Unemployment in the New Europe

Unemployment in the New Europe The movement for European integration has yielded a European Union of fifteen states with a unified monetary system that will eventually embrace more than 370 million people. If current trends continue, an average of one in ten European Union adults will be unemployed. Not surprisingly, the European public ranks joblessness among its primary political concerns. This book focuses on both unemployment and economic unification. It examines the consequences of each and their interconnections. With chapters on the policy implications of European union, on current workings of domestic bargaining institutions, and on how unemployment affects political behavior, this book yields a message with important policy implications: The organized managed economies of Europe should be reformed but not replaced, and a united Europe should be wary of modeling itself on the United States. Nancy Bermeo is an Associate Professor of Politics at Princeton University, where she writes on political economy and regime change and directs the Southern European Research Group at the Center of International Studies.

Cambridge Studies in Comparative Politics General Editor Margaret Levi University of Washington, Seattle Associate Editors Robert H. Bates Harvard University Peter Hall Harvard University Stephen Hanson University of Washington, Seattle Peter Lange, Duke University Helen Milner Columbia University Frances Rosenbluth Yale University Susan Stokes University of Chicago Sidney Tarrow Cornell University

For a list of other books in the series, see the pages following the index.

Unemployment in the New Europe Edited by NANCY BERMEO Princeton University

PUBLISHED BY THE PRESS SYNDICATE OF THE UNIVERSITY OF CAMBRIDGE

The Pitt Building, Trumpington Street, Cambridge, United Kingdom CAMBRIDGE UNIVERSITY PRESS

The Edinburgh Building, Cambridge CB2 2RU, UK 40 West 20th Street, New York, NY 10011-4211, USA 10 Stamford Road, Oakleigh, VIC 3166, Australia Ruiz de Alarcón 13, 28014 Madrid, Spain Dock House, The Waterfront, Cape Town 8001, South Africa http://www.cambridge.org © Cambridge University Press 2001 This book is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 2001 Printed in the United States of America Typeface Janson Text 10/13 pt.

System QuarkXPress [BTS]

A catalog record for this book is available from the British Library. Library of Congress Cataloging in Publication Data Unemployment in the new Europe / edited by Nancy Bermeo. p. cm. – (Cambridge studies in comparative politics) Includes bibliographical references and index. ISBN 0-521-80241-5 (hardbound) – ISBN 0-521-00276-1 (pbk.) 1. Unemployment – European Union countries – Congresses. 2. Full employment policies – European Union countries – Congresses. 3. Europe – Economic integration – Congresses. I. Bermeo, Nancy Gina, 1951– II. Series. HD5764.A6 U484 2001 339.5¢094 – dc21 00–050355 ISBN 0 521 80241 5 ISBN 0 521 00276 1

hardback paperback

Contents

List of Contributors INTRODUCTION

page ix 1

Nancy Bermeo Part I Macroeconomic Contexts and Models 1

UNEMPLOYMENT, JOB CREATION, AND ECONOMIC AND MONETARY UNION

7

David Cameron 2

ORGANIZED MARKET ECONOMIES AND UNEMPLOYMENT IN EUROPE: IS IT FINALLY TIME TO ACCEPT LIBERAL ORTHODOXY?

52

Peter A. Hall 3

GLOBALIZATION, WELFARE STATES, AND EMPLOYMENT: IS THERE A EUROPEAN “THIRD WAY”?

87

Martin Rhodes Part II Unemployment and Domestic Bargaining Institutions: Challenging Some Myths 4

WAGE GROWTH, RECESSION, AND LABOR DECLINE IN THE INDUSTRIALIZED DEMOCRACIES, 1965–1993

121

Bruce Western and Kieran Healy 5

UNEMPLOYMENT AND UNION DENSITY

145

Lyle Scruggs and Peter Lange vii

Contents

Part III Unemployment and Domestic Bargaining Institutions: Three Cases 6

UNEMPLOYMENT AND TRADE UNION STRENGTH IN PORTUGAL

173

Alan Stoleroff 7

TRADE UNIONS, UNEMPLOYMENT, AND WORKING-CLASS FRAGMENTATION IN SPAIN

203

Javier G. Polavieja and Andrew Richards 8

MODELL NEDERLAND: SOCIAL PARTNERSHIP AND COMPETITIVE CORPORATISM IN THE NETHERLANDS

245

Steven B. Wolinetz Part IV Unemployment, Voting, and Political Behavior 9

DESPERATE TIMES CALL FOR DESPERATE MEASURES? UNEMPLOYMENT AND CITIZEN BEHAVIOR IN COMPARATIVE PERSPECTIVE

271

Christopher J. Anderson 10

THE POLITICS OF UNEMPLOYMENT: THE SPANISH EXPERIENCE IN COMPARATIVE PERSPECTIVE

291

José María Maravall and Marta Fraile 11

CONCLUSION: UNEMPLOYMENT, THE NEW EUROPE, AND THE OLD INEQUALITIES

329

Nancy Bermeo Index

viii

355

Contributors

Christopher J. Anderson Department of Political Science, Binghamton University, SUNY

Javier G. Polavieja Juan March Institute, Madrid, Spain

Nancy Bermeo Department of Politics, Princeton University

Andrew Richards Juan March Institute, Madrid, Spain

David Cameron Department of Political Science, Yale University

Martin Rhodes European University Institute, Florence, Italy

Marta Fraile Juan March Institute, Madrid, Spain

Lyle Scruggs Department of Political Science, University of Connecticut

Peter A. Hall Center for European Studies, Harvard University

Alan Stoleroff Instituto Superior de Ciências do Trabalho e da Empresa, Lisbon, Portugal

Kieran Healy Department of Sociology, University of Arizona Peter Lange Department of Political Science, Duke University José María Maravall Juan March Institute, Madrid, Spain

Bruce Western Department of Sociology, Princeton University Steven B. Wolinetz Department of Political Science, Memorial University of Newfoundland, St. John’s, NF, Canada ix

Introduction Nancy Bermeo

For Western Europe in the 1960s, full employment seemed a nearly achieved reality while full integration seemed an unattainable goal. Today, the opposite is true. The movement for European integration has yielded a European Union of fifteen states with a unified monetary system that will eventually connect over 370 million people (OECD 1998: 19). With an independent budget of over a billion dollars and a combined GDP of over 8 trillion dollars, the EU’s economy is only slightly smaller than that of the United States (Cottrell 1999). Its rate of joblessness is, alas, much larger. Europe’s days of nearly full employment began to wane in the mid1970s and never returned. Now, an average of one in ten West Europeans is without work. Unemployment among Europeans under the age of twenty-five hovers around 20%. Over 50% of those who are jobless have been without work for over twelve months (OECD 1999: 242). Not surprisingly, the European public ranks joblessness among its primary political concerns. This book focuses on both unemployment and economic unification.1 It examines the consequences of each and their interconnections. The connections between high unemployment and European Monetary Union are of special concern. The book’s scope is ambitious, and thus the contributors’ essays are wide-ranging. The first three offer, at the empirical level, 1

Most of these essays began as working papers for a Princeton University conference titled “Unemployment’s Effects” held in November 1997. Neither the conference nor this volume would have been possible without the generous assistance of the Luso-American Development Foundation, the Stanley J. Seeger Fund of the Hellenic Studies Program at Princeton University, the Spanish Ministry of Education and Culture, and Princeton University’s Center of International Studies and Center for Regional Studies. Sincere thanks for editorial and research assistance go to Ugo Amoretti, Megan Kennedy, Rogelio Pier-Martínez, and Brad Simmons.

1

Nancy Bermeo

broad overviews of the New European landscape. David Cameron sets the different national patterns of unemployment in relief, drawing our attention to cross-national contrasts and relating statistical peaks and valleys to the chronology of unification. Peter Hall looks at the institutional and economic bedrock beneath the patterns that Cameron describes. He points out the contrasts between the organized managed economy model long associated with continental Europe and the liberal market economy model associated with the United States and questions whether their recent performance trends (in the areas of unemployment and elsewhere) mean that the former should be abandoned. Martin Rhodes focuses on the corporatist and welfare institutions that lie at the foundation of the organized managed economy and analyzes the extent to which the economic liberalization drive associated with unification has changed Europe’s bedrock institutions already. Each of the opening essays thus provides the empirical background against which unemployment and unification can be understood. Yet they offer something more, for their empirical components yield a common general theme: the organized managed economy served Europe well in the past and, with renovation, may yet do so in the future. Though they each work with different materials, Cameron, Hall, and Rhodes all craft cautionary tales about the embrace of neoliberalism. Cameron concludes that “institutions that regulate labor markets” or “provide a locus for bargaining between economic actors” still have a salutary effect on job creation and deserve to be preserved. Hall concedes that the liberal economy has some advantages but insists that the organized managed economy offers a second viable model for economic success. He finds no “grounds for dismantling the organized market economies” and concludes that they are “capable of returning to reasonably low levels of unemployment.” Rhodes ends his argument in a similar place. Though he concedes that “boosting employment will certainly mean redesigning welfare systems,” he finds no valid “a priori justification for an assault on the public sector as such” and concludes that “neither welfare state retrenchment nor labor market deregulation is necessary” for Europe to “remain competitive.” Rather than destroying corporatist institutions, the “completion of a single market and movement to full monetary union are likely to lock bargaining partners more closely.” The essays that follow the opening section, on macroeconomic contexts and models, legitimate the defense of the organized managed economy (OME). Two essays using aggregate data on unemployment and trade 2

Introduction

unions challenge some key arguments used to legitimate neoliberal reform. Bruce Western and Kieran Healy examine whether unemployment dampens real wage growth, as pro-market analysts would predict. They find that the relationship varies greatly across time and institutional context. Prior to the mid-1970s, unemployment did discipline wage growth: they estimate that in liberal market economies (LMEs) a threepoint rise in joblessness lowered wage growth by about one point. In OMEs corporatist bargaining made wages twice as sensitive to unemployment, though these effects were nullified when collective bargaining coverage was extensive or when social democratic parties were in control. After 1974–75, when labor institutions began to weaken, wage sensitivity began to change. In the United States, the model LME, wage behavior is now “largely unrelated” to unemployment, whereas in Europe, the opposite is true. “Where union representation remains extensive and wages are set administratively rather than by the market, wages are now highly responsive to labor demand.” Western and Healy’s focus on the explanatory power of institutional context is mirrored in the essay by Lyle Scruggs and Peter Lange. Scruggs and Lange examine the argument that unemployment correlates with a drop in union density. They find that the effect varies greatly depending on two factors: whether union density is high when unemployment begins to rise and whether unions are responsible for the administration of unemployment benefits. In countries where union density is low or medium or where unions do not administer benefits, unemployment does drive density down. Yet, where density is high, unemployment has no effect, and where unions do administer benefits, unemployment actually increases density. Thus the connection between unemployment and union density – a key form of associational life – is not purely a market one but is heavily mediated by institutional arrangements. The positive effects of the institutions of the organized managed economy are illustrated by the three case studies in Section III. Portugal used its state-owned enterprises to avoid high unemployment during the first decades of its newly consolidating democracy. Unemployment has risen and has had a deleterious effect on union density and bargaining power, as the essay by Alan Stolleroff illustrates. Yet for a variety of institutional reasons, Portugal’s state sector shed labor and privatized only after redundant workers could be absorbed by other sectors of the economy. As a result, Portugal has consistently maintained one of the lowest rates of unemployment in Western Europe. 3

Nancy Bermeo

Spain did not avoid high unemployment as the Portuguese did. Instead, its rate of joblessness was at or near the Continental high for well over a decade. The essay by Javier Polavieja and Andrew Richards shows how the rise in joblessness reduced identification with unions and raised their tolerance for liberalizing reforms. Yet, as José Maravall and Marta Fraile show, Spanish political elites used welfare spending and corporatist bargaining to mitigate unemployment’s most negative effects. Support for the democratic system has remained remarkably high. The Netherlands adds an interesting gradation to the contrasting cases of Portugal and Spain. As Steven Wolinetz illustrates, the Dutch began to struggle with a sharply rising unemployment rate beginning in the late 1970s. Yet, they managed to reverse the trend by 1983, through the revival of older corporatist structures and the creative reform of state spending. In this case, as in each of the others, institutions fundamental to the organized managed economy performed a positive role – either preventing unemployment, reversing unemployment, or softening its impact through compensatory spending. The fact that political institutions were able to play a positive role in each of these cases helps explain Christopher Anderson’s findings in the closing section of the book. Anderson examines the effects of unemployment on the political opinions and behavior of the unemployed and finds that the jobless are neither especially given to political extremism nor especially xenophobic. He also finds that unemployed citizens in the more extensively developed welfare states are significantly more satisfied with their own lives and with the functioning of their own national political systems than the unemployed who live in less protective policy environments. In the book’s conclusion, I show how various European institutions mediate the effects of unemployment and how this mediation has benefits that the ascendant liberal economic models do not afford. Using evidence from our case studies, from Europe more broadly, and from the United States, I end the book with a cautionary tale of my own. References Cottrell, Robert. 1999. Europe: So Far, It Flies. New York Review of Books (April 1999): 66–73. OECD. 1999. Employment Outlook 1999. Paris: OECD.

4

PA R T I

Macroeconomic Contexts and Models

1 Unemployment, Job Creation, and Economic and Monetary Union David Cameron

There once was a time, not long ago, when Europe was close to full employment.1 In the 1960s and early 1970s, after the completion of the postwar recoveries and before the demise of the Bretton Woods exchange rate regime, unemployment rates throughout Europe were generally in the range of 2 to 3 percent or lower (European Commission 1998b: 224–53). It appeared, as Andrew Shonfield proclaimed in Modern Capitalism, that, with few exceptions, governments had – through indicative planning; increased cooperation between business, government, and labor; and the application of Keynesian principles to macroeconomic policy – overcome the job-destroying effects of business cycles and recessions (Shonfield 1969). Now, some three decades later and in the wake of the major recessions of 1974–75, 1980–84, and 1991–94, Europe is afflicted with enduring high levels of unemployment. Throughout the 1990s, the fifteen member states of the European Union (EU) experienced an average rate of unemployment of about 10 percent, an almost fivefold increase from the average for the fifteen states in the 1960s. Even some half-dozen years after the end of the last major recession, and despite the sustained recovery in much of Europe in the last half of the 1990s, the rate of unemployment remained close to double digits in the EU as a whole and in double digits in the eleven member states that formed the euro-zone,

1

Earlier versions of this chapter were presented at conferences or seminars at Princeton University, Harvard University, and the European University Institute. For their helpful comments and suggestions, I wish to thank Jens Alber, Sheri Berman, Nancy Bermeo, Matthew Canzoneri, Peter Hall, Torben Iversen, Peter Kenen, Sofía Pérez, Martin Rhodes, and Thomas Risse.

7

David Cameron

Table 1.1. Growth, Unemployment, and Employment in the European Union, the United States, and Japan, 1961–1999

and it was expected to remain at or close to those levels in the foreseeable future.2 What makes the long-term deterioration in employment in Europe especially notable, of course, is the fact that it has been far more severe than in other advanced economies (see Table 1.1). Thus, while the decrease 2

8

The four member states that did not move to the third and final stage of Economic and Monetary Union on January 1, 1999, are Denmark, Greece, Sweden, and the United Kingdom. Greece subsequently satisfied the conditions for entry to the third and final stage and did so on January 1, 2001.

Unemployment, Job Creation, and Union

in recent decades in the decade-long average rate of growth was only slightly greater in the EU than that experienced in the United States, and far less than the precipitous decrease that occurred in Japan, the average level of unemployment in the fifteen states that now constitute the EU rose far more dramatically than in either the United States or Japan, from slightly more than 2 percent in the 1960s to 4 percent in the 1970s, 9 percent in the 1980s, and more than 10 percent in the 1990s.3 The sustained recovery in much of Europe in the mid- to late 1990s, and the steady, albeit gradual, decline over several years in the rate of unemployment, led many observers to believe that, with sufficient time, the number of unemployed in Europe would return to more tolerable levels. That optimistic scenario was called into question by the looming shadow of the global financial crisis that began in 1997. But by the second quarter of 1999, after having experienced a slowdown in the rate of growth, most of the European economies had resumed their recovery and could look forward to a continuing decline in the rate of unemployment in the near future. Despite that recovery, however, the rate of unemployment in the eleven member states constituting the euro-zone remained above 10 percent, and the European Central Bank (ECB), as well as the International Monetary Fund and the OECD, predicted that it would remain above that level in 2000.4 This chapter examines the enduring problem of high unemployment in Europe, beginning with a detailed description of its magnitude and distribution. While Europe, as a whole, has become a region of enduring high unemployment, the levels of unemployment vary widely across the member states and, within the member states, across both time and space. I note some of the more important bases of variation. Next, I consider why, notwithstanding the general upward trend in unemployment throughout Europe in recent decades, some of the EU 3

4

For discussions of the high levels of unemployment in Europe in the 1980s and 1990s, see, e.g., Alogoskoufis et al. (1995); Therborn (1986); Bean, Layard, and Nickell (1987); Blanchard and Summers (1987); Lindbeck and Snower (1988); Summers (1990); Layard, Nickell, and Jackman (1991; 1994); Lindbeck (1993); Bean (1994); Benoit-Guilbot and Gallie (1994); Bentolila and Blanchard (1990); OECD (1994; 1997; 1999); Blanchard, Jimeno, et al. (1995); Drèze and Malinvaud (1995); Saint-Paul (1995; 1996); Henry and Snower (1996); Dolado and Jimeno (1997); European Commission (1997a; 1998a); Nickell (1997; 1999); Snower and de la Dehesa (1997); Blanchard and Fitoussi (1998); Rubery et al. (1998); IMF (1999). The ECB’s forecast appeared in its June 1999 monthly report. Those of the European Commission, the IMF, and the OECD are reported in Financial Times, September 10, 1999, 4.

9

David Cameron

member states have been more successful than others in creating large numbers of new jobs and in resisting the upward trend – and, indeed, in a few instances, have even been able to reduce the rate of unemployment over the long term. I conclude with a brief consideration of how the historic advent of the third and final stage of Economic and Monetary Union (EMU) is likely to affect employment and unemployment in the Union.

Unemployment in Europe across Time and Space Figure 1.1 presents the rates of unemployment in the European Union, the United States, and Japan over the period 1970–98. These data illustrate the extent to which and the point at which the EU became a region of high unemployment in relation to the other major developed economies. While Europe was, prior to the first of the two OPEC oil price shocks of the 1970s, a region of relatively low unemployment, especially when compared with the United States, the rate of unemployment rose sharply in the wake of those shocks and reached double digits by the mid-1980s. Then, after dropping in the late 1980s, the rate of unemployment rose to even higher levels in the early 1990s and for the rest of the decade remained at or close to the peaks registered in those years. In contrast, after having been more than twice as high as the European rate in the early 1970s and then moving upward in a series of sharp increases in the early 1970s, the mid-1970s, and the early 1980s, the rate of unemployment in the United States began a long downward movement, interrupted only briefly in 1991–92. As a result,

Figure 1.1 Unemployment Rates in the European Union, the United States, and Japan, 1970–1998. Source: IMF (1999: 88). 10

Unemployment, Job Creation, and Union

by the end of the 1990s, the rate of unemployment in the United States was some half-dozen percentage points lower than that in the EU. Figure 1.2 presents the unemployment rate in the EU since 1970 along with the annual change in total employment. Not surprisingly, the two are related; the long-term upward drift in unemployment was driven by the three major recessions of the mid-1970s, early 1980s, and early 1990s that, collectively, resulted in the net loss of some nine million jobs. Especially during the recessions of the early 1980s and early 1990s, when total employment dropped in both instances for three consecutive years, the result was an increase of several percentage points in the EU-wide rate of unemployment, from less than 6 percent to about 10 percent in the early 1980s and from less than 8 percent to about 11 percent in early 1990s. The upward trend in the rate of unemployment in Europe portrayed in Figures 1.1 and 1.2, and the fact that the rate was higher after each major recession than it was prior to each recession, calls into question the notion that there is a fixed “natural,” or equilibrium, or “structural” rate of unemployment in Europe. It lends credence, instead, to the argument of

Figure 1.2 The Annual Change in Total Employment and the Rate of Unemployment in the European Union, 1970–1996. Source: European Commission (1997a: 13). 11

David Cameron

Blanchard and Summers (1986) and others that unemployment is characterized by hysteresis: that is, rather than remaining unchanged over time, the equilibrium, or “natural,” or “structural” rate changes, in part because of the effect of prior levels of unemployment. Moreover, the pattern of movement in the upward drift of unemployment suggests that cyclical and structural unemployment are not easily disaggregated and that waves of cyclical unemployment such as occur in major recessions result in higher levels of structural unemployment after the recessions, in effect raising the “natural” or equilibrium level. That being the case, it follows that the more frequent and closely spaced the major synchronized recessions in Europe, the more likely it is that the level of structural unemployment will increase over time, just as, conversely, the more infrequent and separated the major recessions (as in the United States since the early 1980s), the more likely it is that the level of structural unemployment will decrease over time. The fact that the long upward drift in unemployment portrayed in Figures 1.1 and 1.2 was punctuated by the deepening recessions of the mid1970s, early 1980s, and early 1990s suggests an important macroeconomic fact of life in Europe. Each of those major recessions derived from processes and changes in the international economic and geopolitical environments – processes and changes that were largely exogenous to the EU and, for that reason, largely beyond its control. For example, the major recession of the mid-1970s followed, and was to a large extent triggered by, the American inflation of the late 1960s and the breakdown of the Bretton Woods exchange rate regime, the Yom Kippur War, and the fourfold increase in oil prices by the Organization of Petroleum Exporting Countries (OPEC). Likewise, the major synchronized recessions of the early 1980s were, in large part, triggered by the effects on oil production and prices of the Iranian Revolution and the Iran-Iraq War. And the major synchronized recessions in Europe in the early 1990s were, in large part, triggered by the upward pressure on European interest rates that followed increases in German interest rates in the late 1980s and early 1990s, especially those which occurred in 1990 and 1991 in the wake of unification. To some extent, of course, the latter increases were endogenous. On the other hand, they were driven in no small measure by the unification-related increases in transfer payments, government spending, budget deficits, and money supply and can therefore be viewed as the product of exogenous forces – specifically, those forces in Central and Eastern Europe and the former Soviet Union that gave rise to the historic events of 1989–91. 12

Unemployment, Job Creation, and Union

While the two most recent synchronized recessions in Europe were, like the first in the mid-1970s, largely the product of exogenous forces, it would be wrong to conclude that those recessions, and the sharp increases in unemployment that resulted, were entirely beyond the control of European governments. Indeed, it would appear that their effects were amplified by national and EU policies – in particular, by EU initiatives in the domain of monetary and exchange rate policy and the responses of member states to those initiatives. In the early 1980s, for example, soon after the creation of the European Monetary System (EMS) in 1979, the norms that governed German macroeconomic policy – the perceived need to maintain stable prices and the willingness to deploy a tight monetary policy and an austere fiscal policy in order to maintain stable prices – were internalized in the socialist governments of France and Spain and subsequently generalized to the rest of the European Community. As they were, the rate of unemployment in both countries, and in the Community as a whole, increased sharply.5 Likewise, the combined effects of the EMS crisis of 1992–93, which, in the absence of a revaluation of the German mark and a broad realignment of all of the exchange rates, resulted in a plethora of individual realignments and upward shifts of interest rates in a number of member states, contributed to the further erosion of employment in the early 1990s. And that contractionary effect was subsequently amplified when member states sought, in the mid-1990s, to bring their budget deficits, inflation rates, and debt levels in line with the “convergence criteria” stipulated by the Treaty on European Union for qualification for the third and final stage of EMU.6

Job Creation in the European Union and Elsewhere In addition to presenting the decade-long averages since 1960 of the rates of economic growth and unemployment in the EU, the United States, and Japan, Table 1.1 presents the decade-long average annual rates of change in the total number of persons employed. Note the marked difference between the United States and the EU in the extent to which new jobs have been created over the past four decades. The total number of 5

6

On French policy under Mitterrand and the Socialist-dominated government in the early 1980s, see Cameron (1996). On the policy of the Spanish Socialist government that came to power in 1982, see Pérez (1997; 1999a). The criteria are described in Article 109j(1) and Protocol 6 of the treaty. See European Communities (1992).

13

David Cameron

employed in the fifteen EU member states increased by a very modest amount – on average, by 0.3 to 0.5 percent a year in the 1960s, 1970s, and 1980s, and not at all in the 1990s. In sharp contrast, the total number employed in the United States increased by roughly 2 percent a year between 1960 to 1990 and by 1.5 percent a year in the 1990s. Figures 1.3 and 1.4 illustrate the cumulative effect of the marked differences in the rates of job creation in the EU, the United States, and Japan

Figure 1.3 Number of Persons in Millions Employed in the European Union, the United States, and Japan, 1970–1998. Source: IMF (1997b: 39).

Figure 1.4 Percentage of Working-Age Population Employed in the European Union, the United States, and Japan, 1975–1997. Source: European Commission (1998a: 25). 14

Unemployment, Job Creation, and Union

in recent decades. Figure 1.3 shows the number of persons employed over the past three decades. Since 1970, the American economy has added roughly 50 million new jobs, net of losses, to the 80 million that already existed, an increase of more than 60 percent. The Japanese economy added roughly 15 million new jobs, net of losses, to the 50 million that existed in 1970, an increase of 30 percent. The EU, on the other hand, added barely 10 million new jobs, net of losses, to the 140 million that existed in 1970, an increase of only 7 percent over the three decades! As a result of their higher rates of job creation over the past three decades, both the United States and Japan achieved substantial increases in the proportions of the working-age population employed, whereas in the EU the proportion actually decreased. Thus, between 1975 and 1997 the employment rate in Japan increased from about 69 percent to about 75 percent. In the U.S., the increase was even greater, from about 62 percent to nearly 75 percent (Figure 1.4). But in the EU, the employment rate decreased, from about 64 percent in the mid-1970s to 60 percent in the mid-1980s, where it has remained ever since, with the exception of a transitory improvement in the late 1980s and early 1990s. Clearly, the high rate of unemployment is only one facet of the employment problem in Europe. Another equally important aspect of the problem is the failure to create a sufficient number of new jobs to increase the employment rate of its working-age population.

Variations in Unemployment and Employment within the European Union Table 1.2 presents standardized measures of unemployment in the fifteen current member states of the European Union over the period 1961–99, as reported by Eurostat and the OECD. These rates illustrate the sharp increase in unemployment that occurred in most of the member states in the decade after the first OPEC oil shock, as well as the continued erosion of employment that occurred in almost all of the current member states in the 1980s and 1990s. The data indicate that the rates of unemployment dropped in most of the member states in the late 1990s. But those decreases were, for the most part, very modest, and as a result the rates of unemployment in most countries remained, despite the economic recovery in the mid- to late 1990s, close to the historic peaks registered earlier in that decade. Indeed, in several countries – Belgium, Germany, Greece, France, Italy, Austria, and Sweden – the rates of unemployment recorded in 1997–99 were equal to or exceeded those of the early 1990s. 15

David Cameron

Table 1.2. Percentage of Civilian Labor Force Unemployed in the European Union, 1961–2000 1961–73

1974–85

1986–90

1991–96

1998

1999

2000

Ave.

Ave.

Ave.

Ave.

Belgium

2.0

7.7

8.7

8.9

9.5

9.0

8.4

Denmark

0.9

6.4

6.4

8.3

5.2

5.2

4.9

Germany

0.7

4.2

5.9

7.6

9.4

8.7

8.3

Greece

4.5

3.8

6.6

8.5

10.7

10.4

10.0

Spain

2.8

11.3

18.9

21.1

18.8

15.9

13.9

France

2.2

6.4

9.7

11.3

11.8

11.3

10.3

Ireland

5.6

10.6

15.5

14.0

7.6

5.8

4.9

Italy

5.2

7.0

9.6

10.6

11.8

11.3

10.8

Luxembourg

0.0

1.7

2.1

2.6

2.7

2.3

2.1

Netherlands

1.3

7.1

7.4

6.4

4.0

3.3

2.6

Austria

1.6

2.5

3.4

3.6

4.5

3.7

3.4

Portugal

2.5

6.9

6.1

5.9

5.2

4.5

4.4

Finland

2.3

4.8

4.3

13.7

11.4

10.2

8.9

Sweden

2.0

2.4

2.0

7.6

8.3

7.2

6.5

U.K.

2.0

6.9

9.0

9.3

6.3

6.1

5.8

EU

2.3

6.4

8.9

10.1

9.9

9.1

8.5

Euro-ll

2.5

6.6

9.4

10.7

10.9

9.9

9.2

Source: European Commission (2000: 164–5). Note: The data reported here are seasonally adjusted and use Eurostat’s standardized definition of unemployment. The data are identical to the standardized measures of unemployment reported by the OECD.

There are, of course, some significant exceptions to the tendency for unemployment in the late 1990s to have remained at or close to the decade-long peaks. In several member states – most notably, Denmark, Spain, Ireland, the Netherlands, Portugal, and the United Kingdom – the 16

Unemployment, Job Creation, and Union

level of unemployment by 1999 was significantly lower than the average level registered not only in the first half of the 1990s but also in the late 1980s. Indeed, the levels of unemployment in Denmark, Ireland, the Netherlands, Portugal, and the United Kingdom in 1998–99 were lower than the levels registered in 1974–85 as well! Although situated in a regional context marked by low growth and deteriorating employment, those countries – above all, Ireland and the Netherlands, but Denmark, Portugal, and the United Kingdom as well – evidently were able to create enough new jobs over a period of years to lower the long-term, noncyclical, or structural, or “natural,” rate of unemployment. The trends in unemployment in selected EU member states over the period since 1970 resemble the long gradual upward trend in unemployment in the EU as a whole (Figure 1.5a). But in contrast to the trends in Germany, France, and Italy, and the EU as a whole, the rates of unemployment in Ireland, Denmark, the Netherlands, and the United Kingdom dropped significantly in the mid- and late 1990s. In Ireland, for example, unemployment dropped some 7 to 8 percentage points from the peak rates of 17 percent in the mid-1980s and 15 percent in the early 1990s (Figure 1.5b). If there are substantial differences in rates of unemployment among the EU member states, there are also substantial variations within many of the member states, which should caution one about drawing conclusions about unemployment solely on the basis of national averages. Perusal of a map of the EU with the rates of unemployment for the various regions of the member states reveals areas having unusually high rates of unemployment and other areas with unusually low rates of unemployment (European Commission 1998a: 31). In recent years a few of the smaller states – most notably, Austria, Denmark, and the Netherlands (and Luxembourg, of course) – have enjoyed relatively low rates of unemployment, as have substantial portions of northern and northeastern Italy, southern and southwestern Germany, southeastern England, and northern Portugal. But in other regions – most notably, southern Spain, where the rate has exceeded 35 percent; southern Italy and eastern Germany, where unemployment has averaged 20 percent or more in recent years; and northern Sweden, central and northern Finland, southern France, northern France, and parts of Wallonia – the unemployment rates have been unusually high. These geographic concentrations of unusually high levels of unemployment, alongside concentrations of low unemployment, suggest that unemployment in Europe depends, to a considerable degree, upon regionally concentrated 17

18

Figure 1.5a The Rate of Unemployment in Selected EU Member States, 1970–1998. Source: European Commission, European Economy no. 68.

19

Figure 1.5b

David Cameron

processes of economic development and underdevelopment – such as the expansion of the service sector (as in the London, Paris, and Milan regions), on one hand, or deindustrialization (as in eastern Germany, southern Belgium, and northern France) or contraction of the primary sector (as in southern Spain, southern Italy, northern Sweden, and northern and central Finland), on the other (see Iversen and Wren 1998). Substantial differences exist in rates of employment as well. Table 1.3 shows the average annual change in total employment in the EU member

Table 1.3. Average Annual Percentage Change in Total Employment in the European Union, 1961–1999

20

Unemployment, Job Creation, and Union

states decade by decade since 1960. These data highlight the differences in the average annual rate of job creation (net of job losses), which are masked by the overall EU-wide rates of change. Thus, for example, while the aggregate annual rate of net job creation in the EU in the 1990s was zero – indicating that, on average, as many jobs were lost as were created each year of the decade throughout the EU – and while certain member states – Finland and Sweden, which experienced unusually severe recessions in the early 1990s, Germany in the wake of unification, and Italy – experienced significant contractions in employment, several member states experienced significant increases in employment that ran counter to the EU-wide trend. Most notably, average annual increases in Luxembourg, the Netherlands, and Ireland in the 1990s were altogether exceptional – not only for the 1990s but also in comparison with any previous decade. The sustained job creation in the 1990s in Ireland, Luxembourg, the Netherlands, and several other member states stands in marked contrast not only to the sharp contractions in Finland, Sweden, Germany, and Italy but also to the relative stagnation in job creation experienced by Belgium, France, and Britain. A comparison of the cumulative effect of those differences in annual rates of job creation during the late 1980s and 1990s (Table 1.4) indicates that the aggregate rates of job creation in the EU as a whole mask significant differences among the member states. Thus, for example, while the total number of employed persons dropped by almost 10 percent in Finland after 1985 and by almost 10 percent in Sweden in the 1990s alone, and by 4 to 7 percent in Germany and Italy in the 1990s, the number of employed persons in Ireland, the Netherlands, and Luxembourg increased substantially in the 1990s. Table 1.5 presents data on the proportion of the working-age population – that is, persons aged 15 to 64 – employed in the fifteen EU states in 1985, 1991, and 1998. As noted earlier in regard to Figure 1.4, approximately 60 percent of the working-age population in the EU was employed in the 1980s and 1990s. That rate, and its stability over time, stand in marked contrast to the employment rates in both the United States and Japan, where the rates increased over the three decades and by the late 1990s were in the vicinity of 75 percent. Some of the EU member states – most notably, Denmark, Germany, Austria, Finland, Sweden, Portugal, and the United Kingdom – had rates above the EU average during much of the period (although the severe recessions in Finland and Sweden in the 1990s resulted in substantial decreases in the employment rates of both countries). However, others – most notably, Italy, Spain, Greece, and 21

David Cameron

Table 1.4. Cumulative Percentage Change in Total Employment in the European Union, 1985–1998 1985–91

1991–98

1985–98

Belgium

5.9

3.7

9.8

Denmark

2.0

4.9

7.0

Germany

10.2

-7.0

na

Greece

1.2

9.2

10.5

Spain

18.5

4.7

24.1

France

4.1

0.8

5.0

Ireland

6.1

29.5

37.3

Italy

4.1

-4.1

-0.1

Luxembourg

21.9

21.0

47.5

Netherlands

14.6

12.2

28.6

Austria

7.4

2.6

10.2

Portugal

11.3

0.6

11.9

Finland

-4.0

-5.8

-9.5

Sweden

3.6

-9.5

-6.4

U.K.

7.9

2.6

10.7

EU

7.7

-0.4

na

Source: European Commission (1999a: 127–42). Note: The data for Germany in 1985–91 are for the pre-1990 Federal Republic only.

Ireland – had relatively low employment rates and remained below the EU average throughout the period. What is most interesting, perhaps, is that several member states experienced significant increases in their rate of employment between the mid-1980s and late 1990s. In particular, Belgium, Spain, Ireland, the Netherlands, Portugal, and Britain experienced increases of at least several 22

Unemployment, Job Creation, and Union

Table 1.5. Percentage of Working-Age Population (15–64) Employed in the European Union, 1985–1998 Change, 1985–98

1985

1991

1998

Belgium

53.1

54.7

57.5

4.4

Denmark

77.4

77.6

78.9

1.5

Germany

63.1

65.9

61.5

-1.6

Greece

57.3

56.6

57.2

-0.1

Spain

44.1

49.7

50.2

6.1

France

62.0

62.9

60.8

-1.2

Ireland

51.4

53.5

60.5

9.1

Italy

53.0

53.6

51.7

-1.3

Luxembourg

58.0

57.7

58.6

0.6

Netherlands

57.7

62.2

68.3

10.6

Austria

67.3

69.7

70.1

2.8

Portugal

63.5

66.2

68.9

5.4

Finland

75.2

75.7

65.1

-10.1

Sweden

80.3

82.8

70.3

-10.0

U.K.

66.2

72.4

71.4

5.2

EU

59.8

62.4

61.1

1.3

Source: European Commission (1999a: 127–42).

percentage points or more in their rates of employment, in marked contrast to Germany, France, Italy, Greece, Finland, and Sweden, in all of which the rate decreased. Some of the countries that experienced significant increases had initial employment rates that were well below the EU average, so the increases are not surprising. But several others – most notably, the Netherlands, Portugal, and Britain – experienced significant increases despite the fact that each was already close to or above the EU average in 1985. 23

David Cameron

Gender, Age, and Unemployment Table 1.6 presents the rates of unemployment in the EU, the United States, and Japan in 1998, disaggregated by gender and age. Notwithstanding conventional stereotypes, there appears to be a much more marked gender division in unemployment in the European Union than in either the United States or Japan. Thus, the rate of unemployment among women in the EU was roughly three percentage points, or 33 percent, higher than that among men, whereas in the United States and Japan the rates were nearly identical for men and women. (See Rubery, Smith, Fagan, and Grimshaw 1998.) Table 1.6 also presents the rates of unemployment in 1998 of those in the active labor force who were under the age of 25. In general, these rates are twice those of the labor force as a whole in the EU, a little more than twice as large in the United States, and a little less than twice as high in Japan. Thus, in the EU, almost 20 percent of those in the labor force who were under the age of 25 were unemployed. And as with the labor force as a whole – and, interestingly, again in contrast to the situation in the United States and Japan – the rate of unemployment of young European women exceeds that of young men by several percentage points. Table 1.7 presents the rates of unemployment in the EU member states in 1998, disaggregated by gender and age. What is perhaps most note-

Table 1.6. Unemployment by Age and Gender in the European Union, the United States, and Japan, 1998 (percentage of active labor force)

24

Unemployment, Job Creation, and Union

Table 1.7. Unemployment in the European Union by Age and Gender, 1998 (percentage of active labor force)

worthy, and indeed sobering, is the high rate of unemployment of the young, particularly of young women. In some countries, more than 30 percent of those under the age of 25 who are in the labor force are unemployed, and in some countries, more than 40 percent of the women under 25 who are in the labor force are unemployed. Notwithstanding the 25

David Cameron

(relative) success of a few member states such as Denmark, Germany, Luxembourg, the Netherlands, and Austria in keeping the unemployment rate for the young in the single digits and close to the overall rate, in most of the EU those under the age of 25 – especially young women – confront the prospect and reality of unemployment to a much greater degree than do their peers in the United States and Japan. Thus, for example, in Spain, Italy, and Greece, more than 30 percent of all young persons in the labor force, and more than 35 percent of all young women in the labor force, were unemployed in 1998. In France, Belgium, and Finland, the rates of unemployment among young persons and young women in particular were somewhat lower but still very high. This, we should note, was the case despite the fact that the rates of participation in the labor force were not unusually high in these countries and, indeed, except for France and Finland, were significantly lower than the EU average.

Long-Term Unemployment One of the most problematic aspects of European unemployment has been and remains the very high rate of long-term unemployment (see BenoitGuilbot and Gallie 1994). Table 1.8 presents the proportion of all those unemployed in 1997 in each of the member states who were out of work for more than one year. For the EU as a whole, approximately one-half of all of the unemployed had been unemployed for more than a year. (Approximately one-third had been unemployed for more than two years.) That figure is far greater than the rate in the United States, where typically about 10 percent of the unemployed have been without work for at least a year. As with unemployment in general, the experience of long-term unemployment was more frequent among European women than men (and also among the younger and older cohorts of the labor force compared with those aged 25–55). And as with the other facets of employment and unemployment, there are marked variations within the EU. In Germany, Greece, and Ireland, the proportion of the unemployed who had been unemployed for more than one year exceeded 50 percent, and in Belgium and Italy the proportion exceeded 60 percent! Several of the member states – Denmark, Austria, Finland, and Britain, as well as Sweden in regard to women – appear to have been markedly more successful than the others in reducing the number of persons unemployed for more than a year in relation to the total number of persons unemployed. The relative success of those countries may reflect their propensity and ability to create low26

Unemployment, Job Creation, and Union

Table 1.8. Percentage of All Unemployed Who Had Been Unemployed for More Than One Year, 1998 All

Men

Women

Belgium

61.1

59.7

63.9

Denmark

27.5

25.6

30.8

Germany

52.1

49.4

55.9

Greece

55.2

44.9

61.5

Spain

50.0

44.2

54.5

France

41.9

41.4

42.0

Ireland

56.1

63.3

46.9

Italy

68.0

67.0

68.5

Luxembourg

na

na

na

Netherlands

47.5

50.0

46.2

Austria

29.7

28.2

32.7

Portugal

43.1

41.5

42.2

Finland

28.1

33.3

23.3

Sweden

37.3

40.7

33.8

U.K.

33.3

38.6

23.6

EU

49.0

47.7

49.2

Source: European Commission (1999a: 127–42).

skilled jobs in the public service sector: in both Denmark and Sweden, roughly 30 percent of all jobs are in the public sector; Austria and Finland have substantial and labor-intensive public sectors as well.7 However, their apparent “success” may, of course, also reflect their ability to induce,

7

On the role of public service sector employment in alleviating unemployment, see, e.g., Hemerijck and Schludi (1999).

27

David Cameron

through the limitations placed on unemployment benefits, the long-term unemployed to leave the active labor force.

Part-Time Employment One possible explanation for the considerable variation in the extent and duration of unemployment across the EU involves the existence of opportunities in some countries but not others for persons to work in part-time jobs. Table 1.9 presents data on the proportion of those employed in 1998 Table 1.9. Part-Time Employment in the European Union, 1998 (percentage of total employed) All

Men

Belgium

15.7

3.5

33.3

Denmark

22.3

10.9

35.8

Germany

18.3

4.7

36.4

Greece

6.0

3.3

10.6

Spain

8.1

3.0

17.2

France

17.3

5.7

31.6

Ireland

12.3

5.4

23.2

Italy

7.3

3.4

14.1

Luxembourg

9.4

1.9

22.2

Netherlands

38.8

18.1

67.9

Austria

15.8

4.4

30.3

Portugal

11.1

6.2

17.4

Finland

11.7

6.9

16.9

Sweden

23.9

9.2

40.7

U.K.

24.9

8.8

44.9

EU

17.4

5.9

31.8

Source: European Commission (1999a: 127–42).

28

Women

Unemployment, Job Creation, and Union

who were working part-time. What is especially noticeable is that women have much higher rates of part-time employment than men. In the EU as a whole, and in Austria, Belgium, Denmark, Germany, and France, more than 30 percent of all working women work in part-time jobs. In Sweden and the United Kingdom the figure exceeds 40 percent. And in the Netherlands, it approaches 70 percent! In contrast, in several member states with unusually high rates of overall unemployment and unemployment of young persons and young women in particular – for example, Greece, Spain, and Italy – the rates of part-time employment are well below the average for the EU. In certain member states – most notably, the Netherlands and Denmark, but also Sweden, the United Kingdom, and Finland, all of which have either unusually low levels of unemployment (the Netherlands, Denmark, the United Kingdom) or unusually low rates of long-term unemployment (Denmark, Sweden, Finland) – a relatively large number of men work part-time. While the relationships are by no means overwhelming, it would appear that the existence of opportunities for parttime employment may help to explain why some member states have lower levels of overall unemployment, unemployment among women, and longterm unemployment.

The Political, Economic, and Institutional Sources of Employment Over the last quarter century, most of the member states of the European Union experienced dramatic increases in unemployment and entered the twenty-first century with rates of unemployment that were relatively high compared with those in the United States and Japan or those in Europe two or three decades earlier. But the EU-wide trends in unemployment and employment mask considerable divergence among the member states. Thus, while a number entered the new century with unusually high rates of unemployment that remained close to or above the peak levels experienced in the synchronized recession of the early 1990s, at least a few were able to create sufficient numbers of new jobs to lower substantially their rates of unemployment in the 1990s and, indeed, over the longer term. Here we consider why some of the European states were more successful than others in dampening the rate of unemployment and, in some instances, significantly reducing the noncyclical, structural rate over the longer term. While the overall rate of unemployment in the EU in the late 1990s was in the vicinity of 10 percent, certain member states – most notably, 29

David Cameron

Luxembourg, the Netherlands, Austria, Portugal, Denmark, the United Kingdom, and Ireland – enjoyed significantly lower rates (see Table 1.2). Indeed, by mid-2000, the rate of unemployment was below 5 percent in Denmark, Portugal, and Austria, and below 3 percent in the Netherlands and Luxembourg. Similarly, while the rate of unemployment in several countries either remained virtually unchanged or even increased in the late 1990s (Table 1.2), in others – again, Denmark, Ireland, the Netherlands, Portugal, and the United Kingdom, as well as Spain and Finland – it dropped significantly in the late 1990s. And in some of the member states – especially the Netherlands and Ireland but also in Portugal, Denmark, and the United Kingdom – the rate of unemployment in the late 1990s was significantly lower than it had been in the late 1970s and early 1980s. While the rate of unemployment among young persons was roughly double the rate for all persons and about 20 percent across the EU as a whole in the late 1990s (Table 1.7), in some member states – most notably, Austria, Luxembourg, Denmark, the Netherlands, Portugal, and Germany – the rate was much lower, in fact below 10 percent. And while the rate of unemployment of young women was especially high, in some member states – again, Austria, Denmark, Germany, Luxembourg, the Netherlands – the rate was unusually low, below 10 percent. Likewise, while roughly one-half of the unemployed in Europe were unemployed for more than a year, and while that figure exceeded 60 percent for men and/or women in some member states, in others – most notably, Denmark, Austria, and Finland – it was substantially lower, around 30 percent. Employment rates within the EU varied as well. While the total number of persons employed declined in the 1990s by .4 percent, some of the member states – most notably, Luxembourg, Ireland, and the Netherlands, but also Spain – experienced large cumulative increases in the total number of persons employed after the mid-1980s (Table 1.4). And while the employment rate in the EU as a whole remained more or less constant over the past two decades, some of the member states – most notably, the Netherlands and Ireland but also Belgium, Spain, Portugal, and the United Kingdom – experienced significant increases that, at least in the case of the Netherlands, Portugal, and the United Kingdom resulted in employment rates substantially above the EU average (Table 1.5). Taken together, the various measures reported here suggest that some of the EU states – particularly, the Netherlands, Denmark, Ireland, and Portugal, but also Britain, Austria, and Luxembourg – were able to counter the long-term trends contributing to higher levels of unemploy30

Unemployment, Job Creation, and Union

ment, fewer persons employed, and smaller shares of the working-age population employed in the EU. Why was that the case? Why did the citizens of those states enjoy lower rates of unemployment and higher rates of job creation and employment than did the citizens of other EU member states? Political factors may have played a role in enabling some of the member states to reduce the rate of unemployment below the levels recorded in the late 1970s and early 1980s. There appears to be a systematic association between the partisan composition of a country’s government and the ability to effect a significant reduction in the rate of unemployment over the long term. In particular, it appears that most if not all of the states effecting significant reductions over the long term shared a common political attribute: prolonged control of government by conservative and center-right parties and/or coalitions in the 1980s and early 1990s followed by control by left-of-center parties and/or coalitions in the mid- to late 1990s. This is not to say, of course, that the Thatcher governments in Britain, or those headed by Lubbers in the Netherlands, or Schlüter in Denmark, or Cavaco Silva in Portugal, assigned a high priority to the reduction of unemployment; they were concerned, instead, with stabilizing prices and balancing budgets, and, indeed, the rate of unemployment rose dramatically in all of those countries during the period those governments were in office. Nevertheless, the reality of high unemployment, and the threat of still higher levels of unemployment – at a time when the government was controlled by parties that were not perceived as sympathetic to labor and was not only attempting to change labor market regulations and institutions but signaling its commitment to price stability and a tight money policy by linking the currency to the German mark in the EMS – caused the economic actors representing employees to give greater priority to the retention of existing jobs and the creation of new jobs. Those changes in institutions, policies, and behavior were important in initiating a shift in the rate of unemployment over the long term. But important as they were, it was only some time later, after the conservative parties had been removed from office and replaced by parties and coalitions of the center-left and the labor market-deregulating policies of the former had been complemented by the job-creating policies of the latter, that the rates of unemployment dropped to levels substantially below those experienced in the recent past. The magnitude of the decrease in the rate of unemployment and increase in employment in Ireland in the late 1980s and 1990s suggests an additional, economic reason why some member states have been more 31

David Cameron

successful than others in protecting and expanding the ranks of the employed. Over the past decade, Ireland experienced truly exceptional rates of economic growth – roughly 7 percent a year throughout the 1990s, and in excess of 10 percent a year, on average, in 1995–99 (Economic Commission 1998b: 236). The Irish case is, of course, something of an anomaly in the European context. Ireland’s truly exceptional rates of growth reflected, no doubt, the lagged effect of equally exceptional levels of investment ranging between 24 and 30 percent of GDP from the early 1970s to early 1980s. Those levels reflected, in large part, a quite exceptional degree of foreign direct investment as Ireland became a production and export base for non-European firms wishing to move into the European market.8 With such high rates of investment and growth sustained over a long period, it is hardly surprising that the level of unemployment was dragged down from 16 percent in 1985–88 and 15 percent in 1992–93 to the range of 6–8 percent in 1998 and 1999, and that the number of persons employed, and the employment rate, increased dramatically in the 1990s. Nevertheless, however anomalous, the Irish experience does testify to the very strong impact that sustained high rates of growth have on employment. What is true for Ireland regarding the close association between growth and employment is true more generally in the EU. There is a near-perfect association between the annual change in employment and the rate of economic growth in the EU, lagged two quarters, over the period 1975–97 (see European Commission 1998a: 29). Likewise, one finds a strong statistical relationship between rates of growth and levels of employment and unemployment across the member states. For example, there is a correlation of r = 0.80 between the rate of economic growth in 1995–99 and the average annual percent change in the number employed in 1995–99, indicating that the member states with the highest rates of growth experienced the highest average increases in employment. Conversely, there is a correlation of r = -0.68 between the rate of growth in 1995–99 and the increase in unemployment between 1991–96 and 1998–99, indicating that the member states with the highest rates of growth experienced the greatest decreases in unemployment in the 1990s. Ireland is not the only country in which high rates of unemployment dropped sharply in the wake of sustained high rates of growth. Through8

For a discussion of the Irish government’s active courtship of foreign direct investment, and its sectoral location, see IMF (1997b: 62–63).

32

Unemployment, Job Creation, and Union

out the 1980s and 1990s, Spain experienced exceptionally high levels of unemployment. Indeed, between 1982 and 1999, the rate of unemployment never dropped below 16 percent.9 Nevertheless, as in Ireland, a sustained high rate of growth in the mid- to late 1990s, although not nearly as high as that in Ireland, contributed to a substantial drop in the rate of unemployment. Thus, the rate dropped from more than 24 percent in 1994 to less than 14 percent in 2000, an even larger decrease than in Ireland.10 As important as Spain’s sustained high rate of growth was in driving down the rate of unemployment in the late 1990s, that reduction was also the result of significant changes in the existing institutions and rules pertaining to the labor market. In 1997, under the aegis of the new conservative government headed by José María Aznar, the employers confederation (CEOE) and the two labor confederations (the CC.OO. and the UGT) agreed to modify some of the rules governing the Spanish labor market.11 The unions agreed to a significant reduction in the amount of compensation that companies had to pay to permanent workers who were laid off, the employers accepted longer-term (up to four years and renewable) fixed-term work contracts, and both sides agreed to continue negotiating on expanding the use of part-time employment. In return, the government committed itself to a large-scale jobs program that involved increased spending on a variety of “active” measures such as training, employment, and vocational programs – programs that in the first three years after the 1997 accord contributed to the creation of more than one million new jobs (Financial Times, “Survey: Spain,” IV, June 8, 1998; June 12, 1998, 2; August 3, 1998, 13; and “Survey: Spain,” May 20, 1999, III). The decrease in the rate of unemployment in Spain in the late 1990s was unusually large. But the institutional means by which that reduction

9

10

11

See, e.g., Bentolila and Blanchard (1990); Blanchard et al. (1995); Boix (1997); Dolado and Jimeno (1997); Pérez (1997; 1999a); Perez-Diaz (1998). In 1994–96, Spain’s average annual rate of economic growth was between 2 and 3 percent. In 1997–99, its average annual rate of growth was between 3 and 4 percent. See European Commission (1998b: 82). On the reappearance in the 1990s of national-level bargaining in Spain, see Pérez (1999b). Pérez notes that the stage was set for the 1997 negotiations by a 1994 agreement between employers and the unions to shift some aspects of the regulation of labor markets to the collective bargaining process.

33

David Cameron

was initiated – national bargaining between employers and labor over issues pertaining to the regulation and flexibility of the labor market – was by no means unusual or exceptional. Indeed, as Martin Rhodes and others have noted, the existence of high and rising levels of unemployment in Europe in the 1980s and 1990s caused new patterns of bargaining to appear in a number of countries (see, e.g., Van Ruysseveldt and Visser 1996; Visser and Hemerijck 1997; Rhodes 1998; Streeck 1998; Casey 1999; Hemerijck and Schludi 1999; Pérez 1999b; Rhodes chap. 3 in this volume). While reminiscent in certain respects of earlier forms of “corporatist” bargaining – for example, in the negotiation of national “frame” agreements – in most of the countries the new arrangements were more decentralized and gave greater latitude to those bargaining at the level of sectors and firms (see, e.g., Cameron 1984; Calmfors and Driffill 1988). One result of the more decentralized, flexible, and somewhat ad hoc character of these new bargaining arrangements was a greater degree of variation across sectors and firms in wage settlements and a greater degree of autonomy from government. Nevertheless, the new bargaining patterns and arrangements did retain some features of the older, more explicitly “corporatist” forms of bargaining – for example, in involving governments in agreements with the economic bargaining partners that enabled them to reach agreements with each other. And in one important respect, the governments remained deeply involved, because, to a greater extent than was the case when the earlier forms of “corporatism” flourished in the pre-OPEC era, the essential quid pro quo for concessions in wage negotiations often was a government commitment to create jobs, through reductions in the work week or retirement age, or the easing of restrictions pertaining to part-time employment, or the creation of public sector jobs, or the implementation of targeted training and retraining programs. That changes in labor market and bargaining institutions have affected levels of unemployment is supported not only by the recent experience of Spain but by the experiences of several of the states that now enjoy unusually low levels of unemployment. Indeed, all of the smaller states that stood out among the EU member states at the end of the 1990s for their relatively low levels of unemployment – Ireland, the Netherlands, Denmark, Luxembourg, Austria, and Portugal – witnessed the appearance of new patterns of collective bargaining that, while in some instances drawing upon preexisting “corporatist” traditions, involved in all cases a greater degree of decentralization and flexibility, a lesser degree of formal institu34

Unemployment, Job Creation, and Union

tionalization, and a central focus on the negotiation of wage moderation in return for the creation of jobs.12 The success of Ireland and the Netherlands – the two member states which experienced the largest increases in the rate of employment after the mid-1980s – in creating new jobs at a time of high unemployment throughout Europe reflected in large part the impact of these new forms and patterns of collective bargaining over wages and work. Of course, both Ireland and the Netherlands have well-educated, skilled work forces and, for reasons pertaining to culture and location, have been especially attractive sites for foreign investment. In addition, they are, like most EU states, small open economies, and their high degree of trade dependence has sensitized economic actors to the trade-offs between wages and work.13 In both, and in the other states too, institutional mechanisms have been created through which the major economic actors – the labor confederations and the employers associations – have, with the informal support of their governments, negotiated agreements that, reminiscent of the earlier corporatist bargains, resulted in wage moderation in return for commitments by employers and governments to enact a variety of job-creating measures. Perhaps the best-known, indeed paradigmatic, instance of the newer, more decentralized form of bargaining between employers and labor occurred in the Netherlands (see in particular Visser and Hemerijck 1997; also Hemerijck and Schludi 1999; de Beus 1998; Van Ruysseveldt and Visser 1996; Wolinetz, chap. 8 in this volume). Acting in the face of unemployment rates that were already in double digits and were continuing to rise, the Dutch labor confederation (the FNV) agreed at Wassenaar in 1982 and in later negotiations to moderate its demands for wage increases and accepted government measures to delink public and private sector wages, de-index certain social benefits, and introduce stricter criteria for eligibility for disability and other social benefits. In return, labor received 12

13

We might note at this point the intriguing difference between the Portuguese experience and that of its Iberian neighbor in regard to the rate of unemployment in the 1980s and 1990s. For discussions of the contrast, see Bermeo (1994); Blanchard and Jimeno (1995); Blanchard and Portugal (1998); Glatzer (1998). In the mid-1990s, the ratio of exports to GDP was roughly 75 percent for Ireland and 55 percent for the Netherlands. Among the other member states, Luxembourg’s ratio exceeded 90 percent, that of Belgium exceeded 70 percent, and those of Denmark, Portugal, Austria, Finland, and Sweden were all in the range of 30–40 percent. See IMF (1997a); European Commission (1998a). For an early discussion of openness, see Cameron (1978).

35

David Cameron

a commitment from the employers association (the VNO-NCW) to preserve existing jobs and create new ones, accept a shorter work week, and institute more flexible work arrangements that would allow substantial part-time and temporary employment, and a commitment from the government to introduce new employment schemes that would subsidize employment of the unskilled and long-term unemployed (often referred to as “Melkert” jobs after the social affairs minister who introduced the schemes). (See Financial Times, September 23, 1997, 1; October 23, 1997, IV; Visser and Hemerijck 1997.) In the absence of investment ratios and foreign investment of the magnitude enjoyed by the Irish, one must conclude that it was, above all, the sustained wage restraint in the Netherlands from 1982 onward – reflected in unusually small increases in real compensation and unit labor costs – coupled with the widespread adoption of more flexible and part-time working arrangements that fueled the exceptional growth in employment – including part-time employment – that by the end of the 1990s had lowered the rate of unemployment to less than 4 percent.

Discussion: Unemployment in the Era of EMU On January 1, 1999, eleven member states of the EU jointly entered the third and final stage of Economic and Monetary Union. As they did so, they presided over economies that, for the most part, suffered from low rates of economic growth and high rates of unemployment, and in most the rates of unemployment showed little evidence or likelihood of dropping significantly in the foreseeable future. By its constituting documents, the European Central Bank, which is charged with sole responsibility for monetary policy in the euro-zone, is independent of political instruction and singularly committed to the objective of maintaining stable prices. In addition, the Stability and Growth Pact agreed by the member states in 1997 severely curtailed their ability to use fiscal policy to stimulate the economy.14 That being the case, to what extent, if at all, will EMU allow the participating member states to change the pattern of repeated, synchronized recessions, relatively low rates of economic growth, and high 14

The “pact” is designed to keep the budget deficits of the member states participating in the single currency below 3 percent of GDP. It stipulated a series of actions, including the imposition of sanctions, in the event a member state incurs a budget deficit in excess of 3 percent that was not “exceptional and temporary.” For the terms of the “pact,” see European Commission (1997b: 41–42, 49–53).

36

Unemployment, Job Creation, and Union

rates of unemployment that has characterized so much of Europe in recent decades? While it is obviously much too early to answer that question, we can speculate about the likely effects of EMU on unemployment in Europe (see Cameron 1997, 1998b).

Structural Effects of EMU There can be no doubt that EMU has already had, and will continue to have, far-reaching structural consequences for the economies of the participating member states. Especially for firms involved in financial services, but also for firms in a variety of other sectors as well, the irrevocable locking of the exchange rates of the participating currencies and eventual introduction of the euro will bring greater transparency in producer costs and enable consumers to compare prices more meaningfully and to be more fully informed in making choices among competing producers. That, in turn, is likely to lead to some downward pressure on prices, and price harmonization at the low end, which will presumably be of greatest benefit to the producers who can sell profitably at the lowest price throughout the euro-zone. Conversely, smaller, high-cost producers who have heretofore benefited by operating behind various forms of national protection, including the ability to price goods in national currencies, may be hurt by that downward pressure on prices. As important as those changes in the competitive position of firms will be, both for the firms themselves and consumers, the aggregate effect of those changes on levels of unemployment is likely to be mixed. The competitive advantages enjoyed by the low-cost producers should result in increases in employment. But any such gains may be offset, in part or in full, by losses of jobs in the high-cost firms that are at a competitive disadvantage in the more transparent euro-economy. No doubt competitive pressures on firms will be intensified by the greater transparency of costs and comparability of prices across the euro-zone. But it is not immediately apparent that this will have any significant effect on the overall rates of unemployment in the euro-zone and its participating member states. Beyond the employment-generating and employment-contracting firm-level effects of greater transparency, the advent of EMU and adoption of a single currency in most member states will undoubtedly transform certain sectors of the European economy. The financial services sector, in particular, is likely to be transformed – and has, indeed, already experienced dramatic change in such countries as France, Italy, and Spain 37

David Cameron

– as some banks and other financial institutions move to create economies of scale through mergers and acquisitions and other adaptations to the larger euro-market. The consolidation that has already begun and will surely continue in financial services will undoubtedly cause a good deal of unemployment within the sector in the short term. On the other hand, over the longer term that change may contribute to the process of job creation and a reduction in the current high levels of unemployment. As the sector experiences consolidation, and increasing concentration in a smaller number of larger institutions offering the full range of financial services throughout the EU, financial markets in Europe will inevitably become deeper, more liquid, and more highly integrated (see European Central Bank 1999). That increased liquidity and depth should make it less difficult for small and new firms to obtain capital – which is potentially of great consequence for employment since most new jobs in Europe (and elsewhere) are created in small and new firms. On balance, then, the structural changes generated by EMU may quite possibly contribute to a reduction in the level of unemployment in the euro-zone over the long term through their contribution to economic growth and, in particular, by providing the institutional prerequisites for the development of new and small firms. But at least in the short term, while the process of consolidation and integration is taking place, that process could actually cause local increases in unemployment.

Interest Rates and Investment While the advent of the third and final stage of EMU has already begun to trigger a series of changes at the firm and sectoral level in the European economy, it is by no means obvious that EMU will contribute (except indirectly and through those firm and sectoral changes) to the attainment of a higher rate of economic growth and a lower level of unemployment. It would, of course, be unreasonable to expect EMU to have that effect; after all, EMU was not designed and implemented in order to achieve a higher rate of growth and lower level of unemployment but, rather, for other, quite different purposes (see, e.g., Kenan 1992, 1995; Dyson 1994; Padoa-Schioppa 1994; Cameron 1995, 1997, 1998b; Eichengreen 1997; McNamara 1998; Dyson and Featherstone 1999). That has not stopped many from assuming that, in addition to its other effects, EMU would result in an increase in the rate of growth and a decrease 38

Unemployment, Job Creation, and Union

in the level of unemployment in Europe, and, indeed, the European Commission itself claimed as much in the comprehensive analysis of the economic effects of EMU it published in 1992 (see Emerson et al. 1992). The authors of the commission’s 1992 analysis, and many others, assume that eliminating exchange rate uncertainty through the creation of a single currency will reduce transaction costs within the single market and increase the risk-adjusted rate of return on investment, thereby stimulating higher levels of investment, which in turn will generate higher rates of economic growth and, as a result, lower rates of unemployment. Moreover, the existence of low rates of inflation in the member states participating in EMU, the improvement in their public finances, and the credibility of their commitment to maintain those policies for the foreseeable future will, it is assumed, allow interest rates to remain at lower levels than at present, thereby further stimulating investment, growth, and employment. There can be no doubt that the stimulation of higher levels of investment represents a necessary (if not sufficient) condition for raising the long-term rate of employment in Europe. Relatively high levels of investment over the long term tend, all else being equal, to generate relatively high rates of economic growth, and those high rates of growth, in turn, tend to drag down the level of unemployment.15 From that perspective, the long upward drift in the rate of unemployment over the past three decades in Europe can be viewed as the result of a long downward drift over that period in both the proportion of GDP that was invested and the long-term rate of economic growth.16 Perhaps, then, if the downward drift in investment over the past several decades were to be reversed, the rate of economic growth would increase and, in the wake of that increase, the rate of unemployment would decrease. As plausible as it may be to assume that EMU will contribute to higher levels of investment, and in turn higher rates of growth and employment, 15

16

Data published by the commission suggest a very strong inverse relationship over time in the EU between the ratio of investment to GDP and the level of unemployment, as well as a very strong positive relationship between annual change in gross fixed capital formation and growth, and a very strong positive relationship between the rate of growth and employment. See European Commission (1998b: 206–7). The proportion of GDP allocated to gross fixed capital formation in the fifteen current members of the EU dropped from 23.2 percent in 1961–70 to 22.8 percent in 1971–80, 20.1 percent in 1981–90, and 19.1 percent in 1991–97. See European Commission (1998b: 237, 255).

39

David Cameron

it is by no means certain that this will in fact happen. Among other things, despite the near-perfect association over time already noted between growth and job creation, it appears the two have become to some extent “decoupled,” meaning that growth can occur without the creation of new jobs (and even the loss of existing jobs), through increases in productivity obtained by substituting capital for labor. And even if growth and employment were not “decoupled,” it is not obvious that the elimination of transaction costs through the irrevocable locking of the exchange rates of the currencies, and their eventual replacement by the euro will actually generate any additional investment, just as it is not obvious that the existence of those transaction costs actually deterred any investment in the eleven countries of the euro-zone in the period prior to the locking of exchange rates. Moreover, while it is plausible to think that the low rates of inflation, small public deficits, and commitment of the EMU participants to those policies will enable them to enjoy lower interest rates than existed prior to the run-up to the third stage, it is important to note that for most of the EMU participants interest rates had already reached very low levels – in the range of 3.0–3.5 percent – in the mid-1990s. Of course, several EMU participants with traditionally high interest rates – most notably, Italy, Spain, Portugal, and Ireland – experienced significant reductions in the run-up to EMU. But since those reductions occurred in a series of steps in 1996–98, it is likely that much, if not all, of their stimulative effect had already been felt prior to the advent of EMU and in the first few months of 1999.17

Monetary Policy under the European Central Bank By Articles 105 and 107 and Protocol 3 of the Treaty on European Union, the European Central Bank (ECB) is singularly committed to the attainment and maintenance of price stability, free of all political instruction (see European Communities 1992). Even if that treaty-derived commitment to price stability did not exist, it would be reasonable to expect that, especially in its early years, when the credibility of its commitment to maintain the value of the new currency would depend, to a large extent, upon its behavior, the ECB would maintain rates sufficiently high to prevent 17

While the EMU-related reduction in interest rates contributed to high rates of growth and sharp drops in unemployment in Spain, Ireland, and Portugal in the late 1990s, they did not have that effect in Italy.

