Political Parties and Legislative Party Switching
This page intentionally left blank
Political Parties and Legislat...
72 downloads
1307 Views
2MB 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
Political Parties and Legislative Party Switching
This page intentionally left blank
Political Parties and Legislative Party Switching Edited by William B. Heller and Carol Mershon
POLITICAL PARTIES AND LEGISLATIVE PARTY SWITCHING
Copyright © William B. Heller and Carol Mershon, 2009. All rights reserved. First published in 2009 by PALGRAVE MACMILLAN® in the United States—a division of St. Martin’s Press LLC, 175 Fifth Avenue, New York, NY 10010. Where this book is distributed in the UK, Europe and the rest of the world, this is by Palgrave Macmillan, a division of Macmillan Publishers Limited, registered in England, company number 785998, of Houndmills, Basingstoke, Hampshire RG21 6XS. Palgrave Macmillan is the global academic imprint of the above companies and has companies and representatives throughout the world. Palgrave® and Macmillan® are registered trademarks in the United States, the United Kingdom, Europe and other countries. ISBN: 978–0–230–60755–2 Library of Congress Cataloging-in-Publication Data is available from the Library of Congress. A catalogue record of the book is available from the British Library. Design by Newgen Imaging Systems (P) Ltd., Chennai, India. First edition: July 2009 10 9 8 7 6 5 4 3 2 1 Printed in the United States of America.
Contents
List of Figures and Tables
vii
Preface
xi
Part I 1
2
The Importance of Party Switching
Introduction: Legislative Party Switching, Parties, and Party Systems William B. Heller and Carol Mershon Integrating Theoretical and Empirical Models of Party Switching William B. Heller and Carol Mershon
3
29
Part II Party Switching and Representation 3
Switching Equilibria Norman Schof ield
4
Party Switching and the Procedural Party Agenda in the US House of Representatives Timothy P. Nokken
5
Party Switching in Brazil: Causes, Effects, and Representation Scott Desposato
55
81
109
Part III Party Switching, Party Competition, and Policy Making 6 Party Group Switching in the European Parliament Gail McElroy and Kenneth Benoit 7
Legislator Preferences, Party Desires: The Impact of Party Switching on Legislative Party Positions William B. Heller and Carol Mershon
147
173
vi 8
Contents Timing Matters: Incentives for Party Switching and Stages of Parliamentary Cycles Carol Mershon and Olga Shvetsova
201
Part IV Party Switching and the Dynamics of Party Systems 9
Competition for Power: Party Switching and Party System Change in Japan Junko Kato and Kentaro Yamamoto
233
10 Party Switching, Party Systems, and Political Representation Marcus Kreuzer and Vello Pettai
265
11
287
Conclusions William B. Heller and Carol Mershon
List of Contributors
295
Index
299
Figures and Tables
Figures 1.1 2.1 3.1 3.2 3.3 3.4 3.5 4.1 4.2 4.3 4.4 4.5 5.1 5.2 5.3 5.4 5.5 5.6 6.1
The legislative context for party switching The four stages of the switching game Activist coalitions in the United States Party positions in the Israeli Knesset 1996 A local Nash equilibrium in the Knesset in 1996 Party positions in the Knesset in 2003 Party positions in the Knesset in 2006 Total roll calls by vote types, 83rd to 107th Congresses Party polarization, 1879 to 2006, distance between parties on 1st DW-NOMINATE dimension DW-NOMINATE scores for 96th to 107th Congresses on final passage votes: 104th Congress party switchers DW-NOMINATE scores for 96th to 107th Congresses on Amendment votes: 104th Congress party switchers DW-NOMINATE scores for 96th to 107th Congresses on procedural votes: 104th Congress party switchers Party membership equilibria strategies—public goods without party control Party membership equilibria strategies—private goods without party control Party membership equilibria strategies—public goods with party control Party membership equilibria strategies—private goods with party control Party switching by month in the 49th and 50th Brazilian Chamber of Deputies Ideology and switching in the 50th Brazilian legislature European party groups and national members on the general left-right scale
15 39 59 67 68 70 71 91 96 96 97 97 115 116 117 118 128 139 159
viii 6.2 7.1 8.1 8.2 9.1 9.2 9.3 9.4 9.5 9.6 9.7 10.1 10.2 10.3 10.4
Figures and Tables Relationship of first factor to independent measures of left-right positions Parties with overlapping membership Switching behavior during the parliamentary cycle Mean weekly switches per 100 MPs (All and SMD), by substage in term, Russia and Italy Changes in distribution of seats among parties in Japanese house of representatives, 1993–2005 Changes in Shapley–Shubik index for Japanese parties, 1993–2005 Expected office payoffs to legislators ([SSI*100]/seats), by party, Japan 1993–2005 Comparison of party policy positions in Japan: The 1996 expert survey Comparison of party policy positions in Japan: The 2000 expert survey Comparison of party policy positions in Japan: The 2003 expert survey Comparison of party policy positions in Japan: The 2005 expert survey Types of party switching Levels of party system institutionalization Types of party Effects of hopping
162 175 203 212 243 245 246 255 255 256 256 268 269 277 280
Tables 1.1 1.2 1.3 3.1 4.1 4.2 4.3 5.1
The definition and measurement of switches and types of switches 9 Illustrations of the incidence of switching (% MPs ever switched) in established and new democracies 11 The incidence and impact of switching 13 Seats, votes, and valences in the Knesset 67 House defectors and change in aggregate DW-NOMINATE scores, 1953–2004 90 House defectors and the statistical significance of changes in aggregate voting, 1953–2002 93 House party defectors’ change in DW-NOMINATE scores, 1953–2002 99 Party switching rates in the Brazilian Chamber of Deputies 110
Figures and Tables
ix
5.2 Typologies of legislative party systems 5.3 Payoffs of party membership 5.4 Conditional logit model of affiliation decisions 5.5 Impact of selected party characteristics on predicted probability of membership 5.6 Impact of transaction costs on probability of switching 5.7 Party agreement scores 5.8 Switching and party agreement scores 5.9 Party influence measures—Dimension 1 5.10 Party influence measures—Dimension 2 6.1 Political party groups in the European Parliament pre-2004 election 6.2 Party group volatility in the European Parliament 6.3 Dimension reduction of policy contestation 6.4 Conditional logit model of party choice 6.5 Conditional logit model of individual MEP switching, third Parliament 7.1 Expectations: Party positions as a function of party-member positions 7.2 Party positions as a function of party-member positions 7.3 Changes in the size of party groups, X and XI Legislatures, Italy 7.4 Changes in the size of party groups, XII and XIII Legislatures, Italy 8.1 Case selection: Countries and legislative terms 8.2 Operationalizing stages of the parliamentary cycle 8.3 Mean weekly switches per 100 MPs, Italy and Russia 8.4 Mean monthly switches per 100 MPs, 35 country-terms 8.5 Switching to create new parliamentary parties, 37 country-terms 9.1 Illustrations of the decisive majority 9.2 Switchers from the NFP to the DPJ and the LDP between 1996 and 2000 9.3 Switchers from the LP to the DPJ between 2000 and 2003 9.4 DPJ legislators in October 2003: Inswitchers and loyalists
111 115 129 130 130 134 135 138 138 150 153 161 165 167 186 187 192 192 206 208 210 217 221 236 252 253 258
This page intentionally left blank
Preface
Political parties and democratic politics go hand in hand. Since parties matter, it matters too when legislators and other elected politicians change party affiliation. The phenomenon of party switching becomes puzzling when the switch occurs during a politician’s term of office. Why would a sitting legislator who won his seat under one party label jump to another during a legislative term? What are the effects of such switches? This book investigates these and related questions. Scholarship arises from curiosity, advances through some form of collaboration, and typically requires financial support as well. In the late 1990s, the coeditors of this book first discussed the curious frequency of legislative party switching in Italy, when we met for dinner at one of the annual meetings of the American Political Science Association. In our collaboration, one thing has led to another, including this book. We first coauthored several conference papers on the phenomenon of party switching, and then, in 2003, we secured support from the National Science Foundation for collaborative research (NSF SES-0339920 to Heller and SES-0339877 to Mershon, for “Collaborative Research on Legislative Party Switching: Integrating Theoretical and Comparative Empirical Analyses”). This book represents the culmination of that project and of the joint work of the 12-member Research Work Group on Legislative Party Switching, which we recruited and co-chaired. We and the other contributors to the volume are convinced that the study of elected politicians’ changes and choices of party affiliation can enrich the understanding of political parties, legislative politics, policy making, and democratic representation. The first two chapters and the concluding chapter perhaps emphasize this conviction most strongly. All chapters, however, seek to demonstrate the analytical purchase to be gained from considering party switching as one of the
xii
Preface
possible results of the strategic interactions between legislators and party leaders. All share as a premise and as a theme the notion that taking party switching seriously sheds new light on issues of enduring importance in political science. The contributors to this volume investigate a wide range of empirical settings, from Western Europe to the Baltics, from the United States to Japan, from Russia and Romania to Brazil. At the same time, the research integrates formal and empirical approaches to the study of politicians’ decisions on party affiliation. To identify this book as the culmination of our work is to point to prior stages of collaboration. The Party Switching Research Group (PSRG) held its first meeting at Trinity College, Dublin, Ireland, July 2004, and its second meeting at the University of Virginia, Charlottesville, Virginia, July 2005. The group also organized panels, featuring papers from several of the contributors here, at the 2005 Annual Meeting of the International Studies Association and the 2005 Annual Meeting of the Midwest Political Science Association. Both meetings of the PSRG qualified as true workshops, with wide-ranging conversations, informal presentations, and active debate and discussion among all participants. We welcomed and profited from the participation of several political scientists who were not strictly speaking members of the PSRG, including graduate students and, in Dublin, Daniela Giannetti from the faculty of the Università di Bologna. This book and the research that underpins it benefited greatly from the contributions of several key players. The first is Michael Laver, who as a member of the PSRG prepared papers for Dublin and Charlottesville. As then-chair of the Department of Political Science at Trinity College, moreover, Laver also was a generous and gracious host for our Dublin meetings. As it turned out, Laver did not write a chapter for the book, but his voice in the conversations and collaboration remains. We also acknowledge the very helpful comments of John Aldrich, offered at several stages in this project, and the thoughtful remarks of Dave Clark, Daniela Giannetti, Dan Gingerich, Mikhail Filippov, Michael McDonald, Lucio Renno, Keith Poole, and Katri Sieberg. We extend our deep thanks to all of these individuals. We are grateful in addition to a number of research assistants who in one way or another have made it possible for us to pursue research and writing on legislative party switching: Nina Barzachka, Robin Best, B.J. Bloom, Susan Brewer, Adriana Buliga-Stoian, Daniela Coleman, Lindsay Flynn, Weiwei Hu, Miriam Hurley, Rado Iliev, Drew Kurlowski, Sam Seeley, Jon Shoup-Mendizabal, and Julie VanDusky. All have facilitated our work and have earned our thanks.
Preface
xiii
We have been fortunate to work with Emily Hue, Asa Johnson, Farideh Koohi-Kamali, and Toby Wahl at Palgrave Macmillan. We greatly appreciate their expertise and advice in shepherding the book through to publication. We dedicate this book to our families, for their love, joy, and ability to pull us away from our work. Will thanks Tere, Norma, and Andrea for their patience and willingness to put up with the various travels (and other travails) involved in bringing this book to fruition. It has been a learning experience all around, and their sacrifices and support are deeply appreciated. They may never read it, but they are present on every page. Carol dedicates the book to Lyle and Harriette, whose example and inspiration remain. Their presence too echoes throughout the book.
This page intentionally left blank
I
The Importance of Party Switching
This page intentionally left blank
1 Introduction: Legislative Party Switching, Parties, and Party Systems William B. Heller and Carol Mershon
1.1 Political Parties and the Question of Party Switching That political parties are fundamental to the functioning of modern democracies is well known. Politicians build their careers within parties, parties convey information to voters about candidate preferences, and parties provide labels that identify candidates to voters. When voters choose candidates for office, they delegate decision making on public policy to parties and to party-identified representatives. Repeated elections give voters the opportunity to hold parties responsible and accountable for policy decisions and outcomes. Parties thus are indispensable elements of democratic delegation and representation (Cox 1997; Filippov, Ordeshook, and Shvetsova 2004; Powell 2000; Schattschneider 1942; Stokes 1999). Honing in on the legislative arena highlights the vital role of parties. In the United States, the majority party in each chamber organizes that chamber, filling all legislative offices and managing the flow of legislation (Cox and McCubbins 1993; 2005). Parties manage the business of the legislature in parliamentary systems too, albeit generally through their control of the executive (Cox 1987). This control, along with the tight discipline commonly ascribed to parliamentary parties (Bowler, Farrell, and Katz 1999), means that majority parties (or coalitions) in parliamentary systems not only manage the legislative process, but also determine legislative outcomes (see, e.g., Baron 1998; Heller 2001; Huber 1996; Laver and Schofield 1990; Martin and Vanberg 2004; 2005). Given the essential place of parties, it seems reasonable to expect legislators to stick to the party labels under which they have won election.
4
William B. Heller and Carol Mershon
In many democracies, much of the time, this expectation is upheld, as is generally the case in the established parliamentary systems of Western Europe (Bowler, Farrell, and Katz 1999). In some settings, however, it is not uncommon for elected legislators to abandon one party and enter another, even during the legislative term. For instance, approximately one-fourth of the members of the Italian lower house switched parties at least once during the 1996–2001 legislature (Heller and Mershon 2005; 2008), and more than one-third of the Brazilian MPs elected in 1986 had transferred from one party to another by late 1990 (Mainwaring and Pérez Liñán 1997). Observers of politics have recorded legislators hopping parties relatively frequently in such diverse contexts as Fourth Republic France (MacRae 1967), Hungary and the Czech Republic (Àgh 1999), Russia between 1993 and 1995 (White, Rose, and McAllister 1997), Papua New Guinea and India (Miskin 2003), and the United States during what political scientists deem to be periods of realignment (Canon and Sousa 1992; for additional examples of abundant switching, see section 1.3 and Mershon and Shvetsova 2007). The null hypothesis—that politicians hold fast to the party affiliations they at first choose—evidently fails to hold uniformly. Compared to the null, even an annual average of a single switcher, sustained more than a century—the record for both chambers combined in Australia since federation (see Miskin 2003, 4)—and the two-score switches in the House and Senate since 1947 (Nokken 2000) command attention. Why do politicians switch parties, and to what effect? That overarching question motivates the research in this book. Despite its manifestation across countries and continents, party switching remains an understudied phenomenon. This is not to say that students of legislative and party politics have neglected it entirely, but rather that they have underestimated the breadth and depth of its significance. In line with the common sense intuition that legislators should stick with the parties that got them elected, scholars have tended to treat switching as anomalous (and undesirable) behavior symptomatic of some underlying system-wide condition, such as electoral realignment (Canon and Sousa 1992) or a weakly institutionalized party system (Mainwaring 1999; Mainwaring and Torcal 2005). Switching can be these things, but it also is much more. First, switching can occur even in the most stable of systems. Second, even when party switching (presumably) indicates some underlying disease of the body politic, its importance goes far beyond any utility it has as a diagnostic tool. Simply put, party switching has vital normative, theoretical, and substantive implications.
Introduction
5
The next section here elaborates on the implications of party switching. Section 1.3 provides a definition and typology of switching, used in common by all contributors to this volume, and also surveys the incidence of switching. The fourth section takes stock of extant research on switching and the place of switching within research on legislative and party politics. This introductory chapter closes with an overview of the other chapters that make up the book.
1.2 The Implications and Importance of Party Switching The normative concerns raised by party switching center on accountability, responsibility, and representation. At least at first blush, the link between voters and politicians is undermined if voters elect a politician as a candidate of one party and he or she jumps to another during the legislative term, without consulting his or her constituents at the polls. All the same, analysts should be wary of assuming or arguing that party switching impairs democratic representation. If representatives are in essence seat holders for their parties, if voters choose representatives based solely or at least primarily on their party affiliation, and if switchers adopt the priorities of their new parties, then switching indeed amounts to a betrayal of democratic representation. These are big “ifs,” however, and to the extent that they do not obtain, it is possible that switching might even improve representation, not undermine it. The validity of the presumption that legislators are seat holders for their parties depends on how voters choose and view their representatives, which in turn depends in part on electoral and legislative rules, as well as on the party system. If voters vote purely on the basis of party label, without reference to the identities of individual candidates, then sitting legislators are indeed agents of their parties and only indirectly of voters. Such party-focused voting seems most likely in systems with closed-list electoral rules (Carey and Shugart 1995), but even here there is room for individual candidates to entice some voters to choose one party over another, or at least to go to the polls rather than stay at home on election day (e.g., on Spain see Bruneau et al. 2001). In this context, party switching is a betrayal of democratic representation as long as party positions remain fixed or at least faithful to party voters: if parties drift, whether because their membership changes (e.g., as a result of elections, by-elections, or
6
William B. Heller and Carol Mershon
switching), their leadership changes, or for some other reason, then party members who want to remain faithful to the voters who elected them might have no recourse but to seek a new party home (Crewe and King 1995; Hopkin 1999; Mair and Marsh 2004; Miskin 2003).1 In chapter 7, we cast doubt on the notion that party positions are fixed when we show that party positions in legislative voting are determined in some measure by the preferences of party members, and changes in party membership occasion changes in party positions. Switchers thus cause movements in party positions that potentially could make parties—or at least policy making—more representative, rather than less so. The third assumption, that switchers adapt to their party positions, does seem to hold in at least some cases. It is borne out by evidence from the US House (Ansolabehere, Snyder, and Stewart 2001; McCarty, Poole, and Rosenthal 2001; and cf. Nokken this volume) and Brazil (Desposato this volume), where low levels of observed party unity make expectations of party influence over legislator behavior seem a priori unrealistic. Which is more important for outcomes, changes in party positions or changes in legislators’ behavior, is an empirical question that hinges in part on party discipline and in part on how other legislative players respond to changes in individual legislator or party positions (Best and Heller 2005). Finally, the significance of switching also might depend on the degree to which it affects outcomes. The normative implications of party switching, in short, must be disentangled with care. They are likely to depend greatly on the institutional context and on the array of individual and party preferences in the legislature, as well as the extent to which voters care about “descriptive” or “symbolic” representation alone (e.g., the simple presence of ethnic minorities in the legislature) versus the ability to influence political and policy outcomes. On the theoretical front, party switching provides new and powerful leverage on prominent, longstanding questions in the study of politics. Principal among these is the issue of “why parties” (Aldrich 1995; McElroy 2003, 21), which subsumes questions of what parties are, what they do, and how they influence (and are influenced by) individual legislators. Research on switching illuminates not only parties and party discipline but also the dynamics of party systems. Students of legislative politics typically—albeit usually implicitly— assume parties to be stable entities. Work on government coalitions in parliamentary democracies routinely rests on the same assumption.
Introduction
7
In probing the conditions under which this assumption fails to hold, examination of party switching can “lift the lid” on parties (Laver 1998, 22) and set out guidelines for linking analytically the microfoundations of parties to legislative organization and governance, and to interparty competition at and between elections. Party switching opens a new window onto party systems, enabling the analyst to see and treat partisan identities not as fixed and exogenous but as fluid and endogenous, to examine the processes whereby politicians choose party identities at one election, reevaluate them during the life of the legislature, and potentially revisit partisan identities again at the next election. Systematic thinking about party switching thus affords novel insight into the dynamics, evolution, and institutionalization of party systems (cf., e.g., Laver and Benoit 2003; Mainwaring and Torcal 2005). Substantively, party switching matters because it can at least potentially alter policy bargaining in the legislature and even government composition. Party switchers in Spain (see Tomás Mallén 2002) have brought down local governments, for instance, and Senator Jeffords’s defection from the Republican Party in 2001 changed the balance of power in the US Senate, giving the Democrats a degree of agenda authority that elections had denied them. Similarly, in Canada, one MP’s May 2005 defection from the opposition Conservative Party to join the minority Liberal cabinet helped the government survive a vote of no confidence (Economist 2005b; 2005a). Absent such obvious swings in the legislative balance, party switching still can affect the legislative bargaining context by changing parties’ seat shares and policy positions, with possible repercussions that include the allocation of ministerial portfolios among government parties (since a coalition party’s share of cabinet portfolios is roughly proportional to its contribution to the seats controlled by the government; Browne and Franklin 1973; Mershon 2001; 2002; Schofield and Laver 1985; cf. Mershon 2008), the allocation of legislative committee seats among parties (Yoshinaka 2005), parties’ ability to impose voting discipline on their members (Heller and Mershon 2008; see also, Best and Heller 2005), and, as a consequence of all of this, policy outcomes. The basic point is that party switching is neither as rare nor as idiosyncratic as conventional wisdom suggests. Indeed, focusing on party switching highlights the observed stability in parties and party systems as an equilibrium, the result of legislator and party leadership strategies developed partially in response to the possibility of switching. The contributors to this volume take seriously that possibility. To
8
William B. Heller and Carol Mershon
investigate it, of course, they need common terminology and analytical instruments, to which we now turn.
1.3
The Phenomenon of Party Switching
1.3.1
Definition and typology
All contributors to the book adopt the term “switch” as the umbrella label for any recorded change in party affiliation on the part of a politician holding or competing for elective office. As table 1.1 indicates, operationalizing this basic definition depends on the type of register of switches that is available. For instance, membership lists may be maintained at the level of the parliamentary group and made available relatively infrequently, as for the European Parliament (EP) and Russia. Or the national legislature may provide chronologies of each move executed, each day, as for Canada and Italy. Additional possibilities exist. The point is that the empirical analysis of switching hinges on the nature of the records at the disposal of the researcher. We identify different types of switches in four ways, as table 1.1 exhibits. First, switches may be distinguished by direction. In the United States, Senator Jeffords exited the Republicans in 2001, effecting an “outswitch,” without officially carrying out an “inswitch” into the Democrats. Even as an Independent, as noted, Jeffords handed majority control of the Senate to the Democrats. Operationalizing this distinction depends as well on party and legislative records. If inswitches and outswitches are not recorded separately, information on one must be inferred from information on the other. Second, switches may be classified by their impact on the number of legislative parties (or groups or factions or fractions). Both the raw number and the effective number of parties2 are useful, since they tap changes in the legislative bargaining context somewhat differently; such changes in context matter, in turn, because they might affect the calculations of potential switchers. Whereas switches across two existing parties leave intact the number of parties, the number of parties changes with other types—fission (i.e., the splintering of a single party into two or more distinct parties), fusion (where two or more parties merge into one), and start-up (where switchers from multiple parties join to create a new one).3 Third, switches may be distinguished by how isolated or close in time they are relative to other moves. When individual switches are isolated in time, it makes sense to consider them as independent from
Table 1.1
The definition and measurement of switches and types of switches
Label
Definition
Measure and/or comment
Umbrella label: Switch
Change in party affiliation
Measure is recorded change in affiliation; data depend on how (and how often) legislature or party records change
Formal adoption of new label by MP after having another label Abandonment of one label in favor of another
Data depend on whether inswitch and outswitch are recorded separately
1—Type by direction Inswitch
Outswitch
2—Type by impact on N parties Both the raw number and the effective number of parties hold interest Across existing parties Fission Fusion Start-up
Move leaves intact raw N parties Existing party splits to create two or more new Two or more extant parties merge to create one new Move founds new party with MPs from multiple parties
3—Type by relative timing Legislative (or party) records determine how fine-grained temporally data can be Solo Simultaneous Near-simultaneous
Move isolated from others in time Same day (week, etc.) as other moves Close in time to other moves
Coordination of moves (or, related but distinct, signaling) may or may not occur
4—Type by degree of (apparent) choice Rule-driven move
Move forced by parliamentary rules
Voluntary
Any move not driven by rules
Example of rule is minimum of 20 MPs in parliamentary group Includes moves recorded as expulsions
10
William B. Heller and Carol Mershon
each other. When legislators’ changes of party affiliation occur simultaneously or nearly so, by contrast, their moves might well be coordinated. For instance, switches might cluster in time as legislators respond to the initiatives of a political entrepreneur starting up a new party, or they might come in response to actual or expected partysystem changes that switching could provoke (cf. the conclusions and contributions by Kato and Yamamoto, Kreuzer and Pettai, Mershon and Shvetsova, and Schofield of this volume). With or without deliberate coordination, near-simultaneous switches on the part of one set of actors may signal to others a change in shared aims. Switches also could be orchestrated as one legislative party helps another meet a minimum threshold for official recognition—and rights—as a parliamentary party group. Of course, clusters of switches could be coincidental, or, more likely, a common (but essentially individual) response to some exogenous shock. Fourth, switches may be distinguished by the degree to which the actors involved are able to exercise free choice. Parliamentary rules governing the formation, membership, and importance of legislative parties can lead to forced switches, as when a party’s membership dips below some rule-defined threshold and its label disappears, forcing its members to switch into some other established group, to merge with other MPs to form a new group, or to fall into the mixed group or become independents. Another variant of involuntary moves might seem to come in the form of expulsions; yet since elected politicians can be assumed to exercise some degree of foresight, it is reasonable to view expulsions as in essence stemming from voluntary choice. On the flip side, laws, party rules, or interparty agreements may formally constrain MPs’ ability to switch parties: For example, in 1982, the Spanish Cortes prohibited MPs from joining groups (other than the mixed group) later than five days after the start of the legislative session (Sánchez de Dios 1999, 151), and in 1998, Spanish parties signed a “pact” designed to discourage switching at lower levels of governance (Santero 1998).
1.3.2
Incidence
As we and other contributors to the literature on party switching have indicated, the incidence of switching varies across countries— that is, across institutional contexts and party systems. Even where switching is rare, it is not utterly absent and its occurrence varies over time. Against stereotype, as table 1.2 shows, the phenomenon
Introduction
11
is not confined to new democracies or weakly institutionalized party systems. Table 1.2 includes but is not limited to the country cases examined in our book. Tables 1.2 and 1.3 below are brief and simple, yet ambitious and unique in the field. To our knowledge, nothing like them exists in the available literature; we thus construct them from some two dozen sources.4 The contributors to this volume go well beyond the observation of how many MPs switch parties in a legislative term, focusing on the motivations for, and the effects of, switching. In so doing, they take switching as a product of individual choice (whether coordinated or not), with consequences that can range from the level of the individual switcher to the party system as a whole.
Table 1.2 Illustrations of the incidence of switching (% MPs ever switched) in established and new democracies System
Term 1: % switched
Term 2: %
Term 3: %
Term 4: %
Australia Brazil Britain Canada Denmark* EP* France Germany Hungary Italy Japan New Zealand* Romania Russia South Africa Spain Turkey Ukraine United States
1975–1977: 3 1991–1994: 39 1974–1979: 1 1993–1997: 2 1966–1968: 3 1989–1994: 16 1997–2002: 4 1969–1972: 2 1990–1994: 13 1988–1992: 27.6 2000–2003: 7 1993–1996: 12 1992–1996: 11 1993–1995: 33 1999–2004: 6 1982–1986: 1 1961–1965: 22 1998–2002: 56 1991–1993: 0.2
1977–1980: 0 1994–1998: 33 1979–1983: 5 1997–2000: 4
1980–1983: 0 1998–2001: 36 1983–1987: 0 2000–2004: 9
1983–1984: 0
2002–2007: 10 1972–1976: 0.4 1994–1996: 6** 1992–1994: 33.7
1987–1992: 1 2004–2006: 2 1994–1998: 2
1976–1980: 0.2
1980–1983: 1
1994–1996: 34.4
1996–2000: 32.1
1996–1999: 6 1996–2000: 17
2000–2004: 10
2004–2009: 2** 1986–1989: 12 1965–1969: 21
1989–1993: 1 1969–1973: 23
1993–1996: 0.3 1973–1977: 10
1993–1995: 0
1995–1997: 1
1997–1999: 0.5
* Unicameral legislature; all other data pertain to lower houses. ** First two years of four-year term (Hungary); first 2.5 years of five-year term (South Africa). Sources: Ágh (1999, 172, 182); Australia, Parliament (2007); Booysen (2006, 735); Butler and Butler (2000, 248–249); Canada, House of Commons (2006); Bille and Pedersen (2004, 216); Corstange (2000); Desposato (2006, 69; this volume); France (2007); Heller and Mershon (2008); Kato and Yamamoto (2005, Table 2); Left Socialist Party, Denmark, 1997; McElroy (2003, 4); Mershon and Shvetsova (2007; 2008a); Miskin (2003, 17, 31); Nokken (this volume); Nokken and Poole (2004, 555); Schindler (1999, 926–929); South Africa, Parliament (2008); Thames (2005, 9, 24); Tomás Mallén (2002, 207–216); Turan (1985, 23).
12
William B. Heller and Carol Mershon
1.4 Extant Research on Switching: Lessons and Open Questions The forces that drive individual MPs’ decisions on party membership have attracted the attention of analysts of party switching. In particular, the contributors to this volume agree on the centrality of legislator ambition in choices and changes of party affiliation. Together, the available theoretical models and empirical findings (including contributions to this volume) highlight office perks, policy influence, and electoral advantage as motives for “jumping ship” (Aldrich and Bianco 1992; Desposato 2006; Heller and Mershon 2005; 2008; Laver and Benoit 2003; Mejía Acosta 1999; Mershon and Shvetsova 2005; 2008a; 2008b; Reed and Scheiner 2003; on multiple aims for parties, see Müller and Strøm 1999; Strøm 1990). Hence, switching not only is widespread but also is the product of strategic behavior, of a calculus of cost and benefit on the part of the individual legislator who faces incentives and constraints in his or her institutional environment (e.g., Desposato 2006; this volume; Heller and Mershon 2005; 2008; ch. 2 this volume; Kato and Yamamoto this volume; Mershon and Shvetsova 2008a; 2008b; this volume). More to the point, as strategic behavior switching has a profound impact on policy and politics. For instance, as illustrated in table 1.3, switchers in Canada have strengthened the hand of a new Premier, and those in France have forced government concessions on the budget (Huber 1996, 21, 154–159). Mobile German MPs have ruptured a governing coalition (Keesing’s Contemporary Archives 1957, 15757– 15759); their counterparts in Brazil and the Ukraine have enlarged the presidential majority (Ames 2001, 191–192, 273–276; Thames 2005, 9, 24), and in Japan have endowed the Liberal Democrats with a majority (Kato and Yamamoto this volume). Switchers in Spain have created “dysfunctionalities” in committee (Tomás Mallén 2002, 296–303); those in the EP and the United States have earned rewards via committee assignments (McElroy 2003; Mershon and Shvetsova 2008a; Yoshinaka 2005). The literature currently furnishes evidence on relationships between MPs’ ambition and goals, on the one hand, and institutional variables, on the other. For example, in the United Kingdom, the prime minister’s use of cabinet office to reward loyalists, which developed in the 1890s (Cox 1987), inhibited switching in phases of unidimensional party competition (Mershon and Nokken 2008). Under open-list proportional representation (PR) in Brazil, the automatic
Table 1.3
The incidence and impact of switching: Illustrations from established and new democracies
System
House
% Switched (MPs in house)
Australia Brazil Britain Canada Canada Denmark EP France Germany India Italy Japan New Zealand Russia South Africa Spain Spain Ukraine United States
lower lower lower lower lower unicameral unicameral lower lower state-level lower lower unicameral lower lower lower upper lower lower
n.a. (75–150) 36 (513) 5 (635) 9 (301) 1 (308) 2 (175) 15 (518) 7 (577) 7 (487) m.d. (30–403) 25 (630) m.d. (480) 6 (120) 33 (450) 6 (400) 12 (345) 11 (260) 56 (450) 1 (435)
Total Switches 97 262 29 36 4 4 81 41 49 2,700 277 40 7 342 23 55 38 527 5
Term(s) Covered
Example of Effect of Switching
1901–2003 1998–2002 1979–1983 2000–2004 2004–2006 1994–1998 1989–1994 1988–1993 1953–1957 1957–1973 1996–2001 2000–2003 1996–1999 1993–1995 1999–2004 1986–1989 1986–1989 1998–2002 1995–1997
one split led to rival’s “hegemony” switchers pipe more pork to win electoral challenge to Labour strengthened hand of new PM government survived confidence vote SDP moved rightward switchers shape policy in committee government concessions on budget government composition changed in 1956 ~ 45 state governments fell fall of government in 1998 LDP gained majority in 2001 minority government ruled from 1997 to 1999 switchers grab agenda advantage law on switching in constitutional court “dysfunctionalities” in committee “dysfunctionalities” in committee presidential majority enlarged switchers shape policy in committee
n.a. = not applicable. m.d. = missing data (not reported or cannot be constructed from source). Sources: Ames (2001, 46–47); Belford (2005); Bille and Pedersen (2004, 216); Booysen (2006, 735); Canada, House of Commons (2006); CBC News (2003a; 2003b); Corstange (2000); Crewe and King (1995, 104, ch. 23); Desposato (2006, 69); Economist (2005b; 2005a); Huber (1996, 21, 154–159); Kato and Yamamoto (2005, 14–16, Table 1); Keesing’s Contemporary Archives (1957, 15757–15759); Lalvani (2005, 133–134); McElroy (2003, 4); Mershon and Heller (2003, 18); Mershon and Shvetsova (2008a, 104, 122–123); Miskin (2003); Nokken (this volume); Nokken and Poole (2004, 555); Thames (2005, 9, 24); Tomás Mallén (2002, 207–209, 296–303).
14
William B. Heller and Carol Mershon
renomination of MPs has lowered the electoral costs of switching (Desposato 2003; 2006). Under open-list PR in Poland, incumbents switch to the opposition when economic performance is poor, as part of (typically successful) bids for reelection (Zielinski, Slomczynski, and Shabad 2005). One lesson from the extant literature that, in our view, too often remains implicit or is even ignored is that observed switching is the proverbial tip of the iceberg. To appreciate the point, note that the key insight underlying thinking on the strategic doctrine of mutual assured destruction was that the infrequency of the use of nuclear weapons testified not to the absence of a credible nuclear threat, but rather to the strength of the threat and the consequent effectiveness and power of mutual deterrence (Waltz 1990). Analogously, if—under some conditions—changes of party affiliation are infrequent (Norway provides a real-world example of stable legislative party membership), this observation can reflect the presence, effectiveness, and power, not the absence, of MPs’ strategic threats to switch or reconsider party affiliation, and of party leaders’ strategic moves to keep their legislative followings intact. For Britain, Crewe and King (1995, esp. 479) explicitly analyze 40 potential recruits to the SDP who remained in the Labour Party.5 The contributions to this volume are also explicitly premised on the notion that observed switching is the start, not the end, of the story. As we discuss in chapter 2, parties (i.e., party leaders in one form or another) seek to manipulate potential switchers, deterring or encouraging their moves (depending on context) with carrots or sticks, buying them off or threatening or imposing sanctions. MPs as a result decide whether to switch in a continuous calculus of affiliation that under some circumstances has the potential to change parties and possibly even legislative party systems (see Best 2008; Mershon and Shvetsova 2007). These can be viewed as transactions costs (Desposato 2006), but also as the product of strategic interaction. To elaborate, the transactions costs that dampen switching may be viewed as largely exogenous constraints. At one extreme, rules on ballot access might deprive switchers of their prerogative to run at the next election. At the other extreme, rules might enable any incumbent, regardless of switching behavior, to be automatically renominated (the case in Brazil until 2002, as stressed by Desposato). Internal legislative rules may lower the attractiveness—and hence the probability—of switching if, for example, parties control committee seats, so that MPs risk losing treasured committee positions if they
Introduction
15
change parties (see Desposato 2006). In Spain, for instance, legislative rules restrict switching to a very short span of time (Sánchez de Dios 1999; Tomás Mallén 2002). From the perspective we take in this book, as noted earlier, any observed stability in parties and party systems must be understood as an equilibrium result of legislator and party leadership strategies developed partially in response to the possibility of switching. Chapter 2 captures these strategic interactions in a four-stage game. For now, Figure 1.1 depicts these relationships among voters, legislators, party leaders, and the legislative party system and highlights the central role of institutions in mediating and structuring their interactions. The four basic elements of legislative politics illustrated in Figure 1.1 have long attracted political scientists’ attention. Our focus on party switching allows us to locate them all in the same theoretical framework and subject them to consistent, coherent empirical investigation. Although normative and positive studies of representation focus on the relationship between voters and their representatives, in the bottom left quadrant of the figure (e.g., Barnes 1977; Kohno 1997; Przeworski, Stokes, and Manin 1999), for instance, and research into elections and electoral rules investigates the links occupying either the upper left or the lower left quadrant (Cox 1997; Duverger 1964; Grofman and Lijphart 1986; Riker 1982), each type of study tends to refer to the other as an afterthought at best. Scholars on the US Congress take the lead in examining relationships between legislators and party leaders (on major debates in the vast literature, see Cox and McCubbins 1993; 2005; Krehbiel 1991; for prominent examples of comparative work, see Cox 1987; Döring 1995; Müller
Legislative Party System
Voters
Institutions
Legislators
Figure 1.1
The legislative context for party switching.
Party Leaders
16
William B. Heller and Carol Mershon
and Strøm 1999; Strøm 1990), but rarely—and understandably, given their tight focus on politics in the United States—do they examine legislative party systems.6 Studies of parties and party systems (the upper right quadrant in the figure), for their part, tend to leave voters very much in the background and, as noted, view parties and party systems as stable (the classic case, as well as the classic example, is Lipset and Rokkan 1967). These studies do not focus specifically on legislative party systems and thus neglect the possibility of legislative party switching as a source of change in party systems broadly defined. More broadly, the place of party switching is underestimated in all four well-tilled fields represented in the quadrants of figure 1.1. Just as each of the four quadrants in the figure has attracted substantial scholarly attention, so too has the role of institutions in the different quadrants. Different electoral rules yield varying incentives for sitting legislators to care about what their constituents want, for instance, hence fundamentally affecting the relationship between represented and representatives. Legislative and party rules, as well as electoral rules, define how closely a sitting legislator will be identified as a servant of his or her party versus an agent of his or her constituents. And party and legislative rules, within the parameters described by individual legislators’ independence, on one hand, and the legislative party system on the other, define party leaders’ ability unilaterally to set party objectives, induce the compliance of the party rank and file, and generally make a mark on policy. Other types of rules also have important consequences for political processes and outcomes—for example, whether a legislature is unicameral or bicameral (see, e.g., Druckman, Martin, and Thies 2003; Druckman and Thies 2001; Heller 1997; 2001; 2007; Tsebelis and Money 1997), or whether the system is presidential or parliamentary (Huber 1992; Moe and Caldwell 1994)—precisely because they structure how much actors with influence in decision making can cleave to their own desires versus follow someone else’s lead. As the contributions to this volume make abundantly clear, these institutional effects are crucial to legislators’ decisions about their party affiliation. This aspect of the impact of institutions on legislative politics is, however, essentially ignored by most scholars; the resulting understanding of legislative, party, electoral, and constitutional rules thus is incomplete. Once we see switching as the product of strategic calculation, we have to consider the role of reelection prospects in switchers’ calculations. After all, legislative careers grind to a halt for politicians who fail to win elections. In this vein, Grose and Yoshinaka (2003) find
Introduction
17
that switchers in the US Congress are less often reelected than are nonswitchers. Schmitt (1999) shows the same in a comparison of switchers and nonswitchers in the Brazilian lower house. Mershon and Heller (2003) also find that deputies who switch parties in Spain win reelection at lower rates than their less-mobile peers; Heller and Mershon (2005, note 9) observe that switchers are both less likely than nonswitchers to run for reelection and to be reelected if they do run. The easy inference would be that switching often carries an electoral cost. That inference should not be embraced too hastily, however, for this issue requires counterfactual thinking and more. If the switcher had stayed put, would he have faced some form of demotion (placement low on the party list, movement to a district where the party is historically weak) that would have increased the probability of electoral defeat had he not switched? Who among switchers and nonswitchers decides not to run again? What kinds of candidates challenge those who have switched versus those who have not? Of course, the electoral prospects of switchers depend also on factors beyond their own behavior or control, not least among which are national or local trends that affect party performance (and also could motivate switching). As Kato and Yamamoto emphasize in this volume, the electoral prospects of switchers hinge as well on the nominating strategies of the parties (of the party leaderships) that have accepted them as new members. The proper baseline for analyzing the consequences of switching is what would have happened in the absence of switching. It is important, in other words, to consider counterfactuals; reelection is just one instance—albeit perhaps the most obvious—where this awareness of what might have been is vital. The key point is that the choice to switch parties is strategic, and the unobserved decision not to switch—we cannot, after all, easily tell who among nonswitchers might have entertained the idea of moving to a different party (but see Crewe and King 1995)—also is strategic. This treatment of strategy, choice, and chance leads naturally to the next section’s preview of the overall architecture of the book.
1.5 Map of the Book All of the chapters here build on a common theoretical foundation and demonstrate the common aspects of party switching across countries and systems, even though no two chapters address precisely the
18
William B. Heller and Carol Mershon
same aspect of the phenomenon and its consequences. The shared foundation, laid out by Heller and Mershon in this chapter and the next, begins with the implications of switching, a definition of the phenomenon, and a survey of its incidence. The discussion culminates in the next chapter, as we set up what is in principle a multistage game between party leaders, party rank and file, and voters. The game, which is complex and probably intractable in its general expression, suggests the different and sometimes conflicting goals of political actors. The remaining chapters in this volume investigate how political actors pursue these varying goals in different contexts. Part two of the book, “Party Switching and Representation,” focuses on how party switching both affects and reflects voters’ indirect influence over policy. In this vein, Norman Schofield examines party switchers’ motivations by leveraging the general observation that party positions tend to be heterogeneous in policy space. Schofield not only explores an integrated theory of party strategy but also uses examples of party switching in the Israeli Knesset and from US politics to demonstrate that party systems can be stable as long as party positions are insensitive to switching. Schofield further identifies conditions for switching: under PR rules, a party leader with high valence may switch so as to adopt centrist policies; and under plurality rules, the conflicting demands of activist groups may lead politicians to switch party. For party positions to be insensitive to switching, parties must in some sense be immune to influence from new members who might want to rearrange party priorities or goals. Switchers, in other words, have to adapt to the positions of their new parties, and not vice versa. In separate chapters focusing on the United States and Brazil, Timothy Nokken and Scott Desposato show that switchers do indeed change their behavior to fit the positions of their new parties. For the US Congress, previous analyses (Nokken 2000; Nokken and Poole 2004) had found that switchers change their behavior at times of high ideological polarization. In “Party Switching and the Procedural Party Agenda in the US House of Representatives,” Nokken focuses on procedural votes, which are less publicly visible than final passage votes on bills, but arguably equally or more important for policy outcomes, to show that party switchers consistently toe their new party’s line in procedural and amendment votes, even when they evince relatively little change of behavior in final passage votes. Along the same lines, Scott Desposato demonstrates in “Party Switching in Brazil: Causes, Effects, and Representation” that party switchers in Brazil alter their
Introduction
19
behavior to align themselves with their new parties. Desposato’s argument, presented in two distinct steps, turns on parties’ ability to control the flow to their members of the benefits of membership in the legislature, including but not limited to policy outcomes as well as ballot access for reelection. To the extent that they control resources important to legislators, parties can make themselves more or less attractive to a potential switcher, depending on the switcher’s contribution to the party and its existing members. It follows straightforwardly that parties that can manipulate the incentive to switch also can manipulate individuals’ incentives to toe the party line in legislative voting. Taken together, Nokken and Desposato’s results lend strong support to Schofield’s key condition for a structurally stable party equilibrium. Hence, switching might not threaten representation as long as voters condition their ballots on choice of parties, not of individual candidates who then could change both party and position. The chapters in the third part, “Party Switching, Party Competition, and Policy Making,” examine how party switching affects and is affected by interparty competition. In their investigation of switching among party groups in the EP, Gail McElroy and Kenneth Benoit explore the relationships among the officially designated party groups in the EP, national parties, and the incentives of individual Members of the European Parliament (MEPs). Drawing on original datasets of party group affiliations in the EP at the level of both national parties and individual MEPs, McElroy and Benoit isolate ideology and party policy as the central influences on choices of EP group affiliation. In the next chapter, William Heller and Carol Mershon use party switchers’ movements to gain analytical leverage on the determinants of party positioning. They show, using roll-call data from the Italian parliament, that party positions are sensitive to party switching, but in a one-sided manner. Starting with the notion that party positions ought to reflect the preferences of their members, Heller and Mershon find that party positions change when old members switch out, but not when new members switch in. They conclude that outswitchers probably often defect under pressure, and that their exits allow their now-former parties to adjust their positions in ways that they could not have otherwise. Inswitching, by contrast, appears to have no effect on party positions, a finding that marches with those of Nokken and Desposato’s findings, just described. The final chapter in this section, Carol Mershon and Olga Shvetsova’s “Timing Matters: Incentives for Party Switching and Stages of
20
William B. Heller and Carol Mershon
Parliamentary Cycles,” emphasizes party affiliation as a question of strategic choice that recurs throughout the legislative term. Analyzing Italy, Romania, Russia, United Kingdom, and the United States, Mershon and Shvetsova identify heightened switching for office benefits, policy advantage, and vote seeking at distinctive moments in the parliamentary cycle. The emergence of this fundamental commonality across the five different systems—alongside important contrasts— supports the notion that the timing of a legislator’s switch reveals the predominant motivation for the switch. The final section of the volume, “Party Switching and the Dynamics of Party Systems,” treats the possibility that party switching can change parties’ legislative weights and, consequently, entire legislative party systems. The first chapter in this section, Junko Kato and Kentaro Yamamoto’s “Competition for Power: Party Switching and Party System Change in Japan,” focuses on the institutional context of party switching. Where legislative institutions create seat-share thresholds for party influence, switchers can push a party to increased influence (or, when they are outswitchers, to reduced influence); a party that is close to such a threshold has a strong incentive to seek to attract inswitchers, thereby increasing its overall legislative influence. Kato and Yamamoto distinguish between a majority capable of passing bills in the plenary and one that holds enough seats to control the legislative agenda, particularly in committees. In their chapter on “Party Switching, Party Systems, and Political Representation,” Marcus Kreuzer and Vello Pettai argue that the degree of party system institutionalization shapes the frequency, forms, and consequences of party switching. To build their argument, they consider different kinds of party switching behaviors, from individual party hopping, party fissions, and party fusions to the creation of completely new parties. By comparing a broad array of party systems, from transitional to overinstitutionalized and uncompetitive, Kreuzer and Pettai emphasize party switching as an important and understudied mechanism for political change. Party switching by sitting legislators presents an empirical curiosity and a theoretical opportunity. It seems odd on its face that a politician whose career and successes are tied up in a single party might consider abandoning that party for another. It is thus all too easy to view party switching as symptomatic of special circumstances, as for instance in new democracies (e.g., Àgh 1999; Grofman, Mikkel, and Taagepera 2000; Mainwaring 1999), in party systems that are relatively unconsolidated or undergoing some upheaval (Castle and
Introduction
21
Fett 2000; Mejía Acosta 1999; Reed and Scheiner 2003; Turan 1985), or where parties are relatively weak (as in the United States; see Ansolabehere, Snyder, and Stewart 2001; McCarty, Poole, and Rosenthal 2001; Nokken 2000). As the contributions to this volume show, studying party switching yields far more than an understanding of who switches and why: It allows us to analyze parties and party systems, as well as policy making, in terms of interdependent, individual decisions. The next chapter takes up the task of modeling interdependent individual decisions as structured by institutions.
1.6
Notes
1. The Belgian case illustrates a possibility more complex than policy “drift.” If a few political entrepreneurs activate an additional dimension of conflict, and virtually all parties move to define positions in the new policy space, legislators can switch in the effort to serve faithfully their constituencies (e.g., Dewachter 1987). For a formal analysis of the activation of suppressed dimensions, with illustrations from the United States, see Miller and Schofield (2003; they touch on switching as well, 256; 2008; Schofield 2006; Schofield and Sened 2006; and cf. Schofield’s chapter in this volume). 2. The effective number of parliamentary parties is a widely used measure that takes into account both the raw number and the seat strength of legislative parties (e.g., Cox 1997; Ordeshook and Shvetsova 1994). 3. Start-up parties are distinct from parties created by fission because they draw legislators from multiple parties; they are distinct from fusion because the parties that lose members do not disappear. 4. For a cross-national comparison of switching, see Chang (2006). Mershon and Shvetsova (2008b) analyze more than 3,800 observations from 9 countries. 5. Crewe and King identify potential recruits on the basis of several sources: a key vote in the Commons (October 1971 on what was then the European Economic Community, EEC), interviews, signed support for reforms of party structure (in September 1980), and press reports. 6. Americanists do refer to different party systems in the United States (see Key 1964; Sundquist 1983), but not in the same sense that students of comparative politics define them.
1.7 References Àgh, Attila. 1999. “The Parliamentarization of the East Central European Parties: Party Discipline in the Hungarian Parliament, 1990–1996.” In Party Discipline and Parliamentary Government, ed. Shaun Bowler, David M. Farrell, and Richard S. Katz, 167–188. Columbus: Ohio State University Press. Aldrich, John H. 1995. Why Parties? The Origin and Transformation of Political Parties in America. Chicago: University of Chicago Press.
22
William B. Heller and Carol Mershon
Aldrich, John H., and William T. Bianco. 1992. “A Game-Theoretic Model of Party Affiliation of Candidates and Office Holders.” Mathematical and Computer Modelling 16 (8/9): 103–116. Ames, Barry. 2001. The Deadlock of Democracy in Brazil. Ann Arbor: University of Michigan Press. Ansolabehere, Stephen, James M. Snyder, Jr., and Charles Stewart III. 2001. “The Effects of Party and Preferences on Congressional Roll Call Voting.” Legislative Studies Quarterly 26 (4): 533–572. Australia, Parliament. 2007. Parliamentary Handbook of the Commonwealth of Australia: Historical Information on the Australian Parliament. http://www. aph.gov.au/library/handbook/historical/index.htm. Last accessed August 8, 2007. Barnes, Samuel H. 1977. Representation in Italy: Institutionalized Tradition and Electoral Choice. Chicago: University of Chicago Press. Baron, David P. 1998. “Comparative Dynamics of Parliamentary Governments.” American Political Science Review 92 (3, September): 593–609. Belford, Aubrey. 2005. Interview: The Great Labor Schism. Pulse Books, 6 May. http://www.vibewire.net/3/node/4279. Best, Robin. 2008. “Equal Access for All? Electoral Institutions and the Dynamics of Party System Size in Western Democracies, 1950–2005.” Paper presented at the annual meeting of the Midwest Political Science Association, Chicago, IL. Best, Robin, and William B. Heller. 2005. “Safety in Numbers? Seat Shares and Discipline in Legislative Parties.” Paper presented at ECPR Joint Sessions of Workshops, Granada, Spain, April 14–19. Bille, Lars, and Karina Pedersen. 2004. “Electoral Fortunes and Responses of the Social Democratic Party and Liberal Party in Denmark: Ups and Downs.” In Political Parties and Electoral Change, ed. Peter Mair, Wolfgang C. Müller, and Fritz Plasser, 207–233. London and Thousand Oaks, CA: Sage. Booysen, Susan. 2006. “The Will of the Parties versus the Will of the People? Defections, Elections, and Alliances in South Africa.” Party Politics 12 (6): 727–746. Bowler, Shaun, David M. Farrell, and Richard S. Katz, eds. 1999. Party Discipline and Parliamentary Government. Columbus: Ohio State University Press. Browne, Eric C., and Mark N. Franklin. 1973. “Aspects of Coalition Payoffs in European Parliamentary Democracies.” American Political Science Review 67 (2, June): 453–469. Bruneau, Thomas C., P. Nikiforos Diamandouros, Richard Gunther, Arend Lijphart, and Leonardo Morlino. 2001. “Democracy, Southern European Style.” In Parties, Politics, and Democracy in the New Southern Europe, ed. P. Nikiforos Diamandouros and Richard Gunther, 16–82. Baltimore and London: Johns Hopkins University Press. Butler, David, and Gareth Butler. 2000. Twentieth Century British Political Facts 1900–2000. 8th ed. London: Macmillan. Canada, House of Commons. 2006. “Members of the House Who Crossed the Floor of the House of Common or Who Changed Parties 1867 to Date.” Canada, House of Commons, Library of Parliament.
Introduction
23
Canon, David T., and David J. Sousa. 1992. “Party System Change and Political Career Structures in the U.S. Congress.” Legislative Studies Quarterly 17 (3): 347–363. Carey, John M., and Matthew Soberg Shugart. 1995. “Incentives to Cultivate a Personal Vote: A Rank Ordering of Electoral Formulas.” Electoral Studies 14 (4, December): 417–439. Castle, David, and Patrick J. Fett. 2000. “Member Goals and Party Switching in the U.S. Congress.” In Congress on Display, Congress at Work, ed. William T. Bianco, 231–241. Ann Arbor: University of Michigan Press. CBC News. 2003a. Bloc MP Crosses over to Liberals, December 3. http://www. cbc.ca/canada/story/2003/12/02/bloc031102.html. Last accessed August 14, 2006. ———. 2003b. MacKay Slams Brison for Joining Liberals, December 10. http:// www.cbc.ca/canada/story/2003/12/10/brison_031210.html. Last accessed August 14, 2006. Chang, Chuan-hsien. 2006. “Politics of Defection: Reinvestigating Influence of Electoral System over Party System.” Paper presented at the James F. Jakobsen Conference 2006, University of Iowa, Iowa City, IA. Corstange, Daniel. 2000. “Denmark: The Party System from 1963 to 2000.” In International Comparative Political Parties Project, ed. Kenneth Janda. http://www.janda.org/icpp/index.htm. Last accessed June 28, 2007. Cox, Gary W. 1987. The Efficient Secret: The Cabinet and the Development of Political Parties in Victorian England. Cambridge: Cambridge University Press. ———. 1997. Making Votes Count: Strategic Coordination in the World’s Electoral Systems. Cambridge: Cambridge University Press. Cox, Gary W., and Mathew D. McCubbins. 1993. Legislative Leviathan: Party Government in the House. Berkeley: University of California Press. ———. 2005. Setting the Agenda: Responsible Party Government in the U.S. House of Representatives. Cambridge: Cambridge University Press. Crewe, Ivor, and Anthony King. 1995. SDP: The Birth, Life, and Death of the Social Democratic Party. Oxford: Oxford University Press. Desposato, Scott W. 2003. “The Impact of Party Switching on Legislative Behavior in Brazil.” Paper presented at the annual meeting of the American Political Science Association, Philadelphia, PA, August 28–31. ———. 2006. “Parties for Rent? Ambition, Ideology, and Party Switching in Brazil’s Chamber of Deputies.” American Journal of Political Science 50 (1): 62–80. Dewachter, Wilfried. 1987. “Changes in Particratie: The Belgian Party System from 1944 to 1986.” In Party Systems in Denmark, Austria, Switzerland, the Netherlands and Belgium, ed. Hans Daalder, 285–363. New York: St. Martin’s. Döring, Herbert, ed. 1995. Parliaments and Majority Rule in Western Europe. New York: St. Martin’s. Druckman, James N., Lanny W. Martin, and Michael F. Thies. 2003. “Influence without Confidence: Upper Chambers and Parliamentary Government
24
William B. Heller and Carol Mershon
Formation.” Paper presented at the annual meetings of the Midwest Political Science Association, Chicago, IL, April 3–6. Druckman, James N., and Michael F. Thies. 2001. “The Importance of Concurrence: The Impact of Bicameralism on Government Formation and Duration.” Paper presented at the annual meetings of the Midwest Political Science Association, Chicago, IL, April 19–22. Duverger, Maurice. 1964. Political Parties. Translated by B. North, and R. North. Cambridge: Methuen. Economist. 2005a. “After the Vote.” May 26. ———. 2005b. “Belinda’s Leap.” May 21, 40. Filippov, Mikhail, Peter C. Ordeshook, and Olga V. Shvetsova. 2004. Designing Federalism: A Theory of Self-Sustainable Federal Institutions. Cambridge: Cambridge University Press. France, Assemblée Nationale. 2007. “Modifications apportées à la composition de l’Assemblée Nationale, XIe legislature.” http://www.assemblee-nationale. fr/11/qui/modifications-11leg.asp. Last accessed May 11, 2007. Grofman, Bernard, and Arend Lijphart, eds. 1986. Electoral Laws and Their Political Consequences. New York: Agathon Press. Grofman, Bernard, Evald Mikkel, and Rein Taagepera. 2000. “Fission and Fusion of Parties in Estonia, 1987–1999.” Journal of Baltic Studies 31 (4): 329–357. Grose, Christian R., and Antoine Yoshinaka. 2003. “The Electoral Consequences of Party Switching by Incumbent Members of Congress, 1947–2000.” Legislative Studies Quarterly 28 (1): 55–75. Heller, William B. 1997. “Bicameralism and Budget Deficits: The Effect of Parliamentary Structure on Government Spending.” Legislative Studies Quarterly 22 (4): 485–516. ———. 2001. “Making Policy Stick: Why the Government Gets What It Wants in Multiparty Parliaments.” American Journal of Political Science 45 (4, October): 780–798. ———. 2007. “Divided Politics: Bicameralism, Parties, and Policy in Democratic Legislatures.” Annual Review of Political Science 10: 245–269. Heller, William B., and Carol Mershon. 2005. “Party Switching in the Italian Chamber of Deputies, 1996–2001.” Journal of Politics 67 (2): 536–559. ———. 2008. “Dealing in Discipline: Party Switching and Legislative Voting in the Italian Chamber of Deputies, 1988–2000.” American Journal of Political Science 52 (4): 910–925. Hopkin, Jonathan. 1999. Party Formation and Democratic Transition in Spain: The Creation and Collapse of the Union of the Democratic Center. New York: St. Martin’s. Huber, John D. 1992. “Restrictive legislative Procedures in France and the U.S.” American Political Science Review 86 (3): 675–687. ———. 1996. Rationalizing Parliament: Legislative Institutions and Party Politics in France. New York: Cambridge University Press. Kato, Junko, and Kentaro Yamamoto. 2005. “Competition for Power: Party Switching as a Means for Changing Party Systems in Japan.” Paper presented
Introduction
25
at the meeting of the Research Work Group on Legislative Party Switching, University of Virginia, Charlottesville, VA. Keesings Contemporary Archives. 1957. Bristol: Longman. Key, V. O. 1964. Politics, Parties, and Pressure Groups. New York: Thomas Y. Crowell. Kohno, Masuru. 1997. Japan’s Postwar Party Politics. Princeton, NJ: Princeton University Press. Krehbiel, Keith. 1991. Information and Legislative Organization. Ann Arbor: University of Michigan Press. Lalvani, Mala. 2005. “Coalition Governments: Fiscal Implication for the Indian Economy.” American Review of Political Economy 3 (1): 127–163. Laver, Michael. 1998. “Models of Government Formation.” Annual Review of Political Science 1: 1–25. Laver, Michael, and Kenneth Benoit. 2003. “The Evolution of Party Systems between Elections.” American Journal of Political Science 47 (2, April): 215–233. Laver, Michael, and Norman Schofield. 1990. Multiparty Government: The Politics of Coalition in Europe. Oxford: Oxford University Press. Left Socialist Party, Denmark. 1997. “Documents and History.” English version last updated December 3, 1997. http://www.venstresocialisterne.dk/uk_index. htm. Last accessed June 27, 2007. Lipset, Seymour Martin, and Stein Rokkan. 1967. Party Systems and Voter Alignments: Cross-National Perspectives. New York: Free Press. MacRae, Duncan. 1967. Parliament, Parties, and Society in France 1946–1958. New York: St. Martin’s. Mainwaring, Scott P. 1999. Rethinking Party Systems in the Third Wave of Democratization: The Case of Brazil. Stanford, CA: Stanford University Press. Mainwaring, Scott P., and Aníbal Pérez Liñán. 1997. “Party Discipline in the Brazilian Constitutional Congress.” Legislative Studies Quarterly 22 (4): 453–483. Mainwaring, Scott P., and Mariano Torcal. 2005. “Party System Institutionalization and Party System Theory after the Third Wave of Democratization.” Unpublished paper, Kellogg Institute, University of Notre Dame. Mair, Peter, and Michael Marsh. 2004. “Political Parties in Electoral Markets in Postwar Ireland.” In Political Parties and Electoral Change, ed. Peter Mair, Wolfgang C. Müller, and Fritz. Plasser. Thousand Oaks, CA: Sage. Martin, Lanny, and Georg Vanberg. 2004. “Policing the Bargain: Coalition Government and Parliamentary Scrutiny.” American Journal of Political Science 48 (1, December): 13–27. ———. 2005. “Coalition Policymaking and Legislative Review.” American Political Science Review 99 (1): 73–106. McCarty, Nolan M., Keith T. Poole, and Howard Rosenthal. 2001. “The Hunt for Party Discipline in Congress.” American Political Science Review 95 (3): 673–687. McElroy, Gail M. 2003. “Party Switching in the European Parliament: Why Bother?” Paper presented at the annual meetings of the Midwest Political Science Association, Chicago, IL, April 3–6.
26
William B. Heller and Carol Mershon
Mejía Acosta, Andrés. 1999. “Indisciplina y deslealtad en el congreso ecuatoriano.” Iconos (6, January): 13–21. Mershon, Carol. 2001. “Contending Models of Portfolio Allocation and Office Payoffs to Party Factions: Italy, 1963–79.” American Journal of Political Science 45 (2, April): 277–293. ———. 2002. The Costs of Coalition. Stanford, CA: Stanford University Press. ———. 2008. “Legislative Party Switching and Executive Coalitions.” Unpublished paper, University of Virginia, Charlottesville, VA. Mershon, Carol, and Olga Shvetsova. 2005. “Electoral Cycles and Party Switching: Opportunistic Partisan Realignment in Legislatures.” Paper presented at the annual meeting of the Midwest Political Science Association, Chicago, IL, April 6–10. ———. 2007. “Institutional Determinants of Party System Change and Party Volatility in Legislatures.” Unpublished paper, University of Virginia, Charlottesville, VA. ———. 2008a. “Parliamentary Cycles and Party Switching in Legislatures.” Comparative Political Studies 41 (1): 99–127. ———. 2008b. “Party Switching in Sitting Parliaments and the Midterm Effect.” Paper presented at Annual Joint Sessions of the European Consortium for Political Research, Rennes, France. Mershon, Carol, and Timothy P. Nokken. 2008. “Party Formation and Changes of Party Affiliation among Legislators: The United States and Great Britain in the Nineteenth and Early Twentieth Centuries.” Paper presented at the annual meetings of the Midwest Political Science Association, Chicago, IL, April 3–6. Miller, Gary, and Norman Schofield. 2003. “Activists and Partisan Realignment in the United States.” American Political Science Review 97 (2): 245–260. ———. 2008. “The Transformation of the Republican and Democratic Party Coalitions in the US.” Perspectives on Politics 6 (3): 433–450. Miskin, Sarah. 2003. “Politician Overboard: Jumping the Party Ship.” Unpublished paper, Parliament of Australia: Parliamentary Library—Politics and Public Administration Group. Moe, Terry M., and Michael Caldwell. 1994. “The Institutional Foundations of Democratic Government: A Comparison of Presidential and Parliamentary Systems.” Journal of Institutional and Theoretical Economics 150 (1): 171–195. Müller, Wolfgang C., and Kaare Strøm, eds. 1999. Policy, Office, or Votes: How Political Parties in Western Europe Make Hard Decisions. Cambridge: Cambridge University Press. Nokken, Timothy P. 2000. “Dynamics of Congressional Loyalty: Party Defection and Roll Call Behavior, 1947–1997.” Legislative Studies Quarterly 25 (3): 414–444. Nokken, Timothy P., and Keith T. Poole. 2004. “Congressional Party Defection in American History.” Legislative Studies Quarterly 29 (4): 545–568. Ordeshook, Peter C., and Olga V. Shvetsova. 1994. “Ethnic Heterogeneity, District Magnitude, and the Number of Parties.” American Journal of Political Science 38 (1, February): 100–123.
Introduction
27
Powell, G. Bingham, Jr. 2000. Elections as Instruments of Democracy: Majoritarian and Proportional Visions. New Haven, CT: Yale University Press. Przeworski, Adam, Susan C. Stokes, and Bernard Manin, eds. 1999. Democracy, Accountability, and Representation. Cambridge: Cambridge University Press. Reed, Steven R., and Ethan Scheiner. 2003. “Electoral Incentives and Policy Preferences: Mixed Motives behind Party Defections in Japan.” British Journal of Political Science 33 (3): 469–490. Riker, William H. 1982. “The Two-Party System and Duverger’s Law: An Essay on the History of Political Science.” American Political Science Review 76 (4): 753–766. Sánchez de Dios, Manuel. 1999. “Parliamentary Party Discipline in Spain.” In Party Discipline and Parliamentary Government, ed. Shaun Bowler, David M. Farrell, and Richard S. Katz, 141–162. Columbus: Ohio State University Press. Santero, F. Javier. 1998. “Pacto de los partidos contra el transfuguismo en los ayuntamientos.” El Mundo July 8. http://www.elmundo.es/1998/07/08/ espana/08N0026.html. Last accessed July 29, 2008. Schattschneider, E. E. 1942. Party Government. New York: Rinehart. Schindler, Peter. 1999. Datenhandbuch zur Geschichte des Deutschen Bundestages 1949 bis 1999. Band I. Berlin: Deutscher Bundestag. Schmitt, Rogério. 1999. “Migração partidária e reeleição na Câmara dos Deputados.” Novos Estudos CEBRAP 54 (July): 127–146. Schofield, Norman. 2006. Architects of Political Change: Constitutional Quandaries and Social Choice Theory. Cambridge: Cambridge University Press. Schofield, Norman, and Michael Laver. 1985. “Bargaining Theory and Portfolio Payoffs in European Coalition Governments, 1945–83.” British Journal of Political Science 15 (2): 143–164. Schofield, Norman, and Itai Sened. 2006. Multiparty Democracy: Parties, Elections and Legislative Politics in Multiparty Systems. Cambridge: Cambridge University Press. South Africa, Parliament of the Republic. 2008. “State of Parties in the NA.” http:// www.parliament.gov.za/live/content.php?Category_ID=148. Last accessed July 29, 2008. Stokes, Susan C. 1999. “Political Parties and Democracy.” Annual Review of Political Science 2: 243–267. Strøm, Kaare. 1990. “A Behavioral Theory of Competitive Political Parties.” American Journal of Political Science 34 (2, May): 565–598. Sundquist, James L. 1983. Dynamics of the Party System: Alignment and Realignment of Political Parties in the United States. Washington, DC: Brookings Institution Press. Thames, Frank C. 2005. “Parliamentary Party Switching in the Ukrainian Rada, 1998–2002.” Paper presented at the annual meetings of the Midwest Political Science Association, Chicago, IL.
28
William B. Heller and Carol Mershon
Tomás Mallén, Beatriz. 2002. Transfuguismo parlamentario y democracia de partidos. Madrid: Centro de Estudios Políticos y Constitucionales. Tsebelis, George, and Jeannette Money. 1997. Bicameralism. Cambridge: Cambridge University Press. Turan, Ilter. 1985. “Changing Horses in Midstream: Party Changers in the Turkish National Assembly.” Legislative Studies Quarterly 10 (1): 21–34. Waltz, Kenneth N. 1990. “Nuclear Myths and Political Realities.” American Political Science Review 84 (3, September): 731–745. White, Stephen, Richard Rose, and Ian McAllister. 1997. How Russia Votes. Chatham, NJ: Chatham House. Yoshinaka, Antoine. 2005. “House Party Switchers and Committee Assignments: ‘Who Gets What, When, and How?’ ” Legislative Studies Quarterly 30 (3): 391–406. Zielinski, Jakub, Kazimierz M. Slomczynski, and Goldie Shabad. 2005. “Electoral Control in New Democracies: The Perverse Incentives of Fluid Party Systems.” World Politics 57 (3): 365–395.
2 Integrating Theoretical and Empirical Models of Party Switching William B. Heller and Carol Mershon
2.1 Introduction Party switching is a relatively common yet little studied phenomenon. Observers have remarked on the presence of switching in various circumstances and settings, but with a very few exceptions (Aldrich and Bianco 1992; Desposato 2006; Heller and Mershon 2005; 2008; Laver and Benoit 2003) scholars have not seen party switching as theoretically interesting. They have instead treated party switching as an idiosyncratic phenomenon, entirely dependent on context, and essentially sui generis in each occurrence. The contributions to this volume represent an attempt to address what we see as a gap between extant empirical accounts of switching and the substantial leverage that a theoretically driven approach to switching can provide. Taken together, the chapters examine the contexts, causes, and consequences of party switching. Although most of the chapters focus on one or a few country cases, and each chapter examines only a piece of a larger set of strategic interactions in which switching occurs, each does so in explicitly theoretical terms. We indicated in the introduction that party switching takes place in the context of strategic interaction between individual members of parliament (MPs) and the leaders of the MPs’ parties and of other parties. The decisions that these actors make reflect their interaction and also a larger context that includes the legislative party system, on one hand, and voters, on the other. In this chapter, we lay out a theoretical framework for analyzing party switching that permits isolation of the separate elements and consequences of the strategic interactions entailed in party switching, with a clear focus on individual decisions, while placing those decisions explicitly in their larger
30
William B. Heller and Carol Mershon
context. We do not construct a fully specified, overarching switching game. Rather, we present the framework of a four-stage game—only two stages of which involve strategic interaction—as a foundation for examining switching in specific cases and environments. Building on this foundation, the chapters that follow contribute, individually and in the aggregate, to a general understanding of the causes and consequences of party switching. Our sketch of a switching game, along with the pieces of it fleshed out by the other chapters in this volume, suggests a potential for carrying the study of party switching even farther than we and our colleagues do here. One intriguing possibility, for example, is that legislative party switching should come in cascades, as every switch both provides new information about (cf. Heller and Mershon 2005) and alters the legislative environment. Such changes could in turn induce at least some legislators who previously had not considered switching to reevaluate their positions. To the extent that such reevaluation leads to more switches, it is easy to imagine circumstances where just a few initial switches could trigger many more. The result should be cascades of switching. Although an in-depth examination of the possibility of cascades lies beyond the scope of this book, in the conclusions we briefly discuss how they might occur and what sorts of impacts they might have. To construct a game-theoretic framework for legislative party switching, we must first identify the bases for our inquiry. In the next section, we concentrate on the key assumptions undergirding both this volume and the project out of which it grew. The most important of these is that legislator behavior, for those who switch and those who do not, is motivated by ambition. In the third section, we lay out the pieces and logic of our four-stage switching game. We then highlight several of the questions that the game brings to the fore, and that the contributors to this volume explore in greater detail. We end this chapter with some brief observations about the analytical purchase afforded by submitting party switching on the part of elected politicians to rigorous analysis.
2.2 Precursors to Theory Party switching is fundamentally an individual phenomenon, as the introduction indicated. Even when legislators move en masse from one party to another, or to form a new party, each legislator’s move generally is the product of an individual choice.1 The study of party
Theoretical and Empirical Models
31
switching thus imposes a basic theoretical perspective, grounded in a focus on individual behaviors, strategies, and choices. This perspective entails two underpinning assumptions. First, individual politicians seek to achieve their personal goals and desires, and their behavior is motivated by ambition directed at achieving them. Second, party switching is a tool that individual legislators can use as they seek to achieve their own ambitions. The ability to move from one party to another is not unconstrained, of course—a MP cannot move into a party that will not accept him or her, for example—but it does give MPs the possibility of improving their lot by changing their party affiliation.
2.2.1
Individuals and their choices
We begin with individual choices, as we must if we are to model legislators’ decisions to change their party affiliations. Our focus expands rapidly beyond the individual making the choice, however, as both the decision to switch and the alternatives available to a would-be switcher depend heavily on context. An individual’s choice, after all, depends on the set of feasible alternatives to doing nothing—and, in some cases (e.g., for an MP at risk of expulsion from his or her party), doing nothing is itself not a feasible alternative. In the introduction, we observed that party switching seems odd in light of parties’ importance to legislators.2 The first question to ask, therefore, is: why would a sitting legislator want to change his or her party affiliation? To answer this adequately, we need to know not only how they are faring in their current party, but also what alternatives are open to them and how they could expect to fare in each of those. We need, in other words, to ask where a would-be switcher would go. The answer depends on other parties. Thus, we need to know whether other parties will accept switchers and, if so, why—that is, what would they have to gain from accepting defectors from other parties? A party’s willingness to accept new members depends on a number of factors, not least of which is its position in the legislative party system. On one hand, the legislative party system matters because it defines the bargaining context for policy making, agenda control, and, in parliamentary and semi-presidential systems, government formation. On the other hand, the party system matters because it defines the range of options available to voters at election time and, hence, it provides the parameters, in combination with the electoral
32
William B. Heller and Carol Mershon
system, that define any given legislator’s probability of reelection. Further, the institutionalization of parties and the party system might play an important role in determining how any given MP views the attractiveness of membership in any given party as well as the costs of switching from one party to another. A key, unanswered question in this regard, however, is whether relatively low levels of party and party system institutionalization lead to higher levels of party switching, or whether, inversely, frequent party switching retards or even reverses such institutionalization (for discussion of this issue, see Kreuzer and Pettai this volume). Switching is motivated by and occurs in the context of these factors. They in turn are dynamic, and every switch can in principle change them. If larger parties are generically more attractive than smaller ones, for example, then every time one MP switches all other MPs implicitly have to reevaluate their own prospects. Every move by an MP changes the legislative weight of at least one party (though compensating moves might change weights back). More subtly, every move by an MP is likely to change the observed ideal point of at least one party (see Heller and Mershon in chapter 7 here), hence altering its attractiveness both to other potential switchers and to its remaining members. We stress, in other words, the interdependence of decisions to switch and of switches themselves (cf. Aldrich and Bianco 1992). A related and more general point is that whether and when an MP switches is in large part a consequence of the information available to her. Information about voter preferences (e.g., from subnational election results) might motivate an MP to reevaluate his or her expectations about party vote shares in the next parliamentary elections (and thus postelection seat shares), for example. If MPs have information about their counterparts’ or colleagues’ plans to switch parties—through rumor, inference, or straightforward announcement—shared expectations of moves could lead to simultaneous or near-simultaneous jumps by several legislators, either as a group or in some kind of interparty shuffle. Party repositioning, whether calculated or incidental to switches or some other occurrence, can move parties’ observed policy preferences relative to the ideal points of potential switchers. Or observed moves by well-known or otherwise powerful politicians might change an MP’s priors about how attractive parties are for voters as well as for MPs interested in legislative influence. Note that a group of ideologically compatible MPs might see forming a new party as an attractive option, particularly if they believe that voters
Theoretical and Empirical Models
33
are looking for a change, or that a new party could induce still other MPs to switch into it. This line of reasoning suggests that switching cascades are indeed likely, as noted earlier. Clearly, the possibility of switching means that what has commonly been considered static is not. Switching introduces a dynamic to legislative politics and to party politics more broadly, a dynamic that is progressive and might not permit an equilibrium. Every time an MP changes party affiliation, there is a reallocation—often subtle, but not necessarily so—of bargaining power across parties and even across individual MPs. This reallocation changes the bargaining context, wherein rank-and-file MPs and party leaders alike reevaluate their situations and their options (cf., Filippov, Ordeshook, and Shvetsova 2004). Moreover, as Mershon and Shvetsova point out in chapter 8 (see also Mershon and Shvetsova 2008), the incentives facing sitting MPs are themselves dynamic over the life of a legislature, so at some level MPs have to reevaluate their situations even in the absence of visible changes in the party system.
