Mexico Beyond NAFTA: Perspectives for the European Debate (Routledge Studies in the Modern World Economy)

Mexico Beyond NAFTA With the European Monetary Union well underway, Europe is looking at nearby countries and cultural...

Author: Lionella Punzo

11 downloads 642 Views 3MB Size Report

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

Report copyright / DMCA form

DOWNLOAD PDF

Mexico Beyond NAFTA

With the European Monetary Union well underway, Europe is looking at nearby countries and culturally closer continents to deﬁne its strategies for the future. The recent ﬁnalization of a general agreement indicates that Mexico, and Latin America in general, will play a key role in such scenarios. In this book, chapters by leading Mexican economists are matched with reactions from European colleagues. They offer a novel viewpoint on the critical assessment of the NAFTA process and its implications for the economies of the two continents. The analysis of the critical experience of integration between unequal economic partners covers a wide range of issues, including: • • •

dollarization and ﬁnancial markets integration dynamics of industrial structures and international specialization ﬁscal federalism and related issues, including those related with labor mobility.

The book may thus also contribute to a better understanding of the issues involved in the process of European Union enlargement. Technical analyses are complemented by non-technical expositions, so that the book will be of interest to researchers and policy makers alike, on both sides of the Atlantic. Martín Puchet Anyul is Professor of Quantitative Methods, Faculty of Economics, UNAM, and has taught at CIDE (Mexico), Universidad de la República (Montevideo) and University of Siena. He has published a book: Sistemas de Cuentas Económicas, México: UNAM 1996; and a number of papers. Lionello F. Punzo has taught at UCLA, UNAM and IAS (Vienna). In 1996 he visited ISER, Osaka University, and he is currently Professor of Economics at Siena University. His works include The Dynamics of a Capitalist Economy (with Richard M. Goodwin), Polity Press and Westview Press 1987, Economic Performance (edited with B. Böhm), Physica Verlag 1992, and European Economics in Transition (edited with O. Fabel and F. Farina), Macmillan 2000.

Routledge Studies in Development Economics

1

Economic Development in the Middle East Rodney Wilson

2

Monetary and Financial Policies in Developing Countries Growth and stabilization Akhtar Hossain and Anis Chowdhury

3

New Directions in Development Economics Growth, environmental concerns and government in the 1990s Edited by Mats Lundahi and Benno J. Ndulu

4

Financial Liberalization and Investment Kanhaya L. Gupta and Robert Lensink

5

Liberalization in the Developing World Institutional and economic changes in Latin America, Africa and Asia Edited by Alex E. Fernández Jilberto and André Mommen

6

Financial Development and Economic Growth Theory and experiences from developing countries Edited by Niels Hermes and Robert Lensink

7

The South African Economy Macroeconomic prospects for the medium term Finn Tarp and Peter Brixen

8

Public Sector Pay and Adjustment Lessons from ﬁve countries Edited by Christopher Colclough

9

Europe and Economic Reform in Africa Structural adjustment and economic diplomacy Obed O. Mailaﬁa

10 Post-apartheid Southern Africa Economic challenges and policies for the future Edited by Lennart Petersson

11 Financial Integration and Development Liberalization and reform in Sub-Saharan Africa Ernest Aryeetey and Machiko Nissanke 12 Regionalization and Globalization in the Modern World Economy Perspectives on the Third World and transitional economies Edited by Alex F. Fernández Jilberto and André Mommen 13 The African Economy Policy, institutions and the future Steve Kayizzi-Mugerwa 14 Recovery from Armed Conﬂict in Developing Countries Edited by Geoff Harris 15 Small Enterprises and Economic Development The dynamics of micro and small enterprises Carl Liedholm and Donald C. Mead 16 The World Bank New agendas in a changing world Michelle Miller-Adams 17 Development Policy in the Twenty-First Century Beyond the post-Washington consensus Ben Fine, Costas Lapavitsas and Jonathan Pincus 18 State-Owned Enterprises in the Middle East and North Africa Privatization, performance and reform Edited by Merih Celasun 19 Finance and Trade in Developing Countries Edited by Rohinton Medhora 20 Contemporary Issues in Development Economics Edited by B.N. Ghosh 21 Mexico Beyond NAFTA Perspectives for the European debate Edited by Martín Puchet Anyul and Lionello F. Punzo 22 Economies in Transition A guide to China, Cuba, Mongolia, North Korea and Vietnam at the turn of the twenty-ﬁrst century Ian Jeffries 23 Population, Economic Growth and Agriculture in Less Developed Countries Nadia Cuffaro 24 From Crisis to Growth in Africa? Edited by Mats Lundahl

Mexico Beyond NAFTA Perspectives for the European debate

Edited by Martín Puchet Anyul and Lionello F. Punzo

London and New York

First published 2001 by Routledge 11 New Fetter Lane, London EC4P 4EE Simultaneously published in the USA and Canada by Routledge 29 West 35th Street, New York, NY 10001 Routledge is an imprint of the Taylor & Francis Group This edition published in the Taylor & Francis e-Library, 2001. © 2001 Editorial material and selection, Lionello F. Punzo and Martín Puchet Anyul; individual chapters, the authors. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library. Library of Congress Cataloging in Publication Data Mexico beyond Nafta: perspectives for the European debate/edited by Martín Puchet Anyul and Lionello F. Punzo. p. cm. 1. Monetary policy–Mexico. 2. North America–Economic integration. 3. Mexico–Commerce. 4. Free trade–North America. 5. European Union countries–Economic integration. I. Puchet, M. (Martín) II. Punzo, Lionello F. HG665.M486 2001 332.4'972–dc21 00–051745 ISBN 0-415-24386-6 (Print Edition) ISBN 0-203-18783-0 Master e-book ISBN ISBN 0-203-18906-X (Glassbook Format)

Contents

List of tables List of ﬁgures Acknowledgments Preface Notes on contributors

xi xiv xv xvi xxiii

PART I

Paths to convergence and monetary union(s) 1

Currency boards and monetary unions: The road ahead or a cul de sac for Mexico’s exchange rate policy?

1 3

DAVID IBARRA AND JUAN CARLOS MORENO-BRID

1.1 1.2 1.3 2

Advantages and limitations of currency boards: an overview A monetary union: a necessary step for Mexico’s sub-regional integration? Final considerations

Is a convergence path of monetary policies possible under the North American Free Trade Agreement?

4 11 14

21

LUIS MIGUEL GALINDO AND HORACIO CATALÁN

2.1 2.2 2.3 2.4 2.5

Theoretical framework The P* model Empirical evidence The estimated P* model Conclusions and economic policy issues

Comments FRANCESCO FARINA

22 25 26 30 35 46

viii

Contents

PART II

Mexico’s foreign sector, the exchange rate, trade and productive structures 3

The external relationship of the Mexican economy with the United States, and its implications beyond trade

51 53

MARÍA ELENA CARDERO

3.1 3.2 3.3 4

A new export–import model The transborder integrated labor market Conclusions

Exchange rate and competitiveness

53 66 72 78

PABLO RUIZ-NÁPOLES

4.1 4.2 4.3 4.4 5

Theoretical approaches to the exchange rate determinants Relative unit labor costs and real exchange rate ULC, wages and productivity Conclusions

Vertical integration and comparative advantages

78 80 89 90 102

FIDEL AROCHE REYES

5.1 5.2 5.3

Methodology Some tests on vertical integration and revealed comparative advantage Summary and conclusions

Comments

103 106 112 115

BODO HILGERS AND JÜRGEN MECKL

PART III

Sectorial performance in manufacturing, industrial organization and the innovation system during the trade liberalization process

121

6

123

Integration and polarization: Mexico’s economy since 1988 ENRIQUE DUSSEL PETERS

6.1 6.2 6.3

Neoliberalism, export-oriented industrialization and liberalization strategy The impact of liberalization strategy Conclusions

123 128 141

Contents ix 7

Liberalization policies and Mexico’s competitiveness: Are technological capabilities upgraded or downgraded?

147

MARIO CIMOLI

7.1 7.2 7.3 7.4 7.5 7.6 7.7 8

Import substitution period The transition period The Exchange Rate Based Stabilization and the new incentive framework The Mexican industrial performance and the corresponding analysis The globalization of Mexican industry: towards a modernized assembly plant Learning capabilities within ﬁrms and sectoral speciﬁcities Conclusions

Trade integration and sectorial productivity

148 150 151 153 157 160 168 174

FLOR BROWN AND LILIA DOMÍNGUEZ

8.1 8.2 8.3 8.4

The new Mexican industrial model Manufacturing productivity 1984–94 Determinants of productivity Conclusions

Comments: International openness, dualism and growth

175 177 180 186 190

JEAN-LUC GAFFARD AND MICHEL QUÉRÉ

PART IV

Regional divergence and structural policies 9

Structural divergence and the dynamics of dualism: Lessons from Mexico before and after NAFTA

197 199

MARTÍN PUCHET ANYUL AND LIONELLO F. PUNZO

9.1 9.2 9.3 9.4 9.5 9.6 9.7

Some of the old, new and newer perspectives on integration Before or after NAFTA? Structural change from the growth perspective Structural change through a growth cycle approach Structural dynamics in the long and the medium run Dualism and growth traps Economic policies for?

201 204 206 210 213 216 233

x

Contents

10 Regional policies in the North American and European integration agreements

243

SÁRAH MARTÍNEZ PELLÉGRINI

10.1 10.2 10.3 10.4 10.5 10.6

The two treaties and their objectives The institutional framework developed by the agreements Policies involved in the agreements Regional policies Institutional change Final remarks

11 The ﬁscal federal system in Mexico: Fiscal insurance with moral hazard

243 246 247 249 256 258

262

JAIME SEMPERE AND HORACIO SOBARZO

11.1 11.2 11.3 11.4 11.5 11.6

Determinants of moral hazard in the federal system Mexican ﬁscal federalism The institutional context Redistribution An example of moral hazard: the bailout of the states Some proposals to reduce moral hazard

Comments

262 264 266 270 277 278 281

OLIVER FABEL AND BODO HILGERS

Index

287

Tables

2.1 2.2 .2.3 2.4 2.5 2.6 A.2.1 A.2.2 A.2.3 A.2.4 A.2.5 A.2.6 A.2.7 A.2.8 A.2.9 A.2.10 A.2.11 A.2.12 A.2.13 A.2.14 A.2.15

Unit root tests Statistics of the Johansen procedure for yMX, yEU, yCN 1986(1)–93(4) Statistics of the Johansen procedure for yMX, yEU, yCN 1986(1)–98(4) Statistics of the Johansen procedure for P* for Mexico 1986(1)–98(4): pt, m2t, yt, rt Statistics of the Johansen procedure for P* for the United States 1986(1)–98(4): pt, m2t, yt, rt Statistics of the Johansen procedure for P* for Canada 1986(1)–98(4): pt, m2t, yt, rt Rate of growth and variance of the GDP in each quarter Inﬂation rates mean and variance (pt) Monetary aggregate mean and variance (m2t) Real monetary aggregate mean and variance (m2tpt) Nominal interest rate mean and variance (rt) Real interest rate mean and variance (rrt) Nominal exchange rate mean and variance (st) Real exchange rate mean and variance (srt) Characteristic roots of the Johansen procedure 1986(1)–98(1) Characteristic roots of the Johansen procedure 1986(1)–94(4) Co-integration tests for the P* for Mexico 1986(1)–98(4): pt, m2t, yt, rt Characteristic roots of the Johansen procedure for Mexico Standardized values of the Johansen procedure for Mexico: (pt, m2t, yt, rt) Co-integration tests for the P* for the United States 1986(1)–98(4): pt, m2t, yt, rt Characteristic roots of the Johansen procedure for the United States

29 30 30 31 31 31 37 37 37 38 38 38 39 39 39 39 40 40 40 40 41

xii Tables A.2.16 Standardized values of the Johansen procedure for the United States: (pt, m2t, yt, rt) A.2.17 Co-integration tests for the P* model for Canada 1986(1)–98(4): pt, m2t, yt, rt A.2.18 Characteristic values for the Johansen procedure for Canada A.2.19 Standardized values of the Johansen procedure for Canada: (pt, m2t, yt, rt) 3.1 Changes in tariff structure 1985–8 3.2 Global supply and demand, annual structure and variations 3.3 Share of foreign trade in the Gross Domestic Product 3.4 Percentage share of Mexico’s trade with Canada and the United States 3.5 Sectorial shares of Mexico’s foreign trade in merchandise percentage shares 3.6 Industry exports and imports 3.7 Share of MEI in manufacturing’s GDP, by selected branches: percentage share 3.8 Balance of payments: current account and capital 3.9 Number of Mexican emigrants to the United States by year, in thousands 3.10 Population born in Mexico but residing in the United States, in thousands A.3.1 Growth of Gross Domestic Product, by economic sectors 4.1 Unit labor cost 4.2 Nominal and real exchange rates 4.3 Real exchange rate and relative Unit Labor Costs Index numbers 4.4 Real Wage Rate indexes 4.5 Productivity indexes A.4.1 Unit root tests sample 1973–95 A.4.2 Modelling LE by OLS sample A.4.3 Equation tests results A.4.4 Cointegration analysis 1973–95 A.4.5 VAR tests results 5.1 Total and bilateral trade composition between the US and Mexico by sector of origin 1990 5.2 Vertically integrated coefﬁcients 5.3 Vertically integrated sectors by net trade 5.4 Revealed relative abundance of inputs 6.1 Main macroeconomic variables 6.2 General tendencies of the manufacturing sector (1988–98) 7.1 Macroeconomic indicators

41 41 41 41 55 56 58 60 61 62 62 67 71 72 75 81 82 83 90 92 95 96 96 98 99 107 108 110 111 129 136 151

Tables 7.2 7.3 8.1 8.2 8.3 8.4 8.5 10.1 11.1 11.2 11.3 11.4 11.5 11.6

Trade balance A taxonomy of sectoral technological capabilities Labor productivity Labor productivity and exports Labor productivity according to size of establishment: annual growth rate Large establishments Small establishments Technical assistance provided by BECC 1995–7 Distribution of resources Revenues and Expenditures 1993 Revenues 1993 Recipients and contributors of the federal system Index of Municipal Privation, 1990 Estate debt as a percentage of GDP, 1993

xiii 154 162 177 178 179 184 185 255 271 273 274 275 276 277

Figures

2.1 4.1 4.2 4.3 4.4 4.5 6.1 6.2 6.3 6.4 9.1 9.2 9.3 9.4 9.5 9.6a 9.6b 9.7 9.8a 9.8b 9.8c

Rates of real growth of the GDPs of Mexico, Canada and the United States Nominal and Real Exchange Rates 1968–82 Nominal and Real Exchange Rates 1983–95 Real Exchange Rate and Relative Unit Labor costs Real wage rates indexes Productivity indexes Inﬂation rates 1980–98 Employment of micro, small and medium ﬁrms (MSMF) and big ﬁrms Share of maquila and main 300 exporting ﬁrms over total exports and employment 1993–8 Trade balance/GDP Mexico–US: rates of productivity growth (1971–97) Mexico: observed and simulated rates of productivity growth Mexico–US: investment per employee, rates of growth (1971–97) FS one sector by Mexico and US FS for the long run: Mexico and US FS for the medium run: Mexico FS for the medium run: US FS for the long run: South, North Center and Italy FS for the medium run: South Italy FS for the medium run: North Center Italy FS for the medium run: Italy

27 84 85 86 91 93 131 133 134 137 205 208 209 214 217 218 220 222 224 227 230

Acknowledgments

Martín Puchet acknowledges the Mexican CONACYT for its support to the sabbatical year project #990217: ‘Dinámica y cambio estructural en las integraciones económicas’. The research project # IN06597 of DGAPAUNAM: ‘Métodos de análisis de la integración y estructuras económicas de América del Norte’ supported several activities within the research reported in this book. Lionello F. Punzo’s contribution was made possible by CNR grants #97.05179.CT10 and #98.03803.PS10, within the National Project: ‘Sistemi tecnologici, valutazione della ricerca e politiche per l’innovazione’.

Preface Martín Puchet Anyul and Lionello F. Punzo

The year 1999 saw a prompt recovery of the Latin American economies from the recession of the year before, in particular the partly unexpected positive performance of Mexico. There, the economic situation now looks much brighter and attractive for investors and industrialists alike, from Europe as well as Latin America itself. The quick improvement is due to a favourable international conjuncture (with a prompter than expected recovery of the East Asian economies which caused the turmoil initially), high commodity prices (of which Mexico is a major exporter, particularly of crude oil) and the continuing boom of the US economy, which keeps increasing Mexican exports (of manufacturing, as well as agricultural products, though on a less relevant scale). Of course, the response of Mexico was comparatively better. The economy was set on a 3.7 percent expansion, after the 5.8 percent of 1998, and with a forecast 4.8 percent for the year 2000; the annual inﬂation rate to June 2000 is down to 10.04 percent (with an end-of-year forecast of 9.2 percent); the peso-to-dollar exchange rate is fairly stable, and ﬁnally the rate of open unemployment seems to have settled around an almost incredible 3 percent. These results have been attributed to far-reaching, strengthening effects of reforms and other structural changes induced by the enforcement of the NAFTA agreements. NAFTA has become a center of attention, as an exemplary success story. This book is appearing, therefore, at an important time not only in the history of Mexico but, as we will see, also in the history of the European Union. Being written largely by Mexican economists and edited and commented on by their European counterparts, such coincidence is present in the authors’ minds. At the time of writing this preface, two events have just altered deﬁnitively Mexico’s political scene. First, a so-far tranquil change of ruling party, after 71 years, is underway. On the other hand, the entering into full force of the so-called Global Agreement between the EU and Mexico1 marks the beginning of a new era in their general relations. It is interesting to notice the coincidence. The international press has viewed Vicente Fox’s victory as among the healthy effects of the NAFTA agreements which, together with the EU treaty, has been an

Preface

xvii

accomplishment of the previous presidencies (respectively, Salinas’ and Zedillo’s). The political changeover, it is said, is the result of the votes of a new generation who have grown with the NAFTA and are investing their lifetime expectations on it. Likewise, this generation will be working in a phase of new relations between the two parties on opposite sides of the Atlantic Ocean. Thus, we are now seeing the early fruits of a long, manifold process of structural change, one step of which has been the signing of the NAFTA agreement in 1993. The present situation seems to be only a temporary end or halfway house: while closing a phase, it is opening another one, with some quite unknown perspectives. 1. The ﬁnalization of the Global Agreement2 with the EU is the result of a risk management policy for Mexico’s portfolio of political and economic alliances. This had a fast acceleration in the last year of the previous administration, with presidential (Chirac’s and Zedillo’s) visits to both sides of the Atlantic and the conference of the Hibero-speaking countries in Spain and Portugal. However, political proximity between Europe and Mexico has been steadily increasing in the recent years, in parallel with the progress of the integration processes produced by NAFTA and the EU. The EU, through its operative arm, the European Commission, was recognizing a global change of policy towards Latin American countries. A major turn around is feasible, and advisable, now that Monetary Union is apparently underway. It is, however, based within an articulated strategy of counter-balancing the construction of a massive European fortress with bilateral and multilateral agreements. This strategy is but the continuation of a subtler policy developed so far by Europe, to ‘keep doors open’ and at the same time to compete with other major countries’ strategy of playing different cards simultaneously on the international stage, e.g. regionalism at the same time as global liberalization, continental integration with EU, NAFTA, MERCOSUR and the like, at the same time as keeping the WTO negotiations alive. The 1997 South East Asia crisis has spurred such multiplayer, multicountry games. Hence, there is a play going on between Mexico and the EU, but also between the US and the EU, and they are played basically around FDIs. In fact, prudence suggests diversiﬁcation while the long-term prospects still remain uncertain, and the Union has attracted in 1999 the greatest share of US industry-related investment (US $46 billion). The Global Agreement may, of course, reﬂect European fears of losing ground on the North American continent. The EU is Mexico’s second biggest commercial partner after the US and, similarly, Mexico’s second largest direct foreign investor. European FDIs have increased by 252 percent between 1993 and 1996, and this requires a legal framework to protect them and their long-term prospects. Likewise, the trade balance between the EU and Mexico might still not be terribly signiﬁcant in terms of ﬁgures, with only 1 percent of total EU exports and 0.6 percent of

xviii

Preface

imports involving Mexico; and 9 percent of Mexico’s exports and 5 percent imports involving the Union. Still, the expectations of greater exports to Mexico have increased recently as a result of the devaluation of the Euro against the US dollar. The market may still be small but it is full of opportunities. However, the progress of NAFTA has caused EU shares to decline. With the possibilities offered by the complementarity in export–import structures between the two areas, and of using Mexico as a platform to access the US market, this evolution is clearly seen in Brussels with some preoccupation. But the ﬁnalization of the Global Agreement shows similar worries on the part of Mexico. Zedillo’s active presence on the European scene in the last semester of his mandate demonstrates his intention to diversify the country’s foreign dependence, reducing the relative weight of the US partner by encouraging European FDI that could also help ﬁnd new export niches in the European markets. Broader political effects can also be anticipated for the future. So far relations have been quite unbalanced in favour of the richer and stronger neighbour, which could propose rules and regulations expecting they were accepted as such (as is reﬂected in the ‘imperfect’ NAFTA agreements, as has often been bitterly remarked). It is clear that what we can expect is a continuation of the policy of rendering the country more independent, politically as well as economically. As noted in the press,3 from now onwards relations between the two will be on a more equal footing, as a result of the recent political changeover but also of the enlarged option set provided by such treaties as the one with the EU. A substantial increase of FDI ﬂow from the EU would also have other important economic effects. It would contribute to Mexico’s macroeconomic stability by generating ﬁnancial inﬂows that are necessary to equilibrate an external balance systematically in the red. But an associated effect would perhaps be more important: by concentrating on sectors that are not privileged by US investment, they may reduce the structural fragility of the economy. Therefore, FDIs from Europe can become a strategic element in the country’s future developments, at the same time as making it even more interesting for European investors. Correspondingly, it will become more important for Europe to have ﬁrst-hand information about that area. 2. Still, in Europe, the principal source of information about Mexico and its economy ﬂows from US mass media and specialized literature, where the focus tends to be on NAFTA (and the US). This book presents a set of contributions and assessments that are founded upon a distinct perspective. Altogether, they propose to evaluate NAFTA against the backdrop of the long sequence of facts that have accompanied the integration process in North America. But, ambitiously enough, they also propose themes for reﬂection on the European side of the Atlantic. Hence, the subtitle emphasizes the European perspective, with a two-fold

Preface

xix

meaning. On one side, as has already been said, it stresses the increasing involvement of the EU into the events of Latin America, and in Mexico in particular. On the other side, it proposes themes for a comparison of the integration process between unequal economies, typical of NAFTA, with the likely evolution of the Central Eastern European countries, after their accession to the EU around 2003 (or 2005). Sizes are comparable, and some problems are, perhaps, similar. Mexico has the same population as the group of countries to be admitted to the EU in the ﬁrst round (100 million people); standards of living and wage differentials are comparable; production structures are not dissimilar, with a great reliance on agriculture, compared with the dominant economies, and weaker industrial and service sectors. However, the set of issues is roughly the same. First of all, it is the issue of labor mobility, which is at the very core of public attention in the US and Mexico (and recent borderline events have shown their radical dimensions). But the same problem is also becoming increasingly important in many of the European countries (signiﬁcantly those with a long border with the CE countries, and in Italy) where fears of an ‘invasion’ are being aroused by politically vested interests (while the continent as a whole is badly in need of a net inﬂux of young, whether highly specialized or unskilled,4 workers). Likewise, other sensitive areas where potential conﬂicts can arise are agriculture and the environment (strong in the European case), not to forget illicit trafﬁcking. Beyond the strict needs of the economies linger the problems of political democracy, civil rights and institutional stability, that have all ﬁgured high on the list of pre-conditions set by the EC for moving closer to Mexico and the CE countries alike. Of course, NAFTA, the process(es) it has initialized, as well as those that have led to it, have been the key factors in Mexico’s recent history. Mexico is now a patchy country (probably even more than before). Geographical and sectorial differences (the country’s ‘two-fold dualism’) have worsened; fairness in income distribution does not seem to have improved; the average real income is still below the 1982 level. Still, the country is on the move, and it actually operates at the forefront of the set of emerging economies. Compared with other Latin American countries, its position is well established. Moreover, accepting the risks involved in establishing free trade with the two largest economies of the globe has put the country in a better condition to face the challenges of globalization. This book is about Mexico. The contributions presented hereafter are oriented, mainly, to the decision makers, journalists, experts and academics who, in Europe, are interested in the new EU–Mexico agreement and its short- and long-term effects. In this light, the authors’ aim is to highlight the fundaments as well as the obstacles to the economic integration between the economies of Mexico and the US by resorting to ﬁrst-hand information about economic facts (from ‘insiders’ on the Mexican side) with an effort to reduce the ideological bias that tendentially plagues available interpretations. The process of realization of NAFTA is interesting not

xx

Preface

only for what it means in itself. It may also become a key reference tool to anticipate the limits as well as the possibilities inherent in an agreement intended to reduce economic and political distance between one country (Mexico), and perhaps through it the whole North American continent, and the European Union, a political and economic reality of a magnitude deﬁnitely never seen before. Thus, the authors of the various chapters are specialized in the different aspects of the Mexican economy and the integration process with the US. Hence, also the decision to have, in each part of the book, European economists responding to the sometimes varied interpretations expressed by providing views from the European perspective. The majority of the existing literature on NAFTA stresses its importance in warranting the continuity of the Mexican reform plans in economic policy that have been implemented since 1982. Moreover, it focuses on the assessment of the effects of such agreement and other related political arrangements, and measures the economic evolution of the partners. Somehow, in contrast to the dominant perspective, in this book the dynamics of economic and in particular production structures, the pattern of international trade, their tendencies and behaviors are the key objects of study, and they are proposed as the factors accounting for institutional as well as policy changes. The points of view of the different authors can at times be macro or micro, or even ‘meso’ (to refer to their chosen levels of aggregation); likewise, they may analyze the whole country, some sectors or regions of it. However, all of them consistently follow the line of reconstructing the long- or medium-term evolution – the last ten years or more. This common perspective takes as its principal reference the European tradition to understanding processes of integration. 3. In detail, the four parts articulating the structure of the book focus upon distinct and important aspects of the relationships between Mexico and its principal partner. Their common leitmotif can be described, in synthesis, as an analysis of the multifaceted process of integration between economically unequal (size-wise) partners, who also have different dynamic structures, technologies and institutions. Like all processes of mutual integration, this too encompasses much more than simple market and production integration. Highly aggregative analysis, so fashionable in mainstream growth empirics, shows its limits for it can only render and map out differences related to size, rather than, say, to composition and structure. It has thus to be integrated and enriched with vistas provided by more detailed, disaggregated analytical perspectives. Chapters in Part 1: Paths to Convergence and Monetary Union(s), report on the much-debated subject of the feasibility of a monetary union (dollarization and/or currency board) between Mexico and US. The issue is discussed in the light of policy experiences and in a political economy perspective as well as in more technical terms. The initialization of the last-but-one step towards monetary union in Europe, with the creation of

Preface

xxi

a conceivable alternative to the dollar, has given further momentum to opinion movements in Latin America pressing for a similar measure. The perils of the creation of two competing giant currency areas, with its shortcomings for the process of world trade and ﬂow liberalization, are evident to the informed audience. In Part 2: Mexico’s Foreign Sector, the Exchange Rate, Trade and Productive Structures, focus is placed upon the link between trade and productive structures. The determination of the exchange rates by relative labor costs and the involved effects upon level and composition of the external balance is analyzed from different perspectives, while the observed repercussions of trade liberalization on the movements of production factors and structures show both the role and limits of NAFTA. This theme lends itself to immediate, though certainly not easy, comparison with likely scenarios that will emerge in the EU following the integration of thirteen neighbouring Central Eastern European countries, whenever this will take place. Part 3: Sectorial Performance in Manufacturing, Industrial Organization and the Innovation System During the Trade Liberalization Process, illustrates some of the functional and dynamic linkages between trade liberalization, macroeconomic evolution and the new tendencies in the manufacturing industry. Mexico here offers a whole set of case studies. The forms of industrial organization and the principal features of the institutional innovation system are studied following several interpretation lines. Finally, Part 4: Regional Divergence and Structural Policies assembles papers dealing with differential regional development driven by the integration process and focusing on the necessity of differential policies to modify the tendencies towards greater inequality. The uneven patterns of structural change are here at the core of the descriptive part, and dynamic dualism is proposed as the interpretative scheme. Again hot topics in the EU are so-called social cohesion and regional disparities that form background and motivation for a volley of EU structural policies. Their redeﬁnition goes together with the issue of ﬁscal federalism. The former is the basic concern for the ﬁnal stage of European uniﬁcation, the latter has become the battle cry for many a political movement across the old continent. They offer ideal themes for comparison across the two continents.

Notes 1 The full title is Economic Partnership, Political Co-ordination and Cooperation Agreement, scheduled to take effect on 1 July, 2000. (The Mexican elections are scheduled for 2 July.) It is superseding the 1991 Framework Cooperation Agreement, as well as the Interim Agreement on Trade and Trade-Related Matters, in force since 1 July 1998. Negotiations started in 1998 and were completed on schedule at the end of 1999. The Global Agreement was signed with Mexico on 8th December 1997. The part of the Global Agreement that remains to be

xxii Preface ﬁnalized is the Free Trade Agreement. At the time of writing: ‘The chapters on services, investments and IPR the negotiation of which was ﬁnalized in November 1999, must await completion of the ratiﬁcation procedures by Member states and Mexico of the EU–Mexico Economic Partnership, Political Co-ordination and Co-operation Agreement signed with Mexico in December 1997. This is expected to take place during the French Presidency’ (IP/00/703, Brussels, 3 July 2000). 2 For a brief description: ‘In terms of coverage, this package will provide EC operators with more rapid preferential treatment than Mexico has ever before granted to any of its preferential partners and will place them in a much better position to compete on the Mexican market which is strategically important and has a signiﬁcant growth potential. All industrial goods will be free of tariffs by 2007. In trade volume, 52 per cent of EC exports will enter the Mexican market duty free by 2003 and for the remaining 48 per cent a maximum duty of 5 per cent will be applied by 2003. Whilst preserving EC sensitivities for agriculture and ﬁshery products, the package negotiated for these products will grant quick and full market access for the EC’s most important export products. For services EC operators will be granted better access than that currently enjoyed by Mexico’s other preferential partners and in particular the USA and Canada. This package will be completed by a commitment to liberalise investment and related payments. Substantial access to the Mexican procurement market, similar to NAFTA, will also be secured. Finally, the package will include substantive disciplines for competition matters, the protection of intellectual property and an effective dispute settlement mechanism.’ Communication from the Commission to the Council and the European Parliament, accompanying the ﬁnal text of the draft decisions by the EC–Mexico Joint Council (COM(2000)9, Brussels, 18 January 2000, extract). A useful information site is: http://www.europa.eu.int/comm./ external_relations/mexico and for the complete version of the Agreement, see: http://www.europa.eu.int/comm./trade/bilateral/mexico/fta.htm 3 The Economist, 8 July, 2000. 4 See The Economist, 6 May, 2000.

Contributors

Fidel Aroche Reyes is a full-time professor on the graduate programme of the Faculty of Economics of the National Autonomous University (UNAM), in Mexico City. He has a Ph.D. from Queen Mary and Westﬁeld College, University of London. His research interests include: economic structures, structural change, economic integration, algorithms, applied input–output models and methods. Flor Brown is currently a lecturer on the graduate programme of the Faculty of Economics, of the National Autonomous University of Mexico (UNAM). She has a Ph.D. from the Metropolitan Autonomous University (UAM) in Mexico. She has published on industrial productivity, technical change and industrial linkages and networks in Mexican industry, with special emphasis on the automotive industry. María Elena Cardero is a full-time professor on the graduate programme of the Faculty of Economics, UNAM. She has a Ph.D. in economics from the National Autonomous University of Mexico (UNAM). She is the leader of the NAFTA Unit within the Foreign Ministry; a member of the Mexican negotiating team for the Energy Chapter; and was appointed as General Coordinator of Economic Affairs of the Foreign Ministry. At present, she is Director General of Organisms of Economic Cooperation and Development at the Ministry of Foreign Affairs. She is currently on leave, working for the Ministry of Foreign Affairs. Recently she has conducted research on trade and ﬁnancial issues and their impact on economic growth. Horacio Catalán is currently a lecturer at the Faculty of Economics, UNAM. He has a Masters Degree in economics from UNAM. He has written several articles about monetary policy in Mexico. He is currently working at UNAM, teaching applied econometrics at the Faculty of Economics. Mario Cimoli has been Professor of Economics at the University of Venice (Ca’ Foscari) since 1990 and is currently Economic Affairs Ofﬁcer at the Economic Commission for Latin America and the Caribbean (ECLAC) of the United Nations. He obtained a Ph.D. at the SPRU (University of Sussex) and has held a number of visiting appointments in

xxiv

Contributors

different universities and institutions (Udine University, UNAM and CONACYT in Mexico). He has published many papers and books, including: Developing Innovation Systems: Mexico in the Global Context, (Pinter Publishers, 2000); Technology, Growth and Development (with J. De la Mothe, Pinter Publishers, 2000); ‘Technological paradigms, patterns of learning and development: an introductory roadmap’ (with G. Dosi in Journal of Evolutionary Economics, 1995), ‘A generalized technology gap trade model’ (with L. Soete in Economie Appliquée, 1992), ‘Technological gaps and institutional asymmetries in a North–South model with a continuum of goods’ (Metroeconomica, 1988). Lilia Domínguez is Professor in Industrial Economics on the graduate programme in the Faculty of Economics, UNAM. She is an economist with an MA in economics from the University of East Anglia, UK, and doctoral studies in the New School for Social Research, New York. She has researched on industrial organization, new technologies and productivity in Mexican manufacturing and entrepreneurial behavior in relation to pollution abatement and the possibility of green innovations. Enrique Dussel Peters has been an associate professor on the graduate programme in the Faculty of Economics, UNAM, since 1993. He has a BA and MA at the Free University of Berlin, and a Ph.D. from the University of Notre Dame. His main ﬁeld of research is on industrial organization, Latin America’s and Mexico’s industrial and macroeconomic development, and NAFTA. Publications include, with Michael Piore and Clemente Ruiz Durán, ‘Pensar globalmente y actuar regionalmente. Hacia un nuevo paradigma industrial para el siglo XXI’ (México: Editorial JUS/UNAM/Fundación F. Ebert, 1997); and Polarizing Mexico: the Impact of Liberalization Strategy (Boulder: Lynne & Rienner, 2000, forthcoming). Oliver Fabel was, until 1999, Professor of Management, in particular labor and industrial organization, at the Otto-von-Guericke-University, Magdeburg, and subsequently Professor of Management, in particular ﬁrm policy formation, at the University of Konstanz. His main research interest focused on the theory of the ﬁrm and incentives and risk shifting in organizations. Francesco Farina has taught at the Universities of Naples, Perugia and Toronto, and is presently Professor of Economics at the University of Siena. He is the author of books and articles on monetary economics, social choice and economic integration. He is co-editor of Ethics, Rationality, and Economic Behaviour (Clarendon Press, 1996) and of European Economies in Transition. In search of a New Growth Path (Macmillan, 2000). Jean-Luc Gaffard is Professor of Economics at the University of NiceSophia Antipolis and senior member of the Institut Universitaire de France. His major research interest is in industrial economics, economics of innovation, and dynamic economics. He has published many papers in scientiﬁc journals and several books, among them, The Innovative Choice.

Contributors xxv An Economic Analysis of the Dynamics of Technology (with Mario Amendola) (Oxford: Blackwell, 1988) and Out of Equilibrium (with Mario Amendola) (Oxford: Clarendon Press, 1998). Luis Miguel Galindo is Professor of Quantitative Methods, special chair Narciso Bassols at the Faculty of Economics, UNAM. He has a Ph.D. in economics from the University of Newcastle Upon Tyne and has published several articles in referee journals on the topics of monetary economics and macroeconomics. He has also worked as an advisor for various ﬁnancial institutions in Mexico. Bodo Hilgers ﬁnished his doctoral thesis with the title ‘International trade and involuntary unemployment’ in February 1999 at the chair of Professor Albert Schweinberger Ph.D. Since then he has been working at the chair of Professor Fabel as Assistant Professor. His research ﬁelds are labor-market theory, contract theory and its applications to organizations, especially universities in a general equilibrium framework. David Ibarra has been Minister of Finance, Director of a number of banking institutions including Nacional Financiera (Mexico’s leading Development bank) as well as Chief Economic advisor at PEMEX (Mexico’s oil company). He is currently a special advisor to the UN Economic Commission for Latin America and the Caribbean (ECLAC). He has published extensively on Mexico’s macroeconomic and political issues. Among his most recent publications are ‘Politica y economia: semblanza y ensayos’ (1999), ‘México: transición o crisis: las contradicciones de la politica economica y el bienestar social’ (1996), ‘El nuevo orden internacional’ (México, Editorial Aguilar, 2000). Sárah Martínez Pellégrini is the Dean for Academic Affairs at El Colegio de la Frontera Norte, Tijuana, Baja California, Mexico, where she has been Director of the Graduate Programs and Public Administration departments. Her ﬁelds of work are regional development, regional policies and international economic integration in which she teaches at MA and Ph.D. levels. Jürgen Meckl is Associate Professor at University of Konstanz. He has the following academic degrees: Diplom-Volkswirt, University of Regensburg (1988), Master of International Economics, University of Konstanz (1989), Dr. rer. pol., University of Konstanz (1993), Habilitation, University of Konstanz (2000). His research activities include: international trade theory, macroeconomic theory. Juan Carlos Moreno-Brid is currently a Research Associate at Harvard’s University David Rockefeller Center for Latin American Studies. He conducted his postgraduate work at the Faculty of Economics and Politics at the University of Cambridge, UK. He has been a consultant for various international organizations and, before joining Harvard in 1995, he was an advisor to the UN’s ECLAC. He has numerous publications on issues related to macroeconomic policies of Mexico and other Latin American countries.

xxvi

Preface

Martín Puchet Anyul is Professor of Quantitative Methods, at the Faculty of Economics, UNAM. He has a Ph.D. in economics from UNAM. His research interests include: national accounts, structural and applied input–output analysis and economic dynamics. Lionello Franco Punzo is Professor of Economics, in the Department of Political Economy, University of Siena, Director of the Degree Course in Economics of Tourism, University of Siena at Grosseto. He has been a visiting professor in a number of universities and research centers abroad. He is currently involved in IDEE, a EU research project ﬁnanced as part of the Targeted Socio-Economic Research within the Fourth Framework Programme for Science and Technology. His interests are in the ﬁeld of the study of structural change from the point of view of industrial and production dynamics, mathematical modelling of economic dynamics, and mathematical methodology. Michel Quéré is a research fellow at CNRS-IDEFI/LATAPSES. His main ﬁeld of research is economics of innovation and, more speciﬁcally, the analysis of R&D strategies of companies, inter-ﬁrms’ partnerships and science–industry relationships. Pablo Ruiz-Nápoles is Professor of Economics, Faculty of Economics, UNAM. He has a Ph.D. in economics from New School for Social Research. His ﬁelds of research are international trade and policy and economic theory. Horacio Sobarzo is a professor at the Center of Economic Studies (CEE), El Colegio de Mexico. He is Director of the Center for Economic Studies at El Colegio de México. He has a Ph.D. from Warwick University. His main ﬁelds are public ﬁnance and international trade. Previous research work includes NAFTA impacts in an applied general equilibrium model, and ﬁscal decentralization. Jaime Sempere is a professor at the CEE, El Colegio de México. He obtained his Ph.D. in economics at the European University Institute (Firenze, Italia) in 1994, and his research ﬁeld is local public ﬁnance.

Part I

Paths to convergence and monetary union(s)

1

Currency boards and monetary unions The road ahead or a cul de sac for Mexico’s exchange rate policy?1 David Ibarra and Juan Carlos Moreno-Brid

The world, now that commercial frontiers have been erased, is en route to the elimination of monetary divisions. The process is already underway and it seems that, once again, it will catch us unprepared. The reason for this lies in the innate, or imposed, global instability which burdens governments and economic agents operating in the markets of Mexico and Latin America, overwhelming them with uncertainty. These uncertainties have become more acute in recent years, as a result of the rapid and unrestricted movement of global ﬁnancial capital in amounts which exceed, many times over, the international reserves of the majority of countries in this region. Confronted by unsatisfactory, contradictory or utopian solutions proposed by economists and other experts, there have been lapses of behavior that border on the irrational or that clutch at elusive miraculous cures. Under present circumstances, to ﬁx the nominal exchange rate by decree tends to mirror an ideological posture: displaying complete distrust in the abilities and goodwill of governments, together with a blind faith in the markets’ self-correcting and optimizing mechanisms. Both extremes are false. Some state interventions and regulations are indispensable, and dangerous market disarrays can exist, as exempliﬁed by the multiple ﬁnancial crises occurring today and in the past. The apologists of a world without frontiers regard the formation of a zone of monetary integration among members of the North American Free Trade Agreement as imminent. In their view, the creation of a currency board – eliminating, once and for all, any variation between the peso and the dollar – would be even simpler. The cost of surrendering monetary sovereignty is deemed to be of little account. This course of action, they afﬁrm, would wipe out inﬂation at a stroke, interest rates would fall to levels comparable to those of the First World, the uncertainty surrounding the exchange rate would stop blurring investment decisions, foreign trade transaction costs would be substantially reduced and the fragile banking system would, at last, be put back on a healthy track. The cases of the single European currency, the West African Monetary

4

David Ibarra and Juan Carlos Moreno-Brid

Union,2 and the currency boards in Bosnia, Brunei, Djibouti, Hong Kong, Estonia, Lithuania or Argentina3 are promoted as irrefutable proof of the viability of proposals for monetary reform, without any reference to the speciﬁc nature of each case. What is overlooked, due to candor, ignorance or convenience, are the conditions that must be fulﬁlled and the price that must be paid to ensure even the minimal success of such proposals. The purpose of this paper is to examine these questions as they apply to the Mexican economy. The work is organized in three sections. Following this introduction, section 1.1 analyses the viability and the necessary requirements to establish a currency board, with particular reference to crucial characteristics of Mexico’s productive and ﬁnancial structure. Section 1.2, after a brief review of the progress made in the commercial integration between Mexico and the United States, analyzes the extent to which a monetary union between them would boost, or perhaps hinder, Mexico’s economic development. There are two main aspects to this analysis. The ﬁrst is a study of the intermediate goals established by European countries, long before their monetary union was formalized, with respect to harmonizing macroeconomic strategies and creating institutional means to supervise and regulate their labor and capital markets. The second aspect is to reveal certain elements of Mexico’s institutional organization and the operational dynamics of its labor, goods, and services markets, as well as its banking system, which are far from fulﬁling essential requirements to establish a viable monetary union with the United States. The work closes (section 1.3) with a series of ﬁnal considerations. Here, based on the foregoing analysis and emphasizing the profound asymmetries between the Mexican and the United States’ economies, it is concluded that for Mexico, at present and in the near future, the best option is a ﬂoating exchange rate. The alternatives of ﬁxing the nominal exchange rate of the peso against the dollar, via a currency board or monetary union, will end up critically deteriorating Mexico’s international competitiveness and, more seriously, making domestic output and employment more vulnerable to the impact of future external shocks. Such negative effects, besides hindering Mexico’s insertion into a path of rapid and sustained economic growth, could damage its internal social cohesion and may even threaten its political stability.

1.1 Advantages and limitations of currency boards: an overview The deﬁning characteristic of a currency board is the legal obligation on the part of the monetary authority to maintain international reserves in excess of the total amount of coins and bills in circulation. Thus, it is a system that displaces and narrows down the traditional functions of the central bank, transforming it into a mere convertibility window forced to

Currency boards and monetary unions

5

redeem, without limit, all monetary liabilities at the stipulated exchange rate, dispensing with any attributes of being lender of last resort beyond the limits imposed by the amount of foreign reserves. In effect, it is a radical, extreme means of returning to the gold standard; that is, of freezing the nominal exchange rate and eliminating its use as an instrument to adjust to external disequilibria.4 Such redirection of exchange rate policy towards the single-minded achievement of the goal of price stabilization has proved, without doubt, to be effective in reducing inﬂation. But this success is invariably accompanied by the progressive loss of international competitiveness which, sooner rather than later, erupts into a balance of payments crisis. The Mexican economy has experimented on several occasions with ﬁxed nominal exchange rates. Indeed, its evolution over the past 40 years has been marked by episodes during which macroeconomic policy has had as its primary – and, on occasion, unique – objective to eliminate inﬂation. In this quest the peso–dollar nominal exchange rate was ﬁxed at levels that gradually undermined the growth potential of the tradable sector and, therefore, effectively slowed down the overall rate of economic expansion. During these episodes the goal of economic growth was perennially subordinated to the goal of reducing annual inﬂation to a single digit; price stabilization that, it may be added, was at most achieved in a transitory manner. The results have proved disastrous. The existence of severe structural imbalances was overlooked and, rather than being resolved, they became more acute, thereby making it more difﬁcult for the Mexican economy to achieve a strong and persistent expansion. The ﬁrst necessary condition to create and sustain a currency board is to have sufﬁcient foreign reserves to fulﬁll convertibility requirements, as well as to gain the requisite internal and external credibility.5 This condition is not easily fulﬁlled if, in addition to M-1, other closely linked monetary substitutes are included.6 However, of greater importance is its implicit obligation, not only to surrender monetary autonomy, but also to subordinate ﬁscal policy to sustain the ﬁxed exchange rate; an economic objective that would have the utmost priority.7 In other words, everything would be subject to the ﬁxing of the exchange rate. This was possible during the ﬁrst years of the gold standard only due to the political alienation of a great part of the population at the time, as well as a lack of a clear understanding of the relationship between monetary regulations and economic activity.8 Once these conditions began to disappear, the gold standard collapsed in the face of pressures from structural imbalances in foreign trade and from domestic social upheaval. Within such a system, external shocks would force the government to expand – or to reduce – its expenditure or taxes in order to modify the level of economic activity in accordance with the goal of maintaining unaltered the nominal exchange rate. Simultaneously, it is likely that the strategy of outward-oriented growth and the satisfaction of long-time

6

David Ibarra and Juan Carlos Moreno-Brid

postponed popular demands would enter in contradiction with the obligations assumed in establishing the currency board. 1.1.1 Economic policy, productive structure and external adjustment in Mexico The Mexican government’s room to maneuver economic policy was reduced when it relinquished protectionist practices, and was further restricted by deregulation and ﬁnancial liberalization, the weakening of development banks, and by the gradual elimination of virtually all industrial policies.9 In contrast, it was necessary to deal simultaneously with multiple social demands and with the extremely complex problems associated with the transition from an interventionist and protectionist model to economic liberalization. At present, the main policy measures still available to the government to achieve external balance are the, more or less, free ﬂoat of the exchange rate, together with the manipulation of interest rates which, by the way, is left to be determined by market forces. When you implement an exchange rate policy of managed ﬂoating geared at eliminating domestic inﬂation (as Mexico did in the ﬁrst half of the 1990s), manipulation of domestic interest rates remains as the principal – almost the only – policy instrument to try to achieve external balance. In this case, the use of interest rates in response to other objectives – for example to facilitate the microeconomic adjustment of businesses or sectors to changing conditions of foreign competition – would frequently be impossible or, at the very least, incongruent.10 Consequently, in practice it would be necessary to resort to ﬁscal policy or to the use of relatively outmoded instruments such as the legal reserve requirements or the so-called ‘short-circuits’ by Banco de Mexico in its open market operations. In practice, the adjustment would fully depend on these instruments, because the use of ﬁscal policy for shortterm macroeconomic stabilization purposes has become particularly difﬁcult in Mexico due to the shrinking share of public expenditure in the economy, and the structural and transitional impediments to increase ﬁscal revenues.11 The implicit logic in ﬁghting inﬂation via an appreciation of the real exchange rate, and in increasing domestic interest rates to attract foreign capital with which to close the trade gap, led to the current plight of the banking system and of small and medium-sized manufacturing and agricultural enterprises. The fall in prices – but not in all costs – provoked by the abrupt liberalization of foreign trade, made it impossible for many producers, particularly those of tradable goods and services, to comply with the repayment schedules on their loans. Bank portfolios deteriorated and credit renewal was suspended. In the choice between defending the exchange rate parity or defending banks and businesses, government policy opted for the former, eventually

Currency boards and monetary unions

7

provoking the 1994–5 crisis, at which point no level of domestic interest rates proved high enough to attract the required ﬂow of foreign capital. This policy transferred, and will continue to do so, an important part of the costs of structural adjustment onto the shoulders of the banks and their debtors, while simultaneously undermining the level of competitiveness of domestic producers of tradable goods for export or for internal consumption; a situation which clearly contradicts the essence of any strategy for outward growth. Indeed, the recent currency crises in Brazil, Ecuador, and in certain Asian economies have exposed the futility – and the enormous economic and social costs – of insisting on raising domestic interest rates to defend an over-valued exchange rate peg. These policies deepen the recession, postpone the adjustment of relative prices and the reallocation of resources towards the production of tradables. Ultimately, they exacerbate the loss of international reserves and culminate in even more drastic depreciation of the exchange rate than those initially required to correct the balance of payments disequilibrium. The potentially severe impact of the high mobility of international capital has led many experts – even within the International Monetary Fund – to acknowledge that developing economies may beneﬁt from measures which tax, or temporarily restrict, the entry of short-term capital.12 Chile’s experience deserves special attention, as it was through the imposition of measures to levy taxes on the entry of short-term foreign capital that it was able to alter the composition of its foreign capital inﬂows without necessarily affecting its volume, thereby signiﬁcantly reducing the negative impact of the so-called ‘Tequila Effect’ prompted by Mexico’s 1994–5 balance of payments crisis. China’s rigorous controls on its capital account have enabled it to isolate itself considerably more from the pernicious consequences of volatile foreign capital ﬂows – which wreaked havoc among its neighboring economies. These policies and restrictions represent the opposite pole to the adoption of currency boards and monetary unions now regarded by some as guarantees of economic stability and growth.13 Put differently, an underlying condition for the success of a currency board is a robust banking system – capable of not having to rely on the central bank or the government as lenders of last resort – and a fundamentally robust business sector that, together, can shoulder the ﬁnancial obligations and grant credibility to the proposal to change the responsibility of the central bank. Are these requirements being satisﬁed in Mexico? Have we forgotten the episodes of 1976, 1982 and 1994, when balance-of-payments crises erupted once it became impossible to sustain the exchange rate parity because the conditions to do so on a permanent basis had not been created; or, more recently, those related with the rescue of the banking system?14 To ﬁx the exchange rate is equivalent to off-loading part of a multiplicity of functions onto the manipulation of interest rates. This would

8

David Ibarra and Juan Carlos Moreno-Brid

necessitate highly ﬂexible prices and wages. Indeed, any adverse shock in the balance of payments, besides requiring a cut in domestic expenditure, would have to be accommodated by a reduction in prices and wages to keep the nominal exchange rate unaltered. Given the inertia in Mexico’s labor markets, and the dynamics of monopolistic competition in a vast part of the market for goods and services, achieving the required reduction in prices and wages to absorb the external shock would surely lead to a fall in production and employment even more severe and protracted than those deemed necessary under a ﬂoating exchange rate regime. Consequently, a currency board would imply greater sacriﬁce on the part of workers and small and medium-sized businesses – through the fall in their prices and their failure to thus beneﬁt from possible gains in productivity. This would be tantamount to placing an even more unjust share of the sacriﬁces inherent to the macroeconomic adjustment to negative external shocks on these two particular segments of the population and the economy. It should also be noted that, at present, two other key requisites for adopting a ﬁxed exchange rate regime are lacking. The ﬁrst is related to Mexico’s imbalance in the external accounts which, far from being solved, has become more pronounced.15 Trade liberalization, despite its undeniable merits, actually weakened domestic inter-industrial links and, consequently, brought about a higher propensity to import inputs, consumption goods, machinery and equipment. This increase in import propensity is of major concern as, notwithstanding the commendable boost in exports of several manufacturing industries – favored by NAFTA – it is still proving impossible to generate the amount of foreign currency required to ﬁnance a robust and sustained growth of the Mexican economy without exerting excessive pressure on the current account of the balance of payments. Thus, the foreign exchange constraint, endemic to the Mexican economy, is now gaining renewed and unexpected strength. This is evidenced by the fact that foreign purchases are expanding in real terms at a rate almost three times higher than domestic output, thereby placing a tight limit on the sustainable rate of growth (3–4 percent annually) of the Mexican economy. Such a growth rate does not solve the problem of absorbing the excess labor force, nor will the growth in productivity be necessarily compatible with an adequate dissemination of the beneﬁts of economic development. Thus, having discarded the instruments of trade protection and of regulation of foreign capital ﬂows, there is no other alternative than to adopt a system of ﬂexible exchange rates or to use (subject to the risks indicated) the interest rate to correct the growing imbalances in the balance of payments, even though this entails costs expressed via increased price instability.16 Like it or not, the effective alleviation of the balance of payments constraint implies strengthening the domestic content of production. The growth of exports in itself is not enough. It is necessary that increasing quantities of imports must be efﬁciently substituted so as to bolster the

Currency boards and monetary unions

9

foreign trade balance, and thus ensure that the outward-looking macroeconomic strategy indeed acts as an engine of growth for Mexican output and employment. This does not refer to utopias. Through similar means, and without jeopardizing their insertion in the global economy, Japan and the Asian Tigers succeeded for decades in balancing their external accounts while still being able to accumulate a huge amount of foreign reserves. China, based on similar policies, has systematically achieved high rates of investment and economic growth. The second requirement, also difﬁcult to fulﬁll, consists of bringing Mexico’s inﬂation in line with that of its main trading partners.17 Although the adoption of a currency board could indirectly assist in achieving this objective – by creating favorable expectations – it would fall short of eliminating the root cause of the problem.18 Mexico is faced by institutional deﬁciencies which make it rather risky to try to control inﬂation by any means other than those implying low economic growth or unemployment. First of all the rigidities in the labor and goods markets, and the fact that the manufacture of non-tradable goods registers lower productivity gains and higher increases in prices – relative to activities subject to foreign competition – tend to exert inﬂationary pressures in the Mexican economy. Moreover, the transition towards a strategy for outwardoriented growth implies huge reallocation of resources and employment from stagnating sectors to new, dynamic activities; particularly those geared towards exports. This same strategy subjects businesses to a merciless competitive discipline, which forces them to reduce costs, employment, and expenses or to invest in new equipment. It also forces workers to accept greater job insecurity or to surrender previously acquired privileges. All this cannot be achieved instantly; it takes time to gain concrete form. Rigid positions must be countered and the defensive attitudes of sectors, businesses, unions or workers whose interests are affected in the transition from one model to another must be overcome. The shift in the exchange rate and the rise in interest rates, the elimination of subsidies and, in general, the alterations in relative prices which invariably accompany the transformations underway will most likely face resistance, as they induce uneven price and wage movements which signiﬁcantly alter the distribution of income.19 Accordingly, one of the lubricants for the economic transition can be found in the acceptance of higher than single-digit rates of inﬂation, which will slowly but surely transfer income between sectors and economic agents, thereby tempering the underlying social tensions and smoothing the adjustment in the productive structure. By virtue of the conﬂicts between objectives and those structural or transitional faults which exist in a world where vast volumes of foreign capital can be mobilized in an instant, it is doubtful that nations will be able to permanently peg their nominal exchange rate. Central banks not only suffer from a loss of credibility where their discipline in ﬁghting

10

David Ibarra and Juan Carlos Moreno-Brid

inﬂation is concerned, but their ability to resist the destabilizing pressures exerted by the external ﬂows of capital is also questioned.20 The adoption of a currency board, therefore, invites speculative attacks akin to that which affected the Mexican economy in 1994–5, or that which was bearing down on Brazil at the beginning of 1999.21 In addition, this decision is equivalent to wearing a straightjacket: the cost of turning back or of subsequently altering the exchange rate increases exponentially as it strikes at the very heart of economic agents’ expectations, justiﬁed or not.22 In the best of circumstances, freezing the exchange rate would generate temporary gains by reducing the levels of uncertainty, the rates of interest, and the pace of inﬂation. In the long run, however, as long as no real cures emerge for the structural imbalances, it would be illusory and costly to expect the exchange rate parity to be ﬁrmly anchored by a currency board. In pursuit of this pseudo-stability it is quite likely that economic growth and employment would be sacriﬁced in the expectation that increases in productivity could strengthen the foreign trade accounts and boost the reconstruction of Mexico’s productive apparatus.23 This would be a slow, treacherous and perhaps nonviable course, as it would multiply the losses in income and production and accelerate processes of marginalization and ungovernability; ultimately increasing the probability of detonating an additional balance of payments crisis. It is regrettable that miraculous cures or immediate solutions to the crises that plague the external economic relations and hinder a smooth process of capital formation are not forthcoming. Oversights in Mexico’s transition towards a globalized pattern of economic development have resulted in an enormous and long overdue task that is far from being accomplished: rebuilding our productive structure and institutional organization. Moreover, we lack the national and international regulations to control the unsettling ﬂows of short-term capital that exacerbated the recent ﬁnancial crises. The shortcomings of the institutional framework that emerged from Bretton Woods that supports – but by no means regulates – the global dynamics of commercial and ﬁnancial ﬂows have led to a polarization of opinions concerning the exchange rate policies that should be adopted by developing economies. One side argues that the free ﬂoat – or at least one managed with great ﬂexibility – exchange rate regime is the only option compatible with the volatility of international ﬁnancial movements. This perspective acknowledges that the variability in the exchange rate is by no means an ideal situation for any economy. But it considers that, given the current architecture of the international ﬁnancial system, the alternatives lack viability and imply the imposition of an excessive burden of economic and social costs on the population in the event of adverse external shocks. On the other side are those who, as previously indicated, regard the instability of the exchange rate as the principal obstacle to economic development. They consider that overcoming such a situation is worth

Currency boards and monetary unions

11

more than the potential loss of monetary and exchange rate independence. Thus, they recommend that a currency board be adopted or, in extreme, that a monetary union be established that would lead to the unrestricted use of the dollar as the sole regional currency. A closer examination of the radical proposal, currently being advocated by some, to establish a zone of monetary integration between Mexico and the United States, and possibly Canada, requires immediate consideration of its soundness, viability and its prerequisites; clearly its implementation would make it absolutely essential to correct the aforementioned structural imbalances.24

1.2 A monetary union: a necessary step for Mexico’s sub-regional integration? The Mexican economy is highly integrated with the United States in its commercial and ﬁnancial transactions. Over 80 percent of Mexico’s imports and exports take place with its northern neighbor, and a vast proportion of its direct foreign investment is carried out by US ﬁrms. The macroeconomic reforms instituted in Mexico since 1985 have served to diversify the sectorial composition of the country’s foreign trade. Indeed, since then, and motivated in part by the international debt crisis and the collapse of the global oil market, Mexico has broadened its export base. Whereas it previously depended almost exclusively on oil exports, its foreign revenues now mainly derive from manufacturing exports; a successful transformation which has permeated almost every branch of Mexico’s industrial activity. While there is still a long way to go in the sense that the lion’s share of exports is still concentrated among relatively few companies and products, the dynamic reinsertion of Mexico’s economy into the international markets is undeniable and irreversible. This globalization is also evidenced by the greater import content in domestically produced manufactured goods. The import coefﬁcient in Mexico’s consumption has increased in all industrial activities, mainly due to a greater use of inputs and ﬁnished goods coming from the United States. Certain important entrepreneurs in Mexico’s export industries claim that the growing commercial and ﬁnancial interaction between both countries is a sufﬁcient basis to immediately move forward to establish a monetary union with the United States. This conclusion, however, is based on weak assumptions. The formulation of a project of this nature to be viable would necessitate a far greater degree of harmony and concordance in the economic dynamics of both countries. Prerequisites that, if unfulﬁled, would culminate in a costly and failed exercise.

12

David Ibarra and Juan Carlos Moreno-Brid

1.2.1 Lessons for Mexico from the European Monetary Union experience It may be convenient to start by examining the formation process of the European Monetary Union. Its key objective, well-known but perhaps not always made explicit, was to establish the foundations for a political community of nations that could once again lead Europe to the world’s economic pinnacle. In North America, this desideratum has not been contemplated or, at least, has not been suggested as a framework for political reference by any of the parties concerned in the sub-regional monetary union. As far as the control of monetary and exchange rates matters is concerned, for many years (since the late 1970s) Europe experimented with systems of policy coordination that took shape in the creation of the European Monetary System and the Exchange Rate Mechanism.25 The next stage consisted of determining, negotiating and establishing rigid criteria for key national economic policies to ensure the feasibility of a zone of monetary integration with a single currency and to create the necessary regional institutions that would later assume the functions of a central bank serving the community of European nations. The ﬁrst step to eliminate all remaining restrictions on capital ﬂows among the countries of the Community was taken in 1990. The European Monetary Institute was created in 1994 as a precursor of a common central bank. Finally, in early 1999, the foreign exchange rate parities of each country were ﬁxed and the common currency and the European System of Central Banks were launched.26 These institutional reforms were accompanied by vast efforts to harmonize individual economic strategies, which also took many years to bear fruit. The ﬁrst criterion stipulated that the consumer price index must not rise 1.5 percentage points above the average rate in the three countries of the Community with the lowest inﬂation (in 1998, in the case of Mexico and the United States this differential reached seventeen percentage points). Second, it was agreed that the interest rate for long-term government bonds must not exceed by more than two percentage points the average ﬁgure in those same three European countries (in this case a comparison between Mexico and the United States would reveal enormous differences – greater than twenty percentage points in the case of active rates). A better position is revealed when comparing European criteria on ﬁscal deﬁcits (no more than 3 percent of GDP) to that of Mexico (1.5 percent in 1998) or of the United States (surplus). Fourth, the total gross debt of the European governments must not exceed 60 percent of their GDP, a criterion which Mexico would satisfy even when incorporating the liabilities associated with the Bank Savings Protection Fund – Fobaproa – (totaling approximately 50 percent of GDP). Finally, where exchange rates are concerned, the countries have to keep them within the bands determined by the European Exchange

Currency boards and monetary unions

13

Rate System for at least two years prior to the Union, without resorting to any devaluation. In addition, as noted above, community norms were put in place to ensure the unrestricted movement of capital, greater facility for free transit of labor, and a common bill of social rights.27 The Pact for Growth and Stability, wherein these stipulations are recorded, also includes supervisory and regulatory measures for the common policies, as well as appropriate sanctions.28 By simple comparison it is easy to conclude that, for Mexico, the possibility of establishing a monetary zone with the United States is still a long way off, as neither the objectives and the necessary steps to achieve them, nor the timeframe or the required institutional framework have been negotiated or even identiﬁed. In practice, any rushed or inadequately prepared attempt to adjust the parameters of the Mexican economy to coincide with the standards that currently exist in the United States would most likely imply a descent into another painful process of adjustment, which would only pile on top of those experienced over the past two decades. Within any zone of monetary integration that encompasses several countries, the pace of economic expansion and job creation in countries burdened by trade deﬁcits depends upon the willingness of trade surplus countries to further expand their economic activity. Otherwise, adjustments would have to be made via reductions in production and employment in the economies laboring under trade deﬁcits.29 Thus, it may be argued that only a miracle would make prices in the United States rise exprofeso so that Mexico’s balance of payments reaches a stable equilibrium, including its transactions with third-party countries. In addition, in the event that inﬂation rates in Mexico and the United States should ever be similar, there is no guarantee whatsoever that this condition would be permanent. In fact, the inﬂationary processes in developing economies tends to be caused by deep-seated factors beyond the mere inﬂuence of external prices reacting to adjustments in the exchange rate. They may include distributive struggles, ﬁscal imbalances, the need to erode the excessive internal debt30 and the heterogeneous evolution of productivity in different areas of economic activity, together with inertia in the pricing mechanisms of goods, services and public utilities. These factors can impede keeping comparable rates of inﬂation within a monetary union, even supposing that the member nations started from similar levels of domestic inﬂation. If the increase in prices of Mexican products should continue at a faster pace than those of its competitors, a ﬁxed peso–dollar parity imposed by a monetary union would exacerbate the trade imbalance and would ultimately trigger another economic crisis. Under such circumstances, the monetary union would prove to be a burden for the tradable sector, ultimately increasing unemployment and under-employment and provoking more closures of businesses and factories throughout the country.

14

David Ibarra and Juan Carlos Moreno-Brid

1.3 Final considerations If only it had been possible to foresee the costs of the abrupt opening of the domestic market to foreign competition and of ﬁnancial liberalization; not with any intent to avoid Mexico’s insertion into the new global economic order but, rather, to have been better prepared, with greater care and without the gravely unequal distribution of the costs associated with this change. Hopefully, the lesson has been learned, and the same mistakes will not now be repeated regarding monetary and exchange rate matters. One should be forewarned that the imbalances which may subsequently appear after the formation of the proposed monetary union would prove even more difﬁcult to correct because it won’t be possible any more to use domestic monetary instruments or external ones (such as the exchange rate), on top of the binding constraints on ﬁscal policy’s room to maneuver. These ﬁnal considerations make it necessary to highlight signiﬁcant differences between the economies of Mexico and the United States, the most obvious of which is their relative size and political importance. Inevitably, any readjustments in the relevant institutional organization would not occur in the United States, but would take place south of the border. It is illusory to think that the Federal Reserve would modify its practices to give precedence to Mexico’s speciﬁc problems.31 In particular, it is unthinkable that the free movement of labor within the zone of monetary integration – Mexico United States – will be established, and thus facilitate equilibrium among the labor markets and ease the regional adjustments required to confront diverse economic phenomena (such as that existing in the European Economic Community and between the different regions of the United States). It should be stressed that no developing economy – irrespective of its monetary or exchange rate policies – is isolated from the impact of abrupt changes in the global arena that modify, suddenly and drastically, the demand for its exports or its access to international capital markets. Consequently, however credible a future binational monetary union may seem, its establishment does not guarantee that the Mexican economy will no longer face external shocks that will force it to adjust, as rapidly as possible, the relative prices of its tradable versus its non-tradable goods and services. The types of external shocks to be confronted and the degree of efﬁciency exercised in modifying domestic relative prices (essentially, the real exchange rate) will affect economic activity and employment. This, in itself, would test the viability and strength of the monetary union’s institutions. In fact, since Mundell32 it has been known that, within a zone of monetary integration, the absorption of external shocks is easier when they exert symmetrical effects upon its various regions.33 If the effects diverge, there is a risk of disintegration for the union or of a marked difference in

Currency boards and monetary unions

15

the distribution of the costs of adjustment. The asymmetries that abound between Mexico and the United States would make the implementation of a monetary union between them a difﬁcult task. As long as the dollar is the principal reserve currency, the United States counts on permanent and automatic ﬁnancing for its trade deﬁcit, thus relieving it from any obligation to correct its external imbalance. At the same time, this grants great autonomy to the Federal Reserve in the manipulation of interest rates, whether to ﬁght unemployment or the overheating of the United States economy. The rest of the world thus ﬁnds itself forced to absorb this increased liquidity either by piling up its reserves of American dollars or by accommodating an expansionary adjustment in its macroeconomic policy. Mexico’s necessary adjustment would have to be achieved either by accumulating greater foreign debt – within narrowed limits – or by reducing its economic growth and employment, or through a combination of both, given the imperfect integration of the capital and labor markets between the two countries. This, in itself, would impede the Mexican government from conducting the counter-cyclical manipulation of interest rates. Many other asymmetries exist between both economies. A sudden alteration in world oil prices has an almost inverse effect on the economies of Mexico and the United States. A cut in oil prices would result in smaller export and tax revenues for Mexico,34 translating into lower rates for sustainable economic growth, and thus inducing deﬂationary policies. In the United States the effects tend to be quite the opposite. The same could be said regarding the repercussions of the recent exchange rate depreciations in Southeast Asia on the ﬂows of trade and ﬁnance towards Mexico versus the United States. In Mexico the cost of adjusting to such external shocks, which by themselves would tend to deteriorate the well-being of a great majority of its population, could be even more severe due to the rigidities which the monetary union would a fortiori impose upon the shift of relative prices of tradable and non-tradable goods and services. Besides the mobility of factors of production, within the United States there are other compensatory mechanisms that tend to reduce the negative impact of contractionary shocks, be they of a regional or a national nature. These include unemployment insurance, ﬁscal measures, health and welfare beneﬁts and other speciﬁc means of support for depressed regions or those affected by adverse circumstances. In the European Community, the social support system is even more advanced and also has established measures to beneﬁt its less-developed or depressed regions within the zone of monetary integration. Mexico lacks these compensatory devices, and the core of its social security systems is barely able to protect a minority of workers employed in the modern sector of the economy. The unexpectedly wide debate provoked by the proposals in favor of a North American monetary union or the adoption of a currency board reﬂects, more than the actual soundness of their arguments, ideological

16

David Ibarra and Juan Carlos Moreno-Brid

dogma or the simple desperation of businessmen, bankers and even workers to ﬁnd rapid remedies to cure the instability of the ﬁnancial system and prices in the Mexican economy. The realities are, however, stubborn. If the desired objective is to be achieved – a question in itself open to debate in light of other national priorities – it is necessary to work to create the consensual conditions required for its attainment. It is not sufﬁcient to sit back and wait in expectation of a miracle, or to accept ﬂash-in-the-pan measures that cannot be sustained even in the short run, or to experiment with systems irrationally copied from other countries. All in all, for Mexico, the free ﬂotation of its exchange rate does not currently appear to face any preferable or less costly alternative. No doubt, the huge magnitude and unrestricted mobility of the ﬂows of ﬁnancial capital in this world without frontiers, and the widespread reluctance to regulate them, explain the numerous failed attempts to rebuild a stable system of exchange rates subject only to occasional adjustments of a structural nature.

Notes 1 Comments by Martin Puchet and Francesco Farina, as well as editorial assistance by Pari Kasliwal, are gratefully acknowledged. The opinions here expressed by the authors do not necessarily coincide with those of the institutions to which they are currently afﬁliated. 2 Created in 1912 among the former British colonies of Nigeria, Ghana, Sierra Leone and Gambia, it served for many years as a model to establish monetary unions. During the 1930s, such unions had proliferated among British colonies in Asia, the Caribbean and the Paciﬁc islands. See Schuler (1996). 3 See Hanke et al. (1993), Hanke and Schuler (1993), Humpage and McIntire (1995), and IMF (1997). 4 In the same manner that there is no truly clean ﬂoat of the exchange rate, currency boards too do not entirely eliminate the lender of last resort function of the central bank. 5 One of the most recent analyses of the convenience of renouncing the future use of certain economic policy instruments, in pursuit of gaining credibility, can be found in Giavazzi and Pagano (1998). 6 M-3, which includes banknotes, checking accounts and short-term bank and non-bank instruments, amounted to approximately 1,150 trillion pesos in December of 1998, that is, 110 trillion dollars. 7 The three main objectives of monetary policy are: a) to guarantee external convertibility, b) to autonomously determine domestic interest rates to counteract inﬂationary pressures and recessions, and c) to ensure a stable evolution of the exchange rate. Given that, at most, only two of these goals may be systematically achieved by it, the currency board would have to sacriﬁce the use of monetary policy as an instrument of stabilization of the business cycle (see Mundell and Swoboda 1988). 8 According to various experts, three circumstances under the gold standard explain the subordination of all social objectives to that of maintaining exchange rate convertibility. In the ﬁrst place, Keynesian theoretical structures linking macroeconomic performance to central bank policies did not yet exist. Second, the ﬂexibility of labor and other markets in the past century allowed

Currency boards and monetary unions

9 10 11

12 13

14

15 16 17 18

19 20

21 22

17

reductions in aggregate demand – induced by payments deﬁcits – to take place through price cuts rather than higher unemployment. Finally, until the close of the nineteenth century, workers’ organizations did not wield sufﬁcient power to ensure protection of their interests. As Eichengreen argues: ‘The credibility of the government’s commitment to convertibility was enhanced by the fact that workers who suffered most from hard times were ill-positioned to make their objections felt. In most countries, the right to vote was still limited to men of property . . . Labor parties representing working men remained in their formative years’ (Eichengreen 1996: 31). See also Bagehot (1874), Whale (1939), Bloomﬁeld (1959), Trevor and Floyd (1992), and Eichengreen and Flandreu (1996). See Clavijo and Casar (1994), and Ibarra (1994). Tinbergen’s principle should be recalled: each policy instrument should be associated with the accomplishment of one, and only one, objective. Among the members of the OECD, the Mexican economy has one of the lowest tax revenues as a proportion of its GDP (OECD 1998). This reﬂects, not so much a lax commitment to levy taxes, but the slowdown of the modern segment of the economy and the fact that the informal sector does not or cannot pay taxes. Such traits introduce a recessionary bias into ﬁscal policy from two sides, as it must defend the exchange rate and also counteract the downward trend in ﬁscal revenues inherent in the quasi-stagnation that has, on average, marked Mexico’s economic performance in the last 15 years. See Ul-Haq et al. (1996), and Krugman (1998). In Argentina, the currency board is far from resolving in itself the destabilizing cycles of emerging economies or the structural faults of its economic system. Faced by external shocks, the authorities must implement adjustment measures that affect, above all, the real variables – production, employment, income – which subsequently induce changes in the structure and level of prices depending upon the ﬂexibility of the different markets. The liabilities of the Fondo Bancario de Proteccion al Ahorro (FOBAPROA) that the government has been forced to assume as public debt add up to a total of 60 to 65 billion US dollars. On the other hand, it may be useful to recall that, in the wake of the so-called ‘Tequila Effect’, Argentina – notwithstanding its currency board – had to seek the support of the World Bank to avoid the collapse of a large portion of its banking system. See Moreno-Brid (1999, 2000). See Aghevi et al. (1991), Garber and Svensson (1995), and Obstfeld (1995). In 1998, inﬂation in Mexico was approximately ten times greater than in the US. Although in Argentina interest rates initially fell, they remained at a high level (11 percent in real terms during the second semester of 1998), while the trade deﬁcit exceeded six billion dollars (Economic Commission for Latin America and the Caribbean 1998). See Ibarra (1996). A key element in establishing the regime of pegged exchange rates of the Bretton Woods system can be attributed to the generalized prevalence of controls on capital movements. Financial liberalization radically altered this situation, and explains the growth in the number of countries that have adopted a system of free-ﬂoating rates of exchange. See Krugman (1992), Eichengreen et al. (1994), and Agenor and Flood (1994). Argentina’s exchange rate regime had to resort to many of the old Central Bank practices to resist the onslaught of the 1995 crisis: it reinstalled a system of insurance for deposits, the Central Bank once again assumed the functions of lender of last resort – inappropriate for a currency board – and banks’ legal

18

23 24

25 26 27

28

29 30

31 32 33 34

David Ibarra and Juan Carlos Moreno-Brid reserve requirements were reduced as well as the punitive interest rates to noncomplying banks. The government and the Central Bank assumed 7.3 billions dollars in foreign debt to rebuild their reserves and to support the banking system. Even so, confronted by the increase in unpaid loans, approximately 25 percent of banking institutions were forced to close. See Blanchard and Muet (1993). Gary Becker and James Tobin, both winners of the Nobel Prize in economics, have diametrically opposed opinions concerning a possible Mexico–United States monetary union, and the impact it could have on the Mexican economy. Becker considers, without a doubt, that its impact would be favorable, while Tobin states that: ‘To have a ﬁxed exchange rate, especially in small countries, is one of the biggest and most ridiculous mistakes . . . ’ (Milenio, 8 February, 1999, Mexico, D.F., p. 44). Of course, the actions designed to pave the way for the formation of the European zone of integration commenced almost 50 years ago. Sufﬁce to recall the creation, in 1950, of the European Payments Union. See Begg (1997), Obstfeld (1997), and Eichengreen and Frieden (1998). Even so, critics of the European process of integration point out that consensual criteria have left countries defenseless in their ﬁght against recession for a good part of the 1990s, and that national ﬁscal policies are incapable of compensating the economic ﬂuctuations and shocks that affect countries asymmetrically. They also add that the design of a common monetary policy, centered exclusively upon achieving stability, can translate into permanent unemployment and into a worsening of the so-called democratic deﬁcit (Martin and Ross 1999). Although the common currency has only recently been introduced, new demands for policy coordination loom on the horizon. One deals with the divergent cyclical position of national economies. Here, a reduction in interest rates can beneﬁt some economies, but may overheat others. From this can be inferred the inherent difﬁculty for national ﬁscal policies to act countercyclically and, at the same time, satisfy the other convergence criteria. In the same manner, it may seem urgent to achieve sufﬁcient ﬂexibility in the labor market to ensure adequate creation of jobs, but there is resistance to initiatives that weaken the national welfare states. Additionally, the adoption of the dollar as the sole currency would imply a ﬁscal loss for Mexico as it would surrender its seigniorage revenue. The liabilities assumed by the government under the Bank Savings Protection Fund (Fobaproa) in its rescue of the banking system will increase the domestic public debt from 24 percent to 50 percent of GDP. This ﬁgure does not include debt originating in the non-performing loans of the development banking system or the public programs that promote low-cost housing, nor does it consider the extra budgetary ﬁnancing mechanisms of public enterprises – mainly Comisión Federal de Electricidad and PEMEX – or the accrued interest payments on FOBAPROA. If it did, the total for public debt would be much higher. A. Greenspan has recently declared that the Federal Reserve serves only the United States. See Mundell (1961) and Kenen (1969). See Monticelli and Papi (1996). The asymmetry of an oil shock is more pronounced, given the high degree of dependence of Mexico’s public sector on ﬁscal revenues from oil exports.