40

Unemployment, Job Creation, and Union

inflationary increases in the money supply and depreciation of the value of the euro – even at the cost of dampening the rate of economic growth and perpetuating high levels of unemployment. That would be all the more likely since the credibility of the euro as a strong currency and the ECB as the guarantor of its value will have to be established not only in international financial markets but also in the national political markets of the participating member states, some of whose publics view the imminent loss of their national currencies in early 2002 with skepticism (see European Commission 1999). While one would expect the ECB to follow the strictures of the treaty, paying close attention to changes in various measures of growth in money, credit, and prices, and pursuing policies designed to maintain stable prices even at the cost of continued low growth and high unemployment, an intriguing discrepancy appeared in the bank’s early existence between this conventional view of how it should and would act, and its actual behavior. In October 1998, for example, the ECB’s Governing Council decided that its monetary policy strategy would consist of achieving price stability – defined as an annual rate of change in the harmonized index of consumer prices of less than 2 percent – by comparing changes in a specific monetary aggregate (M3) with a quantitative reference value – initially, 4.5 percent – that was judged to be consistent with price stability.18 What is notable about this strategy is the latitude it provides the bank. By defining price stability in terms of a range of inflation from 0 to 2 percent, the bank can allow the current rate of inflation in the euro-zone to increase without being obliged to tighten policy, as long as the rate of inflation remains within the 0–2 percent range. On the other hand, by defining price stability in terms of a range that begins at 0, the ECB can raise interest rates even when the inflation rate is negligible. (Indeed, it did just that in November 1999, when it increased its benchmark rate by 0.5 percentage point, to 3 percent, despite the fact that the inflation rate was well within the target range.) And by targeting both the rate of inflation and a broad monetary aggregate (M3), the ECB gave itself some additional flexibility; for example, it can hold back on raising interest rates in the face of rising prices if the money supply is for some reason contracting, and, conversely, it can hold back on raising rates even if the money 18

The council’s decisions were described in its press releases following the meetings of October 13, 1998, and December 1, 1998, available at www.ecb.int.

41

David Cameron

supply is increasing at a rate that exceeds its target as long as prices are not increasing (a situation that existed during much of 1999).19 After adopting that strategy, the ECB pursued a policy that cannot be understood simply in terms of its implementation of its treaty-mandated objectives. In December 1998, it orchestrated a coordinated reduction in the interest rates of the eleven member states that were about to enter EMU, and in April 1999, it lowered its benchmark rate from 3.0 to 2.5 percent. The combined effect of the sharp reductions in interest rates that had occurred in Italy, Spain, Ireland, and Portugal in the run-up to EMU, the small reductions in rates that took place in the other participating member states prior to January 1, 1999, and the reduction in the benchmark rate in April resulted in a euro-zone interest rate during the first year of EMU that was well below a weighted average of the rates that prevailed in the eleven member states prior to the advent of EMU. And more to the point, after April 1999, the bank’s benchmark rate remained at or near the historic lows (in nominal terms) for rates for more than six months, despite increases in the M3 measure of money supply that exceeded the bank’s targeted growth rate, large increases in private sector credit, and a drop of more than 10 percent in the value of the euro in relation to the dollar. The apparent discrepancy that appeared in 1999 between the treatydefined objectives of the ECB and its actual conduct can perhaps best be understood as the inevitable disjunction between the treaty, on one hand, and the reality of a new institution which exists and operates in a complex political and economic context, on the other. For although the ECB is, by the terms of the Treaty on European Union, highly immunized from and formally independent of political direction, it nevertheless exists as one supranational institution among several, in an institutional and political context in which other interested actors are able to make their preferences in regard to monetary policy known – and in voices loud enough that even the well-insulated ECB sitting in its tower in Frankfurt can hear. As Lohmann (1998: 445) argued in regard to German monetary policy, “the specifics of how a central bank is embedded in a larger political system matter a great deal.”

19

Beyond the flexibility in policy, there is in addition some likelihood, with monetary policy set by the ECB, rather than the national central banks, while wage policy remains the subject of bargaining among economic actors at the national level, that the signaling effect of monetary policy in that bargaining will be dampened or distorted. On this possibility, see Hall and Franzese (1998) and Iversen (1998).

42

Unemployment, Job Creation, and Union

Among the most important of the interested actors whose voices will be heard, even by the well-insulated ECB, are the governments of the participating member states, meeting in the Euro-12, the informal forum consisting of the finance ministers of the member states participating in the euro-zone, as well as in the councils of the EU (e.g., the Ecofin council of ministers and the European Council of heads of state and government).20 As Oskar Lafontaine’s outspoken advocacy, beginning in the fall of 1998 after the German election, in support of a reduction in interest rates demonstrated, political leaders in the participating member states can and undoubtedly will continue to articulate their views and preferences in regard to monetary policy, interest rates, growth, unemployment, and other matters, irrespective of what the treaty says about the ECB’s independence from political direction, not only within the confines of EU institutions but also in public.21 Likewise, the European Parliament, with its powers and standing enhanced by the Treaty of Amsterdam and its ability to force the resignation of the Santer Commission in March 1999, and exercising an oversight role based on the periodic reporting to its Monetary Affairs Subcommittee by the ECB president, will undoubtedly continue to constitute an important venue for examining, and occasionally challenging, the performance of the ECB. As a result, however hawkish the ECB may sound about inflation, however much it drags its feet in moving rates down, and however great its propensity to raise rates (as evidenced by the several increases that occurred in late 1999 and 2000), it may in fact exercise its discretion in ways that do not always correspond to the conventional view that it is, and will always be, concerned only with maintaining stable prices. Most of the time it will undoubtedly behave in a manner consistent with the objectives specified for it in the treaty and not concern itself with the matter of unemployment. But occasionally – especially, perhaps, when the member states make their views known within the institutions of the EU – it may act in such a way as to give the economy a little stimulative boost. That may not be of much consolation to those who are unemployed and those who would 20

21

For a discussion of the stormy negotiations leading up to the formation of what is now the Euro-12, see Cameron (1997, 1998a, 1998b). The ECB’s coordinated reduction in rates to 3 percent in December 1998 muted Lafontaine’s calls for a reduction, but by February 1999, as growth estimates in Germany continued to drop, he renewed his campaign. Lafontaine abruptly resigned as finance minister in March; in its first meeting after his resignation, the Governing Council lowered the benchmark rate to 2.5 percent.

43

David Cameron

prefer a monetary policy that would promote growth and employment. But it is a bit more than they could reasonably expect from a bank that adhered rigidly to its treaty mandate.

EU-Level Employment Policy EMU and unemployment are inextricably linked. Whether or not they are causally linked is a matter of debate and speculation, but they certainly are politically intertwined. That was made obvious at the Amsterdam meeting of the European Council in June 1997 when, in order to obtain the support of the new Socialist-dominated French government for the Stability and Growth Pact, the council issued a resolution on growth and employment declaring it was “imperative to give a new impulse for keeping employment firmly at the top of the political agenda of the European Union [and] to strengthen the links between a successful and sustainable economic and monetary union, a well-functioning internal market and employment” (European Commission 1997b: 42–44). The council also reaffirmed the importance of employment and the need to reduce the high levels of unemployment, especially among the young, the long-term unemployed, and the low skilled, reiterated the need for a positive and coherent approach to job creation, noted with approval the inclusion in the treaty of a new title on employment, and announced that it would meet in the fall of 1997 to consider the issue of jobs and unemployment (see European Commission 1997b: 42–44). In November 1997, the European Council met in Luxembourg for its “jobs summit” and agreed on a modest package of proposals to address the issue of “structural” – that is, noncyclical – unemployment. The leaders adopted a joint action plan that included, among other things, the submission to the council of annual national action plans to reduce unemployment that would be reviewed by the commission and council. They agreed that all those under the age of 25 who were unemployed should be offered new employment within six months and all others who were unemployed should receive new jobs within a year,22 and that henceforth job retraining should be offered to at least 20 percent of the unemployed – twice as many as were then being offered such retraining. And they called upon the European Investment Bank to raise an extra Ecu 10 billion to finance small and medium-sized businesses, new technology ventures, and 22

All but Spain agreed these goals should be met within five years.

44

Unemployment, Job Creation, and Union

labor-intensive projects in the service sector, including health and education, and called upon the member states to simplify their regulatory rules for small businesses and to reduce taxes and nonwage costs associated with employment (see Financial Times, November 22–23, 1997: 2). However well-intended the Luxembourg “jobs summit” was, a one-day meeting of leaders and the adoption of various schemes for “national action plans” and consideration of “best practices” will not, by themselves, reduce the high levels of unemployment throughout the EU. Nevertheless, the “jobs summit” may at least have concentrated greater attention on the issue of how and why job creation occurs and, in so doing, called attention to a host of “supply-side,” microeconomic issues, involving unemployment benefits, tax levels and employer incentives, labor market regulations and rigidities, wage and nonwage costs, and job training and retraining, that, if addressed coherently from the perspective of increasing employment, could conceivably have a discernible effect in reducing the high level of structural unemployment, above and beyond whatever reduction occurs through the cyclical recovery.23 During the first half of 1999, the German government made the reduction of unemployment its chief priority during its presidency of the EU, and its desire to bring the national and European experiences together in an “employment pact” kept attention focused on the issue as EMU was launched (Financial Times, January 19, 1999: 3). Prospects for such a “pact” were, of course, dashed when Lafontaine abruptly quit as Germany’s finance minister midway through its presidency. But Germany did at least gain support at the Köln meeting of the European Council for its proposal that (taking a leaf from its own “alliance for jobs”) the EU undertake a tripartite macroeconomic dialogue with employers and labor aimed at creating jobs. Whether such talks will take place and prove to be effective in alleviating unemployment, and in that regard will constitute on the EU level an analogue to the newer forms of bargaining that have been witnessed in recent years in many of the member states, remains to be seen. In any case, the attention given to unemployment and employment by the European Council at Amsterdam, Luxembourg, Köln, and, most recently, at its special meeting on employment in Lisbon during the Portuguese

23

According to the ECB, unemployment in Europe is “overwhelmingly structural.” The IMF shares that view, estimating that approximately three-quarters of all unemployment in the euro-zone is structural. See Financial Times (October 2, 1997: 3); Economist (October 4, 1997: 54, 57); IMF (1999: 89).

45

David Cameron

presidency in the first half of 2000, is of great significance, for it reflects the extent to which EMU has not only raised the salience of unemployment but moved employment policy to the top of the EU’s agenda.24

Conclusion Over the past three decades, Europe has been afflicted with high rates of unemployment. With few exceptions, the member states of the EU suffer rates of unemployment that are considerably higher than those experienced two or three decades ago and considerably higher than those experienced by the United States and Japan. They also seem unable, again with a few notable exceptions, to create large numbers of new jobs, especially jobs for young workers, and to put back to work those who have lost jobs. And large geographically concentrated regions of unusually high unemployment persist. This chapter has examined the enduring problem of high unemployment in Europe. It considered how the rate of unemployment has varied across time and space in the European Union, and how it varies in regard to age, gender, and region. After describing those temporal and spatial variations in some detail, it considered why it was that, notwithstanding the general upward trend in unemployment throughout Europe in recent decades, some EU member states were able to create sufficient numbers of new jobs to resist, and even reverse, the upward trend. Among the factors that enabled Ireland, the Netherlands, Denmark, Portugal, and others, to create enough new jobs to reduce the rate of unemployment over the long term were, in addition to sustained high rates of economic growth which contributed to the reduction of unemployment in some countries, political-institutional features that influenced the collective behavior of economic actors with regard to wage-setting and working conditions and the propensity of governments to enact job-creating policies. Thus, unusually high levels of unemployment in the 1980s and 1990s in countries that were governed at the time by conservative parties or coalitions and are highly dependent upon trade, and in which economic actors are therefore especially sensitive to possible trade-offs between wages and

24

The Luxembourg and Köln summits are best viewed as two in a series of summits dating back to Brussels in December 1993 and continuing through Lisbon in March 2000, in which the EU’s political leaders elaborated a new and far-reaching European Employment Strategy. For a discussion of this “strategy” and its effects, see Cameron (2000).

46

Unemployment, Job Creation, and Union

work, appear to have induced those actors to participate in new forms and patterns of collective bargaining that resulted in agreements and policies that, when sustained and extended by subsequent left-of-center governments, significantly reduced the rate of unemployment over the long term. In concluding, the chapter considered several aspects of Economic and Monetary Union that may affect the rate of unemployment in the eurozone as a whole and in the participating member states in the foreseeable future. The structural changes generated by EMU, which may indeed be quite significant, may have both employment-creating and employmentdestroying effects, probably the latter before the former. Contrary to the hopes of its institutional proponents, EMU may not result in any significant increase in levels of investment and growth, and employment, over the long term, apart from those that derive indirectly from the structural changes. On the other hand, because the ECB is embedded in a complex and densely articulated web of national and supranational institutions, it may on occasion act in ways that are more nuanced, and more sensitive to issues of growth and employment, than one would predict on the basis of its constituting documents. In the wake of EMU, the issues of employment, unemployment, and macroeconomic policy are no longer the exclusive concern of national governments but have moved to the top of the policy agenda of the European Union. Whether that will result in a coordinated national and supranational policy effort that significantly reduces the high rates of unemployment that afflict so much of Europe remains to be seen. References Alogoskoufis, George, Charles Bean, Giuseppe Bertola, Daniel Cohen, Juan Dolado, and Gilles Saint-Paul. 1995. Unemployment: Choices for Europe. London: Centre for Economic Policy Research. Bean, Charles R. 1994. European Unemployment: A Survey. Journal of Economic Literature 32: 573–619. Bean, Charles R., Richard Layard, and Stephen J. Nickell, eds. 1987. The Rise in Unemployment. New York: Basil Blackwell. Benoit-Guilbot, Odile, and Duncan Gallie. 1994. Long-Term Unemployment. London: Pinter. Bentolila, Samuel, and Olivier J. Blanchard. 1990. Spanish Unemployment. Economic Policy 10: 233–81. Bermeo, Nancy. 1994. Comments. In John Williamson, ed., The Political Economy of Policy Reform. Washington, D.C.: Institute for International Economics. Blanchard, Olivier J., and Jean-Paul Fitoussi. 1998. Croissance et chômage. Paris: Documentation française. 47

David Cameron Blanchard, Olivier J., and Juan F. Jimeno. 1995. Structural Unemployment: Spain versus Portugal. American Economic Review 85: 212–18. Blanchard, Olivier J., Juan Francisco Jimeno, Javier Andres, Charles Bean, Edmond Malinvaud, Ana Revenga, Gilles Saint-Paul, Dennis J. Snower, Robert Solow, David Taguas, and Luis Toharia. 1995. Spanish Unemployment: Is There a Solution? London: Centre for Economic Policy Research. Blanchard, Olivier J., and Pedro Portugal. 1998. What Hides behind an Unemployment Rate: Comparing Portuguese and U.S. Unemployment. Working Paper 6636. Cambridge, Mass.: National Bureau of Economic Research. Blanchard, Olivier J., and Lawrence H. Summers. 1987. Hysteresis in Unemployment. European Economic Review 31: 288–95. Boix, Carles. 1997. Spain’s Economy in the Last Forty Years: Between Growth and Unemployment. South European Society & Politics 2: 139–45. Calmfors, Lars, and John Driffill. 1988. Bargaining Structure, Corporatism and Macroeconomic Performance. Economic Policy 6: 13–61. Cameron, David R. 1978. The Expansion of the Public Economy: A Comparative Analysis. American Political Science Review 78: 1243–61. ———. 1984. Social Democracy, Corporatism, Labour Quiescence, and the Representation of Economic Interest in Advanced Capitalist Society. In John H. Goldthorpe, ed., Order and Conflict in Contemporary Capitalism. Oxford: Oxford University Press. ———. 1995. Transnational Relations and the Development of European Economic and Monetary Union. In Thomas Risse-Kappen, ed., Bringing Transnational Relations Back In: Non-State Actors, Domestic Structures and International Institutions. Cambridge: Cambridge University Press. ———. 1996. Exchange Rate Politics in France, 1981–83: The Regime-Defining Choices of the Mitterrand Presidency. In Anthony Daley, ed., The Mitterrand Era: Policy Alternatives and Political Mobilization in France. London and New York: Macmillan and New York University Press. ———. 1997. Economic and Monetary Union: Underlying Imperatives and Third-Stage Dilemmas. Journal of European Public Policy 4: 455–85. ———. 1998a. EMU after 1999: The Implications and Dilemmas of the Third Stage. Columbia Journal of European Law 4: 425–46. ———. 1998b. Creating Supranational Authority in Monetary and Exchange Rate Policy: The Sources and Effects of EMU. In Wayne Sandholtz and Alec Stone Sweet, eds., European Integration and Supranational Governance. Oxford: Oxford University Press. ———. 2000. The Europeanization of Employment Policy. Paper presented at the Max Planck Institute for the Study of Societies. Köln, May 30. Casey, Bernard. 1999. The Impact of Social Partnership and Protection on Economic Performance. Paper presented at the Conference on Globalization, European Economic Integration and Social Protection. Florence: European University Institute, March 11–12. de Beus, Jos. 1998. The Relationship between the European Model of Society and the Dutch Model of Consensus Society in the EMU. Paper presented at the

48

Unemployment, Job Creation, and Union Workshop on Economic and Monetary Union and the European Model of Society. Waltham, Mass.: Brandeis University, December 11–12. Dolado, Juan J., and Juan F. Jimeno. 1997. The Causes of Spanish Unemployment: A Structural VAR Approach. European Economic Review 41: 1281–307. Drèze, Jacques, and Edmond Malinvaud. 1995. Growth and Employment: The Scope of a European Initiative. European Economic Review 38: 489–504. Dyson, Kenneth. 1994. Elusive Union: The Process of Economic and Monetary Union in Europe. London: Longman. Dyson, Kenneth, and Kevin Featherstone. 1999. The Road to Maastricht: Negotiating Economic and Monetary Union. Oxford: Oxford University Press. Eichengreen, Barry. 1997. European Monetary Unification: Theory, Practice and Analysis. Cambridge, Mass.: MIT. Emerson, Michael, Daniel Gros, Jean Pisani-Ferry, Alexander Italianer, and Horst Reichenbach. 1992. One Market, One Money: An Evaluation of the Potential Benefits and Costs of Forming an Economic and Monetary Union. Oxford: Oxford University Press. European Central Bank. 1999. Possible Effects of EMU on the EU Banking Systems in the Medium to Long Term. Frankfurt: ECB. European Commission. 1997a. Employment in Europe 1997. Luxembourg: European Communities. ———. 1997b. European Economy: 1997 Broad Economic Policy Guidelines, No. 64. Luxembourg: European Communities. ———. 1998a. Employment in Europe 1998. Luxembourg: European Communities. ———. 1998b. European Economy: 1998 Broad Economic Policy Guidelines, No. 66. Luxembourg: European Communities. ———. 1999a. Eurobarometer. Report # 51. Luxembourg: European Communities. ———. 1999b. European Economy: 1999 Broad Economic Policy Guidelines, No. 68. Luxembourg: European Communities. ———. 2000. European Economy: 2000 Broad Economic Guidelines, No. 70. Luxembourg: European Communities. European Communities. 1992. Treaty on European Union. Luxembourg: European Communities. Glatzer, Miguel. 1998. Portugal and Spain: Social and Labor Market Policy Reform under the Shadow of EMU. Paper presented at the Workshop on EMU and the European Model of Society. Waltham, Mass.: Brandeis University, December 11–12. Hall, Peter A., and Robert J. Franzese, Jr. 1998. Mixed Signals: Central Bank Independence, Coordinated Wage Bargaining, and European Monetary Union. International Organization 52: 505–35. Hemerijck, Anton, and Martin Schludi. 1999. Policy Adjustment under International Constraints: Sequences of Challenges and Responses. Paper presented at the Conference on Globalization, European Economic Integration and Social Protection. Florence: European University Institute, March 11–12. Henry, S. G. B., and Dennis J. Snower, eds. 1996. Economic Policies and Unemployment Dynamics in Europe. Washington, D.C.: IMF.

49

David Cameron International Monetary Fund (IMF). 1997a. Direction of Trade Statistics Yearbook 1997. Washington, D.C.: IMF. ———. 1997b. World Economic Outlook, May 1997. Washington, D.C.: IMF. ———. 1999. World Economic Outlook, May 1999. Washington, D.C.: IMF. Iversen, Torben. 1998. Wage Bargaining, Central Bank Independence, and the Real Effects of Money. International Organization 52: 469–504. Iversen, Torben, and Anne Wren. 1998. Equality, Employment, and Budgetary Restraint: The Trilemma of the Service Economy. World Politics 50: 507–46. Kenen, Peter B. 1992. EMU after Maastricht. Washington, D.C.: Group of Thirty. ———. 1995. Economic and Monetary Union in Europe: Moving beyond Maastricht. Cambridge: Cambridge University Press. Layard, Richard, Stephen Nickell, and Richard Jackman. 1991. Unemployment: Macroeconomic Performance and the Labour Market. New York: Oxford University Press. ———. 1994. The Unemployment Crisis. New York: Oxford University Press. Lindbeck, Assar. 1993. Unemployment and Inflation. Cambridge, Mass.: MIT. Lindbeck, Assar, and Dennis J. Snower. 1988. The Insider-Outsider Theory of Employment and Unemployment. Cambridge, Mass.: MIT Press. Lohmann, Suzanne. 1998. Federalism and Central Bank Independence: The Politics of German Monetary Policy, 1957–1992. World Politics 50: 401–46. McNamara, Kathleen. 1998. The Currency of Ideas: Monetary Politics in the European Union. Ithaca: Cornell University Press. Nickell, Stephen. 1997. Unemployment and Labor Market Rigidities: Europe versus North America. Journal of Economic Perspectives 11: 55–74. ———. 1999. Labour Market Institutions and Economic Performance. Paper presented at the Conference on Globalization, European Economic Integration and Social Protection. Florence: European University Institute, March 11–12. Organisation for Economic Co-Operation and Development (OECD). 1994. OECD Jobs Study: Investment, Productivity and Employment. Paris: OECD. ———. 1997. Implementing the OECD Jobs Strategy: Member Countries Experience. Paris: OECD. ———. 1999. OECD Employment Outlook, June 1999. Paris: OECD. Padoa-Schioppa, Tommaso. 1994. The Road to Monetary Union in Europe: The Emperor, the Kings, and the Genies. Oxford: Oxford University Press. Pérez, Sofía A. 1997. Banking on Privilege: The Politics of Spanish Financial Reform. Ithaca: Cornell University Press. ———. 1999a. From Labor to Finance: Understanding the Failure of Socialist Economic Policies in Spain. Comparative Political Studies 32: 659–89. ———. 1999b. From De-Centralization to Re-Organization: Explaining the Return to National-Level Bargaining in Italy and Spain. Manuscript. Perez-Diaz, Victor. 1998. The “Soft Side” of Employment Policy: The Spanish Experience. West European Politics 21: 200–23. Rhodes, Martin. 1998. Globalization, Labour Markets, and Welfare States. In Rhodes and Yves Mény, eds., The Future of European Welfare: A New Social Contract? London. Macmillan. 50

Unemployment, Job Creation, and Union Rubery, Jill, Mark Smith, Colette Fagan, and Damian Grimshaw. 1998. Women and European Unemployment. London: Routledge. Saint-Paul, Gilles. 1995. Some Political Aspects of Unemployment. European Economic Review 39: 575–82. ———. 1996. Exploring the Political Economy of Labour Market Institutions. Economic Policy 23: 265–315. Shonfield, Andrew. 1969. Modern Capitalism. Oxford: Oxford University Press. Snower, Dennis J., and Guillermo de la Dehesa, eds. 1997. Unemployment Policy: Government Options for the Labour Market. Cambridge: Cambridge University Press. Statistical Office of the European Communities. 1999. Eurostatistics: Data for ShortTerm Economic Analysis, 7-1999. Luxembourg: European Communities. Streeck, Wolfgang. 1998. The Internationalization of Industrial Relations in Europe: Problems and Prospects. Politics and Society 26: 429–59. Summers, Lawrence H. 1990. Understanding Unemployment. Cambridge, Mass.: MIT Press. Therborn, Göran. 1986. Why Some Peoples are More Unemployed Than Others. London: Verso. Van Ruysseveldt, Joris, and Jelle Visser, eds. 1996. Industrial Relations in Europe: Traditions and Transitions. London: Sage. Visser, Jelle, and Anton Hemerijck. 1997. “A Dutch Miracle”: Job Growth, Welfare Reform and Corporatism in the Netherlands. Amsterdam: Amsterdam University Press.

51

2 Organized Market Economies and Unemployment in Europe IS IT FINALLY TIME TO ACCEPT LIBERAL ORTHODOXY?

Peter A. Hall

An old specter is haunting Europe: the specter of liberal orthodoxy.1 I refer to the view that the only way for a nation to secure high levels of economic growth and employment is to develop an economy built around perfectly competitive markets, an ideal type that implies weak trade unions, substantially deregulated financial markets, and interfirm relations based on highly competitive relationships mediated by legal contracts rather than long-term collaborative arrangements. Of course, this is an Anglo-American orthodoxy, developed since the eighteenth century by British and American economists, whose ideals have been implemented most extensively and with great success in the economies of the United States and the United Kingdom. Whenever the continental European economies have experienced economic problems, the Anglo-American model has been urged on them as a solution. It inspired much of the advice they received about postwar reconstruction and was revived again during the 1970s to indicate how Europe should respond to “stagflation” (OECD 1977; Olson 1982; Ruggie 1982; Keohane 1984; Milward 1984). Its policy prescriptions call for the deregulation of markets in labor and capital, retrenchment in social policy, and reductions in the state’s role in the economy. 1

This chapter draws on many conversations and some joint writing with David Soskice, from which it benefits immensely. For comments on a previous version presented to the 1998 Conference of Europeanists, I am grateful to Chris Allen, Suzanne Berger, Nancy Bermeo, Andreas Busch, Orfeo Fioretis, Torben Iversen, Richard Locke, Kathleen McNamara, John Stephens, and William Wallace.

52

Organized Market Economies and Unemployment

However, such prescriptions pose serious dilemmas for most European nations. Many have powerful trade union movements that resist efforts to deregulate labor markets (Calmfors and Driffill 1988). European value systems reflected in the strength of Social Democratic and Christian Democratic parties are inimical to the allocation of resources entirely by competitive markets (Esping-Andersen 1990; Polanyi 1944; Van Kersbergen 1995). Many European firms have developed strategies whose effectiveness depends on the nonmarket coordination that high levels of regulation make possible (Soskice 1991, 1994b). Accordingly, European nations have historically resisted wholesale movement toward an AngloAmerican model of capitalism (Wood 1997; Thelen 2000). This is not surprising. More surprising is that these nations have had successful records of economic performance, marked by rates of growth, productivity increase, and unemployment that have often been superior to those of the American or British economies (see Table 2.1). From the efforts of political economists to explain how nations that do not embrace economic liberalism still secure economic performance has emerged an important literature on comparative capitalism (see Hall 1999). One of its central contentions is that there are at least two viable ways of organizing a capitalist economy.2 I will refer to one type as that of a “liberal market economy” (LME) and the other as an “organized market economy” (OME) (Soskice 1991). Liberal market economies instantiate the ideals associated with perfect competition. Images of an organized market economy have evolved over time. In the 1970s, it was associated with the concept of “neocorporatism” built around coordinated wage bargaining (Lehmbruch and Schmitter 1979; Cameron 1984; Goldthorpe 1984; Calmfors and Driffill 1988; Alvarez, Garrett, and Lange 1991) and later defined in terms of distinctive financial market arrangements, employer associations, and production strategies (Zysman 1983; Soskice 1991; Streeck 1992). The key point is that, as late as the 1980s, many Northern European nations still secured satisfactory levels of economic performance by maintaining organized market economies. In recent years, however, high levels of unemployment have called this observation into question. In Sweden, rates of unemployment that averaged only 2 percent from 1964 to 1979 tripled during the 1990s, and 2

To speak of such a dualism is a slight simplification, since virtually all analysts acknowledge some variation within these types, and some emphasize additional categories of variation (see Shonfield 1969; Zysman 1984; Katzenstein 1985; Scharpf 1990; Iversen 1994).

53

Peter A. Hall

Table 2.1. Standard and Equilibrium Rates of Unemployment, 1964–1999 ( percentage)

Germany saw unemployment rise from 3 percent in the 1970s to 12 percent in the 1990s.3 This experience has revived liberal orthodoxy: a chorus of commentators now argue that the European nations must deregulate financial and labor markets, cut back social programs, and reduce 3

I focus here on “unemployment” rather than on “employment,” even though measures of the former are highly sensitive to official definitions and cross-national comparisons must be made with caution, because unemployment is the more intense social problem of high political relevance.

54

Organized Market Economies and Unemployment

government intervention in order to resolve their unemployment problems.4 The creation of fifty million jobs in the United States since 1970 when the European Union created only ten million has lent force to such arguments. In a new context, comparative political economy faces an old question: can the European nations secure acceptable employment performance again without abandoning the institutions of an organized market economy? Is there still a viable alternative to market liberalism? This essay proposes some tentative answers to these questions – tentative because full exploration of the issues would demand more space and evidence than this chapter can provide. I seek simply to provide a framework for approaching these issues, one distinguished by its analytical stance. Most investigations assume the problem is one of establishing whether the European economies can become more like the American. By contrast, I argue that there are at least two models on which effective economic performance has been based and investigate the problems and potential specific to the model that is often neglected, namely, the organized market economies of Europe. Only if we consider the distinctive features of such economies, can we arrive at a precise diagnosis of the challenges they face and their potential for coping with them. The next section provides an overview of the two models of the political economy. The third provides an account of how the organized market economies were once able to secure low levels of unemployment. In the fourth section, I argue that changes in context have undermined the feasibility of the techniques used in the past. The fifth section explores the capacity of these economies to surmount the new challenges they face, and the sixth explores the institutional strains generated by these challenges. Although Europe contains a variety of economic models, I concentrate here on nations that can be called organized market economies, such as Germany, Sweden, Austria, Norway, the Netherlands, Denmark, and Finland, where the dilemmas are most acute (cf. Soskice 1998).

Two Models of the Political Economy I begin by reviewing the principal differences between liberal and organized market economies, drawing on the models developed by many scholars, with an emphasis on the recent formulations of Soskice (1991, 4

This view is widely reflected in the press (Financial Times, October 27, 1997: 14) and in the publications of international organizations (OECD Jobs Study 1995).

55

Peter A. Hall

1998; Hall and Soskice forthcoming; see also Zysman 1983; Goldthorpe 1984; Katzenstein 1985; Hall 1986; Sharpf 1990).5 These see the firm as the central actor in the economy, responsible for the key decisions that ultimately aggregate into overall levels of economic performance. Each firm faces a variety of coordination problems associated with the recruitment, compensation, and training of labor, securing access to finance and technology, and managing relations with suppliers, clients, and employees. Thus, the principal challenges facing the firm are relational, and its core competencies turn on the strategies devised to manage these relations (see Milgrom and Roberts 1992; Aoki 1994). The defining feature of a liberal market economy is the degree to which its firms rely on the price signals arising from competitive markets and formal contractual relations to resolve such coordination problems. Here, firms typically secure finance via arm’s-length relations in which equity markets play a large role and bank lending turns heavily on cash flow or collateral. The access of most firms to finance depends heavily on their short-term profitability. Labor is recruited from relatively unconstrained labor markets where trade unions are often weak. Wage determination lies primarily at the firm or plant level, and management retains substantial prerogatives over firing. Access to production inputs and technology is generally secured by competitive bidding among suppliers, licensing, or hiring technical personnel from fluid labor markets. Vocational training is the responsibility of the worker or individual firm and emphasizes general skills. By contrast, an organized market economy is defined by the extensive degree to which it relies on institutions other than market mechanisms to resolve the coordination problems facing firms, including encompassing trade unions, works councils, employers associations, cross-shareholding, and other linkages between firms. These institutions allow firms to engage in high levels of “nonmarket coordination” with each other and their employees that have the character of strategic interactions, by providing facilities that allow the actors to exchange information in a context of credibility, to monitor the behavior of others more closely than market relations allow, to sanction deviation from agreed courses of action, and to deliberate about changes to such arrangements. Thus, corporate finance is usually provided on terms that are more sensitive to a firm’s long-term 5

For a more extended discussion of the differences between these types of political economies, see Hall and Soskice (forthcoming).

56

Organized Market Economies and Unemployment

strategy than its short-term profitability by banks or other firms with extensive capacities for network monitoring of firm behavior. Equity markets featuring high levels of cross-shareholding inhibit hostile takeovers and facilitate the exchange of inside information among firms. Labor is recruited on markets that are dominated by powerful trade unions or employer associations often accompanied by institutions that limit managerial prerogatives over layoffs and work reorganization. Firms typically secure access to inputs and technology via long-term collaborative arrangements whose viability depends on collective capacities for standard-setting and the enforcement of implicit contracts. Vocational training is often accomplished through collaborative schemes that encourage the development of industry-specific skills. Although this is a schematic account identifying two ideal types, it allows us to identify some of the distinctive strengths and weaknesses of each economy (see Hall and Soskice forthcoming). The principal strengths of liberal market economies flow from the capacities they vest in corporate managers to redeploy resources quickly, both within the firm and across sectors. Labor can be recruited, laid off, or redirected with relative ease. Well-developed equity markets push capital quickly from spheres offering low returns to those promising higher ones. As a result, these economies tend to be propitious sites for radical innovation of the sort that involves very new technologies or entirely new product lines and high-risk investments (Soskice 1994a). Firms in such economies also tend to be effective at controlling labor costs. The central weaknesses of liberal market economies flow from the low levels of support they provide for the long-term relationships associated with incomplete contracting among firms or between firms and their employees (see Milgrom and Roberts 1992), and the absence of such arrangements in some spheres inhibits their development in others. For instance, the salience of short-term profitability to finance limits the capacity of firms to offer long-term employment contracts, which may reduce the willingness of employees to share their private knowledge or to acquire firm- or industry-specific skills. Accordingly, companies in liberal market economies often find it difficult to raise some kinds of skill levels, to sustain new projects through a downturn in demand, or to develop the kind of long-term relations with customers and suppliers that facilitate incremental innovation in products and production processes. The strengths and liabilities of organized market economies are virtually the mirror-image of those in liberal market economies. Such 57

Peter A. Hall

economies draw their strengths from institutional infrastructures that support long-term relationships alongside market relations. Thus, firms enjoy superior capacities for vocational training and competitive advantages in sectors that demand highly skilled labor (Finegold and Soskice 1988). The availability of long-term finance allows firms to develop “implicit contracts” with employees that encourage them to share knowledge with management (Aoki 1994). Thus, quality control is easier, and close relations with suppliers facilitate continuous innovation. Organized producer groups provide mechanisms for aligning wage growth with productivity growth that do not depend on periodic deflations (cf. Soskice 1990; Golden 1993). The weaknesses of organized market economies flow largely from the slow speed with which they redeploy resources. Financial systems that rely heavily on network reputations and discourage hostile takeovers can be slow to shift funds from areas of low return to those offering higher ones. Where layoffs are more difficult, it is harder for firms to cut costs and riskier for them to invest in new sectors or technologies where failure might leave them holding labor they cannot shed. Although long-term relationships inside and among firms offer gains from reorganization that might not otherwise be possible, realizing those gains may be a slow process demanding the renegotiation of skill categories, task assignments, and resource allocation with multiple actors. This portrait implies that national economies derive comparative institutional advantages from the coordinating capacities embedded in national institutional infrastructure, and there is considerable evidence that these give rise to national patterns of corporate strategy and economic performance (Knetter 1989; Burgess and Knetter 1996; Casper 1998; Lehrer 1997b).

Explaining Employment Performance in Organized Market Economies To understand whether organized market economies can reduce the high levels of unemployment they have recently experienced, we need to explore how they coped with unemployment problems in the past. Why were they able to secure relatively low levels of unemployment for so many years? It should be obvious that the answer to this question cannot be drawn entirely from liberal orthodoxy. Many organized market economies have 58

Organized Market Economies and Unemployment

long had powerful trade union movements and generous social benefits that militate against the kind of labor-market adjustment on which liberal economies rely. We need an alternative explanation for their economic success. It can be found in the presence and operation in these economies of three relatively integrated structural capacities: the capacity to restrain wages via the coordination of wage bargaining, the capacity to secure continuing productivity gains via incremental improvements in product and production processes, and the capacity to mop up “excess” labor via distinctive social policy regimes. Each capacity is structural in that it derives from the institutional infrastructure of the political economy, and together they have been central to the employment strategies of Northern Europe. Although liberal orthodoxy normally associates powerful trade unions with labor-market rigidities, a large literature shows how powerful trade unions and employers associations of the sort commonly found in organized market economies can generate the coordinating capacity to produce nominal and real wage moderation. Powerful trade unions seem to be a drawback only in the absence of institutional structures for the coordination of wage bargaining (cf. Przeworski and Wallerstein 1982; Calmfors and Driffill 1988; Scharpf 1990; Soskice 1990; Alvarez et al. 1991; Golden 1993; Dore, Boyer, and Mars 1994; Hall and Franzese 1998; Iversen 1998b). In such systems, however, wage restraint is sustainable over the long term only if it is leavened by some gains in real wages, and these are possible without a decline in international competitiveness and corresponding unemployment only if the nation can generate sustained increases in productivity. Accordingly, the structural capacity for continuous productivity gains present in these economies is a crucial complement to systems of coordinated wage bargaining. Powerful trade unions and works councils militate against the use of layoffs to secure such gains, but institutional structures that provide long employment tenures and active worker participation in the management of the production process enhance the ability of firms to exploit the private knowledge of the workforce and secure continuous improvements in product and production processes (cf. Milgrom and Roberts 1992; Aoki 1994; Soskice 1994a). Industry-wide trade unions and employer associations are also conducive to the development of training systems that give firms access to highly skilled labor and the levels of productivity it offers (Culpepper forthcoming). In short, the interaction of these two structural features have allowed many organized market economies to maintain relatively low rates of 59

Peter A. Hall

unemployment in the context of substantial real wage increases offset by productivity growth. We can think of this dynamic as the historic “employment machine” of the organized market economies. It is reflected in rates of increase in labor productivity that have been consistently higher than in liberal market economies, maintaining their competitiveness despite the higher rates of real wage growth there that liberal critics often emphasize.6 However, this employment machine has a drawback: it encourages firms to respond to rising real wages by economizing on labor. To some extent, this is offset by structural capacities for enhancing productivity but, over time, real wage increases will bias firms toward more capital-intensive forms of production that militate against job growth (Berthold, Fehn, and Thode 1999). Thus, although these economies provide a core labor force with high levels of job security and wages, they are often not as effective as liberal market economies at expanding employment. They face the potential for excess supply in the labor market. Accordingly, the third leg of the triad behind the operation of these economies has been a set of social policy regimes designed to limit the size of the active labor force or to mop up excess labor there. Among the organized market economies of Europe, two kinds of social policy regimes address these problems differently (Esping-Andersen 1990, 1999). One regime, characteristic of the Nordic economies, uses the expansion of public sector employment, notably in services such as day care, health care, and education, to absorb excess labor. Here, social democratic parties that attach high value to access to work and hold few reservations about high levels of government spending have expanded both the public sector and the workforce (Wren 2000). Generous maternity benefits, day-care provision, and regulations governing absenteeism have drawn women into the workforce and increased employment in public services. The other approach, typical of what Esping-Andersen (1990) calls the “conservative” welfare states of Germany, Austria, and the Netherlands, uses generous early retirement programs and maternity or disability 6

In the OMEs listed in Table 2.1, labor productivity increased by 4.8 percent from 1960 to 1973, by 1.9 percent from 1973 to 1979, and by 2.1 percent from 1979 to 1993, compared with an annual average of 3.1, 1.3, and 1.8 percent in LMEs over these periods; and unit labor shares averaged 64 percent in both kinds of economies in 1993 (OECD 1996a,b; Tronti 1997). Annual increases in the real hourly wage in manufacturing averaged 4.4 percent from 1960 to 1973, 1.3 percent from 1974 to 1982, and 1.6 percent from 1983 to 1990 in these OMEs compared with increases of 2.9 percent, 1.1 percent and 0, respectively, in the LMEs.

60

Organized Market Economies and Unemployment

benefits to encourage many to exit or remain outside the labor force, thereby reducing the total numbers seeking work and classified as unemployed. Christian Democratic parties that embrace the traditional family structure and eschew a large public sector have been influential proponents of this approach. The differences between the two regimes show up in total employment, which tends to cover 70–75 percent of the adult population in the social democratic regimes and only 50–65 percent in nations with conservative social policy regimes. In different ways, however, these two social policy regimes have suppressed unemployment in the organized market economies for some years (Mares 1997; Ebbinghaus and Manow 1998; Huber and Stephens 1999; Scharpf and Schmidt 2000).

Changes in the Context for Successful Employment Performance If we could simply extrapolate this analysis of the past into the future, the outlook for the organized market economies of Europe would be good. These economies have had distinctive capacities for maintaining low levels of unemployment. However, the situation is neither so simple nor so promising. The successful operation of this “employment machine” has depended on the presence of several contextual features that may no longer obtain in (1) the production regime, (2) the social regime, and (3) the macroeconomic policy regime.7

The Production Regime The employment machine of the OMEs was especially feasible when employment was concentrated in the industrial sector because, as Iversen and Wren (1996; cf. Schettkatt 1992) point out, industrial production takes a form that has long provided greater scope for rapid productivity gains than the agricultural or service sectors. Indeed, the industrial sector has been notably large in many organized market economies (see Table 2.2). In recent years, however, the size of the industrial sector has been declining, and the service sector now accounts for most of the growth in

7

Although there is an increasing amount of scholarly work pertinent to these issues, I want to note that my own thinking about them has been especially influenced by the analyses of my colleague, Torben Iversen, whose work points toward precisely this kind of problematic.

61

Peter A. Hall

Table 2.2. Percentage of Employment in the Agricultural, Industrial, and Service Sectors, 1960–1994

economic activity and employment in the developed world (EspingAndersen 1996; Iversen and Wren 1996). This is significant because many analysts suggest that the rate of increase of productivity is lower in the service sector than in the industrial sector. Although we should be cautious about taking this as a fact that can be projected into the future because productivity in services is difficult to measure and new technologies may improve it considerably, the slow rate of productivity growth in services must be taken seriously. Two important implications follow. First, if the European economies are to create a significant number of new jobs, they will have to do so in services. Second, if productivity increases in services cannot match high 62

Organized Market Economies and Unemployment

and rising real wages, the organized market economies may not be wellequipped to create jobs there. Their employment strategies may have been more suitable for an industrial than a postindustrial economy. This shift in the sectoral structure of the economy may confer relative advantages on liberal market economies instead, whose institutions provide more support for low-wage labor and for corporate strategies that do not depend so heavily on continuous increases in productivity.

The Social Regime The social policy regimes that once underpinned the employment strategies of the organized market economies are also coming under pressure from various directions. The conservative regimes are challenged by an aging population that is reducing the numbers of workers available to support those who are retired. As a result, these welfare states are facing a fiscal crisis that could cause benefits to fall by half or social charges to double to support the number of pensioners expected in 2020. This prospect calls into question strategies that have encouraged exit from the labor force or early retirement to reduce levels of unemployment (see Scharpf and Schmidt 2000). Long-run relief is available from cultural trends that are bringing more women into the labor force; but in the short term this influx also swells the ranks of the unemployed, especially when women have difficulty entering traditional positions in the core economy or seek jobs that existing labor regulations do not support. Because they have higher rates of total employment, many nations with social democratic policy regimes face less acute challenges from demographic change. But most have reached the limits to which employment in the public sector can be expanded, and there is some question whether these nations can maintain the high tax rates that finance public sector employment in the face of the competitive deregulation that more intense global competition may inspire (see Glyn and Rowthorn 1988; Genschel 1999). In sum, the social policy component of the employment strategies utilized by organized market economies in the past may no longer be as viable in the coming years.

The Macroeconomic Policy Regime Despite the tendency of liberal economics to attribute unemployment to labor-market rigidities flowing from government regulation and strong 63

Peter A. Hall

trade unions, there is substantial evidence that the character of macroeconomic policy can also affect it. The sharp increase in unemployment in Europe during the 1990s coincided with tight monetary policies that originated with the German Bundesbank and were then transmitted throughout Europe by the fixed exchange rates of the European Monetary System followed by the transition to monetary union, whose convergence criteria put severe restrictions on the fiscal and monetary policies of many of the European nations. The American job expansion was fueled by buoyant levels of aggregate demand, fed by the Reagan-era deficits and later an accommodating monetary policy, while levels of demand in Europe stagnated. Several economists have argued that a significant portion of European unemployment can be explained as the effects of macroeconomic policy, and there is some evidence for this in the righthand columns of Table 2.1, which adjust for differences in demand by adding the current account deficit (as a percentage of GDP) to the current rate of unemployment (Fitoussi and Phelps 1986; Fitoussi 1997; Soskice 1999, forthcoming; see also Blanchard 1999). They show that equilibrium rates of unemployment have generally been lower in the organized market economies than in their liberal counterparts. However, macroeconomic policies can matter in a number of other ways as well, notably in maintaining the capacity of wage-bargaining systems to deliver the wage moderation that is conducive to low levels of unemployment. Martin (1979) argues that the effective operation of the Rehn-Meidner model in Sweden depended on restrained fiscal policies. Iversen (1998b) suggests that centralized bargaining systems work best under an accommodating monetary policy. And several analysts argue that bargaining systems coordinated at the industry level function most effectively under nonaccomodating policies (Franzese 1994; Hall and Franzese 1998; Iversen 1998, 1994; Soskice and Iversen 1999). Over time, the organized market economies of Europe developed macroeconomic regimes that were conducive to the efficient operation of the institutions organizing their political economies. Thus, the advent of Economic and Monetary Union (EMU) has important implications. Even if the new European central bank pursues accommodating policies – which is far from certain given its independence – the context within which national industrial relations systems operate has changed because national monetary policy is no longer possible. Nations with centralized bargaining systems can no longer count on accommodating monetary regimes, and bargaining has shifted to the industrial level in 64

Organized Market Economies and Unemployment

several of them (Iversen 1998a). Those in which bargaining is coordinated at the industrial level no longer have central banks that can respond directly to national wage negotiations so as to enhance wage discipline across sectors, as the Bundesbank once did, since the European central bank must consider rates of inflation across the continent (Soskice 1997; Hall and Franzese 1998). Thus, a change in macroeconomic regime is forcing many of the organized market economies to reorganize the institutions and practices on which they have long relied for wage coordination. At a more general level, EMU also puts pressure on these nations to modify their labor market regulations. Since its members can no longer deploy a national monetary policy to respond to economic shocks, which may often be asymmetric across the union, the burden of adjustment to such shocks must be borne more heavily by labor markets; and many nations are considering reforms to labor regulations and social policies designed to render wages more flexible and labor more mobile. Some countries, such as the Netherlands, have devised coordinated responses to this challenge that reinforce the organized character of their economies, but others may be pushed toward deregulation. In sum, important developments in the production regime, social regime, and macroeconomic regime have altered the context for unemployment strategy in ways that call into question the viability of the practices utilized in past decades by the organized market economies to keep unemployment low. Some adjustment in strategy is inevitable. Must it entail wholesale deregulation? To address this question, one must look at the challenges that changes in the contemporary world now pose for these nations.

Economic Challenges at the Turn of the Century The most prominent dilemmas generated by recent economic and political developments are those associated with (1) a technological and managerial revolution, (2) the pressures of globalization, and (3) the rise of the service sector. After outlining each, I assess the capacities of the organized market economies to deal with them.

Coping with a New Technological Revolution A certain amount of technological change is always taking place. However, economic history is punctuated by moments when such changes are so 65

Peter A. Hall

far-reaching that they transform the business of many sectors at the same time, creating an industrial revolution. Innovations associated with steam power, iron production and steel-making, electrical goods, and the internal-combustion engine were far-reaching enough to do so (Landes 1969). At the close of the twentieth century, developments in microprocessing, telecommunications, and biotechnology again seem to have this character: they are transforming the nature of products, production, and distribution across many sectors simultaneously. Similarly, new approaches to the organization of business, such as the invention of the assembly line or multidivisional form, occasionally display such power that they transform corporate structures and strategy around the world (Chandler 1962, 1974). We are seeing a revolution here as well, based on a new set of practices associated with just-in-time inventory systems, team methods of production, quality circles, new forms of subcontracting, and customized production strategies, all demanding extensive reorganization of relations inside firms and among them (Womack, Jones, and Roos 1990). Many of these managerial techniques have been adopted around the world under the impetus of that most powerful motor for institutional isomorphism, the force of competition in international markets, which has been intensified by growing international interdependence; and the new microprocessing and telecommunications technologies have rendered many of these practices more feasible. These developments have serious implications for employment. First, they mean that many of the jobs available in coming years will be in entirely new sectors devoted to the development of such technologies in semiconductors, software, and bioengineering. Can the organized market economies create jobs in such sectors? Second, to remain competitive, firms in all sorts of sectors will have to build these technologies into their production and distribution processes. Can firms in the organized market economies do so? Finally, firms will also need to adopt the new managerial techniques to compete effectively. Can enterprises in the organized market economies make effective use of these techniques? The first of these challenges poses the greatest problems for organized market economies. The orientation of their financial markets to network monitoring as well as their lengthy employment tenures and highly consensual systems of decision making make it more difficult for such economies to shift resources – whether in the form of labor or capital –

66

Organized Market Economies and Unemployment

rapidly across sectors or into radically new product lines.8 Although some European nations have made inroads where they already had a foothold, as in pharmaceuticals and telecommunications, many have not been able to take full advantage of the potential for employment growth in such high-growth sectors as semiconductors or biotechnology (Ziegler 1997). Nor have firms in the organized market economies of Europe been adept at the radical innovation that drives growth in these new sectors (Hall 1997; Hall and Soskice, forthcoming). As liberal orthodoxy suggests, the structures of a liberal market economy are better suited to such tasks. Since managerial prerogatives are high in such economies, it is easier for firms to embark on new ventures and to reorganize to accommodate such changes (cf. Lehrer 1997b). Financial markets with high levels of capital mobility and venture capital and fluid labor markets make it easier for firms in such economies to assemble the capital and labor needed to move into radically new pursuits and more feasible because they know they can divest themselves of those assets if the endeavors do not prove profitable. This encourages risk taking, especially in new technologies, and the strength of the American economy in many sectors dominated by radical innovation provides support for this view (Soskice 1994a). However, the situation of the organized market economies is far from entirely bleak. In many respects, they are better placed than their liberal counterparts to meet the other two challenges that this new technological revolution poses. With a highly skilled workforce oriented toward continuous improvement in products and production processes, these economies can readily incorporate new technology into existing operations (Soskice 1994a). Although such changes often require renegotiation of skill categories and job tasks because of the extensive involvement of employees in corporate decision making, once this has been accomplished, that involvement means that firms can often count on more extensive cooperation from their employees in making the new technology effective than their Anglo-American counterparts secure. Moreover, for full effectiveness, many of these new managerial techniques rely on incomplete or relational contracting of the sort that has 8

As Soskice (1994b) notes, the “group-based” business organization of the organized market economies in Asia seem to be much more conducive to rapid movements of resources of this sort than are the “industry-based” organized market economies of Europe.

67

Peter A. Hall

long been practiced in organized market economies and for which their institutions and dense business networks provide ample support (Casper 1997). This is not surprising given that a number of these techniques originated in Japan, itself an organized market economy. Conversely, as many Japanese transplants have found, it can be more difficult to implement such practices in a liberal market economy, where there are few supporting institutions for them. Nations such as Germany and Sweden have long had high levels of quality control and product customization of the sort that the new managerial revolution emphasizes. In short, although the organized market economies of Europe are likely to be slower to shift resources into some of the new sectors created by this technological revolution, in other respects, they are well equipped to take advantage of it.

Coping with Globalization Every age has a term that encapsulates the ideals and anxieties of the time. In Europe today, that term is surely “globalization.” It captures the hopes of some for a new transnational community and the fears of many that international forces beyond the control even of governments will destroy their jobs and culture. As such, the term has become highly ambiguous. I use it to refer to the processes that follow from recent increases in the flows of goods and capital across national borders, inspired by declining transportation and communication costs, falling trade barriers, the expansion of international financial markets, and the growing accessibility of foreign markets. Although these processes have affected all nations, their impact has been strongly felt in Western Europe, where development of a single European market and democratization to the south and east have intensified competition and opened up new production sites. Firms now face greater incentives to locate abroad and to rationalize their operations so as to compete more effectively in global markets.9 What impact will globalization have on the organized market economies of Europe? Liberal orthodoxy offers a clear-cut answer: globalization is likely to shift employment away from such economies unless they deregulate, truncating social programs and labor regulations. Much of the political discomfit in Europe today arises because this diagnosis offers the organized market economies a highly unpalatable choice: dis9

Here I treat what some call “Europeanization” and “globalization” as components of the same process. This section draws on Hall (1997).

68

Organized Market Economies and Unemployment

mantle the institutions many there associate with social progress or face higher levels of unemployment. However, there are good grounds for questioning core elements of the conventional image of globalization on which this diagnosis rests. First, it regards firms as essentially similar across nations and subject to the same incentives. Second, it associates firm competitiveness almost exclusively with labor costs. From this, it follows that many firms will be tempted to move their production offshore in search of cheaper labor. Third, it posits a particular political dynamic. Governments will come under increasing, and potentially irresistible, pressure from business, backed by threats to exit the national economy, to deregulate so as to lower domestic labor costs, reduce rates of taxation, and render domestic markets more flexible. What resistance there is to such steps will come from trade unions seeking to protect the wages of their members and social democratic parties trying to preserve social programs. Thus, the model postulates a frontier defined by a trade-off between unemployment and liberal policies and predicts that any nation’s position on it will be determined largely by the amount of political resistance that labor and the Left can mount to such proposals (Ohmae 1991; Reich 1994). In the short term, this model forecasts substantial woe for organized market economies, as strong labor movements and social democratic parties resist deregulatory reform and delay it, raising levels of unemployment. In the long term, it predicts that economic institutions and public policies will converge across nations driven by processes of competitive deregulation. Thus, contemporary views of globalization contain a “convergence hypothesis” analogous in force, but considerably less sanguine in social implications, to the one generated forty years ago by theorists of industrial society (cf. Graubard 1964; Kerr 1973; Berger and Dore 1996). To date, those who have challenged this view have generally tried to show that the internationalization of trade and finance is not as unprecedented or extensive as many suppose or that globalization does not mean that nation-states have been superseded because the architects of the international regimes facilitating it are national governments (cf. Boyer 1996; Cohen 1996; Wade 1996). There is some validity to both views. However, there are even stronger grounds for challenging the conventional image of globalization, rooted in the observation that the institutional differences between liberal and organized market economies create comparative institutional advantage. Theories of comparative economic 69

Peter A. Hall

advantage have been central to international economics for over a century, but most turn on national differences in factor endowments (Stolper and Samuelson 1941). More recently, it has been suggested that comparative economic advantage may also derive from the institutional structures of a nation (Nelson 1993; Zysman 1994; cf. Porter 1990). This intuition is congruent with the observation of “endogenous-growth” theorists that economic growth in most nations cannot be explained entirely by changes in the level of capital, labor, and technology there (Romer 1986; Grossman and Helpman 1994). But most studies postulating comparative institutional advantages remain vague about how they are generated. Recent analyses of the differences between liberal and organized market economies, however, specify the sources of comparative institutional advantage with much more precision (Soskice 1998; Hall and Soskice forthcoming). They suggest that firms derive distinctive advantages from the institutional structures available for various kinds of coordination in the economy. Some kinds of activities can be pursued more efficiently where there are plentiful mechanisms for nonmarket coordination of the sort found in organized market economies, whereas others can be accomplished more efficiently in liberal market economies where interfirm and intrafirm relations are mediated more heavily by market mechanisms (Hall 1997). Soskice (1994a) has shown, for instance, that the two kinds of economies are conducive to quite different types of innovation. Such a perspective calls into question the core postulates of the conventional view of globalization. It rejects the proposition that firms are essentially similar across nations, suggesting instead that firms will develop nationally distinctive strategies to take advantage of the institutional infrastructure present in their economy. Although many assume that all companies behave like American corporations, firms in different types of economies tend to pursue quite different strategies (Knetter 1989; Schettkat 1992; Burgess and Knetter 1996; Lehrer 1997b; Soskice 1998). Second, this approach suggests that firms in organized market economies may not be as mobile as conventional views of globalization imply. Other things being equal, companies will prefer lower labor costs; but many firms in organized market economies compete on quality as well as price and derive a significant portion of their competitive edge from the institutional infrastructure supporting various kinds of nonmarket coordination there. They may be reluctant to give up that infrastructure simply for the sake of cheaper labor, and few nations offering low-cost labor also

70

Organized Market Economies and Unemployment

provide this kind of institutional support. Indeed, it is firms in liberal market economies, which tend to compete more heavily on price and depend primarily on market mechanisms to mediate their relations with other actors, that will be more likely to leave the national economy in search of low-cost labor. Lane (1997) finds that German firms are more reluctant to move their operations abroad than British firms. Third, while this approach admits that many firms will move components of their operations abroad when opportunities to do so increase, it suggests that the dynamic driving this movement may not conform to conventional views. Instead of roaming the world for cheap labor, firms are just as likely to engage in institutional arbitrage; that is, to use new opportunities for movement to exploit the benefits available from variation in the institutional frameworks of political economies. Firms may move to liberal market economies in order to secure the opportunities for radical innovation that they offer, as when German banks locate their merchant banking in London and European pharmaceutical companies place laboratories in the United States; but firms may also move toward organized market economies to secure access to the skilled labor, high levels of quality control, and other institutional advantages they provide, as when General Motors locates its new engine plant in Dusseldorf. Finally, this perspective calls into question the political dynamic that is conventionally associated with globalization, namely, one which pits labor against capital as business interests press governments for greater deregulation. We are likely to see such conflict in liberal market economies where business is interested in deregulation because it sharpens the effectiveness of the market mechanisms on which firms there depend for coordination. In organized market economies, however, large portions of the business community may resist deregulation in order to preserve the institutional infrastructures supporting the kind of nonmarket coordination on which they rely for competitive advantages. Here, the pressures of globalization may even unite business and labor interests in defense of arrangements that both find useful, as Wood (1997), Thelen (forthcoming), and Swenson (1989) have found in detailed case studies. Thus, political pressure to deregulate in organized market economies may be less substantial than conventional wisdom postulates. In short, while firms in organized market economies can be expected to move some of their operations and assets abroad as they search for market access and market share in a globalizing economy, this need not

71

Peter A. Hall

produce domestic deregulation or unemployment. Just as the American economy has long benefited from the reach of its multinational enterprises, so may the Europeans (Doremus et al. 1998).

Coping with the Rise of the Service Sector Equally important challenges confronting the organized market economies stem from the rise of the service sector (see Table 2.2). If the Europeans are to create significant numbers of jobs, they will have to do so in services. This development has breathed new life into liberal orthodoxy. Demand for many kinds of services, especially in retailing, personal or domestic services, tourism, and restaurant work, is highly price-elastic and positions there are often associated with low wages and low rates of productivity growth. Accordingly, even sophisticated analysts have begun to speculate that job creation in the service sector may require widespread acceptance of low wage rates, higher levels of income inequality, reductions in social benefits to lower the reservation wage or nonwage labor costs, and the expansion of part-time or temporary employment (Iversen and Wren 1996; Scharpf 1997; Esping-Andersen 1999). These proposals pose profound challenges to organized market economies where powerful trade unions and political parties committed to social equality have long sought to increase wage floors and reduce income differentials. To the many Europeans who believe that economic progress means shorter working hours, more egalitarian wage structures, and greater job security, such measures seem retrograde steps. At best, efforts to introduce them threaten the social consensus that underpins the operation of many political economies. At worst, they threaten the regulatory foundations on which the capacity for nonmarket coordination has been built. However, there are several reasons for believing that the European economies may be able to create service-sector jobs without wholesale deregulation. First, despite the emphasis of some on personal services and retailing, the service sector is a broad one that includes business services in accounting, finance or advertising, health care, and education, where many positions demand high skills and the productivity that can justify reasonably high wages. Employment can be created in many spheres of services without substantially increasing income inequality or job insecurity. Second, as Esping-Andersen (1996) points out, the route taken toward service sector expansion in the United States, where low wages and 72

Organized Market Economies and Unemployment

labor-market flexibility have generated jobs in retailing, personal, and food services, is not the only possible one. The Nordic nations have already expanded employment in health, education, and social services, not by lowering wages but by expanding public provision of such goods (Iversen and Wren 1998). There are some trade-offs here: the tax rates required to fund job expansion in these sectors may squeeze disposable income enough to limit demand for personal services and job growth there, but this route to service sector expansion is available to all the organized market economies of Europe at least on a modest scale. The economies in most difficulty are those led by Christian Democratic parties reluctant to expand the public sector enough to create service employment there but committed to generous social programs and labor regulations that discourage job creation in private services by raising the reservation wage and limiting labor flexibility (Esping-Andersen 1996; Scharpf and Schmidt 2000).10 However, there may be ways for them to expand the service sector. As more women enter the workforce in these nations, the demand for services that housewives once performed is growing, as is the pool of women available for service sector work, often on a part-time basis. The result is likely to be some expansion in service sector employment. Moreover, selective regulatory changes that stop well short of largescale deregulation, such as those that permit the extension of shop hours and part-time employment, can improve service sector growth. As Scharpf (1997) has suggested, by reducing the social security taxes that make up almost half of labor costs in many of these economies, notably on positions at the bottom of the wage distribution, the cost of labor could be reduced enough to encourage job expansion without radically altering wage distributions or the disposable income of workers. Of course, the difficult issue is how to replace the government revenue lost through such initiatives. Although higher taxes on assets or incomes might be used to make up the shortfall, it can be difficult to mobilize political consent for such measures. However, the government of France has already begun to take such steps, coupled with more means testing of social benefits, and there is substantial room for them in many other nations (Levy 2000). 10

It should be noted that, while this stance is characteristic of most Christian Democratic parties, some, such as those in Italy and Spain, have been more tolerant of a large public sector, and the views of all such parties are in flux today.

73

Peter A. Hall

Finally, the oft-repeated premise that productivity in services cannot grow fast enough to support real-wage increases of the sort that coordinated bargaining systems demand may well be incorrect. The technological innovations of the new industrial revolution in microprocessing and telecommunications lend themselves to applications in the service sector and may facilitate more rapid productivity increases there in the coming years. If so, the organized market economies will find it easier to accommodate expanded service sector employment without substantial modifications to their wage or industrial relations structures. Job creation should be possible in the organized market economies even within the context of postindustrialism.

Institutional Convergence and the Politics of Adjustment Familiar processes of hysterisis make it difficult for nations to reduce rates of unemployment once they have risen, especially when the incidence of long-term unemployment is high, as it is in many European economies; and some features of these economies, including their capital intensity, high nonwage labor costs, and levels of employment protection, mean that the process of lowering unemployment rates may be more protracted than it would be in liberal market economies (cf. Blanchard and Summers 1986; Berthold et al. 1999; Blanchard 1999). In general, adjustment processes of this sort should be slower in organized market economies where the presence of institutionally entrenched producer groups means that many facets of adjustment must be negotiated and cannot be effected simply by changing relative market prices. This is a cost borne by economies that promote extensive forms of nonmarket coordination, high real wages, and extensive social protection. For these reasons, any decline in unemployment in continental Europe is likely to be gradual. However, I have argued that the organized market economies of Europe have more resources, institutional and otherwise, for coping with the challenges posed by technological revolution, globalization, and postindustrialism than many conventional analyses allow. The implication is that they can do so without radically transforming the shape of their economies, although the effort is bound to entail selective deregulation. I want to conclude by considering this issue more closely: can such economies implement limited reforms to cope with these challenges without dissolving into liberal market economies and thereby erasing the alternative economic model they have developed? 74

Organized Market Economies and Unemployment

There are some grounds for skepticism on this point implicit even in the analytical framework developed here. Perhaps the most important turn on the institutional complementarities that can be found among the subspheres of the economy. Two institutions are complementary when the presence of one raises the returns available from the other; and such complementarities are often found within organized market economies (Milgrom and Roberts 1992, 1995). Financial institutions that provide capital on terms independent of short-term profitability render labormarket institutions associated with long job tenures more feasible, for instance, and the latter make corporate strategies and structures based on implicit contracts with employees more productive (Aoki 1994; Hall and Soskice forthcoming). Since organized market economies often have tight institutional linkages of this sort, deregulatory reforms to institutions in one sphere of the economy may put significant pressure on institutions in other spheres that can snowball into wider deregulation across the economy. The prospects for this kind of dynamic in Europe stem primarily from initiatives to deregulate financial markets (cf. Streeck 1997). Powerful trade unions and differences of opinion among the member states of the European Union (EU) have limited deregulatory initiatives in the spheres of social policy and labor markets. But the EU has actively encouraged deregulation of financial markets and many of the large European banks have been receptive as they seek market share in increasingly global capital markets. More firms have turned to these markets for finance where it is often provided on Anglo-American terms that stress financial transparency, short-term profitability, and the corporate strategies associated with “shareholder value” (Ziegler 1998; Vogel 1999). Financial deregulation not only threatens the dense networkmonitoring systems associated with house-bank relationships, crossshareholding, and close interfirm collaboration. Through institutional complementarities, it also raises the prospect of widespread changes in labor-market practices. Without access to patient capital, many firms would have difficulty maintaining long-term employment contracts, and more liberal layoff strategies could precipitate changes in industrial relations systems touching works councils and wage coordination. Financial-market deregulation could be the wedge that drives large-scale deregulation in organized market economies. Although this dynamic is a lively possibility, to see it as inevitable would be premature. The number of German firms that have sought a listing on 75

Peter A. Hall

American stock exchanges remains negligible, and well-publicized efforts to increase shareholder value have often had little effect on the distinctive modes of operation of European firms. There are signs that, while the large banks are internationalizing, a dual system is emerging in which other financial institutions continue to maintain close relationships to industry networks and cross-shareholding remains high in many parts of Europe (Griffin 1997; Deeg and Pérez 1998; Ziegler 1998). The organized market economies may be able to sustain some financial reform without widespread effects on other spheres. The other challenge to institutional stability in the organized market economies is more sociological. Effective nonmarket coordination depends on more than the presence of appropriate institutions. Since there may be multiple equilibria on which the actors can coordinate even in the presence of one set of institutions, effective coordination on one of them also depends on shared understandings among the actors, the reputations they cultivate with each other, their capacities to work out ongoing problems, and some level of consensus about appropriate goals. In sum, achieving effective equilbria in such settings also requires an appropriate social underpinning (Streeck 1992, 1997). However, high levels of unemployment, rapid technological change, globalization, and the movement toward a postindustrial economy threaten the norms and understandings that underpin organized market economies by placing new issues on the agenda and disrupting longstanding compromises among social actors. Greater international interdependence intensifies cleavages between the traded and sheltered sectors of the economy (Frieden 1991; Pontusson and Swenson 1995). Efforts to expand the low-wage service sector can unsettle hard-won compromises about wage equality and social benefits. High levels of unemployment can deepen insider-outsider conflicts and call into question the good faith of those who bargained peaceably under full employment. Indeed, although hysterisis is normally seen as an economic phenomenon, there may be analogous processes of political hysterisis whereby rising levels of unemployment intensify conflict among the relevant producer groups to such an extent that it becomes more difficult for them to negotiate adjustments that would bring unemployment down. Rothstein (1998) argues that some of Sweden’s recent problems stem from precisely such a breakdown in trust among the key social partners. High levels of unemployment also destabilize collaborative systems of vocational training as firms facing low levels of demand decline to train and those seeking 76

Organized Market Economies and Unemployment

apprenticeships find there are no positions for them. In such circumstances, it can also be difficult to maintain the effectiveness of coordinated wage bargaining, as skilled workers with substantial organizational power protect their wages at the expense of unemployed outsiders (Culpepper forthcoming). In short, organized market economies may not operate as effectively when unemployment is high as when it is low, and frustration with the difficulties of negotiating adjustment when settled understandings have been disrupted may fuel initiatives to deregulate them. Whether these difficulties will precipitate breakdowns in coordination and a reversion toward deregulation may turn on the effectiveness of existing institutions for resolving conflict in the political economy and on the quality of the leadership available (Thelen 2000). Such problems put heavy political demands on nations. However, there are reasons for cautious optimism on this front. The actors in organized market economies do not operate on a tabula rasa. In most cases, they have past experience in resolving such problems and deliberative institutions for doing so. Moreover, their key producer groups operate from bases of organizational power entrenched and substantial enough to remind their counterparts that reaching agreement is likely to be less costly than trying to impose a settlement unilaterally. The capacity of the “social partners” in these political economies to find negotiated adjustment paths is ultimately founded on a finely tuned balance of power that remains robust even when consensus founders.11 Recent experience bears out this observation. Despite strains, bargaining over wages and working conditions remains relatively coordinated in all of the organized market economies (Lange, Golden, and Wallerstein 1995; Regini 1995; Iversen 1998a). Where there have been major changes, most entail a shift downward in the level at which bargaining is conducted, as from the peak to the sectoral level in Sweden. However, recent challenges have also revived negotiations at the peak level in nations such as the Netherlands, where unusually broad agreements covering labor regulations and social policy as well as wages have been reached in order to expand employment (Rhodes 1997; Visser and Hemerijck 1997). The pace and extent of such agreements will vary across nations. They may be harder to secure in economies where the trade unions are domi11

Of course, socioeconomic developments can alter this balance of power in significant ways by shifting the interests of the actors and thus the opportunity costs of alternative courses of action, but even those disadvantaged by such shifts often retain residual bases of organizational power (Swenson 1999).