2.2.2 Ambition We hold as a strong maintained hypothesis the notion that politicians of all stripes are motivated by the desire to gain or retain office. The logic underpinning this claim, so ubiquitous in political science that it might be held as a law of politics, is based on the assumption that politicians seek office to achieve things that they could not achieve—or could not achieve to equal effect—out of office. The goal could be anything, including but probably not limited to a yen for public service, a drive for power, a vision of “good” public policy, or a profitable career. The key is that these are goals best achieved in office and, in some cases, unattainable out of office.3 A politician’s desire to gain and hold office is, in short, a simple consequence of what she seeks to achieve. It is at once both product and tool of, in a word, ambition. Even the MP who seeks only to hold a simple, unremarkable legislative seat, we suspect, does so not because he or she wants the seat per se, but rather it is the most attractive career option open to him or her. As we see it, where there are goals, there is ambition—that is, the drive to attain, or at least to seek to attain, those goals. From this perspective, choices to change party affiliation are decisions about how best to realize ambitions (or, less propitiously, how best to avoid seeing goals thwarted and ambitions dashed).
34
William B. Heller and Carol Mershon
One of our basic premises is that politicians, legislators included, are ambitious. Their ambitions are personal, and while they might overlap with and be served by the ambitions of their party leaders and other copartisans, it would be a mistake to equate an MP’s goals with those of his or her party. We assume as well that individuals have basic preferences over outcomes in both policy and politics.4 Individuals can act strategically to achieve the best possible feasible outcome,5 but they also pay a cost for acting against their raw (i.e., untinged by strategic or practical considerations) preferences. That is, acting strategically is on one hand psychologically or intellectually more difficult—hence, more costly—than acting naively. On the other hand, to the extent that an MP is beholden to an audience of voters who does not appreciate the complexities of strategic decision making, strategic choices might appear to represent an abandonment of basic principles or interests, for which the unsophisticated audience might impose punishment. (As Nokken suggests in this volume, legislative procedures such as those in the US Congress might allow politicians to behave strategically while appearing to be both consistent and sincere.)
2.2.3 Party unity In the quest to realize their ambitions, MPs (and their parties and party leaders) face tradeoffs. Primary among these is the tension between party unity and each individual’s desire to vote against the party line, whether because his or her goals conflict with those of the party on some issue, or because the desire to vote according to his or her raw preferences is particularly strong. For MPs whose goals are close to each other and to those of their party leaders, instances of conflict should be relatively few; for MPs farther from the center of the party, they should be more common. Because policy making is multidimensional, however, and a party’s policy positions to some extent represent an aggregation of its members’ policy preferences (see Heller and Mershon this volume, chapter 7), it is unlikely that any one legislator would agree with the party on all issues. Consequently, maintaining party unity might be difficult. Party unity is important because party labels are important. Party labels give voters information-laden cues about candidates (Snyder and Ting 2002), but the amount of information they can transmit depends in part on party unity. When a candidate bears the label of a given party, the label tells voters how the candidate is likely to vote
Theoretical and Empirical Models
35
across a wide range of issues; the more unified are members of the party in legislative voting, the clearer is the information conveyed. A voter who cares about party programs might want to vote for the candidate or party whose program he or she likes best, for example.6 To the extent that party programs are unclear, so that the party label imparts little information, voters will have a hard time choosing among them. In essence, the less meaningful party labels are, the more voters who care only about programs must vote either by coin flip or not at all. Parties can uphold unity in three ways: negative agenda control, cohesion, and discipline. Negative agenda control—the party leadership’s ability to keep unwanted bills off the legislative calendar—is effective, albeit hard to observe. In the US House, for example, Rohde (1991; see also, Campbell, Cox, and McCubbins 2001; Cox and McCubbins 2001; 2005) suggests that the majority party seeks to ensure that the House never considers issues that would split the party. Party cohesion jibes with the idea of a party as a group of like-minded individuals joined for a common purpose (e.g., Bowler, Farrell, and Katz 1999; Krehbiel 1993; Rasch 1999). Even like-minded individuals disagree from time to time, however, and some purposes are more common than others. When cohesion fails, unity can be maintained through party discipline, which imposes costs on MPs who fail to toe the party line. Party leaders control an array of disciplinary tools, including but not limited to ballot access (or the right to use the party label; Cox and McCubbins 1994), committee positions (D’Onofrio 1979; Kiewiet and McCubbins 1991), advancement within the party and, by extension, increased influence over party policy positions and access to perks of legislative office (Bowler, Farrell, and Katz 1999). Inasmuch as party labels are important (which in turn depends on how voters view candidates; see Cox 1987; 1997), the ideal situation for an individual MP is to be in a highly unified party that does not constrain his or her own voting behavior. Accepting the premise that all members of a party are likely to take exception to the party’s chosen position on some issues at least some of the time, this means that every MP would like to be able to vote against the party position, without consequences, even as all his or her copartisans toe the party line. This is of course the basis for a fundamental collective dilemma and is generally untenable. On any given vote, each MP’s incentives to deviate from the party line depend on institutions as well as her personal preferences. Institutions influence how much voters care about individual MPs
36
William B. Heller and Carol Mershon
(as opposed to parties) and their voting habits—and, as a result, how keenly individual MPs feel the need to stick to certain positions that might bring them into conflict with their party. And institutions are central to leaders’ ability to punish members of their rank and file who buck the party. Rules that make party switching relatively difficult, as for example in Spain (Mershon and Heller 2003; Sánchez de Dios 1999) can make party switching less common than it might otherwise be, though as Miskin (2003) points out, such rules do not necessarily fulfill their intended normative objectives. Rules that allow party leaders to determine who gets on the ballot and who does not— and where on the party list a candidate lies—give party leaders overweening power to determine MPs’ political careers (Gallagher and Marsh 1988). The wider the array of tools available to party leaders for disciplining their rank and file, in general, the fewer incentives individual MPs have to buck their party line (because it is likely to be more costly in the end). At the end of the day, discipline is the credible threat of punishment. It works when party leaders can block or at least hamper the pursuit of one or more of the aims routinely attributed to legislators: policy, office (e.g., a cabinet post) and legislative perks, and reelection (Müller and Strøm 1999; Strøm 1990). The tools of discipline derive from the party’s control over necessary steps along the paths of MPs’ ambition. Leaders’ ability and willingness to impose discipline on their party rank and file are fundamental for party unity. The more MPs find themselves forced either to vote against their preference (whether strategic or naïve), however, the more they are likely to resent the constraints of discipline. The greater their resentment, the greater their incentive to search for and take advantage of alternatives that offer a better net benefit.
2.2.4
The link between ambition and discipline
From this vantage point, it becomes clear that ambition and discipline are closely linked, and that in combination they play strongly into an MP’s consideration of the possibility of changing party affiliation. Other scholars have honed in on the importance for switching of ambition: McElroy’s (2003) analysis is explicitly ambition theoretic, for instance, and Desposato’s (2006) assessment of the impact of electoral, institutional, ideological, and patronage concerns on legislators’ propensity to change parties hinges on legislator ambition. Desposato finds, consistent with Aldrich and Bianco (1992), that
Theoretical and Empirical Models
37
switchers aim to enhance their prospects both for reelection and legislative influence. Other studies suggest ambition as the driving force behind switching. Turan (1985, 32), for instance, finds that switchers in Turkey seem to serve longer in the legislature and to “accomplish other goals,” while Sánchez de Dios (1999; see also, Àgh 1999; Tomás Mallén 2002) views switching as a means of jockeying for position and political influence. Studies that show that legislators who change parties also change the way they vote (e.g., Ansolabehere, Synder, and Stewart 2001; McCarty, Poole, and Rosenthal 2001; Nokken 2000) implicitly consider ambition, since party-induced changes in voting behavior likely are linked to party discipline, and discipline works because legislators are ambitious. Ambition is a quality possessed by individuals. Discipline is a threat wielded by party leaders who possess the wherewithal to thwart individual MPs’ ambitions. Individual politicians’ ambitions turn on holding on to office, influencing policy making, and advancing their political careers (on the primacy of these goals, see Müller and Strøm 1999; Strøm 1990). Party leaders, for their part, seek to maximize the integrity of the party label, party influence on coalition formation, and party influence in policy formation; and they use discipline to attain those ends. Leaders’ goals, which are to some extent incompatible—involvement in coalition bargaining implies accepting close identification with other parties, for instance, which erodes the integrity and uniqueness of the party label—converge with leaders’ personal desires, which match those of any other MP but are easier to realize in the leadership than in the party rank and file (see Cox and McCubbins 1993, Ch. 9). Taking discipline and ambition together, analysis is in principle straightforward. Leaders simply need to impose enough discipline to generate strong party unity; their only constraint is the costs of monitoring their party’s members and expending whatever resources are required to exact punishment. In equilibrium, if leaders calculate correctly, we should see MPs defecting from the party line only very rarely, and only when they gain more from defecting than they lose under the lash of discipline. The possibility that an MP might move to another party adds a new dimension of complexity: absent the possibility of switching, MPs and their leaders engage in simple cost-benefit analyses; if switching is possible, their analyses become strategic. A leader has to ask himself or herself whether a given level of discipline will induce some members to leave the party, and if so what consequences that will have for achieving their party’s goals;
38
William B. Heller and Carol Mershon
an MP might consider or appear to consider switching in the face of harsh discipline at least in part in an effort to convince his or her leaders to make discipline less onerous. Ultimately, the greater the role of discipline in determining MP behavior, the net benefits an MP realizes (i.e., considering the costs he or she pays) from membership in a party, the less attractive is that party to him or her. The lower the attractiveness of the party, the greater the probability not only of voting against the party line but also of abandoning the party altogether (cf. Heller and Mershon 2008). Party switching thus arises at least in part out of the interaction of leaders and rank-and-file MPs. We now turn to characterize that interaction as a game.
2.3
The Legislative Switching Game
As strategic interaction between potential switchers and their party leaders, party switching occurs in a legislative context defined by the legislative party system, on one hand, and, on the other hand, an electoral context. Voters look at politicians and parties and legislative outcomes as they choose how to vote (strategically or naively), and the results at the polls then feed back into the legislative party system. The legislative party system in turn sets the stage for strategic interaction between potential switchers and party leaders. And so on: one round of stages forms the foundation for the next. The key is that party switching is one of the tools available to legislators who want to maximize their political fortunes, their political influence, and their ability to achieve their policy goals. Legislators choose to switch in reaction to decisions made by party leaders and in anticipation of voters’ behavior the next time they go to the polls. In this light, we portray legislative switching as part of a four-stage game, as shown in Figure 2.1. In the first stage, Nature exogenously and unstrategically sets the key parameters of the game (cf. Laver and Benoit 2003) by choosing a legislative party system. Strategic behavior begins at the second stage, when parties (more accurately, party leaders) choose a level of discipline to impose on their members. The level of discipline, along with party ideology and legislative bargaining weight, is a key element of a party’s attractiveness to potential switchers. The parties’ choices at the second stage in effect are designed to condition switching decisions in the third stage. The legislator’s decision is at heart a simple one, but it is complicated by multiple and sometimes conflicting electoral, office (or career), and policy
Theoretical and Empirical Models Nature chooses legislative party system * number of parties * party weights (seat shares) * party ideologies 1
Legislators choose to switch or not to maximize * probability reelection * policy goals * career goals 3
Party leaders set discipline to maximize * integrity of party label * influence in coalition negotiations * influence in policy formation 2 Figure 2.1
39
Voters vote to minimize distance between own ideal points and expected policy outcomes 4
The four stages of the switching game.
goals. These objectives are facets of political ambition, which drives politicians of all stripes, whether party leaders or faces in the legislative crowd, to use whatever tools are available to them to further their own interests. It is thus ambition that drives both parties’ decisions on discipline and legislators’ decisions on party affiliation. Voters enter the game at the final stage. Concerned only about policy outcomes, they vote on the basis of their perceptions of parties’ policy preferences and legislative bargaining weights.7 They are sophisticated in the sense that they implicitly consider how their votes affect each party’s input in policy making (i.e., the politics of legislative coalitions, which we do not model here). As players in the final stage of the switching game, however, voters register their sincere preferences, given expectations about each party’s influence and the value of their own votes. We present this sequence as a framework, rather than as fully specified game. In the manner of game-theoretic analysis in search of subgame-perfect equilibria, however, we begin at the end and work backward. We assume that the players in the game have perfect information, and that each player’s ideal point is common knowledge (and does not change). The usual spatial-modeling assumptions apply.
2.3.1 Stage four: Voters decide At the end of the day, it is voters who decide how parties and their candidates fare at the polls. We assume that voters care most strongly
40
William B. Heller and Carol Mershon
about policy outcomes. (As a practical matter, we suspect that even voters who evince a strong preference for types or qualities of process—e.g., “fairness” or “democracy”—would do so only as long as decision-making processes yield outcomes that do not consistently damage their interests.) How a voter decides—for example, whether he or she simply votes naively for the party whose ideal point is closest to her own, or backs the party he or she believes most likely to move policy outcomes closer to his or her own ideal point—is beyond the scope of this project. In any case, if party switching affects party positions as well as their seat shares, voter decisions should take switching into account.8
2.3.2 Stage three: Legislators decide Whether legislators decide to switch parties depends fundamentally on what they want to achieve. Scholars generally agree that parties want some combination of votes, office, and policy (Strøm 1990; Müller and Strøm 1999; and see in particular Strøm 1994). Parties are not qua parties reasoning entities, and while for some analytical purposes it might make sense to assume that they are (as in the field of coalition studies, cf. Laver and Schofield 1990), this is not the case here. Parties ultimately are vehicles for their members, and it is vital to keep in mind that those members, not the parties themselves, want to win votes, occupy office, and make policy (Aldrich 1995; McDonald 2004, 14). To analyze legislators’ party-affiliation choices, we begin with the assumption that legislators (and politicians in general) are ambitious, as already emphasized. Ambition dictates that they seek the means to their desired ends that yields the highest expected benefit; and they have chosen political careers because they can best realize their goals in the political arena. The reelection imperative so often attributed to politicians follows directly from this assumption. We can only conjecture about the goals that political ambition serves, but two possibilities stand out: power (for its own sake or the sake of the perks of office) and policy. Both lead politicians to seek increasing policy influence and, we believe, motivate sitting legislators’ choices of party affiliation. A sitting legislator’s decision whether to change his or her partyaffiliation status unfolds in a process similar to economic models of job searches. Where job search models focus on monetary remuneration, however, we highlight the expected utility of choice of
Theoretical and Empirical Models
41
party affiliation, where rewards come in the currency of power and policy. Analysis is complicated by several factors, of which we flag two: 1. the legislator in question’s state before taking action (i.e., a legislator can be in some party or, possibly, in no party9); and 2. the possible range of alternatives available to a legislator, which might be limited by rule-bound constraints on switching or formation of legislative party groups. Those alternatives include a. staying put; b. switching to a different, established party; c. switching to or participating in the formation of a new party; d. switching to independent status (if not independent already); and e. exiting the legislature, whether to pursue new options outside politics or to continue a political career in a different capacity, for example, to take subnational or supranational office.
As noted earlier, the legislator’s objective function consists of three components, each playing on different aspects of ambition. The first and most basic is votes or, more accurately, continuance in the legislature. This depends on two things: nomination (including list position for party-list ballots), controlled by local or national party leaders or party activists, and party electoral strength, which we take as (expected) vote share. The second is offices controlled by the legislature, for which it is best to belong to a party in the legislative majority or with a good chance of getting into the legislative majority. The third is policy. Policy considerations are complicated by several factors, including 1. The need to be part of a majority coalition to make policy, which makes party membership attractive and party unity desirable (Laver and Shepsle 1999; see also, Cox 1987). We assume, in line with these observations, that membership in a large party is more desirable than membership in a small party, all else constant. 2. The likelihood that a legislator’s personal policy preferences will diverge somewhat from those supported by his or her party, which suggests a. Legislators sometimes would like to vote against their parties. b. Notwithstanding the desire to break ranks with his or her party, each legislator should value party unity as a collective good for members of the party; that is, membership in a party with relatively high observed unity should be more valuable than membership in a party with less observed unity, all else equal.
42
William B. Heller and Carol Mershon 3. A legislator who is a member of a party is better able than a legislator not in the party to move that party’s observed ideal point toward his or her own. Even so: a. A party’s ability to influence policy outcomes and a legislator’s likely ability to influence the party’s observed preferences are independent (or perhaps inversely related). b. There might well be tradeoffs between a legislator’s ability to influence policy outcomes (as a member of an influential party) or enjoy enhanced career opportunities and his or her desire to be in a party whose members’ policy preferences converge with his or her own (McDonald 2004, 14).
A legislator’s calculation of the value of staying where he or she is, whether as a party member or a lone independent, depends on how he or she evaluates his or her own situation and the available alternatives to it. The criteria for evaluating options depend on both long- and short-term concerns. In the long term, legislators care about positioning themselves for reelection and so worry about both their party’s chances and their chances within the party. In the short term—that is, during the legislature and between elections—legislators care about policy and office. These two considerations are inseparable much of the time: holding legislative office yields policy influence, and a legislator with policy influence (e.g., as a leading member of a party whose size or position makes it pivotal or nearly so) is an attractive candidate for office. Party unity is a collective good for party members, not least because it gives parties leverage for bargaining over policy (Laver and Shepsle 1999). It also contributes to voters’ ability to distinguish among parties, and it underpins voter confidence in the solidity of party platforms. As a collective good, however, it is likely to be undersupplied absent some mechanism—such as disciplinary threats wielded by party leaders—to motivate members to vote with their party but against their personal preferences.10 The final element of the switching calculus, therefore, is party discipline, which acts on an MP’s policy influence when party leaders can block or roll back an MP’s advancement within the party, withhold the benefits of the party label (i.e., expel the MP from the party, see, e.g., Cox and McCubbins 1994), or affect the MP’s possibilities for reelection through control over ballot access and list placement (cf. Gallagher and Marsh 1988). It is of course difficult to distinguish between behavior motivated by individual preferences and that motivated by partisan pressures (cf. Heller and Mershon 2008). Party switching offers leverage on the
Theoretical and Empirical Models
43
question; Timothy Nokken and Scott Desposato, in their contributions to this volume, focus on party switchers to examine the extent (and form) of party influence on individual legislators’ behavior. The range of disciplinary instruments available to party leaders will vary by both leader and party. The range of disciplinable offenses against the party also vary by party, by party leader, and—as we will discuss in relation to party strategies in the face of potential switching—by legislator. In all cases, however, discipline keeps in line MPs who might otherwise break ranks with their party; discipline works in this way by threatening to impede their ability to realize their ambitions. Legislators face two kinds of discipline. On one hand, there is discipline designed to induce desired behavior on the part of legislators. On the other hand, there is discipline designed to ensure that the kinds of people who advance within the party are acceptable to party leaders. The former consists of punishments (or threats thereof); the latter consists of obstacles to advancement. The two types of discipline are in principle separate, though they can easily fold together, and we suspect that legislators prefer that they be closely linked.11 Behavioral discipline acts both on MPs who vote against the party and so incur high costs from discipline unleashed, and on MPs who would prefer to vote against the party but do not because of fear of discipline. Every time an MP votes contrary to his or her personal preferences he or she pays a cost in benefits foregone equivalent to the difference in his or her utilities over the alternative he or she prefers versus the alternative his or her party prefers. The elements of the legislator’s choice are varied and, in many cases, dynamic. If larger parties are generically more attractive than smaller ones, for example, then every time one MP switches all other MPs implicitly have to reevaluate their own prospects. An MP will consider changing his or her state—for example, switching into a new party, exiting the legislature, or becoming an unaffiliated legislator (legislative or constitutional rules permitting)—only when he or she expects to benefit more from moving than from staying put. If legislators can easily compare their expected utility from different states with their current one—and we believe that in the confines of a legislature they can—then searching for alternatives, like applying for a new job in some professions, should be essentially costless (Hey and McKenna 1979). Changing states, however, is not costless: first, most legislators who move risk losing some or all of their investments in resources specific to the state from which they move, resources
44
William B. Heller and Carol Mershon
such as seniority, intraparty relationships, and constituent support; second, switchers muddy their reputations with colleagues, possible future colleagues (in or out of the legislature), and voters (Desposato 2006; Hey and McKenna 1979). The general point is that whether and when an MP switches depend on his or her expectations and hence on the information available to him or her.
2.3.3 Stage two: Parties decide All strategic play in the legislative switching game begins at the second stage, after Nature’s initial move. At the second stage, each party sets the parameters and average overall levels of discipline, that apply to its members. As noted, parties are collectivities of individual politicians. It is party members, not parties themselves, who are ambitious and who benefit when their party can influence policy and attract votes (Aldrich 1995, 4). Parties can do neither, however, if they cannot overcome the collective action problems endemic to enterprises that require distinct individuals with disparate interests to work together for their common good. For that, they have party leaders (Luebbert 1986). “Party” decisions are in actuality party leaders’ decisions. Leaders are constrained by checks in the decision-making process (Heller, Keefer, and McCubbins 1998; Strøm 2000) the need to balance keeping their rank and file reasonably happy (and so retain control of the leadership), the need to attract voters, and the need to bargain over policy with counterparts in other parties (see Cox and McCubbins 1993). They seek to maximize their own ambitions with regard to reelection, office, and policy influence; to that end, they need jointly to maximize the party’s ability to influence policy, which in turn is closely related to its ability to win votes at the next election and its observed coherence (in legislative voting and in the behavior, pronouncements, and reputations of its members). To the extent that the party succeeds in these terms, its members—and in particular its leaders—benefit. Where decision-making power lies in a party can be hard to identify. Rules might specify who holds specific agenda powers, but if such powers are “nested” in relationships defined by other sets of rules (Tsebelis 1990), including norms or tradition (Greif 2006), then the exercise of formal agenda powers is in itself strategic (cf. Heller 2001; 2007). Heller and Mershon address this issue with respect to the question of how party positions are set in chapter 7 of this volume, leveraging switching-induced changes in party composition to
Theoretical and Empirical Models
45
see how the preferences of the parties’ legislative rank and file affect party policy positions. There is an added benefit to party leadership, in that leaders can use their influence in the party to move its observed ideal point closer to their own. To that end, they can seek to attract new members who move the party in the desired direction and they can apply harsh discipline (even to the point of expulsion from the party) for members who impede the party’s movement in the direction their leader wants to take it (Heller and Mershon 2008). Discipline, broadly construed as a combination of opportunities for advancement and threats of punishment, serves both these purposes. At the second stage, therefore, the party leadership has to set both behavioral and career discipline to (a) induce desired behavior, present a unified face to voters, and maintain as strong a position as possible in bargaining over policy with other legislative parties; (b) retain party members and possibly attract switchers; and (c) move the party’s observed ideal point as close as possible to the leadership ideal point. Whatever its level and however strictly it is enforced, discipline plays directly on legislators’ ambitions. The threat of punishment for failing to behave according to criteria set by party leaders motivates legislators to act in the party’s (or party leader’s) interests, even when they would rather not, to achieve longer-term career goals. For parties that are concerned with retaining members and attracting switchers as well as maintaining party unity, at least part of the decision on setting discipline is conditioned by the distribution of voter preferences and the locations and weights of other parties. Extremist parties, for example, can impose strong discipline on their members because, first, other parties are too distant to make switching likely and, second, the lack of nearby competitors means that they do not have to blur their labels to attract wavering voters (cf. Shepsle 1972). The focus on party switching spotlights the basic fact that parties are made up of individuals. This observation, while in itself neither controversial nor new, tends to be treated as an afterthought in studies of party behavior. By putting party switching front and center and treating it as a phenomenon of theoretical interest, the contributors to this book have gained—and offer to other analysts—new perspective not only on individual legislators’ behavior, but also on a broad range of political action, from party discipline to party systems. The choices that parties and their members make define party systems and, more generally, both the context and the content of political decision making. The contributions of Norman Schofield (chapter 3)
46
William B. Heller and Carol Mershon
and Junko Kato and Kentaro Yamamoto (chapter 9), for example, take the interaction of party choices with voter choices (Schofield) or legislator ambition (Kato and Yamamoto) to characterize the evolution of party systems. Along similar lines, Kenneth Benoit and Gail McElroy (chapter 6) use party switching in the European Parliament to look at the development of a supranational political space. Their work suggests—contra the conventional wisdom that European voters pay little attention to European Union politics and policy making when they go to the polls—that the day might not be so far off when local, national, and supranational questions all impinge strongly on European voters’ voting behavior in every election, much as local, state, and federal questions affect vote choices in the United States.
2.4 Conclusion We characterize party switching as an aspect of strategic interaction among legislators. In so doing, we embed legislators’ decisions on party affiliation in a framework that ties them to the decisions both of party leaders and of voters, which are in turn contingent on the legislative party system. We also emphasize the effectiveness—indeed, the necessity—of focusing on individual actors to understand this larger picture. We do not examine party switching to understand it alone. True, party switching is interesting in itself, normatively, substantively, and theoretically, all the more so because it is not the idiosyncratic anomaly it has often been seen to be. More important, it affords new and powerful leverage on questions that have long attracted the attention of political scientists interested in such diverse areas as voting behavior, legislative organization, and party systems. The components of the framework presented in this chapter can be unpacked and applied in diverse ways by other theoretical and empirical researchers, as demonstrated by the analyses in the chapters that follow. Informed by the framework in this chapter, the contributors to this volume shed new light on the relationships among voters, legislators, party leaders, and the legislative party system.
2.5
Notes
1. Some moves are not voluntary, as discussed in the introduction. 2. The importance of parties goes beyond their utility to their individual members, of course. As students of collective dilemmas have shown, however, members of collectivities have little incentive to take the good of the
Theoretical and Empirical Models
3.
4.
5.
6.
7.
8.
9.
10.
11.
47
collectivity into account unless their behavior as individuals contributes directly to that good and unless they realize positive benefits as individuals from it (Olson 1965; Ostrom 1990). We recognize that there might be reasons for people who do not in fact want to hold office to present themselves as candidates. The publicity afforded by a political campaign could be useful for any number of reasons, for example. Those who are motivated by these non-office benefits of running are likely to be scarce in the universe of sitting legislators, however, and so we ignore them and their motivations here. Ultimately, we ascribe to the belief that preferences over political outcomes derive from preferences over policy outcomes. We do not wish to belabor the point, however, and allowing for preferences over both does not affect our argument. A strategic individual is one who makes decisions with an eye to outcomes, rather than just the choices immediately available, and in full awareness that other individuals are doing the same. In many electoral systems, party control of nomination processes and of candidate rankings on the ballot induces voters to view parties, not individual candidates, as the entities they elect. The information available to voters, however, varies across national settings and in particular across established and new democracies (cf. Cox and McCubbins 1997; Mershon and Shvetsova 2008; this volume; Moser 2001) with consequences for legislators’ and party leaders’ behavior that we explore. We could complicate things greatly by including voter considerations of coalition games. We do not—although we do think that in some circumstances voters might actually think in those terms (or at least parties do, as evidence from Germany on voting for FDP suggests; Riker 1982; Anderson 1995). In some legislatures, members not formally affiliated with a party group belong to a formally established “mixed group”; in others, they have the status of independents, which could be disadvantageous. In Norway, for example, independents receive only half the nonsalary budget resources that they would get as members of a party (personal communication from Hans Brattestå, Secretary General of the Norwegian Storting, May 26, 2004). Personal preferences might derive from constituencies, particularly in single member district (SMD) or proportional representation (PR) systems with strong local nominating authorities. For our purposes, the difference between such electorally driven preferences and a legislator’s “own” preferences is simply that in most cases electorally driven preferences probably are harder for a party to suppress. The two types of discipline are linked in one way, inasmuch as punishments for breaking party discipline are meted out in terms of blocked ambition. Good behavior is not necessarily a guarantee of good career prospects, however.
48
William B. Heller and Carol Mershon
2.6 References Àgh, Attila. 1999. “The Parliamentarization of the East Central European Parties: Party Discipline in the Hungarian Parliament, 1990–1996.” In Party Discipline and Parliamentary Government, ed. Shaun Bowler, David M. Farrell, and Richard S. Katz, 167–188. Columbus: Ohio State University Press. Aldrich, John H. 1995. Why Parties? The Origin and Transformation of Political Parties in America. Chicago: University of Chicago Press. Aldrich, John H., and William T. Bianco. 1992. “A Game-Theoretic Model of Party Affiliation of Candidates and Office Holders.” Mathematical and Computer Modelling 16 (8/9): 103–116. Anderson, Christopher. 1995. Blaming the Government: Citizens and the Economy in Five European Democracies. Armonk, NY: M.E. Sharpe. Ansolabehere, Stephen, James M. Snyder, Jr., and Charles Stewart III. 2001. “The Effects of Party and Preferences on Congressional Roll Call Voting.” Legislative Studies Quarterly 26 (4): 533–572. Bowler, Shaun, David M. Farrell, and Richard Katz. 1999. “Party Cohesion, Party Discipline, and Parliaments.” In Party Discipline and Parliamentary Government, ed. Shaun Bowler, David M. Farrell, and Richard S. Katz, 3–22. Columbus: Ohio State University Press. Campbell, Andrea C., Gary W. Cox, and Mathew D. McCubbins. 2001. “Agenda Power in the U.S. Senate, 1877 to 1986.” In Party, Process, and Political Change: New Perspectives on the History of Congress, ed. David Brady and Mathew D. McCubbins, 146–165. Stanford, CA: Stanford University Press. Cox, Gary W. 1987. The Efficient Secret: The Cabinet and the Development of Political Parties in Victorian England. Cambridge: Cambridge University Press. ———. 1997. Making Votes Count: Strategic Coordination in the World’s Electoral Systems. Cambridge: Cambridge University Press. Cox, Gary W., and Mathew D. McCubbins. 1993. Legislative Leviathan: Party Government in the House. Berkeley: University of California Press. ———. 1994. “Bonding, Structure, and the Stability of Political Parties: Party Government in the House.” Legislative Studies Quarterly 19 (2): 215–231. ———. 1997. “Political Structure and Economic Policy: The Institutional Determinants of Policy Outcomes.” In Political Institutions and the Determinants of Public Policy: When Do Institutions Matter? ed. Stephen Haggard and Mathew D. McCubbins. San Diego: University of California. ———. 2001. “Agenda Power in the House of Representatives.” In Party, Process, and Political Change: New Perspectives on the History of Congress, ed. David Brady and Mathew D. McCubbins, 107–145. Stanford, CA: Stanford University Press. ———. 2005. Setting the Agenda: Responsible Party Government in the U.S. House of Representatives. Cambridge: Cambridge University Press. Desposato, Scott W. 2006. “Parties for Rent? Ambition, Ideology, and Party Switching in Brazil’s Chamber of Deputies.” American Journal of Political Science 50 (1): 62–80.
Theoretical and Empirical Models
49
D’Onofrio, Francesco. 1979. “Committees in the Italian Parliament.” In Committees in Legislatures, ed. John David Lees and Malcolm T. Shaw, 61–101. Durham, NC: Duke University Press. Filippov, Mikhail, Peter C. Ordeshook, and Olga V. Shvetsova. 2004. Designing Federalism: A Theory of Self-Sustainable Federal Institutions. Cambridge: Cambridge University Press. Gallagher, Michael, and Michael Marsh, eds. 1988. Candidate Selection in Comparative Perspective: The Secret Garden of Politics. London: Sage. Greif, Avner. 2006. Institutions and the Path to the Modern Economy: Lessons from Medieval Trade. Cambridge: Cambridge University Press. Heller, William B. 2001. “Political Denials: The Policy Effect of Intercameral Partisan Differences in Bicameral Parliamentary Systems.” Journal of Law, Economics, and Organization 17 (1): 34–61. ———. 2007. “Divided Politics: Bicameralism, Parties, and Policy in Democratic Legislatures.” Annual Review of Political Science 10: 245–269. Heller, William B., and Carol Mershon. 2005. “Party Switching in the Italian Chamber of Deputies, 1996–2001.” Journal of Politics 67 (2): 536–559. ———. 2008. “Dealing in Discipline: Party Switching and Legislative Voting in the Italian Chamber of Deputies, 1988–2000.” American Journal of Political Science 52 (4): 910–925. Heller, William B., Philip Keefer, and Mathew D. McCubbins. 1998. “Political Structure and Economic Liberalization: Cases from the Developing World.” In The Origins of Liberty: Political and Economic Liberalization in the Modern World, ed. Paul W. Drake and Mathew D. McCubbins, 146–178. Princeton, NJ: Princeton University Press. Hey, John D., and Chris J. McKenna. 1979. “To Move or Not to Move?” Economica 46 (182): 175–185. Kiewiet, D. Roderick, and Mathew D. McCubbins. 1991. The Logic of Delegation: Congressional Parties and the Appropriations Process. Chicago: University of Chicago Press. Krehbiel, Keith. 1993. “Where’s the Party?” British Journal of Political Science 23: 235–266. Laver, Michael, and Kenneth A. Shepsle. 1999. “How Political Parties Emerged from the Primeval Slime: Party Cohesion, Party Discipline, and the Formation of Governments.” In Party Discipline and Parliamentary Government, ed. Shaun Bowler, David M. Farrell, and Richard S. Katz, 23–48. Columbus: Ohio State University Press. Laver, Michael, and Kenneth Benoit. 2003. “The Evolution of Party Systems between Elections.” American Journal of Political Science 47 (2, April): 215–233. Laver, Michael, and Norman Schofield. 1990. Multiparty Government: The Politics of Coalition in Europe. Oxford: Oxford University Press. Luebbert, Gregory. 1986. Comparative Democracy: Policy-Making and Government Coalitions in Europe and Israel. New York: Columbia University Press. McCarty, Nolan M., Keith T. Poole, and Howard Rosenthal. 2001. “The Hunt for Party Discipline in Congress.” American Political Science Review 95 (3): 673–687.
50
William B. Heller and Carol Mershon
McDonald, Michael D. 2004. “Parties in Democracy, Democracy in Parties: Lessons I Learned from Ian Budge and the CMP Data.” Binghamton. Mershon, Carol, and Olga Shvetsova. 2008. “Parliamentary Cycles and Party Switching in Legislatures.” Comparative Political Studies 41 (1): 99–127. Mershon, Carol, and William B. Heller. 2003. “Party Switching and Political Careers in the Spanish Congress of Deputies, 1982–1996.” Paper presented at the annual meetings of the Midwest Political Science Association, Chicago, IL, April 3–6. Miskin, Sarah. 2003. “Politician Overboard: Jumping the Party Ship.” Unpublished paper. Parliament of Australia: Parliamentary Library—Politics and Public Administration Group. Moser, Robert G. 2001. Unexpected Outcomes: Electoral Systems, Political Parties, and Representation in Russia. Pittsburgh: University of Pittsburgh Press. Müller, Wolfgang C., and Kaare Strøm, eds. 1999. Policy, Office, or Votes: How Political Parties in Western Europe Make Hard Decisions. Cambridge: Cambridge University Press. Nokken, Timothy P. 2000. “Dynamics of Congressional Loyalty: Party Defection and Roll Call Behavior, 1947–1997.” Legislative Studies Quarterly 25 (3): 414–444. Olson, Mancur. 1965. The Logic of Collective Action: Public Goods and the Theory of Groups. Cambridge, MA: Harvard University Press. Ostrom, Elinor. 1990. Governing the Commons: The Evolution of Institutions for Collective Action. Edited by J. E. Alt and D. C. North. Cambridge: Cambridge University Press. Rasch, Bjørn Erik. 1999. “Electoral Systems, Parliamentary Committees, and Party Discipline: The Norwegian Storting in a Comparative Perspective.” In Party Discipline and Parliamentary Government, ed. Shaun Bowler, David M. Farrell, and Richard S. Katz, 121–140. Columbus: Ohio State University Press. Riker, William H. 1982. “The Two-Party System and Duverger’s Law: An Essay on the History of Political Science.” American Political Science Review 76 (4): 753–766. Rohde, David W. 1991. Parties and Leaders in the Postreform House. Chicago: University of Chicago Press. Sánchez de Dios, Manuel. 1999. “Parliamentary Party Discipline in Spain.” In Party Discipline and Parliamentary Government, ed. Shaun Bowler, David M. Farrell, and Richard S. Katz, 141–162. Columbus: Ohio State University Press. Shepsle, Kenneth A. 1972. “The Strategy of Ambiguity: Uncertainty and Electoral Competition.” American Political Science Review 66 (2): 555–568. Snyder, James M., Jr., and Michael M. Ting. 2002. “An Informational Rationale for Political Parties.” American Journal of Political Science 46 (1): 90–110. Strøm, Kaare. 1990. “A Behavioral Theory of Competitive Political Parties.” American Journal of Political Science 34 (2, May): 565–598. ———. 1994. “The Presthus Debacle: Intraparty Politics and Bargaining Failure in Norway.” American Political Science Review 88 (1): 112–127.
Theoretical and Empirical Models
51
———. 2000. “Delegation and Accountability in Parliamentary Democracies.” European Journal of Political Research 37 (3): 261–289. Tomás Mallén, Beatriz. 2002. Transfuguismo parlamentario y democracia de partidos. Madrid: Centro de Estudios Políticos y Constitucionales. Tsebelis, George. 1990. Nested Games: Rational Choice in Comparative Politics. Berkeley: University of California Press. Turan, Ilter. 1985. “Changing Horses in Midstream: Party Changers in the Turkish National Assembly.” Legislative Studies Quarterly 10 (1): 21–34.
This page intentionally left blank
II
Party Switching and Representation
This page intentionally left blank
3 Switching Equilibria* Norman Schof ield
Abstract Models of elections typically indicate that all parties, in equilibrium, will adopt positions at the electoral center. This situation, however, is seldom observed. This chapter outlines how an integrated theory of party strategy may be constructed based on a theorem for existence of Local Nash Equilibrium (LNE) in a stochastic electoral model where political agents have differing valence, or electorally perceived quality. In a model based on exogenous valence, it is shown that, under proportional representation, a high valence political leader may be motivated to switch party so as to control a centrist policy position. Under plurality rule, political agents may be forced to switch party as a result of the conflicting demands of activist coalitions.
3.1
A Model of Electoral Competition and Party Switching
A fully developed formal theory of politics would connect the nature of the electoral system, the motivations of parties concerning policy and perquisites, and the process of government formation, in a way which makes sense of the empirical phenomena. A number of attempts have been made to model the motivations of parties in a game-theoretic framework. For example, one class of models is based on the Downsian framework, where parties compete via policy declarations to the electorate to maximize the number of seats they obtain. Since parties are assumed in these models to be indifferent to policy objectives, viewing policy solely as a means to gain power, symmetry would suggest that one equilibrium would be the situation where all parties declare the same position. In the standard spatial model
56
Norman Schofield
(Downs 1957; Hotelling 1929), convergence of this kind is a typical result. There is substantial evidence that convergence by all parties to an electoral center is an extremely unlikely phenomenon (Schofield and Sened 2006). Partly as a result of this theoretical difficulty with the formal electoral model, and also because of the need to develop empirical models of voter choice, this chapter focuses on a “stochastic” vote model. In such a model, the behavior of each voter is characterized by a vector of choice probabilities determined by the candidate positions (Hinich 1977; Lin, Enelow, and Dorussen 1999). Again, a standard result in the class of stochastic models is that all parties converge to the electoral origin when the parties are motivated to maximize vote share or plurality (in the two party case) (Banks and Duggan 2005; McKelvey and Patty 2006). Recent formal analysis (Schofield 2006; 2007) shows that the convergence result need not hold if there is an asymmetry in the electoral perception of the “quality” of the political agents (Stokes 1992). The average weight given to the perceived quality of a political agent is called the agent’s valence.1 In empirical models, adding valence to the model increases the statistical significance of the model (Schofield and Sened 2006). In these empirical models, valence is usually assumed to be exogenous, in the sense of being independent of the agent’s position. The early empirical work by Poole and Rosenthal (1984) on US presidential elections included these exogenous valence terms and noted that there was no evidence of candidate convergence. This chapter discusses this general model of elections based on the assumption that there are two kinds of valence. The first kind is an intrinsic or exogenous valence, which for a political agent j is denoted lj. As in empirical work, we assume that lj is held constant at the time of an election and so is independent of the agents’s current position. 2 The second kind of valence is known as activist valence. When agent j adopts a policy position zj, in the policy space, X, then the activist valence of the agent is denoted μj(zj). Implicitly, we adopt a model originally due to Aldrich (1983a; 1983b) and Aldrich and McGinnis (1989). In this model, activists provide crucial resources of time and money to their chosen political agent, and these resources are dependent on the agents policy position.3 The agent then uses these resources to enhance its image before the electorate, thus affecting its overall valence. Although activist valence is affected by agent position, it does not operate in the usual way by influencing voter choice through the distance between a voter’s preferred policy position,
Switching Equilibria
57
say xi, and the agent position. Rather, as agent j’s activist support, μj(zj), increases due to increased contributions to the agent in contrast to the support μk(zk) received by agent k, then (in the model) all voters become more likely to support agent j over agent k. 4 The problem for each political agent is that activists are likely to be more extreme than the typical voter. By choosing a policy position to maximize activist support, the agent will lose centrist voters. The agent must therefore determine the “optimal marginal condition” to maximize vote share. Theorem 1, presented in the following text, gives this as a (first order) balance condition. Moreover, because activist support is denominated in terms of time and money, it is reasonable to suppose that the activist function will exhibit decreasing returns. For example, in an extreme case, a political agent who has no activist support at all may benefit considerably by a small policy move to favor a particular interest group. On the other hand, when support is very substantial, then a small increase due to a policy move will little affect the electoral outcome. For this reason it is reasonable to assume that the functions themselves are concave, so their Hessians are everywhere negative-definite. Theorem 1 points out that when these functions are sufficiently concave, then the vote maximizing model will exhibit a Nash equilibrium. Because analysis of the stochastic model with activist valence is quite complex, this chapter also presents Theorem 2, for the case when activist valence is nonexistent. This theorem, valid for the case when only exogenous valence is relevant, gives the necessary and sufficient conditions for convergence of political agents to the electoral center. In applying these result to the question of the switching of the positions and party affiliations of political agents, Theorem 2 is used to examine the Israeli Knesset in the elections betweem 1996 and 2006. These elections can be considered paradigmatic examples of political positioning under an electoral system that is highly proportional. To understand these elections, we assume that parties adopt positions that are similar to their vote maximizing positions, as given by the stochastic vote model with exogenous valence. To further examine the nature of coalition bargaining, we use the idea of the core, presented earlier in Laver and Schofield (1990), to argue that a dominant party, located at the center of the policy space, can control the formation of government. In the postelection phase, the positions of the parties are assumed to be given, as is the distribution of seats. This distribution defines a set of winning coalitions. The compromise sets
58
Norman Schofield
of these various winning coalitions may intersect, to define a single policy point, the core. If the core is stable under small perturbations in the positions of the parties, then it is said to be structurally stable. If a party’s position is at the structurally stable core, then we shall call this party the core party. Under these circumstances, it is argued that the core party will dominate the formation of government. If the compromise sets do not intersect, then they will define a domain in the policy space, known as the heart. This model of the heart can then be used to characterize the general pattern of coalition bargaining and government formation. Because the coalition situation is very different depending on whether the core is empty or nonempty, it is argued that a party leader will have a very strong incentive to seek the central core, but only when the party leader has a very high valence. It is consistent with this argument that the policy move by Ariel Sharon, the leader of Likud in the Israel Knesset, in the months before the March 2006 election, was possible precisely because Sharon had such a high exogenous valence. To make such a policy move, Sharon had to switch out of the party, Likud, and create a new party called Kadima (“Forward”). It is implicit in this argument that political moves by other low valence parties would have had an insignificant effect on the political configuration. In contrast, I suggest that the role of activists in a plurality electoral system provides extensive scope, in principle, for party switching. To apply Theorem 1, first consider presidential elections in the United States. To illustrate the idea, consider figure 3.1, and suppose that the economic dimension alone is relevant for political policymaking. We assume that there is an electoral distribution of voter ideal points, whose mean we take as the electoral origin. If we ignore activism for the moment, then the results of the following section show that there are two very different possibilities, depending on the parameters of the model. If the exogenous valences, ldem and lrep, of the Democrat and Republican candidates are sufficiently similar, then both candidates will position themselves at the electoral origin, and both will gain approximately 50 percent of the vote. In this case, potential activists are unlikely to be motivated to contribute to the parties. On the other hand, if the valences differ sufficiently, with ldem > lrep, say, then the lower valence candidate will vacate the origin to increase vote share. For purposes of exposition, we may suppose that conservative economic activists have the preferred position, E. If the Republican candidate moves away from the origin, to a position
Switching Equilibria
59
Social Liberal Social Liberal indifference curve
B
E
R us h
D R*
nc e Bala
loc
Pro-Capital
fo us
ala er B t a w Gold
C
Economic conservative indifference curve
loc us
to n
D*
rB
Pro-Labor
Clin
L
John
so nB
S Contract curve between economic and locus social liberals ce an al locus nce a l a
e nc
Contract curve between economic and social conservatives
Social Conservative
Figure 3.1
Activist coalitions in the United States.
similar to R, then conservative economic activists would be induced to support this candidate. The asymmetry induced by this support will cause liberal economic activists at L to support the Democratic candidate. Then R will be pulled further toward E, while D will be pulled toward L. Moreover, if the marginal effect of activists for the Republicans is greater than for the Democrats, then the optimal candidate positions, R and D, will satisfy |R| > |D|. This model implies that once the convergent equilibrium is destroyed because of some exogenous change in parameters, and activists become motivated to support the appropriate parties, then convergence can never be recreated. Note that, in terms of the model, there is no reason why R should be to the right, and L to the left. However, once the move is made in one direction or the other, then activist support will tend to reinforce the left-right positioning of the parties. This simple marginal calculation becomes more interesting when there is a second “social” dimension of policy. Consider the initial positions R and D, on either side of, and approximately equidistant from the origin, as in the figure. Both Social Conservatives, represented by C, and Social Liberals represented by S, would be indifferent between both parties. A Democratic candidate by moving to position D* will
60
Norman Schofield
benefit from activist support by the social liberals, but will lose some support from the liberal economic activists. Note that the figure is based on the supposition that activists are characterized by ellipsoidal indifference contours, reflecting the different saliences (or emphases) they put on the policy axes. The contract curve between the two activist groups, centered at L and S, represents the set of conflicting interests or “bargains” that can be made between these two groups over the policy to be followed by the candidate. Miller and Schofield (2003) showed that this contract curve between the economic and social liberals is a catenary whose curvature is determined by the eccentricity of the utility functions of the activist groups (See also Schofield and Miller 2007; Miller and Schofield 2008). We therefore call this contract curve the Democratic activist catenary. In the same way, the Republican activist catenary is given by the contract curve between economic conservative activists, positioned at E, and social conservative activists, positioned at C. Theorem 1 gives the balance condition that a pair of positions, such as (D*, R*) in figure 3.1, constitute a Nash equilibrium for the vote maximizing game between the agents. These equilibrium positions will lie on what are called the balance locii of the agents. These balance locii will depend on the exogenous valences of the two candidates, as well as on the relative willingness of the various activist groups to provide support. For example, because the exogenous valence of Goldwater was much lower than Johnson in the 1964 election (Schofield, Miller, and Martin 2003), we can infer that the balance locus of Goldwater was further from the origin than that of Johnson (as illustrated in figure 3.1). The figure also presents plausible balance locii for Bush and Clinton in the elections of 2000 and 1992, respectively. This model of activist support for presidential candidates (Miller and Schofield 2003; 2007; Schofield and Miller 2007) provides a more complex model of candidate positioning than the more simple model with exogenous valence. It suggests that candidates must juggle the opposed interests of competing activist groups. In recent US elections, the issues that concern voters and activist groups have involved the so-called war against terror, immigration, stem cell research, and so forth. These crosscutting issues create the possibility that a candidate for congressional office will jump or switch party, when the candidate believes that the party is overly dominated by certain activist groups whose policy positions are in opposition to the candidate. For example, a traditional Republican, such as John Danforth of Missouri,
Switching Equilibria
61
with a policy position at E, may feel that his party is dominated by activists at C, possibly causing him to switch to a Democrat party that has adopted policies close to S. This model is applied in section 3.4 to discuss such recent examples of party switching by candidates in congressional and gubernatorial elections. The general conclusions that can be drawn from the formal models are (i) Under proportional rule, and with exogeneous valence alone, there will be a centripetal tendency toward the electoral center that will make itself felt only for political agents with high valence. In particular, if a party leader has a very high valence, such as Sharon in Israel, then this leader will have an incentive to leave his old party so as to be able to take up a dominant position at the core of the political configuration. Low valence party leaders, in contrast, will be subject to a centrifugal tendency toward the electoral periphery. In a situation where the core is empty, then policy repositioning by low valence party leaders will have a negligible effect on the political configuration. (ii) In an electoral system based on plurality rule, such as the United States, activist groups have a significant effect on candidate positioning. Policy positions adopted by candidates will involve compromises between the more extreme positions preferred by activists, and the centrist positions that win votes. In a two-dimensional context, candidates for office may be obliged to switch parties (and choose new policy positions) to gain activist support from groups that previously opposed the candidate. In particular, when policy concerns are polarized, and activist groups play a major role, then party switching may induce a dramatic realignment of the two-party system.
3.2 Local Nash Equilibrium in the Stochastic Electoral Model The purpose of this section is to present a model of positioning of parties in electoral competition so as to account for the generally observed phenomenon of nonconvergence. The model presented here is an extension of the multiparty stochastic model of Lin, Enelow, and Dorussen (1999) and McKelvey and Patty (2006), constructed by inducing asymmetries in terms of valence. The basis for this extension is the varied empirical evidence that valence is a significant component of the judgments made by voters of political agents. The stochastic model essentially assumes that political agents cannot predict
62
Norman Schofield
vote response precisely, but can estimate an expected vote share. In the model with valence, the stochastic element is associated with the weight given by each voter, i, to the average perceived quality or “valence” of the political agent. At a particular election, valence will be due to the “integral” of the agent’s competence in the past and expectations about future competence. Valence may also be effected by the resources available to the agent, and in particular the degree of positive media exposure that the agent can exploit. There are a number of possible choices for the appropriate game form for multiparty competition. The simplest one, which is used here, is that the utility function for agent j is proportional to the vote share, Vj, of the agent. With this assumption, we can examine the conditions on the parameters of the stochastic model that are necessary and sufficient for the existence of a Pure strategy Nash Equilibrium (PNE) for this particular game form. This chapter uses calculus techniques to estimate optimal positions. As usual with this form of analysis, we can obtain the necessary conditions for the existence of local optima, or local pure strategy Nash equilibria (LNE). Clearly, any PNE will be a LNE, but not conversely. However, in the models considered here, it is evident that the necesary condition fails, leading to the inference that PNE are typically nonexistent. Although the true utility functions of party leaders are unknown, the comparison of LNE, obtained by simulation of empirical stochastic models, with the estimated positions of parties in the various polities that have been studied suggest that the valence model provides a close first order approximation to the political game. Definition 1. The Stochastic Vote Model M(l, μ, b; C) with Activist Valence for a policy space X of dimension w. (i) Each of the parties in the set P = {1, . . ., j, . . ., p} chooses a policy, zj ∈ X to declare. Let z = (z1, . . ., zp) ∈ Xp be a typical vector of party policy positions. Given z, each voter, i, is described by a vector ui(xi, z) = (ui1(xi, z1), . . ., uip(xi, zp)), where uij(xi, zj) = lj + μj(zj) − bxi − zj2 + eij. = u*ij ( xi , z j ) + ⑀ ij .
(
)
(1) (2)
Here u*ij xi , z j is the observable component of utility. The term, lj, is the intrinsic or exogenous valence of party j, while the function
Switching Equilibria
63
μj(zj) is the component of valence generated by activist contributions to agent j. It is assumed that the resources contributed by activists are used by the political agent to enhance its image through advertising and the like. The term b is a positive constant, called the spatial parameter, giving the importance of policy difference defined in terms of a metric induced from the Euclidean norm, · , on X. The vector ei = (ei1, . . ., eij, . . ., eip) is the stochastic error. It is assumed the multivariate cumulative distribution of each ei is identical. This distribution is denoted by C. (ii) The exogenous valence vector
l = (l1, l2, . . ., lp)
satisfies lp ≥ lp−1 ≥ . . . ≥ l2 ≥ l1.
(iii) Voter behavior is modeled by a probability vector. The probability that a voter i chooses party j at the vector z is rij(z) = Pr[[uij(xi, zj) > uil(xi, zl)], for all l ≠ j] = Pr [⑀il − ⑀ij < u*ij (xi , z j ) − u*il (xi , z j ), for all l ≠ j ].
(3)
(4)
Here Pr stands for the probability operator generated by the distribution assumption on e. (iv) The expected vote share of agent j is Vj z
1 ij z . n i N
(5) p
The differentiable function V : Xp → \ is called the party profile function. Definition 2. Equilibrium Concepts. (i) A strategy vector z* z1* , !, z*p X p is a local strict Nash equip librium (LSNE) for the profile function V : Xp → \ iff, for each agent j e P, the policy z*j strictly locally maximizes Vj ( ! z*j −1 , −, z*j +1 ! ) . (ii) A strategy vector z* is a weak Nash equilibrium (PNE) iff, for each agent j, z*j weakly but globally maximizes Vj ! z*j1 , , z*j 1 !, .
64
Norman Schofield
The most common assumption in empirical analyses is that C is the Type I extreme value distribution (sometimes called Gumbel). The theorems presented in this chapter are based on this assumption. This distribution assumption is the basis for much empirical work based on multinomial logit (MNL) estimation. Dow and Endersby (2004: 111) also suggest that “researchers are justified in using MNL specifications.” Definition 3. The Type I Extreme Value Distribution, C. The cumulative distribution, C, has the closed form C(h) = exp [– exp [–h]]. It follows from this that for each voter i, and agent j, the probability that a voter i chooses party j at the vector z is
ρij(z) = [1 + ∑k≠j[exp(f k)]] –1
(6)
where f k = lk + μk(zk) − lj − μj(zj) + bxi − zj2 − bxi − zk2 . It can be shown that the first order condition for z* to be a LSNE is that it be a balance solution. Definition 4. The balance solution z* = (z1* ,!, z*j ,!, z*p ) for the model M(l, μ, b; C) is given by the set of equations, j = 1, . . . p: n ¯ 1 d j ¡ ␣ij xi z*j ° 0 ¡¢ i 1 °± 2 dz j
(7)
2 ¯ ij ij ° ␣ij ¯ ¡¡ 2 ° where ¡¢ °± ¡ ij ° ¢ i ij ±
(8)
and [ρij] = [ρij(z*)] is the matrix of choice probabilities at z*. n The vector i 1 ␣ij xi . is the weighted electoral mean of agent j. The bracketed term on the left of expression (7) is termed the marginal electoral pull of agent j and is a gradient vector pointing toward the weighted electoral mean of the agent. Thus, the weighted electoral d mean is that point where the electoral pull is zero. The vector dz is called the marginal activist pull for agent j. j
j
Switching Equilibria
65
The following theorems are proved in Schofield (2006, 2007). See also Schofield (2008). Theorem 1. Consider the electoral model M(l, μ, b; C) based on the Type I extreme value distribution, and including both exogenous and activist valences. (i) The first order condition for z* to be an LSNE is that it is a balance solution. (ii) If all activist valence functions are highly concave, in the sense of having negative eigenvalues of sufficiently great modulus, then the balance solution will be a PNE. □ The following theorem classifies the model M(l, b; C), when all activist functions are identically zero. To state the theorem, the coordinates are first transformed, so that in the new coordinates, x* n1 xi 0. The vector z 0 = (0, . . ., 0) i is termed the joint origin in this new coordinate system. Whether the joint origin is an equilibrium depends on the distribution of voter ideal points. These are encoded in the electoral covariance matrix. Definition 5: The electoral covariance matrix, ∇ is the w by w matrix r 1,!, w th r , s ¯ ¢ ± s1,!, w where ξr = (x1r, x2r, . . ., xnr) is the n vector of the r coordinates of voter ideal points, and (ξr, ξs) refers to scalar product. 2 Write s = 1n ( s , s ) for the electoral variance on the sth axis and 1 n
w
1 w s , s trace
n s1 s 1 for the total electoral variance. The formal model just presented, and based on C is denoted M(l, b; C).
2 s2
Theorem 2. (i) At the vector z0 = (0, . . ., 0) the probability ij(z0) that i votes for agent j is independent of i and is given by
¯
¢¡
±°
1
j ¡¡1 exp ¢¡k j ¯±° °° kv j
(9)
(ii) The Hessian of the vote share function of the low valence party, 1, at the vector z0 is C1 = 2[b(1 − 2r1)∇ − I],
(10)
66
Norman Schofield where I is the ω by ω identity matrix.
(11)
(iii) The condition for the joint origin to be a LSNE in the model M(l, b; C) is that C1 has negative eigenvalues. □ Corollary 1 Assume X is two dimensional. Then, for the model M = M(l, b; C), the sufficient condition for the joint origin to be a LSNE is that the coefficient c(l, b; C) = 2b[1 − 2r1]trace(∇) is strictly less than 1. Corollary 2 In the case that X is w-dimensional, the necessary condition for the joint origin to be a LSNE is that c(l, b; C) < w.
3.3
Empirical Analyses for Israel
3.3.1 The Israeli Knesset in 1996 and 2006 To illustrate Theorem 2, consider the case of Israel in 1996. Figure 3.2 shows the estimated positions of the parties at the time of the 1996 election in a two dimensional policy space (see Schofield and Sened 2006 for details). Table 3.1 presents summary statistics of the 1996 election, together with valence estimates based on a stochastic model derived from a voter survey (Arian and Shamir 1999). It is readily shown that one of the eigenvalues of the NRP is positive. Indeed it is obvious that there is a principal component of the electoral distribution, and this axis is the eigenspace of the positive eigenvalue. The formal analysis indicates that low valence parties should position themselves on this eigenspace as illustrated in the simulation given in figure 3.3. As table 3.1 indicates, the lowest valence party was the NRP with valence −4.52. The spatial coefficient is b = 1.12, and the electoral variances on the two axes are 1.0 and 0.732 respectively. Computations for the model M(C) give the following: 1 0. 1 e 4.15 4.52 e 3.14 4.52 1.0 0.591¯ ° ¡ ¡0.591 0.732° ¢ ±
NRP
1.24 1.32¯ ° CNRP 2 1.12 I ¡ ¡1.32 0.64° ¢ ± c , ; ⌿ 3.88.
2
Yahadut Shas 1 Moledet
Religion
Likud 0
Olim
Labor
Ill way
−1
Tzomet
Meretz Dem-Arab communists
−2 −1
−2
0
1
2
Security Figure 3.2 Party positions in the Israeli Knesset 1996.
Table 3.1
Seats, votes, and valences in the Knesset 1996
Party Others Left Meretz Labor Center Likud Shas NRP Moledet Yahadut a b
1996
1996
National Vote % Sample Vote % Seats 7.3 7.6 27.5 3.2 25.8 8.7 8.0 2.4 3.7
0 6.0 44.0 1.8 43.0 2.0 5.1 1.8 1.8
9 9 34 11 32 10 9 2 4
1996
1999 2003
Valence
Seats
Seats
– 0 4.15 –2.34 3.14 –2.96 –4.52 –0.89 –
10 10 28 18 23 17 5 4 5
9 6 21 15 47a 11b 6 – 5b
Israel Beitenu obtained 7 seats in 2003, which I have included with Likud. To simplify Figure 3.4, I have combined the seat totals for Shas and Yahadut.
68
Norman Schofield 2
1
1
Religion
4
0 3
6 5
2
7
−1
−2 −2
−1
0
1
2
Security Key: 1=Shas, 2=Likud, 3=Laor, 4=NRP, 5=Moledet, 6=Third Way, 7=Meretz Figure 3.3
A local Nash equilibrium in the Knesset in 1996.
The eigenvalues of C NRP are 2.28 and −0.40, giving a saddlepoint, and a value for the coefficient, c(l, b; C), of 3.88. This exceeds the necessary upper bound of 2.0, required for an equilibrium at the joint origin. We immediately infer that all local equilibria are nonconvergent. The major eigenvector for the NRP is (1.0, 0.8), and along this axis the NRP vote share function increases as the party moves away from the origin. The minor, perpendicular axis is given by the vector (1, –1.25) and on this axis the NRP vote share decreases as the party moves away from the origin. Figure 3.3, gives one of the local equilibria in 1996, obtained by simulation of the model. The figure makes it clear that the vote maximizing positions lie on the principal axis through the origin and the point (1.0, 0.8). Five different LSNE were located: in all cases, the two high valence parties, Labor and Likud, were located at almost precisely the same positions. The only
Switching Equilibria
69
difference between any two of these various equilibria were that the positions of the low valence parties were somewhat perturbed. The simulations of vote maximizing positions were compatible with the predictions of the formal model based on the extreme value distribution. All parties were able to increase vote shares by moving away from the origin, along the principal axis, as determined by the large, positive principal eigenvalue. In particular, the simulation confirms the logic of the above analysis. Low valence parties, such as the NRP and Shas, in order to maximize vote shares must move far from the electoral center. Their optimal positions will lie either in the northeast quadrant or the southwest quadrant. In contrast, since the valence difference between Labor and Likud was relatively low in the elections of 1988, 1992, and 1996, the relevant eigenvalues on the major axis, for their Hessians at the origin, were also low (but still positive) and their optimal positions would be relatively close to, but not identical to, the electoral mean. Simulation for the elections of 1988, 1992, and 1996 were compatible with this theoretical inference. It should be noted that the positions of Labor and Likud are particularly closely matched by their positions in the simulated vote maximizing equilibria. Clearly, the configuration of equilibrium party positions will fluctuate as the valences of the large parties change in response to exogenous shocks. The logic of the model remains valid however, since the low valence parties will be obliged to adopt relatively radical positions to maximize their vote shares. It is important to note, however, that the positions of the low valence parties in figure 3.2 are similar, but not identical, to their estimated vote maximizing equilibrium position as given in figure 3.3. This strongly suggests that an appropriate model of party positioning assumes that parties are concerned with policy, and adopt positions with a view toward the coalitions that may form after the election. One way to express this inference is as follows: The close correspondence between the simulated LSNE based on the empirical analysis and the estimated actual political configuration suggests that the true utility function for each party j has the form Uj (z) = Vj (z) + δj (z), where δj (z) may depend on the beliefs of party members about the postelection coalition possibilities and on the support of activist coalitions. This suggests that in any political equilibrium of the electoral game, a party can position itself at the electoral origin only when its leader has a very high valence in contrast to all other competing political leaders.
70
3.3.2
Norman Schofield
Recent Switches in the Israeli Knesset
This section discusses recent changes in the political configuration in Israel. Figure 3.4 provides a schematic representation of the Knesset based on the party positions after the election of 2003. The figure shows Labor with 21 seats, after Am Ehad, with 2 seats, joined Labor in 2003, while Likud has 40 seats after being joined by Olim, with 2 seats. Although Barak, of Labor, became prime minister in 1999, he was defeated by Ariel Sharon, of Likud, in the election for prime minister in 2000. Schofield (1999) and Schofield and Sened (2006) propose a set known as the heart to describe the set of policies that may occur as a result of coalition bargaining. The figure indicates that there are three plausible governments: Likud, Shinui, and a religious party; Likud and Shas/Shinui; and a coalition of the left, led by Labor, with support from Shas/Shinui. These coalition possibilities are encapsulated in the idea of the heart (the triangular domain in figure 3.4). Notice that this domain is insensitive to small changes in location or strength of these parties. Given the vector of leader exogenous valences at the time of the 2003 election, it is reasonable to suppose that the
Shas (16)
Religion
NRP (6) Likud (40) Labor (21) Shinui (15) Meretz (6) Arab (9) Security
Figure 3.4
Party positions in the Knesset in 2003.
Israel Beiteinu (7)
Switching Equilibria
71
election result would give rise to an empty core. Notice that the party positions given in figure 3.4 are consistent with the empirical and simulation analysis presented earlier. In particular, it is obvious that neither Likud nor Labor are located at the electoral origin.5 Now consider figure 3.5, which presents estimates of the party positions at the election of March, 2006. In a highly publicized move before the election, Sharon left the Likud Party and signaled a strong move to the left by allying with Shimon Peres, the former leader of Labor and the author of the Oslo accords, together with a number of other senior Labor Party members, to form the new party, Kadima. This move positioned Kadima at the origin of the electoral space at (0, 0), as shown in figure 3.5. Following Sharon’s stroke in January 2006, Ehud Olmert took over as leader of Kadima and was able to win 29 seats. Likud took only 12 seats, while the 4 parties on the upper right of the figure won 38 seats. One surprise of the election was the appearance of a Pensioners’ party with 7 seats. The consequence of this sequence of events is that Kadima secured itself a weakly dominant position in the Knesset. The Kadima position in figure 3.5 can be readily be seen to be a policy core (Laver and Schofield 1990). To
Religion
Shas, Yahadut (12+6)
Labor (19)
NRP (9)
Kadima (29) Likud (12) Pensioners (7)
Arab (10)
Meretz (5)
Security Figure 3.5
Party positions in the Knesset in 2006.
Israel Beiteinu (11)
72
Norman Schofield
see this, note that the convex hull of the policy positions of parties in any winning coalition (with more than 61 seats) must include the position of Kadima. Notice, however, that the core position of Kadima is structurally unstable, in the sense that a small move by Labor on the vertical axis creates the possibility of a counter-Kadima coalition, involving Labor and Likud. As long as Labor is situated at the position given in figure 3.5, then Kadima will be able to veto any attempt to construct a winning coalition government excluding Kadima. The simulations of the 1996 election in Israel demonstrated that Likud was previously unable to capture the core. By moving Labor to the left, Peretz created the opportunity for Sharon to out maneuver him, allowing Sharon to strategically move to a position that would increase the probability that he would control the core. Because Sharon’s own party members would not support him in this move, he had to leave Likud and form Kadima. The formal analysis presented earlier suggests that the location of Kadima at the electoral origin was a vote maximizing position for the party because of the high valence difference between Sharon and Peretz.6 It thus appears that Sharon’s shift in party and his change of policy has led to a fundamental transformation in the political configuration, from the coalition structure without a core (that had persisted since 1996), to a new configuration, associated with the center, core party, Kadima. Even though Kadima is estimated to be a core party after the 2006 election, Olmert needed the support of Labor to be able to deal with the complex issue of fixing a permanent border for Israel. The debacle in Lebanon severely weakened Olmert’s popularity, and in October 2006, the 61 members of the Kadima-Labor governing coalition voted to bring Israel Beiteinu into the coalition. The April 2007 report on the failure of the government during the war with Lebanon in summer 2006 has threatened the stability of the government. Barak won the election for the Labor party leadership on June 12, 2007, and became minister of defense in the Kadima-led government on June 18, 2007, while Shimon Peres became president. On February 3, 2008, Avigdor Lieberman, chairman of Israeli Beiteinu, announced that the party would quit the government because of disagreement over issues such as Jerusalem, the refugees, and the nature of a future Palestinian state. Barak agreed to remain in the coalition, thus maintaining the majority of the coalition. However, as of July 2008, Olmert faces charges of corruption, and he only avoided a vote of no-confidence by agreeing to an election for leader of Kadima in the following September. Athough the Kadima government is structurally unstable, in the sense used here, it has opened negotiations with both Syria, over the
Switching Equilibria
73
question of stability in Lebanon, and with Hamas, using Turkey and Egypt as intermediaries. In the February, 2009 election, the new leader of Kadima, Tzipi Livni, gained 28 seats, while Netanyahu, leader of Likud, took 28 seats. A coalition government still has not formed as of mid-March 2009.
3.4
Party Switches in the United States
As figure 3.1 suggests, social conservative activists have come to prominence in US presidential elections. As social conservatives gain control of the Republican political machinery in a given state or electoral district, socially moderate Republican candidates may be forced into the Democratic Party simply because they can no longer hope to win a Republican primary. The normal ambition of politicians transforms socially liberal Republicans into moderate Democrats. The result is increased party polarization on the social dimension and decreased party differences on the economic dimension. A case in point is John Moore, a longtime executive with Cessna Aircraft in Wichita; a pro-business conservative, he was nevertheless unlikely to win a Republican primary for any statewide position due to his “softness” on social issues. He consequently converted in 2002 and was elected as the Democratic lieutenant governor. Moore retired in 2006, and the open position brought about an even more dramatic development. Mark Parkinson officially switched parties in time to run for the lieutenant governor’s position. Parkinson, a former Republican Party Chairman for the state of Kansas, was elected lieutenant governor on the Democratic ticket. Others in Kansas are going the same route. In 2004, Republican Cindy Neighbor switched parties to run for the state legislature, opposed to a social conservative who had defeated her in the primary in 2004. She was elected in 2006 (Milburn 2006). These ballot box conversions are not limited to ambitious Kansas moderates. Perhaps the most striking and visible such conversion was that of Jim Webb of Virginia. Webb is a much-decorated Vietnam war veteran who had been Reagan’s Secretary of the Navy. As recently as 2000, he supported Republican Allen to be the US Senator from Virginia. In 2006, however, he was a Democrat running against Allen. Traditional New Deal Democrats were aghast; but Webb defeated Allen, and his presence in the party moves the Democratic center of gravity to the right on economic policy. Each such switch makes further switches more likely. Although Kansas has been seen as a state in which the Democratic Party is all
74
Norman Schofield
but defunct, the conversion of a small number of socially moderate Republicans to the Democratic Party could easily restore a healthy two party competition in Kansas. But in the process, each individual conversion changes what it means to be a Democrat (cf. Heller and Mershon this volume). Increasingly, a Democrat is an economic moderate or conservative who is strongly liberal on social issues—not (as in the New Deal), a strong economic liberal whose Democratic affiliation is a response to class conflict. These observations are not meant to advocate any particular strategy for either party. Rather, they suggest that partisan change continues to have a certain inevitability about it, despite the fond wishes of entrenched party activists. Each partisan realignment has occurred despite the opposition of existing party activists. Populist Democrats in the 1930s, who earlier supported William Jennings Bryan, were suspicious of the ethnic industrial laborers that the New Deal brought into the party. In the same way, traditional Republican activists were aghast when their candidate Rockefeller was booed for criticizing Goldwater-style radicalism at the 1964 convention (Branch 1998, 402). Partisan realignment is a dynamic process because of the destabilizing influence of vote-maximizing candidates who see opportunities to win elections even at the cost of generating some hostility within the ranks of the preexisting activist cadres. As a result, partisan identities are always changing, even though there is a tendency to see them as fixed and immutable. The Republican Party in 1868 was the post– civil war party of racial equality through strong national government. The Republican Party in 1948 was the party of the balanced budget and civil libertarianism. Neither of these identities proved to be immutable, and the current identities of both parties are again in flux. The departure of even a small number of pro-business social liberals from the Republican Party—like Jeffords of Vermont or Parkinson in Kansas—has inevitable effects on both parties. Each such departure increases the proportion of social conservatives in the Republican Party, making it easier for social conservatives to dominate both the party primaries and the activists who give the party its image to the nation. This in turn makes it even more difficult for social liberals to hope for a successful career within the GOP. Voters, as well as activists and candidates, adjust. If they are concerned about women’s rights or the separation of church and state, they are less likely to vote as Republican and more likely to shift to independent or Democratic status.
Switching Equilibria
75
At the same time, symmetrical adjustments are made in the Democratic Party. Just as Strom Thurmond’s conversion to the Republican Party helped trigger a long list of similar conversions by socially conservative Democrats, so each socially liberal Republican who converts to the Democratic party makes the social issue that triggered the conversion a more salient aspect of the Democratic identity. The Democratic Party and Republican Party become internally more homogeneous as regards economic policy, and more polarized with respect to social policy. Thus, as social polarization increases between the parties, the economic differences will slowly disappear. As pro-business social liberals join the Democratic Party, it will become increasingly difficult to imagine that party going back to a New Deal identity. Just as the New Deal Democratic Party consisted of segregationists and labor unions united on an antibusiness platform, the emerging Democratic Party will find itself united at a social liberal position, with a centrist position on economic policy. The proportion of Democrats who adopt a traditional antibusiness stance will be reduced. A simple electoral calculus by candidates will tend to move them to a Clinton-style moderate position on economic policy—advocating (among other things), a more inclusive policy toward immigrants, and a more enthusiastic commitment to stem cell and related medical research.
3.5 Concluding Remarks The theorems presented here, together with analysis of spatial policy maps (Schofield 2008), suggest the following set of tentative hypotheses about the nature of political competition. The examples from Israel and the United States allow for further hypotheses about the role of party switching in affecting coalition formation, under electoral systems based on proportional representation, and in influencing political realignment in plurality electoral systems. 1. The pure spatial model of direct democracy indicates that the occurence of a core, or unbeaten alternative, is very unlikely in a direct democracy using majority rule, when the dimension of the policy is at least two (Schofield 2008). However, a social choice concept known as the heart, a generalization of the core, will exist, and converges to the core when the core is nonempty (Schofield 1999). 2. A legislative body, made up of democratically elected representatives, can be modeled in social choice terms. Because party strengths will be
76
Norman Schofield
3.
4.
5.
6.
7.
8.