Currency boards and monetary unions

19

Bibliography Agenor, P. and Flood, P.R. (1994) ‘Macroeconomic policy, speculative attacks and balance of payments crises’, in Van der Ploeg, F. (ed.) The Handbook of International Macroeconomics, Oxford: Basil Blackwell. Aghevi, et al. (1991) ‘Exchange rate policy in developing countries: some analytical issues’, IMF Occasional Paper 78. Bagehot, W. (1874) Lombard Street, London: Kegan Paul. Begg, D. (1997) ‘The design of EMU’, IMF Working Paper 97/99, IMF: Washington. Blanchard, O. and Muet, P. (1993) ‘Competitiveness through disinﬂation’, Economic Policy 10: 11–56. Bloomﬁeld, A. (1959) Monetary Policy Under the International Gold Standard 1880–1914, New York: Federal Reserve Bank of New York. Clavijo, F. and Casar, J. (eds) (1994) La Industria Mexicana en el Mercado Mundial: Elementos para una Política Industrial, Mexico: Fondo de Cultura Económica. Economic Commission for Latin America and the Caribbean (1998) Balance Preliminar de la Economía de América Latina y el Caribe, Santiago de Chile. Eichengreen, B. et al. (1994) Speculative Attacks on Pegged Exchange Rates, Berkeley: University of California. Eichengreen, B. (1996) Globalizing Capital: A History of the International Monetary System, Princeton: Princeton University Press. Eichengreen, B. and Flandreu, M. (1996) ‘The geography of the Gold Standard’, in Braga, J. et al. (eds) Currency Convertibility: The Gold Standard and Beyond, London: Routledge. Eichengreen, B. and Frieden, J. (eds) (1998) Forging an Integrated Europe, Ann Arbor: University of Michigan Press. Garber, P. and Svensson, L. (1995) ‘The operation and collapse of ﬁxed exchange rate regimes’, in Grossman, G. and Rogoff, K. (eds) Handbook of International Economics, Vol. III, Elsevier Science B.V., 1865–912. Giavazzi, F. and Pagano, M. (1998) ‘The advantage of tying one’s hands: EMS discipline and central bank credibility’, European Economic Review 32, June: 1055–82. Hanke, S.H. et al. (1993) Russian Currency and Finance: A Currency Board Approach to Reform, London: Routledge. Hanke, S.H. and Schuler, K. (1993) ‘Currency boards and their relevance for Latin America’, World Bank Discussion Papers 27: 3–21. Humpage, O.F. and McIntire, J.M. (1995) ‘An introduction to currency boards’, Federal Reserve Bank of Cleveland Economic Review 2–11. Ibarra, D. (1994) ‘Es aconsejable una política industrial en México?’, Examen 66, November. Ibarra, D. (1996) ¿Transición o Crisis?, Mexico: Aguilar. IMF (1997) World Economic Outlook, Washington. Kenen, P. (1969) ‘The theory of optimum currency areas: an eclectic view’, in Mundell, R. and Swoboda, A. (eds) Monetary Problems of the International Economy, Chicago, Ill.: University of Chicago Press. Krugman, P. (1992) Currencies and Crises, Cambridge, Mass.: MIT Press. Krugman, P. (1998) ‘Saving Asia: its time to get radical’, Fortune Investor 7, September.

20

David Ibarra and Juan Carlos Moreno-Brid

Martin, A. and Ross, G. (1999) ‘The Euro and democracy’s devaluation’, Current History 98, 627: 171–5. Monticelli, C. and Papi, L. (1996) European Integration, Monetary Coordination and Demand for Money, Oxford: Clarendon Press. Moreno-Brid, J.C. (1999) ‘Mexico’s economic growth and the balance-of-payments contraint: a cointegration analysis’, International Review of Applied Economics 13(2): 149–60. Moreno-Brid, J.C. (2000) Essays on Economic Growth and the Balance-ofPayments Constraint, with Special Reference to the Case of Mexico. Ph.D. Dissertation, Faculty of Economics and Politics, University of Cambridge, UK. Mundell, R. (1961) ‘A theory of optimum currency areas’, American Economic Review 51, September: 657–64. Mundell, R. and Swoboda, A. (1988) Managing Exchange Rates, London: Royal Institute of International Affairs. Obstfeld, M. (1995) ‘International currency experience: new lessons and lessons relearned’, Brookings Papers in Economic Activity 1: 119–220. Obstfeld, M. (1997) ‘Europe’s gamble’, Brookings Papers on Economic Activity 2: 241–300. OECD (1998) OECD Economic Surveys 1997–98. Mexico, Paris. Schuler, K. (1996) ‘Should developing countries have central banks?: Currency quality and monetary systems in 155 countries’, Research Monograph 52, London: Institute of Economic Affairs. Trevor, D. and Floyd, J. (1992) Canada and the Gold Standard: Balance of Payments Adjustment 1871–1913, Cambridge: Cambridge University Press. Ul-Haq, M. et al. (eds) (1996) The Tobin Tax: Coping with Financial Volatility, Oxford: Oxford University Press. Whale, B. (1939) ‘Central banks and the state’, Manchester School 10: 38–49.

2

Is a convergence path of monetary policies possible under the North American Free Trade Agreement?1 Luis Miguel Galindo and Horacio Catalán

Mexico, the United States and Canada have been members of the North American Free Trade Agreement (NAFTA) since 1 January 1994. Since then the ﬂow of trade and capital movements between these countries has increased in spite of the Mexican ﬁnancial instability. This situation has raised the issue of the pros and cons of further ﬁnancial integration which might include a monetary union or the use of the dollar as the main currency for the three NAFTA members. It is, however, still far from clear as to what are the problems and the necessary pre-conditions that need to be addressed and solved before a major step towards monetary policy coordination or even a monetary union can be taken. Under these circumstances the European economic integration is a necessary reference. But one should not forget that the initial conditions of each country participating in the European integration were very different from the conditions among the members of NAFTA. In fact, the divergence in the initial conditions among the NAFTA members are arguably among the largest in the world. The economic diversity among NAFTA countries suggests that a convergence process which leads to a monetary union is only possible through the articulation of an explicit strategy in this direction. This plan should be based on similar trends in economic growth and a credible coordination of monetary and ﬁscal policies that explicitly acknowledge the important differences between the member countries. This would incorporate the fact that Mexico represents an emerging ﬁnancial market which is subject to problems of market volatility, speculative attacks and currency substitution. The main objective of this essay is to analyse some of the possibilities and problems of monetary policy integration between Mexico, the United States and Canada in the context of a convergence or divergence process of these economies. The essay is divided into three sections. The ﬁrst provides a general theoretical framework. The P* model included here was selected due to its current use for monetary policy purposes by the Federal Reserve in the United States. The second section includes the empirical

22

Luis Miguel Galindo and Horacio Catalán

evidence and, ﬁnally, the third section presents the main conclusions and some comments about monetary policy.

2.1 Theoretical framework Monetary policies among the NAFTA members represent one of the main factors needed in order to achieve a convergence process of ﬁnancial variables and, in the long run, of real variables. However, this convergence process crucially depends on the future evolution of the real economies and therefore monetary policies should, under these circumstances, only improve the convergence paths but not reverse the direction of the process. In this sense, monetary policies should have an important role in a long-term perspective, but before this can happen they must initially fulﬁl certain common patterns such as a convergence of interest rates or a monetary aggregates evolution. The endogenous growth models, developed from the classical works of Romer (1986) and Lucas (1988), claim the existence of different possible forms of convergence (Galor 1996). The ﬁrst group of studies suggests a convergence process where poor countries grow faster than rich countries. This process is known as absolute convergence of and is normally associated with the relative convergence of which occurs as a consequence of a reduction on the volatility of the rates of growth among different countries (Barro and Sala-i-Martin 1995, Durlauf 1996). This set of models assumes similar technologies and preferences among countries. Baumol (1986), Barro (1991) and Mankiw, Romer and Weil (1992) have found empirical evidence in favour of this hypothesis after adjusting for saving rates and population growth and recognizing the existence of random shocks which introduce oscillations through the convergence path. However, with persistence of different per capita income levels among countries over long periods of time, the concept of conditional convergence is raised. Recent estimations, which include a more representative sample of countries, reject the existence of absolute convergence (Sala-iMartin 1996). There is also evidence of important differences in technology among nations, which is considered to be one of the leading factors for economic growth (Bernard and Jones 1996). These results suggest that the convergence process exists only when the technology and behaviour parameters are equal, and thus steady states are also similar. The empirical evidence, using a sample of 110 countries, suggests the existence of a convergence process which is adjusted for technologies, political factors, savings and population trends (Barro and Sala-i-Martin 1995, Sala-iMartin 1996). The convergence rate is estimated at about 2 per cent a year by Sala-i-Martin (1996). Recent empirical research (Quah 1996) suggests that per capita incomes in the world went from a normal distribution to two normal distributions with different means or ‘twin peaks’. This is also known as the existence of

Is a convergence path of monetary policies possible?

23

a convergence club of countries. In this sense, the poor countries tend to converge to a low per capita income mean while the rich countries converge towards a high per capita income mean (Galor 1996, Quah 1996). Still, some countries manage to go from one club to another. The convergence process basically depends on factors which explain why a country belongs to a certain club or coalition. This situation implies the existence of multiple equilibria or steady state solutions (Galor and Ryder 1989). In the extreme case, it could be argued a country’s membership in a given coalition leads to altered economic policy which implies a realignment of most of its macroeconomic and ﬁnancial variables. In this sense, the convergence process is a result of a previous decision to belong to one club of countries and not the other (Quah 1996). This is consistent with the idea that international capital mobility reduces the relevance of the initial conditions. It is also important to include variables such as income distribution, fertility and human capital, in order to establish the steady state and, consequently, the club or coalition of each country (Galor 1996). It is worth noticing that the monetary convergence process in the European case went through rather different phases. The ﬁrst phase (1987–92) was based on a managed exchange rate mechanism characterized by several exchange rate adjustments which converged to a more stable system. The second phase involved a ﬁxed exchange rate regime though with some adjustments (Farina 1998). A third current phase includes the well known harmonization conditions where each country’s public deﬁcit should not exceed 3 per cent of GDP and the gross public debt should be no more than 60 per cent of GDP, or must show rapid decline to this target level. Furthermore, there is also a ﬁxed exchange rate system with inﬂation and interest rates targets (Masson 1996). It is also expected that the convergence process will be accompanied by more ﬁscal coordination in order to prevent investors from taking advantage of tax rates disparities. Simultaneously, the coordination of monetary policy included several institutional steps such as the creation of the European Monetary Institute and, later, the coordination of monetary policies in the European Central Bank where a ﬁxed nominal exchange rates system became a basic tool with the German mark as its anchor (Artis 1996). NAFTA does not include a coordination programme or a convergence path of monetary policies among its members. The economic, social and political differences among Mexico, Canada and the United States are, no doubt, important barriers for a possible monetary and ﬁnancial integration. Thus an analysis of convergence among NAFTA members must ﬁrst consider the conditions of similar economic patterns or trends in these countries. Most economic variables such as output, prices, monetary aggregates, exchange rates and interest rates are normally non-stationary series with unit roots. Stochastic convergence among variables with unit roots

24

Luis Miguel Galindo and Horacio Catalán

requires that the difference between them is stationary. In this sense, stationary residuals in a regression between these variables implies cointegration and therefore the presence of a stable long-term relationship and the existence of common trends that can be interpreted as a slow convergence process (Daniel 1997, Parikh, Bailey and Lovatt 1997). The presence of more co-integrating vectors between the variables is also considered as evidence of a stronger convergence process, because it implies the existence of more common trends. The common trends in the gross domestic product (GDP) of the NAFTA members can be analysed by using the Johansen (1988) procedure which is based on a vector auto-regressive model (VAR) deﬁned as: y1t 01 i y1ti1 i y2ti i w3ti u1t y2t 02 i y1ti i y2ti1 i w3ti u2t wt 03 i y1ti i y2ti i w3ti1 u3t

(1)

Lower case letters represent the logarithm of the series and the lag length depends on the statistical properties of the series. A more general representation of equation (1) in a matrix form (Cuthbertson, Hall and Taylor 1992) is: Xt 1Xt1 . . . kXtk Dt ut

(2)

where Xt is a vector that includes all the endogenous variables and Dt is a vector including the constant, trend and dummy variables. Reparameterizing the VAR of equation (2) in its error correction form (Johansen 1988) yields: Xt 1Xt1 . . . k1Xtk1 Xtk Dt ut

(3)

where Xtk is the long-term solution including all the error correction mechanisms and therefore the common trends. In this case the exogenous shocks are not necessarily symmetric because they have different weights in each economy. The importance of these effects in each country are given by the coefﬁcient of the Johansen (1988) procedure. Recent empirical evidence conﬁrms that most economic time series are non-stationary, which indicates the existence of exogenous shocks that include permanent and transitory effects. For example, Blanchard and Quah (1989), King, Plosser, Stock and Watson (1991), Gali (1992) and Daniel (1997) argue that most of the variance on the US GDP is due to permanent shocks associated with productivity or oil price changes. There is also evidence of a convergence process in the GDP of industrialized countries adjusted for the effects of different kinds of shocks. This is the case, for example, of the GDP of different regions of the United States, Japan and Western Europe. Henceforth these nations show evidence of an

Is a convergence path of monetary policies possible?

25

absolute convergence process on and a relative convergence of (Barro and Sala-i-Martin 1995). Also, Bértola and Porcile (1998) ﬁnd that the GDP of Argentina, Brazil and Uruguay are co-integrated. This evidence is used by the authors to argue in favour of a convergence process guided by technical diffusion. Daniel (1997) and Parikh, Bailey and Lovatt (1997) ﬁnd the presence of common trends, using the Johansen procedure, in the European Union, the United States, the United Kingdom and Japan, which is also interpreted as a convergence process. MacDonald and Taylor (1991) and Parikh, Bailey and Lovatt (1997) ﬁnd evidence in favour of a convergence process in exchange rates and monetary aggregates among the members of the European monetary union but not for non-members. The existence of a convergence process in GDP or ﬁnancial variables inside a common trade market is a necessary condition for common monetary policy. In effect, the existence of common trends in the GDPs suggests that the monetary policies must concentrate on the adjustment of the paths but not on reversing the direction of economic growth in order to match it with other countries.

2.2 The P* model The P* model is used by the Central Banks of the United States, Germany and France, among others, to forecast the price level that is consistent with the expansion of monetary aggregates. This model assumes that one monetary aggregate has a stable long-term relationship with potential output and this information can be used as an indicator of the long-term trend of prices. This model is derived directly from the quantity of money identity: MtVt PtYt

(4)

where Mt represents the monetary aggregate, Vt the velocity of circulation, Pt the price index and Yt real income. The equilibrium price level that is consistent with the actual value of the monetary aggregate, potential output and a constant or stable velocity of circulation is deﬁned, from equation (4) as: P*t (MtV *t)/Y *t

(5)

Equation (5) can be represented in logarithms as: P*t 1y*t 2mt 3v*t

(6)

And it is expected that 1 1, 2 = 1 y 3 = 1 (Hall and Milne 1994). The P* model identiﬁes the price level that is consistent with the monetary aggregate using the long-term relationship between the monetary aggregate and the potential output as a long-term indicator of the trend in

26

Luis Miguel Galindo and Horacio Catalán

prices (Hallman, Porter and Small 1991). Equation (6) offers a simple rule to forecast the future behaviour of prices and allows consistency between the inﬂationary target of the Central Bank and the expansion path of the monetary aggregate. Equation (6) is used to obtain the P* that can be compared with the observed price behaviour. The gap between P and P* indicates the existence of inﬂationary or deﬂationary pressures in the economy. The estimation of the P* models for Mexico, Canada and the United States provides a way to compare the effect of monetary expansion on prices in each case and can be used as a base line to investigate the effects of possible monetary policy coordination. The P* models can allow a comparison between two possible options of monetary policies that are the two extremes for open economies: 1) the central bank controls the monetary supply and lets the exchange rate move freely, or 2) it ﬁxes the exchange rate and thus the monetary aggregate becomes an endogenous variable. While this model is only an initial point for the comparison of the effects of monetary policies, it should be taken as an explanation of the price behaviour in these countries.

2.3 Empirical evidence 2.3.1 Mean and variance analysis The data set includes quarterly series from 1983(1) to 1998(4). The sample selection was made due to difﬁculties encountered in the construction of compatible and comparable time series for the three countries. The series are real income (yt), the consumer price index (pt), the monetary aggregate in nominal (m2t) and real terms (m2p)t , the interest rate for the respective government bonds in nominal (rt) and real terms (rrt), and the exchange rate also in nominal (st) and in real terms (srt). ‘MX’ represents Mexico, ‘US’ United States and ‘CN’ Canada. A graphical analysis of the levels and rates of growth of GDP of Mexico, the United States and Canada indicates that there is a possible common trend in their paths despite the high seasonal dependence in the Mexican case. The rates of growth of GDPs (Figure 2.1) conﬁrms that the Mexican series is rather volatile, while the series of the United States and Canada tend to move together. Hence, it is possible to argue in favour of an important convergence process between Canada and the United States either in levels or in rates of growth; meanwhile, the Mexican economy is converging to this trend with a highly variable behaviour. A basic indicator of a convergence or divergence pattern is given by the mean and the variance of the series before and after the beginning of the Trade Agreement in January 1994. This artiﬁcial division of the series tries to capture possible differences in behaviour due to the impact of NAFTA. The average rate of growth and variance of GDP gives weak evidence in favour of NAFTA considering that the average growth rate of the three

Is a convergence path of monetary policies possible?

27

Figure 2.1 Rates of real growth of the GDPs of Mexico, Canada and the United States.

countries increased after 1994. There is also weak evidence of a convergence process considering that the Mexican economy grew faster than the other two, and that the differences between these rates tends to decrease (Table A.2.1 in the Appendix). Nevertheless, the variance in the Mexican case is still high and should be smaller in order to consider the existence of a strong convergence process. The differences in the inﬂation rates are still large, mainly because of Mexico’s high increase in prices and their variability (Table A.2.2). The reduction of the Mexican inﬂation rate from the ﬁrst period (1986–94) to the second period (1994–8) associated with a similar reduction in the volatility of the series, appears to indicate a convergence pattern. However, the differences in inﬂation rates are still rather large between Mexico and the other two countries. Furthermore, the mean difference between the United States and Canada increases during the second period (Table A.2.2). These results suggest that, in order to argue in favour of convergence in nominal terms and to discuss the option of monetary policy coordination amongst NAFTA members based on similar inﬂation rate targets, it is still necessary to substantially reduce the inﬂation rate and its volatility in Mexico. The high inﬂation level has negative effects on the behaviour of the exchange rate, nominal interest rates and the monetary aggregates, generate periods of violent instability for these variables (Tables A.2.3, A.2.5 and A.2.7). This is not surprising considering the fact that the path of nominal interest rates is associated (co-integrated) with the inﬂation rate (for the US, see Mishkin 1992) and the behaviour of the nominal

28

Luis Miguel Galindo and Horacio Catalán

monetary aggregate in Mexico is basically dominated by the price index due to the high inﬂation rate. Mexico’s nominal exchange rate also shows very volatile behaviour linked to the abrupt changes of purchasing power parity and the uncovered interest rate parity condition (Galindo 1995). The differences in real monetary aggregate and real interest rates are smaller in the second period but there are still very important differences, particularly for the latter variable (Tables A.2.4 and A.2.6). The real interest rate, despite a convergent trend, is still very high in Mexico in comparison with the other two countries. This difference is relevant because it is most probably associated with the risk of holding pesos, or country risk. In this situation it is important to recognize that the only way to decrease the real interest rate differential is by reducing of the risk of devaluation of the Mexican peso which is most probably associated with the inﬂation differential rates among other things. Finally, during the second period the real exchange rate (Table A.2.8) also shows some convergence in Mexico and Canada with respect to the United States, which can be used in the future as an anchor to reduce the risk of holding pesos considering the uncovered interest rate parity. It is worthwhile to mention that the empirical evidence from Europe indicates that the exchange rate regime is fundamental to determining the volatility of real exchange rates (Stockman 1983, Mussa 1986). The main results of this section give weak evidence in favour of a convergence process, especially in real terms. However, the high and variable inﬂation rate in Mexico imposes important restrictions on monetary policy coordination because it increases the volatility of other variables such as the nominal exchange, and the nominal and real interest rates and hence raise the risk premium of holding pesos. This situation affects the real interest rate differentials, with negative effects on economic growth. 2.3.2 Co-integration analysis The unit root tests are summarized in Table 2.1. The Dickey–Fuller (1981) and Phillips–Perron (1988) tests indicate that all the series are nonstationary but that there are important differences in the order of integration of the variables in each country. The price index is I(2) in Mexico and probably also in the United States but it is I(1) in Canada. The nominal monetary aggregate is I(2) in all cases, the nominal interest rate is I(1). The real interest rates, real exchange rates and real income are I(1) in all cases. However, the evidence is not conclusive for the United States and Canada in real income because the series might be I(2). These results show the difﬁculties in ﬁnding a convergence pattern among these economies for, despite the tendency for all series to grow, its intensity can be different. Only when all the series are characterized by non-stationary behaviour, will it be possible to study the existence of stochastic convergence through the presence of common trends.

1.45[0.15] 0.50[0.61] 0.56[0.57] 1.83[0.07]

3.50[0.00]** 3.09[0.00]** 2.87[0.00]**

2.16[0.03]* 1.43[0.16] 3.00[0.00]** 2.19[0.03]*

1.46[0.15] 0.80[0.43] 0.38[0.70]

0.10[0.92] 0.13[0.88] 0.75[0.45] 1.65[0.11]

RRMX RRUS RRCN

sMX sCN srMX srCN

0.71[0.48] 0.83[0.41] 0.63[0.53]

0.94[0.35] 8.62[0.00] 5.60[0.00]

ADF(4) Augmented Dickey–Fuller with four lags. PP(4) Phillips–Perron test with four lags. ** rejection at 1 per cent of significance; *rejection at 5 per cent of significance. Period: 1986(1)–98(4).

4.40[0.00]**

1.54[0.13] 0.95[0.35] 1.23[0.22]

3.02[0.00]** 2.63[0.01]* 3.33[0.00]**

0.76[0.45] 0.61[0.54] 0.72[0.48]

rMX rUS rCN

4.28[0.00]** 4.20[0.00]**

2.29[0.03]* 0.81[0.42] 1.34[0.18]

1.90[0.06] 2.09[0.04] 1.95[0.06]

6.97[0.00] 8.52[0.00] 7.4[0.00]

yMX yUS yCN

3.32[0.00]** 5.03[0.00]** 2.97[0.00]**

1.37[0.18] 0.57[0.56] 1.76[0.08]

5.36[0.00] 11.6[0.00] 0.61[0.54]

0.75[0.45] 1.28[0.21] 0.93[0.36]

3.35[0.00]** 4.72[0.00]**

1.63[0.11] 0.53[0.59] 2.96[0.00]**

Levels

m2MX m2US m2CN

2ª difference ∆∆

1ª difference ∆

PP(4)

0.89[0.40] 1.66[0.10] 0.62[0.54]

Levels

ADF(4)

pMX pUS pCN

Variable

Table 2.1 Unit root tests

3.72[0.00]** 6.67[0.00]** 6.37[0.00]** 6.14[0.00]**

8.62[0.00]** 10.4[0.00]** 10.9[0.00]**

7.73[0.00]** 5.09[0.00]** 7.73[0.00]**

17.8[0.00]** 2.24[0.02]* 3.92[0.00]**

3.76[0.00]** 2.61[0.01]* 1.63[0.11]

1.70[0.09] 2.73[0.00]** 7.16[0.00]**

1ª difference ∆

12.8[0.00]**

10.8[0.00]** 12.4[0.00]**

17.4[0.00]** 11.9[0.00]** 8.2[0.0]**

6.45[0.00]** 14.8[0.00]**

2ª difference ∆∆

30

Luis Miguel Galindo and Horacio Catalán

Table 2.2 Statistics of the Johansen procedure for yMX, yEU, yCN 1986(1)–93(4) Ho:ran kp

T log(1l)

Tnm

95%

Tra(1lp1)

Tnm

95%

p 0 p1 p2

28.5** 22.13** 8.84

17.81 13.83 5.52

22.0 15.7 9.2

59.47** 30.97** 8.84

37.17* 19.36 5.52

34.9 20.0 9.2

T log(1) maximum eigenvalue test; Tra(1p+1) trace test; * ** ( ) rejection of the null hypothesis at 5 per cent (1 per cent); p number of co-integrating vectors; the VAR includes 4 lags.

Table 2.3 Statistics of the Johansen procedure for yMX, yEU, yCN 1986(1)–98(4) Ho:ran kp

T log(1l)

Tnm

95%

Tra(1lp1)

Tnm

95%

p 0 p1 p2

45.63** 22.39** 14.41**

18.25 8.956 5.765

18.25 8.956 5.765

82.43** 36.80** 14.41

32.97 14.72 5.765

34.9 20.0 9.2

T log(1) maximum eigenvalue test; Tra(1p+1) trace test; * ** ( ) rejection of the null hypothesis at 5 per cent (1 per cent); p number of co-integrating vectors; the VAR includes 4 lags.

The Johansen (1988) procedure (Tables 2.2 and 2.3) indicates the existence of common trends in real income between Mexico, the United States and Canada. Furthermore, the number of common trends increases after the NAFTA agreement. This may be interpreted as evidence in favour of a weak convergence process between these economies. Arguably the nonstationary relationship among the series represents the productivity shocks (Daniel 1997) that have different impacts on each country, or the existence of asymmetric responses to external shocks. This situation limits the possibility of an optimal currency area under NAFTA. Nevertheless the disparities in productivity might be reduced in the near future considering that the world has changed, from one technological leader during the 1960s, to a situation of equal technological rivals (Dollar and Wolff 1993) which allows more competition and increases the pressures to improve technological methods. It is worth noticing that the convergence process in productivity is accompanied by different specialization patterns among nations. In this sense the convergence in real terms among NAFTA members will include the development of complementary production based on national comparative advantages.

2.4 The estimated P* model The estimation of the P* model indicates the existence of a long-term relationship between the price index, real income, monetary aggregate and the interest rate for the three members of NAFTA (Tables 2.4, 2.5 and 2.6). These co-integration relationships can be interpreted as price

Is a convergence path of monetary policies possible?

31

Table 2.4 Statistics of the Johansen procedure for P* for Mexico 1986(1)–98(4): pt, m2t , yt , rt Ho:ran kp

T log(1l)

Tnm

95%

Tra(1lp1)

Tnm

95%

p0 p1 p2 p3

32.04* 18.92* 8.739 3.772

22.18 13.1 6.05 2.612

22.18 13.1 6.05 2.612

63.48** 31.43** 12.51 3.772

43.94* 21.76 8.661 2.612

39.9 24.3 12.5 3.8

T log(1) maximum eigenvalue test; Tra(1p+1) trace test; * ** ( ) rejection of the null hypothesis at 5 per cent (1 per cent); p number of co-integrating vectors; the VAR includes 4 lags.

Table 2.5 Statistics of the Johansen procedure for P* for the United States 1986(1)–98(4): pt, m2t , yt , rt Ho:ran kp

T log(1l)

Tnm

95%

Tra(1lp1)

Tnm

95%

p0 p1 p2 p3

32.79** 17.22* 3.249 0.2044

22.7 11.92 2.25 0.1415

23.8 17.9 11.4 3.8

53.46** 20.67** 3.454 0.2044

37.01* 14.31 2.391 0.1415

39.9 24.3 12.5 3.8

T log(1) maximum eigenvalue test; Tra(1p+1) trace test; * ** ( ) rejection of the null hypothesis at 5 per cent (1 per cent); p number of co-integrating vectors; the VAR includes 4 lags.

Table 2.6 Statistics of the Johansen procedure for P* for Canada 1986(1)–98(4): pt, m2t , yt , rt Ho:ran kp

T log(1l)

Tnm

95%

Tra(1lp1)

Tnm

95%

p0 p1 p2 p3

23.05 20.02* 7.717 4.559*

13.03 11.31 4.362 2.577

23.8 17.9 11.4 3.8

55.34** 32.29** 12.28 4.559*

31.28* 18.25 6.938 2.577

39.9 24.3 12.5 3.8

T log(1) maximum eigenvalue test; Tra(1p+1) trace test; * ** ( ) rejection of the null hypothesis at 5 per cent (1 per cent); p number of co-integrating vectors; the VAR includes 4 lags.

equations and, therefore, allow to obtain the equilibrium price which is consistent with a certain level of monetary aggregate for each nation. These results are consistent with previous work of Galindo (1997), Catalán (1998) and Liquitaya (1998) for Mexico, and Hallman, Porter and Small (1991) for the United States. The impact of the monetary aggregate on the price level for each country can be evaluated through the P* model. The long-run solutions for the P* model for Mexico, the United States and Canada are presented in equations (7), (8) and (9) respectively. These results stress the important differences evident between the countries. The P* model for Mexico has

32

Luis Miguel Galindo and Horacio Catalán

smaller coefﬁcients for the monetary aggregate and the real income in comparison with the United States and Canada. The main consequence is that a similar increase in the monetary aggregate in the three countries will have asymmetric effects on prices. A rise of 10 per cent in the monetary aggregate implies a growth in prices in Mexico of around 4 per cent, while generating increases of 16 per cent and 18 per cent in the United States and Canada respectively. In this sense, a monetary expansion in Mexico has a smaller impact on the price than in the United States and Canada. Arguably, this is a consequence of a long history of inﬂation and problems of market inefﬁciency and price stickiness in the Mexican economy. On the other hand, an output increase in the United States or Canada has a much bigger negative impact on prices. Therefore, a similar growth in the three countries is consistent with different monetary expansion and price level targets. This situation indicates that the coordination of monetary policy among NAFTA members is difﬁcult because there are important differences in the impact of monetary aggregates on prices even in the case of real convergence. Furthermore, the Central Bank authorities must make a choice, before any monetary union is created, between a policy that aims to control the money supply, leaving the exchange rate free or to ﬁx the exchange rate while the monetary aggregate becomes endogenous (Alogoskouﬁs 1994). The ﬁrst option, of controlling domestic money supply, allows the inﬂation rate to be determined internally and leads to a endogenous exchange rate even under an income convergence process. That is, targeting similar monetary supply growth rates across countries would imply different national inﬂation rates which will be reﬂected, eventually, in corresponding adjustments in the exchange rates. Thus similar monetary targets will lead to different inﬂation rates and on to the necessity of a ﬂexible exchange rate system. On the other hand, coefﬁcient value for the United States and Canada suggests that those countries can easily coordinate their monetary policies even on the basis of monetary targets. The alternative option is to impose a ﬁxed exchange rate system by targeting similar inﬂation rates which implies different growth rates of the monetary aggregates. Under this regime the domestic inﬂation tends to equal the international inﬂation rate (Mundell 1963). Assuming the coefﬁcients of the P* models, the policy would allow a Mexican monetary expansion that is larger than in the other two countries. This situation requires, under a ﬁxed exchange rate system, some sort of subsidy from the United States and Canada to Mexico. This is necessary since, while the relative price of the currencies remains constant, Mexico would expand its monetary aggregates faster than the other two countries. Mexico: pt 0.8167m2t 0.4773yt 0.2280rt

(7)

Is a convergence path of monetary policies possible?

33

United States: pt 1.604m2t 0.956yt 0.409rt

(8)

Canada: pt 1.856m2t 1.21yt 0.642rt

(9)

The selection of the appropriate monetary policy under NAFTA is, therefore, anything but an easy task considering the substantial differences between Mexico and Canada and the United States. The main results of this analysis suggest that monetary integration of Canada and the United States is an almost natural phenomenon, but the addition of Mexico creates additional challenges. A commitment to a ﬁxed exchange rate regime by NAFTA would certainly provide advantages in credibility, in particular for Mexican monetary policy. It will also allow a larger expansion of the Mexican monetary aggregates which might be consistent with a target of high economic growth. However, these results also indicate that the coordination of monetary policies based on a ﬁxed exchange rate regime, is still premature and even rather risky until the moment at which some critical points are properly addressed. Among the problems that a further coordination of monetary policies would create are: 1

Reducing on the variance of Mexican series, in particular the inﬂation rate, is an obvious requirement before further monetary integration can be undertaken. Exchange rate stability provides an anchor to price stability, but is by no means a sufﬁcient condition. Eichengreen (1994) has found, in effect, that the history of this century does not give conclusive evidence to support a direct link between a ﬁxed exchange rate system and a reduction of the volatility of the series. For example, the volatility of GDP was 50 per cent larger during the gold standard than under Bretton Woods, and also in the European Monetary System convergence to the German inﬂation rate during the ﬁxed exchange rate regime (1987–92) was slower than in the previous phase (1979–86) (Farina 1998). This evidence suggests that a ﬁxed exchange rate system prevails under a stable environment because of a reverse causation from general stability and domestic policies to the relative price of currencies.

2

The large economic differences among the NAFTA members also suggest a possible incompatibility of the exchange rate regime with the targets of domestic policy in each country, in particular in the Mexican case. The evidence suggests (Eichengreen 1994) that a ﬁxed exchange rate tends to link output ﬂuctuations which, in turn, might narrow the

34

Luis Miguel Galindo and Horacio Catalán possibilities for fast Mexican economic growth. Additionally, the existence of asymmetric shocks between these countries suggests that exchange rate management might be an optimal regime in order to compensate for them (Alogoskouﬁs 1994). All this makes the conﬁguration of an optimal currency area rather difﬁcult.

3

Credibility and speculative attacks. The announcement of a monetary coordination policy among the NAFTA members would enhance the credibility of a ﬁxed exchange rate regime and would thereby give reputation gains and reduce the transition costs to a low inﬂation rate in Mexico. However, the actual differences in inﬂation rates might be a source of speculative attacks against the Mexican peso. The economic literature suggests that a large expansion of the ﬁnancial system under weak economic foundations, large current account deﬁcits, and a real exchange rate appreciation are all factors which increase the possibilities of speculative attacks. Krugman (1979) argues that excess domestic credit leads to a sudden speculative attack which can ﬁnally destroy a ﬁxed exchange rate regime. Sachs, Tornell and Velasco (1996) also claim that an expansion of the ﬁnancial system on a weak foundation is a very important factor to explain speculative attacks. Agenor and Flood (1994) point out that large deﬁcits and real exchange rate appreciation are all empirical regularities before a currency crash in the cases of Argentina, Chile and Mexico. In sum, the empirical evidence basically suggests that speculative attacks are more likely to occur in a ﬁxed exchange rate economies that have high inﬂation rate and weak foundations as expressed in a current account or ﬁscal deﬁcit, accelerating credit expansion and limited ﬁnancial regulation. These features are all present in the Mexican case. A ﬁxed exchange rate agreement under the NAFTA framework would thus trigger the potential risk of speculative attacks. In order to reduce this risk, it is necessary to ﬁrst achieve convergence paths on inﬂation rates. Otherwise, the option will be to raise Mexican real interest rates to compensate the increasing risk of holding pesos. This situation will have negative effects on production, leading eventually to a real adjustment process in the Mexican economy.

4

Currency substitution. There is an increasing literature on the empirical relevance of the currency substitution2 effect. According to Gresham’s Law, bad money substitutes for good money as the means of payment (Giovannini and Turtelboom 1994). Individuals prefer to keep the good money and spend the bad money. The US dollar plays the role of the secure asset in NAFTA and thus this currency would eventually become the basic reference for important transactions and savings.

Is a convergence path of monetary policies possible?

35

The empirical evidence supports the existence of currency substitution in Mexico in particular during periods of exchange rate crises and high inﬂation rates (Ortiz 1983, Ramirez-Rojas 1985, and Rogers 1992). This situation would lead to a reduction of the role of the Mexican peso and would increase pressures for the creation of a monetary union. However, this union would only maintain the Mexican peso as a credible and useful currency provided there exists monetary and ﬁnancial stability in Mexico. In this sense, a monetary union created on the basis of full coordination of monetary policies with ﬁxed exchange rates seems risky and problematic. However, there is a lot of room for improvement regarding the coordination of monetary policies among NAFTA members that in the future may lead to a monetary union. This coordination should consist of several steps in order to address the large differences between Mexico on one side, and Canada and the United States on the other. The current situation allows a monetary union between the latter two, but the inclusion of Mexico with its high and variable inﬂation rate presents an important obstacle. A ﬁxed exchange rate regime between the three members would be subject to important speculative attacks despite the large credibility gains. In order to avoid such a situation it is indispensable, at a ﬁrst step, to signiﬁcantly reduce the Mexican inﬂation rate and the risk of holding pesos. This can be achieved by using some measures of monetary coordination such as a common defence of the exchange rate, explicit commitments to monetary targets and similar patterns of nominal interest rates in order to try to reduce the high real interest rate differentials. A second stage must impose exchange rate bands and convergence paths on monetary policies. The success of this exchange rate regime would depend on strong fundamentals in the Mexican economy such as small current account and ﬁscal deﬁcits, and a sound ﬁnancial system with strict bank supervision. It is only after these ﬁrst two stages that it is possible to consider a monetary union based on a ﬁxed exchange rate regime, applying convergence conditions on the ﬁscal and current account deﬁcits and acceptance by the United States and Canada to support the Mexican monetary expansion. The implementation of these stages requires a ﬁrm decision and strong political and economic commitment from the NAFTA members that would lead to achievement of an inclusive and useful monetary union for all three nations. Otherwise the currency substitution syndrome and speculative recurrent attacks will dominate the Mexican monetary policy in the coming years.

2.5 Conclusions and economic policy issues The main results of this essay indicate that there is weak evidence in favour of a convergence process associated with the beginning of NAFTA. This evidence is provided by the behaviour of the mean and variance of

36

Luis Miguel Galindo and Horacio Catalán

GDP, inﬂation rate, real and nominal values of the monetary aggregate, interest rate and the exchange rate. However, the Mexican inﬂation rate is still very high which imposes certain trajectories on the future path of other variables, such as real and nominal exchange rates, and this also increases the risk of devaluation which has to be countered by high real interest rates. If more coordination of the monetary policies among NAFTA members is seriously considered, then as a ﬁrst step the Mexican inﬂation rate has to be substantially reduced. The P* models indicate that there is a long-term relationship between the price index, the monetary aggregate, real income and interest rate in the three member countries. However, the impact of the monetary aggregates and income are very similar in the United States and Canada and differ sharply from the Mexican case. These discrepancies can be partially explained by the history of inﬂation in Mexico, but still represent an obstacle to coordinate monetary policies. That is, a similar inﬂation target implies very different behaviour in the monetary aggregate, while targeting a similar growth of monetary aggregate must imply quite different inﬂation paths in these countries. Under these conditions a ﬁxed exchange rate system would be subject to very strong pressures that will eventually lead to large realignments in the relative values of the currencies. It seems that the most plausible strategy is to accept a stronger Mexican monetary expansion, targeting similar inﬂation rates together with a full commitment to this policy from the three member countries. However, under ﬁxed exchange rates this would imply a subsidy from Canada and the United States to Mexico. Monetary policy coordination among the NAFTA members, given the differences between their economies, requires an explicit policy that includes several steps, probably with more restrictions and limited possible exits than in Europe. In this process the behaviour of the exchange rate system is crucial, as it has also been in Europe. The commitment to the monetary union must be total in order to prevent speculative attacks. The increasing dollarization of the Mexican economy is a natural trend which needs to be properly addressed with an explicit policy. Otherwise Mexico might ﬁnd it is extensively using the dollar but without any of the advantages of a monetary union. In this sense, the two main risks in the Mexican situation arise if the members: a

b

disregard the existence of a currency substitution problem while maintaining a weak peso. Under these conditions the dollarization phenomena would rapidly increase, and thereby reduce the degrees of freedom of the Mexican Central Bank. create a monetary coordination or even a monetary union, without the necessary pre-conditions, political commitment and credibility. This carries the risk that they will be obligated to abort the project with high costs for all the participants.

Is a convergence path of monetary policies possible?

37

APPENDIX

A.2.1 Mean, variance and co-integration of the key variables Table A.2.1 Rate of growth and variance of the GDP in each quarter 1986(1) to 1993(4) Statistic

Mexico

United States

Canada

µ σ2

0.7896 29.1459

0.5363 0.3167

0.4502 0.6155

1994(1) to 1998(4) Statistic

Mexico

United States

Canada

µ σ2

0.9491 30.9287

0.8565 0.2399

0.6893 0.2070

Table A.2.2 Inﬂation rates mean and variance (∆pt) 1986(1) to 1993(4) Statistic

Mexico

United States

Canada

µ σ2

9.8572 94.6448

0.9058 0.2777

0.9373 0.3132

1994(1) to 1998(4) Statistic

Mexico

United States

Canada

µ σ2

5.4466 15.5069

0.6057 0.1170

0.2926 0.1659

Table A.2.3 Monetary aggregate mean and variance (∆m2t) 1986(1) to 1993(4) Statistic

Mexico

United States

Canada

µ σ2

0.104410 0.008685

0.010454 0.000081

0.020075 0.000138

1994(1) to 1998(4) Statistic

Mexico

United States

Canada

µ σ2

0.057546 0.001704

0.011693 0.000096

0.004310 0.000045

38

Luis Miguel Galindo and Horacio Catalán

Table A.2.4 Real monetary aggregate mean and variance ∆(m2tpt) 1986(1) to 1993(4) Statistic

Mexico

United States

Canada

µ σ2

1.7122 64.3596

0.1516 1.3490

1.0962 1.0414

1994(1) to 1998(4) Statistic

Mexico

United States

Canada

µ σ2

0.6337 24.5456

0.5742 1.4016

0.1364 0.4903

Table A.2.5 Nominal interest rate mean and variance (rt) 1986(1) to 1993(4) Statistic

Mexico

United States

Canada

µ σ2

47.0963 1220.1925

6.3816 4.3994

8.8041 7.1102

1994(1) to 1998(4) Statistic

Mexico

United States

Canada

µ σ2

28.9370 220.5274

5.4515 0.3016

5.2080 1.9439

Table A.2.6 Real interest rate mean and variance (rrt) 1986(1) to 1993(4) Statistic

Mexico

United States

Canada

µ σ2

9.805 2.40389

2.581 4.04529

5.265 0.05382

1994(1) to 1998(4) Statistic

Mexico

United States

Canada

µ σ2

6.178 1.78164

3.120 0.02059

3.889 0.03242

Is a convergence path of monetary policies possible?

39

Table A.2.7 Nominal exchange rate mean and variance (∆st) 1986(1) to 1993(4) Statistic

Mexico

Canada

µ σ2

7.4081 125.8429

0.1468 4.5061

1994(1) to 1998(4) Statistic

Mexico

Canada

µ σ2

6.6994 200.4803

0.7738 5.1457

Table A.2.8 Real exchange rate mean and variance (∆srt) 1986(1) to 1993(4) Statistic

Mexico

Canada

µ σ2

1.3438 23.8457

1.1037 44.5462

1994(1) to 1998(4) Statistic

Mexico

Canada

µ σ2

1.9397 200.7972

2.3050 20.5137

Table A.2.9 Characteristic roots of the Johansen procedure 1986(1)–98(1) (yMX, yUS, yCN) 0.58963 0.49913 0.24147 Table A.2.10 Characteristic roots of the Johansen procedure 1986(1)–94(4) (yMX, yUS, yCN) 0.897855 0.673577 0.513552

40

Luis Miguel Galindo and Horacio Catalán

A.2.2 P* models Mexico 1986(1)–98(4) Table A.2.11 Co-integration tests for the P* for Mexico 1986(1)–98(4): pt, m2t, yt, rt Ho:rankp

T log(1l)

Tnm

95%

Tra(1lp1)

p0 p1 p2 p3

32.04** 18.92** 8.739** 3.772

22.18 13.1 6.05 2.612

22.18 63.48** 13.1 31.43** 6.05 12.51** 2.612 3.772

Tnm

95%

43.94* 21.76 8.661 2.612

39.9 24.3 12.5 3.8

T log(1) maximum eigenvalue test; Tra(1p+1) trace test; * ** ( ) rejection of the null hypothesis at 5 per cent (1 per cent); p number of co-integrating vectors; the VAR includes 4 lags; Period: 1986(1)–98(4).

Table A.2.12 Characteristic roots of the Johansen procedure for Mexico (pt, m2t, yt, rt) 0.460025 0.305013 0.154688 0.069978 Table A.2.13 Standardized values of the Johansen procedure for Mexico: (pt, m2t, yt, rt) pt

m2t

yt

rt

1.000 2.2750 1.2065 29.1000

0.8167 1.0000 1.2669 27.12200

0.4773 0.05152 1.0000 14.7620

0.22803 0.005025 1.2565 1.0000

United States 1986(1)–98(4) Table A.2.14 Co-integration tests for the P* for the United States 1986(1)–98(4): pt, m2t, yt, rt Ho:ran kp

T log(1l)

Tnm

95%

Tra(1lp1)

Tnm

95%

p0 p1 p2 p3

32.79** 17.22** 3.249** 0.2044

22.7 11.92 2.25 0.1415

23.8 17.9 11.4 3.8

53.46** 20.67 3.454 0.2044

37.01 14.31 2.391 0.1415

39.9 24.3 12.5 3.8

T log(1) maximum eigenvalue test; Tra(1p+1) trace test; * ** ( ) rejection of the null hypothesis at 5 per cent (1 per cent); p number of co-integrating vectors; the VAR includes 4 lags; Period: 1986(1)–98(4).

Is a convergence path of monetary policies possible?

41

Table A.2.15 Characteristic roots of the Johansen procedure for the United States (pt, m2t, yt, rt) 0.467667 0.28191 0.060577 0.00392221 Table A.2.16 Standardized values of the Johansen procedure for the United States: (pt, m2t, yt, rt) pt

m2t

yt

rt

1.0000 0.83235 2.7890 17.294

1.6042 1.0000 2.7514 60.738

0.95654 0.43529 1.0000 47.784

0.40946 0.10712 0.05270 1.0000

Canada 1986(1)–98(4) Table A.2.17 Co-integration tests for the P*model for Canada 1986(1)–98(4): pt, m2t, yt, rt Ho:rankp

T log(1l)

Tnm

95%

Tra(1lp1)

Tnm

95%

p0 p1 p2 p3

23.05 20.02* 7.717 4.559*

13.03 11.31 4.362 2.577

23.8 17.9 11.4 3.8

55.34** 32.29** 12.28 4.559*

31.28 18.25 6.938 2.577

39.9 24.3 12.5 3.8

T log(1) maximum eigenvalue test; Tra(1p+1) trace test; * ** ( ) rejection of the null hypothesis at 5 per cent (1 per cent); p number of co-integrating vectors; the VAR includes 5 lags; Period: 1986(1)–98(4).

Table A.2.18 Characteristic values for the Johansen procedure for Canada (pt, m2t, yt, rt) 0.394093 0.352849 0.154437 0.09435 Table A.2.19 Standardized values of the Johansen procedure for Canada: (pt, m2t, yt, rt) pt 1.0000 3.6412 2.2409 66.768

m2t 1.8559 1.0000 0.67833 36.854

yt

rt

1.2090 3.1678 1.0000 14.051

0.64221 1.4568 0.11672 1.0000

42

Luis Miguel Galindo and Horacio Catalán

A.2.3 Database All the series are quarterly from 1986(1) to 1998(4). Mexico PMX Mexican consumers’ price index (base 1994 100), Bank of Mexico. M2MX Monetary aggregate (M2) in millions of Mexican pesos, Bank of Mexico. YMX Real Gross Domestic Product PIB in millions of 1993, INEGI. RMX CETES of 90 days of the average of the last month of the quarter, Bank of Mexico. SMX Exchange rate from pesos to United States dollars in order to pay external obligations. The data is taken from the last day of each quarter, Bank of Mexico. SRMX Real exchange rate using the following formula srt st(p*/p), p* is the United States consumer price. RRMX The Mexican real interest rate approximated by RRt (Rtpt). United States (US) PEU Consumers’ price index base (1982–84 100), Federal Reserve. M2EU Monetary aggregate M2, in thousands of millions of US dollars, Federal Reserve. YEU GDP in thousands of millions of US dollars at prices of 1990, FMI and OCDE (the series is already seasonally adjusted). REU Interest rate of CD of three months, we use the interest rate of the last month of each quarter, Federal Reserve. RREU real interest rate approximated by RRt (Rtpt). Canada (CN) PCN Consumers’ price index base (1990 100), FMI and Bank of Canada. M2CN Monetary aggregate M2, in thousands of millions of Canadian dollars, Bank of Canada. YCN GDP in thousands of millions of Canadian dollars at prices of 1990, FMI and OCDE. The series is seasonally adjusted. RCN Interest rate of CD of 90 days, we use the interest rate of the last month of each quarter, Bank of Canada. SCN Exchange rate of Canadian dollars to US dollars, end of period of each quarter. IMF and Bank of Canada. SRCN real exchange rate estimate using the identity of srt st(p*/p), p* is the US consumers’ price index. RRCN= Real interest rate given by RRt (Rtpt).

Is a convergence path of monetary policies possible?

43

Notes 1 We appreciate the comments of Carlos Guerrero, Fidel Aroche and Maria Elena Cardero. The usual disclaimer applies. This paper was done as part of the research project ‘The effects of the monetary policy and capital movements in the Mexican economy’ PAPIIT: IN-304197. 2 This term deﬁnes a situation where foreign currency substitutes the domestic currency in the roles of general equivalent, transfer of wealth across time and space and store of value.

Bibliography Agenor, P.R. and Flood, R.P. (1994) ‘Macroeconomic policy, speculative attacks, and balance of payments crises’, in Ploeg, F.V.D. (ed.) The Handbook of International Macroeconomics, Oxford and Cambridge: Blackwell, 224–50. Alogoskouﬁs, G. (1994) ‘On inﬂation, unemployment and the optimal Exchange Rate Regime’, in Ploeg, F.V.D. (ed.) The Handbook of International Macroeconomics, Oxford and Cambridge: Blackwell, 153–91. Artis, M. (1996) ‘Alternative transitions to EMU’, Economic Journal 106, July: 1005–15. Barro, R. (1991) ‘Economic growth in a cross section of countries’, Quarterly Journal of Economics 106: 407–44. Barro, R.J. and Sala-i-Martin, X. (1995) Economic Growth, New York: McGraw Hill. Baumol, W.J. (1986) ‘Productivity growth, convergence, and welfare: what the long-run data show’, American Economic Review 76, 5: 1072–85. Bernad, A.B. and Jones, C.I. (1996) ‘Technology and convergence’, Economic Journal 106, July: 1037–44. Bértola, L. and Porcile, P. (1998) ‘Cambio institucional, tecnología y convergencia/divergencia económica: Argentina, Brasil y Uruguay 1870–1990’, Investigación Económica 224: 137–84. Blanchard, O.J. and Quah, D. (1989) ‘The dynamic effects of aggregate demand and supply disturbances’, American Economic Review 79: 655–73. Catalán, H. (1998) ‘El Modelo P*: un Instrumento de Análisis de la Relación Agregado Monetario y Nivel de Precios’, Working Paper, Mexico: Maestría en Ciencias Económicas. Cogley, T. (1990) ‘International evidence on the size of the random walk in output’, Journal of Political Economy 98: 501–18. Cuthbertson, K., Hall, S. and Taylor, M. (1992) Applied Econometric Techniques, London: Philip Allan. Daniel, B.C. (1997) ‘International interdependence of national growth rates: a structural trend analysis’, Journal of Monetary Economics 40: 73–96. Dickey, D. and Fuller, W.A. (1981) ‘Likelihood ratio statistics for autoregressive time series with unit root’, Econometrica 49, 4: 1057–77. Dollar, D. and Wolff, E.N. (1993) Competitiveness, Convergence and International Specialization, Cambridge, Mass: MIT Press. Durlauf, S.N. (1996) ‘On the convergence and divergence growth rates’, Economic Journal 106, July: 1016–18. Eichengreen, B. (1994) ‘History of the International Monetary System: implications for research in International Macroeconomics and Finance’, in Ploeg,

44

Luis Miguel Galindo and Horacio Catalán

F.V.D. (ed.) The Handbook of International Macroeconomics, Oxford and Cambridge: Blackwell. Farina, F. (1998) ‘The Economics of the European Monetary Integration and the Convergence Problem’, Working Paper, September, Siena: University of Siena. Gali, J. (1992) ‘How well does the IS-LM model ﬁt post-war US data?’, Quarterly Journal of Economics 107: 709–38. Galindo, L.M. (1992) ‘La hipótesis de Fisher en la economía Mexicana, 1985–1990’, Monetaria XVIII, 2: 195–208. Galindo, L.M. (1993) ‘Los determinantes de corto y largo plazo del consumo en México: un análisis con mecanismo de corrección de errores y cointegración’, Investigación Económica 206: 177–208. Galindo, L.M. (1995) ‘Notas sobre el tipo de cambio en México’, Investigación Económica 212: 113–24. Galindo, L.M. (1997) ‘El modelo P* como indicador de la Política Monetaria en una economía con alta inﬂación’, El Trimestre Económico LXIV, 2: 221–39. Galindo, L.M. (1998) ‘Un Modelo de Inﬂación para México: un Enfoque Heterodoxo’, Working Paper, Facultad de Economía, UNAM. Galor, O. (1996) ‘Convergence? Inferences models from theoretical models’, Economic Journal 106, July: 1056–69. Galor, O. and Ryder, H.E. (1989) ‘Existence, uniqueness and stability of equilibrium in an overlapping-generations model with productive capital’, Journal of Economic Theory 49: 360–75. Giovanini, A. and Turtelboom, B. (1994) ‘Currency substitution’, in Ploeg, F.V.D. (ed.) The Handbook of International Macroeconomics, Oxford and Cambridge: Blackwell, 390–436. Hall, S.G. and Milne, A. (1994) ‘The relevance of the P-star analysis to UK Monetary Policy’, Economic Journal 104, May: 597–604. Hallman, J.J., Porter, R.D. and Small, D.H. (1991) ‘Is the price level tied to the M2 monetary aggregate in the long run?’, American Economic Review 81: 841–58. Johansen, S. (1988) ‘Statistical analysis of co-integrating vectors’, Journal of Economics, Dynamics and Control 12: 231–54. King, R.G., Plosser, C.I., Stock, J.H. and Watson, M.W. (1991) ‘Stochastic trends and economic ﬂuctuations’, American Economic Review 81: 819–40. Krugman, P. (1979) ‘A model of balance of payment crises’, Journal of Money, Credit and Banking 3: 311–25. Liquitaya, J.D. (1998) ‘Dinero, producto, tasas de interés y precios’, Investigación Económica 225: 99–128. Lucas, R. (1988) ‘On the mechanics of economic development’, Journal of Monetary Economics 22: 3–42. MacDonald, R. and Taylor, M.P. (1991) ‘Exchange rates, policy convergence, and the European Monetary System’, Review of Economics and Statistics 73: 553–7. Mankiw, N.G., Romer, D. and Weil, D. (1992) ‘A contribution to the empirics of economic growth’, Quarterly Journal of Economics 107: 407–37. Masson, P.R. (1996) ‘Fiscal dimensions of EMU’, Economic Journal 106, July: 996–1004. Mishkin, F.S. (1992) ‘Is the Fisher effect for real? A re-examination of the relationship between inﬂation and interest rates’, Journal of Monetary Economics 30: 195–215.

Is a convergence path of monetary policies possible?

45

Mundell, R.A. (1963) ‘Capital mobility and stabilisation policies under ﬁxed and ﬂexible exchange rates’, Canadian Journal of Economics 29: 475–85. Mussa, M. (1986) ‘Nominal exchange rate regimes and behaviour or real exchange rates: evidence and implications’, in Brunner, K. and Meltzer, A.H. (eds), Carnegie Rochester Conference Series on Public Policy 25, Amsterdam: North Holland. Neusser, K. (1991) ‘Testing the long-run implications of the neoclassical growth model’, Journal of Monetary Economics 27: 3–36. Ortiz, G. (1983) ‘Currency substitution in Mexico: the dollarization problem’, Journal of Money, Credit and Banking 15: 174–85. Parikh, A., Bailey, D. and Lovatt, D. (1997) ‘The European Monetary System and various tests of policy convergence’, The Journal of International Trade and Economic Development 6: 43–62. Phillips, P.C.P. and Perron, P. (1988) ‘Testing for unit root in time series regression’, Biometrica 75: 335–46. Quah, D.T. (1996) ‘Twin Peaks: growth and converge in models of distribution dynamics’, Economic Journal 106, July: 1045–55. Ramirez-Rojas, C.L. (1985) ‘Currency substitution in Argentina, Mexico, and Uruguay’, IMF Staff Papers 32, 4: 629–67. Rogers, J.H. (1992) ‘Convertibility risk and dollarization in Mexico: a vector autoregressive analysis’, Journal of International Money and Finance 11: 188–207. Romer, D. (1986) ‘Increasing returns and long run growth’, Journal of Political Economy 94: 1001–37. Sachs, J.D., Tornell, A. and Velasco, A. (1996) ‘Financial crises in emerging markets: the lessons from 1995’, Brooking Papers of Economic Activity 1: 147–215. Sala-i-Martin, X. (1996) ‘The classical approach to convergence analysis’, Economic Journal 106, July: 1019–36. Stockman, A. (1983) ‘Real exchange rates under alternative nominal exchange rate systems’, Journal of International Money and Finance 2: 147–66.

Comments Francesco Farina

The question addressed by Chapters 1 and 2 is to verify the costs and beneﬁts for Mexico of a possible passage to coordination of monetary policies among the countries forming NAFTA, ranging from a ﬁxed exchange rate agreement to a currency board, to a monetary union. Both papers agree that – due to the large inﬂation differential with the US – the ﬂoating exchange rate with the dollar is still the best option for Mexico. Surrendering monetary sovereignty would be too costly, because high inﬂation and the high propensity to import inputs put the Mexican economy at the risk of trade account imbalances. The external constraint, coupled with low monetary policy credibility and the weak structure of the banking system which the recent crisis has dramatically conﬁrmed, would expose the currency to speculative attacks. However, the fact that a wide inﬂation differential with the US coexists with a huge wage differential at the disadvantage of the Mexican workers suggests that the ﬁrst objective of monetary and government authorities must consist in reforms to remedy the backwardness of the productive structure. The present ﬂoating exchange rate with the dollar has proved unable to push the Mexican economy to struggle for efﬁciency. Then, the question becomes whether a ﬁxed exchange rate agreement would be capable to ﬁrmly put the Mexican economy on the track of a rapid disinﬂation, at the same time allowing the country to continue its export-led expansion which is the pre-condition for reforms in the real and ﬁnancial sectors. Although the integration process of the Mexican economy with the two North American countries presents very speciﬁc features, with NAFTA being a unique example of free trade between an underdeveloped and two developed countries, it may be useful to compare this Center-NorthAmerican integration process with the path from the Common Market to the common currency followed by the countries belonging to the European Monetary Union. A ﬁrst observation is that, just as the ﬁxed exchange rate agreement (European Monetary System) which has been in place in Europe from 1979 to 1997 has been centered on a leader country, Germany, a similar

Comments 47 agreement in America would be ruled by the United States. Yet, an important difference immediately shows up. During the period of ﬁxed exchange rates which followed, to the ﬁrst phase of adjustable parities and preceded the collapse of the system in 1992, the disinﬂation processes of the EMS peripheral countries (the economies outside the ‘core’ of the system, represented by Germany and the so-called DM zone including the Benelux and Denmark) have certainly beneﬁted from the increase in Central Bank reputation and monetary policy credibility due to the DM pegging. However, not only has the decline of inﬂation differentials with Germany been very slow, but those countries have been obliged to follow the tight monetary policy dictated by the Bundesbank, thus bearing the deﬂationary effects of high interest rates even in presence of high unemployment and a depressed business cycle. Moreover, the scrutiny by international ﬁnancial markets of the soundness of monetary and ﬁscal policies, which has become severe with the completion of the capital markets’ liberalization process at the end of the 1980s, was at the origin of the risk premium which pushed the nominal and real interest rates in most of Southern European countries at even higher levels, thus opening wide interest differentials with Germany. Therefore, the peripheral EMS economies have suffered from the deﬂationary bias due to the depressionary effect of high interest rates on aggregate demand (Farina 2000). The prolonged slack in demand, stemming in Europe from the functioning of the EMS, caused a substantial increase in unemployment rates, which was then aggravated by the wage and job protection in the labor markets. The launch of the European Monetary Union does not seem to improve growth rates in Europe. Due to the continuation of prudent money creation by the ECB and to the deﬂationary effects of the Stability Pact (especially in countries where the public budget is burdened by heavy interest payments as a consequence of their large public debt), aggregate demand continues to stagnate. In the absence of expansionary macroeconomic policies, the sluggishness in Europe of the adjustment mechanism through wage and price ﬂexibility contributes to transform the lack of jobs for the new generations and the job losses among the unskilled in a larger amount of structural unemployment in all the age sections of the labor force. A ﬁxed exchange rate agreement among NAFTA countries would not be plagued by this trade-off between disinﬂation and growth, that is by the lack of the expansion which would be needed in order to foster the demand for labor after the huge increase in unemployment provoked by a long period of tight money creation. Quite on the contrary, also due to high ﬂexibility in labor markets, this regime is unlikely to be undermined in America by a deﬂationary bias and by the consequent rise in the unemployment rate. Given that the Federal Reserve is in the position to balance current deﬁcits by issuing dollars, and has traditionally rendered the monetary stance much more expansionary than the Bundesbank’s one, a

48

Francesco Farina

tighter linkage of the Mexican monetary policy to the dollar should not impose a sacriﬁce of growth to low inﬂation. In fact, the expansion of the Mexican economy is sustained by a large amount of exports (27 per cent of the GDP), and is certainly beneﬁting from geography, with 80 per cent of these exports going to NAFTA markets, mainly towards the United States. This trend is bound to continue, as Mexican exports are pulled by the steadily high demand in the US and Canada, and despite the slowdown provoked in 1996–7 by the ‘peso crisis’ the intra-regional export shares of the NAFTA countries improved from 48.0 per cent at the inception of the agreement in 1994 to 51.0 per cent in 1998 (IMF 1999). The US exports to Mexico mainly consist of capital goods to build new plants in Mexico and of intermediate goods which are shipped to Mexican ﬁrms in order to be ﬁnished there by low-wage workers and then reexported to the US. There is no doubt that ‘investment diversion’ and specialization in speciﬁc stages of the productive process determine an upgrading of the Mexican manufacturing sector. Most of the new productive capacity created is made up of unskilled workers, so that a structural change in the form of a raise of the productivity growth rate predicted by the ‘endogenous growth’ models may not be substantial. Positive externalities are taking place anyway, such as the diffusion of technological knowledge and the boost to the formation of clusters of industries, with the birth of subsidiary ﬁrms beneﬁting from the technological interdependence with the US machinery utilized in these plants. Yet, the transition to a more efﬁcient and modern productive structure cannot be led only by this peculiar version of the theory of ‘comparative advantages’. The limit of this expansionary process, led by the US imports, is the lack of a clear strategy aimed at transforming the integration process with the US economy from an ‘engine for growth’, where the Mexican economy functions as the subsidiary manufacturing partner of the US industries, in a modernization process of the industrial structure more according to the efﬁcient allocation of internal resources than to the mere displacement of the US manufacturing. To pursue this objective, drastic cuts in production costs are required in order that the inﬂationary pressures could be curbed. The problem is that the present ﬂoating exchange rate of the peso, while allowing Mexico to cope in the short run with the competitiveness losses due to the much lower productivity in manufacturing compared to the US, is not pushing the economic system towards the ‘change of regime’ of low inﬂation. The EMS functioning has pushed Central Banks and governments of the peripheral EMS countries to achieve monetary stability, improve productivity and re-equilibrate the public budget. A ﬁxed exchange rate agreement among the NAFTA countries would boost a faster disinﬂation process in Mexico and would compel monetary and government authorities to start structural reforms both in the real and the ﬁnancial sectors. As a consequence of the increase in monetary policy credibility, interest rates would

Comments 49 fall so as to allow an expansionary process of the Mexican economy, led also by the internal components of aggregate demand, thus more efﬁciently exploiting the negative wage differential with the US. The authors of the two chapters defend the ﬂexible exchange rate with the dollar because of the higher probability of ﬁnancial crises to which Mexico could be exposed in a ﬁxed exchange rate regime. But the threat of ﬁnancial turmoil is not a point in favour of ﬂexible exchange rates. It is well known that a country cannot simultaneously maintain ﬁxed exchange rates, a liberalized capital market and a monetary policy oriented towards the internal objectives of macroeconomic performance. In the case of Mexico, this ‘inconsistent trinity’ is even more compelling, because liberalized capital markets can conﬂict with monetary policy autonomy, whatever the foreign exchange regime. Even in a regime of ﬂexible exchange rates, the credibility problem this unstable country is plagued with can bring ﬁnancial markets to distrust the Central Bank, no matter if monetary policy is oriented towards the fulﬁllment of the balance of payments constraint and not to fuelling higher demand. With a ﬂoating currency, authorities are supposed to be more prone to rely on short-term remedies. The ‘bandwagon effect’, that is the endogenous formation of expectations of an expansionary maneuver triggered by devaluation and public deﬁcit, can lead ﬁnancial markets to launch speculative attacks against a currency. A ﬂoating peso then exposes Mexico to a depreciation process vis-à-vis the dollar (with increases in import prices which might enlarge the inﬂation differential with the US even further) and does not shield Mexico from the loss of sovereignty stemming from the dollarization of the economy. The problem is that, in the presence of fully liberalized capital movements, a ﬁxed exchange rate agreement among the NAFTA countries in Mexico might take the shape of a currency board, with the Mexican economy constrained by the amount of dollars entering in the country in exchange for the Mexican exports. Similarly to the underdeveloped Asian countries, Mexico faces the problem of how to exploit the advantage of large capital inﬂows permitted by the uniﬁcation of ﬁnancial markets without suffering from the distorted functioning of full liberalization, where huge amounts of capital rapidly move across borders in search of the highest short-term revenue. Therefore, the narrow path towards a much lower inﬂation rate consists, for Mexico, in entering a ﬁxed exchange rate agreement but protected from the destabilizing effects of open capital markets. The recent defeat of the radical IMF approach to the opening of developing countries to the international markets might allow Mexico to take advantage of the more pragmatic intellectual climate and accelerate the passage from the common market to a tighter monetary policy coordination. The terms of the bargain could be the commitment to start the structural reforms in exchange for a progressive switch to ﬁxed exchange rates, through an initially large band and taxes both on the entrance and on the exit of short-term capital ﬂows.

50

Francesco Farina

Bibliography Farina, F. (2000) ‘The economics of the European Monetary Integration and the convergence process’, in Faber, O., Farina, F. and Punzo, L.F. (eds) European Economies in Transition, London: Macmillan. IMF (1999) World Economy Outlook, October, II, 55 and V, 4.

PART II

Mexico’s foreign sector, the exchange rate, trade and productive structures

3

The external relationship of the Mexican economy with the United States, and its implications beyond trade María Elena Cardero

The North American Free Trade Agreement (NAFTA) created expectations and questions regarding the further integration of a developing country with two of the world’s most developed economies, one of which is today the most powerful. Five years have passed since the Agreement was implemented. There is curiosity about results that has stimulated papers on numerous topics having to do with the various impacts of NAFTA. This chapter attempts to answer some of the questions posed. I analyze some changes that have been brought about in Mexico’s foreign sector dealings with the United States in two areas: ﬁrst, the modiﬁcations of Mexico’s external sector and the role that the maquiladora industry, the ﬁnancial opening, and the 1995 crisis have had on them. These aspects are taken up in the ﬁrst part of the study. Then, together with the trade, productive and ﬁnancial phenomena, and as part of this complex and long process of integration between Mexico and the US, the binational labor market on both sides of the border is taken up in the second half of the chapter. The ﬁnal section presents some of the chapter’s conclusions. In general, NAFTA continues the long historic process of economic integration of Mexico’s productive sectors with the US economy. With its implementation, the process has accelerated and deepened in areas such as investment, government procurement, intellectual property rights, and procedures for dispute settlement. Yet, two essential aspects of the relationship remain outside this Agreement: the labor market and the failure to acknowledge the economic asymmetries that exist between Mexico and its two partners. These matters will have negative effects in the relationship if not promptly dealt with.

3.1 A new export–import model Between 1985 and 1998, the Mexican foreign sector underwent profound changes that began with a critical modiﬁcation of the pattern of growth, moving from an economy based on import substitution to one whose central objective was to achieve export-led growth.

54

María Elena Cardero

3.1.1 Opening and changes in the Mexican foreign sector In this context several policies were implemented that affected the domestic productive structure and the share of the foreign sector in major macroeconomic aggregates. Today the foreign sector has greater weight in these aggregates. In particular, the country’s export proﬁle has been modiﬁed, taking on an export–import conﬁguration. These changes occurred at great speed and were pushed along by the rapid entrance of both portfolio and direct foreign investment. This new export–import model, weakly linked to domestic production, accelerated and deepened with the signing of NAFTA.

3.1.1.1 Changes in economic policy The productive apparatus that had grown during the import substitution model was opened rapidly in the 1980s. Additionally, in 1982 privatization of state companies,1 the state ﬁnancial apparatus and other sectors also began. Policies of public spending were substantially modiﬁed, reducing the budget deﬁcit to the point that, within a few years, a ﬁscal surplus was obtained; the ﬁght against inﬂation was given priority, a ﬁnancial opening was undertaken and the process of deregulating the economy began. In 1986, the country’s foreign accounts were severely affected by the fall in oil prices2 leading to severe exchange rate modiﬁcation and a great deal of speculation. Inﬂation reached a historic high of 159 percent in December 1987, the highest since the 1910–20 Mexican Revolution. To hold down prices, a monetary policy was introduced that assumed the existence of inertia as a component of the inﬂationary spiral. So far neither the decrease in the ﬁscal deﬁcit, nor the stranglehold on growth and severe reduction of wages, were able to control prices. Thus the efforts made at stabilizing the economy strayed from ‘orthodox’ policy and took on a more heterodox look (Aspe 1991). An agreement (‘concertación’) was signed among diverse social sectors and implemented by numerous pacts in which the prices of public goods were raised, the pace of privatization of state enterprises3 quickened, public spending reduced, especially in investment, while the principal prices in the economy were kept ﬁxed, i.e., the exchange rate, salaries,4 and prices of products with a signiﬁcant impact on the economy. In 1987, a decision was taken to accelerate the pace of the foreign opening, even beyond that agreed to the previous year with Mexico’s entry to the GATT (General Agreement on Tariffs and Trade). It was hoped that the prices of imports would become an anchor for prices of domestic products. In a few years Mexico went from being highly protectionist to being one of the most open nations in the world. While, in 1982, 100 percent of the value of imports was controlled, by 1991 only 9.1 percent was under some type of control, dropping even further by 1993 to 5.0 percent. In the ﬁrst

The external relationship of the Mexican economy 55 Table 3.1 Changes in tariff structure 1985–8 Month

Year

Average tariff

Coverage import permits

December June December June December June December

1985 1986 1986 1987 1987 1988 1988

28.5 24.0 24.5 22.7 11.3 11.0 11.0

47.1 46.9 39.8 35.8 25.4 23.2 22.9

Source: Secretaría de Comercio y Fomento Industrial.

stage (1983 to June 1985), 16.4 percent of imports were freed from a permit requirement. The decree of 25 July, 1985 accelerated liberalization, as 64.1 percent of imports were freed from the permit requirement, and to compensate for the fall in non-tariff barriers, the tariff was increased to an average 25.2 percent. By 1987 liberalization covered 73.2 percent of imports, the number of tariff levels were reduced to seven and the maximum tariff rate was set at 40 percent, with a weighted average tariff of 11 percent. In the ﬁnancial sphere, to cover foreign debt commitments between 1983 and 1987, a spectacular net transfer of resources was made abroad, amounting to about 8 percent of annual GDP.5 This led to efforts to restructure the schedule of payments abroad, which culminated in the Brady Plan that allowed postponement of payments. 3.1.1.2 Changes in the foreign sector’s share in supply and demand As a result of the changes mentioned, the weight of the external variables in the make up of domestic supply shifted. While in 1989 imports made up 11.7 percent of the total supply, ten years later it had doubled to 22.5 percent. Starting in 1994, but especially in the past few years, foreign demand has grown at a higher rate than internal demand, thus becoming the more dynamic component. Financed by the ﬂood of foreign resources, private consumption rose steadily until 1993. The drop that year was partially held in check by the rise in public consumption, and in 1994 both rose signiﬁcantly. In 1995, with the outbreak of another economic crisis, private consumption shrank 9.5 percent (the component of durable goods fell 32.8 percent; nondurable goods fell by 8.3 percent, and services fell by 4.1 percent). During the ﬁrst quarter of 1997, private consumption recovered except in nondurable goods, which again fell by 4.2 percent. Further, the recovery failed to reach the poorest households, i.e., consumers of basic products who represent 5/6 of domestic private consumption. Investment grew slightly between 1988 and 1992 and fell in 1995, barely

100.0 86.9 13.1 100.0 61.9 9.6 14.8 1.5 12.2

1990

5.6 4.2 18.0 5.5 7.3 2.2 5.8 24.7 5.7

6.8 5.1 19.7 6.8 6.4 3.3 13.1 2.3 5.3

Annual variation

100.0 88.3 11.7 100.0 62.1 9.9 14.0 1.6 12.4

1989

5.7 4.2 15.2 5.7 4.7 5.4 11.3 0.8 5.1

100.0 85.7 14.3 100.0 61.3 9.6 15.6 1.4 12.2

1991

5.9 3.6 19.6 5.9 4.7 1.9 10.8 41.6 5.0

100.0 83.9 16.1 100.0 60.6 9.2 16.3 1.8 12.1

1992

1.9 2.0 1.9 1.9 1.5 2.4 2.5 14.4 5.1

100.0 83.9 16.1 100.0 60.3 9.3 15.6 2.0 12.8

1993

7.0 4.4 20.5 7.0 4.6 2.9 8.4 20.0 17.4

100.0 81.8 18.2 100.0 58.9 8.9 15.8 2.4 14.1

1994

7.4 6.2 12.8 7.4 9.5 1.3 29.0 73.8 30.0

100.0 83.2 16.8 100.0 57.8 9.5 12.1 0.7 19.9

1995

8.1 5.1 22.8 8.1 2.2 0.7 16.4 27.0 18.2

100.0 80.0 19.1 100.0 54.6 8.7 13.1 1.9 21.7

1996

9.8 6.8 22.7 9.8 6.5 2.9 21.0 39.6 10.7

100.0 78.8 21.2 100.0 52.9 8.1 14.4 2.3 22.3

1997

7.3 4.8 16.5 7.3 5.4 2.2 10.3 4.5 12.1

100.0 77.2 22.8 100.0 52.7 7.6 14.9 2.3 22.5

1998

Source: National Institute of Geography, Statistics and Information (INEGI). SCNM base line year: 1993, GFFC means Gross Formation of Fixed Capital.

