1.8

⫺5.2

⫺0.3

2.0

1.4

1.3

1.5

1.6

3.5

3.0

⫺0.9

1.7

Italy Netherlands

⫺1.6

⫺1.6

0.4

Norway Sweden

decade mean

4.5

1.6

580

DATA APPENDIX

Rates of status transformation (dependent labor force as a percentage of the active population) 1880– 1890

1890– 1900

1900– 1910

1910– 1920

1920– 1930

Austria

5.8

⫺1.0

Belgium

1.2

Denmark

0.4

10.0

0.5

⫺2.4

⫺3.5

7.2

2.5

⫺1.3 ⫺2.4

⫺1.0

0.3

0.2

7.4

⫺1.6

1.2

1.2

Finland France Germany

1930– 1940

1940– 1950

11.0

⫺6.4

4.5

⫺2.3

⫺3.5

2.0

Ireland

1950– 1960

1960– 1970

country mean

3.8

6.1

8.0

4.2

3.1

1.0

1.0

7.9

1.6

6.3

0.3

5.0

4.1

2.3

1.2

6.0

14.4

7.6

13.2

6.9

⫺1.1

⫺0.3

7.3

8.9

5.5

1.9

1.2

1.1

1.1

6.8

6.4

2.8

1.6

5.3

4.6

4.7

5.1

4.3

1.0

8.2

⫺3.8

⫺5.5

2.0

2.0

10.4

3.6

2.1

Netherlands

0.6

1.4

4.0

⫺1.4

⫺7.7

4.8

3.7

5.8

1.4

Norway

1.3

2.4

4.9

⫺6.3

4.3

6.9

9.5

2.7

3.2

Italy

Sweden

1.1

⫺0.6

Switzerland Un. Kingdom decade mean

1.3

3.7

7.4

9.0

1.0

7.0

4.4

7.4

6.0

5.0

11.0

⫺1.2

4.8

5.9

1.6

3.6

5.9

4.1

4.5

2.4

⫺2.0

5.3

⫺11.5

6.3

6.2

⫺1.1

⫺3.8

0.2

2.2

1.7

2.9

⫺0.3

2.2

4.6

5.8

5.3

3.0

Rates of working-class constituency growth (workers in mining, manufacturing, utilities, building, and transport as a percentage of the active population) 1880– 1890

1890– 1900

1900– 1910

Austria

1.7

0.9

2.9

Belgium

2.4

4.8

Denmark

⫺3.5

2.3

Finland

1910– 1920

1920– 1930

1930– 1940

1940– 1950

1950– 1960

1960– 1970

country mean

2.3

2.3

0.9

3.3

2.3

⫺0.5

1.8

4.6

3.7

⫺0.5

⫺2.3

⫺1.6

⫺1.6

⫺4.0

0.6

2.1

0.4

0.4

6.4

0.9

2.0

⫺0.7

1.1

0.9

1.2

1.5

4.2

6.8

3.3

3.0

3.0

France

0.6

⫺0.8

2.2

3.1

0.1

0.7

3.4

3.9

⫺1.4

1.3

Germany

4.8

3.6

2.4

2.3

⫺17.2

14.9

⫺0.4

4.1

⫺3.2

1.3

2.9

2.3

5.1

3.7

5.2

3.8

⫺3.3

⫺3.4

4.4

1.7

2.3

40.4

1.9

8.6

⫺1.1

1.3

2.3

2.8

1.6

2.5

0.0

⫺5.7

3.1

3.0

⫺7.1

0.3

Ireland Italy Netherlands

DATA APPENDIX

581 (cont.)

1880– 1890 Norway Sweden

4.2

1890– 1900

1900– 1910

4.1

0.4

4.9

5.8

1910– 1920

1920– 1930

1930– 1940

1940– 1950

1950– 1960

5.9

⫺0.1

6.2

1960– 1970

country mean

4.1

3.2

0.0

0.0

2.1

1.9

2.7

2.7

0.1

0.2

3.2

Switzerland

9.9

⫺2.3

1.0

2.0

⫺1.0

5.4

0.4

⫺5.0

1.3

Un. Kingdom

4.0

⫺0.5

0.4

⫺7.9

5.7

5.7

⫺3.5

⫺14.8

⫺1.4

3.0

2.0

2.6

⫺1.0

2.6

3.0

2.0

⫺2.3

1.4

decade mean

1.2

Rates of working-class homogenization (workers in mining, manufacturing, utilities, building, and transport as a percentage of the total working class) 1880– 1890

1890– 1900

1900– 1910

1910– 1920

1920– 1930

1930– 1940

1940– 1950

Austria

4.5

4.3

7.3

⫺0.6

⫺0.7

11.4

Belgium

2.6

8.6

11.8

3.0

7.1

⫺2.2

Denmark

16.0

4.4

4.9

4.2

0.7

5.5

5.4

Finland France

1.3

5.3

5.1

5.0

2.8

Germany

3.3

9.9

4.6

4.5

⫺7.5

1950– 1960

1960– 1970

6.5

5.2

1.4

4.4

1.4

1.4

⫺4.7

3.2

9.6

4.4

7.6

2.4

6.0

9.5

13.3

6.1

2.0

7.0 (3.9)

10.3

⫺0.8

8.2

1.7

Ireland

4.4

Norway Sweden

10.9

Switzerland

0.0

0.9

7.1

12.0

0.0

1.6

8.7

2.3

4.1

5.6

2.7

2.7

1.8

⫺5.2

9.9

9.8

⫺8.6

2.6

8.1

1.2

7.7

5.1

8.5

0.3

7.2

1.4

4.9

9.8

9.6

7.4

1.2

8.4

6.6

2.0

2.7

6.5

10.4

1.8

⫺1.7

⫺1.0

⫺2.1

10.1

⫺1.8

⫺4.5

1.4

Un. Kingdom decade mean

3.8 n.a.

Italy Netherlands

country mean

n.a. 6.1

6.6

5.0

4.1

2.4

4.8

5.3

5.4

⫺0.4

It is important to emphasize that the means by decade are country means, that is, they are the mean value of the thirteen (or fewer) European countries.

4.3

582

DATA APPENDIX

CULTURAL HETEROGENEITY DATA RELIGION AND LANGUAGE The data concerning the religious and linguistic composition of the European population are taken from Flora et al. (1983: 55–85). For some countries the information is not available in census data: for instance, the religious composition of the United Kingdom and the Protestant minorities in France, Belgium, and Italy. Similarly, no statistics about linguistic minorities exist for Italy, France, and Denmark. When this systematic source was not available, I resorted to the estimates provided by Urwin (1980: 193–194). LITERACY The ideal statistic for literacy estimation is the percentage of the people able to read (or read and write). Such data are available for only a ﬁve countries: Austria, Finland, France, Ireland, and Italy. Other, not strictly comparable, data for Belgium, Germany, the Netherlands, and Sweden concern the literacy level of army recruits and married people. I have utilized the estimates of Flora (1973: 245) on the percentage of the population able to read and write. An alternative source is the statistics concerning the number of pupils enrolled in primary education as a percentage of the 5–14 age group. The comparability problems are described in Flora et al. (1983: 55, 70–85). Enfranchisement a. The electorate as a percentage of the population twenty years of age and older: 1830–1880 Year 1831

Au

Be 1.9

De

Fi

Fr 0.8

Ge

Ir

1834

1.9

Sw

Sz

UK 3.8

2

5.9

0.8 2.1

9.8

0.9

1838 1839

No

9.9

1835 1837

Ne

1.7

1832 1833

It

10 0.9

6

DATA APPENDIX

583

a. (cont.) The electorate as a percentage of the population twenty years of age and older: 1830–1880 Year

Au

Be

De

Fi

Fr

Ge

Ir

It

Ne

1841

Sw

Sz

UK

9.7

1842

1

1844

9.4

1846

1.1

1847

1.8

1848

3.1

1849

2.7

6.8 9.4 ⬃30

36.3 25.7

43.4

1850

9.3

1852

25.6

1853

25.1

1854

25.7

42 4.6

1856 1857

No

9

8.9 3.3

1858

40.8 25

1859

8.9

1860 1861

25.3

3.4

1862

8.8

1863 1864

41.2 3.6 24.6

1865

3.5

1866

25.3

1866

25.5

8.8

6.7

1867

8.3

(35)

1868

8.6

1869 1870 1871

25.8

42

7.4

3.7

14.5 3.5

43.7 33

7.7

5

8.5 14.9

584

DATA APPENDIX

a. (cont.) The electorate as a percentage of the population twenty years of age and older: 1830–1880 Year

Au

Be

De

1872

26

1873 10.31

26.5

Fi

Fr

Ge

Ir

It

Ne

No

Sw

Sz

UK

9.8 8.4

1874

36.2

3.6

1875

10.2

1876

26.7

42

3.8

1877

41.8 36.9

1878

37.4

1879 10.41

8.3

10.5

26.9

8.3

1880

3.8

5.4

b. The electorate as a percentage of the population twenty years of age and older: 1881–1917 Year

Au

Be

De

1881

27.1

1881

27.1

Fi

Fr

Ge

41.6 36.2

Ir

It

Ne

No

8.2

1882

Sw

Sz

10.7 38.7 16.4

12.1

9.4

1883

16.5

1884 1885 13

3.9 27.8 1

36.8

10.9 38

41.3

11.4

1886

29.0 5.7

1887

28.3

37.3

10.1 38.1

1888

11.8 11.8

1889 1890

29.4

41.8

28.9

37.4

15.2

10.4 38.3

1

1891 12.9

11.5 12.6 3.9 29.3

1893 1894

29.3

27.4 14.1

1886/ 1887

1892

UK

37.8

16.6

10.7

41.8 37.3

11.3 16.4

29.3

DATA APPENDIX

585

b. (cont.) The electorate as a percentage of the population twenty years of age and older: 1881–1917 Year

Au

Be

1895

De

Fi

Fr

Ge

29.5

Ir

It

Ne

No

Sz

11.8

UK 28.9

1896 13.4/ 35.72

10.8 38.2

1897

11.7 20.9 16.6

1898

30.0

42.0 37.8

1899 1900

Sw

11.5 38.0 37.7

12.3

34.8

28.5

1900/ 14.2/ 1901 34.12 1901

29.0

8.3

1902

21.2 43.2

1903

29.1

1904

12.7 37.9 38.3

9

34.4 13.5 24.4

1905

14.0 37.4

1906

43.7

1907 37.9

76.2

1908

75.9

1909

29.8 75.6

1910

30.1 75.5 43.4

1911 38.0 1912 1913

35.2 38.3

15.8 37.5 15.0 25.7 58.5 28.7

75.7

32.5 37.0

38.2

38.7 30.1 77.8

1914

60.2 42.0 27.6

42.8

1915

28.5

32.8 36.3 77.1

1916

75.4

1917

74.5

32.3 38.6

586

DATA APPENDIX

c. The electorate as a percentage of the population twenty years of age and older: 1918–1944 Year

Au

Be

1918

De

Fi

Fr

Ge

69.1

1920 90.1

74.1 43.4 97.9 74.0

No

Sw

UK 74.8

40.1

95.1

33.0 52.5

73.4

77.5

1923 90.0

86.9 87.9 80.7

40.3 74.5

97

1924

75.1

79.6 73.5 39.9 98.5

87.4 88.2

45.2

1926

81.8

75.6 40.4

82.0

1927 92.6

74.8

1928

95.4

88.4

40.0 97.9

88.5 40.7

45.3 80.6 76.5

1930 89.9

82.1

75.5

98.5

95.5 89.6

1931 1932

Sz

39.3 80.4

48.8

45.5

1922

1929

Ne

74.1

1919 85.9 43.8

1925

It

74.2

1918

1921

Ir

41.0 97.0 54.0 82.0

1933

39.6 98.4 93.7 75.9

97.6 94.9

89.0 82.9 91.0

1934 1935 1936

83.9 45.6

42.4 97.4 77.1 40.1

92.6 90.1

1937

95.0

1938

95.0

1939

85.2

45.1 84.6 77.8

42.3

1940

90.6

1941 1942 1943 1944

85.4

96.9 97.2

43.4 91.8

DATA APPENDIX

587

d. The electorate after World War II Year

Au

Be

1945 69.4 1946

De

Fi

Fr

Ir

It

45.5

95.4 95.0 89.2 95.6

96.0

95.7

42.9 97.6 89.7

90.5

97.3 97.4

98.0

97.8

96.3 42.3 97.0 87.9

1957

93.1 93.9

90.5 97.5 97.9

96.7 88.2

95.3 97.3

96.6

1959 95.2

96.0 91.3

1960

40.8 97.5

93.2

94.9

94.4

97.2 97.9

1962 96.4

96.9

97.3 86.4

1963

98.7 90.2

1964

38.4

97.0

93.5

92.9

95.6 98.2

95.7

96.0

97.1 95.4

95.3

1967

93.6 92.8 96.2

85.5

1969

95.7 99.0

99.8

97.1

1971 97.3 94.3 97.0 1972

94.7 99.8

99.0

38.2

98.9 93.8 99.5

1970 98.7

1973

95.6

94.1

1956 94.1

1968

96.3

96.0 97.3 83.0

1954

1966 96.4

99.6

91.4 88.2

1953 93.6

UK

43.7

1952

1965

Sz

95.0 90.0

98.9

1951

1961

Sw

87.0

1949 89.3 91.5

1958

No 91.2

88.0

1948

1955

Ne

86.7 96.5 88.3

1947

1950

Ge

98.8 87.5

80.8

98.9 99.8 98.1

99.8

99.4 96.3

588

DATA APPENDIX

d. (cont.) The electorate after World War II Year

Au

1974 1975 98.0

Be

De

Fi

Fr

Ge

Ir

It

Ne

No

Sw

Sz

94.3

UK 99.8

98.5 99.8

83.5

Notes on Austria: 1 The percentage refers to the combined electorate of the Second and Fourth Curiae, which represented the urban and rural male electorate and can, therefore, be added together. The electorate of the First and Second Curiae, representing the large landowners and the chambers of commerce and trade, respectively, included only about 5,000 and about 550 electors and can be left out, with no major distortion. 2 The ﬁrst ﬁgure corresponds to the Third and Fourth Curiae as indicated in note 1. The second ﬁgure corresponds to the electorate of the Fifth Curia added in the 1897 reform. In this curia, there was universal and equal suffrage for male citizens over age twenty-four. However, citizens of the ﬁrst four curiae gained a second vote in the Fifth Curia. The two ﬁgures cannot be summed; instead, they should be subtracted. The total electorate over the population twenty years of age or more should be between the ﬁrst and second ﬁgures, but much closer to the second ﬁgure.

SOURCES OF PARTY MEMBERSHIP FIGURES Membership ﬁgures for the left parties are scattered throughout a variety of often conﬂicting sources. From the 1960s on, a systematic and reliable comparative time series is available, thanks to the research project on party organization directed by Mair and Katz (1992). The database with absolute ﬁgures is presented in Katz and Mair (1992). Other data collections are either national (the Dutch collection of Koole and Voerman [1986: 115– 176]), limited in comparison (for Scandinavia, Sundberg [1987]), or general collections of secondary sources, such as Beyme, [1985]). See also Katz (1990). For the 1880–1960 period, I have relied both on national experts’ advice and on secondary sources. For their help in collecting national data and for their useful suggestions, I would like to thank Wilfried Dewachter, Peter Gerlich, Pertti Pesonen, Gunnar Sjo¨blom, and Henry Valen. The following secondary sources were used: • Austria: 19l3–1932 period: Leichter (1964); 1945–1977 period: Berichte an den Bundesparteitag der SPO (1945 ff.). Systematic, detailed historical data are presented in Maderthaner and Mu¨ller (1996), which became available too late to be used in this work.

DATA APPENDIX

589

• Belgium: 195l–1963 period: Kendall (l975); 1966–1975 period: Rowies (1977). • Denmark: Data collected by Thomas (1977) from Bertold, Christiansen, and Hansen (1954–1955); personal communications from SD Party Headquarters, Medlemsudviklingen 1944–1974. • Finland: Figures provided by Pertti Pesonen. • France: The ﬁgures for the membership in the Socialist Party are many and conﬂicting. I have used the following: 1905–1954 period: Duverger (1967); 1956–1970 period: Hurtig (1971); 1971–1975 period: Criddle (1977) and ofﬁcial ﬁgures declared at the Party Conference of Nantes (Le Point, June 27, 1977). Different ﬁgures are provided by Braunthal (1945 ff.) and by Simmons (1970). The Socialist International Yearbook ﬁgures, although not coincident with the Duverger-Hurtig ﬁgures, are very similar. By contrast, Simmons’s ﬁgures are consistently lower than the Duverger-Hurtig ﬁgures for the postwar period. • Germany: 1906–193l period: Braunthal (1945 ff.); 1946–1975 period: Paterson (1977) and Jahrbu¨cher der SPD. • Ireland: 1964–1978 period: ofﬁcial party ﬁgures collected by Peter Mair. • Italy: 1896–1922 period: ofﬁcial party ﬁgures in Cannarsa (1950); 1945–1967 period: ofﬁcial party ﬁgures in Cazzola (1970); 1970, 1972, and 1973: PSI (1975); 1974 and 1975: ofﬁcial party ﬁgures in PSI (1976). • The Netherlands: 1895–1939 period: Vorrink (1945); 1945–1946 period: Braunthal (1945 ff.); 1947–1976 period: Wolinetz (1977); 1977: Daalder (1987b) and Daalder and Schuyt (1986). • Norway: Figures provided by Henry Valen. For nonelection years, the following sources have been used to the extent that they coincided with the ﬁgures for election years provided by Valen: Heidar (1977), Duverger (1967), and Torgersen (1962). For the period after 1950, Heidar’s ﬁgures do not coincide with Valen’s ﬁgures, and I have used the latter. • Sweden: Ofﬁcial party ﬁgures in Scase (1977). • Switzerland: 1902–1954 period: Duverger (1967). • United Kingdom: Annual report of the National Executive Committee of the Labour Party from Pelling (1976b). The party ranks were open to individual members in 1918, but the ofﬁcial party ﬁgures for individual members are available only from 1928 on. In addition to these sources, additional information is drawn from Sworatowski (1973) for a number of small communist parties.

590

DATA APPENDIX

SOURCES OF TRADE UNION MEMBERSHIP FIGURES Special thanks are due to Jelle Visser, Hans Hirter, and William K. Roche for their help in collecting the trade union membership data. Trade union membership ﬁgures come from two main sources: Visser (1987) and Bain and Price (1976). Neither of them covers all countries included in this work. Jelle Visser made available to me his Trade Union Membership Database (Amsterdam: Department of Sociology, March 1992), where additional data are available covering Ireland, Finland, and Belgium. Other supplementary information has been drawn from various sources: for Finland, Tyo¨ela¨ma¨n Suhteet (1990) (made available by Jelle Visser) and Suviranta (1987); for Ireland, Roberts (1958–1959: 95), MacCarthy (1977: 622) and, Roche (1990) and (1992); for Switzerland 1881–1913: Balthasar, Gruner, and Hirter (1988: 66–68, 155, 158).