9.
disparate, a large, centrally located party may be located at a core position. Such a party, even in a situation with no majority party, will be able to dominate the formation of coalition government, as suggested by the example from Israel. A more typical situation is one with no core party. In such a case, the legislative heart can give an indication of the nature of bargaining between parties as they attempt to form a winning coalition government. This theory of legislative behavior takes as given the position and strengths of the parties. Because a centrally located party may dominate coalitional bargaining, and because such a party should be able to garner a large share of the vote, there would appear to be a strong centripetal tendency in all electoral systems based on proportional representation. However, this tendency will be most relevant for a party whose leader has a high exogenous valence. Estimates of party positions indicate that parties adopt quite heterogenous positions.7 This suggests that there is a countervailing or centrifugal force that affects all parties other than those with very high valence. Although core parties can be observed in a number of polities with electoral systems based on proportional rule, the dominance of such center parties can be destroyed, particularly if there is a tendency to political fragmentation and social conflict. The stochastic model suggests that only a political leader with a high exogenous valence is able to move to a central position to control the core. Although the model with exogenous valence is able, in some circumstances, to predict the existence of a core party, it is suggested that the model with activist valence gives a better understanding of the extensive dispersion of positions of the parties.8 This suggests that party location can be modeled as a balancing act between the centripetal electoral pull, and the activist centrifugal pull. This implies that on occasion a party leader is subject to a conflict between the more extreme policy preferences of activists for his party, and the centripetal tendency implied by his high exogenous valence. A switch of policy position by a leader, such as Sharon in Israel in 2006, can then be rationalized in terms of the centripetal tendency implied by high valence. Under proportional electoral methods, there need be no strong tendency forcing activist groups to coalesce in order to concentrate their influence. Thus proportionality in the electoral system, as in Israel and many European polities (see Benoit and Laver 2006; Schofield 2008), leads to activist fragmentation and thus party fragmentation. Belgium is a good example, as discussed in Schofield (2008). More generally, it can be expected that low valence political leaders may switch parties particularly in situations where they perceive that they can gain increased resources from activist goups as a result of switching.
Switching Equilibria
77
10. Under plurality rule, small parties face the possibility of extinction. Unlike the situation in a polity based on proportional rule, an activist group linked to a small party in a plurality polity has little expectation of influencing government policy. Thus activist groups face increasing returns to scale, in terms of the relationship between the size of the political coalition that they influence on the one hand and political power and representation on the other. The activist model of elections presented in this chapter suggests that when there are two dimensions of policy, then there can be a number of principal, opposed activist groups. The nature of the electoral contest generally forces these activist groups to concentrate support for at most two parties, as in the United States. 11. Thus, in the United States, plurality rule induces the two party system, through this effect on activist groups. Although the two party configuration may be in equilibrium at any time, the tension within the activist coalitions induces a slow rotation, and thus political realignment. Candidates for political office, at the federal and local level, will be forced to balance the centripetal electoral effect against the centrifugal effect induced by activist groups. 12. On occasion in the United States, polarization within the two dimensions of policy may induce conflict within and between activist groups. These effects may become so pronounced that the two party system may become less stable.9 The illustrations presented in this chapter suggest that switching between parties can become extreme, leading eventually to a political realignment. 13. The well-known relationship between proportional representation and the degree of political fragmentation may be accounted for indirectly as a consequence of the logic forced on activist groups and the effect induced on the political configuration.
3.6
Notes
* This chapter is based on work supported by NSF grant 0715929, and a research grant from the Weidenbaum Center, Washington University. The chapter was revised while Schofield was a visiting fellow at ICER, Turin. Versions have been presented at conferences at Trinity College, Dublin, at the University of Virginia, Charlottesville, and at Emory University. The figures are reproduced from Schofield and Sened (2006) with permission of Cambridge University Press. 1. Stokes (1963: 373) used the term valence issues to refer to those that “involve the linking of the parties with some condition that is positively or negatively valued by the electorate.” As he observes, “in American presidential elections . . . it is remarkable how many valence issues have held the center of the stage.” Stokes observation is validated by recent empirical work on many polities, as well as a study on the psychology of voting (Westen 2007).
78
Norman Schofield
2. However, current exogenous valence will depend on the agents previous policy choices, and the outcome of these choices. As an illustration, George Bush’s low overall valence, as of July 2008, is a consequence of policy choices in the past that turned out to be imprudent. 3. For convenience, we assume that μj(zj) is only dependent on zj, and not on zk, k ≠ j, but this is not a crucial assumption. 4. In other words, it is not the source of the resources that matters, just the amount. 5. Jeong et al. (2008) estimate the 2003 election. Their results indicate that Sharon had a very high valence in 2003. 6. We may hypothesize that Sharon was able to position Kadima at the center precisely because of his high valence. 7. This is clearly indicated by the estimates of party position presented in Benoit and Laver (2006). 8. The empirical analyses in Schofield and Sened (2006) for Italy, Netherlands, and the United Kingdom support this inference. 9. Of course, there have been presidential elections, such as 1860, where activist polarization has been so strong that there were four competing candidates.
3.7 References Aldrich, John H. 1983a. “A Downsian Spatial Model with Party Activists.” American Political Science Review 77 (December): 974–990. ———. 1983b. “A Spatial Model with Party Activists: Implications for Electoral Dynamics.” Public Choice 41 (January): 63–100. Aldrich, John H. and Michael McGinnis. 1989. “A Model of Party Constraints on Optimal Candidate Locations.” Mathematical and Computer Modelling 12 (4–5): 437–450. Arian, Asher and Michal Shamir. 1999. The Election in Israel: 1996. Albany: SUNY Press. Banks, Jeff S. and John Duggan. 2005. “Probabilistic Voting in the Spatial Model of Elections.” In Social Choice and Strategic Decisions, ed. David AustenSmith and John Duggan. Heidelberg: Springer. Benoit, Kenneth and Michael Laver. 2006. Party Policy in Modern Democracies. London: Routledge. Branch, Taylor. 1998. Pillar of Fire. New York: Simon and Schuster. Dow, John K. and Jay Endersby. 2004. “Multinomial Probit and Multinomial Logit: A Comparison of Choice Models for Voting Research.” Electoral Studies 23 (March): 107–122. Downs, Anthony 1957. An Economic Theory of Democracy. New York: Harper and Row. Hinich, Melvin. 1977. “Equilibrium in Spatial Voting: The Median Voter Result Is an Artifact.” Journal of Economic Theory 16 (2): 208–219. Hotelling, Harold 1929. “Stability in Competition.” Economic Journal 39 (March): 41–57.
Switching Equilibria
79
Jeong, Gyung-Ho, Itai Sened, Yael Shomer, with Yanai Sened. 2008. “Two Decades of Israeli Legislative Politics: 1988—2006.” Typescript: Washington University in Saint Louis. Laver, Michael and Norman Schofield. 1990. Multiparty Government: The Politics of Coalition in Europe. Oxford: Oxford University Press. Reprinted 1998. Ann Arbor: University of Michigan Press. Lin, Tse Min, James Enelow, and Han Dorussen. 1999. “Equilibrium in Multicandidate Probabilistic Spatial Voting.” Public Choice 98 (January): 59–82. McKelvey, Richard D. and John W. Patty. 2006. “A Theory of Voting in Large Elections.” Games and Economic Behavior 57 (October): 155–180. Milburn, John. 2006. “Former GOP Chairman Parkinson Switches Affiliation to Democrat.” Kansas City Star, May 30. Miller, Gary and Norman Schofield. 2003. “Activists and Partisan Realignment in the U.S.” American Political Science Review 97 (May): 245–260. ——— 2008. “The Transformation of the Republican and Democratic Party Coalitions in the U.S.” Perspectives on Politics 6 (3): 433–450. Poole, Keith and Howard Rosenthal. 1984. “U.S. Presidential Elections 1968– 1980: A Spatial Analysis.” American Journal of Political Science 28 (May): 283–312. Schofield, Norman. 1999. “The Heart and the Uncovered Set.” Journal of Economics Suppl. 8: 79–113. ——— 2006. Equilibria in the Spatial Stochastic Model with Party Activists. Review of Economic Design 10 (December): 183–203. ——— 2007. “The Mean Voter Theorem: Necessary and Sufficient Conditions for Convergent Equilibrium.” Review of Economic Studies 74 (July): 965–980. ——— 2008. The Spatial Model of Politics. London: Routledge. Schofield, Norman and Gary Miller. 2007. “Elections and Activist Coalitions in the United States.” American Journal of Political Science 51 (July): 518–531. Schofield, Norman and Itai Sened. 2006. Multiparty Democracy: Elections and Legislative Politics. Cambridge: Cambridge University Press. Schofield, Norman, Gary Miller, and Andrew Martin. 2003. “Critical Elections and Political Realignments in the USA: 1860–2000.” Political Studies 51 (June): 217–240. Stokes, Donald. 1963. “Spatial Models of Party Competition.” American Political Science Review 57 (June): 368–377. ———. 1992. “Valence Politics.” In Electoral Politics, ed. Dennis Kavanagh. Oxford: Clarendon Press. Westen, Drew. 2007. The Political Brain. New York: Perseus Books.
This page intentionally left blank
4 Party Switching and the Procedural Party Agenda in the US House of Representatives* Timothy P. Nokken
4.1 Introduction: Party Defection in the US Congress The importance of party affiliation in the United States may seem difficult to establish. Legislative parties in the United States are rather loose confederations of individuals who share a common party label but lack the discipline and uniformly high levels of party cohesion of their counterparts in many parliamentary systems. As a consequence of the primary system for selecting candidates to carry the party label in legislative election campaigns, party leaders lack an important tool for disciplining their rank and file. Moreover, the sizable contingent of independent and unaligned voters in the electorate may contribute to the relatively weak legislative parties in the United States (Wattenberg 1994, 1998). Given relatively loose legislative party discipline and relatively fluid party identification in the electorate, one might expect to observe frequent party switching among strategic political actors as they try to capture the benefits associated with membership in whichever party enjoys greater support among voters. Yet shifts in party affiliation among US House and Senate members have been rare. Specifically, only 38 Senators and 160 House members switched parties over a 163-year period (Nokken and Poole 2004).1 The observed dynamic of party defection among members of Congress (MCs) in the United States stands in marked contrast to switching in a number of other countries. The most obvious difference concerns relative frequency: Party switching in the United States is far less common than in several other systems. For example, in
82
Timothy P. Nokken
the 1996–2001 Italian Chamber, nearly one-fourth of all legislators jumped parties at least once and total switches numbered 277; and in the 1993–1995 Russian Duma, nearly one-third of all legislators switched at least once, for a total of 342 switches (Mershon and Shvetsova 2008). Thames (2007) reports a total of 527 switches in the Ukranian Rada from 1998 to 2002. The contrast with the United States is stark indeed: The number of party switches in each of the countries just cited, for a single legislative term lasting five years or less, exceeds the total number of congressional party switches over the entirety of American history (see Poole and Nokken 2004; for more cross-national comparisons, see the introduction and other chapters in this volume). The variation in the incidence of party switching influences the focus of scholarly research on switching. Where switching is frequent, scholars tend to ask why and when legislators switch parties (e.g., Castle and Fett 1996, 2000; Desposato 2006; Heller and Mershon 2005; King and Benjamin 1986; Thames 2007). The best predictive model of party defection in the United States, on the other hand, is a naïve model that predicts no one switches parties.2 Nonetheless, the premise of this study is that, while the act of switching parties in US history is rare, the set of party defectors serves as an analytically useful group of legislators with which to investigate how congressional party affiliation constrains member behavior. Precisely because the vast majority of members retain a single party label throughout their careers, scrutiny of switchers affords evaluation of how changing party affiliation influences behavior. Hence, a common approach to studying party switching in the American context is to use a quasi-experimental research design to assess the behavioral consequences of party switching, conducting two sorts of comparisons. The first comparison, across time, examines pre- and post-switch roll-call behavior for individual party defectors. The second, across MCs, juxtaposes the roll-call behavior of party switchers with that of members who do not change parties. The theoretical motivation for such analyses is that party affiliation is one of the primary factors that shapes the roll-call behavior of MCs. Parties often provide clearly defined positions for their members within particular policy areas, thereby placing constraints on the positions their members can credibly take and still assert status as a loyal Democrat or Republican. If party affiliation does constrain the voting behavior of MCs, then significant shifts in behavior should be more likely among party defectors than among those who do not change parties.
Party Switching and the US House Agenda
83
I exploit a natural experimental design to discern how party affiliation influences roll-call behavior. I use the DW-NOMINATE scaling procedure to calculate MCs’ ideal points on final passage, amendment, and procedural votes from 1953 to 2002 (83rd to 107th Congresses). I then utilize a technique implemented by Poole (2005), Goodman (2004), and Nokken and Poole (2004) to evaluate the magnitude and direction of ideological change in the voting behavior of party defectors. By distinguishing among three categories of vote types, I am able to ascertain whether the aggregate roll-call measures typically used to evaluate the behavioral implications of party defection mask changes within important subsets of roll calls. More specifically, some switchers may not exhibit dramatic behavioral changes on final passage votes in an effort to protect their ideological reputation among voters. They may, however, be more likely to exhibit notable changes in the less visible, but highly partisan, procedural and amendment votes. The chapter proceeds as follows. In the next section, I discuss previous works investigating party defection in the United States, and develop several hypotheses regarding party switching and rollcall behavior. The third section describes the natural experimental research design I use to analyze the behavioral consequences associated with party switching and also addresses the scaling of voting behavior and categories of votes. The fourth section reports evidence on the effects of switching, examining both aggregate vote scores and voting behavior for three categories of votes. The fifth section concludes the chapter.
4.2 Rare Bird Sightings: Analyzing Party Defection in the US Congress 4.2.1 Two Paths in Analyzing Party Switching in the United States In the study of congressional party defection, scholars generally take one of two paths. The first path analyzes the act of switching itself and seeks to isolate the factors that lead individuals to switch parties. The varying electoral fortunes of the parties, for instance, may induce office holders to switch parties. Aldrich and Bianco (1992) derive a formal model to identify settings in which incumbent legislators may have strong electoral incentives to switch parties. They contend that party switching is best modeled in gametheoretic terms whereby individual-level decisions regarding party affiliation fundamentally depend on the decisions made by other
84
Timothy P. Nokken
players. 3 Grose and Yoshinaka (2003) devise an empirical test of the Aldrich and Bianco model, finding that MCs who switch parties face relatively stiff electoral competition and see a reduction in their vote share in both the primary and general elections following their change in party affiliations. Grose (2004) extends the analysis and investigates the electoral consequences for state legislators who switch parties. Turning to career trajectories within Congress, Yoshinaka (2005) finds that MCs who switch parties are relatively likely to be rewarded for their trip across the aisle, since switchers are more likely than nonswitchers to benefit from violations of seniority in committee assignments; the implication is that MCs who expect such benefits would be induced to switch. King and Benjamin (1986) identify the MCs who switched parties from 1789 to 1984 and conclude that switches tended to coincide with such major changes in the macropolitical environment as military conflicts, economic conditions, and partisan control of key institutions. A member’s ideological position relative to the two parties also appears to influence the decision to switch. In particular, ideologically cross-pressured members—those at or near the liberal (conservative) end of the distribution of Republicans (Democrats) in the House or Senate—may find that they more closely resemble members of the other party, and switch parties to take advantage of the better ideological fit (Castle and Fett 1996, 2000). Recognizing the cross-pressures on many Southern Democrats, the Republican Party in the 1980s began actively recruiting Democratic office holders at all levels of government to join the Republicans (Canon 1992; Canon and Sousa 1992). The second broad category of inquiry focuses on the consequences that party switching has for roll-call behavior. The notion here is that studying the set of party switchers affords analytical leverage on the effects of party on roll-call behavior, and thus contributes to the “party versus preferences” debate. Traditionally, those studying congressional behavior have simply asserted that party affiliation in part has played a causal role in MCs’ voting behavior. By virtue of affiliating with a party, legislators could be led to support the party’s stated position despite possessing policy preferences that diverge from the rest of their party. Krehbiel (1993) offers a forceful challenge to that conventional wisdom, countering that individual-level policy preferences are the primary determinants of both roll-call voting behavior and party affiliation. In a nutshell, if an individual tends toward the conservative side of the ideological spectrum, he or she
Party Switching and the US House Agenda
85
will likely affiliate with the Republican Party and cast roll-call votes that reflect both partisan and personal preferences. To prove the existence of significant partisan effects, one must be able to demonstrate that pressures emanating from partisanship, namely pressure from party leaders, cause MCs to support the party’s position on roll-call votes, even when doing so runs contrary to their own policy preferences (Krehbiel 1993). Nokken (2000) addresses the question of how party affiliation has influenced roll-call behavior by analyzing the voting behavior of the set of 20 MCs (16 House members, 4 Senators) who switched parties while in office from 1947 to 1997. Using member support scores on Democratic leadership and agenda votes (Cox and McCubbins 1993), he compares the pre- and post-switch behavior of defectors to determine whether they change their voting behavior and, if they do, whether the change took place gradually over time or transpired at the time of the switch. He also compares the voting behavior of switchers with that of a control group of “proximates”—members who have maintained their party affiliation, compiled roll-call records similar to those of switchers, and supported party leadership so as to generate leadership support scores within an interval of plus or minus five points of an individual switcher’s score during that switcher’s first term in Congress. He found that party switchers dramatically shift their voting behavior at the time they switch, and do so in a direction consistent with expectations: MCs who have left the Democratic Party to join the GOP have moved to the right, while Republicans who have become Democrats have shifted to the left. Furthermore, such shifts are limited to party defectors, with proximates showing no evidence of statistically significant changes. These changes in roll-call behavior among party switchers suggest that party affiliation carries with it a code of conduct. Strong incentives exist for compliance with such a code even in a newly entered party, since deviations from the party’s stated course may result in leadership sanctions (Cox and McCubbins 1993) or electoral punishment, especially in primaries (Brady and Schwartz 1990; Grose 2004; Grose and Yoshinaka 2003). Party switching does not, however, always lead to noticeable changes in roll-call behavior. To determine how individual-level behavior covaries with changes in party affiliation, Nokken and Poole (2004) treat each party switcher as if he or she were two distinct members, and calculate pre- and post-switch DW-NOMINATE scores.4 They find that instances of significant behavioral shifts resulting from party
86
Timothy P. Nokken
defection tend to be concentrated in eras of high levels of ideological polarization. Oppenheimer and Hatcher (2003) also analyze the rollcall behavior of switchers over a long span of time and arrive at similar conclusions. In addition, they conduct interviews with a number of switchers serving in Congress. The interviewers showed the members the change in NOMINATE scores pre- and post-switch, and then asked the members about the behavioral consequences of their party defection. A number of switchers asserted that they made no significant changes to their roll-call voting behavior, and that any observed shifts likely resulted from changing votes on procedural matters.5
4.2.2
Party Switching and Legislative Behavior: Hypotheses
Both sets of findings make sense. First, during polarized periods, switchers have no real options but to make dramatic changes in voting behavior if they are to fit in with their new party. On the flip side, instances of partisan overlap are especially likely during periods of low polarization, reflecting the presence of conservative Democrats and liberal Republicans (cf. Krehbiel 1993, 239). During times of ideological overlap, members can switch parties without having to alter their behavior dramatically, because partisan labels do not closely map onto specific ideological predispositions. The era from World War II through the 1950s is a classic example of just such a setting, since it prominently featured the Conservative Coalition, the alliance of Southern Democrats (known as “Boll Weevils”) and Republicans who opposed the more liberal Northern wing of the Democratic Party on a number of roll-call votes. Most of the Southern Democrats of this era were conservatives whose ideology often resembled their Republican colleagues, but who remained Democrats because they could never win reelection in the Solid South by running on a Republican label. Second, MCs may go to great lengths to construct an ideological reputation among constituents (Dougan and Munger 1989). Variation from a well-defined vote history (Asher and Weisberg 1978) could increase the likelihood of electoral defeat (Lott and Bronars 1993). Consequently, members rationally could choose to maintain a consistent voting record on final passage votes regardless of party affiliation, reasoning that constituents would likely observe their decisions on final passage votes in particular and that their attempts to explain public flip-flops on policy could lead to electoral difficulties. Although one might well expect voting behavior on final passage measures to remain relatively stable for party switchers—especially
Party Switching and the US House Agenda
87
during periods of intraparty heterogeneity—roll-call behavior within the other categories of votes might change more dramatically, as implied by the interview responses reported in Oppenheimer and Hatcher (2003). To constituents, party switchers continue to compile a consistent record on substantive policy proposals, the set of votes most salient to the electorate. It is less likely that constituents notice or care about members’ votes on procedural matters, even though such votes ultimately carry with them significant policy consequences. Although constituents may fail to observe changes on procedural votes, a switcher’s colleagues certainly will take note. Sinclair (1981, 406) quotes one member on the importance of procedural support for a Democratic budget resolution as follows: “ ‘If you’re going to vote no on final passage of the budget resolution, help us out at this point by voting no on . . . any other Republican substitute.’ And that argument seemed to hold with a lot of people that were willing to help us out by voting no on any Republican substitute, even if they were going to vote no on the [final] resolution.” Cox and McCubbins (2002; 2005) argue that the majority party leadership is able to control the legislative agenda so as to preserve majority party cohesion. Of particular interest is the notion that the majority can exert negative agenda control—that is, has the power to keep issues off the floor that would split the majority. The procedural votes accompanying major legislation form an important component of such potentially divisive issues. Consistent with the notion of negative agenda control, Cox and McCubbins find only a handful of instances in which legislation passed over the opposition of a majority of the majority party. Given the importance of the procedural aspects of negative agenda control, switchers of all stripes should make significant moves toward their new party on procedural matters. Thus, party switchers might enjoy the best of both worlds: they might please their colleagues in their new party with shifts on the more electorally innocuous procedural votes, while pleasing their constituents by maintaining a consistent voting record on final passage policy votes. The discussion thus far generates a series of testable hypotheses about the nature of party influence on roll-call behavior. The first and most general hypothesis is that if party labels provide substantively important signals about ideological type, then changes in party affiliation should lead to changes in roll-call voting behavior and in particular should align the MCs’ votes more closely with their new party. The underlying notion is that individual MCs’ party affiliation affects the content of roll-call voting behavior. Recognizing this, MCs who
88
Timothy P. Nokken
change party should also change their votes on roll calls. Switching, though, might not have a constant effect across the entire set of rollcall votes considered by MCs if the effect of party switching on voting behavior varies across different types of vote. I depart from most of the past work on the behavioral effects of party defection to determine how a change in party labels influences switchers’ voting behavior on specific categories of votes. The second hypothesis is that members should exhibit higher levels of party loyalty on procedural and amendment votes than on final passage votes. Moreover, precisely because of the greater loyalty on procedural and amendment votes, switchers should make a dramatic move across the ideological spectrum on procedural votes at the time they switch parties. As shown in the following text, my analyses allow me to determine whether switchers modified their roll-call behavior when they switched and whether changes are limited to MCs who switch from one party to another or extend to those MCs who trade independent status for a party label. The third hypothesis is that switchers’ votes on final passage should be relatively unlikely to exhibit significant changes, except during phases of polarization. If party affiliation is a key determinant of voting on procedural matters, then it is plausible that ideology is a key determinant also of policy votes. Party switching would thus carry with it policy-based behavioral changes that reflect the switcher’s new party label. Yet, one might expect such changes to be smaller than the changes observed with procedural votes, since final votes are relatively visible to the electorate. This behavior should vary over time, however. In particular, dramatic changes in policy-based final passage votes should appear during times of polarization, giving rise to the setting that fosters the components of conditional party government (Aldrich and Rohde 2000, 2001; Rohde 1991).
4.3 Party Effects and Party Switchers: Research Design To investigate how changes in party affiliation influence roll-call behavior, I identify the set of House party switchers from the 83rd to the 107th Congress (1953 to 2002). I use Rohde’s roll-call voting data set (2004) to classify roll calls into final passage, amendment, and procedural votes. I use a modified version of Poole’s DW-NOMINATE procedure to estimate ideal point coordinates for all House members
Party Switching and the US House Agenda
89
for the three subsets of votes and thus parallel the aggregate-level analyses in Nokken and Poole (2004).6 I limit the analysis to House members from 1953 to 2002 for two reasons. First, the House, unlike the Senate, is a highly structured institution with a complex system of restrictive rules that require members to vote on a number of important procedural matters in the course of normal legislative operations. For example, for a bill to reach the floor of the House via “normal” procedures, the House must first adopt the rule issued by the Rules Committee that specifies the terms of debate and the number and types of amendments (if any) that will be allowed. Votes on rules are tremendously important for the majority party leadership because they help ease the passage of legislation the leadership prefers and erect barriers in the path of proposals it might oppose; these effects for the most part operate unheeded by constituents.7 The second reason concerns data availability. Rohde’s roll-call voting data set (2004) codes House votes by type from 1953 to 2002 (the 83rd 107th Congresses). Though somewhat limited as to time, the Rohde data set provides exhaustive and valuable categorizations of all House votes. For 1953 to 2002, table 4.1 lists a total of 25 instances of party switching involving 24 individuals.8 The direction of switching favored the Republicans: On 5 occasions Republicans became Democrats, and on 15, Democrats joined the Republicans; the remaining 5 cases involved switches to or from minor parties or Independent status. To be sure, switchers constitute a very small percentage of the total number of members who served in the House over this period. As stressed, however, the small number of observed switches is not problematic given my purpose of evaluating the behavioral implications associated with changes in party labels. After identifying the set of switchers, I use the Rohde data set to disaggregate roll calls into final passage votes, procedural votes, and amendment votes. Appendix A details the vote types included in each of the three categories.9 Figure 4.1 plots the total numbers of votes in each category, which vary markedly over time and evince a general upward trend.10 Of some concern is the relative paucity of amendment and procedural votes from the earlier congresses in my data. The 83rd to 91st Congresses all held less than 50 recorded roll calls on amendment votes.11 Starting in the 92nd Congress (1971), when the House implemented a rule allowing recorded votes in the Committee of the Whole, the number of roll calls on amendments increased sharply (Bach and Smith 1988).12 Overall, there are enough votes within each category to generate meaningful ideal point estimates.
90
Timothy P. Nokken
Table 4.1 House defectors and change in aggregate DW-NOMINATE scores, 1953–2004 Old Party
Congress Name
Party
85 86 89 92 93 93 93 94 96 97 97 97 98 100 100 101 104 104 104 104 104 106 106 106 107 108 108
REP IND. DEM DEM REP DEM IND. DEM REP REP DEM IND. DEM DEM DEM DEM DEM DEM DEM DEM DEM DEM DEM REP DEM DEM IND. DEM DEM
Dellay Alford Watson Reid Jarman Moakley Riegle Peyser Atkinson Foglietta Gramm Stump Ireland Grant Robinson Watkins Deal Hayes Laughlin Parker Tauzin Forbes Goode Martinez Goode Hall Alexander
New Party DWDWNOMINATE Congress Party NOMINATE 0.009 0.012 0.427 –0.225 0.280 –0.463 –0.459 –0.125 –0.016 –0.406 0.477 0.561 0.153 –0.166 0.013 –0.045 0.008 0.038 0.051 0.219 0.134 0.111 0.263 –0.123 0.671 0.109 –0.051
85 87 90 93 94 94 94 96 97 98 98 98 99 101 101 105 104 104 104 104 104 106 106 106 107 109 109
DEM DEM REP DEM REP DEM DEM DEM REP DEM REP REP REP REP REP REP REP REP REP REP REP DEM IND REP REP REP REP
–0.318 0.197 0.377 –0.432 0.251 –0.415 –0.399 –0.362 0.090 –0.499 0.602 0.714 0.433 0.232 0.284 0.395 0.381 0.365 0.510 0.438 0.350 –0.155 0.529 0.397 0.822 0.351 0.443
Note: Names in Boldface indicate those members who switched whose roll-call behavior shifted in the expected direction following their switch. Source: DW-NOMINATE data available at ftp://pooleandrosenthal.com/junkord/HL01108A1_ PRES.DAT
I utilize a procedure developed by Poole to compare the magnitude of a switcher’s change in voting behavior with the change demonstrated by a set of common members serving in the same Congresses. I thus determine whether the switcher exhibited larger variance in his or her DW-NOMINATE score than the nonswitchers. I do not detail the scaling procedure here, but point interested readers to Goodman (2004), Nokken and Poole (2004), and Poole (2005) for further discussion of the scaling methodology (see also note 4 above).
Party Switching and the US House Agenda 700
91
Final Passage Amendment
600
Total Votes
500 400 300 200
Procedural
100
83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 10 0 10 1 10 2 10 3 10 4 10 5 10 6 10 7
0 Congress Figure 4.1
Total roll calls by vote types, 83rd to 107th Congresses.
4.4
The Effects of Party Switching
I first assess my hypotheses in the light of aggregate data, and then turn to more finely grained evidence. To anticipate the chief findings, the aggregate analysis confirms the established wisdom on the effects of switching, as expressed in the first hypothesis, and the disaggregated analysis supports the second and third hypotheses.
4.4.1. Switching and Aggregate Voting Scores For the aggregate view, consider table 4.1 once more, which registers, for every House switcher from 1953 to 2002, the last Congress in which the MC served with his or her original party label and DW-NOMINATE score from that Congress, along with the Congress in which he or she switched parties and the corresponding DW-NOMINATE score.13 According to the first hypothesis on the direction of changes in voting behavior, Democrats (Republicans) who leave to join the Republicans (Democrats) should move to the right (left). Indeed, as table 4.1 displays, all 22 switchers who moved from one major party to the other shifted in the direction of the party entered. The absolutely uniform behavior of those switching from
92
Timothy P. Nokken
one party to another indicates the purchase of the first hypothesis, which speaks specifically to the importance of party labels. Of the three MCs who switch directly from Independent status to a party label, only one votes more in line with the new party. Virgil Goode (VA), with two switches, is a case apart: If we ignore his service as an Independent and treat him as defecting directly from the Democrats to the Republicans, his voting confirms the first hypothesis, shifting notably to the right.14 Thus, table 4.1 illustrates that party defection generates significant behavioral changes at the time a member crosses the aisle, as previous research has shown (Nokken 2000; McCarty, Poole, and Rosenthal 2001; Nokken and Poole 2004). Table 4.2 shows not only the direction but also the statistical significance of shifts in the voting behavior of party switchers (Nokken and Poole 2004). The rightmost column isolates those members whose votes shift in the expected direction (marked with a +). MCs whose names appear in boldface exhibit statistically significant behavioral changes. To identify such shifts, Nokken and Poole (2004) calculate the distance between a switcher’s pre- and post-switch first dimension DW-NOMINATE score. Next, they compare the magnitude of the switcher’s shift to the changes for all MCs serving in the two corresponding Congresses (the number available for comparison is arrayed to the right of the switcher’s name). Members are then rank-ordered by magnitude of change, and categorized as evincing a statistically significant change if the switcher’s behavioral shift was in the proper direction (as addressed in the last column) and if it placed him in the 95th percentile of all MCs serving with the switching MC (the penultimate column). To clarify, Dellay’s votes shift in the correct direction (to the left) as he moves from the Republicans to the Democrats. The magnitude of this change on the first dimension is greater than 100 percent of the other 351 members serving in the 2 congresses involved. Overall, these data on aggregate roll-call voting provide support for the first hypothesis. Fourteen of the 16 switchers whose votes realign in the expected direction exhibit statistically significant changes; neither of the two whose votes deviate from expectations show significant change. Yet the MCs claiming various degrees of independent status deserve a closer look. It is also important to probe more deeply into the sources of statistically significant voting changes among other switchers. Do MCs move across the aisle to adopt a new set of positions on matters of both procedure and substantive policy? Identifying whether behavioral shifts tend to take place across all roll-call votes or vary
Table 4.2 House defectors and the statistical significance of changes in aggregate voting, 1953–2002
Old Party Last Year 1957 1959 1965 1971 1973 1973 1973 1975 1979 1981 1981 1981 1983 1987 1987 1989 1995 1995 1995
Last Cong 85 86 89 92 93 93 93 94 96 97 97 97 98 100 100 101 104 104 104
New Party Party
Republican Ind. Democrat Democrat Republican Democrat Ind. Democrat Republican Republican Democrat Independent Democrat Democrat Democrat Democrat Democrat Democrat Democrat Democrat Democrat
First Year 1957 1961 1967 1973 1975 1975 1975 1979 1981 1983 1983 1983 1985 1989 1989 1997 1995 1995 1995
First Cong 85 87 90 93 94 94 94 96 97 98 98 98 99 101 101 105 104 104 104
Party Democrat Democrat Republican Democrat Republican Democrat Democrat Democrat Republican Democrat Republican Republican Republican Republican Republican Republican Republican Republican Republican
Switcher’s Name Dellay Alford Watson Reid Jarman Moakley Riegle Peyser Atkinson Foglietta Gramm Stump Ireland Grant Robinson Watkins Deal Hayes Laughlin
N Compared
Pct
Dir
351 369 359 361 339 339 339 292 354 351 351 351 390 395 395 168 360 360 360
100 23 77 99 9 41 99 100 89 89 99 89 100 100 100 100 100 100 100
+ 0 – + – 0 + + + 0 + + + + + + + + +
Continued
Table 4.2
Continued Old Party
Last Year 1995 1995 1999 1999 2001
Last Cong 104 104 106 106 107
New Party Party
Democrat Democrat Republican Democrat Independent
First Year 1995 1995 1999 1999 2002
First Cong 104 104 106 106 107
Party Republican Republican Democrat Independent Republican
Switcher’s Name
N Compared
Pct
Dir
360 360 0 0 0
99 100
+ +
Parker Tauzin Forbes Goode Goode
Note: Names appearing in Boldface exhibited significant behavioral changes in voting behavior following their party switch. Key to changes in direction (Dir): + Major party to major party switches, behavioral change in proper direction. – Major party to major party switch, behavioral change in wrong direction. 0 Switch to or from minor party. Source: Nokken and Poole (2004, 555).
Party Switching and the US House Agenda
95
among specific subsets of votes will help clarify the role that party affiliation plays in both the legislative and the electoral arena.
4.4.2 Policy or Procedure? Sources of Behavioral Change among Party Switchers The conventional wisdom on the consequences of party defection in the recent past is that switchers dramatically change their roll-call behavior at the same time they change party labels. What has yet to be addressed is whether members make a wholesale change among all categories of votes or whether the changes are concentrated among a few types of roll calls. I tackle this question by analyzing MCs’ preand post-switch behavior for three important categories of roll-call votes: final passage votes, amendment votes, and procedural votes. The second hypothesis here holds that members may well be mindful of the risks associated with “flip-flopping” on important final passage votes but exhibit large behavioral changes in their votes on amendments and procedural matters. Indeed, the party leadership in the House often grants members the latitude to break from the party line on final passage votes in return for supporting the party’s position on important procedural and amendment votes. In other words, parties are effective at asserting discipline on those votes that precede final passage and that best characterize US legislative parties as procedural coalitions. I compare the roll-call behavior of party switchers with their nonswitching colleagues, using a modified version of the DW-NOMINATE scores on three sets of roll-call votes: final passage, amendment, and procedural votes. Given the findings so far, noticeable shifts across all types of votes are likely. That said, investigating the changes within the three subsets of votes is still revealing. As the second hypothesis suggests, switchers provide the analytic leverage to determine whether the parties are successful in guiding the voting behavior of their members on procedural matters, even if they are unable to convert them on substantive final passage votes. As figure 4.2 documents, levels of partisan polarization increased sharply from 1953 to 2002, allowing evaluation of the third hypothesis on the effect of polarizing forces on roll-call behavior (Fiorina, Abrams, and Pope 2006; McCarty, Poole, and Rosenthal 2006). In a first cut at evaluating changes in behavior across subsets of votes, figures 4.3–4.5 display the DW-NOMINATE scores on final passage, amendment, and procedural votes for the five MCs
1
Distance between the Parties
0.9 House
R = .93 0.8
0.7
Senate
0.6
0.5
18 7 18 9 8 18 3 8 18 7 9 18 1 9 18 5 9 19 9 0 19 3 0 19 7 1 19 1 1 19 5 1 19 9 2 19 3 2 19 7 3 19 1 3 19 5 3 19 9 4 19 3 4 19 7 5 19 1 5 19 5 5 19 9 6 19 3 6 19 7 7 19 1 7 19 5 7 19 9 8 19 3 8 19 7 9 19 1 9 19 5 9 20 9 03
0.4
Figure 4.2 Party polarization, 1879 to 2006, distance between parties on 1st DW-NOMINATE dimension. Source: http://voteview.com/Polarized_America.htm#POLITICALPOLARIZATION
Final Passage Vote DW-NOMINATE Scores
0.7 0.6 0.5 0.4 0.3 0.2 Parker 0.1 0 −0.1
Deal
Hayes
Tauzin
Laughlin
−0.2 96
97
98
99
100
101
102
103
104
104
105
106
107
Congress
Figure 4.3 DW-NOMINATE scores for 96th to 107th Congresses on final passage votes: 104th Congress party switchers.
0.6 Amendment Vote DW-NOMINATE Scores
Deal 0.5 0.4 0.3 0.2
Parker Tauzin
0.1 0 −0.1
Laughlin
−0.2
Hayes
−0.3 96
97
98
99
100
101
102
103
104
104
105
106
107
Congress
Figure 4.4 DW-NOMINATE scores for 96th to 107th Congresses on Amendment votes: 104th Congress party switchers.
Procedural Vote DW-NOMINATE Score
0.6
Deal
0.5 Parker
0.4 0.3 0.2 0.1 0
Tauzin
−0.1 Hayes
−0.2
Laughlin
−0.3 96
97
98
99
100 101 102 103 104 104 105 106 107 Congress
Figure 4.5 DW-NOMINATE scores for 96th to 107th Congresses on procedural votes: 104th Congress party switchers.
98
Timothy P. Nokken
who switched from the Democrats to Republicans during the 104th Congress (Billy Tuazin [LA], Jamie Hayes [LA], Nathan Deal [GA], Mike Parker [MS], and Greg Laughlin [TX]). The three plots show a nearly identical pattern across vote types. In accord with the partisan-based argument underlying the first hypothesis, each MC dramatically changes his roll-call voting behavior at the time he or she switches parties, with pronounced and positive increases in DW-NOMINATE scores expected of a Republican MC. This focus on the single Congress with the most switching from 1953 to 2002 yields suggestive preliminary evidence in support of the third hypothesis, since high levels of polarization prevailed during the 104th (Nokken and Poole, 2004). A fuller examination is needed, however. For a more thorough analysis of the voting behavior of party defectors, I compare their roll-call behavior with that of their colleagues who do not change party affiliation. My analysis of the changes in switchers’ DW-NOMINATE scores across the three subsets of votes mirrors closely that of Nokken and Poole’s (2004) of the aggregate scores, discussed earlier. For those MCs who switched between Congresses, I use the difference between the scores in the first Congress with the new party label and the last Congress with the old party label. For those who switched during the course of a Congress, I treat the individual as two distinct members and calculate the difference in the pre- and post-switch scores. To determine the relative magnitude of the changes in voting behavior for the switcher, I compute the difference for all other members who served with the member across the two Congresses. I then rank order the set of members and calculate the percentile score for each switcher for the three subsets of votes. Unlike Nokken and Poole (2004), I make no assertions about levels of significance but simply report the percentile for each switcher. Table 4.3 compares voting choices on final passage, amendment, and procedural votes. For each type of vote, the right column on voting behavior identifies those switchers who move in the “right” or “wrong” direction; those trading Democratic for Republican labels should move to the right, and those making the opposite switch should move left. As the table 4.3 details, of the 20 major party switchers, 17 move in the expected direction on final passage votes, 19 on amendment votes, and 18 on procedural votes. Whereas it is difficult to determine the “correct” direction of change for most MCs who switch to or from Independent status,
Table 4.3 House party defectors’ change in DW-NOMINATE scores, 1953–2002, for three categories of votes Old Party
Last Year
Last Cong
Old Party
1957 1959 1965 1971 1973 1973 1973 1975 1979 1981 1981 1981 1983 1987 1987 1989 1995 1995 1995 1995 1995 1999 1999 2000 2001
85 86 89 92 93 93 93 94 96 97 97 97 98 100 100 101 104 104 104 104 104 106 106 106 107
Rep Ind. Dem Dem Rep Dem Ind. Dem Rep Rep Dem Ind. Dem Dem Dem Dem Dem Dem Dem Dem Dem Dem Dem Rep Dem Dem Ind.
New Party
First Year Cong 1957 1961 1967 1973 1975 1975 1975 1979 1981 1983 1983 1983 1985 1989 1989 1997 1995 1995 1995 1995 1995 1999 1999 2000 2002
85 87 90 93 94 94 94 96 97 98 98 98 99 101 101 105 104 104 104 104 104 106 106 106 107
Vote Category Final
New Party
Switcher’s Name
Dem Dem Rep Dem Rep Dem Dem Dem Rep Dem Rep Rep Rep Rep Rep Rep Rep Rep Rep Rep Rep Dem Ind Rep Rep
Dellay Alford Watson Reid Jarman Moakley Riegle Peyser Atkinson Foglietta Gramm Stump Ireland Grant Robinson Watkins Deal Hayes Laughlin Parker Tauzin Forbes Goode Martinez Goode
Key to changes in direction (Dir): + Major party to major party switches, behavioral change in proper direction. – Major party to major party switch, behavioral change in wrong direction. 0 Switch to or from minor party.
Amend
Procedure
N Compared
Pct
Dir
Pct
Dir
Pct
Dir
386 369 359 361 340 340 340 293 355 352 352 352 390 397 397 171 349 349 349 349 349 393 393 393 392
98 94 9 83 38 79 89 91 79 73 7 47 97 99 98 100 100 96 99 53 93 99 37 100 35
+ 0 + + – 0 – + + 0 – + + + + + + + + + + + 0 + 0
98 94 78 93 26 32 50 96 38 89 66 82 100 100 100 99 100 99 100 99 96 91 75 100 76
+ 0 + + + 0 – + + 0 + + + + + + + + + + + + 0 + 0
92 15 76 98 85 28 75 99 89 54 100 96 98 100 100 98 100 99 99 92 95 99 97 100 100
+ 0 – + – 0 + + + 0 + + + + + + + + + + + + 0 + 0
100
Timothy P. Nokken
Goode (D, I, and R, VA) is, once more, a case apart. Elected to the House as a Democrat in the 105th Congress, Goode defected and as an Independent caucused with the Republicans during the 106th Congress, and finally completed his partisan migration by joining the Republicans in the 107th Congress. The expectation would be that Goode should move to the right. Indeed, adoption of Independent status led Goode to compile more conservative voting records in all three categories of votes. His subsequent shift to the Republican Party saw him continue his rightward moves on final passage and procedural votes, understandably so since formal affiliation with the Republican Party made him eligible for party-controlled benefits and since strong evidence of party loyalty is an important prerequisite for career advancement. It is noteworthy that Goode received an assignment to the Appropriations Committee after officially affiliating with the Republican Party. I now turn to evaluating the magnitude of changes, using the percentile scores reported to the right of table 4.3 to compare switchers and nonswitchers. According to the second hypothesis, switchers should evince the largest changes on procedural and amendment votes. I first count switchers for whom a majority of nonswitchers exhibited larger changes in roll-call behavior across the two corresponding Congresses than did the switcher. On final passage votes, 6 of the 25 party switchers registered smaller changes than a majority of the nonswitching members; in particular, the voting records of Phil Gramm (D to R, TX) and Albert Watson (D to R, SC) placed them at the 7th and 9th percentile, respectively. This result is not a complete surprise since the last vestiges of the Conservative Coalition existed when Gramm and Watson joined the Republicans. On amendment votes, only 3 switchers fell below the 50 percent threshold: John Jarman (OK), John Moakley (MA), and Eugene Atkinson (PA). Only 2, Thomas Alford (AR) and John Moakley (MA), failed to exceed the 50 percent mark on procedural votes. Note that Alford and Moakley held Independent status before their switches. I next focus on switchers whose change in voting behavior placed them at or above the 90th percentile. On final passage votes, 12 of the 25 switchers placed at or above the 90th percentile, and the relevant figure was 15 on amendment votes and 18 on procedural votes. This finding conforms with the second hypothesis. Finally, I compare each switcher’s percentile rank on final passage votes with his ranks on amendment and procedural votes. For 10 cases, a member’s percentile ranks on one or both of the categories
Party Switching and the US House Agenda
101
of amendment and procedural scores exceeded the rank for final passage votes. For two, switchers’ ranks were identical across categories. Near-identical percentiles appear for 10 switchers (in chronological order of switch, Dellay, Peyser, Ireland, Grant, Robinson, Watkins, Hayes, Laughlin, Tauzin, and Forbes). Only for three switchers (two of whom traded Independent status for party labels) do percentile ranks on final passage votes surpass their ranks on amendment and procedural votes. The evidence in favor of the second and the third hypotheses becomes even stronger when the timing of the (near-) identical voting behavior is considered. All but 2 of the 12 instances total of (near-) identical behavior occurred after the mid-1970s, when partisan polarization increased sharply. It is true that important changes in roll-call behavior associated with party defection take place on the less visible but still crucial procedural and amendment votes, as consistent with the second hypothesis and more generally with the notion of legislative parties as procedural coalitions. Moreover, in accord with the third hypothesis, the differences between the three categories of votes fade in phases of partisan polarization, as votes on final passage are aligned with party label.
4.5 Conclusions It is somewhat ironic that quantitative studies of party switchers in the United States focus squarely on the outliers in the data. My premise is that investigating these seemingly anomalous individuals is of theoretical interest to scholars of legislative politics. Despite the “small N” problem, the study of party switchers allows analysts of US politics to understand more fully how party labels and partisan structures within the legislative arena shape (or fail to shape) the behavior of individual MCs. Lacking individual-level variation on party affiliation makes it difficult to ascertain just how membership in a party may influence an individual. Even where switches are rare, they offer insight into broad questions in politics. Here I extend my previous work on the roll-call behavior of party switchers to probe more deeply into how roll-call behavior shifts when House members change parties. Extant research provides persuasive evidence that party switchers, especially those in recent times, have made significant changes to their roll-call behavior upon joining a new party. Given the tendency for individual MCs to stake out ideological turf and strive to maintain a consistent voting record,
102
Timothy P. Nokken
such dramatic changes are somewhat surprising. I seek to ascertain whether party switchers change their roll-call behavior across the board. Do they change their votes on substantive policy as much as on procedural matters? Are differences observed regardless of the degree of partisan polarization? By disaggregating roll-call votes into specific categories, I take a first step toward answering these questions. Although the switchers studied here exhibit sizable shifts in their voting behavior across all types of votes, changes on amendment and procedural votes often exceed those observed on final passage votes. Moreover, this tendency is most pronounced in phases of relatively low polarization. When polarization is high, greater similarities appear in switchers’ roll-call behavior on final, amendment, and passage votes. These results shed light on several major issues in legislative and party politics. Legislative scholars have focused on the construction of congressional agendas and the central role that rules and procedures play in agenda-setting in an effort to understand the importance of partisan institutions. Party coalitions seek to gain a majority in the House and, with majority status, seek to create and implement rules and procedures favorable to the majority. The parties engage in bitter fights because they realize that by controlling procedures, they gain an upper hand on substantive policy matters. Policy accomplishments thus depend on a cohesive majority on procedural matters. Scholars have noted the increasingly cohesive parties on procedural matters and illustrated that cohesion can be translated into partisan victories on policy matters (Aldrich and Rohde 2000; Cox and McCubbins 2002, 2005; Rohde 1991). If these are apt depictions and if US legislative parties are best characterized as procedural coalitions, parties may well start to welcome new members despite how those members vote on final passage as long as they lend their enthusiastic support to party procedural positions. The findings here are consistent with that perspective and show that investigating member behavior on amendment and procedural votes is a fruitful area of inquiry. Perhaps most broadly, the differences I have documented in representatives’ voting behavior across distinct categories of votes carry implications for the real-world workings of representation in US politics. Especially in periods of low partisan polarization, legislators who switch parties appeal to (perceived) continuities in their constituents’ policy preferences by maintaining continuity on final passage votes. Although this pattern fits nicely with the hoary assumption that
Party Switching and the US House Agenda
103
politicians are reelection seekers, it also forces us to reevaluate our understanding of representation. If, as Cox and McCubbins (2005) argue, the heart of policy making lies not in final passage votes, but rather in procedural votes on rules and amendments (and, above all, legislative leadership), then the changes in party switchers’ behavior that I have demonstrated here could overwhelm the more easily observed continuities on final passage. Switching in the US Congress, rare as it is, carries important consequences for the representation of the policy preferences of US voters. More generally, the disjunction in switchers’ behavior between final-passage and procedural votes lends credence to the claim that “party matters” even in the United States and invites taking MCs’ claims of independence with a healthy dose of skepticism.
4.6 Appendix A: Roll-Call Vote Types 4.6.1 Final Passage Amendments to the Constitution Final Passage/Adoption of a Bill Final Passage/Adoption of a Conference Report Final Passage/Adoption of a Bill Final Passage/Adoption of Conference Report Final Passage/Adoption of Resolution Final Passage/Adoption of Joint Resolution Passage/Adoption of a Bill under Suspension of the Rules Passage/Adoption of a Joint Resolution under Suspension of the Rules Final Passage/Adoption of Concurrent Resolution Passage/Adoption of a Concurrent Resolution under Suspension of the Rules Passage/Adoption of a Resolution under Suspension of the Rules Passage over Presidential Veto Adoption of First Part of Resolution Adoption of Second Part of Resolution Suspension of Rules for Conference Report Motion to Suspend the Rules and Concur
4.6.2
Amendment Votes
Straight Amendments (includes en bloc and amendments in the nature of a substitute) Amendments to Amendments
104
Timothy P. Nokken
Substitute (to an amendment) Motion to Table Amendment Amendment to Amendment to Substitute Perfecting Amendment Amendment to Substitute Perfecting Amendment to Substitute Suspension of Rules to Amend Bill Amendment to Special Rule
4.6.3 Procedural Votes Budget Waivers Motion to End Debate Motion to Rise from the Committee of the Whole Motion to Disagree Passage of Rules (Special Rule) Motion to Recede Motion to Order Previous Question (Note: Previous Question on Special Rule is 99) Election of Speaker Motion to Recommit (Note: Recommit to Conference is 72) Motion to Instruct Conferees Motion to Recede and Concur (also includes motion to concur) Previous Question on Special Rules Source: Rohde’s Vote Type Codebook (2004) for Rohde/PIPC Roll Call Database (83rd to 107th Congress).
4.7
Notes
* I would like to thank Mike Crespin, Maria Escobar-Lemmon, Chuck Finocchiaro, Craig Goodman, Keith Poole, Lucio Renno, and the Research Group on Legislative Party Switching for helpful comments and suggestions. I would also like to acknowledge Keith Poole and David Rohde for providing data utilized in this chapter. Special thanks to Keith Poole for his generous help in calculating the DW-NOMINATE scores utilized in this analysis. I also want to recognize Carol Mershon and William Heller for their work in creating the Research Group on Legislative Party Switching and spurring me to think hard about placing party switching in the United States within a larger comparative framework. 1. Nokken and Poole (2004) define a party switch to include nearly any instance of a member adopting a new party label during three stable partisan eras in the House and Senate (1793–1813, 1827–1849, and 1877–1998). They use Martis’s (1989) party codes to identify all MCs whose party label changed
Party Switching and the US House Agenda
2. 3.
4.
5. 6. 7.
8.
9. 10.
11.
12.
13. 14.
105
over the course of their careers. Oppenheimer and Hatcher (2003) present an alternative list of switchers derived from the use of more restrictive definitions of switching. I acknowledge Brian Gaines for elucidating on this predictive model of party defection in the United States. Aldrich and Bianco argue that choices of party affiliation could be considered an additional component of the strategic politician theory developed by Jacobson and Kernell (1983). DW-NOMINATE scores are generated from an individual’s roll-call behavior by using a multidimensional scaling procedure. The scores range from –1, the most liberal value, to +1, the most conservative. For a member switching parties during a legislative session, we split the member’s roll-call record at the time the switch was announced, essentially treating him or her as a new MC upon changing parties. Procedural votes are especially important in the House of Representatives, as discussed in the following text. See Poole (2005) and Goodman (2004) for similar analyses of roll-call behavior in other settings. A number of scholars emphasize the importance of procedural matters in the House. See Sinclair (2007) on the majority party’s use of “unorthodox” procedures to pass legislation and both Roberts and Smith (2003) and Theriault (2007) on the contribution of procedural matters to the high levels of polarization in Congress. Virgil Goode (VA) switched twice, as reported. Two MCs appear in table 4.1 but are not counted in the totals just cited: Ralph Hall (TX) and Rodney Alexander (LA) both switched from the Democrats to the Republicans in the 108th. These switches came too late to allow the calculation of subsetspecific DW-NOMINATE scores. My decisions on which vote types to assign to the respective categories essentially mirror those of Crespin, Rohde, and Vander Wielen (2002). The mean number of final passage votes was just more than 308, with a minimum of 73 and maximum of 629. Amendment votes ranged from 10 to 590 and averaged just more than 243, whereas procedural votes ranged from 39 to 259 and averaged just more than 138. Generating ideal point estimates for the early congresses with such a small number of votes may well produce scores with large standard errors, hence, questionable precision (personal communication with Keith Poole). The explosion of recorded votes in the Committee of the Whole helped trigger the Rules Committee’s effort to use creative rules so as to regain control of the amendment process in the Committee of the Whole. The data are available at http://www.voteview.com/dwnomin.htm, file HL01108A1_PRES.DAT. It is reasonable to characterize Goode’s switch as a straight move from the Democrats to the Republicans since he caucused with and received committee assignments from the Republicans while serving as an Independent.
106
Timothy P. Nokken
4.8 References Aldrich, John H., and William T. Bianco. 1992. “A Game-Theoretic Model of Party Affiliation of Candidates and Office Holders.” Journal of Mathematical and Computer Modelling 16: 103–116. Aldrich, John H., and David W. Rohde. 2000. “The Consequences of Party Organization in the House: The Role of the Majority and Minority Parties in Conditional Party Government.” In Polarized Politics: Congress and the President in a Partisan Era, Ed. Jon R. Bond and Richard Fleisher. Washington, DC: CQ Press. ———. 2001. “The Logic of Conditional Party Government: Revisiting the Electoral Connection.” In Congress Reconsidered, 7th Edition, ed. Lawrence C. Dodd and Bruce I. Oppenheimer. Washington, DC: CQ Press. Asher, Herbert B., and Herbert F Weisberg. 1978. “Voting Change in Congress: Some Dynamic Perspectives on an Evolutionary Process.” American Journal of Political Science 22 (May): 391–425. Bach, Stanley, and Steven S. Smith. 1988. Managing Uncertainty in the House of Representatives: Adaptation and Innovation in Special Rules. Washington, DC: Brookings. Brady, David, and Edward P. Schwartz. 1990. “Ideology and Interests in Congressional Voting: The Politics of Abortion and the U.S. Senate.” Public Choice 84 (July): 25–48. Canon, David T. 1992. “The Emergence of the Republican Party in the South, 1964–1988.” In The Atomistic Congress: An Interpretation of Congressional Change, ed. Allen D. Hertzke and Ronald M. Peters, Jr. Armonk, NY: M.E. Sharpe. Canon, David T., and David J. Sousa. 1992. “Party System Change and Political Structures in the U.S. Congress.” Legislative Studies Quarterly 17: 347–363. Castle, David, and Patrick Fett. 1996. “A Predictive Model of Congressional Party Switching.” Manuscript. Beaumont, TX: Lamar University. ———. 2000. “Member Goals and Party Switching in the US Congress.” In Congress on Display, Congress at Work, ed. William T. Bianco. Ann Arbor: University of Michigan Press. Cox, Gary W., and Mathew D. McCubbins. 1993. Legislative Leviathan: Party Government in the House. Berkeley: University of California Press. ———. 2002. “Agenda Power in the US House of Representatives, 1877–1986.” In Party, Process, and Political Change in Congress: New Perspectives on the History of Congress, ed. David W. Brady and Mathew D. McCubbins. Stanford, CA: Stanford University Press. ———. 2005. Setting the Agenda: Responsible Party Government in the U.S. House of Representatives. New York: Cambridge University Press. Crespin, Michael H., David W. Rohde, and Ryan J. Vander Wielen. 2002. “Variations in Party Voting in the House of Representatives, 1953–2000.” Paper presented at the annual meetings of the Southern Political Science Association, Savannah, GA.
Party Switching and the US House Agenda
107
Desposato, Scott. 2006. “Parties for Rent? Careerism, Ideology, and Party Switching in Brazil’s Chamber of Deputies.” American Journal of Political Science 50 (1): 62–80. Dougan, William R., and Michael Munger. 1989. “The Rationality of Ideology.” Journal of Law and Economics 32 (April): 119–142. Fiorina, Morris P., Samuel J. Abrams, and Jeremy C. Pope. 2006. Culture War? The Myth of Polarized America. New York: Pearson Longman. Goodman, Craig. 2004. Ideological Stability in Congress: Experiments in Roll Call Voting. Doctoral Dissertation, University of Houston. Grose, Christian R. 2004. “Is It Better to Join the Majority? The Electoral Effects of Party Switching by Southern State Legislators, 1972–2000.” American Review of Politics 25: 79–98. Grose, Christian R., and Antoine Yoshinaka. 2003. “The Electoral Consequences of Party Switching by Incumbent Members of Congress, 1947–2000.” Legislative Studies Quarterly 28 (February): 55–75. Heller, William B., and Carol Mershon. 2005. “Party Switching in the Italian Chamber of Deputies, 1996–2001.” Journal of Politics 67 (May): 536–559. Jacobson, Gary C., and Samuel Kernell. 1983. Strategy and Choice in Congressional Elections, 2nd Ed. New Haven, CT: Yale University Press. King, Gary, and Gerald Benjamin. 1986. “The Stability of Party Identification among US Representatives: Political Loyalty, 1789–1984.” Paper presented at the Midwest Political Science Association Convention. Krehbiel, Keith. 1993. “Where’s the Party?” British Journal of Political Science 23 (April): 235–266. Lott, John R., Jr., and Stephen G. Bronars. 1993. “Time Series Evidence on Shirking in the U.S. House of Representatives.” Public Choice 76 (June): 125–150. Martis, Kenneth C. 1989. The Historical Atlas of Political Parties in the United States Congress 1789–1989. New York: Macmillan. McCarty, Nolan M., Keith T. Poole, and Howard Rosenthal. 2001. “The Hunt for Party Discipline in Congress.” American Political Science Review 95 (September): 673–687. ———. 2006. Polarized America: The Dance of Ideology and Unequal Riches. Cambridge, MA: MIT Press. Mershon, Carol, and Olga Shvetsova. 2008. “Parliamentary Cycles and Party Switching in Legislatures.” Comparative Political Studies 41 (1): 99–127. Nokken, Timothy P. 2000. “Dynamics of Congressional Loyalty: Party Defection and Roll Call Behavior, 1947–1997.” Legislative Studies Quarterly 25 (August): 417–444. Nokken, Timothy P., and Keith T. Poole. 2004. “Congressional Party Defection in American History.” Legislative Studies Quarterly 29 (November): 545–568. Oppenheimer, Bruce I., and Andrea C. Hatcher. 2003. “Congressional Party Switchers, 1876–2002: The Effect of Party and Constituency on Strategic Behavior.” Paper presented at the annual meetings of the Midwest Political Science Association, April 3–6, Palmer House Hilton, Chicago, IL. Poole, Keith T. 2005. Spatial Models of Parliamentary Voting. Cambridge and New York: Cambridge University Press.
108
Timothy P. Nokken
Roberts, Jason M., and Steven S. Smith. 2003. “Procedural Contexts, Party Strategy, and Conditional Party Voting in the US House of Representatives, 1971–2000.” American Journal of Political Science 47 (April): 305–317. Rohde, David W. 1991. Parties and Leaders in the Postreform House. Chicago: University of Chicago Press. ———. 2004. Roll Call Voting Data for the United States House of Representatives, 1953–2002. East Lansing: Compiled by the Political Institutions and Public Choice Program, Michigan State University. Sinclair, Barbara. 1981. “The Speaker’s Task Force in the Post-Reform House.” American Political Science Review 75 (June): 397–410. ———. 2007. Unorthodox Lawmaking: New Legislative Processes in the U.S. Congress. Washington, DC: CQ Press. Thames, Frank C. 2007. “Searching for the Electoral Connection: Parliamentary Party Switching in the Ukrainian Rada, 1998–2002.” Legislative Studies Quarterly 32 (May): 223–256. Theriault, Sean M. 2007. “Procedural Polarization in the US Congress.” Manuscript. Austin: University of Texas. Wattenberg, Martin P. 1994. “Dealignment in the American Electorate.” In Parties and Politics in American History: A Reader, ed. L. Sandy Maisel and William G. Shade. New York: Garland. ———. 1998. The Decline of American Political Parties, 6th ed. Cambridge, MA: Harvard University Press. Yoshinaka, Antoine. 2005. “House Party Switchers and Committee Assignments: Who Gets ‘What, When, How?’ ” Legislative Studies Quarterly 30 (August): 391–406.
5 Party Switching in Brazil: Causes, Effects, and Representation Scott Desposato
5.1 Introduction Brazil has a long history of frequent party switching by sitting legislators. Although there are other countries with higher rates of switching, most are transitioning democracies with evolving party systems. In such contexts, switching goes hand in hand with less-developed party brands, weaker voter partisan attachments, and evolving norms and procedures of recruitment and hierarchy. The Brazilian case appears to represent a stable long-term equilibrium, with fairly steady rates of switching over the last 20 years: approximately one-third of deputies have switched in each legislature since democratization. Table 5.1 shows switching rates in the last several legislatures. During the 49th legislature (1991–1994), there were 262 incidents of switching, for an average switching rate of .52 per legislator-term. That figure slips slightly to .41 in the second period, and rises to .51 in the third.1 Switching was prohibited for most of Brazil’s authoritarian period (1964–1985), but scholars of previous periods have noted high rates of switching in the 1960s and even as far back as the 1800s (Graham, 1990): Switching apparently was common during the Second Republic (1946–1964), and there is evidence as well of frequent switching during the First Republic (1889–1930). As in the other country cases discussed in this volume, switching in Brazil is not viewed favorably by citizens, pundits, or scholars. Observers decry the lack of backbone or consistency in switchers, and look with envy at highly disciplined stable systems. More fundamentally, switching potentially jeopardizes accountability and representation. Voters may support a candidate based on his party label or coalition membership—switching to another party appears a
110 Table 5.1
Scott Desposato Party switching rates in the Brazilian Chamber of Deputies Legislative Session
Number of Switches Switching Rate Number of Switchers
49th
50th
51st
52nd
262 .52 198
212 .41 169
262 .51 183
135 .54* 130
* current term, projected rate.
complete betrayal of the electoral compact between voter and elected official. Further, these potential problems are exacerbated in a system with vote-pooling, such as Brazil’s. Under vote pooling, a vote for one candidate may help elect another on the list, who may then switch party. Without reliable information about party membership, how can votes have any meaning? In this chapter, I explore the causes and effects of party switching in Brazil. To tackle the first question, I offer a general model of why party switching may take place, discuss implications for Brazil, and test it empirically using data from two legislative sessions. For the second question, I examine the impact of party switching on party cohesion, legislators’ voting behavior, and representation. I conclude by examining the impact of recent developments, most notably a supreme court ban on party switching imposed in March 2007.
5.2 Causes of Party Switching Why switch party? I aim not only to answer this question for the Brazilian case, but also to offer a generalizable model that can capture different kinds of payoffs to membership, different kinds of party structure, and endogenous party leadership decisions. My approach is similar to the models adopted in other studies of party switching, but has one important difference. Most work reduces switching to a fundamentally binary decision—the choice to switch or not to switch. This approach implies a logit or probit empirical model. I adopt a multichoice modeling strategy, examining the specific choice of party destination—not just the decision to switch or not. 2 In multiparty systems, switchers may join any of many alternative parties, and not all destinations are the same. To fully
Party Switching in Brazil
111
understand the motives behind a switch, I argue that we cannot just look at a binary choice set of switch or not switch. Much of a switcher’s calculus is about the potential payoffs of changing party, and some of those payoffs are specific to destinations. For example, consider a study of the role of pork in party switching. If we were modeling binary switching, we might observe higher switching rates out of opposition than government parties. This appears to be evidence that pork matters for party affiliation. However, if we did not know which party switchers joined, we would not fully understand their motives. For instance, if all switchers simply move from one opposition party to another, that would be evidence that belonging to the governing coalition is not important—or that the governing coalition does not accept switchers. The key point is that we should not just model the binary decision to switch or not switch in multiparty systems. Instead, we must model the full set of alternative choices and their associated payoffs to understand actors’ incentives. A Typology of Party Systems To begin, I propose conceptualizing legislative party systems according to the nature of the benefits that party membership provides and the extent to which parties control access to such benefits, as summarized in table 5.2. The benefits of party membership may be classified either as private or as club goods. Private goods are both rival and exclusive, meaning that their consumption reduces the amount of that good available to others, and others can be excluded from enjoying the same good. Examples in legislatures often include pork, committee assignments, and nominations for elected office. Club goods fall between private and public goods. They are nonexclusive only for club members, all of whom automatically enjoy equal access. For example, one club good is the electoral value of a party label. All members automatically enjoy (or suffer) a party’s reputation or performance in government.3 Table 5.2
Typologies of legislative party systems Payoff Club
Type Private
I III
II IV
Party Control? No Yes
112
Scott Desposato
The second dimension of my typology is the extent to which political parties control access to the benefits of membership. In the case of private goods benefits, a party might make all decisions about distribution. For example, in most closed-list systems, ballot positions for upcoming elections may be considered a private good controlled by the party leaders. In other systems, private goods need not be controlled by parties. For instance, nominations in the United States are still effectively private goods but are controlled by voters in primary elections, not party leaders. With club goods, parties cannot restrict members’ enjoyment of benefits. Again, all members enjoy a party’s brand name or reputation. Parties do control possession of club goods, however, when they can restrict membership to a subset of legislators, repelling or expelling unwelcome members. The framework proposed by Aldrich and Bianco (1992) is a special case of type II, where politicians compete in a single-member district for a private good (a nomination or seat) without any interference from party leaders. Legislators switch party to maximize their expected career utility, a function of the resources they receive from their party of choice, less a switching transaction cost. Transaction costs can take many forms, codified in law or imposed by informal institutions. For example, Nepal, Ecuador, and Japan have passed laws whereby party switchers are expelled from office, losing their mandates.4 Alternatively, transaction costs can be imposed by voters. Where the electorate or party militants are very partisan or otherwise use parties as information cues, but voters can influence the choice of candidate, switchers will usually have little credibility and great difficulty attracting votes or campaign support. Parties will offer resources and invite switchers when the benefits of welcoming a new member exceed the costs of resources offered. The value of a new member will vary greatly across systems, depending on political institutions and voters’ characteristics, but at least two benefits to accepting new members are obvious. First, in all legislatures, size matters. On average, larger parties have more political influence, more cabinet positions, better committee assignments, and more pork. The relationship between size and resources may occasionally be linear, when resources are distributed proportionately. In such a case, each additional member brings equal value in party resources. But in winner-take-all systems and multiparty coalition governments, the value of an additional member depends on whether he or she changes the balance of power. Second, switchers may bring with them electoral support for their new party. Especially popular
Party Switching in Brazil
113
politicians, often celebrities, may significantly increase a party’s vote share. Such a benefit, however, depends on the electoral context. Voting must be personalistic enough that voters will continue to vote for a switcher and the electoral system must pool votes, so that other party members will directly benefit from the influx of popular support (on pooling, see, e.g., Carey and Shugart [1995]). Formalizing the Model When formalized, this logic makes predictions about the patterns and frequency of switching across systems. I illustrate with a simple model, following Aldrich and Bianco (1992). Let a legislature have two parties, A and B, and two legislators, 1 and 2. Legislator 1 starts in party A; legislator 2 starts in party B. Each legislator has an additional exogenous characteristic, gi, that is their contribution or value-added to a club good of the party they join. For example, a very popular politician might bring votes to a new party. The contribution gi need not be positive; a scandal-ridden legislator might have a negative value-added in a party. Parties have exogenous resource endowments uj and endogenous resource endowments f(Gj). Whereas uj may be a private or club good, f(Gj) is a club good enjoyed by all members. The latter’s value is determined by f, a strictly increasing function, and Gj, the sum of the gi of all members of party j: Gj = ∑gij. When possible, parties exercise control over access to resources by restricting membership. Specifically, members of each party cast simultaneous votes to determine which legislators will be allowed to join or stay, and which will be refused admission. Party members—themselves legislators in my model—cast these votes to maximize their own expected payoffs. Legislators switch party or stay put to maximize their expected utility, a function of the resources they will receive in their party of choice, less any transaction costs. Legislator i’s utility associated with membership in party j is: uij = αijuj + f(Gj) − TiIj≠home where uj is party j’s resources, αij is legislator i’s share of party j’s resources, f(Gj) is the value of the public good contribution of all party members to party j, Ti is a transaction cost, and Ij≠home is an indicator variable coded “1” if party j is not i’s current party, “0” otherwise. The precise form that payoffs and f(Gj) takes will vary with
114
Scott Desposato
the specific goods that contribute to parties’ value for and contributions to legislators, and legislators’ value-added for parties. The game is played as follows: 1. If parties have control over membership, they simultaneously decide whether to accept or reject each legislator i via majority vote of current members. 2. Legislators make membership decisions—switching or staying put— and receive payoffs. 3. The game ends.
For the simple case with one legislator per party, the payoffs of membership for each legislator are summarized in table 5.3. Legislators start in the upper-right hand cell: (A, B). Note that where parties have control over access, a decision to reject a member eliminates that legislator’s options. For example, if Party A, controlled by legislator 1, decides to exclude legislator 2, the first column from table 5.3 is eliminated, and the game is reduced to legislator 1 choosing between Party A and Party B. Virtually, any standard game can emerge from combinations of parameter values, functional forms, and levels of party control. When legislators can freely join any party, that is, parties cannot restrict access, the most common results include the following: • Nobody Move: When T is big, both legislators stay put in their current party. • All Aboard!: When the payoff differential |uA − uB | is big enough, both legislators join the resource-rich party. • Battle of the Sexes: When f(g 1 + g 2) is big enough and all gi positive, both legislators want to be in the same party, but each prefers that the other switch and pay the transaction cost Ti. Similarly, when u is a private good or both gi are less than zero, that is, each legislator has a negative value-added, legislators may play BOTS (call this variant BOTS2) but prefer to be in opposite parties. • Fashion: One legislator prefers to share a party with the other, but the other prefers to be alone. This result emerges when one legislator has a sufficiently large positive gi and the other a sufficiently negative one, and leads to mixed-strategy equilibrium. • Get out! Without party control, a legislator with a negative gi , starting in a resource-poor party, could join the other party and force the other legislator to abandon ship. The first legislator could then keep all the resources of the wealthy party for himself or herself.
Party Switching in Brazil Table 5.3
115
Payoffs of party membership Legislator 2
Legislator 1
Party A
Party B
Party A Party B
α1AuA + f(g 1, g 2), α2AuA + f(g 1, g 2) − T uB − T + f(g 1), uA − T + f(g 2)
uA + f(g 1), uB + f(g 2) α1BuB + f (g 1, g 2) − T, α2BuB + f(g 1, g 2)
Plot III: T=0, G=(1,–1) 5
4
4
3
3
θA
θA
Plot I: T=0, G=(1,1) 5
2
2
1
1
0
0 0
1
2
3
4
5
0
1
2
3
4
5
θB
θB Plot II: T=2, G=(1,1)
Plot IV: T=2, G=(1,–1)
4
4
3
3 θA
5
θA
5
2
2
1
1
0
0 0
1
2
3 θB
Figure 5.1 control.
4
5
0
1
2
3
4
5
θB
Party membership equilibria strategies—public goods without party
In the case where parties can reject new members, the range of parameters that lead to switching equilibria is much smaller, and some are eliminated entirely. For example, for BOTS2, Fashion, and Get Out!, the resource-rich party will simply reject unwanted new members.5
116
Scott Desposato
Figures 5.1 to 5.4 show predicted equilibria for different types of goods u, different degrees of party control, and different values of T and gi, with α = 1 for club goods and ␣ = n1j for private goods, where nj is the number of members of party j. Each of the four figures corresponds to one of the ideal system types of table 5.2. Within each figure, the four plots show equilibria for different values of T and G, as labeled. Combinations of uA and uB that lead to All Aboard! are black, those u that lead to Nobody Move are white, and u that lead to mixed or other switching equilibria are grey. Thus systems with switching have large shaded regions; systems where switching is rare have large blank regions. Code to produce similar graphs and calculate payoffs for different types of goods and values of all parameters is available on my Web site (http://swd.ucsd.edu).
Plot III: T=0, G=(1,–1)
5
5
4
4
3
3
θA
θA
Plot I: T=0, G=(1,1)
2
2
1
1 0
0 0
1
2
3
4
5
0
1
2
3
4
5
θB
θB Plot II: T=2, G=(1,1)
Plot IV: T=2, G=(1,–1)
4
4
3
3 θA
5
θA
5
2
2
1
1
0
0 0
1
2
3 θB
Figure 5.2 control.
4
5
0
1
2
3
4
5
θB
Party membership equilibria strategies—private goods without party
Party Switching in Brazil Plot III: T=0, G=(1,–1)
5
5
4
4
3
3
θA
θA
Plot I: T=0, G=(1,1)
117
2
2
1
1 0
0 0
1
2
3
4
5
0
1
2
3
4
5
θB
θB Plot II: T=2, G=(1,1)
Plot IV: T=2, G=(1,–1)
4
4
3
3 θA
5
θA
5
2
2
1
1
0
0 0
1
2
3 θB
Figure 5.3 control.
4
5
0
1
2
3
4
5
θB
Party membership equilibria strategies—public goods with party
Comparing the figures generates predictions about the conditions under which we will observe party switching, and the direction of such switching. The most important of these predictions include: • Increasing transaction costs decreases switching. Within all of the figures, moving from T = 0 to T = 2 significantly reduces switching. For example, in figure 5.1, compare plot I with II. In I, there is always unconstrained switching. In the plot immediately below that one, the range of u that leads to switching falls to 64 percent. A similar vertical comparison of plots in almost all figures yields different percentage changes, but identical results—the range of u that leads to switching falls in every case. • Increasing resource differentials |uA − uB | leads to All Aboard! Both legislators join the well-endowed party, with important qualifications. Without party control, increasing party resource differentials lead to switching as legislators play All Aboard or other strategies, for both
118
Scott Desposato
kinds of goods. The range of u that leads to switching is much larger for club goods than for private goods, because club goods are nonrival. With party control and club goods, the impact of resource differentials is unchanged. An important exception arises with party control and private goods. In this scenario, all switching is less likely as party resource differentials increase. When parties control valuable private goods, they will not share them unless a new member brings a gi that exceeds the cost of sharing. In the simple illustration, parties only accept switcher i when gi > 2j ⋅ • Negative gi decrease switching; positive gi increase switching. Where parties control access and gi is positive, party j will accept new member i, except when uj is a private good and 2j > gi . But where gi is negative, parties will always exclude that member if they can. Where parties cannot restrict membership, a negative gi reduces the range of All Aboard! and increases the ranges of mixed-strategy and other equilibria discussed earlier. Plot III: T=0, G=(1,–1) 5
4
4
3
3
θA
θA
Plot I: T=0, G=(1,1) 5
2
2
1
1 0
0 0
1
2
3
4
5
0
1
2
3
4
5
θB
θB Plot II: T=2, G=(1,1)
Plot IV: T=2, G=(1,–1)
4
4
3
3 θA
5
θA
5
2
2
1
1
0
0 0
1
2
3 θB
Figure 5.4 control.