Supply GDP Imports Demand Private consumption Governmental consumption GFFC Variation of inventories Exports

Supply GDP Imports Demand Private consumption Governmental consumption GFFC Variaton of inventories Exports

1999

Structure

Table 3.2 Global supply and demand, annual structure and variations (based on prices in 1993)

13.9

5.8 3.7 12.8 5.8 4.3 1.0 5.8

24.6

100.0 75.3 24.7 100.0 51.3 7.5 14.8

The external relationship of the Mexican economy 57 recovering its 1988 level until 1997. A large part of the resources went to servicing foreign debt, leading to a substantial fall in construction and a lack of investment in machinery and equipment, which affected production and the ability to maintain and create jobs. 3.1.1.3 The new components of the foreign sector One of the outstanding characteristics of the structural change of the Mexican economy is the growing weight of the foreign sector, nowadays fully a quarter of national aggregate value. While in 1988 exports contributed 16.6 percent to GDP, in 1998 that share had risen to 31.2 percent, while the share of imports went from 15.2 percent to 33.3 percent of GDP. Surprisingly, this change occurred from one year to the next. In 1995, when GDP fell by more than six points, the weight of imports in the GDP grew by 5.8 points (see Table 3.3), and exports by 12.2 points. Among relevant changes in the export structure we have: A decline in oil income. Oil income from foreign sales declined continuously, but recovered slightly in 1996 due to an increase in the prices of crude. Agriculture’s export declined. The share of the agricultural sector was paltry during this period as opposed to the 1950s, 1960s and 1970s, when it was the principal source of hard currency for the Mexican economy. Today agriculture barely contributes 1 percent of exports as a share of GDP. A sharp increase in manufacture’s exports. These increased from an average of 10.4 percent of GDP to almost 12 percent in 1994, and then doubled its share in just a year to 24 percent of GDP in 1995. During the period under study, manufacturing imports and exports became determining factors in the make up of the foreign sector. Manufacturing imports increased by 30 percent between 1988 and 1993, a period during which the wide-scale opening in the Mexican economy occurred, now comprising more than 90 percent of total imported value. As a result of the heavy intra-industrial and maquiladora trade, purchases of some foreign intermediaries became a ﬁxed percentage of total imports. This is the case of metallic products, and machinery and equipment that make up 67 percent of exports and are, on average, 57 percent of the value of manufacturing imports. The strong surge in input imports. Linked to exports, input imports surged, doubling between 1993 and 1995 – that is, from a year before to a year after the implementation of NAFTA – making up 20 percent of domestic production. Imports concentrated in the so-called export-associated ﬁrms. The increase in foreign sales of the export-associated ﬁrms rests on the import of intermediary inputs which, added to their purchases of capital goods, elevated these ﬁrm’s imports to 16.5 percent of GDP during the ﬁrst half

16.6 5.5 11.1 3.6 3.2 0.4 13.0 0.9 0.4 11.7 5.5 6.2 15.2 4.2 11.0 1.0 12.0 4.2 7.7 – – 2.2 – – 1.4 1.3 0.1

TOTAL EXPORTS2 Maquiladoras Transformation Petroleum Crude oil Other3 Non-petroleum Agricultural Extrative Manufacturing –Maquiladoras –Transformation TOTAL IMPORTS (FOB) Maquiladoras Transformation Consumption goods Intermediate goods Maquiladoras Transformation Associated with exports4 Non-associated with exports4 Capital goods Exporting ﬁrms4 Non-exporting ﬁrms4 TRADE BALANCE Maquiladoras Transformation

14.5 5.1 9.4 3.2 3.0 0.2 11.2 0.7 0.2 10.2 5.1 5.2 15.6 4.2 11.4 1.6 11.8 4.2 7.7 – – 2.1 – – 0.2 1.3 –1.2

1989 15.5 5.3 10.2 3.8 3.4 0.4 11.6 0.8 0.2 10.6 5.3 5.3 15.9 3.9 11.9 1.1 11.3 3.9 7.4 – – 2.6 – – –0.3 1.3 –1.7

1990 13.6 5.0 8.5 2.6 2.3 0.3 11.0 0.8 0.2 10.0 5.0 5.0 15.9 3.7 12.1 1.9 11.3 3.7 7.5 – – 2.7 – – –2.3 1.3 –3.6

1991 12.7 5.1 7.6 2.3 2.0 0.2 10.4 0.6 0.1 9.7 5.1 4.6 17.1 3.8 13.2 2.1 11.8 3.8 7.9 – – 3.2 – – –4.4 1.3 –5.7

1992 12.9 5.4 7.5 1.8 1.6 0.2 11.0 0.6 0.1 10.3 5.4 4.9 16.3 4.1 12.2 2.0 11.6 4.1 7.5 6.1 5.5 2.7 0.5 2.2 –3.4 1.3 –4.7

1993 14.4 6.2 8.2 1.8 1.6 0.2 12.6 0.6 0.1 11.9 6.2 5.7 18.7 4.8 13.9 2.2 13.3 4.8 8.5 7.4 5.9 3.1 0.5 2.6 –4.3 1.4 –5.7

1994 26.8 10.5 16.3 2.8 2.5 0.3 23.9 1.4 0.2 22.4 10.5 11.9 24.5 8.8 15.5 1.8 19.6 8.8 10.8 13.7 5.9 2.9 1.0 1.9 2.4 1.7 0.7

1995 29.2 11.0 17.6 3.5 3.2 0.3 25.2 1.1 0.1 24.0 11.0 13.0 27.1 9.1 17.6 2.0 21.4 9.1 12.3 14.8 6.6 3.3 1.1 2.1 2.0 1.9 n.s.

1996 27.4 11.2 16.2 2.8 2.6 0.3 24.6 1.0 0.1 23.6 11.2 12.3 27.3 9.0 18.3 2.3 21.2 9.0 12.2 14.8 6.5 3.8 1.2 2.5 0.2 2.2 –2.0

1997 28.3 12.8 15.5 1.7 1.5 0.2 26.6 0.9 0.1 25.5 12.8 12.8 30.2 10.0 19.9 2.7 23.3 10.2 13.1 16.4 7.0 4.2 1.4 2.9 –1.9 2.5 –4.4

1998

Notes 1 Calculations based on the new GDP series at current price from INEGI, base year: 1993. SHCP was used to convert dollars to pesos. 2 Includes a revaluation due to an update of the value registered in the customs schedules and prevailing in the international market, especially with agricultural goods, ﬁsh and seafood. 3 Includes natural gas derivates from petroleum and petrochemicals. 4 Data are given from the year in which record taking started. 5 Preliminary data.

Source: Ernesto Zedillo (1998), Fourth Report of the Government; and INEGI BDI database.

1988

Category

Table 3.3 Share of foreign trade in the Gross Domestic Product (percentages)1

27.7 12.7 15.0 1.6 1.4 0.2 26.1 1.2 0.1 24.8 12.7 12.1 28.6 10.0 18.5 2.3 22.1 10.0 12.1 16.0 6.1 4.1 1.4 2.7 –0.9 2.6 –3.5

19995

The external relationship of the Mexican economy 59 of 1998. However there was a relatively minor weight of intermediate imports of industries not linked to exports, which went from 5.5 percent in 1993 to 6.6 percent of GDP in the ﬁrst half of 1998, and to which we must add their purchases of capital goods, resulting in 9.4 percent of GDP. Constant drop in domestic value added. The domestic integration of manufacturing exports fell from 91.4 percent in 1983 to 77 percent in 1987 and 39 percent in 1994 (Vázquez 1995). The most recent calculations of the degree of domestic integration in manufacturing exports, measured by subtracting the sector’s total exports from imports associated with them and dividing by the former, show similar results: domestic value added fell from 41.3 percent in 1993 to 38 percent in 1995. The most signiﬁcant change with the greatest impact was the shift from an import substitution model to an export–import model. This shaped new structural linkages in which, in order to maintain the pace of exports, there is a growing need to purchase goods abroad, causing a de-coupling of the foreign sector from national production.6 3.1.1.4 Geographic concentration of foreign trade Among the objectives that were expected from opening trade doors and linking Mexico to the world economy was the geographic diversiﬁcation of foreign trade. This did not occur and, in fact, Mexico’s trade with the United States became even more important. In the year of NAFTA’s startup, Mexico–US trade rose from 69 percent of total exports in 1990 to 85 percent in 1994. Further, United States has been Mexico’s principal supplier, although not so high in percentage terms as in the case of exports. But the tendency is towards a growing consolidation of 75 percent of Mexico’s purchases abroad. The goods traded between the two countries are extremely varied and cover practically all economic sectors. Much of the trade is intraindustrial. Globerman (1992) estimates that Mexico’s index of intraindustrial trade (IIT) with the United States went from 43 in 1980 to 63 in 1987, thus reﬂecting the advances in trade liberalization. Estimates by Esquivel (1991) show an IIT of 27 in 1981 and 54 in 1990. Máttar and Shatán (1993), who analyzed auto parts, electronics and petrochemical sectors, ﬁnd that intraindustrial trade therein rose signiﬁcantly from 24 in 1982 to 79 in 1990. Mexico is the United State’s principal market in Latin America, acquiring more than half of US exports to the region. While Mexico imports more than 75 percent of its total imports from the United States, in all other countries, except Venezuela and Chile, imports from the US are in general no higher than 30 percent. Insofar as Mexico’s trade with Canada is concerned, although it has also tended to grow, particularly since the start of NAFTA, it is clear that trade

60

María Elena Cardero

Table 3.4 Percentage share of Mexico’s trade with Canada and the United States Exports

Imports

Year

Total

Canada

US

Total

Canada

US

1990 1991 1992 1993 1994 1995 1996 1997 1998 19991

100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

1.7 2.6 2.2 3.0 2.4 2.5 2.3 2.0 1.3 1.7

68.6 79.4 81.0 83.0 85.2 83.4 83.9 85.4 87.8 88.2

100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

1.5 1.4 1.7 1.8 2.0 1.9 2.0 1.8 1.8 2.1

65.5 73.7 71.2 71.1 71.7 74.3 75.4 74.7 74.4 74.2

Source: INEGI, Foreign Trade Annual Surveys and the Bank of Mexico. Note 1 Preliminary data.

between the two countries is comparatively limited. Canada is an important supplier of milk to Mexico, while the latter sells diverse agricultural products to the former. There is also a certain amount of intraindustrial and maquiladora trade among both countries, affected by the presence of the United States, and the distribution that takes place in the US with regard to its principal suppliers of maquiladora articles. The European Union (EU) is Mexico’s second partner with around 6 percent of total trade. Mexico’s trade balance with this regional market is permanently in the red. The principal trade partner in the region is Germany, followed by Spain, France, Italy and the United Kingdom. The United States is the primary source of foreign investment in Mexico’s productive sector, with almost 60 percent of direct foreign investment accumulated in 1997. However seven countries within the EU make up 21.3 percent of Mexico’s foreign investment, with the UK and German capital outweighing Japanese and Canadian capital. The prospect of a Free Trade Agreement being negotiated between the EU and Mexico is part of the growing competition between the EU and the US for the Latin American market in a strategic geopolitical and economic environment. Since Latin America is the second fastest growing region in the world, after the onset of the Asian crisis, and particularly with the recent revision of its particular development paradigm, the region has become more attractive. Yet the EU’s interest in Latin America revolves basically around a small number of countries. The region’s production is heavily concentrated in a few of them. While 80 percent of the GDP on the American continent is generated by the United States, 15.2 percent is produced in Brazil, Mexico, Canada and Argentina. Brazil is larger than Canada and

The external relationship of the Mexican economy 61 Mexico but has a per capita GDP below that of Canada’s. The remaining 4.8 percent is produced by the other countries of the region. Trade topics are important to Europe, but so are the spaces which would be opened in a wide-scale trade agreement, such as the treatment accorded to investment, the large-scale telecommunications market, government procurement, and protection to intellectual property, among others. In the case of Mexico, as in the rest of Latin America, the idea is to strengthen links with Europe as a counterweight to the economic and political inﬂuence of the United States, attract capital from abroad, and to improve its trade position in the European market. 3.1.2 The maquiladora export industry During the period under study, the wide-scale opening of the maquiladora export industry (MEI) has taken on an important role, as today it produces almost half of total Mexican exports. While export activity of traditional productive sectors closely follows the pattern of the general economy, the maquiladora export industry’s performance is largely independent of the economy and of nonmaquiladora production. Geographically, maquiladora labor has ceased being exclusively a border phenomenon and is now increasingly extended to non-border states, to the point that today the non-border maquiladora labor force is a third of all maquiladora labor. At the national level, the impact of the MEI on GDP is very small, generating only 1.8 percent of GDP and 6.03 percent of manufacturing GDP. Table 3.5 Sectorial shares of Mexico’s foreign trade in merchandise percentage shares

EXPORTS Maquiladora Transformation Agricultural Manufacturing Extraction industries IMPORTS Maquiladora Transformation Agricultural Manufacturing Other sectors

1980

1988

1993

1996

1997

1998

19991

14 86 8.4 34.1 57.7

33.1 66.9 5.4 40 21.3

42.2 57.8 4.8 81.9 13

38.2 61.5 3.7 84.4 11.7

40.9 59.1 3.4 86.5 9.8

45.2 54.8 3.2 90.3 9.4

46.6 53.4 3.0 89.4 9.3

8.3 91.7 7.2 88.2 4.6

27.8 72.2 5.9 64.5 29.6

25.8 74.8 4 94.2 1.8

34.1 65.9 5.2 90.7 4.1

33.1 66.9 3.8 92.5 0.8

33.9 66.1 3.4 92.9 0.8

35.5 64.5 2.8 93.8 0.6

Source: Ernesto Zedillo (1998), Fourth Report of Government; and Bank of Mexico. Note 1 Preliminary data.

62

María Elena Cardero

Table 3.6 Industry exports and imports (average annual growth rates) Exports

Imports

Years

Totals

Transformation Maquila

Totals

Transformation Maquila

1980–2 1983–7 1988–93 1994–7

15.5 1.6 11.1 22.0

17.0 2.1 7.9 23.5

10.2 12.3 18.4 11.4

11.8 10.2 19.3 7.6

5.9 18.2 16.6 19.8

6.3 18.2 16.1 21.1

Source: Author’s calculations based on report from the Bank of Mexico.

Yet its repercussions in terms of employment is signiﬁcant since it employs almost a third of all labor – near a million jobs – in manufacturing. Its minor impact on domestic production reﬂects the low level of wages paid and the reduced weight that these wages have on the value of total product. Although the maquiladora sector’s weight in manufacturing GDP is, on average, 6.03 percent, this ﬁgure is higher in 11 branches (see Table 3.7). Rearranging the branches of manufacturing’s nine divisions, we ﬁnd that the core of exports within the MEI is comprised of metallic products, and machinery and equipment, since 70 percent of MEI’s exports have their origin in division VIII; 11 percent come from division II, textiles and apparel; 7 percent from division IX, medical and dental equipment and instruments; and 4 percent from division V, chemical substances, petroleum derivatives, rubber and plastic products. In some branches, for example branch 54 of division VIII, spare parts, electronic apparatus and equipment, televisions, radios, and modular sets, the MEI makes up 40 percent of total exports from the maquiladora sector and 84 percent of total sales abroad of the entire branch. Table 3.7 Share of MEI in Manufacturing’s GDP, by selected branches: percentage share Average of all manufacturing

6.03

Branches 54 55 57 27 59 52 26 30 53 50 51

Electronic equipment and apparatus Electrical equipment and apparatus Car bodies, motors and auto parts Dress items Other manufacturing industries Machinery and non-electrical equipment Other textile industries Other products of wood and cork Household appliances Other metallic products except machinery Non-electrical machinery and equipment

Source: Gómez Luna (1998).

34.24 28.43 20.36 20.24 19.79 16.99 15.13 10.90 8.46 7.81 6.24

The external relationship of the Mexican economy 63 In the most important maquiladora sectors, such as electronic equipment and apparatus, household appliances and auto parts, motors, car parts and accessories, imported inputs make up 90 percent of production value, a fact that favors the United States, since around 91 percent of the imports for the maquiladora industry come from there. The largest share of production from the maquiladora plants is sent to the US market. Mexico is today by far the principal supplier of maquiladora production in the United States, making up some 41 percent of total maquiladora trade done by the United States and 61 percent of the value of components used in imported products that return as ﬁnished products to the US under the HTSPSP regime.7 The Dominican Republic and Malaysia are other maquiladora suppliers to the US market, with 6 percent and 5 percent respectively of the value of imports with US content that enter the US under import tariff 9802.00.80. Even though, by 1996, US imports from Mexico under the HTS PSP provision had risen to US$28 billion, this ﬁgure may underestimate by US$14 billion the true amount, since companies were unwilling to pay the US$400 fee to register in customs under this provision, even though they were in strict compliance under ninety-eight import tariff provision. If we sum up these two categories we ﬁnd that, of the US$74 billion that the United States recorded in 1996 as imports from Mexico, 57.4 percent had US content, either produced by Mexican maquiladoras, assembly plants or export platforms. The US’s preference for the Mexican maquiladora is due to Mexico’s excellent competitiveness in wages, the closeness of markets, infrastructure, complementarity of productive structure and the possibility that managers can live in the United States and cross the border to work. Among other reasons, this is why more than a million daily border crossings occur between the two countries, while it is the only place in the world after the fall of the Berlin Wall that a barrier has been raised in order to contain the ﬂow of people. Mexico is the leading supplier and maquiladora processor of many of the economic branches that work with mature technologies such as automobiles, apparel, automotive wiring, TVs, auto parts, medical instruments, measuring instruments and internal combustion engines. Trade in parts, materials and machinery from the US to Mexico for assembly and ﬁnishing employs Mexican labor for US industry working inside Mexican territory. It is a type of labor that does not create social problems for the United States, unleashes no racist or xenophobic feelings, and is linked to the economy of the US in order to supply its internal consumption and foreign demand. It is evident that most jobs created by these industries are to be found in the US while companies employ a labor force in Mexico whose wages are sixteen times less than those paid on the other side of the border for similar work.

64

María Elena Cardero

On 1 January, 2001, according to NAFTA the (Mexican) maquiladora industry as such will cease to exist insofar as its effect on trade with the US and Canada. Then the maquiladora industry will be obliged to refund to the maquiladoras from outside the North American area only the taxes equivalent to the lowest existing tariff among the three countries. This implies that Mexico will be unable to maintain any special terms which might favor or encourage investment to locate within its boundaries that differ from those in the other countries. 3.1.3 The shape of an ‘emerging ﬁnancial market’ in Mexico Export–import expansion in recent years has brought with it an explosive rise in the trade deﬁcit, whose accumulated total between 1990 and 1994 was greater than US$49 billion. In order to meet this expansion and the maturing of foreign debt obligations, the Mexican stock market was ‘internationalized’ between December 1989 and July 1993. The purpose of the measure was to link it to capital movements and international ﬁnancial centers and to compete for foreign resources. This opening went unnoticed compared to the attention given to NAFTA, but it nonetheless had a resounding impact on the Mexican economy with similar or greater effects than trade liberalization. Internationalization of the Mexican stock market had two stages. In the ﬁrst stage, in December 1989, the market was opened to foreign portfolio investment; the venturing abroad of Mexican ﬁrms, stocks and intermediaries was supported in international markets, and institutional relationships were established with foreign authorities, stock markets and multilateral agencies of cooperation. Among the measures beneﬁting foreign investment was the establishment of a neutral investment mechanism to permit foreign equity to buy Series ‘A’ stocks, traditionally set aside for Mexican investors, giving patrimonial rights to corporations. Restrictions were lifted so that investors could buy government bonds, meaning that all market instruments were now freely available for purchase; foreign investors were allowed to buy equity of stock brokerage ﬁrms, banks and insurance companies, and terms were granted for capital repatriation in keeping with the implementation of an accessible and uniﬁed tax. In July 1993, a second stage began with the reform of the Stock Market Law, in which linkages between the stock markets were deepened, thus allowing entrance of foreign stock issuers and equity into the local market, in order to carry out intermediary and public offering activities (Moreno 1993). Starting in 1997, a legal and regulatory framework was drawn up for the International Bidding System (IBS) and the Mexican Stock Market (MSM), as a mechanism to list and operate within the MSM environment those stocks not publicly offered in Mexico and which are transacted in foreign stock markets recognized by the National Banking and Stock

The external relationship of the Mexican economy 65 Commission (Foreign Recognized Markets), or whose issuers had received recognition from the Commission (Recognized Foreign Issuers). The purchase of shares in the stock market can be undertaken through the following mechanisms, among others: • • • •

American Depository Receipts (ADR) Neutral Investments Freely tradable stocks and Investment funds bided on abroad (Mexico Fund).

Offerings of ﬁxed income instruments can be done through: • • • • •

Bonds Promissory notes Commercial stocks Fixed income funds located abroad Government bonds.

These modiﬁcations were technically approved in order to be recognized by the ﬁnancial authorities of the United States (Securities and Exchange Commission), and so that the Mexican stock market might be considered as a ‘designated foreign market’. This also permitted certain Mexican paper, such as the Ajustabonos, Bondes, Cetes and nineteen stock series to enter the market for effects of equity, as deﬁned by the Net Capital Rule. Access to international capital made it easier for Mexican ﬁrms to undertake their stock market operations abroad. In March 1999, 63 percent of their operations in stock markets were taking place outside Mexico. Of the 49 Mexican issuers with stock listed both in Mexico and the United States, 24 were experiencing greater placements abroad. Further, while there are funds from Mexico whose behavior has been erratic, the Mexico Fund earned 16.8 percent from January to March 1999, Mexico Equity and Income increased by 10.79 percent and the Emerging Mexico Fund grew by 5.42 percent.8 In the internal ﬁnancial market, regulations increasingly were related to open market operations, and the government’s ﬁnancing mechanisms went from using banks’ required reserves to competing for local resources. In spite of high interest rates, the coefﬁcient of private savings fell continually from 18 percent of the GDP in 1988 to 12 percent in 1994,9 thus encouraging increased foreign debt on the part of the private sector, which increased its share of the total foreign debt by over four times in just six years, going from 8.5 percent to 37.4 percent of the total. With these legal modiﬁcations, Mexico became an ‘emerging market’ of international private funds which ﬂowed into the stock market and which also entered as speculative capital gravitating around the arbitrage of the local currency.

66

María Elena Cardero

At an early stage the entry of more than US$71 billion in portfolio investment in just ﬁve years (1990–4) created a ‘wealth effect’ that became a credit bubble, encouraging consumption and underwriting a spectacular amount of imports. This increase in foreign demand (Table 3.8) further weakened Mexico’s productive structure, and further increased ruptures in the productive chains, while increasing de-industrialization. In 1993 much of the equity that entered the Mexican ﬁnancial market began to scale up its arbitrage operations by adjusting its portfolios and moving to other more secure assets, such as bonds with rate-of-exchange coverage, which in turn moved the CETES10 rate upwards. At the beginning of 1994 the external deﬁcit, maturing foreign debt and the exchange rate and ﬁnancial uncertainty combined with higher interest rates in the US. Then, PRI’s candidate for president, Luis Donaldo Colosio, was assassinated; all these events led to a speculative attack. The response of the Central Bank was to increase its public debt issues in foreign currency, eventually taking the Tesobonos to over US$33 billion in 1994. Subsequent political events generated new speculative attacks to which the government responded with monetary policies of raising interest rates and increasing public debt valued in foreign currency. It all ﬁnally ended in a profound ﬁnancial and economic crisis at the end of 1994, and a drop in the GDP of 6.2 percent in 1995. Given that much of the money invested was from US pension funds, the US government quickly put together a loan in excess of US$20 billion, backed by future oil income. Thanks to this package, private foreign capital began ﬂowing once again to portfolio investments,11 only to drop again in 1997 and 1998 as a result of the Asian ﬁnancial crisis.

3.2 The transborder integrated labor market Within the North American Free Trade Agreement the most signiﬁcant missing party in the negotiations between Mexico and its northern neighbors was the labor force. Trade in goods, services and capital was liberalized, but the other productive factor, labor, was subject solely to a series of lateral regulations on general labor conditions between the three countries. This reﬂected the wish of Canada and the United States to avoid Mexico undertaking some sort of disloyal competition in labor matters, which would displease labor unions, due to possible job losses in Canada and the US. Yet by so doing, the existence of a fundamental aspect of Mexico–US relations was ignored, with important consequences in matters of economic integration, linked to the presence of a transborder labor market. This labor market is not a recent phenomenon. In fact, the US takeover of Mexican territory during the nineteenth century left a part of the Mexican population living under a new ﬂag and in a new land, but with permanent links to the other side of the border.

2,375.6 5,821.3 2,609.5 405.0 30,691.5 35,171.0 28,082.0 34,766.0 1,163.1 3,175.9 591.0 4,346.1 3,879.8 3,526.8 2,880.0 3,175.5 999.8 451.3 493.3 s.d. 999.8 142.0 3,288.8 819.3 1,740.8 996.2 1,548.0 176.9 1,754.1 1,170.2

1989

1991

1992

1993

Note 1 Preliminary data.

1994

1995

1996

1997

1998

19991

7,451.1 14,646.7 23,438.6 23,399.2 29,661.9 1,576.6 2,330.4 7,448.4 15,726.4 14,012.8 882.4 7,279.1 15,933.8 13,480.5 18,463.7 7,088.5 6,530.9 623.6 7,913.5 5,360.4 40,710.9 42,687.5 46,195.6 51,886.0 60,882.2 79,541.6 95,999.0 110,431.4 117,459.6 136,703.4 41,593.3 49,966.6 62,129.4 65,366.5 79,345.9 72,453.1 89,468.8 109,807.8 125,373.1 142,063.8 8,297.2 24,507.5 26,418.8 32,482.3 14,584.2 15,405.6 4,069.2 15,762.7 17,464.5 14,141.8 16,996.7 25,507.1 20,866.9 36,084.8 20,254.2 22,763.3 10,410.4 9,046.9 17,033.0 16,781.5 6,003.7 17,514.7 22,433.9 33,308.1 19,154.7 –188.4 22,603.9 17,866.6 10,731.6 22,358.1 2,633.2 4,761.5 4,392.8 4,388.8 10,972.5 9,526.3 9,185.5 12,829.6 11,310.7 11,568.1 3,370.5 12,753.2 18,041.1 28,919.3 8,182.2 9,714.7 13,418.5 5,037.1 579.2 10,790.0 1,994.5 6,332.1 4,783.1 10,716.6 4,083.7 519.2 2,800.6 3,215.3 665.6 3,769.2 s.d. 3,406.4 8,146.9 7,405.7 2,225.3 13,859.6 907.5 584.8 214.1 131.4 1,376.0 3,014.8 5,111.1 10,797.0 6,323.8 3,625.7 9,710.4 1,237.0 127.7 6,889.4 10,993.0 7,992.4 1,567.0 2,776.7 1,099.5 22,951.7 12,193.5 8,819.7 6,301.4 5,576.6 6,212.0 140.5 3,990.4 3,383.4 1,564.2 24,784.7 12,441.5 9,543.3 438.2 8,477.5 4,781.0 8,132.9 2,423.4 6,160.1 2,663.7 1,833.0 248.0 723.6 5,863.2 2,900.9 8,699.5 999.6 5,551.9 3,602.5 5,670.0 7,357.7 6,341.2 6,715.8 431.5 2,639.7

1990

Source: Ernesto Zedillo (1998), Fourth Report of Government; and Bank of Mexico.

CURRENT ACCOUNT Trade balance EXPORTS IMPORTS CAPITAL ACCOUNT Liabilities Foreign investment Direct Portfolio Stock market Stocks local currency Stocks foreign currency Net debt Public sector Private sector Assets

1988

Table 3.8 Balance of payments: current account and capital (Millions of dollars)

68

María Elena Cardero

Halfway through the twentieth century, due to a lack of labor in the United States, the Bracero Program was created and lasted into the early 1960s, when the US refused to renew it. The migratory phenomenon, initially due to movements of the agricultural population, changed through the ensuing years as a result of various intertwined economic, social, cultural and demographic factors. Today this labor force is made up of both workers who cross the border temporarily or remain in the US, and by those in the maquiladoras who work for foreign, principally US, concerns, but on the Mexican side of the border. Among the factors contributing to this pattern of labor are the slow growth of the Mexican economy and its inability to absorb the workingage population; the wage difference between both countries; the demand for Mexican labor in the United States in agriculture, industry and the services sectors; the migratory tradition of going northward; and the workings of complex social and family networks that link places of origin and destination. After the end of the Bracero Program in the 1960s, the migratory ﬂow was mostly circular, composed principally of adults and young people from the rural sector who worked seasonally in agriculture for six to eight months, returning home. Today this proﬁle has changed (National Population Council 1997): • •

•

•

• •

the circular ﬂow of migrants has tapered off in favor of groups who tend to prolong their stay in the US; there is an increasing job and sectorial diversiﬁcation among migrants (rural migrants are no longer the majority, neither in terms of their origin or their destination); among temporary workers, 95 percent are men in their productive years (70 percent are between 12 and 34 years old), with more than six years schooling; half are married; two-thirds have previous migratory experience; on average they remain eight months in the United States with a monthly income of US$900, US$500 of which they remit to their families in Mexico; three of every ﬁve migrants are from urban locations; there is growing regional diversiﬁcation of the ﬂow, which means that apart from the traditional areas, migrants now come from other areas of Puebla, Hidalgo, Mexico City and Morelos; the main destinations are the states of California, Texas, Illinois and Arizona; due to the militarization along the border and the construction of the wall, the exit routes have shifted from Tijuana–California to Piedras Negras, and Nuevo Laredo, Texas.

Indirect measurements estimate that the loss of Mexican population due to international migration has increased since 1970. Estimates of these drops are:

The external relationship of the Mexican economy 69 • • •

Between 1.20 and 1.55 million during the 1970–80 decade; Between 2.10 and 2.60 million during the 1980–90 decade; 1.5 million between 1990 and 1995.12

In other words, the new annual ﬂow has grown by ten times, from an annual average of around 26,000 people during the sixties, to close to 300,000 migrants a year in the ﬁrst ﬁve years of the 1990s. In the Binational Mexico–United States Study on Migration, it is estimated that the number of Mexicans born in Mexico and residing in the USA ﬂuctuates between 7.0 and 7.5 million. Of these, between 4.7 and 4.9 million are authorized migrants and between 2.3 and 2.4 million live in the US without legal documents. The annual ﬂow of temporary workers between 1990 and 1995 was calculated by Mexican ﬁgures at 277,000 persons and by sources in the United States at 315,000. This average would seem to be growing, since the Survey of Migration on the Northern Border (SMNB)13 from March 1993 to March 1994 gave the temporary migrant ﬁgure of 508,000. During 1995 the number of migrants decreased to 387,000 and from mid-1996 to mid-1997 the ﬁgure rose to 424,000. There are two reasons behind this transborder labor market. On the one hand is undoubtedly supply, the primary cause of which is, as mentioned, the tremendous economic gap between the two countries. This gap has not closed over time but has widened due to the basic inequality in income distribution and the great disparity in wages that makes it extraordinarily attractive for many Mexicans to cross the border. This Mexican labor force14 is increasingly heterogeneous, and is now made up of urban workers, many of whom had jobs in Mexico, and who are better qualiﬁed, with more schooling, in their prime working years, better trained, young, whose education and health expenses were covered for the most part by Mexican ﬁscal resources. Along with this supply there is an increasing demand for labor in the United States, in a labor market segmented by region, sector and yearly seasons, differentiated and ﬂexible in terms of salary and beneﬁts. In general 95 percent of these temporary workers get a job, that is, there are job offers for them, and many of the employment opportunities then become relatively stable, leading workers to remain in the US, while others return, continuing the circular cycle based on a closely-knit social network. This demand for labor is not based solely on the need to ﬁll posts but also has to do with US demographic trends, now growing older, with a signiﬁcant reduction of working-age groups. The US Department of Labor estimates that by the year 2000 there will be a decrease of 6 percent in the number of young men from 16 to 24 years of age and a decrease of 15 percent in the 24–34 year-old group.

70

María Elena Cardero

When the economic gap between the two countries is narrowed, and Mexico experiences sustained growth for a long time, the migratory ﬂow will be reduced, but for the short to mid-term, estimates of the number of migrants are not expected to diminish. The changes in agriculture are particularly critical. This sector today employs one out of every four Mexican workers. Economic reforms and trade liberalization have brought a depression of basic agricultural goods. A rise in production for export of garden crops, vegetables and fruits has been insufﬁcient to counter the downward trend on the production of basic items, bringing with it distribution shocks, ups and downs in the labor market and a rise in migratory pressures.15 Based on the nature of the phenomenon and the demographic tendencies of the Mexican population, Tuirán (1998) has drawn up ﬁve alternative scenarios, based on relevant variables (GDP, non-agricultural salary structures between the United States and Mexico, the rates of open unemployment and per capita annual fund remittances). He has simulated changes over a 34 year period, assuming two basic conditions: those that alleviate migratory pressures and those that aggravate them. He considers two economic growth scenarios: 1

2

An economy with accelerated growth (similar to the 1954–70 ‘stabilizing development’ period), with a positive impact on employment and real wages. Under this scenario the economy grows at an average annual rate of 5 percent between 1997 and 2015, falling to 3 percent by 2030. The wage ratio differences between the United States and Mexico would drop from 4.87 in 1996 to 2.71 in 2015 and to 1.0 in 2030. The 5.5 percent rate of open unemployment in Mexico in 1996 would be reduced to 2.6 percent in 2030, while the rate of unemployment in the United States of 5.4 percent in 1996 would rise to 9.7 percent in 2030. A slow-growth economy. This scenario assumes an erratic economic performance during the coming decades in which the GDP growth would fall from 5.09 percent in 1996 to 1.5 percent in 2000 and remain constant until 2030 (which is the annual rate of growth posted between 1982 and 1996); the rate of open unemployment of 5.5 percent (in 1996) would rise by 2030 to the highest posted during the 1970–96 period (8 percent) and the rate of unemployment in the US, of 5.4 percent in 1996, would decrease by 2030 to 4.9 percent. The wage ratio in 1996 (4.87) would increase to 10.3 in 2015 and 14.60 in 2030.

These basic scenarios are combined with two alternative hypotheses of the future behavior of remittances: 1) the per capita amount remains constant at US$586, and 2) the per capita amount rises from US$586 to US$979. Thus ﬁve scenarios emerge:

The external relationship of the Mexican economy 71 Table 3.9 Number of Mexican emigrants to the United States by year, in thousands (1996–2030, by hypothesis) Year

1996 2000 2005 2010 2015 2030

Constant rates of emigration 346 369 389 403 418 434

Hypothesis A

B

C

D

346 365 380 389 398 396

346 375 405 430 456 503

346 369 387 399 411 412

346 378 411 439 466 514

Source: Tuirán (1998), 292.

A B C D E

a growing economy with constant per capita remittances; a low-growth economy with constant per capita remittances; a high-growth economy with increasing per capita remittances; a low-growth economy with increasing per capita remittances; an economy in which the migratory rates estimated for 1996 remain constant.

The results of the exercise (Table 3.9) show that in hypotheses A and C the rates of emigration tend to drop or grow slightly, but the absolute numbers increase as a result of an ever-increasing population and the weight that demographic inertia continues to have. On the other hand, under scenarios B and D the migratory pressures increase, reﬂecting the increase in the rates and the numbers. The constant-rates hypothesis generates results half way between scenarios A, C and B, D. The differences that emerge between scenarios A–D are striking. The results indicate that even in relatively optimal economic conditions, emigration will continue its course over the short and mid-terms, spurred on by the scale that the migratory phenomenon takes on, and its cumulative effects over time. The positive effect on the migratory movement will occur over the long run if sufﬁciently high economic growth is maintained to modify the structural conditions in which the Mexican labor market functions. As a result of the persistent migratory ﬂows towards the United States, a projection can be made that the population, born in Mexico and residing in the US, will rise signiﬁcantly from 7.0 million in 1996 to 8.2 million in 2000, 11.2 million in 2010 and 16.9 million in 2030, under the constant migration hypothesis (Table 3.10). If, to the previous ﬁgures of seven million Mexicans resident in the United States, we add the annual ﬂow of migrants of 300,000 persons and

72

María Elena Cardero

Table 3.10 Population born in Mexico but residing in the United States, in thousands (1996–2030, by hypothesis) Year

1996 2000 2005 2010 2015 2030

Constant rates of migration 7,033 8,173 9,654 11,150 12,647 16,885

Hypothesis A

B

C

D

7,033 8,167 9,619 11,063 12,480 16,322

7,033 8,186 9,722 11,322 12,971 17,945

7,033 8,173 9,652 11,137 12,607 16,646

7,033 8,193 9,753 11,389 13,082 18,199

Source: Tuirán (1998), 292.

the one million Mexicans that presently work in maquiladoras with mainly US materials, components and machinery, we have a labor market in excess of eight million Mexicans, linked to a much higher number of jobs in the United States and which together make up a highly-integrated labor market, based on a productive apparatus whose structure and characteristics are determined in the US.

3.3 Conclusions During the last 15 years, the Mexican economy has undergone farreaching changes that were implemented with the sudden opening of the economy in the trade and ﬁnancial spheres. The objective was to modify the previous patterns of growth. This pattern gave way to a rising foreign trade deﬁcit generated by a protected productive structure with scarce possibilities of competing in foreign markets and highly concentrated in the US market. The indicators of change reviewed here show that the new exporting model has not had the desired results. Insofar as the impact on general economic growth is concerned, the results of the past few years have been disappointing, particularly if they are compared to what occurred in previous decades, say the 1960s and 1980s, when the country grew at an average annual rate above 6 percent, per capita GDP grew over 3 percent and per capita consumption expanded over four times. Between 1986 and 1998, the years of a wide-scale opening and economic change, GDP growth was barely over 3.03 percent per year, the per capita GDP fell by 1 percent, consumption contracted one-and-a-half times, and the economic apparatus was hardly able to create half of the jobs needed to keep up with the population increase (see Annexed Table). The export production apparatus became very important during these years and had the peculiarity of experiencing a sudden increase in its share of domestic value added from the year NAFTA started until the violent

The external relationship of the Mexican economy 73 crisis of 1995. This very sudden modiﬁcation occurred as a result of the collapse of production centered on the domestic market, as well as the fall in consumption and wages. The trade and ﬁnancial opening, aimed at changing structural determinants of growth in the Mexican economy, molded a new productive structure and accelerated the integration of the Mexican foreign sector with the US in the areas of production, trade, ﬁnancial services and the labor market. The signing of the trade agreement between the United States and Canada speeded up the already existing tendencies in trade between Mexico and its northern neighbor. It modiﬁed certain patterns and further concentrated the regional destination of Mexican trade, but failed to generate improvements in the realm of trade and production. The relationship between both economies is a peculiar form of integration because it links economies of vastly different sizes, substantially different structures and speciﬁc geographic regions in both countries. It is a profoundly asymmetrical integration which has polarized Mexico’s productive structure, and has increasingly affected overall growth. This explains the disappointing behavior of the Mexican economy and also the dis-articulation of some industrial productive structures, as well as the downturn of agricultural production. Although foreign trade became the most dynamic sector of the economy, it did not do so by widening domestic plant production, but rather by substituting domestic production for imported production, which has undoubtedly affected job creation and family income. The destruction of the productive chains formed during the import substitution phase, which supplied most of the domestic market, caused a signiﬁcant downturn of that market. Later this deepened further by the entry of portfolio investments and direct investment, bringing on a sort of ‘Mexican disease’ in which the sudden entry of hard currency created a ‘wealth effect’ which, in a context of opening, went exclusively to ﬁnance the massive purchase of imported goods. Thus the dis-articulation of the productive structure deepened, producing an effect similar to the sudden availability of natural resources, which brings on the so-called ‘Dutch disease’. A growing proportion of Mexican foreign trade – both for reasons of imported content and for its share of exports, as well as its ﬁnancial implications – is now concentrated in a small number of sectors and ﬁrms, particularly the automotive and transport vehicle sectors, that are at the center of intense intraindustrial and maquiladora trade. The maquiladoras are weakly linked to the domestic economy, this being an unusual way of verifying the Heckesher-Ohlin model of productive specialization based on factor endowment, in which Mexico places the labor force at its borders, and the ﬁrms on the other side of the border produce the remaining inputs and goods to put together the ﬁnished article.

74

María Elena Cardero

A crucial aspect of the integration of Mexico’s foreign sector with the United States is the regional labor market that operates on both sides of the border. Although these labor relations have had a long history, NAFTA negotiations between Mexico and its two partners did not touch on the topic of labor, since the United States refused to recognize the existence of this not yet legalized labor market, and treats migration only as a national issue. Recently the migratory labor phenomenon has grown, in part as a result of the slow performance of the Mexican economy, the accelerated foreign opening, the collapse of the agricultural sector, and the growing disparity of income levels in both countries. Further, this labor force meets a speciﬁc and time-varying demand of the United States labor market. Labor employed in the maquiladoras must be accounted for together with the migratory labor force that, either in a circular ﬂow or by permanent residence, legal or not, enters the United States to work. It is an integrated labor market subject today to increasingly frequent labor human rights violations. In summary, the disappearance of a productive system based on import substitution, and replacement by one in which the export sector has taken the leading role, has had severe problems as a model of growth. The disarticulation of the previous structure was not replaced by a new structure, but rather was substituted by purchases from abroad. This affected other aspects of job creation in the Mexican economy, leading to increases in the informal labor market and an increased scramble for work opportunities on the US side of the border. The economic integration that NAFTA formalized implies a total liberalization of trade in goods, services and capital, but does not address the costs and beneﬁts of labor integration and the association with an underdeveloped country. This affects the growth of the Mexican economy in its entirety, as well as its long-term ability to generate jobs and offer better living conditions for its population. In today’s environment, should this pattern of growth continue unmodiﬁed (assuming that to overcome these conditions the lesser developed country needs certain policies, resources and mechanisms, similar to what occurred with European integration), the future of Mexico and the regional relationship is pessimistic, and could be darkened by greater violence and violations of human rights along the Mexico–US border.

3.8 1.7 11.4 1.8 4.6 0.4 1.0 3.1 5.9 0.3 2.6 1.7 4.2 3.1 10.6 4.7 0.8 6.0 5.2 8.9

Total Agriculture, forestry, hunting and ﬁshing Mining Industrial manufacturing Food, beverages and tobacco Textiles, garments, leather and shoes Wood and wood products Paper, printing and publishing Chemistry, oil, and oil derivatives Non-metallic minerals Basic metals industry Machinery and equipment Other industries Construction Electricity, gas and water Commerce, restaurants and hotels Transport, communications and warehouses Financial services, insurances and real estate Community services (education, health, etc.) Inputted banking

1.0 1.3 0.5 2.0 1.2 1.5 1.8 2.8 3.0 4.0 4.5 2.9 1.6 0.1 5.2 1.1 1.8 4.1 0.3 2.6

1983–7 3.8 2.0 1.4 4.3 4.6 2.0 0.1 4.5 2.9 4.8 1.8 6.3 8.0 5.1 2.6 4.4 4.2 4.5 3.1 7.0

1988–93 3.9 2.2 2.0 3.5 3.7 1.3 0.5 3.2 1.7 4.7 2.6 5.5 6.3 6.4 2.7 4.9 5.0 4.9 2.7 8.6

1989–94

1994–6 0.7 2.8 2.6 2.7 1.7 4.1 0.7 3.2 2.8 2.3 11.2 4.9 1.4 8.3 3.5 5.9 1.3 0.1 0.7 8.0

1994–5 6.2 1.8 2.7 4.9 0.0 6.3 7.8 7.6 0.9 11.7 4.1 10.3 10.2 23.5 2.1 15.5 4.9 0.3 2.3 10.7

5.1 2.0 2.9 8.0 4.6 8.0 4.4 6.0 5.5 5.6 8.2 14.4 8.9 6.9 4.0 6.3 8.3 2.9 2.2 4.0

1995–9

Source: From 1980 to 1987, estimates on the basis of INEGI data; from 1988 to 1999 Mexican National Accounts System (SCNM). Base year 1993, INEGI.

1980–2

Concept

Table A.3.1 Growth of Gross Domestic Product, by economic sectors: Average rates of growth, base year 1993

APPENDIX

76

María Elena Cardero

Notes 1 The state’s productive, ﬁnancial and service apparatus fell from 1,090 entities beginning in 1984 to 258 in 1994. 2 The average price of a barrel of Mexico’s oil fell from US$26.90 in 1984 to US$11.90 in 1996. This fall was estimated as have cost the country US$8.5 billion, equivalent to 6.5 percent of GDP, 40 percent of total income from exports and 26 percent of government income (Cárdenas 1996). 3 In 1990 income of the public sector from the sale of enterprises amounted to 4.32 percent of the government’s income and 1.19 percent of GDP. In 1991 these percentages increased to 14.4 percent and 3.83 percent respectively. Since these were one-time income ﬂows, a fund was created to cover contingencies such as, for example, the fall in oil prices. Later it was decided to use these resources to pay back the government’s domestic debt of twenty billion pesos in 1992 and US$7.2 billion of foreign debt. 4 Between 1982 and 1987, the general minimum wage was reduced by 44.6 percent in real terms and contract wages fell by 40.5 percent. 5 In 1983 alone, the transfer was equivalent to 12.34 percent of GDP. The total of resources transferred between 1982–7 was US$53.6 billion. In spite of these transfers the foreign debt rose from US$91.5 billion in 1982 to US$107.47 billion at the end of 1987. 6 Interestingly, the signiﬁcant weight of manufacturing imports is not new, since in 1976–8 these imports were 89 percent of total imports, a percentage that rose to 90.2 percent between 1988–90 and 91.7 percent between 1994–6. What is new is the weight these imports have in total exports. The Program of Industrial Policy and Foreign Trade, implemented by the Secretaría de Comercio y Fomento Industrial, begun in May 1995, recognizes this phenomenon by stating that ‘the opening reduced the integration of several domestic productive chains, insofar as the ﬁrms that participate in these chains as clients or suppliers lost ground to competition from abroad. Relinking these productive chains now in an environment of opening to foreign competition is indispensable to encourage planning and long-term investment of domestic ﬁrms’. 7 This information comes from the United States International Trade Commission (1997). 8 According to El Financiero newspaper, March 9, 1999, p. 3, only six issues of the 30 most traded stocks in the foreign market surpassed a return of 6.04 percent. The New York Stock Exchange is where most Mexican ﬁrms trade. 9 Between 1990 and 1994 total savings amounted to 22.5 percent of GDP, of which domestic savings were 17.1 percent and foreign savings 5.4 percent. For 1995–6 total domestic savings as a share of GDP fell to 20.2 percent; the domestic portion was 19.7 percent and foreign 0.5 percent (Secretaría de Hacienda y Crédito Público 1997). 10 Certiﬁcates of Deposit, Treasury Bills. 11 Figures on capital ﬂow of the Balance of Payments do not coincide with those of Secretaría de Comercio y Fomento Industrial, whose data are signiﬁcantly lower. 12 Indirect measurement taken in the United States from the April 1990 Census and the Current Population Survey (CPS) of September 1994 give a net ﬁgure of 1,560 million, with an annual average of 340,000 migrants. 13 This survey is undertaken jointly by the National Council on Population (CONAPO), the Secretary of Labor and El Colegio de la Frontera Norte. To date it has had various stages. The ﬁrst began in March 1993 and ﬁnished 12 months later; the second took the entire year of 1995; the third began in mid1996 and ﬁnished in mid-1997. 14 The age composition of the population affects the future supply of labor.

The external relationship of the Mexican economy 77 CONAPO, in a study that focuses on the population between 15 and 44 years old, the age of most migrants, estimates that starting in 1990 the increase in this age group stabilized at around 1.3 million, in 1996 it will be 1.1 million, 897,000 in the year 2000, 667,000 in 2005 and 435,000 in 2010 (National Council on Population 1997). 15 Hinojosa and Robinson (1992) estimated that the changes in Mexican agriculture could lead to the displacement of 1.4 million rural workers, of which 800,000 could migrate internally and 600,000 could go the United States. Levy and van Wijnbergen (1991) estimate that around 40 percent of the workers who are involved in corn production (some 700,000) could be displaced and emigrate from rural areas.

Bibliography Aspe, P. (1991) ‘Estabilización macroeconómica y cambio estructural: la experiencia de México (1982–1988)’, in Bazdresch, C. et al. (eds) México: Auge, Crisis y Ajuste, Mexico: Fondo de Cultura Económica. Cárdenas, E. (1996) La Política Económica de México 1930–1994, Mexico: Fondo de Cultura Económica. Esquivel, G. (1991) ‘Comercio intraindustrial México–Estados Unidos 1981–88’, Faculty of Economics Licenciatura Thesis, Mexico: UNAM. Globerman, S. (1992) ‘North American trade liberalization and intra-industry trade’, Welwehshajtliches Archivo 128: 487–97. Gómez Luna, M.E. (1999) ‘El impacto del TLC en el empleo femenino en la industria maquiladora de exportación’, Working Paper. Hinojosa, R. and Robinson, S. (1992) ‘Diversos escenarios de la integración de Estados Unidos y México: enfoque de equilibrio general computable’, Economía Mexicana 1, January–June, 71–144. Levy, S. and van Wijnbergen, S. (1991) ‘Labor markets migration and welfare: agriculture in the Mexico–US Free Trade Agreement’, Working Paper, Washington: World Bank. Mattar, J. and Shatán, C. (1993) ‘El comercio intraindustrial e intraﬁrma México–Estados Unidos: autopartes, electrónico, y petroquímico’, Comercio Exterior 43, 2: 103–25. Moreno, L.M. (1993) ‘La internacionalización del Mercado de Valores’, El Mercado de Valores 18, December, 24–7. National Population Council (1997) La Situación Demográﬁca en México, Mexico: Secretaría de Gobernación. Salomón, A. (1996) ‘Evolución de las exportaciones en el período 1973–95’, Comercio Exterior 46, 11. Secretaría de Hacienda y Crédito Público (1997) Programa Nacional de Financiamiento del Desarrollo 1997–2000, Mexico. Tuirán, R. (1998) ‘La migración mexicana a Estados Unidos; tendencias presentes y desafíos futuros’, in Pellicer, O. and Fernández de Castro, R. México y Estados Unidos; las rutas de la cooperación, Mexico: Instituto Matías Romero-ITAM. United States International Trade Commission (1997) ‘Production sharing: use of US components and materials in foreign assembly operations, 1993–96’, US: Imports Under the Production-Sharing Provisions of Harmonized Tariff Schedule Chapter 98, December. Vázquez Tercero, H. (1995) ‘Medición del ﬂujo de divisas de la balanza comercial de México’, Comercio Exterior 45, 8.

4

Exchange rate and competitiveness Pablo Ruiz-Nápoles

The opening of the Mexican economy in the 1980s and the signing and operation of the North American Free Trade Agreement (NAFTA) in the 1990s has turned into economic integration. In this close economic relationship, a very important issue for the Mexican economy is the current account balance of the balance of payments, which has almost always been negative since the late 1940s; the main determinant of the current account deﬁcit being the trade deﬁcit. Moreover, economic crises in the recent past in Mexico (1976–7, 1981–2 and 1994–5), have all been caused by foreign exchange crises, expressed as foreign exchange shortages brought about by an unmanageable current account deﬁcit and massive capital ﬂight. All through the recent economic history of Mexico the issue of competitiveness has been fundamental to explain the economy’s ups and downs and the limits to its growth. In the context of a free-trade world and the integration to the US economy, this question becomes crucial for the current and future situation of the Mexican economy. Competitiveness is usually associated with comparative advantages in trade, and country’s real exchange rate, and its regime and policies. In this chapter we analyze the measurement and determining factors of competitiveness for the Mexican and US economies. The main proposition is that the real exchange rate of the Mexican peso with respect to the US dollar is determined, in the long run, by the relative unit labor costs of the two countries, and this, in turn, has important implications for policy.

4.1 Theoretical approaches to the exchange rate determinants Conventional theory relies on the exchange rate as the adjusting mechanism to achieve and maintain some degree of competitiveness. In a general equilibrium neoclassical framework, competitiveness between countries is determined by the comparative cost principle; and the real exchange rate is determined by the Purchasing Power Parity (PPP) doctrine. Thus, any country would always ﬁnd at least one industry in which it is competitive.

Exchange rate and competitiveness 79 If the exchange rate is adequately managed to achieve and maintain such competitiveness, foreign trade will tend to be balanced (Friedman 1953). Moreover, the adjusting mechanism is assumed to be automatic if the exchange rate is absolutely ﬂexible and free trade prevails. According to the PPP doctrine, a currency becomes overvalued when the domestic inﬂation rate exceeds the rate of inﬂation in its trading partners and this, presumably, is what produces a loss of competitiveness and consequently a trade deﬁcit in the balance of payments. Thus, to correct the deﬁcit, the nominal exchange rate should be adjusted to the real exchange rate, either directly by the authorities in a ﬁxed exchange rate regime, or through the market mechanism in a ﬂexible exchange rate regime (Krugman 1978, Krueger 1983, Dornbusch 1987, 1988). If the theory is correct, then, in the long run, the nominal exchange rate should freely move around the real exchange rate as deﬁned by PPP. Most empirical models test to what extent this theory is true; that is, they test the market efﬁciency in adjusting from one exchange rate level to the other, under assumed free market and competition (Levitch 1985). However, the PPP theory has some important ﬂaws. On the one hand, it involves the use of price indexes (Dornbusch 1988); which, while expressing the effects of supply and demand conditions in general, do not necessarily show the degree of competitiveness of the economy, but only the general price level. The cost structure would in this case be more revealing (Ofﬁcer 1975, 1980). On the other hand, the evidence shows that, in the 1970s and 1980s, depreciation policies have not signiﬁcantly corrected current account deﬁcits in most economies. Instead, they have had a very strong impact on domestic inﬂation rates (Thygesen 1979, Brailovsky 1981, De Grauwe et al. 1985). In the case of Mexico, studies of the trade pattern in manufacturing show that some of the assumptions of comparative costs theory and perfect competition do not apply in the manufacturing sector (Boatler 1974). Inter-industry trade is declining while intra-industry trade is growing and economies of scale are very important in explaining competitive advantages (Casar 1989). One modern alternative to comparative cost theory is the consideration of absolute advantages, based on technology differences, as major determinants of trade ﬂows and competitiveness. According to this approach, known as the ‘technology gap’ argument, the international competitiveness of a country, or industry, is primarily based on its absolute advantages in terms of product technology and labor productivity (Dosi et al. 1990, Cimoli 1994). With regard to exchange rate determination, since the 1970s there are theories that emphasize unit labor costs as determinants of the real exchange rate. These are essentially of two types: one within the neoclassical tradition (Ofﬁcer 1975), and the other based on Ricardo’s theory of relative prices (Aglietta and Oudiz 1984, Shaikh 1991, 1995). In Aglietta’s model, the real exchange rate, deﬁned as the ‘nominal

80

Pablo Ruiz-Nápoles

exchange rate adjusted unit labor costs for traded goods’, is estimated by econometric techniques, using data of various countries. The estimation shows that this real exchange rate is determined by three structural variables: the ratio between the two countries of income growth, income distribution, and national specialization (Aglietta and Oudiz 1984). Shaikh departs from a theory of international trade based on absolute advantages relating relative prices between countries to relative labor inputs costs. The real exchange rate is thus determined by three main factors: real unit labor costs, terms of trade, and relative prices. Insofar as real wages and total labor requirements are determined by local factors within each country, these in turn determine the terms of trade. As such, the terms of trade are not free to move so as to bring about automatic balance of trade adjustment. Persistent foreign trade disequilibria will, therefore, be normal, as long as competitive differences are themselves persistent (Shaikh 1991, 1995).

4.2 Relative unit labor costs and real exchange rate Based on the above propositions, we analyzed the exchange rate and relative unit labor costs of the Mexican and the US economies for the period 1970–96. 4.2.1 Unit Labor Costs Unit Labor Cost (ULC), deﬁned as the labor cost per unit of output, can be calculated in different ways. However, due to the lack of some of the required information, we calculated only the direct unit labor costs of the manufacturing sector. This sector is used for the calculation of ‘real effective exchange rates’ for various countries in the International Monetary Fund publications (IMF 1985). In this calculation it is assumed that the manufacturing sector represents the majority of tradable goods, which is justiﬁed for all the industrialized – and for some of the semi-industrialized – countries. In the case of Mexico, manufacturing trade has grown in importance and currently represents 80 percent of total exports and 90 percent of imports. Unit labor costs of manufacturing are also calculated as a competitiveness indicator (Capdevielle and Alvarez 1981). Thus unit labor costs were calculated for each country using the following equation: L W U . Q L

(1)

where: 0 U 1, U ULC, L total number of workers per year, W = total wages and salaries paid per year, and Q Gross Domestic Product per year, all measured in the corresponding currency at current prices.

Exchange rate and competitiveness 81 Table 4.1 Unit labor cost Year

Mexico

US

Mex./US

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996

0.3733 0.3670 0.3684 0.3596 0.3646 0.3715 0.3883 0.3638 0.3483 0.3486 0.3284 0.3372 0.3319 0.2600 0.2350 0.2295 0.2199 0.2004 0.2009 0.2239 0.2350 0.2452 0.2685 0.2748 0.2773 0.2142 0.2037

0.6346 0.6103 0.6045 0.6077 0.6277 0.5962 0.5869 0.5765 0.5804 0.5869 0.6099 0.5950 0.5943 0.5813 0.5712 0.5839 0.5775 0.5592 0.5457 0.5408 0.5441 0.5471 0.5451 0.5330 0.5143 0.5083 0.5157

0.5882 0.6013 0.6095 0.5918 0.5808 0.6232 0.6615 0.6310 0.6002 0.5939 0.5385 0.5667 0.5585 0.4473 0.4114 0.3931 0.3808 0.3584 0.3682 0.4141 0.4319 0.4481 0.4926 0.5156 0.5392 0.4214 0.3951

Source: Author’s calculations based on data from INEGI, Mexico and US Department of Commerce.

The results are shown in Table 4.1. It must be noted that each ULC is a proportion calculated in each country’s currency. The ratio of ULC of Mexico to the US was also calculated. This will be called the Relative Unit Labor Costs (RULC). In the case of the US, the unit labor costs were observed to decline almost continuously from 1980 to 1995, then increasing in 1996. Meanwhile for Mexico, ULC declined from 1976 to 1987 and then rose up to 1993, they declined sharply again in 1995 and 1996. Relative ULC went down constantly from 1976 to 1987 and then up from 1988 to 1994. In 1995 and 1996 they declined again. 4.2.2 Real exchange rate The real exchange rate of the Mexican peso to the US dollar is deﬁned as the number of pesos per dollar, deﬂated by the ratio of consumer prices in Mexico to consumer prices in the US, each measured in its

82

Pablo Ruiz-Nápoles

corresponding currency. It is an annual average rate, calculated according to the equation: er e (P/P*)

(2)

where: er real exchange rate pesos per dollar, e nominal exchange rate pesos per dollar, P Mexico’s Consumer Price Index, in old pesos, and P* US Consumer Price Index, in dollars. The results of this calculation are shown in Table 4.2. All variables are measured in index numbers base 1970 100. Measured in this way, the real exchange rate index shows an overvaluation of the peso in respect to the dollar, when its value is below 100; that is Table 4.2 Nominal and real exchange rates Year

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997

Exchange Rate

Consumer Price Indexes

Real exchange rate

Old Pesos

Index

Mexico

US

Index

12.5 12.5 12.5 12.5 12.5 12.5 15.4 22.5 22.7 22.8 22.9 24.5 56.4 120.0 167.8 256.9 611.3 1,366.7 2,250.2 2,453.1 2,810.2 3,012.4 3,091.5 3,111.3 3,382.4 6,396.5 7,579.9 7,908.7

100.0 100.0 100.0 100.0 100.0 100.0 123.5 180.6 182.1 182.4 183.6 196.1 451.2 960.7 1,342.6 2,055.6 4,890.8 10,933.8 18,002.2 19,625.3 22,481.9 24,099.4 24,732.6 24,890.8 27,059.6 51,172.3 60,639.5 63,269.5

100.0 105.5 110.8 124.1 153.6 176.8 204.8 264.0 310.2 366.7 463.2 592.9 941.9 1,901.5 3,146.2 4,962.9 9,242.7 21,427.1 45,888.4 55,069.6 69,746.8 85,553.0 98,820.5 108,457.2 116,011.9 156,616.0 210,457.1 253,865.3

100.0 104.3 107.7 114.4 127.0 138.6 146.6 156.1 168.0 186.9 212.3 234.2 248.6 256.6 267.6 277.1 282.3 292.7 304.6 319.3 336.6 350.8 361.5 372.1 381.8 392.5 404.0 413.5

100.0 98.9 97.2 92.2 82.7 78.4 88.4 106.8 98.6 93.0 84.1 77.5 119.1 129.6 114.2 114.8 149.4 149.4 119.5 113.8 108.5 98.8 90.5 85.4 89.1 128.3 116.4 103.1

Source: Author’s calculations based on data from Banco de Mexico, S.A. and the US Department of Commerce.

Pablo Ruiz-Nápoles 83 the case in the periods 1971–6, 1978–81 and 1991–4. When the value is above 100, the peso is undervalued and that corresponds to the years 1977, 1982–90 and 1995–7. See Figures 4.1 and 4.2. This illustrates the argument of section 4.1 above. Now, for expositional purposes, we deﬁne E 1 / er (P/e P*)

(3)

where: E is the real exchange rate of the Mexican peso to the US dollar, expressed in dollars. The purpose of using this measure of the real exchange rate is to observe more clearly the relation between the real exchange rate and relative unit labor cost. The relative unit labor costs and real exchange rate, both expressed as index numbers, base 1977 100, called RULCI and RERI respectively, are shown in Table 4.3 and in Figure 4.3. The base year of 1977 was chosen Table 4.3 Real exchange rate and relative Unit Labor Costs Index numbers Year

RERI Base 1977100

RULCI

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996

106.80 108.01 109.83 115.88 129.15 136.25 120.76 100.00 108.27 114.83 126.94 137.85 89.69 82.39 93.53 93.06 71.48 71.51 89.37 93.85 98.44 108.07 118.06 125.07 119.92 83.27 91.74

93.21 95.29 96.58 93.78 92.04 98.75 104.84 100.00 95.12 94.11 85.34 89.81 88.50 70.88 65.19 62.30 60.35 56.79 58.35 65.62 68.44 71.01 78.06 81.71 85.45 66.79 62.61

Source: Author’s calculations.

68

69

70

71

72

73

75

Nominal

74

Figure 4.1 Nominal and Real Exchange Rates 1968–82.

5

15

25

35

45

55

65

Pesos/Dlls

76

Real

77

78

79

80

81

82

83

84

85

86

87

88

Figure 4.2 Nominal and Real Exchange Rates 1983–95.

0

1,000

2,000

3,000

4,000

5,000

6,000

Old Pesos/Dlls 7,000

Nominal

89

90

Real

91

92

93

94

95

RERI

RULCI

70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95

Figure 4.3 Real Exchange Rate and Relative Unit Labor costs.

0

20

40

60

80

100

120

140

0

10

20

30

40

50

60

70

80

90

100

110

Exchange rate and competitiveness 87 because it was the ﬁrst time in 23 years that the exchange rate was set free, so its level was determined by market forces. Also, in that year the foreign trade ﬂows were balanced. In order to avoid possible confusion, note that, in this case the Real Exchange Rate Index (RERI) shows a real appreciation of the peso when it is above 100 and a real depreciation when it is below 100. We see that in the periods 1970–6, 1978–81 and 1991–4 the peso was overvalued in real terms according to this measure, while it was undervalued in real terms in the periods 1982–90 and 1995–6. The RERI reached a peak in 1975 as the peso became overvalued before the 1976 depreciation, then it went down in 1976 and 1977. In the next period, as the nominal rate was stabilized, the real index went up again to reach its highest level in 1981 before the big depreciation of 1982. It goes down again as a result of a continuous depreciation policy, from 1983 to 1987. From 1988 on, the real exchange rate index rose as the peso was appreciated in real terms. The undervaluation margin, estimated as almost 30 percent in 1987, disappeared between 1990 and 1991, and by 1993 the overvaluation was of about 25 percent. The small depreciation of the peso at the beginning of 1994 and the big depreciation at the end of that year made the average go down again. In 1995 and 1996, the index was below 100, which means the peso was undervalued in real terms, but the tendency showed in the ﬁgure indicates it was moving upwards again. 4.2.3 Long-run trend and short-run movements The Relative Unit Labor Cost Index (RULCI) line in Figure 4.3 shows a trend very similar to the Real Exchange Rate Index (RERI) but smoother. Both lines together seem to support the proposition that the trend of the real exchange rate is associated with unit labor costs. However, when modelling real exchange rates, authors introduce more than one variable, in order to take into account the effect of imports and exports on the real exchange rate (Aglietta and Oudiz 1984, Krugman 1992, ch. 3). National income has two different but equally important effects on the real exchange rate. On the one hand, the level of income determines the level of imports. Imports, in turn, are the main determinant of foreign exchange’s demand and, therefore, they have an important inﬂuence on the exchange rate. This implies a negative relation between the level of income and the real exchange rate, according to the way E is measured in equation (3). When income rises, so do imports, this puts an upward pressure in the nominal exchange rate e in equation (3), which will result in a decrease of E. This means a real depreciation of the currency, in the short run, unless other factors, like a net capital inﬂow and/or a policy measure, hold the nominal exchange rate e constant.

88

Pablo Ruiz-Nápoles

On the other hand, since the national income (or gross domestic product) includes exports, when the economy grows and exports rise, the expected relationship between income and real exchange rate is a positive one. This increase in exports implies an increase in the supply of foreign exchange that puts a downward pressure on e in equation (3). In the case we are analyzing, we have included the level of real national income as a variable that has had these two important effects on the real exchange rate. But considering the various exchange rate policies followed, the dynamics of exports and the inﬂow of short-term capital, during the period, we expect the relationship to be positive. Thus, the real exchange rate is determined by relative unit labor costs and real national income. This proposition is formulated in the following equation: E f (U, Y)

(4)

where, E real exchange rate index dollars/pesos, U relative unit labor costs index, Mex/US, Y GDP of Mexico. Equation (4) was estimated by econometric techniques, and the procedure and results are presented in the Appendix. The estimation shows that relative unit labor costs determine the trend for the real exchange rate. The main determining variable in the long run is relative unit labor costs while the other variable determines the variations around it. It was also found that the relation is not linear. Relative unit labor costs act as center of gravity of the real exchange rate, in the long run. 4.2.4 Some implications for policy A nominal depreciation of the currency by itself does not necessarily imply a real depreciation, since a depreciation produces a short-run domestic inﬂation which offsets its effect on the real rate. In that case a real depreciation can only be produced by a reduction of unit labor costs. These, in turn, have two determinants: wages and productivity. Thus, in order to reduce unit labor costs, either the real wage rate is to be reduced or productivity raised, or both. Moreover, this has to be done at a faster rate than the country’s main competitor(s) in the world market, so as to improve the country’s competitive position. A nominal depreciation of the currency, together with a recessionary policy, may in fact produce a reduction of real wages and, thereby, reduce unit labor costs in the short run, assuming constant productivity. This can be accomplished either by directly reducing the wage rate or by reducing the level of employment, to a given output level (Corden 1977, Dornbusch 1980). However, a lasting reduction in unit labor costs can only be produced by the use of more productive techniques of production. This will have an

Exchange rate and competitiveness 89 effect in the input–output matrix, and in the output–labor ratio. If we assume that wages do not rise as fast as the productivity of labor, unit labor costs will decline when productivity increases. This would imply a real depreciation of the currency, even with a constant nominal exchange rate. As a consequence of this real depreciation, the trade balance will tend to improve, as will the competitive position of the country.

4.3 ULC, wages and productivity ULC in Mexico’s manufacturing sector are, on average, about half of the corresponding ULC in the US, (see Table 4.1), From the mid-1970s until 1987, relative ULC declined almost constantly, while from 1988 to 1994 the tendency was reversed and the relative ULCs were rising. This increasing tendency was due to the combined effect of decreasing ULC in the US and increasing ULC in Mexico. From equation (1) we decompose the ULC formula into two elements: W/L, the wage rate and L/Q, the inverse of the output–labor ratio, that is, the productivity of labor. As mentioned above, for the ULC to decline, either the wage rate declines with constant productivity or the productivity of labor rises with a constant wage rate, or both. In this case, the decline of relative ULC during the 1980s was mostly the result of the strong decline of Mexico’s ULC which was mainly due to one particular policy: the systematic reduction of the real wage in Mexico since 1981. The real wage rate in manufacturing declined in Mexico from 1981 to 1988, while in the US the real wage rate was rising from 1981 to 1986, as can be seen in Table 4.4 and Figure 4.4. We can distinguish two periods: from 1976 to 1982, in which Mexico’s real wages were moving higher relative to US real wages; and from 1983 on, when they moved lower. However, in the period from 1987 to 1994, the gap between the two was reduced since the real wage rate in Mexico recovered in this period mainly due to a reduction in the rate of inﬂation. The productivity of labor is measured by: LP Q/L

(5)

where: Q GDP in manufacturing at constant prices in each country’s currency, and L number of workers in manufacturing. Both variables are normalized as index numbers base 1977 100. The results are shown in Table 4.5 and Figure 4.5. This measure of labor productivity shows an important increase in Mexico’s manufacturing sector beginning in 1983, which helped to bring about the ULC decline. However, this increase was slower than the increase in labor productivity in the US, especially after 1986, as can be observed in Figure 4.5. Thus, as soon as wages started to recover, in 1988, the tendency was reversed and Mexican ULC rose very fast.

90

Pablo Ruiz-Nápoles

Table 4.4 Real Wage Rate indexes (Base 1977100) Year

Mexico

USA

Mexico/USA

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996

76.84 78.28 82.65 83.39 86.63 90.96 98.96 100.00 97.94 100.12 96.98 99.60 96.56 74.87 70.30 70.54 64.99 64.40 64.07 68.36 70.78 74.55 80.88 83.48 86.80 74.75 68.63

94.41 95.81 99.03 98.91 96.06 96.26 98.62 100.00 100.15 98.13 95.23 94.44 95.04 97.67 98.99 101.16 103.52 102.82 103.33 101.68 100.25 100.22 102.25 101.97 103.01 103.50 104.24

81.39 81.70 83.46 84.31 90.18 94.49 100.35 100.00 97.79 102.02 101.84 105.46 101.60 76.65 71.02 69.73 62.78 62.63 62.01 67.23 70.60 74.39 79.10 81.86 84.26 72.23 65.83

Source: Author’s own calculations based on data from US Bureau of Labor Statistics and INEGI.

4.4 Conclusions When we analyze the period between the two last foreign exchange crises, we ﬁnd that between 1982 and 1987 relative ULC declined, and this meant a real depreciation of the Mexican peso making the economy more competitive in those years. This decline was essentially based on a relative wage decrease. Between 1988 and 1994, this tendency was reversed by a relatively higher increase in productivity in the US, so that the relative ULC rose, eliminating the undervaluation margin by 1991. The Mexican government tried to improve the level of competitiveness of the manufacturing sector, based mainly on the reduction of real wages through a devaluation–inﬂation process. It worked for half a decade. But

76

74

72

70

Figure 4.4 Real wage rates indexes.

50

55

60

65

70

75

80

85

90

95

100

105

82

80 Mexico

88

86

84

United States

96

94

92

90

78

75

80

85

90

95

100

105

110

92

Pablo Ruiz-Nápoles

Table 4.5 Productivity indexes (Base 1977100) Year

Mexico

USA

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996

82.95 83.12 87.36 90.63 91.81 95.33 97.42 100.00 105.54 108.31 107.54 109.28 108.50 107.68 110.80 113.88 109.98 112.12 115.61 120.90 127.33 133.05 138.94 145.14 149.44 149.60 152.40

84.54 88.40 90.55 91.65 90.57 95.72 96.27 100.00 100.49 101.49 100.07 104.98 106.28 114.78 120.81 125.19 126.07 135.25 141.50 140.53 140.58 140.81 145.43 150.55 160.94 170.00 173.14

Source: Author’s own calculations based on data from US, BEA and Mexico, INEGI.

since the labor productivity in manufacturing did not rise as fast as in the US, once the rate of inﬂation was reduced using the nominal exchange rate as an anchor, the tendency was reversed and Mexico’s competitive edge was gradually eliminated. Under conditions of unrestricted opening of the economy and maintaining the nominal exchange rate within narrow margins, the trade balance showed an unprecedented deﬁcit, which by 1994 was out of control. During the 1994 foreign exchange crisis, some economists argued that it was caused by the reluctance of the Mexican government to depreciate the peso in the amount required by the accumulated rate of the relative price ratio Mex/US, that is, according to the PPP doctrine. From our approach, this opinion was not quite correct. In the ﬁrst place, a nominal depreciation always produces inﬂation, which very soon eliminates the competitive price margin initially obtained by the depreciation, so at the end the real exchange rate remains

76

74

70

Figure 4.5 Productivity indexes.

50

60

70

80

90

100

110

120

130

140

150

160

170

180

72

Base 1977=100

82

80

Mexico

88

86

84

United States

96

94

92

90

78

94

Pablo Ruiz-Nápoles

unchanged, in the average. This can easily be seen by recalling equation (2) er e (P/P*), in which the real exchange rate is by deﬁnition equal to the nominal exchange rate divided by the price ratio. So that when e is raised, er goes up too, unless the price ratio changes. However, experience shows that almost always when there is a depreciation of the currency, domestic prices rise as a consequence of such depreciation. Assuming that foreign prices remain unchanged in the short run, the effect of the nominal depreciation on the real exchange rate will be partially or totally offset by domestic inﬂation unless, of course from our approach, the relative unit labor cost is reduced in the process. Secondly, depreciation does not eliminate the cause of the trade deﬁcit, which is always the basic cause of a foreign exchange shortage. This was, in effect, overlooked by policy makers, based on some ‘modern approach’ (Corden 1993). What, in fact, reduces the deﬁcit in the short run is the recession that follows the depreciation as a ‘complementary’ policy, because it reduces imports as well as some domestic demand for exportable goods. However, when the economy emerges from the recession, the underlying tendencies of imports, exports and the trade balance reappear. It is, of course, much worse if the nominal exchange rate is artiﬁcially maintained, away from the relative price ratio, because then some speculative capital movements also come into play. From the point of view of the unit labor costs approach, a nominal depreciation of the currency has only a temporary and limited effect on competitiveness. A lasting improvement of competitiveness and, therefore, of the trade balance requires the use of more productive and cost reducing techniques. In the case of Mexico, being in competition with the US for the whole North American market, differences in the level and the rate of change of productivity in manufacturing with respect to the US can be explained by the existing technological differences between the two countries. These differences persist whatever the exchange rate policy. They are always expressed in lower costs, and wage differentials cannot offset them completely and permanently. The opening of the Mexican economy made these differences show up more clearly with some rather dramatic results. The different exchange rate policies applied since 1983 have failed in improving Mexico’s economic performance because the policy makers’ target has always been to reduce relative wages, and this process has shown its limits. The other much more important process of increasing the relative productivity of labor has not been successful. Mexico’s experience has been unique in the sense that an export-led growth strategy was never before tried in a country that was in the middle of an economic crisis and starting a process of trade opening, deregulation and privatization (Sachs 1988). Mexico was exactly in that situation and it imposed serious limits to such a strategy.

Exchange rate and competitiveness 95 APPENDIX

A.4.1 Econometric estimation In order to specify equation (4) for estimation by econometric techniques, we ﬁrst get the following equation: LEt 1 LEt2 2 LUt 3 LYt ut

(A.1)

where: LEt logarithm of the Real Exchange Rate Index (pesos per dollar), at time t, LUt logarithm of the Relative Unit Labor Costs Index of manufacturing, Mex/US, LYt logarithm of the GDP of Mexico, in constant pesos. In the decision whether to use ﬁrst differences or levels for this equation, we adopted the criterion that since no serial correlation is present in the estimated equation there was no need to specify the equation in ﬁrst differences. A.4.1.1 Unit root As part of a cointegration anaysis, we applied unit root tests to all variables involved, The results obtained by Econometric Views (Lilien et al. 1995: v.2) are shown in Table A.4.1; they indicate that all variables are I(1), that is, they all are non-stationary (Dickey and Fuller 1979, Dickey et al. 1984). However, this is not a problem if the series are cointegrated, so

Table A.4.1 Unit root tests sample 1973–95 (critical values: 1% 2.666, 5% 1.957, 10% 1.624) Variable

t-ADF

Lag

Prob.

P–P

Lag

LE LE LU LU LY LY ∆LE ∆LE ∆LU ∆LU ∆LY ∆LY ∆∆LY

0.45394 0.42967 0.85066 0.91207 0.98820 3.51070 4.4721* 3.5297* 2.6009* 2.8200* 1.8084*** 2.0896** 4.6120*

1 0 1 0 1 0 1 0 1 0 1 0 1

0.6545 0.6716 0.4046 0.3716 0.3343 0.0020 0.0002 0.0019 0.0167 0.0100 0.0849 0.0484 0.0002

0.44483 0.41555 0.83082 0.84504 2.80488 3.02684 3.35389* 3.58548* 2.77242* 2.85893* 1.95589*** 2.03442** 5.14348*

2 1 2 1 2 1 2 1 2 1 2 1 2

(*), (**), (***) rejection of null hypothesis at 1%, 5% and 10% level of signiﬁcance; ADF Augmented Dickey–Fuller test; P–P Phillips–Perron test.