INDEX OF COALITION POTENTIAL The index of coalition potential is calculated for the socialist parties, the communist parties, and the combined total left. I have used a modiﬁed version of the Shapley formula. The original index and the modiﬁcations are documented in Bartolini (1998). I acknowledge the research assistance of Stefano Bocconi, who developed the modiﬁed version of the Shapley– Shubik index, and of Massimiliano Miglio, who actually calculated the index.

CLASS CLEAVAGE INCLUSIVENESS AND DISTINCTIVENESS To estimate how many members of the working class voted for the left parties (INCLUSIVENESS) and how many left voters were members of the working class (DISTINCTIVENESS), I have consulted the survey evidence of the 1950s and 1960s. Survey data normally refer to individual parties and have to be recalculated for the whole left. I have taken the category ‘‘workers’’ or ‘‘working class’’ as deﬁned by each national source, and I have not performed any cross-country standardization of the notion. This means that the data are biased for any systematic difference in the survey deﬁni-

DATA APPENDIX

591

tion of what a working-class member is. The following notes document my sources and my ﬁnal choices. • Austria: For 1969, I have taken the percentage of workers voting left (73%) from Haerpfer and Gehmacher (1984) and the percentage of left voters who are workers from Pelinka (1978: 416). For a more detailed analysis of the IFES surveys of 1969, 1972, and 1977, see Haerpfer and Gehmacher (1984) and Gehmacher (1974: 55–75). • Belgium: The level of class inclusiveness in 1968 (47%) is taken from Hill (1974: 83); the value for class distinctiveness (60–61%) is taken from Delruelle, Evalenko, and Fraeys (1970: 49). Dogan (1960: 28– 29, Tables 1 and 2) produces different ﬁgures for the 1952–1954 period – 63% and 68%, respectively. • Denmark: A 1957 Gallup poll reported in Worre (1979: 68–82) indicates that 80% of workers supported left parties. Dogan (1960) reports 73% for 1953–1955. Berglund and Lindstro¨m (1978: 108) report that in the mid-1960s 85% of workers voted for the left (2% for the Communists, 10% for the new radical left, and 73% for the Social Democrats). On the basis of 1971 data, Borre (1980: 253) reports the ﬁgure of 75%. I have taken the intermediate ﬁgure of 80%. Dogan (1960) estimates the percentage of left voters who are working class as 76% in 1952–1953; Borre (1980: 258) estimates it as 77% in 1971. • Finland: Pesonen (1968) and Allardt and Pesonen (1967: 342) report that about 68% of workers supported left parties in 1958 and that 82% of those supporting left parties were workers. Based on the same survey, however, Dogan (1960) reports widely diverging ﬁgures: 80% and 50%, respectively. Berglund and Lindstro¨m (1978) estimate that 80% of workers voted left in the mid-1960s. Finally, the 1975 survey analyzed in Pesonen and Sa¨nkiaho (1979: 121) indicated that 60% of workers voted for left parties and 60% of those supporting the latter were workers. I have taken the 1958 ﬁgures. • France: Various sources provide different ﬁgures, but all within a relatively narrow range. Converse and Pierce (1986: 154–156) estimate between 48% and 51% of workers voting for the left. Their calculations are based on subjective social class and on the manual–nonmanual dichotomy. Stoetzel (1955: 118–119) provides an estimation of about 62%, Dogan (1960) offers the value of 67%, and Adam, Bon, Capdeville, and Mouriaut (1971: 200) give the ﬁgure of 57% based on a 1969 survey. I have taken an average value of 55%. An IFOP poll of 1952 (Sondages, 1952, n. 3) gives the ﬁgure of 48% for the left voters

592

•

• •

•

•

DATA APPENDIX

being workers; a second survey for 1956 (Sondages, 1960, n. 4) gives the ﬁgure of 48–49%. Dogan (1960) suggests 54–55%. Successive polls in 1967, 1968, and 1973 consistently report ﬁgures between 42% and 47% (see Charlot [1973: 55 ff.]). I have chosen the average ﬁgure of about 45%. Germany: A complete list of survey results estimating class inclusiveness is provided by Hildebrandt and Dalton (1978: 72). The ﬁgures are 59% for 1953, 61% for 1957, and 58% for 1961. Dogan (1960) offers a lower estimate of 48%, and Urwin (1974: 147) arrives at the same ﬁgure. This suggests an average level of 53–54%. As for the distinctiveness of the left, Linz (1959: 195, 200) suggests that between 68% and 70% of manual workers supported the left in the late 1950s, while Dogan (1960) suggests the ﬁgure of 66%. A ﬁgure of 68% has been used. Ireland: In 1969, the ﬁrst Gallup poll indicated that 27% of workers voted for the Labour Party, while 73% of the labor voters were workers. Peter Mair has provided the row data. Italy: For 1953, Lipset (1966: 415) reported that roughly 69–70% of Italian workers voted for the left. A Doxa survey of 1964 reports a ﬁgure of 68%. Dogan (1960) suggests the ﬁgure of 65%. Later election surveys give lower estimates, between 55% and 60%, as for instance in Barnes and Pierce (1971), Barnes (1974) and (1977: 59), and Sani (1977: 111). A prudent estimate for the 1950s–1960s should therefore be set at around 60%. The percentage of left voters who are workers is set at about 50–55% by Dogan for 1953 (1960), at 40% by Barnes (1977), and at about 33–35% by Bruschi and Pacini (1978). The ﬁgures in this case diverge widely. I have decided to take an intermediate value of 45%. Netherlands: The survey material reviewed by Andeweg (1982: 93) indicates that in 1966 about a third of the Dutch workers voted for left parties, and this proportion was roughly conﬁrmed in studies in 1971, 1972, and 1977. Dogan (1960), however, reports the ﬁgure of 50%. I have used the intermediate value of 41%. Dogan’s estimate for the percentage of left voters being workers in the 1950s is 53–54%. The Dutch Parliamentary Election Study of 1971 suggests the ﬁgure of 67%. The two surveys of the 1980s (quoted in Daalder [1987b: 226–227]) indicate ﬁgures of 53% and 52%, respectively. I have opted for a ﬁgure of 55%. Norway: The 1957 survey analyzed by Rokkan (1967: 427–431) indicated that 79% of Norwegian workers voted Labor, giving, however, no information on the Communist vote. The 1968 survey discussed by

DATA APPENDIX

593

Valen and Rokkan (1974: 334) puts at 76% the workers’ vote for the three left parties. Dogan (1960) indicates that 70% of the left parties’ support came from working-class voters in the mid-1950s. • Sweden: Dogan (1960) sets at 74% the percentage of workers voting left; Stjerquist (1966: 127) indicates the ﬁgure of 68% on the basis of the 1964 survey The 1968 election survey, analyzed in Sa¨rlvik (1974: 401–402), reports the ﬁgure of 77%. Berglund and Lindstro¨m (1978: 108) argue that in the mid-1980s 84% of workers voted left (78% Social Democrats and 6% Communists). I have used the value of 77%. The same three sources offer the following estimation of the percentage of workers within the left electorate: 73%, 71%, and 66%. I have used 71%. • Switzerland: Kerr (1974: 11, 14) indicates that 40% of Swiss workers voted left in 1972 and about 33% of the left vote was cast by workers. • United Kingdom: The large set of British survey studies is difﬁcult to compare due to the different categories used to identify the working class. For the percentage of workers voting left, there are the following estimates: 1955: 75%; 1959: 71% (Goldthorpe, Lockwood, Bechhfer, and Platt [1971: 12]), based on a sample of industrial workers; 1958: 62% (Abrams [1958]); 1963: 72–73% (Butler and Stokes [1970: 76]), based on subjective class assignment; in the period 1945–1965 six surveys have values ranging between 57% and 64%, based on the manual–nonmanual distinction (Heath, Jowell, and Curtice [1985: 30%]); in the period 1964–1966 two survey have values ranging from 70% to 72%, based on a strictly deﬁned working class (Heath, Jowell, and Curtice [1985: 32–33]). I have decided to take a value of 72– 73%. As far as the percentage of Labor’s vote cast by workers, two sources agree on an estimate of 66–67% for the early 1950s (Bonham [1953: 130] and Abrams and Rose [1961: 18]). Both ﬁgures are based on self-reported class.

LIST OF VARIABLES IN THE ELECTION DATA FILE NUMBER OF CASES (ELECTIONS): 361 COUNTRY ELECTION YEAR DECADE PERIOD:

594

DATA APPENDIX

1 2 3 4

before 1918 1918–1944 1945–1965 1966–1985

ELECTORAL DATA ELECTORA TURNOUT TURNOUTE SDVOTE CPVOTE OLVOTE TLVOTE TLSEATS SDSEATS CPSEATS OLSEATS

Electorate as % population ⬎ 20 years Total votes as % population ⱖ 20 years Total votes as % of electorate % valid votes for Social Democrats % valid votes for Communist Party % valid votes for other left parties % valid votes for whole left % of seats for whole left % of seats for Social Democrats % of seats for Communist Party % of seats for other left parties

SOCIOECONOMIC DATA AGRI INDU SERVI INDI INDIWFW DEPE DEPAAP INAGAP INAGAA INAGIN WOAGAP WOAGAA WOAGWO WO1AP WO2AP WO3AP

Labor force in agriculture as % of active population Labor force in industry as % of active population Labor force in service as % of active population Independent, self-employed, and family workers Independents without family workers Dependent labor force Dependents in public administration as % of active population Independents in agriculture as % of active population Independents in agriculture as % of active in agriculture Independents in agriculture as % of independents Workers in agriculture as % of active population Workers in agriculture as % of active in agriculture Workers in agriculture as % of workers Workers as % of active population Workers in mining, manufacturing, construction, utilities, transportation as % of active population Workers in mining, manufacturing as % of active population

DATA APPENDIX

WO2WO WO3WO EMRATE RELIGION LANGUAGE LITERACY ILLITERA U2 U5 U20 U100

595

Workers in mining, manufacturing, construction, utilities, transportation as % of workers Workers in mining manufacturing as % of workers Employers-self-employed as % of active in manufacturing Religious F ⫻ 100 Linguistic F ⫻ 100 Pupils in primary schools as % of 5–14 age group % unable to write and read Urbanization: % in localities ⬍ 2,000 inhabitants Urbanization: % in localities ⬍ 5,000 inhabitants Urbanization: % in localities ⬍ 20,000 inhabitants Urbanization: % in localities ⬍ 100,000 inhabitants

GOVERNMENTAL DATA MAFTLAST MBEFNEXT LEFT A LEFT B LEFT C COMPAR A COMPAR B COMPAR C SOCIAL A SOCIAL B SOCIAL C GOVPOW

No. of months elapsed since previous elections No. of months before next election Coalition potential index of the left: blind version Coalition potential index of the left: spatial version Coalition potential index of the left: spatial version with jump Coalition potential index of the Communist Party: blind version Coalition potential index of the Communist Party: spatial version Coalition potential index of the Communist Party: spatial version with jump Coalition potential index of the Socialist Party: blind version Coalition potential index of the Socialist Party: spatial version Coalition potential index of the Socialist Party: spatial version with jump Governmental power of Socialist Party in the legislature. Based on an index comprising: (1) socialist presence in absence form cabinet; (2) parliamentary status of cabinet; (3) socialist status in cabinet; (4) duration in months. This index results from an elaboration done in a different ﬁle, taking as a unit all cabinets from 1918 to 1966.

596

DATA APPENDIX

PARTY SYSTEM DATA FRAC FRACLEFT TNP NUMPART MAJORNP TLDOMIN

Index of fractionalization Index of left block fractionalization Total number of parties Number of parties for volatility calculations Number of parties over 2% Total left dominance: distance from strongest or second-strongest party SDDOMIN SD dominance: distance from strongest or secondstrongest party STRNSOC % of seats of strongest nonsocialist party TPSNS Type of strongest nonsocialist party NNEWP No. of parties founded after 1959 VOTENEWP Vote % of nnewp SEATNEWP Seat % of nnewp NP2PC No. of parties ⱕ 2% of votes VOTEP2PC Vote % of np2% SEATP2PC Seat % of np2% NP5PC No. of parties ⱕ 5% of votes VOTEP5PC Vote % of np5% SEATP5PC Seat % of np5% NP10PC No. of parties ⱕ 10% VOTEP10P Vote % of np10% SEATP10P Seat % of np10% BPVOTE Vote % of biggest party BPSEAT Seat % of biggest (seat) party IIBPVOTE Vote % of second-biggest party IIBPSEAT Seat % of second-biggest (seat) party TV Total volatility by election BV Left block volatility WBV Within-bloc volatility PBV BV as % of TV PRINROSE Index of proportionality: Rose PRIN2 Index of proportionality: %seat ⫺ %vote: 2 PRINLIJ Index of proportionality: Lijphart RELITYPE Type of religion 1.00 Heterogeneous 2.00 Homogeneous Catholic 3.00 Homogeneous Protestant RELIMOB Type of religious mobilization

DATA APPENDIX

1.00 2.00 3.00 4.00

597

Before socialists in Catholic homogeneous countries (Be, Ir) Early before or parallel to socialists in heterogeneous countries (Ge, Ne, Sz) Late after socialists in homogeneous Catholic countries (Au, It, Fr) Homogeneous Protestant countries (GB, Scan)

PARTY ORGANIZATION DATA SDMEMELE SDMEMVOT CPMEMELE CPMEMVOT PDELE PDVOT TUD LTUD LTUMELE LTUMVOT BARG 1.00 late 2.00 medium 3.00 early TUSEQUEN 1 party ﬁrst 2 parallel 3.00 tu ﬁrst PASEQUEN 1.00 tu ﬁrst 2.00 parallel 3.00 party ﬁrst TUTIMING 1.00 tu late 2.00 tu mid 3.00 tu early PATIMING 1.00 party late 2.00 party mid

Social Democratic Party membership as % of electorate Social Democratic Party membership as % of Social Democratic Party voters Communist Party membership as % of electorate Communist Party membership as % of Communist Party voters Sdmemele ⫹ cpmemele Sdmemvot ⫹ cpmemvot Trade union as % of dependent labor force Left trade union as % of dependent labor force Left trade union members as % of electorate Left trade union members as % of left votes Bargaining timing

Party–trade union consolidation sequence

Trade union–party consolidation sequence

Comparative trade union consolidation timing

Comparative party consolidation timing

598

DATA APPENDIX

3.00 party early TULENGHT Length of trade union consolidation 1.00 long 2.00 mid 3.00 short PALENGHT Length of party consolidation 1.00 long 2.00 mid 3.00 short LINKAGE Party–trade union linkages 1.00 contingent 3.00 interlocking MODE Party and trade union mode of representation 1.00 fragmented 3.00 corporate INSTITUTIONAL VARIABLES ELECTSYS Type of electoral system 1 proportional 2 majoritarian 3 mixed ELSYSCHA Year of Electoral system change FRANPATT Franchise timing/tempo pattern 1.00 early-intermediate/gradual (Ir, Ne, Uk) 2.00 early/sudden (Dk, Fr, Ge, Sz) 3.00 intermediate-sudden/interm. (Au, Be, No) 4.00 late/sudden (Fi, It, Sw) STATENESS Z-score sum of tax ratio, tax centralization, central administration personnel, military personnel, police, and judiciary personnel REPRESS Level of repression 01 to 13: rank ordering of countries REPOBSTA Representation obstacles 01 to 13: rank ordering of countries PARLA Parliamentarization (responsible government) 01 to 13: rank ordering of countries COMBINED INDICES FOR THE GENERAL MODEL INDURB

Pattern of industrialization and urbanization

DATA APPENDIX

599

1.00 uncoupled growth (all other cases) 2.00 medium growth in both (Au, Dk, Fr) 3.00 accelerated growth in both (It, Fi, Sw) MOBSOC Combined index (ZSindiwfw ⫹ ZSemrate ⫹ ZSwo2wo ⫹ ZSwo2ap ⫹ ZSindurb) CULTETHE Country cultural heterogeneity (ZSlanguage ⫹ ZSreligion ⫹ ZSillitera) ININ Institutional integration. Sum of 13 countries rank ordered on three variables (stateness, repression, representation obstacles) ELEFRAPA Electorate level and pattern of enfranchisement (ZSelectorat ⫹ ZSfranpat) SEQ Sequence liberalization, universal suffrage, Socialist Party foundation 1.00 RG-SDF-US 2.00 RG-US-SDF 3.00 US-RG-SDF 4.00 US-SDF-RG 5.00 SDF-US-RG ORGDEN Organizational density ltumele ⫹ pdele PAORCON Party organizational consolidation: paseq ⫹ patiming ⫹ palength (from 3.00 ⫽ low, low, low to 9.00 ⫽ high, high, high) PATURELA Party–trade union relationship ⫽ mode ⫹ linkage 2.00 Contingent–noncorporate 4.00 Interlocking-noncorporate/noninterlocking-corporate 6.00 interlocking and corporate ORGVAR ZSorgden ⫹ ZSpaorcon ⫹ ZSpaturela

REFERENCES

Autori, Vari [various] (AA. VV.), 1955. Compendio dei risultati delle consultazioni popolari dal 1840 al 1954. Rome: Istituto Poligraﬁco dello Stato. Abendroth, W., 1965. Sozialgeschichte der europa¨ischen Arbeiterbewegung. Frankfurt am Main: Suhrkamp Verlag. Abrams, M., 1958. Class Distinctions in Britain. In The Future of the Welfare State. London: Conservative Political Center. Abrams, M., and R. Rose, 1961. Must Labour Lose? London: Penguin Books. Abramson, P. R., 1971. Social Class and Political Change in Western Europe: A Cross-National Longitudinal Analysis. Comparative Political Studies 4: 131–155. Adam, G., F. Bon, J. Capdeville, and R. Mouriaut, 1971. L’ouvrier franc¸ais en 1970. Paris: Colin. Agnelli, A., 1969. Questione nazionale e socialismo. Contributo allo studio del pensiero di K. Renner e O. Bauer. Bologna: Il Mulino. Alapuro, R., 1981. Origins of Agrarian Socialism in Finland. In Torsvik (1981: 274– 283). 1988. State and Revolution in Finland. Berkeley: University of California Press. Alapuro, R., and E. Allardt, 1978. The Lapua Movement. The Threat of Rightist Take-over in Finland 1930–32. In Linz and Stepan (1978: 122–141). Alber, J., 1982. Vom Armenhaus zum Wohlfahrtsstaat. Analysen zur Entwicklung der Sozialversicherung in Westeuropa. Frankfurt: Campus Verlag. Alford, R., 1963. Party and Society: The Anglo-American Democracies. Chicago: Rand McNally. Alford, R., and R. Friedland, 1974. Nations, Parties and Participation: A Critique of Political Sociology. Theory and Society 1: 307–328. Allardt, E., 1964. Patterns of Class Conflict and Working Class Consciousness in Finnish Politics. In Allardt and Littunen (1964: 97–131). 1968. Past and Emerging Political Cleavages. In Stammer (1968: 66–74). 1970. Types of Protest and Alienation. In Allardt and Rokkan (1970: 45–63). 601