4
5
0
1
2
3
4
5
θB
Party membership equilibria strategies—private goods with party
Party Switching in Brazil
119
Adding more legislators and more parties increases the complexity of the model considerably, but the basic implications remain unchanged. How Does It Work in Brazil? In many ways, Brazil fits all the prerequisites for relative frequent party switching. There are relatively low transaction costs, payoffs that are either club goods or uncontrolled private goods, large resource differentials, and positive value-added for most legislators. Effectively, most Brazilian deputies fall somewhere in figures 5.1 and 5.3, Plots I and II. The parameters of the system create and structure many opportunities for defection from parties. I turn now to consider the specifics of the Brazilian case and the incentives for party-switching. Transaction Costs Until recently, transaction costs were manageably low for many Brazilian legislators. The legal environment was very accommodating: there were no legal prohibitions on party switching, except immediately before elections for those seeking reelection.6 Until recently, all incumbents were guaranteed renomination in the next election under the candidato nato or “birthright” rule.7 Further, personalistic voting means that switchers can take their voters with them. Partisanship in Brazil is relatively low (Samuels, 2006), and most switchers I interviewed expected no or little repercussions in terms of electoral support. Many seemed to think that voters simply did not care about party membership. In one interview, the legislator’s receptionist was not even aware that he had switched parties! Some legislators, however, face significant transaction costs for switching. A few Brazilian parties have aggressively cultivated partisan followings, building grassroots organizations, promoting cohesive policy platforms, and encouraging partisan, not personal, votes in campaigns. Most notable of these is the Workers’ Party, or PT, but some smaller parties (including the Green Party) also have more ideological popular followings (Ames, 2001; Keck, 1995; Mainwaring, 1999; Samuels, 2006). And even catch-all parties may have a few committed partisan supporters. Legislators with such constituencies cannot take their voters with them when changing party—and should be much more likely to stay put. Benefits of Party Membership Brazilian parties can facilitate access to pork, provide ideological or policy benefits, and can directly enhance electoral success. One of the
120
Scott Desposato
unique features of the Brazilian system is that the primary benefits of party membership are either club goods or uncontrolled private goods, facilitating a fluid party system. I discuss the nature of benefits and parties’ role in their distribution in what follows. Pork A major benefit of party membership is privileged access to state resources. In the case of Brazil, scholars have widely acknowledged the importance of state resource distribution for deputies’ careers (Abrúcio, 1998; Ames, 2001, 1995a, b; Leal, 1977; Mainwaring, 1997, 1999; Pereira and Mueller, 2004; Samuels, 2003). Government resources can be channeled into pork projects that provide lucrative construction contracts and needed infrastructure. Cushy jobs in the bureaucracy reward campaign workers and maintain politicians’ support networks for the next election. Resources might also be used in less legal ways—for direct campaign activities or even vote-buying. Public resource distribution is controlled by the executive branch, not political parties. For legislators, access to public funds and resources is maximized by joining the governing coalition and voting for the executive’s proposals. Brazilian Presidents typically use state resource distribution to build multiparty supermajority coalitions. Supermajorities are required because legislators, even in the governing parties, are hesitant to support controversial economic reforms that impose significant costs on their constituents.8 Finally, there are stark differentials in access to resources across parties. Legislators in opposition parties are unlikely to receive any government projects for their constituents. Two empirical implications follow. First, government parties should retain and attract members, given large resource differentials and open markets for votes and pork. Second, since pork is a rival good, governing parties’ attractiveness may decline as the coalition grows large, and the spoils are widely shared. A similar logic applies to subnational coalitions. Brazil’s decentralized federalism grants subnational political actors—especially state governors—control over resources that national legislators covet (Abrucio, 1998; Mainwaring, 1999; Samuels, 2003). State governors’ influence should only be enhanced by progressive ambition in Brazil. Many national deputies aspire to state and local elected and appointed positions. Governors directly control key state appointments and can facilitate election to subnational office by providing campaign resources, endorsements, and selective pork distribution.
Party Switching in Brazil
121
Consequently, governismo—a tendency to join the governing coalition to enhance resource access—ought to extend to subnational governments, making state governors’ parties more attractive than state opposition parties. Electoral Payoffs Brazilian parties play a limited, but important, role in elections. In many political systems, parties provide many electoral resources: campaign workers, financing, consulting services, and meaningful labels to cue voters. By contrast, most Brazilian parties play little or no active role in legislative campaigns—parties do not provide candidates with volunteers, campaign finance, or consulting services, all resources that candidates must generate on their own. Voters generally do not use party labels as information cues; most citizens cast votes based on candidates’ reputations, not based on their partisanship (Mainwaring, 1999, 122). But party membership still can have an important effect on electoral outcomes. A feature of the Brazilian electoral system, open-list proportional representation (OLPR), makes election easier in some parties than in others and affects legislators’ affiliation incentives.9 The OLPR vote-pooling mechanisms allow the minimum number of votes required for election to vary across party and state. For example, a single, very popular candidate could draw millions of votes, earning several seats for their party. Because of vote pooling, his or her extra votes could make election easier for other candidates within their party. Consequently, legislators have incentives to switch party to take advantage of easier electoral quotas. Parties where election is easier should retain members; parties where election requires more votes should lose members. This should vary with legislators’ own electoral strength. Marginal legislators should be concerned about their party’s electoral quotient. Politicians that barely passed the electoral threshold should seek another party where they could be elected with a safer margin. At the other extreme, strong vote-getters do not need to pay any attention to quotients; their election would be assured in any party. In interviews, legislators have described these calculations: comparing their own expected vote share with the electoral quotient in each party, and switching to enhance reelection prospects. Electoral quotas are effectively uncontrolled club goods. Enjoying a party’s quota just requires joining the party. Ballots are unordered, so legislators do not compete for list position. And every party has a surplus of nominations: parties can nominate more candidates than
122
Scott Desposato
there are seats in every district, and yet rarely find enough willing candidates. Consequently, legislators should adopt membership strategies to take advantage of party quota differentials. Ideological Payoffs A third component of party affiliation payoffs is ideological. Membership in a party of like-minded legislators has several payoffs. Political parties can play an important role in helping like-minded legislators advance a policy agenda. In addition, a political party acts as a “brand name,” providing information to voters about a legislator’s policy positions (Aldrich, 1995; Snyder Jr. and Ting, 2001). Further, membership in an ideologically incompatible party can have significant costs. Legislators will be less able to advance their preferred agenda, potentially angering voters and frustrating the legislator’s own policy goals. Parties might also punish their less compatible members by giving them fewer advancement opportunities, refusing to support their proposals, or even preventing them from speaking on the floor. As a result, legislators should prefer membership in ideologically proximate parties and avoid parties far from their own ideal points. Ideological compatibility is clearly a club good. All legislators enjoy (or suffer) their (in-)compatibility with their copartisans. Differentials in ideological agreement should prompt legislators’ switching to, or staying in, the most compatible party. Pork or Policy? Legislators may have to balance competing desires for state resources and ideological compatibility. In such situations, their behavior should reflect constituents’ demands. Legislators from less-developed regions should prioritize access to state resources to deliver the public infrastructure their constituents need. Legislators from more-developed regions should be more responsive to ideological concerns, as basic infrastructural needs have been satisfied. This logic implies interacting pork and ideology differentials with an index of development. Legislators’ Value-Added: gi With important exceptions, most deputies have a positive impact on f(G) for most parties. In the Brazilian Chamber of Deputies, size matters. Most resources are allocated proportionally: committee assignments and chairships, as well as chamber leadership positions are a direct function of party size. In addition, for members of the governing coalition, commanding additional votes usually translates into
Party Switching in Brazil
123
additional cabinet positions and government jobs when negotiating with the executive branch. The important exceptions include the several Brazilian parties that rely more on their ideological cohesion and policy label to generate votes than on their capacity to generate pork. For such parties, ideologically inconsistent legislators have a negative gi and should be denied admission. Summary of Predictions The preceeding discussion offers the following hypotheses regarding party membership in Brazil: • H1: Increasing transaction costs decreases switching. Legislators with more partisan voters will be less likely to switch party. • H 2: Increasing resource differentials increases switching into resourcerich parties. State resources: Legislators stay in or switch into governing parties, but the attractiveness of such parties may decline as coalition size grows. Electoral Opportunity: Legislators stay in or switch into parties with easy electoral thresholds. Ideology: Legislators stay in, or switch into ideologically compatible parties. Pork or Policy: Legislators from less developed regions are most responsive to resource access; legislators from developed regions are more responsive to ideological concerns. • H3: Parties exclude legislators with negative value-added. Switchers will be welcomed when their value-added is positive but excluded when ideologically too inconsistent with a party.
Methods The formal model can be operationalized empirically as a discrete choice problem. Legislators pick the party that maximizes their expected payoff, a function of party resources less transaction costs. In a multiparty environment, we can write this as Uijt = MAX(Ui1t, Ui2t , Ui3t, . . ., Uimt), where uijt is legislator i’s utility of membership in party j at time t. When party j is her current party, she stays put. When j is some other party, we observe a switch. I write legislator i’s utility associated with membership in party j as uijt = b0 + b1 Xijt − b2TiIj≠home + b3gi + eijt,
124
Scott Desposato
where Xijt is a matrix of the covariates described earlier, and b1 is a vector of coefficients. If we assume that the eipt are independent and identically distributed following an extreme-value distribution, then this becomes a standard conditional logit model, also known as McFadden’s choice model. Though rarely used by political scientists, the conditional logit has been widely used by economists to explore individuals’ choices between products, transportation options, and, most comparable to this project, migration patterns.10 Measurement To test my hypotheses, I collected data on party membership, constituencies, and party characteristics for all deputies during the 49th and 50th legislatures (1991–1994 and 1995–1999, respectively). Estimating coefficients requires measuring the benefits that each legislator expects to receive from each party. The unit of analysis is the legislator-month, that is, each legislator’s membership choice during each month in office.11 Dependent Variable: Membership at time t The dependent variable follows standard practice for choice models.12 Yijt is legislator i’s choice of party during month t, coded “1” if the legislator chose party j and “0” otherwise. Electoral Threshold I measure electoral payoff differentials with Tdiff. I first calculated Threshold for each legislator-party combination, coded “1” if legislator i would have been elected in party j, and “0” otherwise. Finally, I calculate the threshold differential as Tdiff = Thresholdij − Thresholdik, where i references the legislator, j references the prospective party, and k references the current party. This measure captures both the electoral value of a party and the marginality of the legislator. Some legislators would have been elected in any party, so party membership does not affect their electoral prospects directly. For such deputies, Tdiff=0 for all parties. For weaker legislators, party strength matters; whereas some parties reduce electoral prospects, others increase them, as captured in the negative or positive values of Tdiff.
Party Switching in Brazil
125
State Resources Measuring the pork and other resource benefits of party membership in Brazil can be quite difficult. Because Brazil uses multimember district elections with high district magnitude, scholars cannot easily tie government expenditures to individual legislators. One alternative would be to use budget amendments, but amendments are frequently ignored by the executive branch and are only a small part of government expenditures. In fact, the executive branch at all levels of government has very strong control over budget development, expenditures, and other components of resource distribution (Schneider, 2001). Most real access to financial resources is gained through individual legislators’ lobbying of the president and ministers, consistent with patterns observed in other countries. How can we measure access to government resources? Ames (2001) creatively used meetings with ministers as a proxy for pork access. As no such data are available for my time period, I use coalition membership as a crude, but unbiased, proxy for access to resources. • Cabinetijt is coded “1” if party j was part of the president’s coalition at time t, and “0” otherwise. • Governorij is coded “1” if prospective party j was in the governor’s office in legislator i’s home state, and “0” otherwise.
Ideology I test the impact of ideological compatibility differentials on party membership using Ideological Distance ij, the relative proximity of prospective party j.13 Ideal points are estimated using the WNOMINATE method (Poole and Rosenthal, 1997). Pork or Policy? Which of these payoffs matters most to legislators will depend on the underlying preferences of their constituents. As a proxy for preferences, I use Educationi, the percentage of legislator i’s constituents with four or more years of education. I interacted this variable with Cabinet, Governor, and Ideological Distance. Education is highly correlated with other demographic indicators of development. A wellknown pattern in Brazilian politics, as in many other countries, is that pork, patronage, and clientelism are more common in less-developed regions, and policy promises are likely to be the currency of politics in more developed regions. Cabinet and Governor should have
126
Scott Desposato
their strongest impact in low education regions; Ideological Distance should have its strongest impact in high education regions. Transaction Costs Precisely identifying the partisanship of each legislator’s vote share is impossible, given the lack of detailed survey data and the high district magnitude in Brazil. We can, however, measure the degree to which each party has a partisan following using list votes. Under OLPR, voters have two choices in Brazil: to cast votes for individual candidates, or to cast votes for the overall party list. The great majority (more than 90 percent in 1994) cast personal votes, but this varies substantially across party. In the center-right Liberal Front Party (PFL), only approximately 1 percent of the votes were list votes, while the Worker’s Party (PT) received almost 34 percent list votes. I use two measures of partisanship: District Partisanshipi and National Partisanshipj. For each, I take advantage of the structure of the Brazilian ballot, on which, again, voters can vote for individuals or for the party list. District Partisanshipi is the percentage of legislator i’s electoral base that cast party list votes. Specifically, I calculated an average of list voting in the areas where each legislator received votes, weighted by the proportion of votes earned in each area. This measures partisanship among each legislator’s subconstituency within a district. National Partisanshipi is the percentage of list votes received by the legislator’s party, based on national returns. Together, these proxy partisanship: the first captures characteristics of the voters in i’s electoral base; the second controls for the size of a party’s following in aggregate. Because both variables were highly skewed, I used their natural logs in all models. Legislators’ Value-Added: gi All legislators make a small positive contribution to all parties, but deputies that are ideologically incompatible will have a negative effect on the party’s policy label. This implies that a deputy i’s value-added in party j is gi = c −
␥i − j j
where c is the constant value-added that all parties receive, γi is legislator i’s ideal point, and tj is party j’s centroid, and sj is the standard deviation of the ideal points of party j’s members. Thus, being
Party Switching in Brazil
127
far from a diverse catch-all party (large sj) is not problematic for a legislator; the second term is small and gi is positive. But being far from a highly cohesive and unified party (small sj) is problematic: gi is negative. Despite the clarity of this logic, the value-added cannot be operationalized separately from payoffs. The first term, c, is a constant. The second term is too highly correlated with ideological compatibility to be estimated separately. Effectively, this means that the Ideological Distance term must do double duty in the model, capturing both legislators’ payoffs and their value-added. Controls Building on existing work on party switching, Brazilian politics, and quantitative methods, I included several additional variables as controls: • PRN controls for the stigma associated with the Party of National Reconstruction (PRN), as President Collor was impeached for corruption in 1992. PRN is a dummy variable coded “1” for that party from the time the scandal broke until the end of the relevant legislature (August 1991–December 1994), and “0” otherwise. • Home is a dummy variable identifying legislator i’s current party, effectively a constant controlling for the baseline frequency of switching. • Committeeit is coded “1” if legislator i was on a key committee, or a committee leader, at time t. Committee should have a negative impact on switching: committee assignments are controlled by party leaders and may be lost by switching. • Party Leaderit is coded “1” if legislator i was a leader or a vice leader of her party within the legislature at time t, and “0” if she was not. Party leadership provides prestige, additional staff and office resources, and influence over the legislative agenda. Deputies in leadership positions should be less likely to switch. • Incumbent is a dummy variable coded “1” for incumbents and “0” for first-term legislators. More experienced deputies should be less likely to change party. • I also included several time-specific variables to control for key moments and differences in the legislative cycle, and to allow the underlying hazard rate of switching to vary during the legislative session (Beck, Katz, and Tucker, 1999; Mershon and Shvetsova, 2008a, b). As seen in figure 5.5, switching is common in the early periods of the session during organization and committee assignments. Switching is also common before elections—both national (1994 and 1998) and subnational (1992 and 1996). To run for reelection or for a new office, candidates had to finalize their preelection party membership months before the election, and these deadlines were lengthened after 1994. This explains the spike in early 1992 and then at the end of 1995 before mayoral
128
Scott Desposato elections (October 1992 and October 1996), and also spikes in early 1994 and at the end of 1998 before the national presidential and legislative elections (October 1994 and October 1998).
Results Tables 5.4, 5.5, and 5.6 show coefficient estimates and predicted probabilities for models of party affiliation. Overall, the results show that my affiliation model fits the Brazilian case quite well. The supposedly chaotic and antiparty behavior of Brazilian legislators in actuality reflects calculated and rational career moves. In particular, the data suggest that legislators are seeking distributive, ideological, and some electoral opportunities, but that institutional resources are not highly valued. Further, party switching patterns vary with constituency types. Characteristics of the electorate raise and lower transaction costs, and shift legislators’ priorities between ideology and distributive functions of parties. I now discuss more specific results. First, there is evidence that legislators use Brazilian parties as electoral vehicles. Brazil’s electoral rules can make election easier in some parties than in others, and party affiliation patterns take these rules into account. The coefficient on Threshold is positive and significant: legislators choose parties where vote-pooling will facilitate reelection, and avoid parties with thresholds that are too high. Table 5.5 shows
Number of Switches
30
20
10 5 0 1991
1992
1993
1994
1995
Number of Switches
Period
30 20 10 0 1996
1997
1998
1999
Period
Figure 5.5 Party switching by month in the 49th and 50th Brazilian Chamber of Deputies.
Party Switching in Brazil Table 5.4
Conditional logit model of affiliation decisions Model 1
Payoffs Threshold Cabinet Cabinet*Educ. Governor Gov.*Educ Ideol Distance Ideol Dist.*Educ Transaction Costs District Partisanship National Partisanship Controls Committee Incumbent Party Leader National Campaign Election Deadline Organize Period 50th Leg. PRN Home N LL
129
Model 2
Coef.
SE
Coef.
SE
.88 1.14 –.80 1.51 –1.86 –.79 .15
.32** .28** .53 .35** .69** .37* .64
.89 .73
.32** .09**
1.62 –2.09 –.71
.34** .67** .13**
–.40 –.49
.07** .07**
–.40 –.49
.07** .07**
–.02 –.19 –.45 –.89 1.88 .91 .83 –1.30 12.51
.12 –.13 –.16** –.25** .19** .27** .19** .31** .51**
–.46 –.88 1.89 .91 .81 –1.33 12.62
.16** .25** .19** .27** .18** .31** .51**
43,286 –1981.47
43,286 –1983.76
Note: Cubic spline estimates not shown. * = p-value < .05. ** = p-value < .01.
the substantive impact: legislators more than twice as likely (p = .71) to choose a party with an easy electoral threshold as compared to one where additional votes are needed for reelection (p = .29). Second, legislators value ideological compatibility when choosing between parties. Deputies do not want to join or stay in a party that is far from their own policy preferences; the coefficient on Ideol Distance is negative and significant. The impact is substantial: deputies are approximately five times more likely to pick an ideologically proximate party than one at the opposite end of the political spectrum (p = .86 versus p = .14). In other words, the farther a deputy is from a party, the more likely he or she is to switch away if he or she
Table 5.5 Impact of selected party characteristics on predicted probability of membership (95 percent Confidence Intervals in Parentheses) Electoral Threshold Below
Above
0.29
0.71
(.18, .42)
(.58, .82)
Cabinet Membership Opposition
Cabinet
0.32
0.68
(.29, .37)
(.63, .71)
Ideological Distance Far
Close
.14
.86
(.07, .23)
(.77, .93)
State Coalition Education Low High
Opposition
Governor
0.21 (.14, .31) 0.53 (.40, .66)
0.79 (.70, .86) 0.47 (.34, .60)
Table 5.6 Impact of transaction costs on probability of switching P(Switch) National
District Low High
Low
High
0.061 0.013
0.010 0.002
Party Switching in Brazil
131
is a member, and the less likely he or she is to switch to that party if he or she is not. This finding is evidence that legislators do care about ideological consistency and compatibility, consistent with Heller and Mershon (Heller and Mershon 2005; 2008). These results would hardly be surprising in most legislatures; they are potentially controversial for Brazilianists. Brazil’s largest parties have been categorized as “inchoate” and “catch-all.” The natural question is: Why should voting coalitions matter for legislators in such parties? The importance of roll-call compatibility may reflect legislators’ own desire to be ideologically consistent, pressure from party leaders, or pressure from constituents. Which factor is at work cannot be parsed out here, but the key point is that Brazilian legislators do value parties as voting coalitions. Third, the analysis confirms the importance of distributive capacity for legislators’ careers; the coefficient for Cabinet is positive and significant. Legislators in parties with access to federal government largesse are more likely to stay put, and switchers are more likely to head for such parties. When choosing between two otherwise identical parties, legislators are about twice as likely to pick a cabinet party than an opposition party: pCabinet = .68 versus POpp = .32. This is consistent with existing work on the importance of patronage in Brazilian politics (Amorim Neto, 1998). Governor is positive and significant, and the interaction of Education and Governor is negative and significant, as predicted. In impoverished and less-developed areas, voters prioritize governismo—following the state governor in exchange for access to state resources. In such areas, legislators in a gubernatorial party are more likely to stay put, and switchers more likely to make such a party their destination. This finding shows that decentralized Brazil’s state politics spill over into the national arena, affecting party affiliation decisions and coalition formation. These results echo previous research on the impact of federalism on national politics in Brazil (Abrúcio, 1998; Mainwaring, 1999; Samuels, 2003), and yet also establish an important additional point. Gubernatorial influence varies with constituency type. A deputy with a low-education constituency is approximately three times more likely to pick a state governor’s party over a state opposition party (p = .79 versus p = .21). But among high-education constituencies, legislators are effectively ambivalent between state opposition and governing parties: p = .53 versus p = .47, respectively.14 In impoverished and less-developed areas, where voters prioritize governismo,
132
Scott Desposato
legislators align with the state governor in exchange for access to state resources. In more developed areas, voters’ demands for localistic goods decline. This decline reduces the relative attractiveness of stateruling parties and increases the relative importance of ideological and other factors in affiliation decisions. The other interaction effects were not significant. The Education *Cabinet interaction is negative, as predicted, but its effect is not significant. It may be that the benefits of membership in executive coalitions vary across governmental levels. Perhaps national cabinet membership provides payoffs to legislators with all kinds of constituencies, while gubernatorial affiliations are most important in backward, clientelistic contexts. The other possibility is collinearity between Governor and Cabinet status—parties in the national Cabinet also tend to be parties winning gubernatorial elections. I also predicted that the interaction of Distance and Education would be negative, reflecting an increased attention to ideological consistency as voters’ preferences change. However, the interaction effect is small, has the wrong sign, and is not significantly different from zero. Two interpretations are possible. The first is that my argument about voters does not apply to voting coalitions; perhaps all legislators are equally concerned with ideological consistency, regardless of their constituents’ preferences. The other interpretation is that legislators with more educated constituents have already self-selected into the parties that are ideologically most proximate. Finally, the results emphasize the importance of voters in creating strong and stable party systems. Legislators with more partisan constituencies are much less likely to ever switch party. Table 5.6 reports the probability that a legislator switches party during a four-year legislative term as a function of partisanship. In all cases, the choice set is restricted to two identical parties, so these probabilities can be interpreted as the impact of random, unobserved, within-party personality conflicts and struggles that cause switches. The hypothetical party list voting levels are set to 0.6 percent (low) and 30 percent (high), which is slightly narrower than the actual range of the data. Legislators in less partisan parties, with personalistic constituencies, have a .061 probability of switching during the term, even when there are no payoff differentials. At the other extreme, legislators with more partisan electoral bases are very unlikely to switch, with a predicted probability of switching of just .002 over a four-year period. The control variables operate mostly as expected. Incumbency may deter switching—its coefficient is negative—though it does not reach
Party Switching in Brazil
133
conventional levels of significance. Party leaders, not surprisingly, are significantly more likely to stay put than nonleaders. The findings also have important implications for the parties’ internal organization of the Brazilian legislature. Scholars have argued that party leaders’ control of the mechanisms of the Chamber of Deputies enables them to enforce discipline and maintain party cohesion. Of all resources in the Chamber, party leaders have the most direct control over committee assignments. Apparently, however, this control is nearly worthless. Good committee assignments do not deter switching; the coefficient on Com is not distinguishable from zero. The time-specific variables did shape switching frequency. Legislators were more likely to switch party around campaign filing deadlines, and significantly more likely to stay put during campaigns. The coefficients on Organize and Deadline are positive and significant, indicating that switching is more frequent during the first part of each session when coalitions, committees, and leadership are forming, and immediately before new election filing deadlines.
5.3 Effects of Party Switching Party switching in Brazil is apparently highly rational and structured by formal and informal institutions. But how does this affect policymaking and representation? Do switchers change their policy positions to match their new parties’ platforms? Or do switchers stick with their original campaign promises to constituents? The answers to these questions are important in contradictory ways for the functioning of democracy. On the one hand, if switchers change their policy positions, this means that political party membership is in some sense meaningful, in that switchers can offer their new party labels as information shortcuts for voters. On the other hand, from the voters’ perspective, if switchers change their behavior, this implies breaking campaign promises and abandoning policy platforms previously presented to voters. To answer these questions, I examine roll-call votes for evidence of changes in policy positions before and after switching party. I use two approaches. First, I examine basic measures of party agreement before and after switching. Second, I build and estimate a spatial model of behavior.
5.3.1 Switching and Agreement Scores We can observe basic behavioral patterns by simply comparing the extent to which switching legislators vote with their old and new
134
Scott Desposato
parties, before and after switching. Define a legislator L’s agreement with party P, LP, as: LP
n
I
i 1 Pi Li
n
.
Pi and Li are the party’s and legislator’s positions on bill i, respectively, so I Pi = Li is an indicator variable coded “1” if the party and legislator agree, and “0” if they take differing positions. Effectively, this is the percentage of times that a legislator votes with a party leader’s recommendation. Table 5.7 compares basic party agreement scores for the 49th and 50th legislatures. These can be interpreted as the proportion of the time that a randomly chosen legislator voted with the majority of his or her party. For both switchers and nonswitchers, overall agreement is fairly high: more than .70 for the first period, and more than .80 for the second period. For both legislatures, there is a very small difference between the two groups; nonswitchers have slightly higher agreement scores than switchers. This is consistent with the previous section: switchers tend to be those who have ideological disagreements with their parties, and these disagreements are partly what make them switchers in the first place. Now focus only on switchers. If parties do not matter, then legislators should vote with their old party just as much after switching as before; if parties do influence behavior, then legislators should increase their voting with their new party, and decrease their voting with their old party. Table 5.8 reports pre- and post-switch agreement scores. For each legislature, the top left cell shows mean agreement scores while in their old parties; the bottom right cell shows the same while in their new parties. The off-diagonals show agreement with the new party before switching and agreement with the old party after switching. That is, they capture the extent to which a deputy votes with his or her old party after leaving it, and the extent to which a Table 5.7
Party agreement scores Legislature
Switchers Nonswitchers
49
50
.75 .76
.83 .84
Party Switching in Brazil Table 5.8
135
Switching and party agreement scores Legislature 49th
Old Party New Party
50th
Before
After
Before
After
.76 .60
.64 .73
.82 .72
.78 .84
deputy votes with his or her new party before even joining it. There are clear and obvious changes in legislators’ behavior, lending support to the party influence hypothesis. In the first term, before changing, deputies voted with their party’s positions approximately 76 percent of the time; after changing, they voted with their old party significantly less, just 64 percent of the time. Similarly, they voted with their new party just 60 percent of the time before switching and 73 percent of the time after switching. Agreement scores increased across the board for the 50th legislature, but the same basic pattern emerges, suggesting significant changes in voting behavior by legislators. Switchers do modify their voting behavior to align more closely with their new party after changing, providing evidence of party influence over legislators.
5.3.2 A Spatial Test The agreement measures provide compelling support for the party influence hypothesis, but do not include any spatial component. In this section, I build a simple spatial model to test for party influence by using standard spatial methods. A standard random utility model of voting is as follows. Let ui be legislator i’s ideal point, bj be the location of a “yes” vote on bill j, and αj be the location of a “no” vote on bill j. Legislator i votes “yes” on bill j if f(ui − bj + eij) < f(ui − αj + eij) and “no” if f(ui − bj + eij) > f(ui − αj + eij), where eij is an iid random variable and f describes legislators’ utility functions. Typically, e is distributed iid normal or extreme
136
Scott Desposato
value, generating a probit or logit-like model, and f is quadratic or exponential.15 In the above model, ui is effectively a combination of legislators’ own preferences and pressures from party organizations and other actors. We can further separate the u into legislator- and party-specific components: ui = (1 − α)Li + αPj, where 0 ≤ α ≤ 1. In this case, α measures the relative influence of parties over legislators’ observed ideal points, which reflect both party (P) and personal (L) preferences.16 When α is at its minimum value (α = 0), all of legislators’ behavior reflects their own preferences. When α is at its maximum (α = 1), all of legislators’ behavior reflects party preferences. In a system with stable party membership, this is a purely academic exercise. With only one measure of u per legislator, ui is an irreducible function of legislators’ own preferences, party influence, and other possible pressuring actors. Without additional information or strong assumptions, we cannot distinguish between party and legislators’ preferences. Although both ideal point and party influence parameters could be jointly estimated from a single model (see Desposato [2006], e.g.), for simplicity’s sake I adopt a two-step process, estimating ideal points, then using ideal points to estimate α, as follows. Step one: estimate ideal points for legislators and parties. For legislators, I separately estimate each observed legislator-party combination. In other words, if a legislator switches from the Democratic Worker’s Party (PDT) to the Liberal Front Party (PFL), I estimate two ideal points for him or her: one for his or her votes cast while in the PDT (uiPDT), and another for his or her votes cast while in the PFL (uiPFL). To estimate party locations (P), I use two methods. First, I estimate each party’s ideal point as the median of its nonswitching members. Second, I estimate a separate party ideal point by counting publicly taken party positions as votes. When party leaders have declared a party position before roll-call votes, I count that position as a party vote for the purpose of estimating a party location.
Party Switching in Brazil
137
Step two: estimate α with a simple linear model: ui = αPk + (1 − α)Li + eik This is a fixed-effects model, where the dependent variable is switcher i’s ideal point while in party k, the covariate Pk measures party k’s ideal point, α is the coefficient of interest, and a fixed effect is inserted to capture (1 − α)Li.17 The core hypotheses to be tested are: • H 0: α1 = 0; α2 = 0 Legislators are uninfluenced by parties and executives; their votes reflect only their own preferences or those of their constituents (see footnote 16). The observable implication of this hypothesis is that there are no significant changes in the estimated ideal points of switchers. • H1: α1 > 0 Legislators’ roll-call votes are influenced by their parties. Switchers move closer to the party they switch into. Legislators’ roll-call votes are influenced too by the executive branch. At the extreme, if the president has influence but not the parties, we will observe changes in ideal points for legislators switching into or out of cabinet parties, but not for those switching between parties within the cabinet circle or within the circle of the opposition.
5.3.3 Data To test these hypotheses, I used all roll-call votes from the Chamber of Deputies, 1989–1998. I used two sources. First, I collected all roll-call votes directly from the Secretaria Geral da Mesa in Brasilia. Second, I compared these results with Argelina Figueiredo and Fernando Limongi’s dataset of roll-call votes. Results were similar for both datasets. Ultimately, I report results from their data because the Figueiredo-Limongi dataset includes party and presidential positions on key votes. I use the Poole and Rosenthal (1997) WNOMINATE scores to summarize legislators’ voting behavior.
5.3.4 Results Tables 5.9 and 5.10 show results from estimating only party effects while excluding executive pressures. The tables report estimated values for α using both measures of party preferences discussed earlier for three legislatures (1987–1999, inclusive). Each value of α was estimated in a model including fixed-effects for each legislator; estimated fixed effects are not shown. The first table shows estimates from the
138
Scott Desposato
Table 5.9 Party influence measures—Dimension 1 Legislature 48th
^ α Party Leader Party Median
49th SE **
.44 .54
.11 .10**
^ α
50th SE **
.44 .49
.09 .09**
^ α
SE
.75 .81
.03** .03**
Note: Deputy-specific fixed-effects not shown. * = p-value < .05. ** = p-value < .01.
Table 5.10
Party influence measures—Dimension 2 Legislature 48th
^ α Party Leader Party Median
.37 .60
49th SE .08** .09**
50th
^ α
SE
^ α
SE
.22 .24
.10** .11**
.64 .71
.04** .05**
Note: Deputy-specific fixed-effects not shown. * = p-value < .05. ** = p-value < .01.
first dimension WNOMINATE scores; the second table shows second dimension estimates. Both tables reveal consistent and significant party influences over legislative behavior; switchers on average do change their voting to better match their new party’s preferences. Party influence is significant on both dimensions using all measures for all legislators, across all models. In general, the median-based estimates are larger than the party leadership-based estimates, but the two tend to move in sync. Further, the results suggest that Brazilian parties are in fact getting stronger, with party influence rising from approximately .50 in the 48th legislature (first dimension) to more than .75 in the 50th legislature. This suggests some ongoing consolidation of the political system, which marches with the evolution of the broader environment during these periods. The first two terms were fairly chaotic and transitory, with some doubt as to the survival of the young democratic regime. Brazil suffered hyperinflation, especially in the first
Party Switching in Brazil
139
term, with occasional sabre-rattling by the military and rumors of coup attempts. In the second, President Collor was impeached for corruption, his party imploded, and the vice president who replaced him (Itamar Franco) had only weak ties to the existing political parties. Finally, in the third term, during the presidency of Fernando Henrique Cardoso, economic reform stabilized the economy, and Cardoso was able to construct a stable governing coalition. This increasing party influence might be seen as good news by some, but the implications for representation are contradictory. On the one hand, the evidence of party influence is encouraging; it suggests that, to the extent that political parties do have influence over their members, voters can use the party identification of candidates as a proxy for a set of issue positions. On the other hand, knowing that switchers will likely change their behavior in a new party is troubling. You might vote for a candidate from one party representing one set of ideals, but after a party switch, he she might support an entirely different set of positions. How can we reconcile the two perspectives? Figure 5.6 displays patterns of switching in two dimensions. Each point represents a legislator-party, an ideal point estimated using the roll-call votes each legislator cast while in one or another party. The first dimension captures a classic left-right dimension and accounts for most of the roll-call voting behavior; it is easily the predominant 1.0
Dimension 2
0.5
0.0
−0.5
−0.1 −0.1
−0.5
0.0
0.5
Dimension 1 Figure 5.6 Ideology and switching in the 50th Brazilian legislature.
1.0
140
Scott Desposato
dimension, with an aggregate proportional reduction in error (APRE) score of more than .85. The second dimension distinguishes parties on a number of issues, including government and opposition within ideological tendencies, but has much weaker predictive power. The lines link before and after locations for all switchers during the 50th legislature. Some deputies do switch from one extreme to another, from the far left to the far right. Most of the lines are vertical, however, meaning that switchers did not move far on the first dimension. Instead, most switched within the same center-right coalition, often from parties that were not included in the governing coalition, to parties with cabinet positions and redistributable pork. These vertical lines represent changes in behavior on secondary issues, not on the primary system conflict. This pattern also integrates the results from the preceeding section (switchers do not choose distant parties) with the results from this section (switchers change to be more like their new party). Legislators do change their behavior to match their new parties, and they are more likely to switch if alienated ideologically. Yet most switchers are not moving very far, usually within the large center-right coalition. Finally, combined with the interactive results, the implication is that party switching is not inconsistent with representation, for most Brazilian voters. Legislators who switch tend to avoid moving “far” ideologically, but they will make some changes. Many of these switchers are motivated by the pursuit of government resources, similar to earmarks in the United States. Especially in less developed areas, constituents prefer delivery of developmental local public goods to die-hard ideological stands.
5.4 Conclusion As indicated at the outset, the Brazilian system is in a critical transitional moment. In March 2007, the Brazilian Supreme Court ruled that elected offices belong to parties, not to individual candidates. The immediate effect was to call into question the status of the many deputies who had switched since taking office on February 1, 2007. Since then, virtually no deputies have changed party. The likely impact of this exogenous shock is unclear. For the short run, this ruling has put an end to party switching. Within the legislature, this could strengthen or weaken parties, and could strengthen or weaken representation in Brazil. For parties, removing the “exit” option may mean that legislators will be more vulnerable to leaders’ demands and directions, thus increasing party cohesion. On the other hand, if
Party Switching in Brazil
141
party leaders are less powerful, the ruling could reduce party cohesion by forcing ideologically diverse politicians to stay together, rather than moving on to parties that better suit their or their constituents’ policy opinions. These divergent possibilities also imply different representational outcomes. If parties are more cohesive, voters may have clearer choices and greater certainty of the impacts of their votes. Further, these elite changes might enhance issue-based voting at the mass level. On the other hand, less cohesive parties could dramatically increase voter uncertainty about the ideological impact of a vote, and further discourage issue voting. A possible secondary product of this scenario is that legislative elections will be increasingly personalistic and pork-oriented, but legislators will not be able to switch party in pursuit of pork. Numerous other scenarios are possible. Switching could reappear as a preelection phenomenon. A legislator who wants to run for executive office but is denied his or her party’s nomination could switch party and resign to run under a different label. Alternatively, the ban might be overturned. Indeed, a bill to relegalize switching has already been introduced. No matter which scenario materializes, Brazil will provide a useful natural experiment in the effort to assess the impact of a ban on switching on representation and governability.
5.5
Notes
1. Halfway through the current legislative period, there have been 135 switches for a projected rate of .57 by January 2006. 2. Note that in a two-party system, the two approaches are identical. Switching party in the United States nearly always means moving from Democrat to Republican, or Republican to Democrat. Even in the United States, however, switchers have more than two choices (switch/not switch). They may become independents or join a third party as some recent switchers have done (e.g., Lieberman). 3. Club goods may not be entirely nonrival, however: each additional member may slightly degrade the quality of the club good. For example, as a party’s size grows, it may become a catch-all party, and its ideological label may lose meaning. New members with very similar ideological positions will have minimal degrading effects; legislators with distinctly different or even contradictory positions may reduce the value of the party label dramatically. 4. More recently, party leaders in Ecuador informally agreed to allow illegal party switching to continue so that a minority president could increase his legislative coalition size. The rule in Japan applies to legislators elected through proportional representation, and only prohibits switching to a party that competed in their same district. 5. The market for party membership has clear parallels with labor markets, with three important differences. First, endogenous parties, not firms, make
142
6. 7. 8. 9.
10.
11.
12. 13. 14. 15.
Scott Desposato personnel decisions. Second, personnel decisions are made to maximize members’ utility, not profits. Third, instead of wages, legislators receive payoffs that may be private or club goods. At this point, the model begins to bear a striking resemblance to the market for political science faculty, in which personnel decisions are made by majority vote to maximize departmental quality and payoffs are partly private goods (wages) and partly club goods (reputation). Candidates for office must stay in a single party in the months leading up to the election. This provision was recently eliminated by a court decision but was in effect for the period covered by my dataset. For the executive, supermajorities may well be cheaper, as they allow the president to pay lower costs on a vote-by-vote basis to buy a majority. Under OLPR rules, citizens cast a single vote, usually for an individual candidate. Seats are distributed to parties based on the total of votes received by all of the parties’ candidates. Mandates are distributed to candidates based on the total number of votes each receives. See Ames (2001, 1995a, b), and Mainwaring (1999) for more details on OLPR. See McFadden (1973), Maddala (1983), and Long (1997). Alternative choice models can be implemented with more flexible error distribution assumptions, but are only tractable for smaller choice sets and datasets. See Glasgow (2001) for more details. An alternative, and possibly more accurate, specification would incorporate parties’ valuation of legislators separate from deputies’ choices—a model where parties first decide whether to accept a legislator, then legislators choose from the choice set available to them. I have a working paper demonstrating how to estimate such a model, but it proved intractable for this dataset. Why is it necessary to use a time-based measure of affiliation decisions? First, many of the key variables in the model—cabinet formation, committee assignments and leadership, party leadership, and even the party of the president—vary over time, as do many of the control variables. During one period, legislator i’s party might have cabinet access. A month later, after a cabinet reshuffle, legislator i might be in the opposition. Second, the actual membership of the Chamber of Deputies is constantly in flux as legislators leave and substitutes take their places. Using the legislator-month as the unit of analysis accounts for the varying amount of time that each legislator actually serves in the Chamber. See Maddala (1983), Glasgow (2001), Long (1997), and Ben-Akiva and Lerman (1985). Specifically, Distanceip is calculated as DistDif fij = |γi − tj |, where γi is legislator i’s ideal point, tj, is the centroid of prospective party j. All discussions of interaction effects’ significance account for the covariance between the base and interactive terms. See Londregan (2002) and Poole and Rosenthal (1997) for additional discussion.
Party Switching in Brazil
143
16. For this project, I lump constituents and special interest influence with legislators’ own personal ideological positions into a single variable, L. I discuss relaxing that assumption in the conclusion. 17. Consequently, we cannot directly estimate 1−α, but can only get at it indirectly through the estimate of α.
5.6 References Abrúcio, Fernando. 1998. Os Barões da Federação: Os Governadores e a Redemocratização Brasileira. São Paulo: Editora Hucitec. Aldrich, John H. 1995. Why Parties? The Origin and Transformation of Political Parties in America. Chicago: University of Chicago Press. Aldrich, John H., and William T. Bianco. 1992. A Game-Theoretic Model of Party Affiliation of Candidates and Office Holders. Mathematical Computer Modeling 16(8–9): 103–116. Ames, Barry. 1995a. Electoral Rules, Constituency Pressures, and Pork Barrel: Bases of Voting in the Brazilian Congress. Journal of Politics 57(May): 324–343. ———. 1995b. Electoral Strategy under Open-List Proportional Representation. American Journal of Political Science 39(May): 406–433. ———. 2001. The Deadlock of Democracy in Brazil. Ann Arbor: University of Michigan Press. Amorim Neto, Octavio. 1998. Of Presidents, Parties, and Ministers: Cabinet Formation and Legislative Decision-Making Under Separation of Powers. Ph.D. thesis, University of California, San Diego. Beck, Nathaniel, Jonathan N. Katz, and Richard Tucker. 1999. Taking Time Seriously: Time-Series-Cross-Section Analysis with a Binary Dependent Variable. American Journal of Political Science 42(October): 1260–1288. Ben-Akiva, Moshe, and Steven R. Lerman 1985. Discrete Choice Analysis: Theory and Application to Travel Demand. Cambridge, MA: MIT Press. Carey, John M., and Matthew Soberg Shugart. 1995. Incentives to Cultivate a Personal Vote: A Rank Ordering of Electoral Formulas. Electoral Studies 14(December): 417–439. Desposato, Scott W. 2006. The Impact of Party Switching on Legislative Behavior in Brazil. Working Paper. Glasgow, Garrett. 2001. Mixed Logit Models for Multiparty Elections. Political Analysis 9(Spring): 116–136. Graham, Richard. 1990. Patronage and Politics in Nineteenth-Century Brazil. Stanford: Stanford University Press. Heller, William B., and Carol Mershon. 2005. Party Switching in the Italian Chamber of Deputies, 1996–2001. Journal of Politics 67(May): 536–559. ———. 2008. Dealing in Discipline: Party Switching and Legislative Voting in the Italian Chamber of Deputies, 1988–2000. American Journal of Political Science 52 (4): 910–925. Keck, Margaret E. 1995. The Workers’ Party and Democratization in Brazil. New Haven, CT: Yale University Press.
144
Scott Desposato
Leal, Victor N. 1977. Coronelismo: The Municipality and Representative Government in Brazil. New York: Cambridge University Press. Londregan, J. B. 2002. Appointment, Reelection, and Autonomy in the Senate of Chile. In Legislative Politics in Latin America, ed. Scott Morgenstern and Benito Nacif, 341–376. New York: Cambridge University Press. Long, J. Scott. 1997. Regression Models for Categorical and Limited Dependent Variables. Number 7 in Advanced Quantititative Techniques in the Social Sciences Series. Thousand Oaks, CA: Sage. Maddala, G. S. 1983. Limited-Dependent and Qualitative Variables in Econometrics. Cambridge: Cambridge University Press. Mainwaring, Scott. P. 1997. Multipartism, Robust Federalism, and Presidentialism in Brazil. In Presidentialism and Democracy in Latin America, ed. Scott P. Mainwaring and Matthew Soberg Shugart, 55–109. New York: Cambridge University Press. Mainwaring, Scott. P. 1999. Rethinking Party Systems in the Third Wave of Democratization: The Case of Brazil. Stanford: Stanford University Press. McFadden, Daniel. 1973. Conditional Logit Analysis of Qualitative Choice Behavior. In Frontiers in Econometrics, ed. Paul Zarembka, 105–142. New York: Academic. Mershon, Carol, and Olga Shvetsova. 2008a. Parliamentary Cycles and Party Switching in Legislatures. Comparative Political Studies 41(1): 99–127. ———. 2008b. Party Switching in Sitting Parliaments and the Midterm Effect. Annual Joint Sessions of the European Consortium for Political Research, Rennes, France. Pereira, Carlos, and Bernardo Mueller. 2004. The Cost of Governing: Strategic Behavior of the President and Legislators in Brazil’s Budgetary Process. Comparative Political Studies 37(7): 781–815. Poole, Keith T., and Howard Rosenthal. 1997. Congress: A Political-Economic History of Roll Call Voting. New York: Oxford University Press. Samuels, David. 2003. Ambition, Federalism, and Legislative Politics in Brazil. New York: Cambridge University Press. ———. 2006. Sources of Mass Partisanship in Brazil. Latin American Politics and Society 48(Summer): 1–27. Schneider, Aaron. 2001. Federalism against Markets: Local Struggles for Power and National Fiscal Adjustment in Brazil. Ph.D. thesis, University of California, Berkeley. Snyder Jr., James M., and Michael M. Ting. 2001. An Informational Rationale for Political Parties. American Journal of Political Science 46(January): 90–110.
III
Party Switching, Party Competition, and Policy Making
This page intentionally left blank
6 Party Group Switching in the European Parliament* Gail McElroy and Kenneth Benoit
6.1
Two Levels of European Party Competition
Party politics in the European Union (EU) is characterized by competition at two different levels. At the national level, political parties contest national, regional, and local elections. At a transnational level, national parties also compete in elections, for seats in the European Parliament (EP), the legislative body of the EU. National political parties remain the basic organizational unit at both the national and European levels, controlling access to European ballots in each member state, representing policy, and organizing election campaigns. National political parties also form the primary constituent units of the party groups in the EP, coming together to form transnational political entities. These EP party groups have considerable influence over office and policy at the European level, and their goals may be either compatible or conflicting with the interests of national member parties (Hix, Noury, and Roland 2007). EP party groups also formulate their own policy positions and seek to control important offices and legislative goods at the European level. Nonetheless, they remain fundamentally conglomerations of national level parties (Kreppel 2004, 977). A key question thus concerns the calculus of affiliation by which national political parties join or switch between EP party groups. On what basis do national political parties become members of EP party groups? Furthermore, given the fact that individual members of the EP (MEPs) and national parties both frequently change their party group affiliations in the EP, what motivates this form of switching? Directly addressing these questions is the main objective of our chapter. Although our examination is preliminary in the sense that we infer the motivations for changing party groups
148
Gail McElroy and Kenneth Benoit
from a static examination of affiliation, the implications of our explanation of switching should form the basis for further, fuller examination along more dynamic lines and encompassing a longer period. Party competition in the EU has received a great deal of scholarly attention in recent years, yet significant lacunae exist in explaining the full dynamic of competition at the national and European levels. Political competition in Europe is traditionally studied at the national level. Studies of voter behavior and outcomes in European elections have mainly confirmed the national orientation of European elections, consigning EP elections to “second-order” status, mainly extending national political contestation in symbolic referendums on governing party support (e.g., Marsh 1998; Reif and Schmitt 1980). Other work examines national party positions on European issues, focusing on which national factors explain party support for EU integration (e.g., Dietz 2000; Marks and Wilson 2000; Marks, Wilson, and Ray 2002; Ray 2003; Taggart 1998), whether party positions toward the EU are independent or integrated with positions on traditional leftright political issues (e.g., Marks and Steenbergen 2004), or whether national political parties have changed their policies to become more favorable to the EU (Daniels 1998). A final research strand focuses on the rise and organization of party groups or “europarties” at the European level (Kreppel 2002; Raunio 1997) and the direct effects of these new organizations on national parties (e.g., Bardi 2002; Damm 1999; Dietz 1997; Niedermayer 1997). The dynamic of affiliation with EP party groups and switching by national parties and MEPs between them has received relatively scant attention, however. While party groups in the EP have been analyzed in terms of voting cohesion (Hix, Noury, and Roland 2005; Kreppel and Tsebelis 1999), the policy cohesion of their constituent national members has never come under systematic empirical scrutiny. And while McElroy (2003) has examined individual MEP switching among party groups, no study to date has explored the basis on which national political parties decide to affiliate or disaffiliate with EP party groups. The conventional wisdom, usually no more than an unexamined assumption, is that national party affiliation in transnational party groups, including the EP, is driven by historical tradition and a party’s set of core principles, usually rooted in enduring cleavages and represented by party family (see Klingemann, Hofferbert, and Budge 1994, 24). Yet party family explanations of EP party groups cannot explain several striking outcomes. First, national parties (as well as individual MEPs) quite frequently switch EP party groups,
Party Group Switching in the EP
149
while remaining constant in terms of party family. Second, national party members of EP party groups clearly have heterogeneous positions on multiple dimensions of policy, a fact that cannot be captured by broad categories of party family. Finally, many national parties do not belong to clearly definable party families, either because their organizational basis does not correspond to the classic cleavages identified by Lipset and Rokkan (1967)—a situation characterizing many of the parties from the 12 post-communist member states—or because they are formed primarily around new (single) issues such as European integration or immigration. Our chapter investigates legislative party switching in the EP and in the process sheds light on the foundations of political parties in the EU. Our chapter proceeds as follows. The second section describes the organization of political competition at the national and European levels, and highlights the mechanisms available to EP party groups for maintaining policy coherence among their members. This discussion culminates in our hypotheses. Third, we address the compilation of the data used to test our hypotheses. Next, we characterize the European policy space, comparing the association of policy dimensions at the national and European levels. The fifth section examines the politics of EP party group choice and affiliation at the party level, and the sixth section evaluates the evidence on switching at the individual level.
6.2 The Organization of Party Politics in the EP 6.2.1
Political Groups in the EP
Just as with parties in national parliaments, political or party groups in the EP form the backbone of legislative organization and decision making, acting as the gatekeepers to legislative amendments and the vast bulk of interruptive and procedural motions in the assembly. Unlike national parties, political groups in the EP do not form governments; even so, political group size determines membership of committees and all other key parliamentary decision-making bodies. Furthermore, the legislative function of the EP has steadily grown in importance. Estimates suggest that as much as 60 percent of national legislation originates at the European level (Nugent 2003), an indication of the growing significance of the European arena for domestic policy. Since the inception of the EP in 1957, political groups have been officially recognized in the EP’s rules of procedure, and they have
150
Gail McElroy and Kenneth Benoit
received financial support for administrative costs from the parliamentary budget. Seven political groups currently exist in the EP, representing almost 150 different national political parties from the 27 member states. These political groups vary significantly in their degree of institutionalization, as evidenced by age. For instance, the 2 largest groups, the European People’s Party (EPP) and the Party of European Socialists (PES), have existed for more than 50 years, whereas others, such as Independence and Democracy (iD) are much younger. At the time of writing (May 2008), the largest party, the EPP, has representatives from all 27 member states, comprising a total of 287 representatives from 46 different national political parties. In contrast, the Union for a Europe of Nations (UEN) has just 42 members from only 6 states of the EU while iD has a mere 23 members from 9 different member states. Our study of party group choices focuses on a slightly earlier phase, near the historic 2004 elections following the expansion of the EU to 25 member states. Table 6.1 provides details of the party groupings in the EP just before these elections, along with the number of national member parties they had. Over time, party groups in the EP have become increasingly cohesive and powerful. Levels of voting cohesion have risen across Table 6.1
Political party groups in the European Parliament pre-2004 election
EP Party Group
Label
Seat %
Seats
Member Parties
European People’s Party Party of the European Socialists European Liberal and Democrat Reform Party European United Left/Nordic Green Left Greens Union for a Europe of the Nations Group for a Europe of Democracies and Diversities Other Total
EPP PES ELDR
37.5 29.5 8.4
294 232 66
51 32 28
GUE Verts UEN EDD
7.0 6.0 3.8 2.2
55 47 30 17
19 20 11 5
5.6 100.0
44 785
16 182
Other
Note: These figures represent the standing in early June 2004, just before the 2004 European Parliament Elections. The 785 member total is composed of the 626 members from the 15 preenlargement states and the temporary members representing the accession states since May 1, 2004. Source: European Parliament official Web site (http://wwwdb.europarl.eu.int/ep5/owa/p_meps2. repartition).
Party Group Switching in the EP
151
parliamentary sessions, especially for the three largest political groups, despite increases in the size of the EP and the number of member states (Hix, Noury, and Roland 2005; Raunio, 1997). The increasing importance of the political groups in the EP likely reflects the concomitant increase in powers of the parliament itself. Once a purely consultative institution, the EP is now, under the co-decision procedure, a co-legislator with the European Council in a wide variety of policy areas, covering everything from health, to consumer protection to the environment. In elections to the EP, party groups adopt policy platforms and issue manifestos as do the national political parties. The largest groups translate their election manifestos into most European languages, for dissemination among voters of all member states. The ELDR, for instance, translated its 2004 European election manifesto into 17 languages, and the EPP into 19 languages. National parties are not obligated to adopt the entirety of the EP group’s positions when formulating their own manifestos either for national or European elections, yet the availability of the EP party group manifestos makes any such differences quite clear and potentially problematic (Gabel and Hix 2004). Once in the EP, MEPs elected by national parties are organized into EP party groups, regularly issued voting instructions and, at least for the two largest groups, subject to a party whip. Party groups have mechanisms to punish MEPs who vote against the party group instructions, whether on individual initiative or (more likely, see Hix 2004) because of national party instructions. Party groups may demote members from important committee or rapporteur positions as punishments (McElroy 2002) or at an extreme may even expel MEPs from their party groups.1 Strong divergences in policy positions between national member parties and the EP party groups could therefore cause persistent conflicts between national party and EP party group voting positions. Hix, Noury, and Roland’s (2007, 137) analysis of roll-call votes from the 1999–2004 parliament, for instance, showed that approximately 7 percent of MEPs voted against their party group to vote with their national party. Most of the time (89 percent), however, MEPs vote in line with their party group (Attinà 1990; Hix, Noury, and Roland 2007). The high and increasing cohesion of MEP voting behavior suggests that the policy cohesion of national party members of EP party groups— policy cohesion measured independently from voting behavior, that is—should also be high.
152
Gail McElroy and Kenneth Benoit
6.2.2
National Party Membership in EP Party Groups
At the national level, political parties in Europe continue to fulfill all of the classic functions of political parties, including representation, government formation, and articulation of policy platforms. Yet parties not only contest national elections but also contest EP elections. The overwhelming majority of viable candidates for the EP are recruited and sponsored by national political parties, who act as the principal gatekeepers to the European electoral arena (Mair 2000, 38). National parties also join EP party groups, usually as a bloc, although individual MEPs may also switch from one group to another.2 In EP elections, most national parties issue European election manifestos targeted specifically at their national electorates. These national party European manifestos stand alongside the European election manifestos of the EP party groups with which they are affiliated. The existence of these two policy platforms for the same election highlights the tension experienced by national parties. We expect that national parties, faced with this tension, will attempt to minimize the incongruence between their policy positions represented in the national and European arenas. When it comes to EP party group policy positions on issues of the greatest domestic importance, national parties will attempt to steer these, when possible, toward their own most preferred policies. National parties will also be motivated to become members of party groups whose overall policy platforms are closest to their own. National party policy positions change, sometimes dramatically, over time. The Comparative Manifestos Project has studied manifesto content in detail and has documented significant movement in party positions by major parties since World War II (Budge et al. 2001). If political parties affiliate with EP party groups on the basis of policy congruence, then movement of national party positions implies that EP party group membership should also change over time. Indeed, studies of party switching in the EP have shown just this: EP party group composition is quite dynamic. Not only are EP groups characterized by a high frequency of party switching, but also much of this switching occurs when national political parties decide to change EP group affiliations, bringing all their MEPs along with them. Table 6.2 documents notable party switches in recent EP sessions. For example, the Social Democratic Party (PSD) in Portugal switched from the Liberal (LDR) party group to the EPP in 1994.3
Party Group Switching in the EP
153
Table 6.2 Party group volatility in the European Parliament: Notable examples of national party switching, 1989–2004 Party
Country
From
To
Year
Partido Popular Scottish National Party Izquierda Unida MSI Parti communiste français Socialistisk Folkeparti Agalev CDS Ulster Unionist Party D’66 Conservative Party Konservative Folkepartei PDS UDF PSD Vlaams Bloc Front National PSDI CDU-PCP Partido Nacionalista Vasco Socialistisk Folkeparti Liberalu˛ demokratu˛ partija Valstiecˇiu˛ ir Naujosios demokratijos partiju˛ sa˛junga UDF CDS/PP PEV Lega Nord Bloco de Esquerda League of Polish Peasants
Spain UK Spain Italy France Denmark Belgium Spain UK Netherlands UK Denmark Italy France Portugal Belgium France Italy Portugal Spain Denmark Lithuania Lithuania
ED RDE COM DR COM COM ARC NI DR NI ED ED GUE LDR LDR DR DR S CG Greens GUE/NGL ELDR EPP
PPE ARC GUE NI CG GUE V LDR PPE LDR PPE-ED PPE-ED S PPE PPE NI NI RDE GUE ALDE Greens UEN UEN
1989 1989 1989 1989 1989 1989 1989 1989 1989 1989 1992 1992 1993 1994 1994 1994 1994 1994 1994 2004 2004 2004 2004
France Portugal Portugal Italy Spain Poland
EPP UEN V NI NI NI
ALDE EPP GUE/NGL ID GUE ID
2004 2004 2004 2004 2004 2004
ARC/ARE: Arc en Ciel (Rainbow Group) CG: Coalitions des Gauches COM: Communists DR: Droites Européenes (Front National, Die Republikaner) ED: European Democratic Group ELDR (LDR/ALDE): European Liberal and Democratic Reform Group EPP (PPE): European People’s Party FE: Forza Europa (Forza Italia) GUE: Confederal Group of the European United Left NI: Unattached members PES (S): Party of European Socialists RDE: Group of the European Democratic Alliance TDI: Technical Group UEN: Union for a Europe of Nations
154
Gail McElroy and Kenneth Benoit
6.2.3
Individual MEP Affiliations with EP Party Groups
Although the majority of MEPs become members of the EP party groups with which their national parties have declared an affiliation, individual MEPs may decide to switch their affiliations in the EP from one party group to another. This form of switching is not uncommon. In the Third Parliament (1989–1994), for instance, just less than 20 MEPs (almost 4 percent of all MEPs) chose to move individually to another party group from the one in which they commenced the legislative session. Previous studies of party switching have focused mainly on decisions taken by individual legislators, and previous work on the EP has indeed targeted this level (McElroy 2003). Switching is typically ascribed to office- and policy-seeking motivations, resting on the assumptions, respectively, that legislators are concerned primarily with winning elections (Downs 1957) and legislators aim chiefly at achieving policy outcomes (McKelvey and Schofield 1987). Most scholars now accept that there is some mix of these motivations in the political calculus of most actors, though there is no consensus on the precise nature of the trade-off (Strøm 1990). Previous studies of switching suggest that the electoral connection is paramount; for instance, Aldrich and Bianco’s (1992) model posits that members choose the party that maximizes their prospects of reelection. In the EP, however, electoral incentives appear to be extremely weak or even nonexistent. First, again, elections to the EP are not run on the basis of European issues but are “second-order elections” (Hix 2005; Reif and Schmitt 1980; Reif 1997). The elections serve as essentially informal referendums on national government performance; competition takes place between national political parties rather than the political groups of the EP.4 In addition, citizen knowledge on the activity of the EP and its political groups is very restricted, as has been documented in numerous surveys among voters in member states. Third, EP party groups do not control ballot access for candidates, a power instead retained by political parties at the national level. Fourth, EP party groups have only limited access to resources that can be used selectively to reward individual members for switching—and that feature prominently in accounts of switching in Brazil for instance (Ames 2001). Finally, the fact that most party switchers within the EP do not actually change their national party affiliation suggests that electoral incentives do not provide a pressing motivation for defection. Switchers generally run on the
Party Group Switching in the EP
155
same national ticket in European elections both before and after they switch in the EP. For the 1989–1994 EP, a mere 10 percent of switches also involved a switch of political party at the national level. There are thus essentially two choices involved in the individual-level “calculus of candidacy” for the EP: the choice to affiliate with a particular national party and subsequently the choice to affiliate with a particular political group once in the parliament. The traditional electoral motivations may prevail in the first choice but do not appear to apply to the second. Given that incumbent MEPs nonetheless do switch from one group to another, the obvious question is: why do they bother? Previous work (McElroy 2003) suggests that both ideology and career advancement within the parliament play a role. EP party groups can offer a range of selective benefits internal to their groups, such as membership in the party bureau as well as party-group controlled committee assignments and chairmanships. Individual MEPs are also powerfully motivated by policy concerns, however, and previous research on the EP (McElroy and Benoit 2007; Hix, Noury, and Roland 2007) has yielded strong evidence that, for MEPs and their party groups, politics at the transnational level is primarily an extension of domestic politics, especially left-right politics on traditional issues. Thus, MEPs should also be motivated to switch to join a party grouping whose median policy preferences are relatively closely aligned with their own, or at least deterred from switching to a group whose median policy preferences are relatively far from their own.
6.2.4 Hypotheses This discussion may be recapitulated in two overarching testable hypotheses. Hypothesis 1. For national parties, the fundamental driving force behind affiliation and switching in EP party groups should be the desire to maximize policy congruence between their own positions and those of their group. National political parties should affiliate with the EP party groups whose policy platforms are closest to their own on important national policy dimensions. Pressures from the demands of contesting European elections should push national parties toward coherence in the policy platforms they develop for domestic and EP elections; such programmatic coherence in turn should strengthen the criterion of policy proximity as a motivation for choice of party group once in the EP. Moreover, EP party groups should attempt to maintain a degree of policy coherence among their national party members.
156
Gail McElroy and Kenneth Benoit
Hypothesis 2. Policy concerns should feature prominently in individual MEP decisions to switch between party groups. MEPs who switch should seek groups whose legislative activities and voting instructions are more compatible with the individual MEPs’ own preferences. Conversely, individual MEPs should avoid switching into groups whose ideological positions are less compatible with their own, although individual switchers might also be motivated by the promise of career advancement.
6.3 Data: Policy Positions of Parties, Party Groups, and MEPs To evaluate our hypotheses, we have collected data at both the national party and individual MEP level from four principal sources. First, we have gathered data on national party affiliation into party groups in the EP. These data were drawn from the EP party groups in the Fifth Parliament (1999–2004) just before the 2004 elections, as detailed in McElroy and Benoit (2007) and portrayed in table 6.1. This snapshot in time captures the round of enlargement on May 1, 2004. Thus, represented in the parliament at this time are such political parties as the German Christian Democratic Union (CDU), which has had members in the EP since the first constituent meeting of the assembly, and also the Eastern European parties that have only recently made affiliation decisions. In total, 182 national parties were represented in 7 political groups in early June 2004. Second, we gathered data on MEP switching at the individual level, collected from the party listings in the Medlemsfortegnelse (List of Members) published (at irregular intervals) each year. This part of our analysis is restricted to the Third EP (1989–1994). Two time points were chosen, membership immediately after the 1989 EP elections and membership in June 1993. The key advantage of selecting 1993 is that this was the year in which the co-decision procedure came into effect, a watershed moment in the history of the EP. There were 10 parties for each member to choose in 1989 and 8 in 1993.5 Third, for party positions on policy we draw on expert surveys of the policy positions of European party groups conducted by the authors from April to June 2004, and on expert surveys of the policy positions of European national parties reported in Benoit and Laver (2006). The survey of the EP party group positions is detailed in McElroy and Benoit (2007). In this survey of experts, conducted just before the June 2004 elections to the newly expanded EP, each party group was scored on a 20-point scale presenting two extremes of policy.
Party Group Switching in the EP
157
Substantive policy dimensions covered in the survey included a set of four “core” dimensions deployed in every country in the Benoit and Laver data: increase spending versus reduce taxes; liberalism versus conservatism on social policy; pro-growth versus pro-environmental policy; and pro- versus antideregulation. Also included were a set of three questions about EU integration addressing party stances on the scope of EU authority, national control versus federalism, and approach to collective security. A full list of the question wordings and dimensions appears in McElroy and Benoit (2007). Following the Benoit and Laver (2006) methodology, we also measured the relative policy salience of each dimension to each political group. The results were therefore easily comparable to the Benoit and Laver data, which we employed for all measures of party policy at the national level for the EU-25 countries, covering a total of 146 of the 182 EU member state parties affiliated with EP party groups in June 2004. Finally, we draw on a number of individual-level data gathered on MEPs from the Third Parliament. In addition to measures of legislative privileges, we also estimated the ideal points of legislators by analyzing their roll-call votes using Poole and Rosenthal’s (1991) NOMINATE procedure.6 Roll-call vote analysis is common in the EP as a means of estimating party voting cohesion as well as the ideal points of parties and MEPs (see for instance Hix, Noury, and Roland 2007). Roll calls represent roughly one-third of all votes taken in the EP and the percentage of such votes has increased dramatically over the course of the past three decades. In addition to positions estimated from roll-call voting, we also compiled a set of career data on each MEP in the Third Parliament, tapping such variables as whether an MEP was in a leadership position in his or her party and what level of seniority the MEP had attained within his or her political group. As we have already pointed out, our examination should be treated as preliminary inasmuch as the data we have gathered is largely based on snapshots of party group affiliation in the EP at particular time periods. Nonetheless, to the extent that individual MEP switching has received almost no attention, and national party switching none, in previous comparative work, the largely static data analyzed here should form a valuable first step in establishing more completely the basis for a fully dynamic model of party group switching in the EP by individual members as well as national parties. As we show in the following text, the basis for affiliation and switching appears to be motivated most strongly by general policy concerns, and this finding applies to both levels we examine.
158
Gail McElroy and Kenneth Benoit
6.4 Characterizing the European Policy Space Our overarching argument is that ideological divides at the national level and the preferences for policy that they imply drive legislative affiliation in the EP. For this argument to be sustained, the structure of party competition and policy contestation should be broadly similar at the national and EU levels. Before we examine the determinants of party group switching, therefore, we must establish that national and European parties inhabit the same general policy space. Although some have expected that the policy space in the EP will reflect divisions over the institutional and political development of the EU itself (Thomassen and Schmitt 1997), we believe that the more the Union integrates, the more it should be concerned with the issues that define national party competition, just as would be the case for any other federal entity. Indeed, Hix (2004) suggests the closest analogy for parliamentary parties in the EP is parties in the US Congress, with different US states represented by state delegations formed into political parties at the federal level. We therefore take as a maintained hypothesis that the political groups in the EP will structure the legislative behavior of their constituent national parties in the same fundamental political space that exists at the national level. If national and European parties do in fact inhabit a common policy space, two observations should hold. First, we should observe that the policy space at the national and transnational levels has a basically similar structure. In other words, issues relating to the “European” level should be fully integrated into national level competition. Second, policy competition at the European level should reflect policies formulated and contested at the national level, in a similar dimensional space.
6.4.1
Consistency of National and European Policy Positions
Our first comparison concerns the match between policy positions held by national member parties and those held by their party groups in the EP. For the general left-right dimension, figure 6.1 depicts the correspondence between party groups and member parties. Each dot represents the mean left-right policy position, and each bar, a 95 percent confidence interval. The mean positions for the EP party groups come from the EP expert placements, while the mean positions for the national member parties come from the mean expert placements of the parties in each party group. On the general left-right
Party Group Switching in the EP
159
EP EDD EP UEN Natl UEN EP EPP Natl EPP EP ELDR Natl ELDR EP PES Natl PES EP Verts Natl Verts EP GUE Natl GUE 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Left−Right Dimension (1=Left, 20=Right) Figure 6.1 European party groups and national members on the general left-right scale.
dimension of policy, the party groups are ranked in the same order at both levels, with the rankings between party groups and national party means statistically indistinguishable for every party grouping. The left-right space appears to consist of four sets of party groups. First, the GUE and Verts appear on the far left, with a small degree of overlap in their 95 percent confidence intervals. Second, the PES inhabits the moderate center left. Third, two political groups, the ELDR and the EPP, form a grouping with a substantial degree of overlap just right of the center. Finally, the UEN and the EDD occupy the solid right position, with their positions being statistically indistinguishable. Similar results (not shown) were observed for Taxes versus Spending, the most important economic dimensions of policy, and for EU Authority and Immigration, the two most salient transnational issues. The most important exceptions on these dimensions were for the GUE, whose position was more pro-Integration than the mean of its national member parties, and the UEN and EDD, whose positions were considerably more extreme on antiintegration than the mean of their national member parties. On the dimensions of environmental policy and social liberalism, we also observed excellent correspondence between party group positions and the means of their national member party positions.
160
Gail McElroy and Kenneth Benoit
6.4.2
Common Structure of Policy Contestation
One implication of the “Europeanization” of political contestation is that the organization of policy contestation between the European and national arenas should demonstrate convergence. In other words, we expect to observe that political dimensions are associated with one another in the same ways across both arenas, as European-level political contestation becomes an extension of political competition at the national level. The expert surveys provide a means to map the organization of policy contestation, using the technique of factor analysis to analyze the correlation of issue positioning with underlying factors. By estimating the number of these factors and analyzing their principal components, we can compare political contestation in both the national and European arenas. Previous research on the dimensionality of policy competition in the European political space draws mixed conclusions about whether the EU policy space is unidimensional or consists of two or possibly more dimensions. Some scholars describe the EP policy space as unidimensional, with traditional left-right or “regulation” issues dominating (Kreppel and Tsebelis 1999; Tsebelis and Garrett 2000) or with geopolitical pressures defining the principal axis of competition (Hoffman 1966; Moravcsik 1998). Other analysts depict the European policy space as consisting of two dimensions, a left-right dimension composed of economic and sociopolitical issues from the domestic arena and an orthogonal dimension of EU integration versus national sovereignty (Hix and Lord 1997). Hence, scholars disagree on whether positions on EU integration are significantly correlated with those on left-right issues (e.g., Gabel and Hix 2004; Hooghe and Marks 2001) or instead positioning on the two dimensions is independent. In table 6.3, we report the results of a principal components factor analysis that groups and separates the constituent policy dimensional scorings into orthogonal factors. For each policy variable, we observe the positions of 153 national parties as well as the positions of the 7 EP party groups. The results lend strong support to our interpretation of the structure of EU policy contestation as not only twodimensional but also similar at the national and transnational party levels. The bottom part of the table provides the varimax-rotated factor loadings for the constituent policy, with the higher rotated loadings for each of the two factors highlighted in boldface. Unlike many exploratory factor analyses, these results show a remarkably
Party Group Switching in the EP
161
Table 6.3 Dimension reduction of policy contestation using principal components factor analysis Factor 1 2 3 4 5 6 7 8 9
Eigenvalue
Cumulative Proportion
3.23 1.46 0.65 0.34 0.23 0.07
0.54 0.78 0.89 0.95 0.99 1
Rotated Factor Loadings Factor
Taxes vs. Spending Deregulation Environment Social Immigration/Nationalism EU Authority N
1
2
0.92 0.96 0.78 0.33 0.46 –0.28
0.15 0.01 0.31 0.76 0.77 0.78 160
Note: Rotated scores use varimax rotation; variables are party mean scores. Boldface in panel 1 highlights that the two factors explain a total of 78% of the variance in the items. Boldface in panel 2 distinguishes between those items with high loadings on the first factor and those with high loadings on the second.
unambiguous result: Left and right issues cluster into two orthogonal component sets, one related to classic left-right, “domestic”-level issues from national party politics, and the second related clearly to EU and “transnational” issues such as immigration (although also including a high loading from social liberalism). The inclusion of immigration with the second “EU” factor is not surprising, since immigration and migration are closely associated with the expansion of the EU among older member states and with the issue of nationalism among the postcommunist countries. Since the policy space is structured in the same way in both the national and European arenas, we expect national and europarties to be located along the same underlying latent dimensions of policy. That this is in fact the case we observe in figure 6.2, which plots the left-right scores for each party and party group against the first factor
162
Gail McElroy and Kenneth Benoit
20
CZ−NKL PL−UPR HU−MI...P IT−MSFT
EE−Ref
BE−FN BE−VB CY−NEO LV−TP NL−LPFCY−DISI SE−M EE−ResP SK−OKS ES−PP SI−NSiSK−KDH IT−LN NL−SGP LV−JL UK−Con NL−VVD LV−TB/LNNK SE−KD SK−ANO CY−ADIK LT−TS LT−LiCS SK−SKDU IT−FI CZ−US FI−KOK GR−ND CZ−ODS PL−PiS
LU−ADR HU−FKGP IT−AN PL−LPR FI−PS CZ−RMS HU−FIDESZ SK−SNS
EE−EKRP
15
EDD UEN
SI−SDS BE−VLD CY−EDI ES−PNV BE−N−VA SE−FPEE−Isam SI−SLS FI−SFP HU−MDF NI−UUP SK−SMK LU−DPPL−PO NL−CDAES−CiU LV−LPP LU−CSV LT−LKD PL−AWS LT−LKDS NI−DUP BE−MR IT−UDC BE−CD&V NL−CU SE−C CZ−MDS CY−KOP FI−KESK IT−Pann LT−LDP MT−NP CZ−SNK PL−UW CZ−KDU LV−ZZS UK−Lab SK−HZDS CY−DIKO BE−CDH EE−RL GR−PASOK NL−D66 SI−SNS IT−It.Val. NI−APNI HU−CENTRUM
Left-Right Score
FI−KD
PPE ELDR
10 SE−MP GR−KKE LU−G
5
NL−GL
SE−V
LU−DL
EE−M??d LT−NS/SL MT−MLP EE−Kesk SI−SMS PL−PSL IT−SDI NL−PvdA NI−SDLP FI−SDP EE−E
FI−VIHR CZ−SZ
V
HU−SZDSZ
PSE
PL−S
FI−VAS BE−PS IT−Green EE−ESDTP ES−IU BE−Gro! BE−Eco LV−PCTVL IT−PDCI NL−SP CY−AKEL CZ−KSCM HU−MUNK¡S IT−RC SK−KSS
SK−SDL’
SI−ZLSD
GUE
PL−UP
0 −3
−2
−1
0 First Factor
1
2
Figure 6.2 Relationship of first factor to independent measures of left-right positions.
scorings from table 6.3; the left-right scores derived from expert placements performed independently are arrayed on the vertical axis, and the more specific policy placements listed in table 6.3, on the horizontal axis. The results are clear: the first underlying factor in EU party policy contestation, explaining .54 of the observed variance in policy positioning (table 6.3), consists of positioning on classic left-right policy issues. Furthermore, the EP party group scorings on this dimension lie squarely along its main axis. Only the UEN and the EDD are slightly off the main fitted line, a finding probably explained by the salience of the second dimension (EU authority) for both parties. From these results we draw two conclusions. First, political contestation in both the European and national arenas occurs on two broad dimensions. The first represents the classic national policy issues associated with left and right, namely economic policy, as well as a bundle of relatively new issues such as the environment and immigration. The second dimension consists of support for EU integration, relating directly to the authority and institutions of the EU itself, as well as immigration policy and (partially) social liberalism. Our results thus provide strong support for the model of policy competition based on two orthogonal dimensions. These two latent factors explain more
Party Group Switching in the EP
163
than three-quarters of the variance in party positions on specific policy dimensions in all of the models we tested. The second major conclusion has to do with the remarkable similarity in results between the national and European arenas. The organization of policy contestation in the European arena almost perfectly matches that found in the national arenas, with only slight differences regarding the relative association of immigration and social liberalism with the two factors. The European political space, in other words, bundles the same (primarily economic) left-right issues contested at the national level. These results strongly support the idea of a similar policy space shared by the national and domestic arenas in EU politics. Having established that the policy space at both the national and EU levels is broadly similar, we address patterns of affiliation and switching in the EP, first at the party group and then at the individual MEP levels. We expect that policy compatibility will largely drive party group affiliation in the EP and explain MEP patterns of switching.