77

Peter A. Hall

nated by a core labor force focused on industrial production and inclined to resist deregulatory moves in order to protect their existing privileges. The German trade unions, for instance, refused to countenance the development of a low-wage economy in the east after unification and have resisted more recent proposals to trim social benefits and deregulate shop hours in order to create employment (cf. Webber 1994; Locke and Jacoby 1995; Manow and Seils 2000). In such economies, more rapid progress is likely to be made on schemes to preserve jobs in the core industrial sectors by shifting work-time and working conditions of the sort seen in the German chemical and automobile industries. Conversely, economies with a large service sector, substantial wage dispersion, and more heterogeneous trade unions are more likely to secure broad agreements that reform social regimes, taxation systems, and labor regulations in order to expand employment, since trade unions of this type will speak more strongly for the interests of those on the margins of the core labor force. Such agreements have been prominent in Italy, Spain, and Portugal (Rhodes 1997; Pérez 1999). However, the Dutch case indicates that, with sufficient government pressure, social pacts of this sort can be forged even in organized market economies; and the experience of the Netherlands, Denmark, and Germany suggests that selective deregulation can be implemented in such economies without unraveling the coordination processes that underpin their distinctiveness (Visser and Hemerijck 1997). There are many ways to improve the functioning of these economies without impairing their coordinating capacity, and, as Levy (2000) points out, most have social policy regimes that are susceptible to incremental reform. In sum, there is no doubt that the organized market economies face significant challenges and have to make changes to meet them, but the pressures for convergence to a liberal model seem far from inexorable. The outcomes should vary across nations in response to political as much as economic factors.

Conclusion I have argued that the liberal views permeating economic commentary about Europe today, while telling in some points, misdiagnose the problems of the European economies because they tend to assume there is only one viable way to organize an economy when there are at least two routes to economic success. As a consequence, these analyses overstate the problems stemming from globalization and the rise of the 78

Organized Market Economies and Unemployment

service sector and fail to appreciate the comparative institutional advantages that organized market economies retain even in the face of such challenges. The result is often a blanket endorsement of deregulation that ignores the corrosive effects it can have on the existing strengths of these economies. To understand economic and political developments in Europe, one needs to acknowledge the distinctive character of organized market economies in terms that appreciate both their weaknesses and strengths. A growing body of work leads me to conclude that these economies have considerable resources for coping with the challenges they face. They are well positioned to adapt to the managerial revolution and to diffuse new technology. They bring an important set of comparative institutional advantages to global competition. Although many are not as well placed to promote job growth in the service sector as most liberal market economies, their prospects for doing so are substantial. It may take some time to lower rates of unemployment again in these economies, as well as selective deregulation and other adaptations in their institutional frameworks. However, the problem does not seem to require convergence to a liberal model. Indeed, the pressures for convergence, while substantial, seem far from inexorable. Even in the face of contemporary challenges, there remain at least two viable models for economic success, whose fate will turn as much on the politics as on the economics of adjustment. References Alvarez, Michael, Geoffrey Garrett, and Peter Lange. 1991. Government Partisanship, Labor Organization and Macroeconomic Performance. American Political Science Review 85, 2 (June): 539–56. Aoki, M. 1994. The Japanese Firm as a System of Attributes: A Survey and Research Agenda. In M. Aoki and R. Dore, eds., The Japanese Firm: Sources of Competitive Strength. Oxford: Clarendon Press. Berger, Suzanne, ed. 1982. Interest Politics in Western Europe. New York: Cambridge University Press. Berger, Suzanne, and Ronald Dore, eds. 1996. National Diversity and Global Capitalism. Ithaca, N.Y.: Cornell University Press. Berthold, Norbert, Rainer Fehn, and Eric Thode. 1999. Falling Labor Share and Rising Unemployment: Long-Run Consequences or Institutional Shocks? Working paper, University of Wurzburg. Blanchard, Olivier. 1999. The Role of Shocks and Institutions in the Rise of European Unemployment. Working Paper W7282. Cambridge, Mass.: National Bureau of Economic Research. 79

Peter A. Hall Blanchard, Olivier, and Lawrence Summers. 1986. Hysterisis and the European Unemployment Problem. NBER Macroeconomics Annual, 15–77. Boyer, Robert. 1990. The Regulation School: A Critical Introduction. New York: Columbia University Press. ———. 1996. The Convergence Hypothesis Revisited: Globalization but Still the Century of Nations? In Suzanne Berger and Ronald Dore, eds., National Diversity and Global Capitalism. Ithaca, N.Y.: Cornell University Press. Burgess, Simon, and Michael M. Knetter. 1996. An International Comparison of Employment Adjustment to Exchange Rate Fluctuations. Working Paper 5861. Cambridge, Mass.: National Bureau of Economic Research. Calmfors, Lars, and John Driffill. 1988. Bargaining Structure, Corporatism and Macroeconomic Performance. Economic Policy 5 (October): 13–61. Cameron, David. 1984. Social Democracy, Corporatism, Labour Quiescence, and the Representation of Economic Interest in Advanced Capitalist Society. In John A. Goldthorpe, ed., Order and Conflict in Contemporary Capitalism. New York: Oxford University Press, 143–78. Carlin, Wendy, and David Soskice. 1990. Macroeconomics and the Wage Bargain. Oxford: Oxford University Press. ———. 1997. Shocks to the System: the German Political Economy under Stress. National Institute Economic Review 159 (January): 57–76. Casper, Steven. 1997. The Legal Framework for Corporate Governance: Explaining the Development of Contract Law in Germany and the United States. Paper presented to a workshop on Varieties of Capitalism, Wissenschaftszentrum, Berlin, June. ———. 1998. National Institutional Frameworks and High-Technology Innovation in Germany: The Case of Biotechnology. Manuscript. Chandler, Alfred. 1962. Strategy and Structure. Cambridge, Mass.: MIT Press. ———. 1974. The Visible Hand. Cambridge, Mass.: Harvard University Press. Cohen, Eli. 1996. La Tentation Hexogonale. Paris: Seuil. Culpepper, Pepper. Forthcoming. Employers Associations, Public Policy, and the Politics of Decentralized Cooperation. In Peter A. Hall and David Soskice, eds., Varieties of Capitalism: The Foundations of Comparative Institutional Advantage. Oxford: Oxford University Press. Deeg, Richard, and Sofía Pérez. 1998. International Capital Mobility and Domestic Institutions: Corporate Finance and Governance in Four European Cases. Dore, Ronald, Robert Boyer, and Zoe Mars, eds. 1994. The Return of Incomes Policy. London: Pinter. Doremus, Paul, William Keller, Louis Pauly, and Simon Reich. 1998. The Myth of the Global Corporation. Princeton, N.J.: Princeton University Press. Ebbinghaus, Bernhard, and Philip Manow. 1998. Studying Welfare-State Regimes and Varieties of Capitalism: An Introduction. Paper prepared for the Conference on Varieties of Welfare Capitalism, Max Planck Institute for the Study of Societies, Köln. Eichengreen, Barry. 1994. Institutions and Economic Growth: Europe after World War II. Paper presented to a conference on Comparative Experience of Economic Growth in Postwar Europe, Oxford. 80

Organized Market Economies and Unemployment Esping-Andersen, Gosta. 1990. Three Worlds of Welfare Capitalism. Princeton, N.J.: Princeton University Press. ———. 1996. Workers without Work. In Gosta Esping-Andersen, ed., Welfare States in Transition. Beverly Hills: Sage. ———. 1999. Social Foundations of Postindustrial Economies. Oxford: Oxford University Press. Finegold, David, and David Soskice. 1988. Britain’s Failure to Train: Explanations and Some Possible Strategies. Oxford Review of Economic Policy 4 (November). Fioretos, Karl-Orfeo. 1997. Anchoring Adjustment: Globalization, National Production Regimes and the Domestic Sources of Multilateral Institutions. Ph.D. dissertation, Columbia University. Fitoussi, Jean-Paul. 1997. Substitutabilities vs. Complementarities between Structural and Macroeconomic Policies. European University Institute Working Paper. 97/39. Florence. Fitoussi, Jean-Paul, and Edmund S. Phelps. 1986. Causes of the 1980s Slump in Europe, Brookings Papers on Economic Activity 2: 487–513. Franzese, Jr., Robert J. 1994. Central Bank Independence, Sectoral Interest and the Wage Bargain. Harvard Center for European Studies Working Paper. ———. Forthcoming. Monetary Policy and Wage Bargaining: Macro-Institutional Interactions in the Traded, Public and Sheltered Sectors. In Peter A. Hall and David Soskice, eds., Varieties of Capitalism: The Foundation of Comparative Institutional Advantage. Oxford: Oxford University Press. Frieden, Jeffry. 1991. Invested Interests: The Politics of National Economic Policies in a World of Global Finance. International Organization 45 (1991): 425–51. Garrett, Geoffrey. 1997. Partisan Politics in the Global Economy. New York: Cambridge University Press. Genschel, Philipp. 1999. Tax Competition and the Welfare State. Max Planck Institute for the Study of Societies Discussion Paper. Köln. Glyn, Andrew, and Robert Rowthorn. 1988. European Unemployment, Corporatism and Structural Change. American Economic Review 78 (May): 194–99. Golden, Miriam. 1993. The Dynamics of Trade Unionism and National Economic Performance. American Political Science Review 87 (June): 439–54. Goldthorpe, John A., ed. 1984. Order and Conflict in Contemporary Capitalism. New York: Oxford University Press. Graubard, Stephen, ed. 1964. A New Europe? Boston: Beacon. Griffin, John. 1997. National Capitalisms? Toward a Comparison of Tenuous States from a Modern Perspective of Ownership Politics. Paper presented to the American Political Science Association. Washington, D.C. Grossman, Gene, and Elhanan Helpman. 1994. Endogenous Innovation in the Theory of Growth. Journal of Economic Perspectives 8, 1 (Winter): 23–24. Hall, Peter A. 1986. Governing the Economy. New York: Oxford University Press. ———. 1994. Central Bank Independence and Coordinated Wage Bargaining: Their Interaction in Germany and Europe. German Politics and Society 31: 1–23. 81

Peter A. Hall ———. 1997. The Political Economy of Adjustment in Germany. In Frieder Naschold et al., eds., Okeonimische Leistungsfahigkeit und Institutionelle Innovation. Berlin: Sigma. ———. 1999. The Political Economy of Europe in an Era of Interdependence. In Herbert Kitschelt et al., eds., Continuity and Change in Contemporary Capitalism. New York: Cambridge University Press, 135–63. Hall, Peter A., and Robert J. Franzese, Jr. 1998. Mixed Signals: Central Bank Independence, Coordinated Wage Bargaining and European Monetary Union. International Organization 52, 3 (Summer): 505–35. Hall, Peter A., and David Soskice. Forthcoming. Introduction. In Peter A. Hall and David Soskice, eds., Varieties of Capitalism: The Foundations of Comparative Institutional Advantage. Oxford: Oxford University Press. Huber, Evelyne, and John D. Stephens. 1999. Welfare State and Production Regimes in the Era of Retrenchment. Occasional Papers, School of Social Science, Institute for Advanced Study, Princeton University. International Monetary Fund. 1996. World Economic Outlook (May). Washington, D.C.: IMF. Iversen, Torben. 1998a. Power, Flexibility and the Breakdown of Organized Wage Bargaining. Comparative Political Studies 28: 399–436. ———. 1998b. Wage Bargaining, Central Bank Independence and the Real Effects of Money, International Organization 52, 2 (Summer): 469–504. Iversen, Torben, and Anne Wren. 1998. Equality, Employment and Budgetary Restraint: The Trilemma of the Service Economy. World Politics. Katzenstein, Peter. 1985. Small States in World Markets. Ithaca, N.Y.: Cornell University Press. Kennedy, Neale O. 1997. Unemployment in the EU and the U.S.: A Macroeconomic Perspective. European University Institute Working Paper 97/44. Florence. Keohane, Robert O. 1984. The World Political Economy and the Crisis of Embedded Liberalism. In John H. Goldthorpe, ed., Order and Conflict in Contemporary Capitalism. New York: Oxford University Press, 15–38. Kerr, Clark. 1973. Industrialism and Industrial Man. 2d ed. Cambridge, Mass.: Harvard University Press. Knetter, Michael. 1989. Price Discrimination by U.S. and German Exporters. American Economic Review 79 (March): 198–210. Landes, David. 1969. The Unbound Prometheus. Cambridge, Mass.: Harvard University Press. Lane, Christel. 1995. Industry and Society in Europe. Aldershot: Edward Elgar. ———. 1997. European Companies between Globalization and Localization: A Comparison of the Internationalization Strategies of British and German MNCs. Paper presented to the Seminar on the State and Capitalism since l800, Harvard University. Lange, Peter, Miriam Golden, and Michael Wallerstein. 1995. The End of Corporatism? Wage Setting in the Nordic and Germanic Countries. In Sanford Jacoby, ed., Workers of Nations. Oxford: Oxford University Press. 82

Organized Market Economies and Unemployment Lehmbruch, Gerhard, and Philippe Schmitter. 1982. Patterns of Corporatist PolicyMaking. Beverly Hills: Sage. Lehrer, Mark. 1997a. The German Model of Industrial Strategy in Turbulence: Corporate Governance and Managerial Hierarchies in Lufthansa. Industry and Innovation 4, 1. ———. 1997b. Comparative Institutional Advantage in Corporate Governance and Managerial Hierarchies: The Case of European Airlines. Ph.D. dissertation, INSEAD, Fontainbleau. Levy, Jonah. 2000. France: Directing Adjustment? In Fritz Scharpf and Vivien Schmidt, eds., Welfare and Work in the Open Economy. New York: Oxford University Press, 308–50. Locke, Richard, and Wade Jacoby. 1995. The Dilemmas of Diffusion: Social Embeddedness and the Problems of Institutional Change in East Germany. Manuscript. Manow, Philip, and Eric Seils. 2000. Adjusting Badly: The German Welfare State, Structural Change and the Open Economy. In Fritz Scharpf and Vivien Schmidt, eds., Welfare and Work in the Open Economy. New York: Oxford University Press, 264–307. Mares, Isabela. 1997. Business Coordination and Social Policy Development: The Case of Early Retirement. Paper presented to a workshop on Varieties of Capitalism, Wissenschaftszentrum, Berlin. June. ———. Forthcoming. Business and Social Policy: A Theoretical Framework for the Understanding of Firms’ Roles in the Process of Social Policy Development. In Peter A. Hall and David Soskice, eds., Varieties of Capitalism: The Foundations of Comparative Institutional Advantage. Oxford: Oxford University Press. Martin, Andrew. 1979. The Dynamics of Change in a Keynesian Political Economy: The Swedish Case and Its Implications. In Colin Crouch, ed., State and Economy in Contemporary Capitalism. London: Croom Helm. Milgrom, Paul, and John Roberts. 1992. Economics, Organisation and Management. Englewood-Cliffs. N.J.: Prentice-Hall. ———. 1995. Complementarities and Fit: Strategy, Structure and Organizational Change in Manufacturing. Journal of Accounting and Economics 19: 179–208. Milward, Alan. 1984. The Reconstruction of Western Europe, 1945–51. London: Methuen. Nelson, Richard, ed. 1993. National Innovation Systems. New York: Oxford University Press. Nickell, Stephen. 1997. Unemployment and Labor Market Rigidities: Europe vs. North America. Journal of Economic Perspectives 11, 4 (Summer): 55–74. Ohmae, Kenichi. 1991. The Borderless World. New York: Harper. Olson, Mancur. 1982. The Rise and Decline of Nations. New Haven, Conn.: Yale University Press. Organisation for Economic Co-operation and Development (OECD). 1977. Towards Full Employment and Price Stability. Paris: OECD. ———. 1994. The OECD Jobs Study. Paris: OECD. ———. 1996a. Economic Outlook. 60. Paris: OECD. ———. 1996b. Economies at a Glance: Structural Indicators. Paris: OECD. 83

Peter A. Hall Pérez, Sofía. 1999. The Resurgence of National Social Bargaining in Europe: Explaining the Italian and Spanish Experiences. Estudios Working Papers 1999/130. Instituto Juan March, Madrid. Polanyi, Karl. 1944. The Great Transformation. Boston: Beacon. Pontusson, Jonas, and Peter Swenson. 1995. Labor Markets, Production Strategies and Wage-Bargaining Institutions. Comparative Political Studies 29, 223–50. Porter, Michael. 1990. The Competitive Advantage of Nations. London: Macmillan. Przeworski, Adam, and Michael Wallerstein. 1982. The Structure of Class Conflict in Democratic Capitalist Societies. American Political Science Review 76: 215–38. Regini, Marino. 1995. Uncertain Boundaries. Cambridge: Cambridge University Press. Reich, Robert. 1994. The Work of Nations. New York: Knopf. Rhodes, Martin. 1997. Globalization, Labour Markets and Welfare States: A Future of Competitive Corporatism. In Martin Rhodes and Yves Meny, eds., The Future of European Welfare. London: Macmillan. Romer, Paul. 1986. Increasing Returns and Long-Run Growth. Journal of Political Economy 94: 1002–37. ———. 1994. The Origins of Endogenous Growth. Journal of Economic Perspectives 8, 1 (Winter): 3–22. Rothstein, Bo. 1998. Breakdown of Trust and the Fall of the Swedish Model. Paper presented to the Seminar on the State and Capitalism since 1800, Harvard University, December. Ruggie, John Gerald. 1982. International Regimes, Transactions and Change: Embedded Liberalism in the Postwar Economic Order. International Organization 36: 379–415. Salant, Walter. 1977. Worldwide Inflation. Washington, D.C.: Brookings Institution. Scharpf, Fritz. 1990. Crisis and Choice in European Social Democracy. Ithaca, N.Y.: Cornell University Press. ———. 1997. Employment and the Welfare State: A Continental Dilemma. Paper presented to the Center for European Studies, Harvard University. Scharpf, Fritz, and Vivien Schmidt. 2000. Welfare and Work in the Open Economy. New York: Oxford University Press. Schettkat, Ronald. 1992. The Labor Market Dynamics of Economic Restructuring. New York: Praeger. Shonfield, Andrew. 1969. Modern Capitalism. New York: Oxford University Press. Siebert, Horst. 1997. Labour Market Rigidities: At the Root of Unemployment in Europe. Journal of Economic Perspectives 11, 3 (Summer): 37–54. Snower, Dennis, and Guillermo de la Dehesa. 1997. Unemployment Policy: Government Options for the Labour Market. Cambridge, Mass.: Cambridge University Press. Soskice, David. 1990. Wage Determination: The Changing Role of Institutions in Advanced Industrialised Countries. Oxford Review of Economic Policy 6, 4: 1–23. 84

Organized Market Economies and Unemployment ———. 1991. The Institutional Infrastructure for International Competitiveness: A Comparative Analysis of the U.K. and Germany. In A. B. Atkinson and R. Brunetta, eds., The Economies of the New Europe. London: Macmillan. ———. 1994a. Innovation Strategies of Companies: A Comparative Institutional Analysis of Some Cross-Country Differences. In W. Zapf, ed., Institutionenvergliech un Institutionendynamik. Berlin: Sigma. ———. 1994b. Finer Varieties of Advanced Capitalism: Industry versus Group Based Coordination in Germany and Japan. Paper presented to the Varieties of Capitalism Conference, Poitiers, September. ———. 1997. The Political Economy of European Monetary Union. Discussion Paper, Wissenschaftszentrum, Berlin. ———. 1998. German Technology Policy, Innovation and National Institutional Frameworks. Journal of Industry Studies. ———. 1999. Divergent Production Regimes: Coordinated and Uncoordinated Market Economies in the 1980s and 1990s. In Herbert Kitschelt et al., eds., Continuity and Change in Contemporary Capitalism. New York: Cambridge University Press. ———. 2000. Macroeconomic Analysis and the Political Economy of Unemployment. In Torben Iversen, Jonas Pontusson, and David Soskice, eds., Unions, Employers and Central Banks. New York: Cambridge University Press. Stolper, Wolfgang Friedrich, and Paul A. Samuelson. 1941. Protection and Real Wages. Review of Economic Studies 9: 58–73. Streeck, Wolfgang. 1992. Social Institutions and Economic Performance. Beverly Hills: Sage. ———. 1994. Pay Restraint without Incomes Policy: Institutionalized Monetarism and Industrial Unionism in Germany. In Ronald Dore, Robert Boyer, and Zoe Marn, eds., The Return of Incomes Policy. London: Pinter. ———. 1997. German Capitalism: Does it Exist? Can It Survive? In Colin Crouch and Wolfgang Streeck, eds., The Political Economy of Modern Capitalism. London: Routledge. Swenson, Peter. 1989. Fair Shares. Ithaca, N.Y.: Cornell University Press. ———. 1999. Varieties of Capitalist Interests and Illusions of Labor Power: Employers in the Making of the Swedish and American Welfare States. Paper presented to a Conference on Distribution and Democracy, Yale University, New Haven. Thelen, Kathleen. 1991. Union of Parts. Ithaca, N.Y.: Cornell University Press. ———. 2000. Why German Employers Cannot Bring Themselves to Abandon the German Model. In Torben Iversen, Jonas Pontusson, and David Soskice, eds., Unions, Employers and Central Banks. New York: Cambridge University Press. ———. Forthcoming. Varieties of Labor Politics in the Advanced Industrial Democracies. In Peter A. Hall and David Soskice, eds., Varieties of Capitalism: The Foundations of Comparative Institutional Advantage. Oxford: Oxford University Press. Tronti, Leonello. 1997. Benchmarking Employment Performance and Labour Market Policies. Brussels: European Commission. 85

Peter A. Hall Van Kersbergen, Kies. 1995. Social Capitalism: A Study of Christian Democracy and the Welfare State. London: Routledge. Visser, Jelle, and Anton Hemerijck. 1997. A Dutch Model. Ann Arbor: University of Michigan Press. Vogel, Steven K. 1999. The Crisis of German and Japanese Capitalism: Stalled on the Road to the Liberal Market Model. Manuscript. Wade, Robert. 1996. Globalization and Its Limits: Reports of the Death of the National Economy Are Greatly Exaggerated. In Suzanne Berger and Ronald Dore, eds., National Diversity and Global Capitalism. Ithaca, N.Y.: Cornell University Press. Webber, Douglas. 1994. The End of Solidarity. West European Politics. Womack, James P., Daniel T. Jones, and Daniel Roos. 1990. The Machine That Changed the World. New York: Macmillan. Wood, Stewart. 1997. Capitalist Constitutions: Supply-Side Reform in Britain and Germany 1960–90. Ph.D. dissertation, Harvard University. ———. Forthcoming. From Employer Preferences to Public Policy. In Peter A. Hall and David Soskice, eds., Varieties of Capitalism: The Foundations of Comparative Institutional Advantage. Oxford: Oxford University Press. Ziegler, Nicholas. 1997. Governing Ideas: Strategies for Innovation in France and Germany. Ithaca, N.Y.: Cornell University Press. ———. 1998. Corporate Governance in Germany: Toward a New Transnational Politics. Paper presented to the American Political Science Association, Washington, D.C. Zysman, John. 1983. Governments, Markets and Growth. Ithaca, N.Y.: Cornell University Press. ———. 1994. How Institutions Create Historically Rooted Trajectories of Growth. Industrial and Corporate Change 3, 1.

86

3 Globalization, Welfare States, and Employment IS THERE A EUROPEAN “THIRD WAY”?

Martin Rhodes

“Globalization” may now be a buzzword, but it reflects an important if as yet poorly understood reality.1 A “new international division of labor” has emerged, with extensive and expanding trading networks between countries, as well as new production networks (or “chains”) established by multinational companies (MNCs). Tariffs have declined to negligible levels among the industrialized countries, and successive negotiation rounds of the General Agreement on Tariffs and Trade (GATT) have increased the openness to trade of the developing countries as well. Meanwhile, the latter have become major exporters of manufacturing goods. Their share of world manufacturing output grew from 5 percent in 1953 to 20 percent in 1994. The world trade–GDP ratio has doubled to around 15–20 percent since the 1950s, and the emergence of global markets has been accompanied by the growth of global production and intra-industry trade. Currency transactions, meanwhile, doubled between 1989 and 1992, to $1.2 trillion per day – a figure much larger than the currency reserves of the world’s central banks (Garrett 1998a: 790). As a result of these and associated developments, it has been argued that the role of the nation-state in the international system has been fundamentally transformed. The “old” social contract – and its diverse institutionalization in Europe’s welfare states – was very much part of the “golden age,” during which there was a coincidence of national sovereignty (preserved with the help of exchange and trade controls), uninterrupted growth, and the Keynesian welfare state (Ruggie 1982). In the new “global 1

Part of the research on which this chapter is based was funded by the European Commission (DG5) and originally presented in Rhodes and van Apeldoorn (1998a).

87

Martin Rhodes

era,” so the argument goes, the state no longer has the capacity to defend competitiveness within economic borders or to sustain national social contracts and workers’ rights and entitlements. For those who accept that globalization is far from complete, the real culprit in constraining European social policy and welfare options is European integration, not just as a vector of globalization, but as the source of regulatory competition among the member states. National governments in the past could set taxes, establish labor regulations, and influence the framework for corporate governance. But an integrated European economy makes it fairly easy, it is argued, for companies to circumvent national jurisdictions. Thus, by shifting production abroad, moving their tax burden round the world, and transferring their corporate entities to other stock exchanges, firms can undermine the power of governments. The consequences for employment and labor market regulation may be profound: international and intraEuropean competition are fragmenting labor markets, upsetting the traditional balance of power between capital and labor, and demanding a far-reaching transformation of the conditions of employment if high levels of unemployment are to be reduced. For many neoliberals (and social democratic pessimists), these developments signal the death of the European “social model” based on long-standing commitments to low levels of inequality, solidaristic welfare policies, and protective systems of labor market regulation. The demise of the social partnerships that underpinned these arrangements is also expected to ensue. I maintain that that this argument is mistaken in a number of ways. It exaggerates the extent of globalization and its impact on the European economies and ignores the importance of domestically generated challenges to the institutional status quo. Furthermore, arguments that European economies are undergoing reorganization along neoliberal lines underestimate the capacity of their institutional arrangements for adapting and responding to internal and external pressures for adjustment. In particular, they ignore widespread evidence that the organized capitalist economies of Europe are already engaged in the search for new solutions that preserve commitments to solidarity and high levels of labor market regulation while also seeking to make welfare states more efficient and employment systems more flexible. As a result, changes in production systems in response to new competitive pressures do not necessarily entail a radical retrenchment of welfare states or deregulation of labor markets. Nor is there much evidence that countries with institutionalized social partnerships are abandoning them in favor of unilateral state or employer decision making. Meanwhile, 88

Globalization, Welfare States, and Employment

those countries where social partnership or corporatism have been nonexistent or weak are strengthening or putting such mechanisms for social consensus creation in place. This is not to argue that European countries are converging on a new domestic institutional equilibrium or that their welfare state and employment problems will thereby be solved. But if there is a European “Third Way” at the dawn of the twenty-first century, it will take the form of diverse social, economic, and institutional compromises between states and markets in which capitalism is cautiously liberalized but not unbound (Rhodes and van Apeldoorn 1998b). Part 1 of this chapter explores the scope for such an enterprise among the multiple pressures allegedly exerted by globalization. Part 2, which introduces the notion of “competitive corporatism,” argues that institutions for negotiated adjustment remain resilient and can assist in producing policies that aim to reconcile social justice and growth.

The Challenge of Globalization: Myth or Reality? Supporters of what can be called the “strong globalization” thesis point out that, while the nation-state clearly remains important, something clearly has changed in the relationship between the domestic and international political economies. The strains on social systems seem to reflect this. Support for strong welfare states appears to have declined (among elites if not among publics). Not only have cutbacks in social provision and the liberalization of labor markets taken place in many countries, but they are often justified in similar market-supporting terms. Furthermore, some would argue, the era of egalitarian and redistributive policies is clearly over. For the traditional project of the postwar Left now faces considerable obstacles, given the uncontainable power of global financial markets, the veto threat of multinational companies, and the emergence of a transnational neoliberal consensus, backed by MNCs and international organizations such as the OECD, International Monetary Fund (IMF), and World Bank. Even if union membership remains high in certain countries and social democratic parties are in power, this does not mean that the social democratic “project” can move forward or that the neoliberal onslaught can be halted. In response, a “globalization skeptic” would reply that many of the difficulties national policy makers have experienced with macroeconomic management – balancing budgets, mobilizing sufficient revenue to fund programs, and so on – have more to do with internal fiscal difficulties caused by recession than “globalization” tendencies. Prolonged economic 89

Martin Rhodes

recession coupled with demographic changes have required governments to raise more taxes to support the increasing number of jobless and citizens retiring from productive work. These pressures on the public purse are occurring at a time of stagnant living standards (Weiss 1997: 14). Pierson (1997) would add that welfare states have generated many of their own difficulties. By helping improve living standards and life spans, they have created new social needs that social services were not originally designed to meet. Meanwhile, other problems stem from the increasingly postindustrial nature of many advanced societies. As Rowthorn and Ramaswamy (1997) point out, this trend has little to do with trade but much to do with the natural maturation of industrial economies. Nevertheless, as in all academic debates, there is a middle ground to be found between the extremes. For this reason, it is worth taking the globalization arguments seriously while also ranging against them the appropriate counterclaims. The principal arguments concerning the impact (existing or potential) on European welfare states and labor markets are trade and competition, the globalization of finance, tax competition, and the role of multinational companies. I briefly survey these arguments below, considering in each case their importance for our understanding contemporary European developments.

Trade and Competition For every gloomy thesis there is an optimistic counterpoint. Free-trade optimists argue that trade integration with the South leads to efficiency gains as it releases workers in the North from low-productivity sectors where its firms can no longer compete. Rising imports of labor-intensive manufactures from the South may hurt workers in declining sectors, but this problem is seen as localized and temporary and can easily be dealt with by making labor markets more flexible. Imports to the North provide the South with the foreign exchange to pay for more sophisticated northern exports, also raising employment in the northern high-productivity sector. In sum, as a result of trade integration, employment in the North will shift from low- to high-productivity sectors, improving overall efficiency and aggregate welfare. If the optimists are right, there should be no long-term adjustment costs for advanced-country labor markets. The pessimists disagree. Wood (1994, 1995) argues that as skilled labor is relatively abundant in the North and unskilled labor abundant in the South, the latter will tend to specialize in the production (and exportation) of unskilled, labor-intensive products, while the North will specialize in 90

Globalization, Welfare States, and Employment

skill-intensive items, exporting them to the South. Thus, as trade integration proceeds, the relative demand for unskilled labor in the North will fall, leading to either lower wages (where wages are flexible, as in the United States) or higher unemployment (where they are rigid, as in Europe). The evidence is that in Europe the rising unemployment rate has been accompanied by a rising vacancy rate; unemployment is concentrated among unskilled or low-skilled workers; and thus unemployment seems to be caused at least in part by a rising mismatch in skills. This argument has not gone unchallenged (e.g., Sachs and Schatz 1994; Burtless 1995). One of the most influential counterarguments concerns the nature of the global market for labor. As Freeman (1995) argues, the notion that trade with the South is responsible for growing wage inequality and unemployment in the North not only assumes equal access to technology – so that trade is determined solely by factor endowments – but also a fully integrated, single global labor market in which demand and supply set the price of all labor. In fact, labor is not homogeneous: better qualifications command a return to human capital and protect many workers in industrialized countries from lower pay. The way capital is used also protects workers against leveling down: better infrastructure, including communications and research and development all raise returns to labor. Specialization in the production of different goods among traditional countries and their new trading partners reduces downward wage pressures in the former (see Thygesen, Kosai, and Lawrence 1996). There is also an argument that trade is much less important than technological change in producing a pervasive shift in labor demand from lower to more highly skilled workers (Slaughter and Swagel 1997). While there may be an interaction between globalization and technological bias, the bias in technological progress toward the skilled predates the globalization of the last decade or so. Does this mean that we can discount the trade and competitiveness argument? Not entirely, for it is inconceivable that trade competition is unconnected to technological change or that, in combination, they have few implications for labor markets, employment, and social policy. Snower (1997) links trade competition, skill-biased technological change, and “the organizational revolution” in explaining problems in sustaining the welfare status quo, greater wage dispersion, and rising unemployment in the industrialized countries. Four developments are seen as crucial: the reorganization of firms into flatter hierarchies; flexible, computerized manufacturing, allowing production decentralization and the adoption of “lean” 91

Martin Rhodes

and “just-in-time” methods; dramatic changes in the nature of products and in seller-customer relations; and the breakdown of traditional distinctions between “skilled” and “unskilled” workers as employees are given multiple responsibilities. By making existing jobs less secure, and making traditional skills – and those who possess them – redundant, these developments are likely to create greater reliance on unemployment insurance, public support for education and training, and a wide variety of welfare state services. This may generate what Snower calls “the quicksand effect” – the phenomenon whereby welfare structures designed for a different era become weighed down and generate negative effects, destroying incentives and making redistributive policies inefficient, while the productivity of welfare services declines and their cost increases. This argument, if correct, gets around the frequently made objection to the trade competition thesis that only around 14 percent of European trade is with the rest of the world and is therefore relatively unimportant. For such trends could also be driven by increases in intra-European trade (now at 70 percent of Europe’s total, compared with only 9 percent with Asia) (Fligstein 1998). But what are the policy implications? As for welfare structures, many countries are now engaged in policy innovations to reduce the quicksand effect of social policy provisions by tackling the costs of transfer-heavy welfare states and redesigning benefit formulas so as to make them both financially sustainable and more employment friendly. As for employment, the outright deregulation of labor markets is clearly not the answer (a topic I return to below). For while a moderate degree of wage flexibility might help contain the high rates of unemployment in Europe, the main emphasis has to be on the unskilled or those whose skills are becoming redundant. As Scharpf (1997, 1999) argues, the big employment shortfall in many parts of Europe is in the services sector and especially in personal services that use much less skilled labor and where tax wedges may deter new recruitment. Upgrading the skills of the workforce, and hence gradually shifting the relative supply of skilled and less-skilled labor, is also important. One way to do this is to change the relative price of skilled versus unskilled labor by greater wage differentiation, while avoiding greater inequality by engineering changes in tax and transfer systems. Another is to lower high indirect labor costs differentially, so that these costs are reduced most sharply for the lowest paid and then phased in gradually for those earning more. As argued in the next section, this is precisely the type of reform that many European countries are now

92

Globalization, Welfare States, and Employment

engaged in. But it requires a complex set of trade-offs and a negotiated approach to adjustment.

The Globalization of Finance For many, the globalization of financial markets is the main culprit in explaining the reduced autonomy of national governments and their inability to respond in progressive ways to new social policy and employment challenges. Many analyses of economic policy autonomy depart from the so-called Mundell-Fleming theorem or the “unholy trinity” which states that exchange rate stability, capital mobility, and domestic monetary independence cannot be achieved simultaneously (Andrews 1994; see also Frieden 1991; Webb 1991; Goodman and Pauly 1993; Perraton et al. 1997). The argument is that in the 1960s – when capital mobility was still limited – payments imbalances arising out of divergent macroeconomic policies were still small enough to be manageable. Now that capital is internationally mobile, this is no longer so: external positions can only be stabilized by coordinating the aforementioned policies themselves. If capital is highly mobile across borders, then some policies will be rewarded while others will be punished. Thus, if a deficit country decides to pursue an expansionary monetary policy in relation to other countries, it will experience pressure either to devalue or to defend its currency (depleting its foreign exchange reserves) (Andrews 1994). The main implication for social policy and employment is that in an environment of international austerity, it becomes increasingly costly for individual states to pursue an expansionary monetary policy as part of an effort to stimulate growth and employment (Eatwell 1995). A second argument concerns the shifting balance of power. Now that capital controls have been lifted, new and privileged interests have been embedded in the policy-making process to the extent that reinstating them has become very difficult. Meanwhile, new coalitions have been formed “revolving around the differential effects of greater openness,” a reaction in part to the changing opportunity costs created by the benefits of participation in the international economy (Keohane and Milner 1996a). Frieden (1991) argues that because capital mobility has given investors a wider set of choices, governments are forced to provide more favorable conditions to capitalists. Increased financial integration has increased the social and political power of capital – particularly among capitalists with

93

Martin Rhodes

mobile or diversified assets (such as multinational companies [MNCs]) – and disfavors those with assets tied to specific locations and activities such as manufacturing. Those most affected by such developments are purportedly left or social democratic governments devoted to policies to bolster employment growth and defend welfare spending. A regression analysis of the interactive effects of economic internationalization and left labor power on economic policies for fifteen OECD countries carried out by Garrett (1995, 1998b) concludes that the “propensity to deficit-spend” – which he calls “the political-economic sine qua non of social democracy” – has not been constrained by increasing trade and capital mobility but rather strengthened with greater internationalization. On the other hand, he also points out that the integration of financial markets has put a premium on leftwing policies in terms of higher interest rates, something that in the future might still pose problems for social democracy. As Keohane and Milner (1996b) also suggest, there is evidence that financial market integration or capital mobility has a potentially more detrimental effect on the policymaking autonomy of left labor governments than trade integration. All of this might suggest that the game is up for countries wishing to defend or reinforce social solidarity in an era of global finance. But as Glyn (1995) has argued, the real issue here is “credibility” with the markets. Globalization has often simply been made the scapegoat for the failure of certain nation states (Britain in the 1970s, France in the early 1980s, Scandinavia in the early 1990s) to control their domestic sources of social conflict and spending – an argument also applied to the Scandinavian social democracies by Notermans (1993). After all, if budgetary capacity and price stability permit it, there is no reason why increases in welfare spending or equitable modifications to the welfare status quo should be penalized and retrenchment rewarded. The real task is to build or rebuild domestic coalitions and arrangements containing trade-offs that make such policies “credible” with the markets (as well as with national capital). Again, as argued below, although not easy, this is precisely the path being taken by many European countries.

Tax Competition Tax competition should also be taken seriously as a challenge, although again not necessarily as an ineluctable force making national governments impotent. Tanzi (1996) and Rodrik (1997) – among others – have argued 94

Globalization, Welfare States, and Employment

that governments can no longer maintain let alone expand their generous welfare state–progressive taxation mix owing to increasing tax competition, and that the future of the welfare state requires shifting the tax burden from mobile (firms and financiers) to immobile (labor) asset holders, thereby emasculating its redistributive effects. Globalization thus induces a process of “tax degradation” – and a tendential “race to the bottom” – by enabling individuals and especially firms to exploit the existing differences between national tax systems. Could this be the principal mechanism eroding systems of social and income protection? Research (e.g., Swank 1992) has shown that between the 1960s and mid-1980s, those economies most exposed to international markets exhibited a tendency (conditioned by partisan control of government and party policy preferences) to respond to ebbs in investment policies with reductions in levels of corporate taxation. Left-wing governments seemed the most sensitive, especially with regard to investment in new productive capacity. One explanation for this is that left-wing governments need to compensate for the negative effects (on the investment climate) of their other policies, for example, an extensive social security system, redistributive policies, and labor market regulation. Thus social democratic governments may be caught in a contradiction: in trying to boost employment-creating investment they must also shift the tax structure to the disadvantage of immobile employees. More recently, however, Swank (1998) and others have argued against the claim that increasing international capital mobility has generated a shift away from capital taxation, and that governments have continued to rely on corporate taxes for a significant, and even in some cases, a growing, share of revenue. The considerable changes that have occurred in the structure and rates of taxation have coincided with attempts to make the overall changes revenue-neutral or otherwise protect the revenue needs of the state. Thus, rate cuts have been offset by broadening the tax base and eliminating investment-related allowances, credits, and exemptions (Swank 1998: 678–79). Nevertheless, the implications for policy are arguably still important. Clayton and Pontusson (1997) argue that whatever the direct distributive consequences of falling rates of corporate tax, they make life harder for social democratic governments by reducing the scope of using the tax code to boost investment and employment. Thus, income taxes have become less progressive, corporate tax rates have fallen, and the burden of taxation has tended to shift toward indirect taxes such as value-added taxes. 95

Martin Rhodes

With the flattening of corporate tax schedules, governments have reduced their ability to influence the volume or character of private sector investment in line with egalitarian policy objectives. For as Ganghof (2000) argues, while revenues from capital income taxation may not have fallen, on average, they have also not increased. This contributed to the fact that the average total tax ratio of the eighteen most advanced (and largest) OECD countries has stagnated since the mid-1980s and has not kept up with increasing public expenditures. Tax competition is not the main cause of the financing problems of the welfare state – if only because taxes on capital contribute only a minor share of revenue in most countries. But it may have constrained policy responses by making some forms of revenue raising more costly at a time when the maturation of welfare programs (see the next section) has increased spending. One solution in the face of greater capital mobility in the future might be to tax more heavily spending with an unavoidable physical presence, namely, property.2 Another – perhaps the most popular alternative – would be to shift the burden to consumption, including levies on energy use, pollution, and other “green” taxes. But as Scharpf (1997) argues, this would be regressive and also destructive of low-productivity service employment. There is thus a considerable need to rethink a range of fiscal policies and their relationship with social security expenditure. Again, this is inimical neither to negotiated paths of adjustment nor an equitable recasting of welfare states. Labor market reform and responding to high levels of unemployment require simultaneous changes to social security, tax, and pensions systems: and where it is logical to negotiate these reforms through corporatist-style bargaining, this is now being done. I return to this point later.

The Power of Multinationals What of workers’ rights and entitlements? The strong globalization thesis argues that these are being driven down by regulatory competition, as transnational companies shift production abroad and pitch the lower labor costs and wages of developing-country workers against their better-paid industrialized country counterparts. While there is evidence that some MNCs shift production to take advantage of lower labor and other social costs, this argument underestimates the complexity of MNC strategies and overestimates the footloose nature of these companies and their produc2

See “Disappearing Taxes,” Economist, May 31, 1997, 17–19.

96

Globalization, Welfare States, and Employment

tion process. In fact, the idea that there has been a massive shift to lowcost locations is not borne out by statistics. Weiss (1997: 10) presents convincing evidence that MNCs do not by and large invest where wages and taxes are lowest: between 1967 and 1991, the total world stock of foreign direct investment (FDI) located in high-wage and relatively high-tax countries increased by 12 points to 81 percent. Moreover, as Ruigrok and Van Tulder (1995) convincingly demonstrate in a study of the internationalization strategies of the world’s 100 largest companies, not one of these can be dubbed truly “global,” “footloose,” or “borderless.” In addition, “domestic bargaining arenas in all parts of the world impose huge social, political and financial constraints, thus setting the outer margins of a core firm’s internationalization strategy.” For key actors in the development and diffusion of new technologies, the “national rootedness” of MNCs remains a vital determinant of where future innovation takes place (Pauly and Reich 1997: 24ff.). The power of MNCs to “arbitrage” diverse national structures and force deep, structural convergence across diverse societies is, they argue, “chimerical.” Nevertheless, even if it does not relocate all parts of its production and design process, the internationalized firm can use the threat of exit to modify contractual relations in the labor market, with important implications for employment and, indirectly, social policy. German firms, for example, have used locational threats increasingly to weaken the power of unions and force concession bargaining (Mueller 1996). As a result, German companies and unions are now agreeing on patterns of flexible working that were unthinkable just five years ago, and many of these are reached locally with company works’ councils, thereby circumventing national accords. But the national union, IG Metall, is also sometimes involved, and most of the deals reached modify only marginally the status quo. Further-reaching changes might be expected in countries such as Ireland where in 1996 foreign (mainly American) companies accounted for 47 percent of manufacturing employment, 66 percent of the sector’s gross output, 77 percent of its net output, and 84 percent of industrial exports (OECD 1999: 54). Nevertheless, multinationals in Ireland have by and large complied with that country’s industrial relations rules and have not sought to counter the advances made under its various social pacts, including most recently (May 1999) an agreement between unions and employers to counter a growing trend for union derecognition. In the end, the role of domestic institutions and actors will be critical in conditioning not just the embeddedness of the MNC in its home base, 97

Martin Rhodes

but also its strategies of foreign direct investment and the relations it forms in new production locations – including those with workers and unions. The critical point here is that, as Lange and Scruggs (1997) show, a strong union context is unlikely to be undermined by FDI, although MNC investment will clearly not assist in building a strong union culture or organizational base where it does not already exist. Not only do open economies seem able to sustain strongly unionized environments, but corporatist bargaining structures remain strong in those economies increasingly exposed to international forces by growing economic integration. As shown by the persistence of centralized or coordinated bargaining in Scandinavia (Wallerstein and Golden 1997), and by the emergence and consolidation of new forms of social partnership elsewhere in Europe (Rhodes 1998a), the effect of internationalization may well be the opposite of what is generally assumed. As argued below, far from disaggregating organized interests and allowing business to break free from constraining policy arrangements, it seems in many instances to be reinforcing them.

Reforming European Welfare and Employment Systems: A Future of “Competitive Corporatism”? The discussion to this point has tried to put the globalization arguments in perspective. Its conclusions are broadly sympathetic with Pierson (forthcoming), who argues that “although it is important to recognize the linkages between international and domestic developments . . . such links are likely to be more modest, complex and bi-directional than is commonly suggested.” However, as argued earlier, these links do exist. Together, trade competition, the liberalization of exchange control and the internationalization of financial markets, tax competition, and the power of multinationals may be reducing the autonomy of the state, changing the balance of power between capital and labor and making welfare politics more problematic. Traditional national bargains on welfare and labor market regulation are therefore under stress. But the influence of these factors has to be considered in conjunction with domestic sources of strain and within the context of the national institutions that continue to filter external pressures for change.

Internal Pressure and the “New Politics” of Welfare and Employment Although apparently “in crisis” owing to the combination of factors mentioned above, welfare states have changed relatively little in recent years. 98

Globalization, Welfare States, and Employment

According to an OECD investigation of 15 cases of fiscal consolidation involving 11 countries, in all but one (Ireland 1986–89), governments pursuing consolidation increased revenues as a percentage of GDP and in 9 of the 15 cases, revenue increases accounted for the majority of fiscal consolidation. Transfer spending fell in only 7 cases, and generally by very modest amounts. Only 3 of 15 cases reduced transfer spending by more than 1 percent of GDP, while 6 achieved that large a cut in public consumption, and 7 in investment spending (Pierson 1997; OECD 1996). In fact, to date, there is little sign of a major onslaught on European welfare states, except for Britain under the Thatcher/Major governments – and even there retrenchment in spending was restrained. Support for the welfare state remains high among publics everywhere, and institutional stickiness creates tremendous resilience against pressures for change. The most important changes in Europe have been increased managerialism (in line with a new and spreading orthodoxy of public sector management); attempts to make certain benefits (especially unemployment support) more “incentive compatible”; a marginal degree of privatization (mainly in health, and mainly in Britain), and some decentralization; and attempts (largely successful) to control budgetary expansion (see Therborn 1997 for a survey). This hardly amounts to “neoliberal” convergence. As far as the scale of welfare expenditure (public spending as a proportion of GDP) is concerned, the pattern across continental Europe still conforms by and large with Esping-Andersen’s (1996) image of a “frozen landscape.” However, in institutional terms, governments have been introducing changes within their systems to deal with perceived cost pressures by experimenting with finance, the organization of provision, and access to benefits. While this has not undermined the core of the major programs for social protection, important changes are now occurring, in many cases as the result of complex new bargains. A recasting of the traditional welfare bargains of Europe is essential, for the “frozen landscape” metaphor indicates sclerosis rather than a stable and sustainable equilibrium. Pierson (1998) has construed the crux of the problem in terms of “irresistible forces” (postindustrial pressures) meeting “immovable objects” (strong public support and veto points). Thus, the relative growth of the service sector implies lower productivity growth and entails either greater public spending or increased wage inequality if new jobs are to be created. The maturation of governmental commitments and population aging demand reforms to health care provision and old age pensions (in 1992 these accounted for 80 percent of all social protection 99

Martin Rhodes

outlays in the European Union) if costs are not to escalate and employment creation is not to be undermined by higher direct taxation and/or payroll taxes. Then there is the unemployment problem itself. The task of responding to persistent high rates of unemployment across the European continent (compared with high rates of job creation in the liberal welfare states of the United Kingdom and United States) reveals how different national systems have to face the sometimes adverse consequences of existing social contracts. In the past, each “welfare regime” created its own employment pathology. In Scandinavian countries, the distributional costs of generous social contracts were met by those in employment who paid high taxes for an overdeveloped public sector to soak up the potentially unemployed. In continental Europe, governments, employers, and labor unions more or less agreed that the price of adjustment should be shouldered by the unemployed, comprised largely of younger, female, and older workers. In southern Europe, an acute “inside-outsider” problem has developed as a result of the fragmentation and disparities in the income support system for those without work, with large differences in the level of protection given to core and marginal workers. All three continental regimes are haunted by the specter of “the service sector trilemma” (Iversen and Wren 1998) in which the goals of employment growth, wage equality, and budgetary constraint come increasingly into conflict (see also Esping-Andersen 1996; Scharpf 1997, 1999). If service sector employment is to be generated in the private sector, this entails adjustments to wage and nonwage costs for the less skilled and potentially greater wage inequality. Creating such employment through the public sector (the traditional Scandinavian solution) entails increased budgetary pressure at a time when deficits and higher taxes are definitely out of fashion. The alternative to doing neither would appear to be continuing if not rising unemployment. The severity of Europe’s unemployment problem is producing a consensus that part of the solution will come from changes to the regulations that cover both employment and the funding of welfare. If income inequalities are not also to increase, such changes must be accompanied by other innovations such as reforms to taxation (e.g., the introduction of negative income taxes) and social security arrangements alongside a reregulation of the sheltered sectors of continental economies, to price people with low or nonexistent skills into work. But electoral incentives, “institutional stickiness,” and the veto points created by powerful vested interests devoted to defending transfer-heavy welfare states and their redistributive 100

Globalization, Welfare States, and Employment

outcomes make anything other than incremental reform very difficult. Reforms to health care systems, pensions, and labor markets all require a careful process of adjustment if social cohesion as a governing principle of these systems is not to be sacrificed and if core constituencies and their representatives (welfare professions, the labor movement, citizens) are not to erect insuperable impediments to change. However, advocates of the globalization thesis would argue that the disorganizing effects of greater competition, post-Fordism, porous economic borders, and the enhanced power of capital make such consensus-based solutions impossible. Are they correct?

The Scope for Negotiated Adjustment As shown by recent experiences, potential blockages in the process of reform are being avoided in some countries by the creation of new coalitions behind the reform agenda, most notably through new types of concertation and negotiation. Indeed, one of the most important, but until recently largely neglected, aspects of the new politics of welfare in Western Europe concerns the transition in some countries toward new nationally negotiated social pacts. I have referred to these as “competitive corporatism” (Rhodes 1998a) to distinguish them from traditional forms of social corporatism and to signal their competitive rationale. Those in Ireland, the Netherlands, Portugal, and Italy have attracted most attention, although others can be identified in Scandinavia (albeit against the background of a more institutionalized corporatist tradition), while still others appear to be emerging in more hostile political environments (Spain). According to much analysis of state capacity and the potential for coalition building in contemporary Europe, such innovative forms of corporatism should not be possible. Grahl and Teague (1997: 418) argued that “there is no reason to doubt that, as a strategic programme for the resolution of employment issues, neo-corporatism is moribund – defeated on the ground by the actual evolution of employment relations before reluctant abandonment by its academic proponents.” Lash and Urry (1987) put forward a persuasive thesis concerning the decomposition of traditional institutions in the transition to post-Fordism, and their recomposition on a more flexible, decentralized basis. Streeck and Schmitter (1991) argued that a combination of the business cycle effect and European integration would remove the logic that had underpinned successful corporatism. An integrated European economy, with less room for discretionary national 101

Martin Rhodes

macroeconomic policies, would reduce the incentives for unions to organize collectively and deliver wage restraint in return for package deals or side payments. Gobeyn (1993: 20) argued that “contemporary economic realities . . . make corporatism largely unnecessary. Market forces alone can presently achieve labor discipline and wage demand moderation.” Indepth analyses of the experiences of particular countries apparently confirmed the unidirectional nature of such developments and the role and responsibility of globalization behind them. Thus, Kurzer concluded her 1993 study of Belgium, Austria, Sweden, and the Netherlands by stating that social concertation was no longer feasible: “high capital mobility and deepening financial integration prompt governments to remove or alter institutions and practices objectionable to business and finance” (Kurzer 1993: 244–45). It is also argued (and not just by right-wing economists) that the labor market rules and welfare state organization linked to such concertation have now become dysfunctional for growth; and that institutions that underpin them are in the grip of such deep uncertainty that the importation of the United Kingdom’s neoliberal deregulation policies to the rest of Europe should not be discounted (see Grahl and Teague 1997; Teague and Grahl 1998). Much of this type of argument flies in the face of the evidence. The fact of the matter is that, at the dawn of the twenty-first century, all of the countries referred to by Kurzer – plus many of their neighbors – retain critical elements of policy concertation between capital, labor, and the state and/or various degrees of centralized or coordinated wage bargaining. This tendency to exaggerate and extrapolate the crisis and collapse of corporatism is due in part to the “Swedish syndrome.” Just as Sweden was long seen as the example par excellence of the corporatist welfare state, so more recently it has been considered a harbinger of things to come. For the collapse of Swedish centralized wage bargaining in the 1980s suggested the imminent demise of a “model” combining a strong, decommodifying welfare state with low levels of income dispersion. If Sweden could not survive globalization and the increased services orientation of the economy, then which country could? In fact, corporatism is alive and well and living in some unsuspected places. But why should this be the case, given that – at least until recently – even the “high priest” of corporatism, Philippe Schmitter (1989), was convinced of its decline? There are a number of reasons. First, arguments predicting a widespread transformation of the European economies were simplistic and ignored three essential limits on radical change in both 102

Globalization, Welfare States, and Employment

policies and institutions: path dependence, the prevailing distribution of organizational power, and the “efficiency” deriving from the established links between competitiveness, systems of macroeconomic regulation, and the microstructures of industries and firms (see Gourevitch 1996). Path dependence and resistance to change derive from the fact that forms of labor market regulation are deeply embedded in national systems of law and collective bargaining. The fact that the existing distribution of power in continental systems (with the exception of France) is located in the sustained strength of both employers’ and trade union organizations, often organizationally linked to political parties in coalition governments, clearly limits and conditions the scope and content of reform and socioeconomic transformation. Second, far from being a “fair weather” creature, as many had argued, corporatism has come into its own in many European countries during the recent years of recession and high levels of unemployment. This suggests that the “business cycle” theory of corporatism (reiterated in Schmitter and Grote 1997) is flawed in postulating phases of decline followed by renaissance in traditional corporatist practices. The truth is that many of the factors that were supposed to topple the corporatist edifice have in fact supported it through periods of renewal and modification, leading sometimes to substantial transformation rather than return to a preexisting equilibrium. A fragmentation of interest associations has occurred in some cases as socioeconomic change and restructuring has undermined traditional memberships. There have also been attempts to decentralize systems of wage bargaining, or to differentiate wages from national pay norms at lower levels. But generally, unions and employers in continental Europe have acted to prevent these developments from endangering either national bargains or the institutional structures that underpin them. Employers’ organizations and trade unions are both aware of the costs involved in the outright loss of public goods such as pay coordination and the dangers of free riding. For decentralized systems are notoriously prone to wage drift and inflation – which can only be countered by a rising level of unemployment – and international competition makes wage cost containment an imperative (Rhodes 1998a, 1998b). Nor, contrary to expectations, has flexible specialization seriously threatened corporatist arrangements or collaborative industrial relations: first because its extent has been exaggerated and second because where it has been implemented it has actually increased the need for cooperative labor relations and the collective provision of vocational skills. As Negrelli 103

Martin Rhodes

(1997) notes in the Italian case and elsewhere, modern human resources management can be combined with traditional industrial relations practices, creating possibilities for mutual enrichment between these two different systems for organizing the workplace. This complementarity derives from the beneficial trade-off between job security and flexibility at work, producing positive effects for industrial adjustment. It is also wrong to assume that employers have uniformly attempted to use restructuring to weaken trade unions: although this has certainly been the case in Britain, industrial relations remain distinctively consensual in countries with a tradition of social partnership. Despite its difficulties, German Mitbestimmung, for example, retains its functional value. As in the case of pay bargaining, it is not so much that employers are constrained by unions to participate in these arrangements, but that they are generally aware of the great sense in doing so in terms of their own interests. As Thelen (2000) argues, most German employers are acutely aware of the costs to themselves of a decentralization strategy that seeks a radical shift in the balance of power with labor. For similar reasons, other countries – to be discussed in the next section – which seek to follow a “high road” (i.e., highwage/high-productivity path) of economic adjustment may consciously seek to develop similar coordinating capacities in their own economies. As for economic performance, continental systems retain enormous economic strength, and their large and successful export sectors and trade surpluses demonstrate that their regulatory systems – including labor market rules – are far from dysfunctional, at least for their traditional growth paths. Moreover, there is considerable empirical evidence for a strong and positive correlation between such “cooperative economies” and high rates of productivity and investment growth, whereas “conflictual economies” (conventionally including Canada, the United States, and Great Britain) have traditionally lagged behind in both respects (for a survey and analysis, see Gordon 1996). It is certainly the case, though, that the continental European countries have performed poorly in terms of employment creation in recent years. Fluctuations in the European economic cycle have left larger numbers unemployed whenever there has been an upturn in the cycle (“hysteresis”). This suggests that employment creation lags behind growth and that the fruits of new growth are not evenly shared between insiders and outsiders (see Ormerod 1998) – a situation much more acute in some countries than others. An acute insider/outsider dualism in the labor markets of some Southern European countries, for example, has been created by over104

Globalization, Welfare States, and Employment

protective regulations for those in full-time, standard employment (see Siebert 1997). But combinations of incremental reform in labor market rules and social security systems, plus certain policies encouraging a redistribution of work (some forms of work sharing, for example) can help mobilize those sections of the unemployed workforce left behind by a return to higher levels of growth. In addition, there may have to be some selective deregulation of the labor market to enhance flexible (i.e., parttime or temporary) service sector employment, and this will form an important component of labor market strategies for many continental countries. But as Hall (this volume) argues, there is no reason to expect that this will push Europe’s organized, cooperative economies down the slippery slope to Anglo-Saxon–style deregulation and inequality. Indeed, periods of high unemployment and painful restructuring in the trough of the cycle seem to have bolstered the search for consensual solutions. The best way of tackling the unemployment problem is via negotiation, not the unilateral imposition of looser regulation à la Thatcher, coupled with awareness that not all protective employment regulations deter employment creation. Of particular importance is the way in which optimal forms of labor market regulation require collaborative industrial relations as well as corporatist bargains to cement them. As Nickell (1997) has demonstrated, a number of protective measures that are generally assumed to impede employment creation may in reality have little effect. These include employment protection measures and general labor market standards, generous unemployment benefits (as long as they are accompanied by strict benefit durations and measures to help the jobless back into work), and high levels of unionization and union coverage (as long as they are offset by high levels of coordination in wage bargaining). Selective deregulation, leading to an expansion of part-time employment, has been achieved in the Netherlands, for example, within the context of a broad social pact sustaining the broader system of labor market regulation, while also minimizing the impact on disparities in real income (Visser and Hemerijk 1997; Van den Ploeg 1997).