96

Pablo Ruiz-Nápoles

that the residuals in the estimated equation are I(0), that is, the residuals are stationary (Cuthberson et al. 1992). A.4.1.2 Ordinary least squares In equation (A.1) all original variables are expressed in natural logarithms, and there is no intercept. The results obtained by Pc Give 8,0 (Doornik and Hendry 1996) are shown in Table A.4.2, and the equation tests results are shown in Table A.4.3. According to the results obtained, the coefﬁcients are all signiﬁcant; the R squared is high. The Durbin Watson statistic is useless in this case since we are including a lagged explanatory variable. The use of logarithms for the dependent and the two explanatory variables, imply that the parameters for LU and for LY are the elasticities of E with respect to U and Y respectively. The signs and values of the parameters are as expected. The tests’ results showed no problems of serial correlation, heteroscedasticity, normality, stability or functional form. The Augmented Dickey–Fuller test for cointegration shows that the residuals of the equation are I(0), that is stationary. By substituting the parameters obtained into equation (A.1) we get the equation: LEt 0.544478 LEt2 1.186404 LUt 0.130593 LYt

(A.2)

Table A.4.2 Modelling LE by OLS sample (1973 to 1995) Variable

Coefﬁcient

Std. Error

t-value

t-prob

Part. R2

LEt2 LU LY

0.544478 1.186404 0.130593

0.165665 0.152444 0.031226

3.286617 7.782560 4.182224

0.0037 0.0000 0.0005

0.3507 0.7518 0.4665

R2 0.999564; Standard error of reg. 0.104053; Durbin Watson 1.67; sum of squared Res. 0.216541 for three variables and 23 observations.

Table A.4.3 Equation tests results Test for Error autocorrelation Serial correlation Normality Jarque-Bera Heteroscedastic errors Heteroscedastic errors Stability Unit root for resid. Stability

LM(1) ARCH(1) χ2 White Xi2 Xi*Xj RESET(1) t-ADF (1) CUSUM

F(1,19) F(1.18) – F(6.13) F(9.10) F(1.19) (1 % 2.682) CUSUM Q

Statistic

P value

0.5336 0.3133 1.4823 1.2237 1.0500 1.1419 3.4613 Passed

[0.474] [0.582] [0.476] [0.355] [0.466] [0.298] [0.002]

Exchange rate and competitiveness 97 Then, by making LEt LEt2 and substituting into equation (A.2) we get the Solved Static Long Run equation: LE 0.768159 LU 0.084555 LY (SE) (0.06096) (0.01738)

(A.3)

A.4.1.3 Cointegration analysis Despite the various statistical tests regarding the validity of the OLS estimation, recent developments in econometrics put new emphasis on some problems, which OLS techniques cannot solve; an important one is the possibility of spurious correlation between non-stationary variables (Charemza and Deadman 1997). This problem can be dealt with by means of a group cointegration analysis following the Johansen procedure (Johansen 1991). This method is applied to the group of time series corresponding to the variables involved in the equation estimated. In this case, since the three variables involved in equation (A.1) are all non-stationary, the cointegrating analysis was considered pertinent. In order to carry out this analysis, a vector autoregressive (VAR) model for the three variables involved is deﬁned as: LEt 0,1 iLEti iLUti jLYti u1t LUt 0,2 iLEti iLUti jLYti u2t LYt 0,3 iLEti jLUti iLYti u3t i 1, 2

(A.4)

This VAR(2) is then estimated following the Johansen procedure. The results obtained by PcFiml 9,0 (Doornik and Hendry 1996) are shown in the Table A.4.4. The reported Likelihood Ratios (LR) statistics (in particular the ‘trace’ LR) compared to the critical values at 5 percent level of signiﬁcance indicate that both hypotheses of zero and at most two, cointegrating vectors should be rejected, but not the existence of one cointegrating vector (the LR statistic is not in the rejection region). The conclusion is that there is precisely one cointegrating vector in the estimated model. Therefore the rank of the () matrix is one. The ﬁrst element of the ﬁrst row in the standardized matrix corresponds to the dependent variable so this gives, as a result, the following long run relationship: LE 1.1071 LU 0.15041 LY

(A.5)

The equation is meaningful from an economic point of view. The signs and values of the parameters are correct. The parameter of LU which is the elasticity of E associated with U is positive and its value is around one as expected, and the parameter of LY which is the elasticity of E associated

98

Pablo Ruiz-Nápoles

Table A.4.4 Cointegration analysis 1973–95 Eigenvalue m

Loglik

For rank

0.576091 0.263693 0.214203

200.299 210.169 213.689 216.461

0 1 2 3

Ho:rank p p 0 p 1 p 2

T log (1mp) 19.74 7.04 5.544*

using Tnm 14.59 5.204 4.098*

95% 21.0 14.1 3.8

Ho:rank p p 0 p 1 p 2

T Σlog (1mi) 32.32* 12.58 5.544*

using Tnm 23.89 9.302 4.098*

95% 29.7 15.4 3.8

Standardized β eigenvectors LE LU 1.00000 1.10710 9.14800 1.00000 0.29886 1.38310

LY 0.15041 4.20850 1.00000

Standardized α coefﬁcients LE 1.10590 LU 0.03216 LY 0.30095

0.020811 0.0063529 0.0049424

0.26382 0.15834 0.05370

Long run matrix Po α* β, rank 3 LE LE 1.21750 LU 0.13760 LY 0.23969

LU 0.88034 0.26095 0.25398

LY 0.00990 0.18991 0.02923

Number of lags used in the analysis: 2; variables entered unrestricted: constant. * rejection of null hypothesis at 5% level.

with Y is positive and very low, also as expected. The statistical tests applied to the estimated model reported no problems of autocorrelation, or heteroscedasticity. Also, the normality test is not rejected in the ﬁrst and third equations (LE and LY); in the case of the LU equation, the Jarque-Bera statistic is rejected. The ﬁrst equation in the system that gives us (A.5) shows satisfactory tests results. A.4.1.4 Exogeneity In equation (A.1) and its corresponding estimated long run equations (A.3) and (A.5), LU and LY are utilized to estimate LE. In an econometrics deﬁnition of exogenous variables, this means that past and present

Exchange rate and competitiveness 99 values of LU and LY are utilized to estimate and forecast LE, and it is assumed that LE cannot contain ‘information’ to estimate or forecast LU and LY. These variables are the logarithms of the real exchange rate, E, unit labor costs, U, and real product, Y. We are implying not only that unit labor costs and real product explain the real exchange rate but that real exchange rate cannot explain unit labor costs or real product. That is, we are assuming that LU and LY are the exogenous variables, while LE is the endogenous variable in the model. This division is not assumed in cointegration analysis so it must be tested (Charemza and Deadman 1997). In particular, what seems important, in this case, is the exogeneity of LU. The appropriate test, in this case, is the test for weak exogeneity of variable LE in the VAR. This is done, following Johansen (1992), by calculating the maximum likelihood ratio between the unrestricted VAR, involving all variables, and one in which it is assumed that one of them is exogenous, that is, a restricted VAR. This ratio is: p (1l1) T 冱 log x2 (p,b) (1l2) i

冤

冥

where: T number of observations, p number of cointegrating vectors, 1 maximum eigenvalue of the restricted VAR, 2 maximum eigenvalue of the unrestricted VAR, and b number of parameters. The distribution follows a 2 distribution (Johansen 1992). The ratio was estimated with the results of the unrestricted VAR reported in Table A.4.4. and the estimation of a restricted VAR in which, it is assumed, one cointegrating vector and that ’s associated with the equation of LU are zeros. The reported statistic for one cointegrating vector was 2 (1) 0.021462 [0.8830]. This indicates that the null hypothesis cannot be rejected; that is, weak exogeneity cannot be rejected. This means that LU is exogenous to the system and so it can be utilized to estimate LE.

A.4.2 Data sources For Mexico: GDP and labor data, Instituto Nacional de Geografía, Estadística e Informática (INEGI), Cuentas Nacionales; exchange rate and price data, Banco de México, S,A, Indicadores Económicos and Informe Anual. Table A.4.5 VAR tests results Test Error autocorrelation Heteroscedasticity Normality Jarque-Bera Heteroscedastic errors

LM (12) F(2,14) ARCH (1) F(1,14) χ2 (2) White F (12,3)

LE

LU

LY

0.54807[0.0590] 0.25444[0.6218] 1.34260[0.5111] 1.11050[0.5306]

0.06747[0.9351] 0.03018[0.8646] 7.20940[0.0272] 0.39838[0.8918]

0.85777[0.4452] 0.42350[0.5257] 1.46090[0.4817] 0.58401[0.7826]

100

Pablo Ruiz-Nápoles

For the US: GDP and labor data, Bureau of Economic Analysis, US Department of Commerce, Survey of Current Business; price data, Bureau of Labor Statistics.

Bibliography Aglietta, M. and Oudiz, G. (1984) ‘Real exchange rates and structural transformations’, in Malgrange, P. (ed.) Contemporary Macroeconomic Modelling, Oxford: Basil Blackwell. Boatler, R.W. (1974) ‘Las predicciones de la teoría del comercio internacional y el crecimiento de las exportaciones manufacturadas de México’, El Trimestre Económico 164: 811–41. Brailovsky, V. (1981) ‘Exchange rate policies, manufactured exports and the rate of inﬂation’, Working Paper, Institute for Industrial Planning, Ministry of National Resources and Industrial Development, Mexico. Capdevielle, P. and Alvarez, D. (1981) ‘International comparisons of trends in productivity and labor costs’, Monthly Labor Review, December. Casar, J.I. (1989) Transformación en el Patrón de Especialización y Comercio Exterior del Sector Manufacturero Mexicano 1978–1987, Mexico: ILET. Charemza, W.W. and Deadman, D.F. (1997) New Directions in Econometric Practice: General to Speciﬁc Modelling, Cointegration and Vector Autoregression, Cheltenham, UK: Edward Elgar, 2nd edition. Cimoli, M. (1994) ‘Technological gaps, specialization and growth: a structuralist evolutionary view’, unpublished, Venice: University of Venice. Corden, W.M. (1977) Inﬂation Exchange Rates and the World Economy, Chicago: University of Chicago Press, 2nd edition. Corden, W.M. (1993) ‘¿Importa la cuenta corriente? El punto de vista tradicional y el moderno’, Economía Mexicana Nueva Epoca II, 1. Cuthberson, K., Hall, S.G. and Taylor, M.P. (1992) Applied Econometric Techniques, London: Philip Allan. De Grauwe, P., Janssens, M. and Leilaert, H. (1985) ‘Real exchange rate variability from 1920 to 1926 and 1973 to 1982’, Princeton Studies in International Finance 56, Princeton: Princeton University. Dickey, D.A. and Fuller, W.A. (1979) ‘Distributions of the estimators for autoregressive time series with a unit root’, Journal of the American Statistical Association 74: 427–31. Dickey, D.A., Hasza, D.P. and Fuller, W.A. (1984) ‘Testing for unit roots in seasonal time series’, Journal of the American Statistical Association 79: 355–67. Doornik, J.A. and Hendry, D.F. (1996) PcGive 8,0: An Interactive Program for Modelling Econometric Systems, London: International Thomson Publishing. Dornbusch, R. (1980) Open Economy Macroeconomics, New York: Basic Books. Dornbusch, R. (1987) ‘Exchange rate economics: 1986’, Economic Journal 97: 1–18. Dornbusch, R. (1988) ‘Purchasing Power Parity’, in Exchange Rates and Inﬂation, Cambridge, Mass.: MIT Press. Dosi, G., Pavitt, K. and Soete, L. (1990) The Economics of Technical Change and International Trade, New York: New York University Press. Friedman, M. (1953) ‘The case for ﬂexible exchange rates’, in A.E.A. Readings in International Economics, 1968, 413–17.

Exchange rate and competitiveness 101 International Monetary Fund (1985) International Financial Statistics, Supplement on Exchange Rates, Supplement Series 9, Washington. Johansen, S. (1991) ‘Estimation and hypothesis testing of cointegration vectors in Gaussian vector autoregressive models’, Econometrica 59: 1551–80. Johansen, S. (1992) ‘Testing weak exogeneity and the order of cointegration in the UK money demand data’, Journal of Policy Modeling 14, 3: 453–63. Krueger, A. (1983) Exchange Rate Determination, Cambridge: Cambridge University Press. Krugman, P.R. (1978) ‘Purchasing Power Parity and exchange rates’, Journal of International Economics 8: 397–407. Krugman, P.R. (1992) Currencies and Crises, Cambridge, Mass.: The MIT Press. Levitch, R.M. (1985) ‘Empirical studies of exchange rates: price behavior, rate determination and market efﬁciency’, in Jones, R.W. and Kenen, P. (eds) Handbook of International Economics, Elsevier, II, ch.19. Lilien, D.M. et al. (1995) Econometric Views v.2.0, Quantitative Micro Software. Ofﬁcer, L.H. (1975) ‘The Purchasing Power Parity theory of exchange rates: a review article’, International Monetary Fund Staff Papers XXIII, 1. Ofﬁcer, L.H. (1980) ‘Effective exchange rates and price ratios over the long run: a test of the Purchasing-Power-Parity theory’, Canadian Journal of Economics XIII, 2. Sachs, J.D. (1988) ‘Políticas comerciales y de tipo de cambio en programas de ajuste orientados al crecimiento’, Estudios Económicos 3, 1: 77–108. Shaikh, A. (1991) ‘Competition and exchange rates: theory and empirical evidence’, Working Papers 25, Graduate Faculty, New School for Social Research. Shaikh, A. (1995) ‘Free trade, unemployment and economic policy’, in Eatwell, J. (ed.) Global Unemployment, Loss of Jobs in the ’90s, New York: Sharpe, 59–78. Thygesen, N. (1979) ‘Exchange rates experiences and policies of small countries: some European examples of the 1970’s’, Essays in International Finance 136 1–30, May, Princeton University.

5

Vertical integration and comparative advantages Fidel Aroche Reyes

The US and Mexico have been increasing their trading volumes for decades. The commercial relationships between Canada and Mexico have been rather small. The exchange of goods, services and factors between these countries is a natural result of vicinity as well as differentials in the range of goods each economy produces, among other determinants (Frankel et al. 1995). Some authors also include different resource endowments (Estrada and Kushida 1983). However, such arguments do not sufﬁce to explain the size of trade between Mexico and Canada; institutional practices might also be relevant. However, if the theory of international trade would include transport costs as a determinant of trade between two countries, the distance between Canada and the US and between the US and Mexico may become an important factor in the picture (Krugman 1991). In a sense, trade volumes as well as the directions of ﬂow imply that these countries are integrated to a certain degree, since they supply and demand goods in each other’s markets. Those ﬂows might be of importance for the circular ﬂow of the economy (Pearce and Turner 1990). The theory of international trade postulates that a country enjoys comparative advantages in the production of those goods that require intensively factors which are abundant within the country (Chacholiadis 1992). Therefore, international trade can be explained by differentials in the factor endowments between different countries if consumers show comparable preferences. Accordingly, foreign trade will increase a nation’s welfare if that enhances specialization. The simplest version of the Hecksher-Ohlin (HO) theorem applies to a model of two countries, two commodities and two factors. That theorem has been discussed theoretically and tested empirically in numerous ways (Deardorff 1984), perhaps the starting point has been the paper by Leontief (1953) where the well-known Leontief paradox was ﬁrst explained. Nevertheless, despite the abundant literature on the subject, the implications of the HO theorem in a world with many countries, many goods and many factors remains a point of debate even at a textbook level (Chacholiadis 1992).

Vertical integration and comparative advantages 103 The factor content version of the HO theorem follows Leontief’s methodology within the input–output (IO) model; this version states that a country tends to export the services of its abundant factors embodied as factor contents in the goods it sells abroad, (Vanek 1968, Leamer 1980). This version of the theorem, called HOV, has also been tested repeatedly (Deardorff 1984). Unfortunately, it is still unclear whether the multi-factor content theorem of international trade is empirically valid (Maskus 1985). Moreover, the prevalence of intra-industry trade, especially between countries of similar levels of development, is often cited as an empirical reason to reject the validity of the orthodox theory (Helpman and Krugman 1989). Davis (1995), however, argues that by introducing some aspects of technical differences between trading countries à la Ricardo, the HOV theorem retains its relevance. The aim of this chapter is to assess to what extent the US–Mexican trade can be explained by a version of the multi-factor content theorem or by the vertical integration hypothesis. The main focus is the use of intermediate inputs, rather than the use of factors (labour and capital). These two countries apparently exchange similar goods, while their size differentials make the weight of bilateral trade to be somewhat unbalanced. Nevertheless, due to striking similarities in the productive structures of both economic systems, the structural role of foreign trade in both economies is quite similar (Aroche Reyes 1996). The remainder of the chapter is divided into three sections. The ﬁrst one states the methodology employed. The second presents an exercise of revealed factor abundance and comparative advantage; this section also discusses the correspondence between trade performance and comparative advantage. The last section offers some ﬁnal considerations.

5.1 Methodology Perhaps a good point of departure for a discussion concerning the factor–content theorem is the concept of sectoral vertical integration, which, despite its widespread use, has been seldom explicitly discussed (Pasinetti 1977, Haltia 1994). It seems Sraffa was the ﬁrst modern economist who made the problem explicit, in the context of the multisectoral models and Pasinetti (1977) offers an exact deﬁnition. Coase (1972) has also deﬁned the concept of vertical integration, but in regard to the ﬁrm and its optimal size. If one considers an economic system as a set of interrelated producer sectors, it is easy to deduce that it will actually function along a circular ﬂow. In other words, commodities within the system will be produced by means of commodities (intermediate inputs) together with non-produced goods (e.g. capital, labour, and resources) combined in certain proportions by means of a given technology. Such productive processes, in turn, give way to factor incomes, which allow households and ﬁrms to buy the

104

Fidel Aroche Reyes

produced goods. Finally, those transactions enable ﬁrms to acquire the inputs and factors to continue production. When international trade is involved, the set of relationships between ﬁrms and between producers and households can involve transactions between agents located in different countries. Assuming that each industry produces a single commodity, employing a speciﬁc technology, it is possible to deﬁne a technical coefﬁcient matrix A. Each column j 1, . . .., j, . . ., n, for n commodities n industries, in that matrix represents the technically determined amounts of intermediate inputs necessary to produce one unit of each good j. In terms of the conventional input–output (IO) model, this system can be represented by the following equation (Miller and Blair 1985): (IA)1 · f

(1)

where f is the ﬁnal demand vector and is the total output vector. If the model is to be reinterpreted in physical terms, f will be the vector of those commodities available for consumption and investment, and is a vector of the gross production (intermediate and ﬁnal) necessary to satisfy ﬁnal demand. Pasinetti (1977) deﬁnes the output matrix of the vertically integrated sectors 1, . . ., j, . . ., n as: (IA)1 · f T

(2)

f being the diagonal matrix of ﬁnal demand and each column j in T displays the amount of goods (from sectors 1, . . ., j, . . .., n) absorbed by sector j, directly and indirectly, in order to produce the output destined to ﬁnal demand fj. Each column in T is independent from one another; hence, T is a matrix of virtually independent subsystems, each one producing towards ﬁnal demand. A highly vertically integrated sector draws direct and indirect inputs, as well as factors embodied in those inputs, in higher proportions per unit of output, than a less integrated sector. Second, if the aim of the productive activity is to satisfy ﬁnal demand, it is not difﬁcult to infer that whenever ﬁnal demand in an integrated sector changes, that change will have bigger effects in the economy as compared to those impacts on less integrated industries. Likewise, it is feasible to derive expressions to estimate sectoral integration according to speciﬁc ﬁnal demand categories. Those exercises yield estimates of potential impacts in the output level derived from changes in these demand types. For example, in order to measure sectoral integration in regard to exports, the following equation results: (IA)1 · x

(3)

Vertical integration and comparative advantages 105 x standing for the diagonal matrix of sectoral exports. Each column in matrix represents the produced inputs vector absorbed, directly and indirectly, by each industry 1, . . .., n, to produce the exported goods. Assuming that x is given in value, it is feasible to deﬁne the sum over the columns of as a row vector. That yields the sectoral production needed to produce the bill of exports. Alternatively, x could be also replaced by the net exports vector, deﬁned as the difference between exports and imports of similar goods (similarity by sector of origin); also, matrix A could be substituted by the matrix of domestically produced inputs per unit of output (AI), from which equation (3) could be rewritten as: (IAI)1 · xm *

(3')

The column sum over matrix *, resulting from this alternative equation shows the internal sectoral production value, needed to produce the bill of net exports, net of imported inputs. Normalizing this vector results in a coefﬁcients vector of the relative vertical integration by industry in regard to net exports, from which a sectoral hierarchy can be derived. In line with these exercises, it is not difﬁcult to rewrite the conventional IO model in terms of exports x or in terms of net exports xm. For instance, equation (1) can be rewritten as: (IA)1 · x

(4)

Equation (4) expresses the physical production needed, directly and indirectly as inputs, in order to satisfy the external demand. Replacing matrix A with matrix AI, a new equation is deﬁned; this one indicating the amount of intermediate domestically produced inputs embodied in exports. Naya (1967) employs a similar expression to calculate the natural resource product requirements in the US exports. Further, in this equation, vector x could also be substituted by the export coefﬁcients vector; the result would weigh the input content of exports by sectoral output. If vector x is substituted by vector xm in equation (4), the result can be taken as the necessary production, demanded as inputs, to satisfy net exports. According to the HOV theorem, countries would tend to export the services of those abundant resources within their territories, embodied in the goods actually traded (Vanek 1968). Hence, ranking the entries in the resulting vector would reveal the goods that are exported more intensively embodied in other exported goods. Equation (4) will be rewritten as: (IA)1 · (x m)

(5)

is the vector of the country’s net exports of embodied resources services. In order to test comparative advantage, it is necessary to consider

106

Vertical integration and comparative advantages

relative resource abundance in terms of the world’s resources (Leamer 1980, Deardorff 1984). Leamer (1980) has shown that if all countries allocate expenditures identically, equals the country’s share in the world’s resource availability. Hence, following the HOV theorem, if a country is well endowed with a given resource, it will be a net exporter of that as an embodied input in net trade, even if the country’s direct exports do not directly account for that good. However, if a country shows a trade surplus, it may also be a net exporter of those resources with which it is poorly endowed. It has been suggested then, that the appropriate test for resource endowments should compare the ratios embodied in production and consumption (Deardorff 1984). Yet, normalizing and ranking its entries should sufﬁce to indicate where comparative advantage lies.

5.2 Some tests on vertical integration and revealed comparative advantage The trade composition (exports and imports), for both countries for 1990 by 16 industrial tradable sectors of origin, is showed in Table 5.1. The table splits the ﬁgures for bilateral commerce. It appears that both countries exchange a very similar range of goods. However, the Mexican ﬁgures are somewhat more concentrated in fewer activities. The largest exports and imports by industry of origin for both economies were 3 (food, beverages and tobacco), 8 (chemicals), 13 (non-electric machinery), 14 (electric machinery), 15 (transport equipment) and 16 (other manufactures). According to the vertically integrated sectors hypothesis, whenever ﬁnal demand of an industry grows, that will induce growth in the sector’s output, together with its intermediate demand; this, in turn, will expand output in other industries and even in the overall economy. The level of sectoral integration will determine the size and shape of such effects. Table 5.2 shows the results for this experiment on sectoral integration for the US and Mexican data. This table contains the relative weights of the column sums of matrix T, which shows the independent subsystems of production towards ﬁnal demand, as in equation (2) above. In other words, the results can be taken as a sectoral hierarchy of potential impacts deﬁned by the relative degree of vertical integration. Such results are not so different from that of the total linkage coefﬁcients in Yotopoulos and Nuggent (1973), which are in effect the relative weight of sectoral demands for inputs (direct and indirect) in the total intermediate demand in the economy (Haltia 1994). From the vertically integrated hierarchy in this table, it appears that in both economies the industries 17 (construction), 19 (retail) and 22 (other services) are the highest vertically integrated sectors in 1987 and 1990. That is not surprising, because it has long been established that construction and services are normally highly linked to ﬁnal demand, consume large amounts of inputs relative to output and play similar structural roles

Agriculture and rural industries Mining Food, beverages and tobacco Textiles Leather and products Plywood and wood Paper Chemicals Rubber and plastic Non-metallic mineral prods. Non-ferrous mineral prods. Iron, steel and other metals. Non-electric machinery Electric machinery Transport equipment Other manufactures

1 Total

1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Sector

862 2648 41619 11030 3251 6798 12408 46330 14240 2902 7268 11030 76001 46906 64643 45042

0% 1% 5% 6% 3% 1% 3% 18% 2% 1% 2% 5% 14% 13% 16% 10%

1954 3121 26179 30585 14475 5586 13048 87291 11959 4940 9139 22654 68576 66623 81323 47805

0% 0% 8% 3% 1% 1% 4% 9% 6% 1% 2% 6% 15% 21% 14% 10%

115 113 2209 976 234 275 1074 2536 1592 224 573 1686 4193 5841 3932 2803

1% 1% 8% 3% 1% 1% 1% 20% 1% 1% 2% 5% 8% 26% 12% 8%

445 426 2507 1008 290 211 223 5903 355 396 487 1399 2439 7990 3693 2399

1% 384 2% 659 7% 2731 2% 613 0% 154 0% 147 1% 197 30% 11419 1% 540 1% 396 2% 638 3% 1126 7% 2689 2% 896 9% 3299 33% 12511

0% 149 1% 335 10% 3967 3% 1053 1% 266 1% 199 3% 1086 10% 4003 4% 1380 1% 289 1% 521 5% 1879 14% 5499 8% 3330 4% 1652 35% 13596

Imports

1% 376 1% 443 8% 2417 1% 353 0% 111 0% 140 1% 157 21% 6457 1% 287 1% 338 1% 419 3% 858 7% 2214 2% 657 10% 3014 40% 12160

Exports

0% 118 1% 218 8% 2460 2% 607 1% 183 1% 179 3% 866 9% 2703 4% 1107 1% 188 1% 415 4% 1234 13% 3751 7% 2018 4% 1243 41% 11962

Imports

Trade with US

100% 392978 100% 495258 100% 28376 100% 30171 100% 38400 100% 39203 100% 30400 100% 29252

0% 1% 11% 3% 1% 2% 3% 12% 4% 1% 2% 3% 19% 12% 16% 11%

Exports

Exports

Exports

Imports

Total trade

Trade with Mexico

Total trade Imports

MEXICO

US

Table 5.1 Total and bilateral trade composition between the US and Mexico by sector of origin 1990 (Millions of US dollars)

108

Fidel Aroche Reyes

Table 5.2 Vertically integrated coefﬁcients

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

Sector

US

Mexico

Agriculture and rural industries Mining Food, beverages and tobacco Textiles Leather and Products Plywood and wood Paper Chemicals Rubber and plastic Non-metallic mineral products Non-ferrous mineral products Iron, steel and other metals Non-electric machinery Electric machinery Transport equipment Other manufactures Construction Electricity Retail and hotels Transport Financial services Other services

0.0090 0.0010 0.0677 0.0265 0.0037 0.0124 0.0167 0.0448 0.0015 0.0017 0.0014 0.0054 0.0363 0.0247 0.0834 0.0278 0.1180 0.0203 0.1507 0.0591 0.0526 0.2354

0.0519 0.0090 0.1684 0.0322 0.0100 0.0095 0.0067 0.0255 0.0057 0.0098 0.0011 0.0122 0.0104 0.0155 0.0417 0.0078 0.1128 0.0045 0.2125 0.0753 0.0172 0.1605

Total

1.0000

1.0000

in many countries (Chenery and Watanabe 1958). Despite that, however, the vertical integration hierarchies are different in each country. Among the sectors producing tradable goods, the most vertically integrated industries in the US are 3 (food, beverages and tobacco), 8 (chemicals), 13 (nonelectric machinery) and 15 (transport equipment), whereas in Mexico those are 1 (agriculture), 3 (food, beverages and tobacco), 4 (textiles) and 15 (transport equipment). In fact, the Spearman correlation coefﬁcient of the two columns in Table 5.2 is 0.24 considering the 16 tradable sectors only; the coefﬁcients for all 22 sectors is higher, but still non–signiﬁcant. All this supports the hypothesis that industrial vertical integration patterns are unequal in these economies. As a consequence, identical ﬁnal demand expansions in each country would produce quite different results. It will be through the more integrated industries where expanded demand will have more signiﬁcant effects over the economy. On the other hand, if an industry is vertically integrated, it means that the economy is well endowed with inputs demanded by that particular industry; thus they are widely employed in productive activities. Contrasting Tables 5.1 and 5.2 yields a further result: namely that, in general, the largest US exports correspond to highly vertically integrated sectors. Besides, at least for 1990, that economy presents a trade surplus in

Vertical integration and comparative advantages 109 highly vertically integrated sectors, except for chemicals. Nevertheless, there would not be a perfect correlation between the hierarchies of the vertically integrated sectors and that of the bigger export industries. However, any such relationship, would be less perfect considering the Mexican data for the highly vertically integrated sectors – in 1990 the economy presents a trade surplus only in agricultural goods. Hence, it seems plausible to say that trade patterns for the Mexican economy are not as functional for sustained growth. Besides, this economy would be wasting some trade gains derived from either integrating exporting industries or making integrated industries more export oriented. The study of vertical integration can be further extended to analyse speciﬁc aspects of the economic structure, or it can be useful to assess vertical integration according to given ﬁnal demand categories such as exports or net exports, as discussed above. The theory of international trade assumes that a country tends to export the services rendered by inputs and factors abundant in the country, embodied in the actual exported goods, while importing goods whose production demand locally scarce resources. In other words, from the viewpoint of the vertical integration analysis, international trade involves the exchange of embodied inputs in the goods actually exchanged. Therefore, ranking vertically integrated sectors according to net exports reveals those industries whose trade has greater effects over the productive activity of the economy as a whole. Table 5.3 shows the results for this experiment employing the US and the Mexican data for 1990. It should be noticed though, that the experiment has been performed employing IO tables of total transactions, i.e. domestically produced and imported inputs are combined, due to the lack of a domestic transaction only matrix for the US, which to the best of my knowledge, the Bureau of Economic Analysis (BEA) of the United States Department of Commerce does not publish. From Table 5.3, it can be seen that sectors 4 (textiles), 5 (leather and products), 8 (chemicals), 12 (iron, steel and other metals), 14 (electric machinery) and 15 (transport equipment) are the highest vertically integrated in net exports for the US economy. As for Mexico, sectors 8 and 15 are by far the highest vertically integrated sectors in net exports. In other words, in each economy, those industries have the highest total requirements of inputs to satisfy net external demand. Net exports of these sectors are important to explain the level of production in the system as a whole. Another feature worth considering at this point is the relatively uniform distribution of the entries in the US vector, as compared to the Mexican one. This is clearly related to the manner total trade (exports plus imports) distributes in each country, as it has been deﬁned in this chapter (trade of goods by sector of origin), which can be also analysed by the weight of sectoral balance in the total trade cf. Table 5.1. Both countries seem to have sectors that are ‘negatively vertically

110

Fidel Aroche Reyes

Table 5.3 Vertically integrated sectors by net trade

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

Sector

US

Mexico

Agriculture and rural industries Mining Food, beverages and tobacco Textiles Leather and products Plywood and wood Paper Chemicals Rubber and plastic Non-metallic mineral products Non-ferrous mineral products Iron, Steel and Other Metals Non-electric machinery Electric machinery Transport equipment Other manufactures Construction Electricity Retail and hotels Transport Financial services Other services

0.0118 0.0036 0.1688 0.2096 0.1115 0.0123 0.0078 0.4028 0.0235 0.0183 0.0208 0.1104 0.0658 0.1718 0.1790 0.0231 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

0.4646 0.6234 3.2658 1.1031 0.2789 0.1202 2.1712 20.0912 2.0641 0.2246 0.2971 2.0005 6.6878 5.8261 5.1995 2.3827 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

1.0000

1.0000

Total

integrated’, although there are fewer of such sectors in the US column. Results displayed in Table 5.3 are in fact normalized column sums of matrix * above (equation 3'), whose columns show the input absorption in order to satisfy net exports. When the vector is normalized, negative entries compensate positive ones. Sectors displaying negative coefﬁcients demand a higher value of imported components directly and indirectly than their exports. The actual balance of directly traded commodities can be of any sign (all four negative entries are related to positive balances in the US results and nine out of ten are negative in the Mexican data in Table 5.1), but clearly, such industries would be more tightly related to foreign input suppliers. Therefore, increased ﬁnal demands for these goods result in increased imports of the inputs necessary for production. Another aspect of these results is that it again seems likely to expect that the US productive sector is more tightly integrated than its Mexican counterpart. Besides, a good proportion of activities in that economy show comparable levels of vertical integration, whereas the Mexican industries seem to be integrated unevenly and many are dependent upon intermediate imports. Finally, crossing the information in Table 5.3 and Table 5.1, it turns out that some of the largest export activities in both countries show negative coefﬁcients of vertical integration, which would mean that

Vertical integration and comparative advantages 111 efﬁciency in those exporting industries is achieved by re-exporting intermediate imports. Estimating a country’s comparative advantage, in the sense of the HOV theorem, involves measuring the share of that country in the world’s resources (Haltia 1994), as stated above. Leamer (1980) showed that such a share, under given conditions, could be measured by equation (5). Table 5.4 shows results for an experiment on the US and Mexican data, where it has been assumed that outputs are produced by means of inputs, leaving aside the role of factors in the productive process. This exercise can actually be regarded as a means to estimate the weight of intermediate inputs in net exports, whereas Leamer focused on factors. In terms of Leamer’s paper, however, the present results can also be understood as a measurement of the relative abundance of produced inputs necessary to produce outputs. A highly ranked sector in this table will be one whose produce is embodied in net exports – as intermediate inputs – in a relatively higher proportion. Figures in Table 5.4 show that the US’s exports embody higher proportions of textiles, chemicals, electric machinery and transport equipment. On the other hand, Mexico’s net exports contain higher proportions of chemicals, minerals, transport equipment and electricity, as well as Table 5.4 Revealed relative abundance of inputs

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

Sector

US

Mexico

Agriculture and rural industries Mining Food, beverages and tobacco Textiles Leather and products Plywood and wood Paper Chemicals Rubber and plastic Non-metallic mineral products Non-ferrous mineral products Iron, steel and other metals Non-electric machinery Electric machinery Transport equipment Other manufactures Construction Electricity Retail and hotels Transport Financial services Other services

0.0106 0.0663 0.0697 0.1405 0.0606 0.0029 0.0158 0.2483 0.0168 0.0150 0.0275 0.0872 0.0225 0.1058 0.0933 0.0166 0.0114 0.0287 0.0466 0.0320 0.0128 0.0806

0.2266 1.5497 1.6818 0.7058 0.1692 0.1352 1.2789 12.3996 1.2512 0.1358 0.1922 1.8775 4.0160 3.6530 3.0207 1.5018 0.0000 0.5323 0.1626 0.0696 0.0432 0.1009

1.0000

1.0000

Total

112

Fidel Aroche Reyes

services. It is in those sectors where these countries show comparative advantage. It is worth noticing that the rest of the Mexican sectors display negative results, while in the US economy negative entries are not as frequent. Such entries can be understood as corresponding to industries producing inputs that are displaced by competitive intermediate imports. In other words, the economy as a whole imports competitive inputs embodied in direct imports in larger quantities than it exports. Thus, a negative result implies that domestic production of inputs in that industry is not competitive in the international market. It seems that in the Mexican economy, many more industries are unable to compete successfully with imports as input suppliers. A ﬁnal point concerns the relevance of comparative advantage, as deﬁned in this chapter and the trade patterns of the two countries included in the analysis. It would seem reasonable to say that there is not a one-toone correspondence between the largest exporting sectors and those enjoying the higher comparative advantage in either the US or Mexico. Nevertheless, the former is closer to fulﬁlling this condition. Such a result is related to the more tightly integrated sectoral patterns shown in this economy, since this exercise emphasizes the role that each input plays in the economic structure.

5.3 Summary and conclusions Focusing attention on the sectors producing tradable goods, it seems that the highest vertically integrated sectors in both economies are not the same. That means that identical ﬁnal demand expansions will yield different results in each economy. Moreover, the US exporting industries are more closely correlated to the highly integrated activities, for which trade seems to be more functional to growth, since this economy will experience relatively bigger impacts from expanding trade. These considerations are strengthened considering that US–Mexico exchange a similar range of goods. In both countries, sectors 8 (chemicals) and 15 (transport equipment) are among the highest vertically integrated sectors in net exports and, at the same time, rank among the biggest exporting industries. Nevertheless, there is not necessarily a one-to-one correspondence between the highest vertically integrated industries and the big exporting ones. This implies that, in both national cases, intermediate imports help to achieve efﬁciency in foreign markets. Finally, it has been established that both the US and the Mexican net exports are intensive in different resources, except for chemicals and transport equipment. Once again, it seems that there is not a perfect correlation between industries enjoying comparative advantages, as deﬁned in this chapter, and the biggest exporters. However, given that the US industries are better integrated, these are considerably better positioned to enjoy the gains of bilateral trade.

Vertical integration and comparative advantages 113 APPENDIX

Data sources The data employed have various sources. First, the Make of Commodities by Industries and the Use of Commodities by Industries Tables for 1987, published by the Bureau of Economic Analysis (BEA) of the US Department of Commerce in 1994. Those matrices are disaggregated into 79 commodities and 79 industries. From those tables, a square 79 industry by industry IO matrix was derived for each year, following the standard method (Miller and Blair 1985). The resulting matrix was then aggregated into 22 sectors in line with the aggregated Mexican matrix. The Mexican source is the IO table for 1990, which is an update prepared by CIESA (Consultoría Internacional Especializada S.A.), based on the IO table for 1980, which is the latest published by INEGI (Instituto Nacional de Estadística, Geografía e Informática), the federal ofﬁce devoted to producing ofﬁcial data. Both matrices are disaggregated into 72 industries, so the 1990 one was further aggregated into 22 sectors. Finally, the bilateral trade data is derived from information provided by the Economic Commission for Latin America, an agency of the United Nations. This database classiﬁes transactions of goods according to the two-digit Harmonized System by industry of origin. This information was classiﬁed and aggregated into 16 sectors producing tradable goods; the trade in services is not considered.

Bibliography Aroche Reyes, F. (1996) ‘Integration in North America, trade and productive structures. Some aspects of integration between trade partners’, Mimeo, Mexico: Facultad de Economía, UNAM. Chacholiadis, M. (1992) Economía Internacional, Mexico: McGraw-Hill. Chenery, H. and Watanabe, T. (1958) ‘International comparisons of the structure of production’, Econometrica 26: 487–521. Coase, R. (1972) ‘Industrial organization: a proposal for research’, in Fuchs, V. (ed.) Policy Issues and Research Opportunities in Industrial Organization, New York: National Bureau of Economic Research. Davis, D.R. (1995) ‘Intraindustry trade: a Hecksher–Ohlin–Ricardo approach’, Journal of International Economics 39: 201–26. Deardorff, A. (1984) ‘Testing trade theories and predicting trade ﬂows’, in Jones, R. and Kenen, P. (eds) Handbook of International Economics, Amsterdam: North Holland. Estrada, N. and Kushida, S. (1983) ‘Estructuras productivas y comercio exterior. La integración de dos economías: México y Estados Unidos’, CIDE, Estados Unidos, Cuadernos Semestrales 13: 155–214. Frankel, J., Stein, E. and Wei, S. (1995) ‘Trading blocs and the Americas: the natural, the unnatural and the supernatural’, Journal of Development Economics 47: 61–95.

114

Vertical integration and comparative advantages

Haltia, O. (1994) The Forest Sector in Economic Development. A Comparative Study of Brazil, Chile and Finland, Ph.D. Thesis, Queen Mary and Westﬁeld College, University of London. Helpman, E. and Krugman, P. (1989) Trade Policy and Market Structure, Cambridge, Mass.: The MIT Press. Krugman, P. (1991) ‘The move towards free trade zones’, in Policy Implications of Trade and Currency Zones, Federal Reserve Bank of Kansas City. Klein, L. and Salvatore, D. (1995) ‘Welfare effects of the North American Free Trade Agreement’, Journal of Policy Modelling 17: 163–76. Leamer, E. (1980) ‘The Leontief Paradox reconsidered’, Journal of Political Economy 88: 495–503. Leontief, W. (1953) ‘Domestic production and foreign trade: the American capital position reexamined’, Proceedings of the American Philosophical Society, reedited in Input-Output Economics, Oxford University Press, 2nd edition. Maskus, K. (1985) ‘A test of the Hecksher–Ohlin–Vanek theorem: the Leontief commonplace’, Journal of International Economics 19: 201–12. Miller, R. and Blair, P. (1985) Input–Output Analysis: Foundations and Extensions, Englwood Cliffs: Prentice-Hall. Naya, S. (1967) ‘Natural resources, factor mix and factor reversals in international trade’, American Economic Review 57: 561–70. Pasinetti, L. (1977) ‘The notion of vertical integration in economic analysis’, in Pasinetti, L. (ed.) Essays on the Theory of Joint Production, London: Macmillan. Pearce, D. and Turner, R. (1990) Economics of the Natural Resources and the Environment, New York: Johns Hopkins. Sraffa, P. (1960) Production of Commodities by Means of Commodities, Cambridge: Cambridge University Press. Thomson, A.J. (1993) ‘The anticipated sectoral adjustment to the Canada–United States Free Trade Agreement: an event study analysis’, Canadian Journal of Economics 26: 253–71. Vanek, J. (1968) ‘The factor proportion theory: the N-Factor case’, Kyklos 21: 749–55. Yotopoulos, P. and Nuggent, J. (1973) ‘A balanced growth version of the linkage hypothesis: a test’, Quarterly Journal of Economics LXXXVII: 157–71.

Comments Bodo Hilgers and Jürgen Meckl

The three chapters in this section by Fidel Aroche Reyes, María Elena Cardero and Pablo Ruiz-Nápoles cover three distinct aspects of the economic integration in North America that culminated recently in the North American Free Trade Agreement (NAFTA). Special emphasis is devoted to the perspectives of the Mexican economy, by far the smallest and less developed country participating in the NAFTA integration process. From these analyses, there are several lessons that we can learn for the process of economic integration in Europe – speciﬁcally, the integration of East European economies like Poland, Hungary, the Czech Republic, the Baltic States and Turkey into the European Union. The stage of economic development of Mexico is comparable to the average stage of economic development of the potential EU candidate states. Thus, we might expect that these countries will face similar problems to Mexico during their integration process into the EU. Fidel Aroche-Reyes analyzes to what extent the US–Mexican trade can be explained by a version of the multi-factor-content theorem or by the vertical-integration hypothesis. Based on Pasinetti’s (1977) deﬁnition of the concept of vertical integration Aroche-Reyes ﬁnds that, despite the fact that trade between the US and Mexico has been growing both in volume and weight for both countries, it affects both countries in very different ways. He concludes that the main part of trade is of an intraindustry nature because of the fact that similar goods are exchanged. Additionally, the author ﬁnds that the sectors with the highest degree of vertical integration are not the same in both economies and thus the expansion of trade will have different structural effects everywhere. For some industries like chemicals and transport equipment, Aroche-Reyes ﬁnds that the degree of vertical integration is positively related to the sectors’ export activities. In the light of the differences of the US and Mexican economies with respect to size and structure and the degree of economic development, the ﬁndings of the author are not really surprising. Remarkable, however, is the fact that between such different economies trade is rather of an intraindustry than of interindustry nature. From the perspective of

116

Bodo Hilgers and Jürgen Meckl

neoclassical trade theory (Heckscher–Ohlin model), trade ﬂows, and gains from trade, are caused by price differentials in autarky reﬂecting differences in national relative factor endowments. Thus, we should expect huge trade ﬂows occurring between very different countries. Of course, this refers to intersectoral trade. It is well known, however, that more than 70 per cent of international trade occurs between industrialized nations with similar economic characteristics. This is referred to as intrasectoral trade and can be explained by vertical integration. But inter- and intraindustry trade should have different feedback effects on the involved countries, especially on a small trading partner like Mexico. Aroche Reyes explains trade between Mexico and the USA – two countries with very different economic characteristics – by a vertical integration hypothesis and by intrasectoral trade. These ﬁndings are of great interest both for the potential East European member states and for the EU member states. In accordance with neoclassical trade theory in public discussion, it is often argued that joining the EU means, for the less developed East European states, the opportunity to export relative labor-intensive goods and services like agricultural products. Unfortunately, these sectors have only minor growth potentials and thus joining the EU will bring about a welfare loss for the East European states, at least in the long run. Likewise some EU member states with important agricultural sectors (e.g. Portugal, Spain, Austria and Ireland) fear that they could be swamped by cheap East European products and, thus, the income of owners of speciﬁc agricultural factors will deteriorate. As is well known, income distribution topics are the cause for many crises within the EU and have resulted in the foundation of the socalled ‘structural funds’. Due to Aroche-Reyes, the evidence of Mexico teaches us that such a dismal picture isn’t a necessary consequence of economic integration. The chapter by María Elena Cardero concisely describes the development in Mexico’s production structure and its trade relationships to the US since the 1980s. These years have seen a major increase in the openness of Mexico’s goods and factor markets, a process which had its last stage within the NAFTA agreement in 1994. She devotes special emphasis to the development of a common ﬁnancial market, and to the evolution of the binational labor market primarily caused by the emergence of the maquiladora industry. Her analysis emphasizes that integration involves major structural adjustments in production structures that proceed only slowly over time. As a result, the complete gains from integration cannot be realized instantaneously, and expectations with respect to the beneﬁts of integration may well be disappointed in the short run. This last result should be taken as a clear warning for the candidate states that want to enter the EU. There may be a long and critical phase of adjustment involving the loss of many jobs in certain shrinking sectors and with high demand for labor mobility and labor-market ﬂexibility. The latter require-

Comments 117 ment becomes even more severe when economic integration also involves international labor mobility. From the perspective of trade theory, the dismal conclusion on the gains from integration comes as no surprise. This theory emphasizes that we must distinguish between level effects and growth-rate effects. But the predictions about the effects of integration on the level of GDP and on the growth rate of GDP can differ substantially. According to the static neoclassical trade model, integration should have a positive level effect, provided that there are no further distortions on goods and factor markets. This predicted level effect, however, is a long run effect with the speed of adjustment determined by the intersectoral mobility of factors. The full effect is not realized until the process of structural adjustment comes to an end. During this transitional process to the long run equilibrium, adjustment of sectoral production and GDP may well be characterized by nonmonotonic behavior. This point has been emphasized by, e.g., Blanchard (1997) in the similar context of transformation of East European economies. As far as the empirical analysis of Cardero is concerned, we must be very careful when interpreting the data for the years 1994–8. The observations on structural change indicate that the adjustment process is far from its conclusion. As a result, there is no reason to believe that the level effect of trade liberalization has already been fully realized. Additionally, we have some reservations about the methodological approach with respect to Cardero’s empirical analysis of level effects. If we suppose that the static equilibrium allocation can be described by the neoclassical trade model, Cardero’s approach of using prices of 1993 for discounting should even result in a decline of GDP after NAFTA came into effect. This is simply because, at 1993 prices, the allocation in 1993 maximizes GDP. As a result, the allocations after NAFTA (with different relative prices) cannot maximize GDP for 1993 prices. Hence, Cardero’s ﬁndings of an immediate decline in GDP growth may well be attributed to a negative level effect. With respect to long run growth-rate effects, the predictions of trade theory are by no means clear cut (see, for example, Grossman and Helpman 1991). This, again, is hardly amazing because we all know that at least the results derived from models of endogenous growth are in many cases ambiguous and, furthermore, not too robust with respect to departures from the assumptions on market structures. On the other hand, according to the neoclassical growth theory, market integration should have only transitory effects on the growth rate. Again, we have reservations about Cardero’s measurement of growth effects. In our view, growth effects are long run effects and cannot be discovered adequately by inspection of data over a time interval of three or four years. In the light of these arguments, the apparently dismal picture of the success of NAFTA should not be bewildering. Just as these ﬁndings result

118

Bodo Hilgers and Jürgen Meckl

in a disappointment of the high hopes connected with NAFTA, they should serve to dampen exaggerated expectations within the candidate states for integration into the EU. There may be a long and winding road before the gains from integration will be fully realized, and the necessary reallocation processes may involve costly and time-consuming adjustments especially in labor markets. Pablo Ruiz-Nápoles focuses on the determination of the real exchange rate between the USA and Mexico. His empirical investigation covers the period from 1970 to 1997 which is characterized by three main ﬁnancial and exchange rate crises due to foreign exchange rate shortages: autumn 1976, 1982 and 1994. Interestingly, all three crises emerged in very different circumstances with respect to exchange-rate regimes and policies, and degree of openness of the Mexican economy. The author explains the development of the real exchange rate with an interesting hypothesis: unit-labor costs determine the real exchange rate. Of course, this reverses the usual causality stressed by neoclassical trade theory. For the case of Mexico and the USA, an application of neoclassical trade theory implies that Mexico behaves like a small open economy for which world market prices are given. Thus, due to zero proﬁt conditions, national factor prices are determined by technology and world market prices at each instant. This factor-price dependence holds not only in perfect general equilibrium models, but also in imperfect competitive models where positive proﬁts are possible as long as markups are given exogenously. The continuous depreciation of the Euro with regard to the US dollar has led to enormous pressure on political decision makers to speed up reforms, especially in European labor markets. Economists have pointed out that the depreciation of the Euro has revealed the structural problems of European economies in contrast to the USA. From our view this indicates that world market conditions are a serious determinant of national factor prices even if there are monopoly structures like in most European labor markets. Obviously, our above reasoning is completely static. Differences in factor-price dynamics would require differences in productivity growth or differences in markup determination. The author points out differences in productivity growth but unfortunately not differences in markups. Furthermore, the analysis lacks a sound theoretical basis for the reversed causality with respect to real exchange rate determination. This is essential since the hypothesis implies that the process of wage formation is completely independent of the determination of the real exchange rate. Especially with respect to the long run perspective stressed by the author, it is hard to see that the process of wage formation in a small open economy can be independent of world market conditions. The assumed causality has a further implausible implication, namely that markups are determined by wages in one way or another. We are not aware of any theory that explains this hypothesis.

Comments 119 We are not fully convinced of the interpretation of the empirical analysis. The estimation of equation (4) does not necessarily imply the causality proposed by the author. If there is any kind of factor-price dependence in a neoclassical sense this would mean that world markets determine the wage formation process especially in the long run and then it follows that differences in wages measured by the author are caused by differences in productivity growth in the USA and Mexico. This possibility is also mentioned by the author and so coincides with both interpretations of causality.

Part III

Sectorial performance in manufacturing, industrial organization and the innovation system during the trade liberalization process

6

Integration and polarization Mexico’s economy since 1988 Enrique Dussel Peters

The main goal of this chapter is to understand the changes in the Mexican government’s economic strategy since the end of the 1980s, both conceptually and with respect to its implementation. It is crucial to understand the government’s justiﬁcation as well as the debate about alternative policies. A simple criticism of ‘neoliberalism’, as analyzed, is neither correct nor sufﬁcient. The chapter will also discuss in detail some of the effects of liberalization strategy since 1988, particularly regarding manufacturing. The chapter is divided into three sections. Section 6.1 brieﬂy discusses the conceptual core of liberalization strategy, its economic policy goals and its distinction from ‘neoliberalism’. Section 6.2 considers the impact of liberalization strategy since 1988 on several aspects: macroeconomy, manufacturing and one speciﬁc region (electronics in Jalisco). Section 6.3 presents conclusions drawn from the overall discussion.

6.1 Neoliberalism, export-oriented industrialization and liberalization strategy At least since the end of the 1960s a group of authors – particularly Balassa, Bhagwati and Anne Krueger – initiated a new school of thought based on neoclassical economics, known as ‘export-oriented industrialization’ (EOI). These authors have had a tremendous impact on Latin American policy makers since the 1980s and have little relationship with ‘neoliberal’ authors such as Friedrich von Hayek and Milton Friedman. Since the 1940s, the latter school has had a strong impact, particularly among South America dictatorships during the 1960s and 1970s, in Argentina, Brazil and Chile, among others. Several authors have discussed in-depth the impact of these policies (O’Donnell 1973, Foxley 1988 and Valdés 1995). What are some of the conceptual differences between these schools?1 In brief, it is worth stressing that neoliberalism, as developed by Von Hayek and Friedman, is based on a stark theoretical polarization, i.e. counter to post-1930s communism, but also against other social movements linked to Keynesianism, socialism, Leninism, Trotskyism, and other ‘isms’, and in

124

Enrique Dussel Peters

general against ‘constructive rationalism’ (Hinkelammert 1984). Moreover, neoliberalism’s aggressive stance against totalitarianism justiﬁes authoritarian systems against totalitarianism; even against democratic societies. As a result, the selection process among individuals and societies is of critical importance for neoliberalism. Neoliberal arguments are in some cases conceptually close to Fascism. The priority given to economic freedom over political freedom (Friedman 1962) is most relevant in this context. Thus, neoliberal thought proposes a theoretical, historical, economic and cultural system based on the market, the freedom of individuals, and private property. These arguments pave the way to evaluate prior societies and the ‘evolution of the human being’ based on the development of market freedom, and to conclude that capitalism and authoritarian political systems are compatible for the evolution of economic freedom. On the other hand, many Latin American policy makers – most of them studying in the top US ‘Ivy League’ colleges – were in direct contact during their studies and other experiences with EOI, which has been academically predominant in the US since at least the 1970s. EOI’s authors have also been importantly inﬂuenced by associations with multilateral institutions such as the World Bank and the International Monetary Fund. The starting point of EOI is its critique of import-substituting industrialization (ISI) adopted by most developing countries, and not communism or socialism, since various forms of state interventions generate rentseeking behavior and do not favour the classes that were to modernize the countries. EOI analyzes empirically the proposition that integration into the world market of any economic unit (ﬁrm, region, nation or group of nations) through exports will encourage economic growth and development. It also urges a macroeconomic market-friendly environment (Balassa 1981, Bhagwati and Krueger 1985, Krueger 1997). Neoliberalism goes far beyond just legitimizing market forces, an issue that was raised as early as the eighteenth century by Adam Smith. Neoliberalism is much more aggressive and may be considered antidemocratic, since it supports the elimination of individuals/groups/societies that are not able to integrate or adapt within a society based on economic freedom (Hayek 1981). Neoliberalism even goes far beyond economics as it proposes a revision of science, history, social development and cultural evolution. EOI, by contrast, is much more ‘economicist’, as it does not deal with issues that are not strictly economic. A discussion on supporting authoritarian regimes and the debate between economic and political freedom, for example, goes far beyond the scope of EOI. The latter is relevant from several perspectives. Given the conceptual and policy differences between neoliberalism and EOI, it is possible to argue the pros and cons of neoliberalism in Latin America, and particularly in Mexico since the 1980s. In most of the cases, critiques of ‘neoliberalism’ during the 1990s assume somehow – generally without clear historical and conceptual distinctions – that the economic policies of

Integration and polarization 125 Salinas de Gortari and Pinochet, for example, are similar. This, again, is conceptually and historically misleading. Moreover, most discussions and critiques to ‘neoliberalism’ do not analyze the speciﬁcities of EOI and its implementation, and particularly the related search for alternatives. In Mexico, as well as in most of Latin America, the ‘critiques’ and ‘alternatives’ to neoliberalism seem to be worse than Don Quixote de la Mancha, since there are not even windmills to tilt against: all important political parties during the 1990s, including PRI, PAN and PRD, have distanced themselves from neoliberalism, and even former president Salinas de Gortari, rather cynically, has published his own alternatives to ‘neoliberalism’ (Salinas de Gortari and Mangabeira Unger 1999). In what follows we deﬁne the implementation of EOI in Mexico speciﬁcally, as a liberalization strategy. While such labeling is relatively irrelevant, it is important to deﬁne in conceptual and historical terms the new development strategy, both in order to evaluate it and to propose alternatives. 6.1.1 Liberalization strategy in Mexico since 1988 Mexico followed an import-substituting industrialization model since at least Lázaro Cárdenas (1934–40) that attempted to develop national and ‘priority’ sectors of Mexico’s economy, as well as its domestic activities in general, in order to modernize the economy and society through industrialization. In spite of signiﬁcant successes for the ISI-period,2 the crisis of ISI emerged as a result of the incapacity of the private manufacturing sector – which was enhanced through direct and indirect state interventions throughout the period – to generate forward and backward linkages. The increasing trade deﬁcit of the sector, added to an authoritarian political system, caused increasing current-account deﬁcits and overall balance of payment difﬁculties throughout the 1970s and 1980s. Mexico’s crisis in 1982, which initially resulted from the private and public sectors’ inability to service foreign debt, did not merely reﬂect ‘solvency’ or ‘liquidity’ crisis, but the underlying unsustainability of the ISI strategy. Oil revenues and massive international credits were not sufﬁcient to ﬁnance the crisis of ISI since the late 1960s (Brailovsky et al. 1989, Ros 1991). The speciﬁc international conditions, particularly of the US, did not allow the roll-over of old international credits after 1982. Paradoxically, it was the demand for capital by the US economy that increased interest rates and changed capital ﬂows to the US and other OECD nations. This resulted in the widespread inability to service external debt after 1982, causing the ‘international debt crisis’ of the 1980s. Moreover, in 1979–80 the two-fold increase in oil-prices caused an exaggerated estimation of future oil revenue (Gurría Treviño 1993), whereas prices began to fall in 1982 and eventually collapsed in 1986. The Salinas administration became the starting point of the liberalization

126

Enrique Dussel Peters

strategy in 1988. The charm of export-oriented industrialization, the widespread contacts of Mexican policy makers with US academic institutions and government ofﬁcials, in which export-oriented industrialization was the conceptual mainstream, encouraged the implementation of liberalization strategy. Mexico’s liberalization strategy was consolidated by a series of economic pacts (Pactos Económicos), the ﬁrst one in December 1987. These pacts – which included wage ceilings and allowed for an ex post indexing of wages – were negotiated jointly by union ofﬁcials, the government and the private sector. The pacts became the centerpiece of the new strategy under the Salinas administration, which Zedillo has continued with few changes since 1994. The major pillars and guidelines of this strategy of liberalization, in sharp contrast to ISI, are as follows (Zabludovsky 1990, Córdoba 1991, Aspe Armella 1993, Gurría Treviño 1993, Martínez and Fárber 1994, Zedillo 1994, Dussel Peters 2000): 1

2

3

4

5

Macroeconomic stabilization would be used to induce the process of microeconomic and sectorial growth and development, i.e., all sectorial and speciﬁc policies were to be abolished in favor of neutral policies. Signiﬁcant savings were expected for the abolition of direct and indirect subsidies. The government’s main priority would be to stabilize the macroeconomy. Since 1988, the government has viewed controlling inﬂation rates3 (or relative prices) and the ﬁscal deﬁcit, as well as attraction of foreign investments – the main ﬁnancing source of the new strategy, since oil revenues and foreign credits would be insufﬁcient – as the main macroeconomic variables or priorities of liberalization strategy, backed by restrictive money and credit policies by the Banco de México. The exchange rate would be used as a nominal anchor to control the inﬂation rate, i.e. since the inﬂation control was one of the priorities of macroeconomic stabilization, the government would not allow devaluation, which would impact on inﬂation through higher prices on imported inputs. Through the reprivatization of the banking system beginning in the mid-1980s, and privatization of state-owned industries (paraestatales), the private sector was to lead Mexico’s economy out of the ‘lost decade’ of the 1980s through exports. The major import liberalization process, initiated at the end of 1985, was supposed to help reorient the private manufacturing sector toward exports as a result of cheaper international imports. Finally, government policies toward the labor unions were of utmost signiﬁcance. As reﬂected in the pactos, only a few (governmentfriendly) labor unions were deemed acceptable to negotiate inside ﬁrms and with the government, while the rest were declared illegal.

Integration and polarization 127 This process, which involved violent disruptions of independent labor unions, has, since 1987, made national wage negotiations in Mexico possible within the framework of the successive economic pacts. In this liberalization strategy it was of crucial importance to secure export channels. Otherwise, liberalization strategy would be doomed to fail, i.e. NAFTA and other negotiated trade agreements, including with the European Union, are functional and necessary for the liberalization strategy. After the crisis of 1994–5 – the worst since the international crisis of 1929–33 – the government had not implemented new policies or changes to the liberalization strategy up to 1999. It stressed that the ‘political and criminal events’ (Banco de México 1995: 23) as well as the ‘errors’ of December 1994 were responsible for the outbreak of the crisis. Most remarkably, the government, even up to 1998, had not been able to present a clear analysis and lessons from the crisis. In the best of cases, and this perspective has been picked up internationally, the Mexican crisis has been addressed as a ‘ﬁnancial crisis’. The Zedillo administration has been relatively coherent with liberalization strategy and particularly with its macroeconomic priorities (control of inﬂation and of the ﬁscal deﬁcit, as well as attraction of foreign investments). After the crisis and beginning with the Zedillo administration, the Plan Nacional de Desarrollo 1995–2000 (PEF 1999) stresses the need to increase national savings through exports and foreign investments, within a framework of ‘fundamental macroeconomic equilibrium’ (PEF 1995: 145). Initial departures from this liberalization strategy, particularly regarding industrial policy,4 were abandoned and viewed as unnecessary after the apparent economic recovery of the Mexican economy since 1996. It is in the context of the crisis of December 1994 that liberalization strategy posits that a ‘second generation’ of reforms is required: ‘macroeconomic stability and the removal of allocation distortions will be necessary, but certainly not sufﬁcient . . .’ (Edwards and Burki 1995: 9), i.e. liberalization strategy will have to be deepened in order to achieve success. This view acknowledges that the strategy has already been successful regarding macroeconomic priorities, as well as productivity and export growth, but still requires the removal of profound distortions, particularly in the labor market and regarding social issues and social security. This view stresses that the direction of reforms is generally correct, but they have to further the privatization of education and social security, abolish labor market restrictions and minimum wages, among others, that generate perverse incentives and hinder growth (World Bank 1995a, b). From this perspective, liberalization strategy, and independently of the 1994 crisis, still has a long way to go. Up to the end of 1998, the Mexican government has continued, with few exceptions, coherently with liberalization strategy. Emphasis

128

Enrique Dussel Peters

continues on the private export-oriented sector as the basis for economic growth, on privatization and import liberalization, as well as the abolition of most subsidies – culminating at the beginning of 1999 with the abolition of subsidies for tortillas and most commodities of the ‘basic food basket’ – as well as on services and credits. In addition, there is a focus on making labor laws more ﬂexible and on the falling tendency of real wages. Perhaps the most signiﬁcant incoherence of liberalization strategy in its own terms has been the massive public bailout of the ﬁnancial sector, with estimated costs of around 20 percent of GDP in 1999. The privatization of public banks at the beginning of the 1990s resulted in a boom of credits for consumption goods and real estate. Given the positive expectations that liberalization strategy generated under the Salinas administration, both nationally and internationally, and high real interest rates, the crisis of December 1994 resulted in massive bad loans for the recently privatized ﬁnancial sector. The government, however – and in contradiction to its policy regarding social issues, small and medium enterprises, subsidies and industrial policy, among other issues – decided to bailout the banking sector and to socialize its losses.

6.2 The impact of liberalization strategy The following section will examine the impact of liberalization strategy from different perspectives. The ﬁrst part will analyze the macroeconomic impact of liberalization strategy. The second and third parts will include the impact on manufacturing and on a speciﬁc region and sector in Mexico: electronics in Jalisco. 6.2.1 Macroeconomic impact It is relevant to stress that, strictly from liberalization strategy’s perspective, several contradictions arise, including:5 1

2

Given the obsession of liberalization strategy with inﬂation – since neoclassical theory and EOI emphasize that relative prices send signals to producers and consumers to efﬁciently allocate their resources – the real exchange rate will have a tendency to overvalue. However, this inherent real exchange overvaluation will have a negative impact on interindustry trade and will generate incentives for overall imports. The government will have to increase real interest rates, compared to other international markets in order to attract foreign investments. This policy, however, will also have a negative impact on domestic investment.

Table 6.1 reﬂects some of the most relevant tendencies in the Mexican economy since the implementation of liberalization strategy. First of all, it

2.6 0.3 2.2 4.4 84.1 70.9 17.4 5.4 12.0 20.1 20.5 0.4 63.7 62.4 1.3 9.9 35.9 35.1 0.8 1.5 5.7 0.5 0.5 1.0 76.0 34.7

8.3 5.3 14.7 4.7 100.0 100.0 24.8 10.7 14.1 25.5 20.5 5.0 29.8 –– –– 3.1 24.9 35.3 10.4 11.4 4.2 2.1 2.2 0.1 57.5 19.2

1985

22.0

1.3 –0.8 0.9 3.6 76.4 53.7 18.5 4.4 14.1 22.6 21.3 1.3 51.7 69.5 17.8 9.7 42.1 44.5 2.4 1.2 6.6 5.9 3.2 2.7 58.9

1988

34.3

3.4 1.3 2.9 3.0 73.9 49.4 17.2 4.2 13.0 22.9 20.3 2.6 19.7 45.0 25.3 5.0 48.1 53.9 5.8 3.2 6.9 2.8 2.5 0.3 49.5

1989

22.5

5.2 3.3 4.8 2.8 71.5 43.1 17.9 4.3 13.6 23.1 20.3 2.8 29.9 34.8 4.9 2.6 56.1 63.5 7.5 8.4 10.3 0.2 3.7 3.9 39.3

1990

31.2

4.3 2.2 3.0 2.6 73.6 40.7 18.7 4.1 14.6 23.3 18.7 4.7 18.8 19.3 0.5 0.2 58.1 72.7 14.6 25.1 18.1 18.7 6.6 12.1 37.2

1991

46.3

3.7 1.9 1.6 2.8 77.5 39.4 19.6 3.8 15.8 23.3 16.6 6.7 11.9 15.6 3.7 1.4 61.7 86.1 24.4 27.0 19.3 24.4 5.2 19.2 32.4

1992

40.4

1.8 0.0 1.1 3.4 79.2 38.9 18.6 3.8 14.8 21.0 15.2 5.8 8.0 15.0 7.0 –0.3 67.8 91.2 23.4 33.8 24.3 33.3 4.9 28.4 30.4

1993

46.1

4.6 2.7 2.5 3.7 81.6 38.8 19.4 3.8 15.6 21.7 14.7 7.1 7.1 14.1 7.0 0.7 78.4 108.0 29.7 15.6 6.1 23.1 14.9 8.2 48.7

1994

39.9

5.1 3.4 3.4 5.5 60.0 30.3 17.2 3.7 13.5 23.3 22.7 0.6 27.7 31.4 3.7 0.5 115.5 117.8 2.3 4.1 17.5 23.1 8.9 14.2 43.1

6.2 8.0 2.9 6.2 69.7 33.3 16.1 3.3 12.8 19.8 19.3 0.5 52.0 48.4 3.6 0.8 97.0 98.6 1.6 10.5 15.7 –0.3 9.4 9.7 59.4 35.7

1996

1995

Notes 1 Preliminary. 2 Treasury bills (CETES), 28 days, average of period. 3 Refers to total income less total expenditures of public sector. 4 Billion $US.

Sources: Author’s estimations based on data from CEPAL, INEGI and Banco de México.

GDP growth rate GDP per capita growth rate Employment growth rate Open unemployment rate Real wages (1980 100), total economy Real wages(1980 100), minimum wages Gross ﬁxed investment/GDP Public Private Gross savings/GDP National Foreign Inﬂation Nominal interest rate2 Real interest rate Financial deﬁcit/GDP3 Exports of goods and services4 Imports of goods and services4 Current account4 Capital account4 International reserves4 Foreign investment4 Direct4 Portfolio4 Total foreign debt\GDP Foreign debt service of public sector\ exports of goods and services

1980

Table 6.1 Main macroeconomic variables

32.7

6.8 4.8 3.8 3.7 58.7 30.1 19.1 3.6 15.5 26.4 24.6 1.8 15.7 19.8 4.1 1.4 131.5 139.0 7.4 15.4 28.0 17.9 11.8 5.0 34.2

1997

––

4.8 3.0 –– 3.3 57.0 29.5 19.3 2.3 17.0 22.5 18.8 3.7 18.6 33.0 14.4 2.2 140.6 156.4 15.8 16.2 30.1 11.5 10.2 1.3 ––

19981

130

Enrique Dussel Peters

has to be stressed that this strategy, in its own terms, has been relatively successful in the control of inﬂation and the ﬁscal deﬁcit, as well as the attraction of foreign investments.6 Inﬂation has fallen signiﬁcantly since 1988 and, with the exception of the 1994–5 crises, has remained under 30 percent since 1989 (Table 6.1, Figure 6.1). Similarly, the ﬁscal deﬁcit has been controlled since 1988, particularly if compared to the 1980s. Similarly, foreign investments for 1988–98 have soared impressively, amounting to a cumulative $160 billion. Mexico has probably been one of the most successful nations since 1988 in attracting foreign investments. Finally, exports have been, without doubt, the main motor of accumulation of Mexico’s economy. The annual average growth rate of exports was 11.5 percent, more than three times higher than GDP growth for the period. Moreover, exports have substantially increased their weight in GDP since 1988: from levels below 20 percent of GDP during the 1980s to over 30 percent since 1995 for total economy. Nevertheless, it is important to expand this rather primitive perspective on macroeconomics to include some other traditional macroeconomic variables. Table 6.1 reﬂects for 1988–98 that: 1 2

3

4

Growth of GDP and GDP per capita, with strong oscillations for the period, have been signiﬁcantly below levels achieved during ISI. Gross savings as a percentage of GDP have declined since 1988 and have only been able to increase after the crisis of 1994–5. Until 1994, foreign savings increased substantially. Moreover, the ratio of gross ﬁxed investments on GDP since 1988 have not been able to achieve levels similar to those of the 1980s and, most importantly, private gross ﬁxed investment has been unable to counter-balance the fall of public gross ﬁxed investments. Although foreign debt as a percentage of GDP and foreign debt service have fallen since 1988, they still present relatively high levels, above 30 percent in 1997, and represent a threat to the overall economy. International or domestic instabilities, with an impact on capital inﬂows, can quickly bring back the issue of Mexico’s foreign debt. As discussed earlier, exports have shown outstanding dynamism since 1988. However, so have imports. The trade balance as a percentage of GDP, not including maquiladora activities, increased from 0.52 percent in 1988 to 6.94 percent in 1994, and fell again in 1995–6 as a result of the crisis. Thus, and rather ironically, export-oriented industrialization since liberalization strategy has resulted in an importoriented industrialization.

Several other issues are relevant for understanding Mexico’s deep macroeconomic changes since its liberalization strategy. On the one hand, real wages, both for the overall economy and minimum wages, have fallen

0

20

40

60

80

100

120

140

160

180

80

Source: Table 6.1.

Figure 6.1 Inﬂation rates 1980–98 (average annual percentage).

19

81

19

82

19

83

19

84

19

85 19

86 19

87 19

88 19

89 19

90 19

91 19

92 19

93 19

94 19

95 19

96 19

97 19

98 19

132

Enrique Dussel Peters

dramatically since 1988, and represent in 1998, an estimated 57 percent and 29.5 percent of 1980, respectively. This tendency is strongly related to the economy’s inability to generate employment and to absorb the increasing economically active population (EAP). For 1988–96 EAP increased by 9.4 million (PEF 1999), while formal employment increased by 4.2 million, i.e. 5.2 million or 55.06 percent of EAP did not ﬁnd a formal job during this period (Dussel Peters 1998). These tendencies are complementary to data on employment generation by micro, small, medium and big ﬁrms.7 As reﬂected in Figure 6.2, micro, small and medium ﬁrms (MSMF) have been hardest hit during the liberalization strategy with respect to employment generation, but also regarding the generation of new establishments in manufacturing. Employment share of MSMF fell from 49.79 percent in 1988 and levels above 50 percent during the 1990s to 42.81 percent in 1998, and MSMF, in contrast to other periods, only generated 26.29 percent of total employment for 1988–98. Firm level data for exports show some other relevant aspects of the economy. Mexico’s exports, the motor of accumulation since 1988, have not only been highly concentrated, but also reﬂect an increasing economic polarization. Between 264 and 312 ﬁrms, added to maquiladora exports, account for 93.65 percent of total exports in average during 1993–8 (Figure 6.3). As a result, the rest of Mexican ﬁrms – more than 2.8 million according to the last Census of 1998 – participate only marginally in export activities. It stands out that majority owned foreign ﬁrms – between 54 and 68 ﬁrms for the period – have substantially increased their export share. In spite of these tendencies, the main exporting ﬁrms and maquiladoras – around 3,300 ﬁrms in 1998 – account for only 5.59 percent of total employment in average for the period. This is probably one of the most striking features of Mexico’s export-oriented sectors in general: their inability to absorb the increasing economically-active population. 6.2.2 Impact on manufacturing It is in this general context that the general tendencies of manufacturing sector can be appreciated more in depth. Table 6.2 reﬂects some of the general features and tendencies of Mexico’s manufacturing sector since 1988, including: 1

2

Manufacturing’s GDP growth oscillated strongly for 1988–98 and its AAGR was 4.6 percent, higher than overall economic growth of 3.4 percent. As a result, the share of manufacturing’s output in total GDP increased slightly for the period and has remained at around 21 percent of total economy’s GDP since 1995. The share of manufacturing’s employment in total employment has fallen signiﬁcantly, from 12.61 percent in 1988 to around 12 percent in 1997. This tendency not only reﬂects a higher capital intensity of the

1989

1990

1991

Source: own calculations based on IMSS.

1988

1992

1993

1994

1995

1996

Source: Author’s calculations based on IMSS.

Figure 6.2 Employment of micro, small and medium ﬁrms (MSMF) and big ﬁrms (Manufacturing sector only).

Figure 6.2 Employment of micro, small and medium firms (MSMF) and

0%

20%

40%

60%

80%

100%

1998 Big MSMF

1997

300-employment

300-exports

Maquila-employment

Maquila-exports

Source: Dussel Peters (2000).

Figure 6.3 Share of maquila and main 300 exporting ﬁrms over total exports and employment 1993–8.

0.00

10.00

20.00

30.00

40.00

50.00

60.00

Integration and polarization 135

3

4

5

6

sector (net capital stock/employment) than the rest of the economy, but also limitations in generating employment. In terms of productivity, manufacturing has outperformed the rest of the economy. Labor and capital productivity has increased by 27.2 percent and 15.02 percent during 1988–97, both at signiﬁcantly higher growth rates than the total economy. Manufacturing has probably performed most successfully in terms of exports. Not including maquiladora activities, manufacturing’s share has increased from 63.41 percent in 1988 to more than 80 percent of total exports of goods in 1998.8 However, and in spite of the GDP and export growth, manufacturing has not been able to overcome its most severe structural limitation since import-substitution: its high trade deﬁcit. Thus, while exports have been increasing, so have imports, resulting in a high, increasing and unsustainable trade deﬁcit. This lack of endogenous growth conditions, which has deepened since liberalization in 1988, reﬂects that manufacturing increasingly requires imports to allow for GDP and export growth. Thus, the trade deﬁcit increased sharply, from $6.2 billion in 1988 to $32.6 billion in 1994, and fell in 1995 as a result of the crisis. Since the apparent recovery of Mexico’s economy in 1996, the trade deﬁcit in manufacturing has again increased substantially. It is important to recall that this trade deﬁcit by no means generates an automatic mechanism for its own ﬁnancing (Banco de México 1995). On the contrary, Mexico’s economy and society have to ﬁnance these deﬁcits by different means, either by achieving a trade surplus in other sectors (such as oil, agriculture, tourism or other services) or by attracting foreign investments through high real interest rates. Independent of its absolute value, it is also important to relate the trade deﬁcit to manufacturing’s GDP, i.e. as a coefﬁcient that reﬂects the penetration of net imports. From this perspective, the trade balance/GDP coefﬁcient increases from 15.40 percent in 1988 to 44.90 percent in 1994. These high levels had never been reached in Mexico since the 1960s (Figure 6.4). After the crisis of 1995, and as a result of GDP growth in manufacturing and the economy as a whole, the trade balance/GDP coefﬁcient has picked up again.

The latter tendencies have also been corroborated at the branch level (Dussel Peters 2000):9 1

The most dynamic export-oriented branches during 1988–96 (automobiles, electronic equipment, other textile industries, household appliances, machinery and electric equipment and soft drinks and ﬂavorings) signiﬁcantly increased their share of manufacturing’s GDP from 10.61 percent in 1988 to 18.28 percent in 1996. The latter tendencies are dominated by the automobile and electronic

1989

1990

1991

1992

1993

1994

1995

1996

1997

19983

Notes 1 Annual growth rate. 2 Percentage over total economy. 3 Estimations.

Source: Authors’s calculations based on INEGI (Sistema de Cuentas Nacionales and BDINEGI) and Banco de México.