6.5 Explaining EP Party Group Affiliation To examine the basis on which national political parties choose to affiliate with political groups in the EP, we compare the distance between national political parties and their EP party groups. According to the first hypothesis, a fundamental logic of policy coherence drives national party affiliation at the European level. Ideally, evidence on party group switching by national parties would illuminate the dynamic calculus of affiliation. We offer this more limited static analysis as a preliminary (albeit unprecedented in the field) step. Even though the policy positions of EP party groups are generally close to their national members’ median positions, as shown, considerable heterogeneity within europarty grouping appears. In each EP party group, there are clearly some national parties whose positions are far from their party group’s position. Some of the political groups do indeed cover a reasonably wide range of the political spectrum and clearly contain national parties within their ranks that are less ideologically compatible than others. It is quite possible, however, that a national party may be quite far from its political group mean on a given dimension but that EP group may still have the nearest EP party mean to the national party’s own ideal point. In what follows we map and explore this question using a discrete choice model.
164
Gail McElroy and Kenneth Benoit
To test the determinants of group affiliation, we have fitted a conditional logit (CL) regression model of party group choice to the absolute distance between a national party’s policy position and the policy position of the EP group on the same dimension. The estimation is based on a discrete choice model, with each national party choosing among all political groups in the EP as potential objects of affiliation. We utilize a conditional choice model as it is the most preferred technique in many situations in the social sciences with unordered dependent variables. In particular, the conditional choice model is preferable when a choice among alternatives involves functions of the alternatives rather than characteristics of the individual making the choice. The CL model allows the choice set to vary from case to case. In our model, the independent variable is purely choice-related, consisting of the absolute difference between the national party’s mean point (on a given dimension) and the mean policy position of the EP political group under consideration. The variable thus takes on different values dependent on the response category for each national party. In addition, there are two control variables. The first is the seat share of the political groups. This measure is included to capture the concept that electorally successful national parties (regardless of the absolute number of MEPs they elect) may be predisposed to association with the larger political groups in the EP for reputational and influence reasons (Maurer, Parkes, and Wagner 2008). The second variable is the national party position on the decentralization dimension, a dimension not included in the expert survey on the positioning of the EP political groups and yet potentially important in affiliation decisions (Bomberg 1998). Table 6.4 presents the results of the CL estimation of EP party group choice, for all of the national parties represented in the EP in June 2004. The coefficients on the model indicate the overall effect of the six choice-specific characteristics. To illustrate, Taxes Distance measures the absolute distance between an observation (a national party) and a given EP political group on this dimension. As shown, all of these variables are in the expected direction and five of the six are significant (EU Authority, Left-Right, Social, Environment, and Deregulation) at the 0.05 percent level. Using only the measures of policy distance on the six dimensions chosen—and excluding any national party, party family, electoral, seat share, or other variables of any kind—our simple model correctly predicted 69 percent of the party group memberships of the national parties.
Party Group Switching in the EP
165
Table 6.4 Conditional logit model of party choice, parties represented in EP early June 2004
Variables
Maximum Likelihood Conditional Logit Coefficients
Left-Right distance Taxes distance Social distance Environmental distance EU distance Deregulation distance Vote Share PES Vote Share ELDR Vote Share Verts Vote Share GUE Vote Share UEN Vote Share EDD Decentralization PES Decentralization ELDR Decentralization Verts Decentralization GUE Decentralization UEN Decentralization EDD PES ELDR Verts GUE UEN EDD N Total parties 2LL
–0.28 (–0.48, –0.08) 0.05 (–0.20, 0.30) –0.35 (–0.49, –0.20) –0.28 (–0.48, –0.09) –0.12 (–0.24, –0.01) –0.36 (–0.58, –0.15) 0.10 (0.03, 0.18) –0.01 (–0.07, 0.05) –0.21 (–0.45, 0.02) 0.00 (–0.12, 0.13) –0.11 (–0.22, –0.01) –1.21 (–2.61, 0.19) –0.05 (–0.45, 0.35) –0.19 (–0.44, 0.07) –0.54 (–1.07, –0.01) –0.24 (–0.74, 0.26) 0.37 (–0.01, 0.75) –0.08 (–0.54, 0.37) –0.82 (–4.74, 3.10) 3.10 (0.44, 5.77) 7.03 (2.15, 11.91) 2.62 (–1.98, 7.21) –3.71 (–7.51, 0.10) 3.88 (–2.24, 9.99) 1,036 148 –187.87
Note: 95% confidence intervals in parentheses, with bold coefficients statistically significant at the .05 level.
The coefficients on the CL model indicate the overall impact of the party-specific characteristics on the likelihood of choosing an EP group. This means that as the absolute distance of the national party’s mean increases from that of the European political group in question, the less likely this EP group is to be chosen. In particular, as the distance between a national party and an EP political group increases on the left-right dimension, the less likely this party will affiliate with it (negative sign on the L-R variable). The evidence in table 6.4 demonstrates that increasing the distance by one unit for a given political group choice on the Left-Right dimension decreases
166
Gail McElroy and Kenneth Benoit
the odds of choosing that group by a factor of .80 (20 percent). It is worth reiterating that these effects depend only on traits specific to the EP political group. Regardless of the individual characteristics of a national party, it will choose an EP political group consistent with its own policy positions.
6.6
Switching by Individual MEPs
Turning now to switching by MEPs, table 6.5 summarizes the results of a mixed CL model of individual affiliation choices. In this model, the observational unit is the individual MEP, who at two time points chooses to affiliate with one of the political groups. The mixed model, unlike the model reported in table 6.4, contains both choice-specific and unit-specific variables. The choice-specific variables capture whether a political group is represented in the Parliamentary Bureau or is one of the two major parties.7 Most importantly for the second hypothesis here, the distance between an individual MEP’s ideological preferences and the position of a given political group is tapped in the Distance variable. As this policy distance increases, again, an MEP should be less likely to affiliate with a given group. The unitspecific variables are dummies to capture whether a given MEP is a senior member of his or her party in terms of tenure or a party president or vice president (party Leader). The coefficients on the CL part of the model indicate the overall effect of choice-specific (political group) characteristics on the propensity of an MEP to choose one particular political group over another. First, note that the Distance coefficient, we see that it is strongly significant and in the expected direction. As the distance between an MEP and a party increases, the less likely he or she is to choose it (negative sign on the Distance variable). For example, holding all other variables constant, an increase in distance of 0.05 would decrease the probability of choosing an alternative party by a factor of exp (–15.15*0.05) = 0.46. Second, again holding all else constant, the odds of choosing a party that is in the bureau is exp (1.47) = 4.3 times that of choosing a party not represented in the bureau. Third, the odds of choosing one of the two largest parties are exp (0.81) = 2.2 times those of choosing a small party. It is worth reiterating that these effects depend only on the partyspecific traits. Regardless of the individual characteristics of an MEP, he or she will prefer a large party to a small party, a party in the bureau to one outside of the bureau and a party that is ideologically consistent with his or her preferences.
Party Group Switching in the EP
167
Table 6.5 Conditional logit model of individual MEP switching, third Parliament Variable
Coefficients
Distance Major Party Parliamentary Bureau Session Party Leader* EPP Seniority* EPP Party Leader* ELDR Seniority* ELDR Party Leader* Greens Seniority* Greens Party Leader* ED Seniority* ED Party Leader* ARC Seniority* ARC Party Leader* DR Seniority* DR Party Leader* RDE Seniority* RDE Party Leader* CG Seniority* CG Party Leader* GUE Seniority* GUE EPP ELDR Greens ED ARC DR RDE CG GUE Number of cases 2LL Pseudo R 2
–15.15*** 0.81*** 1.47*** 0.20 0.51* 0.1 2.30*** –0.34 3.62** –1.00*** 0.38 0.19*** 1.25* 0.20 0.66 –0.3 2.33*** 0.21 1.47* 0.18* 0.99 –0.37** –0.48 –0.75*** –2.11*** 0.63 –0.06 1.30 –0.05 –0.56 –1.30* 8,591 –652.14 0.73
Note: * significant at the p<.05 level, ** significant at the p<.01 level, *** significant at the p<.001 level.
As for the unit-specific variables in the model, the coefficients for party leader are significant for a majority of the parties under consideration and in the expected direction for all parties. These estimates indicate that, for instance, the odds of a party leader in the EPP choosing the EPP (versus PES) are exp (0.51) = 1.66 times those of a
168
Gail McElroy and Kenneth Benoit
nonleader in the same party. These odds increase to exp (1.47) = 4.3 when we consider the Left Unity Party (CG). Seniority is only significant for four of the parties and works in both directions. Senior members of the CG and ED are less likely to defect to the PES than are members with no seniority in these parties. However, the coefficients reveal that the relationship is in the opposite direction for the GUE and the Green party; senior members of these parties are more likely to defect than junior members. To illustrate, these estimates indicate that the odds of an MEP choosing ED over PES is exp (0.19*5) = 2.5 for a senior member of the ED. For the GUE group, however, a member with similar seniority rank would be 1/exp (–.37*5) = 6.3 times more likely to choose PES. Although these mixed results may reflect the absence of a seniority norm in some parties in the EP, more research is clearly needed in this area before firm conclusions can be reached on the effects of seniority on individual switching. Overall, these results are encouraging. As in the model of national party choice of EP political group, ideological distance from parties emerges as a very important factor in accounting for party choice among individual MEPs. Moreover, the evidence suggests that factors such as holding party posts may deter an MEP from switching. Large parties, holding all other variables constant, are also favored in the calculus of affiliation in the EP.
6.7
Conclusion
In this chapter, we examine party choice and party switching in the EP from a number of angles. We have established that national parties affiliate with political groups primarily on the basis of ideological compatibility. A similar logic applies to individual MEP switches. In sum, policy matters. The results strongly indicate that proximity of a party group to a national party’s policy platform determines the likelihood that the national party will be affiliated with that EP group. Parties tend not to affiliate with party groups that are farther from their own preferred positions. In addition, our findings show that political contestation, as measured by issue bundling and the calculus of party group affiliation, is similar in both the national and European arenas. The evidence discloses that, as posited, policy concerns drive MEP choices and changes of party affiliation; career considerations also play into
Party Group Switching in the EP
169
MEP decisions to switch. Our findings at the individual level and the party system level reinforce each other: Because MEPs’ careers are grounded in their national parties, the logic of switching implies that the policy spaces occupied by national and EU parties must be the same, or at least comparable. Although we have successfully and systematically demonstrated that policy underlies national party affiliation with transnational EP party groups—a first in the literature—this part of our picture has been essentially static rather than dynamic. Further research into the calculus of party affiliation in the European arena should focus on changes over time, especially national party switching between groups from 1979 to 2004. The authors are currently collecting these data, but the practical challenges are enormous. We have confirmed that the area is of sufficient theoretical and substantive importance, however, that the full dynamic deserves to be investigated.
6.8
Notes
* Support for this project was provided by the Institute for International Integration Studies, Trinity College, Dublin and by the European Commission Fifth Framework (project number SERD-2002–00061). We thank Marina McGale, Alex Baturo, and Slava Mikhailov for research assistance. 1. Expulsions are not common but do happen. A recent example is the case of Daniel Hannan, a member of the UK Conservative Party. Hannan was ejected from the EPP in early 2008 for attempting to use parliamentary procedures to obstruct the passage of the Lisbon Treaty. 2. For instance, French MEPs from Union pour un Mouvement Populaire divided between the EPP and UEN in the 1999–2004 Parliament. 3. Preliminary evidence undertaken by the authors suggests that upward of 40 percent of all national member parties have switched EP party groups at one point or another. 4. In addition, turnout is frequently very low. In the United Kingdom, for example, turnout in the 1999 European elections was an abysmal 24 percent, compared with 71 percent in the national election of 1997. 5. Further information on these data is included in McElroy (2008). The Third Parliament represents the only full legislative period in which the EP does not expand in size due to enlargement. Enlargements tend to change the dynamics of the institution in key ways for which we wish to control. 6. The estimation is based on a random sample of 500 roll-call votes for the parliamentary period 1989–1994. The data were collected from the Official Journal of the European Communities, Series C. Following Poole and Rosenthal (1991), for each legislative period, we include every legislator who cast at least 25 votes. In addition, votes where less than three percent of those voting voted against the majority are excluded.
170
Gail McElroy and Kenneth Benoit
7. The Bureau is the regulatory body responsible for parliament’s budget and for internal administrative issues. Voting members include the parliament’s president and the 14 vice presidents.
6.9
References
Aldrich, John H., and William T. Bianco. 1992. “A Game Theoretic Model of Party Affiliation of Candidates and Office Holders.” Mathematical and Computer Modeling 16: 103–115. Ames, Barry. 2001. The Deadlock of Democracy in Brazil. Ann Arbor: University of Michigan Press. Attinà, Fulvio. 1990. “The Voting Behaviour of the European Parliament Members and the Problem of Europarties.” European Journal of Political Research 18 (5): 557–579. Bardi, Luciano. 2002. “Parties and Party Systems in the European Union: National and Supranational Dimensions.” In Political Parties in the New Europe: Political and Analytical Challenges, ed. Kurt Richard Luther and Ferdinand Müller-Rommel, 293–321. Oxford: Oxford University Press. Benoit, Kenneth, and Michael Laver. 2006. Party Policy in Modern Democracies. London: Routledge. Bomberg, Elizabeth E. 1998. Green Parties and Politics in the European Union. London: Routledge. Budge, Ian, Hans-Dieter Klingemann, Andrea Volkens, Judith Bara, and Eric Tanenbaum. 2001. Mapping Policy Preferences: Estimates for Parties, Electors, and Governments. Oxford: Oxford University Press. Damm, Sven Mirko. 1999. “Die europäischen politischen Parteien. Hoffnungsträger europäischer Öffentlichkeit zwischen nationalen Parteien und europäischen Fraktionsfamilien.” Zeitschrift für Parlamentsfragen 30 (2): 395–423. Daniels, Philip. 1998. “From Hostility to ‘Constructive Engagement.’ The Europeanisation of the Labour Party.” West European Politics 21 (1): 72–96. Dietz, Thomas. 1997. Die grenzüberschreitende Interaktion grüner Parteien in Europa. Opladen: Westdeutscher Verlag. ———. 2000. “Similar but Different? The European Greens Compared to Other Trans-national Party Federations in Europe.” Party Politics 6 (2): 199–210. Downs, Anthony. 1957. An Economic Theory of Democracy. New York: Harper and Row. Gabel, Matthew, and Simon Hix. 2004. “Defining the EU Political Space: An Empirical Study of the European Election Manifestos.” In European Integration and Political Conflict, ed. Gary Marks and Marco R. Steenbergen, 93–119. Cambridge: Cambridge University Press. Hix, Simon. 2004. “Electoral Institutions and Legislative Behavior: Explaining Voting Defection in the European Parliament.” World Politics 56 (January): 194–223. Hix, Simon, and Christopher Lord. 1997. Political Parties in the European Union. Basingstoke: Macmillan.
Party Group Switching in the EP
171
Hix, Simon, Abdul Noury, and Gérard Roland. 2005. “Power to the Parties: Cohesion and Competition in the European Parliament, 1979–2001.” British Journal of Political Science 35 (2): 209–234. ———. 2007. Democratic Politics in the European Parliament. Cambridge: Cambridge University Press. Hoffmann, Stanley. 1966. “Obstinate or Obsolete? The Fate of the Nation State and the Case of Western Europe.” Daedalus 95 (3): 862–915. Hooghe, Liesbet, and Gary Marks. 2001. Multi-Level Governance and European Integration. Lanham, MD: Rowman and Littlefield. Klingemann, Hans-Dieter, Richard L. Hofferbert, and Ian Budge. 1994. Parties, Policies, and Democracy. Boulder, CO: Westview Press. Kreppel, Amie. 2002. The European Parliament and Supranational Party System: A Study in Institutional Development. Cambridge: Cambridge University Press. ———. 2004. “Moving in the Other Direction? The Impact of Domestic Party System Change on Italian MEPs.” Journal of European Public Policy 11 (6): 975–999. Kreppel, Amie, and George Tsebelis. 1999. “Coalition Formation in the European Parliament.” Comparative Political Studies 32 (8): 933–966. Lipset, Seymour M., and Stein Rokkan. 1967. “Cleavage Structures, Party Systems, and Voter Alignments: An Introduction.” In Party Systems and Voter Alignments: Cross-National Perspectives, ed. Seymour M. Lipset and Stein Rokkan, 1–64. New York: Free Press. Mair, Peter. 2000. “The Limited Impact of Europe on National Party Systems.” West European Politics 23 (4): 27–51. Marks, Gary, and Carole J. Wilson. 2000. “The Past in the Present: A Cleavage Theory of Party Response to European Integration.” British Journal of Political Science 30 (3): 433–459. Marks, Gary, and Marco Steenbergen. 2004. European Integration and Political Conflict. Cambridge: Cambridge University Press. Marks, Gary, Carole Wilson, and Leonard Ray. 2002. “National Political Parties and European Integration.” American Journal of Political Science 46 (3): 585–594. Marsh, Michael. 1998. “Testing the Second-Order Election Model after Four European Elections.” British Journal of Political Science 28: 591–607. Maurer, Andreas, Roderick Parkes, and Markus Wagner. 2008. “Explaining Group Membership in the European Parliament: The British Conservatives and the Movement for European Reform.” Journal of European Public Policy 15(2, March): 246–262. McElroy, Gail. 2002. “Committees and Party Cohesion in the European Parliament.” EPRG Paper, No 8. Paper prepared for presentation at the 2001 annual meeting of the American Political Science Association, San Francisco, California, August 30. ———. 2003. “Party Switching in the European Parliament: Why Bother to Defect?” Paper prepared for presentation at the 2003 annual meeting of the American Political Science Association, Philadelphia, Pennsylvania, August 28–31.
172
Gail McElroy and Kenneth Benoit
McElroy, Gail. 2008. “Intra-Party Politics at the Trans-national Level: Party Switching in the European Parliament.” In Intra-Party Politics and Coalition Governments in Parliamentary Democracies, ed. Daniela Giannetti and Kenneth Benoit, 205–226. London: Routledge. McElroy, Gail, and Kenneth Benoit. 2007. “Party Groups and Policy Positions in the European Parliament.” Party Politics 13 (1): 5–28. McKelvey, Richard D., and Normal Schofield. 1987. “Generalised Symmetry Conditions at a Core Point.” Econometrica 55 (4): 923–933. Moravcsik, Andrew. 1998. The Choice for Europe, Social Purpose and State Power from Messina to Maastricht. Ithaca, NY: Cornell University Press. Niedermayer, Oskar. 1997. “Parteien auf der europäischen Ebene.” In Parteiendemokratie in Deutschland, ed. Oscar W. Gabriel, Oskar Niedermayer, and Richard Stöss, 443–458. Opladen: Westdeutscher Verlag. Nugent, Neill. 2003. The Government and Politics of the European Union. Basingstoke: Palgrave. Poole, Keith T. and Rosenthal, Howard. 1991. “Patterns of Congressional Voting.” American Journal of Political Science 35 (1), 228–278. Raunio, Tapio. 1997. The European Perspective: Transnational Party Groups in the 1989–1994 European Parliament. Aldershot: Ashgate. Ray, Leonard. 2003. “When Parties Matter: The Conditional Influence of Party Positions on Voter Opinions about European Integration.” Journal of Politics 65 (4): 978–994. Reif, Karlheinz. 1997. “Reflections: European Elections as Member State Second-Order Elections Revisited.” European Journal of Political Science 31 (1): 115–124. Reif, Karlheinz, and Hermann Schmitt. 1980. “Nine Second Order National Elections: A Conceptual Framework for the Analysis of European Election Results.” European Journal of Political Research 8 (1): 3–44. Strøm, Kaare. 1990. “A Behavioral Theory of Competitive Political Parties.” American Journal of Political Science 34 (2): 565–598. Taggart, Paul. 1998. “A Touchstone of Dissent: Euroscepticism in Contemporary Western European Party Systems.” European Journal of Political Research 33 (3): 363–388. Thomassen, Jacques, and Hermann Schmitt. 1997. “Policy Representation.” European Journal of Political Research 32 (2): 165–184. Tsebelis, George, and Geoffrey Garrett. 2000. “Legislative Politics in the European Union.” European Union Politics 1 (1): 9–36.
7 Legislator Preferences, Party Desires: The Impact of Party Switching on Legislative Party Positions William B. Heller and Carol Mershon
7.1 Introduction Democratic politics and political parties go hand in hand. Politicians win elections and hold office as members of parties (Epstein 1967). For their part, political parties organize legislatures and manage the passage of policy (Aldrich 1995; Cox and McCubbins 1993; 2005). Legislators’ political identities are tightly linked to their party affiliations, even where parties are seen as relatively weak vis-à-vis individual politicians. In this light, party switches, particularly when executed by sitting legislators, are curious and perhaps even bizarre. As the first chapter in this book emphasized, there is on one hand the motivational question: why would a legislator decide to change his or her party affiliation during a legislative term? As we also highlighted at the outset, there is on the other hand the practical question of policy consequences: what difference does party switching make? We take up the latter issue here by asking how party switching by sitting legislators affects the preferences of legislative parties. To grasp the interplay of individual and party preferences, we need to address what individual members of parliament (MPs) get out of party affiliation. To this end, we first briefly examine the literatures on parties qua legislative actors, individual legislative behavior, and party switching. We next sketch a general model of party switching, which we use as a guide to tackling the question of the relationship between individual and party preferences. The logic of the model leads us to several testable hypotheses, which we examine in the fourth section. The final section discusses the broader implications of the research.
174
William B. Heller and Carol Mershon
7.2 Legislators and Parties Legislative parties are routinely assumed to behave as unitary actors (Laver and Schofield 1990). Even students of US politics, where parties are notoriously unable—or perhaps unwilling—to impose unity on their members, often have treated parties as units conditional on their members having relatively homogeneous preferences (e.g., Aldrich and Rohde 2000; Cooper, Brady, and Hurley 1977; Rohde 1991). Party switching is so rare in the United States that the widespread assumption that legislative parties are stable in size tends to escape notice. Analysts of European parliamentary systems, when investigating the formation and duration of governments, customarily treat parties as unitary actors with legislative weights that remain fixed throughout each legislative term. Party preferences are likewise regarded as fixed during legislative terms; in line with this standard assumption, the oft-used Manifesto dataset on party positions is composed of measurements taken at election time (Budge et al. 2001). The treatment of parties as unitary actors is a pleasant fiction, convenient for analysis but also recognized as misleading under some circumstances (Laver 1999; Laver and Schofield 1990; Laver and Shepsle 1990; 1999; Kiewiet and McCubbins 1991). The assumption that parties are fixed in weight and preferences also is a useful fiction, although not as clearly recognized as such (but see Heller and Mershon 2005a; forthcoming; Laver and Benoit 2003; Mershon and Heller 2003; Strøm 1994). On one hand, parties can change in reaction to exogenous events, as when many Communist parties altered their names and their platforms in the wake of the collapse of the Soviet Union. On the other hand, legislators can reevaluate their party affiliations in response to new information—for example, from opinion surveys or subnational elections—and thus can sometimes decide to change parties (Heller and Mershon 2005a). A party that attracts switchers without losing members to rival parties, obviously, increases its seat share, which in turn might make it more attractive to other potential switchers (Laver and Benoit 2003). Party switching often is seen as a pathology of political ambition (see, e.g., Mejía Acosta 1999; Sánchez de Dios 1999; Tomás Mallén 2002; Turan 1985) or a symptom of political systems in flux (see, e.g., Àgh 1999; Reed and Thies 2000). Whatever the reasons for switching—and, as stressed in chapter 2, analysts agree that ambition underlies them all (cf., e.g., Aldrich and Bianco 1992; Desposato 2006; Heller and Mershon 2005a; Laver and Benoit 2003;
Legislator Preferences, Party Desires
175
McElroy 2003)—it will affect party size and the distribution of partymember preferences under all but the most restrictive conditions. We explore party switchers’ motivations elsewhere (Heller and Mershon 2005a; 2008; Mershon and Heller 2003). As a party gains new members through switching, the preferences of its rank and file likely grow more heterogeneous. Preference heterogeneity might not increase if the “raw” preferences1 of members of different parties overlap, as illustrated in figure 7.1, where minority-party members j, l, n, or p could move into the majority without affecting the total range of majority-membership preferences. If membership preferences do not overlap, by contrast, switching necessarily increases the preference heterogeneity of the receiving party (as long as no members balance entries by exiting to join a different party2) and might (but will not necessarily) cause a corresponding decrease in the preference heterogeneity of the sending party. The proximate effects of party switching are clear. Along with changes in party size and heterogeneity, switching also should affect the aggregate distribution of party-member preferences. For example, if any of minority-party members j, l, n, or p in the case illustrated in figure 7.1 were to switch into the majority party, party size would increase and the range of members’ preferences would remain unchanged, but members’ preferences would be slightly skewed to the right related to their pre-switch distribution. If such a switch into the party were balanced by an exit on the left, party size also would remain unchanged, but the change in preference distribution might be even more pronounced. The import of party switching for any change in the distribution of preferences depends on whether and how such changes are expressed in party preferences or unity. On one hand, increased preference heterogeneity within the party could lead to less party unity, as argued for the US case in diagnoses of “conditional party government” (Aldrich and Rohde 2000; Rohde 1991; Krehbiel 1993; 1998; 1999b; 1999a; 2000). Yet greater preference heterogeneity could be advantageous for parties, particularly if it allows them to appeal to a wider audience (Shepsle 1972; but see, Alesina and Cukierman 1990). It also could force parties to use more resources to attract voters (because they Majority party f g h
a b c d e Figure 7.1
i
j
Majority-minority overlap k l m n o p q r
Parties with overlapping membership.
Minority party s
t
u v w x
y
176
William B. Heller and Carol Mershon
have to substitute something else for the diluted appeal of ideology; see Cox 1987), make vote outcomes less certain (cf. Best and Heller 2005), and disappoint legislators—and voters—who care about policy outcomes (Heller and Mershon 2005a). On the other hand, switches that move a party’s center of gravity (in terms of membership preferences) might presage changes in the party’s observed ideal point. It seems reasonable, after all, to posit some connection between the preferences of parties and the preferences of their MPs. Moreover, if a legislator who joins a party can pull it toward his or her own ideal point, that should affect potential switchers’ calculations. Of course, if intraparty delegation regimes imbue party leaders with potent tools for imposing discipline, the observable impact of switching will be whatever party leaders want it to be (Cox and McCubbins 1993; 2005; Kiewiet and McCubbins 1991). We have just affirmed a link between a party’s preferences and its MPs’ preferences. We also think it uncontroversial from either a normative or a positive viewpoint that there should be some relationship between a party’s preferences, as manifested in the behavior of its members in legislative votes, and the preferences of its members. Indeed, one of the most common claims about parties is that they aggregate preferences. And yet how they do so and to what effect are unclear. In pursuit of clarity, we turn now to explore and test some of the implications of changes in party membership relating to party preferences.
7.3
Leveraging Party Preferences
It makes sense to hold party sizes and preferences constant, as long as party memberships do not change. Party memberships do change, however. For example, new members can enter the legislature via byelections or as replacements for legislators who move into executive offices (where executive and legislative offices are legally incompatible). Further, and perhaps more notably, sitting legislators sometimes switch parties. It is easy to say that parties provide representation, in essence by aggregating citizen (or, more accurately, voter) preferences into policy programs. How they do this is not simple. At one extreme, legislators can represent their districts by voting for any and all legislation that benefits at least some substantial portion of their constituents. If all legislators behaved this way, parties would be irrelevant and policy outcomes would most benefit the constituents of the median legislator
Legislator Preferences, Party Desires
177
(or legislators, if the median changes with every issue). At the other extreme, legislators toe their party line on all votes. To the extent that constituencies are heterogeneous, such strong party voting suggests that most if not all such loyalist legislators will have to vote against the interests of their districts on some bills. This suggests that whom parties represent and how representative they are depends on how they aggregate preferences. Political parties that face no competition need neither aggregate preferences nor provide representation. Where parties compete for votes, office, and policy influence, however, a party that does not represent its constituents risks losing to a challenger who promises to do better. Similarly, a legislator who fails to represent the interests of those who put him or her in office risks being replaced at election time. An individual legislator, therefore, would prefer to be in a party that allows him or her at least to stake out positions that his or her constituents appreciate. Inasmuch as voters care about and cast their votes on the basis of policy outcomes, legislators also should value party unity both to signal a clear position to voters (Cox 1987) and to maximize bargaining weight in the legislature (Cox and McCubbins 1993; 2005; Laver and Shepsle 1999). To underscore the point: If clarity of party labels is important, then legislators should want their party to send clear signals about its positions. This suggests a desire to present a unified front to voters, which in turn requires high levels of party unity in legislative voting. 3 This observed unity means that it is possible to talk about party ideal points (Laver and Schofield 1990, among many others), but it does not imply that the preferences of party members do not matter. Rather, party ideal points should be some function of member ideal points, and a party’s ideal point should change as its membership changes. Legislators are elected on the basis of how electoral rules aggregate voters’ preferences (see, e.g., Cox 1997; Ordeshook and Shvetsova 1994; Riker 1982; Saari and Sieberg 2001). Party ideal points—which might be more or less clear depending on how tight is party unity— are determined on the basis of how internal party rules aggregate partisan legislators’ preferences. It thus makes sense that party ideal points should be sensitive to changes in party membership. How sensitive depends on how the aggregation works. There are essentially three possibilities for preference aggregation in a legislative party, depending on how party rules structure decision making. First, decision making could be geared to achieve consensus, resulting in choices that are tolerable to all party members,
178
William B. Heller and Carol Mershon
but possibly ideal for none. The outcomes of this kind of bargained decision making likely would approximate something like a mean of party-member preferences—something, in spatial terms, inside the party’s Pareto hull (Tsebelis 1999; 2000; 2002). Second, if the kinds of issues that define party positions are unidimensional, majority rule would privilege the median member of the party (Black 1958).4 (Decision rules also could bias party positions away from the median, but in a way that privileges some other clearly identifiable member— for example, a supermajority rule that demands the support of the two-third member.) The third possibility is for party leaders basically to set party positions on their own. Rules that privilege leadership preferences do not necessarily permit leaders to dictate their own preferences to their followers, because leaders are constrained by the need to retain rank-and-file support. Such rules do allow a leader to pull the party’s position relatively close to him or her own, but how close and how consistently depends on the distribution of rank-and-file preferences, and how party rules allow those preferences to be expressed. Assuming that incentives exist for at least some members to vote against the party line at least some of the time (as discussed earlier) maintaining unity requires use of the instruments of discipline (for a discussion of the sources of party unity, see Bowler, Farrell, and Katz 1999). Party unity is a collective good, and it is party leaders’ job to maintain it (Kiewiet and McCubbins 1991). Members will accept and tolerate leadership threats and punishments for this purpose as long as they value party unity, and as long as punishments are not too onerous. When punishments get to be too harsh, followers might be able to replace (or otherwise castigate) their leaders (Calvert 1987; Cox and McCubbins 1993; Shepsle and Bonchek 1997, ch. 14) or switch into a different party (Heller and Mershon 2005b; forthcoming; Laver and Benoit 2003; and cf. Hirschman 1970). Party switching is one of the tools available to legislators who want to maximize their own political fortunes and influence and maximize their ability to achieve their own policy goals. The choice to switch has two sides. On one side, as just suggested, it is a reaction to circumstances internal to the party. For example, the party program might be disagreeable (i.e., a legislator might switch because he or she essentially had miscalculated policy distances at the outset); perceived chances for reelection might be unacceptably low, perhaps because of a decline of the party in the polls (Heller and Mershon 2005a; cf. Zielinski, Slomczynski, and Shabad 2005) or a
Legislator Preferences, Party Desires
179
poor ballot list position (Aldrich and Bianco 1992); or opportunities for career advancement within the party might be inadequate (cf. Schlesinger 1966). On the other side, there must be something that makes the switcher’s “target” party appealing (cf. chapters by Benoit and McElroy and Desposato, this volume). A legislator who often finds himself or herself at odds with his or her party’s policies might want to move to a more compatible party, for example. In a more strategic vein, a legislator might switch in hopes of being able to influence his or her target party’s position—and that party’s leadership might try to encourage or dissuade him or her, depending on whether his or her entry would help or hinder leadership goals for the party. Possible enticements for potential switchers include promises of rapid advancement in the party (and a consequent stronger ability to influence party decisions), promises of relative freedom from the pressures of party discipline, and a guaranteed attractive ballot list position (see, e.g., Svåsand, Strøm, and Rasch 1997, 96). We examine some of the consequences of switching for individuals in other work (see, e.g., Heller and Mershon 2002; 2005a). Our focus in this chapter on party positions precludes closer analysis of individual legislators’ behavior, though we do draw some tentative inferences about party leaders’ strategies vis-à-vis individual members. In practical terms, the party leadership’s role in influencing decision rules, setting the agenda, and implementing and enforcing choices make it likely that the leadership will have a hand in determining party stances on important issues. Nonetheless, whether party rules favor leadership preferences or some other position—for example, a party member privileged by where his or her preferences fall relative to his or her copartisans, or a bargaining solution that places the party position at what is essentially the mean of the ideal points of its members—party ideal points should be sensitive to changes in members’ ideal points. This observation opens the door to several hypotheses. H1 Mean (or biased mean) ideal point hypothesis: The party ideal point should be close to some (possibly biased) average of the ideal points of its legislative members.
Possible sources of bias include the party leadership’s relationship with or dependence on external actors, as for example the role of trade union organizations in shaping the Labour party’s program in the United Kingdom (until Tony Blair won election as party leader
180
William B. Heller and Carol Mershon
in 1995 [Seyd 1998]) or an institutionalized requirement that representative organs of the mass membership be consulted on major decisions, such as the Irish Labour party rule, dating from 1977, that a special conference must approve party decisions to enter governing coalitions (Marsh and Mitchell 1999). This hypothesis implies that party ideal points often should be sensitive to changes in party membership, even if the median party member remains unchanged, and under some conditions could be insensitive to changes in the identity of the median member. H2 Median (or biased median) member hypothesis: The party ideal point should mirror the ideal point of some legislator identifiable by her position in the distribution of party-member ideal points.
If party decision-making processes include voting by members of the party’s legislative caucus, then the decision rule (e.g., majority or some kind of supermajority) should privilege specific members, just as majority-rule voting privileges the median voter. If the identity of the members thus privileged should change, then the party ideal point should change as well. This hypothesis implies that party ideal points should be sensitive only to changes in party membership, including the entry and exit of switchers, that affect the identity of the privileged member. H3 Leadership control hypothesis: The party ideal point should be close to the leadership ideal point, with a variance that depends on the location of rank-and-file ideal points.
To the extent that party leaders, by virtue of their position, wield more influence than rank-and-file members in party decision making, party positions should be disproportionately responsive to leaders’ preferences. Rank-and-file preferences matter because leaders are not dictators—if they wish to retain their leadership positions, they have to balance their own ambitions with those of their followers and the good of the organization (see the discussion of the “leadership dilemma” in Shepsle and Bonchek 1997; cf. Müller and Strøm 1999). The variance around leadership ideal points could well differ across parties, depending on how leaders are chosen (Kiewiet and McCubbins 1991), whether they face challenges from ambitious followers, and so on. That said, the party position should be relatively insensitive to changes in party membership, varying only to the extent
Legislator Preferences, Party Desires
181
that the leadership has to cater to the interests of members who are relatively distant from the established party position or who are in some sense pivotal and thus can sway leaders who need their support (cf. Best and Heller 2005). Because party leadership is interested in maintaining (or moving) the party ideal point as close as possible to its own, leaders might want to be strategic about how they treat potential in- and outswitchers. The possibility that party switchers could affect both party size and party ideal points (either prejudicially or beneficially from the perspective of leadership) suggests three subsidiary hypotheses: H3a Leaders should try to attract switchers who would help move party positions closer to their own. H3b Leaders should encourage outswitching by members who, by virtue of their own preferences and contrary to the preferences of most of their copartisans, constrain leaders from moving their parties in desired ways. H3c To the extent that inswitches improve a party’s legislative bargaining position (because they increase the party’s seat share and hence legislative weight), ceteris paribus, leadership should want to attract inswitchers even if they have no effect on the party ideal point. By the same token, leadership should seek to discourage balanced outswitching. H4 Fixed parties hypothesis: Party positions are independent from the positions of individual members, including leaders, and should not change with changes in legislative membership.
Our first three hypotheses arise from the assumptions that individual politicians use political parties as vehicles to gain and retain political power (e.g., Cox and McCubbins 1993; 2005; Müller and Strøm 1999), and that the majority legislative party (or majority coalition of legislative parties) operates in turn as a legislative cartel (Cox and McCubbins 1993; 2005; cf. Katz and Mair 1995). Thus, as the identities of the politicians in a party change, so too does the party. This kind of legislative party (see Katz and Mair 1994; 1995) is but one conception of what parties are, however. Scholars have identified several other types that should in principle be much less responsive to the preferences of their elected members. For instance, Duverger (1972) defined cadre parties as those dominant in the nineteenth-century era of limited suffrage and those organized and maintained by a small circle of elites who could ensure their political influence without holding
182
William B. Heller and Carol Mershon
office themselves. Alternatively, mass membership parties, whether in their heyday in the first half of the twentieth century or in their more attenuated contemporary form, serve as relatively fixed bases into which politicians sort themselves (Duverger 1972; Katz and Mair 1994; 1995; Scarrow 2000). Elected politicians in such parties should have little influence as legislators on their parties’ policies. Finally, catch-all parties, driven by the leadership’s concern to appeal to a broad and diverse electorate (Kirchheimer 1966), should not allow for a tacking of ideological sails to accommodate a shifting parliamentary membership. In sum, if parties are organized primarily as something other than legislative cartels (Cox and McCubbins 1993; 2005), we should not see party positions responding to changes in legislative-party membership. This is in essence our null hypothesis.
7.4
Evidence: Switches and Preferences
Hypotheses in hand, we now spell out our research design and measures and proceed to assess our expectations. Italy 1988–2000 is an especially apt setting for our empirical study because it offers pronounced variation in the nature and rates of party switching over time. From 1988 to 1994, switching in the Italian Chamber stemmed primarily from party splits. It was only after 1994 that solo switching became relatively common. Indeed, nearly one fourth of the MPs in the 1994–1996 and the 1996–2001 Legislatures switched parties at least once. In addition, extant research on switching in Italy concentrates on the causes of the phenomenon, be they systemic (Verzichelli 1996) or individual (Heller and Mershon 2005a), not the effects. Though many scholars have sought to diagnose the continuing evolution of the Italian party system (e.g., Bartolini and D’Alimonte 1995; D’Alimonte and Bartolini 1997; 2002), none, to our knowledge, has probed the impact of party switching on party policy.
7.4.1 Data and Measures To evaluate our hypotheses, we need to compare changes in party ideal points over time to see whether and how those changes are related to changes in the ideal points of legislative party members. This means, in turn, that we need estimations of member and party ideal points that are directly comparable without one being derived from the other. We obtain such estimates by calculating two-dimensional Optimal Classification (OC) scores (Poole 2000; 2005) for Italian
Legislator Preferences, Party Desires
183
deputies and their parties on non-secret roll-call votes in the Italian Chamber of Deputies. Our base dataset of voting in the Italian parliament begins in 1988, near the start of the 10th legislature (July 2, 1987–February 2, 1992), and ends in spring 2000, roughly one year before the mandated end of the 13th legislature on March 9, 2001. The more than 2.9 million legislator-vote observations in the base dataset include all final votes on legislation, all votes to convert decree laws, all votes where the relevant government minister and the relevant committee disagree on the advisability of passage (a disagreement observable only on nonfinal votes), and all votes where the outcome goes against the responsible minister’s recommendation (e.g., the vote was in favor but the government was opposed). 5 Each deputy has six possible casts: yea, nay, “voted,”6 abstain, absent, and “on mission” (absent with cause). We excluded all “voted” casts, as they reveal no information about preferences, and coded “on mission” and “absent” as missing data. We kept certain abstentions, as detailed in the following text. With these data, we calculated OC scores separately in each legislature for all deputies who voted at least 10 times in a single party in votes where the losers comprised at least one-half percent of those voting.7 We counted abstentions as nays when two conditions obtained: first, the vote did not pass; and second, the number of abstentions exceeded the difference between yeas and nays—that v v v is, 1 yeat 1 nayt 1 abstaint , where v is the total number of votes at legislative vote t (see Heller and Mershon 2008). For calculating OC scores for parties, we counted a party as voting for a measure when a plurality of a party’s MPs who were present at a vote voted in favor of it; conversely, we counted a party as voting against a measure when a plurality of the party’s MPs present at a vote voted against it. Otherwise, we treated the party on a given vote as missing data, since we could not determine where the party as a whole stood on the proposition. We then treated each party as another legislator for the purpose of calculating comparable OC scores for parties and sitting legislators.8 Our unit of analysis for calculating OC scores was deputy-party for sitting legislators and party-switch period for parties.9 This means that any deputy who changed parties during the course of the legislature in essence became a new person in the dataset, with a new ideal point. We treated each party as a new entity every time its membership changed due to party switching. We also treated parties as new entities if they changed their names.10 We calculated OC scores in two dimensions, using the Communist Party
184
William B. Heller and Carol Mershon
and its hardline successor (RC) for the left anchor and the Lega Nord for the “up” anchor. Each observation in the full dataset is a deputy-party combination for every recorded vote. Our hypotheses focus not on individual deputies, however, but rather on whether and how party positions change when their membership does. We therefore collapsed the dataset by legislature, party, and switchperiod, yielding 31,962 observations over the 4 legislatures in the dataset; collapsing further to a single party observation per switchperiod left a total of 386 observations, which after accounting for observations dropped in the calculation of OC scores and the differencing required to produce our dependent variable (see the following text) reduced further to 179 usable observations (35 in the 10th Legislature, 34 in the 11th, 37 in the 12th, and 73 in the 13th). The ideal points of deputies remain as averages.
7.4.2
Test
Each of the three principal hypotheses calls for a specific kind of direct test. The second hypothesis suggests that party ideal points should mirror the ideal point of whoever occupies some particular spatial position in the distribution of party members—for example, the two-third members. To test this, we need to identify whether parties’ OC scores closely track the scores of members in specific positions (such as the median) relative to their copartisans. This direct test is conceptually simple but computationally difficult, so we leave it for future work. The third hypothesis, that party leaders manage party ideal points, also is conceptually simple. A direct test of this hypothesis consists of identifying party leaders, comparing their ideal points with those of their parties, and checking whether changes (if any) in party ideal points make party leaders better or worse off. We face information shortfalls, however: Available data for the Italian Chamber sometimes identify more than one parliamentary group leader per group, and we lack full data for the tumultuous 11th Legislature, discussed in the following text. There is, moreover, the question of how frequently party leaders vote. In the United States, for example, the Speaker of the House (and leader of the majority) normally votes only to break ties. Preliminary analysis suggests that party group leaders in the Italian Chamber do not vote as often as their rank and file (cf. Camera dei Deputati 1987; Nocifero and
Legislator Preferences, Party Desires
185
Valdini 1992, 24–34). We therefore leave direct testing of our third hypothesis for future work as well. This leaves the first hypothesis that party ideal points should respond to changes in party membership. Note that this hypothesis is not entirely incompatible with either the second or the third hypotheses. Both the first and second hypotheses envision party ideal point sensitivity to shifts in party membership: in the first hypothesis, the party ideal point is simply an aggregation of the ideal points of its members, as established by party decision-making rules, which changes when membership changes in such a way as to affect the mean; in the second hypothesis, the relationship between the position of parties and their members reflects intraparty decision-making processes that both define and privilege pivotal actors (cf. Krehbiel 1998), whose identities can change when the membership of the legislative party changes. Although changes that would be observable if H1 is correct might not register in the context of H2, and vice versa, many changes in party membership and position that would be consistent with both hypotheses. Similarly, the third hypothesis envisions party positions as unchanging as long as leaders’ identities and the preferences of party members able to check leaders’ decisions are stable. Inasmuch as switches can affect the preferences of those who can check leaders, by affecting members’ aggregate preferences (as defined by party decisionmaking processes), in line with the first hypothesis, or by affecting the identity of pivotal members, in line with the second hypothesis, they can open the door to leader-instigated changes in party position. Thus, while we focus primarily on the first hypothesis, our analysis should provide insight, albeit indirect, on the second and third hypotheses. To test the first hypothesis, we need to see whether and how a party’s ideal point changes when it gains or loses members. To this end, we set our dependent variable as the change in party ideal point across switchperiods: ∆PartyOC it = PartyOC it – PartyOCit–1, where PartyOCit is party i’s first-dimension OC score at vote t. Our key independent variables are two: (1) inswitchedit , the difference between the mean positions of MPs who switched into party i in switchperiod t and of continuing members (inswitchedit = inswitchersOCit – continuersOCit–1); (2) and outswitchedit , the difference between the mean positions of MPs who switched out of party i in switchperiod t and of continuing members (outswitchedit = outswitchersOCit – continuersOCit–1). In both formulations, the calculated variable is
186
William B. Heller and Carol Mershon
weighted by the relevant subgroup’s weight vis-à-vis the entire legislative party. The underlying assumption is that party members’ ideal points constitute constraints on party ideal points, and adding or subtracting members from the overall legislative membership changes the boundaries of the constraint. Our first hypothesis leads us to expect a positive coefficient for inswitchedit , as new members move the party’s position toward their own (inswitchers to the left of the party’s ideal point would move the party to the left, and inswitchers to the party’s right would move the party to the right). For outswitchedit , by contrast, we expect a negative coefficient as the party’s center of gravity moves away from the ideal points of former members. Coefficients statistically equivalent to zero would constitute support for the null hypothesis. Table 7.1 summarizes our expectations. Our expectations under hypotheses 2 and 3 are not as crisp as those under H1 and the null, since, again, we lack data to test the former directly. For the second hypothesis, in general—looking at a relatively large number of switches and assuming that most switches are not balanced—we expect results for the effect of switching that are substantively identical to those under H1. The third hypothesis yields expectations for coefficients that are conditional both on how much authority party leaders have to determine party positions and on where parties are relative to where their leaders want them to be. Even if we assume that party leaders can set party ideal points, we suspect that they can do so only up to a point. As agents for party membership, leaders are constrained by their principals, who also are their followers. A leader should therefore be willing to accept only those inswitchers who help to move the party closer to where the leader wants (as suggested in H3a) or, particularly if the party already is where the leader wants Table 7.1
Expectations: Party positions as a function of party-member positions
Dependent variable: ptyOCit Coefficients H1
H2
H3
H4 (null)
inswitchedit
positive negative
positive (but small) or zero negative
zero
outswitchedit
positive (unless balanced) negative (unless balanced)
zero
Legislator Preferences, Party Desires
187
it, only those who do not influence the party ideal point at all (as in H3c). This yields an alternative expectation for inswitchers, that is, the effect of inswitching on party ideal points should be null or very small. The effect of outswitching, in line with H3b, should be stronger. From this perspective, switching out of a party is not simply the mirror image of switching in. Every MP who leaves the party is in essence one less constraint on the leader. Indeed, party leaders might want to encourage some outswitching to reduce the number or severity of constraints they face. Thus, not only would we not expect to see balanced outswitching,11 but we would expect party ideal points to respond much more strongly to outswitching than to inswitching. Crucially, our test of H1 also tests these expectations. We test our first hypothesis with simple OLS regressions, one for each of the four legislatures included in our data. We do not pool data across legislatures, in part because of differences in switching across terms—switching in the 10th and 11th Legislatures largely entailed party splits generating simultaneous moves, while in later terms, solo and (near-) simultaneous moves both were common—and in part because the 1993 electoral reform constitutes a fundamental change in the institutional context of switching. The results of the estimations are shown in Table 7.2. Overall, the results in Table 7.2 support the basic contention of our first hypothesis that party positions are defined by their
Table 7.2
Party positions as a function of party-member positions
Dependent variable: ∆ptyOC it Coefficients by legislature (standard errors in parentheses)
inswitchedit outswitchedit constant Observations R2 Root MSE
10
11
12
13
–0.617 (3.851) 0.407 (0.600) –0.008 (0.025) 35 0.0034 0.15596
22.441 (24.656) –103.651** (6.326) 0.011 (0.018) 34 0.2992 0.16788
10.924* (5.647) –12.029** (0.661) 0.011 (0.027) 37 0.3153 0.15744
8.622 (7.671) –11.123** (0.494) 0.000 (0.017) 73 0.1261 0.1489
Note: * = p < 0.05; ** = p < 0.01 (one-tailed test).
188
William B. Heller and Carol Mershon
legislative members. Our hypothesis predicts positive and negative coefficients for inswitched and outswitched, respectively. As can be seen in Table 7.2, both the coefficients have the expected signs in the 11th through the 13th legislatures, although the coefficient for inswitched is significant at acceptable levels only in the 12th legislature. The coefficient for outswitched is both negative and significant, as expected, in the 11th through the 13th legislatures. For the 10th Legislature, the coefficients of interest are statistically and substantively insignificant, and both are signed in the wrong direction. At first blush, and leaving aside for the moment the 10th legislature, the findings in Table 7.2 both support and undermine our first hypothesis. The results for outswitched are fully consistent with predictions, again except for the 10th legislature; the results for inswitched, by contrast, appear fairly weak. The results for inswitched take on a different meaning when viewed through the lens of our third hypothesis, however: one of the implications of the third hypothesis (and H3c in particular) is that party leaders might welcome inswitchers but at the same time ensure that inswitchers do not add constraints to leadership influence over party positioning. By this reasoning, as suggested in Table 7.1, inswitchers’ influence over party positions should be minimal at best. Our results thus support both the contention that party positions are products of intraparty processes for aggregating the preferences of their legislative members (as in H1) and the claim that party positions are set by party leaders exploiting their agenda-setting advantage to direct those same processes (as in H3). Honing in on estimations for individual terms, and still deferring discussion of the anomalous 10th, two distinctions emerge. First, the 11th Legislature (1992–1994) stands apart in that the coefficients for both variables of interest are quite a bit larger than in the 12th or the 13th legislatures (1994–1996 and 1996– 2001, respectively). Second, once more, the 12th Legislature is distinctive for its significant coefficient on inswitched. The 11th Legislature was engulfed by the vast corruption scandal that broke open in Italy in 1992. Fully one quarter of the MPs elected in 1992 were under official investigation for abuse of office by late 1993, and the bulk of the MPs embroiled in the scandal came from Italy’s longstanding governing parties (Ricolfi 1993). Reflecting these pressures, the party that had ruled Italy continuously since 1946, also the largest party since 1946, the
Legislator Preferences, Party Desires
189
Christian Democrats (DC), split in early 1994. Most switching in the 11th Legislature thus owed to the dissolution of the DC: 145 MPs who switched into a parliamentary group other than the gruppo misto (and for whom we have OC scores) moved out of the DC and into one or the other of the newly formed Christian Democratic Center (CCD, 13) or Popular Party (DC-PPI, 132). Under the pressures of scandal, too, parliament approved electoral reform in August 1993. We think it quite reasonable that party switching and intraparty decision making in the 12th Legislature rendered inswitchers’ impact on party positions more than minimal. The elections to the 12th Legislature were the first under the hybrid electoral laws introduced in 1993. The largest party in this term, Forza Italia (FI), was founded only two months before the 1994 elections by a (former) nonpolitician who after the elections became prime minister. Other political entrepreneurs created new parties, also responding to the new electoral rules, and established parties welcomed new recruits. Hence, an extraordinary 70.8 percent of MPs took their seats for the first time in 1994 (Verzichelli 1996, 141). The upheaval in the Italian party system was so profound that some analysts pointed to the inauguration of Italy’s “Second Republic,” although the 1948 Constitution was still in place (e.g., D’Alimonte and Bartolini 2002; Ricolfi 1993). It makes sense that much switching in 1994–1996 reflected MPs’ learning and adjustment to the new rules and what qualified as, in essence, a new party system. It makes sense as well that in this uncertain context party leadership would be especially likely to allow party positions to accommodate inswitchers’ ideals. Amid the continuing evolution of the Italian party system, the findings for the 13th Legislature comport with the first hypothesis as shaded by the third hypothesis, whereby intraparty preference aggregation leads a party away from outswitchers’ ideals, and leadership influence limits the impact of inswitchers. The 1996 elections reaffirmed the trend to bipolar competition under the hybrid laws and tilted the balance of power in favor of the center-left bloc, so that the first center-left coalition took office in post–World War II Italy and post-Communists remained in the executive throughout the 1996–2001 term. Switching abounded during this term, as multiple party fissions and startups occurred and a total of 277 MPs changed affiliation, the largest number of switchers ever registered per term in the Italian Republic. Even under these conditions, again, our reasoning holds.
190
William B. Heller and Carol Mershon
It is time to take up the anomalous 10th Legislature. Switching in this term was dominated by the rupture of Italy’s second largest party, the perennial opposition Italian Communist Party (PCI). In 1991 the PCI split in two, with the majority forming the Party of the Democratic Left (PDS, 154 MPs, of whom we have OC scores for 144) and the dissidents, Communist Refounding (RC, 14). After the initial split, there were very few switches involving the former PCI, with a few exits from the PDS and one set of entries for the RC. In the face of the historic exogenous shock of the collapse of communist regimes, outswitchers terminated the PCI even though the mean position of that party lay relatively close to their own ideals. Inswitchers into the PDS and RC accepted that their new parties’ positions brought them away from their own ideals as once defined (and the technical requirements of calculating OC scores [see note 7] meant that an observation for the RC was dropped). Although almost all of the switchers in the 10th Legislature served in the PCI and its successors, a few other MPs of varying stripes also switched in the 10th, and their repeated changes of affiliation generated fully 20 party-per-switchperiod observations—more than half of the 35 observations in the estimation for this term. Of the 20 observations, 12 pertained to 3 parliamentary groups with low policy coherence: the Radicals and the Greens, which each tolerated internal variation on the traditional left-right dimension and lacked definition on the second, centralist-regionalist, dimension; and the motley Mixed Group, which made no effort to aggregate the preferences of its members. With the end of the cold war prompting the bulk of switching among individual MPs, the behavior of a few MPs feeding more than one-half of the party-per-switchperiod observations, and one-third of those observations represented by ideologically loose groups, it is understandable that our hypotheses fall flat for the 10th Legislature. Yet overall our logic accounts quite well for the impact of party switching on party positions in what amount to fundamentally distinct legislative party systems. Except for the 10th Legislature, our data strongly support a nuanced interpretation of our first hypothesis. Not only are the signs and relative magnitudes of the principal coefficients of interest “correct,” but we find strong circumstantial support for the third hypothesis. These effects tell us that a member of parliament who moves into a new party probably cannot expect to be able to pull that party’s ideal point toward his or her own. Legislators who do move, however, probably do so under pressure, as the exits
Legislator Preferences, Party Desires
191
appear to allow parties to readjust their positions much more than they otherwise might. This suggests that much switching stems not so much from MPs seeking to improve their lots as from MPs fleeing inauspicious situations (cf. Aldrich and Bianco 1992).
7.5 Conclusions The upheaval and transformation of Italian politics during the span we study present a stiff challenge to our hypotheses. Our premise is to take switching seriously. We seek to discover whether a general argument can hold even as conditions vary across legislative terms and even under extreme conditions, and can still illuminate politics in a given place. It is remarkable how well our reasoning identifies the effects of party switching in what amount to different party systems operating under different electoral laws. True, we see anomalies in the 10th Legislature and, along with confirmatory evidence, some distinctions across other terms. Yet the favorable findings overall carry weight precisely because Italy is a hard case. We understand party switching as an aspect of strategic interaction among legislators. It is a particularly important one, for the choice of party affiliation shapes a legislator’s ability to affect policy, the people with whom he or she interacts day by day, his or her roles and appointments within the legislature, and likely his or her career trajectory. Switching is just one part of a larger dynamic, so that what often appears to be unchanging—legislative party systems, party ideologies, party membership—in fact or potentially is in constant flux. As we conceive of it, this “larger dynamic” is essentially a fourstage game like that sketched out in chapter 2. On the face of it, the one hypothesis that we test in this chapter skips the first and second stages of that game, suggesting that parties are little more than black boxes. Considered more carefully, however, the assertion that a party’s legislative behavior responds to the desires of its members suggests that even the most reelection-focused deputy (cf. Müller and Strøm 1999; Strøm 1990) to some extent contributes to party policy making. This should be heartening for those who see in their elected representatives no shred of policy interest—because interested or not, every party member plays a part in determining his or her party’s position, whether to keep it where it is or move it one way or another. In this sense, legislative parties aggregate policy democratically, and present democratically constituted policy packages to voters.
192
William B. Heller and Carol Mershon
7.6 Appendix Table 7.3
Changes in the size of party groups, X and XI Legislatures, Italy
Party Group DP/DP-COM RC PCI/PDS SI Network Greens FE-Radicals PSI PSDI PRI PPI DC CCD Lega PLI MSI Mistoª
X Legislature (1987–1992)
XI Legislature (1992–1994)
N MPs Start 8 – 156 20 – 13 13 94 17 20 – 233 – [1ª] 11 35 7
N MPs Start – 35 106 – 12 16 7 92 16 27 0 205 0 55 17 34 6
N MPs End 11 – 148 19 – 16 8 100 12 20 – 233 – [1ª] 11 34 16
N MPs End – 33 105 – 12 16 6 90 15 26 179 0 24 50 17 34 21
Note: Counts based on authors’ voting dataset. Cross-checks performed against D’Alimonte and Bartolini 2002; Pasquino 1996; Verzichelli 1996. For each legislature depicted, all groups but the Misto are listed top to bottom in the left-right order commonly accepted by Italianists. Dashes indicate the complete absence of a group during the legislature. Zeros show that a group at some point existed during the legislature; it either was created or collapsed during the term. ª For the X and XI Legislatures only, voting data track organized components within Mixed Group; thus, for example, the Lombard League’s single MP in the X Legislature is tracked within the Misto. For the entire period here, Chamber rules (Article 14) stated that any group under 20 MPs must dissolve and enter the Misto, save special circumstances; the Chamber leadership began to apply the rules strictly only in the XII Legislature.
Table 7.4
Changes in the size of party groups, XII and XIII Legislatures, Italy
Party Group RC Prog-PDS Dem PPI PD
XII Legislature (1994–1996)
XIII Legislature (1996–2001)
N MPs Start 39 164 0 33 –
N MPs Start 34 171 – – 66
N MPs End 24 166 21 26 –
N MPs 2000 20 162 – – 57 Continued
Legislator Preferences, Party Desires Table 7.4
193
Continued
DemU RI FLD UDEur CCD FI Lega AN-MSI Mistoª
– – 0 – 27 112 114 109 31
– – 29 – 40 110 75 107 29
0 26 – 0 30 122 59 92 27
21 0 – 21 0 110 46 89 98
Note: Counts based on authors’ voting dataset. Cross-checks performed against D’Alimonte and Bartolini 2002; Pasquino 1996; Verzichelli 1996. For each legislature depicted, all groups but the Misto are listed top to bottom in the left-right order commonly accepted by Italianists. Dashes indicate the complete absence of a group during the legislature. Zeros show that a group at some point existed during the legislature; it either was created or collapsed during the term. ª For the X and XI Legislatures only, voting data track organized components within Mixed Group; thus, for example, the Lombard League’s single MP in the X Legislature is tracked within the Misto. For the entire period here, Chamber rules (Article 14) stated that any group under 20 MPs must dissolve and enter the Misto, save special circumstances; the Chamber leadership began to apply the rules strictly only in the XII Legislature.
7.6.1 AN CCD DC Dem DemU DP DP-COM FE FI FLD Lega Misto MSI PCI PD PDS/DS PLI PPI PRI Prog-PDS
Key to Group Acronyms (alphabetical)
National Alliance (reformed Neo-Fascists) Christian Democratic Center Christian Democrats Democrats Democrats-Olive Tree Proletarian Democracy Proletarian Democracy-Communists European Federalists (Radicals) Forza Italy (Go, Italy) Liberal-Democratic Federation Northern League (name changes observed) Mixed Group Neo-Fascists Communists Popular Democrats Democratic Left (reformed Communists) Liberals Popular Party Republicans Progressive Alliance-Democratic Left
194
William B. Heller and Carol Mershon
PSDI PSI RC RI SI UDEur
7.7
Social Democrats Socialists Communist Refounding Italian Renewal Independent Left (elected on PCI party list) Democratic Union for Europe
Notes
1. “Raw” preferences are preferences untinged by strategic or practical considerations. 2. If minority-party member r were to switch into the majority, but majority member a were to switch out—assuming legislators are uniformly distributed—rank-and-file preferences in the majority party would shift to the right, but the range of preferences would be unchanged. 3. We assume that voters pay attention to parties in legislatures. Given that parties compete for votes, it seems reasonable to suppose that a party that failed to maintain unity would see that lack advertised to voters by its rivals. 4. Note that if issues are multidimensional, the conditions for a median (core) are highly constrained (Plott 1967) and, when no multidimensional core exists, the kinds of processes that privilege a median voter in one dimension can lead to undesirable outcomes (Saari and Sieberg 2001; see also, Schofield, Grofman, and Feld 1988). 5. Data selection criteria were determined in the Camera dei deputati’s servizio parlamentare. We were unable to obtain data on nonfinal votes where government and the relevant committee recommendations agreed and the outcome jibed with the government’s stated preferences. 6. Before October 13, 1988, most votes were secret. Since 1988, secret voting is allowed only on votes concerning individual deputies and a limited range of other issues (Regolamento, art. 49). 7. OC scores locate MPs in policy space by assuming that MP votes reveal sincere preferences and summarizing MP votes (Poole 2000; 2005; cf. Benoit and Laver 2006; Gabel and Huber 2000; Spirling and McLean 2007). A minimum number of votes is required because the reliability of estimates of revealed preferences declines with the number of observed votes. In the same vein, there is little information about preferences to be had from unanimous or near-unanimous votes. The 10-vote minimum resulted in the dropping of some 24,662 observations from the original file. 8. Thanks to Mat McCubbins for suggesting this method of obtaining party scores. Thanks also to Keith Poole for providing the program for calculating OC scores and Adriana Prata for walking us through it. For a detailed discussion, see Heller and Mershon 2008. 9. Switchperiod is a party-specific counter that increases by 1 for party i in every vote that at least one deputy votes as a new member of i or at least one deputy who was in i votes in party j≠i.
Legislator Preferences, Party Desires
195
10. Party name changes sometimes might reflect changes that should not be expected to affect party positions. To the extent that this is the case, our data are biased against finding evidence of systematic changes in party preferences in line with our argument. 11. We should not expect to see balanced outswitching in equilibrium. Under extraordinary circumstances or when a party is dissolving, of course, all bets are off.
7.8 References Àgh, Attila. 1999. “The Parliamentarization of the East Central European Parties: Party Discipline in the Hungarian Parliament, 1990–1996.” In Party Discipline and Parliamentary Government, ed. Shaun Bowler, David M. Farrell, and Richard S. Katz, 167–188. Columbus: Ohio State University Press. Aldrich, John H. 1995. Why Parties? The Origin and Transformation of Political Parties in America. Edited by Benjamin I. Page. Chicago: University of Chicago Press. Aldrich, John H., and David W. Rohde. 2000. “The Logic of Conditional Party Government: Revisiting the Electoral Connection.” In Congress Reconsidered, ed. Lawrence C. Dodd, and Bruce I. Oppenheimer, 269–292. Washington, DC: CQ Press. Aldrich, John H., and William T. Bianco. 1992. “A Game-Theoretic Model of Party Affiliation of Candidates and Office Holders.” Mathematical and Computer Modelling 16 (8/9): 103–116. Alesina, Alberto, and Alex Cukierman. 1990. “The Politics of Ambiguity.” Quarterly Journal of Economics 105 (4): 829–850. Bartolini, Stefano, and Roberto D’Alimonte, eds. 1995. Maggioritario ma non troppo. Bologna: Il Mulino. Benoit, Kenneth, and Michael Laver. 2006. Party Policy in Modern Democracies. London: Routledge. Best, Robin, and William B. Heller. 2005. “Safety in Numbers? Seat Shares and Discipline in Legislative Parties.” Paper presented at ECPR Joint Sessions of Workshops, Granada, Spain, April 14–19. Black, Duncan. 1958. The Theory of Committees and Elections. Cambridge: Cambridge University Press. Bowler, Shaun, David M. Farrell, and Richard Katz. 1999. “Party Cohesion, Party Discipline, and Parliaments.” In Party Discipline and Parliamentary Government, ed. Shaun Bowler, David M. Farrell, and Richard S. Katz, 3–22. Columbus: Ohio State University Press. Budge, Ian, Hans-Dieter Klingemann, Andrea Volkens, Judith Bara, and Eric Tannenbaum. 2001. Mapping Policy Preferences. Estimates for Parties, Electors and Governments, 1945–1998. Oxford: Oxford University Press. Calvert, Randall L. 1987. “Reputation and Legislative Leadership.” Public Choice 55 (1–2): 81–119. Camera dei Deputati. 1987. I deputati e senatori del decimo Parlamento repubblicano. Rome: La Navicella.
196
William B. Heller and Carol Mershon
Cooper, Joseph, David W. Brady, and Patricia A. Hurley. 1977. “The Electoral Basis of Party Voting: Patterns and Trends in the U.S. House of Representatives, 1887–1969.” In The Impact of the Electoral Process, ed. Louis Maisel and Joseph Cooper, 135–167. Beverly Hills, CA: Sage. Cox, Gary W. 1987. The Efficient Secret: The Cabinet and the Development of Political Parties in Victorian England. Cambridge: Cambridge University Press. ———. 1997. Making Votes Count: Strategic Coordination in the World’s Electoral Systems. Cambridge: Cambridge University Press. Cox, Gary W., and Mathew D. McCubbins. 1993. Legislative Leviathan: Party Government in the House. Berkeley: University of California Press. ———. 2005. Setting the Agenda: Responsible Party Government in the U.S. House of Representatives. Cambridge: Cambridge University Press. D’Alimonte, Roberto, and Stefano Bartolini, eds. 1997. Maggioritario per caso. Bologna: Il Mulino. ———. 2002. Maggioritario finalmente? La transizione elettorale 1994–2001. Bologna: Il Mulino. Desposato, Scott W. 2006. “Parties for Rent? Ambition, Ideology, and Party Switching in Brazil’s Chamber of Deputies.” American Journal of Political Science 50 (1): 62–80. Duverger, Maurice. 1972. Party Politics and Pressure Groups. New York: Crowell. Epstein, Leon D. 1967. Political Parties in Western Democracies. New York: Praeger. Gabel, Matthew J., and John Huber. 2000. “Putting Parties in their Place.” American Journal of Political Science 44 (1): 94–103. Heller, William B., and Carol Mershon. 2005a. “Party Switching in the Italian Chamber of Deputies, 1996–2001.” Journal of Politics 67 (2): 536–559. ———. 2005b. “Switch or Stick? Formal and Empirical Perspectives on Legislative Party Switching.” Paper presented at the annual meetings of the American Political Science Association, Washington, DC, September 1–4. Heller, William B., and Carol Mershon. 2008. “Dealing in Discipline: Party Switching and Legislative Voting in the Italian Chamber of Deputies, 1988– 2000.” American Journal of Political Science 52 (4): 910–925. Hirschman, Albert O. 1970. Exit, Voice and Loyalty. Cambridge, MA: Harvard University Press. Katz, Richard S., and Peter Mair. 1995. “Changing Models of Party Organization and Party Democracy: The Emergence of the Cartel Party.” Party Politics 1 (1): 5–28. ———, eds. 1994. How Parties Organize: Change and Adaptation in Party Organizations in Western Democracies. London, Thousand Oaks, and New Delhi: Sage. Kiewiet, D. Roderick, and Mathew D. McCubbins. 1991. The Logic of Delegation: Congressional Parties and the Appropriations Process. Chicago: University of Chicago Press. Kirchheimer, Otto. 1966. “The Transformation of Western European Party Systems.” In Political Parties and Political Development, ed. Joseph
Legislator Preferences, Party Desires
197
LaPalombara and Myron Weiner, 177–200. Princeton, NJ: Princeton University Press. Krehbiel, Keith. 1993. “Where’s the Party?” British Journal of Political Science 23: 235–266. ———. 1998. Pivotal Politics: A Theory of US Lawmaking. Chicago: University of Chicago Press. ———. 1999a. “Paradoxes of Parties in Congress.” Legislative Studies Quarterly 24 (1): 31–64. ———. 1999b. “The Party Effect from A to Z and Beyond.” Journal of Politics 61 (3): 832–840. ———. 2000. “Party Discipline and Measures of Partisanship.” American Journal of Political Science 44 (2): 212–227. Laver, Michael. 1999. “Divided Parties, Divided Government.” Legislative Studies Quarterly 224 (1): 5–29. Laver, Michael, and Kenneth Benoit. 2003. “The Evolution of Party Systems between Elections.” American Journal of Political Science 47 (2): 215–233. Laver, Michael, and Kenneth A. Shepsle. 1990. “Government Coalitions and Intraparty Politics.” British Journal of Political Science 20: 489–507. ———. 1999. “How Political Parties Emerged from the Primeval Slime: Party Cohesion, Party Discipline, and the Formation of Governments.” In Party Discipline and Parliamentary Government, ed. Shaun Bowler, David M. Farrell, and Richard S. Katz, 23–48. Columbus: Ohio State University Press. Laver, Michael, and Norman Schofield. 1990. Multiparty Government: The Politics of Coalition in Europe. Oxford: Oxford University Press. Marsh, Michael, and Paul Mitchell. 1999. “Office, Votes, and then Policy: Hard Choices for Political Parties in the Republic of Ireland, 1981–1992.” In Policy, Office, or Votes: How Political Parties in Western Europe Make Hard Decisions, ed. Kaare Strøm and Wolfgang C. Müller, 36–62. Cambridge: Cambridge University Press. McElroy, Gail M. 2003. “Party Switching in the European Parliament: Why Bother?” Paper presented at the annual meetings of the Midwest Political Science Association, Palmer House Hilton, Chicago, April 3–6. Mejía Acosta, Andrés. 1999. “Indisciplina y deslealtad en el congreso ecuatoriano.” Iconos (6): 13–21. Mershon, Carol, and William B. Heller. 2003. “Party Switching and Political Careers in the Spanish Congress of Deputies, 1982–1996.” Paper presented at the annual meetings of the Midwest Political Science Association, Palmer House Hilton, Chicago, April 3–6. Müller, Wolfgang C., and Kaare Strøm, eds. 1999. Policy, Office, or Votes: How Political Parties in Western Europe Make Hard Decisions. Cambridge: Cambridge University Press. Nocifero, Niccolò, and Sergio Valdini. 1992. Il palazzo di vetro. Il lavoro dei deputati italiani nella decima legislatura. Florence: Vallecchi Editore. Ordeshook, Peter C., and Olga V. Shvetsova. 1994. “Ethnic Heterogeneity, District Magnitude, and the Number of Parties.” American Journal of Political Science 38 (1): 100–123.