602

REFERENCES

Allardt, E., and Y. Littunen, eds., 1964. Cleavages, Ideologies and Party Systems. Helsinki: Academic Bookstore. Allardt, E., and P. Pesonen, 1960. Citizens’ Participation in Political Life: Finland. International Social Science Journal 12: 27–39. 1967. Cleavages in Finnish Politics. In Lipset and Rokkan (1967: 325–366). Allardt, E., and S. Rokkan, eds., 1970. Mass Politics. Studies in Political Sociology. New York: Free Press. Almond, G., 1957. The Appeals of Communism. Princeton: Princeton University Press. Anckar, D., 1992. Finland: Dualism and Consensus Rule. In Damgaard (1992: 152– 191). Anderson, E. N., and P. R. Anderson, 1967. Political Institutions and Social Change in Continetal Europe in the Nineteenth Century. Berkeley: University of California Press. Anderson, P., 1974. Lineage of the Absolutist State. Thetford (Norfolk, England): Thetford Press. Andeweg, R. B., 1982. Dutch Voters Adrift: on Explanations of Electoral Change 1963–1977. Leyden: Leyden University. Andler, C., 1932. Vie de Lucien Herr (1864–1926). Paris: Rieder. Andreasi, A., 1981. L’anarco-sindacalismo in Francia, Italia e Spagna. Milan: La Pietra. Angress, W., 1972. Stillborn Revolution: The Communist Bid for Power in Germany, 1921–1923. Port Washington, NY: Kennikat Press. Arfe´, G., 1965. Storia del socialismo italiano (1982–1926). Turin: Einaudi. Aron, R., 1964. La lutte des classes. Nouvelles lec¸ons sur les socie´te´s industrielles. Paris: Gallimard. 1971. Le socialisme franc¸ais face au marxisme. Paris: Grasset. 1989. Lec¸ons sur l’histoire. Paris: Editions de Fallois. Atkinson, M. M., and W. G. Coleman, 1989. Strong States and Weak States: Sectoral Policy Networks in Advanced Capitalist Economies. British Journal of Political Science 19: 47–67. Bain, G. S., and R. Price, 1976. Profiles of Union Growth. A Comparative Statistical Portrait of Eight Countries. Oxford: Blackwell. Bairoch, P., 1982. International Industrialisation Levels from 1750 to 1980. The Journal of European Economic History 11: 269–333. Baldwin, P., 1988. How Socialist Is Solidaristic Social Policy? Swedish Post-War Reform as a Case in Point. International Review of Social History 33: 121– 144. Ball, A. R., 1987. British Political Parties: The Emergence of a Modern Party System. London: Macmillan. Ballini, P.-L., 1985. Le elezioni politiche nel Regno di Italia. Appunti di bibliograﬁa, legislazione e statistiche. Quaderni dell’osservatorio elettorale n. 15: 143–220. 1988. Le elezioni nella storia d’Italia dall’Unita` al Fascismo. Bologna: Il Mulino. Balthasar, A., E. Gruner, and H. Hirtner, 1988. Gewerkschaften und Arbeitgeber auf dem Arbeitsmarkt: Streiks, Kampf ums Recht und Verha¨ltnis zu anderen Interessengruppen, Band 1 of Arbeiterschaft und Wirtschaft in der Schweiz 1880– 1914. Bern: Chronos.

REFERENCES

603

Banﬁeld, E. C., 1958. The Moral Basis of a Backward Society. Glencoe, IL: Free Press. Barnes, S. H., 1967. Party Democracy: Politics in a Socialist Federation. New Haven: Yale University Press. 1974. Italy: Religion and Class in Electoral Behaviour. In Rose (1974: 171–225). 1977. Representation in Italy: Institutionalised Traditions and Electoral Choice. Chicago: University of Chicago Press. Barnes, S. H., and M. Kaase, eds., 1979. Political Action. Beverly Hills, CA: Sage. Barnes, S. H., and R. Pierce, 1971. Public Opinion and Political Preferences in France and Italy. Midwest Journal of Political Science 15: 643–660. Bartolini, S., 1983. The Membership of the Mass Party: The Social Democratic Experience 1889–1978. In Daalder and Mair (1983: 177–220). 1993. On Time and Comparative Research. Journal of Theoretical Politics 5: 131–167. 1998. Coalition Potential and Governmental Power. In Lane and Pennings (1998: 40–61). Bartolini, S., and P. Mair, 1990. Identity, Competition and Electoral Availability. The Stabilisation of European Electorates. Cambridge: Cambridge University Press. Bauman, Z., 1972. Between Class and Elites. Manchester: Manchester University Press. Bendix, R., 1964. Nation-Building and Citizenship: Studies of Our Changing Social Order. Berkeley: University of California Press. 1974. Work and Authority in Industry. Berkeley: University of California Press. 1984a. Citizenship of the Lower Classes. In Bendix (1984b: 91–107). 1984b. Force, Fate, and Freedom: On Historical Sociology. Berkeley: University of California Press. Bendix, R., and S. M. Lipset, eds., 1966. Class, Status, and Power: Social Stratification in Contemporary Perspective. New York: Free Press. Benjamin, R. W., and J. H. Kautsky, 1968. Communism and Economic Development. American Political Science Review 62: 110–123. Berelson, B., P. F. Lazarsfeld, and W. N. MacPhee, 1954. Voting. Chicago: Chicago University Press. Berger, S., 1994. The British Labour Party and the German Social Democrats 1900– 1931. Oxford: Clarendon Press. Berglund, S., and U. Lindstro¨m, 1978. The Scandinavian Party System(s). Lund: Studentlitteratur. Bergounioux, A., and B. Manin, 1979. La socialdemocratie ou le compromis. Paris: Presses Universitaires de France. Bernstein, E., 1899. Die Voraussetzungen des Sozialismus und die Aufgabe der Sozialdemokratie. Stuttgart: J. H. W. Dietz. Berresford Ellis, P., 1972. A History of the Irish Working Class. London: Victor Gollancz Ltd. Berry, D., 1969. Party Membership and Social Participation. Political Studies 17: 196–207. Bertold, O., E. Christiansen, and P. Hansen, 1954–1955. En bygning vi reyser: Den politiske arbejderbevaegelses historie i Danmark. Socialdemokratiet arsberentninger 1924–1943. 2 vols. Copenhagen: Fremad.

604

REFERENCES

Beyme, K. von, 1973. Die parlamentarischen Regierungsysteme in Europa. Munich: Piper. 1985. Political Parties in Western Democracies. Aldershot, England: Gower. Bieder, H. J., 1981. Gewerkschaften in Krieg und Revolution. Arbeiterbewegung, Industrie, Staat und Milita¨r in Deutschland 1914–1920, 2 vols. Hamburg: Hans Christians Verlag. Birnbaum, P., 1988. State and Collective Action: The European Experience. Cambridge: Cambridge University Press. Birnbaum, P., and B. Badie, 1982. Sociologie de l’Etat. Paris: Grasset. Biucchi, B. M., 1976–1977. Switzerland, 1700–1914. In Cipolla (1976–1977: 627–655). Blackbourn, D., and G. Eley, 1984. The Peculiarities of German History. Oxford: Oxford University Press. Blackmer, D., and S. Tarrow, eds., 1975. Communism in Italy and France. Princeton: Princeton University Press. Blainpain, R., ed., 1987. International Encyclopedia for Labour Law and Industrial Relations. Deventer, The Netherlands: Kluwer. Blumer, H., 1960. Early Industrialism and the Labouring Class. The Sociological Quarterly 1: 5–14. Bobbio, N., 1994. Destra e sinistra. Ragioni e significati di una distinzione politica. Rome: Donzelli. Bomsdor, E., 1981. The Distribution of Power in Specific Decision-Making Bodies. In Holler (1981: 283–288). Bonham, J., 1953. The Middle Class Vote. London: Faber and Faber. Borre, O., 1980. The Social Bases of Danish Electoral Behaviour. In Rose (1980: 241– 282). Bosio, G., 1951. La diffusione degli scritti di Marx e di Engels in Italia dal 1871 al 1892. Societa` 7: 268–284, 444–477. Boudon, R., 1979. De´terministes sociaux et liberte´ individuelle. In Effets perverses et ordre social. Paris: Presses Universitaires de France, 187–252. Bourquelot, F., 1972. Les salarie´s agricoles et leurs organisations. In Tavernier, Gervais, and Servolin (1972: 533–556). Bowditch, J., 1951. The Concept of Elan Vital: A Rationalisation of Weakness. In Earle (1951: 32–43). Braunthal, J., 1945ff. Yearbook of the International Socialist Labour Movement. Amsterdam: Socialist International. Breitmen, R., 1981. German Socialism and Weimar Democracy. Chapel Hill: University of North Carolina Press. Breuilly, J., 1985. Liberalism or Social Democracy. A Comparison of British and German Labour Politics c. 1850–1875. European History Quarterly 15: 3– 42. 1992. Labour and Liberalism in Nineteenth Century Europe: Essays in Comparative History. Manchester: Manchester University Press. Brezzi, C., 1979. I partiti democratici cristiani in Europa. Milan: Teti. Brown, B., 1969. The French Experience of Modernization. World Politics 21: 366– 391. Brunn, N., 1991. Law and Society in the Global Village: Toward Collaboration and Comparative Research. Helsinki: Swedish School of Economics and Business Administration.

REFERENCES

605

Bruschi, A., and M. Pacini, 1978. Indagine sulle cultura politiche in Italia. Turin: Fondazione Agnelli. Bu¨cher, K., ed., 1903. Die Großstadt. Dresden: von Zahn und Jaensch. Buell, R. L., 1920. Contemporary French Politics. New York: Appleton. Buiting, H., 1990. The Netherlands. In Linden and Rojahn (1990: 57–84). Bull, E., 1922. Die Entwicklung der Arbeiterbewegung in den drei skandinavischen La¨ndern 1914–1920. Archiv fu¨r die Geschichte des Sozialismus und der Arbeiterbewegung 10: 329–361. Bu¨sch, O., ed., 1980. Wa¨hlerbewegung in der Europa¨ischen Geschichte. Berlin: Colloquium Verlag. Butler, D., 1990. Electoral Reform and Political Strategy in Britain. In Noiret (1990b: 451–463). Butler, D., and J. Cornford, 1969. United Kingdom. In Meyriat and Rokkan (1969: 330–351). Butler, D., and D. Stokes, 1970. Political Change in Britain. London: Macmillan. Callesen, G., 1990. Denmark. In Linden and Rojahn (1990: 131–160). Campbell, P., 1958. French Electoral Systems and Elections 1789–1957. London: Faber. Cannarsa, S., 1950. Il socialismo ed i XXV congressi nazionali del Partito Socialista Italiano. Florence: Societa` Editrice Avanti!. Caracciolo, A., 1957. Stato e societa` civile. Problemi della unificazione italiana. Turin: Einaudi. Carlson, B., 1969. Trade Unions in Sweden. Stockholm: Tidens Fo˜rlag. Carocci, G., 1975. Storia d’Italia dall’unita` ad oggi. Milan: Feltrinelli. Cartiglia, C., 1976. Rinaldo Rigola e il sindacalismo riformista in Italia. Milan: Feltrinelli. Casertano, A., 1911. Il diritto di voto. Naples: Pierro. Castairs, A., 1980. A Short History of Electoral Systems in Western Europe. London: Allen & Unwin. Castles, F. G., 1977. Scandinavian Social Democracy: Achievements and Origins. Berlin: ECPR. 1978. The Social Democratic Image of Society: A Study of the Achievements and Origins of Scandinavian Social Democracy in Comparative Perspective. London: Routledge & Kegan Paul. ed., 1982. The Impact of Parties: Politics and Policies in Democratic Capitalist States. London: Sage. Cazzola, F., 1970. Il partito come organizzazione; Studio di un caso – il PSI. Rome: Edizioni del Tritone. Cella, G. P., and T. Treu, eds. 1989. Relazioni industriali: Manuale per l’analisi dell’esperienza italiana. Bologna: Il Mulino. Cervelli, I., 1978. La Germania dell’800. Un caso di modernizzazione conservatrice. Rome: Editori Riuniti. Charlot, J., ed., 1973. Quand la gauche peut gagner . . . Les e´lections le´gislatives des 4– 11 mars 1973. Paris: Editions Alain Moreau. Chehabi, H. E., and A. Stepan, eds., 1995. Politics, Society, and Democracy: Comparative Studies. Boulder, CO: Westview Press. Chubb, B., 1974. The Government and Politics of Ireland. London: Oxford University Press. Cipolla, C. M., 1969. Literacy and Development in the West. London: Penguin Books.

606

REFERENCES

ed., 1976–1977. The Emergence of Industrial Society. The Fontana Economic History of Europe, Vol. 4. Hassocks, Sussex, England: Harvester Press/New York: Barnes & Noble. Clapham, J. H., 1936. Economic Development of France and Germany 1815–1915. London: Cambridge University Press. Clark, M., 1977. Antonio Gramsci and the Revolution That Failed. New Haven: Yale University Press. Clarke, P., and J. Q. Wilson, 1961. Incentive Systems: A Theory of Organizations. Administrative Science Quarterly 6: 126–166. Clarke, P. F., 1971. Lancashire and the New Liberalism. Cambridge: Cambridge University Press. Cofrancesco, D., 1990. Sinistra. In Grande Dizionario Enciclopedico, 4th ed., Vol. 18: 883–884. Turin: UTET. Cole, G. H. D., 1954. Socialist Thought: Marxism and Anarchism 1850–1890. London: Macmillan. Coleman, J. C., 1965a. Education and Political Development. Princeton: Princeton University Press. 1965b. Education and Political Development. In Coleman (1965a: 3–32). Colletti, L., 1972. Bernstein e il marxismo della Seconda Internazionale. In Ideologia e societa`. Bari: Laterza, 61–147. Committee on Financing of Political Parties, 1976. Report. London: Her Majesty’s Stationery Ofﬁce. Commons, J., 1962. Labor Movements. In Encyclopaedia of the Social Sciences. New York: Macmillan, Vol. 8: 681–685. Congre`s de Tours (de´cembre 1920). Naissance du parti communiste. Edition critique des principaux de´bats, 1964. Paris: Rene´ Juillard. Converse, P. E., and E. Pierce, 1986. Political Representation in France. Cambridge, MA: Belknap Press of Harvard University Press. Converse, P. E., and H. Valen, 1971. Dimensions of Cleavage and Perceived Party Distances in Norwegian Voting. Scandinavian Political Studies 6: 107– 152. Conway, M. M., and F. B. Fiegert, 1968. Motivations, Incentive Systems and the Political Party Organization. American Political Science Review 62: 667– 676. Cornford, J., 1964. The Adoption of Mass Organization by the British Conservative Party. In Allardt and Littunen (1964: 400–424). Correspondance Friedrich Engels et Paul et Laura Lafargue, 1868–1895, 1956. Paris: Editions Sociales. Cortesi, L., 1973. Le origini del PCI. Il PSI dalla guerra di Libia alla scissione di Livorno. Bari: Laterza. Crewe, I., and A. King, 1995. The Birth, Life and Death of the Social Democratic Party. Oxford: Oxford University Press. Criddle, B., 1977. The French Parti Socialiste. In Paterson and Thomas (1977: 25– 66). Croce, B., 1938. Come nacque e come morı` il marxismo teorico in Italia. Rome: Laterza. Cronin, J. E., 1980. Labor Insurgency and Class Formation: Comparative Perspectives on the Crisis 1917–1920 in Europe. Social Science History 4: 125– 152.

REFERENCES

607

Crouch, C., 1993. Industrial Relations and European State Traditions. Oxford: Oxford University Press. Daalder, H., 1966a. Parties, Elites, and Political Developments in Western Europe. In La Palombara and Weiner (1966: 43–77). 1966b. The Netherlands: Opposition in a Segmented Society. In Dahl (1966a: 188– 236). 1967. Cabinet–Parliament Relations and Party Systems, unpublished manuscript. 1973. Building Consociational Nations. In Eisenstadt and Rokkan (1973: 4– 31). 1981. Consociationalism, Center, and Periphery in the Netherlands. In Torsvik (1981: 181–240). ed., 1987a. Party Systems in Denmark, Austria, Switzerland, the Netherlands and Belgium. London: Frances Pinter. 1987b. The Dutch Party System: From Segmentation to Polarization – and Then. In Daalder (1987a: 193–284). 1990. Ancient and Modern Pluralism in the Netherlands: Harvard University: Center for European Studies, Working Paper Series No. 22. 1995. Paths Towards State Formation in Europe: Democratization, Bureacratization and Politicization. In Chehabi and Stepan (1995: 113–130). Daalder, H., and G. A. Irwin, eds., 1989. Politics in the Netherlands. How Much Change? London: Frank Cass. Daalder, H., and P. Mair, eds., 1983. Western European Party Systems: Continuity and Change. London: Sage. Daalder, H., and C. J. M. Schuyt, eds., 1986. Compendium voor politiek en samenleving in Nederland. Brussels: Samsom Uitgeverij bv Alphen aan den Rijn. Dahl, R. A., ed., 1966a. Political Opposition in Western Democracies. New Haven: Yale University Press. 1966b. Some Explanations. In Dahl (1966a: 448–486). 1971. Polyarchy, Participation and Opposition. New Haven: Yale University Press. Dalton, R. J., P. A. Beck, and S. C. Flanagan, 1984. Electoral Change in Advanced Industrial Democracies. In Dalton, Flanagan, and Beck (1984: 3–22). Dalton, R. J., S. C. Flanagan, and P. A. Beck, eds., 1984. Electoral Change in Industrial Democracies: Realignment or Dealignment? Princeton: Princeton University Press. Damgaard, E., ed., 1992. Parliamentary Change in the Nordic Countries. Oslo: Scandinavian University Press. De Bakker, B., and M. Claeys-Van Haegendoren, 1973. The Socialist Party in the Party System and in Organized Socialism in Belgium. Res Publica 15: 237–247. De Fraja, G., 1988. Il potere dei partiti italiani. Politeia 3: 11–17. De Kiriaki, A. S., 1885. Della riforma elettorale. Saggio di diritto costituzionale e legislazione comparata. Rome: Forzano. De Sanctis, F., 1876. Viaggio elettorale. Un racconto. Naples: Morano. Delruelle, N., R. Evalenko, and W. Fraeys, 1970. Le comportement politique des e´lecteurs belges. Brussels: Institut de Sociologie de l’ Universite´ Libre de Bruxelles. Delsinne, L., 1936. Le mouvement syndical en Belgique. Brussels: Castaigne.