GDP1 –.00 7.90 6.80 3.40 4.20 –0.70 4.10 4.90 10.80 10.00 7.40 GDP2 23.86 21.90 20.80 20.59 20.25 19.04 18.76 20.86 21.54 21.71 21.97 Employment1 –.00 4.40 3.40 1.00 2.20 2.10 2.10 5.30 6.90 8.70 5.00 Employment2 12.61 12.79 12.62 12.38 12.44 12.05 11.50 11.21 11.60 12.14 12.00 Labor productivity (1988 100) 100.00 98.54 103.77 107.17 106.82 110.02 117.72 119.21 125.37 127.21 –.00 Labor productivity (total economy 100) 136.48 134.22 138.04 141.13 139.05 141.72 147.78 156.30 158.65 156.66 –.00 Capital productivity (1988 100) 100.00 111.09 123.98 127.54 129.53 120.41 116.67 105.48 115.02 –.00 –.00 Capital productivity (total economy 100) 43.25 45.96 48.60 49.81 50.76 48.12 46.44 46.23 49.50 –.00 –.00 Exports goods1 –.00 6.40 8.70 6.10 9.30 10.80 24.10 35.20 22.20 13.10 11.40 Exports goods2 63.41 60.70 57.96 63.80 64.78 69.84 73.22 77.49 76.45 79.06 84.08 Exports goods (1988 100) 100.00 106.43 115.73 122.81 134.23 148.67 184.56 249.47 304.73 94 385.03 –.00 22.30 24.60 22.90 23.60 0.40 20.30 –27.70 24.10 28.12 14.50 Imports goods1 Imports goods2 90.33 91.18 92.70 94.12 93.69 94.77 94.40 93.71 91.94 93.59 94.00 Imports goods (1988 100) 100.00 122.32 152.41 187.24 231.47 232.32 279.44 201.92 250.64 321.11 359.30 Trade balance (million $US) 6,184.00 10,610.00 15,076.00 20,568.00 29,425.00 27,702.00 32,596.00 7,465.00 11,516.00 20,536.00 23,00000. Trade balance /GDP 15.40 23.70 30.16 34.84 43.70 39.24 44.90 13.69 17.70 26.44 30.00

1988

Table 6.2 General tendencies of the manufacturing sector (1988–98) (does not include maquiladora activities)

1981

1982

1983

1984

1985

1986

1987

1988

1989

1990

Source: Author’s calculations based on INEGI (not including maquiladoras).

Figure 6.4 Trade balance/GDP (percentages).

Source: own calculations based on INEGI (not including maquiladoras).

50

40

30

20

0 1980 10

10

20

1991

1992

1993

1994

1996

1997

Total economy Manufacturing

1995

138

2

3

4

5

6

Enrique Dussel Peters equipment sectors, with an average annual growth rate of 12.8 percent and 9.9 percent for 1988–96, respectively. As noted earlier, export-oriented branches generate employment that is much less than expected given their export and GDP performance. Their share increased from 14.68 percent in 1988 to 17.39 percent in 1996. The automobile sector, on the other hand, generated 1.30 percent of manufacturing employment during 1988–96 and 0.08 percent of total economy’s employment for the period. Real wage tendencies present some of the most outstanding features of export-oriented branches. If comparing the respective branch’s real wages with manufacturing, particularly export-oriented branches as a group, present real wage levels similar to the rest of manufacturing, while other non-export-oriented branches present in 1996 levels more than 20 percent above manufacturing. These tendencies, without doubt, require more in-depth analysis. However, they reﬂect that there is per se no positive association between export-oriented activities and real wages in Mexico’s manufacturing sector. These tendencies also reﬂect the quality of new employment, considering that real wages have fallen dramatically in Mexico for the period.10 Labor and capital productivity have also increased substantially for the period. However, export growth has probably been the most outstanding feature of manufacturing, even if maquiladora activities are excluded. Export-oriented branches increased their share of total manufacturing from 18.64 percent in 1988 to 40.96 percent in 1996. Only the automobile sector increased its share from 11.33 percent of manufacturing’s exports to 29.55 percent for the period. Imports, on the other hand, have been at least as dynamic for the period and have generated an accumulated trade deﬁcit for manufacturing of $161 billion for 1988–96. For the sector as a whole, imports increased by 179.4 percent during 1988–94 and fell as a result of the crisis of 1994–95. Moreover, export-oriented branches have reported the most relevant import dynamism for 1988–96. As a result, practically all manufacturing branches generate a trade deﬁcit. The automobile industry is probably the only important exception for the period, generating a trade surplus of $39.6 billion. However, if we include the balance of the motors and autoparts sector, which imports most of the inputs of the automobile sector, the trade surplus only amounts to $3 billion for 1988–96. These tendencies result in an increasing trade balance/GDP coefﬁcient, with negative sign, since 1988, and accounting for 44.90 percent (or a trade deﬁcit of 32.6 billion for manufacturing in 1994). As a result of the crisis and falling domestic demand, the coefﬁcient fell in 1995, but has picked up again as soon as the economy resumed growth. Thus, one of the most relevant features of manufacturing is its lack of endogenous growth conditions; i.e. even the most dynamic

Integration and polarization 139 export-oriented sectors since liberalization require increasing imported inputs to allow for a given growth of GDP and exports. This has profound impacts on Mexico’s economy regarding employment generation, overall linkages to other sectors and regions, as well as learning processes, among other variables. 6.2.3 Regional aspects Analysis of regional development has, so far, received little attention in Mexico. Some authors (Dávila Flores 1999, Dussel Peters 1999, Ruiz Durán 1999) have stressed that, in step with macroeconomic and manufacturing tendencies, there are polarization tendencies, as a few regions, particularly those in the Northern border linked to foreign investments inﬂows and maquiladora activities, have increased their dynamism in terms of GDP and GDP per capita since 1988. Similarly, the traditional economic political centers of the country – particularly Mexico City – have regained weight. However, most of regions, particularly in Southern Mexico – with the exception of Quintana Roo as a result of tourism – have not integrated to export-oriented activities and overall economic development. This part will brieﬂy evaluate the relatively successful electronic sector in Jalisco. The goal of the analysis will be to present, in summary, the industrial organization that has emerged in this speciﬁc sector in order to elucidate prior issues raised for manufacturing and the overall economy. 6.2.3.1 The electronic industry in Jalisco11 The electronic industry in Mexico, as well as the automobile sector, has been one of the most successful branches of Mexico’s manufacturing sector since 1988. As discussed earlier, electronic equipment stands out in terms of GDP, export, labor and capital productivity growth, among other variables. Exports AAGR increased by 22.9 percent during 1988–96, but trade balance/GDP coefﬁcient also increased from 110.89 percent in 1988 to 143.94 percent in 1996. Thus, similar to most of Mexico’s manufacturing, this particular branch presents an increasing lack of endogenous growth since liberalization strategy. In the particular case of Jalisco, electronics reﬂect an impressive growth in terms of GDP since 1980 and its share over Mexico’s electronic industry increased from 2.93 percent in 1980 to above 12 percent after 1995. Total exports have also increased dramatically, accounting for a growth rate above 100 percent for 1994–7 to an estimated $6.5 billion in 1997. The United States (63 percent of total exports in electronics in 1997) is the main recipient. The three main exporting ﬁrms in Jalisco (IBM, Motorola and Kodak) are all electronic ﬁrms. IBM, SCI Systems, Motorola and Lucent Technology accounted for 94.9 percent of total exports in electronic products of Jalisco in 1996.

140

Enrique Dussel Peters

The electronic industry in Jalisco has specialized in computer products such as PCs, laptops, printers, telephones, ﬂoppy disks, semiconductors, cables, beepers and other electronic components and ﬁnal goods. According to input–output matrixes, the electronic industry in Jalisco accounts for the lowest national and regional integration level in 1996 (CEED/UDG 1997). Until 1997, the electronic cluster in Jalisco included more than 70 ﬁrms, 28,000 direct and more than 100,000 indirect jobs, as well as 53 percent of Jalisco’s exports. In this context, what are the main characteristics of the electronic industry in Jalisco? 1

2

3

After the initial establishment and expansion of IBM in Jalisco in the mid-1980s, a signiﬁcant cluster of electronic ﬁrms initiated activities in Jalisco during the 1990s. These two generations also had different reasons for establishing in Mexico. While the ﬁrst generation was attracted by cheap labor power, proximity to the US and various government programs, the second generation, in addition, include NAFTA, macroeconomic and foreign direct investment policies and the Latin American market. The second generation, moreover, is also clearly a result of a process of ‘cumulative causation’, i.e. new subcontracting and specialized ﬁrms established in Jalisco as a result of the already existing cluster and demand in the region. In general, Jalisco’s electronic industry reﬂects a ‘squeezed’ or ‘funnellike’ value-added structure. This is a result, on the one hand, of few existing regional and national suppliers, and on the other, second and third-tier ﬁrms are foreign-owned and import most of their inputs, including basic raw materials. With few exceptions, the electronic industry in Jalisco has specialized in assembly activities. Most ﬁrms operate formally or informally as maquiladoras, with a low level of linkages with the rest of the economy and domestic value-added.12 Why has this industrial organization evolved? Some research (CEPAL 1998) shows that ‘ﬁrst generation’ electronic ﬁrms in Jalisco generated supplier systems and ﬁrms for products that were ‘necessary’, i.e. processes and products such as packaging, and plastic injection, among many others, which are too expensive to be imported substantially, particularly due to their high volume. In this case, transnational corporations themselves developed the supplier ﬁrms, supporting them through engineers, technology and, in some cases, even ﬁnancially. Besides such ‘necessary’ products and processes, the second generation of ﬁrms have become established in Jalisco with a system based on imported supplies, i.e. they negotiate contracts with a group of TNCs to supply them, with most of their inputs imported. For this reason such an industrial organization generates rather perverse structures. Since TNCs will continue importing raw materials, components and parts if they do not ﬁnd them in

Integration and polarization 141

4

5

the local or national market, while potential regional and national suppliers simply acknowledge too wide a technological and ﬁnancial gap for the potential buyers. Few regional and national attempts have been made to overcome this rational, but perverse, industrial organization. Few ﬁrms in Jalisco operate as original equipment manufacturers (OEM) and are viewed as an important step forward in the valueadded chain in electronics, particularly in those processes that were deﬁned as ‘necessary’ for client ﬁrms. However, it is also important to highlight that client ﬁrms maintain a strict and overall control of the OEM ﬁrms. Strict speciﬁcations regarding production inputs, suppliers, raw materials, machinery and international organization of the ﬁrm do not only reﬂect this direct control of the client-ﬁrm, but also present the difﬁculties to open new windows for suppliers in the region. Until 1999, few government efforts have been undertaken to overcome these structural constraints. At the national level, the government continues with horizontal industries policies in order not to affect any speciﬁc activity, sector, and/or region and not to contradict macroeconomic aspects of liberalization strategy. At the regional level, the opposition government since 1995 has developed several programs to address some of these difﬁculties. But given Mexico’s political and economic centralist structures, regional governments have few resources to face these kinds of structures.13

The pre-existing structures reﬂect the difﬁculty to overcome this rational, but perverse industrial organization. The development of Jalisco’s electronic cluster would have been unimaginable and impossible without policies implemented since liberalization, with an important impact on exports and employment in the region. Moreover, Jalisco’s electronic exports have boomed during the 1990s and a few regional and national suppliers have integrated into the electronic cluster in Jalisco. However, the electronic cluster in Jalisco has generated a structure with little learning effects and little potential for generating endogenous (regional and national) growth conditions in the medium and long run.

6.3 Conclusions The ﬁrst part of this chapter stresses the relevance of historical and conceptual clariﬁcation and the misleading use of ‘neoliberalism’ for understanding most of Latin America’s development strategy since the 1980s. In the Mexican case, neoliberalism is not related to export-oriented industrialization and particularly to liberalization strategy.14 This is theoretically relevant, but particularly important for viewing economic and political issues and to raise alternatives to the development strategy since the

142

Enrique Dussel Peters

1980s. Otherwise, critiques to ‘neoliberalism’ will be worse off than Don Quixote de la Mancha tilting at windmills centuries ago. Mexico’s liberalization strategy began with the series of Pactos Económicos in December 1987. As analyzed, the liberalization strategy involves export-oriented industrialization and assumes that the integration within the world market of any economic unit has to be established through exports, which will have a positive impact on productivity, economic growth and overall development. In aggregate terms the liberalization strategy has been relatively successful in its own terms, i.e. it has managed to control macroeconomic conditions – control of inﬂation and ﬁscal deﬁcit, as well as in attracting foreign investments – while achieving enormous successes in export growth and productivity. It is important to stress that liberalization strategy in Mexico has not been an overall failure, as some critics allege. This strategy has been able to encourage intraﬁrm trade and linkages, and particularly with the US through US ﬁrms. They have been able to generate signiﬁcant global commodity chains and transnational networks. These ﬁrms, including Mexican grupos (Garrido 1998), have been able to generate a worldwide and North American regional integration process. NAFTA and other trade and investment agreements have been signiﬁcant in this respect. Automobiles and auto parts, electronics, maquiladoras in general, garments, among many other branches, reﬂect these tendencies. These activities have been able to advance some of macroeconomic variables in aggregate terms. However, liberalization strategy and worldwide and US integration since 1988 has also resulted in increasing economic polarization.15 Exportorientation has been accomplished only by a small group of branches and ﬁrms, while most Mexican ﬁrms remain separate from these activities. Polarization, from this perspective, means an increasing concentration of a relatively small number of ‘successful’ ﬁrms, branches and regions. However, these ﬁrms, branches and regions – although successful in terms of exports, GDP and productivity – have not been able to generate endogenous growth conditions, i.e. to link their activities to regional and national territories. Further, they have generated a ‘perverse’ industrial organization that led to Mexico’s crisis of 1994–5 and that has not been able to solve Mexico’s economic, social or regional polarization. On the one hand, export-oriented activities have generated an increasing trade deﬁcit, which has to be ﬁnanced by other sectors and classes of Mexico’s society. Increasing dependency on foreign investments reﬂects the ﬁnancial fragility of the liberalization strategy. Moreover, as a result of global commodity chains and overall productivity gains, the share of exportoriented ﬁrms, branches and regions to overall employment and employment generation, as well as to real wages, has been dismal. The case of the electronic industry in Jalisco shows that, like most of manufacturing, it has not been able to generate learning processes and supplier systems, both regionally and nationally, for economic sustainability and overall

Integration and polarization 143 economic and social endogeneity. Being successful mainly in their own terms, they do not solve the problems of a nation of more than 95 million inhabitants. Under these structures and industrial organization, a continuation of the liberalization strategy and integration to the world market, particularly to the US, will further polarize Mexico’s economy and society. The generation of unsustainable economic conditions in manufacturing – since it requires increasing net imports to grow in terms of GDP and exports – will also deepen macroeconomic unsustainability, while the system remains fragile. From this perspective, critiques of ‘neoliberalism’ for Mexico’s strategy since 1988 is rather useless. Rather, it is necessary to discuss, in detail, the theoretical and policy implications of export-oriented industrialization and of the existing liberalization strategy. Altogether different theoretical and policy paradigms are necessary to suggest alternatives. Otherwise, imagine that liberalization strategy policy makers acknowledge the need of ‘interventions’, but admit at the same time that there are no resources to ﬁnance these programs, as a result of liberalization strategy’s priorities. Or, even worse, they design policies to make Mexico’s export-oriented ﬁrms, branches and regions more competitive. However, as discussed, Mexico, in the aggregate, is already ‘competitive’. Polarization, from this perspective, can only be overcome through an alternative to liberalization strategy’s rationality and growth path.

Notes 1 For an in-depth analysis see Hinkelammert (1984), Gómez (1995), Gutiérrez (1998), Dussel Peters (2000). 2 During 1940–81, GDP and GDP per capita increased annually by 6.1 percent and 3.3 percent, respectively, while employment also accounted for a positive development in absorbing most of the economically active population (Dussel Peters 1997). 3 As Aspe Armella (1993) stresses, lowering the inﬂation rate was the crucial targeted variable since high inﬂation rates, caused in general by domestic demand, but particularly by inertial tendencies of real wages, did not allow for improvements in the ﬁscal deﬁcit during 1982–7. 4 The Programa de Política Industrial y Comercio Exterior (1995–2000), for example, stressed that industrial policy could not be left to spontaneous market forces, but required an ‘active industrial policy’ (PEF 1996: 33). 5 For a full discussion, see Dussel Peters (2000). 6 Without doubt, it can be argued that these macroeconomic variables are not sufﬁcient for evaluating macroeconomic aspects, as reﬂected in any neoclassical textbook. However, issues such as employment, real wages, investments and savings, among other variables, are a consequence of the macroeconomic priorities according to EOI and liberalization strategy. 7 Data provided by Instituto Mexicano del Seguro Social (IMSS). This data only includes employment in manufacturing. 8 If we include maquiladora activities as part of the manufacturing sector, its export share increased from 59.71 percent in 1988 to 90.57 percent in 1998. 9 Mexico’s National Accounting System is divided in 73 branches, out of which

144

10

11 12

13

14 15

Enrique Dussel Peters

49 refer to manufacturing. This data does not include maquiladora activities and includes only the period 1988–96. Even in sectors such as automobiles, where real wages have historically been higher than for the rest of the economy and manufacturing, real wage levels, compared to manufacturing, fell from 246.28 percent in 1988 to 188.82 percent in 1996. See CEPAL (1998) and Dussel Peters (2000). Added to these studies, most of the qualitative analysis is a result of interviews with more than 20 computerrelated ﬁrms in Jalisco during 1997–9. According to ofﬁcial sources, the regional value-added of the sector is around 20 percent. However, these estimates include inputs of ﬁrms such as SCI, among many others, which imports most of its components and parts. Our estimates of value-added of this sector are signiﬁcantly below 5 percent. The Productive Chain for Electronic Industry (Cadena Productiva de la Industria Electrónica, CADELEC), has probably been the most important regional institution since 1995. CADELEC, supported by the regional government and electronic ﬁrms, attempts to develop supply chains based on the demand of existing ﬁrms in Jalisco. As discussed, there are similarities between EOI and neoliberalism. It could even be argued that EOI is an extension of neoliberalism. This, however, has to be analyzed in depth, which, so far, has not been done. It is not difﬁcult to imagine that economic polarization also results in social and political polarization. Social turmoil and guerrilla movements in regions such as Chiapas and Oaxaca, some of the poorest regions in Mexico, and an increasing per capita income gap with other regions since 1988 (Dussel Peters 1999), reﬂect some of these tendencies.

Bibliography Aspe Armella, P. (1993) El Camino Mexicano de la Transformación Económica, Mexico: Fondo de Cultura Económica. Balassa, B. (1981) The Newly Industrializing Countries in the World Economy, New York: Pergamon Press. Banco de México (1995) Informe Anual 1994, Mexico: Banco de México. Bhagwati, J. and Krueger, A. (1985) ‘Exchange control, liberalization and economic development’, in Bhagwati, J.N. and Grossman, G. (eds) Dependence and Interdependence, Cambridge: The MIT Press, 68–79. Brailovsky, V., Clark, R. and Warman, N. (1989) La Política Económica del Desperdicio, Mexico: UNAM. CEED/UDG (Centro de Estudios Estratégicos para el Desarrollo y Universidad de Guadalajara) (1997) Jalisco. Matriz Insumo-Producto, Mexico: UDG. CEPAL (1998) La Subcontratación como Proceso de Aprendizaje: el Caso de la Electrónica en Jalisco (México) en la Década de los Noventa, Santiago de Chile: CEPAL (LC/R.1808). Córdoba, J. (1991) ‘Diez lecciones de la reforma económica en México’, Nexos 158: 31–49. Dávila Flores, A. (1999) Globalización Económica y Diferencias Regionales en la Industria Manufacturera en México, to be published. Dussel Peters, E. (1997) La Economía de la Polarización. Teoría y Evolución del Cambio Estructural de las Manufacturas Mexicanas (1988–96), Mexico: Editorial JUS/UNAM.

Integration and polarization 145 Dussel Peters, E. (1998) ‘Recent structural changes in Mexico’s economy: a preliminary analysis of some sources of Mexican migration to the United States’, in Suárez-Orozco, M.M. (ed.) Crossings: Mexican Immigration in Interdisciplinary Perspectives, Cambridge, Mass.: Harvard University/David Rockefeller Center for Latin American Studies, 55–74. Dussel Peters, E. (1999) ‘Polarization of the Mexican economy: structural and regional aspects’, in Bailey, J. (ed.) Impacts of Trade and Integration on Subregions in Mexico and the United States, to be published. Dussel Peters, E. (2000) Polarizing Mexico: The Impact of Liberalization Strategy, Boulder, Colorado: Lynne & Rienner. Edwards, S. and Burki, S.J. (1995), Latin America After Mexico: Quickening the Pace, Washington, D.C.: World Bank. Foxley, A. (1988) Experimentos Neoliberales en América Latina, Mexico: Fondo de Cultura Económica. Friedman, M. (1962) Capitalism and Freedom, Chicago: The University of Chicago Press. Garrido, C. (1998) ‘El liderazgo de las grandes empresas industriales mexicanas’, in Péres, W. (coord.) Grandes Empresas y Grupos Industriales Latinoamericanos, México: Siglo XXI, 397–472. Gómez, R.J. (1995) Neoliberalismo y Seudociencia, Buenos Aires: Lugar Editorial. Gurría Treviño, J.A. (1993) La Política de la Deuda Externa, Mexico: Fondo de Cultura Económica. Gutiérrez, R.G. (1998) Etica y Economía en Adam Smith y Friedrich Hayek, Mexico: Universidad Iberoamericana. Hayek, A. (1981) Recht, Gesetzgebung und Freiheit (three volumes), München: Verlag Moderne Industrie. Hinkelammert, F.J. (1984) Crítica a la Razón Utópica, San José/Costa Rica: Departamento Ecuménico de Investigaciones. Krueger, A.O. (1997) ‘Trade policy and economic development: how we learn’, American Economic Review 87, 1: 1–22. Martínez, G. and Fárber G. (1994) Desregulación Económica (1989–93), Mexico: Fondo de Cultura Económico. O’Donnell, G. (1973) ‘Modernization and bureaucratic authoritarianism: studies in South American politics’, Institute of International Studies of the University of Berkeley. Modernization Series 9, Berkeley and Los Angeles: University of California Press. PEF (Poder Ejecutivo Federal) (1995) Plan Nacional de Desarrollo 1995–2000, Mexico. PEF (Poder Ejecutivo Federal) (1996) Programa de Política Industrial y Comercio Exterior (1995–2000), Mexico: PEF. PEF (Poder Ejecutivo Federal) (1999) Quinto Informe de Gobierno. Anexo, México: PEF. Ros, J. (1991) Mexico’s Trade and Industrialization Experience since 1960: a Reconsideration of Past Policies and Assessment of Current Reforms, paper presented at UNU/WIDER conference on Trade and Industrialization Reconsidered, August 31–September 3. Ruiz Durán, C. (1999) ‘Territorialidad, industrialización y competitividad local en el mundoglobal’, in Ruiz Durán and Dussel Peters, E. Dinámica Regional y Competitividad Industrial, Mexico: UNAM, Fundación F. Ebert and Editorial JUS.

146

Enrique Dussel Peters

Salinas de Gortari, C. and Mangabeira Unger, R. (1999) ‘Hacia un mercado sin neoliberalismo’, Reforma, January 20 and 21, 8A ff. and 18A ff. Valdés, J.G. (1995) Pinochet’s Economists. The Chicago School in Chile, Cambridge: Cambridge University Press. World Bank (1995a) Bureaucrats in Business, Washington, D.C: World Bank. World Bank (1995b) Labor and Economic Reforms in Latin America and the Caribbean, Washington D.C: World Bank. Zabludovsky, J. (1990) ‘Trade liberalization and macroeconomic adjustment’, in Brothers, D.S. and Wick, A.E. (eds) Mexico’s Search for a New Development Strategy, Boulder: Westreview Press, 173–97. Zedillo, E. (1994), La Propuesta Económica de Ernesto Zedillo. Palabras de Ernesto Zedillo Ponce de León, Candidado del Partido Revolucionario Institucional a la Presidencia de la República, in the forum ‘Crecimiento económico para el bienestar familiar’, Mexico City, June 6.

7

Liberalization policies and Mexico’s competitiveness Are technological capabilities upgraded or downgraded? Mario Cimoli

Ever since the 1940s when Mexican industry was recognized as the ‘engine’ for economic growth, manufacturing sectors have been growing under a regime of intensive protection; meanwhile, its orientation maintained a strong inward bias, at least until the 1982 ﬁnancial crisis. In contrast, the more recent period has seen a major shift of regime, involving both the macroeconomic setting and the set of incentives facing the economic actors. From the beginning of this liberalization period, industry has experienced a profound structural transformation, one of the major consequences being a steady process of internationalization based on a previously untested external performance. The observed industrial structure seems to be an overlapping mix of diverse inﬂuences that occurred during different phases of the nation’s development. Consequently, this analysis will look through the lenses of Mexico’s historical development. We will then try to interpret the recent structural transformation as expressed in the performance of various industrial sectors in a response to the different sets of stimuli that varied across three main historical periods. These different modes of development are analysed by their respective macroeconomic and institutional environment, the speciﬁc sets of industrial incentives and the regulatory regimes. The periods considered are Import Substitution Industrialization (ISI) in place from the 1930s to the mid-1970s, the transitional phase that lasted until 1988, and the last period characterized by the adoption of the exchange rate as a nominal anchor for ﬁghting high inﬂation rates (Exchange Rate Based Stabilization policy) and the further promotion of liberalization process. Sections 7.1, 7.2 and 7.3 move on to describe the main features that characterize each of these distinct periods. The issue at stake here is the liberalization process; indeed, our attention here is focused on the effects of the process of the economic and ﬁnancial opening and the incentives provided by the policies that accompanied it. Particularly, our aim is to understand whether the set of incentives that are responsible for the increased competitiveness of Mexican exports has upgraded or, rather, downgraded the local technological capabilities for endogenous innovation. We are interested in the different modes of

148

Mario Cimoli

technological efforts and learning capabilities of manufacturing ﬁrms, as deﬁned by R&D efforts, linkages within the local productive and institutional networks, and the ﬁrm’s ability to solve both technical and organizational problems. A deﬁnition of technological capability which captures the main features of innovation activities is associated with a complex of interactions and efforts that enable a country’s enterprises to efﬁciently adapt, improve and create technologies. While the ﬁrm is the fundamental unit of technological activity, the national capability is larger than the sum of the individual ﬁrms’ capabilities.1 The overall national capability is seen as the joint outcome of three levels of analysis presented in this work: iii the national level: the set of social relationships, rules and political constraints into which microeconomic behaviours are embedded (which has been extensively studied, together with the ﬁrst level, by evolutionary/institutional analyses);2 iii the meso-economic level of networks of linkages between ﬁrms and other organizations, which enhances each ﬁrm’s opportunities to improve its problem-solving capabilities;3 and iii the ﬁrm and sectoral levels: in which companies are seen as repositories of the knowledge embodied in their operational routines, which is modiﬁed over time by their higher level rules of behaviour and strategies (Cimoli and Della Giusta 2000). Section 7.4 reviews the recent debate over the effects that liberalization policies have on industrial competitiveness and the insertion in the international arena. In order to interpret these debates, Sections 7.5 and 7.6 describe and analyse the main stylized facts that characterize Mexico’s technological capabilities. Section 7.7 closes this chapter with some general observations and conclusions.

7.1 Import substitution period The ISI period is characterized by external protection and an active government role in promoting an inward-oriented industrialization pattern (i.e. directed mainly to domestic demand). Trade protection, credit and ﬁscal policies seem to have provided a favourable environment for the development of Mexican industry during the ﬁrst stages of import substitution, positively inﬂuencing the overall GDP growth rates. A long period with a stable exchange rate, price control in key sectors, and small budget deﬁcits encouraged national and foreign investment ﬂows toward the ‘infant’ sector. This model of industrialization, accepted broadly throughout Latin America, was characterized by strong protectionism coupled with policies to facilitate the entry of foreign technology. The industrial sector during this phase was characterized by the coexistence of private

Liberalization policies and Mexico’s competitiveness 149 domestic ﬁrms that produced consumer and intermediate goods, public sector enterprises that provided infrastructure, and MNEs in the relatively more advanced areas of manufacturing. A few very general comments about the ISI strategy can be made from literature on development: when development is seen as ‘essentially and ultimately a matter of learning and searching, of trial and error, in a context of continuous change’ (Bruton 1989), import substitution might provide a potentially powerful tool in order to shelter an ‘infant industry’ while such learning occurs. However, this is true if and only if the nation creates an environment in which learning occurs. Again in Bruton’s words, ‘there must be strong inducements, to search, to experiment, to test – to learn’. When applying this conceptual framework to Mexico’s experience, the ISI period appears to have been characterized by continuous support for technological efforts and learning that resulted from the combination of factors described above. The incentives involved in importing embodied technology within a closed environment stimulate the ability of adaptation, permitting it to develop into the more substantial activity of repair and improvement of the acquired technology (the so-called ‘opening up’ of technology packages), which have provided the following essential step in the process of building endogenous technological capabilities. Therefore, the overall result was that the industrial structure built under the incentives and institutional signals provided under the ISI phase appeared to have been able to support technology ﬂows and the transmission of technological information within networks of ﬁrms and institutions. These observations are also conﬁrmed by Katz (1987, 1995 and 1997) and Cimoli (1999), who stresses that an essential feature of technical change in closed economies has been endogenous in-house learning (producing incremental units of technical knowledge ‘on which to adapt and upgrade product designs, production processes and labour organization technologies’) and ‘disembodied’ forms of technical progress accounting for large parts of the observed productivity growth. Yet such activities have been insufﬁcient to reach the international technological frontier, essentially because the latter was shifting dramatically outwards due to the change in the technological paradigm that was taking place (see Freeman and Perez 1988).4 Finally, the learning pattern which we are interested in evaluating was marked, in essence, by efforts to adapt imported technology: the introduction/imitation of foreign machinery and equipment characterized the Mexican industrial sector since the ﬁrst stages of ISI, when labour intensive and supplier activities (such as textiles, foods, etc.) dominated the industry. This pattern continued during the 1960s when MNEs and foreign ﬁrms established themselves in sectors such as automobiles and car components, electronics, etc., and in the 1970s these efforts corresponded to the promotion of resource intensive sectors (petrochemicals in particular).

150

Mario Cimoli

On the other hand, the unfavourable environment for developing local R&D (high costs and the weak support of public sector) limited this activity to sectors dominated by oligopolistic ﬁrms that exploited their period of protection to increase their production capacity (such as glass makers and breweries).

7.2 The transition period Broadly speaking, the transition period was characterized by the abandonment of the ISI model. Beginning in 1976, major shifts in many macroeconomic indicators (the exchange rate and interest rates, public sector deﬁcit) were, in fact, accompanied by attempts to induce progressive change in the overall path of development. This change produced a confused set of incentives, particularly as far as industrial policy was considered, and this process accelerated in the second half of the 1980s (Ros 1981, Solís 1991). The transition period, which began with the 1976 devaluation and extended until the late 1980s, was characterized by macroeconomic turbulence, increasing inﬂation and massive capital outﬂow, particularly after the 1982 debt crisis (see Table 7.1). During this phase, the need to restore stability represented a major constraint, and the set of incentives for manufacturing, which combined trade liberalization, large devaluations, export promotion and sharp reductions in domestic demand appears to have been the result of a lack of coordination between industrial and macroeconomic policies. The adjustment following the 1982 crisis occurred through a contraction in domestic demand and large devaluations which helped in the recovery of the trade balance, particularly through highly price sensitive export commodities (such as chemicals, foods and textiles), but also via the reduction of domestic absorption. Low rates of growth and a disinvestment process (reﬂected in falling growth rates in net capital stock as well as in gross capital formation) which characterized manufacturing performance during the period, affected the sector’s ability to increase efﬁciency and productivity.5 The effectiveness of industrial policies and incentives is more difﬁcult to evaluate. However, the ability of speciﬁc areas (chemicals, automobiles and motors) to turn to the world market in the mid-1980s does appear related to export promotion measures during the late 1970s, and to the initial phases of the economic and ﬁnancial opening process allowed during the ﬁrst stage of the liberalization. In particular, the automobile sector’s performance is inﬂuenced by investment ﬂows from foreign companies during the late 1970s (Moreno-Brid 1994), which was favoured by the preferential regime of liberalization. The chemicals sector also beneﬁted during the same period from strong public sector incentives such as subsidized inputs related to the effort to exploit petrochemical opportunities, export incentives, protection, etc. (see Máttar 1994).

Liberalization policies and Mexico’s competitiveness 151 Table 7.1

Macroeconomic indicators

Years

GDP 1 rate 2

Inﬂation rate 3

Real exchange

Trade balance 4

1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998

– 8.5 0.5 3.5 3.4 2.2 3.1 1.7 1.3 4.1 5.2 4.2 3.5 1.9 4.5 6.2 5.1 6.8 4.8

– 28.70 98.87 80.80 59.17 63.74 105.75 159.17 51.66 19.70 29.93 18.79 11.94 8.01 7.05 51.97 27.70 15.72 18.61

100 87 251 150 122 160 111 94 93 92 80 72 67 64 63 80 72 66 64

254.8 323.1 587.2 1175.3 1098.7 699.8 418.3 732.3 217.5 33.8 73.5 606.6 1482.8 1127.9 1566.0 500.9 527.1 52.0 714.1

Source: INEGI (1985 and 1995) and Indicadores Económicos de Coyuntura. Notes 1 Average annual growth rate. 2 Consumer prices index. 3 Index (1980=100). 4 Average monthly balance (current millions of US dollars).

7.3 The Exchange Rate Based Stabilization and the new incentive framework The Exchange Rate Based Stabilization period (ERBS 1988–94) is, in this sense, less problematic in that the pursuing of a speciﬁc industrial policy was subordinated to macroeconomic priorities. In 1988 after the sexenio (six-year presidential term transition) and as a consequence of the 1986–7 crisis, the new administration opted for a ﬁrmer set of measures, founded on the idea that macro-economic stabilization and macroeconomic reform (the process of apertura) should be implemented simultaneously in order to reinforce each other. Accordingly, in 1988, Mexico established an exchange rate based stabilization programme, using the exchange rate as a nominal anchor as it was believed to be an effective instrument in ﬁghting high inﬂation. The overall effectiveness was subject to the condition that an in-depth economic reform, trade and ﬁnancial liberalization, and control of public sector deﬁcit would also be implemented.6

152

Mario Cimoli

According to the new administration, a credible shift of the macroeconomic environment toward stabilization (low inﬂation and low public sector deﬁcit) and the adjustment of relative prices should have attracted investments and cheap imported inputs, thus inducing a process of sectoral and microeconomic structural change. Manufacturers, in particular, were expected to experience a new export orientation through increased technological efforts and learning capabilities. 7.3.1 Industrial and trade policies The new macroeconomic context was reﬂected in a more ‘horizontal’, nonsectoral approach to industrial policy, formally deﬁned in the National Program of Industrial Modernization and Foreign Trade (Pronamice 1990–4): the role of government was redirected towards guaranteeing market functioning for resource allocation, thus providing a favourable environment in terms of macroeconomic stability and degree of competition. According to Pronamice, no speciﬁc measures had to be taken in order to carry out a planned path of sectoral growth, nor to change the pattern of investment from what market prices would generate.7 Up to 1994, only three sectoral programmes have persisted in sectors dominated by MNEs (automobiles, computers and pharmaceuticals) consisting in partial protection and duty drawback schemes for imports.8 With the Pacto de Solidaridad Economica (Economic Solidarity Pact) and its successive renovations, the trade liberalization process accelerated: in 1991 only 10 per cent of imports (mainly the petroleum and automobile sectors) were still protected by quantitative restrictions. The average tariff rate dropped to 13 per cent, with the maximum rate falling to 20 per cent (nearly 50 per cent of tariff items were subject to a 10 per cent tariff rate) and the number of different tariff rates was reduced to ﬁve. As a result, the dispersion of tariff rates went from 14 per cent in 1986 to 4.5 per cent in 1991 (Ugarte and Pérez Motta 1994.) Although it started as a unilateral process, the apertura at this time was accompanied by a number of measures aimed at creating a context of fair opportunities between domestic and foreign ﬁrms. These included more restrictive laws on property rights (1991), administrative simpliﬁcation for technology transfer (1990), the system against unfair trade practices (1993) and, more importantly, participation in regional trade integration agreements: NAFTA with United States and Canada, and ALADI with major Latin American countries (Clavijo and Valdiveso 1994). NAFTA, which went into effect on 1 January 1994, involves reciprocal trade liberalization between its three members: all tariffs are to be phased out over a 15-year period; further reductions of foreign (direct and portfolio) investment limitations, the establishment of a transitional commission dealing with antidumping and export subsidies and a dispute resolution mechanism to resolve conﬂicts in the application of the agreement (Blanco Mendoza 1994).

Liberalization policies and Mexico’s competitiveness 153 To summarize: promoters of the exchange rate based stabilization (ERBS) argue that a stable and open economic environment encourages the insertion in the international arena, thus contributing to an increase in overall competitiveness (see Aspe Armella 1993). Under this point of view, the horizontal setting for industrial policies (Pronamice) that accompanied the recent Mexican stabilization and supported the liberalization process should provide a more appropriate framework in order to upgrade technological capabilities, efforts and learning abilities. Policy makers assumed that no measure other than letting the market operate through adequate price signals should be taken in order to grant the optimal allocation of resources and technological upgrading.

7.4 The Mexican industrial performance and the corresponding analysis In fact, the implementation of the Pacto led to decreasing inﬂation, huge foreign capital inﬂow, an initial GDP expansion and progressive appreciation of the real exchange rate (see Table 7.2). During the 1988–92 period, manufacturing growth rate averaged 3.7 per cent, increasing both labour and capital productivity, despite the fact that the growth rates for employment and net capital stock decreased. In this phase, manufacturing also accounted for the largest part of the overall trade balance deﬁcit, accumulating US$83.8 billion, as a result of growing internationalization and an over-valued exchange rate. In the new setting, an expansion of domestic absorption contributed to an unfavourable relative price environment (as the price of non-tradable increased relative to that of tradable) and to an increase in imports causing (after the positive post-1982 period) growing trade balance deﬁcits.9 The incentives generated by Mexico’s new macroeconomic setting thus appear to have penalized the tradable sector, as demonstrated by the shift in foreign investments towards services (ﬁnancial services, communication and transportation) and their growing share of total net capital stock, reported in Dussel Peters (1996). According to some recent analyses on Mexican competitiveness, the country’s external performance is considered one of the main ﬁelds where we can appreciate the positive results of the major structural changes introduced after the 1982 crisis. Data on the overall volume and composition of exports during the 1980s and early 1990s are interpreted as an unprecedented success, percentage consisting of a remarkable improvement in Mexico’s strategic position and an overall increase in competitiveness. Within the overall increase of Mexican participation in the world market (export share growing from 1.49 per cent in 1986 to 2 per cent in 1994), manufacturing goods gained a new leading role. During the 1980s, manufactures’ export growth rate averaged 15.3 per cent, while the world market grew 7.5 per cent per year.10 Moreover, according to Ugarte and

15512 18804 18803 19678 20881 18715 13329 17056 16691 17707 20149 19489 19545 20927 23989 32469 38345 41407 48247

1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 19981

2519 2998 2503 3211 4232 4400 4657 5913 8242 9557 10415 11490 13268 15227 18206 20849 23963 28656 36605

Maquiladoras (C) 13.97 13.75 11.75 14.03 16.85 19.03 25.89 25.74 33.06 35.05 34.08 37.09 40.44 42.12 43.15 39.10 38.46 40.90 43.14

21089 25428 15065 10450 13735 15860 13845 15655 22813 26951 31227 36263 44130 45548 54991 48567 58069 69667 90203

19342 23343 13317 7960 10500 12554 10256 11072 16470 19719 23478 27711 34231 34090 40807 31019 38270 46617 59011

1747 2085 1748 2490 3235 3305 3589 4583 6343 7231 7749 8551 9900 11457 14184 17548 19799 23050 31192

Note 1 Preliminary data.

8.28 8.20 11.60 23.82 23.55 20.84 25.92 29.27 27.80 26.83 24.81 23.58 22.43 25.15 25.79 36.13 34.10 33.09 34.58

3058 3627 6240 12439 11378 7255 4141 7313 2120 314 662 5283 11317 9394 12797 4751 4239 396 5351

3830 4540 5486 11717 10381 6160 3073 5984 221 2012 3329 8222 14686 13163 16818 1451 75 5209 10764

772 913 755 721 997 1095 1068 1330 1899 2326 2667 2940 3369 3770 4022 3301 4164 5605 5413

Percentage Total (D) Manufacturing Maquiladoras Percentage Balance Balance Balance non-maquila (E) (F) (A)–(D) (B)–(E) (C)–(F)

Sources: INEGI (1995): Trade Balance of México; Banco de México, Economic Indicators. Note: Deﬂated by The Prices Index of US Manufacture.

18031 21802 21306 22889 25113 23114 17986 22969 24934 27264 30565 30980 32813 36154 42195 53318 62308 70063 84852

Manufacturing non-maquila

Imports

Trade balance (millions of US dollars at constant 1980 prices)

Years Total (A)

Exports

Table 7.2

Liberalization policies and Mexico’s competitiveness 155 Pérez Motta (1994) during the 1985–93 period, the volume of Mexican manufactured exports grew 225 per cent, while the ‘Asian tigers’ increased by 170 per cent and the aggregate industrialized countries by 20 per cent. Assuming that competitiveness can be measured through export performance, various authors associate the post-1982 opening process with the beginning of a new pattern of development in the Mexican economy, a structural change now based on a high degree of internationalization (see Casar 1994 and Máttar 1996). One methodology developed to exploit this concept of competitiveness is the so-called competitive matrix built with the following indicators:11 i

ii

the participation index, i.e. the share of a country’s exports in the world market classiﬁed according to products/sectors of production (Xi /Mi) and the contribution of these products/sectors in the world market (Mi /M).

Assuming this taxonomy, a country’s industrial sectors can be classiﬁed between two points in time on the basis of their export performance relative to that of the corresponding sectors at the international level. In particular, when a domestic sector’s export share is increasing in international markets, it will be considered as ‘Rising Stars’ (RS) or ‘Waning Stars’ (WS), depending on whether at the world level such a sector is dynamic (i.e. the market share is increasing more than the average of other sector’s shares). The competitive matrix can be represented as follows: Participation

RS

WS

MO

RE

Contribution

where MO ‘Missed Opportunities’ in dynamic sectors, and RE ‘Retreats’ from shrinking markets. The matrix has been applied to analyse the Mexican industrial structure and its export performance within the transition period in Casar (1994, 1996), Capdevielle et al. (1997) and Máttar (1996). Although these studies refer to different periods and samples, they seem to agree in detecting not only a growing participation of Mexican production in the world market, but also an increasing orientation towards those markets that are growing more than the average at the international level. The analysis proposed in

156

Mario Cimoli

Capdevielle et al. (1997) compare the performance of selected industrial activities between the 1960s and the 1980s.12 The initial results obtained in the research indicate that the share of sectors that increased their participation in the international market (that is, the sum of RS and WS) during the last decade reached 86 per cent of the activities considered, with the share of sectors classiﬁed as ‘rising stars’ growing from 26 per cent in the 1960s to 75 per cent during the 1980s. This result is conﬁrmed in Máttar (1996),13 where particular emphasis is also placed on the fact that overall exports increased their weight in the GDP by more than 30 per cent (with the share of manufacturing goods progressively overtaking petroleum in importance). Casar (1994, 1996) also strongly supports the thesis of a repositioning of the Mexican economy in the international arena through an analysis of revealed comparative advantages (RCA) and external performance indexes during the 1980–90 period.14 Casar stresses the fact that, within the RS group, various new exporters (such as automobiles, glass and chemicals) during the 1980s demonstrated a remarkable increase in revealed comparative advantages, augmenting their participation at the world level more than the aggregate of exports.15 Although the average RCA index for manufacture in 1990 is still quite low in absolute terms, and despite the fact that the manufacturing sectoral current account deﬁcit increased dramatically during the late 1980s–early 1990s, these improvements in relative terms were considered as symptoms of a clear trend toward a new pattern of specialization in manufacturing production and exports that would positively affect the Mexican industrial structure. The heterogeneity of the structural change in manufacturing during this phase has been underlined by various authors: Casar (1994) cites a growing specialization in automobiles and chemicals, while traditional manufactures such as textiles, metal products, machines and wood were losing importance. At a more detailed level of analysis, Dussel Peters (1996) shows how the areas with the highest rates of growth, and which account for 41 per cent of manufacturing export performance in 1988–92, have been the most dynamic since the 1970s; furthermore, these areas are characterized by being either MNEs (automobiles, electrical equipment), monopolies (basic petrochemicals) or national oligopolies (glass and beer). On this point, the conclusions of Capdevielle et al. (1997) are more cautious, particularly as a more complete concept of competitiveness is considered.16 The authors select industrial activities characterized as those with ‘actual competitiveness’ by combining the indicators of sectoral export performance and trade balance with those of production and rates of growth of productivity. The results seem to evidence that, among sectors that increased their participation in world markets, an important role is played by those that developed under the protection granted by the Import Substitution regime (such as fruit and vegetables, glass and copper

Liberalization policies and Mexico’s competitiveness 157 manufacture), sectors that underwent a slower liberalization (particularly the automobile industry) and in maquiladora activities (see Table 7.2). Capdevielle et al. (1996) thus argue that the maquiladora industry is a central actor in the new Mexican competitiveness. In fact, beginning in 1988, the importance of the maquiladora industries increased steadily: in terms of number of plants (whose annual rate of growth averaged nearly 10 per cent in the 1989–93 period), workers employed (nearly 20 per cent of total employment in manufacturing) and value added. The maquiladoras phenomenon poses a problem in the present perspective, considering the effects of their presence on the dynamics of local industry. As far as the creation of backward linkages is concerned, Brannon et al. (1994) observe that there is a problem in the lack of local sourcing by maquiladoras, which make very limited purchases of Mexican material inputs. There appear to be several reasons for this, among which the authors include: 1

2 3

4 5

the characteristic inward orientation before the liberalization, which hampered their development in terms of both scale of operations and quality of products; inadequate access to credit and ﬁnancial capital for medium and small ﬁrms during the economic crisis of the 1980s; the lack of a skilled human capital base, coupled with the tendency for the existing human resources to move within maquiladoras rather than set up their own companies or enter Mexican ﬁrms; materials shortages and; an insufﬁcient number of national independent consulting ﬁrms.

7.5 The globalization of Mexican industry: towards a modernized assembly plant After the trade reform, Mexico has substantially increased its participation in the world arena, in terms of exports as well as imports (see Table 7.2). Most of the surviving and efﬁcient ﬁrms (both MNCs and large domestic ﬁrms) have increased their exports of ﬁnal commodities (components for automobiles, chemicals, plastic products, glass, beer, electronics, steel, cement, etc.) and imports of intermediate and capital goods. The image that we have is that Mexico is a country in which production activities are highly globalized and that a new specialization in the global chain of production is emerging. Our analysis of Mexican globalization involves both foreign and domestic ﬁrms. First, we will examine the ﬁrms and industries dominated by foreign ﬁrms (FF, subsidiaries and joint ventures); second, we will study the ﬁrms and industries under the control of domestic ﬁrms (NF, including conglomerates and minority foreign participation.) This distinction will be applied to analyse production capacity and technology efforts in recognition of the kind of interaction expected to occur within and between

158

Mario Cimoli

certain types of industries and also between ﬁrms located in Mexico and abroad. As a ﬁrst step, particularly based on the results obtained in Capdevielle, Corona and Hernández (2000), Cimoli et al. (1998), Unger and Oloriz (2000) and Cimoli (2000) we attempt to substantiate the following general stylized facts: 1

2

3

4

The composition of industry is heavily concentrated in traditional activities: the suppliers of traditional goods and scale intensive groups of sectors. In contrast, science based and specialized producers of equipment and capital goods account for a very small fraction of Mexico’s manufacturing GDP: 8.8 per cent and 2.2 per cent respectively (compared with percentages above 15 per cent for each group in the United States). As mentioned earlier, Mexico’s R&D efforts are rather poor in comparison to those at the technological frontier. Moreover, R&D is highly concentrated in the export sectors (automobiles, glass, cement, ofﬁce machinery and computers, electronic equipment, etc.). R&D effort principally focused toward addressing the modernization of production processes and improvements in production organization and product quality. As for sources of new technological knowledge, the vast majority of ﬁrms rely almost exclusively on their internal sources. Regardless of the sectors in which ﬁrms operate, they have not developed co-operative R&D efforts with other ﬁrms and institutions. Furthermore, in none of the technological sectors have ﬁrms made signiﬁcant expenditures on R&D, except for those that are export oriented, and these ﬁrms principally have invested in improving processes, organization or quality. As a matter of fact, the pattern of R&D efforts – which have been scarce and scattered – and other modes of technology transfer have been mainly dominated by a higher integration of imported inputs in most competitive sectors. The interaction between ﬁrms and the institutions that produce knowledge is very poor, a fact which is most keenly felt by those companies belonging to the science-based sector. The results show that domestic ﬁrms consider internal sources of knowledge as more important for their innovative activities than external sources. Within the production system the activities of engineers, technicians and the experience of the labour force constitute the most relevant sources of knowledge, particularly for ﬁrms within the scale-intensive and science-based sectors. According to ﬁrms, users are also an important source of technological knowledge, especially in the specialized suppliers and supplier-dominated sectors. Public sector or university research centres are not a relevant source of information for Mexican ﬁrms. This is a rather remarkable fact in the case of ﬁrms within the sciencebased sector, since this sector is strongly linked with such centres in the more developed countries. In Denmark, for instance, 29 per cent

Liberalization policies and Mexico’s competitiveness 159

5

of ﬁrms claim that universities are important sources of technological knowledge, whereas this percentage is only 7 per cent in Mexico. The FF are the most dynamic agents in terms of their response to the international competitive pressures during the last decade. However, both the FF and NF are more signiﬁcant exporters, activating different channels to incorporate foreign technology and remaining an important source of technology and production inputs. The FF and the larger NF rely extensively on foreign sources of advanced technology, though in doing this they spread the effects of globalization through the more competitive parts of Mexican industry. It naturally follows that technology is not a major concern for the FF or NF in most industries. Technology from the parent MN is available to the FF in Mexico, and most of the NF in traditional industry can import technology. The economic incentives to develop their own capability through R&D and more intensive technology transfers apparently do not exist.

Analysis of the recent Mexican industrial restructuring has shown that a new specialization pattern has emerged with particular and separate roles for the FF and NF. Industry has developed in a highly unbalanced industrial structure that is specialized in a few sectors. The NF dominates most of the traditional and resource based mature industries (mostly in the supplier dominated and scale intensive sectors such as food and beverages, steel, glass, minerals, cement and basic petrochemicals). In addition, the FF has a few sectoral and production chains, in which economic activities are coordinated and integrated by large MNEs across the geographic borders through highly integrated linkages in production and innovation processes (think of the automobile industry), and ‘maquiladora’ type of linkages which are highly integrated with intra-ﬁrm imports rather than with domestic production capacities (as in the modern science based sectors such as computers and telecommunications, which are led by foreign ﬁrms). Nonetheless, all types of ﬁrms have some integration with countries that lead in international trade and technological innovations, thus becoming dependent on imported technology, as well as on imports of the most technologically dynamic ﬁnal and intermediate products. In fact, the main changes are observed in the modes in which sectors and the type of ﬁrms (considering FF and NF) are inter-linked with foreign production networks and sources of technology. In particular, the pattern of R&D efforts and other modes of technology transfer has been mainly dominated by a greater integration of imported inputs in the most competitive sectors. The evidence shows that their local contribution to R&D efforts and interactions with local institutions is scarce and scattered. To summarize: the assembly of exports and import commodities dominates and explains the main pattern of technological ﬂows and sectoral linkages with the domestic economy. There are two main reasons why the

160

Mario Cimoli

dynamics of inter-industry ﬂows simply are not functioning to improve R&D efforts and linkages with the institutional framework: i) maquiladora operations dominate the production of science based components, thus allowing for very limited links and ﬂows to other domestic suppliers of intermediates; ii) the imported equipment used throughout the industrial system replaces the learning capability that could accumulate in specialized domestic suppliers of equipment in a well integrated industrial system.

7.6 Learning capabilities within ﬁrms and sectoral speciﬁcities On the basis of the sectoral analysis and from the results of several surveys17 of different types of ﬁrms, a taxonomy of technological efforts and sectoral learning capabilities is possible. From these, we can note various sectoral speciﬁcities and learning patterns (Table 7.3). Automobiles. This industry developed during the import substitution era, under the leadership of transnational companies, and during the most recent period it continued its privileged position as an industrial policy objective. During import substitution, it maintained a very capitalintensive dynamics, combined with high relative wages, even though production technology was not competitive at an international level. At the present time, in the context of a restructuring of production at an international level, the Mexican automobile industry has demonstrated export capacity while maintaining a relatively stable gap in productivity compared to the United States. Likewise, the composition of domestic production has been transformed in that Mexico now specializes in the most labour intensive segments and imports those inputs that require a greater technological intensity. During the import substitution process, the technological capacity that Mexico acquired was associated with adapting equipment and blueprints, as well as adapting design and products for the local market. Currently, companies are modernizing productive processes, increasing the degree of specialization and the dimension of scale to achieve international standards. The companies are developing local research capacity applied to production and improvement in organizational efﬁciency. They are making a signiﬁcant effort in the development of suppliers and inter-ﬁrm dealings, as well as in human capital. Chemicals. This industry had support in the form of speciﬁc protection destined to stimulate the development of national companies, thus limiting the participation of transnational ﬁrms during the import substitution period. Based on Mexico’s natural resources, production diversiﬁed with active government participation in development and direct production. During this period, an institutional system developed to support such activity, creating specialized institutions and the development of special-

Liberalization policies and Mexico’s competitiveness 161 ized human capital. In the current context, the industry has emphasized efﬁciency over diversiﬁcation of production. This sector makes an intensive use of capital and pays the highest wages of all the Mexican manufacturing. Nonetheless, despite institutional efforts, this industry’s expenditures on R&D are relatively low compared to international parameters. The gap in productivity compared to the United States has expanded for the group of activities in this industry. During the most recent period, efforts to develop technological capacity have been oriented toward R&D and personnel training. Beer. Under the leadership of domestic companies, the brewery industry is characterized by having made signiﬁcant efforts in terms of the technology incorporated in capital goods and human capital. This industry presents signiﬁcant economies of scale that were developed for the domestic market. Due to the characteristics of a homogeneous and tradable product it was relatively quickly and easily re-oriented toward the international market, particularly during the 1982–8 crisis years. This industry’s R&D expenditures are double that of the United States. The gap in productivity with the United States is stable. R&D is oriented toward production efﬁciency, improvements in quality, training of employed personnel and the merger of companies as mechanisms for learning at the international level, thus reducing costs and carrying out ‘external benchmarking’. Glass. Of all the activities we have selected, this is the industry with the best technological performance; furthermore, it is making a signiﬁcant effort in the development of human capital and in R&D expenditures. This sector has increased productivity of labour and all other factors. It also has a high level of patenting compared to the rest of Mexican industry (10 per cent of patents compared to its share of 2 per cent of value added). This industry made a considerable effort in the purchase of materials and equipment. It is the only one of the selected activities that has reduced the gap in productivity compared to the United States (1980–94). The sector is led by a domestic company that is notable for its outstanding efforts in R&D. It has recently modiﬁed its behaviour pattern toward an improvement in productive processes. Electronics. This industry is suffering due to a process that involves a signiﬁcant transformation in the composition of production, and the maquiladora (twin plant) production predominates in this sector, and even production destined for the domestic market showed a signiﬁcant reduction in the level of integration of domestically-made electronics. Prices dropped signiﬁcantly in relative terms, decreasing the importance of the added value in manufacturing. This sector is led by transnational companies, and its efforts are oriented toward a better quality improvement, costs reduction, changes in the organization of production and in training of the plant’s personnel. Textiles and footwear have domestic ﬁrms of diverse size. Both areas

Scaleintensive.

Scaleintensive and sciencebased.

Chemistry

Type of sector

Automotive

Sectors and their characteristics

Locally-owned ﬁrms (big, medium and small ﬁrms).

MNEs (from Japan, USA and Germany) and some local initiatives.

Type of ﬁrm

Oriented to the domestic market. Increasing production capacity. Highly-protected sector and national monopoly (PEMEX).

Highly-protected sector. Oriented to the domestic market. Large assembly plants. Localized in the centre of the country.

Sector’s characteristics (dimension, openness, etc.)

Adaptation of plants and blueprints. Strong engineering national capabilities. Training: effort (increased effort to obtain technical skills of the personnel at the ﬂoor shop).

Adaptation of plants and blueprints. Adaptation of the local market of design and products. Efforts to improve organization of production. Increasing capacity to obtain full production (effort to increase economies of scale). Low-quality standard (producer–user in terms of quantity and prices only).

Technical change in the closed economy and type of learning

Import Substitution

Table 7.3 A taxonomy of sectoral technological capabilities

Higher participation in the international market (increasing exports). Increasing intra-sectoral trade. Increasing capital concentration. Increase of raw material prices produced by PEMEX.

Increase in the number of ﬁrms. Higher participation in the international market (increasing exports into the NAFTA area). More international integration (decrease in the linkages with local ﬁrms). Decrease in the number of domestic producers of parts (lower linkages with domestic ﬁrms). A change in the geographical distribution of ﬁrms. Protected sector. Competitive advantage based on static resource endowments (low labour cost).

Transformation after the crisis: new macropolicy and regulatory framework

Transition 1982–7

Increased R&D oriented on production efﬁciency for the ﬁrms, export oriented (quality, cost reduction, utilization of acquired machineries and equipment). Training: effort (increased effort to obtain technical skills of the personnel at the shop ﬂoor).

Some tentative of local R&D. Increasing production capacity and improving organizational efﬁciency process innovation product quality (increasing economies of scale). Increasing inter-ﬁrm’s learning effort (producer–user interplay based on quality, large participation of multinational producer of autoparts). Training: effort (increased effort to obtain technical skills of the personnel at the shop ﬂoor).

Type of learning: learning effort and inﬂuence of S&T system

New regulatory framework (after 1988) and NAFTA

Brewing

Supplierdominated and scaleintensive.

Locallyowned ﬁrms, duopolistic market.

Oriented to the domestic market. Increasing production capacity. High control of the domestic market through distribution channels (small shops and low cost returnable bottles).

Adaptation of imported capital goods and plants and effort to improve production process. In-house R&D primarily oriented to generate local knowledge (experimentation with new material and process of production). Taking part in conferences. and literature reviews. Interaction with foreign breweries. Training, hiring foreign technicians, engaging technicians on a temporary basis to solve technical problems.

Higher participation in the international market (increasing exports). Increasing economies of scale. Participation in ﬁnancial institutions.

R&D oriented on production efﬁciency (quality, cost reduction, utilization of acquired machineries and equipment). Reduction of the attendance to conferences and literature reviews. Increased interest in the publications of equipment suppliers. Internal and external benchmarking (benchmarking of effectiveness between plants and diffusion of the best internal practices within the ﬁrm). Use of mergers as a mechanism for learning from world players. Investment in equipment. Training: effort (increased effort to obtain technical skills of the personnel at the shop ﬂoor).

Scaleintensive.

Scaleintensive.

Consumer electronics (TV sector)

Type of sector

Glass container

Sectors and their characteristics

Table 7.3 Continued

MNEs (from Japan, USA and Europe).

Locallyowned ﬁrms, oligopolistic market.

Type of ﬁrm

R&D (both the activities of research and development were important). Literature review, patents and information search. Meetings to analyse experiences. Hiring experienced personnel from outside, personnel training. Joint ventures and making visits to competitors’ facilities. Manufacturing glass containers and equipment. Reverse engineering.

Technical change in the closed economy and type of learning

Highly protected sector. Low R&D. Large assembly plants. Localized in the centre of the country. Oriented to the domestic market.

Oriented to the domestic market. Increasing production capacity. Low-cost, returnable bottles used in beer and other soft drinks (possible economies of scale).

Sector’s characteristics (dimension, openness, etc.)

Import Substitution

Increase in the number of ﬁrms. More international integration (decrease in the linkages with local ﬁrms). A change in the geographical distribution of ﬁrms (more ﬁrms along the frontier with the USA: Tijuana and Sonora). Inputs are produced outside the country and ﬁrms belong to international intra-ﬁrm networks. Competitive advantage based on static resource endowments (low labour costs).

Higher participation in the international market. Increasing economies of scale. Participation in ﬁnancial institutions

Transformation after the crisis: new macropolicy and regulatory framework

Transition 1982–7

Effort to improve production efﬁciencies (quality, cost reduction and improvement in the organization of production). Qualiﬁed technician from the EU and Japan. Training: effort (increased effort to obtain technical skills of the personnel at the ﬂoor shop).

R&D oriented on production efﬁciency (quality, cost reduction, utilization of acquired machineries and equipment). External benchmarking. Acquired ﬁrms. Investment in equipment and reverse engineering. Training: effort (increased effort to obtain technical skills of the personnel at the shop ﬂoor).

Type of learning: learning effort and inﬂuence of S&T system

New regulatory framework (after 1988) and NAFTA

Supplier- Locallydominated. owned ﬁrms (small and medium ﬁrms).

Footwear

Oriented to the domestic market.

Highly-protected sector and large contribution to GDP. High employment share in the local industry and small production scale. Oriented to the domestic market.

Technological background of entrepreneur and personnel, local tradition. Based on imported capital goods (investment in imported machinery and equipment) and adaption of plants. Creation of a centre (CIATEC) for technological support in the area.

Adaptation of imported capital goods. Adaptation of plants and production organization. Pirated design. Tentative linkages with local producers of machinery and equipment.

Higher participation in the international market (low labour cost and dumping practice). Effort to vertical and horizontal integration. High labour mobility. High concentration in Leon with an industrial district structure (micro, small and medium ﬁrms).

High impact of macro-crisis Diminution of sectoral investment. Reduction of sectoral employment and number of local ﬁrms in the sector. Family control of the ﬁrms and moderate increase of maquila and joint ventures. Increasing imports of machinery and equipment (domestic provision disappears). Competitive advantage based on static resource endowments (low labour cost). High institutional support from speciﬁc institution and local technology policy. Effort in a programme of quality. Based on imported capital goods (investment imported machinery and equipment) and adaptation of plants. Design imitation.

Joint ventures and foreign ﬁrms increase their efforts on quality control, just-in-time and commercialization. In family ﬁrms, inefﬁcient organization and administration (effort to increase economies of scale and organizational efﬁciency). Pirated designs and imitation (designing, printing and dyeing). Training: effort (increased effort to obtain technical skills of the personnel at the shop ﬂoor).

Sources: Aboites, J. and Soria, M. ‘Intellectual property and patenting technological specialization’, Lara, A. ‘NAFTA, technological regionalisation and specialisation in the consumer electronics – TV sector’, Vera-Cruz, A. ‘The brewing industry’, Dutrénit, G. ‘The glass container industry’, Martinez, A. ‘Learning dynamics in the footwear sector in Mexico’, Corona, J. M. ‘Competitiveness and dynamics of the textile sector’, Lara, A. and Corona J. M. ‘Automotive industry, NAFTA and technological specialisation’, Arvanitis, R. ‘The chemical industry: learning and competitiveness’ in the Report prepared for the European Commission, DGXII (Science, Research and Development), project Science and Technology Policies in Developing and Transition Countries: Reform and Technological Co-operation with Europe.

Supplier- Locallydominated. owned ﬁrms. Labourintensive.

Textiles

166

Mario Cimoli

registered a decrease in their contribution to added value, showing low and decreasing levels of capital intensity. They also pay the lowest wages. Their total and factorial productivity indicators have tended to fall, while the gap in productivity has expanded. At the present time, companies in this sector are making a considerable effort toward quality improvement and marketing, since they are subject to strong competitive pressure. Science and technology policies (S&T). When considering indirect incentives for industry, OECD (1994) observes how the lack of recognition of strategic sectors in industrial, and particularly S&T, policy has characterized the country’s new environment. The National S&T Plan contains rather generic measures and lacks speciﬁc indications. The OECD report suggests that there is a lack of understanding of the nature of technical change as a complex process requiring long-term transformations, especially in terms of the technological culture of entrepreneurs, scientists and technologists on one side, and policy makers on the other. Other scholars who more explicitly advocate a role for S&T policy in the new open environment include Nadal Egea (1995), who stresses that: comparative advantages are neither inherited nor even [are] a function of factor endowments: they are generated and moulded by S&T policy. Becoming an open economy does not entail renouncing S&T policy; on the contrary, the case can be made that open economies require a well-deﬁned S&T policy.18 The incentives provided by the most recent S&T policies and their fundamental role in building technological capabilities will bear fruit only in the long-term. Thus it is difﬁcult to judge their appropriateness as the recognition of the S&T infrastructure is still recent. A review of the literature seems to indicate that, at the institutional level, there are problems of coordination between the different bodies in charge of S&T policy. Moreover, a dichotomy between decision-making power and command over ﬁnancial resources has been emphasized (OECD 1994, Cimoli 2000). The authors who concentrate on Mexican S&T policies underline a weakness in the linkages (that are believed to be essential for the creation of technological capabilities) between S&T infrastructure and industry; a problem which also seems to exist between different industrial sectors and, within the sectors themselves, between different productive stages. In general, the evolution of industry seems to indicate a lack of effort in the creation of networks between ﬁrms – within a sector or between sectors – and the institutions. This situation is related to the type of specialization that has been followed, based on traditional natural resources, the use of less-qualiﬁed labour and the increased dependence on imported capital goods. This has been accompanied by the disappearance of the following sectors: capital equipment and, generally, specialized suppliers. In addition, these ﬁrms have also experienced a diversiﬁcation

Liberalization policies and Mexico’s competitiveness 167 process and an international integration, which reduced the contribution of domestic subcontractors in important scale intensive activities. The macroeconomic and trade policies have had a great deal of inﬂuence on the learning paths of each sector, as has the new regulatory framework introduced by NAFTA. The analysis derived from the evaluation of the different historical phases of Mexico’s development (from import substitution to the new regulatory framework) reveals that a process of modernization has indeed taken place. As part of this process, we see that a large number of ﬁrms in different sectors have disappeared. Moreover, the analysis of the sectors conﬁrms that the country captured a dynamic comparative advantage in reference to efforts related to modernization of production process, increasing learning capabilities – in terms of improving organizational efﬁciency and skills, and foreign technology absorption – and in improving production organization. Such processes can clearly be explained by the requirements to compete in international markets, in particular, quality, standard certiﬁcation, and the like. This modernization process, however, has occurred mostly within the ﬁrms; there has been a low level of communication with institutions and a small response to the prevalent S&T policies. That is, the learning and innovation process in the country responded mainly to price and market signals, and not to tailor-made S&T policies. The policies were, of course, designed with this goal in mind but were not properly articulated and lacked clear incentives, which were frequently adverse. These observations are supported by an important fact: ﬁrms across sectors have increased efforts to attain higher productivity and quality. Think of the case of the chemicals, brewery and glass containers sectors in which R&D activities have been mainly re-oriented to quality, cost reduction, standardization requirements and utilization of imported machinery and equipment. The learning pattern seems to be dominated by two types of ﬁrms: MNEs corporations and large domestic ﬁrms. Both types show higher capabilities to compete in the international market; however, their historical learning patterns are different. MNEs have beneﬁted from large incentives to set up production facilities within the country, due to trade policy and, more recently, to the NAFTA agreement. In particular for ﬁrms from outside the North American region (Japan, Europe, etc.), MNEs maintain strong links with their parent companies with regards to technological ﬂows; their learning path is mainly involved with adapting capital goods equipment, increasing production capacity and improving process innovation and quality standards. Moreover, an effort to increase inter-ﬁrm linkages (producer/user) and personnel training is present. Here we have to distinguish between MNEs in which production involves locally available inputs other than labour force, and the maquiladora industry. For example, one crucial difference between the two is the absorption of specialized human capital in some sectors such as the automobile industry, whereas in the maquiladora industry, reliance on specialized, local human

168

Mario Cimoli

capital is much weaker. On the other hand, the pattern of large domestic ﬁrms cannot be understood without considering their learning efforts during the import-substitution phase. During this period, these ﬁrms developed the required economies of scale to enable them to compete in the international market after the opening of the economy. This has implied the adoption of plans, blueprints and designs in the domestic market, as well as efforts to improve organization and increase production capacity. Take, for example, the case of some large groups in the chemical, brewery and glass containers sectors where companies have not only developed increasing production capacity, but they have also dedicated their R&D activities to reinforce their knowledge-base during the import substitution phase. However, both types of ﬁrms have beneﬁted during the recent period from technological inﬂows incorporated in capital goods, equipment and intermediate commodities. Two distinct patterns can be identiﬁed in the textile and footwear sectors. In textiles, joint ventures and maquila ﬁrms are increasingly gaining importance and compete in the international arena. Here, the learning process is mainly characterized by efforts in training, just-in-time production and better commercialization; in contrast, family ﬁrms, which comprise a large part of this sector, are characterized by inefﬁcient organization and administration. Their main efforts are concentrated in increasing economies to scale and organizational efﬁciency as suppliers in the domestic market. Moreover, the footwear case is one example in which regional technology policies and institutions have supported the ﬁrms’ learning process, thereby increasing their capabilities to obtain better quality as well as a greater integration among the ﬁrms within the sector and by establishing an industrial district.

7.7 Conclusions The liberalization process should provide an effective retooling of economic activities, by combining a favourable environment in terms of relative prices along with an improvement in incentives for technological upgrade. At the present time, it is particularly difﬁcult to evaluate the long-term effects of these policy instruments which have been created only in recent years and are designed to inﬂuence phenomena of an inherently long-term nature. This contrasts with the sometimes very brief horizon of most macropolicies. The country’s geographical position and its participation in the NAFTA are also extremely inﬂuential factors. However, on the whole, the impression derived from the present study is that Mexico is a country in which structural change is mainly focused on the consolidation of its international participation in the international division of labour as a ‘modernized assembly factory’. The system of local technological linkages across local sectors and institutions and technological efforts (R&D activities) that we propose as an

Liberalization policies and Mexico’s competitiveness 169 essential factor in a technological upgrade has been analysed. We have obtained some disappointing conclusions: one result of recent industrial evolution has been the discouragement of the creation of such a network and R&D efforts. This does not mean that the type of specialization which has been pursued is of a traditional type based on natural resources. Indeed, the analysis developed here suggests that Mexico has modernized its successful export sectors. The ﬁrst inhibiting factor is the increased dependence on imported capital goods and intermediate goods. This has been accompanied by the disappearance of certain sectors (metal working and specialized suppliers) that traditionally were associated with the production of incremental knowledge believed to constitute an essential ﬁrst step in the accumulation of technological capabilities. Second, there is a process of de-virtualization and international integration which ﬁrms have undergone, that has reduced the contribution of local subcontractors to important scale intensive activities.19 Third, static comparative advantages – based on cheaper labour cost – most likely provide an explanation for the competitive performance of most ﬁrms. With regards to the correspondence between the dynamics of learning and S&T policies, there does not appear to be any direct relationship between the observed results and the preceding policies, a fact which leads to the supposition that other elements have shaped the pace and direction of learning activities. In particular, the sectors in which technological upgrading has occurred (supported mainly by internal learning capabilities) are characterized by the presence of some large domestic ﬁrms, some of which have been favoured by incentives dating back to the end of the 1970s ISI period (chemical, glass) or, more recently, isolated protectionist practices even in the liberalized economy (automobiles). In this area, MNEs are becoming more important in explaining the current scenario, in which these ﬁrms have developed an escalating capability to adapt by introducing increasing improvement in production organization and quality.

Notes 1 Within this approach, the processes of technology acquisition, adaptation and learning need to be analysed by so separating the role of the different actors interacting in the process of innovation. The most ideally extended diffusion process for technology (the ‘distribution power’ of an economic system, as it came to be named recently) involves many ﬁrms, institutions and individuals accumulating, while they participate in these operations as suppliers, competitors, users, advisors, etc. 2 The foundations which account for the characteristics of national systems of production and innovation develop through to the ideas that ﬁrms are repositories of knowledge, that they are nested in networks of linkages with other ﬁrms and also with other non-proﬁt organizations (networks which enhance the opportunities facing each ﬁrm to improve their problem-solving capabilities). Also, there exists a broader notion (at a wider level of aggregation) of

170

3 4

5 6

7

8

9 10 11 12 13 14

15

Mario Cimoli

embeddedness of microeconomic behaviours into a set of social relationships, rules and political constraints (Granovetter 1985). In as much as it can be interpreted as an externality or economy-wide mechanism for the generation of knowledge, this level has been at the centre of new growth theories. The emerging picture seems to reproduce the general analysis of the learning patterns in Latin American ﬁrms over the ISI period in the different manufacturing sectors provided by Cimoli, Cingano and Della Giusta (1998), who bring together the typical ownership structure and a dynamic version of Pavitt’s taxonomy (Pavitt 1984). Thus, three fundamental factors have emerged, which have shaped and characterized learning: ﬁrst, the ﬁrms belonging to the specialized suppliers and scale intensive sectors concentrate efforts on organizational innovation and a prevalence of low economies to scale in the same industries; second, a predominance of discontinuous over continuous modes of production, essentially determined by the type of evolution undergone by family ﬁrms and subsidiaries of MNEs; third, ﬁrms have a high degree of vertical integration, which corresponds to a reduced presence of subcontractor networks. According to Casar (1994) the rate of growth of labour productivity averaged 0.81 per cent per year in the period 1981–8, while it was 3 per cent in 1970–81. Generally speaking, successively renewed stabilization plans (since 1988 referred to as the Pacto) combined orthodox and heterodox policy measures, such as tight ﬁscal and monetary control and income policy, privatization of public enterprises and de-indexation of nominal prices. Budget consolidation and trade reform in particular were to play a central role in promoting a structural change toward a more integrated and deregulated economy. Sectoral policies (including quantitative restrictions and tariffs, price control, minimum national content requirements) remained effective in automobiles, pharmaceutical and computer electronics. Nonetheless, as NAFTA went into effect, in 1994, the second and third sectors were abandoned, while for automobiles the programme must expire within ten years. For an accurate description of how a reformist industrial policy should be implemented, see Ugarte and Pérez Motta (1994). Public sector intervention should more generally concentrate on measures such as stimulating R&D expenditure, supporting small companies’ access to capital markets, improving skills formation and simplifying controls on technology transfers. Overall imports increased 46.5 per cent in 1988 and averaged 21.3 per cent until 1992 (with consumption goods accounting for the largest part of the increase). See Casar (1994) and Máttar (1996). The matrix was ﬁrst proposed by Fajnzylberg (1991). In the case of Mexico, imports by OECD countries M are normally used as a proxy of the world market. The industrial activities are reconstructed converting three-digit SITC data, and the selection of export oriented sectors is obtained by ﬁxing a minimum export share of 0.5 per cent as a cut off value. The analysis is there conducted on 236 categories of products (according to Standard International Trade Classiﬁcation) in the period 1986–94. For each country’s sector the revealed comparative advantages index is constructed comparing the sectoral and aggregate participation ratios of Mexican exports to the World Market (RCA (XMi /Wi) / (XM/W)). An index higher than one for a sector indicates specialization, as in that sector the country’s share is higher than the average. The index of the external performance contrasts exports to the amount of imports in each sector and is simply obtained by their ratio X i /Mi. As far as imports are concerned, the average X i /Mi ratio for the sector considered in the sample doubled during the 1981–92 period. In particular, some of

Liberalization policies and Mexico’s competitiveness 171

16

17 18 19

the best performers (i.e. situated well above the manufacture X i /Mi in 1992) at the end of the period had a really poor external performance in 1981. If the achievement of dynamic efﬁciency is the long-term goal to which Mexican industry aspires, not only the system of internal incentives, but that of the incentives provided by the international environment has to be considered as well. According to scholars like Micheli (1993), the desire to develop dynamic comparative advantages based on technological upgrading would contrast with the type of specialization on which the NAFTA is grounded: a sort of standard Hecksher–Ohlin–Samuelson type of specialization, based on static comparative advantages, with Mexico having its strength in cheap and relatively unskilled labour. See Aboites and Soria (1998), Arvanitis (1998), Cimoli, Cingano and Della Giusta (1998), Corona (1998), Dutrénit (1998), Lara (1998), Lara and Corona (1998), Martínez (1998), Vera-Cruz (1998). See the growing evidence on the aggressive S&T policies of developed countries reported by numerous evolutionary writers. Katz (1995) points out that the opening-up of the economy, in Mexico as in other major Latin American countries, appears to be associated with a shift towards assembly-like operations.