The Emergence of “Competitive Corporatism” Thus, corporatism has not gone away, either as a mechanism for coordinating wage bargaining or for negotiating wider policy options. Indeed, recent years have witnessed the preservation of forms of corporatism in countries where it has always been strong, as well as its emergence in those 105

Martin Rhodes

where the traditional prerequisites (e.g., strong, centralized, hierarchically ordered interest associations) have been weak. As for the countries of traditional “social corporatism,” as argued by numerous analysts (Wallerstein and Golden 1997; Dølvik and Martin 1997), Sweden turned out to be the exception. The collapse of the Swedish centralized bargaining system may have had something to do with globalization and the creation of new coalitions within the traded and nontraded sectors. But domestic politics played a particularly important part (specifically, the conflict between employers and unions over new labor laws and wage-earner funds in the late 1970s and early 1980s). By contrast, bargaining remains “coordinated” in various ways in Finland, Denmark, and Norway. In Denmark, where since 1987 wage negotiations have been conducted without confederal participation, the 1990s have seen the emergence of five large bargaining cartels, with even greater controls on plant-level bargaining than in the past. In Norway, centralized negotiations were reestablished in the late 1980s after a period of industry-level bargaining. In 1992, following the attribution of legal status to wage bargains struck between the social partners at the national level, the so-called solidarity alternative (solidaritetsalternativet) was introduced in the form of an informal understanding between the government and the trade union confederation, LO. Its aim – which has been achieved – encourages moderate pay increases in order to strengthen the competitiveness of Norwegian companies and reduce unemployment. Its success in bringing down inflation and unemployment levels has led the Swedish government to find ways of emulating it. In Finland, a reunification of blue-collar unions in the late 1960s, linked with government participation in tripartism, has increased centralization (Wallerstein and Golden 1997). Meanwhile Austria – whose economy has achieved higher levels of openness and tertiarization than even Sweden’s – has also remained highly corporatist, although with considerable flexibility in its wage-bargaining structures (Traxler 1997a; Wallerstein, Golden, and Lange 1997). As for countries in which a defunct corporatist tradition has been revitalized, or a new one put in place, the Netherlands takes pride of place. After an interlude of industrial relations strife, the mid-1980s saw a revival of Dutch corporatist policy making – again with flexible, decentralized bargains within a coordinated structure. This has produced something of a model for advocates of a “third way” between highly regulated continental Europe and the more deregulated economies of Britain and the United States. Dutch social concertation has provided the basis for industrial 106

Globalization, Welfare States, and Employment

relations peace, wage moderation, and an ongoing process of labor market reregulation while preventing an increase in inequality and combating unemployment (now one of the lowest in the OECD). There have also been agreements on social security contributions, work sharing and industrial policy, training, job enrichment, and the development of “entry-level” wages. In 1995, rights for temporary workers were strengthened (the so-called flexicurity agreement) in return for a loosening of dismissal protection for core workers (Visser and Hemerijk 1997). In Ireland, a rather comprehensive social pact negotiated in 1987, 1990, 1993, and again in 1997 has addressed tax, education, health, and social welfare issues in addition to incomes. The emphasis has been on inflationproof benefits, job creation, and the reform of labor legislation in the areas of part-time work, employment equality, and unfair dismissal. Pay rises have been subject to floor and ceiling levels. In return, trade unions have delivered industrial relations harmony. In Italy, negotiations in the early 1990s that initially focused on reforming Italy’s automatic wage-indexation system were extended to include the rationalization of bargaining structures, the reform of union representation in the workplace, improvements to the training system, the legalization of temporary work agencies, assistance for the unemployed to enter the labor market, and, the most significant step, the May 1995 agreement between the unions and the government on pension reform. In Portugal, there have been five tripartite pacts since 1987 focusing on incomes and social and labor market measures, including payrise ceilings and loosening rules on the organization of work, the termination of employment, and the regulation of working hours. The 1996 agreement also covers social security issues, including a minimum wage, a reduction of income tax for low-income groups, and a more favorable tax treatment of health and education benefits and old age pensions. In Spain, a long period of unilateral employer and government decision making (and considerable blockage in welfare and labor market reform) has given way to a new national bargain – the Toledo Pact. This agreement has allowed progress (albeit rather limited) on pension reform and (more significant) innovations in labor market flexibility (Rhodes 1998a). If we take the objectives of these pacts, and consider their final goals, we can construe them analytically in terms of the constitution of two types of coalition – distributional coalitions and productivity coalitions – and the attempt to forge effective and enduring political and functional linkages between them. Although differing, sometimes markedly, from one country to another, these pacts derive precisely from a common attempt to link 107

Martin Rhodes

negotiations over both the formal and informal welfare states (i.e., formal welfare programs and the labor market component of the social wage) with more general policies to bolster competitiveness. Some countries (the Netherlands, Denmark) are simultaneously building both types of coalition. Others have been fairly successful in creating a new productivity coalition (Ireland) but have been less successful – although progress is being made – on the distributional side. Italy is struggling to find the distributional support for further reform of the pensions system (given that half of the membership of all three of its major union confederations are pensioners!), but important gains have been made in negotiating productivity gains in industry. Although slower and perhaps less incisive reformers, all of these countries have thus far avoided the dramatic increases in social inequality that have accompanied the radical reforms imposed unilaterally in the liberal welfare states of the United Kingdom and New Zealand. We can hypothesize that the most enduring examples of competitive corporatism will be those that create the most complete coalitions of both types and forge the most functional links between them – functional, that is, for a politically acceptable and legitimate trade-off between equity and efficiency. Importantly, even those countries without the prerequisites of traditional social corporatism – that is, strong, well-organized, and centralized employers and union movements – can engage in this process of exchange. As the examples of Italy and Ireland show, broadening the bargain to include industrial and structural policy with employment policy, and making the institutionalization of cooperation between the parties an element of negotiation, can provide the functional equivalents of traditional corporatist preconditions (Traxler 1997b). Moreover, partisan politics appears to be becoming less important for such pacts. In the Irish, Portuguese, Spanish, and Dutch cases, centrist, socialist, and center-right parties have governed at various points in the process of building or consolidating concertation. We can also therefore hypothesize that, partly in response to external pressures – particularly the constraints of Economic and Monetary Union (EMU) – the economy is becoming much less central to ideological conflict and party politics in these countries than hitherto (see Rhodes 2000 for a more extensive analysis of these developments). What sort of policies do these new pacts embrace? The most important ideal objectives would appear to be a coordinated decentralization of bargaining structures and an ordered reregulation of the labor market,

108

Globalization, Welfare States, and Employment

accompanied by adjustments to tax and social security systems to avoid a reduction in real incomes or social security protection. Although they tackle a wide range of issues – spanning incomes policy, social security, and tax reform – employment and the labor market provide examples of ways in which new techniques and new modes of negotiation have proven to be critical for policy innovation and policy success. The debate on social protection and employment has focused on several types of reform. These include modifying the funding of welfare by shifting the burden of costs, for example, away from payroll taxes to general taxation; removing tax wedges and eliminating poverty traps; introducing wage subsidies in various forms to employers and “in-work” benefits; and, more specifically, the “activation” of so-called passive benefits. Examples of radical proposals in this area are conditional negative income taxes (conditional on, e.g., evidence of serious job search by an unemployed person); and benefit transfer programs (e.g., providing individuals with vouchers that could be offered by the unemployed to firms that would hire them, reducing correspondingly the amount of public money spent on traditional forms of unemployment benefit). Denmark, the Netherlands, and Ireland have perhaps gone furthest in experimentation in these areas, and in all three cases reforms have been introduced within a general context of policy concertation. Why is this happening? Part of our understanding of these corporatist or quasi-corporatist developments must come from an acknowledgment of the contradictory – although quite logical – tendencies at work. On the one hand, as many who forecast the demise of corporatism correctly pointed out, decentralization in formerly centralized industrial relations systems has been induced by a combination of factors. The new international division of labor within large transnational firms and the introduction by multinationals of “alien” elements into national bargaining arenas has been a disturbing factor. Cross-class “flexibility” alliances between employers and workers have eroded social corporatist systems to some extent, inducing a shift to a more sectoral or company-based form of bargaining. Meanwhile, employers in all systems are searching for greater company- and plant-level flexibility in three areas: internal (or functional) flexibility in the workplace, external (or numerical) flexibility in relation to the wider labor market, and greater pay flexibility at local levels. At the same time, the creation of the single market and the realization of EMU have placed new pressures on wage-cost competition and made

109

Martin Rhodes

competitive devaluation impossible, reinforcing other pressures for greater pay flexibility and productivity bargaining. However, there are also pressures in favor of centralization – as well as high levels of national (and European) employment protection; and some of these derive from the very same forces that are ostensibly “disorganizing” the advanced capitalist economies. As argued earlier, in response to competitive pressures, the diffusion of new forms of “best-practice” human resource management and work organization may require the creation or maintenance of cooperative labor relations and a high-trust firm environment in many countries. And contrary to the general assumption that new types of organizational flexibility inevitably strengthen capital at the expense of labor, employers and workers can become increasingly interdependent in such systems. Well-designed labor market rules also remain essential in this context. The optimal world of internal flexibility is built not by unilateral management action but on teamwork and low levels of hierarchy in firms. It also depends not just on high levels of specific skills but also on the acquisition of more generic skills within national education and training systems. Too high a level of external flexibility – that is, the absence of regulatory constraints on firms – destroys trust and undermines internal flexibility. This trade-off – producing a productive form of “regulated cooperation” – is a critical one for sustaining both competitiveness and consensus in European labor markets. Furthermore, cost competitiveness and monetary stability – as well as “credibility” with international financial markets – require a means of preventing wage drift and inflationary pressures. This has focused the attention of governments on revitalizing incomes policies, a critical component of many if not all of these new national bargains. This imperative, for example, lies behind recent attempts to put in place a new national framework bargain in Spain (Pérez 2000). Thus, rather than disrupting concertation and fragmenting governance in the European labor market, the completion of the single market and movement to full EMU is likely to lock bargaining partners more closely together. This is precisely why social pacts, which have had at their core an incomes policy strategy anticipating monetary union, have been consolidated during the 1990s in countries seeking – and subsequently thereby achieving – EMU membership. Even before the launch of a single currency, the joint search for consensual, yet flexible, solutions to problems of pay, employment, and competitiveness became a paramount objective, changing the behavior of national employers and trade union organization and leading to innova110

Globalization, Welfare States, and Employment

tions, but not outright fragmentation, in wage bargaining systems (Rhodes 1998a). This argument runs counter to predictions that EMU will “disorganize” European industrial relations. Martin (1998) forecasts three “possible futures” for European wage bargaining. First, although coordinated internally, EMU might turn national bargaining arenas into a single uncoordinated bargaining structure across the EMU area as a whole. “Unions in all regions could thereby be drawn into a deflationary vicious circle of labor cost dumping, or competitive internal depreciations (i.e., wage and benefit cuts) in an effort to bolster price competitiveness in lieu of currency depreciation.” Another alternative might be the emergence of crossborder coordination of wage bargaining within particular sectors, such as engineering, given that the German IG Metall has already played a role in coordinating wage bargaining in this sector in the D-mark zone. This is a much more distinct possibility than coordinated bargaining across the EMU countries as whole, but it is much less likely than what Martin (1998: 21) refers to as the “renationalization” of wage bargaining. Although he does not define this precisely, this notion fits precisely with the resurgence of national wage coordination via limited trade-offs between jobs and wages and benefits or broader “social pacts” of the type discussed earlier. As I argue in this chapter, this is not just a more likely future scenario; it is a contemporary reality. Given that coping with EMU convergence and other pressures for centralization have helped put in place or consolidate corporatist arrangements, varying from the fragile to the strong, depending on the country and national traditions, there is no reason to suspect that full monetary union will alter this institutional landscape enormously. If anything, the imperative for incomes policy and collaborative employment and social policy making becomes greater. Of course, as is already evident from the preceding discussion, “competitive corporatism” is far from being a uniform phenomenon. Nor does concertation or corporatist bargaining necessarily confer economic success. There are also a number of criticisms, including most potently the claim that, in the absence of a coordinated expansion of European economy, supply-side innovations and union concessions to labor market flexibility may simply redistribute employment rather than help create it. But there are significant differences in economic and employment performance among countries maintaining large welfare states alongside neocorporatist bargaining systems and powerful union movements, as well as among those with more flexible social pacts. A recent study by Casey and Gold (2000) 111

Martin Rhodes

suggests that the worst performing countries in this category are Austria, Germany, Italy, and Spain – all of which have or have had neocorporatist industrial relations systems with strong “insider-outsider” divisions and are strongly resistant to change. They have proven particularly poor in responding to the needs of “outsiders” – especially the unemployed, younger, female, or older workers – in terms of income security or social protection. By contrast, the Netherlands, Sweden, and Ireland have shown a capacity to adapt, fostering employment creation and economic growth – despite a restrictive fiscal and monetary climate. As their example also shows, by contrast with the liberal welfare states (the United Kingdom, New Zealand, the United States), it is possible to respond to both domestic and international pressures for welfare and labor market change without also sacrificing aspirations for reconciling social equity and justice with growth.

Conclusions The policy implications of the analysis in this chapter is that neither welfare state retrenchment nor labor market deregulation is necessary for an economy to remain competitive in the “global era,” although realizing particular policy objectives – and boosting employment creation – will certainly mean redesigning welfare systems. Nor, it should be added, is there any a priori justification for an all-out assault on the public sector as such. Reviewing econometric cross-national analyses of the impact of the public sector on economic performance, Mosley and Schmid (1993) argue that there is no convincing evidence that this is negative. On the contrary, the public sector, and in particular efficient public services, are an important institutional condition for competitiveness, especially in innovative, highskilled, and high-value-added forms of production. To achieve this goal, social and welfare policies should be part of an institutional ensemble that fosters long-term relations of trust: close links between companies that foster effective systems of training, cooperation in export marketing, R&D, and the provision of long-term finance; relations in firms that stimulate a longer-term product development strategy, ensure a bettereducated and more cooperative workforce and make managers more technically competent and willing to invest in company-specific skills (Soskice 1991; Casper and Vitols 1997). A parallel argument concerns protective labor market regulation. Rather than hampering economic performance because of alleged price distortion, many forms of protective labor market regulation enhance 112

Globalization, Welfare States, and Employment

productivity and are beneficial for economic development. Thus, minimum wages pressure firms into finding ways to raise productivity, whether through technological innovation or through training. As long as they are designed so as not to create or accentuate an insider/outsider dualism in the labor market, employment security regulations will improve the worker’s commitment to the enterprise, creating trust and enhancing forms of work flexibility (see, e.g., Standing 1991, 1999). In sum, the institutions required to build and sustain trust are compatible with internationalization. Not only do open economies seem able to sustain strongly unionized environments, but corporatist bargaining structures remain important in those economies increasingly exposed to international forces. As Lange and Scruggs (1997) explain in their study of the relationship between globalization (measured in terms of trade, financial flows, and foreign direct investment) and unionization rates, there is a contingent effect on labor strength. Unions which have historically done better under economic expansion need not be damaged if new markets for traded goods (involving product market deregulation) are produced in more unionized sectors or if capital flows and inward investments are directed toward such sectors. At the same time, different types of capital and investors are likely to seek different kinds of investment opportunities: as long as they are productive, unionized sectors and national economies may well be as successful – if not more so – than less unionized ones. As I have discussed, far from disaggregating organized interests and collaborative relations with business and government, internationalization has at most modified and in some cases actually strengthened them. As shown by the persistence of coordinated bargaining in traditional corporatist countries and the emergence and consolidation of new forms of social partnership elsewhere, the future for most continental European economies lies neither in the neoliberal approach nor in conventional social corporatism, but in flexible, pragmatic, productivity-oriented social pacts – in other words, in a European “Third Way.” The “new politics” of welfare and employment will revolve around the means and modalities of sustaining such bargains in an era of permanent change. References Andrews, D. M. 1994. Capital Mobility and State Autonomy. International Studies Quarterly 38: 193–218. Burtless, G. 1995. International Trade and the Rise in Earnings Inequality. Journal of Economic Literature 33: 800–16. 113

Martin Rhodes Casey, B., and M. Gold. 2000. Social Partnership and Economic Performance: The Case of Europe. Northampton, Mass.: Edward Elgar. Casper, S., and S. Vitols. 1997. The German Model in the 1990s: Problems and Prospects. Industry and Innovation 4, 1: 1–13. Clayton, R., and J. Pontusson. 1997. Welfare State Retrenchment and Public Sector Restructuring in Advanced Capitalist Societies. Paper presented to the annual meeting of the American Political Science Association, Washington, D.C., August 28–31. Dølvik, J. E., and A. Martin. 1997. A Spanner in the Works and Oil on Troubled Waters: The Divergent Fates of Social Pacts in Sweden and Norway. In G. Fajertag and Ph. Pochet, eds., Social Pacts in Europe. Brussels: European Trade Union Institute/Observatoire Social Européen, 103–34. Eatwell, J. 1995. The International Origins of Unemployment. In J. Michie and J. Grieve Smith, eds., Managing the Global Economy. Oxford: Oxford University Press, 271–86. Esping-Andersen, G. 1996. After the Golden Age? Welfare State Dilemmas in a Global Economy. In G. Esping-Andersen, ed., Welfare States in Transition: National Adaptations in Global Economies. London: Sage Publications, 1–31. Fligstein, N. 1998. Is Globalization the Cause of the Crises of the Welfare States? EUI Working Paper SPS 98/5. European University Institute, Florence. Freeman, R. B. 1995. Are Your Wages Set in Beijing? Journal of Economic Perspectives 9, 3: 15–32. Frieden, J. A. 1991. Invested Interests: The Politics of National Economic Policies in a World of Global Finance. International Organization 45, 4: 425–51. Ganghof, S. 2000. Adjusting National Tax Policy to Economic Internationalization: Strategies and Outcomes. In F. W. Scharpf and V. A. Schmidt, eds., Welfare and Work in the Open Economy: Diverse Responses to Common Challenges, vol. 2. Oxford: Oxford University Press, 597–645. Garrett, G. 1995. Capital Mobility, Trade, and the Domestic Politics of Economic Policy. International Organization 49, 4: 657–87. ———. 1998a. Global Markets and National Politics: Collision Course or Virtuous Circle. International Organization 52, 4: 787–824. ———. 1998b. Partisan Politics in the Global Economy. Cambridge: Cambridge University Press. Glyn , A. 1995. Social Democracy and Full Employment. New Left Review 211: 33–55. Gobeyn, M. J. 1993. Explaining the Decline of Macro-corporatist Political Bargaining Structures in Advanced Capitalist Societies. Governance: An International Journal of Policy and Administration 6, 1: 3–22 Goodman, J. B., and L. W. Pauly. 1993. The Obsolescence of Capital Controls: Economic Management in an Age of Global Markets. World Politics 46, 1: 50–82. Gordon, D. M. 1996. Conflict and Cooperation: An Empirical Glimpse of the Imperatives of Efficiency and Redistribution. Politics and Society 24, 4: 433–56. Gourevitch, P. A. 1996. The Macropolitics of Microinstitutional Differences in the Analysis of Comparative Capitalism. In S. Berger and R. Dore, eds., National Diversity and Global Capitalism. Ithaca, N.Y.: Cornell University Press, 239–59. 114

Globalization, Welfare States, and Employment Grahl, J., and P. Teague. 1997. Is the European Social Model Fragmenting? New Political Economy 2, 3: 405–26. Iversen, T., and A. Wren. 1998. Equality, Employment and Budgetary Restraint: the Trilemma of the Service Economy. World Politics 50, 4: 507–46. Keohane, R. O., and H. V. Milner. 1996a. Internationalization and Domestic Politics: An Introduction. In R. O. Keohane and H. V. Milner, eds., Internationalization and Domestic Politics. Cambridge: Cambridge University Press, 3–24. ———. 1996b. Internationalization and Domestic Politics: A Conclusion. In R. O. Keohane and H. V. Milner, eds., Internationalization and Domestic Politics. Cambridge: Cambridge University Press. Kurzer, P. 1993. Business and Banking: Political Change and Economic Integration in Western Europe. Ithaca, N.Y.: Cornell University Press. Lange, P., and L. Scruggs. 1997. Where Have All the Members Gone? Union Density in the Era of Globalization. Paper presented to the Annual Meeting of the American Political Science Association, Washington, D.C., August 28–31. Lash, S., and J. Urry. 1987. The End of Organized Capitalism. Oxford: Polity Press. Martin, A. 1998. EMU and Wage Bargaining: The Americanization of the European Labor Market? Paper prepared for the 11th International Conference of Europeanists, Baltimore, February 26–28. Mosley, H., and G. Schmid. 1993. Public Services and Competitiveness. In K. S. Hughes, ed., European Competitiveness. Cambridge: Cambridge University Press, 200–31. Mueller, F. 1996. National Stakeholders in the Global Contest for Corporate Investment. European Journal of Industrial Relations 2, 3: 345–66. Negrelli, S. 1997. Social Pacts and Flexibility: Towards a New Balance between Macro and Micro Industrial Relations: The Italian Experience. In Giuseppe Fajertag and Philippe Pochet, eds., Social Pacts in Europe. Brussels: European Trade Union Institute/Observatoire Social Européen, 45–62. Nickell, S. 1997. Unemployment and Labor Market Rigidities: Europe versus North America. Journal of Economic Perspectives 11, 3: 55–74. Notermans, T. 1993. The Abdication of National Policy Autonomy: Why the Macroeconomic Policy Regime Has Become So Unfavorable to Labor. Politics and Society 21, 2: 133–67. Organisation for Economic Co-operation and Development (OECD). 1996. The Experience with Fiscal Consolidation in OECD Countries. OECD Economic Outlook 59, June: 33–41. ———. 1999. OECD Economic Surveys 1998–1999: Ireland. Paris: OECD. Ormerod, P. 1998. Unemployment and Social Exclusion: An Economic View. In M. Rhodes and Y. Mény, eds., The Future of European Welfare: A New Social Contract? London: Macmillan, 21–40. Pauly, L. W., and S. Reich. 1997. National Structures and Multinational Corporate Behavior: Enduring Differences in a Geobreizing World. International Organization 51, 1: 1–30. Pérez, S. A. 2000. The Resurgence of National Bargaining in Europe: Explaining the Italian and Spanish Experiences. Working Paper 1999/130. Centro de 115

Martin Rhodes Estudios Avazados en Ciecias Sociales, Instituto Juan March de Estudios e Investigaciones. Perraton, J., D. Goldblatt, D. Held, and A. McGrew. 1997. The Globalization of Economic Activity. New Political Economy 2, 2: 257–77. Pierson, P. 1997. Skeptical Reflections on “Globalization” and the Welfare State. Paper presented at the International Conference on Socio-Economics. Montreal, July. ———. 1998. Irresistible Forces, Immovable Objects: Post-Industrial Welfare States Confront Permanent Austerity. Journal of European Public Policy 5, 4: 539–60. ———. Forthcoming. Post-Industrial Pressures on the Mature Welfare States. In P. Pierson, ed., The New Politics of the Welfare State. Oxford: Oxford University Press. Rhodes, M. 1998a. Globalisation, Labour Markets and Welfare States: A Future of “Competitive Corporatism”? In M. Rhodes and Y. Mény, eds., The Future of European Welfare: A New Social Contract? London: Macmillan, 178–203. ———. 1998b. Globalization and the Welfare State: The Emergence of Competitive Corporatism. Swiss Political Science Review 4, 1 (Spring): 99–107. ———. Forthcoming. The Political Economy of Social Pacts: “Competitive Corporatism” and European Welfare Reform. In P. Pierson, ed., The New Politics of the Welfare State. Oxford: Oxford University Press. Rhodes, M., and van Apeldoorn, B. 1998a. Globalization, Employment and Social Policy. Report for the European Commission (DG – Employment and Social Affairs). ———. 1998b. Capital Unbound? The Transformation of European Corporate Governance. Journal of European Public Policy 5, 3: 407–28. Rodrik, D. 1997. Has Globalization Gone Too Far? Washington, D.C.: Institute for International Economics. Rowthorn, R., and R. Ramaswamy. 1997. Deindustrialization: Causes and Implications. Working Paper 97/42. Washington, D.C.: International Monetary Fund. Ruggie, J. G. 1982. International Regimes, Transactions and Change: Embedded Liberalism in the Postwar Economic Order. International Organization 36, 2: 379–415. Ruigrok, W., and R. Van Tulder. 1995. The Logic of International Restructuring. London: Routledge. Sachs, J., and H. Schatz. 1994. Trade and Jobs in U.S. Manufacturing. Brookings Papers on Economic Activity 1: 1–84. Scharpf, F. W. 1997. Combating Unemployment in Continental Europe: Policy Options under Internationalization. RSC Policy Paper 3. Robert Schuman Centre, European University Institute, Florence. ———. 1999. The Viability of Advanced Welfare States in the International Economy. Manuscript. Schmitter, Ph. C. 1989. Corporatism Is Dead! Long Live Corporatism! Government and Opposition 24: 54–73.

116

Globalization, Welfare States, and Employment Schmitter, Ph. C., and J. R. Grote. 1997. The Corporatist Sisyphus: Past, Present and Future. Nota di Lavoro, Fondazione Eni Enrico Mattei. Siebert, H. 1997. Labor Market Rigidities: At the Root of Unemployment in Europe. Journal of Economic Perspectives 11, 3: 37–45 Slaughter, M. J., and P. Swagel. 1997. The Effect of Globalization on Wages in the Advanced Economies. Working Paper 97/43. Washington, D.C.: International Monetary Fund. Snower, D. J. 1997. Challenges to Social Cohesion and Approaches to Policy Reform. In OECD, Societal Cohesion and the Globalising Economy: What Does the Future Hold? Paris: OECD, 39–60. Soskice, D. 1991. The Institutional Infrastructure for International Competitiveness: A Comparative Analysis of the U.K. and Germany. In A. B. Atkinson and R. Brunetta, eds., Economics for the New Europe. Basingstoke: Macmillan, 45–66. Standing, G. 1991. Structural Adjustment and Labour Market Policies: Towards Social Adjustment? In G. Standing and V. Tokman, eds., Towards Social Adjustment: Labour Market Issues in Structural Adjustment. Geneva: International Labour Office, 5–51. ———. 1999. Global Labour Flexibility: Seeking Distributive Justice. London: Macmillan. Streeck, W., and P. C. Schmitter. 1991. From National Corporatism to Transnational Pluralism: Organized Interests in the Single European Market. Politics and Society 19, 133–64. Swank, D. 1992. Politics and the Structural Dependence of the State in Democratic Capitalist Nations. American Political Science Review 86, 1. ———. 1998. Funding the Welfare State: Global Taxation of Business in Advanced Market Economies. Political Studies 46: 671–92. Tanzi, V. 1996. Globalization, Tax Competition and the Future of Tax Systems. Working Paper 96,141. Washington, D.C.: International Monetary Fund. Teague, P., and J. Grahl. 1998. Institutions and Labour Market Performance in Western Europe. Political Studies 46: 1–18. Thelen, K. 2000. Why German Employers Cannot Bring Themselves to Dismantle the German Model. In T. Iversen, J. Pontusson, and D. Soskice, eds., Unions, Employers and Central Banks: Wage Bargaining and Macro-economic Regimes in an Integrating Europe. New York: Cambridge University Press, pp. 138–69. Therborn, G. 1997. The Western European Welfare State in Its Hostile World. Working paper 1997/109. Centro de Estudios Avazados en Ciecias Sociales, Instituto Juan March de Estudios e Investigaciones. Thygesen, N., Y. Kosai, and R. Z. Lawrence. 1996. Globalization and Trilateral Labour Markets: Evidence and Implications: A Report to the Trilateral Commission. New York, Paris, and Tokyo: Trilateral Commission. Traxler, F. 1997a. European Transformation and Institution Building in East and West: The Performance of and Preconditions for Neocorporatism. In R. W. Kindley and D. F. Good, eds., The Challenge of Globalization and Institution Building: Lessons from Small European States. Boulder: Westview Press, 145–78.

117

Martin Rhodes ———. 1997b. The Logic of Social Pacts. In Giuseppe Fajertag and Philippe Pochet, eds., Social Pacts in Europe. Brussels: European Trade Union Institute/Observatoire social européen, 27–36. Van den Ploeg, F. 1997. The Political Economy of a Consensus Society: Experience from Behind the Dykes. Economic and Social Review 28, 3: 307–32. Visser, J., and A. Hemerijk. 1997. “A Dutch Miracle”: Job Growth, Welfare Reform and Corporatism in the Netherlands. Amsterdam: Amsterdam University Press. Wallerstein, M., and M. Golden. 1997. The Fragmentation of the Bargaining Society: Wage Setting in the Nordic Countries, 1950 to 1992. Comparative Political Studies 30: 699–731. Wallerstein, M., M. Golden, and P. Lange. 1997. Unions, Employers’ Associations and Wage-Setting Institutions in Northern and Central Europe, 1950–1992. Industrial and Labour Relations Review 50, 3: 379–401. Webb, M. C. 1991. International Economic Structures, Government Interests, and International Coordination of Macroeconomic Adjustment Policies. International Organization 45, 3: 309–38. Weiss, L. 1997. The Myth of the Powerless State. New Left Review (September–October): 3–27. Wood, A. 1994. North-South Trade, Employment, and Inequality: Changing Fortunes in Skill-Driven World. Oxford: Clarendon Press. ———. 1995. How Trade Hurt Unskilled Workers. Journal of Economic Perspectives 9, 3: 57–80.

118

PA R T I I

Unemployment and Domestic Bargaining Institutions: Challenging Some Myths

4 Wage Growth, Recession, and Labor Decline in the Industrialized Democracies, 1965–1993 Bruce Western and Kieran Healy

The current unemployment crisis in Europe strikingly demonstrates that the labor markets there are failing to deliver rising living standards, as they have for most of the postwar period.1 In addition to persistent high unemployment, real wage growth has stagnated across Europe. In the late 1960s and early 1970s, European wages rose by about 4 or 5 percent a year. In the decade from 1983, real wage growth throughout Europe averaged between 1 and 2 percent. Indeed, a general wage stagnation swept across most of the industrialized democracies. Real U.S. wage growth had always been slower than in Europe, and by the 1980s and 1990s, hourly wage rates in the United States were falling by almost 1 percent annually. It is tempting to attribute the broad slowdown in wage growth to the power of market forces. In addition to the persistently high unemployment, wages were buffeted by inflationary shocks and a sustained slowdown in productivity growth. The European wage stagnation thus seems the likely by-product of hostile market forces. The market explanation is further supported by the sheer breadth of the wage slowdown. The United States, where wage setting is significantly more decentralized than in Europe, experienced very similar economic trends. The appearance of common economic conditions despite substantial variation in wage setting suggests that market forces have simply swamped the capacity of labor market institutions to protect wage standards. We present an alternative account of the OECD wage slowdown that investigates the role of institutions and institutional change. In this account, distributional conflict in the labor market is significantly shaped 1

Research for this study was supported by grant SBR95-11473 from the National Science Foundation. We thank Paul DiMaggio, Mike Hout, Paul Pierson, and Michael Wallerstein for useful comments on an earlier version of this chapter.

121

Bruce Western and Kieran Healy

by institutions that provide workers with control over the risks that arise in capitalist economies (e.g., Korpi 1983; Goldthorpe 1984; EspingAndersen 1990). Corporatist institutions, extensive collective bargaining, and supportive social democratic governments shielded workers and their unions from unemployment, inflation, and income insecurity through much of the postwar period. We find evidence that these institutions moderated union demands when unemployment threatened but otherwise facilitated claims on the dividends of economic growth. Labor market institutions allowed wage growth in Europe to outpace that in the United States while protecting those at the margins. Data analysis shows that a historic turning point was reached in the mid-1970s when the institutional logic of postwar labor markets was fractured by a widespread decline in the power of organized labor. This decline curtailed labor’s influence, and wages stagnated in the advanced capitalist countries as a result. This analysis places the European experience in the broader context of OECD wage trends. We report on estimates from a model of real wage growth in 18 OECD countries between 1966 and 1993. We divide the time series into two periods, distinguishing the golden age before the oil shocks triggered by the Organization of Petroleum Exporting Countries (OPEC) from the current era of slow economic growth. This design offers broad variation in labor market institutions and allows a deeper understanding of institutional effects under conditions of protracted recession.

Wage Growth in 18 OECD Countries, 1965–1993 Table 4.1 summarizes trends in manufacturing sector wages that form the dependent variable for this analysis. From the mid-1960s to the 1970s, pay rises accelerated in Europe. Fueled by rapid inflation and the strike waves of 1968–69, wages climbed by about 5 percent a year in nearly all the European countries. An unprecedented economic boom in Japan helped drive the fastest rate of wage growth in the OECD countries – an extraordinary rate of around 8 percent annually. Wages rose more slowly in the Englishspeaking countries. Real earnings grew by about 3 percent a year in Canada and Britain. U.S. wage growth was sluggish by comparison, averaging just under 1.5 percent a year. The OPEC oil shock of 1973–74 ushered in a new period of poor labor market performance across most OECD countries. Mounting unemployment undermined union power and rising inflation eroded real wages. In 122

Wage Growth, Recession, Labor Decline in Industrialized Democracies

Table 4.1. Summary of Annual Percentage Growth in Real Hourly Manufacturing Wage Rates, 18 OECD Countries

France, Germany, and Italy, the rate of real wage growth in the decade following the oil shock was reduced by half. Where wage growth was slow – in North America – real wage declines replaced the small pay rises of the previous decade. In the United States, the real value of wages fell by more than 5 percent in the ten years after 1974. The stagnation and decline of wages in the late 1970s continued through the 1980s and into the 1990s. In Europe, the pace of wage growth slowed significantly. The turnaround in Italian wage trends is particularly striking. Italian wage indexation protected incomes from inflation in the 1970s but was significantly weakened in the 1980s and earnings grew slowly in response. In the United States, real wage decline accelerated despite strong economic performance in the late 1980s. In sum, two types of variation characterize wage movements in the advanced capitalist countries between the mid-1960s and the mid-1990s. First, wage growth was generally slow in the United States compared with Europe. Second, wage growth declined sharply in all countries in the late 1970s. This general wage stagnation persisted into the 1980s. We now turn to explanations that try to account for this variation. 123

Bruce Western and Kieran Healy

Market Explanations Conventional market explanations appear to provide a convincing and general account of wage trends in the industrialized democracies. These explanations focus on the forces of supply and demand in the labor market. Labor demand is reflected in the unemployment rate. On the supply side, productivity growth and inflation are the key variables. When labor demand is weak and unemployment is high, wage growth slows. In theories of competitive labor markets, unemployment restrains wage rises by raising competition among workers for scarce jobs. In unionized labor markets workers are prevented from underbidding wages. In this situation, unemployment increases the threat of layoffs or business failures and unions bargain less aggressively as a result. The negative relationship between real wage growth and unemployment has thus been observed in both unorganized and highly unionized labor markets (Carlin and Soskice 1990: 139; Layard, Nickell, and Jackman 1991: chap. 9; OECD 1994: 3–4). European unemployment is clearly a key suspect in explaining the wage slowdown. High unemployment created constant pressure for wage restraint among many continental unions. For example, German unemployment averaged 1.2 percent between 1965 and 1975 but exceeded 10 percent for most of the 1990s. Similar patterns are found in France and Italy. Such increases in unemployment place a sizable economic constraint on wage growth. The quality of the labor supply also influences wage growth. For neoclassical theories, competitive markets help ensure that workers are paid their marginal product. Productivity growth is thus the engine of wage growth (Hicks 1963: 8). The assumption of competitive markets is not vital, however. In unionized labor markets, employers finance wage rises out of productivity gains, and productivity increases are written into union contracts. In the U.S. auto industry, for example, a wage increase of 3 percent was built into the collective bargaining formula in recognition of “technical progress, better tools, methods, processes and equipment” (Katz 1985: 200). Under very different institutional conditions in Norway and Sweden, productivity trends in the export sector have also established the context for centralized wage negotiations through the postwar period (Flanagan, Soskice, and Ulman 1983). The labor supply is also influenced by inflation. In one approach, workers supply a given quantity of labor in return for a certain real wage. In this view, unexpected inflation causes workers to overestimate the value 124

Wage Growth, Recession, Labor Decline in Industrialized Democracies

of their wages and oversupply their labor (Friedman 1968: 7–11). The excess labor supply then drives down wages. In more concrete terms, inflation can reduce real wages because employment contracts specify nominal rather than real pay rates. Often inflationary expectations are built into union contracts in the form of cost-of-living adjustments. Even in these cases, though, inflation in excess of expectations can still reduce wages. Termed “nominal wage rigidity,” the insensitivity of money wages to prices increasingly devalues real wages as inflation rises (Mitchell 1993).

Institutional Explanations Market forces are only a starting point for analysis because labor movements have acquired a measure of collective control over the risks of unemployment, inflation, and income insecurity in capitalist economies. Corporatist bargaining, sometimes in combination with social democratic governments, has played a key role in controlling the risk of unemployment. Extensive collective bargaining reduces the risks of inflation and income insecurity for workers. Taken together, corporatism, social democratic government, and extensive collective bargaining have substantially altered the influence of market forces on wage growth. Theories of corporatism argue that the effect of unions on wages depends on the centralization of interest representation. In some countries, notably the Scandinavian countries and Austria, labor and business representatives negotiate wage contracts at the national or sectoral level. In corporatist bargaining systems like these, unions negotiate for all workers in the labor market. The impact of wage increases on unemployment affects the central unions’ constituency, so labor leaders weigh heavily the employment effects of their wage demands. Wage moderation is thus a feature of collective bargaining in corporatist countries (Olson 1982; Crouch 1985; Calmfors and Driffill 1988). With decentralized wage talks, unions take a narrower view of their interests, and the employment effects of wage demands are discounted. Instead of protecting jobs at the margins of the labor market, unions in decentralized contexts focus more narrowly on maximizing the wages of union members, whose employment is secure. This argument suggests that the effect of unemployment on wage growth depends on the presence of corporatist institutions. In corporatist countries, wages will be extremely sensitive to unemployment. A small rise in unemployment should cause a large decline in wage growth. In more decentralized labor markets, unions pursue their sectional 125

Bruce Western and Kieran Healy

interests, and wage demands are less responsive to prevailing labor market conditions. Unemployment will have a weaker negative effect on wages in these settings. Research on corporatism also suggests that political parties are important for securing wage moderation for employment goals (Crouch 1985; Headey 1970). Labor and social democratic governments facilitate wage moderation by centralized unions by offering tax and social welfare guarantees. These incentives help maintain living standards while unions restrain their bargaining power. Conservative governments cooperate less with unions, obstructing a coordinated approach to macroeconomic policy. Conservative governments may also be less inclined to reflationary measures such as social welfare spending. Previous research thus associates the combined impact of social democratic parties and centralized unions with low unemployment and strong economic growth (Scharpf 1987; Alvarez, Garrett, and Lange 1992; Hicks 1994). Whether corporatist incorporation of social democratic government affected wage moderation is an open question, however, because social democrats have also actively supported their working-class electorates by maintaining wage standards. In a twist on the corporatist theory, the OECD’s McCracken Report thus apportions some blame to European labor governments for the 1974–75 recession. For McCracken’s analysis, the full employment guarantees of social democratic corporatism removed the market discipline that controlled unions’ wage militancy (McCracken et al. 1977: 50–51, 104). Although perverse from the corporatist perspective, the McCracken analysis does point to social democratic support for union bargaining in the late 1960s and 1970s. During this time, left governments in the corporatist countries backed central union wage policies for low-pay workers. Social democrats also intervened less in industrial relations compared with conservatives, who more commonly preempted collective bargaining with wage freezes (Flanagan et al. [1983] provide evidence for the Netherlands and Denmark). In addition, social democrats allowed large public sector wage increases that spilled over into the economy as a whole. In contrast to the logic of corporatism, these examples suggest social democrats backed unions in their efforts to raise wage standards. Whether social democrats under corporatism facilitate or suppress wage growth is left as an empirical question for the data analysis. This argument for the positive impact of left governments also suggests the importance of unions for raising wages. Although comparative 126

Wage Growth, Recession, Labor Decline in Industrialized Democracies

research typically emphasizes corporatist wage moderation, effective incomes policies that restrict wage growth have been extremely difficult to sustain (Flanagan et al. 1983; Scharpf 1987; Swenson 1989). On the other hand, unions in their traditional role as wage-bargaining agents have widely raised earnings in corporatist and noncorporatist countries. The positive impact of unions on wages takes different forms in different institutional contexts. In many corporatist countries, unions pursued solidaristic wage policies that raised the pay of women and workers in low-wage sectors (Flanagan et al. 1983; Swenson 1989; Rosenfeld and Kalleberg 1991). Corporatist wage agreements were also supplemented by wage drift generated by a second round of local wage talks. In decentralized labor markets where institutional pressures for wage moderation were weak, unions raised earnings significantly above the competitive level. Thus analyses of social survey data find that unionized workers earn more than nonunion workers throughout North America and Europe, but the wage advantage of unionized workers is largest in the United States (Colbjørnsen and Kalleberg 1988; Blanchflower and Freeman 1992). The positive effects of unions on wage growth are captured by the coverage of collective bargaining. Where union contracts extend to only a small minority of workers in the labor force, as in the United States, the influence of unions on aggregate wage movements will be small. The positive impact of unions on average wage growth will be large where union coverage is extensive. The coverage of collective bargaining is not coextensive with union membership. In many countries, collective bargaining agreements are extended beyond the unionized workforce. In Germany, for example, the terms of collective agreements are imposed on nonunion employers by the central employer association. France provides the extreme case: only about 10 percent of its labor force is unionized, but nearly all workers are covered by union rules (Traxler 1996). These extension procedures transmit the gains of union bargaining to the entire labor market and significantly raise the wages of low-pay workers (Freeman and Katz 1995). In contrast to unionization trends, which show significant longitudinal variation, the coverage of collective bargaining is relatively stable over time. Thus Traxler (1996) finds that unionization has generally declined in the OECD countries through the 1980s, but bargaining coverage fell in only a few countries without contract extension procedures. The extent of collective bargaining coverage thus offers a good indication of the enduring – institutionalized – wage-bargaining influence of unions. 127

Bruce Western and Kieran Healy

As well as raising wages, bargaining coverage should also shape the effects of unemployment and inflation. Because it transmits union wage increases, one would expect bargaining coverage to reduce the negative influence of unemployment on wage growth. Extensive bargaining coverage maintains wage standards across the labor market and prevents wages from falling in response to rising joblessness. Furthermore, collective bargaining coverage typically protects the value of wages from inflation. In countries with biennial or triennial bargaining rounds, union contracts usually provide for some type of indexing that raises wages in response to price movements (EIRR 1985; Mitchell 1993). Such terms became extremely common in OECD countries during the period of rising inflation in the mid-1970s. Italy offers a clear example. The Italian system of wage indexation provided quarterly flat-rate adjustments in wages (Erickson and Ichino 1995: 277–81). At the height of the wage spurt in the mid-1970s, the nominal earnings of Italian workers grew by more than 20 percent a year. In countries with annual bargaining rounds, the timely renegotiation of union contracts allows unions to regularly update wage levels. We therefore expect that collective bargaining coverage reduces the sensitivity of wage growth to unemployment but protects real wages from the negative effects of rising inflation.

Wage Growth and the End of the Golden Age The performance of the advanced capitalist labor markets deteriorated dramatically in the 1970s. As we have already noted, not one of the OECD countries escaped the wage stagnation of the 1980s and 1990s. Some countries, such as the United States, suffered large and extended declines in real wages. The golden age of rising prosperity in the 1950s and 1960s was thus replaced by an era of falling living standards with few prospects for improvement. The generality of the market explanations adapts easily to the wage stagnation of the last two decades. Following the first oil shock, a severe slowdown in productivity growth constricted pay rises. A massive increase in unemployment, particularly persistent in Europe, shifted the balance of power in the marketplace from workers to employers. Finally, the purchasing power of pay packets was dissolved by inflationary shocks in 1973–74 and 1979. In contrast to this emphasis on market conditions, we argue that the wage slowdown of the 1980s and 1990s can be traced to labor’s declining 128

Wage Growth, Recession, Labor Decline in Industrialized Democracies

institutional position in the advanced capitalist labor markets. Three developments are important in this argument. First, many of the corporatist bargaining forums of the 1960s and 1970s have been displaced by local bargaining. As a result, centrally orchestrated wage restraint has become uncommon. Second, the bargaining power of unions has diminished. Third, social democratic parties moved away from their traditional support of unions and their working-class constituencies. The extent of corporatist decline is disputed by some researchers. The comprehensive study of Lange, Wallerstein, and Golden (1995; Golden, Lange, and Wallerstein 1993) finds significant continuity in collective bargaining through the 1970s and 1980s in a sample of 16 OECD countries. While the demise of national bargaining in Sweden is often taken as evidence of corporatist decline (Lash 1985), the Swedish case is exceptional. Formal bargaining institutions in Europe were by and large unchanged through the 1980s. Thus measures of corporatism generally show little change over time, and patterns of cross-national variation are enduring. Even in clear-cut cases of institutional change, cross-national variation is large compared with change over time. Despite the emergence of industry bargaining in Sweden, for example, the labor market there still retains many features of corporatism that distinguish it from Britain or the United States. Although formal institutions may show continuity, unions in corporatist countries function differently in the 1980s and 1990s compared with the 1960s (Streeck 1993). Significantly, local bargaining within firms increasingly determines wages and conditions in Europe. For example, industrial relations in Germany still begin with regional bargaining by industry unions, but these have been accompanied by substantial growth in local bargaining through firm-level works councils (Thelen 1993). In Belgium, a ten-year period of decentralized bargaining was replaced by national negotiations in 1986, while enterprise bargaining thrived in the interim (EIRR 1989a, 1989b). Similarly, national bargaining continued in Norway through the 1980s but centrally negotiated wage restraint was circumvented by local wage drift. By the mid-1980s, wage drift accounted for more than 80 percent of all wage increases (Moene and Wallerstein 1993: 393; for further evidence of bargaining decentralization, see also Baglioni and Crouch 1990; Katz 1993). Under these conditions, the corporatist compromises that characterized European wage bargaining in the 1960s and 1970s were undermined by local industrial relations, and few central unions could deliver wage restraint in return for full employment. As might be expected, because the 129

Bruce Western and Kieran Healy

institutional conditions for corporatist wage moderation unraveled, the steep trade-off between unemployment and wages also eroded. Without centralized interest representation in industrial relations, union leaders were less able to dampen wage claims in response to slack labor market conditions in the 1980s and 1990s. The negative impact of unemployment on wages should thus diminish in the corporatist countries in the recent period. The demise of corporatist wage moderation is related to the widespread decline of union bargaining power. Union power sagged under the weight of organizational losses and the growth of the global economy. Unionization rates have fallen in nearly all OECD countries in the 1980s (Western 1995). As a result, the power of the strike is severely weakened, and labor militancy has fallen precipitously (Shalev 1992). The market position of unions has also been threatened by economic competition from low-wage countries (Tilly 1995). Increasing capital mobility raises the threat of employer exit for militant unions. The growth of foreign trade places unionized labor in the North in competition with low-wage exporters of the South (Wood 1994). Although the institutional mechanism for raising wages – collective bargaining coverage – remained stable, the capacity of unions to activate this mechanism has thus declined. Because of declining union power, the effects of collective bargaining coverage that were hypothesized for the golden age should be reversed in the 1980s. Collective bargaining coverage should be associated with slower wage growth in recent times. The positive impact of bargaining coverage on unemployment effects should also diminish as unions succumb to market signals. In addition, indexation mechanisms that widely sheltered wages from inflation in the early 1970s were increasingly viewed as inflationary by employers and governments in the 1980s. Wage indexation was by and large abandoned, and the risk of inflation was transferred from employers to workers (Mitchell 1993; Western 1996). Collective bargaining coverage thus affords much weaker protection from inflation in the 1980s and 1990s than in the earlier period. With this reasoning, the negative impact of inflation on wage growth will increase in countries with extensive bargaining coverage. While labor’s position on the industrial front appears weakened, researchers also point to the “decline of social democracy.” Many governing parties of the left in the 1970s were in opposition through the 1980s (Piven 1991; Pontusson 1995). Those that retained power were unable

130

Wage Growth, Recession, Labor Decline in Industrialized Democracies

to support their traditional working-class constituencies through social welfare and full employment policy. In the corporatist countries, Swedish and Norwegian social democrats are seen to be “rapidly abandoning social democracy and embracing market liberalism” (Moene and Wallerstein 1993: 385). Between 1983 and 1989, social democrats in Sweden cut subsidies to industry, partially privatized state firms, deregulated financial markets, and – following Ronald Reagan’s 1986 tax reform – implemented a large and regressive tax cut. With these policy shifts, employers increasingly targeted labor costs in market competition (Pontusson 1994: 35–38). In Norway, social democratic government between 1986 and 1993 followed a similar line, deregulating housing and financial markets, and generally focusing on price stability over full employment in economic policy. Austerity policies in the late 1980s raised Norwegian unemployment and eroded real incomes (Mjøset et al. 1994: 67–70). Outside of the corporatist sphere, left parties in France, Australia, and New Zealand also turned to policies of deregulation, decentralization, and privatization in the 1980s, relinquishing their historic protection of wage standards (Stilwell 1993; Ross and Jenson 1994; Massey 1995: chap. 3). The neoliberal turn of social democracy through the 1980s damaged support for unions in collective bargaining and subverted the political exchange of public policy for wage moderation. Thus the effects of social democratic government on wage growth are likely to be rather small in the most recent period. These institutional developments mark the emergence of a new political economy of the advanced capitalist labor markets some time in the mid-1970s. As labor lost its institutional hold over the employment relationship, much of the risk in capitalist economies was transferred from employers and the state to unions and workers. With the decline of social democratic corporatism, workers increasingly confronted the threat of unemployment. The erosion of union bargaining power raised the risks of inflation and income insecurity. The empirical implication of this account is fundamentally different from that of the market explanation of the end of the golden age. The market explanation posits a change in the values of key variables – productivity growth, unemployment, and inflation – but the effects of these variables remain the same. If labor movement decline has depressed wage trends, one would expect a basic shift in the effects of economic and institutional variables. The story of labor decline thus implies a structural break in the process of wage determination.

131

Bruce Western and Kieran Healy

A Model of Real Wage Growth and Labor Decline The market explanation suggests that real wage growth depends negatively on unemployment and inflation, but positively on productivity. The institutional explanation says corporatism generates wage moderation, perhaps in combination with social democratic governments. Collective bargaining coverage raises wages by expanding the range of union wage setting. Corporatism and bargaining coverage also help explain the variation in economic effects across countries. Wage growth will be more constrained by the unemployment rate in corporatist countries, but this effect will be offset to some degree by a high level of collective bargaining coverage. Because of wage indexing in collective agreements, extensive bargaining coverage will also insulate wages from rising inflation. These ideas can be formalized with a regression where interaction terms are used to capture the institutionally conditional effects. The variables used for the regression analysis are listed in Table 4.2. In line with theories of institutional effects, the corporatist dummy variable is interacted with the unemployment and left-government variables; bargaining coverage is interacted with unemployment and inflation growth. Data for the analysis are summarized in Table 4.3. Eighteen countries, each with annual time series of twenty-nine years, yield 522 observations for analysis. Collective bargaining coverage is measured by Traxler’s (1996)

Table 4.2. Description of Variables Used in the Regression Analysis of Real Wage Growth

132

Wage Growth, Recession, Labor Decline in Industrialized Democracies

Table 4.3. Means of Time-Series Variables and Institutional Variables Used in Analysis of Real Wage Growth in 18 OECD Countries, 1965–1993

1990 series, and corporatism by a modified version of Crouch’s (1985) classification. In his original formulation, Crouch (1985: 119) observed that the coding of Belgium as noncorporatist and Switzerland as corporatist may be arguable. Belgian unions are centralized compared with the Swiss, and unlike the Swiss, Belgian unions attempted wage moderation to control inflation through the 1960s and 1970s (Molitor 1978; Katzenstein 1985). Because this analysis focuses on the centralized organization of labor, Belgium has been reclassified as corporatist, and Switzerland is treated as noncorporatist. (The appendix provides further details.) To allow for a structural break in the wage-setting regime, the data are split into two periods. The first extends through the first oil shock, from 1965 to 1974. The second covers the current period of slow growth, from 1975 to 1993. Because of institutional changes in wage setting, institutional effects should generally be smaller in the more recent period, and the influence of institutions on economic effects should also decline. 133

Bruce Western and Kieran Healy

Results Table 4.4 reports the regression results. The first column of the table describes wage growth between 1965 and 1974. The market effects must be interpreted in light of the interactions with the institutional variables. In the liberal market economies of the noncorporatist countries, market forces are heavily imprinted on the process of wage determination. A three-point rise in unemployment is estimated to lower wage growth by about one point, while the same rise in productivity growth has a positive effect of about the same magnitude. There is also support for nominal wage rigidity in the weakly unionized labor markets. A five-point rise in inflation is associated with a one-point fall in real wage growth. Table 4.4. Regression Results in a Model of Real Wage Growth in 18 OECD Countries, 1965–1993 (absolute t-statistics in parentheses)

134

Wage Growth, Recession, Labor Decline in Industrialized Democracies

The story becomes more complicated once institutions are taken into account. Consistent with the influence of employment goals in corporatist bargaining, the negative interaction effect for corporatism and unemployment indicates that unemployment heavily constrains wage growth in the corporatist countries. Indeed, corporatist institutions more than double the negative impact of unemployment on wages. However, corporatist wage moderation is offset by left governments and collective bargaining coverage. In contrast to the partisan-politics version of corporatist theory, the interaction effect implies that real wage growth was raised by nearly a full percentage point when social democrats were in power. Bargaining coverage usefully captures the power of unions in their role as wagebargaining agents. The bargaining coverage interaction shows that extensive collective agreements allowed real wages to grow, even when unemployment was high. In countries like Sweden and Germany, where collective bargaining covered about four-fifths of the workforce, the large negative effects of unemployment resulting from corporatist wage moderation are nullified. Bargaining coverage also protected wages from increasing inflation. Estimates show that rising inflation fueled real wage growth in the 1960s, in countries with extensive bargaining coverage. In general, unemployment was low and productivity growth was strong in OECD countries between 1964 and 1975. In light of these conditions, the estimates indicate social democratic corporatism effectively redistributed the gains of strong economic performance to workers in the form of real wage growth. The process of wage determination changed dramatically in the mid1970s. These changes are reported in the second column of Table 4.4. Net effects for the post–oil shock period are reported in the third column. In countries like the United States, with weak industrial relations institutions and labor marginalized in national politics through the 1980s, the impact of market forces reflects the political weakness of organized labor. First, the large positive interaction for unemployment (Table 4.4, column 2) indicates that the negative unemployment-wage slope of the golden age was for the most part eliminated after 1975. The unemployment-wage growth curve in countries with weak labor market institutions is illustrated in Figure 4.1a. The downward sloping unemployment effect of the pre–oil shock period is replaced by an upward sloping unemployment-wage relationship after 1975. Under these conditions in the United States, falling unemployment after recessions in 1983 and 1991 has been insufficient to raise living standards. Second, the positive impact of productivity growth 135

Bruce Western and Kieran Healy

Figure 4.1 Wage Growth and Labor Decline, Before and After 1975. (a) Change in the unemployment effect due to the impact of corporatism; (b) change in the unemployment effect due to the impact of bargaining coverage; (c) change in the unemployment effect in noncorporatist countries with low bargaining coverage; (d) change in the left government effect due to the impact of corporatism.

on wages is reduced by half. Workers have less access to the gains from technological improvements compared with the 1960s. Third, there is also some evidence that workers are more exposed to the risk of inflation after 1975 than before. The estimate for inflation growth indicates the negative influence of price rises on real wage growth has increased by about half. The net effect for the post–oil shock period is statistically significant, indi136

Wage Growth, Recession, Labor Decline in Industrialized Democracies

cating that a one-point rise in the inflation rate lowered real wage growth by about one-third of a point. Wage growth in the institutionalized settings also changed significantly after 1975. The entire logic of labor market regulation was essentially reversed. Figure 1b illustrates the changing unemployment effect in corporatist countries. As corporatist bargaining forums were overtaken by local bargaining arrangements, the steep trade-off between wages and unemployment flattened in the corporatist countries, as hypothesized. The weakening influence of unemployment on wages in corporatist countries offers evidence for the eroding influence of corporatist institutions on the earnings process in the post-1975 period. The power of labor movements to protect living standards through left governments and bargaining coverage also diminishes. While extensive bargaining coverage allowed wages to rise in spite of increasing unemployment in the 1960s, the interaction between bargaining coverage and unemployment shows that joblessness disciplined wage growth in unionized labor markets after 1975. In this case, after 1975 the unemployment-wage relationship tilts downward in countries with high bargaining coverage (Figure 4.1c). Finally, the capacity of left governments in corporatist countries to maintain wage growth appears greatly diminished after 1975 (Figure 4.1d). This evidence suggests corporatist public policy measures to protect wage standards were widely undermined in the 1980s and 1990s. To summarize, these results are strongly consistent with the labor decline theory of the wage slowdown. The formal institutions of social democratic corporatism were unable to maintain rising living standards in the 1980s and 1990s in the way that they had in the 1960s and early 1970s.

Discussion This analysis offers two main findings. First, the earnings process in the industrialized democracies is strongly marked by institutions. Before the mid-1970s, central union representatives dampened wage demands when unemployment was threatening. While this supports earlier research on corporatist wage restraint (Korpi 1983; Crouch 1985), the analysis also shows that extensive unions with supportive social democratic governments raised wages rapidly. With weak and decentralized unions, however, wage growth was slower, the trade-off between wage growth and unemployment was flatter, and living standards were closely tied to the inflation rate. Thus, in general, wages rose more quickly, and 137

Bruce Western and Kieran Healy

with less cost to the disadvantaged, in countries with strong centralized labor movements. Second, the process of wage determination changed fundamentally in the mid-1970s. Although the end of the golden age is often associated with the rise of European unemployment in the 1980s and 1990s, we find evidence for an even more fundamental shift in the European labor markets. To be sure, high unemployment and other recessionary conditions were a significant drag on wage growth in the 1980s and 1990s. Not only did the values of key variables change in a way that hurt wages in the mid-1970s; the causal process of wage determination also shifted. The erosion of corporatism and the decline of union bargaining power loosened labor’s institutional grip on incomes in Europe. Earnings received little institutional protection from the effects of unemployment and inflation, and European workers increasingly bore the risk of market fluctuations. These findings can also be placed in the broader context of European integration. Some researchers argue that expansionary policy was a cornerstone of corporatist wage restraint, but Keynesian expansions are increasingly infeasible in Europe. National policy autonomy has been restricted because capital is more mobile across national borders and monetary policy, under the growing influence of the Bundesbank, is becoming more inflation-averse. Mitterrand’s reflationary experiment of 1982 is often taken as an important watershed in the turn away from an activist role for government in macroeconomic management (Hall 1986). The ensuing capital flight in France demonstrated the risks of countercyclical policy. Deflationary pressure is compounded by the European Monetary System and the move for monetary union. By effectively pegging European currencies to the deutsche mark, the monetary system has diffused low German inflation and the deflationary policies that sustain it (Gomes 1993). Because monetary union membership depended on low budget deficits, spending cuts and the abandonment of traditional demand management have accompanied the process of monetary integration (see Chapter 1 in this volume). In short, European governments are now highly constrained in the tax and spending policies that formerly underwrote the corporatist wage bargain. Because unemployment and wage growth are largely unrelated in the corporatist countries through the 1980s and 1990s, there is evidence that centrally organized wage restraint is currently more difficult to achieve. In addition, other research indicates that high European unemployment is partly attributable to a nonaccommodating monetary policy (see Chapter 138

Wage Growth, Recession, Labor Decline in Industrialized Democracies

2 in this volume). Moreover, if the forces of European integration have disarmed social democratic governments, one would expect to see little difference in the effects of left and bourgeois government on real wage trends. Indeed, the data analysis also shows this to be the case and thus supports the idea of a distinctively European path to wage stagnation. In this scenario, the single-market movement has created significant economic and political obstacles to strong wage growth. The European path is institutionally circumscribed, but this time at the regional, not the national, level. From a more theoretical perspective, this analysis also presents paradoxes to neoclassical accounts of the labor market. After a prolonged program of deregulation in the United States, the forces of supply and demand – productivity growth and unemployment – have little bearing on aggregate movements in the price of labor. Deregulation failed to convey the benefits of good economic times to the average worker. By contrast, European wage setting now strongly bears the imprint of market forces. Because union representation remains extensive and wages are set administratively rather than by the market, wages are now highly responsive to labor demand. Although these results are inexplicable from a purely economic perspective, they can be understood with a richer conception of economic power. The evidence here suggests that power in the labor market reflects not just the conditions of supply and demand. Instead, the capacity of labor market actors to obtain rewards depends on their institutionalized control over economic uncertainties. Institutional influence is pervasive. Institutions not only shape labor’s overall allocation; they also affect the impact of economic forces on the labor market. Institutional control through corporatist bargaining, extensive union regulation, and social democratic government successfully raised workers’ living standards through the 1960s until the mid-1970s. From this point, labor movements were weakened, and wages stagnated or fell in the following decades.

Appendix: Data Sources Real Wage Growth: Annual percentage changes in real hourly earnings in manufacturing are mostly taken from OECD (1996), but national sources have been used in some cases, owing to missing data. Australian data have been supplemented by the weekly award average from the average weekly earnings series for male wage and salary earners (Australian Bureau of Statistics 1988: 294; 1989: 190; 1992: 203). Austrian data are from the 139

Bruce Western and Kieran Healy

monthly earnings in mining and manufacturing series of OECD (1993). Danish and Swedish wage data have been compiled from tables of wages in mining and manufacturing published by the Nordic Council of Ministers and the Nordic Statistical Secretariat (1974: 185; 1983: 239; 1991: 253; 1994: 238). (Swedish data were adjusted to account for the inclusion of overtime and holiday pay in the earliest table.) Finally, the Netherlands series was completed with data for the average hourly earnings of male manufacturing sector workers in tables published by the Netherlands Central Bureau of Statistics (1970: 371; 1971: 313; 1975: 305; 1978: 324; 1980: 344). Where national sources were used to complete OECD series, data were smoothed to eliminate discontinuities, and overlapping national and OECD series were compared to ensure comparability. Unemployment: Standardized unemployment rates have been used where available. Unstandardized figures based on the number of unemployed as a percentage of the civilian labor force have been used for Austria, Denmark, Ireland, and Switzerland. All data come from OECD (1996), except for data for New Zealand and Switzerland, which also come from Layard et al. (1991: 526–29). Productivity Growth: Annual percentage changes in real GDP per person employed are taken from OECD (1996). Inflation: Annual percentage changes in the consumer price index are taken from OECD (1996). Left Government: To obtain the percentage of cabinet seats held by left parties, cabinet representation was coded for every quarter. Where there was a change in cabinet representation, the longest-serving cabinet in the quarter was coded. Annual averages were then taken from the quarterly series. Information about the party composition of cabinets is taken from Woldendorp, Keman, and Budge (1993). These data were updated with tables reported in Koole and Mair (1994). Corporatism: This is a time-invariant, cross-nationally varying dummy variable that indicates corporatist countries, as classified by Crouch (1985). Belgium has been recoded as corporatist, and Switzerland recoded as noncorporatist, as discussed in the text. Collective Bargaining Coverage: This is a time-invariant, crossnationally varying measure of the proportion of the workforce covered by collective agreements in 1990 (Traxler 1996). No data are available for Denmark in 1990, so the 1985 figures are used. Traxler does not report data for either Ireland or Italy, so average union density over the time series (calculated from Visser 1992) has been used for these countries. 140

Wage Growth, Recession, Labor Decline in Industrialized Democracies

References Alvarez, R. Michael, Geoffrey Garrett, and Peter Lange. 1992. Government Partisanship, Labor Organization, and Macroeconomic Performance. American Political Science Review 85: 539–56. Australian Bureau of Statistics. 1988, 1989, 1992. Official Year Book of the Commonwealth of Australia. Canberra, Australia: Australian Bureau of Statistics. Baglioni, Guido, and Colin Crouch, eds. 1990. European Industrial Relations: The Challenge of Flexibility. Newbury Park: Sage. Blanchflower, David G., and Richard B. Freeman. 1992. Unionism in the United States and the Other Advanced OECD Countries. Industrial Relations 31: 56–79. Calmfors, Lars, and John Driffill. 1988. Centralisation of Wage Bargaining and Macroeconomic Performance. Economic Policy 6: 13–62. Carlin, Wendy, and David Soskice. 1990. Macroeconomics and the Wage Bargain: A Modern Approach to Employment, Inflation, and the Exchange Rate. New York: Oxford University Press. Colbjørnsen, Tom, and Arne L. Kalleberg. 1988. Spillover, Standardization, and Stratification: Earnings Determination in the United States and Norway. European Sociological Review 4: 20–31. Crouch, Colin. 1985. Conditions for Trade Union Wage Restraint. In Leon Lindberg and Charles Maier, eds., The Politics of Inflation and Economic Stagnation: Theoretical Approaches and International Case Studies. Washington, D.C.: Brookings, 105–39. Erickson, Christopher L., and Andrea C. Ichino. 1995. Wage Differentials in Italy: Market Forces, Institutions, and Inflation. In Richard B. Freeman and Lawrence F. Katz, eds., Differences and Changes in Wage Structures. Chicago: University of Chicago Press, 265–306. Esping-Andersen, Gøsta. 1990. The Three Worlds of Welfare Capitalism. Princeton, N.J.: Princeton University Press. European Industrial Relations Review (EIRR). 1985. Background to Consumer Price Indices. 132: 20–23. ———. 1989a. Trends in the Level of Collective Bargaining. 183: 24–26. ———. 1989b. Trends in the Level of Collective Bargaining: Part Two. 186: 20–22. Flanagan, Robert J., David W. Soskice, and Lloyd Ulman. 1983. Unionism, Economic Stabilization, and Incomes Policies: European Experience. Washington, D.C.: Brookings. Freeman, Richard B., and Lawrence F. Katz. 1995. Introduction and Summary. In Richard B. Freeman and Lawrence F. Katz, eds., Differences and Changes in Wage Structures. Chicago: University of Chicago Press. Friedman, Milton. 1968. The Role of Monetary Policy. American Economic Review 58: 1–17. Golden, Miriam, Peter Lange, and Michael Wallerstein. 1993. Trends in Collective Bargaining and Industrial Relations in Non-Corporatist Countries: A Preliminary Report. Paper presented at the Annual Meetings of the American Political Science Association, Washington, D.C. 141

Bruce Western and Kieran Healy Goldthorpe, John H. 1984. The End of Convergence: Corporatist and Dualist Tendencies in Modern Western Societies. In John H. Goldthorpe, ed., Order and Conflict in Contemporary Capitalism: Studies in the Political Economy of Western European Nations. Oxford: Clarendon Press, 315–43. Gomes, Leonard. 1993. The International Adjustment Mechanisms: From the Gold Standard to the EMS. New York: St Martin’s. Hall, Peter. 1986. Governing the Economy. London: Polity Press. Headey, Bruce. 1970. Trade Unions and National Wages Policy. Journal of Politics 32: 407–39. Hicks, Alexander. 1994. The Social Democratic Corporatist Model of Economic Performance in the Short- and Medium-Run Perspective. In Thomas Janoski and Alexander Hicks, eds., The Comparative Political Economy of the Welfare State. New York: Cambridge University Press, 189–217. Hicks, John. 1963. The Theory of Wages. London: Macmillan. Katz, Harry C. 1985. Shifting Gears: Changing Labor Relations in the U.S. Automobile Industry. Cambridge, Mass.: MIT Press. Katzenstein, Peter J. 1985. Small States in World Markets: Industrial Policy in Europe. Ithaca, N.Y.: Cornell University Press. Koole, Ruud, and Peter Mair. 1994. Political Data Yearbook, 1994: European Journal of Political Research 26, 3/4. Korpi, Walter. 1983. The Democratic Class Struggle. Boston: Routledge and Kegan Paul. Lange, Peter, Michael Wallerstein, and Miriam Golden. 1995. The End of Corporatism? Wage Setting in the Germanic and Nordic Countries. In Sanford Jacoby, ed., The Workers of Nations: Industrial Relations in a Global Economy. New York: Oxford University Press, 76–100. Lash, Scott. 1985. The End of Neo-corporatism? The Breakdown of Centralised Bargaining in Sweden. British Journal of Industrial Relations 23: 215–39. Layard, Richard, Stephen Nickell, and Richard Jackman. 1991. Unemployment: Macroeconomic Performance and the Labour Market. Oxford: Oxford University Press. McCracken, Paul, Guido Carli, Herbert Giersch, Attila Karaosmanoglu, Ryutaro Komiya, Assar Lindbeck, Robert Marjolin, and Robin Matthews. 1977. Towards Full Employment and Price Stability. Paris: OECD. Massey, Patrick. 1995. New Zealand: Market Liberalization in a Developed Economy. New York: St. Martin’s. Mjøset, Lars, Ådne Cappelen, Jan Fagerberg, and Bent Sofus Tranøy. 1994. Norway: Changing the Model. In Perry Anderson and Patrick Camiller, eds., Mapping the West European Left. London: Verso, 55–76. Mitchell, Daniel J. B. 1993. Keynesian, Old Keynesian, and New Keynesian Wage Nominalism. Industrial Relations 32: 1–29. Moene, Karl Ove, and Michael Wallerstein. 1993. The Decline of Social Democracy. In Karl Gunnar Persson, ed., The Economic Development of Denmark and Norway since 1870. Brookfield, Vt.: Elgar, 385–403.

142

Wage Growth, Recession, Labor Decline in Industrialized Democracies Molitor, Michel. 1978. Social Conflicts in Belgium. In Colin Crouch and Alessandro Pizzorno, eds., The Resurgence of Class Conflict in Western Europe Since 1968. London: Macmillan, 1: 21–51. Netherlands Central Bureau of Statistics. 1970, 1971, 1975, 1978, 1980. Statistical Yearbook of the Netherlands. The Hague: Netherlands Central Bureau of Statistics. Nordic Council of Ministers and the Nordic Statistical Secretariat. 1974, 1983, 1991, 1994. Yearbook of Nordic Statistics. Stockholm: Nordic Council of Ministers. Organisation for Economic Co-operation and Development (OECD). 1993. Main Economic Indicators, Historical Statistics: Prices, Labour and Wages. Paris: OECD. ———. 1994. The OECD Jobs Study: Part II – The Adjustment Potential of the Labor Market. Paris: OECD. ———. 1996. Historical Statistics 1960–1994 on Diskette. Paris: OECD. Olson, Mancur. 1982. The Rise and Decline of Nations: Economic Growth, Stagflation, and Social Rigidities. New Haven, Conn.: Yale University Press. Pontusson, Jonas. 1994. Sweden: After the Golden Age. In Perry Anderson and Patrick Camiller, eds., Mapping the West European Left. London: Verso, 23–54. ———. 1995. Explaining the Decline of European Social Democracy: The Role of Structural Economic Change. World Politics 47: 495–533. Piven, Frances Fox. 1991. The Decline of Labor Parties: An Overview. In Frances Fox Piven, ed., Labor Parties in Postindustrial Societies. New York: Oxford University Press, 1–19. Rosenfeld, Rachel A., and Arne L. Kalleberg. 1991. Gender Inequality in the Labor Market: A Cross-National Perspective. Acta Sociologica 34: 207–38. Ross, George, and Jane Jenson. 1994. France: Triumph and Tragedy. In Perry Anderson and Patrick Camiller, eds., Mapping the West European Left. London: Verso, 158–88. Scharpf, Fritz. 1987. Crisis and Choice in European Social Democracy. Ithaca, N.Y.: Cornell University Press. Shalev, Michael. 1992. The Resurgence of Labour Quiescence. In M. Regini, ed., The Future of Labour Movements. London: Sage, 102–32. Stilwell, Frank. 1993. Wages Policy and the Accord. In Greg Mahoney, ed., The Australian Economy Under Labor. St. Leonard, Australia: Allen and Unwin, 65–84. Streeck, Wolfgang. 1993. The Rise and Decline of Neocorporatism. In Lloyd Ulman, Barry Eichengreen, and William Dickens, eds., Labor and an Integrated Europe. Washington, D.C.: Brookings, 80–101. Swenson, Peter. 1989. Fair Shares: Unions, Pay, and Politics in Sweden and West Germany. Ithaca, N.Y.: Cornell University Press. Thelen, Kathleen. 1993. West European Labor in Transition: Sweden and Germany Compared. World Politics 46: 23–49. Tilly, Charles. 1995. Globalization Threatens Labor’s Rights. International Labor and Working Class History 47: 1–23.