608

REFERENCES

Dent, M., 1993. The Case for an Electoral Pact between Labour and the Liberal Democrats. Political Quarterly 64: 243–251. Derry, T. K., 1968. A Short History of Norway, 2nd ed., London: George Allen & Unwin. Detti, T., 1972. Serrati e la formazione del PCI. Storia della frazione terzinternazionalista 1921–1924. Rome: Editori Riuniti. Deutsch, K., 1961. Social Mobilization and Political Development. American Political Science Review 55: 493–502. Dewachter, W., 1976. Politieke Sociologie. Leuven: Leuven University Press. 1987. Changes in the Belgian Party System Since World War Two. In Daalder (1987a: 311–330). Diani, M., and R. Eyerman, eds., 1992. Studying Collective Action. London: Sage. Dogan, M., 1960. Le vote ouvrier en Europe Occidentale. Revue Franc¸aise de Sociologie 1: 25–44. 1967. Political Cleavages and Social Stratification in France and Italy. In Lipset and Rokkan (1967: 129–196). Dommanget, M., 1969. L’introduction du marxisme en France. Lausanne: Editions Rencontre. Dovring, F., 1964. The Share of Agriculture in a Growing Population. In Eicher and Witt (1964: 78–98). Droz, J., 1972. Socialisme et syndicalisme de 1914 a` 1939. Paris: Centre de Documentation Universitaire. 1973. Einfluss der deutschen Sozialdemokratie auf den franzo¨sischen Sozialismus (1871–1914). Opladen: Westdeutscher Verlag. Drucker, H., ed., 1979. Multi-Party Britain. London: Macmillan. Dubofsky, M., 1966. The Origins of Western Working Class Radicalism, 1850– 1905. Labor History 7: 131–154. Dupeux, G., 1964. La socie´te´ franc¸aise 1789–1960. Paris: Colin. Durkheim, E., 1958. The Suicide. Glencoe, IL: Free Press. (1930. Le suicide. Paris: Presses Universitaires de France). Duveau, G., 1946. La vie ouvrie`re en France sous le Second Empire. Paris: Gallimard. Duverger, M., 1967 (1st ed., 1956). Les partis politiques. Paris: Colin. 1986. Les re´gimes semi-pre´sidentiels. Paris: Presses Universitaires de France. Earle, E. M., ed., 1951. Modern France: Problems of the Third and Fourth Republics. Oxford: Oxford University Press. Easton, D., 1990. The Analysis of Political Structures. London: Routledge. Ebbinghaus, B., 1992. The Transformation of Cleavage Structures in Western European Trade Union Systems: Can We Draw Union Diversity in Rokkan’s Conceptual Map? Limerick: ECPR. Eckstein, H., 1966. Division and Cohesion in Democracy: A Study of Norway. Princeton.: Princeton University Press. Ehrmann, H. W., 1952. French Peasants and Communism. American Political Science Review 46: 19–43. Eicher, C., and L. Witt, eds., 1964. Agriculture in Economic Development. New York: McGraw-Hill. Eisenstadt, S. N., 1966. Modernization, Protest and Change. Englewood Cliffs, NJ: Prentice-Hall.

REFERENCES

609

Eisenstadt, S. N., and S. Rokkan, eds., 1973. Building States and Nations: Models and Data Resources. London: Sage. Elklit, J., 1980. Election Laws and Electoral Behaviour in Denmark until 1920. In Bu¨sch (1980: 366–396). Elvander, N., 1977. Scandinavian Social Democracy: Present Trends and Future Prospects. Berlin: ECPR. Engelmann, F. C., 1966. Austria: The Pooling of Opposition. In Dahl (1966a: 260– 283). Epstein, L. D., 1967. Political Parties in Western Democracies. New York: Praeger. Esping Andersen, G., 1990. The Three Worlds of Welfare Capitalism. New York: Polity Press. Evans, P. B., D. Rueschemeyer, and T. Skocpol, eds., 1985. Bringing the State Back In. Cambridge: Cambridge University Press. Fabbri, M., 1991. Governo e sovrano nell’Italia liberale. Rivista Trimestrale di Scienza dell’Amministrazione no. 3: 103–186. Farrell, B., 1970. Labour and the Irish Political Party System: A Suggested Approach to Analysis. The Economic and Social Review 1: 477–502. 1971. The Founding of Da´il Eireann. Dublin: Gill & Macmillan. Ferner, A., and R. Hyman, 1992. Italy: Between Political Exchange and MicroCorporatism. In Hyman and Ferner (1992: 524–560). Ferrera, M., 1993. Modelli di solidarieta`. Politica e riforme sociali nelle democrazie. Bologna: Il Mulino. Finer, S. E., 1970. Comparative Government. London: Allen and Lane. Fischer, H., ed., 1974. Das politische System O¨sterreichs. Vienna: Europa Verlag. Fleiner-Gerstern, T., 1987. Le conseil Fe´de´ral: Directoire de la Confe´de´ration. Pouvoirs no. 43: 49–63. Flora, P., 1973. Historical Processes of Social Mobilisation: Urbanisation and Literacy 1850–1965. In Eisenstadt and Rokkan (1973: 213–258). 1974. Modernisierungforschung. Zur empirischen Analyse der gesellschaftlichen Entwicklung. Opladen: Westdeutscher Verlag. Flora, P., and J. Alber, 1981. Modernization, Democratization, and the Development of Welfare States in Western Europe. In Flora and Heidenheimer (1981: 37– 80). Flora, P., J. Alber, R. Eichenberg, J. Kohl, F. Kraus, W. Pfenning, and K. Seebohm, 1987. State, Economy, and Society in Western Europe 1815–1875. A Data Handbook. Vol. 2: The Growth of Industrial Societies and Capitalist Economies. Frankfurt: Campus Verlag. Flora, P., R. Eichenberg, J. Kohl, F. Kraus, W. Pfenning, and K. Seebohm, 1983. State, Economy and Society in Western Europe 1815–1975: A Data Handbook. Vol. 1: The Growth of Mass Democracies and Welfare States. Frankfurt: Campus Verlag. Flora P., and A. J. Heidenheimer, eds., 1981. The Development of Welfare States in Europe and America. New Brunswick, NJ: Transaction Books. Fourastie´, J., 1963. Le grand espoir du XXe`me sie`cle. Paris: Gallimard. Foverskov, P., and L. N. Johansen, 1978. The Socialist and Communist Parties in Denmark in a Century: History, Programs and Policy. Unpublished paper. Frey, B. S., 1968. Eine spielttheoretische Analyse der Machtverteilung im schweiz-

610

REFERENCES

erischen Bundesrat. Schweizerische Zeitschrift fu¨r Volkswirtschaft und Statistik 104: 155–169. Fulcher, J., 1991. Labour Movements, Employers, and the State. Oxford: Clarendon. Gaborit, P., 1972. Le PCF et les paysans. In Tavernier, Gervais, and Servolin (1972: 197–222). Gagel, W., 1958. Die Wahlrechtsfrage in der Geschichte der deutschen liberalen Parteien, 1848–1918. Du¨sseldorf: Droste. Galante Garrone, A., 1973. I radicali in Italia 1849–1925. Milan: Garzanti. Galenson, W., 1949. Labour in Norway. Cambridge, MA: Harvard University Press. ed., 1952a. Comparative Labor Movements. New York: Prentice-Hall. 1952b. Scandinavia. In Galenson (1952a: 104–172). 1952c. The Danish System of Labor Relations. New York: Russell & Russell. Gallagher, M., 1982. The Irish Labour Party in Transition 1957–1982. Manchester: Manchester University Press. Galli, G., ed., 1968. Il comportamento elettorale in Italia: Bologna: Il Mulino. ed., 1976a. Esperienze riformiste in Europa. Il socialismo tra il 1918 e il 1934. Naples: Morano. 1976b. Storia del Partito Comunista Italiano. Milan: Il Formichiere. Garbani, P., and J. Schmid, 1980. Le syndicalisme suisse. Histoire politique de l’Union Syndicale 1880–1980. Lausanne: Editions d’en Bas. Gass, S. 1988. Les de´buts du Parti Socialiste Suisse, 1870–1890. In Cent Ans de PSS 1888–1988. Lausanne: Editions d’en Bas. Geary, D., 1981. European Labour Protest 1848–1939. New York: St. Martin’s Press. ed., 1989. Labour and Socialist Movements in Europe Before 1914. Oxford: Berg. Gehmacher, E., 1974. Faktoren des Wa¨hlerverhaltens. In Fischer (1974: 55–75). Gerlich, P., 1987. Consociationalism to Competition: The Austrian Party System since 1945. In Daalder (1987a: 61–106). Gerlich, P., and W. Mu¨ller, eds., 1983. Zwischen Koalition and Konkurrenz. O¨sterreichs Parteien seit 1945. Vienna: Braunmu¨ller. Gerschenkron, A., 1962. Backwardness in Historical Perspective. Cambridge, MA: Harvard University Press. Giddens, A., 1973. The Class Structure of Advanced Societies. London: Hutchinson. Gilissen, J., 1980. Die Entwicklung des Wahlrechts im Benelux-Raum, 1814 bis 1922. In Bu¨sch (1980: 338–365). Girod, R., 1964. Le syste`me des partis en Suisse. Revue Franc¸aise de Science Politique 14: 1114–1133. Goetschy, J., and P. Rozenblatt, 1992. France: The Industrial Relations System at a Turning Point? In Hyman and Ferner (1992: 404–444). Goguel, F., 1950. Les partis politiques en France. In Encyclope´die Politique de la France et du Monde. Paris: Editions de l’Encyclope´die Coloniale et Maritime: 271–359. 1969. Modernization e´conomique et comportement politique. Paris: Colin. 1970. Ge´ographie des e´lections franc¸aises sous la Troisie`me et la Quatrie`me Re´publique. Paris: Colin. Goldstein, R. L., 1983. Political Repression in the Nineteenth Century. London: Croom Helm.

REFERENCES

611

Goldthorpe, J. H., D. Lockwood, F. Bechhofer, and J. Platt, 1971. The Affluent Workers: Political Attitudes and Behaviour. Cambridge, U.K.: Cambridge University Press. Gratton, P., 1972. Le mouvement ouvrier et la question agraire de 1870 a` 1947. In Tavernier, Gervais, and Servolin (1972: 111–136). Graubard, S. R., ed., 1967. A New Europe? A Timely Appraisal. Boston: Beacon Press. Gray, R., 1981. The Aristocracy of Labour in Nineteenth Century Britain. London: Studies in Economic and Social History. Greene, E. N., ed., 1971. European Socialism since World War I. Chicago: Quadrangle Books. Greene, T. H., 1971. The Electorate of Non-Ruling Communist Parties. Studies in Comparative Communism 4: 68–103. 1973. Non-Ruling Communist Parties and Political Adaptation. Studies in Comparative Communism 6: 331–361. Grisalberti, C., 1978. Storia costituzionale d’Italia, 1848–1948. Bari: Laterza. Gruner, E., 1968. Die Arbeit in der Schweiz. Bern: Franche Verlag. Gruner, E., 1977. Die Parteien in der Schweitz. Bern: Franche Verlag. Guillermo, O., 1975. Evaluation of a Presidential Election Game. American Political Science Review 69: 947–953. Haapakoski, P., 1974. Breznevism in Finland. New Left Review no. 86: 29–49. Haerpfer, C., 1983. Nationalrats wahlen und Wa¨hlverhalten 1945–1980. In Gerlich and Mu¨ller (1983: 111–150). Haerpfer, C., and E. Gemacher, 1984. Social Structure and Voting in the Austrian Party System. Electoral Studies 2: 25–46. Halebsky, S., 1976. Mass Society and Political Conflict: Toward a Reconstruction of Theory. Cambridge: Cambridge University Press. Hale´vy, D., 1995 (1st ed., 1935). Visites aux paysans du centre: 1907–1934. Paris: Hachette. Hall, A., 1977. Scandals, Sensation and Social Democracy: The SPD Press in Wilhelmine Germany, 1890–1914. Cambridge: Cambridge University Press. Hamilton, R. F., 1967. Affluence and the French Worker in the Fourth Republic. Princeton: Princeton University Press. Hartenstein, W., and K. Liepelt, 1962. Party Members and Party Voters in West Germany. In Rokkan (1962: 43–52). Harvey, G.-J., 1989. Storia dell’alfabetizzazione occidentale, 3 vols. Bologna: Il Mulino. Hattam, V. C., 1992. Institutions and Political Change: Working-Class Formation in England and the United States. In Steinmo, Thelen, and Lomgstreth (1992: 153–187). Heath, A., R. Jowell, and J. Curtice, 1985. How Britain Votes. Oxford: Pergamon Press. Heidar, K., 1977. The Norwegian Labour Party: Social Democracy in a Periphery of Europe. In Paterson and Thomas (1977: 292–315). Henig S., and J. Pinder, eds., 1969. European Political Parties. London: Allen & Unwin/PEP. Hildebrandt, K., and R. J. Dalton, 1978. The New Politics: Political Change or Sunshine Politics? In Kaase and Von Beyme (1978: 69–96).

612

REFERENCES

Hill, K., 1974. Belgium: Political Change in a Segmented Society. In Rose (1974: 29– 108). Hintze, O., 1962. Soziologie und Geschichte Staat and Vefassung, edited by G. Oestreich. Go¨ttingen: Vandehhoeck and Ruprecht. 1980. Stato e societa`. Bologna: Zanichelli. Hirschmann, A., 1982. Shifting Involvements: Private Interest and Public Action. Oxford: Robertson. Hobsbawm, E. J., ed., 1964a. Labouring Men. London: Weidenfeld and Nicholson. 1964b. The Labour Aristocracy in Nineteenth-Century Britain. In Hobsbown (1964a: 272–315). Hodgson, H. J., 1967. Communism in Finland: A History and Interpretation. Princeton: Princeton University Press. 1973. Finnish Communists and the Opportunism of Conciliation. Studies in Comparative Communism 6: 397–404. Hoffman, S., 1974. The Rulers: Heroic Leadership in Modern France. In Decline or Renewal? France since the 1930s. New York: Viking Press: 63–110. Holler, M. J., ed., 1981. Power, Voting, and Voting Power. Wu¨rzburg: PhysicaVerlag. Holler, M. J., and J. Kellermann, 1977. Power in the European Parliament: What Will Change? Quality and Quantity 11: 189–192. Holton, R. J., 1985. Revolutionary Syndicalism and the British Labour Movement. In Mommsen and Husong (1985: 266–282). Horowitz, D. L., 1976. Storia del movimento sindacale in Italia. Bologna: Il Mulino. Houska, J. J., 1985. Influencing Mass Political Behavior: Elites and Political Subcultures in the Netherlands and Austria. Berkeley: Institute of International Studies, University of California. Huard, R., 1991. Le suffrage universel en France. Paris: Aubier. 1996. La naissance du parti politique en France. Paris: Presses de la Fondation Nationale des Sciences Politiques. Huntington, S. P., 1968. Political Order in Changing Societies. New Haven: Yale University Press. Hurtig, S., 1971. De la SFIO au nouveau parti socialiste. Paris: Colin. Hyman, R., and A. Ferner, eds., 1992. Industrial Relations and the New Europe. Oxford: Blackwell. Inglehart, R., 1977. The Silent Revolution: Changing Values and Political Styles Among Western Publics. Princeton: Princeton University Press. 1984. The Changing Cleavage Structure of Political Cleavages in Western Society. In Dalton, Flanagan, and Beck (1984: 25–69). Irwin, G. A., and J. J. M. Holsteyn, 1989. Decline of the Structured Model of Competition. In Daalder and Irwin (1989: 21–39). Jollivert, M., 1972. Socie´te´s rurales et classes sociales. In Tavernier, Gervais, and Servolin (1972: 79–106). Jost, H. U., 1967. Politisches System und Wahlsystem der Schweiz unter dem Aspekt von Integration und Legitimita¨t. Schweizerisches Jahrbuch fu¨r Politische Wissenschaft 16: 203–219. Joyce, P., 1980. Work, Society, and Politics: The Culture of the Factory in Late Victorian England. New Brunswick, NJ: Rutgers University Press.

REFERENCES

613

Judt, T. R., 1976. La reconstruction du parti socialiste 1921–1926. Paris: Presses de la FNSP. 1979. Socialism in Provence 1871–1914. A Study in the Origins of the Modern French Left. Cambridge: Cambridge University Press. Jutikkala, E., and K. Pirinen, 1962. A History of Finland. London: Thames and Hudson. Kaase, M., and K. Von Beyme, eds., 1978. Elections and Parties: Socio-Political Change and Participation in the West German Federal Elections of 1976. London: Sage. Kaelble, H. Was Prometheus Most Unbound in Europe? Labor Force in Europe during the late 19th and 20th Century. Unpublished manuscript. Kahn-Freund, O., 1972. Labour and the Law. London: Stevens. Kalivas, S. N., 1996. The Rise of Christian Democracy in Europe. Ithaca, NY: Cornell University Press. Kassalow, E. M., ed., 1963. National Labor Movements in the Post-War World. Evanston, IL: Northwestern University Press. Katz, R., 1990. Party as Linkage: A Vestigial Function? European Journal of Political Research 18: 143–161. Katz, R., and P. Mair, eds., 1992. Party Organization: A Data Handbook. London: Sage. Katzenstein, P., 1984. Corporatism and Change: Austria, Switzerland and the Politics of Industry. Ithaca, NY: Cornell University Press. Katznelson, I., 1985. Working-Class Formation and the State: Nineteenth-Century England in American Perspective. In Evans, Rueschemeyer, and Skocpol (1985: 257–284). 1986. Working Class Formation: Constructing Cases and Comparisons. In Katznelson and Zolberg (1986: 3–41). Katznelson, I., and A. R. Zolberg, eds., 1986. Working Class Formation in Western Europe and the United States. Princeton: Princeton University Press. Kautsky, K., 1899. Bernstein und das Sozialdemokratische Programm: Eine Antikritik. Berlin: Dietz. Keman, H., 1988. The Development Toward Surplus Welfare: Social Democratic Politics and Policies in Advanced Capitalist Democracies (1965–1984). Amsterdam: CT Press. Kendall, W., 1975. The Labour Movement in Europe. London: Allen Lane. Kergoat, J., 1990. France. In Linden and Rojahn (1990: 163–190). Kerr, H., 1974. Switzerland: Social Cleavages and Partisan Conflict. London: Sage. 1987. The Swiss Party System: Steadfast and Changing. In Daalder (1987a: 107– 192). Kirby, D., 1971. The Finnish Social Democratic Party 1903–1918. London: Ph.D. thesis, London University. 1986. The Workers’ Cause: Rank and File Attitudes in the Finnish Social Democratic Party 1905–1918. Past and Present no. 111: 144–164. 1990. Finland. In Linden and Rojahn (1990: 523–540). Kirchheimer, O., 1965. Conﬁning Conditions and Revolutionary Breakthroughs. American Political Science Review 61: 964–974. 1966a. Germany: The Vanishing Opposition. In Dahl (1966a: 237–259).