Bibliography Aboites, J. and Soria, M. (1998) ‘Intellectual property and patenting in Mexico’, in Lall, S. (ed.) Science and Technology Policies in Developing and Transition Countries: Reform and Technological Co-operation with Europe, DGXII, UE. Arvanitis, R. (1998) ‘The chemical industry: learning and competitiveness’, in Lall, S. (ed.) Science and Technology Policies in Developing and Transition Countries: Reform and Technological Co-operation with Europe, DGXII, UE. Aspe Armella, P. (1993) El Camino Mexicano de la Transformación Económica, Mexico: FCE. Blanco Mendoza, H. (1994) Las Negociaciones Comerciales de México con el Mundo, Mexico: FCE. Brannon, J., James, D. and Lucher, G. (1994) ‘Generating and sustaining backward linkages between maquiladoras and local suppliers in Northern Mexico’, World Development 22, 12: 1933–45. Bruton, H.J. (1989) ‘Import substitution as a development strategy’, in Chenery, H. and Srinivasan, T. (eds) Handbook of Development Economics, Amsterdam: North Holland. Capdevielle, M., Cimoli, M. and Dutrenit, G. (1997) ‘Specialisation and Technology in Mexico: a Virtual Pattern of Specialisation’, IIASA, WP Interim Report IR–97–016, May 1997 (presented at the workshop on ‘Transfer of Technology, Trade and Development: The Newly Industrialised Economies in the Global Competition’, Ca' Foscari, Venice, 1996). Capdevielle, M., Corona, J.M. and Hernández, C. (2000) ‘The description of the production system and the main sectoral technological patterns’, in Cimoli, M. (ed.) Developing Innovation System: Mexico in the Global Context, New York/London: Continuum-Pinter Publishers. Casar, J. (1994) ‘El sector manufacturero y la cuenta corriente: evolución reciente y perspectivas’, in Clavijo, F. and Casar, J. (eds) La Industria Mexicana en el Mercado Mundial. Elementos para una Politica Industrial, México: FCE.

172

Mario Cimoli

Casar, J. (1995) La Industria Mexicana a Principios de los Noventas, Santiago de Chile: Proyecto CEPAL/CIID, CEPAL. Casar, J. (1996) ‘Un balance de la transformacion industrial en Mexico’, in Katz, J.M., (ed.) Estabilización Macroeconómica, Reforma Estructural y Comportamiento Industrial. Estructura y Functionamento del Sector Manifacturero Latinoamericano en los 90, CEPAL/IDRC – Alianza Editorial. Cimoli, M. (1988) ‘Industrial structures, technical change and the modes of regulation in the labour market. The case of Latin American countries’, DRC Discussion Paper 60, Sussex: SPRU, University of Sussex. Cimoli, M. (ed.) (2000) Developing Innovation System: Mexico in the Global Context, New York/London: Continuum-Pinter Publishers. Cimoli, M., Cingano, F. and Della Giusta, M. (1998) ‘Modes of industrial development, S&T policies and competitiveness in Mexico’, in Lall, S. (ed.) Science and Technology Policies in Developing and Transition Countries: Reform and Technological Co-operation with Europe, DGXII, UE. Cimoli, M. and Della Giusta, M. (2000) ‘The nature of technological change and its main implications on national and local systems of innovation’, in Batten, P. and Martellato, D. (eds) Innovation and Regional Development, Boston/Dordrecht/ London: Kluwer Academic. Clavijo, F. and Valdiveso, S. (1994) ‘La politica industrial de México, 1988–94’, in Clavijo, F. and Casar, J. (eds) La Industria Mexicana en el mercado mundial. Elementos para una politica industrial, México DF: Fondo de Cultura Economica. Corona, J.M. (1998) ‘Competitiveness and dynamics of the textile sector’, in Lall, S. (ed.) Science and Technology Policies in Developing and Transition Countries: Reform and Technological Co-operation with Europe, DGXII, UE. Dussel Peters, E. (1996) La Industrializacion Orientada hacia las Importaciones. Teoria y Evolucion del Cambio Estructural de la Manufactura Mexicana, unpublished. Dutrénit, G. (1998) ‘The glass container industry’, in Lall, S. (ed.) Science and Technology Policies in Developing and Transition Countries: Reform and Technological Co-operation with Europe, DGXII, UE. Fajnzylberg, F. (1991) ‘Inserción internacional e innovaciòn institucional’, Revista de la CEPAL 44: 137–66. Freeman, C. and Perez, C. (1988) ‘Structural crises of adjustment: business cycles and investment behavior’, in Dosi et al. (eds) Technical Change and Economic Theory, London: Pinter. Granovetter, M. (1985) ‘Economic action and social structure: the problem of embeddedness’, American Journal of Sociology 91: 481–510. INEGI (1985) Estadisticas Históricas de Mexico, Vol. I and II. Mexico: INEGI. INEGI (1995) Banco de Datos BDINEGI, Mexico. Katz, J. (ed.) (1987) Technology Generation in Latin American Manufacturing Industries, London: Macmillan. Katz J. (1995) ‘Régimen de incentivos, marco regulatorio y comportamiento económico’, in Katz, J. (ed.) Estabilizacion Macroeconòmica, Reforma Estructural y Comportamento Industrial, CEPAL/IDRC – Alianza Editorial. Katz J. (1997) ‘Structural reforms, the sources and nature of technical change and the functioning of the national systems of innovation: the case of Latin America’, paper presented at the International Symposium ‘On Innovation and Competitiveness in NIEs’, Seoul, Korea, May.

Liberalization policies and Mexico’s competitiveness 173 Lara, A. (1998) ‘NAFTA, technological regionalisation and specialisation in the Consumer Electronics-TV Sector’, in Lall, S. (ed.) Science and Technology Policies in Developing and Transition Countries: Reform and Technological Cooperation with Europe, DGXII, UE. Lara, A. and Corona, J.M. (1998) ‘Automotive industry, NAFTA and technological specialisation’, in S. Lall, (ed.) Science and Technology Policies in Developing and Transition Countries: Reform and Technological Co-operation with Europe, DGXII, UE. Martínez, A. (1998) ‘Learning dynamics in the footwear sector: an industrial district case’, in S. Lall (ed.) Science and Technology Policies in Developing and Transition Countries: Reform and Technological Co-operation with Europe, DGXII, UE. Máttar, J. (1994) ‘La competitividad de la industria quimica’, in Clavijo, F. and Casar, J. (eds) La Industria Mexicana en el Mercado Mundial. Elementos para una Politica Industrial, Mexico: FCE. Máttar, J. (1996) ‘Desempeño exportador y competitividad internacional: algunos ejercicios CAN para Mexico’, Comercio Exterior 46, 3. Micheli, J. (1993) ‘Los paradigmas de la política económica y tecnológica’, in Micheli, J. (ed.) Tecnologia y Modernizacion Economica, Mexico: CONACYTUAM. Moreno-Brid, J.C. (1994) ‘La competitividad de la industria automotriz’, in Clavijo, F. and Casar, J. (eds) La Industria Mexicana en el Mercado Mundial. Elementos para una Politica Industrial, Mexico: FCE. Nadal Egea, A. (1995) ‘Harnessing the politics of science and technology policy in Mexico’, in Bastos, M. and Cooper, C. (eds) Politics of Technology in Latin America, London and New York: Routledge. OECD (1994) Reviews of National Science and Technology: Mexico, Paris: OECD. Pavitt, K. (1984) ‘Patterns of technical change: towards a taxonomy and a theory’, Research Policy 13, 6. Ros, J. (1981) ‘Economía mexicana: evolución reciente y perspectivas’, in Cordera, R. (ed.) Desarrollo y Crisis de la Economía Mexicana, México: FCE. Solís, L. (1991) La Realidad Económica Mexicana: Retrovisión y Perspectivas, Mexico: Siglo Ventiuno Editores. Ugarte, F. and Pérez Motta, E. (1994) La Politica Industrial ante la Apertura, México: SECOFI, NF, FCE. Unger, K. and Oloriz, M. (2000) ‘The globalization of the Mexican system and the role of foreign technology’, in Cimoli, M. (ed.) Developing Innovation System: Mexico in the Global Context, New York/London: Continuum-Pinter Publications. Vera-Cruz, A. (1998) ‘The brewing industry’, in Lall, S. (ed.) Science and Technology Policies in Developing and Transition Countries: Reform and Technological Co-operation with Europe, DGXII, UE. Verdoorn, P.J. (1949) ‘Fattori che regolano lo sviluppo della produttivita del lavoro’, L’Industria, 1, 1: 45–53.

8

Trade integration and sectorial productivity Flor Brown and Lilia Domínguez

Important changes in Mexican economic policy have transformed the economy from the previous import substitution model to one of export promotion by eliminating subsidies, making foreign direct investment and technology transfer regulations more ﬂexible, deregulating some services, privatizing public enterprises1 and opening the country to foreign trade. The objective has been to guide industrial ﬁrms toward strategies that would enhance productive efﬁciency and create more ‘ﬂuid’ behavior in the markets. The trade opening was undoubtedly the most important step, given its profound effects on the economy. Initially, the opening was planned to be implemented gradually; but the signing of the ﬁrst economic pact (Pact of Economic Solidarity) on 14 December 1987, whose objective was to lessen inﬂationary pressures, accelerated the process and within a few months the Mexican economy had become one of the most open in the world. For Mexican manufacturing, the crucial event was the change in the foreign trade regime that eliminated import licenses while dropping tariffs; at the close of 1987 almost all import licenses had been eliminated and tariffs had been slashed to a weighted average of 11.1 percent in 1993. The rules of the game changed and industrial ﬁrms quickly found themselves in a radically different environment. All these changes were enacted in the hope that exports and private investment would take the lead in expanding demand instead of the reliance on public expenditure, as had previously been the case. Increasing export capacity was supposed to boost the rate of output growth over the long run in a context of low inﬂation. Of course these measures would have a major impact on industrial competition, leading to the creation of a new industrial model. In the light of these macroeconomic changes, it is interesting to analyze the behavior of labor productivity and its determinants, particularly the impact brought by changes in foreign trade. The opening of the economy has brought a growing integration of the Mexican economy with the US and Canada. In this paper we will argue that the ﬁrms reorientation to increasing exports forced them to increase output efﬁciency, a result that

Trade integration and sectorial productivity

175

shows up in the overall growth of productivity. Our empirical evidence shows that, during 1984–94, productivity growth was highly differentiated by industrial branch, sector, and ﬁrm size. While some industrial branches showed considerable increases in productivity, others had different results. Something similar occurred in the different strata of ﬁrms grouped together by size. But in addition to the trade opening, other micro- and macroeconomic variables contributed to productivity growth. Before presenting the results of our analysis, Section 8.1 describes the essential features of Mexico’s industrial model. The intent is to describe the micro- and macroeconomic context during our period of study, 1984–94. Section 8.2 analyzes our estimates of the dynamic of manufacturing productivity, based on the Annual Industrial Survey by sector, branch and size of the ﬁrm. The results obtained from an econometric model to analyze the determinants of manufacturing productivity are discussed in Section 8.3. Section 8.4 presents our conclusions and some ﬁnal considerations.

8.1 The new Mexican industrial model A notable characteristic of the new Mexican industrial model is the unprecedented improvement in the quantity and variety of industrial transactions in foreign trade. Approximately until the beginning of the 1980s, fully a quarter of manufacturing trade was undertaken with products in which imports and exports occurred simultaneously in signiﬁcant amounts. Starting in the 1990s, half of all manufacturing foreign trade involved an intra-industrial type of specialization (Casar 1996). As expected, given Mexico’s location, growing foreign trade meant greater integration with the United States and, to a lesser extent, with Canada. Thus while Mexico’s total exports to those countries was 59.2 percent in 1984, by 1994 it was 87.4 percent. Imports from those countries remained at 71 percent throughout the period (NAFIN 1995). A second major change was the modiﬁcation in industrial structure and an upsurge of new forms of organization in the markets. There has been a growing tendency towards technical and economic concentration, and the market structure now has an important differentiation of products as a result of competition. Considering the ﬁve types of market distinguished by Steindl (1952),2 there is a clear movement of the markets towards differentiated oligopolies and concentrated differentiated ones. These increased their share of industrial GDP from 60 percent to 70 percent between 1984 and 1993 (Brown and Domínguez 1997). In these markets the share of foreign ﬁrms in gross output exceeds 30 percent and may reach 80 percent in some branches. GDP growth was superior to the industrial average and a high percentage of manufacturing exports and imports arise from markets. Concentrated oligopolies where leadership is shared among domestic

176

Flor Brown and Lilia Domínguez

and multinational capital – after state ﬁrms ceded this role at the beginning of the 1980s – maintained a constant share of industrial GDP at 15 percent. Their role in foreign trade is of lesser importance than the previous case; the share of exports in these markets’ branches was less than 3 percent, with the exception of the basic metals industry. Lastly, competitive industries and competitive oligopolies, where domestic capital predominates, decreased their share in industrial GDP, due to their slower growth of output. Although imports grew at faster rates than exports in most branches, during the period of this study the difference was much greater, in these markets, particularly in the competitive areas. In addition, export ﬁrms have begun weaving relationships among ﬁrms, with their customers and with their suppliers, thus creating zones where we are beginning to see productive specialization. For example, we can identify zones for the automotive industry that is implementing subcontracting networks in the north of Mexico (Ramírez 1997) and in the center of the country in Aguascalientes and Puebla (Brown 1998). These are incipient processes, but a certain vertical disintegration seems to have occurred in this industry during the period of our study. Thus manufactures that were being produced internally are beginning to be bought from other ﬁrms, either domestic or foreign. We have also begun to detect activity in other sectors that are establishing chains with suppliers, for example in the shoe industry (where a similar movement is underway) in the Bajío area (Brown and Domínguez 1997). Lastly, we should mention advances in the modernization process in various industrial sectors. Evidence from case studies points to technological transformations within a core of ﬁrms geared mainly towards exporting. Modiﬁcations can be detected within the area of production and ﬁrms’ management systems, with marked nuances by branch and size. For example, the introduction of microelectronics based technologies permits greater automatization. The use of new technologies is more frequent in the metal mechanical sector and among ﬁrms with foreign equity, but is not limited to these. In the food and chemical industries there are examples of domestic ﬁrms that have automated gradually, through the introduction of logical programming controls (LPC) (Pozas 1993, Micheli 1994, Domínguez and Brown 1998). But perhaps more extensive than changes in machinery are the changes in organization, implemented at ﬁrst by leading ﬁrms. Now most ﬁrms consider them indispensable, since mere investment in machinery is admittedly insufﬁcient. Following this trend, production and labor are constantly being reorganized to improve productivity, satisfy clients and improve labor relations. Changes include new layouts, production by work groups or production cells, just-in-time systems and total quality systems. Firms that used to handle only large orders, utilizing scale economies with a limited number of models, can now take on smaller orders, deliver them

Trade integration and sectorial productivity

177

with greater frequency and with a greater variety of products. The question is to what degree the evidence from case studies can be detected in the aggregate results for productivity. We address this question in the following section.

8.2 Manufacturing productivity 1984–94 Labor productivity grew at an annual rate of 4.2 percent between 1984 and 1994. Productivity growth was considerably high from 1988 to 1994, as Table 8.1 shows. An important characteristic of the behavior of productivity throughout this period is the difference between industries and branches, as seen by the coefﬁcient of variation in manufacturing (0.984). The sectors with the greatest average annual rates in productivity were: basic metal industries (7.5 percent), metal products, machinery and equipment (5.7 percent), and non-metal mineral products (4.7 percent). The lowest rates of growth were posted by other industries and the textile, apparel and leather industries (0.2 percent and 2.3 percent respectively). Within each industry there is considerable dispersion, as shown by the coefﬁcient of variation that appears in the last column of Table 8.1. The sectors with the greatest coefﬁcient of variation are: textiles, apparel and leather, and metal products, machinery and equipment. These coefﬁcients demonstrate the heterogeneity of labor productivity among the diverse industrial branches. Table 8.2 shows the productivity gap that was posted throughout the period among branches with high growth in their productivity indices, as compared to those with slower growth. High productivity industries tend to concentrate a high percentage of manufacturing exports, as Table 8.2 shows. These industries contributed

Table 8.1 Labor productivity Percentages

Coefﬁcient of variation

1984–7 1988–94 1984–94 1 Food, beverages and tobacco 0.2 Textiles, apparel and leather 0.1 Wood industries 1.0 Paper, printing and publishing houses 6.0 Chemical substances, cork, rubber and plastic 2.8 Non-metal mineral products 3.1 Basic metal industries 2.2 Metal products, machinery and equipment 5.5 Other industries 4.0 Industry total 1.8

6 4 6 3 6 8 11 7 1 6

3.0 2.3 1.9 3.4 2.3 4.7 7.5 5.8 0.2 4.2

0.591 1.494 0.008 0.128 0.625 0.848 0.396 1.422 0.934

Source: Calculations are based on data in ‘Annual Industrial Survey’, INEGI. Note 1 These are long run rates estimated from the following regression: log y c βt ; t time.

178

Flor Brown and Lilia Domínguez

Table 8.2 Labor productivity and exports

Automobiles Basic iron industries Cement Electronic equipment and appliances Animal food Household appliances Fertilizers Non-electrical equipment and machinery Canned fruit and vegetables Beer brewing and malt Total

Labor productivity (percentage above average)

Share of total manufacturing exports (percentage)

1984–94

1984–94

12.66 8.73 8.44 7.18 6.97 6.52 6.27 5.90 5.65 5.51

17.64 6.54 0.91 3.60 0.18 0.37 0.40 9.00 2.15 1.31 42.10

1.48 0.90 0.87 0.53 0.24 0.87 0.89 1.14 4.96

1.43 0.93 6.21 0.74 2.93 1.43 0.19 1.43 0.21 17.53

Industrial branches below average Non-metal products Garments Other chemical products Alcoholic beverages Other manufacturing industries Transport equipment Milled ﬂour and products Leather and footwear Metal structural products Total

Source: Calculations are based on data in ‘Annual Industrial Survey’, INEGI.

on average 37 percent of manufacturing exports from 1984 to 1994, and 50 percent in 1995. In contrast the share of exports of the low-productivity group was 14 percent in the period and 16 percent in 1995. 8.2.1 Productivity in different strata by size of ﬁrms The behavior of productivity in Mexican manufacturing has been very heterogeneous, not only among the diverse industrial sectors but also in different strata classiﬁed by size. In Table 8.3, we can see that giant ﬁrms maintained high growth rates throughout the period; the growth in productivity in the small and medium ﬁrms, by contrast, was highly sensitive to the macroeconomic instability that characterized the ﬁrst sub-period. On average the annual growth rates of labor productivity among giant ﬁrms was double that of the small and medium-sized ﬁrms.

Trade integration and sectorial productivity

179

Table 8.3 Labor productivity according to size of establishment: annual growth rate 1984–7 1988–94 1984–94 Number of Percentage establishments in total value added in the sample Total Small1 (50 to 100 workers) Medium (101 to 250) Large (251 to 500) Very large (more than 500 workers)

4.04 4.33

1.9

692

0.05

3.47 6.41

2.2

542

0.15

0.83 6.92

4.0

374

0.25

3.31 5.39

4.4

388

0.55

Source: Calculations are based on data in ‘Annual Industrial Survey’, INEGI. Note 1 Does not include establishments with less than 25 workers.

In summary, the growth of productivity can be explained by the high rates of growth that appear more often in the group of larger ﬁrms, in particular the giant ﬁrms. If we consider that a high percentage (60 percent) of exports originate from the 100 largest ﬁrms (Expansión 1994), these are likely to come from the group of large and giant ﬁrms. Thus we can state that a relationship exists between the high volume of exports and the growth of productivity in this group of ﬁrms. One of the worst recessions of the Mexican economy occurred in 1995, when industrial GDP dropped 7.7 percent. The question is to what extent this affected the growth of productivity. Starting in 1994, INEGI widened its coverage of the Annual Industrial Survey by including a greater number of ﬁrms and industrial groups.3 Hence, the estimates presented earlier are not totally compatible with those carried out in the new survey. According to the new survey, labor productivity grew by 2.6 percent between 1994 and 1996. In spite of these differences, we can conclude that the macroeconomic instability in these latter years seems to have had less effect on the dynamic of productivity when compared to the 1984–7 period in which the average growth of labor productivity was smaller (1.7 percent). In other words it would seem that ﬁrms undertook a learning process and were therefore able to adjust more rapidly to adverse situations. In sum, between 1984 and 1994, productivity grew rapidly. This fact focuses attention on the determinants of the growth and, particularly, on the inﬂuence of the trade opening in this outcome, topics to be taken up in the following section.

180

Flor Brown and Lilia Domínguez

8.3 Determinants of productivity4 In this section we analyze the determinants of productivity, beginning with the great heterogeneity within manufacturing, and in particular among large and small ﬁrms. The variations in productivity appear to depend on a set of interrelated variables, both micro- and macroeconomic, within which foreign trade variables have their place. Among factors related to the types of ﬁrms and their strategy that can help explain variations in productivity, those that stand out include matters analyzed in the literature on institutional approaches, innovation and industrial organizations such as: equity ownership, the different aspects of technological effort, advertising expenditures and economies of scale. The empirical evidence regarding the relationship among productivity and these variables is not always satisfactory. In the case of Mexican industry, Clavijo (1992) and Venables and van Wijnbergen (1993) show that the country of origin of investment is a factor in explaining how productivity changed. Weiss (1992) estimated an econometric model to deduce the effects on labor productivity of changes in policy, technology and the industrial structure, among other variables. The following variables were not found to be statistically signiﬁcant in his model: index of concentration, the intensity of advertising, and the capital–output ratio. Tybout and Westbrook (1995) discuss the change in efﬁciency and the returns to scale by analyzing the components of productivity. Other studies consider microeconomic aspects such as proﬁt sharing with workers as an incentive for increasing labor efﬁciency, interest paid by the ﬁrm and the proﬁt margin. Oaks (1994) points out the importance of proﬁt margins as a source of internal funds to invest in technology and thus increase efﬁciency. Other authors have pointed to the impact of the macroeconomic environment in innovative capacity, productivity and the ﬁrms’ efforts to introduce technological changes (Tybout et al. 1991, Venables and van Wijnbergen 1993, Casar 1996, Katz 1996). They have noted that lower output growth rates tend to reduce the size of local markets, the volume of production and, further, make it more difﬁcult for ﬁrms to enter international markets. Katz (1985) points out that periods of macroeconomic instability can lead to a search for short-term proﬁts and dampen long-term proﬁt strategies based on technological effort. Here empirical evidence shows the impact of market growth on the behavior of productivity in manufacturing. Casar et al. (1990) shows that the productivity of industrial branches in the 1970s and 1980s was related to market growth. Likewise, Weiss (1992) shows that the increase in output is a key variable in explaining productivity growth, whether it be labor productivity or total factor productivity (TFP). One macroeconomic variable that could explain the behavior of productivity is the difference among countries’ trade policies. According to Dahlman, Ross-Larson and Westphal (1987), the entry of new competitors

Trade integration and sectorial productivity

181

in local markets has prompted ﬁrms to install new technology, make changes in machinery and in the organization of the ﬁrm. Therefore as a country participates more fully in international trade, there should be a positive effect on productivity. Yet up to now there is no conclusive empirical evidence. This chapter approaches the determinants of productivity through an analysis of an interindustrial nature, in the manner of Weiss (1992), to examine macroeconomic variables together with the industry’s structural variables. Although these models do not allow us to estimate structural long-term parameters, they are useful for describing general tendencies.5 The dependent variable of the proposed econometric model is labor productivity (LP), while independent variables included are both those speciﬁc to the nature and strategy of the ﬁrm (ﬁrm size, technological effort, advertising expenditures, proﬁt margin, economies of scale), as well as those related to economic policy (import penetration, exports and market growth). We now present the theory to support the inclusion of this variable in the model speciﬁed. In industrial organization theory, concern has traditionally centered on explaining the determinants of the proﬁt margin (Bain 1956, Weiss 1974, Caves and Uekusa 1976), given the conception that structure determines conduct and results. Yet it is also valid to analyze the impact of proﬁts on other behavioral variables, as these are an expression of the efﬁciency of production, of market control and availability of internal resources. Large ﬁrms that enjoy certain technological advantages and market control can obtain high proﬁt margins to ﬁnance their growth and modernization process with their own resources (Eichner 1976). In other words an interaction exists between structure, conduct and results, as the literature on the new industrial organization maintains (Jacquemin 1992). Economies of scale arise from the technical requirements vis-à-vis the optimal minimum size, the lessening of risks from an enlarged ﬁrm size, and the increase in their bargaining power in factor or product markets. This can lead to savings in costs of purchasing materials and components and can enhance ﬁrms’ proﬁts. Large ﬁrms can also gain advantages in large-scale and collective hiring of labor, they can obtain credit advantages and take risks in introducing new products (Silberston 1972). However, economies of scale do not by themselves guarantee lower unit costs relative to the size of the market. Cost disadvantages must also necessarily exist at the lowest levels of production. These cost disadvantages translate into penalties for ﬁrms with a less-than-optimal size (Gold 1981). Therefore we would expect a differentiated impact between the cost penalties and the productivity of large and small ﬁrms: there would be a positive effect in the former, and a negative one in the latter. Innovative ideas are the result of internal activities that are organized in the ﬁrm by on-the-job experience, i.e., from the technological effort. Schumpeter stated that the innovative process can become more

182

Flor Brown and Lilia Domínguez

important than price competition itself, because it is the most effective means of gaining an advantage over competitors. This variable is linked, in the ﬁrst place, to the ‘hard’ technology component incorporated in the ﬁrm, in the form of the new machinery purchases. Two advantages ensue. First, machinery incorporates technological innovations so as to produce more efﬁciently and, second, the machinery can induce organizational innovations, or at least be an important catalyst for these.6 Thus, a high rate of investment is necessary to realize the advantages of technological change as incorporated in capital. Second, the soft technology component is related to the expense in technological development which also has a positive impact on productivity growth (Dutrénit and Capdeville 1993), and also to the expense of acquiring patents and licenses that can encourage the transfer of disembodied foreign innovations. Other authors point to the effect that technology licenses have in reducing costs by eliminating expenses in research and inefﬁcient development (Gallini 1984). Apart from the discussion of whether advertising expenditures favor consumers or not, it is a fact that ﬁrms must expend resources to raise demand through advertising and sales campaigns (Hay and Morris 1986). This strategy permits ﬁrms to take advantage of economies of scale and learning, which can have a positive impact on productivity. Diverse macroeconomic policies have an impact on the behavior of productivity by, for example, inﬂuencing demand. Verdoorn (1949) pointed out that industrial productivity is related to the growth of output not only because of its economic stimulation, but also due to the learning that takes place with a greater accumulative output. Thus, economies of growth can ensue, as Penrose (1959) mentions, from a set of variables internal to a ﬁrm that make it highly proﬁtable for companies to expand along certain lines. This would suggest a procyclical behavior of productivity. Fay and Medoff (1985) state that in recessions there is a tendency to maintain more workers on the payroll than would be technologically necessary, and therefore when output drops, productivity drops, and vice versa. Firms tend to retain their workers because the cost of varying the labor force in response to changes in demand can imply high transaction costs, in addition to the fact that incumbent workers possess a set of skills that are not easy to replace. For Caballero and Lyons (1992) the procyclical component of productivity has nothing to do with the variation in inputs, but rather arises from externalities, not strictly in the Marshallian sense, but rather from the market, due to the higher frequency of transactions among economic agents during an upturn in the economy.7 As an approximation of market growth, we included the total product for each industrial branch in the estimates of the econometric model. Several authors have posited that an important relationship exists among different strategies of industrial development and productivity growth, especially among those that emphasize supplying the domestic

Trade integration and sectorial productivity

183

market or those oriented to foreign markets. A fundamental change occurred in the trade regime in Mexico between 1984 and 1994 that consisted in lowering the protection that import licenses and tariffs had given domestic production8 (Ten Kate and De Mateo 1989). According to the theoretical analysis, these changes in the foreign trade scheme probably inﬂuenced productivity in a number of ways. First we should mention the increase in internal efﬁciency and the ‘challenge-and-response’ mechanism brought on by the competition that arose when Mexico’s trade doors opened (Nishimizu and Robinson 1986). Firms were forced to adopt new technologies, reduce X inefﬁciency and costs. Therefore, following this argument, a positive relationship can be expected between export expansion and labor productivity (LP). The second possibility has to do with the beneﬁts from widening the market through trade that lead to a drop in unit output costs, by expanding demand through exports. Insofar as the relationship between imports and productivity increases is concerned, there may be contradictory effects. On the one hand import growth can reduce the internal market and dampen interest in increasing productivity. At the same time the trade opening increases competition and thus induces greater efﬁciency. The variables used in estimating the model to analyze the effects of the changes in the trade regime and productivity are exports and the coefﬁcient of import penetration (ratio of imports to domestic demand) in each industrial branch. The model used in the estimates is the following: LP 0 1MAQ 2TEC 3PUB 4MARG 5PEN 6GDP 7IMP 8EXP et where a

b

microeconomic variables are9: Investment in machinery, ofﬁce and computer equipment (MAQ) Technology transfer expenditures (TEC)10 Sales and advertising expenditures (PUB)11 Proﬁt margin (MARG)12 Cost disadvantage for producing at a less than optimal size (PEN)13 macroeconomic variables are14: Gross domestic product: (GDP) Imports penetrations coefﬁcient (IMP) Exports (EXP).

To take into account the heterogeneous nature of the Mexican economy, we carried out an analysis that differentiated between large and small establishments. The econometric model was estimated for each stratum by size. Information from each of the groups of establishments was organized with its respective industrial branches through a panel analysis, due to the

184

Flor Brown and Lilia Domínguez

advantages that are derived from these types of estimates.15 The results from the estimations are listed below. 8.3.1 Large and small establishments Large establishments are those that, in 1988, declared a gross output value of three million dollars or more, while small ones fall below this ﬁgure. As Table 8.4 shows, the estimation of the econometric model for the large group found that the following macroeconomic variables were statistically signiﬁcant: the proﬁt margin (MARG), investment in machinery (MAQ), and advertising expenses (PUB), as well as GDP and exports (EXP). Insofar as the calculation of elasticities is concerned, the impact on the proﬁt margin, the GDP and exports stand out. With respect to the results obtained for small establishments, Table 8.5 shows that the signiﬁcant variables were the gross proﬁt margin (MARG), investment in machinery and equipment (MAQ), expenses in technology transfer (TEC), the GDP, exports (EXP), and imports (IMP). The greatest elasticities also correspond to the proﬁt margin, the GDP and exports. As expected, the results show that the opening of the economy has been an important factor in the behavior of productivity. Other variables, both microeconomic and macroeconomic have also been important. With respect to the microeconomic variables, the hypothesis regarding a link between proﬁt margin and productivity held true in all cases. This result demonstrates the importance of having internal funding for the growth of productivity, and suggests that if ﬁrms had access to alternative sources of funding at competitive rates, they would be able to compensate the general lack of credit and raise productivity. Likewise, empirical results conﬁrm the importance of organized internal activities in the ﬁrm,

Table 8.4 Large establishments Dependent Variable: labor productivity Variables Coefﬁcient T-statistic Probability Elasticity

Constant 2.396 0.399 0.690

MARG. 0.732 14.702 0.000 58.86%

R2 0.64

F-Statistics 123.20

Durbin-Watson

Test Probability

JB 0.05

LM(3) 0.00

LM(21) 0.30

MAQ 0.004 1.563 0.107 0.98%

PUB 0.004 4.158 0.000 1.64%

PIB 0.306 7.479 0.000 31.77%

EXP 0.015 4.128 0.000 3.70%

Arch(2) Arch(3) Reset 0.34 0.52 0.40

Note 410 observations based on 41 industrial branches in a ten-year period.

DUMI 29.315 10.385 0.000

Trade integration and sectorial productivity

185

Table 8.5 Small establishments Dependent Variable: labor productivity Variables Coefﬁcients T statistics Probability Elasticity

Constant 11.205 0.942 0.347

MARG 0.436 8.159 0.000 39.34%

R2 0.368658

F-statistics 18.10

Durbin-Watson stat 1.81

Test Probability

J.B. 0.49

LM(3) Arch(2) ARch(3) Reset 0.56 0.16 0.25 0.07

LM(2) 0.43

MAQ 0.005 3.043 0.003 1.63%

TEC 0.015 1.774 0.078 1.51%

PIB 0.258 2.917 0.004 31.18%

EXP 0.046 3.878 0.000 9.72%

IMP 0.004 2.914 0.004 2.29%

DUMI 19.175 3.961 0.000

Note 230 observations based on 23 industrial branches in ten-year period.

i.e., its technological effort, in particular the purchase of new machinery as a fundamental condition for realizing the advantages of technical change embodied in new capital. From macroeconomic perspectives, the link between GDP and labor productivity, both in large and small ﬁrms, conﬁrms results by Weiss (1992) and Casar et al. (1990). In other words, market growth is a key factor in explaining the dynamic of productivity: it is not possible to conceive of a modernization process without the incentive provided by market growth. As opposed to Weiss (1992) and Tybout and Westbrook’s (1995) inconclusive results of a link between trade opening and productivity, our analysis permits us to gauge this impact. Exports seem to have a greater impact than imports. In any event, although the impact of the variables associated with the opening is not as strong and less systematic than market growth, a positive link between labor productivity and these latter variables shows that an external learning effect exists, brought about by foreign competition, as has been suggested by Leibenstein’s concept of x-efﬁciency. Various interesting differences exist between the two estimated groups that are worth remarking on. Within the group of large establishments, expenses on technological transfer was not a statistically signiﬁcant variable. Two explanations seem plausible. First, such expenses might already have been present as a policy of large ﬁrms before the period of this study, and therefore their impact is not notable in a rise in productivity; and among the small establishments their concern for paying for technology may be recent. Second, insofar as the expense of technology transfer consists of payment for brand names and patents, basic engineering, and technical assistance, the expense among large establishments is skewed towards the former components, while among small establishments the payment of basic engineering and technical assistance is more important.

186

Flor Brown and Lilia Domínguez

Therefore we can expect these expenses to have a greater impact on the increase in productivity in the small establishments. Finally, advertising expenditures are not signiﬁcant in the group of small establishments, which is understandable to a certain degree, given the economies of scale that typically exist therein.

8.4 Conclusions In the light of these macroeconomic changes, in this chapter we have endeavored to analyze the behavior of labor productivity and its determinants. We have shown that the new Mexican industrial model has changed behavior patterns among manufacturing ﬁrms: they are seeking efﬁciency as measured by a clear increase in labor productivity. However, beyond the discussion on the relative size of this increase, most estimates show signiﬁcant differences in productivity performance among industrial branches and different-sized ﬁrms, a fact suggesting that the modernization process has been highly selective. Thus the need to take into account the heterogeneous nature of the Mexican economy, in the analysis of the determinants of productivity. As expected, our econometric estimation shows that the opening of the economy has been an important factor in the behavior of productivity. In particular the association between exports and labor productivity conﬁrms the relation of integration into foreign trade and searching for greater efﬁciency. Nevertheless other variables, both microeconomic and macroeconomic, have also been important: the proﬁt margin, investment in machinery and ofﬁce equipment as well as GDP stand out. The analysis by strata of size also turned up evidence of differentiated impacts of the variables included in the model, such as advertising and technology transfer expenditures and imports. In addition, it should be noted that, while the model is highly satisfactory in the case of large ﬁrms, its capacity to explain the variables when dealing with small establishments is much less. These differences point to a greater heterogeneity in small ﬁrms in terms of their technological abilities and their ﬁnancial resources, which affects their ability to integrate into foreign markets.

Notes 1 The share of gross output in manufacturing of public and state ﬁrms fell from 16.3 percent in 1988 to 10 percent in 1993 (INEGI Industrial Census 1988, 1993). 2 Competitive markets, competitive oligopoly, differentiated oligopoly, concentrated oligopoly and concentrated and differentiated oligopoly. 3 Whereas in the ﬁrst sample the protection for small establishments was very limited and these establishments are picked up well in the second. 4 A full version on this topic is found in Domínguez and Brown (1998). 5 While interindustrial models have been questioned yet, as Schmalensee (1989)

Trade integration and sectorial productivity

6 7 8

9 10 11 12 13

14 15

187

shows, they are useful for describing overall tendencies or ‘stylized facts’. It is in this sense that the results of the proposed model should be considered. Various case studies on the use of new technologies in Mexican industry have shown the complementary nature of new machinery and organizational changes (see Mertens 1995). The authors cite Diamond (1982) who gives evidence regarding the volume of transactions among ﬁrms and their customers as a key factor in the transmission of external factors in the short term. The protection of import license went from 100 percent in 1983 to 9.1 percent in 1991, and the average weighted tariff went from 16.4 percent to 10.5 percent in the same period. Of the 39 industrial branches that make up the study, only 15 had some percentage of output under license in 1988. Estimated from the Annual Industrial Survey of INEGI. The Annual Industrial Survey lists only expenditures of technology transfer. Expenditures in publicity and sales campaigns. Gross product–inputs–the wage bill/gross product. PEN refers to the ratio of labor productivity in the largest ﬁrms producing 50 percent of value added to productivity of the rest of the establishments. The greater the difference in productivity, the more the ratio differs from one, meaning a greater cost penality exists for producing at a less than optimal size. Data at branch level from Cuentas Nacionales (National Accounts), INEGI. Notably such estimates permit us to consider (as opposed to cross section studies) the long-term difference among industries. According to Schmalensee (1989), at a very simple level, industrial averages over a long period of time shed more light on long-term difference than observations from a single year.

Bibliography Bain, J. (1956) Barriers to New Competition, Cambridge: Harvard University Press. Brown, F. (1998) ‘La industria de autopartes mexicana: restructuración reciente y perspectivas’, mimeo, Santiago de Chile: CEPAL. Brown, F. and Domínguez, L. (1997) ‘The dynamics of productivity performance in Mexican manufacturing 1984–90’, The Developing Economies XXXII, 3: 279–98. Caballero, R. and Lyons, R. (1992) ‘External effects in US procyclical productivity’, Journal of Monetary Economics 29: 209–25. Casar, J. (1996) ‘Un balance de la transformación industrial en México’, in Katz, J. (ed.) (1996) Estabilización Macroeconómica, Reforma Estructural y Comportamiento Industrial, Santiago de Chile: CEPAL. Casar, J., Márquez Padilla, C., Marván, S., Rodríguez, G. and Ros, J. (1990) La Organización Industrial en México, México: Siglo XXI. Caves, R. and Uekusa, M. (1976) Industrial Organization in Japan, Washington: Brookings Institution. Clavijo, F. (1992) ‘La eﬁciencia productiva del sector manufacturero mexicano, 1985–90’, mimeo, Oﬁcina de la Presidencia de México. Dahlman, C., Ross-Larson, B. and Westphal, L. (1987) ‘Managing technological development: lesson for the new industrialized countries’, World Development 15, 6: 759–75. Diamond, P. (1982) ‘Aggregate demand management in search equilibrium’, Journal of Political Economy 90: 881–94. Dutrénit, G. and Capdeville, M. (1993) ‘El perﬁl tecnológico de la industria

188

Flor Brown and Lilia Domínguez

mexicana y su dinámica innovadora en la década de los ochenta’, El Trimestre Económico 239. Domínguez L. and Brown, F. (1998) Transición Hacia Tecnologías Flexibles y Competitividad Internacional en la Industria Mexicana, en prensa en Editorial Miguel Angel Porrúa. Eichner, A. (1976) The Megacorp and Oligopoly, Cambridge: Cambridge University Press. Expansión (1994) ‘Las 500 empresas más importantes de México’, Expansión, 26. Fay, J. and Medoff, F. (1985) ‘Labor and output over the business cycle’, American Economic Review 75: 638–55. Gallini, N. (1984) ‘Deterrence through market sharing: a strategic incentive for licensing’, American Economic Review 74: 931–41. Gold, B. (1981) ‘Changing perspectives on size, scale and returns: an interpretative survey’, Journal of Economic Literature 19: 5–33. Hay, D. and Morris, D. (1986) Industrial Economics and Organization, Oxford: Oxford University Press. Jacquemin, A. (1992) ‘What is at stake in the new industrial economics’, in Bianchi, P. and Baldasari, M. (eds) Oligopoly and Dynamic Competition, London: Saint Martin Press. Katz, J. (1985) ‘Domestic technological innovation and dynamic advantages: further reﬂections on a comparative cases studies program’, in Rosenberg, N. and Frischtac, C. (eds) International Technology Transfer: Concepts, Measures and Comparisons, New York: Praeger. Katz, J. (1996) Estabilización Macroeconómica, Reforma Estructural y Comportamiento Industrial, Santiago de Chile: CEPAL. Mertens, L. (1995) ‘Productivity improvement strategies: the case of the Mexican food and metal industries’, International Labor Ofﬁce, EMD/15/E. Micheli, J. (1994) Nueva Manufactura, Globalización y Producción de Automóviles en México, México: Facultad de Economía UNAM. NAFIN (1995) La Economía Mexicana en Cifras, Mexico: Nacional Financiera. Nishimizu, N. and Robinson, S. (1986) ‘Productivity growth in manufacturing’, in Chenery, H., Robinson, S. and Syrquin, M. (eds) Industrialization and Growth, New York: Oxford University Press. Oaks, D. (1994) ‘Mexico, reform and productivity growth’, Working Paper, World Bank. Penrose, E. (1959) The Theory of the Growth of the Firm, Oxford: Basil Blackwell. Pozas, M.A. (1993) ‘Mecanismos de innovación y transferencia tecnológica y sus efectos sobre la organización del trabajo’, in Covarrubias, A. and Lara B. Relaciones Industriales y Productividad en el Norte de México: Tendencias y Problemas, Mexico: Friedrich Ebert. Ramírez, J.C. (1997) ‘La organización justo a tiempo en la industria automotriz del norte de México, nuevos patrones de localización y eﬁciencia’, Economía Mexicana VII, 2. Schmalensee, R. (1989) ‘Inter-industry studies of structure and performance’, in Schmalensee, R. and Willig, R. (eds) Handbook of Industrial Organization, New York: North-Holland. Silberston, A. (1972) ‘Economies of scale in theory and practice’, Economic Journal, Supplement: 379.

Trade integration and sectorial productivity

189

Steindl, J. (1952) Maturity and Stagnation in American Capitalism, Oxford: Basil Blackwell. Ten Kate, A. and De Mateo, F. (1989) ‘Apertura comercial y estructura de protección en México, estimaciones cuantitativas de los ochentas’, Comercio Exterior 39, 4: 497–511. Tybout, J., Melo, J. and Corbo, V. (1991) ‘The effects of trade reforms on scale and technical efﬁciency: new evidence from Chile’, Journal of International Economics 31, 3: 231–50. Tybout, J. and Westbrook, D. (1995) ‘Trade liberalization and the dimensions of efﬁciency change in Mexican manufacturing industries’, Journal of International Economics 39, 1/2: 53–78. Venables, A. and van Wijnbergen, S. (1993) Trade Liberalization, Productivity and Competition: the Mexican Experience, The World Bank. Verdoorn, P.J. (1949) ‘Fattori che regolano lo sviluppo della produtività del lavoro’, L’Industria 1, 1: 45–53. Weiss, J. (1992) ‘Trade policy reform and performance in manufacturing: Mexico 1975–88’, The Journal of Development Studies 29, 1: 1–23. Weiss, L. (1974) ‘The concentration–proﬁts relationship and antitrust’, in Goldschmid, H., Mann, H. and Westo, F. (eds) Industrial Concentration: The New Learning, Boston: Little Brown.

Comments International openness, dualism and growth Jean-Luc Gaffard and Michel Quéré

Dualism is a typical feature of the process of growth of later developing countries: it takes the form of the presence of advanced regions (and sectors) within a backward country. Dualism results from the conditions in which the existence of advanced countries inﬂuences the internal growth process. History, geography and technology are all elements that could explain this phenomenon in the sense that they determine a greater or lower growth rate of output and productivity. However, it should be apparent that dualism is, above all, the expression of a conditioned growth path, essentially characterized by the absence of learning and the lack of gradual acquisition of new technologies, due to out-of-equilibrium coordination failures. Within a structure marked by dualism, only regions that have access to external factors experience a real development process. Nevertheless, this development is conditioned by the development of advanced countries. It may lead to the chronic inability for the country (and its more advanced regions) to be on the frontier of potential productivity gains, and, above all, to effectively capture productivity gains. What matters for such a country is not to be present in some industries, which embody advanced or strategic technologies. It is to master the process of change assimilated to a process of learning of both technical novelty and market conditions; that is, to be able to transform technological advances in increasing returns. In the parlance of Lucas (1993), growth miracles (catching-up processes) are certainly productivity miracles. There is no doubt that the issue of how the maximum productivity growth can be obtained is a crucial one. But it is not one that can be resolved on the basis of technological considerations alone. As a general rule, increasing returns lead to specialization and interdependence. But for this to happen, the market has to be large enough, and speciﬁc (time-dependent) co-ordination problems have to be solved. These problems result from structural changes and, particularly, from the deepening of the vertical division of labor. For the purpose of considering the implications of these changes, it is convenient

Comments 191 to regard ﬁrms as undertaking activities rather than making and selling products, these activities having to do with the discovery and the estimation of future wants, with research, development, and design, with the execution and co-ordination of processes of physical transformation, with the marketing of goods, and so on. (Richardson 1975: 355) As a matter of fact, development essentially requires the consolidation and the co-ordination of these activities within and between ﬁrms. These activities require appropriated knowledge, capabilities, experience and skills. What is at stake is precisely how these capabilities may be acquired. Learning will be all the more effective when the process of change will be more regular. A strong co-ordination between supply and demand, between investment and consumption, which allows a regular growth, is the necessary condition for the economy to gradually acquire new skills and new technologies. On the contrary, co-ordination failures result in market disequilibria, trade unbalancing, huge ﬁnancial or human constraints, which characterize some patterns of evolution and prevent learning from occurring (Amendola and Gaffard 1998). At the macro level, different kinds of development strategies can be considered and have been implemented. One is the import substitution strategy, another is the export led growth strategy. These strategies can be more or less efﬁcient in terms of growth rate. Within a context of free trade, rather than producing high quality goods for the domestic market, a backward economy will import them. As a consequence, learning by doing will be depressed (Stokey 1991). This is a reason why such an economy may choose to implement an import substitution strategy. Then labor force shifts to the production of formerly import goods. This may induce a positive productivity shock; ‘But this is a one time stimulus to productivity and thereafter the mix of goods produced in the closed system can change only slowly, as the consumption mix changes’ (Lucas 1993: 270). Moreover, a so-oriented development is limited by two constraints: the degree of matching between supply and demand for different skills, that addresses the conditions of accumulation of human capital with respect to the new production requirements; the extent of the internal market for the new produced goods. For this kind of strategy to be viable, an appropriate economic policy is required, that is aimed at solving coordination problems that necessarily emerge. An alternative strategy consists of exporting high quality goods. As a matter of fact, rapid productivity growth is strongly connected with international trade and openness, because only a large volume of external demand for high quality goods may allow the economy to be positioned on the frontier of productivity growth (Lucas ibid.). But there is no evidence that such an external demand will match the new domestic supply. As well as in the case of an import substitution strategy, co-ordination problems exist that have to be

192

Jean-Luc Gaffard and Michel Quéré

resolved. As a consequence, once again an appropriate policy making is required. For both development strategies, growth performance is determined less with respect to fundamentals (the potential of productivity growth) than in relation to the way co-ordination issues are dealt with. The idea is that technological capabilities are a result of the process of change and not a pre-condition of it. Whatever the technology potentially available at the beginning of the process, it should be clear that technological capabilities are negatively related to the degree of short- or medium-term economic instability. As for the strategies of development mentioned above, what matters is to know whether, and to what extent, they generate instability. The dynamic regime induced by the implementation of the development strategy will be the key to its success or, on the contrary, to its failure. Instability comes from, and results in, distortions in the structure of productive capacity, and those distortions are the source of different forms of dualism in a backward open economy. The Mexico story as reported by the papers is a good illustration for our purpose. All these contributions address crucial issues regarding the necessary requirements for sustainable economic growth. Two core questions summarize those issues. The ﬁrst one is related to the importance of international integration for the sustainability of Mexican economic development. It can be expressed as follows: should it be viable in the long run that ‘Mexico is a country in which structural change is mostly focused on the consolidation of international participation as a “modernized assembly factory” ’? (Cimoli, this book, 168). The other is related to the requirements for an absence of dualism within a national economy in a context where international integration prevails: should it be possible that the conditions required by international competitiveness can induce anything else than a North–South kind of dualism within the national industrial structure? The three contributions highlight different but complementary aspects of those two core questions. The papers by Dussel Peters and Cimoli are very similar in the sense that they are converging on the different historical periods and economic constraints faced by Mexico in the long run, even if they differ through the emphasis put on different consequences of this shared contextual framework. Since the 1930s, the Mexican economy has experimented with the two different development strategies mentioned above. They both resulted in irregular growth regimes. The dominant import-substitution strategy collapsed in the mid-1970s because of the limits of an active public interventionism (which have resulted in huge ﬁnancial constraints and public sector deﬁcits). A transitory period began, characterized by high inﬂation, massive capital outﬂows, and a huge macroeconomic instability, the main aspect of which was the 1982 ﬁnancial crisis. Remedies to this situation combined large devaluation, trade liberalization and an export-oriented

Comments 193 industrialization, which took place until the 1980s and allowed many multinationals to locate assembly lines (the so-called ‘maquiladora’) in the north part of the country. The latter beneﬁts from the domestic low wage level that allows for a competitive advantage on the US internal market. During the last period, a macroeconomic stabilization has occurred under the pressure of an active public policy (including a policy of ﬁxed exchange rate for the peso) and has promoted the furthering of the liberalization policy. This policy (EOI) also resulted in a ﬁnancial crisis because of the induced increase in national dualism and, consequently, in a dramatic fall in real wages. Although the crisis forced foreign capital out of the country in 1994–5, government strategy has still enhanced the importance of an export-oriented strategy, with the participation of Mexico in the NAFTA agreement, which contributes to secure foreign investments in the country. The Dussel Peters contribution focuses in more detail on the interactive character of public policies in relation to macroeconomic events. It makes explicit structural distortions that result from new economic policy. On the one hand, the average growth of GDP is not actually impressive because, if a few sectors perform in relation to their exports’ possibilities, the rest of the economy bears difﬁculties due to weaknesses in the domestic market. On the other hand, external trade is unbalanced inasmuch as the EOI policy paradoxically induces an increase in imports’ volume. It also insists on the impact of government liberalization policy on sector and regional patterns, and gives signiﬁcant arguments to explain the increase in dualism for the national economy, as well as conclusive comments on the lack of endogenous growth conditions. In the line of this analysis, the Cimoli contribution emphasizes the consequences in terms of technological capabilities. It is especially convincing on the paradoxical aspect of Mexican development. Whereas the successful export strategy allows for technological upgrading, it only concerns a few sectors dominated by large domestic or foreign companies; on the contrary, the rest of the economy exhibits a lack of technological capabilities and R&D efforts. The essence of such a situation has to be found in the very speciﬁc organizational design chosen by export-oriented companies that mainly import the materials and equipment required by the production and only beneﬁt from the relatively low labor cost. The contribution by Brown and Domínguez focuses on a speciﬁc aspect of these historical patterns; that is, the analysis of labor productivity on the 1984–94 period. The latter conﬁrms the complexity of economic growth and its dualistic structural pattern. On the one hand, strong productivity increase concentrates on a few sectors due to a very speciﬁc industrial structure where multinational companies exhibit high local growth rates with a high volume of exports. On the other hand, when those speciﬁc sector patterns are dropped out, a great variance appears in the other industries where companies are not performing so well, be it their technological potentialities, their ﬁnancial or export performances. If the usual

194

Jean-Luc Gaffard and Michel Quéré

relationships between GDP and labor productivity apply for the former population, i.e. that market growth largely explains the dynamic of productivity, a speciﬁc feature of the Mexican economy lies in the dependence of productivity from trade opening. This conﬁrms the central feature of asymmetric dependence with the US economy already mentioned. Alternatively, as far as small companies are concerned, the evolution of labor productivity is much more chaotic, and conclusive results are difﬁcult to be established. The essential interest of the contribution is to map the effects of the dualistic pattern of the economy on labor productivity, which strengthens the fragility of the Mexican economy and threatens the viability of its evolution in the long run. In summary, the chapters conﬁrm that co-ordination failures could be the actual cause of the fragility of the Mexican economy. A common and impressive feeling seems to be shared by the three contributions: the Mexican economy has yet to develop an appropriate infrastructure for a suited and sustainable economic development. The EOI strategy results in an increase of economic dependence from imports. This holds for the assembly lines companies that import raw materials and equipment from their domestic markets (the US essentially), but this also holds for the rest of the Mexican economy, where growth induces an increase in imports because of the lack of the required inputs domestically on the Mexican market. Actually, it seems excessive to consider that the Mexican economy is facing an international integration. What seems to appear is a successful de-location of labor-intensive activities in Mexico from the US, whereas value-added activities are not de-located. This phenomenon cannot be anything but structurally fragile: it is certainly not a guarantee for a sustainable growth. As we already have mentioned, such a situation induces internal imbalances that are a source of macroeconomic instability. In short, the fact that more or less 80 percent of US television sets are now manufactured in Mexico is not a guarantee for the sustainability of Mexican economic growth. On the contrary, this makes explicit the dependence of Mexican growth from the US conjuncture and the recent economic success of the Mexican economy has to be found in the US economic performance much more than in the domestic characteristics of the Mexican economy. Within a structure marked by dualism, trade balance deﬁcit is particularly worrying. If the maquiladora sectors and the energy sector exhibit an important positive trade balance, it is still largely insufﬁcient for compensating for the negative trade balance that characterizes the rest of the economy. As far as the EOI strategy applies, an increase in the dualistic aspect of the Mexican economy is very sensitive. The form taken by external dependence paves the way both for the relative stability of the north of the country, and for the chronic instability of the south, as indicated by the chaotic evolution of productivity levels. As such, the dualistic pattern of

Comments 195 the economy cannot do anything but increase. More, a cyclical downturn in the US economy will result in a generalized macroeconomic instability. The Mexican economy does not seem to be on the road of a sustainable economic development. On the one hand, the answer to the ﬁrst core question addressed by the three chapters is extremely clear: establishing modern assembly factories favored by the NAFTA agreement is not a suitable orientation to maintain long run economic growth. On the other hand, with respect to the second core question, it appears difﬁcult to avoid dualism because of the nature of external relationships that do not contribute to allowing sustainable increasing returns to be obtained. Let us add a few remarks. As we have already mentioned, modern economic theory put into light the importance of trade in order to allow an economy to maintain itself at the technological efﬁciency frontier and produce the most advanced basket of goods in terms of productivity (Lucas ibid.). The Mexican economy does not produce at the efﬁciency frontier, despite its openness. The analysis could suggest that the use of local resources, be they raw materials or equipment goods, would be a necessary condition for enhancing a less dependent, and hence a more regular development. It is clear that the sustainable character of economic growth implies local interdependencies in the industry, intra and inter-sector linkages, local externalities and learning processes, all elements that contribute to generate an endogenous process of development. Facts and ﬁgures in the three chapters commented explicitly on the lack of such elements in the Mexican economy, and pinpoint the need of building them into the perspective of taking advantage of international openness. But establishing these links, which contributes to guarantee a suited co-ordination between consumption and investment over time, must not be confused with developing the productive capacity of a few selected sectors of the domestic economy. Besides, the main reason for US companies to locate assembly lines units in Mexico lies in the relative low cost of labor. This is only possible because of the excess supply of low qualiﬁed labor force. However, the low wage situation cannot be anything but transitory, while no real incentives exist that would allow an increase in individual skills. The absence of learning prevents the economy from effectively catching up to more advanced countries. An active public training policy (be it initial or vocational training efforts) is certainly part of a solution. But the Mexican economy exhibits another paradoxical feature. In a country where the national dimension is a fundamental aspect for politics, economic policy, and more largely the related policies (such as educational or social policies), seem to be guided by short-term objectives in the sense that they seem to be elaborated in reaction to short-term phenomena. The ‘Pronasol’ is a good illustration of this short-term reliance of public policy. Certainly, the reason lies in the liberal vision of the world that the elite of the country has in mind. However, the short-term orientation of public

196

Jean-Luc Gaffard and Michel Quéré

policies (including the social and educational policies) is not aimed at securing the fragile aspects of economic development. Nevertheless, a few efforts toward securing better economic developments do exist but are much more the result of a spontaneous movement than a planned strategy by policy makers. A new situation has recently prevailed: the successful export oriented strategy makes Mexico a platform for penetrating the US market. In particular, European investments have been carried out in this perspective. In breaking the import dependency induced by a maquiladora system and in enlarging the export base for the Mexican economy, they could contribute to diminishing the structural imbalances.

Bibliography Amendola, M. and Gaffard, J.L. (1998) Out of Equilibrium, Oxford: Clarendon Press. Lucas, R.E. (1993) ‘Making a Miracle’, Econometrica 61: 251–72. Richardson, G.B. (1975) ‘Adam Smith on competition and increasing returns’, in Skinner, A. and Wilson, T. (eds) Essays on Adam Smith, Oxford: Oxford University Press. Stokey, N.L. (1991) ‘Human capital, product quality, and growth’, Quarterly Journal of Economics, May: 587–616.

Part IV

Regional divergence and structural policies

9

Structural divergence and the dynamics of dualism Lessons from Mexico before and after NAFTA Martín Puchet Anyul and Lionello F. Punzo

This chapter offers a reconstruction and re-interpretation of the process of integration between Mexico and the US, focusing upon the implied structural dynamics. The results of our approach, based on a device called Framework Space and the related notion of regime dynamics, are compared to existing alternatives, basically constructed on growth theory. Some of the issues emerging in the fast-unfolding NAFTA process may be relevant to Europe, too, in the near future. NAFTA involves a (relatively speaking) small developing economy and a large, developed one. Similarly, central Eastern European countries are queuing up to be admitted into the Union and expect to join soon, an event raising problems and creating challenges that may have the trajectory of the two North American economies as reference points. Some such problems have been on the EU discussion agenda already for some time, in particular those related to the mobility of productive factors, and of labor in particular. Others are also equally relevant, thus causing the uncertainty as to the exact date of their admission. In the same vein, the chapter ends by offering a tentative comparison of aspects of two distinct and, both geographically and historically, far apart processes of economic integration: again the more recent one of Mexico–USA, and an integration process with a longer history, between northern and southern Italy (the Mezzogiorno), that after the political uniﬁcation gathered great momentum in the early 1960s. Some stylized features of Italian regional development may provide hints to understand, at least qualitatively, structural changes currently taking place in North America and anticipate some to come. Thus, the case of southern Italian development is suggested as a mirror for Mexico as it shows the risks of falling into what can aptly be deﬁned as a ‘growth trap’: highly volatile investment and comparatively much lower productivity gains. Such a trap may eventually prevent the economy from moving onto that virtuous cycle of innovation and accumulation that is the greatest promise of economic integration between unequal countries. That the latter outcome may not be as natural an outcome as it is predicted by much of the current

200

Martín Puchet Anyul and Lionello F. Punzo

literature connecting growth with economic integration, is a case for discussion and, more than that, for policy design and action. Integration in the two areas differs in at least one key aspect, of course. Whereas in Mexico and the US, factor mobility is restricted for most categories of labor force and certain types of technology, mobility of both capital and labor has always been fully allowed in Italy (as it was in Germany in the last decade, and it will be the same for the Central European countries after their admission to the EU). This major factor, that does make Italy and the EU different from the NAFTA, is relevant for assessing hypotheses such as the one of neoclassical growth theory, predicting the convergence between countries or regions. This same fact, however, highlights the signiﬁcance of a comparison of some of the fundamental aspects of long-term dynamics of the economies involved. They pertain to the forms of their reciprocal dynamical coupling that alter in a fundamental way the countries’ structures and are, in a sense, complementary to those induced by factor mobility. This illustrates the fact that there are features of structural change that cannot be merely induced by resource reallocation. The chapter focuses upon these features, because they represent a common ground for a comparison of diverse integration processes, where resource reallocation (and in particular labor mobility) greatly varies across and is also rapidly changing over their temporal horizons. Section 9.1 places general issues concerning processes of integration into the debate on economic growth and convergence. It points out how the recent and still hot debate on the latter issue largely ignores the analysis of those structural changes that cannot be (entirely) accounted for by resource reallocation. By contrast, we recall that in the approach of both Latin American structuralists of the 1960s and Italian economists studying north–south dualism, structural transformations accompanying integration processes featured prominently within frameworks built upon the foundations of the theory of development. The latter focused on disequilibrium dynamics and the ﬂuctuating nature of transitions through development phases. Hence, this motivates proposing an alternative, though to a certain extent complementary approach where economic integration is studied as a process of structural change and emphasis is placed on irregular cycles in growth trajectories. This novel approach (described at length in Section 9.4) inserts the long-term trajectories of possibly sectorially disaggregated variables, such as productivity and investment per worker, into the theory of oscillations, with trajectories representing families of growth cycles. This framework develops from the argument below. On the basis of the evolution of labor productivity alone, it was argued earlier (in Section 9.2) that it is difﬁcult to ascertain whether there is (conventional) convergence between the US and Mexico’s economies. The relevant facts are reviewed and the limitations shown, inherent to all analyses that consider countries integration without paying due attention to structural changes.

Structural divergence and the dynamics of dualism 201 In Section 9.3, the integration process in North America is shown to have generated deeper structural changes in the smaller and less developed of the two economies. To substantiate this thesis, our analysis makes use of a speciﬁc deﬁnition of structural change (and related measurement tools) based upon labor reallocation, and it shows its positive contribution as well as its insufﬁciency to account for the growth trajectory of the Mexican economy. Taking the dynamics of capital formation into account separately would not help much, however, for the problem lies with the use of the inadequate, univariate model of conventional growth theories. There, two dimensions are left out: the time evolution of the relation between the relevant dynamic variables, and the rich sectorial dynamics lying behind macroeconomic performance. Within the more general framework set out in Section 9.4, Section 9.5 reviews the main events of recent Mexican and US histories in terms of structural oscillations across growth paths. This highlights how the Mexican economy may be going through a sequence-deﬁned ‘growth trap’ as a result of an accelerated integration process. Finally, Section 9.6 introduces the Italian regional dualism as a benchmark, to show how an integration experience can realize the dynamical coupling of economies by producing structurally divergent trajectories. In highly qualitative terms, the processes of integration examined may be used as examples to ward off the risks implied in certain types of growth behaviors, and to help in designing policies appropriate to reduce them. Our approach to the relation between integration and structural transformation suggests a new view of economic policies that should target regimes of macroeconomic and sectorial dynamics, rather than desirable, but speciﬁc, values of some key variables. Section 9.7 concludes with some policy considerations.

9.1 Some of the old, new and newer perspectives on integration This chapter reviews dynamics induced by a country integration process in the light of (macro-) regional dynamics. It goes over what can be termed the ‘before’ and the ‘after’ of processes of integration: modern Italy being the accomplished result of a (forced) process of economic integration ensuing from the political unification realized towards the middle of the last century; the creation of a North American market being the main, but still only partially realized objective of the NAFTA agreement. This unifying perspective to country integration and regional development is, however, quite distant from the one dominating the literature, though it may be complementary to it. There, two main lines can be recognized. Most of the on-going discussion has been recast into the jargon of growth theory (and theories) whose rebirth in the mid-1980s1 has set up a

202

Martín Puchet Anyul and Lionello F. Punzo

framework centered on the acceptance or rejection of the theoretical predictions of the so-called neoclassical model. In this tradition, trajectories of economies with different per capita output or productivity are explained by different parameter values of an all-encompassing model. Otherwise, they are expected to converge to a (conditionally) unique level and a unique supporting long run path, a steady state. Other kinds of differences (e.g. those between rich and poor economies, or between less developed and developed economies) were also reinterpreted in the same way. While studies on convergence between rich and poor economies and of related phenomena (catching up, lagging behind and the like) became thus central, replacing other perspectives of economic development,2 the idea also emerged of using the same model in conceptualizing the differences observed in regional development within a single country.3 Just like countries, regions are expected to converge to a unique trajectory of sustained growth under the conditions of the neoclassical model, in particular with freely-accessible technologies, a competitive market system and free factors mobility. Accordingly, in the ‘canonical tenet’, if such convergences were not obtained, one would have to look for departures from the restrictive assumptions upon which the predictions rested. Hence, on one side, complex economic, technical or institutional features ended up being subsumed among the causes determining the parameter values of the common growth model or else affecting exogenous variables. On the other, the focus was placed almost entirely upon the issue of factors mobility and hence the optimal allocation of resources. No such role was left for the dynamics of integration in itself to inﬂuence the patterns of ensuing growth. Not everything could be accommodated within that framework, though. In particular, instances of persisting (or, even, increasing) cross-country divergence escaped, but they were often observed. This is among the empirical facts that eventually led to the variety of endogenous growth models we are so familiar with now. Thus, for instance, the need to explicitly relate growth to economic integration and its forms came up in the debate surrounding the creation of the European Monetary Union.4 One politically interesting result, though not theoretically unexpected, shows production concentrating into the more developed areas (countries or regions) as a result of the mere presence of some increasing returns to scale.5 It is now well known that free factors mobility in a structure of competitive markets can only be a necessary, but by no means sufﬁcient condition for growth to diffuse uniformly through a newly-created unique market economy. Much of the so-called new geography is concerned with similar problems and outcomes.6 Although for some time overshadowed by the ﬂourishing growth empirics, the theories of development have approached integration between economies with different sizes and/or development levels from a

Structural divergence and the dynamics of dualism 203 different perspective. They distinguished between problems due to different relative levels of development of national economies and those arising between regions of the same economy, on the basis of the countries’ or regions’ structural conditions. The former pertained to the country’s roles in international trade, production and ﬁnances and the different types of institutions; the latter referred to the conﬂictive or cooperative ways of regions linking up with each other within a national economy. Certain structural analyses, among them those produced by so-called Latin American Structuralism, stressed complementary factors that could enhance development among reciprocally integrating economies. This approach introduced such concepts as ‘dual development’ or ‘development with structural heterogeneity’, on the basis of which various Latin American integration processes from the 1950s to the 1970s were analyzed,7 as well as the Italian dualistic development.8 A quick review of the manifold integration processes treated in such literature shows that economies relate to each other in ways that cannot be synthesized through the observation of single indicators as proposed in the aggregate growth tradition. The key facts that were highlighted were essentially static: ﬁrst of all, the centrality of the sectorial structure of the economies; then, how sectors connect to each other, and ﬁnally (and perhaps principally) how they interact in a variety of forms, or dynamical coupling of networks and architectures. The corresponding notion of economic integration emphasizes that economies, whether they be national or regional economies, connect to each other via the creation of complex networks, the interlinks of which may undergo change at different relative speeds. All processes of economic integration entail fundamentally coupling effects on the sectorial compositions, space distributions, trajectories and rates of growth of the economies involved. In other words, structural changes are brought about that cannot be summarized by otherwise useful aggregative frameworks. They entail much more than convergence or divergence from a unique long run path. To highlight these features of structural change, it is necessary to build a framework that complements these two approaches. In Section 9.4, the ﬂuctuating trajectories of a set of monitoring variables are used to reconstruct structural changes. Of course, the trajectories upon which the theory of growth concentrates are speciﬁc (and, from an empirical point of view, particularly simple) cases of such ﬂuctuations. Likewise, convergence in the growth sense exhibits a special type (monotonic, globally stable) of dynamics that may or may not be observed. It is not there to be observed in the cases of Mexico and the US, nor in the case of Italy. Other, perhaps more interesting, features of their comparative dynamics may be found: rather than monotonic paths, we encounter structural cycles that can be so highly erratic that they are not really cycles. The framework also offers a way of reconceptualizing facts of

204

Martín Puchet Anyul and Lionello F. Punzo

dualism or structural heterogeneity highlighted in the theory of development, as instances of growth paths diverging from one another. By merging notions from the theories of growth and development with those of the economic theory of oscillations and multisectorial dynamics, the approach tries to launch a so-far impossible, but empirically demanded, uniﬁcation. It relies on three variables of growth theory – productivity, investment, employment – and it considers their relative dynamics. Under certain restrictions, we obtain standard results of growth theory: i.e., in the long run, we derive a representation of a (country and/or sector) growth path in the chosen variables whereby one variable (typically, labor productivity) can be regressed against the other (investment per worker) to test for statistically signiﬁcant relations. Our approach also makes an attempt to represent a sort of medium-term dynamical relation and trace out its time evolution. Such framework treats structural changes as changes in the functional relations between variables, as well as changes in the way in which each sectorial or country pair is coupled with all others. In this way we can focus on the qualitative evolution of the growth paths of Mexico and the US as a result of their progressive economic integration. The convergence issue therefore becomes: are they converging to the same pattern or structural cycle?9

9.2 Before or after NAFTA? The following pages will illustrate the previous ideas through a review of statistical evidence. In our view, integration between the Mexican and US economies was taking place before the enactment of the NAFTA agreement and it simply continued afterwards, with greater momentum, of course. We argue for the limits that an aggregative approach exhibits in accounting for such a rich process. These limits descend essentially from the ‘structural dimension’ missing in all conventional growth analysis. It is to try to recover some of the latter, that conventional growth analysis generated a kind of structural approach.10 This is adapted here to the Mexico–US case, basically to ﬁnd out how far one can go with it, and where we have to depart in earnest and establish something new. We will see, in particular, that it cannot account for the trajectory of manufacturing in Mexico relative to the US. Looking for additional variables to capture essential information, we select capital accumulation that our approach coordinates with labor productivity. Technically, the end result is a simultaneous equations, non-linear model for each economy, whose predicted time series are generically irregular ﬂuctuations across regimes. The analysis of the two countries’ regime dynamics will be built piecewise by assembling more conventional growth blocks. In conventional growth analysis, the key dynamics can be reconstructed through the evolution of (economy-wide) labor productivity. Figure 9.1 exhibits this history for Mexico and the US with year-to-year rates, instead of levels.11

Figure 9.1 Mexico–US: rates of productivity growth (1971–97).

206

Martín Puchet Anyul and Lionello F. Punzo

Looking at tendencies implicit in the two paths, it can be concluded that while the trajectory of the US shows a slightly positive growth, the Mexican trajectory is tending towards an increasingly negative growth. Both tendency interpolations, however, do not retain data variability. It is interesting to check these immediate, visual results against the common view that Mexico’s integration into the North American economy would spur growth and place the less developed upon the path of the more developed economy. Against the expectation that the integration process would produce a speeding up of productivity growth and a tendency to stabilize it around that prevailing in the US, from 1970 to 1997, the Mexican economy exhibits systematically a declining productivity growth, with a widening gap from the US economy. This can be interpreted in two different ways that turn out to be complementary in spite of their different theoretical backgrounds. The ﬁrst would argue that the Mexican economy has not managed to fully move towards the new growth trajectory because of the lack of free mobility of its labor force across the border. The alternative would point out that integration was taking place in the manufacturing sectors before NAFTA was even formalized and that NAFTA only reinforced such a tendency. Since those sectors rely upon increasing returns technologies, the observed divergence between productivity growth trajectories exhibit the different specialization patterns in the two countries.12 The preponderant weight of the manufacturing industry makes the dynamics of integration between these two countries clearly different from other processes predominantly involving the primary or the services sector.13 On the other hand, integration itself has substantially modiﬁed both the role and position of the manufacturing sector in Mexico. What changes did sectorial composition undergo before and after NAFTA? What role did they play in productivity growth? Although this may be a limited perspective on structural change, it is relevant to the understanding of the country’s long-term chances.

9.3 Structural change from the growth perspective In the version proposed in this section, the growth approach takes structural change basically to coincide fundamentally with labor reallocation, of all resources across sectors:14 to this process are attributed the observed behaviors of productivity dynamics.15 With this approach, our exercise reconstructs what would have happened if the actual changes in crosssector employment composition in the Mexican economy during the process leading to NAFTA had not taken place, and it compares to what, in fact, we know that happened. A new trajectory emerges from assuming how a certain type of structural change is being incorporated into its evolution. It then becomes possible to compare the real history with a ‘simulated’ or ‘virtual history’ that is built for analytical purposes.

Structural divergence and the dynamics of dualism 207 Two different mean rates of productivity growth (per year) are introduced in order to detect these effects of structural change. One of them considers the initial distribution of employment (gaiv), the other its ﬁnal distribution (gafv). For the economy as an aggregate, such a pair of rates is obtained as the weighted sum of the corresponding rates for its 17 component sectors. (Shares of a sector’s employment in the Total at the initial (or ﬁnal) period, depending on the case, are used in the weighting.) The trajectory that the virtual rate of productivity growth (denominated gascv) would have followed in virtue of the structural change (Figure 9.2) is obtained as the difference between the two rates so computed. The one thing that comes out clearly from Figure 9.2, is that, without structural change as a modiﬁcation of relative sectorial employment, the likely growth performance would have been way below the one on record. (Compare gaiv trajectories, the rate without structural change, to gav trajectories, the actual rate, in order to see the extent of this effect.) Looking at the trajectory that would have taken place if composition had been the same throughout, as for the ﬁnal period, it shows how this speciﬁc form of labor reallocation generated a much higher rate of productivity growth (compare actual gav trajectories and the trajectory constructed with the ﬁnal composition: gafv ). Comparing the hypothetical rate recording structural change, measured as gascv (gafv gaiv) to actual gav, it can be seen that the latter is well approximated by the former. Nevertheless, it is easy to see how, still, this does not fully explain actual productivity behavior. Figure 9.2 shows that, between 1977 and 1988, other factors played a signiﬁcant role.16 It is also important to note that the main structural changes that placed productivity growth around the mean yearly value observed in 1993 have taken place before then. Thus, if we accept this limited notion of structural change, the Mexican economy has transformed its structure, deﬁned in terms of sectorial labor allocation, already before NAFTA.17 However, the gap between real and simulated trajectories suggests that any analysis of the relation between growth and structural change needs to consider other aspects of structural changes as well as other variables. Without going too far from the mainstream, one may think that the rate of productivity growth (gv) and the changes in the sectorial composition of employment are not the only variables relevant for the explanation of long run tendencies. It is important also to take into account what happens to capital accumulation and its cross-sectorial dynamics. The rate of growth of investment per worker (gi) becomes a candidate indicator to monitor the evolution of capital intensity as a vehicle of new technologies. Figure 9.3 plots data regarding the latter rates in the two economies. A simplistic reading, focusing exclusively on tendencies, would fail to see two facts appearing in Figure 9.3: that variability in rates of investment growth is much greater than in rates of productivity growth; and that it is between two and three times greater in the Mexican than in the US economy.18

Figure 9.2 Mexico: observed and simulated rates of productivity growth.

Figure 9.3 Mexico–US: investment per employee, rates of growth (1971–97).

210

Martín Puchet Anyul and Lionello F. Punzo

Moreover, whereas productivity increase (or decrease) lies in a range between (5%, 4%) for Mexico and (1%, 3.5%) for the United States, in accumulation rates it is within (9%, 6%) for the United States and (24%, 16%) for Mexico. Major scale differences as these reveal that economic transformations did not have the same scope or size. Whereas the US has been through large-scale changes in capital endowments, Mexico had to make great efforts to accumulate capital but, also, enormous capital endowments were destroyed before and after the implementation of NAFTA.