143

Bruce Western and Kieran Healy Traxler, Franz. 1996. Collective Bargaining and Industrial Change: A Case of Disorganization? A Comparative Analysis of Eighteen OECD Countries. European Sociological Review 12: 271–87. Visser, Jelle. 1992. Trade Union Membership Database. Unpublished data file. Amsterdam: Department of Sociology, University of Amsterdam. Western, Bruce. 1995. A Comparative Study of Working-Class Disorganization: Union Decline in Eighteen Advanced Capitalist Countries. American Sociological Review 60: 179–201. ———. 1996. Recent Wage Trends in 14 OECD Countries. Paper presented at the meetings of Research Committee 28 of the International Sociological Association. Ann Arbor, Mich. Woldendorp, Jaap, Hans Keman, and Ian Budge. 1993. Parties and Governments in Industrialized Parliamentary Democracies. European Journal of Political Research 24: 1–120. Wood, Adrian. 1994. North-South Trade, Employment, and Inequality: Changing Fortunes in a Skill-Driven World. Oxford: Clarendon.

144

5 Unemployment and Union Density Lyle Scruggs and Peter Lange

For more than two decades, the power and organization of domestic unions have been identified as critical sources of diversity in national policy regimes and outcomes. In the 1970s Schmitter’s discussion of corporatism initiated a sustained discussion of how different systems of “interest intermediation” in the advanced industrial democracies are organized (Schmitter 1979). This discussion eventually came to be focused on trade unions as one of the critical – if not the most critical – elements of such systems. Subsequently, differences in the strength and structure of the labor movement were identified as important factors explaining the social policy regime, the extent and character of industrial and other “supplyside” policies pursued by governments, and differences in macroeconomic performance (e.g., inflation, growth, unemployment) and the particular mixes of them. Particularly relevant for our present discussion, they were also shown to be associated with differences in how different national political economies responded to the economic shocks of the 1970s (Cameron 1984; Bruno and Sachs 1985; Lange and Garrett 1985; Alvarez, Garrett, and Lange, 1991). This association between the strength and structure of organized labor, on the one hand, and responses to international change, on the other, has reappeared as research has come to focus on the effects of globalization on national economies since the beginning of the 1980s. Soskice (1998) maintains that trade union organizational variables are important (though not the only) features distinguishing “organized” and “liberal” market economies, and their characteristic responses to economic internationalization. Iversen (1998) has highlighted the continuing importance of trade union patterns in explaining the effects of central bank independence on economic performance. Garrett and Lange (1991) have shown a 145

Lyle Scruggs and Peter Lange

continued association between labor movement organization and the extent and type of “supply-side” policies pursued by governments. And Pontusson and his colleagues (1996) have found that labor movement characteristics continue to affect patterns of income distribution. Most recently, Garrett (1998), in reviewing the theoretical arguments and empirical evidence (not) supporting the argument that globalization leads to policy convergence, has found correlations between union density (the percentage of trade union members in the dependent labor force) and differences in a wide range of cross-national indicators of government policy outputs and economic outcomes. Taken together, this evidence suggests that even in an era of globalized trade and finance and increasing economic interdependence, differences in the strength and structure of domestic labor movements continue to have a significant impact on differences in domestic policy and national economic performance. If domestic labor movements remain different, it follows, one can expect national policies and patterns of economic performance also to remain so. It is possible, however, that forces are eroding the organizational strength of unions themselves, thereby weakening these mediating structures and hence, with some lag, undermining a traditional pillar of organized market economies. Two changes over the last twenty years are immediately obvious candidates for such a role. One is increasing globalization; the other is increasing unemployment. In an earlier analysis of the argument that globalization is undermining the organizational strength of unions (Lange and Scruggs 1997), we found that globalization is weakly, when at all, related to changes in one critical indicator of the underlying structural strength of trade unions: union density, or the percentage of the national workforce who are members of a trade union. Even when we found discernable relationships, they did not indicate that globalization systematically weakens unions. Thus, to the extent that globalization affects union membership at all, our findings suggest that it promotes divergence more than convergence. In this chapter, we concentrate on the other major change in the political environment in which unions have had to operate since the mid-1970s: the steady rise in unemployment. High and rising unemployment has historically been shown to be associated with declining worker and union militancy as measured by strikes (Ashenfelter and Pencavel 1969). It has also been thought to be associated with declining union organization. The reasons are clear. As unemployment rises, workers feel more vulnerable to the actions of employers, unions are less able to mobilize these vulnerable 146

Unemployment and Union Density

workers in defense of wages and salaries, and employers are emboldened to attack unions and their membership. The threat of job loss makes unionized workers (and their unions) more reluctant to engage in industrial action and serves as a disincentive for workers to join unions in the first place. Similarly, governments, facing the electoral threats posed by higher unemployment, may reduce their support of unions and “inflexible” labor markets. If, as this model assumes, workers and unions, on the one side, and employers, on the other, are locked in an antagonistic struggle, unemployment should be associated with declining union strength. If unemployment increases or higher levels are sustained, one would also expect erosion of union organization. This is the standard perspective of a liberal market view of the economy and economic organizations. Organizational power in the labor market breeds highly inefficient behavior, and thus increasingly undermines national economic prosperity and competitiveness (Olson 1982). However, the perspective emphasizing the benefits of a more organized form of capitalism suggests that the negative effects of organization may be muted or eliminated. Unions, supporting labor market institutions, and a generous welfare state combine to provide insurance against unemployment risks and are thus compatible with economic competitiveness and low unemployment (Calmfors and Driffil 1988; Garrett 1998). Where unions are strong, competition between unionized and nonunionized workers is reduced because most or all workers in most sectors and industries are covered by collective agreements. In such situations, unions have a more “encompassing” interest that considers the unemployed (and nonunion members) when bargaining. In highly organized societies, moreover, previously high union membership loses its distinctiveness as a negative marker for individual workers, and (thus) union organization is less hampered by the threat of employer discrimination. Furthermore, employers may also have weaker incentives to attack unions as institutions because they have generally become “built into” production processes and practices (Regini 1996; Soskice 1998). Finally, given the above conditions, governments, particularly of the Left, are more inclined to support already strong unions, both directly and indirectly, rather than to seek to weaken them. Therefore, at least in countries with “supporting institutions,” the negative association between unemployment and union density would likely be mitigated or reversed. Thus, the central aim of this chapter is to show that institutional arrangements within countries affect the relationship between unemployment and union membership. 147

Lyle Scruggs and Peter Lange

One important institutional feature of the organization of labor markets, which is also (imperfectly) linked to organized capitalism, is the nature of the system of national unemployment insurance. Several scholars have suggested that, in countries where unions administer unemployment insurance systems, increased unemployment should have a positive impact on union density (Rothstein 1992; Western 1997).1 Since workers threatened by possible unemployment believe that union membership would better assure them unemployment benefits, they will join unions when unemployment risks increase. If true, this offers the prospect of a type of “countercyclical” measure that reinforces the strength of worker organizations when they may be most vulnerable. In the remainder of this chapter, we present some data pertaining to these hypotheses. After a discussion of the evolution of unemployment across a broad group of sixteen industrial democracies (including most of Western Europe), we look at the bivariate relationship between unemployment and union density. An exploration of this relationship for all countries, and for subsets based on institutional characteristics, which are indicative of union strength, suggests a highly differentiated impact of unemployment on union density. Finally, we examine a more fully specified statistical model of the determinants of union density change. In that section, we incorporate additional factors likely to affect union density conditional upon some of the institutional features.

Evolution of Unemployment Figure 5.1 shows the movement of (unweighted) average unemployment across sixteen OECD countries from 1964 to 1998.2 The figure reflects the conventional wisdom about the recent history of OECD economies and unemployment. In the period from 1964 to 1974 unemployment was low and stable – varying between 1.9 and 2.6 percent. In only two European countries of the sixteen, Italy and Britain, did unemployment rise above 3 percent between 1960 and 1973. 1

2

The opposite is predicted in countries without such systems, especially where there are not other “supporting” institutions. The countries are Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Italy, Japan, the Netherlands, Norway, Sweden, Switzerland, the United Kingdom, and the United States. Germany is West Germany through 1989 and united Germany from 1990. Unemployment figures are the OECD’s standardized unemployment rates where available and conventional rates where they are not. Almost without exception, the difference between nominal and standard rates is relatively small.

148

149

Figure 5.1

Unemployment: All Countries (percentage).

Lyle Scruggs and Peter Lange

From 1974 to 1983, the period of “stagflation,” average unemployment increased steadily, from 2.6 to 7.4 percent. In Europe, only a few countries – Austria, Norway, Sweden, and Switzerland – avoided large increases in unemployment. It is notable that none of these four countries was an EU member, and thus they were not highly constrained by goals of economic integration. Between 1984 and 1990, following the sustained international economic recovery, average unemployment fell from 7.4 to 5.5 percent. Only in Japan, the United States, and three Western European countries – Finland, Sweden, and Switzerland – did unemployment fall back near its pre-stagflation levels. By 1990, all of the EU countries were still experiencing unemployment at levels not seen prior to the onset of stagflation. Two, Italy and France, actually had higher unemployment in 1990 than they did in 1983. Thus, even accounting for changes in unemployment due to the business cycle, European (particularly EU) unemployment remained very high. Between 1990 and 1993/4, years of recession, unemployment again increased from an average of 5.5 to 8.5 percent. Much of this increase was driven by massive increases in unemployment in Finland (3.2 to 16.8 percent) and Sweden (1.7 to 9.4 percent) and a doubling of unemployment in newly unified Germany (4.2 to 8.4 percent). In the European countries of this group, only Austria, Italy (which already had one of the highest unemployment rates in Europe), and the Netherlands, along with Japan and the United States, managed to reach the mid-1990s without significant increases in unemployment. Following 1993/4, unemployment rates again began to fall, in some places dramatically. Overall, the average unemployment rate went from 8.5 to 6.8 percent. Much of this was due to employment trends in the Nordic countries. Unemployment fell by around 50 percent in Denmark and Norway (10.1 to 5.1 percent and 6.1 to 3.3 percent, respectively) and by at least a third in Finland (16.8 to 11.4 percent), the Netherlands (7.1 to 4 percent), Britain (10.5 to 7 percent), and the United States (7.5 percent in 1992 to 4.5 percent). In Italy and Japan, unemployment was higher in 1998 than its peak in the early 1990s. Overall, what these data suggest is clear: if rising and high unemployment leads to a decline in union density, such effects should appear in the post-1973 period. We direct the bulk of our attention there. The overall average in Figure 5.1 hides significant differences across countries and masks different patterns among countries sharing common 150

Unemployment and Union Density

institutional features in the labor market and the economy more generally. To examine trends along these dimensions, we disaggregate countries into two types of classification. Figure 5.2 shows the evolution of unemployment for countries based on the organization of the unemployment insurance system, that is, Ghent and state.3 In the former, unions operate important (or all) aspects of the unemployment system. This, it is argued, gives unions more of a stake in having low levels of unemployment, and, simultaneously, allows the union to discourage underbidding of bargained wages. In essence, the system permits them to control access to jobs and to monitor the wage rate that those taking those jobs receive (Rothstein 1992). Figure 5.2 shows few and generally short-term differences in the pre-1973 period (although the Ghent countries have slightly lower rates of unemployment). Strikingly, the trends in both Ghent and non-Ghent countries are broadly similar, although unemployment tends to be more volatile – it gets worse in “bad times” but improves more rapidly in “good times” – in Ghent countries. In Figure 5.3, we group countries by their average density levels, under the assumption that this helps to capture the degree of union encompassingness and institutional strength. This, we argue, increases the willingness of unions to pursue strategies that take into account unemployment effects (Lange and Garrett 1985). Those with an average density lower than 33 percent are grouped as low density, those greater than 33 percent but less than 50 percent are grouped as medium density, and those above 50 percent are grouped as high density.4 In the pre-1974 period, the series move in relatively parallel fashion and are tightly bunched at low levels of unemployment. Moderate-density countries have somewhat higher unemployment rates. From 1970 to 1990, the moderatedensity countries consistently have higher unemployment, particularly during the economic recovery of the later 1980s. However, all groups show unemployment levels higher than the pre-1970 period. Though highdensity countries tended to have lower average unemployment through 1990, the series fluctuates more dramatically. This is consistent with many analyses of European countries with highly organized economies: their small size and exposure to international markets makes them sensitive but 3

4

Ghent countries are Belgium, Denmark, Finland, and Sweden. In the other countries, state officials administer unemployment insurance, and it has few or no ties to trade unions. The countries in each group are Low: Canada, France, Germany, Japan, Netherlands, Switzerland, and the United States; moderate: Australia, Austria, Italy, the United Kingdom; high: Belgium, Denmark, Finland, Norway, Sweden.

151

152

Figure 5.2

Unemployment and Insurance Regime (percentage).

153

Figure 5.3

Unemployment and Historical Density (percentage).

Lyle Scruggs and Peter Lange

responsive to the global macroeconomic winds (Katzenstein 1985; Garrett 1998). After the collapse of the Soviet Union and reunification of Germany, unemployment in the high-density countries shot upward dramatically, greatly exceeding unemployment rates in the moderately unionized countries. For many, this change signaled the long-awaited death of small, strong labor, “neocorporatist” countries (Schwartz 1994). However, as in past periods of rising unemployment in the “high-density” countries, economic actors responded quickly, and unemployment, in turn, declined dramatically. These results, particularly for the “middle-density” countries, nevertheless raise an interesting question: could unemployment rates be, at least in part, caused by, or at least strongly associated with, the broader institutional configurations (corporatism, liberalism, or “mixed”)? This is a question to which we return. Three implications of these data should be noted. First, they suggest that the impact of unemployment on density should be larger and also more differentiated across countries grouped by labor’s historical strength in the post-1973 period than before. Second, they suggest that if there is indeed an inverse relationship between unemployment rates and density, the largest would be most clear in the medium-strength cases. Finally, these levels and trends raise the question of the extent to which the unemployment rates are reflective of underlying institutional configurations partly captured by density levels. If this is the case, a causal argument from unemployment to density would be misleading, for both density and unemployment could, at least in part, be endogenous to the broader political economic configurations of institutions and policies. We believe “corporatist” or “organized market” configurations of institutions tend to promote both higher union density and lower unemployment as part of the overall operation of the system (Kitschelt et al. 1998; Soskice 1998). In “corporatist” countries, government, employers, unions, and workers respond to shocks that might undermine union density, or promote unemployment, in ways that blunt such effects (Goldthorpe 1984). This is so because institutional arrangements facilitate (perhaps even require) rapid adjustment. Conversely, in “liberal market economies,” employers and governments seek to take advantage of such shocks at the expense of union members (and equality), and unions are only weakly able to resist them. Unemployment declines, but often “at the expense” of unions. 154

Unemployment and Union Density

Density Trends What of union density? How does it move in this period and how does its movement track with that of unemployment? In Figure 5.4 we show the movements of the averages over all countries for the period 1964–94. This aggregate series has several striking features. First, looking just at density, we see that despite the sustained loss in average density since about 1980 so widely noted in the literature, average union density across these sixteen countries in 1994 was about as high as in 1970 (for a discussion, see Golden, Wallerstein, and Lange 1997). This, of course, masks substantial differences across countries and groups of countries, as we explain shortly, but is worth underlining when there are widespread discussions of a “crisis” of unionism under the impact of globalization and/or other changes in the domestic and international environments of unions. Second, rather than moving inversely, for most of the period the unemployment and density series move in parallel: when unemployment is falling, density is falling, when unemployment is rising, density is rising. Thus, for most of the postwar period, when looked at in the aggregate and without paying heed to national differences, unemployment and density have not been inversely, but rather directly, correlated. This is not always true. In particular, in the early 1980s density began a protracted fall while unemployment was rising. However, density leveled off and even rose slightly in the 1990s, while unemployment lurched upward.5 Third, perhaps most surprisingly, whenever the series move in parallel, trends in density anticipate, rather than follow, trends in unemployment. Density began a sustained surge in the late 1960s just before unemployment rose, and it started falling in the early 1980s approximately five years before unemployment peaked. This pattern, in conjunction with the direct, rather than inverse relationship of the two series, calls for explanation. We cannot offer a lengthy explanation here, but at least two, not mutually exclusive, explanations are possible. First, it may well be that the two series are only weakly related and that other phenomena explain substantial parts of the movements in one, and at times in both. Looking at Figure 5.4, for instance, it is likely that the upward turn in density beginning in 5

Our density data use the dependent labor force (including unemployed) as the denominator. Thus, unlike some measures of union density, ours is not likely to be due to declines in the denominator.

155

156

Figure 5.4

Average Density and Average Unemployment (percentage).

Unemployment and Union Density

the late 1960s is a function of the explosion of militancy across union movements in the advanced industrial democracies that occurred in that period. The sustained low levels of unemployment in preceding years may have contributed to that phenomenon, but its causes are certainly more complex. It is also likely that the sharp upturn in unemployment in the mid-1970s was not caused by the rising trends in density. A far more likely culprit is the first shock triggered by the Organization of Petroleum Exporting Countries and the ensuing recession. It is also likely, however, that unemployment was in part caused by wages that were downwardly rigid. That, in turn, was almost certainly a function of union strength, at least in a number of countries where corporatist practices – explicit bargaining arrangements to elicit wage moderation in exchange for various kinds of social insurance, and enforced by peak-level leaders of unions, employer associations, and government – were not available to dampen such effects through wage regulation. The same could be said of trends in the 1990s. Unemployment was sparked in large measure by a combination of shocks to specific countries. The collapse of the Soviet Union explains much of the increase in Finnish unemployment. Sweden’s economic downturn a couple of years later was caused by entry into the European Monetary System (MS) coterminous with three demand “shocks” coming on top of each other: limits on public spending, response to a “burst bubble” in the real estate sector, and Germany’s deflationary approach to reunification (Soskice n.d.). But the caveat about corporatist countries raises the second possibility: that the aggregate series hides the differences between the types of national institutional systems and that the relationship between unemployment and density differs substantially between the groups. Figures 5.5 and 5.6 allow us to explore this possibility. In Figure 5.5, we show the density series for all countries, flanked, as it turns out, by the series for the Ghent and non-Ghent countries. The differences between the two groups are dramatic: the Ghent countries have had a steady increase in average density, albeit at different rates across time, the non-Ghent countries have had an almost regular decrease. When we recall the unemployment series reported earlier, the steadiness of these trends is remarkable, suggesting that, in both groups of countries, density may be differentially related to unemployment because of this institution, as some have suggested. Figure 5.6 reports the density for the threefold grouping discussed previously. Here the differences across groups are even more striking. The 157

158

Figure 5.5 Density and Insurance Regime (percentage).

159

Figure 5.6

Density by Historic Density Group (percentage).

Lyle Scruggs and Peter Lange

high group has a smooth upward trend, which is largely impervious to fluctuations in unemployment (see Figure 5.3). Conversely, the low-density countries, again seemingly impervious to changes in unemployment, show an almost continuous downward trend in density. Also, consistent with our hypothesis, we see that in the medium-density countries, density exhibits a downward trend both prior to 1968 and after about 1980. The movements across these groups (particularly the high- and low-density cases) are consistent with our general argument that the strategic behavior of workers, unions, employers, and governments will differ markedly on the basis of whether unions have been strong for long periods of time. This observation returns us to the issue of the broader relationship between unemployment and density. To bring more sharply into focus the question of the extent to which unemployment is a major factor in explaining density across all groups or for particular groups, we turn to some multivariate, econometric results. As there are many things that are probably causing union density to go up or down, this type of analysis allows us to take some of those factors into account.6 The approach should also allow us to address more directly the question of how other aspects of the social market economy help determine union density levels.

Econometric Results In order to examine the effects of unemployment on union density while controlling for other economic and political factors that might also affect density, we performed a statistical analysis on the data for the years 1964–94. We estimated the following basic equation: density = b unemployment + b government employment + b government ideology + socioeconomic institutions We have been discussing the relationship between unemployment and density extensively in this chapter, but we should say a word about the other factors in the model.7 All can be seen as important factors 6

7

For example, union membership may be negatively affected by the rate of unemployment and how fast the labor force grows. If both are growing at the same time, it is hard to discern which is related to the change in density. By looking at many countries over many years and the fluctuations in unemployment, labor force growth, and union density in those “country-years,” one can better grasp which factor has which, and how much, effect. The model includes a number of other control variables for economic fluctuations and the degree of globalization. Since these were secondary to the discussion here, we exclude them from the tables.

160

Unemployment and Union Density

differentiating organized market economies (OMEs) and liberal market economies. • Government employment, measured as the percentage of employees in the public sector, is a very straightforward way to measure the degree to which governments intervene to prop up employment. • Government partisanship measures the partisanship of the ruling party/coalition. Governments in which social democratic parties dominate (hold a higher percentage of cabinet posts) are scored lower, and those dominated more by right-wing parties are scored higher; those dominated by centrist parties (or, for example, equal representation of left and right parties) are in the middle. • Socioeconomic institutions measures the effects of those features of the institutional environment that we have been discussing, namely, union strength and the system of unemployment insurance. We also look at the impact of broader institutional aspects of the social- versus liberal-market economy. We use a division of countries into these two groups suggested by Soskice (1998). The institutional features (Ghent/state unemployment insurance, labor institutional strength/weakness and social/liberal market economy) should affect not only the level of union density, but also the way in which the other, noninstitutional factors (specifically unemployment, government ideology, government employment) affect density. Thus, we analyzed the impact of unemployment, government employment, and partisanship on density conditional upon a country having/not having a Ghent system, historically strong unions, or a social (liberal) market economy (SME).8 What are our expectations about how these different factors will affect union density? We expect unemployment to have a negative effect on union density, except where there is a union-run unemployment system. The effect of unemployment in lowering density, however, should be smaller in (a) countries with historically high density and/or (b) social market economies. We expect public sector employment to provide a boost to union density, because public employees tend to face less resistance to organization, and they help maintain solidarity in the private sector by lowering unemployment risks. The effect of public employment should be less pronounced in Ghent countries, since, if public employment prevents 8

This is more formally known as a regression with dummy variable interaction terms (see Hardy 1993). The details of the procedure are excluded from the following discussion, but are briefly discussed in the appendix.

161

Lyle Scruggs and Peter Lange

unemployment, there is less need for unemployment insurance, and thus less incentive to join unions.9 In social market economies, the stronger unions are historically, the less the impact of government employment on density; but the weaker unions are historically, the more public sector employment growth will increase union density. The latter also holds in liberal market economies. Finally, we expect left governments to have a positive impact on density in all cases, though the relative impact will be lower where unions are historically strong or are present in organized market economies. In the latter two contexts, there is likely to be less resistance to unions, and thus less pressure will be placed on them by more right-wing governments. Figure 5.7 presents the results of our regression analysis. More details on the statistical methods used appear in the appendix. Because the data for some variables are limited, the analysis is restricted to the period 1965 to 1994. Simplifying a bit, the results reported in Figure 5.7 indicate the effect of a one-unit change in the independent variable on union density. Thus, the coefficient for the unemployment variable would indicate how much a percentage point increase would change union density, controlling for all of the other factors in the model. Since the coefficients are statistical estimates, they have an estimate error associated with them. Generally speaking, for a coefficient to be considered “statistically significant” from zero, the error associated with the estimate should be about half the value of the estimate itself.

Unemployment In interpreting these results, we first discuss the effects of unemployment variable, since that is of most interest here. We then discuss the impact of government employment and partisanship. The different effects of unemployment on density in Ghent and non-Ghent countries are dramatic and as we expected (panel A). In Ghent countries, the impact of a percentage point increase in unemployment leads to an increase in density of about 3 percentage points. However, this effect is almost, but not quite, statistically different from a zero effect. In non-Ghent countries, we can be confident that a similar increase in unemployment will reduce density by about 2.5 9

Alternatively, since union membership in Ghent countries is somewhat impervious to increased unemployment, public employment will not help Ghent countries as much as non-Ghent ones.

162

163

Figure 5.7

Effects of Institutional Differences on Predictors of Union Density.

Lyle Scruggs and Peter Lange

percentage points. There is thus a discernable difference of 5.5 points in the average effect of changes in unemployment in the two groups of countries. If we examine the difference between historically high-, middle-, and lowdensity countries, there are strong negative effects from unemployment in middle- and low-density countries, and no significant effect in highdensity countries (panel B). Most interesting however, is the fact that the impact of increased unemployment has a greater effect on density in the moderately unionized countries than in weak union countries. Density declines by close to 2 points in middle-density countries when unemployment goes up, but by only 1 point in the weak union countries. This is probably due to the fact that there is a higher base from which to decline in the latter; but the result is still consistent with the argument that unemployment does take its toll on unions where they are not clearly encompassing groups. The effects of being an organized market economy versus a liberal market economy also suggest notable differences in the relationship between unemployment and union density (panel C). Unionization in liberal economies, as expected, suffers considerably from increased unemployment, while unions in social market economies, on average, enjoy a very small positive effect. This result is consistent with the idea that in social market economies, there are fewer (successful) attempts to undermine unions in economic downturns. In general, then, the impact of unemployment on union density tends to be as we predicted earlier. Unemployment seems to have a beneficial effect on union density where unions run unemployment insurance schemes. Moreover, unions in societies characterized by historically strong unions and social market economies are less likely to decline as a result of increased unemployment. However, where unions are only moderately strong, they face much dimmer prospects of thriving in periods of high unemployment. Where unions are weak or a country’s overall political economic institutional environment is liberal, unemployment undermines union density.

Government Employment What do these results tell us about the impact of government employment on density? The results differ considerably depending on which of the three sets of results one examines. They look about as we predicted for 164

Unemployment and Union Density

Ghent and non-Ghent countries; the effects are about the same, but we have more confidence that the effects for the non-Ghent countries are also positive. According to panels B and C of Figure 5.7, the evidence is contradictory. As we projected, the impact of public sector employment on density is actually higher in liberal economies (panel C), probably because the public sector tends to be the last “haven” for unions: unionization there is comparatively low compared with that in other countries, but is much, much higher compared with that in the private sector. On the other hand, the evidence from panel B suggests that the effect of public employment on density is small and insignificant where unionization levels have been lowest historically, and highest where they have been strongest historically. In moderately organized labor movements, as in the highly organized ones, the estimated impact of increased public employment is positive and is almost significant statistically. How can we reconcile these two, almost contradictory, results? One way is to recognize that some of the countries Soskice (1998) classifies as “social market economies” (SMEs) in fact have relatively weak public sector labor movements (e.g., Switzerland and Germany). The notion of using the public sector to assist the labor movement there is not likely, and, even if pursued as a way to reduce unemployment, is not likely to affect density levels much. In contrast, the most highly institutionalized labor movements (they are also SMEs) do have strong public sector unions and did use the public sector as a way to maintain low unemployment and high density. In other words, SMEs operate in very different ways, with quite different structural implications for labor and labor movements.

Paritisanship The results in Figure 5.7 indicate very imprecise and uncertain effects of government partisanship on union density. None of the estimates for any of the groupings of countries is significantly different from zero. The estimated effects are worth mentioning (though they should be interpreted with extreme caution), because they are consistent across the different parceling up of the data. First, all of the estimates suggest a negative effect, as expected. (Recall that a higher score on partisanship means greater control by conservative parties.) Second, in all three panels, it is in the more organized/institutionalized group of countries that partisanship is predicted to matter more. This contradicts our expectations and hints that 165

Lyle Scruggs and Peter Lange

left parties have a beneficial effect on density, even where unions have many institutional advantages.

Conclusion Union strength has been consistently shown to have important effects on the patterns of economic performance and distribution in the advanced industrial democracies. This was certainly the case in the years prior to the acceleration in the global engagement of these countries and appears to be the case subsequently as well. The question is whether factors in the recent period are eroding the institutional strength of unions and thereby weakening their ability to affect how market forces impinge on the wellbeing of different groups within society. In this chapter we have focused on the possible role of unemployment in weakening unions’ strength as measured by their membership. Two findings stand out. First, the overall role of unemployment in affecting the institutional strength of unions is considerable. It does seem to help explain how rates of unionization move in different groups of countries. However, to the extent that unemployment plays a role, its role differs markedly across societies with different kinds of labor market and political economic institutions. The most important one appears to be the Ghent system of unemployment insurance. By providing a kind of “countercyclical” incentive to join or maintain union membership in economic downturns (i.e., rising unemployment), the Ghent system may allow worker organizations to weather storms that might otherwise undermine those organizations. Also important, however, are the impact of union strength historically and other features of the organized market economy on how rising unemployment affects developments in density. We have also suggested that this may reflect broader institutional dynamics linked to different political economic configurations for which the level of union density is only an indicator. Moreover, there is undoubtedly truth in the idea that institutions such as the Ghent system can only be obtained (and retained) by a strong union movement. The implication of this analysis is certainly that the institutional diversity in the political economies of the advanced industrial democracies is likely to be maintained, and may perhaps even grow, under conditions of sustained unemployment. The natural environment facing union movements is like a two-dimensional surface with two resting points: one being 166

Unemployment and Union Density

“full worker organization,” the other “no work organization.” Where unions are strongest, they appear able to retain members, even when labor markets are flaccid and union political power is low. Despite the general fact that labor organizations are weakened by unfavorable economic conditions, high levels of organization provide scope for a revival of strong union influence. Where they are weak, union weakness begets greater weakness, leading toward the economist’s idealized model of a decentralized labor market. This is due in part, we have argued, to barriers to coordination faced by those worker organizations that do exist and the rational behavior of employers. This generally presages a move in such countries toward the eradication of trade unions and is consistent with many claims of the “end of unions” in the conventional wisdom. We should note, however, that it is “conventional wisdom” because it characterizes the main Englishspeaking countries (the United States and Britain), not all industrial countries. In the “liberal” cases, improved labor market conditions are likely to find unions so severely weakened (or discredited) that they will be unable to regain much strength. Those with moderate levels of institutionalized strength, and some of the SMEs have witnessed union declines since the 1980s (or earlier). If unionization rates are an indicator of worker strength, this may suggest that many SMEs will come to look more like liberal countries. On the other hand, some, such as the Netherlands, may have stemmed the tide of a slide to liberal capitalism by reaffirming structured wage moderation (Visser 1997).10 One problem, here, however, is that the agreements have been made on the basis of declining membership. In a sense, the middletier countries appear to have slid somewhat toward the “no organization” state; but this is not a forgone conclusion (yet at least) for some of them. In drawing these conclusions, we should insist upon one caveat. We have examined only union strength as measured by union density. There is no question that a number of other factors affect the political and market influence of unions, among which is their potential for militancy. Nothing we have argued has suggested that the well-established inverse relationship between unemployment and reduced militancy does not hold. To the extent that unemployment is sustained, therefore, we would expect the immediate influence of unions in the labor market and the polity to be 10

Whether and under what terms this is sustainable remains unclear. Dutch unions remain weak, and some have argued that such “last gasps” at neocorporatism without strong unions are simply a well-ordered exit.

167

Lyle Scruggs and Peter Lange

weakened. What does seem clear, however, is that, at least in some countries, such unemployment has not severely undercut the institutional strength of unions. To the extent that is the case, a revival of the labor market may also lead to a revival of union influence.

Appendix The regression estimates used ordinary least squares with panel-corrected standard errors, and correction for heteroskedastic errors. The dependent variable is net union density (as discussed in the text), and a lagged dependent variable specification was used (Beck and Katz 1995). Each institutional variable (unemployment insurance system, historical density category, and social/liberal market economy) and corresponding interaction terms (unemployment, government employment, and government partisanship) were entered in a separate model, along with the following control variables: country dummies, period dummies (1974–79, 1980–84, 1985–90, and 1991–95), measures of financial openness, and trade openness in order to account for effects of internationalization globalization. Including other macroeconomic factors, such as inflation or economic growth, had only marginal estimates. None of the additional regressors were interacted with the institutional variables and were thus assumed to have the same effect. All significance measures assume two-tailed tests. In the text, “significant” means p < 0.05 and “almost significant” means p < 0.09. The lagged dependent variable term strongly suggested that the density series are very “sticky.” The coefficient was 0.95 in the Ghent and social/liberal market economy models and 0.90 in the historical union strength model. Thus, the ultimate effects reported in Figure 5.7 are very long term (>50 years) and approximate. Soskice’s measure of social/liberal market economy excludes France and only takes in northern Italy (in concluding that it is an SME). To avoid confusion, both countries were simply coded “missing,” and thus estimates do not include these two countries.

Data sources Government partisanship – Swank (1998) Government employment – Stephens and Huber (1998) Unemployment – OECD, Quarterly Labor Force Series Union density – Ebbinghaus and Visser (2000), Visser (1991) 168

Unemployment and Union Density

Financial openness – Quinn and Inclan (1997) Trade openness – OECD, National Accounts Results, and the data used in the calculations, are available from the authors. References Alvarez, R. M., G. Garrett, and P. Lange. 1991. Government Partisanship, Labor Organization, and Macroeconomic Performance. American Political Science Review 85: 541–56. Ashenfelter, O., and J. H. Pencavel. 1969. American Trade Union Growth: 1900–1960. Quarterly Journal of Economics 83: 434–48. Beck, Nathaniel, and Jonathan Katz. 1995. Nuisance versus Substance: Specifying and Estimating Time-Series-Cross-Section Models. Manuscript. Bruno, M., and J. Sachs. 1985. Economics of Worldwide Stagflation. Cambridge, Mass.: Harvard University Press. Calmfors, Lars, and John Driffil. 1988. Bargaining Structure, Corporatism and Macroeconomic Performance. Economic Policy 6: 13–61. Cameron, D. R. 1984. Social Democracy: Corporatism, Labour Quiescence, and the Representation of Economic Interest in Advanced Capitalist Society. In J. H. Goldthorpe, ed., Order and Conflict in Contemporary Capitalism. Oxford: Oxford University Press. Ebbinghaus, Bernhard, and Jelle Visser. 2000. Trade Unions in Western Europe since 1945: Organizations and Membership. London: Macmillan. Garrett, G. 1998. Partisan Politics in the Global Economy. New York: Cambridge University Press. Garrett, G., and P. Lange. 1991. Political Responses to Interdependence: What’s “Left” for the Left? International Organization 45: 539–64. Golden, M., M. Wallerstein, and P. Lange. 1997. Unions, Employers Associations and Wage-Setting Institutions in North and Central Europe, 1950–1992. Industrial and Labor Relations Review 50, 3: 379–401. Goldthorpe, J., ed. 1984. Conclusion. Order and Conflict in Contemporary Capitalism. New York: Clarendon Press. Hardy, Melissa. 1993. Regression with Dummy Variables. Newbury Park: Sage. Iversen, T. 1998. Wage Bargaining, Central Bank Independence, and the Real Effects of Money. International Organization 52, 3: 469–504. Katzenstein, P. 1985. Small States in World Markets. Ithaca, N.Y.: Cornell University Press. Kitschelt, H., P. Lange, G. Marks, and J. Stephens, eds. 1998. Continuity and Change in Contemporary Capitalism. New York: Cambridge University Press. Lange, P., and G. Garrett. 1985. The Politics of Growth: Strategic Interaction and Economic Performance in the Advanced Industrial Democracies 1974–1980. Journal of Politics 47: 3. Lange, P., and L. Scruggs. 1997. A Crisis of Unionism? Developments of Trade Union Power in the 1980’s in Postwar Perspective. Paper presented at the 1996 APSA Annual Meeting, San Francisco. 169

Lyle Scruggs and Peter Lange Olson, Mancur. 1982. The Rise and Decline of Nations. New Haven, Conn.: Yale University Press. Pontusson, J., F. D. Rueda, and W. Edwards. 1996. Comparative Political Economy of Wage Inequality in OECD Countries. Paper presented at the 1996 APSA Annual Meetings, San Francisco. Quinn, D., and C. Inclan. 1997. The Origins of Financial Openness: A Study of Current and Capital Account Liberalization. American Journal of Political Science 41 (July): 771–813. Regini, M. 1996. Uncertain Boundaries: The Social and Political Construction of European Political Economies. New York: Cambridge University Press. Rothstein, B. 1992. Labor Market Institutions and Working-Class Strength. In Sven Steinmo, Kathy Thelan, and Frank Longstreth, eds., Structuring Politics: Historical Institutionalism in Comparative Analysis. New York: Cambridge University Press, 35–56. Schmitter, P. 1979. Still the Century of Corporatism. In P. Schmitter and G. Lehmbruch, eds., Trends Toward Corporatist Intermediation. Beverly Hills: Sage. Schwartz, Herman. 1994. Small States in Big Trouble: State Reorganization in Australia, Denmark, New Zealand, and Sweden in the 1980s. World Politics 46: 527–55. Soskice, David. 1998. Divergent Production Regimes: Coordinated and Uncoordinated Market Economics in the 1980s and 1990s. In H. Kitschelt, P. Lange, G. Marks, and J. Stephens, eds., Continuity and Change in Contemporary Capitalism. New York: Cambridge University Press. ———. N.d. Macroeconomic Analysis and the Political Economy of Unemployment. Science Center Berlin. Manuscript. Stephens, John, and Evelyn Huber. 1998. Internationalization and the Social Democratic Model. Comparative Political Studies 31, 3: 353–97. Swank, D. 1998. Funding the Welfare State: Globalization and the Taxation of Business in Advanced Market Economies. Political Studies 46, 3: 671–92. Visser, J. 1991. Trends in Trade Union Membership. OECD Employment Outlook 1991. Paris: OECD, 97–134. ———. 1997. A Dutch Miracle: Job Growth, Corporatism and Welfare Reform in the Netherlands. Amesterdam: Amsterdam University Press. Western, B. 1997. Between Class and Market: Postwar Unionization in the Capitalist Democracies. Princeton, N.J.: Princeton University Press.

170

PA R T I I I

Unemployment and Domestic Bargaining Institutions: Three Cases

6 Unemployment and Trade Union Strength in Portugal Alan Stoleroff

The impact of economic “crisis” and unemployment upon unions became a concern following the oil shocks of the 1970s (see Armingeon et al. 1981; Martin and Ross 1984; Mouriaux 1986). The present interest in the relationship between unemployment and unions in Europe follows an additional series of short-term economic cycles throughout which unemployment has continually increased, reaching unprecedented levels, and union strength has correspondingly – though not universally – tended to decline.1 Union decline has been associated with factors such as the continuously changing composition of the labor force with the decline of industry in relation to services and feminization, rapid technological change displacing traditional manual worker union strongholds, shifts in the international division of industrial labor and industrial restructuring, changes in employer strategies and management practices and changes in the role of the state. Finally, it has also been emphasized that in numerous national cultures the values of workers were changing and attitudes toward unions were thus also changing to their disfavor. In the 1990s the consequences of such changes and mounting economic competition had a snowball effect upon employment, and attention has been increasingly drawn to 1

By union strength I mean among others things unionization, mobilization capacities, bargaining power, organizational stability, and political influence. Numerous authors have been led to speak of the “crisis” of trade unionism. See Caire (1990) for a good short review of this question, somewhat centered on French considerations. In France the crisis of unionism appears to many observers as particularly grave. Note the title of the book by Labbé and Croisat (1992), The End of the Trade Unions? See also Baumard and Blanchot (1994). More balanced work can be found outside of France; see, for example, Visser (1990, 1995) or the collection of essays in Hyman and Ferner (1994).

173

Alan Stoleroff

structurally grounded and long-term unemployment as further impediments to future union development. It is normal to suppose that union strength is dependent upon the conditions of the labor market. Indeed, this generally remains a good rule of thumb – at least with regard to economic downturn and growing unemployment.2 Very few workers are sufficiently insulated from economic insecurity by legal guarantees, institutional immunity, or labor market advantage for high unemployment not to have a depressing effect upon their willingness for collective action. Although economic insecurity may lead workers to join unions as a protective device, it would not mitigate the overall loss in union strength brought about by a recession and unemployment. There is little doubt that unemployment has been an important contributor to the decline in union strength over the past quarter century. Clearly, however, its impact is differentiated by the degree to which national institutions of regulation mediate the effect of such fluctuations in the labor market. With regard to union membership, countries where unions play a role in the administration of unemployment insurance have been significantly more resistant to decline. A corollary question emerges from the persistence of union decline in spite of intermittent phases of increasing demand for labor and lower unemployment. The difficulties faced by unions have to do with employment precariousness and long-term change in the structure and composition of the labor force more broadly (Visser 1994). They are also related to the apparent changes in labor strategies of employers (Kochan, Katz, and McKersie 1986) and whether unions are able to adapt their previous strategies to new circumstances. This chapter therefore focuses on the effects of unemployment on union development in Portugal over the past quarter century. It is my contention that unemployment has not had an undifferentiated and independent effect upon union strength in Portugal. In other words,

2

Union growth is the product of an articulation of factors and theories differ according to the weight placed upon economic, psychological, and political determinants, etc. Within the problematic of union growth, however, the effect of unemployment upon unions occupies an important place (see Dunlop 1987). At given levels of regime stability, shifting economic conjunctures and labor market trends tend to be the most immediate contexts influencing the short-run ebb and flow of trade unionism (see Davis 1941). Visser (1990: 68) continues in the line of Davis, specifying: “it is probably the rate of change of unemployment and the direction of that change, and not the absolute level of unemployment, which has the greatest impact on union growth.”

174

Unemployment and Trade Union Strength in Portugal

the tendency toward declining union strength has not been uniquely associated with rising unemployment. Its effect has very much depended upon the interplay of broader political and economic development and underlying social and labor market trends. I therefore seek to relate the decline of union strength to unemployment and broader labor market change, taking into account the overall political and organizational evolution of trade unionism. As can be seen from Tables 6.1 and 6.2,3 unemployment in Portugal has passed through several distinct phases over the past quarter century. With the revolutionary period, unemployment began an ascending curve, reaching its peak in the second semester of 1978 and first semester of 1979 at over 8%. In 1983 it began to rise again significantly, reaching a new peak of over 10% in the second semester of 1983 and during 1984. Between 1987 and 1991–92 there was a significant decrease in the unemployment rate. After 1993 the unemployment rate began to increase once again. The analysis that follows highlights the impact that the economic crisis and unemployment of the mid-1980s had on Portuguese trade unionism. This period revealed the fragility of industrial relations in the wake of the revolution and democratic transition and introduced the dynamics behind the union decline during the periods of employment growth after 1986 and the downturn after 1992. The analysis also treats the employment growth and decline of unemployment from 1986 to 1992 because, if a linear relationship between economic conjuncture and union growth were to hold, this should have been a period of positive development for trade unionism. I show that it was not. In comparison with other European countries, Portugal has experienced somewhat low unemployment throughout the past decade. If there were a linear relationship between unemployment and union strength, Portuguese unions should have done better than some European counterparts. But all things are not equal, and union strength is dependent upon factors structuring the labor market (such as whether employment growth has taken place in fields and among categories of employees prone to unionization), as well as institutional factors favoring or disfavoring the role of unions in economic regulation. Unemployment alone cannot explain the relative weakness and decline of Portuguese trade unionism that set in during the second half of the 1980s. It is therefore necessary to 3

The two figures correspond to two incompatible series of statistics offered by the Labor Ministry over the years.

175

176

Table 6.1. Unemployment Rates (Full Sense), 1974–1987 (percentage)

Table 6.2. Unemployment Rates (Restricted Sense), 1984–1995 (percentage)

Unemployment and Trade Union Strength in Portugal

place the evolution of Portuguese unionism in the framework of other factors characteristic of employment restricting the development of trade unionsm.

From Democratic Transition to Economic Crisis In the period preceding the revolution, unemployment had been checked by the massive northward emigration of excess population and military mobilization. The unemployment of the revolutionary period was provoked by political uncertainty. Capital flight and dismissals were then followed by a massive and rapid influx of retornados into the active labor force upon decolonization (see Mil-Homens 1984). This led to rising unemployment, which reached the level of 8% by 1978. But in this context and, afterward, in the wake of revolutionary action, trade unions were in their ascendency. Class mobilization sparked and sustained the formative efforts for union organization in defiance of the relative economic insecurity. Unemployment was not demoralizing on a massive scale because it was perceived as an expected consequence, indeed, an instrument of capital’s class struggle, and there was a possibility of socialism – which, many workers thought, would resolve the problem of unemployment once and for all. In any case, even this level of unemployment was sustainable given the tremendous shocks that regime change, decolonization, and nationalization had brought upon the Portuguese economy. The unions emerged from the revolutionary period in a strong position and had become the principal vehicle of working-class struggle with capital. Unionization probably reached its apex in this postrevolutionary period. The unions seemed to be capable of mobilizing workers massively for political struggle, perhaps even more so than strike action. In this period, the unions were able to extend the coverage of collective regulation across the economy. Their considerable bargaining power was reflected in the enterprise and frame agreements obtained in the nationalized firms and sectoral collective bargaining between 1977 and the early 1980s. Initially, the unions even attained substantial bargaining power and, with the contribution of their political influence, were able to increase the share of labor in the national product.4 The influence of labor within or upon the political 4

Union bargaining power was decisively weakened by the balance of payments imbalance which acted as a tight lid upon industrial relations (see Stoleroff 1990). Labor’s postrevolutionary gains in terms of the share of national product were actually quickly dissipat-

177

Alan Stoleroff

parties was very strong.5 The unions in this period were able to assert their strength through social movement and militance. In spite of the fundamental split within the trade union movement in 1978,6 the Communistled Confederação Geral dos Trabalhadores Portugueses (CGTP) was able to mobilize workers in politically motivated strikes against the trend to what was referred to as “capitalist restoration” throughout the right-wing coalition government from 1979 to 1982. This capacity was broken with the economic depression and inflationary crisis of 1983.7 Unemployment shot up from 7.4% in 1982 to 10.5% by 1984. The economic crisis broke the continuity of the union movement that had been on the upswing since the final period of the dictatorship. In the subsequent period, as political uncertainty was eclipsed by economic uncertainty, the structural weaknesses in Portuguese trade unionism became all too visible.

5

6

7

ed. Wages as a share of national income occupied 49.2% in 1973, increasing substantially to 54.6% and 64.7% in 1974 and 1975, respectively. The decrease began as early as 1976 (to 63.8%) and continued to about 50.2% by 1983 (Silva 1985: 270–71). As inflation took off, the state made consistent efforts to hem in wages and labor costs. This effort was behind the corporatistic strategy and austerity policy of the state under the PS/PSD Bloco Central coalition government to create the Conselho Permanente da Concertação Social in 1984. The inverse relation was perhaps even more salient, and a clear political instrumentalization of unions was established. The trade union law originally passed during the revolutionary phase of 1975 had established the Confederação Geral dos Trabalhadores Portugueses-Intersindical Nacional (CGTP-IN) as the sole legal confederation of Portuguese unions. This confederation, known as Intersindical, was formed in 1971 under the old regime as an alliance of opposition unions in the democratic struggle. The dominant influence in the CGTPIntersindical was the Portuguese Communist Party (PCP) which was allied to the revolutionary governments. In the aftermath of the revolutionary phase, a movement of union activists, linked mainly to the Socialist Party (PS), organized around the Carta Aberta document, sought to challenge the organizational domination of the PCP within Intersindical and “democratize” the trade union movement. A fierce political struggle ensued, leading to the political and ideological division of the labor movement. Once the law, which had ensured trade union confederal “unity” was repealed, the door was open for trade union pluralism in Portugal. In 1978 the emergence of a second union confederation, the União Geral dos Trabalhadores (UGT), established a new logic of political and ideological union competition. See Optenhogel and Stoleroff (1985). After rising 3.5% in 1982, GDP fell 0.3% in real terms in 1983 and a further 1.7% in 1984. Domestic demand fell 7% in 1983 and 1984. Unemployment reached 11% officially and the situation of enterprises throughout the country became particularly unstable. Legally recognized bankruptcies increased 28% in 1982/83, 33% in 1983/84 (Expresso, Suplemento 15/11/86). Inflation rose from 16.6% in 1980 to 29.3% in 1984. Many firms delayed payment of their workers’ salaries or ceased paying altogether. The salários em atraso were for many an employer a strategy for dealing with manpower problems in the face of rigid layoff and firing regulations. Real wages in 1983 and 1984 fell by 6 and 11%, respectively, stagnating in 1985 (OECD 1986).

178

Unemployment and Trade Union Strength in Portugal

In 1983 the economy collapsed into a deep depression, during which government policy in the nationalized industrial sector, which was still geared toward maintaining employment, was key to controlling unemployment. The unemployment rate was also attenuated because workers in private sector firms that had gone under would not concede this ruin and would not abandon their posts of employment.8 This was the period of salários em atraso (delayed payment of wages) for which – together with child labor – Portugal then became well known. It was also the period in which precarious and nonofficial employment became noticeable (Cabral 1983; Rodrigues 1988). The economic crisis was decisive for trade union development, and it was a watershed for union strength. The economic crisis of 1983–85 was a decisive blow to the workers’ movement and especially to the class-struggle tendency of the CGTP that had dominated the unions during the democratic transition. It led to a number of involuntary modifications in the course of union development and a decrease in its overall power in Portuguese society, economy, and polity. The economic crisis and the precariousness of employment were decidedly the catalysts of change in the unions’ position, but they did not directly induce changes in union strategy. It was rather the inadequacies and failure of the CGTP’s political and revolutionary strategy and the substitution in the incentives for collective action which resulted in a paradigm shift for union action. The economic crisis coincided with the formation of the Bloco Central coalition government of the Socialist and Social-Democratic Parties led by Mario Soares. The subordination of trade union struggle to political struggle was a by-product of the revolutionary period (Optenhogel and Stoleroff 1985), and the PCP put all its resources, most particularly its capacity to mobilize workers through the CGTP, into bringing down the government. Unemployment, delayed wages, and economic crisis more generally were directly associated with the politics of the government favorable to “capitalist restoration,” and the substitution of the government was held up as the key to the resolution of the immediate problems of the working class. Exacerbated by the crisis-induced deadlock in industrial relations at the enterprise level, this intensified politicization of union

8

This author was witness in this period to situations where workers would fulfill the obligations of the working day, refusing to abandon their posts, day in and day out, although production had been halted by a lack of supplies, so as not to break their employment contract. Of course, they were not paid and the companies were technically bankrupt.

179

Alan Stoleroff

mobilization led to a further displacement of the trade unions’ locus of action from the concrete situation of economic branches and enterprises to politics and the state (Stoleroff 1988a, 1988b). A high level of industrial conflict was therefore maintained during the mandate of the Bloco Central government from 1983 to 1985 (see Table 6.3). In 1984, however, the high frequency of strikes was accompanied by a decrease in participation with a decrease in the average workdays lost per strike. This trend was repeated in 1985. The decreasing density of conflicts was a signal that the politicization of labor action had come up against resistance. Serious economic insecurity – bordering on desperation – exhausted the revolutionary impetus; workers who may have previously been susceptible to the rationale behind the linkage of economics and class politics became preoccupied with the preservation of their jobs and survival. Nevertheless, such political incentives continued to interest a substantial portion of the working class. The experience of the ship-building and repair enterprise Lisnave over the course of the two years of crisis provides a good example of the exhaus-

Table 6.3. Strikes, 1974–1994

180

Unemployment and Trade Union Strength in Portugal

tion of political mobilization in the wake of job insecurity.9 Lisnave was directly affected by the crisis in the ship-building and repair industry and faced serious difficulties. One such problem was salários em atraso. In conformity with the CGTP’s orientation, a number of strikes took place at Lisnave around demands which were only secondarily linked to the enterprise situation. Given the fact that contracts with ship-building enterprises include time stipulations for work done and penalties for work not finished on time, these strikes were alleged – by management and the CGTP’s union adversaries – to have greatly prejudiced the enterprise. The UGT unions, particularly its metalworkers union, were engaged in negotiations seeking to restore the enterprise economically with concessions of greater manpower flexibility. The strikes did not solve the immediate problems of the company’s workers; the late payment of wages continued. In 1982 employment at Lisnave was 7,140; by 1986 it had fallen to 4,511. This decline was accomplished by attrition as well as an early retirement program encouraged by management. Many workers who were favorable toward the CGTP Federation of Metalworkers left the company. Other workers were simply exhausted by the situation. In 1984 a PS-PSD slate made up of candidates identified with the UGT won a majority in the Workers Council at the Rocha shipyard on the Lisbon side of the river. The following year the majority in the Workers Council of the enterprise as a whole likewise passed to the same forces. Strike actions decreased radically from 8 work stoppages in 1985 to none in 1986 and from 24 legal strikes in 1985 to 1 in 1986. It is possible to see in this example of such a rapid change in the political and union alignment of Lisnave’s workforce the failure of a trade union strategy to sufficiently take into account the enterprise locus of union action and the material incentives for collective action (Stoleroff 1988).10 The spectacular decrease in strike action at Lisnave is symptomatic of what was happening more generally in Portugal. In 1986 the Ministry of 9

10

Lisnave’s name became almost synonymous with the Portuguese revolution of 1974–75 because of the vanguard role and militancy of its workers and the importance that the PCP placed upon it strategically. It had also been a center of working-class protest in the period immediately preceding the coup. The following is based upon interviews with unionists of various ideological currents at Lisnave in the spring of 1984; also upon Conceição Laço et al. (1987). Lange and Ross (1982) would emphasize the organizational incentives to membership that the change in economic context provoked. In their language, the economic crisis diminished the efficiency of ideological incentives. Union leadership and strategy were probably not adequate in this situation to reverse the loss of members.

181

Alan Stoleroff

Labor registered a major decrease in the number of strikes from 504 to 363. The explanation for this trend is to be found in the failure of the CGTP’s strategy. It should be noted that strike success, having plummetted in 1982, fell to its lowest point in 1985 (13.3% of strike demands partially or totally accepted). The CGTP unions’ choices, as in the case of Lisnave, did not conform sufficiently to the need of many workers to sense that the unions were doing something to save their jobs. Many workers put a new value upon union cooperation with management rationalization plans. The UGT’s collaborative approach in this context was decisive to its ability to emerge from obscurity in firms like Lisnave. With wage payments still delayed in many enterprises, with numerous enterprises – even public enterprises – threatened with major restructuring, it became increasingly difficult for union militants to ignore the effect of repeated defeats upon their capacity to mobilize.11

Economic Recovery and Restructuring In 1986 the economy began a recovery which, with the entry of Portugal into the European Community, became a short-term boom. (Former Finance Minister Braga de Macedo called Portugal an “oasis” during this period.) In spite of what may abstractly seem like favorable conditions for union growth – namely, a tight labor market – Portuguese unions experienced this recovery more as a continuation of problems that emerged in the crisis than as a new opportunity for their development. This was a period of employment growth and low unemployment. From 1988 to 1991 the average yearly growth in the labor force was 1.7%, reaching 2.4% in 1991. Employment grew at a yearly average of 2.5% in 11

In the wake of such a clear weakening of its mobilizing capacity, the CGTP-IN attempted to adapt its organizational tactics to the problems posed by the defeats within the enterprises by adjustments in its orientation. For example, it launched a campaign in the fall of 1985 to reinforce union organization in the workplaces. However, the reinforcement of the enterprise locus of organization was not effectively a renunciation of the state locus of action deriving from its class, political perspective; it was rather its reaffirmation, though now with more attention to the enterprise basis of mobilization. It was nevertheless in this context that a “pragmatic” element in the leadership of the CGTP began to emerge. The Fifth Congress of the CGTP-IN in June 1986 produced a tacit recognition that a strategy based solely upon class mobilization in the context of increasing job insecurity had become exhausted and that the PSD’s electoral support had risen and Portugal had entered into the European Economic Community. The CGTP-IN found itself in a situation, recognized at the congress, of having to pay more attention to the rules established in the games of micro- and macrobargaining.

182

Unemployment and Trade Union Strength in Portugal

the same period, reaching 3.0% in 1991. Between 1988 and 1990 salaried employment grew at an average rate of 3.6%. The years of reduced unemployment and economic growth following the recovery in 1986 and the entry into the European Community failed to reproduce conditions for union strength that had been attained before the crisis. This is not to say that real wages did not increase for a time in the period after the recovery, for real wages did increase meaningfully in 1987, 1990, and in 1992; or that the political influence of certain unions or union confederations was not significant, for the UGT, in particular, did become a major political actor in the macrolevel institution of social dialogue. But these trends took place in spite of declining unionization, decreasing mobilization capacity, declining political importance, and weakened organization. The unions were unable to recover organizationally from the beatings they had taken during the crisis of the mid-1980s. This problem can be explained in large part by the differentiation of the labor market: on the one hand, it exhibited forms of uncertainty, such as enterprise restructuring and precarious employment, that were functionally equivalent to unemployment in its creation of a psychology of insecurity and, on the other hand, the expectations that accompanied economic growth and tight labor markets in certain branches became individualized. Simultaneously with the recovery, an unfavorable antiunion political and ideological environment emerged. The PSD’s ascent to an absolute majority in the legislative elections of 198712 elicited a strong mandate for neoliberalism and the buttressing of private property and capital to the detriment of the state sector. The PSD government launched a program of “structural reforms,” which entailed the revision of labor legislation, privatization, the reinforcement of private economic groups, the abolition of the agrarian reform, and the revision of the constitution. Its program contained strong antiunion, or more particularly, anti-CGTP implications. Labor law reform and privatization posed direct challenges to the position of the entire labor movement. These trends were accompanied by very significant efforts at restructuring the Portuguese economic system, from a general liberalization with elements of deregulation to privatization of the nationalized sector and enterprise rationalization. The conditions in which employment growth took place mitigated substantially the security that workers could have perceived and thus detracted from the circumstances that could have been favorable to unions. In the 12

See Nataf (1995) for an interesting treatment of electoral and public opinion trends.

183

Alan Stoleroff

first place, employment growth ironically came at the expense of job security owing to the importance of defined term contracts (contratos a prazo). In general fixed-term contracts, requiring annual or semiannual renewal, place workers in an insecure position. In 1985, 67% of new employment contracts were temporary and made up 12% of all jobs. By 1992, 12.1% of salaried employment was of this form. Between 1992 and 1994 this form of employment decreased somewhat, although limited-term contracts went up 6.5% between the second trimester of 1994 and the second trimester of 1995. On the whole, however, so-called atypical employment (including part-time work, limited-term contracts, self-employment, and nonremunerated family labor) continued to rise as a percentage of the active labor force. In the second place, employment growth was concentrated in particular sectors while other sectors saw drastic declines in employment. There has been a marked tertiarization of employment with services growing from 43% of the employed population in 1987 to 55.6% in 1992. From 1988 to 1994 employment in industry as a whole declined from 1,502,400 to 1,393,300. In the same period male employment in industry decreased from 1,065,100 to 953,000 while female employment actually increased, from 437,300 to 446,500.13 Total employment in the services (commerce, restaurants and hotels, transport, posts and telecommunications, finance, public administration, teaching and health, etc.) increased from 1,910,200 in 1988 to 2,367,900 in 1994. Finally, significant changes emerged in the country’s “industrial order” (Lane 1989), that is, the distribution of employment according to size of workplace. These changes reinforced the predominance of small and medium enterprises over large enterprises. In Portugal there is an inherent tendency for stronger industrial relations systems and union presence to be associated with larger enterprises and vice versa. The importance of small and very small enterprises and the disproportional concentration of employment in the strata of smaller firms is a structural factor limiting the development of industrial relations in Portugal. As can be seen from Table 6.4, in 1987 – the first year of Portugal’s entry into the European Economic Community (EEC) – 94.8% of enterprises had fewer than 50 employees, and these accounted for 39.5% of the employed. In 1992, 98.1% of enterprises had fewer than 50 employees, and they accounted for 45.1% of employment. Over the span of the first five years in the EEC, 13

Female industrial employment actually peaked in 1991 at 492,700, having since fallen.

184

Unemployment and Trade Union Strength in Portugal

Table 6.4. Distribution of Enterprises and Employees According to Size of Enterprise, Continental Portugal (percentage)

the proportion of workers increased significantly in the segment of employment where the probability of developed industrial relations is weakest. Moreover, the strata in which industrial relations have the greatest chances of development, that is, the medium and large enterprises, also saw a decline in this initial five-year period. Whereas in 1987 large enterprises of 500 or more employees accounted for 27.5% of employment, in 1992 they accounted for only 21.7% of employment. This clearly significant structural blow to industrial relations in Portugal has been coincident with European integration.14 As a consequence of the recent restructuring, especially the privatization of public enterprises, the number of large and medium-sized firms (those with 500 or more employees or between 200 and 500 employees) has declined considerably in both absolute and relative terms, while the number of smaller firms has increased. The medium-term consequences of privatization should not be underestimated, since the nationalized companies were the stronghold of 14

This trend continues from year to year. According to the Labor Ministry (Quadros de Pessoal, MESS 1994), in 1993 there were 159,047 enterprises with 1,992,142 wage earners. The overwhelming majority of enterprises were small firms (77.7%) that employ less than 10 employees, whereas only 1.7% of the total were enterprises with more than 100 employees. On the other hand, the share in employment of the numerous small firms was only 20.3%, while the bigger firms (more than 100) employed more than twice as many employees (or 41.6%). While there has been continuous growth in the number of enterprises created in Portugal, between 1985 and 1993 the average size of Portuguese enterprises fell from 19 to 14 employees. This has been due to the enormous increase in the number of very small companies as well as the reduction in personnel in the large enterprises. In this context it is notable that there was a decrease in both the average size and the absolute number of large enterprises with 500 or more employees.

185

Alan Stoleroff

industrial relations in Portugal. These “fortresses” of the union movement have been largely dismantled (Stoleroff and Naumann 1994; Stoleroff 1995). As can be seen from the example of two very significant public industrial firms which were restructured for privatization (Table 6.5), the rapid decrease in employment was accompanied by a corresponding decrease in union members, although the percentage of unionized employees remained over 90%. Other trends in the long-term recomposition of the Portuguese labor force unfavorable to union growth came to the fore during the period of economic and employment growth.15 According to the census of 1991, the 1980s saw a significant decrease in the dependent wage-earning labor force in agriculture. The most rapid decrease, a rate of -6.3%, occurred in 1989. The share of agriculture in the dependent labor force declined by almost half. The decline of agricultural employment to 11.5% by 1994 corresponds to a drop of 1% per annum since Portugal joined the European Union.16 The other social trend of note was the important increase of women in the labor force. While the rise of women in the secondary sector in the 1980s amounted to only 1.1%, it was 10.2% in the tertiary sector. On the whole, women increased from 34.0% to 44.2% of the labor force from 1981 to 1992. On the one hand, in the conditions of retracting industrial employment, the decline in the agricultural labor force contributed to the reconstitution of the reserve army of labor and the withering of the rural reserve of semiproletarianization which provided industrial workers with a social cushion. On the other hand, women have been less willing subjects of unionization and economic struggle. Within the enterprises – especially the larger firms where the Portuguese industrial relations system was most developed – management was

15

16

Portugal remained overwhelmingly an agricultural country in 1960. However, industrialization throughout the 1960s was so extensive that by 1970 the industrial labor force had slightly overtaken that of the primary sector. This trend continued and accelerated throughout the 1970s and 1980s, with an increasing preponderance of the service sector in relation to both primary and secondary sectors. By 1992 only 11.6% of the labor force remained in the primary sector. By 1992, the service sector had increased to occupy over a half of the entire labor force, while industry had decreased in proportion over the 1980s. These trends can also be ascertained through the sectoral changes in the composition of the wage-earning labor force. This trend is very likely to continue. From the second quarter of 1995 to the second quarter of 1995, agricultural employment declined by 3.5%, whereas industrial employment fell by 1.0%. Over the decade, manufacturing remained stable with about a third of wage earners, while services, including civil service, increased from approximately 45% to 50%.

186

Unemployment and Trade Union Strength in Portugal

engaged in a restructuring process which entailed the quest for numerical and functional flexibility. Major nationalized enterprises were being primed for sale, and the rush was on to minimize the unprofitable baggage accumulated over a decade and a half of public management through the breakup of conglomerates, downsizing through subcontracting of ancillary services, labor force attrition. The following data regarding labor relations at the company level, from a survey carried out in 1994 (Stoleroff 1995),17 are a mere indication of the restructuring which took place in the late 1980s and early 1990s. Tables 6.6, 6.7, and 6.8 provide some idea of the extent of change in this regard in medium and large Portuguese companies. What is most impressive in the reported data is the extent (not necessarily the quality) of the rationalization effort in Portuguese enterprises in the late 1980s to early 1990s. Hiring practices demonstrate the diffusion of tactics seeking flexibility in the workforce, while personnel policy reflects a generalized attempt to eliminate excessive employment and to qualify and hold on to core workers. Also apparent are significant efforts toward technological renovation, organizational restructuring, and skill development in medium and large Portuguese firms. There has been extensive recourse to employment reduction in spite of the continuation of a certain degree of labor rigidity. The relevant issue here is the attempt to achieve flexibility in the employment of human resources in the workplace. Of the managers interviewed, 38.2% claimed their companies achieved overall flexibility in the use of labor. But it may be more important to distinguish numerical flexibility from functional flexibility, as the former refers to the ability of companies to adjust their labor forces to conjunctural needs either through layoffs, subcontracting, temporary hiring, or the like. Over half the surveyed firms (52.9%) had already decreased their personnel. Owing to Portuguese labor legislation, employment reduction may be a costly affair. As a result, other methods have been used to attain flexibility. The policy of inducing early retirement and other forms of voluntary departure with indemnification was actively practiced in 31.5% of the companies (55.2% of the large companies). The current use of temporary labor contracts was registered in 51.8% of the companies. Temporary labor was used in 24.1% 17

Survey applied to a representative sample of medium and large firms in Portugal conducted with the sponsorship of the Volkswagen Foundation. The responses reported in these tables were obtained from the chief personnel manager in each of the firms.

187

188 Table 6.5. Employment and Unionization in Two Public Sector Industrial Firms, 1986–1992

Table 6.6. Development of Selected Hiring Practices in Medium and Large Companies over the Previous Five Years (percentage)

189 Table 6.7. Development of Selected Personnel Policies in Medium and Large Companies over the Previous Five Years (percentage)

Table 6.8. Elements of Restructuring Achieved in Medium and Large Companies over the Previous Five Years (percentage)

Alan Stoleroff

of the companies. Functional flexibility refers specifically to measures that augment the tasks delegated to individual workers. A large proportion (69.2%) of the companies claim to have developed polyvalence over the five years preceding the survey (87.5% of the large companies). Meanwhile, 31.9% (46.1% of the large firms) affirmed they had succeeded in integration or merging occupational categories in contractual job classifications. On another plane, 40.4% (51.5% of the large and 38.8% of the medium companies) achieved flexibility in labor relations. Among such companies, the most frequent methods mentioned were individualization of the employment relation, the appearance of forms of labor relations outside the terms of collective agreements, and the renunciation of contractual clauses, in descending order of frequency. Only a marginal proportion of companies achieved flexibility on the basis of greater dialogue. Only 11.2% (11.7% of the medium and 7.7% of the large companies) claimed to have “deregulated” of labor relations. Presented in this manner, these percentages do not appear extensive. Nevertheless, deregulation represents a maximal point for the consequences of restructuring. The fact that only a limited percentage of firms has reached this maximum does not signify that there is not a tendency in this direction. The importance of this qualification is indicated by the percentage of managers – 29.3% in the total – who affirm that, although not yet attained, deregulation of labor relations would be desirable in the context of company restructuring. Only half of the managers affirmed that deregulation had not been achieved, nor would be desirable in their company.