614

REFERENCES

1966b. The Transformation of Western European Party Systems. In La Palombara and Weiner (1966: 177–200). Kitschelt, H., 1994. The Transformation of European Social Democracy. Cambridge: Cambridge University Press. Kittell, A. K., 1973. Socialist versus Catholic in Belgium: The Role of Anticlericalism in the Development of the Belgian Left. The Historian 23: 418– 435. Klenner, F., 1951. Die O¨sterreichischen Gewerkschaften, Vol. 1. Vienna: Verlag des O¨GB. Knoellinger, C. E., 1960. Labor in Finland. Cambridge, MA: Harvard University Press. Knopp, K., 1980. Austrian Social Democracy 1889–1914. Washington, DC: University Press of America. Kocka, J., 1989a. Borghesia e societa` borghese nel XIX secolo. Sviluppi europei e peculiarita` tedesche. In Kocka (1989b: 3–67). ed., 1989b. Borghesie Europee del’ottocento. Venice: Marsilio Editori. Koeble, H., 1989. Borghesia francese e borghesia tedesca. 1870–1914. In Kocka (1989b: 127–159). Koole, R. A., and G. Voerman, 1986. Het Lidmaatschap van Politieke Partijen na 1945. In Jaarboek 1985. Groningen: Documentatiecentrum Nederlandse Politieke: 115–176. Kornhauser, W., 1959. The Politics of Mass Society. New York: Free Press. Korpi, W., 1971. Working Class Communism in Western Europe: Rational or Non-rational. American Sociological Review 36: 971–984. 1983. The Democratic Class Struggle. London: Routledge and Kegan Paul. Kossmann, E. H., 1978. The Low Countries 1780–1940. Oxford: Clarendon Press. Kriegel, A., 1969. Aux origines du communisme franc¸ais. Contribution a` l’histoire du mouvement ouvrier franc¸ais. Paris: Flammarion. 1973. Le internazionali operaie (1864–1943). Messina and Florence: D’Anna. 1974. La crise re´volutionnaire (1919–1920): Hypothe`ses pour la construction d’un mode`le’. In Communismes au miroir francais. Temps, cultures et socie´te´s en France devant le communisme. Paris: Gallimard. Kriesi, H., 1990. Federalism and Pillarization: The Netherlands and Switzerland Compared. Acta Politica 25: 433–450. Kuhnle, S., 1978. The Beginning of the Nordic Welfare States: Similarities and Differences. Acta Sociologica Supplement 21: 9–35. La Martine, de A., 1895. Histoire des Girondins, 4 vols. Paris: Furre. Lafferty, W., 1971. Economic Development and the Response of Labor in Scandinavia: A Multi-Level Analysis. Oslo: Universitetsforlaget. 1972. Industrialisation and Labor Radicalism in Norway: An Ecological Analysis. Scandinavian Political Studies 7: 157–175. 1974. Industrialisation, Community Structure and Socialism: An Ecological Analysis of Norway 1875–1924. Oslo: Universitetsforlaget. Landauer, C., 1959. European Socialism, 2 vols. Berkeley: University of California Press. 1961. The Guesdists and the Small Farmer: Early Erosion of French Marxism. International Review of Social History 6: 212–225. Landes, D. S., 1969. The Unbound Prometheus: Technological Change and Industrial

REFERENCES

615

Development in Western Europe from 1750 to the Present. Cambridge, UK: Cambridge University Press. Landsberger, H., ed., 1970. The Church and Social Change in Latin America. Notre Dame, IN: University of Notre Dame Press. Lane, J.-E., and P. Pennings, eds., 1998. Party System Change. London: Routledge. Lane, R. E., 1959. Political Life: Why People Get Involved in Politics. New York: Free Press. Lange, P., 1977. La teoria degli incentivi e l’analisi dei partiti politici. Rassegna Italiana di Sociologia 18: 501–526. La Palombara, J., and M. Weiner, eds., 1966. Political Parties and Political Development. Princeton: Princeton University Press. Laponce, L. A., 1981. Left and Right: The Topography of Political Perceptions. Toronto: University of Toronto Press. Laski, H. J., 1919. Authority in the Modern State. New Haven: Yale University Press. Laulajainen, P., 1979. Some Aspects of the Division of the Finnish Working Class Movement after the Civil War: A Research Note. Scandinavian Political Studies 2: 53–63. Lavau, G., 1968. A la recherche d’un cadre the´orique pour l’e´tude du PCF. Revue Franc¸aise de Science Politique 18: 445–466. Lawson, K., ed., 1980. Political Parties and Linkage: A Comparative Perspective. New Haven: Yale University Press. Lazarsfeld, P. F., B. Berelson, and H. Gaudet, 1948. The People’s Choice. New York: Columbia University Press. Lefranc, G., 1963. Le mouvement socialiste sous la IIIe Re´publique 1875–1940. Paris: Payot. Legendre, P., 1968. Histoire de l’administration de 1780 a` nos jours. Paris: Presses Universitaires de France. Lehmbruch, G., 1966. Proporzdemokratie. Tu¨bingen: Mohr. Leichter, O., 1964. Glanz und Elend der Ersten Republik. Vienna: Europa Verlag. Lenin, V. I., 1939. Imperialism: The Highest Stage of Capitalism. New York: International Publishers. Leonard, D., 1975. Paying for Parties: The Case of Public Subsidies. London: Political and Economic Planning. Lepsius, R., 1973. Parteien und Sozialstruktur: Zum Problem der Demokratisierung der deutschen Gesellschaft. In Ritter (1973: 56–80). 1985. The Nation and Nationalism in Germany. Social Research 52: 43–64. Lerner, D., 1957. The ‘‘Hard-Headed’’ Frenchman. Encounter 8: 27–32. 1958. The Passing of Traditional Societies: Modernizing the Middle East. New York: Free Press. Le´vy-Leboyer, M., 1968. Le processus d’industrialisation: Le cas de l’Angleterre et de la France. Revue Historique n. 230: 281–298. Lewin, L., 1988. Ideology and Strategy: A Century of Swedish Politics. Cambridge: Cambridge University Press. Lewis-Beck, M. S., 1977. Explaining Peasant Conservatism: The Western European Case. British Journal of Political Science 7: 447–464. Libbey, K. R., 1970. Initiatives, Referenda and Socialism in Switzerland. Government and Opposition 5: 307–326.

616

REFERENCES

Lidtke, V. L., 1966. The Outlawed Party: Social Democracy in Germany 1878–1890. Princeton: Princeton University Press. Liebman, M., 1979. Les socialistes belges, 1885–1914. La re´volte et l’organisation. Brussels: Vie Ouvrie`re. Ligou, D., 1962. Histoire du socialisme en France: 1871–1961. Paris: Presses Universitaires de France. Lijphart, A., 1968. Typologies of Democratic Regimes. Comparative Political Studies 1: 3–44. 1971. Class Voting and Religious Voting in the European Democracies. Acta Politica 6: 158–171. 1979. Religious vs. Linguistic vs. Class Voting: The Crucial Experiment of Comparing Belgium, Canada, South Africa, and Switzerland. American Political Science Review 73: 442–458. 1984. Democracies: Patterns of Majoritarian and Consensus Government in TwentyOne Countries. New Haven: Yale University Press. Lilja, K., 1992. Finland: No Longer the Nordic Exception. In Hyman and Ferner (1992: 198–217). Lind, G., 1988. Military State and Bourgeois Democracy. In Stra˚th (1988a: 112–119). Lindemann, A. S., 1974. The ‘‘Red Years’’: European Socialism versus Bolshevism, 1919–1921. Berkeley: University of California Press. 1977. Socialismo europeo e bolscevismo (1919–1921). Bologna: Il Mulino. Linden, van der M., 1988. The National Integration of European Working Classes (1871–1914). International Review of Social History 33: 285–311. Linden, van der M., and J. Rojahn, eds., 1990. The Formation of Labour Movements, 1870–1914: An International Perspective. Leiden: E. J. Brill. Linz, J. J., 1959. The Social Bases of German Politics. Ph.D. thesis. New York: Columbia University. 1973. Early State Building and Late Peripheral Nationalism Against the State: The Case of Spain. In Eisenstadt and Rokkan (1973: 32–116). 1976. Patterns of Land Tenure, Division of Labor, and Voting Behavior in Europe. Comparative Politics 8: 365–412. 1979. Europe’s Southern Frontier: Evolving Trends Toward What? Daedalus 108: 175–209. Linz, J. J., and A. Stepan, eds., 1978. The Breakdown of Democratic Regimes: Europe. Baltimore: Johns Hopkins University Press. Lipset, S. M., 1958. Socialism, Left, Right, East and West. Confluence 7: 173–192. 1959. Democracy and Working-Class Authoritarianism. American Sociological Review 24: 482–501. 1960. Political Man. New York: Doubleday. 1963. Some Social Requisites of Democracy: Economic Development and Political Legitimacy. In Polsby, Dentler, and Smith (1963: 541–569). 1964. The Changing Class Structure and Contemporary European Politics. Daedalus Winter: 271–303. 1966. Elections: The Expression of Democratic Class Struggle. In Bendix and Lipset (1966: 413–427). 1969. Revolution and Counter-Revolution. London: Heinemann. 1981. Whatever Happened to the Proletariat? An Historical Mission Unfulﬁlled. Encounter 56/6: 18–34.

REFERENCES

617

1983. Radicalism or Reformism: The Sources of Working Class Politics. International Political Science Review 77: 1–18. Lipset, S. M., and S. Rokkan, eds., 1967. Party Systems and Voter Alignments. New York: Free Press. Logue, J., 1982. Socialism and Abundance: Radical Socialism in the Danish Welfare State. Minneapolis: University of Minnesota Press. Lorwin, R. V., 1954. The French Labour Movement. Cambridge, MA: Harvard University Press. 1966. Belgium, Religion, Class, and Language in National Politics. In Dahl (1966a: 147–187). ed., 1967a. Labor and Working Conditions in Modern Europe. London: CollierMacmillan. 1967b. Working Class Politics and Economic Development in Western Europe. In Lorwin (1967a: 58–72). 1971. Segmented Pluralism, Ideological Cleavages and Political Cohesion in the Smaller European Democracies. Comparative Politics 3: 147– 187. Lowell, A. L., 1896. Governments and Parties in Continental Europe, 2 vols. Boston: Houghton Mifﬂin. Lucas-Busemann, E., 1976. Zwei Formen von Radikalismus in der deutschen Arbeiterbewegung. Frankfurt am Main: Verlag Roter Stern. Luebbert, G., 1991. Liberalism, Fascism, or Social Democracy. Social Classes and the Political Origins of Regimes in Inter-war Europe. Oxford: Oxford University Press. Lukes, S., 1995. What Is Left? Florence: European University Institute, unpublished manuscript. Lundquist, L., 1982. The Party and the Masses: An Interorganisational Analysis of Lenin’s Model for the Bolshevik Revolutionary Movement. Stockholm: Almqvist & Wiksell International. Luxemburg, R., 1899. Sozialreform oder Revolution. Leipzig: Buchdruckerei und Leipziger Volkszeitung. 1963. Scritti scelti, edited by L. Amodio. Milan: Edizioni ‘‘Avanti!.’’ MacCarthy, C., 1977. Trade Unions in Ireland 1894–1960. Dublin: Institute of Public Administration. MacElligot, T. J., 1966. Education in Ireland. Dublin: Institute of Public Administration. Mackie, T. T., and R. Rose, 1974. The International Almanac of Electoral History. London: Macmillan. Mackintosh, J., 1977. The British Cabinet. London: Stevens and Sons. MacRae, D., 1967. Parliament, Parties and Society in France, 1946–1958. New York: St. Martin’s Press. Madeley, J. T. S., 1977. Scandinavian Christian Democracy: Throwback or Portent? European Journal of Political Research 5: 267–286. Maderthaner, W., and W. C., Mu¨ller, eds., 1996. Die Organisation der O¨sterreichischen Sozialdemokratie 1889–1995. Vienna: Lo¨cker Verlag. Mair, P., 1978. Labour and the Irish Party System Revisited: Party Competition in the 1920’s. Economic and Social Review 9: 59–70. 1979. The Marxist Left. In Drucker (1979: 157–182).

618

REFERENCES

1992. Explaining the Absence of Class Politics in Ireland. Proceedings of the British Academy 79: 383–410. 1997. Party System Change. Oxford: Oxford University Press. Mair, P., and R. Katz, 1992. The Membership of Political Parties in European Democracies 1960–1990. European Journal of Political Research 22: 329–345. Malefakis, E., 1974. A Comparative Analysis of Worker Movements in Spain and Italy. San Francisco: American Historical Association. Mann, H., 1973. Consciousness and Action among Western Working Class. London: Macmillan. 1987. Ruling Class Strategies and Citizenship. Sociology 21: 339–354. 1993. The Sources of Social Power, Vol. 2: The Rise of Classes and Nation-States, 1760–1914. Cambridge: Cambridge University Press. March, A. J. G., ed., 1965. Handbook of Organization. Chicago: Rand McNally College Publishing. Markku, H., 1980. Dialectics between Occupation and Party Structures: Finland since World War II. Acta Sociologica 23: 83–96. Marquand, D., 1981. Classe, coalizione e Partito Laburista. In Piro and Pombeni (1981: 166–176). Marshall, T. H., 1965. Class, Citizenship and Social Development. New York: Anchor Books. Martin, C., and D. Martin, 1977. The Decline of the Labour Party Membership. Political Quarterly 48: 459–471. Martin, W., 1980 (1st ed., 1928). Histoire de la Suisse. Lausanne: Payot. Martin, W. C., and K. Hopkins, 1980. Cleavage Crystallization and Party Linkages in Finland, 1900–1918. In Lawson (1980: 183–203). Marx, K., 1973. Il capitale, 3 vols., Prefazione. Rome: Editori Riuniti. (1955–1956. Capital: A Critique of Political Economy, 3 vols. London: Lawrence and Wishart.) Mastropaolo, A., 1980. Electoral Processes, Political Behaviour, and Social Forces in Italy from the Rise of the Left to the Fall of Giolitti, 1876–1913. In Bu¨sch (1980: 97–124). Matthew, C. H., R. I. McKibbin, and J. A. Kay, 1976. The Franchise Factor in the Rise of the Labour Party. English Historical Review 91: 723–752. Mattl, S., 1990. Austria. In Linden and Rojahn (1990: 293–320). McGrath, W. J., 1974. Dionysian Art and Populist Politics in Austria. New Haven: and London: Yale University Press. McKenzie, R. T., and A. Silver, 1967. The Delicate Experiment: Industrialism, Conservatism and Working-Class Tories. In Lipset and Rokkan (1967: 115–125). 1968. Angels of Marble: Working Class Conservatives in Urban England. Chicago: University of Chicago Press. McKibbin, R., 1984. Why Was There No Marxism in Great Britain? English Historical Review 99: 297–331. Meyriat, J., and S. Rokkan, eds., 1969. International Guide to Electoral Statistics. National Elections in Western Europe. Paris: Mouton. Micaud, C. A., 1955. The Bases of Communist Strength in France. Western Political Quarterly 8: 354–366. Michelet, J., 1847–1853. Histoire de la re´volution franc¸aise. Paris: Chamerot. Michels, R., 1906. Die Deutsche Sozialdemokratie. I. Parteimitgliederschaft und

REFERENCES

619

soziale Zusammensetzung. Archiv fu¨r Sozialwissenschaft und Sozialpolitik. Tu¨bingen: Mohr, 471–556. 1949. Political Parties: A Sociological Study of Oligarchical Tendencies of Modern Democracy. Glencoe, IL: Free Press. Milbrath, L. W., 1965. Political Participation. Chicago: Rand McNally. Millen, B., 1963. The Relationship of the Norwegian Labour Party to the Trade Unions. In Kassalow (1963: 119–141). Miller, D. R., 1973. A Shapley Value Analysis of the Proposed Canadian Constitutional Amendment Scheme. Canadian Journal of Political Science 6: 140– 143. Milward, A. S., and S. B. Saul, 1977. The Development of the Economies of Continental Europe, 1850–1914. London: Allen and Unwin. Mintzel, A., 1990. Political and Socio-economic Developments in the Post-war Era: The Case of Bavaria, 1945–1989. In Rohe (1990a: 145–178). Mitchell, A., 1965. Revolution in Bavaria: The Eisner Regime and the Soviet Republic, 1918–1919. Princeton: Princeton University Press. 1974. Labour in Irish Politics, 1890–1930. Dublin: Irish University Press. Mitchell, B. R., 1975. European Historical Statistics 1750–1975, 2nd ed. London: Macmillan. Mitchell, H., and P. Stearns, 1971. Workers and Protest: The European Labor Movement, the Working Class, and the Origins of Social Democracy 1890–1914. Ithaca, NY: Peacock. Mitrany, D., 1951. Marx Against the Peasants. Chapel Hill: University of North Carolina Press. Mommsen, H., 1979. Zum Problem der vergleichenden Behandlung nationaler Arbeiterbewegungen am Beispiel Ost–und Su¨dosttmitteleuropas. Internationale wissenschaftliche Korrespondenz zur Geschicht 15: 31–34. Mommsen, W., and H. G. Husong, eds., 1985. The Development of Trade Unionism in Great Britain and Germany 1880–1914. London: Allen and Unwin. Moore, B. J., Jr., 1966. Social Origins of Dictatorship and Democracy: Lords and Peasants in the Making of the Modern World. Boston: Beacon Press. 1978. Injustice. The Social Bases of Obedience and Revolt. London: Macmillan. Morandi, R., 1978 (1st. ed., 1945). I partiti nella storia d’Italia. Florence: Le Monnier. Morgan, R., 1969. The Federal Republic of Germany. In Henig and Pinder (1969: 21–67). Moses, J. A., 1982. Trade Unionism in Germany from Bismarck to Hitler, 1869–1933. London: George Prior. Moulin, L., 1981. Riflessioni su alcune particolarita` del movimento operaio belga. In Piro and Pombeni (1981: 192–202). Murphy, J. A., 1975. Ireland in the Twentieth Century. Dublin: Gill and Macmillan. Musson, A. E., 1972. British Trade Unions 1800–1875. London: Macmillan. Mylly, J., 1984. The Emergence of the Finnish Multi-Party System. A Comparison with Developments in Scandinavia, 1870–1920. In Mylly and Berry (1984: 9– 27). Mylly, J., and R. M. Berry, eds., 1984. Political Parties in Finland: Essays in History and Politics. Turku, Finland: Graﬁa Oy.