9.4 Structural change through a growth cycle approach The approach deployed above springs from a growth theoretical framework: it is aggregative (with the analysis of one variable at a time) and sees structural change as a resource re-allocation phenomenon. Other aspects are left out that are worth looking at. These concern, basically, the interdependence between various growth variables and its time evolution. It is the latter that comprises structural change. There are many ways to present the Framework Space, the descriptive statistics device that is used in our analysis of structural dynamics and change.19 The easiest way is to relate it to existing literature on growth, for the FS has coordinates that are growth rates just like in the latter. For expository reasons, our point of entry is the aggregate approach associated with the neoclassical and endogenous theories.20 We argue that, under certain assumptions, our version of the FS is a natural, twofold generalization of the space of such growth models. For the whole economy, output is GDP, hereafter denoted by symbol Y, its ﬁgure being readily calculated as the sum of sectorial Value Added ﬁgures. Growth theory uses level values of aggregative variables, GDP, K for total capital stock, and E and N for total employment and labor force (a proxy for a demographic variable) and so on. From these, levels in intensive form are derived as ratios: y Y/E, k K/E, etc. Finally, we obtain growth rates of such variables to describe the dynamical properties of growth paths. We follow a similar three-layered procedure to derive, from a lower level, a unique higher level description, the opposite route does not lead to unique identiﬁcation but at most to a whole class. In fact, forces determining the values of variables at one level are abstracted from at a higher level, as it is clear from the procedure in standard growth theories. Neoclassical theory implies a ‘prediction’ on the long run behavior of a country’s Y. Under certain conditions, there will be a unique steady state value y* Y*/N, of output per capita, while the long run rate of output growth will be determined according to the equation: g = n , where n dlog N/dt and is the rate of growth of technological progress (TP). Henceforth, due to the other assumptions of the model, the rate of growth

Structural divergence and the dynamics of dualism 211 of output per head (g n) will be driven only by exogenous technological progress. In its absence, it will be equal to zero. With e dlogE/dt (E being actual employment) set equal to n, the steady state involves full employment at the ceiling of potential output, and the welfare index y is also a measure of productivity. Letting gy (g – e), we think of its values as coordinates on a real-line, the productivity growth line. Each value of gy is associated with a family of growth paths21 monitored by output per head. Neoclassical theory maintains that the long run value of gy lies at the origin in the absence of exogenous TP;22 if a country’s observed value is not there, this is due to not being zero.23 The (labor-) productivity line decomposes into the union of an equilibrium set (a unique path, whose location is determined by the value of ) and all other paths, indexed by the associated rate of productivity growth. Stability in the large implies that, eventually, the observed path is the one associated with the exogenously given 0.24 If we choose a reference time dimension (a clock) and identify a set of models inducing a regime partition, the productivity growth line, built from a given and known origin at , becomes a one-dimensional version of the Framework Space. For example, assuming a long run clock, if we introduce the family of endogenous growth models, next to the point-wise Solow’s regime (the class of paths driven by exogenous TP alone, at the origin), we will have all others spanning an endogenous growth regime25 where sustained long run growth may run at values gy* .26 In the exogenous theory, capital accumulation may only inﬂuence the levels attained in the steady state, not their rate of change.27 To explain the latter, nothing else but the knowledge of the actual value of is required. Endogenous theories have to add explanatory variables to account for the possibility (conﬁrmed by recently collected stylized facts) of growing faster. Such theories (and some related, well-known facts recently subjected to hot debate28) could be comparatively reviewed in a suitably generalized dynamical structure, whereby the productivity growth line is embedded into an enlarged space of explanatory variables. The single axis used in neoclassical theory appears not to be sufﬁcient to account for the diversity of the growth phenomenon across countries/regions (a point forcefully made by Baumol, Durlauf, Quah, among others).29 Standard neoclassical theory considers only physical capital to ﬁnd limits to growth in the long run due to the presence of diminishing returns to scale (to capital). To explain increasing returns to scale, fashionable alternatives are measures of human capital or of R&D expenditures, but they are among a host of suggestions trying to explain why other and more variables need to taken into account. From our viewpoint, human-capital explanations are in a family with those non-neoclassical theories where growth is an investment-driven phenomenon. Expanding the variables (or state) space along such a line introduces considerations about the investment pace. With an operation similar to the one performed above to

212

Martín Puchet Anyul and Lionello F. Punzo

obtain rates from levels, we choose to construct and to plot rates of growths of investment per worker on an axis against productivity growth.30 The two axes are made to cross at the origin of the gy axis. The family of neoclassical paths will be located in the origin of this standard, two-dimensional version of the FS. One can then characterize certain growth theories as being based upon the notion of technology embodiment (whether into physical and/or human capital), and discriminate them from those whose ‘predictions’ do not depend upon the pace of capital accumulation. Generalizing this notion, one gets our interpretation of regimes in the FS: in the ﬁrst quadrant we distinguish between paths exhibiting relations of weak or strong functional dependence between the two dynamics, output (productivity) and capital accumulation. Therefore, in the ﬁrst quadrant we talk of investment-driven (or respectively, innovation-driven) growth paths. Either set is a regime by itself, separated from the other by a set deﬁned, for theoretically obvious reasons, the Harrodian corridor. By extending this classiﬁcation outside the ﬁrst quadrant, we obtain altogether the six regimes used in this study and elsewhere, which are numbered clockwise. They are distinguished from each other by the relative strength of the relation between the two state variables as well as by a sign preserving/changing relation.31 In this light, the Framework Space can be thought of as a heuristic device to classify growth phenomena. However, treatment of investment as a key variable brings with it instability due to its typically volatile proﬁle. Growth theory can speak of stable and monotonic dynamics as long as investment dynamics is kept outside the picture (or sufﬁciently smoothed down). This is the result of assuming a long run perspective or ‘clock’ in our jargon. Once we retain volatility in investment behaviors in the medium or short run, we are in the dynamics of oscillations. In this temporal frame, growth dynamics takes the shape of a ﬂuctuating behavior. In the two-dimensional FS, dynamics is typically oscillatory in both growth variables; when it happens to be fairly regular, it follows a generalized growth cycle. But it is normally not of such tame, garden-variety. By predicting long run paths as attractors, growth theories provide a natural way of describing disequilibrium dynamics as deviations. Econometric practice focuses upon equivalent reference paths. Changes to technology and/or demography (generally labor supply) will shift the reference equilibrium in a predictable way; only erratic shocks to ‘Fundamentals’ may shift it in a ‘non-simple’ way.32 Learning the lessons of chaotic and complex dynamics, dynamics in the FS may be simple, but in principle it is richer than that. In particular, it may turn out to be one of systematic disequilibrium if the reference to any particular equilibrium path is used. Attracting sets more interesting than those of growth theories may, therefore, be encountered. In econometrics, any (‘signiﬁcant’) parameter change implies structural change. More in general, structural change is a change in the model. This

Structural divergence and the dynamics of dualism 213 compares with our notion of structural change as a change in regime in the following way. A change in regime is deﬁnitely always a change in the model, but a change of model may not entail a structural change. If our model can be parameterized, say by one key parameter, and we could test its sensitivity to changes in the parameter values, by techniques in the class of those of bifurcation theory, then crossing a bifurcation value would entail a qualitative change in the model. As explained in the footnote 31, the two-dimensional FS with its six regimes is obtained by considering bifurcation values for a two-parameter family of models. Parameterization makes this easy, for a qualitative change in dynamic behavior is almost equivalent to a model change. With more general models, this may not be so easy, but the case can be used to illustrate the idea adequately. Once the notion of the state space partition into dynamic regimes is accepted, there are two sorts of dynamics to keep track of, one across states and a structural dynamics across regimes Over any ﬁxed time interval, we meet a sectorial cloud or a dynamic structure. We speak of structural change meaning a change in the growth model.

9.5 Structural dynamics in the long and the medium run Comparison of trajectories33 in Figure 9.4 between 1970 and 1997 sees the Mexican and the US economies both traveling on a fairly regular structural cycle around the set of balanced growth paths (the 45-line, or Harrodian corridor). However, the difference in variability pointed out earlier for both variables, shows up here too. The US economy moves around a growth path that is relatively closer to the corridor, while the former economy oscillates much further away. In other words, structural oscillations are more pronounced for Mexico, as they involve both state variables. By comparison, the US cycle is closer to a natural accumulation cycle, with productivity following a fairly stable expansion path. The next important fact. While labor productivity grows steadily in the US, Mexico’s corresponding rates are at times negative, showing that aggregate output and employment are not following a coordinated expansion path.34 On the other hand, the pace of investment (per worker) in the US is almost always one of expansion whereas Mexico’s has two marked contraction periods. Thus, if the view is accepted that investment is one vehicle of technological progress, the path followed by the former economy has ensured a steadier absorption of new technologies (the ﬁnal boom being associated with the diffusion of the new economy), which in its turn has supported a likewise steady productivity expansion. Figure 9.3, therefore, shows the dynamic structural counterpart of the on-going US expansion that was clearly prepared by an accumulation process begun in the 1980s. Figure 9.4 for the Mexican economy is built upon a period division to capture its phases of crisis and restructuring.35 And in fact, the 1981 to 1988 phase shows the dynamic effects of the 1982 foreign debt crisis resulting from

214

Martín Puchet Anyul and Lionello F. Punzo

Figure 9.4 FS one sector by Mexico and US.

Structural divergence and the dynamics of dualism 215 the previous accelerated accumulation that had been ﬁnanced by foreign debt at low rates while high prices gave crude oil a disproportionate weight in total exports.36 The sudden change in those conditions was followed by severe restrictions on foreign exchange and, then, a devaluation.37 That crisis initiated a deep process of macroeconomic and structural adjustment,38 with a subsequent extreme drop in accumulation and in productivity growth. The second critical period, 1993 to 1997, began with the abrupt interruption of a new accumulation phase, ﬁnanced by public debt via shortterm instruments in dollars. The current account’s deﬁcit was no longer sustainable and devaluation turned foreign debt into unredeemable liabilities. After NAFTA, integration resulted in an increase of both exports and imports. An increasing ratio of the trade deﬁcit to GDP shows how the structural changes realized between 1988 and 1993 had not yet translated into a greater capacity to generate resources to fuel growth. The option of ﬁnancing this deﬁcit through the entry of short-term capitals soon proved to be highly fragile. Disinvestment and a new contraction in productivity are shown in Figure 9.4. Still, compared with the previous crisis, the investment fall was not that severe; and a path of productivity growth was recovered much more quickly. Therefore, it now seems possible for the Mexican economy to go into a new accumulation phase with an increasing productivity that may place it again into a regime like that of the periods 1970–81 and 1988–93. The strength shown by the economy during the recent crisis in Latin America makes this more likely. In terms of the same clock, the US economy shows a prolonged phase of growth (1970–88) followed by a recession (1988–93) that leads to a new period of growth with renewed and faster accumulation (1993–7). The ﬁrst phase is clearly close to a balanced growth with low rates not exceeding 1 percent, on average. The second phase has negative rates of accumulation (per worker!) but high productivity, generating innovation that reﬂects a major resource creation process with the birth of the new economy. The last period is characterized by a faster accumulation pace, to a certain extent due to the expansion of these new industries. To conclude, ﬁrst, there is an evident contrast between the structural dynamics of the two countries. On one side, there is the cycle of accumulation and innovation in a mature and robust economy that is appreciated by observing the proximity of oscillations to the corridor surrounding the 45 degree line; on the other, we ﬁnd a trajectory with large ﬂuctuations in a highly fragile, developing economy that is reﬂected in the distance from the same corridor.39 Progressive reciprocal integration and its full completion via NAFTA brought about structural changes that cannot be appreciated with aggregate data. High data volatility and the structural fragility just considered give, together, only a glimpse of a more complex scenario lying behind the coarse pictures of macroeconomics. Next, the process of integrating the Mexican economy into the North

216

Martín Puchet Anyul and Lionello F. Punzo

American bloc had already caused far reaching structural changes before NAFTA was enacted. Thus, the argument that NAFTA was signed in order to guarantee the success of institutional reforms and structural adjustment policies that, in fact, had already taken place, misrepresents the actual process. Moreover, integration had far reaching effects, initializing long-term adjustment processes.40 These transcend NAFTA, to a large extent. Such processes deeply overhauled the Mexican economy; the US economy was only partially involved. There, structural changes have many other causes, largely unrelated to the creation of the North American economic bloc. Finally, at any rate, structural changes have not been conﬁned to crosssectorial employment composition, which shows the limits of a pure labor reallocation approach to integration. Figures 9.5 and 9.6 with a 16-sectors breakdown41 of the two economies over the long run, 1970–93, and for the periodization used above, show them going through important changes of dynamical structure.42 The contrast between the histories of the two countries is apparent in either clock or time proﬁles. The US long-term picture in the FS shows a dynamic structure (a collection of sectorial paths) spread in a balanced way within the 45 degree-line corridor. In other words, it conﬁrms the aggregate picture in Figure 9.4.43 The dynamic structure of the Mexican economy is diametrically opposite, and it lies hidden behind its aggregate picture. The paths of sectors are nowhere close to the 45 degree line, they rather line up along the vertical, or accumulation axis. It is this narrow corridor that is relevant for that country: the cross-sector distribution of accumulation (and disaccumulation) rates is much more eschewed than that of productivity.44 From the point of view of accumulation and technological change, Mexico’s sectors spread out into a dualistic structure, those investing faster and those investing less than expanding employment. This is one key to understand the relative performance and structural dynamics of the two countries. The long-term growth of practically all of the US sectors is based on an adequate coupling between accumulation and productivity. This is not the case with Mexico. Differences across the two economies reveal their relative dynamic robustness.

9.6 Dualism and growth traps It is now useful to look at structural (sectorial) dynamics in the perspective of the periods (or phases) used before for the aggregate pictures. Then, another substantial difference can be appreciated. The sectors in the Mexican economy spread out in the same FS corridor noted above, be it the long or the medium run.45 On the contrary, US sectors follow trajectories visiting the regimes in a typical structural cycle of growth, with a remarkably visible, full loop46 (see Figures 9.6a and 9.6b). To interpret this phenomenon, it is useful to resort to certain, long-forgotten perspectives suggested by the theories of development.

Structural divergence and the dynamics of dualism 217

Figure 9.5 FS for the long run: Mexico and US.

218

Martín Puchet Anyul and Lionello F. Punzo

Figure 9.6a FS for the medium run: Mexico.

Structural divergence and the dynamics of dualism 219

220

Martín Puchet Anyul and Lionello F. Punzo

Figure 9.6b FS for the medium run: US.

Structural divergence and the dynamics of dualism 221

222

Martín Puchet Anyul and Lionello F. Punzo

In as far as the long-term behavior along the corridor is mirrored within each of the intermediate periods, the structural trajectory of the cloud of Mexican sectors seems to be prey to a growth trap that has some similarity to the notion of poverty traps in the development literature. The Mexican economy seems to be stuck in such dynamics: more than two decades of high and highly ﬂuctuating accumulation and decumulation patterns accompanied by relatively low productivity growth. How Mexico can come out of it, is the key issue to be considered. Recent events have given clues to the likely outcome. The Mexican economy might be well set on its way into a phase of aggregate growth, as data seem to indicate. Still, as long as its structure remains trapped in the above dynamic design, it is difﬁcult to think that this new period will initialize a virtuous phase in its structural growth cycle. First, though it might look a little daring, for a benchmark compare the long run pictures of Mexico and the US, on one side, with those of Italy represented by its two historically distinct macro-regions, North Center (NC) and the Mezzogiorno (Southern Italy, or MZ). When due attention is paid to the relative different scales, patterns are not much different in a qualitative sense. In the long run, the NC of Italy, very much like the US, clusters around a growth path with modest performance but stable around the 45 corridor (see Figure 9.7).

Figure 9.7 FS for the long run: South, North Center and Italy.

Structural divergence and the dynamics of dualism 223

224

Martín Puchet Anyul and Lionello F. Punzo

Figure 9.8a FS for the medium run: South Italy.

Structural divergence and the dynamics of dualism 225

226

Martín Puchet Anyul and Lionello F. Punzo

Figure 9.8a Continued.

Structural divergence and the dynamics of dualism 227

Figure 9.8b FS for the medium run: North Center Italy.

228

Martín Puchet Anyul and Lionello F. Punzo

Figure 9.8b Continued.

Structural divergence and the dynamics of dualism 229

230

Martín Puchet Anyul and Lionello F. Punzo

Figure 9.8c FS for the medium run: Italy.

Structural divergence and the dynamics of dualism 231

232

Martín Puchet Anyul and Lionello F. Punzo

Figure 9.8c Continued.

Structural divergence and the dynamics of dualism 233 In contrast, very much like Mexico’s, the picture of the dynamics of Mezzogiorno in the long run shows an eschewed distribution of sectorial paths along the vertical axis. In other words, the Mezzogiorno’s growth structure reproduces, in a sort of ‘zoomed in’ form, Mexico’s structural dynamics. In terms of the present qualitative analysis, Italy’s NC becomes a term of comparison with the US. So, the Italian regional dynamics may exhibit some of the effects of the NAFTA process when (and if ever) the US and Mexico will have become macro-regions in the uniﬁed North American market. It has been said already that changes in the weaker region have been geared by those in the stronger one, the asymmetry in the adjustment being great. The same can be seen from the ﬁgures where Italy more resembles NC than the Mezzogiorno. If we now continue with the comparison, we can see how, in very much the same way, the long run pictures for MZ (just like those of Mexico) hide a very much-eschewed medium-term dynamic, with sectors scattered very sparsely but basically along the vertical or accumulation axis. In other words, it is our contention that the Mezzogiorno in Italy has been stuck in a (low) growth trap, at least since the beginning of the 1980s. Italy’s NC, on the contrary, and Italy as a whole follow a fairly regular structural cycle, very much like the US economy. In a low growth trap, the cloud of sectorial growth paths lingers around a region in the regime’s space in such a way that: i) the medium-run dynamics resemble the corresponding long run, and ii) a sustained path of investment per worker is not matched by a comparably sustained path of productivity growth. Therefore, we are in presence of a doubly dualistic structure, each with its own pattern. Italy as a national economy remains split in two dynamical structures, each following a distinct growth ﬂuctuation or cycle, just like Mexico and the US. In this case we can rightly talk of structural divergence, for the emerging dynamical patterns are tendentially different. On the other hand, at the interior of the weaker region (MZ in Italy) or country (Mexico in the NAFTA) a dualistic structure is produced that affects the distribution of sectorial paths, with some sectors performing much better than others. This is appropriately called dynamic dualism, to distinguish it from the older notion in the development literature. It is the possibility of the emergence of one such cross-layered structure, across countries at the same time as across sectors, that is the lesson we want to learn from those two integration processes. The Italian case shows that this, unfortunately, can be a stable arrangement.

9.7 Economic policies for? The issue of the possibility or even necessity of ‘convergence’ takes on speciﬁc signiﬁcance when it is considered in the perspective of the integration processes currently building the European Union, NAFTA (and elsewhere MERCOSUR and ASEAN).

234

Martín Puchet Anyul and Lionello F. Punzo

First of all, it is clear that economic integration is not guaranteed to bring about convergence in the conventional sense of equalization of standards of living or productivity. Several regional histories in Europe (perhaps in contrast to the US case) show this; the Italian full (political and) economic integration over nearly 136 years now, is one of them. Hence, fears about the outcomes of such processes, which are entertained in Europe like elsewhere, are not ill founded. They are based upon an instinctive understanding that successful integration requires convergence in a sense that is much deeper than the one envisaged by the conventional (macroeconomic, or welfare) approach of growth theories. Moreover, that the two notions should go together still gives no indication for an important policy question: do we need to have convergence ﬁrst, and then work for economic integration, or can the latter be promoted under the presumption that ‘it will do the job’ of ensuring (conventional) convergence? Both positions are equally represented among the EU member states as well as the European public, and periodically they re-emerge. The issue of the admission of the Central Eastern European countries can largely be read through this problem. In the Americas, and in the NAFTA area, the view adopted is apparently the latter one, and so far, after an initial dangerous-looking start, it seems to pay off. The structural problems, the trap that our study has unveiled, subject to a control for more recent data, give us a deeper dimension and an appreciation of the fact that things can be more difﬁcult than they look in an aggregative perspective. That is why one more notion of convergence has been introduced and, consequently, there is more than one policy approach to the process of integration. In fact, when, for instance, the feasibility of the European Union or the achievements of the NAFTA are discussed, we are really asking: are the countries involved converging already (or are they able to converge) towards a common way of developing and growing? If they do not (as it often seems to many people on both continents) what sort of policies can be devised to help them to do so? At least in the greater part of the European countries (in contrast to what might be the case for NAFTA or MERCOSUR), standards of living are similar, hence the conventional notion does not have much scope. Still, the perspective of the newlyentering economies emphasizes that problem in the enlarged area together with others, shared with Mexico and the US. That is why the European Union is extending the scope of its special policies to enhance the economic prospects of ‘less favored regions’ (e.g. into the CEE countries), while at the same time, it actively implements general policies to promote conventional convergence among the whole lot. It has been suggested in this chapter, as elsewhere before, that the crucial unifying notion of convergence that is required as a policy foundation must be based upon a deeper understanding of the dynamic behaviors of the economies involved. Economic integration must ‘go along’ with

Structural divergence and the dynamics of dualism 235 convergence to a common model of growth. Using the well-known metaphor of the ‘mechanics of growth’, integration should be viewed as a process shaping a new engine of growth by assembling existing engines, the pre-existing national or regional economies.47 For the weaker countries, this means that they must ﬁnd a new sustainable path. It is here that the growth trap shows up as a real danger. For its unbalanced dynamics of productivity and capital accumulation, a growth trap cannot be a sustainable path. A North American market made up of two economies, one following a structurally virtuous expansion path (as the latest data show, it is deﬁnitely sustainable) and an economy wildly ﬂuctuating has an inbuilt inconsistency that one day will have to be resolved. We can hope that the solution will be at low social cost, but the recent history of the structural cycle in Europe shows that this is not likely to be so. Or else, we can work for it, setting out as a policy target the rendering of that cost as socially bearable. In a sense, this is what the EC has aimed to do so far, with recognizable success. Compared with the UE, NAFTA has no political or economic mechanisms to realize this, and that is why it is an imperfect integration. The challenge is to ﬁnd a common criterion of sustainability and then to use it as the organizing idea of appropriate policies. Here, markets are not the sole suitable institutions, but anticyclical policies still dominate the portfolios of actions considered by many a shortsighted government, in Europe as elsewhere.

236

Martín Puchet Anyul and Lionello F. Punzo

Appendix Databases Mexico sectorial data: INEGI, Sistema de Cuentas Nacionales de México, 1970–1998, and Banco de México, Acervos de capital, 1960–1994. US sectorial data: OECD-Economic Analysis and Statistics Division, The OECD STAN Database 1970–1995. Mexico and US aggregated data: OECD, Annual National Accounts – Volume 1: Main Aggregates and Comparative Tables 1960–1997; Volume 2: Detailed Tables, 1999. Mexico–US sectors: 1 Agriculture, forestry and ﬁshing 2 Mining 3 Food, beverages and tobacco 4 Textiles, wearing apparel and leather industries 5 Wood and wood products 6 Paper and paper products, printing and publishing 7 Chemicals and chemical petroleum, coal, rubber and plastic products 8 Non-metallic mineral products except products of petroleum and coal 9 Basic metal industries 10 Fabricated metal products, machinery and equipment 11 Other manufacturing industries 12 Construction 13 Electricity, gas and water 14 Wholesale and retail trade, restaurants and hotels 15 Transport, storage and communication 16 Finance, insurance, real estate and business services 17 Community, social and personal services

Structural divergence and the dynamics of dualism 237 Correspondence between Mexico–US and Italy sectors: Mexico–US 1 Agriculture, forestry and ﬁshing

Italy 1 Agricoltura, silvicoltura e pesca

2 Mining 9 Basic metal industries

6 Prodotti energetici 13 Minerali e metalli ferrosi e non ferrosi

3 Food, beverages and tobacco

36 Prodotti alimentari, bevande e tabacco

4 Textiles, wearing apparel and leather industries

42 Prodotti tessili e dell’abbigliamento, pelli, cuoio e calzature

5 Wood and wood products 11 Other manufacturing industries

50 Legno, gomma ed altri prodotti industriali

6 Paper and paper products, printing and publishing

47 Carta, prodotti cartotecnici, della stampa e editoria

7 Chemicals and chemical petroleum, coal, rubber and plastic products

17 Prodotti chimici e farmaceutici

8 Non-metallic mineral products except products of petroleum and coal

15 Minerali e prodotti a base di minerali non-metalliferi

10 Fabricated metal products, machinery and equipment

24 Prodotti in metallo e macchine 28 Mezzi di trasporto

12 Construction

53 Costruzioni

14 Wholesale and retail trade, restaurants and hotels

58 Commercio, alberghi e pubblici esercizi

15 Transport, storage and communication

60 Trasporti e comunicazioni

16 Finance, insurance, real estate and business services

69 Credito e assicurazione

13 Electricity, gas and water

74 Altri servizi destinabili alla vendita

17 Community, social and personal services

86 Servizi non destinabili alla vendita

238

Martín Puchet Anyul and Lionello F. Punzo

Notes 1 See Lucas (1988). 2 See a review of these studies in the chapter by Galindo and Catalán in this book. 3 See Barro and Sala-i-Martín (1992) for the cases of the US and Japan, and Sala-i-Martín and Barro (1991) for the European case. 4 For the European debate, see Torres and Giavazzi (1993). 5 See Bertola (1993). 6 See Krugman (1993). 7 See Pinto (1965, 1970), Prebish (1970) part two, and Furtado (1967). 8 See Tamberi and Traù (1999). 9 This notion of convergence is introduced in Punzo (1997) in reference to the European Union. It is clear that the canonical notion is the special case whereby the structural cycle (the attractor) is a point set. 10 Due to Paci and Pigliaru (1997). 11 Convergence implies, in this case, productivities reaching similar levels, and therefore, eventually, their rates of growth will also have to be the same. Taking only the latter, we have to look for evidence in their tendencies, obtained through interpolation, to get closer and to approach a constant common value. 12 As is clear in other chapters of this book (e.g. Cardero’s and Dussel’s) Mexico’s unilateral opening to trade before NAFTA induced a noticeable change in exports towards industrial goods and their concentration in the US market. At the same time, imports of capital and intermediate goods have increased as a proportion of the total and have been geared to the US (see chapters by Dussel and by Cimoli). 13 The European process began with the mining sector, whereas nowadays services are gaining ever-increasing importance. The service sector also played an important role in the transformation of the Italian economy that changed the relations between North and South Italy. 14 This can be compared with the dual view to explain productivity dynamics just recalled, focusing on cross-countries factor mobility. 15 The exercise that follows applies a growth methodology designed by Paci and Pigliaru (1997) for Italy. Two observations must be made here. First, their work focuses on the problem of the productivity growth trajectories across regions, but not across sectors. However, the fact that, as a result of integration, the Mexican economy went through important structural changes at the sectorial level, makes it the more important to measure their effects on the evolution of cross-sector labor productivity. A second observation refers to changes in composition that are important to understand structural transformations. It would be very limited to take only reallocation changes into account, but it enables a comparison with the Italian regions: this will show that, though changing the sectorial mix may not fully explain all sorts of transformations, it did exert a signiﬁcant inﬂuence on the dynamics of the aggregate productivity index. 16 Paci and Pigliaru (1997) obtained similar results regarding the role played by structural change in the trajectory of productivity in the Italian economy. When neither gafv nor gascv are included in the equation, which in their study were estimated on the basis of regional employment composition, the econometric explanation of actual gav rate shows low determination (measured by the R-square or the F-test). On the contrary, when they do appear, the determination coefﬁcient grows and the F-test becomes signiﬁcant. From a regional perspective, it is thus possible to see how the employment structure determined the evolution of productivity. The exercise carried out in this chapter has considered employment’s sectorial structure in explaining the productivity trajec-

Structural divergence and the dynamics of dualism 239

17

18 19 20 21 22 23 24

25

26

27 28 29 30 31

32 33

tory of the Mexican economy. The roles of structural change at the sectorial and at the regional levels are similar. After NAFTA, sectorial employment shares exhibit a dramatic chute for agriculture, forestry and ﬁshing, more than 11 percent, a relatively moderate increase of industrial sectors such as metal products; machinery and equipment; electricity, gas and water, and a more than 2 percent change in the following sectors: construction (5.5 percent); community, social and personal services (5 percent); wholesale and retail trade, restaurants and hotels (2.7 percent); and transport, storage and communication (2.4 percent). The former is veriﬁed by comparing the scales in Figures 9.1 and 9.3; the latter by observing the corresponding variation ranges of each series in Figure 9.3. See Punzo (1995) and (1997), Böhm and Punzo (1992), (1994) and (1999), and Punzo, Abraham and Hutton (1996). Of course there is a multisector tradition as well; and in fact, growth theory was born within the multisectorial approach. Actually a family of growth paths, differing from one another in levels. The predicted steady state is a stationary path. Or else the observed rate of growth of output per head is associated with transitional dynamics. Due to uniqueness and global stability, it is a nonlinear dynamical system with fundamentally dynamical properties typical of a linear one (perhaps, this has been its fundamental appeal!). In particular, in the long run, an attractor in the dynamical system sense, is a monotonic equilibrium path, the simplest of all linear morphologies. It is useful to point this out for later comparison with more complex dynamics that data seems to exhibit. A regime is therefore a class of models, the equivalence class being deﬁned as usual with reference to some property, in this case exhibiting an exogenousonly driven long run path. The introduction of (in the present context, only two) distinct regimes on the gy-line would rationalize the possibility of following different steady state paths (an idea exploited, for example, by Durlauf, who, however, uses it on the y-line). With full employment, this actually implies the uniqueness of the equilibrium solution. We are referring to such things as the issue of convergence, in its manifold qualiﬁcations, also referred to in the text of the chapter. See Baumol, Nelson and Wolf (1994), Durlauf and Johnson (1995), and Quah (1993). This points out that other choices of variables are open. Obviously, we have two bifurcation values for the two parameters: the absolute value of the a coefﬁcient relating to them on each path: in logs, dlogy/dt a dlogi/dt, i I/E, and by whether its sign is positive or negative. If the former is larger than one, we get all regimes above the Harrodian corridor; if the latter is positive, we get those that lie in the ﬁrst and third quadrants. Hence six regimes are generated by the cross product of these two conditions. Of course, here we are locally approximating in the usual way a nonlinear relation between the variables. Whence the consistency between Solow’s and Real Business Cycle theories. Structural dynamics in the Mexican and the US economies is reconstructed through a set of Framework Space graphs: iii a representation of the trajectories of both aggregate economies for the periods: 1970–81, 1981–8, 1988–93, 1993–7 (Figure 9.4); iii a representation of the average rates of growth of 16 sectors (excluding

240

34 35

36 37 38 39 40

41 42

43 44 45 46

47

Martín Puchet Anyul and Lionello F. Punzo

agriculture) in the two economies over the whole period 1970–93 (ﬁgure 9.5); and iii a representation of the average rates of growth of 16 sectors in the two economies over each of the intermediate periods 1970–81, 1981–8, 1988–93 (Figure 9.6). Unfortunately, data for the 16 sectors after 1993 are not yet available. The latter growing faster than the former. It corresponds to economic growth via import substitution in Mexico (1970–81), crisis, macroeconomic adjustment and the opening up of the economy (1981–8), institutional reforms and enforcement of structural adjustment policies (1988–93). See facts and data in the chapter by Cardero. As reconstructed in the chapter by Ruiz. With changes in agricultural land property, the privatization of the parts of the public sector, deregulation of a large number of markets. These facts are described at length by Cardero and by Dussel in this book. Although the ﬁgure range for Mexico is 2.5 times larger than that for the US, there is a quantitative expression of this structural difference. By neoliberal as much as by anti-neoliberal political positions, it is maintained that the NAFTA agreements intended to guarantee the accomplishment of structural and institutional reforms and to tie up, in the long run, Mexico’s economic policy to the objectives of the international ﬁnancial institutions and US political interests. Beyond the issue of their validity, both positions imply that the key changes took place before NAFTA. Both in Whaley (1998) for the former position and in Grinspun and Cameron (1994) for the latter, it is clear that great importance is given to the NAFTA agreements as closing a phase of structural changes crucial for the Mexican economy. Respectively, the pictures of the sectors’ long run and cyclical phases, in terms of regime, or structural dynamics. Note: the scale of the two FS pictures must not be overlooked because it has important implications for interpretation: the long-term FS range for Mexico is ten times larger than that for the US and the FS of the intermediate periods for Mexico is ﬁve times that of the US. Most sectors are between 2 percent above and 2 percent below the line. Moreover, each sector’s productivity and accumulation grow at similar rates and the differences between sectors in both rates usually do not go beyond 3 percent. It should be enough to just compare their ranges: (40, 40) for accumulation rates of growth, (5, 5) for rates of productivity growth. Of course, sectors move about, though their cloud does not, or at least not much. This is the point. First an accumulation and productivity growth phase (197081), then a decumulation phase with productivity growths based on innovation (19818) and, ﬁnally, a new accumulation phase with some sectors remaining in the same regime of the previous phase (198893). See Puchet (1996) and Punzo (1997).

Bibliography Barro, R. and Sala-i-Martín, X. (1991) ‘Convergence across states and regions’, Brookings Papers on Economic Activity 1: 86–125. Barro, R. and Sala-i-Martín, X. (1992) ‘Regional growth and migration: a Japan–US comparison’, Journal of International Economics 6: 312–46.

Structural divergence and the dynamics of dualism 241 Baumol, W.J., Nelson, R.R. and Wolff, E.N. (1994) Convergence of Productivity, New York: Oxford University Press. Bertola, G. (1993) ‘Models of economic integration and localized growth’, in Torres, F. and Giavazzi, F. (eds) Adjustment and Growth in the European Monetary Union, Cambridge: Cambridge University Press. Böhm, B. and Punzo, L.F. (1992) ‘Detecting structural change: a scheme for the comparison of Austria and Italy in the seventies and eighties’, in Clauer, O., Kalmbach, P., Pegoretti, G. and Segnana, M.L. (eds) Technological Innovation, Competitiveness, and Economic Growth Berlin: Duncker and Humbolt. Böhm, B. and Punzo, L.F. (1994) ‘Dynamics of industrial sectors and structural change in the Austrian and Italian economies, 1970–89’, in Böhm, B. and Punzo, L.F. (eds) Economics Performance. A Look at Austria and Italy, Heidelberg: Physica Verlag. Böhm, B. and Punzo, L.F. (1999) ‘Productivity-investment ﬂuctuations and structural changes’, Working Papers Series 2, Industrial Dynamics and Employment in Europe Network IDEFI – CNRS, in Punzo, L.F. (ed.) (2001), Cycles, Growth and Structural Change, London: Routledge. Durlauf, S.N. and Johnson, P. (1995) ‘Multiple regimes and cross-country growth behaviour’, Journal of Applied Econometrics 10: 365–84. Furtado, C. (1967) Teoría y Política del Desarrollo Económico, Mexico: Siglo XXI, 1976. Grinspun, R. and. Cameron, M.A (1994) ‘Restructuring North America: the impact of unequal integration’ in Georgakopoulos, Th., Paraskevopoulos, Ch. C. and Smithin, J. (eds) Economic Integration between Unequal Partners, Aldershot: Edward Elgar. Krugman, P. (1993) ‘Lessons of Massachusetts for EMU’, in Torres, F. and Giavazzi, F. (eds) (1993), Adjustment and Growth in the European Monetary Union, Cambridge: Cambridge University Press. Lucas, R.E. (1988) ‘On the mechanics of economic development’, Journal of Monetary Economics 22: 3–42. Paci, R. and Pigliaru, F. (1997) ‘Structural change and convergence: an Italian regional perspective’, Journal of Structural Change and Economic Dynamics 8: 297–318. Pinto, A. (1965) ‘Concentración del progreso técnico y de sus frutos en el desarrollo latinoamericano’, El Trimestre Económico, January–March. Pinto, A. (1970) ‘Naturaleza e implicaciones de la “heterogeneidad estructural” de la América Latina’, El Trimestre Económico, January–March. Prebish, R. (1970) Transformación y Desarrollo: la Gran Tarea de América Latina, Mexico: Fondo de Cultura Económica-IDB. Puchet Anyul, M. (1996) ‘Análisis de la integración económica: aspectos del caso EUA-México’, in Bizzozero, L. and Vaillant, M. (eds.) La Inserción Internacional del Mercosur: ¿Mirando al Sur o Mirando al Norte?, Montevideo: Arca, 99 y ss. Puchet Anyul, M. (1999) ‘Cambio estructural e integración de la economía mexicana. Un enfoque cualitativo de insumo-producto’, Working Paper, México: UNAM. Punzo, L.F. (1995) ‘Some complex dynamics for a multisectoral economy’, Revue Economique 4, March. Punzo, L.F. (1996) ‘Industrial dynamics and structural change; a framework for the

242

Martín Puchet Anyul and Lionello F. Punzo

analysis of sectoral dynamics in a set of countries’, ISER Working Paper 401, Osaka: Osaka University. Punzo, L.F. (1997) ‘Ciclos estructurales y convergencia durante los procesos de integración económica’, Revista de Economía, Segunda Epoca IV, 2, Banco Central del Uruguay, 3 y ss. Punzo, L.F., Abraham, R.H. and Hotton, S. (1996) ‘Experimenting with a cellular model of Italian economic development’, Working Paper. Quah, D. (1993) ‘Empirical cross section dynamics in economic growth’, European Economic Review 37: 426–34. Tamberi, M. and Traù, F. (1999) ‘Industrializzazione e sviluppo regionale nell’economia italiana’, Rivista di Politica Economica LXX, III, 1: 21–66. Torres, F. and Giavazzi, F. (eds) (1993) Adjustment and Growth in the European Monetary Union, Cambridge: Cambridge University Press. Whalley, J. (1998) ‘Why do countries seek regional trade agreements?’, in Frankel, J.A. (ed.) (1998) The Regionalization of the World Economy, Chicago and London: University of Chicago Press.

10 Regional policies in the North American and European integration agreements Sárah Martínez Pellégrini

This chapter describes the two vastly different integration and development strategies embodied in the North American Free Trade Agreement (NAFTA) and the European Union Treaty (EUT). Despite their obvious differences, both processes create space for regional strategies which can address complex issues related to the well-being of population in both blocs. In fact, regional policies have become a highly promising means of reconciling economic and social development strategies. The importance of discussing the role played by regional policies comes from their being one of the most successful ways to reconcile economic and social strategies. In other words, they could be an answer to the increasing need for integrated approaches for development. The chapter starts with a brief description of NAFTA and the EUT to establish some of the basic differences in the two settings. Then we give an overall description of the regional policy implemented in the European Union and the environmental policy initiative sponsored by NAFTA (which could be taken as a regional proposal). The chapter concludes with some comments on how and why regional policies may provide a possible solution to some of the socio-economic problems found in integrated areas and how the transborder cooperation between Mexico and the US could become a broader program of regional cooperation for development. It is important to underline that our aim is to give an overview which suggests that regional approaches might be an option for development. We are by no means providing an evaluation of the speciﬁc social impacts of particular policies implemented in the Union other than the broad regional policy setting.

10.1 The two treaties and their objectives1 The NAFTA and EU regions comprise about 14 percent of the world’s population, and about 60 percent of the world’s GNP (OECD 1995). This means that both integration efforts can substantially alter the international economic environment just by their relative weight as trade partners in the world.

244

Sárah Martínez Pellégrini

Our discussion also has a qualitative dimension, as the blocs stand for two contrasting policy approaches. The integration models followed by the two groups of countries reﬂect the main driving forces of the respective agreements and their ultimate objectives. The frameworks set by both the NAFTA and EU agreements deﬁne their common policies and cooperation and harmonization limits. The 15 members of the EU agreed to establish an economic and monetary union while the three nations involved with NAFTA chose a free trade area (Secretaría de Comercio y Fomento Industrial 1992, NAFTA art. 101). A free trade area means that ‘member nations remove all trade impediments among themselves but retain their freedom with regard to the determination of their policies vis-à-vis the outside world’ (El-Agraa 1989: 1). So Mexico, Canada and the US will maintain their respective foreign trade policies with the rest of the world. The EU started as a customs union (Tratado de Roma. Acta Unica Europea 1992, Rome Treaty art. 9) and has evolved, through new treaties complementing the initial one, into a common market established 1 January 1994. This new feature implied a stronger commitment, with the real abolition of barriers to the free circulation of goods, capital, services and persons. More recently, by the beginning of 1999, the monetary union was launched between some of the members. In both cases all kinds of internal barriers to trade were meant to disappear, and are on their way toward doing so (NAFTA arts. 102 and 309 and EUT arts. 3, 9, 12 and 30). In the EUT, articles 18 to 29 establish common external tariffs on imports from non-member countries. Both agreements share progressive inclusion policies for new signatories as is clearly seen in recent changes in both groups of nations. For the North American case, it may be considered that the previous treaty between Canada and the US was extended to Mexico. The agreement allows a 15 year period for the less-developed member of the club to progressively dismantle all the tariffs. Annex 302.2 of the agreement establishes that, by the year 2004, 99 percent of all goods will be free of any tax within the three countries territory, and by 2009 the percentage is expected to reach 100 percent of all commodities. The four extensions of the European group have led to speciﬁc integration deals with the new members. In each case the timing and procedures were tailored according to the particular circumstances of the new entrants and their relationship with the incumbent members (competition issues with newcomers caused great delays in some of the cases). Services are treated in very different ways in the agreements. In Europe, service providers enjoy the right to have establishments in any country where they work (EUT arts. 52 to 58) while in North America the need for local establishment of providers is not considered. In both cases, services from any of the members enjoy national treatment in each and all of its partners (NAFTA art. 1202 and Maastricht Treaty art. 60).

Regional policies 245 NAFTA goes beyond the EUT as it makes the national treatment policy explicit for all levels of government. In Europe such ‘institutional oligopolies’ have been one of the biggest obstacles to achieving the Common Market (Cecchini 1988). What is more, NAFTA obliges members to give most favored nation treatment to any service provider if the country where he operates has any agreement of the sort with a third country. The philosophy of the agreement is that providers should receive the most favorable treatment accorded by the host nation to any business of the same branch (NAFTA art. 1203). As far as capital is concerned, after deciding to establish a common market (Maastricht Treaty), the European countries made the move towards free factor mobility, including people and capital. The EUT explicitly forbids ‘every restriction to capital movements among member states or with third countries’ and ‘any restrictions to payments between member countries and third nations’ (art. 73B). NAFTA members agreed to follow the same the rules established by the International Monetary Fund for capital transactions and movements as any other country involved with the IMF Agreement (NAFTA art. 2104). Differences between the two blocs arise when dealing with foreign investment, in spite of their common interest in attracting as much foreign capital as possible. The degree of policy harmonization in the integration model adopted by each group again explains the extent of the common policies. The North American agreement uses the national treatment and most favored nation rules as the guidelines for capital issues among the three partners (NAFTA arts. 1102, 1103 and 1106) and does not specify any need for further harmonization for capital movements between Mexico, Canada or the US and third nations. Each of the three partners is completely free to determine its investment policy following NAFTA’s recommendation of ‘substantially promoting investment opportunities in the territories of the member’ (NAFTA art. 102). The European Union members agreed that ‘in spite of seeking the free ﬂow of foreign capital among the members and third countries . . . the European Council will be able . . . to adopt special measures regarding direct investment ﬂows coming from or directed to third countries’ (EUT art. 73C). From its very start the European Community emphasized workers’ free transit in the area as a basic border liberalization measure for the group. The labor force has been understood as another production factor that will be more efﬁciently used if allowed to ﬂow according to economic requirements. The EUT dedicates a whole chapter to this subject in keeping with the founders’ tradition. Free labor movement is warranted by article 48, which gives European citizens the right to respond to any job offer, and thus establish their residence in any of the countries of the Union during and after their working

246

Sárah Martínez Pellégrini

period. Furthermore, any social assistance and social security payments will accumulate independently of the nation where the entitled person has worked before. This concession achieved by the European Social Chart was one of the hardest proposals for EU members to agree upon. The Maastricht Treaty introduced an interesting concept complementary to labor rights, i.e. Union citizenship. It is the political counterpart to the possibility of economic participation in any of the member states. This ‘condition’ is attained merely by being a citizen of one of the EU members (EUT art. 8). The main aim of the establishment of the European citizenship was to socialize a project, the Union, characterized by critics as a monopoly of economists and politicians and far away from Europe’s people. Within the North American area labor mobility is limited to that ﬂow strictly necessary to conduct business. The only preferential travel considered is temporary trips for business, trade, or investment projects. Workers can also be transferred within a ﬁrm from one country to another. These measures merely complement an agreement limited to free trade.

10.2 The institutional framework developed by the agreements The institutional framework and procedures installed by NAFTA and the EUT clearly reveal their deep differences regarding the underlying integration processes. The European bloc reaches such a degree of integration among its members that it requires well-developed regulations and procedures2 and the existence of supranational institutions.3 The EUT establishes the functioning of ﬁve main institutions as the Union’s institutional backbone: the European Parliament, the European Council, the European Commission, the Court of Justice and the Accounts Court. All of these institutions have evolved in their composition and role in the integration process from their earliest creation to the present day. NAFTA created a few common organizations but it cannot be said that any institutional framework was designed to implement policies. The main reason is that the pursued objective is to increase free trade and not to establish further common policy-making tools. A Free Trade Commission was created to ensure the three members’ adherence to the rules in order to enhance free trade among them (NAFTA art. 2001). This Commission, formed by representatives of all of the members, is supposed to meet at least once a year to follow the evolution of the free trade area towards freer exchanges. To deal with the everyday operation of the agreement, each member country was obliged to create a Secretary (NAFTA art. 2002). The simplicity of the North American structure contrasts sharply with the complexity of the European arrangement. The two alternatives are the

Regional policies 247 main examples followed by other integration initiatives in other parts of the world. The Association of Southeast Asian Nations (ASEAN) is an exception to these models as it is more of a ﬂexible integration system closer to a network organization instead of a more deﬁned structure.

10.3 Policies involved in the agreements4 Economic and monetary policies are especially important when speaking about regional policies because they deﬁne the margin left for territorial approaches and integrated planning initiatives within the international economic integration schemes considered here. Any international integration process will require a certain degree of economic policy coordination among the involved countries. Deeper integration will necessarily increase common decision-making processes. Thus, by the very nature of its integration option, the European Union is impelled towards a strong coordination and harmonization of its members’ national policies. Furthermore, with the monetary union coming about, NAFTA countries are required a much lesser degree of harmonization in order to maximize the beneﬁts from their free trade area. Both NAFTA and the EUT are centered in three basic aspects of economic activity, namely investment, trade and competition. This emphasis derives from the fact that all the countries involved in the agreements have capitalist economies and one of their targets is to promote the three basic aspects in order to improve the market’s functioning. For the EU, monetary and economic policy are fundamental areas of cooperation as the Maastricht Treaty established coordination of both through the European Council (EUT art. 103) in order to establish conditions for the monetary union to come. The Treaty of Rome (1957) – by which the European Economic Union was created – and the Unique European Act illustrate the evolution from suggested cooperation among members towards necessary coordination. The Council has the power to recommend measures for member countries to revise their domestic policies in order to achieve more homogeneous results throughout the European bloc. These recommendations are meant to achieve greater convergence of policy results among nations and regions to assure a deepening of integration. Five convergence criteria were deﬁned as means to achieve a successful monetary union (third phase of the so-called Monetary and Economic Union Plan (MEU)). The Maastricht Treaty (art. 109J and EUT’s Sixth Protocol) deﬁnes them as follows: 1

2

Keep inﬂation under a maximum of 1.5 percent more than the average of the three best national results within the Union, for the 12 months preceding the monetary integration. Maintain public deﬁcit under 3 percent of GDP.

248 3 4 5

Sárah Martínez Pellégrini Limit public debt at a maximum of 60 percent of GDP. Limit currency ﬂuctuation within the narrow band of the European Monetary System (EMS) for at least two years. Have an average nominal long-term interest within a margin of 2 percent above the average for the three countries with lowest inﬂation rates.

Another step towards a common currency was the establishment of the European System of Central Banks (EUT, Third Protocol) as a warranty against high inﬂation rates within the region. All national central banks linked together were much more likely to develop coordinated policies. The system moved forward with the creation, in December 1998, of the Central European Bank (especially relevant for members already into the common currency group). By contrast, NAFTA, as a free trade agreement, does not provide any guidelines in the above mentioned matters. It only contains a general advocacy for cooperation for the mutual beneﬁt of the associated countries with a few speciﬁc concerns further mentioned. The evolution of both associations seems to conﬁrm the idea that mere economic integration tends to bring about enhanced cooperation in some other ﬁelds. The European Union’s case demonstrates this pretty convincingly, even though it could be argued that from the very start its underlying idea was to achieve a complete political integration. Its progressive extension can be interpreted as the result of that increased cooperation with third neighboring countries and among members themselves (each time, apart from widening the Union, changes have launched discussions resulting in deeper integration). There are two facts suggesting that NAFTA could follow a similar pattern. First, the American assistance to Mexico after the 1994 devaluation is hardly understandable unless it is assumed, for some reason, that the US (at least some of its politicians) felt closely involved with their southern neighbor. The second case of interest is the common environmental border policy established through a parallel agreement to NAFTA that is discussed further below. A big polemic within the EU concerns ‘common social policy’. Construed broadly as comprehending labor, environmental, and regional policy, one must consider how it is generally handled by the Union prior to any discussion about regional issues.5 Article 117 of the EUT states the commitment for all the member governments and supranational institutions to promote improved working and living conditions in the area. To achieve this goal, the European Social Charter was signed in 1989 by all members, except the UK, and was ratiﬁed as an annex to the Treaty. The involved countries also created the European Social Fund (ESF) vide article 123 of the Treaty, to improve employment opportunities in the area. The three North American countries subscribed to a more limited

Regional policies 249 commitment dealing with labor issues included as an annex to the main body of NAFTA. It sets up the basis for cooperation among Mexico, Canada and the US in the areas of migration, health and hygiene at work places, productivity, quality and innovation programs and labor statistics and information. As for environmental policy, a somewhat similar pattern occurs for both agreements. The European Council can deﬁne land use and planning, ﬁscal, and energy use policies (EUT art. 130S) in order to contribute to a more rational use of resources, to protect the population’s health and to preserve, protect and improve the environment. The North Americans deﬁned a speciﬁc framework for cooperation in environmental issues only for the US–Mexican border, as will be later described. In general terms the only agreement by the partners is to respect each country’s right to determine its own environmental laws and policies and to apply them. Agricultural policies have been among the more conspicuous topics of negotiation in both the North American and the European case. Competition among members is particularly strong in this sector, which, on the other hand, also has important social and economic consequences. A slowdown or stagnation in agriculture tends to be located in lagging regions, and thus is of special relevance for territorial development proposals. Protectionism and populism are often linked to national positions about rural issues. The European Common Agricultural Policy (CAP), one of the most criticized not only by EU members but also by third countries, regulates agricultural production and commercialization. The objectives of CAP are to contribute to a greater equity in income between rural and urban population, to stabilize markets and assure reasonable prices on the market (EUT art. 39). NAFTA just mentions a commitment by its members to respect GATT rules when providing any kind of aid to this sector. The agreement also contains a timetable for the progressive elimination of import tariffs on Mexican primary sector goods.

10.4 Regional policies From the above it is clear that, in the European case, regional differences have been an important concern from its early stages, while it is not so in the North American case. NAFTA assumes that there is no need for common initiatives for redistribution as national policies and the beneﬁts from free trade will spill over to the less developed areas too. By contrast, in the European Union there is a strong advocacy for common regional development policies to tackle the situation of the lesser developed areas. Article 130A of the EUT says the Union will reinforce its social and economic cohesion in order to ‘decrease disparities among the development level of the regions

250

Sárah Martínez Pellégrini

and the backwardness of the less developed, including rural ones’. To do so, national economic policies must also comply with this objective (EUT art. 130B) while the Union itself fosters the initiative through the six ﬁnancial instruments designed for structural policy purposes. The so-called structural funds are the European Fund for Regional Development (EFRD), the European Social Fund (ESF), the European Fund for Orientation and Warranty-Orientation section (EFOW) and the Fisheries Orientation Financial Instrument (FOFI). Two other sources, governed by different principles, also contribute to cohesion and regional development projects, the Cohesion Fund and the Investment European Bank (IEB). The existence of the structural funds, and their evolution over the past 15 years, are good indicators of the path followed by regional policies within the Union. The way the ﬁrst agreements deﬁned territorial policy objectives was somewhat confusing and lumped them with income redistribution. With time, regional policies have become one of the main tools to achieve greater development, especially in lagging areas. The First Report on Social and Economic Cohesion (Comisión Europea 1997) enhances, more than once, the importance of the 1988 reform of the funds to improve their results. Between 1985 and 1999 the resources for structural policies increased from 3,700 million ECU to 170,000 (Comisión Europea 1997). The same report asserts that 85 percent of these funds were invested in projects pertaining to objectives 1 (70 percent), 2 (11 percent), 5b (4 percent) and 6 (0.5 percent) and only 15 percent of the total was assigned to objectives 3, 4 and 5a.6 These objectives, which are the Union’s classiﬁcation to allot resources from the structural funds, are: •

•

•

•

Objective 1: foster the adjustment and development of less developed regions, deﬁned as the ones with a GDP below 75 percent of the Union’s average. The list of these areas is revised every ﬁve years and initiatives are funded through EFRD (80 percent of its funds go to these projects), ESF, EFOW orientation and IEB. The European Community for Steel and Coal (ECSC) also contributes some resources. Objective 2: restructure regions critically weakened by industrial crisis, deﬁned as those with higher than average unemployment, particularly in industry. This list of regions is updated every three years. Funds working for these areas are SEF, EIB and ECSC. Objective 3: reduce long-term unemployment for people older than 25 and without a job for more than a year. SEF, EIB and ECSC fund programs of this nature. Objective 4: promote young peoples’ access to labor markets. The population targeted here is under 25 years of age and usually looking for their ﬁrst job. The funding, in this case, is by SEF, EIB and ECSC.

Regional policies 251 •

•

Objective 5 is divided in two complementary sections to back the CAP and is funded through the EFOW: 5a) transform and adapt productive structures in agriculture and ﬁsheries, in addition to improving their commercialization. 5b) foster initiatives for development in rural areas with high unemployment in the primary sector, low income levels for rural workers, and low GDP per capita. Objective 6: assistance for scarcely populated regions, with under eight inhabitants per km2. Funded by the EFRD.

The distribution of structural development funds has had a great impact in the emergence of country groupings gathered around common interests. It has changed the nature of negotiations and the relative weight of the ‘North’ and the ‘South’. With the Union’s expansion, the objectives have evolved (the latest example was the creation of the sixth objective when the Nordic countries entered the group) to respond to the increasing need for a strong development policy to face the increasing heterogeneity among member countries. The importance of structural policies has been strengthened by the fact that, on the path to the monetary union, a greater correlation was observed between structural problems and the impossibility of meeting convergence criteria, than between per capita income and possibilities to join the common currency group. Thus it seems of great interest to work further on this hypothesis, and if proved, to use it as an indicator to evaluate regional policies and the relationship between macro and micro aspects. When turning to the results of regional policies in the EU, it is hard to point to undeniable evidence in any direction, but a number of conclusions can be drawn from the European experience that could be of help for other areas. Critics of regional policies argue that European regional policies have not substantially reduced divergences among regions, or even among countries. Using per capita income and unemployment as measures of convergence/divergence, differences were reduced by about 2 percent per year in the long run. Meanwhile the expansions of membership in 1973, 1981 and 1986 more than doubled disparities as the new joiners were Ireland, Greece, Spain and Portugal. While the two percent result is not quantitatively spectacular, it has been argued that the policies should be judged by their ability to start sustainable development processes in declining or underdeveloped areas. The different positions acquired by some of the lagging regions within the Union are one of the lines along which changes are clear. Another is the greater inclusion of many areas, and even of two of the less developed countries – Ireland and Spain – in the international ﬂow of exchanges. A third important aspect is the building up of response capability for some of the regional systems, in other words, the awakening of rigid structures to provide their own solutions to local problems.7

252

Sárah Martínez Pellégrini

To warranty the regions’ representation at the Union’s level, the Maastricht Treaty created a consultative Committee of the Regions (EUT art. 198A). Its members are regional and local representatives and, in spite of receiving all the Commission’s reports, it does not have an active role in regional development policies. For some governments, the creation of the Committee was an attempt by the Union to take sovereignty away from the national level, not only from above, through the common institutions, but also from below through subnational levels of government. This reaction illustrates perfectly the current tension created by the need for institutional change at all levels, not only in the productive system, but to respond to new challenges, mainly the ones related to globalization. As for the other agreement, NAFTA does not mention any cooperation for regional development; the farthest it goes is establishing the bases for later cooperation that will be oriented to enhance the treaty’s beneﬁts (NAFTA art. 102). One explanation for the lower emphasis given to regional policies could be that the average difference in the regional per capita income in the US is about a half of what it represents in Europe (Primer Informe sobre la Cohesión Económica y Social 1996). The trend is for this difference to grow at twice the rate it does in Europe, but start from a lower level. The problem with that standpoint is, of course, that Mexico sharply increases the divergence in the level of development for the North American area. If no corrective measures are proposed on time, this could seriously challenge the economic health and cohesion of the three American partners. There is already collaboration on a speciﬁc regional matter that is worth mentioning: the work of the Border Environment Cooperation Commission (BECC) and the North American Development Bank (NADBank). The environmental issues could be a ﬁrst step towards a bi-national approach to integral development policies, as it deals with some of the problems derived from a basic structural handicap, which is the lack of certain kinds of infrastructure. As the BECC itself states, the purpose of the Commission is ‘to help preserve, protect, and enhance the environment of the border region in order to advance the well-being of the people of the United States and Mexico and achieve sustainable development.’ (BECC 1995: 1). The projects undertaken are located within 100 km or 62 miles on both sides of the border between Mexico and the US and projects outside this area will be considered if the Mexican Secretaría de Desarrollo Social and the US Environmental Protection Agency approve them. The basic condition in the latter case is that the action will solve a problem that affects the 100 km stripe. BECC is concerned with environmental and health issues and should cooperate with the NADBank and other institutions and private sources to improve environmental infrastructure in the border region.

Regional policies 253 The above approach could match most development plans. Two possible bridges with a regional policy approach are deﬁned in the statement of the Commission’s purpose. The ﬁrst is ‘to advance the well-being of the people’, and second is ‘to achieve sustainable development’ (BECC 1995: 1). In general terms, these are the targets of any development initiative and, of course, regional development policies. What is even more, the area on which the program is applicable has been deﬁned according to its distinctive border features. According to endogenous development theories this could be seen as a regional policy instrument. It is endogenous because its starting point is a diagnosis of the conditions in an area, as identiﬁed by local agencies. It also targets development issues as far as it deals with improvement of living standards in a community. And it seems a great example of binational management because it results from the Agreement on Cooperation and Improvement of the Environment in the Border Area (La Paz Agreement), a parallel agreement of NAFTA. What the BECC-NADBank system provides can be summarized as a procedure to obtain funds for what they call environmental infrastructure (that dealing with an environmental or health problem). First there is a certiﬁcation process which requires any applicant to work on the proposal with BECC’s assistance. Then, if all the requirements are met and the project is certiﬁed, it moves on to an evaluation of its funding feasibility by the NADBank. BECC gives preference to projects relating to water pollution, wastewater treatment, municipal solid waste management and related matters. The submission of proposals consists of two steps: a project pre-proposal and a project proposal. The pre-proposal is used to evaluate the applicant’s need for technical assistance for the second step. The proposal involves ‘provision of detailed project proposal information, based upon the proper engineering, environmental, economic, ﬁnancial and social studies’ (BECC 1995: 5). According to the BECC’s 1995 document, the certiﬁcation criteria cover the following eight areas: 1

2

3

General project description: the project should meet the location restrictions already mentioned, and work tasks and budgets must be reasonable. Environment and human health: the projects must address a critical human health and/or environmental need, protect the environment and human health in the affected area, respect all local, regional, state and federal environmental regulations before their operation begins, submit an environmental assessment before certiﬁcation, and comply with applicable international agreements. Technical feasibility: appropriate technology must be used in a cost-effective manner at each stage of the project to achieve its purpose.

254 4 5 6

7

8

Sárah Martínez Pellégrini Economic and ﬁnancial feasibility: the proposal must be sufﬁcient to cover debt amortization and further operation of the facility. Social issues: projects must comply with all cultural resources (namely historical, archeological, and ethnic) regulations. Community participation: all applicants must submit and implement a BECC approved Community Participation Plan. After implementation of the plan the report shall indicate the degree of public acceptance of the project. Operation and maintenance: programs must also include provisions for emergency planning, an occupational health and safety plan, training plan for operation and maintenance personnel and if necessary, a pollution prevention plan, facility closure plan, and postclosure plan. Sustainable development: projects must be based on the principles of sustainable development and thus enhance long-term planning and human and institutional capacity building.

The same BECC document (BECC Project Certiﬁcation Criteria) published in 1995 shows substantial changes. The areas and criteria are redeﬁned as follows: 1

2

3

4

5

General: the project must fall under BECC priority areas, respect the location rules, provide an adequate description of work tasks and conform with International Treaties and Agreements. Human Health and Environment: all proposals should address a critical human health and/or environmental need, comply with applicable environmental and cultural resources laws and regulations and submit an environmental assessment before certiﬁcation. Technical feasibility: appropriate technology must be used to match the local user ability to operate the facilities to avoid dependency on high level of inputs to operate the system. An operation and maintenance plan must exist from the start, including provisions for emergency planning, an occupational health and safety plan, training plan for operation and maintenance personnel and, if necessary, a pollution prevention plan, facility closure plan, and post-closure plan. There should be compliance with applicable regulation and standards (local, regional, state and federal). Financial feasibility and project management: revenues from project operation or other sources must be sufﬁcient to cover debt amortization and further functioning of the facility. The proposed fee/rate model should meet this concern. The applicant should demonstrate his capacity to provide service at a reasonable price and undertake the necessary accounting and ﬁnancial initiatives and reports to keep the facility working. Community participation: all the applicants must submit and imple-

Regional policies 255

6

ment a BECC-approved Community Participation Plan. After implementation of the plan the report shall indicate the degree of public acceptance of the project. Sustainable development: projects must be based on the principles of sustainable development and thus enhance long-term planning and human and institutional capacity building. They must conform to local and regional conservation and development plans, achieve a reasonable degree of natural resource conservation and have a positive impact on community development.

The changes found between these two lists of criteria reﬂect three interesting shifts towards a more integral approach to local development. The ﬁrst is the change of the mandate to comply with ‘cultural resources laws and regulations’ from being the ﬁfth item (called ‘social issues’) in the former list to become part of the second point (‘human health and environment’) in the second list. This implies reinforcement of the systemic approach as cultural and environmental frameworks seem to be considered as connected parts of the context of the project. Another important change is consideration of the project’s ‘technical feasibility’ more broadly to include not only the suitability of the proposed technology but all the plans to operate the infrastructure, along with the respect for local standards. The proposal (infrastructure process) is understood to include technical feasibility as a long-term approach. It goes beyond mere consideration of technology itself and includes its functioning and some of its main possible effects. ‘Operation and maintenance’ included in ‘technical feasibility’ mean a greater effort for consistent planning that integrates all the elements directly related to the infrastructure itself from the very start. The third revision in the second list of criteria is that ‘sustainable development’ is more clearly deﬁned, and the latter version mentions explicitly the need for a ‘positive impact on community development’. This concern suggests the need to evaluate not only environmental impacts, but also the infrastructure projects as development initiatives. During 1995, 1996 and 1997 the technical assistance provided by BECC has been: Table 10.1 Technical assistance provided by BECC 1995–7

No. of Contracts No. of Projects Amount

1995

1996

1997

Total

8 (Mexico: 8)

49 (Mexico: 44 USA: 5)

14 (Mexico: 13 USA: 1)

71

7 $112,758

20 $734,040

11 $360,431

38 $1,207,229

Source: BECC (1997) Annual Report: 11.

256

Sárah Martínez Pellégrini

The data on projects and contracts for the past three years sound promising for Mexico because most of the initiatives were approved for the southern country. Meanwhile the challenge is the next step: meeting the NADBank’s requirements. The bank, as stated before, is BECC’s sister institution. It is a binational ﬁnancial institution that provides a combination of public and private capital for the certiﬁed projects. It pulls together different funding resources and organizes them as a ‘team of players’ (NADBank 1996: 5) trying to encourage private ﬁnancing and risk sharing among public and private actors. The guidelines followed by the bank respect BECC’s requirements (certiﬁcation, impact on the areas 100 km from the border, priority projects dealing with water and solid waste issues, etc.). Yet in spite of this, most of the certiﬁed projects don’t succeed in obtaining funding. At some points the chain breaks down as there is a substantial change from a policy prioritizing development and socio-environmental concerns to a behavior similar to most ﬁnancial institutions other than development banks. The lack of coherence between BECC certiﬁcation and NADBank’s funding approval is a problem that is particularly striking on the Mexican side. Many small communities don’t possess the resources that would make them viable candidates for the loans, in spite of their key role as polluters in the border area (mainly due to the lack of that very same organization needed to become eligible for the bank’s help). As usual with policies, the problem is in the implementation. A closer look shows certain inconsistencies which seriously threaten the project’s success within its own organization as well as external factors that challenge its continuation. The incoherence between the approaches of BECC and the NADBank shows the lack of a real policy choice underlying the creation of these two institutions as noted by some critics of the agreement that initiated them.

10.5 Institutional change Globalization is raising a host of new issues that increasingly fall outside the traditional focus and methods of national policies. That is why it is important to make reference to some aspects of the institutional changes that may be needed to achieve more efﬁcient results, especially from regional policies. The response to the movement towards the establishment of larger units created another tendency focusing on smaller communities. Regional and local authorities, organizations and societies started to demand a certain level of autonomy in solving speciﬁc problems close to their environment. They called for a sufﬁcient scope of action to face everyday obstacles according to their own priorities. Emphasizing regional identities and units has provided a needed sense

Regional policies 257 of security to people who felt lost in an open international context. The human dimension is somewhat regained at regional and local levels as social agents recover ways to have a more active participation in political and economic activity. Moreover, devolution of decision making may be considered to be more the result of an attempt to improve economic efﬁciency than just a political aim in itself. The double-headed thrust mentioned above has introduced changes at three main levels: structural, microeconomic, and spatial. The junction of the three justiﬁes the call for new organizations at every level, and especially in the way politics, policies and the economy are ruled and run. The need for institutional changes – that is to say, a reshaping of the existing institutions – comes directly from the fact that a new distribution of tasks has been taking place in most of the developed and developing countries. It does not mean that central governments have lost their rationale; it is only that within the current context their role is substantially different. The emergence of intermediate institutions is a way for states to share responsibilities in order to achieve better results. As central institutions must each day face more intense and complex relations with other countries, domestic problems have turned out to be an adequate ﬁeld for regional and local actions. From this point of view we could make a distinction between macro-decisions as the domain of central powers, and micro-decisions as regional and local ﬁelds. Macro decisions refer to the design of the directive lines of the countries’ policies or the articulation of the nations’ main goals or priorities. Micro-decisions would then be the speciﬁc implementation of these goals through particular measures ﬁt to each territory’s circumstances. On the one hand, central governments are concerned with general objectives selection; coordination of the different initiatives undertaken within the country, among themselves, and with actions being run in partner or closely-related nations; the provision of adequate infrastructures and services; and the organization of whatever activities that require economies of scale on national levels. On the other hand, regional and local authorities are – or could be – responsible for the organization of exchanges between local entrepreneurs and institutions, both among themselves and with each other; the identiﬁcation and promotion of local endogenous potential to embed development plans in it; giving daily support to local agents’ projects; the detection and training of possible entrepreneurs; and the provision of speciﬁc services to productive activities – as export services and local markets organization – to foster local development and improve local living standards. We are assuming that no matter how each community attempts to achieve it, the general aim pursued is the establishment of sound development processes able to improve population living conditions, no matter whether it is on a local, regional or national level.8

258

Sárah Martínez Pellégrini

The shift from centralized systems to decentralized structures only reﬂects the substitution of one policy by another, not to refashion society entirely. Policies can be changed, even if these do require time; societies must evolve. One remaining problem is that the newly activated objectives have not always been given an institutional space. So, many times, the operating powers clash against each other; they limit, and sometimes even render ineffective, initiatives undertaken by the other entities. Decentralization seeks to make decisions closer to involved people to allow them stronger participation in the design of the communities’ future. It looks for a renewal of public administration so as to increase local and citizen responsibility. The mobilization of regions and territorial groups gives birth to more ﬂexible forms of organizations able to adapt their action to real requirements more accurately and with shorter delays. Local authorities are closer to everyday life and thus better situated to register real needs and the most appropriate means to satisfy them. Furthermore, initiatives promoted at local levels are often more efﬁcient since, having been designed by local agents, they beneﬁt from community support and participation to a much higher degree. Local involvement with the initiatives has been identiﬁed as a crucial success factor in this.

10.6 Final remarks Up to now, it has been difﬁcult to take for granted that regional policies assure regional convergence, even if they seem to be able to diminish structural differences in certain contexts. For the European case at least two of the Cohesion Funds countries – Spain and Ireland – have experienced substantial structural change that could hardly have occurred without the Union’s guidelines. These structural changes must be seen as progress because they have enabled some lagging regions to enter a development process never seen before in spite of their limited achievements in some social aspects.9 There are many points to consider in relation to how regional policies are designed and their efﬁciency increased. But they offer a good alternative to mesh – at least partially – social and economic goals, a basic requirement for any integral approach to development issues. An evaluation of speciﬁc impacts by region and indicator (employment, average income, working conditions, etc.) are basic in order to avoid these policies losing their systemic bias. It is important to underline that development policies, regional as well as national, must be seen as a long-term process with targets being redeﬁned as the social consensus about well-being evolves. Then the evaluation of such initiatives needs a broader approach in terms of the community impact of the actions undertaken. The European Economic Community (EEC) was designed after the

Regional policies 259 Second World War as a means to achieve political cooperation among the European nations in order to reduce the risk of renewed conﬂicts. From the very start, political integration was an unspoken target for many of the project’s fathers. Unspoken because national sovereignty concerns, especially sensitive at a war’s aftermath, banished any such possibility. By 1956 the realizable link among nations rising from destruction was cooperation for economic development which could bring them back to their previous welfare. That could partially explain the evolutionary step that came before stronger political bonds and the EU’s more recent focus on regional development priorities. As far as the Mexican–US border region is concerned, the bi-national approach to environmental problems is an innovative approach that implicitly brings about some of the changes introduced in regional development during the last two decades. That is precisely why it could be a good starting point to talk about regional initiatives in the North American area. The ﬁrst important feature of both cases, in Europe and America, is that regional decision making is not only encouraged, but required. This means a certain institutionalization of a bottom-up mechanism in which local actors have the opportunity to decide about their future and make their own contribution to the development strategies. If we are talking about socializing a development proposal, it is likely that planning capacity building is going to take place in the areas involved with the system. It has happened in Europe where presenting consistent short, medium and long-term plans has become one of the key criteria for regions to obtain structural funds. For some of the Union’s regions this implied totally new practices as they belonged to very centralized nations. The technical assistance program sponsored by BECC acknowledges this problem as it provides planning support for the actors wishing to promote any proposal. An interesting step forward in the North American proposal is that it attempts to involve non-ofﬁcial actors into the system. A second important addition of regional dynamics is that it seems to inculcate forward-looking planning through transborder cooperation and coordination. Regional or local agents can identify other entities with similar problems and proposals and become more efﬁcient in solving them. The association pattern along national lines can be complemented by a more dynamic objective-driven alternative. The third aspect is to ﬁnd possible ways in which cooperation schemes can beneﬁt all the involved parties independent of their development asymmetries. The cooperation between the US and Mexico or among countries in Europe provides local solutions to global problems and could suggest that the tension between the local and the global is necessary to come up with broader answers able to respond to various policies at once. Of course, in these cases, the global framework is provided by the international agreements and the local should be by regional policies.

260

Sárah Martínez Pellégrini

For Mexico the beneﬁts derived from the initiatives fostered by the BECC-NADBank partnership are unquestionable, at least in the short term. The municipalities that succeeded in having their initiatives funded are able to solve current infrustructure problems heavily challenging everday operation. But if the impact of this bi-national cooperation remains restricted to isolated infrastructure projects, even the rich northern neighbor will pay the price of a shortsighted approach. On the one hand, both countries will waste a great opportunity to work out common forward-looking policies to address the border’s environmental crisis. On the other hand, the spillover in terms of planning-capacity building that could enhance further cooperation for regional development will never take place and the asymmetries between the two sides of the border may even increase. Two facts will be of a great help in order to multiply the effects of the BECC-NADBank system. The ﬁrst would be to implement an evaluation process that really applies the idea that projects are not one-time actions, but part of further development strategies. The second would be for the Mexican actors to become aware of the importance of what they might learn during the technical advice phase; as building capacity for overall planning as with the resources they dispose of, it would be impossible for them to implement large-scale qualiﬁcation plans.

Notes 1 This part of the chapter draws heavily on Gutiérrez Vidal and Martínez Pellégrini (1995). 2 For further information see Louis (1991). 3 This derives mainly from the fact that the Union, as a body, is endowed with legislative and judicial powers to protect the general interest of its members. That is the heart of the political integration among the European states as there is a consideration of a common interest exactly as it happens within a single nation. 4 Our aim here is just to revise some of the aspects we identiﬁed as more relevant for our main concern: regional policies. Thus we won’t make an exhaustive revision of all the common policies. For a more detailed account on the policies affected by the treaties, see Gutiérrez Vidal and Martínez Pellégrini 1995. 5 Starting from a broad deﬁnition of development – improvement of the living standards in a region, understood to be an integrated system, and seeking sustainability – both social and regional policies share very close objectives if not the same means and instruments. 6 It is important to note that the last three objectives mentioned are not properly spatially-deﬁned but more functionally or sectorially driven. 7 There is an ample literature on this point, so we will not review it in detail here. A good starting point is the First Cohesion Report published by the European Commission. 8 It could be argued that dealing with international integration even broadens the understanding of a community as it could also be understood as a group of countries. 9 The appearance of the Social Chart and the Cohesion Policies comes from the

Regional policies 261 fact that regional economic success does not automatically turn into a real improvement of general living conditions, therefore it is important that social aspects linked with redistribution and equal participation opportunities are dealt with through speciﬁc initiatives.

Bibliography BECC (1995) Guidelines for Project Submission and Criteria for Project Certiﬁcation, Ciudad Juárez, September. BECC (1997) Annual Report. Cecchini, P. (1988) Europa 1992: Una Apuesta de Futuro, Madrid: Alianza Editorial. Comisión Europea (1997) Primer Informe sobre Cohesión Económica y Social 1996, Luxemburgo: Comision Europea. El-Agraa, A. (1989) The Theory and Measurement of International Economic Integration, London: Macmillan. Gutiérrez Vidal, M. and Martínez Pellégrini, S. (1995) ‘El Tratado de la Unión Europea y el TLC: un análisis comparativo’, Frontera Norte 7, 13: 109–27. Louis, J.V. (1991) El ordenamiento Jurídico Comunitario, Bruselas: Comisión de las Comunidades Europeas. NADBank (1996) What is the North American Development Bank? Washington D.C.: NADBank. OECD (1995) Monthly Statistics of Foreign Trade, Paris, June. Secretaría de Comercio y Fomento Industrial (1992) Tratado de Libre Comercio entre México, Canadá y Estados Unidos, Mexico. Tratado de la Unión Europea (1992), Madrid: Editorial Cívitas. Tratado de Roma. Acta Unica Europea (1992), Madrid: Editorial Cívitas.

11 The ﬁscal federal system in Mexico Fiscal insurance with moral hazard Jaime Sempere and Horacio Sobarzo

The Mexican federal system has been characterized by a high degree of centralism. The more centralized a ﬁscal system is, the more insurance it provides against member states’ idiosyncratic macroeconomic shocks. We can identify the degree of ﬁscal insurance with a state’s degree of independence from the randomness of its speciﬁc ﬁscal system. Thus, the system provides more insurance the higher the tax revenue collection by the central government (and so the system’s capacity to insure), the higher its level compared to the states’ collections, and the more redistribution among its regions. As shown by Persson and Tabellini (1996) the existence of federal ﬁscal insurance generates a ‘moral hazard’ problem. Such moral hazards are reﬂected in the fact that a state government carries out actions that could endanger its ﬁnancial health because it knows that the federation will bail it out by making transfers to counter a fall in public expenditures. Consequences of these actions include inefﬁcient levels of spending and excessive borrowing. We could say that the degree of ﬁscal insurance is inversely related to the degree of ﬁscal responsibility in a federal system. The latter is deﬁned as the degree to which the state’s ﬁscal authority bears the political and economic cost of collecting what is then spent. The associated moral hazard also depends on other factors, such as political control, credibility, etc. The chapter is organized as follows. Section 11.1 presents various determinants of possible moral hazard in a federal system. Section 11.2 describes the main features of the Mexican federal system. Section 11.3 presents some of its institutional characteristics. Section 11.4 assesses the characteristics of its interregional redistribution. Section 11.5 presents one example of a bailout. Finally, Section 11.6 discusses some possible solutions to the lack of ﬁscal responsibility.

11.1 Determinants of moral hazard in the federal system The problem of moral hazard appears when the objectives of the principal and the agent diverge, and when monitoring of the latter’s actions cannot

The ﬁscal federal system in Mexico 263 be done. Also important is the existence (or lack of) of some supervision or control that provides incentives for the right actions. In the case considered here, this external supervision would come from the political sector and the ﬁnancial system. Another important factor discussed in the literature is whether credibility deriving from long run relationships can reduce the moral hazard problem. The institutional characteristics relevant for determining the degree of moral hazard in a federal system are considered below. 11.1.1 Fiscal system’s ﬂexibility The ﬁrst factor (see Eichengreen and von Hagen 1996, and Gamboa 1997) is, obviously, the existence of a state’s ﬁscal system that is ﬂexible and efﬁcient enough for the state to ﬁnance its own obligations. Then, if faced with a certain spending need, the local government can respond by adjusting its ﬁscal parameters with no need to call upon the federal government. Obviously, the state’s ﬁscal ﬂexibility reduces the volatility of tax revenues and so the need for insurance. Less need for insurance means a lower degree of moral hazard. 11.1.2 The political system The second factor is whether the political system can oblige the states to spend in the best possible way. This relates to the idea that the political system or voters can penalize the state government that doesn’t spend their revenues properly. Obviously, this implies the existence of a political system democratic and pluralistic enough so that the removal of an inefﬁcient government is possible. We must also consider the relation between state and federal governments since, without adequate political control at the federal level, the federal government will tend to help those states governed by its same political party. Therefore, plurality and alternation of power can make it harder for such behaviors to occur, and to make the states more responsible. In relation to a state’s public debt, Sanguinetti (1993) shows that, in many cases, the federal government can also help states governed by opposition parties. Nevertheless, the necessary condition for this to happen is that the lenders are members of the federal government’s party. 11.1.3 The discipline of ﬁnancial markets The third factor is whether ﬁnancial markets may impose their discipline over the local governments’ behavior and thereby make them more responsible. This discipline mechanism penalizes over-borrowing and makes credit conditions harder. This is a very important factor in Mexico because of the limits to

264

Jaime Sempere and Horacio Sobarzo

self-ﬁnancing faced by the states. If the ﬁnancial market does not discipline, this can encourage irresponsible borrowing. On this topic, Hernández (1997) says that the former Article 9 of the Fiscal Coordination Law might have inﬂuenced the tendency for excessive lending by banks to state governments. According to this Article 9, federal transfers served as collateral for the credits contracted by the states. This Article has now been modiﬁed to avoid this possibility. 11.1.4 Credibility of the Mexican federal government The question is whether the above mentioned factors are strong enough to force a federal government to credibly commit to stop bearing the consequences of a state’s irresponsible ﬁscal operations. The important issue here is how state governments will be penalized (economically or politically) when they deviate from the responsibility commitment.