The Return of Recession In 1992 employment suffered a significant decline of 1%, and in 1993 it experienced negative growth of -2.0%. The decline continued in 1994 at -0.1% and in 1995 at -0.6%. The annual declines in salaried employment have been greater in this period. Unemployment (restricted sense)18 has increased steadily since the third quarter of 1992, reaching 5.5% in 1993, 6.8% in 1994, 7.2% in 1995, and 7.3% in 1996. Certainly, this unem18

Restricted sense means only those unemployed who have actively sought employment within the thirty days preceding the employment survey.

190

Unemployment and Trade Union Strength in Portugal

ployment rate still remains below the OECD and EU average, but in 1993 Portugal had one of the highest rates of unemployment increase. While the unemployment rate for the European 12 increased from 10.8% to 10.9% between November 1993 and January 1994, in Portugal the rate increased from 5.2% to 5.7%. Between January and March 1994 the Instituto Nacional de Estatística counted 27,000 new unemployed, which resulted in an average loss of 300 jobs per day (Diário de Notícias, March 24, 1994). Finally, in November 1994, 411,942 job seekers were registered in the Centros de Emprego, an all-time high (Expresso, December 24, 1994). The April 1995 figure for job seekers registered with the Centros de Emprego reached 430,425. The future remains very uncertain, although in the past year unemployment has come under some control. As can be seen in Table 6.9 on unemployment until 1992, when the most recent recession took hold, female unemployment was consistently at double the level of male unemployment.19 After 1992 there was a substantial increase in the unemployment rate of men, and hence less disparity between the male and female rates as both increased. The situation of youth unemployment has also deteriorated greatly since 1992. This underscores the importance that recent governments have given to “active labor market policies” and the particular focus on youth. Following 1992, male unemployment increased proportionally, reflecting the impact of industrial restructuring.The CGTP labor confederation has frequently asserted that in Portugal unemployment is taking on new forms above all because of early exit from the labor market through early retirement and negotiated resignations from restructured enterprises. The generation of workers over 55 years of age is most affected. This recent character of unemployment (causing strata with significant remaining productive potential to exit from the labor market) together with the problems of discouragement in job search and 19

Following 1974 there was a steady increase in the active labor force and by the second semester of 1977 there were more people among the unemployed who were looking for their first job than those looking for a new job. According to Rodrigues (1988: 109), it is primarily in the succeeding years that unemployment became strongly influenced by the influx of women and others seeking their first job into the active labor force. The participation of women in the active labor force crept up from 33.2% in 1974 to 39.1% in 1983. This tendency continued until 1982–83, when it was reversed.

191

192

Table 6.9. Unemployment Rate, Portugal, 1988–1996 (percentage)

Unemployment and Trade Union Strength in Portugal

underemployment, attenuate the significance of Portugal’s still low unemployment rate.20 The important development in the labor market in the past decade is the emergence of “structural unemployment.” This concept, though applied universally, in reality signifies specific forms of disarticulation within particular national and regional labor markets. In Portugal the structural unemployment caused by industrial restructuring and the mismatch of supply and demand of existing skills has been especially highlighted (Rodrigues 1991).21 As has already been noted, however, there are other longer-term contributions to structural unemployment: demographic change (general growth and youth entry in the active labor force and its feminization), agricultural decline, sectoral shifts, and the trends in enterprise development brought on by restructuring and privatization. The pace at which these trends are affecting employment and unemployment in Portugal and their form are very much influenced by the particularity of the strains of European integration. In the legislative elections of 1995, the return of recession contributed to the defeat of the PSD and the formation of a government of the Socialist Party. The attenuation of youth unemployment, long-term unemployment, and structural unemployment are high priorities for this government. It is still too early to discern the effect of these efforts. Nevertheless, in its attempt to increase the competitiveness of Portuguese economy and enterprises, this same government has continued elements of the neoliberal approach pioneered by the previous government. It has continued with the privatization process, which has now alienated the 20

21

From 1992 a significant portion of discouraged workers were no longer counted as part of the active labor force. From 1991 to 1992 there was an important decline in the active labor force, which was only slightly reversed in 1994. This decline was generalized across age groups but was particularly sharp among males of the 55–64 age group. The words of the first labor minister of the socialist government which took office in 1995, a prominent labor economist herself, are worth noting in this regard. In a recent interview she affirmed that “Portugal has very serious structural problems in employment.” She made the following remarks about the efforts of her government: “We are succeeding in reducing the unemployment rate, but we continue to have very serious employment problems. About 74% of the employed population has an elementary education or less than that. In Europe this ratio is 35 percent. Almost 40 points less than in Portugal. The Portuguese who haven’t even reached the ninth grade represent 62% of the employed, and only 12% of the employed population have an intermediate or higher education. But another problem exists, which is an authentic paradox: the enterprises don’t have qualified staff and there are qualified people in the labor market who can not find employment” (Público Economia, September 1, 1997).

193

Alan Stoleroff

public sector almost entirely. It has also engaged in labor reforms, some of which, like the legislation to promote flexibility and polyvalence, are considered by the unions to put employment at risk.

Trends in Union Growth and Strength A relative decline in union strength in Portugal throughout the second half of the 1980s and the 1990s is incontestable. It is primarily manifested through the decline in union membership and union density and decreasing frequency of strikes and workers’ participation in them. Other indicators of union strength have either been less demonstrative of the decline or can only be ascertained through a more qualitative analysis. Strength begins in numbers, and there is considerable evidence of a significant and consistent decline in union membership since the crisis of the mid-1980s. At its VII Congress (1993), the CGTP-IN presented a sober analysis of union evolution highlighting various tendencies: an increase in unionization in public administration, a decline in unionization in sectors affected by the economic crisis, a decline in unionization in commerce and other expanding services. The overall tendency was a decline in unionization, which had serious repercussions on union stability. CGTP unions suffered a consistent decrease in dues collection between 1987 and 1990. Figures claimed by the CGTP at its congresses since 1983 indicate a loss of almost 43%, that is, 672,000 members from the unions in its orbit. Unfortunately, the UGT confederation has not been as forthcoming with regard to its fortunes in terms of membership, making it difficult to obtain an overall idea of union membership trends. Membership nevertheless gives only a partial idea of what is happening, and it is necessary to look at union density. There is a problem obtaining confirmed data on density in Portugal. Cerdeira e Padilha (1990) and Cerdeira (1997), basing their attempt to establish union density on the aggregation of union membership with unions’ membership claims,22 claim a rise in overall unionization from 52.4% in the period 1974–78 to 58.8% in the period 1979–84 but find a consistent and substantial decline 22

A controversy surrounds the methodology Cerdeira and Padilha (1990) employed for their study. It is likely that their figures for 1974–78 and 1979–84 are overestimates. See Stoleroff and Naumann (1994, 1998). I use Cerdeira’s figures here in spite of these problems because unfortunately it is still the only publication with figures for union density over the entire period since the advent of free industrial relations. It does not signify an endorsement of her methodology.

194

Unemployment and Trade Union Strength in Portugal

since then. According to Cerdeira, total unionization declined to an average of 44% in the period 1985–90 and to an average of 36% in the period 1991–95. She speculates that at present it may be less than 30%. The International Labour Office recently reported a rate of unionization of 25.6% for Portugal in 1995, a drop of 50.2% in relation to the figure of 51.4% density for 1986 (ILO 1997). Although the apparent decline in unionization coincides with the economic crisis of 1983–85, the trend confirms the lack of a strict causal connection between unemployment and unionization. Unionization continued to decline in the period of decreasing unemployment following 1985. The five-year periods that Cerdeira uses do not correspond with the economic cycles; thus it is impossible to determine the short-term response of unionization to unemployment. Even without a capacity to fine-tune our analysis quantitatively, it is clear that a simple linear relationship does not exist between these two variables. The labor market situation is a context of employment relations and does not affect unionization as unilinearly as it affects union mobilization capacity and bargaining power. Job insecurity might even stimulate unionization – if workers are confident in their right to unionize without repercussions. As I have shown, other factors intervened to maintain the decline in unionization in spite of declining pressure deriving from unemployment. In the evolution of the labor force and employment in Portuguese enterprises, two main factors affected unionization: namely, the overall decrease in wage labor employed in industry, and the decrease in employment in large enterprises. These trends reduced the number of union members, and with the growth of employment since 1985 in the “nonunion” sector, the overall rate of unionization has suffered a mathematical decline. This does not explain everything, however, because industry itself has suffered a major decline in unionization. To this I would add the effects of restructuring and feminization. Privatization took large firms out of the logic of subordination to government employment policy, subjecting them directly to competition, and was frequently associated with organizational restructuring and the adaptation of the model of the “flexible firm.” For example, the subcontracting of noncore activities and the spinoff of these activities into smaller enterprises resulted in deunionization. This has taken place simultaneously with labor force rationalization. The growth of the industrial labor force on the basis of its feminization, especially in the interior of the country, has also not been propitious for unionization. The association of a tyrannical factory regime 195

Alan Stoleroff

based upon fear and pressure has discouraged unionization. Certainly, the context of unemployment facilitates employer repression. Employer rejection of unions and unionization has had a strong impact in Portugal: since unionization is fully subject to voluntary association, the closed shop cannot exist and therefore union membership cannot be an attribute of employment. Each worker must individually take the first step to collective action, which is joining a union.23 The mere constitutional protection for union membership is an extremely weak device and antiunion employers are easily capable of threatening job loss for joining a union. The decline in union membership and density is mirrored by a concomitant decline in the unions’ mobilization capacity. As can be seen from Table 6.3, strike frequency has shown a continuous – if not irregular – tendency to decline from its peak in 1981. The same holds true for participation in strikes, although the trend is not entirely coincident. Although strike frequency is an ambiguous indicator of union strength,24 such ambiguity is not quite pertinent to the analysis of trends in Portugal. In any case, the trend regarding participation in strikes backs up this hypothesis. Strike frequency has been irregular over the years and not always in a linear relationship with unemployment.25 The surge of strikes from 1974 to the end of 1976 was reflective of the working-class offensive in the revolutionary period. It took place in spite of rising unemployment; indeed, unemployment was the product of the economic turmoil in the context of political uncertainty. The frequency of strikes continued to rise in 1979, owing in part to the political tension between the workers movement and the right-wing Aliança Democrática government. Strike frequency reached its peak in 1981, although participation in strikes began to decline. Increasing political competition in the use of the strike began to take hold and the adherence of workers to each strike tended to drop. The high frequency and politicization of strikes attained in 1981–82 continued from 1983 to 1985 23

24

25

The recognition of the importance of this factor motivated the CGTP to invest in a sociological study of workers’ perceptions of unions (see Ribeiro, Leitão, and Granjo 1994). The commonly held notion that increasing union strength is associated with an increasing frequency of strikes may be belied by a countertendency; as unions become more formidable adversaries and recognized partners, there would tend to be a decrease in strikes since neither unions nor employers would be interested in costly tests of strength in routinized collective bargaining. It should be noted that accurate accounting for strikes was not achieved by the Labor Ministry until 1978.

196

Unemployment and Trade Union Strength in Portugal

during the Bloco Central government. However, the increase in conflicts continued to be accompanied by decreasing density. From 1986 on, strike frequency decreased. Participation decreased as well, although with some irregularity at the end of the 1980s. To a certain extent, the fluctuations in the strike statistics of the late 1980s indicate that strike activity was more than ever directly related to economic opportunity – or the lack of it. Strike frequency and density declined with the downward economic curve. They continued to decline and bottom out with the initial changing of slope of the curve, as would be expected in response to the lag between economic upturn and collective perception of the change. Indeed, strike frequency was at its lowest in 1988, the year of the general strike. It then began to increase, and it did so in specific ways coinciding with the opportunities and pressures of the labor market.26 Demands became nominally and really focused upon economic and labor market issues.27 The decline in strikes is also coincident with the institutionalization of tripartite bargaining in the Standing Council of Social Concertation, and from the statistics it appears that conflict was channeled into social dialogue, in other words, that Portuguese neocorporatism has been successful. I think this would be a spurious conclusion. The decline in strikes is indeed an indicator of the decrease in union mobilizing capacity that resulted from the crisis. The signs of moderation in strike activity throughout the rest of the 1980s was a result of 26

27

It is revealing to compare the ratio of strikes in the transport, warehouse, and communications sector with the manufacturing sector. In 1987 the ratio was 18 to 136; in 1988 it went up to a staggering 43 to 94; in 1989 it went back down to 40 to 202; in 1990 it was 33 to 175. It is necessary to look more closely still at the statistics, however. It seems that the strike frequency in the public transport sector fluctuates independently of the manufacturing sector. Strike frequency in the manufacturing sector, more subject to market pressures, may be more indicative of this tendency than the aggregate frequencies. The official cause of strikes in Portugal since the end of the revolutionary period has overwhelmingly been wage demands. The economistic appearance of the predominance of wage demands until the late 1980s, however, concealed their function as an expression of class struggle through the channels of industrial struggle. The CGTP ignored the constraints placed upon collective bargaining by the weakness of the dependent economy. This translated into a rejection of realism in the formation and presentation of wagebargaining demands. The insistence upon demands that could not be met was a means of continued mobilization of workers through strikes. Although greater pragmatism has since come to orient the practice of the CGTP, this routine continues to influence collective bargaining. The CGTP continues to consistently demand wage increases several percentage points higher than the UGT. This is motivated by union competition and the CGTP’s militant objective of seeking gains in real wages.

197

Alan Stoleroff

the incapacity of union leadership to articulate the postcrisis expectations of workers in a situation of growing employment. The trends just discussed have not been strongly reversed with the return of a recessionary climate in the early 1990s. Strike frequency, for example, continued to rise slightly, but strike participation and intensity declined. It remains notable that the highest point of strikes in the 1990s was well below the plateau reached a decade earlier. The extent of coverage of collective bargaining, reflective of the unions’ institutionalized role in economic regulation through collective bargaining, mitigates the previous indicators of the evolution of union strength. It measures the number of workers to whom collective conventions apply whether or not they are union members. By 1984 coverage reached 60% of the wage labor force and has remained around that level ever since.28 There was a short-lived decline below 60% in 1989. Coverage does not seem to fluctuate very much; its changes are fairly unnoticeable. Its minor variations do not seem to accompany unemployment. If anything, it seems to be inversely related to unemployment; that is, it seems to go down those slight bits as unemployment decreases and to go up that slight bit as unemployment goes up. As an indicator of union strength, coverage is not immediately susceptible to variations in the rate of unemployment. On the other hand, the trends in wages, as indicative of union bargaining power, are relatively volatile, closely but not exclusively tied to the economic conjuncture and thereby unemployment. They are nevertheless somewhat dependent as well on a more complex and qualitative evaluation of union strength. The initial significant rise in real wages following the revolution was clearly due to the increment in union and working-class power and its political and social offensive. In the period that followed, these gains were quickly eroded as unions were unable to keep up with inflation. From 1988 to 1990 there was an annual increase in the variation of nominal wage gains. This was a phase in which, unionization and strike frequency were decreasing. Wages were thereby increasing, although union strength was being eroded. From 1991 on, the yearly variation in nominal wage gains has been decreasing. This is due to various factors, but primarily to the government’s wage and inflation-containment poli28

The de facto bargaining that goes on annually between the government and the civil service unions over wages culminates in a unilateral political decision of the government and therefore is not considered a formal collective bargaining agreement. As a result, the statistics on coverage are actually distorted and would be much higher if the civil service workers were considered.

198

Unemployment and Trade Union Strength in Portugal

cies. With the Maastricht Treaty and the commitment to be among the initial countries in the European Monetary Union, there has been increasing pressure on this front. In 1992 unemployment began its most recent rise and concomitantly nominal wage gains, and inflation dropped sharply. By 1994 there was no rise in real wages. The trends are starker if one looks at the enterprise sector, eliminating the effects of employees in public administration. Where workers and employees were more directly subject to economic pressures – without the interference of political decision in wage determination – the rises in nominal wages were consistently lower and real wages were substantially lower.

Conclusion This chapter has demonstrated that unemployment has had a significant if irregular impact upon union strength in Portugal. The rapid and drastic increase in unemployment during the economic crisis of the 1980s all but eradicated the effectiveness of ideological incentives for mobilization on the basis of militant class unionism and forced important changes upon Portuguese unionism. The unemployment of the 1990s pushed job security and employment to the top of the trade union agenda.29 Nevertheless, union decline in Portugal is the result of complex processes of social and political change as well as economic conjuncture. Craypo (1981: 111–12) argues that three conditions are necessary for increasing union bargaining strength: organization of the relevant workforce, elimination of competitive unionism, and sufficient centralization in the bargaining system. The first two are very poorly fulfilled in Portuguese industrial relations, while the third, on its own, is insufficient in the absence of the first two. Given the characteristics of the Portuguese trade unions – the very segmented unionization pattern (together with occupational fragmentation of union organizations) and political competition – the prognosis for trade union strength in the current confrontation with unemployment is not optimistic. The more exclusive and strategic professions have been able to maintain their organization and demand capacity action in the present context – airline pilots are the most noteworthy but doctors in public 29

At its Fifth Congress the CGTP’s discussion of unemployment treated it as an inevitable product of capitalism and unsolvable under capitalism. This line gave way to the quest for full employment from the time of its Sixth Congress.

199

Alan Stoleroff

hospitals, tax department workers, and other such categories of employees are also able to make demands. Among the less elite occupations, the unions of public transport workers also maintain the capacity for strike mobilization with reference to particularistic demands. Unions in private industry are relatively quiescent. Much other trade union action is strictly defensive, being reactive to plant closings and layoffs. It is possible for unions to develop an effective strategy to take advantage of this role to promote unionization.30 As industrial employment continues to shrink, so will the contribution of industrial workers to union ranks. On the other hand, the latent quest for collective representation in public administration and services, where employment is less insecure, remains unfulfilled. (The case of the police is notable.) All the same, it is possible to ask whether Craypo’s conditions remain sufficient for increasing union strength in the present circumstances of high unemployment and precarious employment. It would appear difficult to increase union strength on the basis of a reinforcement of unions’ market capacities alone. It will be interesting to see whether the pressures of unemployment in the context of increasing competition and economic uncertainty will lead the neocorporatist bias emergent in the Portuguese state to develop further. This tendency may lead to the generalization of the public status of the union confederations and branchwide unions, to the exclusion of their role in the workplace. Such a representative function is eventually compatible with low union density. On the other hand, it will be interesting to see whether the unions are able to act politically in ways to mobilize public policy toward new forms of regulation that may alleviate the tremendous burden of insecurity. References Almeida, J. F., A. F. Costa, and F. L. Machado. 1994. Recomposição Socioprofissional e Novos Protagonismos. In A. Reis, ed., Portugal 20 Anos de Democracia. Lisbon: Círculo de Leitores. Armingeon, K., G. Couffignal, A. Cox, D. Labbé, R. Mouriaux, and M. Revelli. 1981. Les syndicats européens et la crise. Grenoble: Presses Universitaires de Grenoble. Bance, P. 1988. Privatisations du patrimoine public et politique économique. Revue d’Économie Industrielle 46. Baumard, M., and M. Blanchot. 1994. Crise du syndicalisme. Paris: Hatier. 30

In this context, the liberalization of the Socialist Party, formerly an important though partial ally of labor, is a strategic disadvantage.

200

Unemployment and Trade Union Strength in Portugal Boyer, R. 1986. La flexibilité du travail en Europe. Paris: Editions La Decouverte. Cabral, M. Villaverde. 1983. A economia subterrânea vem ao de cima: estratégias rurais perante a industrialização e a urbanização. Análise Social 76. Caire, Guy. 1990. Introduction: syndicalisme en crise? (Introduction: unionism in crisis). In Geneviève Bibes and René Mouriaux, eds., Les syndicats européens à l’épreuve. Paris: Presses de la Iondation Nationale des Sciences Politiques. Cerdeira, C., and E. Padilha. 1990. A sindicalização e alguns comportamentos sindicais, Lisbon: MESS (Série C-Trabalho). Cerdeira, M. C. 1997. A Sindicalização Portuguesa de 1974 a 1995. Sociedade e Trabalho 1 (October). Conceição Laço, M., Marío Guerreiro, Paula Mendes, and Paula Santos. 1987. Uma tentativa para a saída da crise: Mudança da política sindical e da gestão e utilização da mão-de-obra na Lisnave. Seminar thesis. ISCTE. Craypo, C. 1981. The Decline of Union Bargaining Power. In Michel Carter and William Leahy, eds., New Directions in Labor Economics and Industrial Relations. South Bend, Ind.: University of Notre Dame Press. Davis, H. B. 1941, 1972. The Theory of Union Growth. Reprinted in W. E. J. McCarthy, ed. Trade Unions. Hammondsworth: Penguin. Dunlop, J. 1987. The Development of Labor Organization: A Theoretical Framework. In Richard A. Lester and Joseph Shister, eds., Insights into Labor Issues. Reprinted in S. Larson and B. Nissen, eds., Theories of the Labor Movement. Detroit, Mich.: Wayne State University Press. Hyman R., and A. Ferner. 1994. New Frontiers in European Industrial Relations. Oxford: Blackwell. International Labour Office (ILO). 1997. World Labour Report 1997–1998: Industrial Relations: Democracy and Social Stability. Geneva: ILO. Kochan, T., R. B. McKersie, and P. Capelli. 1984. Strategic Choice and Industrial Relations Theory. Industrial Relations 23, 1 (Winter). Kochan, T., H. Katz, and R. McKersie. 1986. The Transformation of American Industrial Relations. New York: Basic Books. Kurth, J. 1979. Industrial Change and Political Change: A European Perspective. In D. Collier, ed., The New Authoritarianism in Latin America. Princeton, N.J.: Princeton University Press. Labbé, Dominique, and Maurice Croisat. 1992. La fin des syndicats? Paris: L’Harmattan. Lane, C. 1991. Industrial Reorganization in Europe: Patterns of Convergence and Divergence in Germany, France and Britain. Work, Employment & Society 5, 4 (December). Lange, Peter, and George Ross. 1982. French and Italian Union Development in Comparative Perspective. In P. Lange, G. Ross, and M. Vannicelli, eds., Unions, Change and Crisis: French and Italian Union Strategy and the Political Economy, 1945–1980. London: George Allen & Unwin. Martin, Andrew, and George Ross. 1984. European Trade Unions and the Economic Crisis: Perceptions and Strategies. In Jack Haywood, ed., Trade Unions and Politics in Western Europe. London: Frank Cass. 201

Alan Stoleroff Mil-Homens, António. 1984. Evolução da estrutura do emprego em Portugal após 1974. Economia e Socialismo, Nova Série 63, October-December. Mouriaux, René. 1986. Les syndicats face à la crise. Paris: La Découverte. Nataf, D. 1995. Democratization and Social Settlements: The Politics of Change in Contemporary Portugal. Albany: State University of New York Press. Optenhogel, Uwe, and Alan Stoleroff. 1985. The Logics of Politically Competing Trade Union Confederations in Portugal: 1974–1984. In Eduardo Sousa Ferreira and Walter Opello, Jr., eds., Conflict and Change in Portugal 1974–1984. Lisbon: Teorema. Ribeiro, J., N. Leitão, and P. Granjo. 1994. Visões do sindicalismo. Lisbon: Edições Cosmos. Rodrigues, M. J. 1988. O sistema de emprego em Portugal – Crise e mutações. Lisbon: Dom Quixote. ———. 1991. Competividade e Recursos Humanos. Lisbon: Dom Quixote. Silva, Manuela. 1985. A repartição do rendimento em Portugal no pós Abril 74. Revista Crítica de Ciências Sociais 15/16/17 (May): 269–79. Stoleroff, Alan. 1988a. Estratégia Sindical e Comportamento Organizacional: O Caso da CGTP-IN. Revista Portuguesa de Psicologia 3/4 (September). ———. 1988b. Relações Industriais e Sindicalismo em Portugal. Sociologia: Problemas e Práticas 4 (May). ———. 1990. O Padrão Emergente das Relações Industriais em Portugal: Uma Abordagem Sistémica. Organizações e Trabalho 2 (March). ———. 1992. Corporatism and Class Struggle: The Portuguese Labour Movement and the Cavaco Silva Governments. West European Politics 15, 4 (October). ———. 1995. Elementos do Padrão Emergente de Relações Industriais no Contexto da Restruturação: Alguns resultados de um inquérito nacional às médias e grandes empresas. Organizações e Trabalho 13. Stoleroff, A., and S. F. Casaca. 1996. Produção magra e relações laborais: tendências verificadas num “greenfield” em Portugal. Organizações e Trabalho 15. Stoleroff, A., and R. Naumann. 1994. A Sindicalização em Portugal: A Sua Medida e a Sua Distribuição. Sociologia: Problemas e Práticas 14. ———. 1995. Unions and the Restructuring of the Public Sector in Portugal. In P. Leisink, J. V. Leemput, and J. Vilrokx, eds., The Challenges to Trade Unions in Europe: Innovation or Adaptation. Cheltenham: Edward Elgar. ———. 1998. Contabilidades: Alguns commentários necessários àcerca da publicação “A Evolução da Sindicalização Portuguesa de 1974 a 1995” de Maria da Conçeição Cerdeira. Sociologia: Problemas e Práticas 27. Visser, J. 1990. In Search of Inclusive Unionism, Bulletin of Comparative Labour Relations, 18. Deventer: Kluwer Law and Taxation Publishers. ———. 1994. European Trade Unions: The Transition Years. In R. Hyman and A. Ferner, eds., New Frontiers in European Industrial Relations. Oxford: Blackwell. ———. 1995. Trade Unions from a Comparative Perspective. In J. van Ruysseveldt, R. Huiskamp, and J. van Hoof, eds., Comparative Industrial and Employment Relations. London: Sage.

202

7 Trade Unions, Unemployment, and Working-Class Fragmentation in Spain Javier G. Polavieja and Andrew Richards

The growth and transformation of Spain’s economy since the death of Franco have been remarkable.1 The long period of socialist government from 1982 to 1996 was notable for its extensive program of economic restructuring and deregulation, and Spain’s entry into the European Community (EC) in 1986. During its first term of office (1982–86), the Partido Socialista Obrero Español (PSOE) implemented measures of economic austerity, restructuring, and adjustment, including an immediate devaluation of the currency, the adoption of a restrictive monetary policy, a gradual reduction of the budget deficit, liberalization of trade, deregulation of the financial markets, and a move to a more flexible labor market. Wage increases were limited by means of agreements with the employers and the unions in 1983 and 1984, while the 1984 Law on Reconversion and Industrialization restructured the finances of industrial sectors such as iron and steel, shipbuilding, and textiles. The government also liberalized housing rents, retail shopping hours, and laws governing foreign investment and launched a program of privatization of state-owned companies. Following its reelection in 1986, the PSOE further liberalized and deregulated the economy to increase its levels of productivity and international competitiveness. Given the need to comply with the EC’s requirements prior to the introduction of the European Single Market, the economy was 1

We are grateful to Bernhard Ebbinghaus and Franz Traxler for their comments on an earlier draft of this chapter. For their comments on the current version, we thank Javier Astudillo, Emma Cerviño, Modesto Escobar, Remo Fernández, Marta Fraile, Eric Hershberg, Manuel Jiménez, Pablo Marí, Covadonga Meseguer, Ignacio Molina, Javier Moreno, Guillermo O’Donnell, Luis Ortiz, Antonia Ruiz-Jiménez, and Ignacio Sánchez-Cuenca. We thank Jacqueline de la Fuente for her editorial assistance and Belén Sánchez-Rubio, Jesús Cuellar, and Ignacio Torreblanca for assistance with translation.

203

Javier G. Polavieja and Andrew Richards

further opened up through the liberalization of foreign trade and banking and reform of the stock exchange, capital markets, and the national banking sector. Further measures of privatization involved the total or partial selling off of state-owned companies. Factory closures, deregulation, a reduction of state subsidies, and the repeal of protective legislation combined to reduce (though by no means eliminate) the role of the state and at the same time to encourage the presence of foreign companies and investors (Richards 1999: 167–68). Such measures undoubtedly paid economic dividends. During the second half of the 1980s, Spain enjoyed a spectacular rate of economic expansion: between 1986 and 1990 GDP grew almost 4.5 percent a year, easily exceeding the average for the EC. Spain’s GDP per capita rose from 73 percent of the EC average in 1986 to 77 percent in 1994. By 1992, Spain’s economy was the world’s seventh largest (Richards 1999: 168). Despite these economic achievements, however, Spain was unable to bring down its notoriously high rate of unemployment. The PSOE did not fulfill its pledge to create 800,000 jobs during its first term of office. Instead, with the loss of 500,000 jobs, largely as a result of economic and industrial restructuring, the unemployment rate increased from 16 percent in 1982 to 22 percent in 1985. Even after the economic boom between 1986 and 1990, Spain still suffered the highest jobless rate in Europe, at nearly double the EC average. To a certain extent, the painful effects of economic restructuring and of unemployment itself were offset by the massive inflow of structural funding and investment from the EC and by significant expansion of welfare provision during the 1980s and 1990s. Yet persistent high unemployment has divided Spanish society. While those who held on to their jobs enjoyed rising incomes and standards of living, the unemployed and the temporarily employed fared considerably worse. In 1992, fewer than half the unemployed received any unemployment benefits. Unemployment was especially concentrated among the young; in 1991, almost a third of those under the age of twenty-five were without work (Richards 1999: 168–70, 178). Above all, high unemployment, together with the effects of economic change and labor market reform, has divided the Spanish workforce and provoked a crisis for the trade union movement. Indeed, the unions’ initial support for wage restraint and industrial restructuring soon waned in the face of the government’s inability to reduce unemployment. Claiming that it was the workers who were bearing the brunt of painful economic change, the president of the Unión General de Trabajadores (UGT) accused the 204

Unions and Working-Class Fragmentation in Spain

PSOE as early as 1985 of “bringing to our country nothing other than greater unemployment, greater inequality and greater poverty” (Richards 1999: 169). Since then, the unions have been consistently critical of the inadequacy of measures adopted by the governments of the PSOE (and since 1996, that of the Partido Popular) to reduce unemployment and promote greater job security. However, the contemporary crisis of Spanish trade unionism is not confined to the political marginalization that it has suffered but extends to its own relationship with an increasingly fragmented national workforce. In this chapter, we explore the nature of this crisis. The 1980s and 1990s proved to be an extraordinarily difficult period for the unions in Spain. A combination of changes – economic, industrial, technological, and (in some cases) political – undermined the traditional foundations of trade unionism and, in doing so, restructured its relationship with the national workforce as a whole. In this sense, the “crisis of organized labor” in Spain is broadly typical of that in other advanced capitalist countries. Nonetheless, certain developments within the Spanish labor market during the past twenty years have presented the unions with particularly grave challenges. Persistently high levels of unemployment and a massive growth (since the mid-1980s) of temporary work have created a sharply segmented, or dualized, labor market between insiders (those workers in relatively stable and secure employment), on the one hand, and outsiders (those in a fundamentally insecure and precarious working situation), on the other. This growing division within the national workforce appears to have threatened both the ability and willingness of Spanish trade unions to represent the interests of their historic constituency – the “working class” – as a whole. Here, we assess the degree to which there is a distance opening up between the unions and elements of what is an increasingly fragmented workforce. Specifically, are there significant differences between the way in which insiders and the growing army of outsiders relate to the unions? The chapter is organized as follows. First we present a brief general review of the international trends affecting the fortunes of contemporary trade unionism and relate these trends to continuing debates on the fragmentation of the working class in advanced capitalist societies. Next we examine the nature of the insider-outsider division in the Spanish context. In particular, we analyze the profile and structure of both unemployment and temporary work as a means of determining both the class character and, critically, the interrelated nature of both phenomena. Then we explain 205

Javier G. Polavieja and Andrew Richards

the logic of the unions’ response to high levels of unemployment and temporary work, and we assess how the insider-outsider division has affected the way in which workers relate to the unions. Finally, we highlight issues that merit further research.

Trade Unionism and Working-Class Fragmentation The Undermining of Contemporary Trade Unionism Trade unions throughout the advanced capitalist world have faced multiple challenges during the 1980s and 1990s. Economic, political, industrial, and technological changes have combined, in different ways and to different degrees, to undermine the foundations which had supported trade unions for much of the postwar period. After three decades of growth and consolidation, the 1980s and 1990s were a period of decline and retreat for most national trade union movements. The reasons for the seemingly international retreat of organized labor are complex and numerous and are the subject of much debate. In an adverse economic and political environment, the unionized sections of national workforces have not only been depleted but recomposed and fragmented. Structural economic change, generally higher rates of unemployment, pressures from employers for greater productivity and flexibility, and in some cases (notably Britain and the United States) politically hostile legislation have not only led to membership losses, but served to divide the ranks of what remains of the unionized workforce. Thus the contemporary period has been characterized by declining union membership, declining rates of unionization, declining political and industrial power, and growing divisions – between and within rival national confederations, between national and local unions, between blue-collar and white-collar labor, between unions in declining economic sectors and those attempting to organize in growing sectors, and, indeed, within individual (general or industry-specific) unions. This crisis of solidarity has been accompanied by a crisis of identity as the male, blue-collar, industrial worker has, in the face of the relative rise of white-collar unionism, and the relative increase in the proportion of women trade unionists, lost his prominence as the traditional mainstay of trade unionism. In sum, unionized labor itself is an increasingly beleaguered and heterogenous group beset by division and fragmentation (Richards 1995, 1996).

206

Unions and Working-Class Fragmentation in Spain

At the same time, changes in the structure of the labor market and persistently higher rates of unemployment have led to the emergence of an insider-outsider division within national workforces and have accentuated (by extension) the divide between unionized and nonunionized workforces. Thus high rates of unemployment throughout the advanced capitalist world in the 1980s and 1990s not only sapped the bargaining strength of unions and deprived them of members, but tended to be concentrated in areas characterized, traditionally, by high rates of union density, such as heavy industry and manufacturing. In both absolute and relative terms, the manufacturing sectors of most advanced capitalist countries shrank considerably during the early 1980s and showed few signs of robust recovery thereafter. Levels of employment slumped, while the percentage of workers employed in manufacturing declined steadily throughout the 1980s. Such a trend ensured that unions based in these sectors suffered steady (and in some cases, staggering) membership losses. The decline of manufacturing and heavy industry not only undermined the traditional base of trade unionism, but changed the very structure of employment itself, with further adverse consequences for trade unions. In the 1970s, trade unionism had been underpinned by “well-organized, male full-time workers, employed in large manufacturing plants” (Winchester 1989: 514). In contrast, the 1980s saw an explosion of part-time, temporary, and relatively insecure jobs (especially among women and the young) located in smaller and more dispersed workplaces – all of which, as many authors point out, have been traditionally weak areas of union organization (Coggins et al. 1989; Fulcher 1991; Kern and Sabel 1991; Green 1992; Hyman 1992; Salvatore 1992; Richards 1995). As Hall (1987: 14) writes, “Nonunionized sectors employing many more part-time workers in small establishments and service occupations now constitute the fastest-growing segments of the economy. Hence, women, part-time employees, and workers in services or small industrial establishments, who were formerly marginal to the organized working class, form a growing portion of the labour force.” During the 1980s, therefore, a growing division emerged between unionized and nonunionized sections of national workforces (Hyman 1992: 151). Continuing structural and compositional changes created a core of insiders – relatively skilled, full-time unionized workers – and a periphery of outsiders – those in relatively insecure (temporary and/or part-time), semiskilled, low-paying, nonunionized jobs (Crouch 1986: 7;

207

Javier G. Polavieja and Andrew Richards

Hall 1987: 10–11; Pérez-Díaz 1987: 118; Taylor 1993: 146). General management drives for flexibility, moreover, accentuated the division. In essence, managers have sought to “isolate a core of privileged employees” (Boreham and Hall 1994: 334), involving the “functional flexibility of multi-skilled ‘core’ workers in full-time employment, and the numerical flexibility provided by the employment of trainee, part-time, temporary, or subcontracted workers. Such tendencies divide labor and set up barriers to its effective organization” (Fulcher 1991: 255; see also Lane 1989). Furthermore, pressures of economic decline, managerial aggression, rising unemployment, and the loss of industrial and political muscle tended to promote organizational defense as a key union priority. This undoubtedly led to increased interunion conflict as unions sought to compensate for membership losses by encroaching on other unions’ territory (Winchester 1989: 505–10). More important, though, it exacerbated divisions between the unionized and nonunionized workforces. Hall, for example, notes that management drives for flexibility intensified “the cross-cutting pressures that unions already feel between the demands of their core constituency, who are employed, and the broader penumbra of the working class that is under- or unemployed” (1987: 10). Crouch argues similarly that “if unions primarily represent the secure work force, their cooperation in restructuring may be bought precisely by requiring insecure groups to bear the brunt of adjustment” (1986: 8). Pérez-Díaz also observes that “in hard times [trade unions] have closed ranks and established a clear hierarchy of priorities at the expense of those below” (1987: 118).

Trade Unions and Working-Class Fragmentation In such an environment of division and fragmentation, the crisis of trade unionism revolves around its ability and/or willingness to speak for, or defend the collective interests of, workers as a whole. Is there a collective interest to be articulated and defended, or do unions effectively defend the interests of a relatively privileged section of the workforce? We argue, along with many others, that these questions should be posed and answered in the context of ongoing changes in the class structure of late industrial societies and, more specifically, the (alleged) fragmentation of the working class. Even prior to the 1980s, Hobsbawm (1978) warned in an influential essay that the working class was riven by increasingly serious conflicts of interest. The events of the 1980s merely reinforced his conviction that “workers are crumbling into groups with diverging and 208

Unions and Working-Class Fragmentation in Spain

contradictory interests” (1989: 74). In particular, the demise of blue-collar labor is seen as especially damaging for the foundations of trade unionism, and a dramatic manifestation of the wider process of working-class fragmentation. As Hyman (1992: 153) points out, the old “smokestack” industries are “typically seen as a natural generator of solidaristic collectivism.” With the sharp decline of several traditionally militant occupations, the 1980s bore witness to the disappearance of what has variously been termed the “classic labour movement” (Hobsbawm 1989), the “core working class” (Lash and Urry 1987: 5), or the “quintessential” members of the working class (Hall 1987: 14). This, moreover, was no mere quantitative change, but represented the demise of a set of powerful traditions and cultures within the labor movement. As a result, some authors have pinpointed the 1980s as a decade in which organized labor’s historic role as a class-conscious movement was lost for ever. Touraine, for example, explicitly links the decline of a “class conscious labor movement” to the “disintegration of the ‘classic’ working class . . . the very basis of class consciousness is disappearing and the labor movement is being replaced by interest group unionism” (1986: 161). Yet the implications of the demise of traditional blue-collar labor for working-class solidarity are not clear. On the one hand, some have argued that the decline of blue-collar labor and the concomitant rise of white-collar labor do not signal the demise of the working class per se. Hobsbawm argues that “the problem is not so much objective deproletarianization which has been brought about by the decline of oldstyle industrial labor, but is rather the subjective decline of class solidarity” (1989: 73). In addition, those arguing for the growing “proletarianization” of white-collar labor also dispute the disappearance of the working class per se (Braverman 1974; Aronowitz 1983; Kelly 1988). Others argue that the demise of blue-collar labor is synonymous with the disappearance of the working class itself. According to Gorz, for example, developments in late industrial societies have replaced the working class with “a non-class of non-workers or the ‘neoproletariat’ . . . the old working class . . . is no more than a privileged minority. Most of the population belong to the post-industrial neoproletariat” (Gorz 1991 cited in Giddens 1987: 279; see also Hyman 1992: 154). Other authors have also argued, in different contexts, that the former working class is, in fact, being broken down into competing groups with no common objective class interest. Thus Lockwood (1989) and Marshall and Rose (1988) have criticized the notion that white-collar labor is being “proletarianized,” arguing 209

Javier G. Polavieja and Andrew Richards

instead that it is informed by a very different set of interests. Meanwhile, as Fulcher notes, the emerging core/periphery distinction within late industrial workforces has been generalized to the level of the class structure as a whole by the notion of dualism (1991: 256). At a minimum, this implies that class-based strategies on the part of unions will no longer suffice as a means for bridging the divide between the increasingly beleaguered ranks of organized blue-collar labor and the growing “disprivileged underclass of employees” (Lane 1989: 605). However, great care must be taken in referring to unions’ “class-based strategies” or their alleged “position as the sword and shield of collective interests” (Sabel 1987: 45). Working-class and trade union division is nothing new: “a mythical belief in some previous golden age of proletarian unity and unproblematic trade-union solidarity distorts our perception of current labor-movement dynamics” (Hyman 1992: 166). That is why the historically contingent nature of trade union solidarity has to be acknowledged. History demonstrates that uniting workers in the same industry (let alone different ones), and forging a common class identity, were processes of painstaking construction: “from historical experience we can learn that there are no short-cuts to the identification and (re)definition of interests in a solidaristic manner; it is always necessary to campaign and struggle for (relative) unity among workers and their organizations” (Hyman 1992: 166). And even when established, trade unions have, historically, divided, as well as united, workers. The development of unions tended to perpetuate narrower occupational identities rather than broader class identity: “in embracing particular categories of workers as members and excluding others, each union gives institutional reinforcement to certain perceptions of common interest while presenting obstacles to alternative contours of solidarity” (Hyman 1985: 105). Even in the best of times, traditional blue-collar unionism, based on the skilled, male, manual working class, was exclusionary in nature (Heery and Kelly 1995: 163). In the infinitely worse times of the 1980s and 1990s, such a tendency has been exacerbated – the historic claim of union movements to represent, and speak for, the interests of workers as a whole is now seriously challenged by changes in the structure of employment outlined here. It is in this context that Hall (1987), surveying Europe in the late 1980s, laments the current inability of organized labor to summon up “moral authority” among national workforces as a whole. We now examine the extent to which trade unions have retained such authority in Spain, where very high levels of unemployment and tempo210

Unions and Working-Class Fragmentation in Spain

rary work have ensured that the insider-outsider division has become particularly acute.

Unemployment, Fixed-Term Work, and Labor Market Segmentation in Spain Two-Tier Flexibility in Spain In the ten years following the death of Franco, Spain suffered a crisis of employment which saw the rate of unemployment increase from 5.1 percent in 1977 to 21 percent in 1985 (Rhodes 1997: 114). In 1984, the government of the PSOE undertook a major reform of the labor market through the Reforma del Estatuto de los Trabajadores (Reform of the Workers’ Statute). The reform introduced fixed-term contracts in a bid to reduce unemployment (which had just exceeded the 20 percent level threshold) by making the labor market more flexible. Increasing flexibility promoted job creation in the expansion years. But despite a strong economic recovery during the second half of the 1980s, the unemployment rate itself never fell below 15 percent. When Spain suffered renewed economic recession in the early 1990s, unemployment soared again, climbing in 1994 to 24 percent of the active population, or 3.7 million people. More than a half (54.6 percent) of these had been seeking work for more than a year and almost a third (31.7 percent) for more than two years. At the end of 1998, 2.9 million Spaniards, or 18.2 percent of the active population, were out of work.2 Of these, around 20 percent were first-job seekers. Of the remaining 80 percent with job experience, more than 70 percent were unemployed owing to the termination of their fixed-term contracts. The flexibility strategy implemented in the 1984 labor market reform – which was reinforced through further legal changes in 1992 and 1994 – constitutes a prototypical case of what Esping-Andersen (1998a, 1998b) has labeled a two-tier selective labor market policy. Two-tier selective policies deregulate conditions only for some workers but not for others. In the Spanish case, flexibility through fixed-term employment was exclusively applied to new entrants in the labor market, while workers on permanent contracts continued to enjoy the privileges of rigid employment security 2

Spanish Labor Force Survey, 1998. However, significant regional variations in the unemployment rate ranged, in the fourth quarter of 1998, from 29.4 percent in Extremadura to 10.5 percent in Navarre.

211

Javier G. Polavieja and Andrew Richards

legislation inherited from the Franco regime, making their dismissal comparatively costly for employers (Polavieja 1998). Fixed-term contracts may be used for any activity (temporary or otherwise), may be signed for short periods (six months until 1992, one year since then), and renewed for up to three years (four years since 1993). When the period of the last possible renewal expires, the firm must either offer the worker a permanent contract or dismiss her. Nonrenewal of fixed-term contracts entails very low firing costs and cannot be challenged in court. While the severance pay for permanent contracts varies from twenty to forty-five days’ wages per year of employment, plus legal costs, fixed-term contracts (depending on their type) either bear no firing costs at all or imply only twelve days of wages per year of service (Bentolila and Dolado 1994: 67). The differentiation of employment-termination costs by type of contract, together with the expiration date incorporated into fixed-term contracts, give incentives to employers to effect labor adjustments exclusively through fixed-term work. Fixed-term contracts therefore become both the principal channel of entry into employment and the principal channel of exit from employment to unemployment. The outcome of this process is segmentation by type of contract. The labor market is increasingly divided between a core of permanent workers and a growing periphery of insecurely employed workers who bear the brunt of employment adjustments. While the former group continues to enjoy stable employment trajectories, the labor market histories of those in the latter group are likely to consist of a combination of recurrent unemployment and short-term work. This process can be studied by looking at (1) entries into employment, (2) exits from employment to unemployment, and (3) indirect indicators of labor turnover. Entries into Employment Data from the Spanish Labor Force Surveys (LFS) for the 1987–97 period show that the proportion of fixed-term work among the newly employed – that is, those who were unemployed a year before the survey was carried out – rose from 58.8 percent in 1987 to 83.6 percent in 1991, and further to 87.9 percent in 1997 (second quarters). This meant a dramatic increase in the total proportion of the Spanish workforce on fixed-term contracts. Between 1987 and 1991, the rate of fixed-term work in Spain doubled from 16 percent to 32.6 percent (Bentolila, Segura, and Toharia 1991: 237–38; Jimeno and Toharia 1994: chaps. 1 and 4). By 1995, the rate of fixed-term work had increased to 212

Unions and Working-Class Fragmentation in Spain

surpass the 34 percent level (Polavieja 1998). The sheer scale of fixed-term work in Spain is unparalleled anywhere else in Western Europe ( Jimeno and Toharia 1994: 96).3 The policy on flexibility has meant, therefore, that temporary work has become the principal means of entry into employment in Spain. Yet the rate at which fixed-term contracts have been converted into permanent ones is very low. For example, between the second quarter of 1987 and the second quarter of 1988, when employment was rising, only two out of ten temporary workers were given permanent contracts ( Jimeno and Toharia 1994: 111). Later data show that this rate declined significantly thereafter, remaining at around 10 percent per year in the 1990–95 period (Alba 1991, 1997; Segura et al. 1991; Güell-Rotllan and Petrongolo 1998; Polavieja 1998). The data also suggest that the transition from fixed-term to permanent employment is largely restricted to the most skilled occupations (Bentolila, Segura, and Toharia 1991; Segura et al. 1991; Polavieja 1998). The introduction of fixed-term contracts increased the sensitivity of employment to the economic cycle and facilitated the creation of employment in the growth years. Worldwide economic recovery and Spain’s entry into the European Community in 1986 provided the Spanish economy with a very favorable context for employment growth. Between 1985 and 1991, 1.7 million jobs were created at a rate unknown even during the economic boom of the 1960s (Toharia 1994: 112). Yet by the same token, higher employment sensitivity implied that when the effects of the worldwide economic recession of the early 1990s hit the Spanish economy in 1992, levels of employment fell dramatically (more than one million jobs were lost between 1992 and 1993) and unemployment reached 24 percent of the active population at the beginning of 1994 (Antolín 1995). Fixed-Term Contracts and Unemployment Temporary work has become the principal means of exit from employment at least since 1987, which is the first date for which we have reliable data on fixed-term work (see Argentaria 1993; Jimeno and Toharia 1994; Encuesta Sociodemográfica in Argentaria 1995: 57–59; Polavieja 1998). Original analysis of the LFS data shows that already in the second quarter of 1987, 55.7 percent of the unemployed had been on fixed-term contracts. This proportion rose to 73 3

In 1992, the proportion of the working population in Spain on temporary contracts (32 percent) was nearly four times the EC average of 9 percent (Martínez Lucio and Blyton 1995: 351).

213

Javier G. Polavieja and Andrew Richards

percent in 1991 and since then has never fallen below 72 percent. Meanwhile, the proportion of unemployed coming from permanent employment has dropped consistently. In 1987, 27.9 percent of the unemployed with job experience had been dismissed or had agreed to early or healthrelated retirement, forms of termination of employment contracts exclusively applied to permanent workers. Yet by 1991, the figure had fallen to 12.3 percent. Moreover, despite the massive destruction of employment in the period 1992 and 1994, the proportion of unemployed people coming from permanent contracts increased to only 12.4 percent in 1993 and 13.7 in 1995. In 1997, 73.3 percent of the unemployed came from fixed-term employment, whereas only 13.4 percent of the unemployed had become so following the termination of their permanent contracts (calculated by the authors using LFS data for the period 1987–97). Employment destruction has essentially meant the destruction of fixed-term jobs (Polavieja 1998). Two-Tier Flexibility and the Increase in Job Turnover With the extension of two-tier flexibility, growing numbers of fixed-term workers have moved back and forth between unemployment and temporary work. This has led to the emergence of precarious labor market trajectories characterized by recurrent periods of unemployment and temporary work. Evidence for the existence of such unstable trajectories comes from a variety of indirect indicators obtained from the LFS. One such indicator is the duration of fixed-term contracts. Original analysis of the LFS data shows that the average duration of temporary contracts in the period 1987–97 was around twelve months. Despite the fact that, from 1984 until 1993, the maximum legal duration of fixed-term contracts was three years, and that thereafter the legal limit was extended to four years, the average duration of fixed-term contracts has consistently been less than half the legal maximum. This is a clear indication of labor precariousness4 (see also Bentolila, Segura, and Toharia 1991: 237–38; 4

Other secondary data reinforce our findings regarding the brevity of the fixed-term contract duration. Martinez Lucio and Blyton (1995) calculated that in 1991, nearly 90 percent of job-creation contracts (contratos temporales de fomento al empleo) – which accounted for around 20 percent of all employment contracts registered with the INEM (Instituto Nacional de Empleo) – were for a period of no more than six months. Furthermore, contracts for specific services (contratos para obra o servicio) and casual contracts (contratos eventuales) – neither of which were subject to any specific legal minimum period – accounted for 60 percent of all new registered employment contracts (Martínez Lucio and Blyton

214

Unions and Working-Class Fragmentation in Spain

Polavieja 1998). In sharp contrast, the average duration of permanent contracts is ten years (calculated by the authors using LFS data for the period 1987–97). Another particularly interesting indicator of the spread of precarious labor trajectories is the evolution of the proportion of employed fixed-term workers who were employed in a different job twelve months before the LFS was carried out. An increase in labor turnover should be reflected in an increase in this proportion, since it is reasonable to believe that a short spell in unemployment might have been experienced between the previous job and the job held at the time of the LFS. Data obtained from the LFS confirm this hypothesis. In 1987, 20.2 percent of temporary workers had been employed in a different job twelve months earlier. This proportion increased between 1987 and 1997, reaching 50.1 percent by 1995. Meanwhile, in the case of permanent workers this figure remained more or less constant at around 3 percent (see also Polavieja 1998). Conversely, the proportion of fixed-term workers who had been employed in the same job one year before the survey took place, after increasing from 30.3 percent to 36.2 percent between 1987 and 1991 (the years of economic expansion), dropped to 25.2 percent in 1993, and still further to a mere 11.9 percent in 1997. This is a clear indication of labor precariousness. Yet for permanent workers, the proportion of those who had been employed in the same job twelve months earlier rose from 88.8 percent in 1987 to 93.6 percent in 1991 and remained above the 90 percent level thereafter. The contrast between both figures illustrates dramatically the unequal distribution of employment security between the two types of worker. Two-tier flexibility has led to a new pattern of unstable labor market trajectories of recurring periods of unemployment and temporary work. The rapid spread of this type of precarious trajectory has become one of the main features of the Spanish labor market since 1984. In sum, the particular characteristics of the labor market flexibility strategy implemented in Spain have meant that employment adjustments have been concentrated on fixed-term workers. As a result, the labor market has 1995: 351). Equally, the European Industrial Relations Review (1997) has estimated that by 1996, 70 percent of the temporary contracts signed for a specified fixed term had a duration of less than three months, and only 0.4 percent for a duration of more than one year. Of the fourteen types of contracts available, 80 percent of new contracts signed in 1996 belonged to the three categories of casual work (contratos eventuales), contracts for work or services (contratos por obra o servicio), and part-time contracts (contratos a tiempo parcial) (EIRR 1997: 25).

215

Javier G. Polavieja and Andrew Richards

become segmented along the lines of an insider-outsider division. Flexibility has not solved the problem of unemployment but has confined it largely to a particular group of workers. Today, fifteen years after the 1984 reform, Spain not only has the highest proportion of temporary workers of all OECD countries (over 34 percent) but still has the highest rate of unemployment of all OECD member states (approximately 15 percent) (see OECD 1999).

Type-of-Contract Segmentation and Class The process of segmentation by type of contract in Spain has taken place within all occupational classes. In fact, it is because segmentation can be detected in all classes that one can talk of insider-outsider mechanisms operating in the Spanish labor market (Polavieja 1998). To illustrate this point, we analyze the class composition of unemployment. In terms of determining the class composition of the unemployed, an important qualification has to be made. A very significant component of unemployment in Spain – some 20 percent – is comprised of people who have never worked. (Of these, more than 65 percent are women,5 and more than 75 percent are under thirty years of age [Jimeno and Toharia 1994: 124]). Here, we will not consider this section of the unemployed but focus, instead, on the remaining 80 percent of the unemployed with previous work experience. It is only with the latter category of unemployment that it is possible to establish, unequivocally, the relationship between unemployment and social class. In the case of the unemployed who have never worked, the relationship is much more difficult to infer, both technically and conceptually. Statistical analysis based on the last type of occupation held by the unemployed with work experience enables us to discern which occupa5

The female participation rate in the labor force in Spain increased from 26.9 percent in 1981 to 34.9 percent in 1993, while women’s share of employment increased from 28.76 percent in 1981 to 33.98 percent in 1994 (Richards 1995: 48–49). Unemployment itself has become an increasingly feminine issue – in 1991, women represented almost 52 percent of the unemployed, even though they accounted for only one-third of those employed. Women appear less able to escape unemployment than men. Moreover, their chances of doing so decrease steadily with advancing age ( Jimeno and Toharia 1994: 37, 52). Women are also more likely to experience unemployment because their job trajectories are likely to include periods of inactivity. Once women return to the labor market, they reenter with a fixed-term contract and thus their chances of future periods of unemployment increase considerably.

216

Unions and Working-Class Fragmentation in Spain

Table 7.1. Unemployment by Class Occupational Classes

Employed

Unemployed

Total

Service (I/II) N (hundreds) % of class

36,826 93.0

2,793 7.0

39,619 100.0

Intermediate (III) N (hundreds) % of class

37,663 80.2

9,277 19.8

46,940 100.0

Self-employed (IV) N (hundreds) % of class

15,370 97.3

428 2.7

15,798 100.0

Skilled manual (V/VI) N (hundreds) % of class

27,768 82.8

5,752 17.2

33.520 100.0

Unskilled manual (VII) N (hundreds) % of class

9,437 62.7

5,610 37.3

15,047 100.0

Total N (hundreds) % of class

127,064 84.2

23,860 15.8

150,924 100.0

Source: LFS 1997 (second quarter). Calculated by the authors. Weighted.

tional categories have the highest concentration of unemployment. In Table 7.1, we have cross-tabulated a model of six occupational classes6 with a dichotomous variable of value “0” if the individual surveyed is working, and “1” if the individual is unemployed. These data refer to the LFS for the second quarter of 1997. Table 7.1 shows that unemployment follows a class distribution. Three main groups can be distinguished. The first consists of self-employed and service-class workers (that is, professionals), with rates of unemployment of 2.7 percent and 7 percent, respectively. The second is made up of intermediate (white-collar) workers and skilled manual workers (including here supervisors of manual work) with unemployment rates of 19.8 percent and 17.2 percent, respectively. Third are unskilled manual workers, showing a very high unemployment rate of 37.3 percent. The distribution of fixed-term work by class shows the same pattern: service-class professionals have the lowest rate of fixed-term work (19.7 6

Derived from Goldthorpe’s eleven-position scheme. See Erikson and Goldthorpe (1992).

217

Javier G. Polavieja and Andrew Richards

Table 7.2. Type of Contract by Class Type of Contract Occupational Classes

Permanent

Fixed-term

Total

Service (I/II) N (hundreds) % of class

19,581 80.3

4,806 19.7

24,387 100.0

Intermediate (III) N (hundreds) % of class

24,260 68.4

11,188 31.6

35,448 100.0

Skilled manual (V/VI) N (hundreds) % of class

16,624 61.2

10,545 38.8

27,169 100.0

Unskilled manual (VII) N (hundreds) % of class

3,411 36.8

5,846 63.2

9,257 100.0

Total N (hundreds) % of class

63,876 66.4

32,385 33.6

96,261 100.0

Source: LFS 1997 (second quarter). Calculated by the authors. Weighted.

percent); intermediate and skilled manual workers have medium-level rates (31.6 percent and 38.8 percent respectively); and unskilled manual workers a very high rate (63.2 percent) (see Table 7.2). The data presented so far are consistent with the view that employment insecurity is related to class, being higher in those occupations which do not require special market capacity and where workers are thus largely replaceable. However, LFS data for 1997 also show that the differentiation between insider and outsider workforces occurs within the ranks of all occupational classes. In Table 7.3, class and unemployment have been cross-tabulated again but this time are also differentiated by type of contract for each class category. Table 7.3 shows how the process of segmentation occurs within all occupational classes.7 For example, the unemployment rate of fixed-term 7

Here, we have computed a variable that identifies the type of contract that the employed respondent had when the LFS was carried out and of the last job that the respondent had before losing if currently unemployed. The latter type of information requires a certain degree of recodification since the LFS does not include a question on the type of contract in the last job. It does contain a question regarding the causes of job loss in the last

218

Table 7.3. Unemployment Rates and Average Tenure of Unemployed Respondents’ Last Job in Months by Type of Contract and by Class

Occupational Classes Service (I/II)

Intermediate (III)

Skilled manual (V/VI)

Unskilled manual (VII)

Unemployment Rates Type of Contract Permanent N (hundreds) % Fixed-term N (hundreds) % Total N (hundreds) % Permanent N (hundreds) % Fixed-term N (hundreds) % Total N (hundreds) % Permanent N (hundreds) % Fixed-term N (hundreds) % Total N (hundreds) % Permanent N (hundreds) % Fixed-term N (hundreds) % Total N (hundreds) %

Unemployed Respondents’ Average Job Tenurea

Employed

Unemployed

Total

19,581 97.7

458 2.3

20,039 100.0

124

4,806 75.1

1,590 24.9

6,396 100.0

16

24,387 92.3

2,048 7.7

26,435 100.0

47

24,260 95.1

1,259 4.9

25,519 100.0

93**

11,188 62.1

6,815 37.9

18,003 100.0

14**

35,448 81.4

8,074 18.6

43,522 100.0

16,624 92.7

1,304 7.3

17,928 100.00

151**

10,545 72.4

4,020 27.6

14,565 100.0

15

27,169 83.6

5,324 16.4

32,493 100.0

50

3,411 90.9

343 9.1

3,754 100.0

110

5,846 53.7

5,050 46.3

10,896 100.0

7**

9,257 63.2

5,393 36.8

14,650 100.0

29**

14**

Source: LFS 1997 (second quarter). Calculated by the authors. Weighted. Note: Analysis of variance has been undertaken using Bonferroni’s multiple comparisons of means. The asterisks denotes a significant difference (** significance £ 0.000) with respect to the serviceclass mean tenure of the same contractual category. a Mean duration in months of last job.

219

Javier G. Polavieja and Andrew Richards

workers in the service class is 25 percent. This rate is more than ten times the rate of permanent workers in the same class (2 percent), but also more than three times the rate of permanent workers in the manual working class (7 percent) and almost three times higher than the rate of permanent workers in the unskilled manual class (9 percent). These data suggest the extent to which type of contract, more than occupational class, determines unemployment risks in Spain. Table 7.3 also shows the mean duration in months of the last job of the unemployed respondents belonging to each class and type of contract categories. For example, unemployed service professionals coming from permanent employment show a mean tenure in their last job of 124 months, whereas unemployed service professionals coming from temporary employment show an average tenure of only 16 months. This pattern is typical of all classes. Of course, the shortness of the reported tenure in fixed-term employment was to be expected, given the particular characteristics of the contract and what we know about it so far. What is remarkable, however, is the very long duration of permanent contracts in all classes. For example, the 7 percent of skilled manual workers (previously) on permanent contracts who are unemployed report an average duration in their last job of 151 months, which is even longer (and significantly so) than the average duration of service-class professionals coming from permanent contracts. We also note that formerly permanent unskilled manual workers report an average duration in their last job of 110 months, which is not significantly shorter than the average duration in employment of formerly permanent service-class professionals. The contrast between these figures and those for fixed-term workers in all class categories gives occupation. Among these causes is included the termination of temporary or other types of fixed-term contracts. Logically, we assumed that those workers who declared themselves to be unemployed as a result of the termination of their contracts came from fixed-term work. Conversely, we considered respondents who claimed to be unemployed as a result of regulatory dismissal (despido por regulación), reconversion dismissal (despido por reconversión), dismissal for other causes, regulatory suspension, early retirement, sickness retirement, or other forms of retirement (all of which are forms of termination typical of permanent contracts), to be coming from permanent employment. Unemployed respondents who claimed to be unemployed as a result of the termination of self-activity ( fin de actividad propia), family-related causes, voluntary decision, or initiation of studies were not included in our variable. Of this latter group, respondents who declared the first reason (end of self-activity), are either employers or self-employed and therefore do not concern us here. Respondents who declared themselves to be unemployed for the other three remaining causes (voluntary, family-related, or studies) amount to only 11 percent of all the unemployed.

220

Unions and Working-Class Fragmentation in Spain

us a clear sense of the segmentation process, enabling us to characterize the Spanish labor market as an insider-outsider market. In sum, the Spanish labor market is highly segmented. Employment insecurity in Spain is clearly concentrated among workers on fixed-term contracts who have borne the brunt of employment adjustments. Although employment insecurity follows a class pattern, since both fixed-term employment and unemployment are significantly overrepresented in working-class occupations, we have shown nonetheless how the type of contract determines an individual’s chances of unemployment regardless of occupational class. Spain seems to be a particularly acute example of the general international trend toward labor market dualization to which we referred earlier in the chapter. There now exists a profound segmentation of Spanish workers (Rhodes 1997: 114) between, insiders on permanent contracts with a more stable, secure and long-term job trajectory, on the one hand, and outsiders who are in a much more precarious situation, moving constantly between temporary work and unemployment, on the other. Comparing both groups, we emphasize the following differences: 1. The fact that temporary contracts involve substantially lower firing costs than typical permanent contracts has generated a two-tier system in employment relations (Jimeno and Toharia 1994: 96). There is also strong evidence of wage discrimination against outsiders. The latter, when they work, are less likely to have access to the wage-related benefits (bonuses, incentives, etc.) which insiders enjoy. Moreover, for the same work, outsiders may earn lower wages than those determined by collective agreement between unions and employers (Fernández, Garrido, and Toharia 1991: 74; see also Toharia and Muro 1988; Albarracín 1990, in Recio 1991: 109). Indeed, Jimeno and Toharia (1994: 110) argue that workers on fixedterm contracts earn approximately 10 percent less than permanent workers. 2. Outsiders are not entitled to as many welfare benefits as their securely employed counterparts, especially with respect to unemployment insurance (Rojo Torrecilla 1990: 33; Toharia 1994: 115– 16). 3. Since the dismissal costs of outsiders and their capacity to resist are lower, they suffer harsher working conditions in terms of hours, pace of work, supervision, and productivity demands (Rojo Torrecilla 1990; Recio 1991; Argentaria 1995: 115–16). 221

Javier G. Polavieja and Andrew Richards

4. The accident rate among outsiders is significantly higher than for insiders (Recio 1991). 5. Compared with insiders, outsiders are much less likely to encounter possibilities for promotion and for the acquisition of skills (Rojo Torrecilla 1990). Such differences are, of course, embedded within a specific framework of industrial relations. As a result, the process by which Spanish workers are being segmented may well be affected by the policies of representation adopted by the trade unions. In theoretical terms at least, a union can develop an inclusive strategy that attempts to represent the interests of all workers, both insiders and outsiders. On the other hand, it can adopt policies that tend to privilege insiders (who are within or near their sphere of influence), to the detriment of outsiders who are thereby abandoned to the ongoing process of labor market dualization.8 In the next section, we examine the various dilemmas of Spanish unions in a context of entrenched labor market segmentation.

Insiders, Outsiders, and Trade Unions in Spain The Problems of Trade Unionism in Contemporary Spain How have unions in Spain been affected by the sharp dualization of the labor market between insiders and outsiders? Very high rates of unemployment, the enormous growth of temporary work, and the scale of economic reform and restructuring have presented the Spanish trade union movement with particularly grave challenges. In overall terms, structural changes in the Spanish economy during the 1980s and 1990s have tended to work against the unions. Industrial restructuring in particular has struck most heavily in sectors of the economy where union strength was greatest, such as steel and engineering, mining, shipbuilding and textiles. Large plants have either shed labor or closed down completely, while some regions (for example, Asturias and Galicia) have witnessed the demise of their principal industries (Estivill and de la Hoz 1990: 286–87). The manufacturing sector of the economy – traditionally a source of union strength – also shrank, despite the recovery in levels of employment during the second half of the 1980s. Between 1981 and 1993, the proportion of the Spanish workforce 8

For a detailed examination of these issues, see Iriso (1993).

222

Unions and Working-Class Fragmentation in Spain

employed in manufacturing fell from 24.03 percent to 19.9 percent (Richards 1995: 10). In parallel, the introduction of new technology, the growing prominence of technical and professional staff, and the general shift toward a more service-oriented economy have placed increasing numbers of wage earners into jobs further removed from the direct sphere of production. Some of the most dynamic sectors of the Spanish economy, such as marketing, planning, financial services, advertising, tourism, and foodprocessing, have very little tradition of trade unionism (Estivill and de la Hoz 1990: 287; Jimeno and Toharia 1994: 9, 16–17). The presence of a substantial underground economy (in industries such as toys, shoes, and textiles) and the prominence of small-scale enterprises in the Spanish economy represent further obstacles to unionization (Estivill and de la Hoz 1990: 287; Rigby and Lawlor 1994: 262, 266; Rhodes 1997: 114). In 1994, rates of unionization stood at 23.3 percent in the industrial sector, 12.9 percent in agriculture, 12.3 percent in services, and 10.8 percent in construction (van der Meer 1995: 52–54). Yet at the end of 1996, the service sector employed 63.1 percent of the salaried workforce, and the industrial sector only 23.3 percent.9 In addition to confronting adverse structural change, unions lost power and influence. In the late 1970s and early 1980s, unions were recognized as important and legitimate social actors and received state subsidies (Estivill and de la Hoz 1990: 279–93). In the context of Spain’s transition to democracy, their cooperation with, and support for, measures of economic stabilization were sought by governments and employers alike. Between 1978 and 1986, the unions did, to a remarkable degree, lend critical support to a series of economy-wide agreements relating to wage increases, employment conditions, and the design of the collective bargaining structure and conflict resolution.10 Most notably – and in sharp contrast to the stance they had adopted in the immediate aftermath of Franco’s death – the unions adhered to the policies of wage moderation initiated in the Moncloa Pacts of 1978 and subsequently consolidated in 1980, thereby ensuring that Spain was one of the few OECD countries to register a steady decline in real wage rates and unit labor costs during the

9

10

El País, April 13, 1997. Construction and agriculture accounted for 9.8 percent and 3.9 percent of the salaried workforce, respectively. Five important pacts and agreements were reached during this period. For a chronology, see Jimeno and Toharia (1994: 77).