620

REFERENCES

Nedelmann, B., 1975. Handlungsraum politischer Organisationen. Sozialwissenschaftliches Jahrbuch fu¨r Politik 4: 9–118. 1987. Individuals and Parties: Changes in the Processes of Political Mobilisation. European Sociological Review 3: 181–202. Neidhart, L., 1970. Plebiszit und pluralita¨re Demokratie. Eine Analyse der Funktion des schweizerischen Gesetzesreferendums. Bern: Francke Verlag. Nettl, P. J., 1967. Political Mobilization. New York: Basic Books. 1968. The State as a Conceptual Variable. World Politics 20: 560–592. Neufeld, M., 1974. Italy: School for Awakening Countries: The Italian Labour Movement in Its Political, Economic, and Social Setting from 1800–1960. New York: Greenwood Press. Neumann, S., ed., 1956a. Modern Political Parties. Chicago: University of Chicago Press. 1956b. Toward a Comparative Study of Political Parties. In Neumann (1956a: 395–421). Niehuss, M., 1990. Party Configurations in State and Municipal Elections in Southern Germany, 1871–1914. In Rohe (1990a: 83–106). Nieuwbeerta, P., 1995. The Democratic Class Struggle in Twenty Countries 1945/ 1990. Amsterdam: Thesis Publisher. Nohlen, D., and R. O. Schultze, 1971. Wahlen in Deutschland: Theorie – Geschichte– Dokumente, 1848–1970. Berlin: Walter de Gruyter. Noiret, S., 1990a. Partiti politici e sistema politico in Belgio (1830–1980). Ricerche di storia politica n. 5: 87–128. ed., 1990b. Political Strategies and Electoral Reforms: Origins of Voting Systems in Europe in the 19th and 20th Centuries. Baden-Baden: Nomos Verlagsgesellschaft. Noland, A., 1970. The Founding of the French Socialist Party. New York: Howard Fertig. Nordlingler, E., 1972. Conflict Regulation in Divided Societies. Cambridge, MA: Harvard University Occasional Paper in International Affairs no. 29. Nousiainen, J., 1960. The Structure of the Finnish Political Parties. In Democracy in Finland: Studies in Politics and Government. Helsinki: Finnish Political Science Association, 28–43. Offe, C., 1972. Strukturprobleme des kapitalistischen Staates: Aufsa¨tzung zur politischen Soziologie. Frankfurt: Suhrkamp. Olson, M., 1963. Rapid Growth as a Destabilizing Force. Journal of Economic History 23: 529–552. 1971 (1st ed. 1965). The Logic of Collective Action. Public Goods and the Theory of Groups. Cambridge: Harvard University Press. Olsson, S. E., 1990. Social Policy and Welfare State in Sweden. Stockholm: Archiv Fo¨rlag. Orridge, A., 1975. The Irish Labour Party. Journal of Common Market Studies 13: 484–491. Ostrogorski, M., 1970 (1st ed., 1902). Democracy and the Organization of Political Parties, 2 vols. New York, Haskell House. Paloheimo, H., 1984. Governments in Democratic Capitalist States 1950–1983. A Data Handbook. Turku, Finland: Department of Sociology and Political Science, University of Turku Studies on Political Science no 8.

REFERENCES

621

Panebianco, A., 1982. Modelli di partito. Organizzazione e potere nei partiti politici. Bologna: Il Mulino. (1988. A. Political Parties: Organization and Power. Cambridge: Cambridge University Press.) Pappalardo, A., 1989. I sindacati prefascisti. Ricostruzione e verifica di ipotesi. Florence: Edizioni Polistampa. Parkin, F., 1967. Working Class Conservatism: A Theory of Political Deviance. British Journal of Sociology 18: 278–290. Parri, L., 1987. Staat und Gewerkschaften in der Schweiz (1873–1891). Politische Vierteljahreschrift 28: 35–58. Paterson, W. E., 1977. The German Social Democratic Party. In Paterson and Thomas (1977: 176–212) Paterson, W. E., and A. H. Thomas, eds., 1977. Social Democratic Parties in Western Europe. London: Croom Helm. Pedersen, M. N., ed., 1979. Dansk Politik i 1970’erne. Copenhagen: Samfundsvidens-Kabeligt Forlag. 1980. On Measuring Party System Change: A Methodological Critique and a Suggestion. Comparative Political Studies 12: 387–403. Pelinka, A., 1978. O¨sterreich. In Raschke (1978: 412–432). Pelling, H., 1965. The Origins of the Labour Party. London: Oxford University Press. ed., 1968a. Popular Politics and Society in Late Victorian England. London: Macmillan. 1968b. The Concept of Labour Aristocracy. In Pelling (1968a: 37–61). 1976a. A History of British Trade Unionism. London: Macmillan. 1976b. A Short History of the Labour Party. London: Macmillan. Penniman, H. R., ed., 1977. Italy at the Polls. The Parliamentary Elections of 1976. Washington, DC: American Enterprise Institute. Pepe, G., 1971. Storia della CGL dalla fondazione alla guerra di Libia 1905–1911. Bari: Laterza. Perrot, M., 1967. Les guesdesistes: Controverse sur l’introduction du marxisme en France. Annales: Economies, Socie´te´s, Civilisations 22: 701–710. Pesonen, P., 1968. An Election in Finland: Party Activities and Voters’ Reactions. New Haven: Yale University Press. Pesonen, P., and R. Sa¨nkiaho, 1979. Kansalaiset ja kansanvalta. Juva, Finland: So¨derstro¨m. Piro, F., and P. Pombeni, eds., 1981. Movimento operaio e societa` industriale in Europa, 1870–1970. Venice: Marsilio. Pizzorno, A., 1962. Sviluppo economico e urbanizzazione. Quaderni di sociologia 11: 23–51. 1966a. Amoral Familism and Historical Marginality. International Review of Community Development 15: 55–66. 1966b. Introduzione allo studio della partecipazione politica. Quaderni di sociologia 15: 1–52. 1967. The Individualistic Mobilization of Europe. In Graubard (1967: 281–283). 1969. Elementi di uno schema teorico con riferimento i partiti politici in Italia. In Sivini (1969: 5–40). Pojas, M., 1991. The Swedish Model in Historical Perspective. Scandinavian Economic History Review 39: 64–74.

622

REFERENCES

Polsby, N., R. A. Dentler, and P. Smith, 1963. Politics and Social Life. An Introduction to the Study of Political Behavior. Boston: Houghton Mifﬂin. Portes, A., 1970. Leftist Radicalism in Chile: A Test of Three Hypotheses. Comparative Politics 2: 251–274. 1971a. On the Logic of Post-factum Explanation: The Hypothesis of LowerClass Frustration as the Cause of Leftist Radicalism. Social Forces 50: 26– 44. 1971b. Political Primitivism, Differential Socialization and Lower Class Leftist Radicalism. American Sociological Review 36: 820–835. Potter, A., 1966. Great Britain: Opposition with a Capital O. In Dahl (1966a: 3– 33). Potthoff, P., 1979. Gewerkschaften und Politik zwischen Revolution und Inflation. Du¨sseldorf: Droste. Pre´lot, M., 1939. L’evolution politique du socialisme franc¸ais, 1789–1934. Paris: Spes. Price, R., 1990. Britain. In Linden and Rojahn (1990: 3–24). Procacci, G., 1970. Storia degli italiani, 2 vols. Bari, Italy: Laterza. Prodzyznski, F., 1992. Ireland Between Centralism and the Market. In Hyman and Ferner (1992: 69–87). Przeworski, A., 1985. Capitalism and Social Democracy. New York: Cambridge University Press. Przeworski, A., and J. Sprague, 1986. Paper Stones: A History of Electoral Socialism. Chicago: University of Chicago Press. PSI, 1975. Il Partito Socialista: Struttura e Organizzazione. Venice: Marsilio. 1976. Il partito e il paese: Annuario statistico 1976. Rome: Quaderni della sezione centrale di organizzazione. Rae, D. W., 1971. The Political Consequences of Electoral Laws. New Haven: Yale University Press. Rae, D. W., and M. Taylor, 1970. The Analysis of Political Cleavages. New Haven: Yale University Press. Ragionieri, E., 1976. La storia politico-sociale. In Storia d’Italia dall’unita` ad oggi, Vol. 4, Tome 3, 1774–1785. Turin: Einaudi. Raschke, J., ed., 1978. Die politischen Parteien in Westeuropa. Reinbek bei Hamburg: Rowohlt Taschenbuch Verlag. Rattinger, H., 1981. Measuring Power in Voting Bodies: Linear Constraints, Spatial Analysis, and a Computer Programme. In Holler (1981: 222–231). Rawson, D. W., 1969. The Life-Span of Labour Parties. Political Studies 17: 313– 333. Regini, M., ed., 1992. The Future of Labour Movements. London: Sage. Reymond-Sauvain, P., 1966. Le syndicalisme en Suisse. Geneva´: Editions Ge´ne´rales S.A. Ribeni, L., 1989. Rivoluzione e democrazia in Germania. La rivoluzione del 1918 nella storiograﬁa tedesco-federale degli anni 80. Rivista di storia politica no. 4: 137–154. Rice, G. W., 1973. The Electoral Prospects of Non-Ruling Communist Parties. American Journal of Political Science 17: 597–619. Righart, H., 1986. De Katholieke Zuil in Europa. Amsterdam: Boom. Rimlinger, G. V., 1960. The Legitimation of Protest. A Comparative Study in Labor History. Comparative Studies in Society and History 2: 329–343.

REFERENCES

623

Ritter, G. A., 1963. Die Arbeiterbewegung im Wilhelminischen Reich. Die sozialdemokratische Partei und die freien Gewerkschaften 1890–1900. Berlin: Colloquium Verlag. ed., 1973. Deutschen Parteien vor 1918. Cologne: Kiepenheuer und Witsch. 1985. Die deutschen Parteien 1830–1914. Parteien und Gesellschaft im konstitutionellen Regierungssystem. Go¨ttingen: Vandenhoeck & Ruprecht. 1990. The Electoral Systems of Imperial Germany and Their Consequences for Politics. In Noiret (1990b: 53–75). Robert, J. L., 1980. La scission syndicale de 1921: Essai de reconnaissance des formes. Paris: Centre de recherches d’histoire des mouvements sociaux et du syndicalisme. Roberts, R., 1958–1959. Appendix. Journal of the Statistical and Social Inquiry Society of Ireland 20. Roche, W. K., 1990. A Model Working Class? The Liberal Theory of Industrialization and the Maturation of the Irish Labour Movement. Oxford: Nufﬁeld College, Conference on the Development of the Industrial Society in Ireland. 1992. Modelling Trade Union Growth and Decline in the Republic of Ireland. Irish Business and Administrative Research 13: 86–102. Rohde, P. P., 1973. The Communist Party of Denmark. In Upton (1973: 3–33). Rohe, K., ed. 1990a. Elections, Parties and Political Traditions. Social Foundations of German Parties and Party Systems, 1867–1987. Munich: Berg. 1990b. Political Alignments and Re-alignments in the Ruhr, 1867–1987: Continuity and Change of Political Traditions in an Industrial Region. In Rohe (1990a: 107–144). Rokkan, S., ed., 1962. Approaches to the Study of Political Participation. Bergen: Chr. Michelson Institute. 1966a. Electoral Mobilization, Party Competition and National Integration. In La Palombara and Weiner (1966: 241–266). 1966b. Norway: Numerical Democracy and Corporate Pluralism. In Dahl (1966a: 70–115). 1967. Norway: Geography, Religion and Social Class: Cross Cutting Cleavages in Norwegian Politics. In Lipset and Rokkan (1967: 367–444). 1970a. Citizens, Elections, Parties. Oslo: Universitetsforlaget. 1970b. Nation-Building, Cleavage Formation and the Structuring of Mass Politics. In Rokkan (1970a: 72–144). 1970c. The Mobilization of the Periphery: Data on Turnout, Party Membership and Candidate Recruitment in Norway. In Rokkan (1970a: 181–225). 1970d. The Comparative Study of Political Participation. In Rokkan (1970a: 13– 45). 1970e. Readers, Viewers, Voters. In Rokkan (1970a: 417–431). 1970f. Electoral Systems. In Rokkan (1970a: 147–168). 1970g. The Growth and the Structuring of Mass Politics in Europe: Reﬂections of Possible Models of Explanation. Scandinavian Political Studies Yearbook 5: 65–83. 1973. Cities, States, and Nations: A Dimensional Model for the Study of Contrasts in Development. In Eisenstadt and Rokkan (1973: 73–98). 1974. Entries, Voices, Exits: Towards a Possible Generalization of the Hirschman Model. Social Science Information no. 13: 39–53.

624

REFERENCES

1975. Dimensions of State Formation and Nation-Building: A Possible Paradigm for Research on Variations within Europe. In Tilly (1975: 562–601). 1977. Toward a Generalized Concept of ‘‘Verzuiling’’: A Preliminary Note. Political Studies 25: 563–570. Rokkan, S., and D. Urwin, 1983. Economy, Territory, Identity: Politics of West European Peripheries. London: Sage. Rokkan, S., D. Urwin, F. H. Aarebrot, P. Malaba, and T. Sande, 1987. Centre– Periphery Structures in Europe. Frankfurt: Campus Verlag. Rokkan, S. and H. Valen, 1962. The Mobilization of the Periphery: Data on Turnout, Membership and Candidate Recruitment in Norway. In Rokkan (1962: 111– 158). Rose, R., ed., 1974. Electoral Behaviour: A Comparative Handbook. New York: Free Press. ed., 1980. Electoral Participation: A Comparative Analysis. London: Sage. Rossi, M. G., 1985. Da Sturzo a De Gasperi. Profilo storico del cattolicesimo politico nel Novecento. Rome: Editori Riuniti. Rostow, W. W., 1960. The Stages of Economic Growth. Cambridge: Cambridge University Press. 1978. The World Economy: History and Prospects. Austin: University of Texas Press. Rotelli, E., 1975. Le componenti istituzionali del sistema politico dell’Italia liberale. Alternativa 1: 3–17. Rothstein, B., 1992. Labor Market Institutions and Working-Class Strength. In Steinmo, Thelen, and Longstreth (1992: 57–89). Rowies, L., 1977. Les partis politiques. Brussels: CRJSP. Rucht, D., and T. Ohlemacher, 1992. Protest Event Data: Collection, Uses and Perspectives. In Diani and Eyerman (1992: 76–106). Rueschemeyer, D., E. Huber, and J. D. Stephens, 1992. Capitalist Development and Democracy. Cambridge: Polity Press. Ruitenbeek, H. M., 1955. Het ontstaan van de Partij van de Arbeid. Amsterdam: De Arbeiderspers. Russett, B. M., 1964. Inequality and Instability: The Relation of Land Tenure to Politics. World Politics 16: 442–454. Rydenfeld, S., 1954. Communismen i Sverige. Lund, Sweden: Gleerupska Universitetsbokhandeln. Salles, R., 1977. Le parti communiste de Grande Bretagne et les e´lections. Revue Franc¸aise de Science Politique 27: 407–427. Sani, G., 1977. The Italian Electorate in the Mid-1970s: Beyond Tradition. In Penniman (1977: 81–122). Sani, G., and G. Sartori, 1983. Polarization, Fragmentation and Competition in Western Democracies. In Daalder and Mair (1983: 307–340). Sa¨nkiaho, R., 1971. A Model of the Rise of Populism and Support for the Finnish Rural Party. Scandinavian Political Studies 6: 27–47. Sannerstedt, A., and M. Sjo¨lin, 1992. Sweden: Changing Party Relations in a More Active Parliament. In Damgaard (1992: 100–151). Santarelli, E., 1964. La revisione del marxismo in Italia. Milan: Feltrinelli. Sa¨rlvik, B., 1974. Sweden: The Social Bases of the Parties in a Developmental Perspective. In Rose (1974: 371–434).

REFERENCES

625

Sauvy, A., 1967. Histoire e´conomique de la France entre les deux guerres. Paris: Fayard. Scase, R., 1977. Social Democracy in Sweden. In Paterson and Thomas (1977: 316– 341). Schaffner, M., 1982. Die demokratische Bewegung der 1860er Jahre. Beschreibung und Erkla¨rung der Zu¨richer Volksbewegung von 1867. Basel and Frankfurt: Helbing und Lichtenhahn. Scheuer, S., 1992. Denmark: Return to Decentralisation. In Hyman and Ferner (1992: 168–197). Schmitter, P., 1977. Models of Interest Intermediation and Models of Societal Change in Western Europe. Comparative Political Studies 10: 7–38. Scholten, I., 1980. Does Consociationalism Exist? In Rose (1980: 329–354). Scho¨nhouver, K., 1980. Expansion und Konzentration. Studien zur Entwicklung der freien Gewerkschaften im Wilhelminischen Deutschland 1890 bis 1914. Stuttgart: Klen-Cotta. Schorske, C. E., 1955. German Social Democracy of 1905–1917: The Development of the Great Schism. Cambridge, MA: Harvard University Press. Schorter, E., and C. Tilly, 1974. Strikes in France, 1830–1968. London: Cambridge University Press. Scho¨ttler, P., 1986. Syndikalismus in der europa¨ischen Arbeiterbewegung. Neuere Forschungen in Frankreich, England und Deutschland. In Tenfelde (1986: 419– 476). Schram, R., 1954. Protestantism and Politics in France. Alenc¸on: Corbie`re & Jugain. Schumpeter, J. A., 1951. Social Classes in an Ethnically Homogeneous Environment. In Imperialism and Social Classes. New York: Augustus N. Kelly. 1976 (1st ed., 1943). Capitalism, Socialism and Democracy, 5th ed. London: Allen & Unwin). Sechi, S., 1977. Lenin e il problema della rivoluzione in Europa. Introduction to Lindemann (1977: 11–48). Seiler, D. L., 1977. Clevages, regions et science politique: Applications d’un schema d’analyse au cas de la Suisse et de la Belgique. Revue Canadienne de Science Politique 10: 447–472. Selznick, P., 1960. The Organizational Weapon. Glencoe, IL: Free Press. Sewell, W. H., Jr., 1986. Artisans, Factory Workers and the Formation of the French Working Class 1789–1848. In Katznelson and Zolberg (1986: 45–70). Seyd, P., and L. Minkin, 1979. The Labour Party and Its Members. New Society 20. Shalev, M., and W. Korpi, 1980. Working Class Mobilization and American Exceptionalism. Economic and Industrial Democracy 1: 31–61. Shapley, L. S., 1953. A Value for N-Person Games: Contribution to the Theory of Games. In Tucker and Kuhn (1953: 203–317). Shapley, L. S., and M. Shubik, 1954. A Method for Evaluating the Distribution of Power in a Committee System. American Political Science Review 48: 787– 792. Sharp, S. A., 1994. A Statistical Study of Selective Candidate Withdrawal in a British General Election. Electoral Studies 13: 107–131. Shepis, G. 1958. Le consultazioni popolari in Italia dal 1948 al, 1957: Profilo storicostatistico. Empoli: Edritice Caparrini. Siegfried, A., 1913. Tableau politique de la France de l’ouest sous la Troisie`me Re´publique. Paris: Colin.