198
William B. Heller and Carol Mershon
Pasquino, Gianfranco. 1996. 1945–1996. Archivio della politica in Italia [CD-ROM]. Laterza Multimedia. Plott, Charles. 1967. “A Notion of Equilibrium and Its Possibility under Majority Rule.” American Economic Review 57: 787–806. Poole, Keith T. 2000. “Nonparametric Unfolding of Binary Choice Data.” Political Analysis 8 (2): 211–237. ———. 2005. Spatial Models of Parliamentary Voting. Cambridge: Cambridge University Press. Reed, Steven R., and Michael F. Thies. 2000. “The Consequences of Electoral Reform in Japan.” In Mixed-Member Electoral Systems: The Best of Both Worlds? ed. Matthew Soberg Shugart and Martin P. Wattenberg, 380–403. New York: Oxford University Press. Ricolfi, Luca. 1993. L’ultimo Parlamento. Sulla fine della prima Repubblica. Rome: La Nuova Italia Scientifica. Riker, William H. 1982. Liberalism against Populism: A Confrontation between the Theory of Democracy and the Theory of Social Choice. Prospect Heights, IL: Waveland Press. Rohde, David W. 1991. Parties and Leaders in the Postreform House. Edited by Benjamin I. Page. Chicago: University of Chicago Press. Saari, Donald G., and Katri Sieberg. 2001. “The Sum of the Parts Can Violate the Whole.” American Political Science Review 95 (2): 415–433. Sánchez de Dios, Manuel. 1999. “Parliamentary Party Discipline in Spain.” In Party Discipline and Parliamentary Government, ed. Shaun Bowler, David M. Farrell, and Richard S. Katz, 141–162. Columbus: Ohio State University Press. Scarrow, Susan E. 2000. “Parties without Members? Party Organization in a Changing Electoral Environment.” In Parties without Partisans: Political Change in Advanced Industrial Democracies, ed. Russell J. Dalton and Martin P. Wattenberg, 79–101. Oxford: Oxford University Press. Schlesinger, Joseph A. 1966. Ambition and Politics: Politics in the United States. Ann Arbor: University of Michigan Press. Schofield, Norman, Bernard Grofman, and Scott Feld. 1988. “The Core and the Stability of Group Choice in Spatial Voting Games.” American Political Science Review 82 (1): 195–211. Seyd, Patrick. 1998. “Tony Blair and New Labour.” In New Labour Triumphs: Britain at the Polls, ed. Anthony King, 49–73. Chatham, NJ: Chatham House. Shepsle, Kenneth A. 1972. “The Strategy of Ambiguity: Uncertainty and Electoral Competition.” American Political Science Review 66 (2): 555–568. Shepsle, Kenneth A., and Mark S. Bonchek. 1997. Analyzing Politics: Rationality, Behavior, and Institutions. New York and London: Norton. Spirling, Arthur, and Iain McLean. 2007. “UK OC OK? Interpreting Optimal Classification Scores for the U.K. House of Commons.” Political Analysis 15 (1): 85–96. Strøm, Kaare. 1990. “A Behavioral Theory of Competitive Political Parties.” American Journal of Political Science 34: 565–598. ———. 1994. “The Presthus Debacle: Intraparty Politics and Bargaining Failure in Norway.” American Political Science Review 88 (1): 112–127.
Legislator Preferences, Party Desires
199
Svåsand, Lars, Kaare Strøm, and Bjørn Erik Rasch. 1997. “Change and Adaptation in Party Organization.” In Challenges to Political Parties: The Case of Norway, ed. Kaare Strøm and Lars Svåsand, 91–123. Ann Arbor: University of Michigan Press. Tomás Mallén, Beatriz. 2002. Transfuguismo parlamentario y democracia de partidos. Madrid: Centro de Estudios Políticos y Constitucionales. Tsebelis, George. 1999. “Veto Players and Law Production in Parliamentary Democracies: An Empirical Analysis.” American Political Science Review 93 (3): 591–608. ———. 2000. “Veto Players and Institutional Analysis.” Governance 13 (4): 441–474. ———. 2002. Veto Players: How Political Institutions Work. New York and Princeton, NJ: Russel Sage Foundation and Princeton University Press. Turan, Ilter. 1985. “Changing Horses in Midstream: Party Changers in the Turkish National Assembly.” Legislative Studies Quarterly 10 (1): 21–34. Verzichelli, Luca. 1996. “I gruppi parlamentari dopo il 1994. Fluidità e riaggregazioni.” Rivista italiana di scienza politica 26 (2): 391–413. Zielinski, Jakub, Kazimierz M. Slomczynski, and Goldie Shabad. 2005. “Electoral Control in New Democracies: The Perverse Incentives of Fluid Party Systems.” World Politics 57 (3): 365–395.
This page intentionally left blank
8 Timing Matters: Incentives for Party Switching and Stages of Parliamentary Cycles Carol Mershon and Olga Shvetsova
8.1
Introduction
Virtually all political scientists maintain that democratic political life depends on political parties. Precisely because parties are fundamental to democracy, we and the other contributors to this book probe changes of party affiliation among elected politicians. This chapter is distinctive in that it links the phenomenon of party switching to the stage in the parliamentary cycle. Our goal here is two-fold. On one hand, our research is the first to explore whether and how patterns in switching behavior differ systematically across distinct stages of the legislative term—for instance, one stage devoted to committee assignments, early in the term, and another at the end of the term, dominated by the view of elections on the horizon (Mershon and Shvetsova 2005; 2008a; 2008b; cf. Desposato, this volume). On the other hand, linking changes of party affiliation to the period when benefits are available provides direct confirmation for the main theoretical premise shared by all of this volume’s contributors, namely, that politicians switch parties for opportunistic reasons. In a nutshell, we seek to gain leverage on the question of why legislators switch by asking when they switch. We examine politicians’ choices and changes of party labels during the legislative term, in different stages of the parliamentary cycle. As we define it, the parliamentary cycle includes legislative stages and also the electoral stage that occurs before or during the official campaign for the next legislature and that, assuming backward induction, affects behavior in the term. The next section elaborates on
202
Carol Mershon and Olga Shvetsova
this conception of the parliamentary cycle and hypothesizes that the cycle leaves its imprint on switching behavior. Third, we outline our research design. Fourth, we assess our hypotheses against data from our two primary country-terms, the 1996–2001 Italian Chamber of Deputies and the 1993–1995 Russian Duma, and a total of 35 subsidiary country-terms in Italy, Romania, the United Kingdom, and the United States. The last part draws out the implications of our study.
8.2 The Parliamentary Cycle: Stages, Motives, and Behavior Like other contributors to this book, we assume that legislators switch parties in strategic efforts to attain desired goals. A distinguished tradition holds that parties and politicians are motivated by office, policy, and votes (e.g., Müller and Strøm 1999; Strøm 1990). Hence we assume that members of parliament (MPs) may switch partisan affiliation to obtain parliamentary offices and privileges, achieve preferred policy outcomes, or position for reelection. All of these motives may prove relevant to some degree (cf. Strøm 1990) in each decision to change a party, and thus none should be ignored. Yet we posit that their relative salience differs across specific periods of time, depending on what dominates the parliamentary agenda and hence which payoffs are most prominent, immediate, and available (Mershon and Shvetsova 2005; 2008a). In our approach, individuals can change their choice of parties as they pursue goals that are specific to the stage in the parliamentary cycle. The different stages of the parliamentary cycle hold out different mixes of incentives to MPs, make some incentives more prominent at some times than others, and thus highlight different motivations. Positing variations in the salience of incentives and motivations over time, we distinguish types of switching and expect variations in the frequency of switching according to stage in the parliamentary cycle. Figure 8.1 identifies the stages within the parliamentary cycle where we expect distinctive rewards to switching and distinctive motivations for switching. We assume that the period of greatest salience of each motivation can be discerned at the outset, on the basis of legislative records, for example. By associating changes of partisanship with the periods when they occur, and by classifying periods according to the type of rewards decided, we appraise the relative impact of the aims of office (including legislative perks), policy, and (re)election.
Timing Matters Pre-cycle campaign
Stage A: (Affiliation) Reaffirm affiliation with the electoral party
Taking up seats and announcing group membership
Stage B: (Benefits) Determination of committees, government participation, portfolios
Allocating parliamentary/cabinet offices Parties in elections
Campaign
203
Stage C: (Control of policy) Advantage in agenda control
Stage E: (Elections) Coalition formation to ensure personal reelection
Active policymaking
Positioning for advantage in the next election
Figure 8.1 Switching behavior during the parliamentary cycle.
What we call stage A (for Affiliation) marks the transition from the popular vote to the taking up of legislative seats in the first legislative session. In most systems, MPs have the opportunity to announce affiliation with a label different from their electoral label.1 We posit that MPs are motivated primarily by perks during stage A, for they respond to the availability of goods tied to membership in one parliamentary group or another, as shaped by legislative rules. To be sure, policy motives are not absent during this stage, since office serves as an instrument to affect policy (cf. Laver and Shepsle 1996). Stage B (for Benefits) is when committee seats, committee chairs, and other legislative posts (and, in parliamentary systems, executive portfolios) are allocated. We posit that this division of positions of power heightens office-seeking goals for MPs. Again, policy does not disappear as a concern, for office facilitates the pursuit of policy. But when offices are up for grabs, office is the proximate motivation for MPs’ choices to switch or stay with their original party. In stage C (for Policy Control), the legislative agenda focuses most heavily on policy domains relevant to a broad range of issues. Whereas
204
Carol Mershon and Olga Shvetsova
some analysts, mindful of the executive’s role in policy making, might paint an entire term as akin to a continuous stage C to capture the parliament’s overall contribution to policy decisions, we see a noncontinuous sequence of shorter phases of concentrated attention to the most important policy domains. Given MPs’ heightened focus on policy in these phases, switching in stage C should be aimed at shaping policy choices and securing agenda control. Stage E (for Election) closes the cycle in figure 8.1. In this stage, electoral motivations should come to the fore and switching should chiefly aim at preelectoral positioning. Figure 8.1 simplifies reality, and in some institutional contexts stage E may recur during the cycle, as nonparliamentary elections are held during the term. Although their own seats are not immediately at stake, MPs learn from opinion polls, which proliferate before elections, and from voting returns, which reveal whether a switch is likely to improve their reelection chances (cf. Heller and Mershon 2005). One stage, D (for Dormant), does not appear in figure 8.1. Stage D is simply all periods other than stages A, B, C, and E—a residual set of intervals between the active stages. We expect no distinctive switching by politicians driven by office, policy, or reelection—that is, we expect no switching—to occur during D, since it is more advantageous to switch in active stages. All stages except D can overlap or coincide. Given the general premise that MPs are goal-oriented, the general expectation is that they switch when their goals are best served by such action. The expectation that timing matters leads to more specific testable hypotheses. • Hypothesis 1: The rate of switching should vary over the course of the parliamentary cycle. Switching should be relatively frequent in stages A, B, C, and E. • Hypothesis 2: Attributes of switches should vary by stage. Officedriven switching should predominate in stages A and B, policy-driven switching should predominate in stage C, and vote-driven switching should predominate in stage E. Thus, independents should switch in stages A and B. Notables—MPs who by virtue of seniority or other distinction are most able to strongly influence policy—should switch, if they switch, in stage C. MPs who switch to improve their prospects in the next election should move in stage E. • Hypothesis 3: Stage C switches should aim for clear policy effect. Switchers whose moves are identifiable by timing as policy-driven should seek the center of the policy space, grab agenda advantage, or
Timing Matters
205
break the government. Otherwise, their switches would be irrelevant to policy outcomes. • Corollary to Hypothesis 3: New parliamentary groups formed in stage C should locate at the center of the policy space. Assuming policy goals, political entrepreneurs should found new groups in the most advantageous position in policy space: the center (cf., e.g., McKelvey and Schofield 1987; Schofield this volume). • Hypothesis 0: Null. No or very little variation between stage D and the active stages should emerge in rates of switching and attributes of switches.
8.3 Research Design and Methods Hypotheses in hand, we address our strategy for testing them. First, like others in this field, we define a switch as “any recorded change in party affiliation on the part of a politician holding or competing for elective office” (Heller and Mershon, Introduction this volume, 8). Second, we choose 1996–2001 Italy and 1993–1995 Russia as primary country-term cases, thus affording a rare comparison of two very different countries with strikingly similar electoral institutions. For the terms we study in depth, Italy and Russia both used hybrid laws combining proportional representation (PR), thresholds for PR, and plurality in single-member districts (SMDs). Other institutional similarities included powerful subnational governments (though Italy is not federal) and nonconcurrence between parliamentary and other major elections.2 These similarities are essential to discerning the imprint of timing in the parliamentary cycle on patterns of switching, as formal institutions should strongly affect political actors’ choices in partisan competition. Given the institutional parallels, we would be able to reject our hypotheses with confidence if differences in switching behavior appeared (unless some feature of rules differed too). Similarities in switching would equally clearly corroborate our reasoning. Third, we deliberately select for intensive study the legislative term in Italy and in Russia that qualifies as that term with the highest number of switches for any term to date. Specifically, approximately one-fourth of Italian MPs switched parties at least once during the 1996–2001 term, with the total number of switches in the 630-member Chamber standing at 277. Almost one-third of the MPs in the Russian Duma switched at least once from 1993 to 1995, and switches totaled 342 for the 450-member house. Critics might argue that this choice
206
Carol Mershon and Olga Shvetsova
is problematic, since in the first post–Communist Duma legislators might have still been learning to understand and respond to institutional incentives (Kaminski 2002; Kunicová and Remington 2005). One might also imagine that the Italian deputies under scrutiny were still adapting to a new institutional and partisan environment, especially since the share of neophytes in the 1996–2001 Chamber was relatively high (Verzichelli 1996; Zucchini 2001, 172). Yet we see these possibilities as sources of strength in our design: If we find that, even during the relatively uncertain first Duma and 1996–2001 Chamber, switching varies according to type of incentive dominant in distinct stages, then we are likely to find elsewhere that switching varies by stage in the parliamentary cycle. Whereas practical concerns lead us to restrict the in-depth inquiry to strategically chosen terms, we extend the inquiry to 1963–1972 Italy, Romania, the United Kingdom, and the United States so as to introduce greater variation in electoral laws, executive-legislative relations, and territorial dispersion of power. Whereas the United Kingdom and the United States use plurality electoral laws, Romania uses closedlist PR, and Italy before 1993 used open-list PR for elections to the lower house. The two SMD systems differ in regime type, as do the two PR systems. Table 8.1 also shows variations across countries and across time in powers allocated to subnational governments. Moving beyond the parameters highlighted in table 8.1, the five countries present variation in regime longevity, institutionalization of the party system, and rules on legislative term length. To adopt Table 8.1
Case selection: Countries and legislative terms Electoral Laws List PR
Regime Type
Plurality
Open
Parliamentary
United Kingdom 1970–2005 United States 1950–2000
Italy 1963–1972
Presidential Semi-presidential
Closed
Hybrid Italy 1996–2001 Russia 1993–1995
Romania 1996–2004
Note: Federal and regionalized systems appear in italics. Primary cases, where operationalization of stages is most finely grained, appear in boldface. Number of terms studied in subsidiary cases = 2 Italy, 2 Romania, 6 United Kingdom, 25 United States.
Timing Matters
207
Huntington’s (1991) well-known scheme, the United States and United Kingdom belong to the first wave of democratization, and post–fascist Italy to the second wave. Romania and Russia democratized after the third wave had started in the mid-1970s. Reflecting the timing of democratization, the Romanian and Russian party systems are the most weakly institutionalized of those studied. 3 For the spans covered, the United Kingdom and US party systems are the most strongly institutionalized, and the Italian evinces variation, with great institutionalization in 1963–1972 and diminished institutionalization in 1996–2001. Although the United States is the only country here with fixed terms, Romania has not yet seen an early parliamentary dissolution. Finally, central to our approach to demarcating the boundaries of stages in the parliamentary cycle is the idea that the time of greatest salience of each motivation that induces switching can be readily identified according to the parliamentary record, institutional prescriptions, and events forcing decisions on a key policy dimension. Table 8.2 illustrates how we move from the general principle to specific definitions of stages. For our primary country-terms, we identify stage A (Affiliation) as the interval between election day and the last day that MPs have to announce parliamentary group membership; and we locate stage B (Benefits) from the last day of announcement of group membership to the day that the distribution of all legislative offices (and, in the Italian parliamentary system, executive portfolios) is completed. Stage B can cover multiple substages in a single term, involving, for example, multiple cabinets. For primary cases only, we decompose Stage C (Control of policy) into multiple stretches of time so as to isolate the substages of the most active, controversial policy bargaining, focusing on domains with special weight: budgetary questions and, as applicable, security and foreign policy and constitutional matters (cf. Lijphart 1984; Laver and Shepsle 1996).4 We set exogenous criteria for locating the periods of most intense policy bargaining on these dimensions, attending to such steps as the executive’s transmittal of the annual budget bill to the legislature. Operationalization of stage D (for Dormant) is simple: whatever spans do not qualify as either A or B or C or E become, by default, stage D. Approaches to measuring stage E (Elections) can be several. One set of boundaries would be the start of registration for the ballot for the next legislative elections and election day itself. Yet this approach is inadequate for tapping strategic behavior in anticipation of decisions on lists and candidacies.
Table 8.2 Operationalizing stages of the parliamentary cycle Operationalized Stage Stage Concept
Operationalization shaped by: Italy 1996–2001
Russia 1993–1995
Subsidiary
first 2 months, starting election day
as indicated by other stages
election day to day of group selection for SMD MPs; for PR, 5 weeks after rules specified on group selection given strong presidency, no governing coalition formed; period of allocation of legislative committee posts as reflected in legislative records legislative record, from first vote on finance until finance moves off the agenda; on war: as indicated by key events that open and close policy episode as indicated by other stages
E not observed; parliament dissolved (many nonparliamentary elections in 5-year term; 90 days before, 30 days after election day)
day marking 90 days before election to end of ballot registration 30 days before election (no nonparliamentary elections included)
6 months before election day
LEGISLATIVE A
Affiliation: MPs take up seats and announce group affiliation
electoral laws affecting independents; legislative rules on groups and on start of session
election day to last day MPs must state group membership
B
Benefits: Legislative and executive offices are allocated
relative strength of president vs. premier; number of cabinets per legislative term
C
Control of policy: Policy making dominates agenda
rules on introducing, considering, and approving bills, incl. committee role
with >1 cabinets, B recurs; day groups announced to day legislative and first executive payoffs completed; day Nth cabinet falls to day N+1 cabinet named legislative record, from day executive sends budget to house to day bill passed; day constitutional bill presented to day committee dissolved
D
Dormant: definition of other stages All other than A, (this is residual stage) B, C, and E ELECTORAL E Reelection: rules on dissolution of MPs position for legislature before elections; advantage in next rules on registration for legislative election ballot; rules on election (includes nontiming; duration of term parliamentary elections in some contexts)
Note: All stages save D can overlap or coincide.
for committees: first month that legislature in session; for cabinet(s) in parliamentary systems, any month in which cabinet(s) negotiated any month that legislature is officially in session, as shown in legislative records
as indicated by other stages
Timing Matters
209
Given the evident capacity of Italian and Russian MPs to anticipate electoral competition—well in advance of scheduled parliamentary elections—we use a cutoff date to start stage E, specific to practices in each national setting. As the rightmost column of table 8.2 indicates, a key distinction between the primary and subsidiary country-term cases regards the degree of detail in coding stages. For subsidiary cases, we simply code the first two months of the term, starting with election day, as stage A. For subsidiary cases, too, the first month that the legislature is officially in session becomes stage B; in parliamentary systems, B includes any month(s) in which cabinets form and fall. Lacking particulars on types or intensity of legislative activity, we treat any month that the legislature is in session as entering Stage C. The six months before the next national legislative elections comprise the simplified stage E in subsidiary cases. For both primary and subsidiary cases, D is a residual stage. With this operationalization of stages in subsidiary cases, we may underestimate D relative to primary cases and more generally may introduce error, which works against generating findings favorable to our hypotheses. Whatever support does appear for our argument is thus reinforced.
8.4 Empirical Analysis of Switching For the primary cases of the 1996–2001 Italian and 1993–1995 Russian lower houses, we first investigate the frequency of switching by aggregated stage, distinguishing MPs by mode of election, and next disaggregate stages. We then move to the subsidiary terms in Italy, Romania, the United Kingdom, and the United States.
8.4.1 Switching by Aggregated Stage in 1996–2001 Italy and 1993–1995 Russia Table 8.3 compares switching in Russia and Italy by aggregated stage in the parliamentary cycle. We standardize the measure of frequency of switching by examining mean weekly switches per 100 MPs. Note first, as the top row of the table displays, that the opening weeks of the term, when newly elected deputies must choose parliamentary group affiliation (stage A), exhibit the highest aggregate rate of switching in both Italy and Russia. And in both hybrid electoral systems, it is above all SMD MPs who choose a parliamentary group different from their electoral label.
210
Carol Mershon and Olga Shvetsova
Table 8.3 Mean weekly switches per 100 MPs, by type of stage and by MPs’ mode of election, 1996–2001 Italian Chamber and 1993–1995 Russian Duma Italian Chamber
Russian Duma SMD
Aggregate Aggregate Stage N Weeks All MPs SMD PR N weeks All MPs All
Partya Indepb PR
A B C D E
0.90 15.01 0.07 0.23 0.52 0.15 0.52 2.03 0.56 0.10 0.20 0.10 0.51 0.47 0.33
4 9 109 95 942
Key to stages:
0.36 0.35 0.20 0.10 0.21
0.43 0.25 0.18 0.11 0.21
A: Affiliation
0.18 0.67 0.27 0.08 0.27
6 12 26 53 82
B: Benefits
5.56 0.30 1.09 0.14 0.42
11.04 0.44 1.62 0.18 0.50
C: Control of policy
E: Electoral
D: Dormant, that is, all periods other than A, B, C, and E. a Independents are elected as a subset of SMD MPs only in Russia, as discussed in the text. b The only observable electoral stages in Italy pertain to nonparliamentary elections, whereas the only electoral stage in Russia is associated with the campaign for the subsequent parliament. Sources: For dates and numbers of switches: Camera dei Deputati 2008b; Duma Statistical Services (INDEM) 2000. For detailed documentation needed to code stages: Camera dei Deputati 2008b; INDEM 2000; Istituto Cattaneo 2008; Ministero dell’Interno 2008; Pasquino 1999. Tukey test for differences of means (all MPs) across Russian stages: p < .005; Tukey test for differences of means (all MPs) across Italian dormant and active stages (given overlapping of active stages): p < .10.
Yet in stage A the Russian switching rate is more than 15 times the Italian rate among all MPs, and the Russian rate is more than 25 times the Italian among SMD MPs. The startling contrast stems from the behavior of Russian independents, who won roughly two-thirds of Russian SMDs in 1993, reflecting not only the electoral rules and weakly institutionalized party system but also the compressed campaign, a product of Yeltsin’s decision to call the parliamentary elections on short notice and in conjunction with the constitutional referendum. These independents, having earned SMD seats, engaged in party shopping and hopping in the Russian stage A. Although in Italy the ballot structure inhibited independents from entering the Chamber, all mobile SMD MPs in stage A were center-leftists who in the 1996 race mounted broad appeals and who, once in parliament, opted for the motley Mixed Group instead of joining the legislative group of the largest party sponsoring their list (Di Virgilio 1997; 2002).5 When office benefits are allocated (stage B, on the second row), mean weekly switches per 100 MPs are roughly similar for the full
Timing Matters
211
lower house across Russia and Italy. The occurrence of the event of switching in stage B is almost three times higher among SMD legislators in Russia than among PR ones; independents in particular rushed to move, for joining parties made them eligible for internal legislative office. In Italy, the proportion is flipped, with PR MPs switching approximately three times as often as their colleagues. This pattern largely traced to the choices of a subset of PR MPs who split from their parliamentary party so as to compete for seats on the Committee for Constitutional Reform, as detailed in the following text. Switching in the aggregated policy control stage (C) is more than five times more frequent in Russia than in Italy. The rate of switching in proximity to Italian subnational elections (the aggregated stage E) nearly matches that registered in the Italian policy stage (C). Switching in the Russian electoral stage is double that of the Italian E but less than half that of the Russian C. Mean weekly switches per 100 MPs drop to their lowest rate in the full house for Italy and Russia in the aggregated stage D (dormant), after initial affiliations are announced, when neither benefits are allocated nor major questions of policy control dominate the agenda nor elections loom. In both cases, switching in D dies down. By these data, the null hypothesis fails. Overall, table 8.2 allows for a preliminary assessment of our hypotheses. The basic notion that the rate of switching varies by stage (Hypothesis 1) finds support. Among MPs elected on party labels, switching rates per aggregated stage look roughly alike across Russian and Italian MPs regardless of mandate. It is thus the propensity of the Russian independents to jump parliamentary groups that underpins the statistically significant differences observed so far. Indeed, the fact that Russian independents switch often in stages A and B aligns with Hypothesis 2. Yet Russian independents are also prone to switch in the aggregated policy stage C. Further inspection of the first two hypotheses calls for scrutiny of substages in these primary cases, just as evaluation of Hypothesis 3 requires isolating the most intense substages of policy bargaining.
8.4.2 Switching Disaggregated by Substage Figure 8.2 portrays mean weekly switching rates for all MPs and SMD MPs in each of the substages, ordered chronologically, in the 1993–1995 Duma and the 1996–2001 Chamber. The behavior of PR MPs, as the remaining category, can be imputed. The error bars show
212
Carol Mershon and Olga Shvetsova
standard errors in mean weekly switches per 100 MPs for each substage, and substages with relatively high rates are tagged. Looking first at figure 8.2a, the frequency of switching among Russian SMD MPs in the Duma’s first affiliation stage (stage A.1) is an obvious outlier. This is, by a large margin, the highest rate of switching in any substage for either subgroup of deputies (SMD or
a
Mean Weekly Switches per 100 MPs
A.1 off chart*
10 C.3 8 6 *A.1 is outlier: mean/100 all MPs = 27.8 mean/100 SMDs = 55.6 Vertical axis truncated so as to clarify scale for other stages.
4 A.2
C.2
C.4
E
2 0 0
mean/100 all MPs mean/100 SMDs 4 8 12 16 Substages in Russian Term, Ordered Chronologically
Mean Weekly Switches per 100 MPs
b 2.5 B.3
C.6
2.0 D.5
C.8
1.5 1.0
E.3 A
E.4 D.6
0.5 0.0
mean/100 all MPs mean/100 SMDs 0
8 16 24 Substages in Italian Term, Ordered Chronologically
32
Figure 8.2 Mean weekly switches per 100 MPs (All and SMD), by sequential substage in term, 1993–1995 Russia Duma and 1996–2001 Italian Chamber (SEs in error bars).
Timing Matters
213
PR) in the two legislatures. During the Russian A.1, which lasted only one week, SMD (but not PR) deputies were asked to declare factional membership. Whereas any party qualifying for PR seats automatically received official Duma faction status, additional groups of SMD MPs could be formed and could obtain the same rights enjoyed by factions, as long as the group met the minimum size of 35. Seizing the opportunity to affiliate, SMD legislators switched with abandon, and formed an entirely new faction during A.1. PR MPs could not create new groups during A.1, which ended when the Duma shifted its focus to internal institutional matters, committee assignments, and other office-related votes (stage B, benefits). Only after the extant groups had successfully monopolized committee and leadership posts did they write rules permitting new groups to form, with the approval in March 1994 of the law on registration of Duma factions. This law opened Russia’s second affiliation stage (A.2) and enabled PR MPs to switch. Even in A.2, however, SMD MPs—in particular, those elected as independents—evinced a higher rate of switching. Switching in the single benefits stage was not as great as in the two affiliation stages, but MPs elected as independents provided most of the moves in B as well. Our second hypothesis thus finds additional corroboration. Legislators’ policy disputes dealing with the second major campaign of the Chechen War (classified as stage C.3, near the midterm) generated the second highest mean weekly switches per 100 MPs in the 1993–1995 Duma. Relatively high rates of switching appeared in two other policy stages we define, one involving the first Chechen War and subsuming legislative deliberations on the 1995 budget (C.2) and the other comprised of legislators’ efforts to contribute to managing the Budennovsk hostage crisis (C.4). These stage C switches created new factions and redefined agenda setting, consistent with the third hypothesis. Moreover, as the term unfolded, a number of SMD MPs originally elected as independents emerged as notables within the Duma, able to exert strong policy influence, so that their moves may also be interpreted as supporting the third hypothesis. As figure 8.2b exhibits, the rate of Italian switching is highest in a policy stage near the midterm. In October 1998, during stage C.6, the Prodi government made a vote on the 1999 budget a matter of confidence, lost, and resigned. Communist Refounding (RC), not in the executive but until then routinely in its legislative majority, split on the confidence vote. The dissidents entered the Mixed Group, unable to form a separate legislative party due to rules on minimum size. The pro-government majority of the RC moderated and established
214
Carol Mershon and Olga Shvetsova
a distinct parliamentary group with a new name. The RC’s fission and its effect comport with the third hypothesis and fit the corollary as well. The second-highest peak in Italian switching arose in a benefits stage, B.3, when seats and leadership posts on the large Bicameral Committee on Constitutional Reform were allocated. The Prodi government made constitutional reform one of its top priorities. Since the choice of rules carried profound consequences for policy outcomes (e.g., Riker 1982), appointments to the Bicameral Committee were intrinsically linked to high-stakes policy. The bulk of switches in B.3 involved MPs from the Christian Democratic Center-United Democratic Christians (CCD-CDU), the leftmost group in the center-right bloc. The moves were made en masse, as the CDU portion of what began as the unified CCD-CDU group split off and entered the Mixed Group; the CDU contained an unusually large share of PR MPs. The day after the CDU bolted, its leader was named to the Bicameral Committee, with a seat on two of its four subcommittees. Events in stage B.3 march with our hypotheses 2 (on office and policy motives, bound together, and on notables) and 3 (agenda advantage). Relatively high rates of switching distinguished the run-up to and aftermath of the May subnational elections, stage E.3. To contest the 1998 elections, Senator and former President Cossiga founded a new centrist party (the UDR, Union of Democrats for the Republic), which attracted MPs from the former CDU and CCD, as well as some with center-left origins. As the campaign unfolded and overlapped with a policy stage, Cossiga announced the UDR’s stance on budget legislation and constitutional questions, pursued tightly linked electoral and policy aims, and sought the center of the Italian policy space. His initiatives and the switches he spurred support the second and third hypotheses and the corollary on the location of new groups. After the peak in policy stage C.6, a string of stages with relatively pronounced switching rates appeared. The single week classified as D.5 occurred in October 1998, between decisions on benefits in the D’Alema I cabinet (which replaced the ill-fated Prodi I), and the resumption of committee handling of the 1999 budget; all but one of the switchers in D.5 moved from the UDR to the Mixed Group. Stage D.6 preceded the electoral cycle defined by the June 1999 subnational and European Parliament elections (E.4). In phase D.6, preparing for the 1999 elections, MP and former Premier Prodi launched the center-left Democrats-Olive Tree (Dem-U). Additional moves in stage E.4 enlarged the Dem-U, so that it satisfied the minimum size
Timing Matters
215
requirement (20 MPs) for a separate parliamentary group. Other switches in the wake of the 1999 elections, still in E.4, created the Democratic Union for Europe (UDEur) as an organized component of the Mixed Group and as the successor to the UDR, also led by Cossiga. Chamber membership in the UDEur swelled sufficiently during policy stage C.8 to qualify it as a legislative group, and stage C.8 witnessed debates on finance so fierce that two days after the 2000 budget bill won approval, the government resigned. Together, these episodes feature moves responding to notables’ initiatives and coinciding with policy controversy and electoral campaigns. Given the timing of D.5 and D.6 relative to benefit, policy control, and electoral phases, switching in these two substages of the generally calm D does not upset our hypotheses and even buttresses them. In both Italy and Russia, we uncover evidence on vote-driven switching that aligns with our reasoning. The rate of switching in the Russian reelection stage E is roughly the same as that in the two Italian nonparliamentary electoral stages just discussed. We cannot observe a parliamentary reelection stage for Italy as we do for Russia, since the president dissolved the Italian Parliament before the date of the 2001 election had been set and before the mandated day could be identified for depositing candidate names and party lists. We can apply the 90-day rule of thumb used for nonparliamentary elections, however, and count back from the earliest date of those in question. Not a single MP moved during the preelectoral substage so defined; one moved a few days before this substage started. Thus, in Italy, jockeying for electoral advantage occurred in conjunction with subnational and supranational elections (cf. Heller and Mershon 2005). And since Italian MPs synchronized vote-seeking switching to nonlegislative elections, they had tested the performance of their new electoral vehicles by the time the next parliamentary elections were held. In Russia, instead, relatively many MPs switched immediately before the parliamentary campaign. We attribute this contrast in timing to differences in institutions, the degree of party system institutionalization, the age of the democratic regimes, and the information available to the players in electoral politics. We have adduced evidence in favor of all three hypotheses.6 Yet a more robust appraisal of our hypotheses would extend to varied institutional settings and would include party systems commonly viewed as relatively stable. We thus now turn to our subsidiary country-terms to probe the more general purchase of our reasoning.
216
Carol Mershon and Olga Shvetsova
8.4.3 Extending the Test Table 8.4 examines the frequency of switching in stages of the parliamentary cycle across a total of 35 country-terms in 1963–1972 Italy, Romania, the United Kingdom, and the United States. Not a single switch occurred in stage A (affiliation) in Italy, Romania, and the United Kingdom, whereas the standardized measure of mean monthly switches per 100 MPs is 0.051 for the A stages in the United States. The switchers in the US stage A were almost all elected as independents or on a minor party label; having won their seats, they joined one of the two major parties, with which they may or may not have had a prior affiliation.7 We see this pattern as testament to the impact of electoral systems and modes of candidate selection. It is only in the United States that the use of a primary system in SMDs opens a side door through which a few independents enter the lower house, as, for example, challengers who have lost in primaries run—and win—as independents in a three-way race in the general election. In the other three countries, independents’ access to legislative office is extremely restricted: In the United Kingdom, party organizations choose candidates for SMD races; and in 1963–1972 Italy and in Romania, party-list PR gives party leaders great control over candidate selection (cf. Birnir 2004, esp. 144–146, on Romania; Wertman 1977 on Italy). Hence the contrast emerges between, on the one hand, the completely quiescent stage A in 1963–1972 Italy, Romania, and the United Kingdom, and, on the other, the United States, where the relatively few independents and minor party candidates who win tend to join one of the major parties in stage A so as to avail themselves of the benefits of party affiliation. We do not wish to exaggerate the contrast. Yet the finding here reinforces the lesson drawn earlier from the hybrid electoral systems that switching in stage A is shaped by election rules. The independents permitted in SMD races in Russia switched often in stage A; and the Italian SMD candidates who were forced by law to link to a specific PR party list but whose campaigns stressed the broader themes of the center-left bloc tended to switch to the generic Mixed Group in stage A. Institutional influences also underpin cross-national differences in switching rates for other stages in the parliamentary cycle. In particular, we attribute the variations in switching in stage B (benefits) to variations in electoral laws, legislative institutions, and regime type. First, consider the two PR systems here. The parliamentary regime, 1963–1972 Italy, is distinguished by a relatively high switching rate in
Table 8.4
Mean monthly switches per 100 MPs by type of stage, electoral system, and regime, 35 country-terms List PR Open, Parliamentary Italy 1963–1972
Stage A B C D E
SMD Closed, Semi-Presidential Romania 1996–2004
Parliamentary
Presidential
United Kingdom 1979–2005
United States 1950–2000
Mean/100 MPs
T Mos
Mean/100 MPs
T Mos
Mean/100 MPs
T Mos
Mean/100 MPs
T Mos
0.000 (0.000) 0.933 (3.604) 0.367 (2.118) 0.025 (0.080) 0.011 (0.042)
4 16 99 19 14
0.000 (0.000) 0.172 (0.237) 0.982 (5.348) 0.143 (0.553) 0.258 (0.352)
4 7 77 19 14
0.000 (0.000) 0.000 (0.000) 0.032 (0.149) 0.019 (0.098) 0.022 (0.062)
18 13 247 108 56
0.051 (0.134) 0.069 (0.141) 0.008 (0.041) 0.011 (0.050) 0.003 (0.025)
50 50 450 100 175
Key to stages:
A: Affiliation B: Benefits C: Control of policy E: Electoral D: Dormant, that is, all periods other than A, B, C, and E. Note: Numbers between parentheses are standard deviations. Sources: For numbers and dates of switches: Ballot Access News 2001a, 2001b; Butler and Butler 2006, 112–114; Camera dei Deputati 2008a; Timothy Nokken, personal communications, June 2006 and July 2008; Parlamentul României 2008c; US Congressional Biographical Directory 2008. For stages: Butler and Butler 2006; Camera dei Deputati 2008a; CQ Weekly (various dates, e.g., 11/19/83, 2403; 12/8/84, 3051; 2/2/85, 177; 9/14/85, 1827; 2/17/90, 437; 11/12/94, 3222; 12/3/94, 3427; 12/23/95, 3863; 11/28/98, 3218; 1/16/99, 156–157); Keesing’s Record, various dates; Mershon 2002, 212; Parlamentul României 2008a; United Kingdom Parliament 2008; US House of Representatives, Office of the Clerk 2008; US House of Representatives, Office of the Majority Leader 2008.
218
Carol Mershon and Olga Shvetsova
the aggregated benefits stage. The Italian datum for stage B also enters into a policy (C) stage and rests on a single month—indeed, the single day in 1969 that marked the split of the briefly reunified Socialists into Italian Socialist Party (PSI) and Italian Social Democratic Party (PSDI). The split was precipitated by sharp intraparty disputes over government policy and ruptured the ruling coalition (Di Scala 1988, esp. 161–165). Politicians of the PSI and PSDI merged in 1996 and split in 1969, confident that their implementation and reversal of the fusion carried no adverse consequences for their survival, given PR, and for their long-term participation in the executive, given DC dominance of the party system (Mershon 1994; 2002). In semi-presidential Romania, stages B and C also overlap at times, and it is only during the overlap that any B stage switching occurs. Switches to independent status predominated in this span; in fact, the vice-chair of a ruling party became an independent amid disagreements over the budget that triggered a change in government (BBC 1998). Note that MPs’ behavior was constrained by the standing orders of the Romanian Chamber, which stipulate that MPs cannot enter a parliamentary party that has not already earned seats in elections (Chiva 2007, 203–204; Parlamentul României 2008b). Moves to join (or found) splinter parties or start-ups are impossible under these rules. Switching in the benefits stage for the two SMD systems speaks to the impact of legislative institutions and regime design. In our span of study, British politicians never jumped party when offices in oneparty majority executives were decided or when posts on relatively weak legislative committees were distributed. In the US presidential system, executive positions are obviously not at stake in the legislative stage B. Instead, seats and leadership offices on the most specialized, most powerful set of legislative committees in the world are allocated in the US benefits stage. This stage engenders the highest mean monthly switches per 100 legislators of any stage in the United States. Granted, the mean here stands well below that recorded for Italian and Romanian legislators in some other aggregated stages. The fact remains that if US legislators switch, they are likely to switch when either office benefits are awarded, in stage B, or the perks of party affiliation are available, in A.8 The index of mean monthly switches per 100 MPs per stage reaches its maximum value in our dataset in the aggregated Romanian stage C (policy). This average in turn includes the single greatest monthly observation of switches per 100 MPs within our dataset, in
Timing Matters
219
June 2001, when the Party for Social Democracy in Romania (PDSR) merged with the Social Democratic Party of Romania (PSDR) to create the Party of Social Democracy (PSD). The merger stemmed not just from policy proximity (as the names suggest) but also from institutional incentives and constraints. In part as a response to the summer 2000 increase of the PR threshold (from 3 to 5 percent), the PDSR and PSDR agreed on an electoral alliance for the November 2000 parliamentary elections. The official fusion represented the culmination of the electoral alliance, a culmination delayed by internal party deliberations on the merger (Lovatt and Lovatt 2001; Project on Political Transformation 2008). When we draw within-nation comparisons, switching rates are highest in the aggregated stage C for Romania and the United Kingdom, and reach their second-highest value for Italy. These within-nation maxima reflect the founding of new parliamentary groups, discussed further in the following text. Suffice it to emphasize for now that this pattern supports the third hypothesis, since the new parties aimed for an impact on policy in stage C. The Romanian PSD styled itself a natural “party of government” (Lovatt and Lovatt 2001). The Italian Socialist Party of Proletarian Unity (PSIUP) split from the PSI in a 1964 policy stage to protest Socialist participation in government. The PSI and PSDI merged in a 1966 policy stage so as to participate more effectively in coalition government, and split in an overlapping B and C stage in 1969, given recurring policy disagreements (Di Scala 1988). The British Social Democratic Party (SDP), initiated in a 1981 policy stage, sought to pull votes from both Labour and Conservatives and to identify a middle ground in programmatic terms as well (e.g., Crewe and King 1995, esp. Ch. 23). As the bottom row of table 8.4 exhibits, in stage E the least institutionalized party system, Romania, witnesses the most switching, which echoes the finding for the primary cases in Italy and Russia. Once more, we do not wish to overstate our claims. Yet, we infer from this evidence that legislators are more likely to maneuver for votes by defecting to a new party or by adopting independent status in the run-up to parliamentary elections in party systems where voters lack firm attachments to party labels and where the ideological distinctions among parties are relatively blurry. The relatively weak institutionalization of the Romanian and Russian party systems underlies the flurry of switching that both exhibit in anticipation of near-at-hand parliamentary elections.
220
Carol Mershon and Olga Shvetsova
For no country here do mean switches per 100 MPs in the aggregated stage D bottom out at zero. All, however, evince limited switching in D relative to the rate observed in at least a few of the active stages. It turns out that stage D is indeed relatively dormant, contrary to the null hypothesis and in line with the first hypothesis and our overall reasoning. The evidence from the 35 subsidiary country-terms backs all three hypotheses and also points to the importance of peaks in switching marking the formation of new parliamentary parties. The corollary to the third hypothesis holds that parties established in policy stages should locate at the center of the policy space. We now assess evidence on the corollary from both primary and subsidiary country-term cases.
8.4.4 Switching to Found New Parliamentary Groups As can be seen in table 8.5, 80 percent—12 of 15—of the parliamentary parties created over the course of the 37 terms comprehended in our study were launched in a policy stage, C. This count includes the Italian UDR, established during an overlapping policy and electoral stage. Excluded is the Italian Dem-U, which qualified as a parliamentary group in an electoral stage, E. As detailed earlier, however, the timing of that party’s founding fell soon after the midterm peak in switching in an Italian stage C. Of the 13 new players in legislative decision making just discussed, all but 2—the PSIUP of 1964 and the PSI of 1969—were either centrist or more center-leaning than their predecessor party or parties.9 Consideration of the remaining two parliamentary parties here bolsters the evidence in favor of the corollary. Two new groups in the Russian lower house were created as independents sought to reap the rewards of group affiliation during the tumultuous stage A in 1993–1995 Russia. Both entrants originating in an affiliation stage, when office aims are assumed to dominate, located on the periphery of the policy space. In accordance with the corollary to the third hypothesis, then, when legislative parties were founded during a policy stage, they tended to adopt centrist or center-leaning policy positions. Parliamentary groups established outside policy stages located away from the center. Table 8.5 also presents suggestive evidence on the impact of institutions. The only instances of fission captured in our data occur in the Italian Chamber, which for the 1963–1972 span (and from the end of World War II to 1993) featured a highly permissive form of PR and
Table 8.5 Switching to create new parliamentary parties, by type of stage, electoral system, and regime, 37 country-terms List PR
SMD
Hybrid
Open, Parliamentary
Closed, SemiPresidential
Parliamentary
Presidential
SemiParliamentary Presidential
Stage
Italy 63–72
Romania
United Kingdom
United States
Italy 96– 01
Russia
C
PSIUPa
PSD b
SDPc
No new party
UDR c
Rossc
GC a UDEurc
Stabc
DemUc
LDS c
PSU b PSIa and PSDIa Non-C
LibDemb
No new party
N-96 c Note: Entries are acronyms of parties, ordered chronologically by type of stage and by country. Acronyms in italics indicate a new party that is either centrist or more center-leaning than its predecessor(s). a Fission b Fusion c Start-up Key to party acronyms (year founded): DemU Democrats-Olive Tree (1999) GC Communist Group (1998) LDS Liberal Democratic Union of December 12 (1994) LibDem Liberal Democratic Party (1988) N-96 New Regional Policy—Duma 96 (1993) PSD Social Democratic Party (2001) PSDI Italian Social Democratic Party (1969) PSI Italian Socialist Party (1969) PSIUP Italian Socialist Party of Proletarian Unity (1964) PSU Unified Socialist Party, Italy (1966; name changes not recorded here) Ross Russia (1995) SDP Social Democratic Party (1981) Stab Stability (1995) UDR Union of Democrats for the Republic (1998) UDEur Democratic Union for Europe (1999) Sources: Butler and Butler 2000, 248–249; Camera dei Deputati 2008a, 2008b; Crewe and King 1995, 114, 478, passim; Di Scala 1988; INDEM 2000; Lovatt and Lovatt 2001; Parlamentul României 2008a, 2008c.
222
Carol Mershon and Olga Shvetsova
from 1993 to 2005 used a hybrid system deliberately designed to preserve small parties (e.g., Warner and Gambetta 1994). The products of party fission in Italy could realistically hope to survive, given the electoral rules. Romania witnessed only one fusion over the eight years studied, reflecting the constraints imposed by the legislative standing orders and the incentives to coalesce under the 2000 electoral laws. The United Kingdom experienced one start-up, the SDP, and one fusion, the Liberal Democrats. Under SMD rules, political entrepreneurs should strive to avoid fission; and in the case of the SDP, they succeeded, in that the party drew legislators with multiple affiliations (not only Labour MPs but also, in the 1979–1983 term alone, two MPs with Independent Labour status, one Conservative MP, and one Conservative Peer; see Butler and Butler 2000, 248–249; Crewe and King 1995, 114, 478). The utter absence of new parties in the 1950–2000 US Congress may be interpreted as a consequence not only of the constraints of SMD competition but also of the need for parties to mount nationwide presidential contests. The relative abundance of start-ups in the two systems using hybrid electoral laws also deserves comment. Under the Italian variant of hybrid laws, even small units in electoral competition could hope to earn legislative seats, due both to the mechanics of the linkage between SMD and PR races and to the parties’ strategic construction of electoral cartels and strategic nomination of some candidates from small parties in SMDs (e.g., D’Alimonte and Bartolini 2002; Di Virgilio 1997; 2002).10 The unlinked PR and SMD tiers in the Russian hybrid laws slowed the process of party system institutionalization and invited political entrepreneurs to create and change parties with ease (cf. Shvetsova 2003; 2005).
8.5
Conclusion
The key to the analytical approach here is the notion that if legislators change party affiliations, the timing of their switch reveals its chief motivation: When they switch tells us why they switch. We assume that legislators are strategic actors who view party affiliation as a strategic choice—as a means to an end. In this perspective, any choice to change parties signifies that a new allegiance has become more advantageous than the old one, given changed opportunity structures. Starting from this premise, we expect politicians’ behavior to track and respond to evolving opportunities. To check our expectations, we introduce a partitioning of the parliamentary cycle into distinct stages that are defined by different legislative activities and hence by different
Timing Matters
223
rewards available to MPs. We do not assert that legislators in all legislatures switch automatically as a different type of payoffs becomes most salient in the parliamentary cycle. After all, sticking with the party of origin may well be the most advantageous strategy for many politicians. Acknowledging this, we argue that if MPs switch, they should do so when their move is most likely to bring them a particular type of payoff; and the type, like the switch, depends on time. Attention to the timing of switches brings us to findings that are novel in the literatures on party switching, legislative politics, and political parties. For our two primary country-term cases, 1996–2001 Italy and 1993–1995 Russia, we find that switching patterns are tied to the stages in the parliamentary cycle. We identify in our evidence significant differences in switching rates across stages of the cycle and heightened switching for perks, office, policy advantage, agenda grabbing, and preelectoral jockeying at distinctive stages. Phases of relative calm in the legislative agenda induce relatively low rates of switching. Differences in the timing of vote-driven switching across the two cases can be attributed to differences in the degree of party system institutionalization and thus the information held by elites and voters. In the more institutionalized Italian system, electoral positioning via switching occurs in anticipation of and, to a lesser extent, after subnational elections, as MPs adjust expectations about voters (cf. Heller and Mershon 2005). In Russia, preelectoral positioning occurs during national election campaigns. Driving home the import of party system institutionalization is the rough resemblance, displayed in table 8.3, in partisan Russian MPs (regardless of mandate) and those in Italy. In selecting the two primary country-terms, we deliberately restrict variation in electoral institutions. To submit our hypotheses to a stiffer test, we extend the inquiry to an additional 35 subsidiary countryterms, guided by the criterion of maximizing institutional variation. Our comparison of 1963–1972 Italy, Romania, the United Kingdom, and the United States yields findings that largely comport with the hypotheses we advance. In accordance with the first hypothesis, the rate of switching varies across the stages of the parliamentary cycle. Moreover, institutional variations across countries shape which stage evinces the highest switching rate within each country. The subsidiary cases also furnish support for the second hypothesis. For example, independents (where they win legislative seats, as influenced by electoral laws) tend to move above all in affiliation and benefits stages. Switches in stage C feed into new parties that are centrist or center-leaning. This evidence aligns with the third hypothesis and its corollary.
224
Carol Mershon and Olga Shvetsova
In addition, the broad comparisons we draw permit a tentative inference. PR tends to provide more parties among which to switch and, indirectly, creates more opportunities to switch in parliamentary and semi-presidential systems, since it makes coalition governments more likely and coalitions do not last as long as single-party majorities (see, among many others, Laver and Schofield 1990). To put the point differently, SMD systems tend to have lower switching rates, both because fewer parties exist and because switching tends to be less advantageous. In a two-party system resulting from SMD rules, switches necessarily occur between the minority and the majority. Switching is polarizing and visible, and yet unlikely to be effective—to allow switchers to redirect the policy agenda or win votes. This inference, though provisional, is confirmed in a large-N analysis controlling for multiple institutional effects (Mershon and Shvetsova 2008b). A venerable tradition in political science—indeed, virtually all of the profession—holds that institutions matter. Our study too conveys the import and impact of institutions. What is distinctive here is the emphasis on the timing of changes of party affiliation, which sheds new light on the role of institutions in structuring the strategic choices of elected legislators, choices that to a greater or lesser extent alter the legislative party system in which legislators operate. Taking timing into account enriches research on party affiliation and partisan competition. Some phases of the parliamentary cycle offer legislators relatively many stimuli to reconsidering choice of partisanship. In other phases, such stimuli are in low supply. Not just the quantity but also the quality of stimuli counts: Distinct categories of activity dominating discrete stages of the parliamentary cycle elicit a particularly strong focus on particular goals on the part of legislators. Temporal stages leave their imprint on changes of partisanship in legislatures. Timing matters for switching, parties, and legislative party systems.
8.6
Notes
1. Here too, an Independent might choose to join an organized legislative group. Some legislatures permit an Independent group as well. Note that many scholars have omitted what we call stage A from analytical scrutiny, for research has traditionally focused either on the election or on the legislature after it has convened for its first session. 2. Given these similarities, smaller institutional differences do not undermine the case selection. Whereas Russia permitted independents in SMDs, Italy from 1993 to 2005 required that every SMD candidate be linked to at least one PR party list. In Italy, but not in Russia, SMD wins were compensated
Timing Matters
3.
4. 5. 6.
7.
8.
9.
10.
225
in the PR tier (Katz 2001; Moser 2001). Russia had equal shares of SMD and PR MPs, and in Italy approximately 75 percent of MPs came from SMDs. PR thresholds were 5 percent in Russia and 4 percent in Italy. With Mainwaring (1999; cf. Mainwaring and Scully 1995; Mainwaring and Torcal 2006), we define a highly institutionalized party system as one displaying stable patterns of party competition, strong linkages between parties and voters, substantial legitimacy for parties, and solid party organizations. We do not examine security policy in Italy, since internal warfare appears only in Russia. According to the Chamber’s internal rules (Arts. 14 and 15), the Mixed Group enjoys much the same rights and privileges as do party groups. The evidence also points to midterm peak in switching, which we explore in Mershon and Shvetsova (2008a, 116–118) and Mershon and Shvetsova 2008b. Because we count switches between elections and the convening of Congress, we include switchers whose behavior is not captured in the authoritative Poole-Rosenthal roll-call vote database (available at www.voteview.com; cf. Nokken this volume; Nokken and Poole 2004). That said, because we focus on choices made while politicians hold national legislative office, our data on the United States omit a few switchers examined by Nokken in this volume. Candidates elected as independents or on minor party labels account for 11 of 12 switches in the US stage A and 13 of 16 switches in stage B. (Recall that stages A and B overlap, but do not coincide for the United States or for most other country-term cases as well.) To place parliamentary parties in policy space, we rely on expert judgments (Birnir 2004; Chiva 2007; Crewe and King 1995; Di Scala 1988; Lovatt and Lovatt 2001) and, as available, roll-call voting data (on Italy, see sources cited in Heller and Mershon this volume, ch. 7; on Russia, the Duma statistical service [INDEM]). The well-known Manifesto data (e.g., Budge et al. 2001) are inappropriate for our purposes because they pertain to units in electoral competition, not parliamentary groups founded at particular stages during a term. Moreover, small parties could hope to survive in electoral contests using pure PR, such as EP elections. During the 1996–2001 national legislative term, Italy’s electoral arenas were structured by a total of seven distinct electoral systems (Corbetta and Parisi 1997, 14).
8.7
References
Ballot Access News. 2001a. “Candidates Elected to the House, 1902–2000, Who Weren’t Nominees of the Democratic or Republican Parties.” 17 (3, June 1). http://web.archive.org/web/20030417164754/http://www.ballot-access. org/2001/0601.html#17. Last accessed July 22, 2008. ———. 2001b. “Minor Party Candidates Elected to the House, 1902–2000, Who Were Nominated by Their Own Minor Party, Plus Major Party.” 17
226
Carol Mershon and Olga Shvetsova
(3, June 1). http://web.archive.org/web/20030417164754/http://www.ballotaccess.org/2001/0601.html#17. Last accessed July 22, 2008. BBC (British Broadcasting Corporation). 1998. “Romania’s Coalition Democrats Urge Formation of New Government.” March 13. http://news.bbc.co.uk/2/hi/ world/monitoring/65158.stm. Last accessed July 25, 2008. Birnir, Jóhanna Kristín. 2004. “Institutionalization of the Party System.” In Romania since 1989: Politics, Economics, Society, ed. Henry F. Carey, 138–158. Lanham, MD: Lexington Books. Budge, Ian, Hans-Dieter Klingemann, Andrea Volkens, Judith Bara, and Eric Tannenbaum. 2001. Mapping Policy Preferences. Estimates for Parties, Electors and Governments, 1945–1998. Oxford: Oxford University Press. Butler, David, and Gareth Butler. 2000. Twentieth Century British Political Facts, 1900–2000, 8th ed. London: Macmillan. ———. 2006. British Political Facts since 1979. New York: Palgrave Macmillan. Camera dei Deputati. 2008a. Legislature precedenti. http://legislature.camera.it/. Last accessed July 7, 2008. ———. 2008b. XIII [Tredicesima] legislatura. http://leg13.camera.it/. Last accessed July 9, 2008. Chiva, Cristina. 2007. “The Institutionalization of Post-Communist Parliaments: Hungary and Romania in Comparative Perspective.” Parliamentary Affairs 60 (2): 187–211. Corbetta, Piergiogio, and Arturo M. L. Parisi, eds. 1997. A domanda risponde. Il cambiamento del voto degli italiani nelle elezioni del 1994 e del 1996. Bologna: Il Mulino. CQ Weekly (Online). Various dates. http://library.cqpress.com/cqweekly. Crewe, Ivor, and Anthony King. 1995. SDP: The Birth, Life, and Death of the Social Democratic Party. Oxford: Oxford University Press. D’Alimonte, Roberto, and Stefano Bartolini. 2002. Maggioritario finalmente? La transizione elettorale 1994–2001. Bologna: Il Mulino. Di Scala, Spencer M. 1988. Renewing Italian Socialism: From Nenni to Craxi. New York: Oxford University Press. Di Virgilio, Aldo. 1997. “Le alleanze elettorali. Identità partitiche e logiche coalizionali.” In Maggioritario per caso. Le elezioni politiche del 1996, ed. Roberto D’Alimonte and Stefano Bartolini, 71–136. Bologna: Il Mulino. ———. 2002. “L’offerta elettorale. La politica delle alleanze si istituzionalizza.” In Maggioritario finalmente? La transizione elettorale 1994–2001, ed. Roberto D’Alimonte and Stefano Bartolini, 79–129. Bologna: Il Mulino. Heller, William B., and Carol Mershon. 2005. “Party Switching in the Italian Chamber of Deputies, 1996–2001.” Journal of Politics 67 (2): 536–559. Huntington, Samuel P. 1991. The Third Wave: Democratization in the Late Twentieth Century. Norman: University of Oklahoma Press. INDEM (Information and Analysis System). 2000. INDEM-Statistics 2000 (Full version). Moscow: INDEM Foundation. Istituto Cattaneo. 2008. “Archivi crono. Cronologia dei principali avvenimenti politici e sociali.” http://www.cattaneo.org/index.asp?l1=archivi&l2=crono. Last accessed July 30, 2008.
Timing Matters
227
Kaminski, Marek M. 2002. “Do Parties Benefit from Electoral Manipulation? Electoral Laws and Heresthetics in Poland, 1989–93.” Journal of Theoretical Politics 14 (3): 325–358. Katz, Richard S. 2001. “Reforming the Italian Electoral Law, 1993.” In MixedMember Electoral Systems: The Best of Both Worlds? ed. Matthew. S. Shugart and Martin. P. Wattenberg, 96–122. Oxford: Oxford University Press. Keesing’s Record of World Events. Various dates. www.keesings.com. Last accessed July 3, 2008. Kunicová, Jana, and Thomas F. Remington. 2005. “The Effect of Electoral Rules on Distributive Voting: Some Evidence from the Russian State Duma, 1994–2003.” Presented at the Annual Meetings of the Midwest Political Science Association, Chicago, IL. Laver, Michael, and Norman Schofield. 1990. Multiparty Government: the Politics of Coalition in Europe. Oxford: Oxford University Press. Laver, Michael, and Kenneth Shepsle. 1996. Making and Breaking Governments: Cabinets and Legislatures in Parliamentary Democracies. Cambridge: Cambridge University Press. Lijphart, Arend. 1984. Democracies: Patterns of Majoritarian and Consensus Government in Twenty-One Countries. New Haven, CT: Yale University Press. Lovatt, Catherine, and David Lovatt. 2001. “PSD: A New Party of Government.” Central Europe Review 3 (23, June 25). http://www.pecina.cz/files/www. ce-review.org/01/23/romanianews23.html. Last accessed July 10, 2008. Mainwaring, Scott P. 1999. Rethinking Party Systems in the Third Wave of Democratization: The Case of Brazil. Stanford: Stanford University Press. Mainwaring, Scott, and Mariano Torcal. 2006. “Party System Institutionalization and Party System Theory after the Third Wave of Democratization.” In Handbook of Party Politics, ed. Richard S. Katz and William Crotty, 204–227. London: Sage. Mainwaring, Scott, and Timothy Scully. 1995. Building Democratic Institutions: Party Systems in Latin America. Stanford: Stanford University Press. McKelvey, Richard D., and Norman Schofield. 1987. “Generalized Symmetry Conditions at a Core Point.” Econometrica 55 (4): 923–933. Mershon, Carol. 1994. “Expectations and Informal Rules in Coalition Formation.” Comparative Political Studies 27 (1): 40–79. ———. 2002. The Costs of Coalition. Stanford: Stanford University Press. Mershon, Carol, and Olga Shvetsova. 2005. “Electoral Cycles and Party Switching: Opportunistic Partisan Realignment in Legislatures.” Paper presented at the annual meeting of the Midwest Political Science Association, Palmer House Hilton, Chicago, IL, April 6–10. ———. 2008a. “Parliamentary Cycles and Party Switching in Legislatures.” Comparative Political Studies 41 (1): 99–127. ———. 2008b. “Party Switching in Sitting Parliaments and the Midterm Effect.” Presented at the Annual Joint Sessions of the European Consortium for Political Research, Rennes, France. Ministero dell’Interno. 2008. “Archivio storico delle elezioni.” http://elezionistorico.interno.it/index.php. Last accessed July 30, 2008.
228
Carol Mershon and Olga Shvetsova
Moser, Robert G. 2001. Unexpected Outcomes: Electoral Systems, Political Parties, and Representation in Russia. Pittsburgh: University of Pittsburgh Press. Müller, Wolfgang C., and Kaare Strøm, eds. 1999. Policy, Office, or Votes: How Political Parties in Western Europe Make Hard Decisions. Cambridge: Cambridge University Press. Nokken, Timothy P., and Keith T. Poole. 2004. “Congressional Party Defection in American History.” Legislative Studies Quarterly 29 (4): 545–568. Parlamentul României, Camera Deputat, ilor. 2008a. “Activitate parlamentara˘.” http://www.cdep.ro/pls/steno/steno.home?idl=2. Last accessed June 28, 2008. ———. 2008b. “Organizarea s¸i funct, ionarea camerei deputat, ilor: Grupurile parlamentare.” http://www.cdep.ro/pls/dic/site.page?id=134&idl=1. Last accessed July 10, 2008. ———. 2008c. “Prima pagina.” http://www.cdep.ro/. Last accessed June 28, 2008. Pasquino, Gianfranco. 1999. “Autopsia della Bicamerale.” In Politica in Italia, Edizione 99, ed. David Hine and Salvatore Vassallo. Bologna: Il Mulino. Project on Political Transformation and the Electoral Process in Post-Communist Europe, University of Essex. 2008. “Romania: Legislation.” http://www2.essex. ac.uk/elect/database/indexCountry.asp?country=ROMANIA&opt=leg. Last accessed July 9, 2008. Riker, William H. 1982. Liberalism against Populism: A Confrontation between the Theory of Democracy and the Theory of Social Choice. San Francisco: W.H. Freeman. Shvetsova, Olga. 2003. “Resolving the Problem of Pre Election Coordination: The 1999 Parliamentary Election as Elite Presidential ‘Primary.’ ” In Elections, Parties, and the Future of Russia, ed. Vicki Hesli and William S. Reisinger, 213–232. Cambridge: Cambridge University Press. ———. 2005. “Compromising a Long Lasting Transitional Formula. In Handbook of Electoral System Design, ed. Josep Colomer, 382–397. New York: St. Martin’s. Strøm, Kaare. 1990. “A Behavioral Theory of Competitive Political Parties.” American Journal of Political Science 34 (2): 565–598. United Kingdom Parliament. 2008. “List of Previous Commons Recess Dates.” http://www.parliament.uk/faq/recess.ccfm. Last accessed June 13, 2008. US House of Representatives, Office of the Clerk. 2008. “Official House Information: House Floor Proceedings.” http://clerk.ouse.gov/index.html. Last accessed June 30, 2008. US House of Representatives, Office of the Majority Leader. 2008. “U.S. House Vote Schedule.” http://majorityleader.house.gov/docUploads/2008CALENDAR.pdf. Last accessed June 30, 2008. United States Congressional Biographical Directory. 2008. http://bioguide.congress.gov/biosearch/biosearch.asp. Last accessed July 9, 2008. Verzichelli, Luca. 1996. “I gruppi parlamentari dopo il 1994. Fluidità e riaggregazioni.” Rivista italiana di scienza politica 26 (2): 391–413.
Timing Matters
229
Warner, Steven, and Diego Gambetta. 1994. La retorica della riforma. Fine del sistema proporzionale in Italia. Turin: Einaudi. Wertman, Douglas. 1977. “The Italian Electoral Process.” In Italy at the Polls: The Parliamentary Elections of 1976, ed. Howard R. Penniman, 41–79. Washington, DC: American Enterprise Institute. Zucchini, Francesco. 2001. “La commissione affari costituzionali: Gli effetti paralizzanti del mutamento.” In Parlamento e processo legislativo in Italia. Continuità e mutamento, ed. Gilberto Capano and Marco Giuliani, 153–186. Bologna: Il Mulino.
This page intentionally left blank
IV
Party Switching and the Dynamics of Party Systems
This page intentionally left blank
9 Competition for Power: Party Switching and Party System Change in Japan Junko Kato and Kentaro Yamamoto
9.1 Introduction The reasons that individual legislators change their party affiliations are as varied as the goals they hold dear. Primary among them are electoral, office, and policy considerations. The payoffs from party switching are closely linked to electoral and party systems, as well as the dynamics of party competition. In short, the institutional context defined by electoral rules and party politics influences party switching, which in turn can affect the dynamics of party competition; party switching by individual legislators thus has the potential to alter party systems. Much of the existing literature on party switching focuses on the systemic changes it induces. Laver and Kato (2001), for example, analyze a recent rash of switching in Japan that brought about party breakups and mergers, as well as the formation of new parties. The 1993–1996 period, in which the Liberal Democratic Party (LDP) regained a razor-thin legislative majority, demonstrates an important aspect of inter-electoral change in party systems—that is, that large parties are able to offer benefits and payoffs to attract defectors from other parties. Moreover, accepting party switchers can enable parties to pass the threshold for winning a majority between elections. Another example is provided by Laver and Benoit (2003), who develop the theoretical implications of Laver and Kato (2001) to explore the inter-electoral evolution of party systems in which the largest parties play a key role. To this end, they focus on the dominant party that is free to join either of the mutually exclusive winning coalitions (cf. Einy 1985; Peleg 1981; van Deeman 1989; van Roozendaal 1992), and show, based on computer simulations, that a dominant party
234
Junko Kato and Kentaro Yamamoto
attracts switchers and increases its membership at the expense of the second-largest party. The Japanese case in the 1990s allows further theorizing about inter-electoral party system changes in which party switching by individual legislators and factions ultimately led to the reorganization of parties and changes in the partisan balance of power. The LDP’s predominance, extending from 1955 to 1993, has been replaced by the LDP-centered governing coalition. The LDP was out of office from 1993 to 1994. Since 1994, the LDP has maintained office in coalition (except a short period of a single minority government) and has shut the second-largest party out of power. The analysis presented here is based on two concepts. The first is the notion of a decisive majority (DM), which defines the winning threshold for a party coalition in a particular institutional arrangement. Our definition of a DM is based on the presumption that the institutional constraints at play include the committee system. To distinguish the DM from conventionally understood majorities, we call an absolute majority that surpasses the usual winning threshold the formal majority (FM). The key distinction between an FM and a DM lies in the contrast between the Japanese Diet, with a strong committee system, and the prototypical parliamentary system. In majoritarian parliaments, control over a floor majority yields control over government office, which is in turn synonymous with control over policy; in Japan, although an FM is sufficient to ensure government office, a DM is necessary to ensure control over policy content. This distinction, combined with party switching that can push a party over one or both thresholds, provides useful information about legislator motivations. Switching to a party that is close to an FM is different from switching to one that has an FM but falls short of a DM. Switchers in the latter category either see a DM as the only relevant hurdle for a winning coalition, or they consider both types of majority to be important. In the presence of rules that make policy control feasible only with a DM, switching takes on an added dynamic—switching to the secondlargest party can both deny policy control to the largest party and give the second-largest party the institutional wherewithal to challenge the government. Given recent political changes, Japan provides an apt case both to illustrate the importance of different thresholds in the party coalition game and to explore a variety of incentives for party switching. The distinction between the FM and the DM allows for a reinterpretation
Competition for Power in Japan
235
of the LDP’s power vis-à-vis the second-largest party and thus the extent of its dominance in changing party systems.
9.2
Institutional Constraints and Decisive Majorities 9.2.1 Definition
Simple majority rule is prevalent in contemporary democracies. However, in many parliaments, governing majorities can find themselves constrained by institutions in that they cannot necessarily pass bills on the basis of their control of a simple majority of seats. The relevant constraints include committees whose seats and chairs are allocated proportionately to parties’ parliamentary seat shares (Mattson and Strøm 1995, 276–280). Even committees that reflect the composition of parliament often impose another institutional constraint on legislative organizations in which a simple majority is applied: if bills must be passed in committee before a vote on the floor, an FM is not enough for a governing coalition to safely pass bills because a coalition that has an FM on the floor does not necessarily control all the committees. For this purpose, the majority needs to secure enough parliamentary seats to control a majority in all committees. The number of seats required, a DM, defined in distinction from FM, results not from any additional rule, but rather is a simple mathematical consequence of the committee system. To illustrate this point, a few arithmetical examples suffice. Consider a 100-member parliament, as in Panels 1 and 2 of table 9.1, where seats are allocated on two committees and all legislators must be affiliated with one committee. If the committees have an odd number of seats (e.g., 45 and 55, as on the first row of Panel 2 in table 9.1), it is possible to secure a simple majority in each (23 and 28, for a total of 51), which is the same as having an FM in parliament. However, if the two committees have an even number of seats (e.g., 50 each, as on the second row of Panel 2), a larger number of seats than an FM is necessary to secure a majority in both committees. A hypothetical three-committee system composed of 25-, 35- and 40-member committees, illustrated on the third row of Panel 2, would also require a larger number of seats than an FM (i.e., 52) to control a majority in all committees. A system of three committees made up of 20, 30, and 50 seats, respectively, would necessitate 53 seats to ensure a majority in each committee. A set of
Table 9.1 Illustrations of the decisive majority: Number of seats needed to surpass the winning threshold given different committee sizes and systems Total N Seats on Each Committee, A-G
N Seats Needed to Control Majority on Each Committee
Total N Committees
Total N Seats (across all committees)
Total N Seats Needed for DM
1. Examples of Committees of Varying Size A: 20 11 B: 25 13 C: 30 16 D: 35 18 E: 40 21 F: 45 23 G: 50 26 H: 55 28 2. Hypothetical Committee Systems in a 100-Member Parliament 1 F, 1 H 2G 1 B, 1 D, 1 E 1 A, 1 C, 1 G 2 A, 2 B
1 1 5 1 6 2 1 1
20 ⫻ 1 = 20 25 ⫻ 1 = 25 30 ⫻ 5 = 150 18 ⫻ 1 = 18 40 ⫻ 6 = 240 45 ⫻ 2 = 90 50 ⫻ 1 = 50 55 ⫻ 1 = 55
11 ⫻ 1 = 11 13 ⫻ 1 = 13 16 ⫻ 5 = 80 18 ⫻ 1 = 18 21 ⫻ 6 = 126 23 ⫻ 2 = 46 26 ⫻ 1 = 26 28 ⫻ 1 = 28
2 2 3 3 4
100 100 100 100 100
51ª 52 52 53 54
17
610
320 (52.5% of 610)
17
610
252 b (480 ⫻ 0.525)
3. The Committee System in the 480-Member Japanese HR, 2005 If one committee slot per MP: all A-G (not H), as listed in Panel 1 With multiple slots per MP: all A-G (not H), as listed in Panel 1
Key: FM = formal majority, or simple majority (50% + 1). DM = decisive majority, the majority required to surpass an institutionally defined winning threshold. a FM = DM. b Actual DM in Japanese HR. Source: Modified and adapted from Mukohno 2002, 47, Table 3.1.
Competition for Power in Japan
237
four committees configured as on the last row of Panel 2 would call for a 54-seat decisive majority. In all the above examples—except the first—the DM is larger than the FM. Consequently, considering only the committee system, the number of seats required for a DM is in principle greater than for an FM. Only in special cases does a DM threshold result in the same number as an FM in a parliament.1
9.2.2
In the Japanese Diet
The hypothetical examples show that a DM qualifies as just an important threshold for policy control as an FM in legislatures with committees. However, little scholarly attention has been paid to the concept of a DM as distinct from an FM. In Japan, where one party has long controlled government, the distinction between an FM and a DM looms large. During most of the period from 1955 to 1993, the LDP remained in office without allying itself with other parties. 2 Despite its grip on government, the LDP could be threatened by other parties that could influence the legislative agenda and interrupt and even block government-sponsored bills by taking advantage of the Diet committee system. With only an FM, but no DM, the LDP could not fully control the legislative agenda. During the period of LDP predominance, therefore, observers were interested in whether the LDP had secured a DM, which would allow it to pass legislation though the Diet committee system without relying on legislative cooperation from other parties. In coalition politics since 1993, both thresholds have come to be regarded as measures of the power of the LDP as a core party in a governing coalition. The Diet comprises the House of Representatives (HR) and the House of Councilors (HC), but the analysis here focuses on the HR, since its influence on policies exceeds that of the HC, where the LDP suffered from minority status throughout the period we cover in this chapter. Three points are important to understand the institutional place of committees in the Japanese Diet. First and most important, committees have substantial control over the legislative agenda and bill enactment. All bills are assigned to committees and must pass in committee before being sent to the floor for a vote. Second, committee chairs have agenda-setting power and the casting vote if the committee is evenly split. Hence, a governing party can count on safe passage of all government-sponsored bills if it controls both committee chairs and a majority of seats in all committees. It thus requires a majority
238
Junko Kato and Kentaro Yamamoto
cushion of a few extra seats to counter the power of a chair who is not part of the governing coalition. Third, all HR members must be affiliated with at least one committee, with multiple memberships allowed in the proportional allocation of committee seats. Allocation of chair posts depends on whether chairmanships are considered separately from regular committee seats. Before the 1990s, all committee seats (including chairmanships) were assigned under a proportional rule. When the LDP lost its HC majority in 1989, however, the LDP had to bargain with opposition parties to facilitate passage of bills in both houses. In the interests of compromise demanded by the need to negotiate with a second chamber, the LDP surrendered some of its control over committee agendas. Since 1991, when the LDP and the opposition have agreed on the proportional allocation of chair posts, the proportionality rule has been strictly applied (Kawato 2005, 203). The governing party or coalition, even with a majority on the floor, must always allocate chairmanships proportionally to the opposition. Since a DM emerges as the result of adding committee seat assignments so as to control votes and the agenda in all committees, we use for our real-world Japanese illustration of a DM the 17 standing committees in existence after the merger of ministries and agencies in 2001. As of 2005, there are 480 seats in the HR, and 610 committee positions are open to multiple membership. The bottom panel of table 9.1 indicates how seats are allocated among the 17 committees and how many seats are required to control a majority in each committee. If we were to assume that each MP received only 1 committee position, the figure representing the majority threshold in each committee would add up to the number of seats necessary to attain a DM, that is, 320 (52.5 percent of the total of 610 committee seats). Since multiple committee slots are allowed per MP and committee seats are assigned in proportion to floor seats, the bottom row reports 252 as the decisive majority of HR seats required to ensure that a winning coalition controls all the committees. (For small committees [20 seats in Japan], real-world politics involves rounding up to whole MPs to reach the DM; given such rounding, the DM varies to some degree over time.3) Under the proportional allocation of committee positions (even with rounding), the DM remains as important as ever. The agendasetting power of a committee chair, if granted to the opposition camp, is not strong enough to resist the initiative of the winning coalition. Committee chairs usually follow an agenda set by the Diet steering
Competition for Power in Japan
239
committee in which representatives from the ruling and opposition coalitions negotiate the management of the entire session. Even should an opposition coalition (party) try to take advantage of controlling one of the chairs so as to resist the winning coalition, there is little place for agenda manipulation in the Diet. More important, a coalition with a DM can abandon the proportional allocation of chair posts as long as it is willing to operate in a more confrontational environment. The agreement on proportional allocation has been made on an ad hoc basis and remains precariously in place: opposition parties try to take advantage of their chairmanships, while any government with a DM can threaten to abandon the proportionality rule for committees and revert to simple majority control.4 The DM thus has continued to be important since 1994 under the LDP-centered coalition government. The LDP-centered coalition, therefore, has tried to secure an FM and a DM in elections. Although the LDP gained an FM three times (1997, 2001, and 2003) and a DM once (1999) as a result of inter-electoral party switching, it failed to win an FM in four general elections (1993, 1996, 2000, and 2003). From 1993 to immediately before the 2005 general election, in which the LDP’s landslide victory gave it a majority far exceeding a DM, it was party switching rather than elections that determined whether the governing coalition secured both an FM and a DM.