223

Javier G. Polavieja and Andrew Richards

1980s (Jimeno and Toharia 1994: 71, 76). Both major trade union confederations, the Unión General de Trabajadores (UGT) and the Comisiones Obreras (CC.OO), influenced considerably the contents of the 1980 Workers’ Statute (Martínez Lucio and Blyton 1995: 355). Even more remarkably, the UGT consented to the measures contained in the 1984 Reforma del Estatuto de los Trabajadores, which paved the way for the subsequent massive growth of low-cost, fixed-term contracts (Rhodes 1997: 115). There were, of course, powerful incentives for the unions to support national measures and mechanisms of concertation. Emerging from four decades of dictatorship, the unions saw concertation with government and employers as a means of consolidating democracy in general, of legitimizing themselves as actors in the new regime, and of achieving their own consolidation as organizations (Martínez Lucio and Blyton 1995: 355). Nor could unions afford to ignore the economic context in which they operated – for example, the depths of the economic recession in Spain, and pressure from various groups of workers (especially the young) lay behind the unions’ decision to consent to the expansion of fixed-term contracts in 1984. (Though, as Rhodes [1997: 115] points out, they could not have foreseen at that point that the consequent casualization of employment would prove highly detrimental to their future strength.) Although the unions saw their own cooperation and support as a quid pro quo for gaining both substantive benefits and influence over government employment policy, structural changes in the Spanish economy so weakened the unions that by the mid-1980s the need for either employers or the government to consider the demands of organized labor had diminished considerably. For its part, the PSOE government felt secure enough, with the consolidation of democracy and its own reelection in 1986, to risk its close alliance with the UGT by moving away from the hitherto politically stabilizing framework of concertation and adopting a strategy of relatively strict economic and monetary discipline (Martínez Lucio and Blyton 1995: 355–56; Rhodes 1997: 115; Astudillo 1998). There were serious differences between the government and the unions with respect to the problem of unemployment from the mid-1980s onward. For the government, “active” measures adopted to reduce unemployment have been largely confined to the promotion of fixed-term employment contracts ( Jimeno and Toharia 1994: 128). In contrast, the unions have long advocated a broader range of measures to reduce unemployment, including shortening the standard working week to thirty-five hours, 224

Unions and Working-Class Fragmentation in Spain

reducing the use of overtime, and promoting work sharing. The joint UGT/CC.OO. Propuesta Sindical Prioritaria of 1989 proposed job creation via a shorter working week, earlier retirement, the control and eventual elimination of overtime, greater investment in training and education, and – to enhance job security – tighter controls on temporary employment contracts (the effects of which, in terms of increased job insecurity, were now more than apparent to the unions). These proposals were ignored by employers and the government (Rhodes 1997: 111–12, 116, 119).11

Labor Market Flexibility and the Unions’ Predicament The diverging paths of government and unions, and the latter’s loss of influence, have had important consequences. Resentful and frustrated at their inability to influence government policy toward unemployment, the unions’ support for concertation in the first half of the 1980s gave way to growing hostility and resistance toward government policy in the second half of the 1980s and early 1990s. This became more explicit as the government and employers attempted to implement more radical reforms of the labor market. We have noted how flexibility of the labor market was introduced by diminishing the job security of new entrants into employment (via the use of fixed-term contracts) but without diminishing the job security of those already employed on permanent full-time contracts (see also Jimeno and Toharia 1994: 109). It is the latter group which has become, more recently, a target for further labor market reform. Despite the enormous growth of temporary work over the last ten years, government and employers have increasingly viewed the protection surrounding, and the high dismissal costs associated with, permanent workers as the principal obstacle to greater labor market flexibility. Yet it is from among permanent workers that the unions draw their greatest organizational strength. As a result, the issue of dismissing permanent workers has emerged as a highly charged sticking point in the labor market reform process in contemporary Spain (Estivill and de la Hoz 1990: 295; Rhodes 1997: 118). However, Spanish governments have neither been able nor willing to unilaterally impose major alterations to the existing costs and procedures that protect the core insider workforce. Union agreement to such reforms 11

However, average annual overtime hours have been bargained down over the years, albeit with no obvious impact on the rate of unemployment (Rhodes 1997: 111–12).

225

Javier G. Polavieja and Andrew Richards

has been essential, and while unions ceased to be critical partners for either the state or employers after 1985, they nonetheless retained the ability to resist and sometimes block government policy. Indeed, despite relatively low levels of unionization in Spain, trade unions have wielded significant veto power based on their more widespread support via workplace elections, and their capacity (albeit diminished since the late 1980s) for strike mobilization12 (Rhodes 1997: 118). And the unions have been extremely reluctant to countenance any dismantling of the apparatus protecting the permanent workforce. The reasons for the tenacity with which the unions have defended the rights and conditions of permanent workers are both historical and organizational. In historical terms, the unions’ response to efforts to diminish the job security of permanent workers reflects a legacy of the Franco era. What are now considered by government and employers to be excessive rigidities in the labor market were, in fact, a form of vital (albeit paternalistic) job protection during the dictatorship. In the post-Franco context, however, the unions have used such historic state intervention as the cornerstone of workers’ rights – especially in a democracy characterized by persistent high unemployment, enhanced employer power, and, in the unions’ view, inadequate social welfare provision (Lucio and Blyton 1995: 340–49). In addition, though, there are powerful organizational reasons for the defense of permanent workers, for it is precisely in this sector of a changing labor market that the unions are most solidly represented. As such, defense of their interests “is the guarantee, in the last resort, of their own survival” (Estivill and de la Hoz 1990: 297; Rhodes 1997: 115). However, as many authors have emphasized, “falling back” on their core constituency has had grave implications for the unions’ relationship with the Spanish workforce as a whole. To be sure, unions have persisted in their calls for more radical and interventionist strategies to tackle unemployment, and for greater job security for all workers. Yet whether through choice or force of circumstance, the fact remains that unions have concentrated their efforts on the defense of permanent workers. In doing so, they have steadily lost contact with, and helped promote the alienation of, 12

In 1988, the unions succeeded in forcing the government to retreat from a controversial plan for youth employment. General strikes were organized in May and October 1992 against the modification of entitlements to unemployment benefits, and in January 1994, a general strike brought Spanish industry to an almost complete standstill in protest against what the unions termed an “unprecedented deregulation of industrial relations” (Rhodes 1997: 109, 116, 118).

226

Unions and Working-Class Fragmentation in Spain

a growing proportion of the national workforce. The unemployed, the precariously employed, women, and the young are gradually being left out of the predominantly male, blue-collar culture of the trade union movement (Estivill and de la Hoz 1990: 297–98; Rhodes 1997: 115–16). Hence the unions are implicated in the division between insiders and outsiders and are vulnerable to the charge that regardless of their own programmatic statements, their defense of the “unsolidaristic” employment rights of their core membership has contributed to the effective exclusion of other categories of workers (Estivill and de la Hoz 1990: 298; Martínez Lucio and Blyton 1995: 358). Unions consequently find themselves in a difficult predicament. High unemployment and high levels of temporary work have forced them to retreat to their core constituency. Yet defense of the latter is far from compatible with a general commitment to the reduction of unemployment and the promotion of greater job security. Our evidence strongly suggests that the unions have failed to develop that “single strategic design” (Estivill and de la Hoz 1990: 287) capable of bridging the divide between insiders and outsiders.

Labor Market Segmentation and Workers’ Involvement in the Unions What are the implications of the division between insiders and outsiders for the way in which trade unions in Spain are perceived and evaluated? In this section, we attempt to determine the extent to which the marginalization suffered by outsiders in the labor market has led them to reject the unions. Do “objective” differences between insiders and outsiders produce “subjective” differences in their respective evaluations of the unions, and do these differences affect the behavior of workers in relation to trade union activity? The effects of labor market segmentation have tended to be studied in terms of levels of union affiliation. As such, the principal effect of segmentation has been to provoke a growing division of labor between insiders, with generally high levels of affiliation, and a periphery of outsiders, with low or nonexistent levels of union affiliation. However, this perspective poses serious problems in the case of Spain where, given that trade unionism is characterized as one more of “voters” than of “membership” (Martín Valverde 1991: 24–25; Martínez Lucio 1993: 500–501), the level of union affiliation is not a reliable measure of union support. Thus while it appears clear that the great majority of outsiders are not 227

Javier G. Polavieja and Andrew Richards

union affiliated, and that the great majority of union-affiliated workers are insiders, in Spain – with a level of affiliation of approximately 12 to 15 percent – the great majority of insiders are, in fact, not affiliated to unions. As a result, it is more appropriate to analyze the effects of segmentation on union support as effects on workers’ level of union involvement (implicación sindical).13 By union involvement, we mean that set of attitudes, evaluations, and activities which indicates that a worker feels identified with the unions (that is, feels represented by them) and who is thus interested personally in union activity. The ideal-typical involved worker, therefore, will not just display pro-union sentiments and attitudes, but will act accordingly. Such a worker will vote in union elections when they are convoked in the workplace, will be personally interested in the activities and functioning of the union committees in the workplace, and will attend, if they take place, union meetings on collective action. In some extreme cases, the involved Spanish worker will affiliate with the union. Our conception of union involvement therefore includes both an attitudinal and evaluative component and one of concrete behavior. It is obvious that pro-union behavior requires pro-union attitudes. In addition, though, pro-union behavior needs an appropriate environment. Involved workers, for example, cannot vote in union elections if no such elections take place. Nor can they take an interest in the affairs of the union committee if the company in which they work employs fewer than 49 people (the legal minimum for the establishment of union committees). Nor can they participate in strikes if none are called. Involved, or prounion, behavior therefore requires the institutional and organizational conditions which make possible the direct presence – and thereby the mobilizational capacity – of the unions. Many of these contextual conditions are directly related to the size of the company. The more employees 13

This is not to discount the importance of union membership levels per se. Total union membership rose from 1.697 million in 1990 to 2.166 million in 1993 (before slipping back to 2.127 million in 1994). The net union density rate in Spain rose from 13.32 percent in 1990 to 16.38 percent in 1993 (before falling to 15.82 percent in 1994). Such data challenge the overall thesis of “union decline” but do not contradict the thesis that trade unionism in Spain is increasingly based in, and identified with, the core permanent workforce. Indeed, the recovery in union membership in the early 1990s is based partly on permanent workers who were previously nonmembers sympathetic to the unions actually joining. Union affiliation rates among temporary workers have remained low – in 1992, 34 percent of Spanish workers were temporary, but only 18 percent of the CC.OO.’s membership was made up of temporary workers (Jordana 1996: 215, 216, 218, 220).

228

Unions and Working-Class Fragmentation in Spain

there are in the workplace, the greater are the legal and strategic possibilities for effective union organization. Within the European Union, Spain has the lowest proportion of employees in firms employing more than 500 workers (only 19 percent of salaried workers in 1989), and the greatest proportion of employees in firms employing fewer than 50 workers (53 percent of salaried workers in 1989) (Martínez Lucio 1993: 494–95). This is of fundamental importance in explaining the unequal distribution of trade union presence in Spain. But even in large companies with a strong union presence, the likelihood of a worker participating in union-organized collective action may depend on other contextual factors. We argue that occupying a precarious position in the labor market (precariedad laboral) has a decisive effect on pro-union behavior. Iriso (1993: 427) has argued that the temporary worker, given his or her weak position within the company, may consider participation in union action to be an excessive risk in the face of the employer. It is this weak position in the internal labor market of the employed temporary worker that limits enormously the possibilities of taking part in collective action, regardless of one’s personal attitudes toward, and evaluations of, the trade unions. We therefore argue that labor market precariousness has a direct effect on pro-union behavior that is independent of evaluative questions. A precarious or unstable position in the labor market makes it unlikely that workers who are subjectively involved with the unions will transform such involvement into concrete action. More important, an unstable working situation also affects attitudes toward, and evaluations of, the unions in and of themselves – producing feelings of de-identification, distance, apathy, lack of confidence, and even outright rejection. Trajectories in working life that include temporary work, (long) periods of unemployment, or alternation between temporary work and unemployment can undermine the feeling that the unions represent and defend the common interest of all workers. In a precarious working situation, identification with the unions may be replaced by the feeling that the worker is alone in the face of the employer, or alone in the face of unemployment, and that no one is defending the employee’s interests. Therefore a precarious working situation also diminishes subjective involvement with the unions. In sum, our general hypothesis is that labor market precariousness reduces union involvement because (1) it seriously impedes collective action (subhypothesis 1) and (2) it produces sentiments and attitudes of apathy toward, and rejection of, the unions. The worker in a precarious 229

Javier G. Polavieja and Andrew Richards

situation neither identifies with, nor feels represented by, the unions (subhypothesis 2). Subhypothesis 1: Labor Market Precariousness Diminishes Pro-Union Activity In order to test the hypothesis that the labor precariousness attached to type of contract hinders pro-union activity, we have analyzed the effects of type of contract on union membership, union voting, and participation in strikes or stoppages in the firm. The statistical source for this analysis is a subsample of the employed workforce from the Spanish Survey on Trade Union Activity (CSRSTUA) conducted in 1994 (N = 5,900) (see: CIS2088 1994), to which we have applied logistic regression techniques. Fixed-Term Employment and Trade Union Membership Probably because collective agreements in Spain apply by law to unionized and nonunionized workers alike, and because unions have weak presence and therefore rather limited recruitment capacity in most of the Spanish firms (which are small), the union membership rate is very low. According to the CSRSTUA, the affiliation rate in Spain is 15 percent. If we distinguish by type of contract, though, we see that the membership rate for workers on permanent contracts is 19.2 percent and that 80 percent of all union members have a permanent contract. In contrast, the unionization rate for fixed-term workers is only 8 percent, and only 20 percent of all members are fixed-term workers. However, is this highly significant difference in fact attributable to type of contract or is it the consequence of other possible explanatory variables hidden in the bivariate comparison? In order to answer this question, we have undertaken a logistic regression on the probability of being a union member in Spain, the results of which are presented in the first column of Table 7.4. In our regression model we have introduced as explanatory variables of union membership firm-level factors, individual-level subjective factors, and individual-level “objective” factors. With respect to firm-level factors, the model shows that both the size of the company (measured as number of workers) and the type of ownership (public or private) affect the chances of being a union member. Being in the public sector and in large firms increase the chances of being a union member.14 Membership also depends on ideological factors. Those with a more left-wing outlook and a more 14

The model also controls for industry, yet this variable does not seem to be significant. It is not shown in the table.

230

Unions and Working-Class Fragmentation in Spain

Table 7.4. Logistic Regressions on Different Forms of Collective Action

Models, Explanatory Variables

Probability of Being a Union Member

Probability of Voting in Union Elections (if Convoked)

Probability of Taking Part in Organized Strike or Stoppage

Odds Ratio

Sig.

Odds Ratio

Sig.

Odds Ratio

Sig.

Age Female Class (service) Intermediate Skilled manual Unskilled

1.00 0.89

n.s. n.s.

1.04 1.50

** n.s.

1.00 0.56

n.s. ***

0.45 2.54 2.31

n.s. *** **

3.60 3.95 2.47

** ** **

0.82 0.79 0.45

n.s. n.s. **

Public sector Size of firm (less than 50) From 50–500 workers More than 500 workers

2.25

***

0.86

n.s.

1.59

***

1.90 2.45

** ***

1.03 1.27

n.s. n.s.

1.14 1.79

n.s. ***

Ideology (L-R scale) Evaluation of unions

0.84 1.07

*** ***

0.83 1.05

** * (0.09)

0.79 1.05

*** ***

Employment contract (permanent) Fixed-term

0.57

***

0.47

**

0.72

**

Models’ characteristics

LOGIT ESTIMATES N = 1,037 c2(20) = 161.77 Prob > c2 = 0.0000 Pseudo R2 = 0.1513 Log Likelihood =

LOGIT ESTIMATES N = 342 c2(20) = 40.45 Prob > c2 = 0.0044 Pseudo R2 = 0.1109 Log Likelihood =

LOGIT ESTIMATES N = 1,031 c2(20) = 165.64 Prob > c2 = 0.0000 Pseudo R2 = 0.1160 Log Likelihood =

-453.7134

-162.17917

-631.07653

CLASSIFICATION (Positive p ≥ 0.2) Sensitivity 70.78% Specificity 68.22%

CLASSIFICATION (Positive p ≥ 0.7) Sensitivity 80.75% Specificity 53.25%

CLASSIFICATION (Positive p ≥ 0.5) Sensitivity 68.79% Specificity 65.12%

GOODNESS-OF-FIT P > c2 = 0.3010

GOODNESS-OF-FIT Hosmer-Lemeshow Prob > c2 = 0.2733

GOODNESS-OF-FIT Prob > c2 = 0.2638

Source: CIS 2880 (1994). Subsamples of employed population. (Calculated by the authors.) Notes: Controlling for Industry. n.s. = not significant. *** significance £ 0.001. ** significance £ 0.05. * significance £ 0.09 (significance level in parentheses). The model on union voting presented here comes from a previous model that did not pass the goodness-of-fit test. Analysis of residuals was undertaken and a problematic residual of value -4 was detected and deleted. Once this case is deleted the model achieves a satisfactory fit. Only the level of significance of the union evaluation scale is slightly affected by the elimination of the problematic case, coming a little closer to the 0.05 significance level in the second model presented above. Predicted values have been calculated for both models, and the results are almost totally identical.

231

Javier G. Polavieja and Andrew Richards

positive view of the unions are all the more likely to be union members. In our analysis, ideology has been measured using a ten-point left-right self-placement scale. Union evaluation has been measured using a computed scale that combines different opinions regarding the extent to which unions defend different categories of workers15 with different opinions regarding the characteristics of trade unions as organizations and of their leaders.16 The scale of evaluation of trade unions (of approximately 160 intervals) goes from -10 (most negative evaluation of unions) to +10 (most positive evaluation of unions). It has a Cronbach’s Alpha of 0.85. Finally, the model shows that union membership also depends on individual-level “objective” factors. In a previous model (not presented here) in which type of contract was not introduced, two of these factors were age and gender. The average age of union members in Spain is fortyone years, and the proportion of women is 24.3 percent. However, when controlling for type of contract, the effects of both age and gender disappear (Table 7.4). Type of contract, together with class, are the individuallevel “objective” factors that really matter. Indeed, being a manual worker, as compared with a service-class professional, increases significantly the chances of becoming a union member. Crucially, for the validation of our hypothesis, having a fixed-term contract also significantly reduces the chances of becoming a union member. Thus, even after controlling for firm-level, ideological, and individual subjective and objective factors, having a fixed-term contract hinders union affiliation in Spain (Table 7.4).

15

16

The CSRSTUA contains eight likert-type questions which ask respondents to express the extent to which unions defend different categories of workers (women, young workers, unemployed, permanent workers, fixed-term workers, pensioners, highly paid workers, and poorly paid workers). All answers to these questions were added up in a -16 to +16 index. This index was then divided by 8 to produce a -2 to +2 scale, which was later used as the basis for our union evaluation scale (see note 16). We posed three likert-type questions on the degree of agreement with the following: “If unions disappeared, nothing would really happen”; “Most unions are run by leaders who are detached from society”; “Unions are absolutely necessary to protect the interests of workers.” Finally, respondents were asked to choose between the following: “Generally speaking, Spanish unions are modern organizations which are adapted to today’s society”; “Generally speaking, Spanish unions are old organizations and they should change to adapt to today’s society.” Pro-union responses to these questions were coded either +2 or +1 (depending on the intensity of the response), while antiunion responses were coded -2 or -1. Those who did not answer or answered that they did not know were given a 0. These answers were then added up to the -2 to +2 scale previously obtained from the eight likerttype questions on the extent to which unions defend different categories of workers (see note 15).

232

Unions and Working-Class Fragmentation in Spain

Fixed-Term Work and Union Voting However, in a context in which 85 percent of the workforce is not unionized, the relevance of the analysis of the effects of labor market dualization on affiliation can be rightly questioned. In voter unionism of the type existing in Spain, it is voting behavior rather than affiliation which matters (Martín Valverde 1991: 24–25; Martínez Lucio 1993: 500–501). According to the CSRSTUA, the percentage of workers who voted in the last union elections convoked at their workplaces was 76.8 percent. This figure rises to 79.6 percent in the case of workers on permanent contracts and drops to 61.3 percent for fixedterm workers. In other words, almost 40 percent of fixed-term workers did not vote in the last union elections held at their firms. Again, we have applied logistic regression techniques to further investigate this figure. The second column of Table 7.4 presents the results of a logistic regression model on the probability of voting in union elections. This time, firm-level factors are not significant. Although firm-level factors are of fundamental importance in determining whether elections are convoked or not, once elections take place they do not seem to affect voting behavior. Therefore only individual factors play a role in explaining the decision to vote or abstain in union elections. Age, class, ideology, and type of contract are such factors. The older individual, one who is more left wing, employed as a manual worker and, critically, on a permanent contract, is the one the more likely to vote in a union election when such an election takes place. The model also shows a strong, although not totally significant, correlation between favorable evaluations of unions and union voting.17 In any case, after controlling for other explanatory variables, we find that having a fixed-term contract also reduces the chances of voting in trade union elections. Fixed-Term Employment and Participation in Strikes and Stoppages in the Workplace We have hypothesized that the labor precariousness of fixedterm workers places them in a rather weak position in relation to their employer. Since fixed-term workers are easily – that is, cheaply – replaceable, personal involvement in any form of collective action can be sanctioned rapidly via the termination of the contract. This could also explain why workers who evaluate unions in a positive manner nevertheless show significantly lower rates of union voting and union membership. To further explore this hypothesis, we have examined how type of contract affects the chances of taking part in strikes or workplace stoppages. 17

At a 0.09 level of significance.

233

Javier G. Polavieja and Andrew Richards

The CSRSTUA includes a question on whether respondents have ever taken part in organized strikes or stoppages in their workplaces. Some 45.5 percent of all active respondents report that they have participated in such forms of collective action. This rate increases to 53.3 percent in the case of workers on permanent contracts and decreases to 35.4 percent in the case of fixed-term workers. Again, we have applied logistic regression techniques to investigate this difference further. The results of this logistic regression model are presented in the final column of Table 7.4. Our logistic model shows that the probability of taking part in organized strikes or stoppages again depends on firm-level, ideological, and individual-level “objective” factors. With respect to firm-level factors, the size and the sector of activity of the firm seem to enhance the chances of participation in strikes and stoppages. Workers in the public sector are clearly more likely to participate in strikes and stoppages in the workplace than those employed in the private sector. Equally, workers in large firms (having more than 500 workers) are more likely to take part in strikes and stoppages at their firms than workers in medium and small firms. Ideological variables also play a key role in explaining why some individuals participate and others do not. The more left wing and the more positive about trade unions, the more likely an individual is to take part in organized strikes and stoppages in the workplace. Finally, three individual objective factors also explain participation in organized collective action in the workplace: gender, class, and, again, type of contract. Indeed, our model shows that women, unskilled manual workers, and workers on fixed-term contracts are significantly less likely to take part in organized strikes or stoppages. All the regression models presented in Table 7.4 tend, therefore, to substantiate our thesis. Having a fixed-term contract not only significantly reduces the chances of affiliating to an union, but also reduces the chances of voting in union elections or taking part in organized strikes or workplace stoppages. As expected, the model suggests that a precarious working situation – even for workers sympathetic to the unions – hinders pro-union activity (see predicted values in Table 7.5). Subhypothesis 2: A Precarious Working Situation Diminishes Subjective Involvement with the Unions We have also hypothesized, however, that the labor market precariousness attached to type of contract could reduce subjective involvement in and of itself, eroding workers’ sentiments of collective identity with respect to the unions, and generating instead 234

Unions and Working-Class Fragmentation in Spain

Table 7.5. Average Predicted Probabilities by Class and Type of Contract

Union Affiliation Rate

Union Vote Rate

Strikes and Stoppages Rate

Classes

Permanent Contract

FixedTerm Contract

Permanent Contract

FixedTerm Contract

Permanent Contract

FixedTerm Contract

Service Intermediate Skilled manual Unskilled

0.21 0.23 0.32 0.29

0.10 0.09 0.12 0.13

0.67 0.86 0.86 0.84

0.49 0.64 0.65 0.65

0.65 0.55 0.61 0.46

0.54 0.39 0.46 0.31

Source: CIS 2880 (1994). Subsamples of employed population. Calculated by the authors from the models in Table 7.4.

those of distance, apathy, and even rejection. Workers in the flexible segment of the labor market would thus become increasingly detached from the trade unions. The Spanish Survey on Class Structure, Class Consciousness, and Class Biography (CSCCCB 1991) (Carabaña et al. 1993) asks respondents to place themselves on a 10-point scale according to their degree of identification with trade unions. The identification scale ranges from 0 (“nothing at all”) to 10 (“very much”). Of all other possible indicators found in other surveys, this simple scale seems to be the closest operationalization of our concept of subjective involvement with trade unions. We have undertaken a robust regression on this scale, the results of which appear in Table 7.6. In our regression model we have considered that union identification depends on two structural variables (occupational class and labor market status) and two ideological variables (subjective identification with political parties and ideology). Sex and age have also been introduced in the model, mainly as control variables. Class has been coded according to the same condensed version of Goldthorpe’s class schema used in the previous analyses discussed in this chapter. The labor-market-situation variable has been coded with four different values. If respondents are employed on permanent contracts, they are coded as value 0, which is the reference category in the model; if they are employed on fixed-term contracts, they are coded as value 1; if they are unemployed owing to the termination of their fixed-term contracts, they are coded 2; and if they are unemployed owing to the termination of permanent contracts, they are assigned value 3. Therefore, value 0 corresponds to insiders in the Spanish labor market, 235

Javier G. Polavieja and Andrew Richards

Table 7.6. Robust Regressions on Index of Identification with Unions

Explanatory Variables

Coefficient

P > |t|

Class (service) Intermediate Skilled manual Unskilled

-0.05 0.44 0.43

n.s. ** **

Female Age

-0.05 -0.01

* (0.07) n.s.

Ideology (L-R scale) Identification with parties

-0.25 0.57

*** ***

Labor market situation (insider) Fixed-term Unemployed from fixed-term Unemployed from permanent

-0.50 -0.94 -0.58

*** *** ***

Constant

Model’s Characteristics N. = 2,148 R2 = 0.3896 Adj R2 = 0.3868 Root MSE = 2.31523

2.60

Source: CSCCCB (1991). Calculated by the authors. *** significance £ 0.001. ** significance £ 0.05. * significance £ 0.07 (significance level in parentheses).

and values 1, 2, and 3 correspond to different forms of being an outsider. Finally, ideology has been measured in the usual left-to-right self-assessed 10-point scale used in previous models.18 Our regression analysis leads us to the following findings. First, subjective identification with unions seems to depend on occupational class, being significantly higher among manual workers than among professional and white-collar workers. Second, union involvement also depends on ideological factors. General ideological maps affect the level of subjective identification with trade unions. As expected, workers who define themselves as 18

We have included a measure of political identification (in this case, with parties) on the assumption that both identification with unions and with parties could form part of a common pattern of political involvement (Polavieja 1999). Political identification is measured by the self-assessed degree of identification with political parties on a scale from 1 to 10. Respondents do not have to reveal their party preferences, only the degree to which they identify with a political party. The scale therefore measures identification with parties as political institutions. The model confirms our expectation that a relationship exists between party identification and union involvement – that is, the stronger the identification with a political party, the stronger the identification with unions.

236

Unions and Working-Class Fragmentation in Spain

left-wing supporters are more likely to show higher levels of identification with trade unions than workers on the political right. Finally, the model also confirms our hypothesis that, independently of ideological and structural factors, being an outsider in the labor market diminishes in and of itself workers’ subjective involvement with unions. Employed fixed-term workers, unemployed workers coming from fixed-term employment, and unemployed workers coming from permanent contracts, all show significantly lower levels of identification with unions than their permanently employed counterparts. Among outsider workers, the highest erosion of identification seems to occur among unemployed workers coming from fixed-term work. Yet employed fixed-term workers and unemployed workers coming from stable employment also show significantly lower levels of union identification. All workers in the flexible segment of the Spanish labor market identify less with trade unions than their insider counterparts.

Summary of Findings From our study of the CSRSTUA and the CSCCCB survey data, we conclude that labor market precariousness has clear effects on involvement with the unions. Those workers who are unemployed or who are on temporary contracts demonstrate significantly lower levels of subjective involvement and pro-union behavior than workers on permanent contracts. On the one hand, having a temporary contract reduces the participation of the worker in union activities. This is so in all occupational categories, in the public and private sector, and among both workers who are hostile and those who are sympathetic toward the unions. A precarious working situation in the labor market therefore acts as an “objective” factor impeding collective action among workers. The weakness of their position in the internal and external labor markets ensures that workers in the precarious segment of the labor market – the outsiders – are related much more tenuously to union activities. On the other hand, the dualization of the labor market also has direct consequences for the evaluation of the unions. A precarious working situation reduces significantly the possibilities of subjective identification with the unions. Outsiders, because of their situation in the labor market, feel less ideological affinity for the unions than workers in stable situations. This is so independently of general ideological maps. We argue, therefore, that the process of segmentation of the labor market has had empirically verifiable repercussions on levels of 237

Javier G. Polavieja and Andrew Richards

involvement with the unions. Trade unionism without affiliation, as is the case in Spain, draws its strength from its capacity to mobilize workers – that is, from its capacity to present itself before workers as the legitimate representative of their interests. As a consequence, the decrease in workers’ involvement with the unions represents a debilitating trend of very great significance. If a growing segment of the Spanish workforce is no longer mobilized by the unions, either because it cannot be, or no longer feels, represented, then the unions will have lost an important source of their power.

Conclusions Given the increasing distance between the trade unions and a growing section of the Spanish workforce, the question arises as to whether and how such a distance might be closed. This question, in turn, raises a number of issues which are undoubtedly in need of further research. First, largely structural factors have tended to drive the Spanish trade unions back to the defense of their core constituency within the insider workforce. Yet could the unions, despite the circumstances, have done more to retain widespread support among a changing workforce? As Estivill and de la Hoz note, “internal strains and incompetence” have also contributed to the unions’ problems. Since the advent of democracy, the unions’ analysis of structural changes in the Spanish economy, and of the consequences of economic recession and recovery, has often been “tardy and partial.” It has been quite evident since as far back as the early 1980s that the unions were unprepared to either organize or retain the great influx of new members which they had gained in the late 1970s. Moreover, a key feature of Spanish trade unionism in the 1980s was the often bitter rivalry between the two principal union confederations, the UGT and the CC.OO. Sharp differences existed between the two organizations in their diagnosis of the economic situation and the appropriate tactics and strategy to be adopted, especially in the face of industrial and economic restructuring and with respect to cooperation with government and employers. Indeed, the CC.OO. did not sign several of the concertation agreements of the early 1980s, adopting a much more ideologically charged position on the issues of employment and wages than the UGT. Estivill and de la Hoz argue that interconfederation differences over policies, tactics, and strategy have presented a “disorienting image” to working people, while recurrent hostility between the UGT and the CC.OO. – 238

Unions and Working-Class Fragmentation in Spain

especially at the factory or company level – has alienated many workers (1990: 287–89). Given that it is precisely in the workplace where the unions, in the final analysis, have a direct presence in the daily working lives of employees, the potentially alienating effects of union rivalry and division merit further research. Nonetheless, it is doubtful that greater unity will, in and of itself, narrow the gap between the unions and the workforce as a whole. The UGT and the CC.OO. have, in any case, drawn much closer together in the 1990s, especially with the deterioration of the former’s relationship with the PSOE. Nor has this trend been reversed since the election of the Partido Popular in May 1996. Both union organizations have continued to call for a bolder, broader, and more interventionist range of measures with which to tackle unemployment, and both have stressed the importance of defending and promoting greater job security. These commitments to reducing unemployment and increasing job security are, of course, not inherently “unsolidaristic.” Yet the unions continue to face the predicament discussed in this chapter. Given the realities of the Spanish labor market – high unemployment, high levels of temporary work, and entrenched segmentation between insiders and ousiders – they have struggled to forge a unifying strategy and at the same time evade the charge that in practice they are defending a relatively privileged group of insiders. This continuing predicament was highlighted in April 1997 when the UGT and the CC.OO., along with the employers’ organizations CEOE and CEPYME, signed a new pact on further labor market reform that was hailed by El País as “probably the most important social agreement signed in Spain over the past 15 years” (EIRR 1997: 28). Containing measures to both diminish the protective apparatus surrounding permanent workers and to promote the use of a new type of open-ended employment contract, the agreement sought to ease labor market rigidities and, more indirectly, reduce levels of both unemployment and temporary work by making employers more confident about employing workers on a permanent basis (EIRR 1997: 24). Not surprisingly, the negotations that led to such an agreement were protracted, not least because the employers’ demands (especially with respect to the lowering of dismissal costs) went to the very core of the unions’ dilemma. On the one hand, the unions have come to recognize (albeit very reluctantly) that the easing of so-called rigidities in the labor market may be the only way of reducing, in the long run, levels of 239

Javier G. Polavieja and Andrew Richards

unemployment and temporary work. On the other hand, they have remained nervous about making it easier for employers to fire workers. It is a measure of the pressure that the unions encountered – from employers, from the threat of a politically hostile government to intervene directly, and, indeed, from sections of the workforce itself – that the UGT and CC.OO. signed the agreement. Yet even then, there was significant dissent from within the labor movement as a whole. The respective leaderships of the UGT and the CC.OO. defended the agreement as a means of promoting, above all else, stable employment. In addition, in the light of their own compromises, they made renewed calls for the government to develop a bolder policy on employment. The general secretary of the CC.OO., Antonio Gutiérrez, declared that “there can be no more excuses against unemployment and employment insecurity. It is now the turn of those in charge of economic policy to boost job creation” (EIRR 1997: 28).19 In contrast, dissenting elements within the UGT and CC.OO., and rival minority and regional trade union confederations, argued that the deal would neither reduce unemployment nor boost employment security.20 Only time will tell if this latest agreement on labor market reform will indeed contribute to lowering unemployment, reducing the levels of temporary work, and boosting job security.21 If it does not, the agreement will 19

20

21

See also El País, May 1, 1997; interview with Cándido Méndez, General Secretary of the UGT, El País, April 23, 1997, and Méndez’s article “Una apuesta por la solidaridad,” in El País, May 1, 1997. See EIRR 1997: 12, 28; El País, May 2, 1997. See also the CGT’s undated pamphlet, La Nueva Reforma Laboral. The evidence so far is mixed. In December 1997, Labor Minister Javier Arenas announced that 480,883 open-ended contracts had been signed in the six months after the reform went into force – an increase of 201 percent over the comparable period in 1996. Arenas predicted that 700,000 new open-ended contracts would be signed in 1997 as a whole (EIRR 1998 [288]: 11). In May 1998, one year after the reform came into force, the Ministry of Labor published preliminary findings showing that almost one million openended jobs had been created, with 650,000 of these directly attributable to the reforms (EIRR 1998 [293]: 11–12). In January 1998, however, the main employers’ organization, the CEOE, flatly contradicted the government’s claims, arguing that only 2.5 percent of agreements negotiated in 1997 contained concrete commitments to convert temporary contracts into open-ended ones. Temporary contracts were still the most popular form of contract for employers, with 69 percent of agreements at the sectoral level providing for temporary forms of recruitment, and only 8 percent of agreements providing for specific job-security guarantees (EIRR 1998 [289]: 11). In addition, the number of temporary contracts signed in the six months following the labor market reform also increased, by 28 percent (EIRR 1998 [288]: 11). For these reasons, the unions have remained very concerned about continuing high levels of temporary employment and possible abuses of temporary contracts.

240

Unions and Working-Class Fragmentation in Spain

only have served to underscore, rather than help resolve, the problems of contemporary Spanish trade unionism. References Alba, A. 1991. Fixed-Term Employment Contracts in Spain: Labour Market Flexibility or Segmentation? Working Paper 91–92. Universidad Carlos III, Madrid. ———. 1997. How Temporary Is Temporary Employment in Spain?” Universidad Carlos III de Madrid, Working Paper 97–14. Albarracín, J. 1990. La política económica y el empleo. Paper presented at II Jornadas de Economía Critica, Madrid. Antolín, P. 1995. Movilidad Laboral, Flujos de Desempleo, Vacantes y Comportamiento en la Búsqueda de Empleo en el Mercado de Trabajo Español. Moneda y Crédito 201: 49–85. Argentaria. 1993. Información Estadística sobre Desigualdad Social. I Simposio sobre Igualdad y Distribución de la Renta y la Riqueza. Madrid: Fundación Argentaria. ———. 1995. La Desigualdad en España: Síntesis Estadística. II Simposio sobre Igualdad y Distribución de la Renta y la Riqueza. Madrid: Fundación Argentaria. Aronowitz, S. 1983. Working Class Hero. A New Strategy for Labor. New York: Pilgrim Press. Astudillo, J. 1998. Los Recursos del Socialismo: Las Cambiantes Relaciones entre el PSOE y la UGT (1982–1993). Madrid: Instituto Juan March de Estudios e Investigaciones. Bentolila, S., and J. J. Dolado. 1994. Labour Flexibility and Wages: Lessons from Spain. Economic Policy 18: 54–99. Bentolila, S., J. Segura, and L. Toharia. 1991. La contratación temporal en España. Moneda y Crédito 193: 225–65. Boreham, P., and R. Hall. 1994. Trade Union Strategy in Contemporary Capitalism: The Microeconomic and Macroeconomic Implications of Political Unionism. Economic and Industrial Democracy 15: 313–53. Braverman, H. 1974. Labor and Monopoly Capital. New York: Monthly Review Press. Carabaña, J., et al. 1993. Encuesta Estructura, Conciencia y Biografía de Clase: Informe Técnico. Madrid: IESA/CSIC. Coggins, J., et al. 1989. Trade Unions of the World 1989–1990. 2nd ed. Chicago: St. James Press. Crouch, C. 1986. The Future Prospects for Trade Unions in Western Europe. Political Quarterly 57: 5–17. European Industrial Relations Review (EIRR). 1997, 1998. Various issues. Erikson, R., and J. H. Goldthorpe. 1992. The Constant Flux. Oxford: Clarendon Press. Esping-Andersen, G. 1998a. The Effects of Regulation on Unemployment Levels and Structure. The Evidence from Comparative Research. Manuscript, Dipartamento di Sociologia e Ricerca Sociale, Universita di Trento. ———. 1998b. Conclusions and Policy Implications. Manuscript, Dipartamento di Sociologia e Ricerca Sociale, Universita di Trento. 241

Javier G. Polavieja and Andrew Richards Estivill, J., and J. M. de la Hoz. 1990. Transition and Crisis: The Complexity of Spanish Industrial Relations. In G. Baglioni and C. Crouch, eds., European Industrial Relations: The Challenge of Flexibility. London: Sage. Fernández, F., L. Garrido, and L. Toharia. 1991. Empleo y paro en España. 1975–1999. In F. Miguélez and C. Prieto, eds., Las Relaciones Laborales en España. Madrid: Siglo XXI. Fulcher, J. 1991. Labour Movements, Employers, and the State. Oxford: Clarendon Press. Giddens, A. 1987. Social Theory and Modern Sociology. Cambridge: Polity Press. Gorz, A. 1991. The New Agenda. In Robin Blackburn, ed., After the Fall. The Failure of Communism and the Future of Socialism. London: Verso. Green, F. 1992. Recent Trends in British Trade Union Density: How Much of a Compositional Effect? British Journal of Industrial Relations 30: 445–58. Güell-Rotllan, M., and B. Petrongolo. 1998. The Transition of Workers from Temporary to Permanent Employment: The Spanish Case. Working Paper 98–81. Universidad Carlos III, Madrid. Hall, P. 1987. European Labor in the 1980s. International Journal of Political Economy 17: 3–25. Heery, E., and J. Kelly. 1995. Conservative Radicalism and Nostalgia: A Reply to Paul Smith and Peter Ackers. Work, Employment and Society 9: 155–64. Hobsbawm, E. 1978. The Forward March of Labour Halted? Marxism Today 22: 279–86. ———. 1989. Farewell to the Classic Labour Movement? New Left Review 173: 69–74. Hyman, R. 1985. Class Struggle and the Trade Union Movement. In David Coates, Gordon Johnston, and Ray Bush, eds., A Socialist Anatomy of Britain. Cambridge: Polity Press. ———. 1992. Trade Unions and the Disaggregation of the Working Class. In Marino Regini, ed., The Future of Labour Movements. London: Sage. Iriso, P. L. 1993. Sistemas de Negociación Colectiva y Acción Sindical. Sindicatos y Trabajadores en la Empresa. Madrid: Instituto Juan March de Estudios e Investigaciones. Jimeno, J., and L. Toharia. 1994. Unemployment and Labour Market Flexibility: Spain. Geneva: International Labour Office. Jordana, J. 1996. Reconsidering Union Membership in Spain, 1977–1994: Halting Decline in a Context of Democratic Consolidation. Industrial Relations Journal 27: 211–24. Kelly, J. 1988. Trade Unions and Socialist Politics. London: Verso. Kern, H., and C. F. Sabel. 1991. Trade Unions and Decentralized Production: A Sketch of Strategic Problems in the West German Labor Movement. Politics and Society 19: 373–402. Lane, C. 1989. From “Welfare Capitalism” to “Market Capitalism”: A Comparative Review of Trends towards Employment Flexibility in the Labour Markets of Three Major European Societies. Sociology 23: 583–610. Lash, S., and J. Urry. 1987. The End of Organized Capitalism. Cambridge: Polity Press. 242

Unions and Working-Class Fragmentation in Spain Lockwood, D. 1989. The Blackcoated Worker: A Study in Class Consciousness. 2nd ed. Oxford: Clarendon Press. Marshall, Gordon, and David Rose. 1988. Proletarianization in the British Class Structure? British Journal of Sociology 39: 498–518. Martín Valverde, A. 1991. European Employment and Industrial Relations Glossary: Spain. London and Luxembourg: Sweet and Maxwell/Office for Official Publications of the European Communities. Martínez Lucio, M. 1993. Spain: Constructing Institutions and Actors in a Context of Change. In A. Ferner and R. Hyman, eds., Industrial Relations in Europe. Oxford: Blackwell. Martínez Lucio, M., and P. Blyton. 1995. Constructing the Post-Fordist State? The Politics of Labour Market Flexibility in Spain. West European Politics 18: 340–60. Organisation for Economic Co-operation and Development (OECD). 1999. Employment Outlook. Paris: OECD. Pérez-Díaz, V. 1987. Unions’ Uncertainties and Workers’ Ambivalence. The Various Crises of Trade Union Representation and Their Moral Dimension. International Journal of Political Economy 17: 108–38. Polavieja, J. G. 1998. The Dualisation of Unemployment Risks. Class and Insider/ Outsider Patterns in the Spanish Labour Market. Estudio/Working Paper 1998/128. Madrid: Instituto Juan March de Estudios e Investigaciones. ———. 1999. How Do Labour Market Experiences Affect Political Attitudes? Estudio Working Paper 1999/142. Madrid: Instituto Juan March de Estudios e Investigaciones. Recio, A. 1991. La segmentación del mercado de trabajo en España. In F. Miguélez and C. Prieto, eds., Las Relaciones Laborales en España. Madrid: Siglo XXI. Rhodes, M. 1997. Spain. In Hugh Compston, ed., The New Politics of Unemployment. London: Routledge. Richards, A. 1995. Down but Not Out: Labour Movements in Late Industrial Societies. Estudio/Working Paper 1995/70. Madrid: Instituto Juan March de Estudios e Investigaciones. ———. 1996. Miners on Strike: Class Solidarity and Division in Britain. Oxford: Berg Publishers. ———. 1999. Spain: From Isolation to Integration. In Ronald Tiersky, ed., Europe Today: National Politics, European Integration, and European Security. New York: Rowman and Littlefield. Rigby, Mike, and Tevesa Lawlor. 1994. Spanish Trade Unions 1986–1994: Life after National Agreements. Industrial Relations Journal 25: 258–71. Rojo Torrecilla, E. 1990. El sindicalismo ante los cambios en el mercado y la composición de la clase trabajadora. Política y Sociedad 5: 31–44. Sabel, C. F. 1987. A Fighting Chance. Structural Change and New Labor Strategies. International Journal of Political Economy 17: 26–56. Salvatore, N. 1992. The Decline of Labor: A Grim Picture, A Few Proposals. Dissent (Winter): 86–92. Segura, J., F. Durán, L. Toharia, and S. Bentolila. 1991. Análisis de la Contratación Temporal en España. Madrid: Ministerio de Trabajo y Seguridad Social. 243

Javier G. Polavieja and Andrew Richards Taylor, A. J. 1993. Trade Unions and the Politics of Social Democratic Renewal. West European Politics 16: 133–55. Toharia, L. 1994. Spain: Modernisation of Unemployment. In D. Gallie, C. Marsh, and C. Vogler, eds., Social Change and the Experience of Unemployment. Oxford: Oxford University Press. Toharia, L., and J. Muro. 1988. ¿Es elevado el salario de los jóvenes? Revista de Economía y Sociología del Trabajo 1/2. Touraine, A. 1986. Unionism as a Social Movement. In S. M. Lipset, ed., Unions in Transition: Entering the Second Century. San Francisco: Institute for Contemporary Studies. van der Meer, M. 1995. Trade Union Membership in Spain. Manuscript. Madrid: Instituto Juan March de Estudios e Investigaciones. Winchester, D. 1989. Sectoral Change and Trade-Union Organization. In Duncan Gallie, ed., Employment in Britain. New York: Basil Blackwell.

244

8 Modell Nederland SOCIAL PARTNERSHIP AND COMPETITIVE CORPORATISM IN THE NETHERLANDS

Steven B. Wolinetz

In the mid-1990s officials from labor ministries, trade unions, and employers associations began flocking to the Netherlands.1 The object of their visits was something called the polder model. This referred not to dikes or land reclamation, but rather the modest but steady economic growth (2.9% per annum between 1993 and 1997) and decreasing unemployment which the Dutch were enjoying. Although up to 25% of the adult population remained inactive, unemployment had declined from 10% in 1982 to less than 4.1% in 1998. Reduced unemployment reflected early retirements, reductions in hours of work, and particularly in the 1990s, the creation of full and part-time jobs. Job creation was spurred by wage moderation, new investment, and programs which targeted youth, unskilled or semiliterate workers, immigrants, and refugees. Although many of the new jobs were part-time positions held disproportionately by women, the Dutch record was better than either the EU average or that of neighboring countries. As Table 8.1 demonstrates, growth in real GDP averaged 2.9% in the Netherlands between 1993 and 1997, but only 2.3% in the EU. In the same period, unemployment in the Netherlands dropped from 6.3% to 5.2% and was expected to be 4.2% in 1999. In contrast, unemployment rates hovered around 8–9% in Germany, 11–12% in France, and 10–11% in the EU (OECD 1998a, 1998b). 1

This research is based on interviews conducted from 1984 through 1999 under grants from the Social Sciences and Humanities Research Council of Canada and Memorial University of Newfoundland.

245

Steven B. Wolinetz

Table 8.1. Economic Growth and Unemployment in the Netherlands and Neighboring Countries

The term “model” has typically been reserved for countries like Sweden or Germany. In the former, the ability of the trade unions and the Swedish Social Democrats to define the terms of the postwar bargain and impose their stamp on the welfare state encouraged academics and journalists to talk about a Swedish model. In the latter, economic success and industrial prowess led to a similar designation. Modell Deutschland highlighted the benefits which Germany derived from fiscal rectitude and strict monetary policy during the 1970s (Markovits 1982). The labels attached to the Netherlands in the same period are less flattering. “Dutch disease” referred to ways in which the Dutch were said to have squandered the rents from natural gas on the expansion of the welfare state in the 1970s and more generally to a sick economy. That description fit the Nether246

Modell Nederland

lands in the late 1970s and early 1980s (OECD 1998b). Unemployment rates of 1–2% in the early 1970s shot up to 10% at the end of the decade. However, the figures mentioned in public at the time were much higher – 15–20% of the population.2 Government expenditures appeared to be out of control as well. Public sector wages and benefits were indexed to increases in average private sector wages, giving the government little control over its own expenditures. Increasing deficits and the fact that the government was absorbing more than 60% of the national income in taxes and transfers raised alarm. Critics claimed that the public sector was crowding out private investment. The Dutch economy began to turn around in the 1980s. Change flowed from diverse sources. Recognition that there was a problem was the first step. Although the employment in manufacturing had been declining throughout the 1970s, few were aware of the magnitude of the problem.3 This changed at the end of the decade. The publication of a Scientific Council for Government Policy report, The Place and Future of Dutch Industry (WRR 1980), galvanized discussion by highlighting the decline and calling for a targeted industrial policy. A subsequent commission headed by Gerrit Wagner, a former director of Royal Dutch Shell, called instead for “a new industrial elan” (Advisory Commission on Industrial Policy 1981). The sense of crisis was driven home by the Rijn Schelde Verolme Affair: millions of guilders had been poured into the reorganization of the shipbuilding industry, which was nevertheless failing. In 1982, a center-right cabinet intent on reducing deficits and controlling inflation was formed. The Lubbers cabinet announced plans to freeze wages. Just before the cabinet was to assume office, trade union federations and employers associations struck a deal. Trade unions conceded the elimination of automatic wage indexation if employers would agree to discuss work-time reductions and the creation of part-time jobs. The Wassenaar Accord (after the town where it was negotiated) was announced as the new cabinet took power. 2

3

Figures used in this text are based on ILO/OECD definitions. Standardized unemployment rates are based on the number of workers registered at labor exchanges and seeking more than twenty hours of work per week. Socialist trade unions and the Labor Party focused on the “redistribution of wealth, knowledge, and power” throughout the seventies. Meanwhile, Dutch industry was shedding labor. Industrial employment declined from 1,199,000 persons in 1968 to 980,000 by 1978 (WRR 1980).

247

Steven B. Wolinetz

Changes soon became apparent. Expenditures were cut and deficits trimmed but not eliminated. Links between private sector wages and public sector salaries and benefits were severed and public sector salaries were reduced by 3%. Austerity policies, continued by subsequent cabinets, boosted business confidence and encouraged investment. Unemployment declined, and the Dutch recorded slow but steady growth through the 1980s and 1990s. Decreasing unemployment initially reflected early retirements and reductions in the hours of work supplemented and eventually supplanted by the creation of new part- and full-time jobs.4 By the late 1990s, the Dutch were in a congratulatory mood. Although 25 percent of the population remained inactive, unemployment rates were among the lowest on the Continent.5 Commentators talked about a Dutch miracle, and the term “polder model” became part of their discourse. The shift lured visitors from different parts of the world. Socioeconomic tourism reflected not only the economic turnaround, but also the role which trade unions and employers associations were said to have played in making it happen (Visser and Hemerijck 1997). Sorting out what did or did not happen is complex. The Dutch have been generating jobs and putting people to work. Wage moderation has been a central element (Stekelenburg 1997a; Visser and Hemerijck 1997; OECD 1998b). Social partnership has also been important. Although cooperation was barely visible in the 1980s, officials of trade union federations and employers associations met regularly. Informal contacts, fostered by the Ministry of Social Affairs and Employment, resulted in accords on education and training and long-term unemployment in 1984 and 1986 as well as a pact on environment in 1989. These agreements were different from the central accords which trade unions and employers attempted to conclude in the 1970s. Narrow in scope, the new style accords were advisory and not binding. Nor was the government a party to the agreements. As in 1982, peak associations recommended that bargaining parties (sectoral trade unions and sectoral business or employers associations) include relevant clauses in the collective agreements (CAOs) which they were negotiating. Many did. 4

5

Early exit was frequently used to open positions for others. By the mid-1980s, there were few manual workers over the age of fifty-eight. In 1998, 10% of the working-age population was receiving long-term disability payments. Broad unemployment, defined by the OECD to include subsidized employment, early retirement, and working-age people on social assistance or disability, plus those actively seeking work, totaled 25% of the working-age population (OECD 1998b).

248

Modell Nederland

Cooperation and contacts intensified in the 1990s. Attempts to construct a broader social partnership began in 1989 but only bore fruit in 1992 and 1993. The formation of a center-left government in 1989 provided the impetus for a broader but ultimately unsuccessful agreement. Social partners and the government unveiled a detailed accord, the Common Policy Framework (GBK) when the Lubbers-Kok cabinet assumed office. Unlike earlier agreements, the Common Policy Framework included the government. Trade unions and employers associations agreed to take each others’ demands and perspectives into account when formulating their own positions and, as they had been doing, to continue to be responsible for wage formation. The government agreed to restore links between private sector wages and the minimum wage if it thought this was possible and to stabilize the collective tax burden if government finances permitted it (STAR 1989). The Common Policy Framework lasted only two years. Employers organizations suspended their participation in 1991. Interim budgetary projections and the government’s failure to follow the advice of the Social and Economic Council (SER) convinced them that the government was not keeping its part of the bargain. Stressing that they were not rejecting consultation, the employers sent a letter to the government indicating that they would be “marching in place” – that is, neither withdrawing nor participating in bipartite and tripartite consultations – until further notice. The employers made it clear that their dispute was with the government rather than the trade unions (RCO 1991). No consultation took place that fall. Social partners came together in 1992 and 1993. The Maastricht Accord provided a rationale for the employers to rejoin central discussions. Responding to a suggestion made by the employers, the government asked the SER for advice on the consequences of European Monetary Union (EMU) and the impact of a consultative economy. Anxious to demonstrate that they could act quickly, SER staff and social partners fast-tracked their discussions and produced a report on the necessity of cooperation and consultation to prepare for EMU (SER 1992). The September 1992 shakeout of the European Monetary System provided impetus for further cooperation. Social partners agreed that there should be a three-month pause in negotiations to “catch their breath” and assess its effects. Unions lowered their demands from 4.5% before the pause to 2.5% afterward. Before the pause, the average wage increase had been 3%. Afterward it was 2%. 249

Steven B. Wolinetz

Social partners concluded a broader basic agreement in November 1993. Its aim was to frame the coming round of collective bargaining. Noting the rapidity of change, A New Course argued that “the economic situation is alarming. Medium-term projections indicate that without structural adjustments, economic growth, and the profitability and competitiveness of firms cannot recover. . . . The Netherlands will be forced to cope with steadily increasing unemployment aggravating already low rates of labor participation” (STAR 1993: 1). Economic conditions in the Netherlands, Europe, and the world are deteriorating. A new perspective is needed and new directions should be pursued immediately: “For the remainder of the century, consultations about conditions of employment will focus on ‘tailor-made’ adjustments and diversity. The needs of firms and employees will lead to greater differentiation and more options within regulations” (STAR 1993: 2). Prevailing economic conditions allowed at best limited or nonexistent room for wage increases in many sectors. Where such room exists, consideration should be given to directing a part of that increase to measures that would contribute to greater participation in the labor market. Social partners also recommended that bargaining parties consider profit-based compensation schemes and the creation of entry-level scales above the minimum wage but lower than those prescribed in collective agreements (STAR 1993: 2–3). Endorsing tailor-made contracts was a major step. Previously, trade unions insisted on standard agreements across whole sectors and uniform benefits guaranteed in law. A New Course not only acknowledged the increasing differentiation of the economy, but also endorsed a growing tendency to draft “framework contracts” which allowed details to be specified by works councils (STAR 1993). Like previous accords, recommendations were advisory but were frequently followed in contract negotiations. Since 1993, social partners have produced a prodigious volume of reports and recommendations on topics narrow and broad. These include the 1996 Note on Flexibility and Security (STAR 1996), which recommended guaranteeing rights and benefits of part-time and temporary employees in exchange for shortening dismissal procedures, and Agenda 2002, laying out a bargaining agenda for ensuing years (STAR 1997). Agenda 2002 builds on and extends A New Course. It begins by noting the success of A New Course. The intent of the former was “to use collective bargaining to help boost economic growth and improve profit margins and the competitive 250

Modell Nederland

position of Dutch trade and industry, leading to a rise in employment” (STAR 1997: 2). That has succeeded but there are still too many longterm unemployed. Although the principles agreed to in 1993 remain valid, it is necessary to restate them in light of the needs of workers and employers and changing economic conditions: “Businesses are finding it increasingly necessary to respond quickly to changes in the market and to technological advances. There is consequently a growing demand for multi-skilled employees and versatile labour market participants” (STAR 1997: 3). Workers are also changing. They “require a labour system which is better able to take their personal circumstances into account – circumstances which may in fact change from one phase of life to the next” (STAR 1997: 3). Employees are also more likely to want a voice in their work. To take account of changing needs, collective bargaining parties should not only continue to allow room for subsector and company-specific clauses in collective agreements, but also consider flexible forms of remuneration. (STAR 1997: 4–6). More career guidance is needed. The report goes on to consider problems of job seekers, ethnic minorities, and others on the fringes of the labor market. Agenda 2002 maintains the earlier framework of cooperation, while reworking and updating the 1993 recommendations. This is an ongoing process. In the spring of 1999, social partners were preparing yet another agreement to replace Agenda 2002.

Preliminary Analysis The activity of Dutch social partners is striking because it is not supposed to be happening. Although the Dutch built up a highly institutionalized variant of corporatism after World War II, Dutch corporatism is widely believed to have broken down in the 1970s, when a radicalized and more militant trade union movement was repeatedly unable to conclude central agreements with increasingly professionalized and more resistant employers associations (Akkermans and Grootings 1978; Nobelen 1987). The evidence we have considered suggests that corporatism in the Netherlands was either not quite as dead as many had believed it to be or had undergone a remarkable recovery. This apparent recovery is poorly explained by the literature on corporatism. The earlier literature debated its merits, wrestled with definitions and prefixes, and attempted to ascertain the circumstances under which 251

Steven B. Wolinetz

cross-class alliances occurred. There was agreement that cooperative relations among trade unions, employers, and government were more likely if certain conditions obtained. Typical correlates included the presence of a strong, centralized trade union movement, governing social democratic parties, full employment, and exposure to the international economy (Cameron 1984). Corporatism was also considered to be a regular feature of the smaller democracies (Katzenstein 1985). Unfortunately, debates about these questions were barely resolved when the study of corporatism was abandoned. Changing economic conditions and the difficulties which cross-class alliances faced in the 1980s convinced many that corporatism had been a temporary phenomenon, no longer needed when high unemployment, technological change, and broadened exit opportunities (increased mobility of capital and plant) provided alternate means to discipline the workforce and reduce labor costs. Developments in a broad range of countries have forced further reassessment of this view. Cooperation between capital and labor exists in Austria, Belgium, parts of Scandinavia, and Italy and has developed in Ireland, where it had no previous roots. As Peter Hall (see Chapter 2 in this volume) argues, organized market economies (OMEs) – substituting elements of coordination for market signals – persist because they enjoy comparative advantage over liberal market economies (LMEs) in certain areas. Martin Rhodes (Chapter 3 in this volume) argues that “competitive corporatism” – cooperation between trade unions and employers associations aimed at maintaining and improving competitiveness – may be a more viable though less egalitarian form of corporatism. Rhodes cites the Netherlands as a prime example. Quotations from both A New Course and Agenda 2002 indicate that competitiveness and profitability are prominent in the rationales which trade union leaders and officials of employers associations use to explain their recommendations. Unfortunately, this does not explain how competitive corporatism operates or why it develops. We can gain a better understanding of contemporary corporatism if we examine the polder model and the process through which corporatism was reinvented in the Netherlands. In doing so, two factors should be borne in mind. First emphases on polarization, decentralization, and deadlock in the 1970s in the Netherlands are correct but incomplete. These fail to take account of the institutional context in which socioeconomic bargaining took place or the goals which actors were pursuing and the choices open to them. Second, literature which views corporatism as a temporary phenomenon driven by the needs of a dominant class or a given economic sit252

Modell Nederland

uation is equally incomplete. Missing in both is an appreciation of the impact of institutions, the ways in which they shape choices, and the difference that this can make. Space does not permit an analysis of the earlier literature. Instead, we concentrate on the Dutch case and consider why the Dutch were able to breath life into a stalled system of socioeconomic bargaining, reinventing it and using it as a mechanism to steer an organized market economy. Before we proceed, a note on the actors involved is in order. Pillarization has been a pervasive although changing facet of Dutch society and politics. In order to simplify discussion, we refer mainly to trade unions and employers associations rather than to specific organizations. On the trade union side, the principal association is the Federation of Dutch Trade Unions (FNV), formed in 1976 by merger of the Netherlands Trade Union Federation (NVV), and the Netherlands Catholic Trade Union Federation (NKV). The FNV has 1.2 million members. Other trade union federations represented in the Foundation of Labor and the SER are the Netherlands Christian Trade Union Federation (CNV), with 342,000 members, and the Middle and Higher Personnel with 152,000 members. The FNV normally takes the lead in formal and informal discussions. On the employers side, the principal organizations were the Federation of Netherlands Industry (VNO) and the Netherlands Christian Employers Association (NCW), which merged in 1995 to form the Confederation of Netherlands Industry and Employers (VNO/NCW). Like the FNV, the VNO took the lead in most discussions.

The Polder Model Visitors trying to find out about the polder model may be told that there is no such thing. This is correct if the term “model” implies a uniform set of policies or formulas, applied over and over and readily duplicated elsewhere. The polder model – or more specifically, what the Dutch have been doing – does not consist of specific policies like the Rehn model in Sweden, but rather institutions, relationships, and practices which facilitate the discovery and implementation of policies favoring job creation, competitiveness, and noninflationary economic growth. The institutions involved – the Foundation of Labor (STAR) and the Social and Economic Council – operate differently than in the past. In the two decades following the war, they provided a device for involving trade unions and employers associations in statutory wage regulation. In the 1980s and 1990s, the STAR 253

Steven B. Wolinetz

and the SER were again involved in wage restraint, but on a voluntary basis, more clearly rooted in exchanges among social partners. The aim in both periods was to ensure competitiveness and profitability in order to expand employment. In the postwar period, the principal device was a low-wage strategy: incomes policies which kept labor costs below those of their principal competitors complemented active industrial policies pursued by the Ministry of Economic Affairs (de Hen 1980; Dercksen 1986; ter Heide 1986). The strategies used in the 1980s and 1990s were more diverse. These ranged from efforts to ensure profitability and a favorable climate for business to policies and programs targeted at specific groups. Wage moderation was central. In contrast to the postwar period, the emphasis was no longer on holding all wages down or maintaining a uniform policy across firms or sectors, but rather on dampening wage increases so that profitability and competitiveness were not compromised. This was accomplished not only by wage moderation but also by encouraging differentiation and flexibility in the organization of work and the development of an educated and flexible workforce. Maintaining competitiveness and profitability is only one way to ensure that jobs are available. Other strategies include reductions in work time, the creation of part-time jobs, and a battery of devices intended either to give employers incentives to take on additional workers or provide jobs for specific categories of people. The former include efforts to reduce the social insurance premiums paid by employers in order to lower the cost of taking on labor. The latter include work-experience jobs, youthguarantee programs, efforts to bring starting wage scales closer to the minimum wage, job pools, and “Melkert” jobs6 for difficult-to-place workers (e.g., immigrants and refugees with poor command of Dutch and the long-term unemployed). Policies are developed and given effect in different ways. These include clauses built into sectoral or subsectoral collective agreements, firmspecific programs developed by works councils in the context of existing collective agreements, as well as government programs like the Melkert jobs. Certain elements are constants, such as keeping wage increases below levels which would generate inflation or make industries less competitive, while devoting a portion of available “wage room” to finance these pro6

Named for Ad Melkert, the minister of social affairs and employment when these programs were introduced.

254

Modell Nederland

grams. Other provisions, such as work-time reduction or, more recently, flexibility, may change over time. The initial 1982 accord encouraged sectoral unions to exchange automatic wage indexation clauses, written into collective agreements in the 1970s, for possible reductions in working time and creation of part-time jobs to take up the slack. This was followed by agreements in 1984 and 1986 on youth and long-term unemployment and training for workers. More recently, unions conceded greater flexibility in the organization of work in exchange for further education and training to ensure that workers remain employable despite changing technology and modes of production. The “Flex Accord” guaranteed the rights and benefits of part-time and temporary employees in exchange for easing regulations on dismissal (STAR 1996). Recommendations that bargaining parties tailor agreements to the needs of both workers and firms have been a core element of recent proposals. Policies and programs flow from recommendations made by trade union federations and employers associations to their affiliates in different sectors of the economy. This reflects both the industrial relations system and the uses that have been made of corporate structures established in the aftermath of World War II. Collective agreements are negotiated, as they were under earlier systems of statutory wage regulation, by trade unions and employers associations operating in the major sectors of the Dutch economy. Social partners meet regularly in the Foundation of Labor, which provides a “platform” for discussions between social partners, and the Social and Economic Council. The Foundation of Labor is a private organization founded in 1943 and unveiled when the Netherlands was liberated (Stikker 1965). The SER stemmed from pre- and postwar debates on ordering the economy. Established by the 1950 Law on Public Industrial Organization, the SER was to serve both as the capstone of a series of public industrial organizations, which were to emerge from society, and as the principal advisory organ to the government on social and economic policy. Of its two roles, the second has been more important. Except in agriculture, horticulture, and retail trade, public industrial organization never got off the ground. The SER consists of equal numbers of trade union and employers representatives and independent members appointed by the crown. The latter include the president of the Netherlands Bank and the director of the Central Planning Bureau and academic experts (typically professors of economics, sociology, or industrial relations) balanced according to political affiliation. Crown members can use their 255

Steven B. Wolinetz

expertise to suggest alternatives, facilitate compromise, or tip the balance between capital and labor. The foundation has no counterpart to the crown members. Bipartite recommendations made by the Foundation of Labor may be included in sectoral contracts. Although the foundation is not an advisory organ, the government sometimes asks it for advice. The Dutch government is not formally a party to Foundation of Labor agreements, but the minister of social affairs and employment and other ministers with economic portfolios are aware of what is being discussed and may participate in meetings. Government policies not only influence but may be influenced by agreements concluded in the STAR. Wage increases reflect projections made by the Central Planning Bureau. If necessary, the government may pass legislation that gives effect to matters agreed to in the Foundation of Labor. This was done in 1982, when the Dutch Parliament passed legislation allowing sectoral trade unions and employers associations to reopen signed contracts in order to negotiate the elimination of wage indexation and reductions in working time, and again in 1996, when Parliament adopted the Law on Flexibility and Security recommended by social partners. Government officials and the staffs of STAR and the SER facilitate the process by maintaining contacts, making suggestions, and keeping lines of communication open. In addition, the system is underpinned by frequent contacts among people who regularly work together. All of this tells us a good deal about how the contemporary system operates, but very little about how a corporatist system which had ground to a halt in the 1970s was reinvigorated and regenerated in the 1980s and 1990s.