626

REFERENCES

Simmel, G., 1903. Die Großstadt und das Geisterleben. In Bu¨cher (1903: 185– 206). 1956. Conflict: The Web of Group Affiliation. New York: Free Press. Simmons, H. G., 1970. French Socialists in Search of a Role 1956–1967. Ithaca, NY: Cornell University Press. Simonson, B., 1990. Sweden. In Linden and Rojahn (1990: 85–102). Sivini, G., 1969. Partiti e partecipazione politica in Italia. Milan: Giaffre´. Sjo¨blom, G., 1978. The Swedish Party System in the Post-War Period. Florence: European University Institute, Conference on Recent Changes in European Party Systems. Skocpol, T., 1979. States and Social Revolutions. A Comparative Analysis of France, Russia and China. Cambridge: Cambridge University Press. Skogh, G., 1984. Employers Associations in Sweden. In Windmuller and Gladstone (1984: 149–168). Soares, G. A. D., 1966. Economic Development and Class Structure. In Bendix and Lipset (1966: 190–199). Sombart, W., 1905. Studien zur Entwicklungsgeschichte des nordamerikanischen Proletariats. Archiv fu¨r Sozialwissenschaft und Sozialpolitik 21: 210–236, 308–346, 556–611. 1924. Der proletarische Sozialismus. Jena: G. Fischer. Sorokin, P. A., C. G. Zimmerman, and C. J. Galpin, 1930. A Systematic Source Book in Rural Sociology. Minneapolis: University of Minnesota Press. Sparre, N. S., 1981a. Factional Strife in the Norwegian Labour Party, 1918–1924. Journal of Contemporary History 16: 691–704. 1981b. Labour Insurgencies in Norway: The Crisis of 1917–20. Social Science History 5: 393–413. Spitaels, G., 1974. Le mouvement sysndicale en Belgique. Brussels: Editions de l’Universite´ de Bruxelles. SPO, 1945ff. Berichte an den Bundesparteitag der SPO. Vienna. Spriano, P., 1964. L’occupazione delle fabbriche. Settembre 1920. Turin: Einaudi. Stammer, O., ed., 1968. Party System, Party Organisation and the Politics of the New Masses. Berlin: Free University, Institute for Political Science. Stearns, P., 1971. Revolutionary Syndicalism and French Labour. Rutgers: Rutgers University Press. Steenson, G. P., 1991. After Marx, Before Lenin. Marxism and Socialist WorkingClass Parties in Europe, 1884–1914. Pittsburgh: University of Pittsburgh Press. Steinberg, M., 1971. A l’origine du communisme belge: L’extreˆme gauche re´volutionnaire d’avant 1914. Cahiers Marxistes 8: 3–34. Steiner, H., 1964. Die Arbeiterbewegung Oesterreichs 1867–1889. Vienna: Europa Verlag. Steinmo, S., K. Thelen, and F. Longstreth, eds., 1992. Structuring Politics. Historical Institutionalism in Comparative Analysis. Cambridge: Cambridge University Press. Stephens, J. D., 1979. The Transition from Capitalism to Socialism. London: Macmillan. Stettler, P., 1980. Die Kommunistische Partei der Schweiz 1921–1931. Ein Beitrag zur schweizerischer Parteiforschung und zur Geschichte der schweizerischen Arbeiter-

REFERENCES

627

bewegung im Rahmen der Kommunistischen Internationalen. Bern: Francke Verlag. Stinchome, A. L., 1965. Social Structure and Organisation. In March (1965: 142– 193). Stjernquist, N., 1966. Sweden: Stability or Deadlock. In Dahl (1966a: 116–146). Stoetzel, J., 1955. Voting Behaviour in France. British Journal of Sociology 6: 118– 119. Stra˚th, B., 1983. Workers’ Radicalism in Theory and Practice. Scandinavian Journal of History 8: 261–291. 1988a. Continuity and Discontinuity in Passing from I to II: Swedish 19th Century Civil Society, Culture, Social Formations, and Political Change. In Stra˚th (1988b: 21–42). ed., 1988b. Democratisation in Scandinavia in Comparison. Gothenburg: Gothenburg University, Department of History. Strikwerda, C., 1988. The Divided Class: Catholics versus Socialists in Belgium, 1880–1914. Comparative Studies in Society and History 31: 333– 359. Studer, B., 1993. Le parti communiste suisse. Cahiers d’Histoire du Mouvement Ouvrier n. 9: 5–38. Suleiman, E. N., 1974. Politics, Power and Bureaucracy in France. The Administrative Elite. Princeton: Princeton University Press. Sundberg, J., 1987. Exploring the Case of Declining Party Membership in Denmark: A Scandinavian Comparison. Scandinavian Political Studies 10: 17– 38. Suviranta, A. J., 1987. Finland. In Blainpain (1987: para 275). Sva˚sand, L., 1978. On the Formation of Political Parties: Conditions, Causes and Patterns of Development. Grenoble: ECPR. 1980. Democratisation and Party Formation in Scandinavia. In Bu¨sch (1980: 398–411). Sworatowski, W. S., ed., 1973. World Communism: A Handbook. Washington, DC: Hoover Institution Press. Szabo, E., 1982. Socialism and Social Science: Selected Writings of Ervin Szabo (1877– 1918), edited by G. Litua`n and J. Back. London: Routledge and Kegan Paul. Talos, E., 1981. Staatliche Sozialpolitik in Oesterreich. Rekonstruktion und Analyse. Vienna: Verlag fu¨r Gesellschaftskritik. Tannahill, R. N., 1976. Leadership as a Determinant of Diversity in Western European Communism. Studies in Comparative Communism 9: 349–368. Tanner, D., 1990. Political Change and the Labour Party, 1900–1918. Cambridge: Cambridge University Press. Tarrow, S. G., 1967. Peasant Communism in Southern Italy. New Haven: Yale University Press. 1972. Partito comunista e contadini nel mezzogiorno. Turin: Einaudi. 1977. Between Center and Periphery: Grass-Roots Politicians in Italy and France. New Haven: Yale University Press. Tarrow, S. G., and W. Ascher, 1975. The Stability of Communist Electorates: Evidence from a Longitudinal Analysis of French and Italian Aggregate Data. American Journal of Political Science 19: 475–499.

628

REFERENCES

Tavernier, Y. M., 1972. Le Modef. In Tavernier, Gervais, and Servolin (1972: 467– 495). Tavernier, Y. M., Gervais, and C. Servolin, 1972. L’univers politique des paysans dans la France contemporaine. Paris: Colin. Taylor, C. L., and M. Hudson, 1972. World Handbook of Political and Social Indicators. New Haven: Yale University Press. Taylor, S., 1978. Parkin’s Theory of Working Class Conservatism: Two Hypotheses Investigated. Sociological Review 26: 827–842. Tenfelde, K., 1981. Il movimento operaio tedesco nella Prima Guerra Mondiale. Quellen und Forschungen aus italienischen Archiven und Bibliotheken 61: 216– 247. ed., 1986. Arbeiter und Arbeiterbewegung im Vergleich. Berichte zur internationalen historischen Forschung. Munich: Oldenburg. 1990. Germany. In Linden and Rojahn (1990: 243–270). Terjesen, E. A., 1990. Norway. In Linden and Rojahn (1990: 103–130). Therborn, G., 1977. The Rule of Capitalism and the Rise of Democracy. New Left Review no.103: 3–41. Thomas, H., 1977. Social Democracy in Denmark. In Paterson and Thomas (1977: 234–271). Thompson, E. P., 1963. The Making of the English Working Class. London: Victor Gollancz. Thompson, J. G., 1927. Urbanization: Its Effects on Government and Society. New York: Dutton. Thornley, D., 1974. Historical Introduction. In Chubb (1974: 1–42). Tilly, C., ed., 1975. The Formation of National States in Western Europe. Princeton: Princeton University Press. 1978. From Mobilization to Revolution. Reading, MA: Addison-Wesley. Tilly, C., L. Tilly, and R. H. Tilly, 1975. The Rebellious Century 1830–1930. Cambridge, MA: Harvard University Press. Tilly, R. H., 1980. Kapital, Staat und sozialer Protest in der deutschen Industrialisierung, Vol. 41. Gottingen: Gesammelte Aufsa¨tze. Kritische Studien zur Geschichtswissenschaft. 1989. Comportamento e etica industriale. Indizi di una cultura borghese nella Germania del XIX secolo. In Kocka (1989a: 325–355). Tilton, T. A., 1974. The Social Origins of Liberal Democracy: The Swedish Case. American Political Science Review 68: 561–571. Timoten, P., 1973. Sodan Vaikutukset Poliittiseen Ka¨ytta¨ytimiesen. Politiika 15: 106–128. Tingsten, H., 1937. Political Behavior. London: P. S. King & Son. Torgersen, U., 1962. The Trend Toward Political Consensus: The Case of Norway. Acta Sociologica 6: 159–172. Torsvik, P., ed., 1981. Mobilization, Center–Periphery Structures and Nation Building. Bergen: Universitetsforlaget. Tracy, M., 1964. Agriculture in Western Europe: Challenge and Response 1880–1980. London: Granada. Traxler, F., 1992. Austria: Still the Century of Corporatism. In Hyman and Ferner (1992: 270–297).

REFERENCES

629

Trebilcock, C., 1981. The Industrialisation of the Continental Powers, 1780–1914. London: Longman. Trede, M., ed., 1992. Electricite´ et e´lectrification dans le monde 1880–1980. Paris: Presses Universitaires de France. Tucker, A. W., and H. W. Kuhn, eds., 1953. Contributions to the Theory of Games. Princeton: Princeton University Press. Tucker, R. C., 1967. De-Radicalization of Marxist Movements. American Political Science Review 61: 341–358. Tuominen, A., 1955. The Northern European Communist Parties. Occidente 11: 193–210. Tyo¨ela¨ma¨n Suhteet – Aikasarjola 1907–1988, 1990. Helsinki: Tyo¨ministerio¨. Ulam, A. B., 1979. The Unfinished Revolution: Marxism and Communism in the Modern World, 2nd ed. London: Longman. Upton, A. F., ed., 1973. The Communist Parties of Scandinavia and Finland. London: Weidenfeld and Nicolson. Urwin, D. W., 1974. Germany: Continuity and Change in Electoral Politics. In R. Rose (1974: 109–170). 1980. From Ploughshare to Ballotbox. The Politics of Agrarian Defense in Europe. Oslo: Universitetsforlaget. Urwin, D. W., and F. Aarebrot, 1981. The Socio-Geographic Correlates of LeftWing Voting in Weimar Germany 1924–1932. In Torsvik (1981: 241– 273). Valen, H., and S. Rokkan, 1974. Conflict Structure and Mass Politics in a European Periphery. In Rose (1974: 315–370). Valenzuela, J. S., 1981. Uno schema teorico per l’analisi della formazione dei movimenti operai. Stato e Mercato 1: 447–482. 1992. Labour Movements and Political Systems: Some Variations. In Regini (1992: 53–101). Vallier, I., 1970a. Insulation, and Re-entry: Toward a Theory of Social Religious Change. In Landsberger (1970: 9–35). 1970b. Social Control and Modernization in Latin America. Englewood Cliffs, NJ: Prentice-Hall. Vaussard, M., 1956. Histoire de la de´mocratie chre´tienne. France, Belgique, Italie. Paris: Seuil. Veblen, T., 1994 (1st ed., 1915). Imperial Germany and the Industrial Revolution. In The Collected Works of Thorstein Veblen, Vol. 4. London: Routledge. Verba, S., N. H. Nie, and J. Kim, 1978. Participation and Political Equality: A Seven Nation Comparison. Cambridge: Cambridge University Press. Vilrokx, J., and J. Leemput, 1992. Belgium: A New Stability in Industrial Relations. In Hyman and Ferner (1992: 357–392). Visser, J., 1987. In Search of Inclusive Unionism: A Comparative Analysis. Amsterdam: University of Amsterdam, PhD dissertation. Vorrink, K., 1945. Een Halve Eeuw Beginselstrijd: Overdenkingen over verleden en toekomst bij een historische mijlpaal. Amsterdam: De Arbeiderspers. Vuilleumier, M., 1988. Autour de la ‘‘fondation’’ du Parti Socialiste Suisse. In Les origines du socialisme en Suisse Romande 1880–1920. Lausanne: Association pour l’Etude du Mouvement Ouvrier, Cahier no. 5.

630

REFERENCES

Wald, K. D., 1983. Crosses on the Ballot: Patterns of British Voter Alignments Since 1885. Princeton: Princeton University Press. Waldbrunner, K., 1952. Problems of Nationalization. Socialist International Information no. 2: 13–17. Wallerstein, M., 1989. Union Growth in Advanced Industrial Democracies. American Political Science Review 83: 481–501. Warwich, P., 1977. The French Popular Front: A Legislative Analysis. Chicago: University of Chicago Press. Wearmouth, F., 1937. Methodism and the Working Class Movement in England 1800– 1850. London: Epworth Press. Webb, S., G. B. Shaw, G. Wallace, S. Olivier, W. Clarke, A. Besent, and H. Bland, 1962 (1st ed., 1931). Fabian Essays. London: Allen & Unwin. Weber, E., 1979. Peasants into Frenchmen: The Modernization of Rural France 1870– 1914. London: Chatto & Windus. Weber, M., 1978. Economy and Society: An Outline of Interpretative Sociology, 2 vols. Berkeley: University of California Press (1922. Wirtschaft und Gesellschaft. Tu¨bingen: Mohr). White, J. H., 1981. Catholics in Western Democracies: A Study in Political Behaviour. Dublin: Gill and Macmillan. Willis, K., 1977. The Introduction and Critical Reception of Marxist Thought in Britain, 1850–1900. Historical Journal 20: 417–419. Winch, D., 1966. The Keynesian Revolution in Sweden. Journal of Political Economy 74: 168–176. Windmuller, J. P., 1969. Labor Relations in the Netherlands. Ithaca, NY: Cornell University Press. Windmuller, J. P., and A. Gladstone, eds., 1984. Employers Associations and Industrial Relations. Oxford: Clarendon Press. Witte, E., 1980. De evolutie von de rol der parteijen in het Belgische parlementaire regeringssystem. Res Publica 22: 1–22. Woldendorp, J., H. Keman, and I. Budge, 1993. Political Data 1945–1990: Party Government in 20 Democracies. European Journal of Political Research 24, no. 1, special issue. Wolinetz, S. B., 1977. The Dutch Labour Party: A Social Democratic Party in Transition. In Paterson and Thomas (1977: 342–388) Wood, N., 1959. The Empirical Proletarians: A Note on British Communism. Political Science Quarterly 74: 256–272. World Handbook of Political and Social Indicators, 1964, 1972, 1983. New Haven: Yale University Press. Worlunberg, I., 1990. Kampen om det ideologiska rummet. Umea, Sweden: Umea Universitet. Worre, T., 1979. Forandringer i det danske partisystem sociale grundlad. In Pedersen (1979: 68–82). Wright, G., 1952. Four Red Villages in France. Yale Review 41: 361–372. 1955. Peasant Politics in the Third French Republic. Political Science Quarterly 70: 75–86. 1964. Rural Revolution in France: The Peasantry in the 20th Century. Stanford: Stanford University Press.

REFERENCES

631

Wuorinen, J. H., 1931. Nationalism in Modern Finland. New York: Columbia University Press. Zagoria, D. S., 1973. The Ecology of Peasant Communism in India. Monthly Public Opinion Survey no. 212: 9–16. Zeldin, T., 1979. France: 1848–1945, 2 vols. Oxford: Oxford University Press. Zuckermann, A., 1975. Political Cleavages: A Conceptual and Theoretical Analysis. British Journal of Political Science 5: 231–248. 1982. New Approaches to Political Cleavages. Comparative Political Studies 15: 131–144. Zwahr, H., 1987. Die deutsche Arbeiterbewegung im La¨nder- und Territorienvergleich 1875. Geschichte und Gesellschaft 13: 448–507.