9.3
A Theoretical Framework for Analyzing Party Switching
This section presents a theoretical framework for the analysis of party switching, focusing specifically on a case in which party switching can be observed on a sufficiently large and frequent scale to tip the balance of power among parties. We begin with three assumptions that link individual legislators’ motivations with party capabilities. The first assumption is in essence an elaboration on the usual rationality assumption. Assumption 1. Legislators switch parties to increase their electoral, office (post), and policy payoffs. Although office, policy, and electoral motivations potentially create conflicting goals, legislators give priority to increased payoffs above all, the consideration of an office being in direct proportion to legislators’ expectations that switching will cause a party to exceed a winning threshold. In a parliamentary system, a party surpassing
240
Junko Kato and Kentaro Yamamoto
a winning threshold in the legislature will, in principle, try to control of the executive—to gain incumbent status. Assumption 2. Incumbency status assures advantages for reelection, office, funding, and policy making among legislators who belong to incumbent parties. Thus, legislators tend to change party affiliations if their switching is expected to endow the party they join with incumbent status.
Both office- and policy-seeking models have particular implications regarding which party will secure incumbent status and thus attract more party switchers. In terms of the office-seeking model, since size matters, the largest party cannot be dominant under majority rule unless it holds more than 25 percent of the seats and the second- and the third-largest parties cannot form a winning coalition without it (Laver and Benoit 2003). The largest party is in a better position to attract switchers when it is close to achieving a legislative majority (Laver and Kato 2001). In a one-dimensional policy space, the policy-seeking model shows that the median party has the advantage, although in multidimensional competition it is often hard to identify a median (core) party. Assumption 3. The party that is close to or exceeds a winning threshold is an attractor that absorbs office-seeking legislators. In party competition in a one-dimensional policy space, policy-seeking legislators tend to switch to a party that is expected to be a median party.
In the context of simple majority rule, securing a majority will change a losing coalition into a winning one. Under some conditions, however, additional seats may be required to ensure consistent government policy control. In such cases, a DM is defined by the threshold for overcoming institutional constraints, particularly those that are committee-related. Although attaining an FM is beneficial, a winning coalition does better if it can increase its size to the DM threshold. 5 A party that approaches either an FM or a DM threshold from below stands to benefit from party switching. We present five hypotheses to examine the relationship between these thresholds and party switching. Hypothesis 1. The DM and FM thresholds should be equally important for structuring the dynamics of legislative competition involving both the formation of a winning coalition and party switching.
The more the power balance among parties is likely to change, the more party switching there is likely to be. Hypothesis 1 holds that
Competition for Power in Japan
241
a DM and an FM both are relevant thresholds. Switching could in principle continue until the party with an FM secures a DM. While an FM party welcomes switchers to gain policy control via a DM, the second-largest party can challenge a larger rival that has attained an FM but not a DM. In terms of the two thresholds, parties have different expectations depending on their relative sizes. Hypothesis 2. When the balance of power among parties is likely to change, party switching should take place and the presence of FM and DM thresholds should tend to intensify and prolong switching. Although a party that attains an FM should continue trying to attract switchers to secure a DM, the second-largest party should try to exploit the largest party’s lack of a DM by attracting enough switchers to deprive it of policy control.
Here, the relationship between switchers and parties is reciprocal. The largest and second-largest parties both try to solicit potential switchers, who seek in essence to sell their services to the highest bidder. Switchers will change to either the largest or the second-largest party with the expectation that the party that has attained an FM will also attain a DM, or that the second-largest party will challenge and block the largest party’s influence. Parties use every opportunity to attract switchers and, should the largest party attain an FM but fall short of a DM, more (less) switching to the second-largest party than to the largest one is expected if the largest party is closer to an FM (a DM) than to a DM (an FM). Uncertainty deriving from the two thresholds makes party switching both prolonged and complicated. Moreover, other incentives can be assumed to influence the outcome as well. The implicit assumption in the office- and policy-seeking models is that all legislators have the same motives: office- (policy-) seeking legislators expect others to switch parties for office (policy) benefits. Legislators who do not know what motivates their peers may switch parties based on selective incentives for which they can form more reliable expectations regarding their own payoffs. One selective incentive is reelection. Hypothesis 3. If legislators cannot form definite expectations regarding legislative competition, then switching should increasingly be motivated by selective incentives. Switching should be likely when legislators believe it will enhance their individual reelection chances.
For example, legislators might switch from one party to another in a bid to be nominated in districts with the most advantageous
242
Junko Kato and Kentaro Yamamoto
conditions. Parties also can capitalize on this contingency. They may make competitive offers to attract switchers and possibly even weaken their discipline so as to accept a large number of switchers. Hypothesis 4. A party may try to diversify its position to attract more switchers by trading off the risk of breaking up the party for an advantage in party switching. As uncertainty over the results of legislative competition increases, the party that is trying to exploit this uncertainty should have an incentive to become more tolerant and diversify the policy positions of its members.
Heller and Mershon (2008) argue, with support from Italian legislative voting data, that switchers tend to defect from highly disciplined to less disciplined parties. The common assumption that party positions are fixed, and that policy-seeking legislators’ decisions to switch hinge on those positions, might not stand up to scrutiny (see also Heller and Mershon this volume, ch. 7). If there is a reciprocal relationship between parties and their legislative members, a party can change (or define) its position to attract or retain more legislative members. As uncertainty increases, the relaxation of discipline could attract switchers. We have so far focused on party switching as a means of changing the balance of power among parties, even though elections usually play a more important role (and typically are assumed to play the only role) in altering the partisan balance. But electoral changes also often influence party switching. Hypothesis 5. The patterns in and frequency of party switching should change following parliamentary elections, since elections shape legislators’ expectations.
A given set of election results significantly influences subsequent party switching not only by determining the first distribution of seats, but also by informing potential switchers of which party enjoys the public mandate. It is thus important to assess the impact of party switching on the balance of power among parties.
9.4 The Dynamics of Party Systems: Mergers, Breakups, New Parties, and Party Switching in Japan From 1955 to 1993, the LDP dominated votes, seats, cabinet posts, and the policy agenda. Before 1991, the LDP’s efforts were directed at
Competition for Power in Japan
243
securing a DM, as it almost always controlled an FM. It failed to win a majority in the HC election in 1989, however. Further, it lost an FM in the HR in 1993, when defections moved it into minority status.
9.4.1
Second-Largest Party Challenges to the LDP, 1996–2003
In the 1993 general election, the LDP was unable to recoup the losses incurred by preelection defections. It was excluded from the subsequent coalition government, remaining out of office for the first time in 38 years. Since 1994, when the LDP returned to office, there have been LDP-centered coalitions or minority governments (as of summer 2008). The LDP’s advantage over other parties now hinges on the extent to and ways in which the second-largest party in opposition—first the New Frontier Party (NFP), then the Democratic Party of Japan (DPJ)—has challenged the LDP. Figure 9.1 shows how the distribution of seats changed from 1993 to 2005. Note that the LDP twice repeated a cycle of decline and restoration, losing its FM and then its DM, and then regaining both, as a result of challenges from the second-largest party. In the first cycle of change, the LDP fell short of an FM in 1993, succeeded in securing an FM in 1997, and then regained its DM before the 2000 general election. The 1996 election represented a turning point, when the LDP fell short of an FM, but was quite close to it, as figure 9.1 displays. This weakened the NFP, which broke Number of Seats in the House of Representatives 300
LDP seats SDP/NFP/DPJ* seats FM DM
250 200
* SDP 6/6/1993–6/30/1994 NFP 1/19/1995–9/29/1997 DPJ 4/27/1998–9/11/2005
150
Election day 7/18/1993 10/20/1996 6/25/2000 11/9/2003 9/11/2005
100 50
6/6/1993 7/18/1993 8/9/1993 4/8/1994 6/30/1994 1/19/1995 1/11/1996 9/27/1996 10/20/1996 1/20/1996 9/29/1997 4/27/1998 1/19/1999 10/29/1999 12/28/1999 4/7/2000 5/26/2000 6/25/2000 7/25/2000 4/27/2001 10/28/2001 3/19/2002 12/31/2002 10/25/2003 11/9/2003 11/22/2003 9/27/2004 4/28/2005 9/11/2005
0
Key Moments Legistature 1993–2005
Figure 9.1 Changes in distribution of seats among parties in Japanese house of representatives, 1993–2005.
244
Junko Kato and Kentaro Yamamoto
apart in 1997. The LDP subsequently absorbed more party switchers, exceeded the FM threshold, and increased its membership to the DM threshold before the 2000 general election. Facing the DPJ in the 2000 and 2003 elections, the LDP failed to win a majority, but garnered an FM by attracting party switchers during the subsequent inter-election periods. In contrast to the 1996–2000 period, however, the LDP failed to attain a DM. Compared to the NFP in 1993–2000, the DPJ in 2000–2005 succeeded in mounting a significant challenge to the LDP’s dominance. In 1994, the single nontransferable vote electoral system for HR elections was replaced by a system that includes single-member districts (SMDs) and proportional representation (PR). Elections under the new system were first held in 1996, then again in 2000, 2003, and 2005. The electoral reform and elections under the new system played a role but party switching became as important an influence as elections in determining the balance of power among parties. To understand why the fortunes of the NFP and the DPJ differed as greatly as they sought to wrest power from the LDP, therefore, it is important to examine the dynamics of party switching.
9.4.2 Party Switching and Office, 1993–1999 During the period from the 1993 general election to immediately before that in 2000, the office-related payoffs expected by prospective switchers were high enough to ensure the LDP’s continuance in office (based on the Shapley-Shubik power index, or SSI, for the LDP; see Laver and Kato 2001).6 In addition to the conventional SSI with an FM that Laver and Kato (2001) use, our analysis employs a modified SSI with a DM threshold.7 Figure 9.2 illustrates the changes in the SSIs with an FM and a DM for both the LDP and the secondlargest party—the Social Democratic Party (SDP), the NFP, and the DPJ from April 1998 to September 2005. The LDP lost its FM as a result of preelection defections in June 1993 and did not regain it in the July general election. This reduced the LDP’s expected payoffs, calculated on the basis of the SSI with an FM, much more than might be expected from its loss of seats. While the LDP’s seat share dropped 16 percent from its FM, its SSI decreased from 1 to 0.632 (as exhibited in figure 9.2) because the party was no longer pivotal in all winning coalitions. Although the LDP remained the largest party, its relative decline meant that a smaller party could enjoy disproportionately large payoffs. For example, in 1995, when
Competition for Power in Japan Value of Shapley-Shubik Index 1 0.9 0.8
LDP SSI with an FM LDP SSI with a DM SDP/NFP/DPJ* SSI with an FM SDP/NFP/DPJ* SSI with a DM
0.7 0.6
* SDP 6/6/1993–6/30/1994 NFP 1/19/1995–9/29/1997 DPJ 4/27/1998–9/11/2005
0.5 0.4 0.3 0.2
6/6/1993 7/18/1993 8/9/1993 4/8/1994 6/30/1994 1/19/1995 1/11/1996 9/27/1996 10/20/1996 1/20/1996 9/29/1997 4/27/1998 1/19/1999 10/29/1999 12/28/1999 4/7/2000 5/26/2000 6/25/2000 7/25/2000 4/27/2001 10/28/2001 3/19/2002 12/31/2002 10/25/2003 11/9/2003 11/22/2003 9/27/2004 4/28/2005 9/11/2005
0.1 0
245
Figure 9.2
Election day 7/18/1993 10/20/1996 6/25/2000 11/9/2003 9/11/2005 Key Moments Legistature 1993–2005
Changes in Shapley–Shubik index for Japanese parties, 1993–2005.
the LDP’s SSI with an FM was 0.367, the SDP’s was as high as that of the NFP, at 0.283. Considering that it had the largest share of seats, the level of the LDP’s payoffs, as measured by the SSI, was disproportionately low. This low level was underlined by the formation of the NFP at the end of 1994, and the NFP’s ability to form a winning coalition with the SDP from 1995 to 1996. In the 1996 general election, the LDP failed to secure a majority but increased its seats to a number much closer to the FM threshold, at the expense of the NFP. Its expected office-related payoff consequently increased on a much larger scale than might be expected from the increase in its seat share, which reduced the expected payoffs of the NFP and other parties. The LDP restored its FM without an election in 1997, as the SSI with an FM indicates, and since then it has dominated other parties in a coalition game. The restoration of the LDP’s FM described earlier is an example of legislators trying to form a winning coalition by crossing party boundaries. Party switching can transform a near-majority party into a majority party, as argued in Laver and Kato (2001). The officeseeking model (with a single FM threshold) explains the dynamic of competition between the LDP and the NFP between 1994 and 1997 and, generally, supports the importance of an FM as a winning threshold in influencing legislative competition on the one hand. On the other hand, however, an office-seeking model cannot explain continued switching after the LDP reached an FM. The LDP continued
246
Junko Kato and Kentaro Yamamoto
to attract switchers, and in 1999 reached the DM threshold, that is, 265 out of 500 seats. Figure 9.3 documents that the expected payoffs to individual members, as expressed by the SSI*100/seats (SSI/s in the figure) with an FM, were stagnant and even declined as the LDP surpassed its DM in 1999 and 2000. This implies that a DM, rather than an FM, is the upper limit of the LDP’s size when it comes to increasing payoffs to individuals to stabilize the party’s majority status.
9.4.3
Party Switching, Policy, and the Survival of the DPJ after 2000
The LDP again lost its majority in the 2000 general election, but its seat share—233 out of 480—was close to a majority. As a nearmajority party, the LDP was able to restore its FM majority before the 2003 general election, just as it had after the 1996 general election. In the 2003 general election, it won 237 out of 480 seats and then restored its FM by accepting switchers. The LDP’s inability to reach a DM was a consequence of the DPJ’s persistent challenge to its position. The DPJ, formed immediately before the 1996 general election, had absorbed many members from the NFP (dissolved in 1997) and increased its membership to a little more than one-third of that of the LDP before the 2000 general election. The DPJ not only increased its seats in both the 2000 and the 2003 general elections, but it also accepted party switchers in numbers comparable to those of the LDP. Hence, the LDP, though surpassing an FM,
(SSI*100)/Seats, by Party. 0.450 0.450 0.350 0.300 0.250 0.200
LDP SSI/seats with an FM LDP SSI/seats with a DM SDP/NFP/DPJ* SSI/ seats with an FM SDP/NFP/DPJ* SSI/ seats with a DM * SDP 6/6/1993–6/30/1994 NFP 1/19/1995–9/29/1997 DPJ 4/27/1998–9/11/2005
0.150 0.100 0.050 6/6/1993 7/18/1993 8/9/1993 4/8/1994 6/30/1994 1/19/1995 1/11/1996 9/27/1996 10/20/1996 1/20/1996 9/29/1997 4/27/1998 1/19/1999 10/29/1999 12/28/1999 4/7/2000 5/26/2000 6/25/2000 7/25/2000 4/27/2001 10/28/2001 3/19/2002 12/31/2002 10/25/2003 11/9/2003 11/22/2003 9/27/2004 4/28/2005 9/11/2005
0.000
Election day 7/18/1993 10/20/1996 6/25/2000 Key Moments Legistature 1993–2005
Figure 9.3 Expected office payoffs to legislators ([SSI*100]/seats), by party, Japan 1993–2005.
Competition for Power in Japan
247
was vulnerable to the DPJ’s challenge during the period leading up to the 2005 general election because it fell short of a DM. The contrast with the preceding period is straightforward: vis-à-vis the NFP from 1996 to 2000, once the LDP approached an FM it was able to attract switchers so as to exceed it, and then go on to a DM. However, faced with the DPJ from 2000 to 2005, the LDP surpassed and then lost an FM threshold instead of moving on to a DM. An office-seeking model based on an FM threshold does not explain this contrast.
9.5 Examination of Hypotheses To explain this puzzling contrast, we frame it in terms of the assumptions and hypotheses presented earlier. A party that is close to a winning threshold attracts office-seeking switchers, and a party (coalition) that is expected to occupy a median position absorbs policy-seeking switchers (Assumption 3). Thus, as expected, the LDP regained its FM after the 1996 general election, in which it secured 239 of 500 seats (47.8 percent, 12 seats short of an FM) and drove the NFP to dissolution. However, contrary to expectations, when the LDP was very close to and then exceeded the threshold for an FM between 2000 and 2005, the DPJ had ample opportunity to challenge the LDP. For example, in the 2000 general election, the LDP secured 233 of 480 seats (48.5 percent, 8 seats short of an FM), but could not cement its dominance in policy making until its unexpected landslide victory in the 2005 snap general election. What factors and conditions enabled the DPJ to compete with the LDP where the NFP could not? To address this problem head on, we consider our earlier hypotheses in light of empirical evidence in the following text. We begin with the relevance of a DM threshold.
9.5.1
Hypothesis 1: The FM and DM as Distinct Winning Thresholds
Our first hypothesis states that there are two thresholds for legislative control. The existence of a DM renders an FM neither irrelevant nor somehow less important, even though a DM clearly is vital for policyfocused legislators. The floor of the HR uses a simple majority rule, in which an FM is an important threshold, and an FM party on the floor may be able to secure a majority on some committees.8 Therefore, reaching an FM certainly gives advantage to the largest party. An FM is not sufficient to guarantee full policy control, however, and a party is invincible only if it also attains a DM.
248
Junko Kato and Kentaro Yamamoto
The dynamics of party switching from the 2000 general election until the 2005 general election support the juxtaposition of FM and DM as relevant winning thresholds. Changes in the SSI point to the evolving balance of power among parties, as figure 9.2 indicates. As a result of the 1996 general election, the LDP’s SSI with an FM increased from 0.485 to 0.610, while that of the NFP declined from 0.165 to 0.093 (See figure 9.2 for details). The 2000 general election left the LDP’s SSI with an FM at 0.833, despite the party’s loss of its majority, whereas the DPJ increased its SSI with an FM from 0.000 to 0.033. The LDP’s power was greater—its SSI with an FM was much closer to unity—in 2000 than it had been in 1996. After the 2000 election, the office-seeking model with a single FM threshold predicts that the DPJ should have been less of a challenge to the LDP than the NFP had been after the 1996 election—that is, the LDP should have attracted more party switchers after the 2000 election than it did in 1996. Nonetheless, the increase in seats for the LDP from accepting switchers was smaller after the 2000 than after the 1996 general election, as figure 9.1 exhibits. Figure 9.2 shows that the LDP’s advantage over the DPJ from attracting switchers is much smaller if measured using the SSI with a DM rather than an FM. The office-seeking model of legislator and party motivations fails to explain the LDP’s difficulty in stabilizing its power, or the DPJ’s attempts to replace the LDP-centered governing coalition in the two inter-election periods between June 2000 and September 2005. A model that takes both thresholds into account and adds policy seeking to the mix is more successful. Consider once more figure 9.3, which shows changes in the SSI—both with an FM and a DM—per legislator. Changes in both SSIs for a party as a whole are similar to those for a legislative member of a party, except in the case of the LDP’s supermajority following its 2005 general election landslide victory, when LDP legislators saw their SSI with an FM decrease. The LDP, apparently conscious of the importance both of the FM and DM, tried to attract enough switchers to exceed both thresholds. To restore its FM, it accepted Naoto Kitamura in September 1997 with the best possible electoral considerations. The LDP incumbent had already run in Kitamura’s prioritized SMD, so the party nominated Kitamura and the incumbent, respectively, in the SMD and PR tiers. As we elaborate in the following text, of the 24 members who switched from the NFP to the LDP between the 1996 and 2000 general elections, Kitamura’s was an exceptional case.9 The LDP also
Competition for Power in Japan
249
made appealing offers to switchers as it sought to exceed the DM threshold. In 2001, having surmounted the FM threshold, it began by accepting switchers from a small conservative legislative group called the 21st-Century Club.10 The LDP again guaranteed prospective switchers nomination in their own election districts in the following election and encouraged their switching. Nine of the ten members of the 21st-Century Club who had won in their SMDs in the 2000 general election subsequently switched to the LDP, which appreciated their incumbent status enough to nominate them in the 2003 general election.
9.5.2
Hypothesis 2: Party Competition over Two Thresholds
Figure 9.1 documents that, in terms of seats, the NFP posed a greater challenge to the LDP between the 1996 and 2000 elections than did the DPJ between the 2000 and 2005 elections. The DPJ could cause more problems for the LDP, however, than could the larger NFP. Facing the NFP, the LDP easily exceeded both the FM and DM thresholds, leaving no room for the NFP to scramble for switchers. The DPJ, by contrast, continued to expand even after the LDP had exceeded the FM threshold and thus prevented the LDP from securing a DM. The existence of two thresholds plays an important role in explaining the NFP’s decline despite its size. Figures 9.2 and 9.3 afford comparisons of the SSI with an FM and with a DM. Between the 1996 and 2000 general elections, both SSIs are equally high for the LDP and equally low for the second-largest party, first the NFP and then the DPJ. Yet from the 2000 to the 2005 general elections, the LDP’s SSI with a DM is lower than that with an FM, while the DPJ’s SSI with a DM is higher than that with an FM. This implies that the second-largest party in the latter period (DPJ) could contend for power against an LDP that had exceeded the FM, but not the DM in a way that the NFP could not. The evidence from 2000 to 2005 demonstrates that the secondlargest party was able to capitalize on the coexistence of two winning thresholds and challenge the largest party. Securing an FM is one thing, but securing a DM is quite another. As long as the secondlargest party can compete, in terms of absorbing switchers or retaining members, against the largest party with an FM but no DM, the ability to hold onto an FM cannot be the end of the story. As the case of the NFP illustrates, however, the coexistence of the two thresholds does not always give the second-largest party an advantage over the FM party. The factors and conditions that distinguish
250
Junko Kato and Kentaro Yamamoto
the cases of the NFP and DPJ are outlined in the following three hypotheses that, in turn, examine reelection incentives for switching, the impact of loose party discipline on switching, and the impact of elections on the distribution of power among parties.
9.5.3 Hypothesis 3: Motivation for Switching Because reelection is a prerequisite for acquiring office and implementing policy, legislators should be reluctant to switch parties if they believe doing so might put them at a disadvantage in terms of reelection (e.g., they might be forced to move from their prioritized respective election districts). The new system introduced for HR elections underscores the importance of reelection incentives in party switching. In the mixed system of SMDs and PR that was first used in the 1996 general elections, a party nominates a single candidate per SMD. Legislators who are not nominated in their prioritized SMDs move to the party’s PR lists. Even though the party often compensates such candidates with a higher list position, incumbents prefer to run in SMDs where they have organized support and have secured personal votes that they might have been inherited from a medium-sized district under the former electoral system. Since the LDP, as the largest party, has more SMD (and former district) incumbents and candidates, the prospects for being nominated in prioritized districts are lower for legislators switching into the LDP than into other parties. For this reason, the second-largest party may be able to attract switchers by making better offers for SMD nomination; alternatively, the LDP may plausibly seek a compromise with prospective switchers on the basis of this electoral constraint. Electoral rules determine the range of options parties have for making offers to potential switchers. Since the Japanese system allows candidates nominated in SMDs to be listed simultaneously in a PR list, dual-listed candidates who lose in SMDs may find their fortunes revived as PR winners if their ranking on the party list is high enough to ensure they are seated.11 This system is a double-edged sword for potential switchers. While on the one hand it allows for compromise between switchers and incumbents who compete for nomination in the same SMDs, on the other hand switchers may face incumbents whom they defeated in SMDs but who have been elected to PR positions in their new parties. For nominating switchers, the system gives parties the choice of applying what we term an advantageous, an impartial, or a
Competition for Power in Japan
251
disadvantageous strategy. The advantageous strategy involves accepting switchers’ demands. The impartial strategy is to nominate switchers who have won in their SMDs in the latest election, even at the expense of former party nominees who have lost in those districts and have been elected to PR posts. The most disadvantageous strategy is to put longer-term party members before switchers, regardless of the latest election results. Table 9.2 summarizes the election district changes made by switchers from the dissolved NFP to the LDP and the DPJ, noting the presence or absence of rival incumbent candidates in their prioritized SMDs. Since the NFP did not generally nominate candidates both in SMDs and for PR in the 1996 election, there were no SMD losers who were elected on PR lists (NFP 1996 “PR revival winners”).12 Thus, the NFP switchers who had a vested interest in nomination to the district had all been SMD winners in 1996. The NFP 1996 district winners switched to the DPJ rather than to the LDP when they found that both the LDP and DPJ had incumbent PR revival winners in the same districts (see rows 3 and 8 in table 9.2). Such an inclination for NFP defectors to switch into the DPJ can likely be attributed to the DPJ’s and LDP’s different nominating strategies for inswitchers. While both parties prioritized their incumbents over switchers, the LDP imposed more uncertainty on switchers than did the DPJ. For example, in districts where neither of the parties had winners (as shown in rows 5 and 10), all 12 NFP switchers to the DPJ ran in SMDs, but 5 of the 13 switchers to the LDP gave up SMDs to new nominees. The DPJ, thanks to its more favorable nomination strategy, ultimately attracted more switchers than the LDP. Accordingly, as exhibited on rows 1 and 6 of table 9.2, where there were only LDP incumbents who had lost in the same districts as NFP switchers but had been revived to win in PR, all eight NFP switchers to the DPJ gained nomination for SMD races. Where there were only DPJ incumbents who had lost in the same SMDs but had been revived to win in PR, one switcher to the LDP earned an SMD nomination but three were moved to PR, as displayed on rows 2 and 7. The DPJ was not always so kind to switchers. Table 9.3 summarizes the changes and continuities in ballot location and electoral performance for switchers from the Liberal Party (LP) to the DPJ, drawing comparisons between 2000 and 2003. The DPJ’s insistence on giving priority to winners from SMDs disadvantaged switchers from the LP. Thus, for example, as portrayed on row 10 of table 9.3, the DPJ SMD incumbent pushed onto the PR list one inswitcher from the LP
Table 9.2 Switchers from the NFP to the DPJ and the LDP between 1996 and 2000: Switchers’ performance in 2000 elections, by opponents’ performance in 1996 and 2000 elections and by type of ballot Direction of Switch and Switchers’ Runs and Wins 2000 NFP in 1996 to DPJ in 2000 Ballot Location for NFP Switchers Presence in 2000 of Losing 1996 Nominees from Same District A
SMD 1996 and SMD 2000
1 2 3 4 5
LDP PR revival winners DPJ PR revival winners LDP and DPJ PR revival winners Other party PR revival winners Neither LDP nor DPJ PR revival winners Subtotal
B
SMD 1996 and PR 2000
6 7 8 9 10
LDP PR revival winners DPJ PR revival winners LDP and DPJ PR revival winners Other party PR revival winners Neither LDP nor DPJ PR revival winners Subtotal PR 1996 and PR 2000 PR 1996 and SMD 2000 Total
C D
NFP in 1996 to LDP in 2000
N Winners 2000 T Running 2000
SMD
8 2 1 1 12 24
2 2 1 1 11 17
0 0 1 0 0 1 4 7 36
PR revival
4
1 5
1
0 1 18
4 2 12
N winners 2000 T Running 2000
SMD
2 1 0 2 8 13 0 3 0 0 5 8 2 1 24
Note: Rows omit losers. Since NFP candidates were not as a rule dual-listed, NFP PR revival winners do not appear in the table.
PR revival
2
2 6 10
0
3
0 10
5 8 2 0 10
Competition for Power in Japan
253
Table 9.3 Switchers from the LP to the DPJ between 2000 and 2003: switchers’ performance in 2003 elections, by opponents’ performance in 2000 and 2003 elections and by type of ballot Switchers’ Runs and Wins 2003 LP in 2000 to DPJ in 2003 Ballot Location for LP Switchers Presence in 2003 of Losing 2000 Nominees from Same District A
SMD 2000 and SMD 2003
1 2 3 4 5 6 7
LDP winners LDP PR revival winners DPJ winners DPJ PR revival winners Other party winners Other party PR revival winners No other winners Subtotal
B
SMD 2000 and PR 2003
8 9 10 11 12 13 14
LDP winners LDP PR revival winners DPJ winners DPJ PR revival winners Other party winners Other party PR revival winners Neither LDP nor DPJ PR revival winners Subtotal PR 2000 and PR 2003 PR 2000 and SMD 2003 SMD 2000 and Other District 2003 Total
C D E
N Winners 2003 T Running 2003
7 1 0 0 2 0 3 13
SMD
PR revival
1 1
5
1 3 5
6
0
0
1 1 7
1 7
0 0 1 0 0 0 0 1 0 1 3 18
Note: Rows omit losers.
who subsequently lost due to a low list ranking. Of three LP switchers moved by DPJ leaders to new SMDs in 2003 (row E), two yielded nomination in their home districts to DPJ incumbents, and one had to move even though there was no DPJ incumbent in the SMD. Except for the seven switchers who faced LDP incumbents elected in the same election districts (row 1), it thus seems unlikely that LP switchers moved to the DPJ for electoral considerations.13 This point deserves particular attention, as these moves underscore the DPJ’s success in attracting switchers between the 2000 and 2005 general elections.
254
Junko Kato and Kentaro Yamamoto
More importantly, the NFP ought to have been able to make better offers than the DPJ to switchers. The NFP had far fewer SMD candidates than did the DPJ and could therefore more easily have avoided conflicts between incumbents and potential switchers over SMD nomination. This should have enabled the NFP to attract more switchers than the DPJ from the LDP. For example, in the 1996 election, the NFP presented candidates in 235 out of 300 SMDs (96 of whom—40.85 percent—won), while the LDP fielded 288 candidates, of whom 169 (59.68 percent) won. In the 2000 election, the DPJ backed 80 successful candidates out of the 242 (for a 33.06 percent success rate) who ran under its label, while the LDP contested 271 districts and won in 177 (65.31 percent). In the 2003 election, the DPJ contested 267 districts and won in 105 (39.33 percent), compared to the LDP’s 267 candidates, of whom 168 (62.92 percent) emerged victorious.14 The cultivation of reelection incentives for switchers may help the second-largest party contend for power against the largest party. The DPJ has not always played the game of offering electoral incentives to switchers, however. More generally, it is not obvious that reelection incentives adequately explain the distinct fortunes of the NFP and the DPJ in their contest against the LDP.
9.5.4 Hypothesis 4: Policy Diversity within a Party The DPJ’s tolerance of ideological and policy diversity is another factor that likely has distinguished it from its predecessor, the NFP, and has enabled it to prevent the LDP from achieving stable dominance. Figures 9.4 through 9.7 present the left-right ideological positions of all parties, as well as the distribution of policy positions of the LDP and the second-largest party (NFP and DPJ) on the same onedimensional scale, based on the results of expert surveys after the 1996, 2000, 2003, and 2005 general elections.15 The DPJ has an ideological position distinct from that of the NFP. In 1996, the NFP was just as conservative a party as the LDP and was located at the most conservative end of the left-right policy dimension, on which the DPJ could be a median party if the LDP-centered governing coalition,16 to the right of the DPJ, failed to reach a majority threshold.17 More specifically, figures 9.5, 9.6, and 9.7 show that although the DPJ occupied the center position in the left-right dimension, the party spread out its position in policy dimensions from left to right. The DPJ balanced a more leftist position in one policy against a more conservative
Left-Right Position of All Parties LDP (15.35)
NPH (10.75)
L(0)
JCP (2.75)
SDP (7.25)
DPJ (9.28)
SP (14.07)
R(20)
NFP (16.15)
LDP’s Policy Positions X3 (7.63)
X5 (11.49)
X4 (10.69)
X7 (11.61)
X4 (7.94)
X2 (11.93)
X3 (14.35) X10 (15.65) X6 (17.86) X1 (14.86) X9 (18.26)
X2 (13.75)
X10 (16.14)
NFP’s Policy Positions
X1 (13.64)
X6 (16.85)
X8 (10.32) X7 (12.04) X3 (13.92) X5 (16.37)
X9 (17.74)
Ten Policy Positions : X1: Social X2: Citizen’s Right X3: Environment X4: Decentralization X5: Deregulation X6: National Identity X7: Deficit Bonds X8: Economic (Spending vs. tax) X9: U.S. Affairs
Figure 9.4
X10: Defence
Comparison of party policy positions in Japan: The 1996 expert survey.
Left-Right Position of All Parties LDP (15.08) LP (16.89) R(20)
L(0)
JCP (2.98)
SDP (5.24)
DPJ (9.53) CGP (11.91)
NCP (15.89)
LDP’s Policy Positions X7 (6.69) X8 (8.52)
X5 (10.29) X4 (11.61)
X2 (14.21) X10 (15.98) X6 (17.70) X1 (12.93)
X1 (14.99) X9 (17.78)
DPJ’s Policy Positions X2 (3.83) X1 (4.96) X4 (4.04)
X10 (8.84) X3 (7.44)
X8 (10.25) X6 (10.26)
X7 (12.90) X9 (13.33) X5 (13.14)
Ten Policy Positions : X1: Social X2: Citizen’s Right X3: Environment X4: Decentralization X5: Deregulation X6: National Identity X7: Deficit Bonds X8: Economic (Spending vs. tax) X9: U.S. Affairs
Figure 9.5
X10: Defence
Comparison of party policy positions in Japan: The 2000 expert survey.
Left-Right Position of All Parties CGP (12.22)
L(0)
JCP (3.10)
SDP (5.12)
LDP (15.81)
DPJ (11.59)
R(20)
NCP (16.71)
LDP’s Policy Positions X7 (9.54)
X4 (10.57)
X8 (10.14)
X2 (14.37) X1 (15.77) X9 (17.40)
X5 (12.95) X3 (15.02) X10 (17.00) X6 (17.67)
DPJ’s Policy Positions X2 (8.50)
X3 (10.26)
X5 (12.71)
X1 (8.89) X8 (10.61) X10 (11.57) X6 (14.43) X9 (10.67) X7 (11.86)
X4 (5.03)
Ten Policy Positions : X1: Social X2: Citizen’s Right X3: Environment X4: Decentralization X5: Deregulation X6: National Identity X7: Deficit Bonds X8: Economic (Spending vs. tax) X9: U.S. Affairs
Figure 9.6
X10: Defence
Comparison of party policy positions in Japan: The 2003 expert survey.
Left-Right Position of All Parties CGP (12.38) Daichi (14.28) LDP (16.33) R(20)
L(0)
JCP (2.90) SDP (4.43)
DPJ (11.65) NJP (12.95) PNP (15.86)
LDP’s Policy Positions X4 (8.28)
X7 (12.13) X1 (14.51) X5 (15.62) X8 (11.38)
DPJ’s Policy Positions
X9 (17.54)
X3 (14.28) X6 (16.62) X10 (17.04) X2 (14.39)
X8 (10.26) X5 (12.38) X6 (14.13) X4 (5.47)
X1 (9.45) X3 (9.74) X9 (11.63) X10 (12.57) X2 (9.85) X7 (12.72)
Ten Policy Positions : X1: Social X2: Citizen’s Right X3: Environment X4: Decentralization X5: Deregulation X6: National Identity X7: Deficit Bonds X8: Economic (Spending vs. tax) X9: U.S. Affairs
Figure 9.7
X10: Defence
Comparison of party policy positions in Japan: The 2005 expert survey.
Competition for Power in Japan
257
one in another, an approach expected to cultivate incentives to switch to the DPJ among those legislators whose position on a high-priority policy is close to that of the DPJ, regardless of their overall ideological differences from the DPJ on the left-right dimension. The different policy positions of these parties correspond to their different memberships, particularly with respect to the source of their switchers and should also be interpreted in light of the parties’ leadership. The NFP was formed from parties that had participated in a non-LDP governing coalition, but it excluded the SDP and the Harbinger Party. Exclusion of the SDP underscored the impression that legislators affiliated with the NFP had policy positions that ranged from conservative to center, but that did not stretch beyond the center. This is in sharp contrast to the DPJ, which was formed by selectively accepting switchers from the SDP, especially moderate leftists. The promotion of former SDP members to leading positions has helped stretch the policy positions of DPJ legislators, at the same time that the DPJ has striven to maintain the balance of power among intraparty groups, as switchers arriving from different parties have competed for leadership. This approach contrasts sharply with that of NFP management, whose leadership was dominated by conservative defectors from the LDP who made NFP policies ideologically similar to those of the LDP. After the NFP broke up in 1997, many of its members switched to the DPJ. Before the 2003 general election, those who had served as leaders inside the NFP joined the DPJ via the LP.18 That said, entrants from the NFP have not made the DPJ more like the NFP in either ideological or management terms. By comparing figure 9.5 with figure 9.6, which shows the expert survey following the 2003 general election held approximately six weeks after the LP members’ switch to the DPJ, one can see that the DPJ’s position on the left-right dimension moved only slightly to the right and remained centrist. Ideological tolerance, loose discipline, and double thresholds make the DPJ more attractive to switchers, although at the cost of intraparty disharmony and the recurrent risk of schism. A snapshot of the party’s membership, as in table 9.4, illustrates the point. The DPJ of 2003 included 38 (21.7 percent of the total) former NFP members (of whom 10 switched first to the LP and then to the DPJ), almost 3 times as many as the 14 former SDP members (8.0 percent), whereas 107 members (61.7 percent) had belonged to the DPJ from the start. Switchers constitute a higher proportion of total membership among more experienced members (who have served more than 4 terms), however, with
Table 9.4
DPJ legislators in October 2003: Inswitchers and loyalists, by terms of service in diet DPJ Inswitchers From:
Terms of Service in Diet
NFP (%)
SDP (%)
OtherParties (%)
DPJ Loyalists (%)
Total
≤3 4 5 6 ≥7 Total
18 (13.6) 5 (38.5) 6 (35.3) 1 (33.3) 8 (80.0) 38 (21.7)
0 3 (23.1) 9 (52.9) 1 (33.3) 1 (10.0) 14 (8.0)
7 (5.3) 5 (38.5) 2 (11.8) 1 (33.3) 1 (10.0) 16 (9.1)
107 (81.1) 0 0 0 0 107 (61.1)
132 13 17 3 10 175
Note: Percentages (in parentheses) sum on rows to 100%.
≥ 4 terms NFP 20 SDP 14
43
≥5 NFP 15 SDP 11
≥6 NFP 9 SDP 2
30
13
Competition for Power in Japan
259
fully 32 percent (14 of 43) hailing from the SDP and 46.5 percent (20), from the NFP. Allowing diverse policy positions helps the DPJ strive to contend with the LDP and distinguishes it from the NFP.
9.5.5
Hypothesis 5: The Impact of Elections
Party switching serves to change the balance of power among parties between elections, but the latest election will determine the distribution of seats among parties and, sometimes, even change the subsequent pattern of party switching. A clear-cut example is the contrasting fortunes of the LDP after the 1993 and the 1996 general elections. The LDP failed to regain the seats it had lost through preelection defections in 1993, and its seat share was far from an FM (223 out of 511, or 43.64 percent). Inter-election defections eroded the party’s size to 206 out of 493 (41.78 percent).19 The LDP failed to restore its FM in the 1996 general election, but it was sufficiently large, with 239 out of 500 seats (47.8 percent), to attract switchers that enabled it to exceed the FM threshold without an election in 1997 (Laver and Kato 2001). Again, however, the distribution of power determined by both the 2000 and 2003 general elections facilitated switching to both the DPJ and LDP. These contrasting consequences show that party switching to the largest party between elections is influenced not only by the number of seats won in the prior election, but also by the fluctuation in the number of seats won and lost over the course of several elections. While the NFP’s seats declined in number and those of the LDP remained almost the same (an increase of one seat) in the 1996 general election, the DPJ saw a drastic increase in its seat count in the 2000 and 2003 general elections at the expense of the LDP, as figure 9.1 displays. Ironically, however, the 2005 general election most explicitly demonstrated the effect of elections on party switching. The LDP won 296 out of 480 seats—the second-highest seat share recorded since the end of World War II—and commanded a super majority of twothirds of the seats when combined with the 31 seats of the allied Clean Government Party (CGP). The effects of a simple majority rule in the SMDs contributed to this landslide victory, giving the LDP control of almost 4 times as many seats as the second-largest party, the DPJ, although the LDP had garnered only 30 percent more votes than had the DPJ. This caused the DPJ’s advantage gained in inter-electoral switching to evaporate, while the LDP’s victory had nothing to do with accumulation of gains and losses in party competition since 1993.
260
Junko Kato and Kentaro Yamamoto
The LDP’s good fortune was a result of Prime Minister Jun’ichiro Koizumi’s effective politicization of the privatization of the postal service and of his bold decision to dissolve the HR despite opposition from his own party.
9.6 Conclusion Since 1993, large-scale party switching among Japanese legislators has brought about breakups, mergers, formations, and dissolutions of parties. During this process, the LDP, which had been the dominant party from 1955 to 1993, failed to consolidate its position to the same degree as it had previously. The LDP’s inability to regain its former dominance highlights the existence and importance of a threshold above an absolute majority (FM)—an institutionally derived decisive majority (DM) that it had easily and consistently attained during the period of one-party dominance from 1955 to 1993. The NFP’s dissolution in 1997 underscores the advantage of a near-majority party (the LDP) in attracting switchers at the expense of other parties. On the other hand, the increasing size of the second-largest party (the DPJ) between the 2000 and the 2005 general elections apparently offset the LDP’s largest-party advantage, even as the LDP approached and then surpassed the FM threshold. Our focus on the complex dynamics of party switching, in conjunction with the distinction between the FM and DM thresholds, allows us to tackle this puzzle head on. It also sheds new light on whether parties (and legislators) are office seekers or policy seekers. Further, the two thresholds underpin the emergence of uncertainty over the result of the competition between the largest and the second-largest parties. A variety of factors, including reelection incentives, the extent of party discipline, and election results, explain the party switching of legislators who cope with uncertainty. In particular, the DPJ’s flexible definitions of policy positions, together with the seats it won in elections, helped it compete against the LDP more successfully than its predecessor, the NFP, ever had.
9.7
Notes
1. For example, as table 9.1 suggests, if all legislators are required to belong to a committee but multiple committee memberships are proscribed, an FM would be sufficient for a DM in a parliament with an even number of seats and two odd-numbered committees. 2. Technical exceptions from the LDP’s one-party rule came in 1983–1986, when the LDP governed with its own splinter party, and in the 1970s, when
Competition for Power in Japan
3.
4.
5. 6. 7. 8.
9.
10.
11.
12.
261
the LDP let conservative independents join in its legislative group immediately after an election. Figure 9.1 reports the FM and DM at specific times between June 6, 1993, and November 11, 2005, based on the size of the HR and the total number and sizes of committees. For details and discussion, see Mukohno (2002, 47–50, 66–70). The LDP has successfully repressed opposition attempts to control the agenda by securing the chairmanship of important committees on several occasions. For example, the DPJ had to back down on its insistence on getting a chair post of the HC budget committee because it feared that the LDP would deprive it of the chairmanship of the HR budget committee in retaliation (Asahi newspaper, September 8, 2007). The discussion and argument here remain the same whether we focus on single parties or multiparty coalitions. Based on the logic presented by Shapley and Shubik (1954), the SSI represents a proportion of all the winning coalitions that a party joins. This modified index is easily obtained by changing the winning threshold in the original algorithm. As table 9.1 illustrates, an absolute majority in each odd-sized committee translates into a DM overall, provided that the legislature is composed only of odd-sized committees. (See the first row of Panel 1 in particular.) This implies that in a Diet with 480 seats, attaining an FM allows a winning coalition to control a majority in committees with odd numbers of members, but not in committees with even numbers. The top row of table 9.2 shows that two members had switched from the NFP to the LDP and received LDP nominations for the 2000 general election, although other LDP legislators had been elected in the switcher’s home single-member district in the 1996 general election. One of the switchers was Kitamura. The other, Ichizo Miyamoto, switched and garnered the SMD nomination because the LDP incumbent, Kenzaburo Hara, was expected to retire before the 2000 general election. Although party switching was critical to the LDP in regaining its majority, two supplementary elections in October 2001 added two members to help the party cross the threshold. A chronology of the change in LDP seats clarifies this process. At the end of 2000, the LDP attained 239 seats by accepting five switchers from the 21st-Century Club. Then, following the death of two of its HR members in July and September 2001, the LDP was reduced to 237 members. Its membership then increased to 239 thanks to two switchers from the 21st-Century Club in October 2001. Later, the LDP won both of the supplementary elections held to fill the two vacancies just cited and so restored its majority to 241 members. Another characteristic of the Japanese system is that parties are permitted to give more than one dual-listed district candidate the same ranking on the party PR list. Among candidates with the same ranking, those who win a larger share of SMD votes are given seats, unless the party has already filled its winning seats. There was an exception to this rule. Two NFP candidates who ran in the same districts as the LDP leaders were dual-listed in SMDs and PR (and lost
262
13.
14.
15.
16.
17.
18.
19.
Junko Kato and Kentaro Yamamoto in the SMDs but won in PR). Neither of these two switched to the LDP or the DPJ, and thus they do not appear in table 9.2. Since most LP members formerly belonged to the LDP, one may speculate that they were reluctant to return to the party from which they had defected. Since 1993, however, there have been frequent occasions on which switchers have returned to their original party, in most cases the LDP. In 2003, the DPJ had the same number of candidates as the LDP, although the DPJ’s success rate was lower than that of the LDP. This implies that, while legislators were less likely to face incumbents elected in SMDs when switching to the DPJ rather than to the LDP, both the DPJ and LDP had almost the same number of candidates wanting to run in SMDs. For recent surveys, see Laver and Hunt (1992) and Huber and Inglehart (1995). For a detailed report on the 1996 expert survey, on which figure 9.4 is based, see Kato and Laver (1999a, 1999b); for a report on the 2000 one on which figure 9.5 is based, see Kato and Laver (2003); for a report on the 2003 one on which figure 9.6 is based, see Benoit and Laver (2005). The survey data including the one in 2005 is found at http://www.j.u-tokyo. ac.jp/~katoj/. Since 1994, the LDP has changed its partners from the Japanese Socialist Party (JSP) and the Harbinger Party, to the Liberal Party (LP), the Clean Government Party (CGP), and the New Conservative Party (NCP). When the LP abandoned its alliance with LDP defectors, the party was broken up and defectors from it formed the NCP. The LP’s remaining members subsequently joined the DPJ, and the NCP with the LDP; consequently, as of the time of writing (July 2008), the LDP has allied itself with the CGP (Komeito in Japanese). The governing coalition of the LDP (plus the NCP) and the CGP secured a majority (FM as well as DM) in the 2000 general election and has maintained it up to the time of writing (July 2008). Considering that the LDP alone did not always secure a majority during this period, the CGP, rather than the DPJ, should have absorbed party switchers. It failed to attract switchers because it relies heavily for electoral support on the religious organization Soka Gakkai. Although the party officially has no organizational connection with Soka Gakkai, legislators affiliated with the CGP inevitably rely for electoral support on this organization. This could make prospective switchers more reluctant to make the switch. According to the expert survey conducted after the 2000 general election, the LDP’s position on the left-right dimension (with the extreme left position assigned 1 and the extreme right, 20) was 15.08. The LP’s position was more conservative, that is, 16.89, whereas the DPJ’s was 9.53. This immediately raises the question of whether those LP members who switched to the DPJ were totally indifferent to policy. Consistent with Hypothesis 4, LP switchers may have considered ideological and policy differences as an acceptable cost of joining the loosely disciplined DPJ that they expected would come to power. As figure 9.1 shows, the seat total varies slightly during the course of a term due to vacancies.
Competition for Power in Japan
263
9.8 References Benoit, Kenneth, and Michael Laver. 2005. “Estimating Party Policy Positions: Japan in a Comparative Context.” Japanese Journal of Political Science 6 (2): 187–209. Einy, Ezra. 1985. “On Connected Coalitions in Dominated Simple Games.” International Journal of Game Theory 14 (2): 103–125. Heller, William B., and Carol Mershon. 2008. “Dealing in Discipline: Party Switching and Legislative Voting in the Italian Chamber of Deputies, 1988– 2000.” American Journal of Political Science 52 (4): 910–925. Huber, John, and Ronald Inglehart. 1995. “Expert Interpretations of Party Space and Party Locations in 42 Societies.” Party Politics 1 (1): 73–111. Kato, Junko, and Michael Laver. 1999a. “Theories of Government Formation and the 1996 General Election in Japan.” Party Politics 4 (2): 229–252. ———. 1999b. “Party Policy and Cabinet Portfolio in Japan, 1996.” Party Politics 4 (2): 253–260. ———. 2003. “Policy and Party Competition in Japan after the Election of 2000.” Japanese Journal of Political Science 4 (1): 121–133. Kawato, Sadafumi. 2005. Nihon no kokkai seido to suitor siege (The Japanese Diet and Party Politics). Tokyo: University of Tokyo Press. Laver, Michael, and Junko Kato. 2001. “Dynamic Approaches to Government Formation and the Generic Instability of Decisive Structures in Japan.” Electoral Studies 20 (4): 509–527. Laver, Michael, and Kenneth Benoit. 2003. “Evolution of Party Systems between Elections.” American Journal of Political Science 47 (2): 215–233. Laver, Michael, and W. Ben Hunt. 1992. Policy and Party Competition. New York: Routledge. Mattson, Ingvar, and Kaare Strøm. 1995. “Parliamentary Committees.” In Parliaments and Majority Rule in Western Europe, ed. Herbert Döring, 249–307. New York: St. Martin’s Press. Mukohno, Shinji. 2002. Shugiin—sono shisutemu to mechanizumu (The Japanese House of Representatives: System and mechanism). Tokyo: Toshindo. Peleg, B. 1981. “Coalition Formation in Simple Games with Dominant Players.” International Journal of Game Theory 10 (1): 11–33. Shapley, L.S., and Martin Shubik. 1954. “A Method for Evaluating the Distribution of Power in a Committee System.” American Political Science Review 48 (3): 787–792. van Deeman, A. M. A. 1989. “Dominant Players and Minimum Size Coalitions.” European Journal of Political Research 17 (3): 313–332. van Roozendaal, Peter. 1992. “The Effects of Dominant and Central Parties on Cabinet Composition and Durability.” Legislative Studies Quarterly 17 (5): 5–36.
This page intentionally left blank
10 Party Switching, Party Systems, and Political Representation* Marcus Kreuzer and Vello Pettai
10.1 Introduction The literature on parties and party systems assumes that electoral competition and intraparty policy deliberations are the two primary mechanisms for making parties responsive to public opinion. By implication, this suggests that little else happens between elections to keep politicians responsive to voters. The recent work on party switching, however, has begun to analyze how the reaffiliations of politicians that occur between elections constitute a third, and so far largely overlooked, mechanism of political representation (Kato 1998; Laver and Benoit 2003, 215–216; Zielinski, Slomczynski, and Shabad 2005). The growing literature on party switching has analyzed its import for representation in various ways. Most scholars have concentrated on its representational consequences in the legislative arena and legislative-executive relations by analyzing how switching affects committee assignments, roll-call voting, coalition formation, and ministerial portfolio assignments (Desposato 2006; Grose and Yoshinaka 2003; Heller and Mershon 2005; Mershon and Shvetsova 2008; Nokken 2000). A smaller number of scholars have investigated how party switching affects the larger dynamics of party systems. This work focuses on the complex interactions among party switching, electoral competition, and intraparty deliberations so as to highlight the role of party switching in the overall system of political representation (Kato 1998; Kreuzer and Pettai 2003; Laver and Benoit 2003; Mainwaring 1999; Zielinski, Slomczynski, and Shabad 2005; cf. Kato and Yamamoto and Schofield, this volume).
266
Marcus Kreuzer and Vello Pettai
The goal of this chapter is to explore more fully the systemic effects of party switching. The second section draws an important conceptual distinction, differentiating among forms of party switching and levels of party system institutionalization. Our study rests on the broad assumption that context matters for switching—more specifically, for the frequency of switching, the type of switches, and the consequences of switching—and that the most important element of the political context is the level of party system institutionalization. Thus, switching and its implications are not uniform across the universe of party systems, but vary according to different background conditions. In the third section, we elaborate a set of hypotheses about how the level of party system institutionalization affects the frequency and forms of party switching. In the fourth and most important section, we articulate hypotheses about how the effects of party switching on party systems vary with the level of party system institutionalization. Throughout this chapter, we provide evidence to support the plausibility of our hypotheses. We do not, however, offer a systematic test. Such a test would be difficult for a number of reasons. First, the available cross-national and longitudinal data to get at the relationship between party system institutionalization and switching are still very limited. Second, the endogeneity problem makes reliable testing difficult. Party switching is influenced by numerous other contextual elements that include institutions, historical legacies, socioeconomic factors, media, and resources. It would be virtually impossible to control for or examine all of these factors. Third, many of these influences on party switching affect party systems independent from their impact on switching, thus posing another nettlesome set of methodological problems. Given these formidable challenges, we underscore that the primary purpose of this chapter is not to depict party switching as an explanation for party system change. Rather, we emphasize the importance of switching as a vehicle for change in the hope that future studies of party system change might consider it more systematically.
10.2 Differentiating Party Switching Strategies and Their Contexts The existing work on party switching insufficiently distinguishes among the different forms of party switching and the contexts in
Switching, Party Systems, Representation
267
which they occur. We address this deficiency and thereby lay the ground for a better understanding of the broader systemic effects of party switching. Our discussion revisits the typology of party switching introduced in chapter 1 here and identifies how it might be adopted in the analysis of party system change.
10.2.1
Types of Switches
Party switching is part of a larger repertoire of strategies that politicians use between elections to influence election outcomes. This broader repertoire includes forming temporary electoral alliances, altering electoral procedures, controlling entry of new actors, and coordinating vote-getting efforts across districts. What these strategies have in common is that they are carried out by politicians, take place between elections, and seek to make the existing supply of votes count more effectively (Cox 1997; Kreuzer forthcoming). Party switching belongs to this larger category of what Gary Cox loosely calls “coordination strategies.” We want to focus more narrowly on party switching, however, and disaggregate it into the four basic types outlined in figure 10.1.1 Figure 10.1 displays six organizational outcomes (represented by the double-lined boxes I–VI) and the combination of actors and their choices that lead to each outcome. For ease of exposition, we set aside the size of the party system, which would affect the likelihood of taking one path or another, but not the underlying logic of the available choices: • act collectively to form a start-up party (II), • break away collectively in an instance of party fission to create one or more new parties (III), • defect individually and hop2 to another existing party (IV), or • merge two or more existing parties to form a fusion party (V).
Two points of clarification are in order here. First, in new democracies start-up parties involve novice politicians who are partyless and thus by definition cannot switch. In established democracies, however, political entrepreneurs may found a start-up that attracts already affiliated candidates or legislators from multiple parties; with inswitchers from multiple parties present, such a party would not qualify as an instance of fission.3 Hence we treat all start-ups as including switchers because a new party offers the opportunity for new entrants to forgo affiliation with an existing party and thereby switch away from the
268
Marcus Kreuzer and Vello Pettai
t1: First Time Entry
Political Actors
Affiliate
Do Not Affiliate
Partyless Members
Party Members
No Coordination
Coordination
I. Indep. Entrep.
II. Start-Up Party
Stay Put
Disloyal Members
Loyal Members
Do Not Stay Put
t2: Repeat Entry
Coordination
III. Fission Figure 10.1
No Coordination
IV. Party Hopping
Merge
V. Fusion
Do Not Merge
VI. Org. Status Quo
Types of party switching.
status quo.4 Second, among the four switching outcomes listed earlier, three (II, III, and V) are collective modes of switching, while one (IV) is individual. Among the three collective modes, coordination among individual switchers can result from bandwagoning, formal decisions taken by factions, or a leader strong-arming his followers. We assume that all four switching strategies are voluntary; we thus omit the kind of rule-driven switches discussed in chapter 1.
10.2.2 Levels of Party System Institutionalization Of the many ways to distinguish party systems, the level of institutionalization is most relevant for analyzing party switching. The party literature differentiates between underinstitutionalized, institutionalized, and overinstitutionalized party systems.5 Different authors use different criteria for evaluating levels of institutionalization. We borrow the most frequently used definition, based on the nature of interparty competition and the parties’ organizational characteristics. The
Switching, Party Systems, Representation
269
literature, for example, distinguishes types of competition depending on the number of competitors, the durability of competitors, the degree of ideological differentiation among them, and barriers to entry for new contestants. Party systems become institutionalized to the extent that competition is structured by having a steady number of ideologically differentiated parties (Mainwaring 1999; Mair 1997, 175–194; Sartori 1976). Furthermore, party system institutionalization increases to the extent that party organizations encompass what Richard Katz and Peter Mair (1993; Müller 1993) have labelled the “party in public office” (e.g., executive and parliament), the “party on the ground” (i.e., membership, local party branches, and permanent party staffers), the “party in central office” (i.e., national governance structures independent of the party in public office), and the “party in the state” (e.g., control over patronage, public finance, and public services). Using these admittedly rough indicators of institutionalization, figure 10.2 differentiates among party systems that are underinstitutionalized, institutionalized, and overinstitutionalized. Overinstitutionalized Party Systems. Overinstitutionalized party systems range from highly stable consociational democracies, to oneparty dominant regimes such as 1960s–1980s Japan or 1920s–2000 Mexico, and even to totalitarian Soviet-like regimes. Such party systems have in common high barriers to entry for new parties and result in degrees of competition failure. Organizationally, parties are well entrenched in public office, on the ground, in the central office, and in the state. Incumbent parties in overinstitutionalized party systems regularly use various forms of clientelism and mass mobilization
Level of Institutionalization Under institutionalized 0
Institutionalized 0.5
Over institutionalized Democratic
Auto cratic 1
- Proto-parties & parties in public office
- Party in public office, on the ground, and in central office
Party in public office, on the ground, in central office, and in the state
- Unstructured competition
- Regular competition
- Competition failure
Figure 10.2
Levels of party system institutionalization.
270
Marcus Kreuzer and Vello Pettai
techniques to create and maintain a loyal electorate, manufacture electoral victories so overwhelming that they discourage opposition support, and engage in partisan institutional engineering to split or repress potential opponents (Magaloni 2006, 44–81; Sartori 1976, 217–246; Scheiner 2005, 7–31). Institutionalized Party Systems. In institutionalized party systems, electoral competition is well-structured by stable parties offering clear policy alternatives, and elections are sufficiently contested to produce regular turnover in government. Parties extend their organizational reach beyond parliament as they establish parties on the ground and in the central office. Institutionalized parties have professionalized electoral machines, sophisticated party members at the mass level, extensive links with interest groups, and well-developed governance mechanisms. Underinstitutionalized Party Systems. In underinstitutionalized party systems, electoral competition is unstructured both because barriers to entry for new parties are low and because the organizational weakness of parties contributes to their frequent turnover in office. For example, in late nineteenth-century Europe or during the first, semi-free elections in the former communist dictatorships, parties were little more than names, with minimal staff or organization—proto-organizations that had yet to build the kind of information-laden labels that established parties in western democracies so carefully cultivate (cf., e.g., Snyder and Ting 2002). Candidates’ affiliations were so tenuous that candidates often “belonged” to two different parties. Such proto-parties are rare, however, and most underinstitutionalized parties are usually organizationally anchored in public office, especially parliament. Given their limited logistical infrastructure outside the legislature, these parties have been aptly labelled as being over-parliamentarized (Àgh 1999). Competition in underinstitutionalized party systems accordingly focuses more on the personality of politicians than on reasonably stable and coherent party platforms.
10.3
Patterns of Party Switching
Party switching varies in terms of its frequency (i.e., the percentage of candidates or legislators who switch at a particular moment in time), its sequencing (the distribution of its frequency over time), and the forms it takes. In our work, we have found the level of party system institutionalization is a crucial, but by no means the exclusive,
Switching, Party Systems, Representation
271
determinant of these three dimensions of party switching (Kreuzer forthcoming; Kreuzer and Pettai 2003). Other factors influencing patterns of party switching include, as noted, institutions, historical legacies, and resources. We thus invoke a very broad ceteris paribus clause as we formulate hypotheses about the impact of party system institutionalization on the patterns of party switching. H1: Other things equal, switching patterns should vary across party systems according to the level of institutionalization. H1a: In overinstitutionalized systems, switching frequency should be medium, the sequencing should be wide, and fissions and start-ups should be the dominant forms. H1b: In institutionalized systems, the frequency of switching should be low, the sequencing should be intermittent, and fissions and hopping should be the dominant forms of switching. H1c: In underinstitutionalized systems, the frequency of switching should be high, the sequencing should be ongoing, and all forms should be used.
Few systematic studies exist on party switching in overinstitutionalized party systems. The evidence we have from the Japanese and Italian cases suggests that switching is rare in its sequencing over time; when it does occur, it appears in spurts and involves a significant number of politicians. Heller and Mershon, for example, report that in Italy between 1948 and 1987 solo party switches were extremely rare, but switching exploded to approximately one-quarter of deputies after the corruption scandals erupted in the early 1990s (Heller and Mershon 2005, 241–242; for a similar pattern in Japan, see Kato 1998; cf. Heller and Mershon, Kato and Yamamoto, and Mershon and Shvetsova, this volume). In consociational democracies such as Belgium and the Netherlands, party switching was observed in the 1960s and 1970s when the social cleavages buttressing the parties weakened. It took the form of fissions or start-ups and, in the case of the highly fragmented Protestant camp in the Netherlands, fusions (Deschouwer 1994, 82–87; Koole 1994, 279–282). Autocratic or totalitarian party systems see even fewer party switches, because either the absence of alternatives to the single regime party precludes switching as an option or the incentives to stay with the hegemonic party are so strong (Magaloni 2006, 44–52). There is some evidence, however, that switching takes place during the waning days of autocratic parties. As the Soviet Union neared collapse, for instance, more reform-minded communists broke away from their party to form various reform movements (Grofman, Mikkel, and Taagepera 2000, 341–342; Lieven 1993, 214–315).
272
Marcus Kreuzer and Vello Pettai
The evidence on party switching in institutionalized party systems is more readily available. The well-entrenched party organizations increase the transaction costs of switching, making it a rare event that occurs only intermittently and at low levels of frequency. For example, the United States experienced 20 switches between 1947 and 1997 and the United Kingdom, 37 switches between 1950 and 1996 (Cowley 1996, 217; Nokken 2000, 421; for cross-national overviews, see tables 1.2 and 1.3 in this volume; see also Mershon and Shvetsova, Nokken, and Schofield, this volume).6 The low frequency of such party hopping is consistent with the more general pattern that Peter Mair (1990) reports for post–World War II industrialized democracies. Looking at fusions and fissions, he found that they occurred only once every three elections (e.g., roughly 9 to 12 years) and rarely proved electorally successful (1990, 187). Yet party switching is higher in electorally volatile party systems even if they are otherwise institutionalized. For example, interwar France and Germany experienced 14 and 8 fissions, respectively, a handful of fusions, and an even larger number of individual hops (Kreuzer 2001, 113–133). Thus, institutionalized party systems exhibit a wide range of switching strategies, with hopping and fission emerging as the most common ones. In underinstitutionalized party systems, the allegiances of both politicians and voters to parties are very weak, which contributes to considerable electoral volatility and limits the ability of voters to hold politicians accountable (Mainwaring and Scully 1995, 19–23; Sartori 1976, 257–266; Zielinski, Slomczynski, and Shabad 2005). The weakness of politicians’ party affiliations is so pervasive in many postcommunist democracies that politicians frequently are tagged as political “tourists.” Referring to parties in the Baltics, Anatol Lieven (1993, 215) has noted that “they have no party organization, no structure, no registered membership, no policy-making bodies, and most emphatically no party discipline.” Underinstitutionalized party systems tend to reflect the poor organization and weak power structures of transitional regimes generally. Our data support this impressionistic account (cf. Kreuzer and Pettai 2003). Repeat candidates switch parties frequently, averaging 55 percent between elections. While in Estonia such trends have declined, indicating that the party system is institutionalizing, in Lithuania party switching among repeat candidates has actually grown from election to election, reaching nearly 75 percent between 2000 and 2004. With respect to switching types, roughly half of the
Switching, Party Systems, Representation
273
switching among Baltic politicians involved fusions, whereby parties merged or formed an electoral alliance in advance of the next election. Roughly a third of candidates hopped, 10 percent joined parties resulting from fissions, and 10 percent entered start-ups.
10.4
Systemic Effects of Party Switching
How consequential are these various patterns of party switching for party systems more generally? Again, we emphasize that the dynamics of party systems are influenced by many factors, including institutions, historical legacies, cleavages, media structures, socioeconomic conditions, and international context. The impact of such factors on party systems has been more extensively studied than that of party switching. To redress this imbalance, we focus solely on party switching and once more invoke a very broad ceteris paribus clause to stipulate hypotheses. H2: Other things equal, the effects of party switching should vary across party systems according to the level of institutionalization. H2a: In overinstitutionalized systems, switching, when it occurs, should trigger major political change and in particular should undermine dominant parties. H2b: In institutionalized systems, switching should tend to occur so as to reorient policy within parties and the party system. H2c: In underinstitutionalized systems, the clearest effects of switching should have to do not with policy or with the identity (or existence) of the dominant party but instead with the number of parties (i.e., the degree of party system fragmentation).
The evidence that follows to assess these hypotheses is again more illustrative than systematic. It serves to underscore the plausibility of the hypotheses rather than to offer a rigorous test. Since party switching has the most far-reaching consequences in underinstitutionalized party systems, we devote greatest attention to those effects.
10.4.1 Switching and Corrosion of Overinstitutionalized Parties (H2a) Party switching is extremely rare in overinstitutionalized party systems but when it occurs it becomes a key stimulus for political change. More specifically, party switching permits a modicum of representation in instances when electoral competition fails. Such competitive failures can result from one-party dominance or from
274
Marcus Kreuzer and Vello Pettai
the organizational ossification of parties that have been captured by external interest groups or that lack internal party democracy (Kato 1998; Magaloni 2006, 44–81; Scheiner 2005, 7–31). Party switching functions to undermine collusive, anticompetitive elite strategies. In a study of the Mexican case, Beatriz Magaloni observes that Hegemonic parties are collusive pacts among ruling party politicians to divide the spoils of office among themselves. The party’s central dilemma consists in deterring potential elite challengers, particularly those coming from within. As long as they are given access to government spoils and are rewarded with office, elites will remain united. However, elections pose a fundamental dilemma for hegemonic parties—politicians who are denied the party’s nomination will be inclined to split and challenge the regime as opposition (2006, 79).
In other words, the ability of members of the hegemonic party to hop to another party or to defect in sufficiently large numbers so as to start a new party becomes an important corrective mechanism. Individual politicians, through reaffiliations, fill the representational deficit arising from the ossification of existing parties or more general competitive failures. We stress that switching does not function as a long-term substitute for internal party democracy or interparty electoral competition: it is, rather, a short-term corrective mechanism that works as a catalyst that helps to reactivate them. This reactivating effect results from various corollary effects of party switching. In Mexico, a hegemonic party system with minor opposition parties, switchers in the 1980s increased the electoral credibility of such parties, reduced strategic voting, and thereby improved their chances of winning (Magaloni 2006). In the Baltics, breakaway factions from the old Communist Party of the Soviet Union (CPSU) played a central role first in gaining the independence of the three countries and subsequently in building the postcommunist party systems. Before 1990, the regional communist parties were divided between nationalist and loyal supporters of the all-Soviet party. The breakaway of the nationalist factions triggered the process that led to these countries’ independence from the Soviet Union. The postindependence fission of communists into reformist and hardliner factions was an important impetus toward the emergence of competitive multiparty systems (Commission on Security and Cooperation in Europe 1990, 34–46, 55–56, 66; Taagepera 1990, 305–308). These first party switchers revealed the
Switching, Party Systems, Representation
275
shallowness of the regime’s actual support. They ended up conveying crucial new information to the many potential dissidents who were sitting on the fence. In doing so, they facilitated the dissidents’ coordination to oppose the regime more openly and thus contributed to the cascade-like surge of protest that swept the communists from power (Kuran 1991).
10.4.2
Switching and Policy Change of Institutionalized Parties (H2b)
In established party systems, party switches generally occur intermittently and in conjunction with a party’s significant policy reorientation. Party switching becomes a corrective mechanism through which individual politicians who no longer feel ideologically at home in their party realign themselves with another, better-fitting party (cf. Heller and Mershon, McElroy and Benoit, Nokken, and Schofield, this volume). In the process, switching compensates the representational deficit created when either electoral competition or intraparty deliberations are ineffective. Switching or, just as importantly, the threat to switch is an important lever in internal party policy disputes. Michael Laver and Kenneth Shepsle (1999) write that at the heart of such disputes is the absence of “job market” for party leaders, thus making them more office- and career-seeking than policy-seeking: “Most party leaders have nowhere to go in politics but downhill and are likely to fight very hard to stay on top, even if their parties suffer some collateral damage in the process” (1999, 24). In many instances, the threat of switching is sufficient to constrain party leaders willing to trade policy benefits for career or office benefits. Consequently, switching corrects policy deviations and maintains the ideological coherence of parties. In the early twentieth century, for example, European socialist parties split into Marxist and social democratic parties, and saw their members switching in and out every time party leaders reformed any of their key ideological principles.7 One such vital issue was participation in government. Revolutionaries broke away if socialists joined the government, because they viewed participation in “bourgeois democracy” as a betrayal of their revolutionary principles. Moderates, in turn, broke away if socialists did not participate in interwar bourgeois-led, Popular Front governments seeking to defend democracy against fascism (Kreuzer 2001, 71–90). Immediately after World War I, responses to the Russian Revolution precipitated splits in western socialist parties, as supporters of the
276
Marcus Kreuzer and Vello Pettai
Bolsheviks split to create western Communist parties. Schisms have also occurred on the left throughout the post–World War II era. In 1981, for example, centrist members of the British Labour party broke away to form the Social Democratic Party (SDP) because they objected to the leadership’s leftist orthodoxies that they considered outdated and electorally harmful.8 In 2005, a leftist faction led by Oscar Lafontaine left the German Social Democratic Party (SPD) to form the new Left party, which opposed Gerhard Schroeder’s efforts to adopt more centrist, Third Way policies. This evidence suggests that party switchers often are either moderates responding to their party’s radicalization or radicals protesting their party’s move to the center.9 Not all switching in institutionalized party systems, however, responds to ideological divergence between parties and individual deputies. Occasionally, switching is opportunism, plain and simple. A party that is a few seats short of a parliamentary majority will coax an opposition deputy to jump ship by either offering some bribe or promising a cabinet post. Konrad Adenauer did this repeatedly in early postwar Germany, to incorporate smaller right-of-center parties into his Christian Democratic Union/Christian Social Union (CDU/ CSU) and thereby bolster his thin parliamentary majorities (Cary 1996, 237, 66–67; Kreuzer forthcoming; and cf. Kato and Yamamoto this volume).
10.4.3 Switching and Fragmentation in Underinstitutionalized Party Systems (H2c) Party switching in underinstitutionalized party systems has few of the corrective effects that it does in the other two contexts. Transitional party systems frequently are so inchoate and unsettled that they basically have no representational capacity to correct. We thus posit in H2c that party switching in underinstitutionalized party systems is linked in complex ways to their high levels of fragmentation and weak political parties. In particular, the effects of switching on fragmentation directly depend on the type of switches political actors choose, with some switching strategies increasing fragmentation and others reducing it. We therefore discuss separately the systemic effects of each switching strategy. Fissions. The effects of fissions are straightforward. The defection of politicians from an existing party to a newly forming one
Switching, Party Systems, Representation
277
dilutes the seat concentration of the switchers’ party of origin and thus increases fragmentation. From our evidence on the Baltic states, party fission is the least common form of party switching. Starting such new parties in this context is always risky, as voters have little information about their prospects of winning and such parties face all sorts of logistical obstacles. We thus surmise that the impact of fissions on the overall party system is limited (Kreuzer and Pettai 2003). Fusions. The effects of fusions also are straightforward, but not quite as direct as those of fissions. Party mergers decrease the level of fragmentation, but the degree of the decline depends on the sizes of the parties involved in the merger(s). Figure 10.3 differentiates between four types of mergers: survival mergers, takeovers, mergers among equals, and mergers to establish cartels. We discuss these mergers in the order in which they decrease fragmentation, starting with type exerting the most restricted impact. Survival mergers involve small parties that try to avert their extinction by combining their resources and, they hope, also their voters. They aim for their merger to signal an improvement in their electoral viability to voters and financial backers, thereby reducing strategic voting and strategic investing (and thus enhancing financial contributions). Survival mergers happen quite frequently but their effect on reducing fragmentation is marginal at best. The small size of such parties (roughly 1 to 3 percent of votes) limits the extent to which
Large
Takeover
Party A
Will reduce fragmentation
Cartel
Merger among equals
Will maintain current levels of fragmentation Small
Survival Small
Figure 10.3
Types of party.