11.2 Mexican ﬁscal federalism Mexico is formally a federation in which the three levels of government have some degree of autonomy regarding tax collection and spending decisions. However, in reality this autonomy has traditionally been very limited because of the excessive concentration of functions in the federal government. On average, during the 1980–95 period, the federal government collected 79.9 percent of revenues while the states collected 15.5 percent and municipalities the remaining 4.6 percent. On the spending side, these proportions were 76.1 percent for the federal government, 19.8 percent for the state governments and 4.1 percent for municipal governments. On the revenue side, this concentration of functions in the federal government was based on the so-called National System of Fiscal Coordination (NSFC), originated in 1980.1 Under this system the federal government was designated to collect the broader base taxes, such as the income tax and VAT, and also some others, such as the energy tax (oil and electricity), tariffs, etc. State governments have access to minor taxes. The most important are the payroll tax and the states share in special tax levied on output and services. Finally, since 1984, municipal governments have had jurisdiction over the real property tax which is usually their most important source of revenues. This way of organizing the tax system implies that, in practice, the revenue collecting capability of states and municipalities is fairly small. For example, in recent years, revenues collected by the states have represented less than 3 percent of total public revenues. At the same time, the municipalities’ efﬁciency in collecting the real property tax is conditioned on the high regional heterogeneity in managerial and resources matters, apart from the strong erosion of public ﬁnance over the last two

The ﬁscal federal system in Mexico 265 decades. This implies large differences in the amounts collected by different municipalities. In exchange for their revenue collection, states receive transfers from the federal government (participaciones).2 At the same time, the law establishes that the states must transfer to the municipalities, as shares, at least 20 percent of their total revenues. As argued by Sempere and Sobarzo (1998) the NSFC has managed to make the collecting system more efﬁcient as compared to the pre-1980 existing ﬁscal systems. In fact, while, in 1980, transfers amounted to 13 percent of the so-called ‘shareable ﬁscal collection’,3 since 1995 this percentage was established at 20 percent. On the spending side, performance has been less systematic and highly discretionary. As compared to revenues, until recently there has not existed a clear delineation of functions between government levels. As detailed below, states’ governments have traditionally had very low independence in their spending decisions. To a high degree, this has been so because the most important component of state’s spending comes from the so-called social development agreements (CEDES in Spanish), in which the federal and state governments participate jointly. This instrument normally works on an annual basis, and has not been subject to a transparent allocation mechanism. In practice, this implies that the federal government often has a big inﬂuence not only in the kind of spending, but also in its allocation between states. On the other side, during the last four or ﬁve years, some process of decentralization of state and municipal spending have been initiated. The most important, by their size, are education and health spending. Although these processes have had their difﬁculties, the assignation of resources to state and municipal governments has become more transparent. Despite these recent reforms, it is still correct to say that the decision capability on spending matter is still concentrated at the federal level. In any event, the fact that during the last 15 years the Mexican economy has registered very low economic growth rates has meant that, overall, public revenues as a share of GDP have been reduced. During 1980–95 the GDP growth rate was 1.7 percent, lower than the population growth rate (2.0 percent). As a result, with the exception of the 1988–94 period, non-oil tax revenues have remained below 10 percent as a share of GDP. As a result, states’ governments assign most of their resources to operational spending, on basically wages and administration. Thus their capability to decide over the assignation of spending has been signiﬁcantly reduced. In fact, it can be said that the poor performance of public ﬁnance during the last years is a serious constraint to future decentralization attempts by the federal government. Summarizing what has been said up to now, the NSFC placed the federal government in charge of the broader taxes, in turn making state and municipal governments highly dependent on federal transfers. Although this has meant that their collecting capability is small, the overall

266

Jaime Sempere and Horacio Sobarzo

ﬁscal system has become more efﬁcient. From the spending side, however, the concentration of functions in the federal government is characterized by a lack of transparent rules for the assignment of resources which, combined with a poor performance of national tax collection, has resulted in a ﬁscal system with a low degree of ﬁscal responsibility at the three levels of government. In a way, this explains the inefﬁcient public spending in the country4 and, more generally, various cases of non-fulﬁllment of functions of the three levels of government.

11.3 The institutional context In this section we present a brief summary of the institutional context in which the problem of ﬁscal irresponsibility has arisen. Note that the NSFC in its current form was introduced in 1980, which is why there is no explicit reference to prior years. Recall that the NSFC in its present form was created because of the introduction of VAT. This implied that the federal government became responsible for raising revenue from the widest tax bases. Essentially, the state governments delegated their collection power to the federal government in exchange for receiving federal transfers. In its origin, in 1980, the ﬁscal agreement was that each state would receive, at a minimum, their previous year’s (1979) revenues. Thus the system was designed as a compensatory device. The criterion was simply to compensate the states for the revenue they had stopped receiving since the federal government introduced and collected VAT, substituting other taxes that were their previous sources of revenue. Nevertheless, during the 1980s and the 1990s this compensation criterion lost importance in favor of other criteria. At the present time, the formula used for the distribution of shares to the states is composed as follows: a b c

population (45.17 percent), revenue generation capability (45.17 percent)5, and in inverse relation to the per capita transfers resulting from the two previous criteria (9.66 percent).

Given the sharp regional differences in terms of income and development across Mexico, it is not surprising that there is a discussion over the fairness of the above formula. Since the population component (that reﬂects the need for spending) is important, it is reasonable to justify a redistribution of resources from the high-income to the low-income zones. This will not be discussed here. It will be enough to mention that these reforms have been made possible because the total amount of resources transferred to the states have increased permanently, as mentioned above, from 13 percent of the shareable ﬁscal collection in 1980 to 20 percent in 1996. For the purposes of this chapter, it is important to note two favorable characteristics of the sharing system. First its creation has made the tax

The ﬁscal federal system in Mexico 267 collection system more efﬁcient and, secondly, the share distribution process between states is based on a transparent mechanism. With respect to the shares the states give to municipalities, the legislation establishes that the municipal governments must receive at least 20 percent of state revenues. However, in contrast with the distribution mechanism between the federation and the states, a unique distribution formula does not exist, since each state establishes their own criteria. In practice, not all the states have a formula that establishes transparent mechanisms for redistribution. Even though there are similarities between the federation–state relation with the state–municipalities relation (particularly in the fact that both the states and the municipalities depend heavily on the federal transfers), it is important to note that municipal governments have a higher collection capability because of their jurisdiction over the real property tax collection.6 For the goals of this chapter, which is to try to understand and analyze the ﬁscal responsibility between different levels of government, it is also necessary to understand how revenues and spending work. What follows is a more detailed description of some characteristics of the way in which public spending in Mexico is organized. Public spending can be classiﬁed in different ways. For the purposes of this chapter, it will be divided in ﬁve: a b c d e

own spending, joint spending, decentralized spending, two-part spending and direct federal spending.

Own spending. This refers to spending made exclusively by the states. Its most important characteristic is that it is a non-conditioned spending. According to INDETEC (1997), the states governments have suffered a permanent reduction of this kind of spending. Among the causes, the following are mentioned: i) growing population, ii) growing demand for higher quality services, iii) deterioration of local infrastructure and iv) inﬂation. Undoubtedly all these factors add to the poor performance of public revenues during the last 15 years in Mexico. As a result of this evolution, it seems that states governments have gradually lost their capacity to create infrastructure since their spending has been increasingly forced into operational spending. The proportion of resources devoted to public infrastructure went from 20 percent in 1990 to 13 percent in 1994 (INDETEC 1997). It is important to note the related aspect that, as a result, during the ﬁrst years of the 1990s, state debt accumulated rapidly. Debt service as a proportion of own spending went from 7.2 percent in 1990 to 10.1 in 1994. This point will be further addressed below.

268

Jaime Sempere and Horacio Sobarzo

Joint spending. This refers to situations in which both federal and states governments assign resources to the same function. In essence, this constitutes mostly spending on primary education. Perhaps the most important point to notice about this kind of spending is the large differences in the proportion that different states assign to basic education. On average, during the period 1992–4, state governments assigned 20 percent to basic education while the federal government received the remaining 80 percent. Nevertheless, by state, this proportion varies signiﬁcantly: for example, while Baja California assigned 40.2 percent, Hidalgo just assigned 2.7 percent (INDETEC 1997). This asymmetry implies that, in some years, there have been states which have dedicated as much as 40 percent of their spending budget to education, while other states do not use more than 5 percent.7 It is important to notice that the process of educational decentralization started in 1992; however, the trade union (SNTE) remained a national entity. In practice, this has meant that those states that assign a big part of their budgets to education spending bear a heavier burden of any wage increase. Decentralized spending. This kind of spending is the one that the federal government has been transferring towards the states, conditioned to be used in speciﬁc areas such as education, health, agriculture, etc. Since these decentralization processes are recent, they are hard to evaluate. In any case, up to now this decentralization has caused some costs. Previously such expenditures were offered by the states and the federation separately, but now needs to be offered as a homogeneous service; this has implied that, on some occasions, the states have had to assign more resources to increase salaries and improve equipment. Again, the amount of resources has varied a lot across states. In health spending, for example, the additional states resources in some cases have been up to ten times the level assigned by other states. Moreover, the amounts transferred to the states have not necessarily been related to the different states’ needs, but have rather been determined by the annual availability of resources. Insofar as the decentralization processes have ignored regional differences, it is possible that some degree of non-fulﬁllment has occurred, as was the case with the two previous types of spending. Two-part spending. This is probably the main instrument the states have in order to develop their infrastructure. This kind of spending works via the ‘social development agreements’ (CEDES) and, judging by the amount of resources involved, it has been one of the main mechanisms to get resources to the states and municipalities. Although it is true that formally the allocation of these resources is sketched in the development plans, the assigned amounts by means of this instrument have not been subject to transparent mechanisms. Some authors, like Diaz (1998) and Molinar and Weldon (1994), present evid-

The ﬁscal federal system in Mexico 269 ence to suggest that, to some extent, important political–electoral determinants prevail in the assignment of this type of spending. In general, what happened with this kind of spending is that often the federal government managed to impose its priorities over the states’, thus reducing their degree of autonomy. Although today it is less common, some years ago the amounts assigned to the states occasionally depended on the relationship between the governors and the president. In 1996, some mechanisms were created to make the assignment process clearer. First, the amounts ﬁnanced by the states have been reduced, with a consequent diminution in the pressure the states have to assign resources that aren’t always available. In 1996 23 of the 32 states contributed no more than 15 percent. Second, more resources are assigned directly to the states. And third, the resources assigned for social spending are distributed to the state and local governments according to an explicit formula, thus giving more certainty to the assignment mechanism. During the administration of President Salinas, the so-called Solidarity Program distributed resources without an explicit formula. In 1996 these resources were integrated into the ‘33rd branch’, and were distributed according to an explicit formula, based essentially on poverty indicators. Now the state and municipal governments have a higher degree of certainty not only of amounts, but also of timing; thus these changes have undoubtedly beneﬁted them. In spite of these important changes, some problems remain, such as delays in the transfer of resources, and the lack of administrative capability of some local governments. Speciﬁc difﬁculty is present in the state universities in which both the federal and the state governments take part, not only because of the important weight of federal decisions but also because this weight varies depending on the state. Direct federal spending in the states. This is spending entirely decided and made by the federal government. Decisions regarding amount and structure of spending depend on elements such as availability of resources, federal priorities, etc. In any case, let us just mention that the amounts in this kind of spending are fairly substantial. According to Rodríguez (1997), the amount of resources assigned to ﬁghting poverty during the Salinas administration was about 15 percent of total federal public investment. In summary, we could say the following: ﬁrst, on the revenue side, the federal government is in charge of the most important sources of revenues. This centralization, however, seems to have increased the efﬁciency of the collection process. In the relation between states and municipalities, such a transparent and systematic assignment is rarely present. Second, on the spending side it seems, by contrast, the assignment of resources has been less transparent and systematic. In the last two or three years some attempts at decentralization and clear assignment mechanisms have been put into practice, which have made the processes more

270

Jaime Sempere and Horacio Sobarzo

transparent. Nevertheless, these processes have had their difﬁculties, and perhaps it is too early to evaluate them. In general, it can be said that the rules used to assign resources are clearer today than in the past. In spite of this, the amount of resources controlled by the federal government still remains substantial.

11.4 Redistribution Beyond the strong centralization of functions played historically by the Mexican federal government, it is worth asking about the roles it has played in the regional distribution of resources and in the different degrees of development across regions. It is difﬁcult to answer this question not only because of the typical problem of the causality between public spending and development, but also because the centralization of revenues and spending functions in the federal government makes the exercise even more difﬁcult, since it is not easy to detect the territorial origin (revenues) and destination (spending) of the resources. Fortunately, the recent decentralization measures mentioned in the previous section allow us to further advance in this analysis, at least on the spending side. Although we do not intend to give a deﬁnitive answer, we can at least try to get some idea of the amount of resources the federal government redistributes among regions. As mentioned earlier, more than trying to determine the incidence of public resources in the differential development of the Mexican regions, we try to establish, so far as the available information allows, the resource availability by regions (states) and, comparing revenue generation and spending capability in the states, determine the role played by the federal government in net transfers of public resources among regions. To begin on analysis, Table 11.1 shows the state distribution of the main revenues that come from the federal government, that is, shares and 33rd branch for 1999 in per capita terms. In general, with some exceptions, the share distribution per capita and shares plus resources of the 33rd branch, are fairly homogeneous. If Tabasco and Campeche are excluded (states that have had traditionally high revenues by shares because of the oil component), then the states that receive more resources per inhabitant are Baja California Sur, Colima, Distrito federal and Nayarit. On the other side, the states that receive less are Estado de Mexico, Guanajuato, Puebla and Jalisco. Despite these differences, the distribution seems to be pretty homogeneous. It is curious to note that, in spite of the deterioration of the Distrito Federal situation in comparison to 1980, it still has one of the best relative situations. Although the resources from the 33rd branch make the distribution of public resources more homogenous, this analysis can be deceiving. In con-

The ﬁscal federal system in Mexico 271 Table 11.1 Distribution of resources Estate

Transfers (pesos per person)

Ramo 33 (pesos per person)

Transfers + Ramo 33 (pesos per person)

Aguascalientes Baja California Baja California Sur Campeche Coahuila Colima Chiapas Chihuahua Distrito Federal Durango Guanajuato Guerrero Hidalgo Jalisco México Michoacán Morelos Nayarit Nuevo León Oaxaca Puebla Querétaro Quintana Roo San Luis Potosí Sinaloa Sonora Tabasco Tamaulipas Tlaxcala Veracruz Yucatán Zacatecas TOTAL

1,693.0 1,696.3 2,123.8 2,341.4 1,561.3 2,063.8 1,336.2 1,389.9 2,284.4 1,351.7 1,128.9 1,089.1 1,185.1 1,330.8 1,204.7 1,049.8 1,348.3 1,587.0 1,731.4 1,015.7 1,087.2 1,624.6 1,698.9 1,187.6 1,410.2 1,964.3 3,680.8 1,634.1 1,524.9 1,211.6 1,363.4 1,330.0 1,463.3

1,840.2 1,538.2 3,058.9 2,576.4 1,933.4 1,361.7 1,571.5 2,267.1 1,605.5 2,059.7 1,215.6 2,235.4 1,932.2 1,145.4 1,073.5 1,618.2 1,579.6 2,162.1 1,214.7 1,958.4 1,287.8 1,672.6 2,094.1 1,776.1 1,438.4 1,612.3 1,772.5 1,738.4 1,919.5 1,502.1 1,691.1 1,955.2 1,694.9

3,533.2 3,234.5 5,182.7 4,917.8 3,494.7 3,425.5 2,907.7 3,657.0 3,889.9 3,411.4 2,344.5 3,324.5 3,117.3 2,476.2 2,278.2 2,668.0 2,927.9 3,749.1 2,946.1 2,974.1 2,375.0 3,297.2 3,793.0 2,963.7 2,848.6 3,576.6 5,453.3 3,372.5 3,444.4 2,713.7 3,054.5 3,285.2 3,158.2

Source: SHCP (Ministry of Finance).

trast with shares (that constitute non-conditional spending), the resources from the 33rd branch are conditioned. In its major part, this was previously made by the federal government in the states, basically in the areas of education and health. Although Table 11.1 describes the territorial structure of an important part of states’ revenue, it doesn’t explain whether the states ﬁnance their own expenditures or if they are receiving transfers from other states. In a large part, this is because in the tax collecting system, the main tax revenues (VAT and income tax) are collected by the federal government, and

272

Jaime Sempere and Horacio Sobarzo

no information exists on the territorial origin of their taxes. This has two causes. The ﬁrst one is simply administrative. Many ﬁrms, although they might be operating in different places throughout the country, have their ﬁscal address in big urban areas such as Mexico, Monterrey, and Guadalajara. Thus they pay tax in that ﬁscal address, even though the tax revenue may have been generated in another region.8 The second problem is more conceptual and follows from the interpretation of the VAT as a tax. That is, when imputing this tax territorially, the country’s value-added territorial distribution can be used as a criterion, or the territorial distribution of the ﬁnal consumers, but not the VAT revenue. This is not a recent issue and will not be analyzed here (see Pérez 1991), but it is important to bear in mind when elaborating further. We tried to impute revenue and spending territorially by state, in order to have an idea of the magnitude of transfers across states. These are deﬁned as the difference between revenue generation and spending in a given state.9 For this analysis it is necessary to make some assumptions, which we make explicit below. First, we assume a balanced budget. Thus all the federal expenditure in the states is ﬁnanced with resources generated in the states. This allows us to avoid, for instance, the problem of imputing the ﬁscal deﬁcit by state. Second, the shareable revenue collection was imputed territorially according to the states’ GDP structure. That is, we assume that collection via taxes such as VAT and income tax has a structure similar to the territorial distribution (by states) of GDP. In the case of VAT, this implies imputing it with the value-added territorial distribution criterion, although it is a tax transferred to the ﬁnal consumer.10 Third, the expenses or transfers not assigned to the states were imputed according to each state’s population; this amounted to about 17 percent of revenues. Finally, it was necessary to assume that federal public investment in the states was ﬁnanced with revenues collected in the same year analyzed. This, of course, is not completely correct since it implies mixing capital and current spending. However, if one bears in mind that the evolution of the amount and the territorial distribution of public investment is more or less stable, the problem might not be so serious. It must be mentioned that we took 1993 as one analysis year since this is the most recent year for which GDP by states is available. The year of 1993 is interesting, however, since for this year there is reporting of federal contributions to the states in health, education and poverty.11 Besides, the most important contributions to the states, in addition to shares, are constituted by transfers designated to be used in public education and health. These have suffered few modiﬁcations in their structure, thus making it feasible to assume that the results should be very similar to the ones that obtain today. The results of these estimates are shown in Table 11.2, while the interstate public resources net of transfers as shares of GDP are shown in

378,282 1,121,398 222,642 449,936 834,574 266,161 1,392,179 1,059,287 5,898,830 497,005 1,290,926 851,687 606,823 2,197,706 3,722,225 1,028,911 515,597 376,558 1,719,116 870,289 1,310,507 515,231 292,368 638,249 959,213 1,154,647 1,482,553 1,092,391 348,221 2,252,561 541,519 452,921 36,340,516

1.04 3.09 0.61 1.24 2.30 0.73 3.83 2.91 16.23 1.37 3.55 2.34 1.67 6.05 10.24 2.83 1.42 1.04 4.73 2.39 3.61 1.42 0.80 1.76 2.64 3.18 4.08 3.01 0.96 6.20 1.49 1.25 100

% 527,400 633,300 336,300 1,955,200 1,834,000 287,600 1,060,000 632,300 15,923,700 415,600 838,900 917,900 1,504,100 958,600 2,449,100 730,400 393,700 833,500 1,051,400 1,229,700 551,100 419,100 317,900 397,200 971,200 812,100 2,340,500 1,214,900 287,800 2,453,900 459,200 147,200 44,884,800

*Investment 3

Education4

1.18 259,400 1.41 431,100 0.75 185,300 4.36 227,400 4.09 598,500 0.64 159,900 2.36 947,000 1.41 618,400 35.48 24,216,000 0.93 494,600 1.87 746,800 2.05 1,067,200 3.35 647,500 2.14 917,000 5.46 1,333,500 1.63 0 0.88 0 1.86 314,000 2.34 577,300 2.74 972,700 1.23 876,000 0.93 286,700 0.71 229,800 0.88 670,900 2.16 553,600 1.81 515,500 5.21 456,000 2.71 774,700 0.64 253,400 5.47 1,588,700 1.02 346,300 0.33 415,400 100 41,680,600

% 0.62 1.03 0.44 0.55 1.44 0.38 2.27 1.48 58.10 1.19 1.79 2.56 1.55 2.20 3.20 0.00 0.00 0.75 1.39 2.33 2.10 0.69 0.55 1.61 1.33 1.24 1.09 1.86 0.61 3.81 0.83 1.00 100

% 137,414 159,643 91,849 166,619 280,897 95,420 395,448 245,798 1,039,016 155,720 239,317 409,544 274,470 243,009 386,503 406,973 144,237 106,575 199,399 460,089 239,965 136,469 120,436 215,821 168,343 346,127 323,735 168,993 124,108 300,742 323,501 152,791 8,258,971

Solidarity 5 1.66 1.93 1.11 2.02 3.40 1.16 4.79 2.98 12.58 1.89 2.90 4.96 3.32 2.94 4.68 4.93 1.75 1.29 2.41 5.57 2.91 1.65 1.46 2.61 2.04 4.19 3.92 2.05 1.50 3.64 3.92 1.85 100

%

*Investments abroad and non-geographically allocated investments are excluded. ** Assigned in concordance with the population.

Source: 1 BD INEGI. 2 SHCP. 3 El Ingreso y el Gasto Público en México 1997. INEGI. 4 IV Informe de Gobierno del Presidente Ernesto Zedillo. 1998. Ernesto Zedillo. 1998. 6 BD INEGI.

0.88 2.45 0.47 1.64 2.78 0.63 1.82 2.93 24.06 1.20 3.48 1.98 1.59 6.58 10.52 2.29 1.61 0.73 6.47 1.71 3.22 1.41 1.34 1.77 2.21 2.64 1.48 2.56 0.56 4.94 1.26 0.81 100

Aguascalientes Baja California Baja California Sur Campeche Coahuila Colima Chiapas Chihuahua Distrito Federal Durango Guanajuato Guerrero Hidalgo Jalisco México Michoacán Morelos Nayarit Nuevo León Oaxaca Puebla Querétaro Quintana Roo San Luis Potosí Sinaloa Sonora Tabasco Tamaulipas Tlaxcala Veracruz Yucatán Zacatecas TOTAL

1,261,278 3,511,403 666,560 2,343,586 3,972,897 901,847 2,610,318 4,198,302 34,435,853 1,722,683 4,987,685 2,834,194 2,271,128 9,419,974 15,055,196 3,274,468 2,303,147 1,039,230 9,256,141 2,452,815 4,608,576 2,014,284 1,923,612 2,533,291 3,162,644 3,775,791 2,111,964 3,661,021 796,449 7,065,054 1,809,654 1,156,952 143,138,000

Transfers2

Expenditures 1993 (Thousand of Pesos. Current Prices)

Revenues 1993 (Thousand of Pesos. Current Prices) Federal revenues 1 %

Table 11.2 Revenues and Expenditures 1993

Net

0.91 149,598 2.28 893,134 0.41 219,163 0.67 536,154 2.37 140,550 0.54 28,468 4.01 1,664,439 2.93 1,291,237 9.52 13,781,781 1.61 33,323 4.86 1,289,934 3.19 –793,703 2.27 1,033,656 6.66 4,306,617 12.64 5,650,530 4.33 590,104 1.49 1,071,242 1.01 712,617 3.91 5,240,692 3.71 1,524,613 5.15 1,014,919 1.33 497,675 0.70 878,868 2.43 320,156 2.72 184,008 2.19 685,510 1.88 2,715,395 2.75 81,127 0.95 331,421 7.43 420,768 1.63 55,890 1.51 192,249 100 0.249

%

5 IV Informe de Gobierno del Presidente

108,379 272,828 49,632 80,584 284,376 64,299 480,130 351,280 1,140,088 193,080 581,808 381,566 271,891 797,043 1,513,338 518,081 178,371 121,215 468,233 444,649 616,084 159,110 84,240 290,965 326,280 261,908 224,570 328,910 114,341 889,919 195,025 180,888 11,973,113

**Others6

274

Jaime Sempere and Horacio Sobarzo

Table 11.3 Revenues 1993 Revenues 1993 (Thousand of Pesos. Current prices)

Aguascalientes Baja California Baja California Sur Campeche Coahuila Colima Chiapas Chihuahua Distrito Federal Durango Guanajuato Guerrero Hidalgo Jalisco México Michoacán Morelos Nayarit Nuevo León Oaxaca Puebla Querétaro Quintana Roo San Luis Potosí Sinaloa Sonora Tabasco Tamaulipas Tlaxcala Veracruz Yucatán Zacatecas TOTAL

GDP ESTATE1

NET

NET/GDP %

9,935,842 27,661,436 5,250,894 18,461,836 31,296,899 7,104,395 20,563,046 33,072,552 271,271,933 13,570,609 39,290,995 22,326,650 17,891,045 74,206,802 118,598,838 25,794,958 18,143,275 8,186,643 72,916,189 19,322,297 36,304,528 15,867,729 15,153,451 19,956,259 24,914,048 29,744,180 16,637,207 28,840,064 6,274,110 55,655,682 14,255,737 9,114,004 1,127,584,133

149,598 893,134 219,163 536,154 140,550 28,468 1,664,439 1,291,237 13,781,781 33,323 1,289,934 793,703 1,033,656 4,306,617 5,650,530 590,104 1,071,242 712,617 5,240,692 1,524,613 1,014,919 497,675 878,868 320,156 184,008 685,510 2,715,395 81,127 331,421 420,768 55,890 192,249 0

1.51 3.23 4.17 2.90 0.45 0.40 8.09 3.90 5.08 0.25 3.28 3.55 5.78 5.80 4.76 2.29 5.90 8.70 7.19 7.89 2.80 3.14 5.80 1.60 0.74 2.30 16.32 0.28 5.28 0.76 0.39 2.11

Source: 1 El Ingreso y el Gasto Público en México 1997.

Table 11.3. A positive sign represents a state that is a net contributor, while a negative sign means the state is a net recipient. The extreme cases are worth noting. Those are Nuevo Leon, that transfers 7.1 percent of its GDP, followed by Morelos, Quintana Roo, Jalisco and Estado de Mexico. These are just the most important ones. On the other side, Nayarit, Chiapas, Oaxaca, Hidalgo and Distrito Federal are among the most beneﬁted. Some interesting aspects can be stressed from these results. For example, applying a threshold to the states that receive or transfer more

The ﬁscal federal system in Mexico 275 than 3 percent of their GDP; it results that 9 states transfer 1.8 percent of national GDP or 5.4 percent of the combined GDP of receiving states. Included in the latter group is the Distrito Federal. If the Distrito Federal is excluded, then the GDP transferred by the nine states represents 18.1 percent of the eight biggest recipients’ GDP. In fact, the six states that contribute most transfer to the eight main receivers the equivalent of 16.1 percent of their combined GDP. Taking total transfers among states, we observe that all contributing states give the equivalent of 4.7 percent of the receiving states GDP and 3.9 percent of their own GDP. To have an international reference in mind, according to the OCDE (1998), in Europe, transferring regions give 4 percent of their GDP, and this increases the recipient regions’ GDP by 8 percent. It is also worth stressing that the direction of the transfers suggests that the Mexican system represents redistribution from the rich to the poor states with the notable exceptions of the Distrito Federal. This can easily be seen in Table 11.4, where we group states according to their condition of recipients or contributors and, in parenthesis, we give their indexes of discrimination (in terms of public services) as deﬁned by Conapo (see Table 11.5). With the exception of Guanajuato, Michoacan, Queretaro and San Luis Potosi, the contributing states have a low level of discrimination. In the same way, with the exception of Aguascalientes, Baja California Sur and Distrito Federal, the recipient states are classiﬁed with high discrimination. Now, from the revenue side these results are not surprising since, as was previously mentioned, the share distribution formula has been modiﬁed in Table 11.4 Recipients and contributors of the federal system Recipient states

Degree of poverty

Net contributors

Degree of poverty

Aguascalientes Baja California Sur Campeche Chiapas Distrito Federal Durango Guerrero Hidalgo Nayarit Oaxaca Tabasco Tlaxcala Veracruz Yucatan Zacatecas

Low Low High Very high Very low High Very high Very high Medium Very high High Medium Very high High High

Baja California Coahuila Colima Chihuahua Guanajuato Jalisco Estado De Mexico Michoacan Morelos Nuevo Leon Puebla Queretaro Quintana Roo San Luis Potosi Sinaloa Sonora Tamaulipas

Very low Low Low Low High Low Low High Low Very low Very high High Medium High Medium Low Low

Source: CONAPO.

719,659 1,660,855 317,764 535,185 1,972,340 428,510 3,210,496 2,441,873 8,235,744 1,349,378 3,982,593 2,620,637 1,888,366 5,302,689 9,815,795 3,548,199 1,195,059 824,643 3,098,736 3,019,560 4,126,101 1,051,235 493,277 2,003,187 2,204,054 1,823,606 1,501,744 2,249,581 761,277 6,228,239 1,362,940 1,276,323

Aguascalientes Baja California Baja California Sur Campeche Coahuila Colima Chiapas Chihuahua Distrito Federal Durango Guanajuato Guerrero Hidalgo Jalisco Mexico Michoacan Morelos Nayarit Nuevo Leon Oaxaca Puebla Queretaro Quintana Roo San Luis Potosi Sinaloa Sonora Tabasco Tamaulipas Tlaxcala Veracruz Yucatan Zacatecas

Source: CONAPO 1990.

81,249,645

Country Total

Population

7.06 4.68 5.39 15.40 5.48 9.30 30.12 6.12 4.00 6.99 16.57 26.87 20.69 8.90 9.03 17.32 11.95 11.34 4.65 27.54 19.22 15.37 12.30 14.95 9.86 5.62 12.67 6.86 11.13 18.26 15.86 9.88

33.85 24.01 27.77 44.82 28.18 36.02 62.08 30.86 16.77 39.49 46.91 50.36 45.73 35.92 28.74 48.56 33.98 41.67 23.18 56.70 44.90 36.60 38.80 44.51 37.40 29.08 44.94 32.15 33.60 47.56 47.82 49.13

29.31

2

1

12.44

% of population without complete primary school above 15 years old

% of population illiterate above 15 years old

Table 11.5 Index of Municipal Privation, 1990

10.88 4.89 7.03 24.80 10.81 8.37 42.66 14.10 1.81 32.38 31.11 50.48 42.78 11.98 16.17 24.48 19.60 18.24 4.21 45.49 35.49 34.90 17.29 25.30 18.75 7.72 16.83 6.51 30.67 26.44 30.07 43.65

21.47

3

% of inhabitants in housing without drainage nor WC

4.97 10.49 11.10 15.02 5.24 5.82 34.92 13.23 0.76 13.73 12.50 22.36 22.60 7.87 6.44 13.13 3.96 8.67 3.55 23.81 15.47 15.65 15.39 27.98 8.99 9.24 15.45 15.90 5.64 27.17 9.20 13.30

12.99

4

% of inhabitants in housing without electricity

4.18 19.59 10.17 29.52 7.76 6.61 42.09 11.96 3.33 14.98 16.77 44.03 29.86 13.76 14.72 20.85 11.37 16.11 6.70 42.21 29.22 16.43 10.49 33.77 19.45 8.52 43.82 18.66 8.73 41.91 29.35 24.69

20.92

5

% of inhabitants in housing without drainwater

51.03 45.43 48.60 65.03 50.00 56.03 74.07 47.76 45.58 53.47 59.56 69.64 63.50 48.82 58.61 58.33 56.68 58.91 47.63 69.94 64.48 56.47 63.36 55.91 60.71 54.38 65.64 53.03 67.49 63.43 61.05 57.18

57.09

6

% of housing in stacking conditions

7.45 8.13 13.99 24.16 8.48 21.13 50.90 9.58 2.45 20.04 16.71 49.83 29.90 13.21 12.80 28.70 21.76 21.63 6.10 52.51 30.02 16.75 23.19 29.69 23.68 18.18 13.82 13.81 14.24 35.90 18.60 16.86

20.92

7

% of housing with dust ﬂooring

26.97 11.95 25.60 36.21 17.13 22.39 66.56 25.57 0.32 49.29 40.82 56.13 62.88 23.02 21.30 46.34 26.56 48.76 9.48 69.62 46.45 47.68 33.39 49.90 43.46 25.99 57.52 21.51 38.73 50.39 32.36 62.06

34.38

8

% of population in towns smaller than 5,000 inhabitants

62.53 39.96 54.06 68.07 60.89 50.40 80.08 52.81 60.47 67.65 61.32 67.81 73.70 55.49 62.63 59.98 59.50 53.63 58.74 78.73 72.41 60.46 49.17 71.14 52.61 52.70 65.49 61.07 72.33 71.85 73.61 72.70

63.22

9

% of population with income below 2 minimum wages

0.890 1.345 0.968 0.480 1.053 0.756 2.360 0.872 1.690 0.012 0.212 1.747 1.170 0.768 0.604 0.363 0.457 0.134 1.377 2.055 0.831 0.161 0.191 0.749 0.141 0.860 0.517 0.609 0.036 1.130 0.400 0.568

Index

LOW VERY LOW LOW HIGH LOW LOW VERY HIGH LOW VERY LOW HIGH HIGH VERY HIGH VERY HIGH LOW LOW HIGH LOW MEDIUM VERY LOW VERY HIGH VERY HIGH HIGH MEDIUM HIGH MEDIUM LOW HIGH LOW MEDIUM VERY HIGH HIGH HIGH

Level of marginality

The ﬁscal federal system in Mexico 277 favor of the population component. Nevertheless, from the spending side, even though until recently a transparent allocation mechanism was lacking, the results suggest that the territorial assignation of expenditures hasn’t necessarily opposed the revenue trend.

11.5 An example of moral hazard: the bailout of the states As documented by Gamboa (1997), Mexican states accumulated debt with private and development banks at a 62 percent annual growth rate between 1988 and 1994. In real terms, these credits represented 31.3 percent of the transfers that the states received according to the Fiscal Coordination Law. These percentages aren’t actually very high when referring to countries. However, given the resource generating capacity of the Mexican states, they are very high. To give an idea, in 1990 the states’ debt service was, on average, 70 percent of their own resources (excluding federal transfers). In 1994 this percentage went up to 247 percent (INDETEC 1997). If we include federal transfers, in 1995 the states’ debt service was 42 percent of federal transfers and 75 percent of the operational spending (Hernández 1997). Table 11.6 shows states’ debt as a share Table 11.6 State debt as a percentage of GDP, 1993 State

GDP

Sonora Queretaro México Chiapas Jalisco Baja California Sur Durango Nuevo León Sinaloa Aguascalientes Baja California Hidalgo Veracruz Michoacán Oaxaca Morelos Puebla Guanajuato Zacatecas Coahuila Tamaulipas Tlaxcala

29,744 15,868 118,599 20,563 74,207 5,251 13,571 72,916 24,914 9,936 27,661 17,891 55,656 25,795 19,322 18,143 36,305 39,291 9,114 31,297 28,840 6,274

Millons of pesos Debt

Source: Based on information from the SHCP.

2,961 786 5,861 989 3,330 223 571 2,695 831 331 836 22 253 135 116 129 272 324 97 421 408 90

% 9.95 4.95 4.94 4.81 4.49 4.25 4.21 3.70 3.34 3.33 3.02 0.12 0.45 0.52 0.60 0.71 0.75 0.82 1.06 1.35 1.41 1.43

278

Jaime Sempere and Horacio Sobarzo

of GDP in 1993. We can see a divergence in the way that states became indebted. The increase in interest rates after 1994, and the low degree of ﬂexibility of the states’ public revenue, made the states unable to face their ﬁnancial liabilities. This forced the federal government to bail out the states’ debt through extraordinary transfers and special payment conditions. This ﬁnancial rescue can be related with the ﬁnancial and political factors mentioned in Section 11.2. First, the rescue is necessary, because the states are not capable of increasing their revenue commensurate with the increase in their obligations. Second, states get over-indebted because the ﬁnancial market does not exercise its discipline in the case of states. Third, from Table 11.1 we can see that the most indebted states are those governed by the PRI which would bring the biggest political loss to the federal government (also ruled by PRI) in case they weren’t rescued. Fourth, we could argue that the states got over-indebted because they knew that if they could not service their debt, the federal government would rescue them. This is, in conclusion, a clear case of moral hazard. The signal sent with the ﬁnancial rescue is that the federal government will rescue the states in case they have contracted excessive obligations. These facts provide a precedent for the future because the lack of credibility has been made evident, and if the proper measures are not taken, this could lead to other opportunist behaviors in the future.

11.6 Some proposals to reduce moral hazard The proposals to reduce moral hazard relative to ﬁscal actions of local governments can be made in various ways. The ﬁrst is to relate the ﬁscal objectives of the federal government (the principal) with the objectives of local governments (the agents) using incentive contracts in which transfers are conditional to the fulﬁllment of ﬁscal objectives. The second is the imposition of limits on the public spending of states. Then, the state’s decision would be to choose the best way to spend the available scarce resources, thus limiting the states discretionality. These amount to limits on borrowing (like the ones discussed in Gamboa 1997) or limits to public spending. Limits on borrowing have the problem (Eichengreen and von Hagen 1996) that they could increase the volatility of the states’ economy since borrowing can be seen as an intertemporal insurance device. Consequently, even very small shocks could endanger the states’ economy. This would then imply the need for even higher transfers. On the other hand, exogenous limits to spending can have legal problems to implement. The third proposal is to endow the states with more ﬂexible taxes so they bear the costs of increasing their expenditure via a higher ﬁscal effort. A possibility discussed in Sempere and Sobarzo (1998) is the substitution

The ﬁscal federal system in Mexico 279 of the present Fiscal Coordination System with another in which the states receive a share of the personal income tax collected in their territory. This reduces the need for federal ﬁscal insurance. The ﬁnal proposal would be an improvement of the political supervision that would be led by more perfection in the democratic system.

Notes 1 The NSFC as we know it now, was created in 1980, as a consequence of the introduction of VAT. Before 1980, even though a ﬁscal coordination existed, the states collected 40 percent of the Mercantile Incomes Tax and, in general, the degree of ﬁscal coordination was inferior. 2 On average, during the last years these shares amounted to about 80 percent of total revenues of the states. 3 Such ‘shareable ﬁscal collection’ excludes oil-related revenues. 4 For the cases of education and health see, for example, IDB (1996). 5 It must be mentioned that this criterion takes as a reference the so-called ‘territorially assignable taxes’, basically the special taxes over production and services (IEPS), which are not necessarily collected by the states, but by the federal government. In contrast to VAT and income tax, however, in these taxes it is possible to clearly determine their territorial origin. 6 Unfortunately, with the exception of some urban areas, the real property tax collection has been very low in comparison with international standards, especially in rural areas, where there is a critical shortage of human and administrative resources. 7 The extreme cases are the Distrito Federal and Oaxaca, whose systems are entirely ﬁnanced by the federal government. 8 Note that the Mexican Treasury does not gather this information. 9 Note that we use the term revenue generation and not real collection. The latter is indeed reported by the Mexican Treasury, but it is highly concentrated on the main regions (cities) where ﬁrms register their ﬁscal address, where the tax is reported but not necessarily generated. 10 Ideally, a second exercise should have been done imputing the VAT with the ﬁnal consumer distribution criterion. However, the information required is not available. 11 The last corresponds to what was denominated as the 26th branch during President Salinas’ term.

Bibliography Díaz, A. (1998) ‘Supervivencia politica y distribución de racursos en el sistema federal mexicano: hacia una prueba empírica’, in Sempere, J. and Sobarzo, H. (eds) Federalismo Fiscal en México, México: El Colegio de México. Eichengreen, B. and von Hagen, J. (1996) ‘Federalism, ﬁscal restraints and European Monetary Union’, American Economic Review 86: 134–8. Gamboa, R. (1997) ‘El rescate ﬁnanciero de los gobiernos estatales por el gobierno federal: comparación de los casos de Estados Unidos, Brasil y México’, Documento de Trabajo, Banco de México. Hernández, F. (1997) ‘Es disciplinado el mercado crediticio estatal mexicano? Una arista para el nuevo federalismo’, El Trimestre Económico LXIV: 199–219.

280

The ﬁscal federal system in Mexico

IDB (1996) Economic and Social Progress in Latin America, 1996 Report, Johns Hopkins University Press. INDETEC (1997) ‘Diagnostico de las ﬁnanzas públicas de las entidades federativas’, Mimeo. Molinar, J. and Weldon, J. (1994) ‘Electoral determinants and effects of PRONASOL’, in Cornelius, W., Craig, A. and Fox, J. (eds) Transforming State Society Relations in México: The National Solidarity Strategy, San Diego: La Jolla Center for US–Mexican Studies, University of California, 123–41. OCDE (1995) Mexico, 1995, Paris. Pérez, F. (1991) Informe sobre la ﬁnanciación de las comunidades autónomas, Documento de trabajo, Instituto Valenciano de Investigaciones Ecónomicas, Valencia, España. Persson, T. and Tabellini, G. (1996) ‘Federal ﬁscal constitutions: risk sharing and moral hazard’, Econometrica 64: 623–46. Rodríguez, V. (1997) Descentralization in Mexico, Austin: Westview Press. Sanguinetti, P. (1993) ‘The politics of intergovernmental transfers and local government deﬁcits: theory and evidence’, Estudios Económicos 8: 87–109. Sempere, J. and Sobarzo, H. (1998) Federalismo Fiscal en México, México: El Colegio de México.

Comments Oliver Fabel and Bodo Hilgers

For Germans, the themes discussed by the authors contributing to this section appear all too familiar. In particular, Sempere and Sobarzo’s treatment of the moral hazard problem associated with ﬁscal federalism addresses issues that can be directly related to respective analyses for Germany. This applies to the description of the institutional structure, the empirical constellation, as well as recent theoretical and econometric contributions. However, in order to assess the fundamental problem structure, it appears useful to begin by brieﬂy recalling the underlying principles of ﬁscal federalism. Decentralized government spending can serve to improve allocative efﬁciency by accounting for regionally divergent preference structures concerning the provision of public goods and introducing competition between regional authorities. Both arguments imply that public goods which are almost exclusively consumed by the citizenship of a single federal entity should also be supplied by the respective regional authority. Efﬁciency then also clearly requires that marginal utility of the territory’s citizenship derived from the public goods must determine its contribution to the budget which will be distributed. In fact, the latter only restates the former, given that the different territorial authorities possess only limited ﬁnancing means by themselves. However, these rather clear guidelines for ﬁscal federalism design conﬂict with redistributive goals reﬂecting a national interest in developing the different regions. The German Constitution thus requires that the government must take appropriate steps towards convergence of the living conditions. This constitutional rule binds both federal and state (‘Länder’) governments which must, therefore, agree on a budget redistribution scheme. The German situation is in fact very similar to the respective description provided by Sempere and Sobarzo for Mexico. In the pursuit of the convergence goal, the tax revenue – with only few exceptions – is pooled and collected by the federal government. The state governments then receive approximately the same share as the federal government in Germany. However, the federal–state negotiations must also deﬁne the distribution scheme across states. In the past, and very vividly currently again, this scheme is regularly challenged by state governments which ﬁnd

282

Oliver Fabel and Bodo Hilgers

themselves in the net-contributor position. It has even become common to call upon the Supreme Court to at least claim renegotiations, if not to induce a ruling for ‘fair’ redistributive rules. In recent years, the southern German states – which being traditionally more rural have, in fact, beneﬁted greatly in the past – began to emphasize the moral hazard problem. Now, in the position of net-contributors, they claim that the scheme removes the incentives to work for economic development in states which beneﬁt. Two points should then be noted. First, in Germany the schemes implemented in the past appear to have actually contributed to the convergence of living conditions between the regions. Even without providing detailed statistical information, this becomes evident by simply recalling that the states currently challenging the existing scheme’s incentive effects have received positive net transfers over decades in the past. They only recently ﬁnd themselves in the net-contributor position – a development certainly strongly inﬂuenced, but not exclusively determined by, the German uniﬁcation. Second, the moral hazard problem claimed to provide disincentives for investments in regional development arises due to the necessity to implement the constitutional convergence goal. In particular, it does not reﬂect any intention to provide ﬁscal insurance against random asymmetric shocks. This obviously does not rule out that the scheme may not provide such insurance. However, this must be viewed as a by-product. If ﬁscal insurance induces a positive welfare effect, this only provides an additional argument in favor of ﬁscal federalism, given that the redistributional goal must be pursued on national political grounds. Whether the welfare enhancing effect exists further depends on the analytic perspective – hence, upon whether the economies are on or off an equilibrium path. The latter issue must obviously be subjected to thorough growth analysis. However, given that the problem itself does not arise due to uncertainty and asymmetric information structures, policy conclusions must also be appreciated from the point of view of incentive theory. Problems arising due to private information or lack of contract enforcement may – although analytically often hardly distinguishable – call for rather different implementation mechanisms. Thus, in a situation with asymmetric information, allocative improvements can be realized by conditioning contracts on additional signals. Given an enforcement problem, this would require that contracts written on such signals newly introduce enforceability. Returning to the analysis by Sempere and Sobarzo, programs in which the federal government and single, or groups of state governments team up for speciﬁc purposes can be interpreted as conditioning the contract. Such programs should then be limited in volume and duration. In addition, program success must be measurable by reference ﬁgures. Yet, this only makes sense if the consequences which are ‘ex-ante’ associated with measured program failure or success can actually be enforced through the political process. However, the structure of this process does not reﬂect the

Comments 283 typical distinction of ‘ex-ante’ and ‘ex-post’ periods. Rather, it is an ongoing process. Political authorities will not waive their constitutional right to decide on such issues at any time and, also, to reinterpret the meaning of program success. The solution can then only be to deﬁne clear and transparent general distribution rules for the territorial entities and leave spending decisions to local authorities. This is clearly also the solution emphasized by Sempere and Sobarzo. Thus, their reference to moral hazard induced by ﬁscal insurance in the title is, in fact, somewhat misleading. Transparent rules for the distribution of government funds will close the loopholes of discretionary federal and state government coalitions which conﬂict with national interest in regional economic convergence. Rather than deﬁning possible outcomes, inputs are allocated in a publicly veriﬁable way. The system will then necessarily provide ﬁscal insurance only on a sub-optimal level. Yet, the conditioning of transfers would only introduce more possibilities for bilateral renegotiations. Democratically-elected local authorities will then be held responsible by their electorates. This obviously implies the existence of a functioning competitive political system. In this respect the situation in Germany may, after all, be different from the Mexican case. On both the federal and the state level, ruling parties have more than once experienced defeat by opposing parties. In fact, the new social-democratic government was rather immediately confronted with a loss of majority in the chamber of states through state elections. The conservative coalition parties of the former federal government already had to cope with a minority position in the chamber of states. This chamber has to agree on many federal legal initiatives – not only on the distribution of funds among them. Thus, the disciplinary virtue of political competition can be assumed to be more developed than in Mexico. Analytically, this implies an emphasis on the mechanisms of the political process rather than the moral hazard problem associated with the attempt to provide ﬁscal insurance when investigating the effects of ﬁscal federalism. Only, if constitutional rules calling for the convergence of living conditions between regions could be removed, or did not exist, this conclusion may not hold. Yet, their removal for reasons of attaining economic efﬁciency would not be acceptable for national states. The German uniﬁcation has, in fact, shown the overwhelming importance of this rule for the political decision process. The solidarity principle within the European Union can thus be viewed as a necessary step towards European political uniﬁcation. Hence, the discussion by Sempere and Sobarzo, as well as our brief comments, clearly also apply. In particular, the complete lack of disciplinary pressure through a competitive political process between the European member governments has induced a vast number of speciﬁc regional compensatory programs. The balancing check is then currently only derived from the fact that the member states’ contributions to ﬁnancing the European budget are relatively small. Yet, the European contract

284

Oliver Fabel and Bodo Hilgers

system has, from its very beginning, at least allowed for political uniﬁcation as a development goal and has by now certainly taken ﬁrst steps towards it. In contrast, NAFTA can be characterized as a typical freetrade zone and therefore lacks a comparable political idea. Thus, it is not surprising that the comparative analysis of regional policies in NAFTA and the EU by Pellegrini reveals major differences. NAFTA has – with the Border Environment Cooperation Commission (BECC) and the North American Development Bank (NADBank) – only two institutions dealing with regional policies. These two institutions are aimed mainly at crossborder environmental pollution problems between Mexico and the USA. By contrast, the EU provides at least six institutions which contribute to regional policy and development issues. To compare the regional policies in NAFTA and the EU, it is necessary to understand the different motivations and goals that led to the NAFTA and EU agreements. As emphasized by Pellegrini, NAFTA – as a pure free-trade zone – almost exclusively pursues economic motives. The agreement is guided by a strong belief in the potential economic gains of international free trade in goods and services. Given a well-deﬁned set of assumptions, the Heckscher-Ohlin model shows that international trade in goods and services is a perfect substitute for international factor trade. Hence, NAFTA lacks a comparable agreement about free factor trade between the member states like the EU. In contrast, the EU was, in the ﬁrst place, motivated by the experience of two world wars. Therefore, it is built on the political intention to reduce the potential conﬂicts between European nations to a minimum by political and economic integration. Consequently, regional policies in the EU generally pursue both political as well as economic motives and goals. They deal with all kinds of political and economic regional issues and are not restricted to external crossborder environmental pollution effects. Clearly, political and economic issues cannot be separated from each other. They are very often completely intertwined. The equilibrium allocation of resources and goods in a perfect competitive market system also implies a certain income distribution which may not be compatible with political fairness considerations. Thus, political reasons may even necessitate redistribution schemes upon agreeing on free trade. Conversely, political decisions inﬂuence economic allocation and income distribution. The economic interpretation of the observed differences in the regional policies and regional policy institutions between NAFTA and the EU will then reﬂect the different causalities, however. The US – as the politically and economically dominating member of NAFTA – favors the market allocation and distribution outcome of international free trade. Consequently, the BECC and the NADBank only attempt to internalize the negative external cross-border environmental pollution effects. In Europe there is a less strong belief in the allocative and distributional virtue of market processes. Also, labor markets in Europe appear to exhibit more

Comments 285 frictions than in North America. Therefore, the political integration is seen as a necessary condition for realizing the potential economic gains from international free trade in Europe. Following these arguments, Pellegrini’s interpretation of the US assistance to Mexico during the 1994 devaluation crisis as a ﬁrst step to a deeper political integration within NAFTA appears questionable. This intervention may rather reﬂect a discretionary attempt to guarantee the functioning of the market system and, thus, of the implied allocative and distributional outcomes. Generally, the restriction of regional policies to cross-border pollution issues is perfectly compatible with the HeckscherOhlin world. The factor–price equalization theorem may not hold when markets are missing. Thus, with negative external effects, regional policies may be suitable means to internalize such effects. As emphasized already, inputs should then be allocated in a publicly veriﬁable way to local authorities rather than deﬁning possible outcomes. If instead – following Pellegrini – the convergence of living standards between the regions of the member states constitute the main goal for regional policies in both NAFTA and the EU, two implications for further research immediately follow. First, in the case of the EU, it remains to be shown that regional funds and institutions can foster the convergence process. Second, the case of NAFTA would require thorough investigation into the mechanisms by which market processes can be guided to allocate and distribute economic wealth in a desired way. In the last part of this section, Puchet Anyul and Punzo analyze the interdependence between economic growth and structural change. Economic integration processes generally take place given a background of regional development policies, ﬁscal federalism as well as cross-border environmental effects. For example, regional development policies will change relative prices and, thus, cause an expansion of some sectors relative to others. From a theoretical point of view, structural change in economic growth models often imply multiple equilibrium growth paths. Therefore, it ultimately remains an empirical question which growth path will be reached by real economies. Unfortunately, there are many answers to this question. In particular, rational expectation models only provide one of a variety of possible approaches. If all economic actors are identical with regard to their information processing and learning capabilities, a unique equilibrium growth path will arise. However, according to Puchet Anyul and Punzo, structural change lies at the heart of the integration process between Mexico and the USA within NAFTA, as well as for the member states of the European Union. It is also crucial for understanding regional integration processes as between the north and south (Mezzogiorno) in Italy. Given the corresponding real world experiences, they are thus led to reject the usefulness of the highly aggregated one-sector Solow growth model. Explicitly accounting for the structural effect, the observed economic development

286

Oliver Fabel and Bodo Hilgers

rather appears to be related to a virtual development – in particular, with respect to the growth rate of labor productivity. Pursuing an empirical analysis, Puchet Anyul and Punzo therefore adopt a growth cycle approach. Hence, structural change is perceived as a resource reallocation phenomenon. Critically, it should then be noted that resource reallocation constitutes a necessary, but not a sufﬁcient condition for structural change. Hence, the fact that the long-term growth of nearly all US sectors is based on an adequate coupling between accumulation and productivity is seen to contrast with Mexico’s dualistic sectorial structure. Here one type of sector is investing more and the other type less relative to the employment expansion. Yet, this could also reﬂect a learning process of a developing country in search for their comparative advantages. Obviously, a dynamic concept of comparative advantage appears more suitable in growth analysis. Interestingly both Mexican crises of 1982 and 1994 were initiated by ﬁscal troubles caused by balance of payments deﬁcits and an enormous currency devaluation. The economic development literature adds the observation that Mexico seems to be stuck in a poverty trap. This is characterized by highly ﬂuctuating capital accumulation and low productivity growth. Further, the case of Mexico appears to provide a mirror image of the Eastern European development, or the Mezzogiorno experience in Italy. Thus, structural change should be recognized as a main feature of economic integration processes in general. To a large extent, it is determined by the existence (or non-existence) of a well-developed ﬁnancial system. Although structural change in the Mexican economy began years before the NAFTA agreement in 1994, the agreement itself can thus be interpreted as a political device to stabilize and develop the ﬁnancial system in Mexico. More generally, Puchet Anyul and Punzo emphasize a highly relevant point for policy analysis: economic growth, structural change, as well as the underlying ﬁnancial system and the respective feedback mechanisms, must all be rigorously accounted when discussing economic integration processes.

Index

Agenor, P.R. 34 Aglietta, M. 79–80, 87 agricultural products 60, 116 agriculture 57, 70, 73–4, 135; exports of 7, 57; and migration 70 Alogoskouﬁs, G. 32, 34 Alvarez, D. 80 Amendola, M. 191 Aroche Reyes, F. 103, 115–16 Artis, M. 23 Aspe Armella, P. 54, 126, 153 assembly: plant 63; activities 140 Bailey, D. 24–5 Bain, J. 181 Balassa, B. 123–4 Banco de México 6, 126–7, 135, 236 banking system 46; operation of 6–7; reprivatization of 126 Barro, R. 22, 25 Baumol, W.J. 22, 211 Bernard, A.B. 22 Bértola, L. 25 Bhagwati, J. 123–4 Blair, P. 104, 113 Blanchard, O.J. 24, 117 Blanco Mendoza, H. 152 Boatler, R.W. 79 Border Environment Cooperation Commission (BECC) 252–6, 284 Brailovsky, V. 79, 125 Brannon, J. 157 Brown, F. 175–6, 193 Bruton, H.J. 149 Burki, S.J. 127

Caballero, R. 182 Capdevielle, M. 80, 155–7, 182 capital: inﬂows 49; mobility 7; and monetary policy 7, 49, 125 Cardero, M.E. 115–16 Casar, J.I. 79, 155–6, 175, 180, 185 Catalan, H. 31 Caves, R. 181 Cecchini, P. 245 CEED/UDG 140 Central Eastern European countries (CEE countries) 199 CEPAL 140 Chacholiadis, M. 102 Charemza, W.W. 97, 99 Chenery, H. 108 Cimoli, M. 79, 148–9, 158, 166, 192–3 Clavijo, F. 152, 180 Coase, R. 103 co-integration 24, 28–30, 37, 97–8 Comisión Europea 250 comparative advantages 30, 48, 78–9, 286; revealed cost(s) 78–9, 156; theory of 48 competitive matrix 155 competitiveness 63, 78–9; actual 156; international 4–5, 192; level of 7; losses 48 CONAPO 275 convergence 26–7, 281–3, 285; absolute 22, 25; conditional 22; and economic integration 233–5; in the EU 22; of Mexico and US 203–4; and monetary convergence 21, 23; path 21–3, 34–5; process 21–8, 30, 32, 35; rate 22; relative 22, 25; stochastic 23, 28;

288

Index

convergence cont. to a structural cycle 204 cooperation 60, 259; transborder 259 co-ordination failures 194 Corden, W.M. 88, 94 Córdoba, J. 126 Corona, J.M. 158 currency boards 4–6, 46, 48; and the banking system 7; and crises 7; and speculative attacks 9–10, 35, 46, 48 Cuthbertson, K. 24, 96 Dahlman, C. 180 Daniel, B.C. 24–5, 30 Dávila Flores, A. 139 Davis, D.R. 103 De Grauwe, P. 79 De Mateo, F. 183 Deadman, D.F. 97, 99 Deardorff, A. 102–3, 106 Della Giusta, M. 148 deregulation 6 Diaz, A. 268 Dickey, D.A. 28, 95 Dollar, D. 30 Dominguez, L. 175–6, 193 Doornik, J.A. 97 Dornbusch, R. 79, 88 Dosi, G. 79 dualism 192–3, 199–200; dynamic 233; and the growth process 190; and the growth trap 222; in Italy 199, 222, 233; in Mexico 199, 216, 233–4 Durlauf, S.N. 22, 211 Dussel Peters, E. 126, 132, 139, 153, 156, 192–3 Dutrénit, G. 182 dynamical coupling: and NAFTA 200 economic integration: and convergence 233–5; and structural change 203–4, 234–5, 285–6 econometric model: P* model 21, 25–6, 30–2, 36, 40; of productivity 183–6; VAR 24–5 Edwards, S. 127 Eichengreen, B. 33, 263, 278 Eichner, A. 181 El-Agraa, A. 244

electronics 59 employment and (un)employment 4, 8–10, 13–15, 69–70, 88, 132; and EU policy 250; in manufacturing 132–5; and SMFs 132 environmental policy: in the EU 249; in NAFTA 249 Esquivel, G. 59 Estrada, N. 102 European Monetary System (EMS) 46; and peripheral countries 47 European Union (EU) 60, 115, 127, 199–200, 243–6, 250–2, 283; and labor mobility 245–6; lessons for Mexico 12–13; Monetary Union 12, 25, 46–7, 202; as a treaty (EUT) 243–6 exchange rate 5–7, 9–16, 23, 26, 28, 32–6, 42, 54–5, 66, 78–80; ﬂexible 8, 32, 49; ﬂoating 4, 8, 46, 48; ﬁxed 5, 8, 46–9, 54, 193; nominal 3–5, 8–9, 28, 36; parity 6–7, 9, 12; peg 7; policy 3, 5–6, 10, 14; real 6, 14, 28, 34, 36, 78, 81–7, 118 exchange rate based stabilization 147; period 151 Expansión 179 exports 8–9, 11, 47–8, 57–9, 61–4, 105–12, 135–8; and employment 138; and ﬁrm size 179; and foreign ﬁrms 132; and GDP 79, 130; –import model 53–64; and labor productivity 178 Fárber, G. 126 Farina, F. 23, 33, 47 Fay, J. 182 ﬁnancial markets 47–9, 116; Mexico as emerging 64–6; and moral hazard 263–4; opening 53; structure 4–5 ﬁrms: domestic 157; foreign 157; SME 132 Fiscal Coordination 23; law 264, 277; National System of (NSFC) 264 ﬁscal deﬁcit 12, 54, 126–7 ﬁscal federalism 264–6, 281–3, 285 Flood, R.P. 34 foreign debt 55, 57, 64–5, 130; service of 125 foreign direct investment(s) 11, 54–5, 64, 140

Index 289 foreign exchange 87; constraint 8; regime 49 foreign sector 54–9 foreign trade: geographical concentration of 59–61 Foxley, A. 123 Framework Space (FS): deﬁnition of 212 Frankel, J. 102 Freeman, C. 149 Friedman, M. 79, 123–4 Fuller, W.A. 28, 95 Gaffard, J.L. 191 Gali, J. 24 Galindo, L.M. 28, 31 Gallini, N. 182 Galor, O. 22–3 Gamboa, R. 27–8, 263 Garrido, C. 142 GATT 54, 249 German uniﬁcation: and ﬁscal federalism 281–2 Giovannini, A. 34 globalization 11, 157–9, 256 Globerman, S. 59 Gold, B. 181 Grossman, G.M. 117 growth 190–5; cycle approach 211–13; export-led 53; generalized cycle 212; and inﬂation 48; model 22, 285; neoclassical model 210–11; sustainable rate of 8, 15, 22, 26, 32; trap 199, 216 Gurría Treviño, J.A. 125–6 Hall, S. 24–5 Hallman, J.J. 26, 31 Haltia, O. 103, 106, 111 Harrodian corridor: deﬁnition 212 Hay, D. 182 Hayek, F. 123–4 Hecksher-Ohlin theorem 73, 102, 116, 284–5 Helpman, E. 103, 117 Hendry, D.F. 97 Hernández, C. 158, 264, 277 import substitution 53–4, 59, 73–4,

191–2; and the growth rate 191; industrialization 125–6, 148–50 imports 8, 11, 55–60, 106–12, 130, 135, 157 INDETEC 267–8, 277 industrial model 175–3 industrial organization 140–1 industrial structure 48, 192–3; and specialization 159; and technology 149 industrialization: export oriented debate 124–5, 192–3, 196; import substitution debate 124–5; model of Mexico 125, 148; policy 125–6 INEGI 236 inﬂation 6–7, 9, 27–8, 46–9, 126–8, 130, 142, 147, 150–1, 174, 192, 247–8, 267; differential 46 innovation: deﬁnition 148; and S&T 166–8 Input–Output model 103–6 institutional change 256–8 Institutions: monetary union 14; regional 12 interest rates 28, 47–8, 125, 128, 135, 150, 278; deﬂationary effects of 47 international integration 192, 194 International Monetary Fund (IMF) 48–9, 80, 124, 245 international trade 80, 116, 191, 284–5; theory 102–3 irregular growth regimes: and development strategies 192 Jacquemin, A. 181 Jalisco: and the electronics industry 139–41 Johansen, S. 24–5, 30, 97, 99 Jones, C.I. 22 Katz, J. 149, 180 King, R.G. 24 Krueger, A. 79, 123–4 Krugman, P. 34, 79, 87, 102–3 Kushida, S. 102 labor market 47, 118, 284; binational 53, 116; and migration 68–72, 248–9;

290

Index

labor market cont. and the real exchange rate 83, 118; transborder 66–72 Leamer, E. 103, 106, 111 learning: capabilities 160–1, 285; dynamics of 169; patterns 167–8; processes 142–3, 190–1, 195, 286 Leontief, W. 102; paradox 103 Levitch, R.M. 79 liberalization 193; of capital markets 47; economic 6; ﬁnancial 6, 14; of foreign trade 6, 117; and polarization 142; strategy 125–8; trade 8, 53–4 Lilien, D.M. 95 linkages 64, 106, 125, 142; maquiladora type 159–60; intersectoral 195; structural 59; technological 168–9 Liquitaya, J.D. 31 Lovatt, T.D. 24–5 Lucas, R.E. 22, 190–1, 195 Lyons, R. 182 Maastricht Treaty 244, 247–8, 252 MacDonald, R. 25 Mangabeira Unger, R. 125 Mankiw, N.G. 22 manufacturing: enterprises 6; exports 57, 132–5, 153; industries 8 maquiladora industry 61–4, 116, 157; and exports 132; and R&D 160 Martínez Pellégrini, S. 284–5 Martínez, G. 126 Maskus, K. 103 Masson, P.R. 23 Máttar, J. 59, 150, 155–6 Medoff, F. 182 Mexican disease 73 Micheli, J. 176 Miller, R. 104, 113 Milne, A. 25 Mishkin, F.S. 27 Modernization, National Program of 152 Molinar, J. 268 monetary policy 21–3, 25–8, 32–3, 35–6, 46–9, 54–5 monetary union 46; between Mexico and US 14–15

moral hazard 262, 281–2; determinants of 262–4; an example 277–8; and ﬁscal insurance 283; measures to reduce 278–9 Moreno, L.M. 64 Moreno-Brid, J.C. 150 Morris, D. 182 multi-factor content theorem 103, 115 multinational enterprise 193–4 Mundell, R.A. 32 Mussa, M. 28 Nadal Egea, A. 166 NAFIN 175 NAFTA 8, 21–4, 26–7, 30, 32–6, 244–5; and capital movements 245; coordination of policies 27, 32–6; and environmental policy 252–6; and the EU 243–6, 284–6; impact of 26, 30; initial conditions 21–4; and labor mobility 245–6; and liberalization 127; and regional development 252; as a treaty 243–6 Naya, S. 105 neoliberalism 124–5 networks: of linkages 148; transnational 142 new economy 215; and the structural cycle in US 215 Nishimizu, N. 183 North American Development Bank (NADbank) 256, 284 Nuggent, J. 106 O’Donnell, G. 123 Oaks, D. 180 OECD 166, 236, 243 Ofﬁcer, L.H. 79 oil revenue 57, 125–6, 135 Oloriz, M. 158 Ortiz, G. 34 Oudiz, G. 79–80, 87 Parikh, A. 24–5 Pasinetti, L. 103–4, 115 Pearce, D. 102 Penrose, E. 182 Pérez, F. 272

Index 291 Pérez Motta, E. 152, 155 Perez, C. 149 Perron, P. 28 Persson, T. 262 Phillips, P.C.P. 28 Plosser, C.I. 24 Poder Ejecutivo Federal (PEF) 127, 132 polarization 132–3 policy 7, 32, 36, 191; coordination 12, 21, 34, 49; economic 6, 23, 35; ﬁscal 5–6; government 6, 14; implications for 88–9; industrial 152–3; macroeconomic 5, 15; measures for external balance 6; of monetary coordination 34–5, 46, 49; public 193–6; regional 284–5; redistributive 281–4; trade 152–3 Porcile, P. 25 Porter, R.D. 26, 31 Pozas, M.A. 176 Primer Informe para la Cohesión Económica y Social 252 privatization and (re)privatization 54, 94 productivity 8, 10, 13, 24, 190–1, 194–5; capital 135; determinants of 180–3; and ﬁrm size 178–9; gains 9, 190; growth 118–19; of labour 90–4, 135, 193, 286; in manufacturing 49, 135, 173–4; and SMEs 183–6 public debt 23, 47, 66, 215; and the political system 263 Puchet Anyul, M. 285–6 Punzo, L.F. 285–6 Quah, D. 22–4, 211 Ramírez, J.C. 176 Ramirez-Rojas, C.L. 34 redistribution: interregional 266, 270–7; measures in NAFTA and the EU 249–50 regional development 281–2; analysis of 139 regional policies: in the EU 250–2; impact on technology 168; for labor markets 74; in NAFTA 252–5; in NAFTA and in the EU 284–5

regions 14, 24, 73, 190, 282–3, 285; depressed 15; measures for depressed regions in the EU 250 relations with US economy 59–60 research and development (R&D) sectors 158–61 Ricardo: theorem on international trade 79, 102–3 Richardson, G.B. 191 Robinson, S. 183 Rodriguez, V. 269 Rogers, J.H. 34 Romer, D. 22 Ros, J. 125, 150 Ross-Larson, B. 180 Ruiz Durán, C. 139 Ruiz-Nápoles, P. 115–16, 118 Ryder, H.E. 23 Sachs, J.D. 34, 94 Sala-i-Martin, X. 22, 25 Salinas de Gortari: administration 125–7, 269 Sanguinetti, P. 263 Schatán, C. 59 Schumpeter, J. 182 Science and Technology (S&T) policy 166–8 Secretaría de Comercio y Fomento Industrial 244 sectoral integration: in Mexico and US 106 sectors: deﬁnition of vertically integrated 104; and dualistic structure 216, 286; export-oriented 135–8; industrial 106 Sempere, J. 265, 278, 281–3 Shaikh, A. 79–80 Silberston, A. 181 Small, D.H. 26, 31 Sobarzo, H. 265, 278, 281–3 Solis, L. 150 specialization 30, 102, 190–2; national 80 Sraffa, P. 103 Stability Pact: deﬂationary effects of 47 state debt: service of 277–8 Steindl, J. 175 Stock, J.H. 24

292

Index

Stockman, A. 28 Stokey, N.L. 191 structural change(s) 168; and allocation of resources 206–7; and competitiveness 57, 155; empirical analysis 213–16; and economic integration, 203–4, 234–5, 285–6; policies in the EU 250–1 structural cycle: deﬁnition 203 structural divergence 233 structural dynamics: deﬁnition 213 Structuralism: Latin American 200, 203 structure: doubly dualistic 233 Tabellini, G. 262 Taylor, M.P. 24–5 technological capabilities 30, 192–3; deﬁnition 148; and exports 160; and productivity growth 193; sectoral 162–5 technology 22, 104, 118, 190; gap 79 Ten Kate, A. 183 Tequila effect 7 Thygesen, N. 79 Tornell, A. 34 trade 191–5; deﬁcit 13, 15, 64, 78–9, 94, 135, 194; foreign 5–6, 9–11, 59, 79, 87; gains from 109; inter-industry 79; international, theory of 80, 102–3; intra-industry 57, 59–60, 73, 79, 115–16; neoclassical, theory of 116–18; opening 94, 179, 194; openness 191; patterns 79, 109–12 transition period 150–1 Treaty of Rome (also Tratado de Roma) 244 Tuirán, R. 70 Turner, R. 102

Turtelboom, B. 34 Tybout, J. 180, 185 Uekusa, M. 181 Ugarte, F. 152–3 Unger, K. 158 unit labo(u)r costs 80–1, 169 unit root 23–4; test 28, 95–6 Valdés, J.G. 123 van Wijnbergen, S. 180 Vanek, J. 103, 105 Vázquez Tercero, H. 59 Velasco, A. 34 Venables, A. 180 Verdoorn, J. 182 vertical integration 115–16; and comparative advantages 105–12; methodology of 104–5 von Hagen, J. 263, 278 von Hayeck, F. 123 wages 8; differential 46; formation 118–19; real 70, 80, 88–90, 138, 193 Watanabe, T. 108 Watson, M.W. 24 Weil, D. 22 Weiss, J. 180–1, 185 Weiss, L. 181 Weldon, J. 268 Westbrook, D. 180, 185 Westphal, L. 180 Wolff, E.N. 30 World Bank 127 Yotopoulos, P. 106 Zabludovsky, J. 126 Zedillo: administration 126–7

Mexico Beyond NAFTA: Perspectives for the European Debate (Routledge Studies in the Modern World Economy)

Japanese Production Organizations (Routledge Studies in the Modern World Economy)

Yoga in the Modern World: Contemporary Perspectives (Routledge Hindu Studies)

Mexico: from Montezuma to NAFTA, and beyond

International Perspectives on Temporary Work and Workers (Routledge Studies in the Modern World Economy)

Russia Moves into the Global Economy (Routledge Studies in the Modern World Economy)

The Early Modern European Economy

Mexico And Mexico City In The World Economy

Mexico And Mexico City In The World Economy

Fiscal Deficits in the Pacific Region (Routledge Studies in the Modern World Economy, 44)

Mexico (Modern World Nations)

The Capitalist World-Economy (Studies in Modern Capitalism)

The Economics of Language: International Analyses (Routledge Studies in the Modern World Economy)

Solving the Riddle of Globalization and Development (Routledge Studies in the Modern World Economy)

Rice: The Primary Commodity (Routledge Studies in the Modern World Economy, No 14)

Innovative Fiscal Policy and Economic Development in Transition Economies (Routledge Studies in the Modern World Economy)

Tax Systems and Tax Reforms in Europe (Routledge Studies in the Modern World Economy, 42)

Governing Rapid Growth in China: Equity and Institutions (Routledge Studies in the Modern World Economy)

Tax Systems and Tax Reforms in New EU Members (Routledge Studies in the Modern World Economy)

China in the World Economy (Routledge Studies on the Chinese Economy)

International Water Treaties: Negotiation and Cooperation Along Transboundary Rivers (Routledge Studies in the Modern World Economy)

Trade Liberalization and APEC (Routledge Studies in the Modern World Economy, 43)

Competition Policy Analysis (Routledge Studies in the Modern World Economy, 25)

Musical Ritual in Mexico City: From the Aztec to NAFTA

Macroeconomic Policies and Poverty Reduction (Routledge Studies in the Modern World Economy)

Competition and Trade Policies: Coherence or Conflict (Routledge Studies in the Modern World Economy , No 13)

China and Asia: Economic and Financial Interactions (Routledge Studies in the Modern World Economy)

Managing Fiscal Decentralization (Routledge Studies in the Modern World Economy, 36)

A General Theory of Institutional Change (Routledge Studies in the Modern World Economy)

Economics and Management of Technological Diversification (Routledge Studies in the Modern World Economy, 34)

International Labor Mobility: Unemployment and Increasing Returns to Scale (Routledge Studies in the Modern World Economy)

Mexico Beyond NAFTA: Perspectives for the European Debate (Routledge Studies in the Modern World Economy)

Japanese Production Organizations (Routledge Studies in the Modern World Economy)

Yoga in the Modern World: Contemporary Perspectives (Routledge Hindu Studies)

Mexico: from Montezuma to NAFTA, and beyond

International Perspectives on Temporary Work and Workers (Routledge Studies in the Modern World Economy)

Russia Moves into the Global Economy (Routledge Studies in the Modern World Economy)

The Early Modern European Economy

Mexico And Mexico City In The World Economy

Mexico And Mexico City In The World Economy

Fiscal Deficits in the Pacific Region (Routledge Studies in the Modern World Economy, 44)

Mexico (Modern World Nations)

The Capitalist World-Economy (Studies in Modern Capitalism)

The Economics of Language: International Analyses (Routledge Studies in the Modern World Economy)

Solving the Riddle of Globalization and Development (Routledge Studies in the Modern World Economy)

Rice: The Primary Commodity (Routledge Studies in the Modern World Economy, No 14)

Innovative Fiscal Policy and Economic Development in Transition Economies (Routledge Studies in the Modern World Economy)

Tax Systems and Tax Reforms in Europe (Routledge Studies in the Modern World Economy, 42)

Governing Rapid Growth in China: Equity and Institutions (Routledge Studies in the Modern World Economy)

Tax Systems and Tax Reforms in New EU Members (Routledge Studies in the Modern World Economy)

China in the World Economy (Routledge Studies on the Chinese Economy)

International Water Treaties: Negotiation and Cooperation Along Transboundary Rivers (Routledge Studies in the Modern World Economy)

Trade Liberalization and APEC (Routledge Studies in the Modern World Economy, 43)

Competition Policy Analysis (Routledge Studies in the Modern World Economy, 25)

Musical Ritual in Mexico City: From the Aztec to NAFTA

Macroeconomic Policies and Poverty Reduction (Routledge Studies in the Modern World Economy)

Competition and Trade Policies: Coherence or Conflict (Routledge Studies in the Modern World Economy , No 13)

China and Asia: Economic and Financial Interactions (Routledge Studies in the Modern World Economy)

Managing Fiscal Decentralization (Routledge Studies in the Modern World Economy, 36)

A General Theory of Institutional Change (Routledge Studies in the Modern World Economy)

Economics and Management of Technological Diversification (Routledge Studies in the Modern World Economy, 34)

International Labor Mobility: Unemployment and Increasing Returns to Scale (Routledge Studies in the Modern World Economy)

Recommend Documents