INDEX

Adler, Viktor, 99, 252, 545 Agriculture, question, 466–9 —structure, 156–7, 469–76, 512 —peasantry, 469–75 —productivity, 472–3 —laborers, 475–6 —movements, 476–9, 483–6 Finland, 481–2; France, 483–6; Scandinavia, 479–81 Aksel, Larsen, 103 Allemane, Jean, 252 Alliances, political, 412–5 Andler, Charles, 77 Asquith, Herbert Henry, 421 Axel, Danielsson, 480 Bakunin, Michail, 78–9 Balbo, Cesare, 429 Bauer, Otto, 80–1, 100 Babel, August, 254 Bernstein, Eduard, 73–4, 89, 106, 123, 468 Bismarck, Otto, 257, 388, 532 Bissolati, Leonida,106 Blanc, Louis, 75, 252 Blanqui, Louis-Auguste, 75 Blum, Leon, 76–7, 340, 360, 391, 410 Bonomi, Ivanoe, 106 Bordiga, Amadeo, 106, 108, 358 Boulanger, Georges-Ernest-Jean-Marie, 331, 340, 459 Bourgeois, Le´on, 435

Branting, Hjalmar, 480 Brousse, Paul, 76, 252 Bru¨ning, Henrich, 364 Cabet, Etienne, 75 Cabinet, 358 —access to, 359–64, 380–3 —duration in, 364–7 —ideological completion, 372–5 —format and status, 367–71, 375–6 Scandinavia, 370–1; United Kingdom, 371–2 Cachin, Marcel, 516 Cattaneo, Carlo, 429 Cavour, Camillo Benso, 432 Chautemps, Camille, 360 Churchill, Winston, 364 Cleavage, 15–25 —concept, 15–6, —dimensions, 16–8 —Rokkan’s theory, 14–5 Cleavage, class, 20–2 —agrarian movements, 494–6 —denominational movements, 488–94 —liberalism, 486–8 —organizational density, 303–7 —cultural distinctiveness, 27 —distinctiveness / inclusiveness, 496– 501, 590–3 —structuring, 18–25, 411–2, 546–8 economic development, 553–4 democratization, 554–558 633

634 Cleavage, class (cont.) nation building, 551–3 organizational network, 27 social constituency, 27 state consolidation, 548–51 mobilization, 32–4, 38–41 Austria, 436–8; Belgium, 423–5; Denmark, 418–9; France, 428–31, 434– 5; Finland, 442–4, 447–50, 453–4; Germany, 438–40; Ireland, 444–7, 450–3; Italy, 431–5; Netherlands, 425–6; Norway, 417–8; Sweden, 419– 20; Switzerland, 427–8; United Kingdom, 416–7, 422 Cleavage, religious, 454–7 —ideological clericalism, 458 Belgium,458–9; France, 459–61; Italy, 460–1 —ghetto church, 462 Germany, 462–3; Ireland, 465–6; Netherlands, 463–4; Switzerland, 464–5 —monopoly church, 456–7 Austria, 457–8 Cleavage, rural —Austria, 436–8; Belgium, 423–5; Denmark, 418–9; France, 428–31, 434– 5; Finland, 442–4, 447–50, 453–4; Germany, 438–40; Ireland, 444–7, 450–3; Italy, 431–5; Netherlands, 425–6; Norway, 417–8; Sweden, 419– 20; Switzerland, 427–8; United Kingdom, 416–7, 422 Clemenceau, Georges, 435 Combes, Emile, 435 Communism, 502–8 —economic development, 502–10 —industrialization, 510–2 —rate of economic growth, 513–5 —rural structure, 516–22 —socialist organizations, 522–4, 526–8 —trade unions, 525–6, 528 —institutional openness, 532–3 —socialism political isolation, 536–7 —vote, 528–32, 534–5, 537–45 Connolly, James, 255 Cosgrave, William T., 367 Costa, Andrea, 254 Crispi, Francesco, 332 De Gasperi, Alcide, 367 De Gaulle, Charles, 358, 367 Delacroix, Le´on, 367

INDEX Disraeli, Benjamin, 532 Dovitt, Michael, 450 Engels, Friedrich, 72, 76–7, 80, 123, 484 Ferrari, Giuseppe, 429 Ferry, Jules, 331 Fleurant, Gabriel, 483 Fourier, Franc¸ois-Marie-Charles, 75 Franchise, 206–9, 582–7 —reversals, 220–1 —timing, 209–15 —tempo, 215–20 —and vote, 225–34, 236–9, 309–11, 404– 5 Frossard, Ludovic Oscar, 516 Gioberti, Vincenzo, 429 Giolitti, Giovanni, 97, 332 Gramsci, Antonio, 97, 106, 108 Guesde, Jules, 75–6, 107 Guizot, Franc¸ois, 9 Herr, Lucien, 76–77 Hitler, Adolf, 361, 364, 391, 503 Heterogeneity —cultural, 180–4, 582 —linguistic, 184–7 —religious, 184–7 and vote, 187–92, 199–200, 200–5, 235–9, 309–11, 404–5 Illiteracy, 192–7, 582 —vote, 197–9 Industrialization, 130–1, 169–70, 575–9 —rate of growth, 159–62, 579–81 and vote, 162–3 —sector occupation, 132–7 and vote, 141 —status occupation, 141–2 and vote, 143–5 —working-class, 122–30, 145–8 and vote, 148–51 —working-class concentration, 158 and vote, 159–60 —working-class homogeneity, 151–3 and vote, 153–7 Interest differentiation, 393–7, 440–1 Institutions political —and ideological orientation, 405–10. —development, 397–9 —openness/closure, 399–405 and vote, 400–5

INDEX International —First, 79, 252, 358 —Second, 21, 71, 77, 83, 100, 103, 107, 123, 248–9 —Third, 88, 92, 100, 102–8,115, 260, 357, 402, 516 —Work Community of the Socialist Parties, 100 —Red Trade Unions, 105 Jacquemotte, Jacque, 103 Jaure`s, Jean,76–8, 252, 435, 484 Kautsky, Karl, 7–4, 82, 86, 89, 105 Kuusinen, O., 109 Kuyper, Abraham, 425

635 Marx, Karl, 19, 20, 72, 76–77, 79–80, 84, 123, 178, 252, 410, 467, 468–9 Mauroy, Pierre, 363 Mazzini, Giuseppe, 78, 427, 429 Methodology, 41–3, 44- 8, 48–53, Millerand, Alexandre, 107, 252, 360, 435 Minghetti, Marco, 432 Mobilization, 11–2, 261–3 —political, 12–4, 241–4 —social, 167–72, 200–5, 309–11 and vote, 171, 173–9, 235–9, 404–5 Mondolfo, Rodolfo, 79 Mussolini, Benito, 97, 347, 348 Napoleon III, 459

Labriola, Arturo, 79 Lassalle, Ferdinand, 72, 84, 254 Lazzari, Costantino, 92 Left —cohesion, 97–9, 566–9 —composition, 111–9, 573–4 —deﬁnition, 9–11 —electoral development, 58–66, 109–19 —electoral size, 54–7, 558–62 —ideology, 66–70, 405–10, 562–66 orthodox Marxism, 407–9 Austria, 80–1; Belgium, 84; Denmark 84; Finland, 84–5; Germany, 72–4; Norway, 84; Netherlands, 82– 3; Sweden, 83; Switzerland, 85 socialist eclecticism, 407–9 France, 74–8; Italy, 78–9 trade unionism, 407–9 Ireland, 78; United Kingdom, 71–2 —organizational consolidation and vote, 291–7, 307–9, 404–5 —organizational density and vote, 297– 302, 307–9 —partisan density and vote, 294–5 —union density and vote, 292–3 Liebknecht, Karl, 106 Lenin, Nicolay, 91, 100, 107, 123, 516 Lloyd George, David, 421 Lueger, Karl, 457 Luis Philippe, 9 Luxemburg Rosa, 74, 89, 106, 358 MacDonald, Ramsey, 359, 365, 371 Macmahon, Marie-Edme-Patrice-Maurice, 331, 339 Mann, Tom, 89

Palm, August, 252 Parnell, Charles, 329 Parri, Ferruccio, 367 Parties —communist coalition potential, 385–7, 590 governmental power, 385–7 membership, 268–70, 588–9 Austria, 99–100, 278, 534, 544–5; Belgium, 103, 363, 516; Denmark 102–3, 115; Finland, 96, 108–9, 268–70, 300, 363, 385–6, 482, 506; France, 107–8, 114, 268–9, 356–7, 363, 386, 483, 485, 516–8; Germany, 95, 99, 105–6, 115, 118–9, 120, 269, 516, 521; Ireland, 101, 110; Italy, 106–7, 118, 269, 270, 357, 363, 385–6, 483, 516; Netherlands, 83, 101, 516; Norway, 100, 105, 116; Sweden, 101–2,115, 389, 428; Switzerland, 103–5, 117; United Kingdom, 100–1, 110 —membership, 263–5, 271–7 —socialist and unions, 256–61 coalition potential, 378–80, 383–7, 590 foundation, 244–61, 397–9 and franchise, 250–1 governmental power, 383–8 France, 391; Scandinavia, 388–9; United Kingdom, 390 membership, 265–8, 588–9 Austria, 80–1, 99–100, 229, 253, 256, 260, 267, 278, 302, 334, 359,

636 Parties (cont.) 381, 384, 437, 478, 544–5; Belgium, 85–6, 103, 116–7, 252, 267, 352, 359, 369, 373, 382, 428; Denmark, 84, 102–3, 115, 251, 267, 270, 327–8, 366, 371, 383, 419, 480–1; Finland, 84–5, 108–9, 116, 252, 257, 261, 270, 300, 334–5, 359, 363, 369, 374, 444, 449, 481; France, 76, 90–1, 107–8, 114, 252– 3, 259, 267, 330–1, 355, 357, 360, 365, 369, 374, 385, 391, 410, 435, 484, 516–7, 524; Germany, 60, 72– 4, 76–7, 80, 82, 89–90, 100, 105–6, 115, 254, 256–7, 268, 272–3, 303, 334, 359, 438, 473, 477, 542; Ireland, 101, 117, 229, 255, 257–8, 329, 368, 384–5, 448, 451–3; Italy, 91, 106, 114, 254, 259–60, 332–3, 357, 362, 384–5, 434, 460, 516; Netherlands, 70, 82–3, 101, 114–5, 253, 274; Norway, 28, 105, 116, 231, 252, 256, 265, 267, 269–70, 324, 356, 359, 365, 371, 417–9, 481, 508, 516; Sweden, 83, 101–2, 115, 145, 252, 265, 267, 326–7, 370, 383, 389, 480–1; Switzerland, 103–4, 117, 253–4, 267, 324, 361, 428; United Kingdom, 71–2, 100–1, 111–2, 145, 255, 258, 265, 268, 272, 308, 325, 356, 359, 365, 371– 2, 377, 384, 390, 416, 420–2, 435, 472, 525 Pio, Luis, 327 Pisacane, Carlo, 429 Proudhon, Pierre-Joseph, 75, 252 Radicalization —and regime change, 95–7 —ideological, 87–92 —of class action, 93–4 —Austria, 96–7; Finland 95–6; Germany, 95; Italy, 97 Rattazzi, Urbano, 432 Renner, Karl, 100 Repression, 391–7 —censorship, 320–3 —association, 323 Austria, 333–4; Belgium, 328; Denmark, 327–8; Finland, 334–5; France, 330–1; Germany, 333–4; Ireland, 329; Italy, 332–3; Netherlands,

INDEX 328–9; Norway, 323–4; Sweden, 326– 7; Switzerland, 324; United Kingdom, 325–6 Responsible government, 335–7, 394–5, 398–9 —Austria, 346; Belgium, 338–9; Denmark , 341–2: Finland, 343–4; France, 339–41; Germany, 344–6; Italy, 346–8; Netherlands, 338–9; Sweden, 341–2; Switzerland, 338; United Kingdom, 339 Ricasoli, Bettino, 432 Rigola, Rinaldo, 259 Rochefort, Henry, 252 Rudinı`, Antonio, 347 Saint-Simon, Henri-Claude, 75 Sangnier, Marc, 459, 461 Schumacher, Kurt, 268 Sella, Quintino, 432 Sembat, Marcel, 516 Serrati, Giacinto Menotti, 106 Sonnino, Sidney, 347 Sorel, George, 79, 86 Spaak, P. H., 373 Stalin, Iosif Vissarionovich, 104 Stateness, 313–20 —and repression, 393–7 Stauning, Thorvald, 102 Tanner, Vaino, 109, 344, 359 Thiers, Adolphe, 331 Turati, Filippo, 106 Treves, Claudio, 106 Turnout, 221–5 Unions —foundation, 244–61 —membership, 277–83, 588–9 —density, 284, 289–90 —collective bargaining, 285–90 —Austria, 260, 278, 334; Belgium, 258, 328; Denmark, 327–8; Finland, 96, 108, 257, 261, 334–5; France, 72, 79, 107, 259, 290, 330–1, 391, 435, 469, 484; Germany, 254, 257, 291, 333–4; Ireland, 78, 255, 257, 329; Italy, 97, 259–60, 332–3; Netherlands, 260, 328–9; Norway, 281, 324; Sweden, 83, 326–7; Switzerland, 253, 261, 324; United Kingdom, 78, 89, 101, 255, 325–6

INDEX

637

Urbanization, 163–5 —and vote, 166–7, 169–72

—open, 348–50 —plural, 351–3

Vaillant, Edouard, 252 Voting —curia/estate, 350–1 —indirect, 350 —misrepresentation, 353–8

Waldeck-Rousseau, Pierre Marie, 107, 331, 360, 435 Wright von, Victor, 251 Zinoviev, Gregorij Evseevic, 100

Continuation of series list

Thomas Janoski and Alexander M. Hicks, eds., The Comparative Political Economy of the Welfare State Robert O. Keohane and Helen B. Milner, eds., Internationalization and Domestic Politics Herbert Kitschelt, The Transformation of European Social Democracy Herbert Kitschelt, Peter Lange, Gary Marks, and John D. Stephens, eds., Continuity and Change in Contemporary Capitalism Herbert Kitschelt, Zdenka Mansfeldova, Radek Markowski, and Ga´bor To´ka, PostCommunist Party Systems David Knoke, Franz Urban Pappi, Jeffrey Broadbent, and Yutaka Tsujinaka, eds., Comparing Policy Networks Allan Kornberg and Harold D. Clarke, Citizens and Community: Political Support in a Representative Democracy David D. Laitin, Language Repertories and State Construction in Africa Mark Irving Lichbach and Alan S. Zuckerman, eds., Comparative Politics: Rationality, Culture, and Structure Doug McAdam, John McCarthy, and Mayer Zald, eds., Comparative Perspectives on Social Movements Scott Mainwaring and Matthew Soberg Shugart, eds., Presidentialism and Democracy in Latin America Anthony W. Marx, Making Race, Making Nations: A Comparison of South Africa, the United States, and Brazil Joel S. Migdal, Atul Kohli, and Vivienne Shue, eds., State Power and Social Forces: Domination and Transformation in the Third World Ton Notermans, Money, Markets, and the State: Social Democratic Economic Policies Since 1918 Paul Pierson, Dismantling the Welfare State? Reagan, Thatcher, and the Politics of Retrenchment Marino Regini, Uncertain Boundaries: The Social and Political Construction of European Economies Yossi Shain and Juan Linz, eds., Interim Governments and Democratic Transitions Theda Skocpol, Social Revolutions in the Modern World David Stark and La´szlo´ Bruszt, Postsocialist Pathways: Transforming Politics and Property in East Central Europe

Sven Steinmo, Kathleen Thelan, and Frank Longstreth, eds., Structuring Politics: Historical Institutionalism in Comparative Analysis Sidney Tarrow, Power in Movement: Social Movements and Contentious Politics Ashutosh Varshney, Democracy, Development, and the Countryside

The Political Mobilization of the European Left, 1860-1980: The Class Cleavage (Cambridge Studies in Comparative Politics)

Nationalist Mobilization and the Collapse of the Soviet State (Cambridge Studies in Comparative Politics)

Nationalist Mobilization and the Collapse of the Soviet State (Cambridge Studies in Comparative Politics)

The End of Class Politics?: Class Voting in Comparative Context

The End of Class Politics?: Class Voting in Comparative Context

The European Voter (Comparative Politics)

Procedural Politics: Issues, Influence, and Institutional Choice in the European Union (Cambridge Studies in Comparative Politics)

Procedural Politics: Issues, Influence, and Institutional Choice in the European Union (Cambridge Studies in Comparative Politics)

The Puzzle of Strikes: Class and State Strategies in Postwar Italy (Cambridge Studies in Comparative Politics)

The Puzzle of Strikes: Class and State Strategies in Postwar Italy (Cambridge Studies in Comparative Politics)

Politics in the Pews: The Political Mobilization of Black Churches (The Politics of Race and Ethnicity)

The Mobilization of Political Islam in Turkey

Legislative Politics in Latin America (Cambridge Studies in Comparative Politics)

Legislative Politics in Latin America (Cambridge Studies in Comparative Politics)

Foundations of Comparative Politics (Cambridge Textbooks in Comparative Politics)

Foundations of Comparative Politics (Cambridge Textbooks in Comparative Politics)

Foundations of Comparative Politics (Cambridge Textbooks in Comparative Politics)

The Transformation of European Social Democracy (Cambridge Studies in Comparative Politics)

The Transformation of European Social Democracy (Cambridge Studies in Comparative Politics)

Political Topographies of the African State: Territorial Authority and Institutional Choice (Cambridge Studies in Comparative Politics)

Political Topographies of the African State: Territorial Authority and Institutional Choice (Cambridge Studies in Comparative Politics)

The Architecture of Government: Rethinking Political Decentralization (Cambridge Studies in Comparative Politics)

Comparative European Politics

Presidential Impeachment and the New Political Instability in Latin America (Cambridge Studies in Comparative Politics)

Presidential Impeachment and the New Political Instability in Latin America (Cambridge Studies in Comparative Politics)

Presidential Impeachment and the New Political Instability in Latin America (Cambridge Studies in Comparative Politics)

Presidential Impeachment and the New Political Instability in Latin America (Cambridge Studies in Comparative Politics)

Internationalization and Domestic Politics (Cambridge Studies in Comparative Politics)

Internationalization and Domestic Politics (Cambridge Studies in Comparative Politics)

Encyclopedia of Politics. The Left

Democracy Derailed in Russia: The Failure of Open Politics (Cambridge Studies in Comparative Politics)

The Political Mobilization of the European Left, 1860-1980: The Class Cleavage (Cambridge Studies in Comparative Politics)

Nationalist Mobilization and the Collapse of the Soviet State (Cambridge Studies in Comparative Politics)

Nationalist Mobilization and the Collapse of the Soviet State (Cambridge Studies in Comparative Politics)

The End of Class Politics?: Class Voting in Comparative Context

The End of Class Politics?: Class Voting in Comparative Context

The European Voter (Comparative Politics)

Procedural Politics: Issues, Influence, and Institutional Choice in the European Union (Cambridge Studies in Comparative Politics)

Procedural Politics: Issues, Influence, and Institutional Choice in the European Union (Cambridge Studies in Comparative Politics)

The Puzzle of Strikes: Class and State Strategies in Postwar Italy (Cambridge Studies in Comparative Politics)

The Puzzle of Strikes: Class and State Strategies in Postwar Italy (Cambridge Studies in Comparative Politics)

Politics in the Pews: The Political Mobilization of Black Churches (The Politics of Race and Ethnicity)

The Mobilization of Political Islam in Turkey

Legislative Politics in Latin America (Cambridge Studies in Comparative Politics)

Legislative Politics in Latin America (Cambridge Studies in Comparative Politics)

Foundations of Comparative Politics (Cambridge Textbooks in Comparative Politics)

Foundations of Comparative Politics (Cambridge Textbooks in Comparative Politics)

Foundations of Comparative Politics (Cambridge Textbooks in Comparative Politics)

The Transformation of European Social Democracy (Cambridge Studies in Comparative Politics)

The Transformation of European Social Democracy (Cambridge Studies in Comparative Politics)

Political Topographies of the African State: Territorial Authority and Institutional Choice (Cambridge Studies in Comparative Politics)

Political Topographies of the African State: Territorial Authority and Institutional Choice (Cambridge Studies in Comparative Politics)

The Architecture of Government: Rethinking Political Decentralization (Cambridge Studies in Comparative Politics)

Comparative European Politics

Presidential Impeachment and the New Political Instability in Latin America (Cambridge Studies in Comparative Politics)

Presidential Impeachment and the New Political Instability in Latin America (Cambridge Studies in Comparative Politics)

Presidential Impeachment and the New Political Instability in Latin America (Cambridge Studies in Comparative Politics)

Presidential Impeachment and the New Political Instability in Latin America (Cambridge Studies in Comparative Politics)

Internationalization and Domestic Politics (Cambridge Studies in Comparative Politics)

Internationalization and Domestic Politics (Cambridge Studies in Comparative Politics)

Encyclopedia of Politics. The Left

Democracy Derailed in Russia: The Failure of Open Politics (Cambridge Studies in Comparative Politics)

Recommend Documents