Takeover Party B
Large
278
Marcus Kreuzer and Vello Pettai
their fusion influences the effective number of parties.10 Survival mergers are also rarely effective as the merged parties tend to be factionalized and plagued by leadership conflicts and programmatic differences. Their fusion typically does little to improve their prospects among voters, since the latter already have come to perceive them as electorally marginal. Survival mergers thus at best delay the electoral market-induced decline by one or two electoral cycles but rarely reduce fragmentation in a lasting fashion. A good example comes from Estonia, where the Independent Royalists and the Greens formed an electoral alliance in 1995. During the previous election in 1992, the two parties had scored 7.1 percent and 2.6 percent of the vote, respectively. Worried about their prospects in advance of the 1995 race, they teamed up. In the event, the fused party gained just 0.8 percent of the vote. In takeover mergers, a relatively large party (with 20 percent or more of the vote share) incorporates one or more small, electorally languishing, but ideologically proximate parties. The scenario for such a merger frequently is as follows. A larger party depends on a smaller one for maintaining its parliamentary majority, but the smaller party’s seats are threatened because its vote share risks falling below some electoral law threshold (Kreuzer forthcoming). Such a takeover merger would decrease fragmentation by eliminating the smaller party. In Estonia during the 1992–1995 legislative cycle, the Center Party executed just this move. It absorbed the small Entrepreneurs’ Party, which had just one seat in parliament, since the latter was clearly afraid of dropping out entirely during the next elections. From the limited evidence we have, however, it would appear that such strictly voluntary takeover mergers are rare. A merger of equals involves the merger of small to medium-sized, yet electorally viable, parties. In combining their sizable vote shares, such mergers would significantly reduce fragmentation. In the Baltics, for example, such mergers frequently ensue after two or more parties have formed an electoral alliance that has proved successful. The combined strength of the merged parties increases their legislative bargaining power and allows them more readily to comply with the membership requirements stipulated by various postcommunist party registration laws. In line with our reasoning is the case of two nationalist Latvian parties, For Fatherland and Freedom and the Latvian National Independent Movement, which merged in June 1997 to consolidate their voter base. Previously, each of the parties had polled an average of 8 to 9 percent, but they were unsure as to whether, over
Switching, Party Systems, Representation
279
time, one of them might not fall below the 5 percent threshold for parliamentary representation. In the first election after their fusion, they earned 14.6 percent of the vote (see Kreuzer and Pettai 2003). Cartel mergers mostly are hypothetical. There are few, if any, reasons why two parties, each of which is sufficiently large to hope to approach a parliamentary majority, would want to merge. Such a merger would establish a cartel whose only intent could be to do away with electoral competition altogether. Nonetheless, the category is analytically useful, for it highlights, among other things, the unusual institutional constraints contributing to such recurring legislative vote coalitions as that of the Socialist and Christian Democratic groups (PES and EPP) in the European Parliament (Kreppel 2002; cf. McElroy and Benoit this volume); the PES-EPP vote coalitions, albeit frequent and longstanding, have not culminated in fusion. Overall, fusions at a minimum maintain the current level of fragmentation and at best reduce it. This reductive effect, however, is subject to two qualifications. First, fusions have to be durable. Otherwise, they simply become short-term alliances that only temporarily contribute to a drop in fragmentation. Second, fusions also have to be self-limiting; otherwise, they lead to the sort of collusive behavior found in overinstitutionalized party systems. For example, the recent fusion of parties in Russia holds out the prospect of transforming Russia from an under- to an overinstitutionalized party system within just a few electoral cycles. Hopping. Of all the types of party switching, party hopping has the most complex implications for fragmentation. For starters, the total number of hops need not tell us much about how party hopping affects the overall party system. For example, if half the candidates (or sitting legislators) from party A switch to party B and vice versa, the fragmentation of the party system remains unchanged. To understand the impact of party hopping, we focus, as we did with fusions, on the relative size of the receiving party (gaining inswitchers) and the sending party (losing outswitchers) as well as their net flow of their inswitchers and outswitchers. Figure 10.4 illustrates three possible effects of hopping depending on the characteristics of the sending and receiving party. For ease of exposition, the figure is based on the assumption that sending parties have no new inswitchers and that receiving parties have no defectors. Hopping reduces fragmentation when candidates or legislators defect either from small to large parties or from one large party to
280
Marcus Kreuzer and Vello Pettai Large
(a) (b) Sending Party (c) (a) Small Small
Figure 10.4
Receiving Party
Large
Effects of hopping.
Note: a) reduces fragmentation, b) increases fragmentation, c) maintains status quo.
another. Such switches increase the candidate or seat share of the receiving party and thus reduce fragmentation. They also reduce the electoral viability of the sending party, especially if it is small and faces the risk of strategic voting. Hopping from smaller to larger parties is quite common in transitional party systems, especially if the hopper is a deputy. A larger party is interested in welcoming new deputies because in so doing it increases its parliamentary size and influence. In the Baltic states, large parties often took in electoral candidates from other parties because they had a hard time recruiting a sufficient number of candidates, especially well-qualified ones, to fill their lists. Hopping from one large to another large party is an uncommon occurrence. Especially in established democracies, well-institutionalized parties usually do not welcome defecting candidates or legislators. They have their own recruitment channels and generally decide career advancement within the party on the basis of seniority. New members, especially if they are incumbents, might undermine this seniority principle (cf. Kato and Yamamoto this volume). The barriers to welcoming party hoppers would be somewhat lower in first-past-the-post electoral systems, where inswitching candidates and legislators could add a new seat without threatening another incumbent’s list position. The most likely scenario for party hopping between large parties occurs when the hoppers would sufficiently increase the receiving parties’ parliamentary majority to affect the make up of the government (or, in presidential systems, legislative
Switching, Party Systems, Representation
281
agenda control). That is, hopping is especially likely to occur in situations when two equally large parties are in close competition to form the next executive (or to win control of the legislative agenda in presidential systems). Hopping, in turn, would increase fragmentation if deputies switch from large to small parties. The rise in fragmentation would be proportional to the number of such hoppers. Judging by available evidence, such hopping is rare—understandably so, since switchers would diminish their career prospects by joining a smaller party. The smaller party that such a hopper would join invariably is electorally less safe and has less legislative clout to assure either cabinet appointment or post-electoral career options. Policy-seeking motivations would be the most likely reasons for such switching (cf. Heller and Mershon 2005; this volume). Highly principled politicians valuing policy consistency more than votes or office benefits would join a party whose platform is more congruent with their personal policy preferences. In Germany, for example, a number of nationalist politicians were elected in 1949 with the CDU/CSU when the party had barely articulated its platform. Once it began to adopt moderate center-right positions, many of the more nationalist deputies left for more conservative parties (Kaack 1972, 19). In the Baltics, large parties rarely saw more than 2 to 3 percent of their repeat candidates hop away to another established party. Finally, politicians quite often switch from small to small parties in transitional party systems. Small parties are electorally more marginal and thus are more likely to lose candidates, but at the same time they are also more likely to accept new candidates to compensate for the ones who have defected. Hopping among such small parties is in effect a game of musical chairs, with little to no effect on fragmentation. In the Baltics, this tendency has appeared, although in general small parties have tended to lose more repeat candidates than they have gained (cf. Kreuzer and Pettai 2003).
10.5 Conclusion Party switching has been largely neglected in the party literature, which for too long has treated parties as unitary and fixed actors. The study of changes of party affiliation is a complex enterprise, however, because switching takes different forms and its impact varies depending on the context in which it occurs. Moreover, collecting data on party switching can be extremely time-consuming and
282
Marcus Kreuzer and Vello Pettai
often requires extensive contextual knowledge, which creates significant costs for large-N, cross-national studies. These obstacles notwithstanding, party switching is a crucial and understudied element of party system transformation. We see our working hypotheses as a starting point both for incorporating party switching more fully into theories of party system change and for motivating the extensive data collection that would permit more systematic assessments of its impact.
10.6 Notes * We would like to thank Carol Mershon and Will Heller for their thoughtful comments. Research for this chapter was supported by Estonian Targeted Financing Grants 0182573 and 0180128. 1. Note that this chapter considers switching by nonincumbent candidates vying for legislative office as well as by incumbents. 2. The term party hopping was first coined by Attila Àgh (1999, 179). 3. Recent examples from Israel (the founding of Kadima, see Schofield this volume) and Italy (Mershon and Shvetsova this volume) stand out in this regard. 4. In our research on Germany and the Baltics, we found that most such switchers to start-up parties are repeat candidates (i.e., candidates who run in a previous election but fail to get elected). For examples of already elected deputies in Italy and the United Kingdom who join start-up parties, see Mershon and Shvetsova, this volume. 5. The label for this tripartite division is taken from Andreas Schedler (1995), although it closely resembles the categorizations of other scholars (Duverger 1951; Mainwaring 1999; Mainwaring and Scully 1995; Sartori 1976). 6. Note, however, that the authoritative Butler and Butler compendium (2000, 247–249) gives a total of 82 switches in the UK Commons during the 1950– 1996 span, including the 34 switches occurring in the 1979–1983 term alone (when the Social Democrats were founded). 7. Granted, the European party systems of the first decades of the twentieth century were not as firmly institutionalized as they were in the 1950s and 1960s (on suffrage extensions as solidifying party systems, see, e.g., Lipset and Rokkan 1967). They were, however, institutionalized compared to their counterparts in postcommunist democracies in the 1990s. 8. As Mershon and Shvetsova (this volume) note, the SDP’s capacity to attract a few politicians outside the Labour Party qualified it as a start-up rather than as an instance of fission. 9. This link between switching and representation is a key focus of work on legislative party switching where scholars ask to what extent legislators who have switched realign their voting behavior with the policies of their new party (e.g., Ansolobehere, Snyder, and Stewart 2001; Desposato this volume; Heller and Mershon 2008; this volume; McCarty, Poole, and Rosenthal 2001; Nokken 2000; this volume).
Switching, Party Systems, Representation
283
10. Party-system fragmentation can be measured in terms of the raw number of parties or the effective number of parties, as Heller and Mershon observe in the Introduction to this volume. Which measure makes sense generally depends on the nature of the analysis.
10.7 References Àgh, Attila. 1999. “The Parliamentarization of the East Central European Parties: Party Discipline in the Hungarian Parliament, 1990–96.” In Party Discipline and Parliamentary Government, ed. Shaun Bowler, David Farrell, and Richard Katz, 167–188. Columbus: Ohio University Press. Ansolabehere, Stephen, James M. Snyder, Jr., and Charles Stewart III. 2001. “The Effects of Party and Preferences on Congressional Roll Call Voting.” Legislative Studies Quarterly 26 (4): 533–572. Butler, David, and Gareth Butler. 2000. Twentieth Century British Political Facts, 1900–2000, 8th ed. London: Macmillan. Cary, Noel. 1996. The Path to Christian Democracy: German Catholics and the Party System from Windhorst to Adenauer. Cambridge, MA: Harvard University Press. Commission on Security and Cooperation in Europe. 1990. Elections in the Baltic States and Soviet Republics: A Compendium for Reports on the Parliamentary Elections Held in 1990. Washington: Commission on Security and Cooperation in Europe. Cowley, Philip. 1996. “ ‘Crossing the Floor’: Representative Theory and Practice in Britain.” Public Law 2: 214–224. Cox, Gary. 1997. Making Votes Count. Strategic Coordination in the World’s Electoral Systems. Cambridge: Cambridge University Press. Deschouwer, Kris. 1994. “The Decline of Consociationalism and the Reluctant Modernization of Belgian Mass Parties.” In How Parties Organize, ed. Richard Katz and Peter Mair, 80–108. London: Sage. Desposato, Scott. 2006. “Parties for Rent: Ambition, Ideology, and Party Switching in Brazil’s Chamber of Deputies.” American Journal of Political Science 50 (1): 62–80. Duverger, Maurice. 1951. Les Partis Politiques. Paris: Armand Colin. Grofman, Bernard, Evald Mikkel, and Rein Taagepera. 2000. “Fission and Fusion of Parties in Estonia, 1987–1999.” Journal of Baltic Studies 31 (4): 329–357. Grose, Christiane, and Antoine Yoshinaka. 2003. “The Electoral Consequences of Party Switching by Incumbent Members of Congress, 1947–2000.” Legislative Studies Quarterly 28 (1): 55–75. Heller, William B., and Carol Mershon. 2005. “Party Switching in the Italian Chamber of Deputies, 1996–2001.” Journal of Politics 67 (2): 536–559. ———. 2008. “Dealing in Discipline: Party Switching and Legislative Voting in the Italian Chamber of Deputies, 1988–2000.” American Journal of Political Science 52 (4): 910–925. Kaack, Heino. 1972. “Fraktions- und Parteiwechsler im Deutschen Bundestag.” Zeitschrift für Parlamentsfragen 3 (1): 3–27.
284
Marcus Kreuzer and Vello Pettai
Kato, Junko. 1998. “When the Party Breaks Up: Exit and Voice among Japanese Legislators.” American Political Science Review 92 (4): 857–870. Katz, Richard, and Peter Mair. 1993. “The Evolution of Party Organizations in Europe: The Three Faces of Party Organization.” American Review of Politics 14: 593–617. Koole, Ruud. 1994. “The Vulnerability of the Modern Cadre Party in the Netherlands.” In How Parties Organize, ed. Richard Katz and Peter Mair, 278–302. London: Sage. Kreppel, Amie. 2002. The European Parliament and Supranational Party System: A Study in Institutional Development. Cambridge: Cambridge University Press. Kreuzer, Marcus. 2001. Institutions and Innovation: Voters, Parties, and Interest Groups in the Consolidation of Democracy: France and Germany, 1870–1939. Ann Arbor: University of Michigan Press. ———. n.d. “How Party Systems Form: Path Dependency and the Institutionalization of the Postwar German Party System.” British Journal of Political Science. Forthcoming. Kreuzer, Marcus, and Vello Pettai. 2003. “Patterns of Political Instability: Affiliation Patterns of Politicians and Voters in Post-Communist Estonia, Latvia and Lithuania.” Studies in Comparative International Development 38 (2): 76–98. Kuran, Timor. 1991. “Now Out of Never: The Element of Surprise in the East European Revolution of 1989.” World Politics 44 (1): 7–48. Laver, Michael, and Kenneth Benoit. 2003. “The Evolution of Party Systems between Elections.” American Journal of Political Science 47 (2): 215–233. Laver, Michael, and Kenneth Shepsle. 1999. “How Political Parties Emerged from the Primeval Slime: Party Cohesion, Party Discipline and the Formation of Governments.” In Party Discipline and Parliamentary Governments, ed. Shaun Bowler, David Farrell, and Richard Katz, 23–48. Columbus: Ohio State University. Lieven, Anatol. 1993. The Baltic Revolution: Estonia, Latvia, and Lithuania and the Path to Independence. New Haven, CT: Yale University Press. Lipset, Seymour Martin, and Stein Rokkan. 1967. “Cleavage Structures, Party Systems, and Voter Alignments: An Introduction.” In Party Systems and Voter Alignments, ed. Seymour Martin Lipset and Stein Rokkan. New York: Free Press. Magaloni, Beatriz. 2006. Voting for Autocracy: Hegemonic Party Survival and Its Demise in Mexico. Cambridge: Cambridge University Press. Mainwaring, Scott. 1999. Rethinking Party Systems in the Third Wave of Democratization: The Case of Brazil. Stanford, CA: Stanford University Press. Mainwaring, Scott, and Timothy Scully. 1995. Building Democratic Institutions: Party Systems in Latin America. Stanford, CA: Stanford University Press. Mair, Peter. 1990. “The Electoral Payoffs of Fission and Fusion.” British Journal of Political Science 20 (1): 131–142. ———. 1997. Party System Change: Approaches and Interpretations. Oxford: Clarendon Press.
Switching, Party Systems, Representation
285
McCarty, Nolan M., Keith T. Poole, and Howard Rosenthal. 2001. “The Hunt for Party Discipline in Congress.” American Political Science Review 95 (3): 673–687. Mershon, Carol, and Olga Shvetsova. 2008. “Parliamentary Cycles and Party Switching in Legislatures.” Comparative Political Studies 41 (1): 99–127. Müller, Wolfgang. 1993. “The Relevance of the State for Party System Change.” Journal of Theoretical Politics 5 (4): 419–454. Nokken, Timothy. 2000. “Dynamics of Congressional Loyalty: Party Defection and Roll-Call Behavior, 1947–97.” Legislative Studies Quarterly 25 (3): 417–444. Sartori, Giovanni. 1976. Parties and Party Systems. Cambridge: Cambridge University Press. Schedler, Andreas. 1995. “Under- and Overinstitutionalization: Some Ideal Typical Propositions Concerning New and Old Party Systems.” In Working Paper # 213. Kellogg Institute, University of Notre Dame. Scheiner, Ethan. 2005. Democracy without Competition in Japan: Opposition Failure in a One-Party Dominant State. Cambridge: Cambridge University Press. Snyder, James M., Jr., and Michael M. Ting. 2002. “An Informational Rationale for Political Parties.” American Journal of Political Science 46 (1): 90–110. Taagepera, Rein. 1990. “The Baltic States.” Electoral Studies 9 (4): 303–311. Zielinski, Jakub, Kazimierz M. Slomczynski, and Goldie Shabad. 2005. “Electoral Control in New Democracies: The Perverse Incentives of Fluid Party Systems.” World Politics 57 (3): 365–395.
This page intentionally left blank
11 Conclusions William B. Heller and Carol Mershon
11.1
Taking Stock of Party Switching
This book is motivated by the belief that party switching is an important phenomenon, neither as idiosyncratic nor as noxious as the general perception suggests. The extant literature tends to either neglect party switching or treat it as an aberration. In a sense, this is understandable. At first blush, it seems illogical or self-defeating for a politician whose career and fortunes are bound to a particular party to even consider defecting from that party and joining another. It is all too easy to view party switching as confined to special conditions, as hold for example in new democracies, transitional party systems, or weak party organizations. Yet, as the contributions to this volume make abundantly clear, changes of party affiliation among elected politicians occur in virtually all legislatures and party systems. Party switching varies in frequency but it is ubiquitous. As one consequence of strategic behavior, moreover, it is both systematic and predictable. Further, as emphasized in the introduction, the presumption that switching is anomalous and rare belies the extent of its implications, its causes, and its effects. Party switching is significant not because it is more common and more systematic than the conventional wisdom suggests, but rather because of what it reveals about politics and what we can learn from it. The theoretical, empirical, and normative importance of party switching emerges when it is seen as the product of strategic choice in a larger game. As such, legislators’ decisions on party affiliation must be viewed in the context of their relationships with their parties—or party leaders—and voters; these relationships in turn are contingent on the nature of the legislative party system. For their part, legislative party systems can change in important ways as a consequence of
288
William B. Heller and Carol Mershon
switching (see, e.g., Laver and Benoit 2003; Kato and Yamamoto this volume), hence affecting the legislative bargaining context for interparty interactions and, in turn, the views that voters have of parties, the constraints and incentives facing party leaders, and the factors motivating an individual legislator’s decision to switch or stay put. In a way, the conventional wisdom has it right in that party switching must be considered in light of the circumstances in which it occurs. The context, however, is strategic: Potential switchers take action (or not) depending on the panorama before them, and also on their expectations about whether and how their actions and others’ responses will alter that panorama. The focus throughout this volume on individual actions and motivations thus makes party switching more than an interesting and important phenomenon. The fundamentally strategic nature of party switching by ambitious politicians means that analysis of switching provides new and effective leverage for understanding changes in policy outcomes, the evolution of party systems, and myriad other aspects of the larger political setting. Politics is about choices, and choices are made by individuals, albeit in the context of institutional or institutionally relevant factors like aggregation rules (Riker 1982), party discipline (Heller and Mershon 2008), socioeconomic concerns (Desposato 2006), and electoral rules (Amorim Neto and Cox 1997; Cox 1997; Desposato 2006). Strategic interaction is the purview of game-theoretic analysis, and party switching is indeed part of a larger game, as stressed in chapter 2. The game itself is complex and, we suspect, probably intractable given available analytical technologies. All the same, the contributions to this volume investigate switching in the context of different pieces of the game, and in the aggregate they demonstrate not only how the study of switching can enrich understanding of the political world, but also how the pieces of the switching game depicted in figure 2.1 fit together. That switchers move with an eye to elections, for example, is amply documented in the chapters by Scott Desposato, Carol Mershon and Olga Shvetsova, and Norman Schofield. That switchers move to improve their ability to influence policy making is evident in the chapters by Kato and Yamamoto, Desposato, and Mershon and Shvetsova—and is nicely consistent with Heller and Mershon’s substantive contribution in chapter 7 (and with Heller and Mershon 2005). Gail McElroy and Kenneth Benoit show that party switching is motivated by politicians’ ideology and party policy as well as ambition, a finding that resonates with Heller and Mershon’s chapter. In the process, McElroy and Benoit point to the emergence of a new,
Conclusions
289
pan-European party system that unites national and supranational parties in a single policy space, much like state and national parties in the United States. Timothy Nokken, looking at the United States, shows not only that US parties can and do influence legislators’ voting behavior, but also that legislators can manipulate voter perceptions of that behavior to both adhere to discipline in their new party and present themselves as consistent defenders of constant principles. Hence Nokken’s chapter demonstrates the importance of both achieving policy outcomes and pleasing voters—even when the two might seem incompatible. Marcus Kreuzer and Vello Pettai, finally, lay out the logic connecting switching and to the overall shape of the party system, arguing that context matters but, consistent with every other contribution to this volume, it matters in systematic ways. Simply put, the take-away message is not only that party switching matters both because of how it influences and modifies politics and the political context and because of what it reveals about them. Switching affects the political context, and the political context affects switching. There is more: the causes of switching, how the consequences of switching play out, and how switchers behave all provide insight on political bargaining, party discipline, and the relative (and varying, cf. Mershon and Shvetsova this volume) importance of office, policy, and votes as motivations for legislators (Müller and Strøm 1999; Strøm 1990). Taking switching seriously can open the door to new insights, both static and dynamic, about politics and political systems. For instance, if the relative sizes of legislative parties are a defining element of party systems (as virtually all political scientists agree), and if switching alters party sizes, then individual-level party switching potentially can have a profound effect at the level of the party system. In this vein, the logic of our sketch of a switching game, together with the contributions to this volume, suggests that legislative party switching should be “lumpy,” as switchers respond to changes in information and also to switching-induced changes in the legislative environment. In the next section, we pursue briefly this implication as an example of just one of many lines of reasoning that flows from the logic of our general model of party switching.
11.2 Exploring Implications: The Question of Switching Cascades Party leaders’ decisions can induce (or retard) switching independent of the decision makers’ intent. Switchers’ decisions can do the same.
290
William B. Heller and Carol Mershon
The individual decisions of party leaders and party switchers can affect parties’ legislative weights, attractiveness to voters, and policy preferences. To the extent that they do so, they have system-level consequences that can in turn trigger further, potentially system-altering decisions by other individual legislative actors. It is worth noting that our logic addresses not only changes in parties’ legislative weights (as do Laver and Benoit 2003), but also the formation of new parties (cf. Schofield and Mershon and Shvetsova this volume). As a general rule, party switching affects both party seat shares and policy positions. In the process, and in conjunction with party leaders’ decisions on discipline, it affects the rewards switchers can reap (Heller and Mershon 2008).1 This means (a) that legislators’ choices are interdependent (Aldrich and Bianco 1992); and (b) that at least some switches have pivotal or nearly pivotal effects on the legislative party system (Laver and Benoit 2003; Kato and Yamamoto this volume), thus raising the stakes for other potential switchers. Switching by one or a few legislators thus could lead to a cascade of switching, as later-switching legislators respond to the changes in party weights induced by prior switches. (If legislators can share information, moves that would otherwise look like a trigger followed by a cascade might appear as multiple legislators moving all at once.) One reasonable hypothesis, then, is that switching should be lumpy—observed as en masse moves in reaction to new (or shared) information or as cascades triggered by a small number of key switches. 2 Switches into a party increase its seat share, thereby making it marginally more attractive to other potential switchers. This attraction should increase as more switchers follow the first, enticing more and more legislators to switch. Such inswitching should continue until the marginal benefits that accrue to a switcher drop to a sufficiently low level (whether due to parties becoming less welcoming or to other adjustments in the costs and benefits of switching). For a party that loses members, the dynamic is reversed; the value of the party should decrease for its remaining members, in turn making further exits more likely. Countering this kind of decline is likely to be particularly difficult. We therefore would expect to see flight trends, to the extent that they occur, as particularly damaging to parties and possibly even eroding them to the point that they vanish entirely (particularly where legislative rules require a minimum number of members for a legislative party group). When existing parties are weak (or when they disappear), they leave an attractive prize of potential voters for the party that can
Conclusions
291
attract them. This should create in incentive to form new parties that are positioned to win that prize. For a single legislator to try to pick up those voters might be suicidal in career terms, but if an entrepreneur who sees the opportunity to pick up voters expects his or her own action to motivate others to switch as well—to set off a cascade, in other words—then a move that looks suicidal might in fact be a calculated risk. From this perspective, switchers need not be attracted only to existing parties; they also can be drawn to attractive but underoccupied niches in the party spectrum (cf. Schofield this volume). Switches have the capacity to redefine the legislative party system. When they do, at least some legislators who previously had not considered switching might come to reevaluate their options. To the degree that such reevaluation produces a few more switches, it is easy to envision conditions where those few initial switches could stimulate many more. The result should be cascades of switching, possibly but not necessarily settling into some kind of equilibrium.
11.3 Taking Party Switching Seriously: In Lieu of a Conclusion, a Beginning Extant research on party switching, this volume included, has only begun to indicate the insights to be gained from studying the phenomenon. We cannot predict just how or how much taking party switching seriously can advance our understanding of political parties and legislative politics. We are convinced, however, that the study of party switching has the potential to reorient thinking about parties and party systems by concentrating analytical attention on the strategic, self-interested, ambition-driven decisions of individual politicians who, through their choices and changes of party affiliation, constitute parties and also drive the evolution of party systems. Indeed, the brief treatment of switching cascades just above illustrates that the analysis of switching offers new purchase on the dynamics of party systems. The contributions to this volume build on a varied but largely inchoate literature on party switching. Ours is in essence a first cut at subjecting party switching to the sort of rigorous theoretical and empirical analysis that typifies studies of, for example, coalition formation. Taken together, our findings are interesting, even intriguing. And the findings form a coherent whole: We show how, why, and to what effect ambitious politicians change party affiliation in pursuit
292
William B. Heller and Carol Mershon
of their goals, as constrained by institutions and in response to their environments. The findings and this volume are not and cannot be definitive, however. As is so often the case in early exploration of a new area of research, the questions that remain—and new questions emerge along with every new finding—far outpace the available answers. In a new field such as this, conclusions are a moving target. We have drawn our conclusions, and we have attempted to point the way to where other researchers might go as they consider how to examine party switching so as to gain new leverage on their own questions about legislative politics, parties, party systems, and democratic decision making.
11.4 Notes 1. Seat shares would be unaffected by switching only if each exit from a party were offset by an entrance from another. As Mershon and Heller (this volume) show, switches affect party preferences in line with each switcher’s marginal contribution both to the preferences of party he or she leaves and the party he or she enters. 2. We see no reason to expect legislators’ influence in their parties and in the legislature as a whole to be uniform. Moves by some legislators—notables—probably matter much more than moves by ordinary party rank and file.
11.5 References Aldrich, John H., and William T. Bianco. 1992. “A Game-Theoretic Model of Party Affiliation of Candidates and Office Holders.” Mathematical and Computer Modelling 16 (8/9): 103–116. Amorim Neto, Octavio and Gary W. Cox. 1997. “Electoral Institutions, Cleavage Structures, and the Number of Parties.” American Journal of Political Science 41 (1, January): 149–174. Cox, Gary W. 1997. Making Votes Count: Strategic Coordination in the World’s Electoral Systems. Cambridge: Cambridge University Press. Desposato, Scott W. 2006. “Parties for Rent? Ambition, Ideology, and Party Switching in Brazil’s Chamber of Deputies.” American Journal of Political Science 50 (1, January): 62–80. Heller, William B., and Carol Mershon. 2005. “Party Switching in the Italian Chamber of Deputies, 1996–2001.” Journal of Politics 67 (2): 536–559. ———. 2008. “Dealing in Discipline: Party Switching and Legislative Voting in the Italian Chamber of Deputies, 1988–2000.” American Journal of Political Science 52 (4). Laver, Michael, and Kenneth Benoit. 2003. “The Evolution of Party Systems between Elections.” American Journal of Political Science 47 (2, April): 215–233.
Conclusions
293
Müller, Wolfgang C., and Kaare Strøm, eds. 1999. Policy, Office, or Votes: How Political Parties in Western Europe Make Hard Decisions. Cambridge: Cambridge University Press. Riker, William H. 1982. Liberalism against Populism: A Confrontation between the Theory of Democracy and the Theory of Social Choice. Prospect Heights, IL: Waveland Press. Strøm, Kaare. 1990. “A Behavioral Theory of Competitive Political Parties.” American Journal of Political Science 34 (2, April): 565–598.
This page intentionally left blank
Contributors
Kenneth Benoit is professor of quantitative social sciences at Trinity College, University of Dublin, in the Department of Political Science. His substantive research interests include comparative electoral systems, Eastern European politics, political party systems, and transitions to democracy. His methodological interests include statistical applications, quantitative content analysis, and computer simulations. He has published extensively in leading scholarly journals and written (with Michael Laver) Party Policy in Modern Democracies (London: Routledge, 2006). Scott Desposato is assistant professor of political science at the University of California, San Diego. His research examines legislative representation, including parties, campaigns, and elections, especially in Latin America. William B. Heller is associate professor of political science at Binghamton University. His research focuses on political institutions, primarily legislative, but covers topics as varied as, among others, bicameralism, the evolution of cooperation, coalition unity, and party discipline. Heller has published articles in a number of journals, most recently including the American Journal of Political Science, Annual Review of Political Science, Public Choice, and Legislative Studies Quarterly. Junko Kato is professor of political science at the University of Tokyo. Her research investigates political parties, institutions, and welfare policies. She is the author of The Problem of Bureaucratic Rationality: Tax Politics in Japan (Princeton, NJ: Princeton University Press, 1994) and Regressive Taxation and the Welfare State (Cambridge: Cambridge University Press, 2003), and she has published in an array of professional journals such as the American Political Science Review, British Journal of Political Science, and Party Politics.
296
Contributors
Marcus Kreuzer is associate professor of political science at Villanova University. His work focuses on evolution of party systems in Western and Eastern Europe and origins of electoral systems. He is the author of various articles, including “Institutions and Innovation: Voters, Parties and Interest Groups in the Consolidation of Democracy— France and Germany, 1870–1939” (2001). Gail McElroy is assistant professor in the department of political science at Trinity College, Dublin. Her research interests include legislative organization, European Union politics, and the comparative study of political parties. She has published several articles on party competition in the European Parliament. Carol Mershon is associate professor of politics at the University of Virginia. Mershon’s research examines intraparty competition, the dynamics of party systems, the politics of multiparty government, and legislative politics. Her most recent book is The Costs of Coalition (Stanford University Press, 2002). Her articles have appeared in the American Journal of Political Science, the American Political Science Review, Comparative Political Studies, and the Journal of Politics, among others. Timothy P. Nokken is assistant professor in the department of political science at Texas Tech University. His research focuses primarily on the US Congress. In addition to his studies of congressional party switching, he has written extensively on the consequences lame duck sessions of Congress have on roll-call behavior and legislative agenda construction. His work appears in the Journal of Politics, Legislative Studies Quarterly, Studies in American Political Development, Journal of Theoretical Politics, and American Politics Quarterly. Vello Pettai is professor of comparative politics at the University of Tartu. He has worked on ethnopolitics, party development, and constitutional courts in the Baltic states. He has published in World Politics, Nations and Nationalism, East European Constitutional Review, and Journal of Democracy. Norman Schofield is director of the Center in Political Economy, the William R. Taussig professor of political economy, and professor in the departments of economics and political science. He is currently working on the theory of social choice, political economy, and democracy. His recent publications on democratic theory include articles in The Review of Economic Studies (2007), and (with Gary Miller) in Political Studies (2003), the American Political Science Review (2003), the American Journal of Political Science (2007),
Contributors
297
and Perspectives on Politics (2008); he has published on authoritarian regimes with Micah Levinson in Politics, Philosophy and Economics. His several books include Social Choice and Democracy (1985), Multiparty Government (with Michael Laver, 1990), and Mathematical Methods in Economics and Social Choice (2003). In 2006, he published Architects of Political Change and (with Itai Sened) Multiparty Democracy, both with Cambridge University Press. The Spatial Model of Politics was published in 2007, while Democracy and Tyranny will soon be published. He was elected to the American Academy of Arts and Sciences in 2005. Olga Shvetsova is associate professor of political science at Binghamton University. Her work on institutions and parties appeared in the American Journal of Political Science, Journal of Theoretical Politics, Comparative Political Studies, Electoral Studies, Journal of Democracy, and in other peer-reviewed journals and edited collections. She is a co-author of Designing Federalism (Cambridge University Press, 2004). Kentaro Yamamoto received his PhD in political science from the University of Tokyo. His research explores Japanese party switching and party system change in post–1993 Japan. He explains the dynamics of Japanese party system change using coalition theory.
This page intentionally left blank
Index
Note: Page numbers in italics indicate illustrative material; those followed by a t indicate tables; those followed by an f indicate figures. accountability, 3, 5, 109–10, 272 activist coalitions: electoral competition model, 55–7 plurality rule, 58–61, 77 polarization, 77, 78n9 policy positions, 60–1 switching equilibria, 59f United States elections, 60–1, 77 valence, 56–7, 76 Adenauer, Konrad, 276 affiliation decisions: Brazil, 113–19, 123–33, 129t, 141nn2,5 choices, 105n3, 141nn2,5, 142nn10–11 code of conduct, 85–6 European Parliament (EP), 147–9, 154–5, 163–6, 165t legislative switching game, 40–1, 46 models, 31–3, 40–1, 46, 123–33, 129t, 165t nature of political parties, 174–6 time-based, 124, 142n11, 222–4 United States, 81–3, 101–3 voting behavior, 82–3, 87–8, 91–103 See also benefits of party membership; models of party
switching; motivation for switching; party switching agenda control, 20, 35, 87, 102, 179, 261n4, 280–1 Ágh, Attila, 282n2 Aldrich, John H., 36, 56, 83–4, 105n3, 112, 154 Alexander, Rodney, 105n8 Alford, Thomas, 93–4t, 99t, 100 Allen, George, 73 ambition: legislators, 12, 30–1, 36–40, 178–9, 201–5, 224 motivation for party switching, 12, 174–5, 178–9 party leaders, 275 switching models, 30–1, 33–4, 36–8, 40 Am Ehad Party (Israel), 70 amendment votes: defectors’ vote changes, 99f DW-NOMINATE scores, 97f significance, 103 types of, 103–4 vote numbers, 83, 88–9, 91f, 103–4, 105n10 voting behavioral changes, 95–103 Arab Party, 70f, 71f Atkinson, Eugene, 93t, 99t, 100 Australia, 4, 11t
300
Index
balance of power, 7–8, 12, 57, 60, 76 ballot access. See under elections Baltic States, 272–5, 278, 281, 282n4 bandwagoning, 268 Barak, Ehud, 70, 72 bargaining, 12, 33, 38–9, 70, 72, 76, 102, 288 See also coalitions Belgium, 21n1, 76, 271 benefits of party membership: Brazil, 119–22 elections, 12, 41, 47n3, 111–12, 121–2, 270 parliamentary cycles, 203–5, 207–11, 217t, 218 See also affiliation decisions; motivation for switching; payoffs; pork Benjamin, Gerald, 84 Benoit, Kenneth, 78n7, 147–72 Bianco, William T., 36, 83–4, 105n3, 112–13, 154 bicameral legislatures. See under legislatures Blair, Tony, 179–80 Boll Weevils, 86 Bolshevism, 276 Brazil: First and Second Republics, 109 ruling on elected offices, 140 Brazilian Chamber of Deputies, 109–44 access to state resources, 120–2, 125 affiliation decisions, 113–19, 123–33, 129t, 141nn2,5, 142nn10–11 ballot access, 121, 126 benefits of party membership, 119–22 birthright rule on candidacy (candidato nato), 119, 142n7
causes of party switching, 110–33 club goods, 111–13, 116, 118, 119–21, 141n3 coalitions/government composition, 12, 120–1, 130t, 131 effects of party switching, 133–40 elections, 17, 119, 121–2, 125, 130t, 142n6 equilibria strategies, 113–19, 115f, 116f, 117f, 118f ideology, 119, 122–3, 125, 130t, 139f, 141n3, 143n16 open-list proportional representation (OLPR), 12, 121, 126, 142n9 party agreement scores, 133–5, 134t, 135t party influence, 137–40, 138t party system, 111–13, 111t, 138–40 payoffs, 111–13, 115t, 121–2, 125–6 predictions for party switching, 123, 128–33, 129t, 130t prohibition on party switching, 119, 140–1, 142n6 switching incidence, 4, 11t, 109–10, 110t, 128f, 139–41, 139f, 141n1 transaction costs, 14, 117, 119, 123, 126, 130t value-added of legislator for party, 112–13, 122–3, 126–7, 142n10 vote pooling, 110, 121 voting model, 135–7 British House of Commons: electoral laws, 206 legislative committees, 12 legislative terms, 206t new groups, 222 start-up parties, 282nn4,8
Index switching behavior, 218–19, 282n6 switching rates, 11t, 14, 216–20, 217t, 223–4, 271 trade unions, 179–80 Bryan, William Jennings, 74 Bush, George W., 78n2 cabinet offices, 12, 130t Canada, 7, 8, 11t, 12 candidato nato (birthright rule on candidacy), 119, 142n7 Cardoso, Fernando Henrique, 139 career advancement, 39f, 40, 202, 280 cartels, 277f, 279 cascade switching, 30, 32–3, 289–91 Center Party (Estonia), 278 Center Party (Israel), 67t centrist parties, 76, 78n6, 214, 223, 276 centrist policy positions, 55, 58, 61 Chamber of Deputies (Brazil). See Brazilian Chamber of Deputies Chamber of Deputies (Italy). See Italian Chamber of Deputies Christian Democratic Center (CCD; Italy), 189, 192–3t Christian Democratic Center-United Democratic Christians (CCDCDU; Italy), 189, 192–3t, 213 Christian Democratic Union/ Christian Social Union (CDU/ CSU; Germany), 276, 281 Christian Democrats (DC; Italy), 189, 192t, 218 Clean Government Party (CGP; Japan), 255f, 256f, 259, 262nn16–17 clientelism, 269 coalitions: Brazil, 12, 120–1, 130t, 131 electoral competition model, 57–8 European Parliament (EP), 279
301
Germany, 12, 47n8 government formation, 12, 70, 72, 76, 120–1, 130t, 131, 224 Israel, 70 Japanese Diet, 238–9, 245, 254, 257, 259, 261n6, 262nn16–17 legislative switching game, 39f party unity, 37 pro-capital, 59f procedural, 102 pro-labor, 59f switching equilibria, 59f US Congress, 102 voters’ interests, 47n8 See also activist coalitions cohesion: ideology, 123 party unity, 35, 81, 87, 133, 140–1, 148–9, 151 policy, 151 procedural matters, 102 voting, 148, 150–1, 157 Collor de Mello, Fernando Afonso, 139 committee assignments, 14–15, 35, 100, 111, 265 Communist Party (CPSU), 271, 274 Communist Party (Italy), 183–4 Communist Refounding (RC), 184, 190, 192t, 213–14 conditional logit (CL) models, 124, 129t, 164–8, 165t, 167t Congress (US). See United States Congress Conservative Coalition, 86 Conservative Party (Canada), 7 Conservative Party (UK), 222 consociational democracies, 269, 271 constituent relations: legislators, 16, 87, 102–3, 109–10, 119, 176–7, 191, 194n3 party systems, 176–7, 191 political representation, 265
302
Index
Cossiga, Francesco, 214, 215 Cox, Gary, 87, 103, 267 Crewe, Ivor, 14, 21n5 Czech Republic, 4 D’Alema, Massimo, 214 Danforth, John, 60–1 Deal, Nathan, 93–4t, 96f, 97f, 98, 99t decentralization, 165t decision making: legislators, 40–4 parties, 39f, 44–6 voters, 39–40 See also affiliation decisions; models of party switching; motivation for switching decisive majority (DM), 234–49, 260, 261n3 Dellay, Vincent J., 92, 93–4t, 99t, 101 Dem-Arab communists (Israel), 67f Democratic Party (US): electoral competition models, 58–60 Northern Wing, 86 party switching, 89, 90t, 93–4t power balance, 7 realignment, 73–5 voting behavior, 85, 91–101, 93–4t Democratic Party of Japan (DPJ): elections, 243–8, 259–60, 262n14 inswitchers and loyalists, 258t Left-Right policy, 254–9, 255f, 256f, 260, 262n18 party membership, 257, 259, 262nn16–17 seat distribution, 243t, 260 survival, 246–7, 260 switching behavior, 252t, 253t, 257 switching rates, 251–4, 257 winning thresholds, 249–50
Democratic Union for Europe (UDEur), 215 democratization, 207 Democrats-Olive Tree (Dem-U), 214–15, 220 Denmark, 11t deregulation, 157, 161t, 165t, 255f, 256f Desposato, Scott W., 36, 109–44 diet. See Japanese Diet direct democracy, 75 discipline: and ambition, 30–1, 33–4, 36–8, 40 Japanese Diet, 242, 250, 257, 260, 262n18 party leadership, 14–15, 35–8, 42–3, 45 party systems, 1, 6–7, 176, 178, 288 tools, 34–8, 45, 47n11, 176, 178, 288 Dorussen, Han, 61 Dow, John K., 64 Downs, Anthony, 55 Duma. See Russian Duma Duverger, Maurice, 181 DW-NOMINATE data, 83, 85–6, 88–9, 91–101, 96f, 97f, 105n4 Ecuador, 112, 141n3 Egypt, 73 elections: activist coalitions, 60–1, 77 ballot access, 14, 19, 35–6, 41–2, 121, 126, 154, 179 benefits of party membership, 12, 41, 47n3, 111–12, 270 Brazil, 17, 119, 121–2, 125, 130t, 142n6 by-elections, 176 European Parliament (EP), 151–2, 154, 169n4, 225n10 incumbency/nonincumbency, 282n1
Index institutionalization, 268, 270, 276–81 Israel, 58 Italy, 225n10 Japan, 233, 239–41, 243–8, 259–60, 261nn10,12, 262n14 nominating strategies, 17, 47n6, 111 parliamentary cycles, 203–5, 207–11, 217t party switching, 4–5, 14, 17, 265, 267–8 plurality rule, 206 presidential, 56, 58–60, 77n1, 78n9 primaries, 81, 112, 216 rules, 177, 206, 288 switching rates, 17, 210t, 212f United States, 56, 58–61, 77, 77n1, 78n9 winning thresholds, 130t, 233, 239–41, 245, 247–50, 261n10 See also reelection electoral competition models, 55–73 coalitions, 57–8, 59f Israel, 57–8, 66–73, 67f, 67t, 68f, 70t, 71t Nash equilibrium, 61–6, 68f party switching, 55–61 spatial models, 55–6 stochastic model, 55–7, 61–73 United States, 58–60, 73–5 voter behavior, 56–7 electoral rules. See open-list proportional representation; plurality rule; proportional representation Endersby, Jay, 64 Enelow, James, 61 Entrepreneurs’ Party, 278 environmental policy, 151, 157, 159, 161t, 162, 165t, 255f, 256f equilibria. See switching equilibria Estonia, 272, 278
303
European Liberal and Democrat Reform Party (ELDR), 150–1, 150t, 159, 165t European Parliament (EP): affiliation decisions, 147–9, 154–5, 163–8, 165t, 167t, 168–9 Comparative Manifestos Project, 152 conditional logit (CL) models, 164–8, 165t, 167t elections, 151–2, 154, 169n4, 225n10 expansion, 151 expulsions, 169n1 ideology, 19, 166, 168 languages, 151 Left-Right policy dimension, 148, 155, 158–66, 159f, 161t, 162f, 165t membership lists, 8, 156 national/transnational policy, 151, 158–63, 161t Parliamentary Bureau, 166, 167t, 170n7 party competition levels, 147–9 party groups and organization, 8, 148–56, 150t, 151–3, 155, 175f, 193–4, 289 party position expectations, 186t, 187t policy space, 158–63, 169 switching rates, 11t, 153t, 154–5, 158, 159f, 166–9, 167t, 169nn2–3 vote share, 157, 165t See also coalitions; elections European People’s Party (EPP), 150–1, 150t, 159, 169nn1–2, 279 European Union (EU), 147–9, 157–62, 161t, 165t European United Left/Nordic Green Left (GUE), 150t, 159, 165t, 168
304
Index
exogenous valence, 56, 58, 60, 63, 70, 76, 78n2 extremist parties, 45 final passage votes: DW-NOMINATE scores (US), 96f party defectors’ vote changes, 99f vote numbers, 83, 86, 88, 89, 91f, 103, 105n10 voting behavioral changes, 95–101 See also amendment votes; procedural votes fission switches, 8, 9t, 220, 222, 267–8, 268f, 271–2, 276–7 formal majority (FM), 234–9, 242–9, 260, 261n3 Forza Italia (FI), 189 fragmentation, 76–7, 276–7, 279–81, 283n10 France, 4, 11t, 12, 169n2, 271 fusion switches, 8, 9t, 222, 267–8, 268f, 271–2, 277–9 Gaines, Brian, 105n2 German Christian Democratic Union (CDU), 156 Germany, 11t, 12, 47n8, 271, 276, 281, 282n4 goals, individual/party, 33–4, 40–1 Goldwater, Barry, 60 Goode, Virgil, 92, 93–4t, 99t, 100, 105nn8,14 Goodman, Craig, 83 government formation: coalitions, 12, 70, 72, 76, 120–1, 130t, 131, 224 electoral competition models, 57–8 party switching, 7 Gramm, Phil, 93–4t, 99t, 100 Green Party (Brazil), 119 Greens (Estonia), 278 Greens (Italy), 190, 192t Greens (Verts; EP), 150t, 159, 165t, 168
Grose, Christian R., 84 Group for a Union of Democracies and Diversities (EDD), 150t, 159, 162, 165t Hall, Ralph, 105n8 Hamas, 73 Hannan, Daniel, 169n1 Hara, Kenzaburo, 261n9 Harbinger Party, 257, 262n16 Hatcher, Andrea C., 86, 87, 104n1 Hayes, Jamie, 93–4t, 96f, 97f, 98, 99t, 101 Heller, William B., 3–28, 17, 29–51, 131, 173–99, 201–29, 242, 271, 283n10, 287–93, 292n1 Hix, Simon, 151, 158 House of Commons (UK). See British House of Commons House of Representatives (Japan). See under Japanese Diet House of Representatives (US). See United States House of Representatives Hungary, 4, 11t ideal points: changes in, 177 estimates, 89, 105n11, 125 legislators, 179–80 Optimal Classification (OC), 182–5, 189–90, 194nn7–8 party system, 176, 180–2, 184–7 ideology: legislative switching game, 38 party membership, 83, 87–8, 130t polarization, 86 policy outcomes, 176 switching behavior, 84–5, 139f, 176 immigration/nationalism, 149, 159, 161, 161t, 163 Independent Democratic Party, 90t, 93–4t Independent Labour Party, 222 Independent Royalists, 278
Index independents: entrepreneurs, 268f Russia, 213 switching behavior, 223, 224n1 switching rates, 89, 90t, 225n8 United States, 89, 90t, 92, 93–4t, 98–100, 258n8 voting, 92, 93–4t, 98–100 India, 4 institutional constraints: Japan, 235–9 party-switching contexts, 12–17, 15f, 224 party unity and, 34–6 procedural rules, 16 See also agenda control; elections; decisive majority inswitches, 8, 9t, 19, 181, 185–8, 258t, 279–81, 290 involuntary switches, 10 Israel Beiteinu, 70f, 71f, 72 Israeli Knesset, 66–73 elections, 58 party positions, 66–9, 67f, 68f, 70f, 71, 71f recent switches, 70–3 seat share, 67t Italian Chamber of Deputies: balance of power, 189 Committee for Constitutional Reform, 211, 214 corruption scandals and, 271 elections, 225n10 legislative terms, 206t outswitching, 181, 185–8, 195n11 parliamentary cycles, 208t, 209–11, 213–15 party formation, 189–91, 192–3t, 220–3, 221t proportional representation (PR), 205, 211, 216–18 and Russian Duma, 82, 205–6, 209–20 secret voting, 194n6
305
single-member districts (SMDs), 205, 224n2 switching behavior, 182–91, 187–91, 213–15, 223–4, 271 switching rates, 4, 11t, 82, 205–6, 210t, 212f, 217t, 271 Italian Communist Party (PCI), 190, 192t Italian Social Democratic Party (PSDI), 218 Italian Socialist Party (PSI), 218, 220 Italian Socialist Party of Proletarian Unity (PSIUP), 219, 220 Jacobson, Gary C., 105n3 Japanese Diet, 233–63 balance of power, 240–2, 248, 259 coalitions, 238–9, 245, 254, 257, 259, 261n6, 262nn16–17 committee membership, 238–9, 260n1, 261nn4,8 decisive majority (DM), 234–9, 236t, 242–9, 260, 261n3 discipline, 242, 250, 257, 260, 262n18 elections, 233, 239– 41, 243–8, 259– 60, 261nn10,12, 262n14 formal majority (FM), 234–9, 242–9, 260, 261n3 House of Councilors (HC), 237, 261n4 House of Representatives (HR), 237–8, 243t, 247, 250, 260, 261nn3–4,10 institutional constraints, 141n4, 235–9 Left-Right policy positions, 241–2, 254–9, 255f, 256f, 262n18 motivations for switching, 239–42, 250–4 party system dynamics, 242–7 payoffs, 244–6, 246t, 261n9
306
Index
Japanese Diet—Continued proportional representation (PR), 250–2, 261nn11–12 reelection incentives, 250, 254, 260 seat share, 243t, 245, 260n1, 261n10, 262n19 single-member districts (SMDs), 248–54 SSI (Shapley-Shubik Index) data, 244–7, 245t, 246f, 248, 249 structure/organization, 12, 237–9 switching behavior, 233–4, 240, 242, 252t, 253t, 258t, 259–60, 262n13 switching rates, 11t, 244–6, 250–4, 271 winning thresholds, 233, 239–41, 245, 247–50, 261n10 See also Democratic Party of Japan (DPJ); Japan; Liberal Democratic Party (LDP) Jarman, John, 93–4t, 99t, 100 Jeffords, James, 7, 8, 74 Jeong, Gyung-Ho, 78n5 Johnson, Lyndon B., 60 Kadima Party (Israel), 58, 71–2, 71f, 78n6, 282n3 Kansas, 73–4 Kato, Junko, 233, 233–63, 244, 245 Katz, Richard, 269 Kernell, Samuel, 105n3 King, Anthony, 14, 21n5 King, Gary, 84 Kitamura, Naoto, 248, 261n9 Knesset. See Israeli Knesset Koizumi, Jun’ichiro, 260 Kreuzer, Marcus, 265–85 Labor Party (Israel), 67f, 67t, 68–9, 68f, 70, 70f, 71–2, 71f Labour Party (Ireland), 180 Labour Party (UK), 14, 179–80, 222, 276, 282n8
Lafontaine, Oscar, 276 Latvian National Independent Movement, 278–9 Laughlin, Greg, 93–4t, 96f, 97f, 98, 99t, 101 Laver, Michael, xii, 78n7, 156–7, 233, 244, 245, 275 Lebanon, 72–3 Left Unity Party (CG; EP), 168 legislative parties: catch-all parties, 182 decision-making power, 39f, 44–6 effective numbers, 8, 21n2 electoral strength, 41 hegemonic parties, 274 integrity, 39f legislative switching game, 38–46, 39f, 44–6 majority parties, 87, 89, 102, 175f mass membership parties, 182 membership and policy, 5–6 membership lists, 8 minority parties, 175f name changes, 195n10 pan-European, 289 party-switching contexts and, 3–5, 8–12, 15f, 40, 288–9 positions/ideology, 19, 44, 67f, 68t, 174, 181–8, 275, 292n1 preferences, 175–82, 190–1 procedural rules, 10, 14–15, 16 rules and processes, 35–6, 47n6, 102 start-up parties, 8, 9t, 21n3, 267, 268f, 271, 282nn4,8 value of new members, 112–13 See also cohesion; discipline; party leadership; party membership; party switching; party systems; party unity legislative switching game, 38–46, 39f, 288
Index legislators, 173–99 affiliation choices, 40–1 ambition, 12, 30–1, 36–40, 178–9, 201–5, 224 constituent relations, 16, 87, 102–3, 109–10, 112, 119, 176–7, 191 ideology, 86, 101–2, 275 leadership control, 180–1 legislative switching game, 40–4 motivation for holding office, 33 multidimensional/unidimensional issues, 12, 160, 178, 194n4 party preferences, 176–82 party switching, 15f party system, 174–82, 292n2 personality, 270 policy preferences, 41–2, 47n10, 84–5, 88, 102–3, 155, 292n1 rank-and-file preferences, 178, 180, 194n2 raw preferences, 175, 194n1 voting behavior, 37, 88, 101–2 See also affiliation decisions; benefits of party membership; models of party switching; motivation for switching; party switching legislatures: bicameral, 16 committees, 12 control of agenda, 20 legislative terms, 206t legislative weight, 32, 44 parliamentary systems, 10, 16 party affiliation types, 46, 47n9 presidential systems, 16, 206t, 217t, 221t, 280–1 seat shares, 20, 174, 243t, 245, 260n1, 261n10, 290, 292n1 semi-presidential systems, 206t, 217t, 221t, 224 unicameral, 16 Liberal Democratic Party (LDP; Japan): 21st-Century Club, 249, 261n10 agenda control, 261n4
307
challenges to, 243–4 coalitions, 257, 262nn16–17 decisive majority (DM), 242–4, 261n3 elections, 239, 243–7, 250–4, 259–60, 261n10, 262n14 Left-Right policy positions, 255f, 256f, 262n18 party system dynamics, 242–7 payoffs, 244–6, 248–9, 261n9 policy positions, 254–9 power balance, 12, 248 seat share, 261n10 SSI (Shapley-Shubik Index) data, 244–7, 245t, 246f, 248–9 strength, 12, 233–4, 237, 246–7, 248, 260, 260n2 switching behavior, 252t, 253t switching rates, 250–4, 261n9 winning thresholds, 261n10 Liberal Democrats (UK), 222 Liberal Front Party (PFL; Brazil), 126 Liberal Party (Canada), 7 Liberal Party (LP: Japan), 251, 253, 257, 262nn13,16,18 Lieberman, Avigdor, 72 Lieven, Anatol, 272 Likud Party (Israel), 58, 67f, 67t, 68–72, 68f, 70f, 71f Lin, Tse Min, 61 Lipset, Seymour M., 149 Lithuania, 272 Local Nash Equilibrium (LNE), 55, 57, 61–6 Magaloni, Beatriz, 274 Mainwaring, Scott P., 225 Mair, Peter, 269, 271 majority parties, 87, 89, 102, 175f majority rule, 75 manifestos, 151–2, 174, 225n9 Martis, Kenneth C., 104n1 Marxist parties, 275–6 mass mobilization parties, 269
308
Index
McCubbins, Matthew, 87, 103, 194n8 McElroy, Gail, 36, 147–72, 148, 156, 157, 169n5 McFadden, Daniel, 124 McGinnis, Michael, 56 McKelvey, Richard D., 61 Meretz Party, 67f, 67t, 68f, 70f, 71f Mershon, Carol, 3–28, 17, 29–51, 131, 173–99, 242, 271, 282n8, 283n10, 287–93, 292n1 methodology: Brazil, 110–11 European Parliament (EP), 156–7, 169n6 Italian Chamber of Deputies, 182–4, 194n7 Japanese Diet, 239–42 legislators, 182–4, 194n7 parliamentary cycles, 205–9 Mexico, 269, 274 Miller, Gary, 60 ministerial portfolios, 265 minority party, 175f Miskin, Sarah, 36 Mixed Group (Italy), 190, 192–3t, 210, 213–15, 225n5 mixed groups, 10, 47n9 Miyamoto, Ichizo, 261n9 Moakley, John, 93–4t, 99t, 100 models of party switching, 29–51 affiliation decisions, 31–3, 123–33, 129t ambition and discipline, 30–1, 33–4, 36–8, 40 legislative switching game, 38–46, 39f, 288 motivation, 31–3, 110–12 party unity, 34–6 See also affiliation decisions; benefits of party membership; legislative switching game Moledet Party (Israel), 67f, 67t, 68f Moore, John, 73
motivation for switching: ambition, 12, 55–6, 174–5, 178–9 career advancement, 39f, 40, 202, 280 contexts, 19–21, 30–3, 281, 288–9 electoral incentives, 121, 154–5 Italy, 182 legislative switching game, 42 models of party switching, 31–3, 83–4, 110–12 parliamentary cycles, 201, 202–5, 207, 220–3, 221t party leaders, 55–6 presidential systems, 280–1 switching behavior, 166 See also affiliation decisions; benefits of party membership; legislative switching game; models of party switching; parliamentary cycles mutual assured destruction doctrine, 14 Nash equilibrium, 61–6, 68f national identity, 255f, 256f National Religious Party (NRP; Israel), 67t, 68f, 70f, 71f negative agenda control, 35, 87 Neighbor, Cindy, 73 Nepal, 112 The Netherlands, 271 New Conservative Party (NCP; Japan), 255f, 256f, 262nn16–17 New Deal, 74, 75 New Frontier Party (NFP; Japan): dissolution, 260 elections, 248, 261n12 Left-Right policy positions, 254–9, 255f, 256f seat share, 243t, 245 switching behavior, 252t, 253t, 258t switching rates, 251–4 winning thresholds, 249–50
Index new parliamentary groups, 32–3, 205, 213–15, 219–22, 221t, 267–8 New Party Herald (NPH; Japan), 255f, 256f See also Harbinger Party New Zealand, 11t no-confidence votes, 7 Nokken, Timothy P., 81–108, 83, 85, 89, 92, 98, 104n1 NOMINATE, 83, 85–6, 88–101, 104, 105n4, 105n8, 157, 166–8 W-NOMINATE, 125, 128–33, 137–140 Norway, 14, 47n9 Noury, Abdul, 151 Olim Party, 67f, 70 Olmert, Ehud, 71, 72 open-list proportional representation (PR), 206 See also proportional representation Oppenheimer, Bruce I., 86, 87, 104n1 Optimal Classification (OC), 182–5, 189–90, 194nn7–8 Oslo Accords, 71 outswitching, 8, 9t, 19, 181, 185–8, 195n11, 279–81, 290 overinstitutionalized party systems, 273, 279 Papua New Guinea, 4 Parker, Mike, 93–4t, 96f, 97f, 98, 99t Parkinson, Mark, 73, 74 Parliament (UK). See British House of Commons parliamentary cycles, 201–29 affiliation stage, 203–5, 207–11, 210t, 216–18, 217t aggregated stages, 209–11 benefits, 203–5, 207–11, 217t, 218 country comparisons, 206t, 209–20
309
disaggregated substages, 211–15, 212f dormant stage, 204, 207–11, 217t, 220 elections, 203–5, 207–11, 217t electoral systems, 217t, 218, 221t Italy/Russia comparison, 205–6, 208f, 209–11 legislative terms, 206t legislators’ goals, 202, 204, 224 new groups, 205, 213–15, 220–3, 221t operationalized stages, 207–9, 208t policy control, 203–5, 207–11, 217t, 218–19 proportional representation (PR), 205, 210t, 218 regime types, 206t, 217t, 221t single-member districts (SMDs), 205, 209, 210–11, 210t, 216, 218 switching behavior, 201, 203f, 205, 209–11, 209–20, 210t switching rates, 204, 205–6, 209–15, 210t, 217t, 222–4 See also motivation for switching parliamentary majorities, 280–1 parliamentary systems, 10, 16 partisanship, 74–5, 119, 126, 223–4 Party for Social Democracy in Romania (PDSR), 219 party hopping, 267–8, 268f, 271–4, 279–81, 280f, 281, 282n2 party labels: communist parties, 174 identity, 7, 73–5, 122 integrity, 37, 87, 109–10 party unity, 34–8, 42 significance, 81, 92, 101, 103, 121, 141n3, 177, 270 party leadership: agenda control, 20, 35, 87, 102, 179, 180–1, 261n4, 280–1 decision-making process, 39f, 44–5, 289–90
310
Index
party leadership—Continued exogenous valence, 76, 78n2 legislative switching game, 38–46, 39f motivation for switching, 55 party-switching contexts, 15f, 55 policy preferences, 178–81 switching behavior, 167–8 valence, 55, 61, 76 voting behavior, 184–5 See also ambition; discipline party membership: automatic renomination, 14 equilibria strategies, 115f, 116f, 117f, 118f expulsions, 10 ideology, 83, 87–8, 130t loyalty, 88 payoffs, 111, 115t, 223 predictability, 130t value of new members, 112–13 See also affiliation decisions; benefits of party membership; party switching; party systems party mergers. See fusion switches Party of Social Democracy (PSD; Romania), 219 Party of the Democratic Left (PDS; Italy), 190, 192t Party of the European Socialists (PES; EP), 150t, 159, 165t, 168, 279 party switching, 3–28, 287–92 balance of power, 7, 8, 12 cost-benefit analysis, 37–8, 43–4, 112 definition, 8–10, 104n1 effects, 5–6, 91–101, 112, 134–8, 175–6 existing research, 12–17, 29, 82, 83–8, 101–2 implications and importance, 3–8, 288 legislative context, 15f
phenomenon of, 3–5, 8–11, 29–30 prohibitions, 10, 15, 36, 109–19, 140–1, 141n4, 142n6, 218 recording methods, 8 timing, 222–4, 225n6 voting behavior, 84–8, 91–101 See also legislative switching game; motivation for switching; parliamentary cycles; switching behavior; switching rates; switching types party systems: degree of institutionalization, 4, 6–7, 20, 207, 225n3, 268–71, 269f, 270–3, 276–81, 282n7 effective numbers of parties in, 8, 21n2 equilibrium status, 195n11 fragmentation, 76–7, 276–7, 279–81, 283n10 hegemonic parties and, 274 institutionalized parties in, 275–6, 280 legislative switching game, 38–46, 39f, 44–6 organizational outcomes in, 267–8 overinstitutionalized parties in, 273–5 party-switching contexts, 3–5, 8–12, 15f, 32, 40, 223–4, 266–70, 288–9 realignment, 4, 74–5, 77 switching and, 233–63, 265–85 types, 111t See also elections; legislative parties; switching equilibria party unity: agenda control, 35, 87 disciplinary tools, 34–8 levels, 177–8 majority party, 102 value of, 34–8, 41, 42, 102, 194n3 See also cohesion
Index Patty, John W., 61 payoffs: electoral, 121–2, 124, 223 expectations, 113, 123, 202, 241, 244–6, 246t forms of, 12, 113–14, 119–22, 125, 132, 141n5, 154t, 240 hegemonic parties, 274 motivation for switching, 110–11, 121 parliamentary cycle, 208t, 223 party membership, 115t, 233 pork, 13t, 111–12, 119–23, 125–6, 140–1 See also affiliation decisions; benefits of party membership; policy Pensioners’ Party (Israel), 71f, 71t Peretz, Shimon, 71–2 Pettai, Vello, 265–85 plurality rule, 56, 58–61, 77, 206 Poland, 14 polarization: activist coalitions, 77, 78n9 impact of procedural rules, 105n7 levels, 86, 96f roll-call voting behavior, 88, 95, 98, 101–3 policy: activist coalitions, 60–1 bargaining, 7 centrist positions, 57 consistency, 21n1, 281 government formation, 72 Left-Right dimension (EP), 148, 155, 158–66, 159f, 161t, 162f, 165t Left-Right dimension (Japan), 241–2, 254–9, 255f, 256f, 262n18 legislative switching game, 39f, 40, 41–2 motivation for switching, 12, 190, 281
311
party identity, 7, 73–5, 122 party switching and, 5–6, 133–40, 174–95 party switching effects, 225n9, 282n9, 290 party unity, 37, 102 polarization, 77 voting behavioral changes, 95–101 See also affiliation decisions policy outcomes: ideology, 176 influences on, 6, 7, 12 legislative switching game, 39, 40 party switching impact, 8, 288 preferences, 34, 47n4, 177 switching behavior, 202 political parties. See legislative parties; party systems political representation, 265–85 coordination strategies, 267–8 effects of party switching, 5–6, 133–40, 273–81, 282n9 legislative-executive relations, 265 legislator-constituent relations, 16, 87, 102–3, 109–10, 119, 176–7, 191, 194n3 party-switching strategies, 266–73 See also legislative parties; party systems Poole, Keith, 56, 83, 85, 88–92, 98, 104n1, 157, 169n6, 194n8, 225n7 Popular Front governments, 275–6 Popular Party (DC-PPI; Italy), 189, 192t pork, 13t, 111–12, 119–23, 120–2, 125–6, 140–1 Portugal, 152 postcommunist democracies, 270, 272, 274–5, 278–9, 282n7 Prata, Adriana, 194n8 presidential systems, 16, 206t, 217t, 221t, 280–1
312
Index
primary elections, 81, 112, 216 procedural rules, 16, 89, 102 procedural votes: behavioral changes, 95–103, 99f DW-NOMINATE scores, 97f significance, 103, 105n5 types of, 103–4 vote numbers, 83, 87–9, 91f, 103–4, 105n10 Prodi, Romano, 214 proportional representation (PR): Brazil, 12 closed-list, 206 EP elections, 225n10 Italy, 205, 211, 216–18 Japanese Diet, 250–2, 261nn11–12 new parliamentary groups, 222 open-list, 206 parliamentary cycles, 205, 210t, 218 party fragmentation, 77 party-switching phenomena, 12, 14, 211, 213, 219, 224, 224n2 Poland, 14 policy preferences, 47n10 Russia, 205 vote shares, 76 proto-parties, 270 Pure strategy Nash Equilibrium (PNE), 62 Radicals (Italy), 190, 192t reelection: Brazil, 119 incentives, 36–7, 119, 240–2, 250–4, 260 Italy, 191 Japan, 240–2, 250–4, 260 legislative switching game, 39f, 40–2, 44 legislator behavior, 154, 178, 191, 202, 208t parliamentary cycles, 204, 215 party membership rules, 127, 208t
party switching, 14, 16–17, 31–2, 102, 121, 128–9, 202–4, 203f rates, 17 Russia, 215 United States, 86 See also elections religion, 67f, 68f, 70f, 71f Republican Party (US): electoral competition models, 58–60 Jeffords defection, 7 party switching, 89, 90t, 93–4t realignment, 73–5 Southern Democrats, 84, 86 voting behavior, 85, 91–101 Rockefeller, Nelson, 74 Rohde, David W., 35, 88–9 Rokkan, Stein, 149 Roland, Gérard, 151 roll call votes: European Parliament (EP), 157 Italian Chamber of Deputies, 182–4, 194n5 party affiliation, 82–3, 85–9 party switching, 265 polarization, 95, 98 US Congress, 91–103, 105n12 See also amendment votes; final passage votes; procedural votes Romanian Chamber of Deputies: closed-list PR, 206 electoral laws, 206 legislative terms, 206t new parliamentary groups, 221t, 222 switching prohibitions, 218 switching rates, 11t, 216–20, 217t, 223 Rosenthal, Howard, 56, 157, 169n6, 225n7 Russia: Budennovsk hostage crisis, 213 Chechen War, 213 Russian Revolution, 275–6
Index Russian Duma, 209–20 independents, 213 institutionalization, 279 and Italian Chamber of Deputies, 209–20 legislative terms, 206t new parliamentary groups, 220–2, 221t parliamentary cycles, 208t, 209–11 party membership lists, 8 proportional representation (PR), 205, 213 single-member districts (SMDs), 205, 213, 216, 224n2 switching behavior, 223 switching rates, 4, 11t, 82, 205–6, 210–11, 210t, 212f Sánchez de Dios, 37 Schedler, Andreas, 282n5 Schmitt, Rogério, 17 Schofield, Norman, 55–79, 60, 65, 70, 78n8 Schroeder, Gerhard, 276 seat shares, 174, 243t, 245, 260n1, 261n10, 262n19, 290, 292n1 security policy, 67f, 68f, 70f, 71f semi-presidential systems, 206t, 217t, 221t, 224 Sened, Itai, 70, 78nn5,8 seniority, 280 Shapley-Shubik power index (SSI), 244–7, 245t, 246f, 248–9, 261nn6–7 Sharon, Ariel, 58, 61, 70, 71–2, 76, 78nn5–6 Shas Party (Israel), 67f, 67t, 68f, 69, 70f, 71f Shepsle, Kenneth, 275 Shinui Party (Israel), 70, 70f Shvetsova, Olga, 201–29, 282n8 single-member districts (SMDs): country comparisons, 205, 206 Japanese Diet, 248–54, 261nn11–12 policy preferences, 47n10
313
switching rates, 210–11, 213, 216–18, 222, 224, 224n2 social conservatives, 59–60, 59f, 73–5 Social Democratic Party (SDP; Japan), 243t, 245, 255f, 256f, 258t Social Democratic Party (SDP; UK), 14, 219, 222, 276, 282n8 Social Democratic Party (SPD; Germany), 276 socialist parties, 275–6 social liberalism, 59–60, 59f, 74–5, 159, 161–3, 165t socioeconomic policy, 161t, 255f, 256f, 288 Soka Gakkai (Japan), 262n17 solo switches, 9t South Africa, 11t Southern Democrats (US), 84, 86 Soviet Union, 174, 271, 274 Spain, 10, 11t, 12, 15, 17, 36 spending vs. taxes, 255f, 256f start-up parties, 8, 9t, 21n3, 205, 214–15, 220–2, 221t, 267, 268f, 271, 282nn4,8 stochastic electoral models, 55–73 activist valence, 62–3 balance solution, 64–5 electoral covariance matrix, 65–6 equilibrium concepts, 63–4 extreme value distribution, 64 Israeli Knesset elections, 66–73 party leaders, 76 Stokes, Donald, 77n1 strategic individuals, 34, 47n5 supermajorities, 120, 142n8, 178 switching behavior, 3–28, 173–99, 287–293 aggregate voting scores, 91–101 cartels, 277f, 279 cascades, 30, 32–3, 289–91 collective modes, 268 contexts, 15–17, 15f, 191, 242, 266, 288–9
314
Index
switching behavior—Continued defined, 8, 205 importance, 5–8, 287–8, 291–2 inswitches, 8, 9t, 19, 181, 185–8, 258t, 279–81, 290 levels of party system institutionalization and, 270–3 motivation, 166, 202, 276 near-simultaneous switches, 9t, 10 original status quo, 268f outswitching, 8, 9t, 19, 181, 185–8, 195n11, 279–81, 290 party agreement scores, 133–5, 134t, 135t policy/ideology-driven, 84–5, 88, 139f, 176, 178–91, 225n9, 292n1 post-switch behavior, 85 rule-driven, 9t simultaneous switches, 9t, 10 solo switches (hops), 9t survival mergers, 277–8, 277f switching strategies, 266–70 types, 8–10, 9t, 252t, 253t, 267–8, 268f, 277f United States, 83–8, 95–101, 218, 225nn7–8 voluntary/involuntary, 9t, 10, 14, 19, 21n5, 268 See also parliamentary cycles; party switching; political representation; switching rates switching equilibria, 55–79 activist coalitions, 59f electoral competition model, 55–61 Israel, 66–73 Local Nash Equilibrium (LNE), 55, 57, 61–6 party strategies, 7, 15, 73–5, 115f, 116f, 117f, 118f stochastic models, 55–7, 61–6 United States, 73–5
switching rates: Brazil, 109, 110t, 128f cross-national comparisons, 10–14, 11t, 81–3 data availability, 8 elections, 17, 210t, 212f European Parliament (EP), 11t, 153t, 154–5, 158, 159f, 166–9, 167t, 169nn2–3 independents, 89, 90t, 225n8 Japan LDP, 250–4, 261n9 level of party system institutionalization and, 20, 271–3 new parliamentary groups, 221t parliamentary cycles, 204–6, 209–15, 210t, 212f, 217t, 222–4 presidential systems, 217t semi-presidential systems, 217t single-member districts (SMDs), 210–11, 213, 216–18, 222, 224, 224n2 United States, 4, 11t, 73–5, 81, 90t, 103, 216–20, 217t switching types, 8–10 fission, 8, 9t, 220, 222, 267–8, 268f, 271–2 fusion, 8, 9t, 222, 267–8, 268f, 271–2, 277 party hopping, 8, 9t, 267–8, 268f, 271–4, 279–81, 280f, 282n2 start-ups, 8, 9t, 21n3, 267, 268f, 271, 282nn4,8 takeover mergers (fusion), 277f, 278–9 switch period, 183, 194n9 See also parliamentary cycles Syria, 73 Tauzin, Billy, 93–4t, 96f, 97f, 98, 99t, 101 taxes vs. spending, 161t, 165t
Index Third Way Party (Israel), 67f, 68f Thurmond, Strom, 75 totalitarian regimes, 269, 271 transaction costs, 14–15, 17, 112, 130t, 272 transitional party systems, 276, 280 Turan, Ilter, 37 Turkey, 11t, 73 Tzomet Party (Israel), 67f Ukraine, 11t, 12, 82 underinstitutionalized party systems, 268, 270, 276–81 unicameral legislatures. See under legislatures Union for a Europe of the Nations (UEN; EP), 150t, 159, 162, 165t, 169n2 Union of Democrats for the Republic (UDR; Italy), 214, 220 United Kingdom. See British House of Commons United States Congress: affiliation decisions, 81–3, 87–8, 141n2 committee structure, 12, 100, 105n12 compared to EP, 158 DW-NOMINATE scores, 96f, 97f electoral laws, 206 independents, 89, 90t, 92, 93–4t, 98–100, 258n8 legislative terms, 206t new parliamentary groups, 221t, 222 party unity, 21n6, 102–3 plurality rule, 77 polarization, 77, 95, 96f, 98 political realignment, 77 power balance, 7 presidential elections and, 58–60, 77n1, 78n9
315
switching behavior, 83–8, 95–101, 218, 225nn7–8 switching rates, 4, 11t, 73–5, 103, 216–20, 217t, 223–4, 272 voting behavior, 91–103, 91f United States House of Representatives, 81–108 Appropriations Committee, 100 Committee of the Whole, 89, 105n12 effects of party switching, 88–101 negative agenda control, 35 party switching, 81–8, 99f procedural rules, 89, 105n7 Rules Committee, 89, 105n12 switching rates, 81, 89, 90t voting behavior, 91–101, 93–4t See also amendment votes; procedural votes; roll call votes United States Senate, 7, 8, 81 valence, 55–6, 58–66, 76, 77n1 Virginia, 73–4 voluntary switches, 9t, 10, 46n1 vote pooling, 110 voters: candidate or party choice, 19 electoral competition models, 56–7, 63 information levels, 47nn7–8 legislative switching game, 38, 39–40, 39f, 41, 42 legislator/constituent relations, 16, 87, 102–3, 109–10, 119, 176–7, 191, 194n3, 287 party labels, 34–6, 109–10 party-switching contexts, 15f policy preferences, 176–7 vote shares: European Parliament (EP), 157, 165t
316
Index
vote shares—Continued expectations, 32, 41, 62–5, 121 maximizing, 56–8, 68–9, 113 proportional representation (PR), 76 switching impact, 84, 278 Watson, Albert, 93–4t, 99t, 100 Webb, Jim, 73
winning thresholds, 130t, 233, 239– 41, 245, 247–50, 261n10 Workers’ Party (PT; Brazil), 119, 126 Yahadut Party (Israel), 67f, 67t, 71f Yamamoto, Kentaro, 233–63 Yoshinaka, Antoine, 84