Reinventing Corporatism: The Origins of the Polder Model If democratic corporatism is to work, trade unions, employers associations, and the government must be willing to talk to one another, find compromises, and be able to explain their bargains to others. Leaders of the trade union federations must be able to sell their deals to the leaders of affiliated unions, who in turn must retain the support of their members. Employers organizations are similarly constrained. Nor is the government completely free. Cabinets and the coalitions which support them may not be in agreement. Corporatism attributes special status to certain groups (Offe 1981). In an individualistic society, such an approach is not always 256

Modell Nederland

accepted. In recent years, politicians and others, anxious to assert the primacy of politics, have expressed skepticism about the social partners and the role they play. The ups and downs of Dutch corporatism enable us to examine why political actors opt for more or less cooperation. Several points should be borne in mind. One is that during and after World War II, the Dutch built up a highly institutionalized variant of corporatism which placed its stamp on the principal actors and the choices that they were likely to make. Both the historic build-up of the Dutch trade union movement and employers associations and the legal and institutional framework in which they operated exerted profound influence over the choices that they were likely to make. Second, although Dutch corporatism stalled in the 1970s, this was not solely because of the radicalization and decentralization of the trade union movement, but also because portions of the trade union movement were trying to redefine the terms of the postwar bargain in ways unacceptable to the employers. Third, faced with mounting unemployment and austerity measures which threatened the welfare state, trade unions in the 1980s were forced to back off and do whatever they could to save jobs and preserve benefits. In view of the structure of the Dutch trade union movement, built up to work within a system of central bargaining, one of the most obvious alternatives was to seek bargains with employers and governments. Militant strategies, such as the 1983 strikes against reductions in public sector wages and social benefits, had limited success (Wolinetz 1989). Postwar corporatism provides a useful starting point. The Netherlands relied on a highly institutionalized variant of corporatism in order to administer statutory incomes policies and facilitate postwar recovery and industrialization. Holding wages down ensured investment and complemented the active industrial policies pursued by the Ministry of Economic Affairs (de Hen 1980). Incomes policies were established and implemented by trade union federations and employers associations in the Foundation of Labor on the basis of projections worked out by the Central Planning Bureau. Wage restraint helped the Netherlands industrialize but brought other problems with it. Investors made enormous profits. The postwar recovery was complete by the late 1950s, but trade unions and employers associations (at this time still segmented into Catholic, Protestant, and Socialist organizations) could not agree on how to rework the system of guided incomes policies. While they were arguing, the system spun out of control. 257

Steven B. Wolinetz

Wage restraint proved difficult to sustain in a full-employment economy. “Black” wages paid by employers on the German border triggered wage explosions, subsequently legitimized in 1963, 1964, and 1965. By the end of the decade, trade union federations and employers organizations agreed that it was time to return to free collective bargaining (Windmuller 1969). The Dutch were supposed to do that in the 1970s, but other factors intervened. Workers and their unions became more militant, and power shifted from central offices to sectoral unions, which themselves were quite centralized. Trade unions adopted a more militant posture. Rather than pursuing wage increases, the trade unions attempted to secure nonmaterial demands, such as investment controls, increased codetermination, and sharing of surplus profits. The formation of the left-of-center den Uyl cabinet in May 1973 provided an opportunity. The cabinet was committed to a redistribution of “wealth, knowledge, and power.” However, the first energy crisis intervened. The den Uyl government responded with an enabling act giving it the power to freeze wages and prices. This turned out to be the first in a series of interventions and freezes imposed by successive governments. Unions were anxious to preserve automatic wage indexation which employers repeatedly sought to eliminate. Strikes broke out in 1972–73, when the Industrial Union (metal, electricity, chemicals, general industry) rejected the only central accord successfully negotiated, and again in 1977, when the employers associations tried to end automatic wage indexation. Although the wage measures imposed by the den Uyl cabinet narrowed income differentials, unions found it difficult to secure the broader reforms that they sought. Time was needed to prepare legislation. When it was finally ready, a split in the coalition brought the cabinet down two months short of scheduled elections.

The Road to Wassenaar By 1979, the decline in employment in manufacturing was visible enough that some unions were ready to make concessions in order to preserve jobs. However, the unions federations were not able to complete their retreat until 1982. Although the Industrial Union was painfully aware of job losses, public sector unions and the Food Workers Union were not. A 1979 central accord failed when affiliated unions refused to give officials sufficient latitude to reach an agreement. In the aftermath, the FNV emphasized job sharing, but employers were unwilling to discuss proposals for either the redistribution of work or agreements to guarantee a certain 258

Modell Nederland

number of jobs. Confronted with a corporatist system unable to reach agreement, the van Agt cabinet resorted to wage measures, antagonizing both employers and unions. The stalemate continued until November 1982, when FNV chairman Wim Kok (currently prime minister) seized upon an offer from his VNO counterpart, Chris van Veen, to negotiate anywhere at any time. This led to the Wassenaar Accord. Agreement was possible because (1) the trade unions were willing to allow their leaders to make concessions, (2) the employers associations – with some reluctance on the part of their members – were willing to discuss work-time reductions (though not necessarily to concede them), and (3) the incoming Lubbers cabinet was willing to share limited public space with employers associations and trade unions rather than impose its own wage measure. Let us examine each actor in turn. In the face of mounting unemployment, the trade unions had few alternatives but to seek an accommodation with employers and the government. Attempts to redefine the postwar bargain had failed. Backing off fit the unions’ ideological bent and the resources available to them. Both reflect the ways in which trade unions had been built up over time. Historically, employers’ opposition and religious segmentation had made it difficult for Dutch trade unions to establish themselves in the workplace. Tendencies toward centralization were reinforced by the establishment of a highly institutionalized form of corporatism after World War II. Trade unions had to cultivate their own expertise in order to participate effectively. The employers were willing to engage the trade unions and allow reductions in work time onto the agenda because many preferred to work with the unions rather than confront or attempt to destroy them. This reflects a culture of conciliation and a pragmatic view that unions are a useful interlocutor in the workplace and society as a whole. Even so, not everyone agreed with the concession which allowed reductions in hours to be put on the table. Opponents included the Federation of Metal and Electrical Industries, the counterpart to the Industrial Union, as well as other affiliated organizations and firms, which, though not in agreement, were reluctant to undercut their leaders. The position of the incoming cabinet was also crucial. The 1970 Law on Wage Formation empowered the cabinet to impose wage freezes under certain circumstances. The den Uyl and van Agt cabinets had done so repeatedly. By 1982 a sense of crisis prevailed and key groups were ready for change. The Lubbers government was prepared to impose yet another 259

Steven B. Wolinetz

wage freeze but instead chose not only to accept the deal reached by unions and employers (its support was required to ensure the passage of legislation permitting the renegotiation of signed collective agreements) but also to abolish the provisions of the 1970 Law on Wage Formation allowing the government to intervene. This reflected an explicit choice: rather than reserving the last word for the government, Minister President Ruud Lubbers and Minister of Social Affairs and Employment Jan de Koning let social partners assume responsibility for wage levels. In doing so, Lubbers and de Koning opted to work with trade unions and employers associations rather than around or without them. Reasons for this choice include the possibility of securing cooperation on other issues (which they did not) and the feeling that they had little to lose by sharing some public space with trade unions and employers (Crouch 1993). However, sharing public space did not mean offering social partners carte blanche or drawing them in on a wide range of decisions. Accepting the Wassenaar Accord was easy because unions and employers were prepared to do themselves what the government was going to do anyway. On issues in which there was less agreement, the cabinet was less willing to incorporate social partners.

Beyond Wassenaar The circumstances which we have described – the trade unions’ willingness to concede automatic wage indexation, employers’ willingness to engage the trade unions, and the government’s willingness to accommodate the social partners – were sufficient to make the 1982 accord possible and keep the social partners talking and occasionally reaching agreement on topics of mutual interest, typically strategies to reduce longterm unemployment. Trade union federations remained strongly committed to consultations, but neither the employers association nor the government had a strong interest in doing more than keeping lines of communication open and reaching occasional agreements. The positions of trade unions, employers, and government reflected their objectives and the means available to achieve them. Confronted with high unemployment and a government determined to trim public expenditures, trade unions found themselves on the defensive, in a weakened position, with few alternatives other than engaging others in central discussion. Trade unions were interested in maintaining employment and preserving the welfare

260

Modell Nederland

state. Doing the former required them to engage private sector employers and the government. Doing the latter required them to engage the government and anyone else who would support their cause (primarily opposition political parties). Employers in some sectors were willing to talk about reductions in work time with concomitant loss of income but were less interested in the creation of part-time jobs and the proposed redistribution of work. Trade union officials found that they could realistically expect the replacement of only 25% of the hours saved. Thus, work-time reductions increased profitability and competitiveness by eliminating overstaffing. Trade unions could not force employers to create more jobs. Workers were unlikely to strike in a period of high unemployment to create jobs for unspecified individuals. Unable to do more than help firms to be competitive, unions directed their attention to the government. However, the government’s desire to trim the welfare state meant that government policies were diametrically opposed to what the unions wanted. To gain control of its own expenditures, the government severed linkages between private sector wages and public sector salaries and benefits and trimmed budgets repeatedly to meet deficit-reduction targets. In response, both the FNV and the smaller CNV assumed a broad tribune-like role, defending not only their members but also recipients of benefits. However, the unions’ ability to mobilize occasional demonstrations did little to change government policy or create jobs. The employers’ position was slightly different. The central employers associations needed to engage government and trade unions in order to preserve their raison d’être but had no reason to take bipartite or tripartite talks any further. The first and second Lubbers cabinets were anxious to restore profitability, reduce deficits, and bring public sector expenditures under control. Employers were generally satisfied with government policies and had access to the government through other channels. Although employers had reason to maintain contacts with unions through the Foundation of Labor, they had little use in the 1980s for the Social and Economic Council. Lubbers and Minister of Social Affairs Jan de Koning were anxious to keep lines of communication open but had no reason to encourage a more full-blown consultation or shared decision-making process. In the 1970s corporatist structures had proved unable to achieve consensus or deal with major problems confronting the Netherlands. Despite the Wassenaar

261

Steven B. Wolinetz

Accord, there was little evidence that they were better equipped to do so in the 1980s. Rather than rely on stalemated advisory organs like the Social and Economic Council, the Lubbers government created special commissions to provide alternate sources of legitimacy to consider changes in industrial policy, health care, and the scope and organization of government. Led by prominent industrialists, these commissions typically included people familiar with union and employers’ positions, but those involved served as individuals rather than group representatives. Even if the government did not want to rely on the social partners, efforts were made to ensure that they met regularly. In contrast to the 1970s, when announcements of talks generated expectations of success and reports of failure in the media, contacts in the 1980s were low key: the FNV’s collective bargaining coordinator, the VNO’s director for social affairs, and the official in charge of labor relations at the Ministry of Social Affairs and Employment met for lunch once a month to consider whether there were grounds for more formal discussions. The 1984, 1986, and 1989 accords emerged from this process. Actors, then, chose to maintain but not fully employ corporatist structures in the 1980s. The principal institutions – the Foundation of Labor and the Social and Economic Council – persisted and, deliberately or not, were held in reserve while the government facilitated sufficient contact to maintain a form of “massaged” corporatism. This blossomed into the polder model in the 1990s.

From Massaged Corporatism to the Polder Model The shift from a lighter variant of corporatism, based on informal contacts, in the 1980s to a more broadly based form in the 1990s reflects choices made initially by social partners and only latterly by the government. The first faltering step was the Common Policy Framework concluded in 1989. This requires little discussion. The 1989 fall of the center-right second Lubbers government and its replacement by the center-left Lubbers-Kok altered the situation we have just described. After years of decrying tripartite bargaining, employers were willing to embrace it lest the trade unions, now better positioned, made a deal with a government which might prove friendlier to unions. Unions were willing to go along. The Common Policy Framework was the result. However, the agreement – constructed without an underlying consensus – produced few results. Promises to take account of partners’ positions had little 262

Modell Nederland

impact on a system in which they were in regular contact and did that anyway. Change was provoked by the government and then advanced by the social partners. The most important developments came in 1991, when employers pulled back from central discussions, and in 1992 and 1993, when trade unions and employers associations discovered common ground. Employers pulled back because they felt that the government was not keeping its part of the bargain. However, remaining outside established policy circuits was not in their interest. Employers enlisted the trade unions in the series of exchanges which we have characterized as the polder model. The Lubbers-Kok government helped by provoking unions and employers to come together. This occurred not because Premier Lubbers or his vice-premier, Wim Kok (leader of the FNV at the time the Wassenaar Accord was signed), were hostile, but rather because the Minister of Social Affairs and Employment, Bert de Vries, was impatient with social partners. Anxious to ensure wage restraint, de Vries threatened to intervene in wage formation. Whether these threats were taken seriously is a matter of conjecture because the government, as part of the 1982 Wassenaar Accord, had stripped itself of the legal means to intervene. More important is that de Vries also raised doubts about the automatic extension of collective agreements which, if carried out, threatened the entire system of industrial relations. To make sense of this, we need to explain the process of extension. In 1938, ministers of social affairs gained the authority to extend duly negotiated collective agreements and declare them binding on entire sectors. This practice, routinely carried out by successive ministers, is important to both employers associations and trade union federations. On the employers’ side, extension means that firms cannot evade collective bargaining by hiring nonunion members or establishing parallel firms. Unable to escape, firms have strong reasons to join employers associations; voice replaces exit (Hirschman 1970). The effects on the union side are somewhat different. Workers benefit from collective agreements whether they are union members or not. Free-riding is a problem for trade unions – unions require members to maintain their credibility – but although membership is important, it need not be a paramount concern if unions, individually or collectively, are able to negotiate collective agreements which are then declared binding across entire sectors. Unions can maintain their position in corporatist circuits with low membership densities. 263

Steven B. Wolinetz

Frustrated with the outcomes of collective bargaining, Minister of Social Affairs de Vries threatened to make extension contingent on the content of collective agreements. De Vries’s comments resonated with growing doubts about grants of status or privilege to special groups. Such doubts represent a challenge in a society which has made a virtue of granting status or power to religious or ideologically based groups. Although it would be an overstatement to argue that threats to eliminate or restrict automatic extension of collective agreements or intervention in wage formation led to the polder model, they are one more factor bringing trade unions and employers associations, and eventually the government, together. More important factors are the discovery of common interests reinforced by ongoing experience. This does not rest solely on consensus. Institutions which both limit and weight actors’ choices and force them to view issues in common frames of reference play a major role.

Conclusion The road from Wassenaar to the polder model has not been an easy one. Concertation was possible in the 1990s because the institutions in which principal actors operate narrowed their choices and facilitated the discovery of common interests. Even so, the present system is not a full-blown version of corporatism in which trade unions and employers associations share in the formulation and implementation of social and economic policy, but a far more limited and contingent variant, in which social partners have gained control over one area, wage policy and industrial relations, and may exert influence in others. As Visser and Hemerijck (1997) demonstrate, corporatist decision making does not extend to other areas. When successive governments were unable to gain consent for changes which they desired in social welfare programs, these were imposed. The polder model is thus a limited and contingent form of corporatism, dependent on the willingness of actors to participate and their ability to achieve results satisfactory to the government. The polder model depends on consensus and would have difficulty operating without it. However, that consensus flows not only from ideas or forces which put power behind them (Hall 1993; Visser and Hemerijck 1997) but also from institutions which help give them form. In the Dutch case, institutions established for other reasons happened to be available. Whether or not the Dutch model is transferable remains to be seen. Most participants and observers believe that it is not. Few societies 264

Modell Nederland

have either the same battery of institutions available for use when needed or the same cultural background. If the polder model is to be exported, it will have to be done by finding other means of developing and sustaining a consensus mutually beneficial to capital, labor, and government. We need to know much more about both the role which institutions play in corporatist arrangements and the circumstances under which such systems develop and take hold. Who benefits? Critics such as Erik Jones (1999) argue that the polder model works to the disadvantage of older people, who have been deprived of employment, youth who receive lower entry wages, and women who disproportionately hold lower-paying part-time jobs. Jones is correct in his assessment that some groups or categories benefit more than others, but it is unclear whether alternate arrangements would produce a fairer or more egalitarian outcome or whether the groups he mentions are discontented with their situation. Dutch trade unions and others on the left tried to polarize and operate from a different model in the 1970s, but this proved to be inappropriate, and trade unions found the results – deindustrialization and massive unemployment – unacceptable. Only by backing off were the unions able to find alternate ways to defend the interests of their members and others whom they try to represent. Discontent is harder to measure: despite the temporary success of parties for the aged in the 1994 elections, there is scant evidence of widespread discontent among pensioners. Nor is it clear that women or youth are as disadvantaged as Jones suggests. Part-time employment is not ideal, but it meshes with a tax system favorable to family units with 1.5 wage earners and provides time for other activities. Invariably, we come to an impasse and must decide whether some work is better than no work. In concluding, we might consider whether workers feel disadvantaged with what their unions have served up. The comments of former FNV chairman Johan Stekelenburg to a meeting of the Finland Trade Club in 1997 are interesting in this regard: In recent months, I have frequently been approached by journalists . . . in their mind the only way Dutch business can have been so successful is by exploiting the worker. “What do your members think of it?” [they] ask. . . . The answer will surprise you. Just like anywhere else, workers vote in the Netherlands with their feet. If they do not get what they want, they leave. And if they are satisfied, they will come to you. Over the last eight or nine years the FNV has seen its membership rise by more than 25 percent! The degree of unionization has also grown by 4 or 5%. The Dutch trade union movement is the only one in Europe which can show you such a record. (Stekelenburg 1997b: 2) 265

Steven B. Wolinetz

References Adviescommissie inzake het industriebeleid (Advisory Commission on Industrial Policy). 1981. Een Nieuwe Industrieel Elan (A new spirit for industry). The Hague. Akkermans, Tinie, and Paul Grootings. 1978. From Corporativism to Polarization: Elements of the Development of Dutch Industrial Relations. In Colin Crouch and Alessandro Pizzorno, eds., The Resurgence of Class Conflict in Western Europe since 1968. London: Macmillan, 159–90. Dercksen, W. J. 1986. Industrialisatiepolitiek rondom de jaren vijftig. Assen: van Gorcum. Cameron, David. 1984. Social Democracy, Corporatism, Labour Quiescence and the Representation of Economic Interest in Advanced Capitalist Society. In John Goldthorpe, ed., Order and Conflict in Contemporary Capitalism. Oxford: Clarendon Press, 143–78. Crouch, Colin. 1993. Industrial Relations and European State Traditions. Oxford: Clarendon Press. Hall, Peter A. 1993. Policy Paradigms, Social Learning, and the State: The Case of Economic Policy-Making in Britain. Comparative Politics 24: 275–96. Heide, F. J. ter. 1986. Ordening en Verdeling: Besluitvorming over sociaal-economisch beleid in Nederland 1949–1956. Kampen: Kok Agora. Hen, P. E. de. 1980. Actieve en re-active industriepolitiek in Nederland: De overheid en de ontwikkeling van de Nederlandse industrie in de jaren dertig en tussen 1945 en 1950. Amsterdam: Arbeiderspers. Hirschman, Albert O. 1970. Exit, Voice, and Loyalty: Responses to Decline in Firms, Organizations, and States. Cambridge, Mass.: Harvard University Press. Jones, Erik. 1999. Is “Competitive” Corporatism an Adequate Response to Globalization? Evidence from the Low Countries. West European Politics 22: 159–81. Katzenstein, Peter J. 1985. Small States in World Markets: Industrial Policy in Europe. Ithaca, N.Y.: Cornell University Press. Markovits, Andrei S., ed. 1982. The Political Economy of West Germany: Modell Deutschland. New York: Praeger. Nobelen, P. W. M. 1987. Ondernemers georganiseerd. Een Studie over het verbond van Nederlandse Ondernemingen in de periode 1973–1984. Ph.D. diss. Rotterdam. Organisation for Economic Co-operation and Development (OECD). 1998a. OECD Economic Outlook. December. ———. 1998b. OECD Economic Surveys: The Netherlands, 1997–1998. Paris: OECD. Offe, Claus. 1981. The Attribution of Public Status to Interest Groups: Observations on the West German Case. In Suzanne Berger, ed., Organizing Interests in Western Europe: Pluralism, Corporatism, and the Transformation of Politics. Cambridge: Cambridge University Press. Raad van de Centrale Ondernemingsorganisaties (RCO). 1991. (Council of the Joint Employers Associations, Letter to the Cabinet). 11.823/Bk/AB Sociaal-Economische Raad (SER). 1992. Convergentie en overlegeconomie: Advies inzake convergentie en overlegeconomie. Uitgebracht aan de minister van Sociale Zaken 266

Modell Nederland and Werkgelegenheid, de minister van Economische Zaken en de minister van Financie (Convergence and Consultative Economy: Advice on convergence and consultative economy presented to the Minister of Social Affairs and Employment, the Minster of Economic Affairs, and the Minister of Finance). SER 92-15. The Hague: November 20. Stekelenburg, Johan. 1997a. Aantekeningen voor Johan Stekelenburg, voorzitter van de FNV, ten behoeve van een spreekbuurt over de arbeidsverhoudingen in Nederland tijdens de Nederlandse City Lunch, Woensday, Januari 22. (Notes for Johan Stekelenburg, chair of the FNV for a speech on labor relations in the Netherlands, Netherlands City Lunch, Wednesday, January 22) FNV website http://www/fnv.nl/Toespraak/9701.htm. ———. 1997b. Industrial Relations as a Win-Win Game: Beyond the Prisoners’ Dilemma. Speech delivered to the lunch meeting of the Finland Trade Club, The Hague, May 13. FNV website http://www/fnv.nl/ Toespraak/97051301.htm. Stichting van de Arbeid (STAR). 1989. Gemeenschappelijk beleidskader van het kabinet en de Stichting van de Arbeid d.d. 1 December 1989 (Common Policy Framework of the Cabinet and the Foundation of Labor, dated December 1). The Hague: Stichting van de Arbeid. ———. 1993. Principe-Overeenkomst: Een Nieuwe Course: agenda voor het cao-overleg 1994 in het perspectief van de middellange termijn (Foundation of Labor, basic agreement: a new course: agenda for the 1994 negotiations on collective labor agreements in a medium-term perspective). The Hague: Stichting van de Arbeid. ———. 1996. Nota Flexibiliteit en Zekerheid (A policy document on “flexibility and security”). No. 2/96. The Hague: Stichting van de Arbeid, April 3. ———. 1997. Agenda 2002: Agenda voor het cao-overleg in de komende jaren (Agenda 2002: collective bargaining agenda for the years ahead). No. 13/97. The Hague: Stichting van de Arbeid, December 9. Stikker, D. U. 1965. Men of Responsibility: A Memoir. New York: Harper and Row. Visser, Jelle, and Anton Hemerijck. 1997. A Dutch Miracle: Job Growth, Welfare Reform and Corporatism in the Netherlands. Amsterdam: Amsterdam University Press. Windmuller, John P. 1969. Labor Relations in the Netherlands. Ithaca N.Y.: Cornell University Press. Wolinetz, Steven B. 1989. Socio-economic Bargaining in the Netherlands: Redefining the Postwar Policy Coalition. West European Politics ( January): 79–99 [special issue on the Netherlands; also published in Hans Daalder and Galen Irwin, eds., Politics in the Netherlands: How Much Change? (Frank Cass, 1989)]. WRR (Wetenschappelijke Raad voor Regeringsbeleid). 1980. Plaats en toekomst van de Nederlandse industrie. (Netherlands Scientific Council for Government Policy, Industry in the Netherlands: Its Place and Future, trans. 1982). Rapporten aan de Regering no. 18.

267

PA R T I V

Unemployment, Voting, and Political Behavior

9 Desperate Times Call for Desperate Measures? UNEMPLOYMENT AND CITIZEN BEHAVIOR IN COMPARATIVE PERSPECTIVE

Christopher J. Anderson

European voters have long been concerned about high unemployment rates.1 In fact, during much of the postwar period, European mass publics placed unemployment near or at the top of the list of most important problems facing their country (Anderson 1995). In Eurobarometer 41 (Spring 1994), for example, roughly 25 percent of European voters considered unemployment to be their country’s most important problem (24.9%), followed by issues specific to individual countries (mentioned by 8.9% of respondents), and far ahead of crime, prices, immigration, or the environment. In the same survey, over 80% of respondents said that the unemployment situation in the country had stayed the same or gotten worse, and roughly two-thirds (66.6%) expected the employment situation to stay the same or get worse over the next twelve months. Similarly, Eurobarometer 45, taken in the spring of 1996, reveals high levels of anxiety regarding future unemployment trends throughout Europe. The level of agreement about the problem of unemployment across Europe is remarkable and attests to the fact that the secular trend toward higher unemployment that started in the 1960s has not been reversed; in fact, it has risen to the top of people’s political concerns. Unemployment is not a new phenomenon on the European continent, however. After all, the economic depression of the 1920s and 1930s affected Europe as much as the United States. And although European policy makers and citizens were in the fortunate position of witnessing times of almost full employment for over a decade following World War 1

The research for this chapter was supported by NSF Grant SBR-9818525.

271

Christopher J. Anderson

II, joblessness started to rise again in the second half of the 1960s and has been increasing ever since. Figure 9.1, which graphs unemployment rates in the member states of the European Union, reveals quite clearly that there has been an almost linear upward trend in unemployment in Europe since the 1960s. Although unemployment rates in some countries are higher than in others (see, e.g., Chapter 1 in this volume), it also is true that all European democracies have witnessed a significant increase in joblessness over the past thirty years. The notion that unemployment creates significant political problems and constitutes a salient issue for mass publics and policy makers alike also is hardly novel. In fact, the success of Fascist parties in the 1930s and the breakdown of democracy in Germany both have been attributed in part to high unemployment rates and social problems produced by joblessness. Similarly, the recent rise of right-wing radical parties in countries such as France, Germany, Italy, and the Scandinavian states has been linked to the greater propensity of the unemployed and underemployed to support radical political solutions (Betz 1994; Anderson 1996). Because unemployment is an old problem that has been associated with the failure of democratic processes and undesirable social consequences, social scientists have long been interested in unemployment both as a

Figure 9.1 272

Unemployment Rate in the European Union.

Desperate Times Call for Desperate Measures?

dependent and independent variable. Viewed from the perspective of policy making, researchers have examined the effects of politics – broadly conceived – on economic outcomes, including unemployment (Hibbs 1977; Lange and Garrett 1985; Crepaz 1995). Focusing on partisan preferences for particular economic outcomes, the interplay of government partisanship and mechanisms of interest aggregation, and the role of political economic institutions, scholars have sought to establish the political conditions under which low levels of unemployment can be achieved (see also Chapter 2). In a related vein, a number of scholars recently started to investigate the effects of greater economic interdependence, globalization, and regional integration on employment. While many agree that states’ steering capacities have been constrained by globalization and international trade (see, e.g., Chapter 3; see also Keohane and Milner 1996), it is not altogether clear that this automatically results in a loss of state control or convergence toward neoliberal policies – all of which may be associated with different employment patterns. Moreover, scholars have been debating whether European Monetary Union and the introduction of the euro will produce the desired effects of reducing unemployment in the short run (see, e.g., Chapter 1). What is clear, however, is that more than ever before unemployment as a policy outcome is systematically affected by the interplay of both domestic and international political factors. Viewed from the perspective of voters and focusing on a nation’s unemployment rates as an independent variable, a rich literature has sought to explain voting behavior, government and party support, and trust in the political system with the help of changes in employment trends (Anderson 1995; Norpoth 1996). Usually centering around the question of who deserves blame for rising or high levels of unemployment, this line of research has thoroughly investigated the effects of unemployment (and other economic outcomes) on political behavior and election outcomes both among individuals and across countries (see Lewis-Beck and Paldam 2000). Despite the obvious importance of unemployment for people’s lives and the detrimental effects of poor economic performance on the functioning of political systems, few scholars have examined whether the political behavior and attitudes of unemployed voters differ from those of employed ones, however. This is all the more surprising as unemployed voters traditionally have been considered to be most likely to engage in political acts that may contribute to the instability of democratic regimes and a 273

Christopher J. Anderson

breakdown of civil society. This chapter seeks to help fill this gap by examining the political attitudes and behaviors of employed and unemployed voters and comparing their differences across twelve European states. I first develop some hypotheses about how unemployment affects political behavior, both at the level of individuals and across countries. I then test these hypotheses with the help of Eurobarometer data. In the concluding section I discuss the results and their implications.

Employment, Unemployment, and Political Behavior Social scientists have paid little attention to the question of how the unemployed and underemployed vote or how unemployment affects political behavior more generally. In part, this has been a data problem. To conduct systematic analyses and draw meaningful conclusions at the level of individual voters, one needs a large number of unemployed respondents. Given that the unemployed (and those who admit to being unemployed in surveys) usually constitute only about 5 to 10 percent of any random national sample, it frequently is difficult to draw systematic inferences about how the unemployed behave or what they think. Fortunately, the Eurobarometer data analyzed in the next section allow for the pooling of data across countries and over time and thus permit a more thorough analysis of differences across employed and unemployed citizens. Aside from the problem of data availability, the unemployed have not usually figured prominently on the radar screen of those interested in political attitudes and behavior because unemployed voters tend to constitute only a small minority of the voting-age population. As a result, they may not be politically relevant in the aggregate – especially if, as many assume, they participate even less than employed voters. Given the secular rise in unemployment rates across Europe, however, this question may be due for reconsideration. Given that the unemployed, and especially the long-term unemployed, now constitute a significant share of the overall voting-age population across many European democracies (cf. Chapter 1), it would be important to know more about how they feel and behave in the political arena. In addition to the possible negative political consequences in terms of behaviors and attitudes exhibited by the unemployed themselves, there also may be significant political spillover to the general public. Given that the proportion of unemployed voters has increased over the past two to three decades, the chance that other voters come in contact with and pay attention to them has increased as well. This, in turn, may 274

Desperate Times Call for Desperate Measures?

indeed affect those formulating policy agendas and competing for elected office.

Differences among Employed and Unemployed: Economic-Psychological Research Since the depression of the 1930s, and again since the economically troubled 1970s, social scientists have examined the effects of unemployment on people’s psychological well-being (Bakke 1933; Eisenberg and Lazarsfeld 1938; Brenner 1973; Catalano and Dooley 1983). In particular, they have been interested in how being unemployed affects people’s sense of mental well-being (Brenner 1973; MacFadyen, MacFadyen, and Prince 1996: 293). Those who face personal economic risk, such as being laid off or already having been laid off, are less likely to maintain high levels of self-esteem and are more likely to be depressed and to engage in a variety of detrimental behaviors. Moreover, those who fear unemployment tend to have lower levels of job and life satisfaction. While instructive, earlier research has not considered political attitudes or behaviors as a dependent variable. Extrapolating from its findings, however, one would expect unemployment also to have negative effects on political behavior. Specifically, unemployment should be associated with lower levels of political involvement because of a lack of embeddedness in social milieus, organizations, and social support networks. At the level of individuals and on the basis of human capital notions of behavior, it is well known that those who are unemployed tend to have fewer resources to be engaged in political life. Thus, I also would expect to see lower levels of political participation among those who participate less than fully in the labor market.

Political Science Research on Unemployment and Political Behavior Among the few students of politics who have directly addressed the question of how unemployment affects voters, Schlozman and Verba (1979) examined political participation among unemployed and employed voters in the United States. Drawing on a national sample of employed and unemployed citizens collected during the 1970s, they found that unemployed voters were less likely to participate in politics. However, their research also showed that many of the differences between unemployed and employed voters were enhanced by differences in 275

Christopher J. Anderson

socioeconomic status among these groups. Thus, Schlozman and Verba found that although unemployed voters voted less and were less interested in politics, those differences were attenuated when the socioeconomic status of respondents was controlled for (see Verba and Nie 1972). Schlozman and Verba’s study was groundbreaking because it focused explicitly on comparing employed and unemployed voters. Unfortunately, the study has not been replicated since, and it is simply not known whether similar dynamics can be observed in countries other than the United States or whether unemployment’s negative consequences extend to other aspects of political behavior. Although there has been some comparative research linking young unemployed males to extreme right-wing parties and movements (Betz 1994), such research was not designed to explicitly address the difference between employed and unemployed voters (see also Anderson 1996).

Hypotheses This study picks up where earlier research left off and addresses the question of how unemployment affects political behavior and attitudes in European democracies in the 1990s. It focuses on cross-national differences in how unemployment affects political behavior and attitudes. Specifically, I compare employed and unemployed citizens’ ideological, policy, and political attitudes as well as their involvement in politics. These general attitudes and behaviors measure general underlying satisfaction with one’s life and the way the political system functions, citizens’ active individual engagement with the political process, and where individuals locate themselves politically. Moreover, because voting turnout is the most commonly used indicator of people’s active participation in the political process, the study seeks to replicate Schlozman and Verba’s findings regarding turnout among European voters. I expect to find that unemployment has negative consequences. Specifically, unemployed voters should be less satisfied with their life and the political system than employed voters; unemployed voters should be more extreme in their ideology than employed voters; unemployed voters should be less interested in politics than employed voters; and unemployed voters should be less likely to vote than employed voters. Because I am able to test these hypotheses in a number of European democracies, I also examine whether there are differences across the coun-

276

Desperate Times Call for Desperate Measures?

tries included in the study. Specifically, on the basis of the notion that welfare states are designed to alleviate the ill effects of a market economy – after all, unemployment is one of the prime negative outcomes individuals experience in market economies – I hypothesize that high levels of welfare state development should reduce the expected negative effects of unemployment on political behavior. Unemployed voters should be less negatively affected in more highly developed welfare states than in welfare states that are less extensive. Moreover, the gap between employed and unemployed voters with regard to our measures of attitudes and behaviors should be smaller in more highly developed welfare states. Taken together, these hypotheses state that there should be a gap between employed and unemployed voters with regard to political behavior, but also that welfare states help temper the ill effects of unemployment on voter behavior.

Data and Measures To test these hypotheses, I pooled two Eurobarometer data sets from 1994 (EB 41.1 and 42), which included random national surveys of citizens in the (then) twelve member states of the European Union. The national surveys involved about 1,000 respondents per country (with the exception of Luxembourg with about 500); the entire data set thus yielded roughly 25,000 respondents. Of the total number of respondents, about 7 percent indicated that they were unemployed, with the number varying somewhat across the member states. To analyze citizen attitudes and behavior, I relied on the following variables: life and political satisfaction, political interest, ideological extremism and xenophobia, as well as voting behavior. Life and political satisfaction were measured by the proportions of citizens indicating that they are very or fairly satisfied with their life or the way democracy works (respectively), whereas political interest simply asked respondents to what extent they were interested in politics. The proportions of those who said they were very or fairly interested were used here. Ideology was measured by using respondents’ self-placement on the well-known 10-point left-right scale: I then calculated the proportions of citizens locating themselves at the extreme left and right ends (1–2 and 9–10) of the scale. Xenophobia is a variable that calculated the percentages of citizens saying that there were too many foreigners in the country, and voting turnout, finally, was measured by asking respondents whether they voted in the last national

277

Christopher J. Anderson

election. To distinguish between employed and unemployed respondents, I relied on a question gauging respondents’ occupational status.

Unemployment and Citizen Attitudes Life Satisfaction Because earlier research had found that the unemployed are likely to have lower levels of job and life satisfaction, I tried to replicate these results in the cross-national European context. Thus, I calculated levels of life satisfaction both among the employed and the unemployed. As Table 9.1 shows, employed Europeans consistently have higher levels of life satisfaction than unemployed ones. The gap in satisfaction levels ranges from a low of 5.8 percent in Denmark to a high of 38.2 percent in Germany, with most countries showing a satisfaction gap somewhere in the vicinity of 20 percent. While a majority of the unemployed across Europe (60.9%) says that they are satisfied with their life, two out of five unemployed Europeans say that they are not. The least satisfied unemployed can be found in Greece, whereas 91 percent of all Danish unemployed say that they are satisfied with their life.

Table 9.1. Life Satisfaction among Employed and Unemployed Citizens (percentage satisfied) Country

Employed

Unemployed

Difference

Respondents

Belgium Britain Denmark France Germany Greece Italy Ireland Luxembourg Netherlands Portugal Spain Total population

86.9 89.3 96.8 77.9 86.7 51.8 89.1 89.7 95.7 94.8 71.2 72.6 82.5

58.8 71.3 91.0 50.0 48.5 40.4 58.2 73.5 75.0 62.5 49.0 46.8 60.9

28.1 18.0 5.8 27.9 38.2 11.4 30.9 16.2 20.7 32.3 22.2 25.8 21.6

1,026 1,342 998 1,004 1,194 1,001 1,046 987 496 1,038 996 1,003 12,131

Source: Eurobarometer surveys (1994).

278

Desperate Times Call for Desperate Measures?

Despite nation-specific differences, these figures indicate that being unemployed carries significant personal costs. We can speculate that it is likely that other indicators of personal well-being would show similar gaps between those who have a job and those who do not. Whether these negative consequences of joblessness extend to the political realm as well is investigated next.

Political Satisfaction Because the unemployed are marginalized in the political process, I hypothesized that unemployed voters also would be more cynical about and less involved in the political process. To test this hypothesis with regard to satisfaction with the political system, I examined the differences in satisfaction with the way democracy works among employed and unemployed respondents. The results are shown in Table 9.2. The employed were significantly more satisfied with the way the political system works than the unemployed. There was a political satisfaction gap of 13 percent (55 to 42) across the European public as a whole. Moreover, this gap existed across all EU member states – in most countries, the satisfaction gap was somewhere in the 10–15 percent range. There also

Table 9.2. Democracy Satisfaction among Employed and Unemployed Citizens (percentage satisfied) Country

Employed

Unemployed

Difference

Respondents

Belgium Britain Denmark France Germany Greece Italy Ireland Luxembourg Netherlands Portugal Spain

60.4 52.4 84.0 58.1 56.4 32.2 26.2 69.1 86.1 69.7 50.3 37.2

44.2 41.1 76.5 46.4 31.8 26.6 14.5 58.5 70.6 56.6 34.7 32.2

16.2 11.3 7.5 11.7 24.6 5.6 11.7 11.0 15.5 13.1 15.6 5.0

1,972 2,045 1,990 1,958 3,285 1,987 2,079 1,911 962 2,020 1,904 1,947

Total population

55.5

42.2

13.2

24,629

Source: Eurobarometer surveys (1994).

279

Christopher J. Anderson

were some notable cross-national differences in the magnitude of the satisfaction gap, however: the biggest gap in satisfaction between employed and unemployed was found in Germany (25%), whereas the smallest ones existed in Spain (5%) and Greece (5.6%). Thus, the German unemployed were far more likely to be dissatisfied, compared with their employed compatriots, whereas citizens in the new democracies of southern Europe displayed the smallest differences. The significant political satisfaction gap observed in Germany mirrors the gap in life satisfaction recorded there. Thus, for Germans, being unemployed appears to carry more significant consequences than elsewhere in Europe. These results were consistent with the expectation that the unemployed as losers from economic competition are not full participants in the political economy and are less satisfied with the way the system works. This also indicated that there is spillover from personal experiences into political attitudes: not only are the unemployed less satisfied with their own lives, they also are more dissatisfied with the political process. Thus, although unemployment may be individually experienced, it does translate into systematically lower levels of system support.

Political Extremism and Xenophobia Because previous research has pointed to the possibility that unemployed voters may be more receptive to radical political solutions, an analysis of the differences between employed and unemployed with regard to extreme political views is likely to be revealing. Are unemployed voters more extreme in their ideology than those who have a job? That is, are those out of work susceptible to extreme political solutions or positions? Extremism To answer these questions, I first examined voters’ location along the left-right continuum. To examine whether unemployed voters were more likely to be extreme in their ideology, I calculated the percentages placing themselves in the extreme categories on both ends of the spectrum. Table 9.3 shows the proportions of employed and unemployed voters who placed themselves in the extreme categories (1,2: left; 9,10: right) of the 10-point scale. Although about 15 percent of respondents placed themselves in the extreme categories across the EU member states as a whole, the results show that the overall difference between employed and unemployed across the twelve countries included in the study was small to insignificant (-1.1 280

Desperate Times Call for Desperate Measures?

Table 9.3. Extreme Left-Right Self-Placement of Employed and Unemployed Citizens (percentage) Country

Employed

Unemployed

Difference

Respondents

Belgium Britain Denmark France Germany Greece Italy Ireland Luxembourg Netherlands Portugal Spain

20.7 10.9 10.7 20.2 10.4 24.2 22.3 14.3 9.0 10.7 13.4 20.3

17.4 9.2 8.9 21.3 15.9 18.8 33.7 16.7 18.8 11.5 19.0 17.3

3.3 1.7 1.8 -1.1 -5.5 5.4 -11.4 -2.4 -9.8 -0.8 -5.6 3.0

1,671 1,946 1,926 1,816 2,834 1,734 1,673 1,662 835 1,903 1,539 1,669

Total population

15.4

16.5

-1.1

21,692

Source: Eurobarometer surveys (1994).

percentage points). Thus, this result did not support the hypothesis that unemployed voters are more ideologically extreme than employed ones. In fact, when the cross-national differences are examined more closely, it turns out that employed voters were more extreme in five of the twelve countries (Belgium, Britain, Denmark, Greece, Spain), whereas in the other seven countries (France, Germany, Italy, Ireland, Luxembourg, Netherlands, Portugal) the unemployed were more extreme. Moreover, in seven of the twelve countries, the difference between unemployed and employed voters was within a 3–6 percent range and therefore quite small. The only countries where the difference in extremism was somewhat more sizable were Italy (-11.4) and Luxembourg (-9.8), while the gap was in the 5–6 percentage point range in Germany (-5.5), Greece (5.4), and Portugal (-5.6). In Italy, Luxembourg, Germany, and Portugal the unemployed were more extreme than the employed; in Greece it turned out to be the other way around. Given the small number of respondents in Luxembourg, however, Italy thus turned out to be the only reliable case where the unemployed were significantly more extreme in their political views than the employed. 281

Christopher J. Anderson

Xenophobia A number of studies have found that extremist parties tend to appeal to unemployed and otherwise marginalized voters by scapegoating foreigners. Thus, an additional measure of political extremism – attitudes toward foreigners, or xenophobia – calculated the percentages of citizens saying that there were too many foreigners in the country. The results for employed and unemployed Europeans are shown in Table 9.4. As in the analysis of extreme left-right self-placement, the results (based only on one Eurobarometer) were mixed. Overall, unemployed voters were about as likely to have xenophobic attitudes as employed ones. However, this masked quite considerable variation across countries. While the employed were more xenophobic in Greece and Italy, for example, the unemployed were significantly more likely to have negative attitudes toward foreigners in Belgium and Germany. Overall, however, there was little evidence to support the notion that the unemployed were consistently more extreme than employed voters – with few exceptions. While they were less satisfied with the way the political system works, they were not significantly more extreme than their fellow (employed) citizens in terms of their ideology or policy positions.

Table 9.4. Attitudes toward Foreigners among Employed and Unemployed Citizens (Xenophobia) ( percentage saying “too many”) Country

Employed

Unemployed

Difference

Belgium Britain Denmark France Germany Greece Italy Ireland Luxembourg Netherlands Portugal Spain

56.9 39.0 38.7 52.2 45.2 62.4 57.1 9.7 25.0 41.3 21.7 28.3

63.8 38.0 36.7 53.2 56.0 51.9 52.7 10.1 50.0 46.2 19.6 33.3

-6.9 1.0 2.0 -1.0 -10.8 10.5 4.4 -0.4 -25.0 -4.9 2.1 -5.0

1,003 1,383 1,000 1,000 2,134 1,002 1,067 1,000 502 1,005 1,000 1,000

Total population

40.9

43.5

-2.6

13,096

Source: Eurobarometer surveys (1994).

282

Respondents

Desperate Times Call for Desperate Measures?

Unemployment and Political Involvement If the findings reported by Schlozman and Verba for the United States hold in European democracies, we should find unemployed voters to be less interested in politics and less likely to vote than employed voters. To test these hypotheses, I calculated the percentages of employed and unemployed respondents who indicated that they were very or fairly interested in politics. Table 9.5 shows the results.

Interest in Politics The results show that the employed on average displayed higher levels of political interest than the unemployed; the overall gap across the EU member states was 7 percent (44–37%). Thus, the hypothesis was confirmed when the population as a whole was considered. However, when individual countries are considered, it turns out that this was not uniformly the case. While the unemployed actually were more interested in politics by an 11-point margin in Portugal, there was no discernible difference in Britain, and the difference was small (3%) in Greece. Looking at the other countries that had a disparity in political interest, the results showed that

Table 9.5. Political Interest among Employed and Unemployed Citizens ( percentage interested) Country

Employed

Unemployed

Difference

Respondents

Belgium Britain Denmark France Germany Greece Italy Ireland Luxembourg Netherlands Portugal Spain

29.9 55.2 73.3 43.7 42.6 40.7 39.2 42.4 54.3 51.4 27.8 39.5

20.2 55.7 55.0 36.0 32.4 37.6 25.4 30.8 38.9 43.8 39.0 30.8

9.7 -0.5 18.7 7.7 10.2 3.1 13.8 11.6 15.4 7.6 -11.2 8.7

2,025 2,111 2,000 2,004 3,300 2,000 2,106 1,994 1,001 2,036 1,974 1,992

Total population

43.8

36.6

7.2

25,150

Source: Eurobarometer surveys (1994).

283

Christopher J. Anderson

the gap ranged from a modest 7–8 percent (France, Netherlands, Spain) to a more sizable 15 percent and higher (Italy, Luxembourg). The biggest gap of almost 20 percent (18.7%) was found in Denmark. Overall, the hypothesis that the employed would be more interested in politics than the unemployed was confirmed in nine (or ten if Greece is included) of twelve countries.

Voting Turnout To examine whether employed voters were more likely to vote, I calculated the differences in turnout levels among both types of voters. The results are shown in Table 9.6. Overall, about 81 percent of the employed and about 74 percent of the unemployed said they voted in the last election. This means that the employed were more likely to go to the polls by about a 7-point margin. These results were consistent with those found for political interest, indicating that unemployed voters are less politically involved. However, again there were some interesting cross-national differences. This gap did not exist in Belgium, Italy, Portugal, or Spain, where the unemployed and

Table 9.6. Turnout among Employed and Unemployed Citizens ( percentage voting in last election) Country

Employed

Unemployed

Difference

Belgium Britain Denmark France Germany Greece Italy Ireland Luxembourg Netherlands Portugal Spain

79.8 77.5 88.1 77.1 82.2 88.6 86.1 73.2 91.9 86.3 73.8 78.6

78.8 66.7 82.2 59.6 77.1 75.0 87.3 64.6 100.0 66.7 72.5 79.0

1.0 10.8 5.9 17.5 5.1 13.6 -1.2 8.6 -8.1 19.6 1.3 -0.04

1,003 1,078 1,000 1,000 2,134 1,002 1,067 1,000 462 1,005 1,000 1,000

Total population

81.2

74.4

6.8

13,096

Source: Eurobarometer surveys (1994).

284

Respondents

Desperate Times Call for Desperate Measures?

employed were equally likely to vote, nor in Luxembourg, where the unemployed actually were more likely to vote (the Luxembourg results should be approached with caution given the small number of cases). Yet, there also were sizable gaps in voting participation among employed and unemployed respondents in countries such as the Netherlands (19.6%), France (17.5%), Greece (13.6%), and Britain (10.8%). Thus, two groups of countries emerged from the analysis: one where the employed were no more likely to vote than the unemployed (Belgium, Italy, Portugal, Spain), and one where they were. Overall, the results thus far have shown that unemployed Europeans are somewhat less involved in political affairs than the employed. They participate less, are less interested, and less satisfied with their own life and the way the system works. However, they do not (or have not) become extreme in their political views. Instead, they appear to withdraw from the political process. While this is a general conclusion, there also were some noteworthy cross-national differences. The most prominent exceptions were Italy and Portugal. Italians in general were less interested in politics and less satisfied with the way democracy works. This was the case both among employed and unemployed voters. However, unemployed Italians also were more extreme and voted just as often as employed ones. Unemployed voters in Portugal also constituted an interesting exception because they were more extreme in their political views than employed ones, and they were significantly more interested in politics than the employed, but also just as likely to vote. Thus, both Italian and Portuguese unemployed were more likely to be engaged in the political process than unemployed voters in most other European countries.

Do Welfare States Make a Difference? Are these negative effects of unemployment on political attitudes and political behavior attenuated by public policies and institutions designed to cushion the ill effects of a market economy? After all, the history of the welfare state is closely associated with efforts to alleviate the negative outcomes associated with capitalism. If this is the case, unemployed voters in more highly developed welfare states should be less negatively affected than those in welfare states that are less highly developed. Moreover, the gap between employed and unemployed voters with regard to our measures of attitudes and behaviors should be smaller in more highly developed welfare states. 285

Christopher J. Anderson

Table 9.7. Political Attitudes and Behaviors among Unemployed Citizens by Type of Welfare State Welfare State Attitude/Behavior

More Extensive a

Less Extensiveb

Difference

Democracy satisfaction Life satisfaction Political extremism Xenophobia Political interest Turnout

54.35 64.30 15.63 50.98 37.72 77.40

34.53 56.53 19.12 34.27 36.55 74.18

19.82 7.77 -3.49 16.71 1.17 3.22

Source: Eurobarometer surveys (1994). a Based on 1994 government expenditure on social protection as a percentage of GDP (source: OECD). Includes Belgium, Denmark, France, Germany, Luxembourg, Netherlands. b Based on 1994 government expenditure on social protection as a percentage of GDP (source: OECD). Includes Britain, Greece, Italy, Ireland, Portugal, Spain.

As an indicator of welfare state development, I rely on data collected by the OECD on government expenditure on social protection as a percentage of GDP (1994). To investigate whether welfare state spending – that is, compensation for the losers of labor market competition – is associated with less pronounced negative effects on the unemployed with regard to political attitudes and behaviors, I calculated levels of attitudes/behaviors for states falling into different categories of welfare state spending, namely more or less extensive welfare states. According to the OECD figures, more extensive welfare states include Belgium, Denmark, France, Germany, Luxembourg, and the Netherlands, whereas the less extensive welfare states are Britain, Greece, Italy, Ireland, Portugal, and Spain. Table 9.7 shows the political behavior data for unemployed citizens summarized by type of welfare state, and Table 9.8 shows the differences between employed and unemployed citizens.

The Political Behavior of Unemployed Voters and Type of Welfare State Welfare states do make a difference, it seems, but not always in ways that were expected. As hypothesized, unemployed voters were significantly more satisfied with their own lives and the functioning of the political system in more extensively developed welfare states (Table 9.7). However, there is no significant difference among unemployed voters across differ286

Desperate Times Call for Desperate Measures?

Table 9.8. Differences in Political Attitudes and Behaviors among Employed and Unemployed Citizens by Type of Welfare State Welfare State Attitude/Behavior

More Extensive

Democracy satisfaction Life satisfaction Political extremism Xenophobia Political interest Turnout

14.77 25.50 -2.02 -7.77 11.55 6.83

a

Less Extensive b 10.01 20.75 -1.55 2.10 4.25 5.51

Source: Eurobarometer surveys (1994). a Based on 1994 government expenditure on social protection as a percentage of GDP (source: OECD). Includes Belgium, Denmark, France, Germany, Luxembourg, Netherlands. b Based on 1994 government expenditure on social protection as a percentage of GDP (source: OECD). Includes Britain, Greece, Italy, Ireland, Portugal, Spain.

ent types of welfare states with regard to political extremism, political interest, or turnout. Finally, in contrast to expectations, voters in extensively developed welfare states had significantly more xenophobic views (51–35%). When we compared political behaviors and attitudes among employed and unemployed voters in different kinds of welfare states (Table 9.8), we see that there are few significant differences that could be explained with the help of differences in welfare state spending on social protection. In fact, there are bigger differences among employed and unemployed with regard to democracy satisfaction, life satisfaction, and political interest, for example, in more extensive welfare states – a finding that is very much in contrast to prior expectations. That is, the differences among employed and unemployed with regard to political attitudes and behaviors were not diminished in more highly developed welfare states.

Discussion and Conclusion My investigation of the effects of unemployment on life and political satisfaction, ideology/extremism, political interest, and individual voting behavior across the twelve western and southern states of the European Union indicated that unemployed voters, by and large, were less satisfied with their own life and the performance of the political system, less 287

Christopher J. Anderson

interested in politics, and less likely to vote than employed voters. However, they were not significantly more likely to place themselves at the extreme ends of the left-right spectrum or to espouse extreme political views. Thus, while unemployed voters were more likely to tune out, be disaffected, and participate less compared with employed voters, they were not more prone to take extreme ideological stances. Although there was no obvious pattern in the cross-national differences between the employed and unemployed when their distributions were compared along the political behavior variables, I hypothesized that some of the differences may be driven by the presence or absence of a highly developed welfare state. Specifically, I expected the negative effects of unemployment to be attenuated the more highly developed the welfare state. The statistical analyses did not support this hypothesis. While welfare state spending had positive effects on some of the variables and was negatively associated with political extremism, these effects were found both for employed and unemployed voters. Moreover, the gap in attitudes and behaviors between employed and unemployed voters could not be accounted for with the help of state spending overall or on social protection. The results presented here, although preliminary, have implications for our understanding of European democracy: they point to alienation among unemployed voters, but not extremist attitudes and behaviors. In fact, challenges to system legitimacy are unlikely to come from the growing segment of unemployed or unemployed citizens. Instead, there appear to be signs of disintegration of civil society, with the unemployed less likely to be involved in the political process and less satisfied with their lives more generally. The analysis of differences across different kinds of welfare states showed few significant differences. Part of the reason for these small differences across more or less extensively developed welfare states may lie in the different origins and consequences of unemployment across countries. Specifically, unemployment may be a different economic, political, and social phenomenon in Southern European states, which differ from the rest of Europe in terms of the role of social networks and the role of families in cushioning the effects of unemployment. Such differences may impede efforts to devise EU-wide solutions for the unemployment problem (see Chapter 1). Although only a preliminary step toward a fuller understanding of this phenomenon, the study showed that the effects of unemployment on polit288

Desperate Times Call for Desperate Measures?

ical attitudes and behaviors are worth exploring in the face of continually rising unemployment figures across Europe. After a further mapping of the effects of unemployment on political attitudes and behaviors, future studies may wish to concentrate on the political and institutional as well as the individual-level determinants that drive differences in political behavior among employed and underemployed citizens. It is likely that more sophisticated analyses, which can account for alternative causes of differences in political behavior would show even more marked differences among employed and unemployed citizens. Moreover, it may be worth investigating whether there are other distinctions – such as between unemployed, part-time employed, and fully employed citizens – that are of political relevance. Unemployment has been, and continues to be, the number one political priority among European mass publics and, by implication, for European policy makers, both at the national and European levels. Understanding how it affects real people is an important subject that has yet to be fully explored. References Anderson, Christopher. 1995. Blaming the Government: Citizens and the Economy in Five European Democracies. Armonk, N.Y.: M. E. Sharpe. ———. 1996. Economics, Politics, and Foreigners: Populist Party Support in Denmark and Norway. Electoral Studies 15, 4: 497–511. Bakke, E. W. 1933. The Unemployed Man: A Social Study. London: Nisbet. Betz, Hans-Georg. 1994. Radical Right-Wing Populism in Western Europe. New York: St. Martin’s Press. Brenner, H. M. 1973. Mental Illness and the Economy. Cambridge, Mass.: Harvard University Press. Catalano, R., and D. Dooley. 1983. Health Effects of Economic Instability: A Test of the Economic Stress Hypothesis. Journal of Health and Social Behavior 24: 46–60. Crepaz, Markus. 1995. Consensus vs. Majoritarian Democracy: Political Institutions and Their Impact on Macroeconomic Performance and Industrial Disputes. Comparative Political Studies 29, 1: 4–26. Eisenberg, P., and P. F. Lazarsfeld. 1938. The Psychological Effects of Unemployment. Psychological Bulletin 35: 358–90. Keohane, Robert O., and Helen V. Milner, eds. 1996. Internationalization and Domestic Politics. New York: Cambridge University Press. Hibbs, Douglas. 1977. Political Parties and Macroeconomic Policy. American Political Science Review 71: 1467–87. Lange, Peter, and Geoffrey Garrett. 1985. The Politics of Growth: Strategic Interaction and Economic Performance in the Advanced Industrial Democracies, 1974–1980. Journal of Politics 47: 792–827. 289

Christopher J. Anderson Lewis-Beck, Michael S., and Martin Paldam, eds. 2000. Economics and Elections: Consequences and Conclusions. Special issue of Electoral Studies 19 (2–3). MacFadyen, Alan, Heather MacFadyen, and Nancy Prince. 1996. Economic Stress and Psychological Well-Being: An Economic Psychology Framework. Journal of Economic Psychology 17: 291–311. Norpoth, Helmut. 1996. The Economy. In Lawrence LeDuc, Richard Niemi, and Pippa Norris, eds., Comparing Democracies: Elections and Voting in Global Perspective. Thousand Oaks, Calif.: Sage, 299–318. Schlozman, Kay, and Sidney Verba. 1979. Injury to Insult: Unemployment, Class, and Political Response. Cambridge, Mass.: Harvard University Press. Verba, Sidney, and Normal Nie. 1972. Political Participation in America. Cambridge, Mass.: Harvard University Press.

290

10 The Politics of Unemployment THE SPANISH EXPERIENCE IN COMPARATIVE PERSPECTIVE

José María Maravall and Marta Fraile

This chapter defends two related theses.1 The first is that, if certain institutional conditions are absent, economic policies will follow an unanticipated path, with consequences for employment, wages, and social policies that will contradict the interests of the government and its initial support coalition. To quote Scharpf (1991: 161), “To be successful, an economic strategy must satisfy two requirements. It must be appropriate to the (constantly changing) economic conditions that actually obtain, and it must also be feasible under existing (but equally changeable) institutional arrangements.” There are no universal blueprints that can produce similar effects whatever the institutional setting; and it is obvious that governments can hardly implement policies under conditions of their own choosing. Unexpected outcomes and goals that turn out to be contradictory will force governments to make difficult choices. One of the lessons of the Spanish experience is that, besides the general difficulties faced by social democracy, specific conditions made it particularly difficult to successfully combine wages, jobs, and redistribution. Our second thesis is that if unemployment is high, perhaps as the unintended consequence of the economic policies of the government, then traditional policies of income maintenance and passive protection, focused on the main household provider, will not be easily replaced by alternative ones. In these circumstances, proposals for welfare policy reform will be incompatible with the demands of crucial social democratic constituencies, and governments of such orientation will have no incentives to carry them through, even at the cost of other preferences. This is particularly the case 1

The authors wish to thank, besides the editor, Carles Boix, Julio Carabaña, and Adam Przeworski for their comments on an earlier draft of the chapter.

291

José María Maravall and Marta Fraile

if the reforms reduce present entitlements, increase income inequalities, and erode the “social wage,” even for the sake of an uncertain future employment of spouses, sons, and daughters, and life-long opportunities of higher income. Contrary to arguments that point to political pressures from labor market “insiders” to maintain the status quo (Saint-Paul 1993, 1996), we locate such pressures among “outsiders” – that is, among the unemployed rather than the employed. Social democratic governments will hardly follow such reform proposals if they want to preserve electoral support that is crucial for their survival as long as social policies that mitigate hardship among the jobless can be financed. To support this thesis, we examine the political effects of unemployment on electoral support, citizens’ reactions to policies and to inequality, and the degree to which such reactions influence governments. We draw on survey data to examine the political consequences of vast unemployment. At the time of the Spanish general election of June 1986, over 2.9 million people were out of work. Former prime minister Adolfo Suárez wondered why the unemployed had not taken to the streets. One may indeed wonder why the unemployed did not show their discontent either in the streets or in the ballot box. During the fourteen years of social democratic government, support for the incumbent among the unemployed, although declining, was always greater than that for any of the opposition parties, even when it was eventually defeated in the general election of March 1996. Why did these voters without jobs exonerate the government from responsibility? Were there circumstances that mitigated their hardship? We shall argue that a network of social policies, together with ideology, helped the government retain this support. And if a pattern of policies is electorally important, incentives for reform will be weak. Our purpose is to discuss, given this experience of high unemployment, more general trends and new dilemmas having to do with policies: more particularly, different combinations of wages, employment, and redistribution. We also assess the political viability of proposals for welfare reform, particularly those that present the welfare state, rather than as a remedy, as a cause of vast unemployment, incapable of catering to new social needs. These proposals rest on the assumption that if enough jobs are to be provided by economies, labor markets and social policies will have to be drastically reformed, even at the cost of greater inequality. They therefore treat the two social democratic objectives of redistribution and employment as incompatible. 292

The Politics of Unemployment

The Difficulties of Virtue: The Trade-offs among Wages, Unemployment, and Welfare Our first thesis is that institutional conditions may lead to unintended policy mixes, and to policy outcomes that contradict the interests of a government and its support coalition. We first note the rapid expansion of unemployment in order to shed light on the conditions the new Partido Socialista Obrero Español (PSOE) government faced in 1982. Between 1975 and 1982, unemployment in Portugal rose from 4.4% to 7.2%, while in Spain it climbed from 4.5% to 16.3% (Commission Européenne 1995: Table 3). This increase was unparalleled both regionally in the South and on the continent as a whole. When the Spanish Social Democrats came to power in 1982, they did not question in some fundamental respects a policy paradigm that was to become increasingly dominant in the course of the decade: they thought that employment in an open economy depended in the long run on macroeconomic conditions of low inflation and a balanced budget, as well as on the competitive conditions of the supply side. This view departed substantially from that held by social democracy over the three decades that followed the Second World War; it was increasingly shared, however, by most social democratic parties in the 1980s. The reasons are well known: countries with social democratic governments, Keynesian policies, and neocorporatist formulas were now experiencing high unemployment and seemed unable to achieve the “virtuous” combination of growth cum redistribution. As capital markets became internationalized, national governments lost their autonomy in determining the level of interest rates. This had two dramatic consequences for traditional social democratic policies: first, high interest rates increased the cost of expansive fiscal policy with public deficits; second, only high rates of return of private productive investment could keep unemployment at low levels. The reasons were not just high wages and taxes, as neoclassical diagnosis would insist upon, but, more generally, the opportunity costs of productive investment. Unfortunately for social democrats and unions, in these new circumstances, employment, real wages, welfare transfers, and services could not be expanded simultaneously. The Spanish experience shows that orthodox macroeconomic management, regarding inflation, the budget, or the balance of payments, was an inevitable part of a social democratic economic policy. And any breach of macroeconomic discipline was to have dramatic effects: the only question was who would pay the costs. 293

José María Maravall and Marta Fraile

Although the difficulties for neocorporatism increased, concertation with unions remained a fundamental part of social democratic strategy. If inflation was to remain under control while public expenditures expanded, the only alternatives were wage moderation or a tight monetary policy that raised interest rates, discouraged private investment, and increased unemployment. As Scharpf has put it, “The availability of an incomes policy creates important additional degrees of freedom for macroeconomic policy makers. . . . [For] supply-side problems, economic policy would be well served by close cooperation between government and unions” (1991: 34–35). In order to tackle investment-gap and demand-constrained unemployment, as well as cost-push inflation, wage restraint could moderate prices and promote profits, while fiscal and monetary policies could ensure demand, help productive investment, and finance social policies. Thus, it was either wage moderation, or fewer jobs, or less redistribution. As a vast literature has shown, institutional conditions influence such outcomes. Thus, concertation and wage moderation are facilitated, on the one hand, by high rates of affiliation of the labor force to single, centralized, and encompassing trade unions; on the other, by governments with the disposition and capacity to deliver compensations (Schmitter 1979; Cameron 1984; Lange and Garrett 1985; Wallerstein 1990; Golden, Lange, and Wallerstein 1995, 1997). In such circumstances, the unions will have a rational self-interest in avoiding the macroeconomic consequences of high wage settlements in centralized bargaining. The same argument is valid in harder times: the supply-side-oriented wage policy demanded to restore profits, facilitate investment, and create jobs is easier when unions are encompassing, solidaristic, large, and centralized. Such unions, because of self-interest, will opt for the “self-exploitation” (Scharpf 1991: 254) of employed workers, rather than exploitation by capitalists, if it leads to more jobs and social policies. In contrast, concertation will be difficult when a plurality of small trade unions, representing the interests of “insiders” in the labor market and negotiating at a sectoral level, hardly internalize the consequences of high wage increases and growing unemployment. Unions will also lack incentives to moderate wages when domestic autonomy to decide fiscal and monetary policies has been curtailed and economic conditions are largely determined by exogenous factors (Scharpf 1987: 26–28, 1991: 7–12, 238–55, 269–75; Regini 1995: 37–38). If the level of employment lies largely beyond the control of domestic actors, intertemporal calculations about wages, jobs, and redistribution make little sense. 294

The Politics of Unemployment

The competitive conditions of the supply side could be promoted following different strategies. As Boix (1998) has argued, such strategies reflect the ideological preferences of governments. On the one hand, a government may undertake a “high-added-value” strategy: in order to improve products and production processes, the state will play an active role in promoting the productivity of the factors of production and the competitiveness of the economy; public investment in the education and training of the workforce and in physical infrastructures will be high. If successful, this strategy will simultaneously achieve high levels of wages, welfare provision, and employment. On the other hand, a government may choose a neoliberal competitive route: in that case, it will keep state intervention, public expenditures, and fiscal revenues at low levels; leave economic decisions to markets; eliminate wage rigidities (including minimum wage laws) which prevent labor markets from clearing; suppress generous unemployment compensations which reduce the incentive of searching for a new job; and promote favorable conditions for private investment. If successful, it will achieve high private profits, savings, and investment. The first route is more egalitarian, although the instrument is education rather than the “social wage.” The argument is well known: training may both combine employment and redistribution, as it widens occupational opportunities and eventually diminishes wage differentials; in contrast, a high “social wage” may redistribute income, but at the cost of employment, as high unemployment benefits that extend over time discourage the unemployed from getting new jobs that do not pay high enough wages to act as an incentive. While the distributive consequences of the two strategies are different, both are supposed to lead to new “virtuous circles” of growth and competitiveness. Whatever its distributive consequences, the first (“social democratic”) strategy requires specific institutional conditions. Otherwise, these will have to be created. And during the transition, the evolution of wages is particularly crucial for jobs, because in the short run, other things being equal, the competitiveness of the economy will depend on the price differentials of foreign and domestic goods, and wages will be the crucial cost of production. If price levels are constrained by competition and if wages rise too much, as happened in Spanish industry in the 1980s, firms will be unable to preserve or create jobs. If they are not, prices will go up, as was the case in the Spanish service sector over the same period. But a “social democratic” strategy that consists of wage moderation and limits to the “social wage” will not be easily accepted by trade unions. Resistances will 295

José María Maravall and Marta Fraile

increase when governments cannot guarantee intertemporal payoffs in terms of jobs, wages, and welfare in the future, owing to the high dependence of the performance of their economies on international circumstances. The same will happen when governments attempt to reform overregulated labor markets, in the hope that jobs will more readily follow upturns in the economic cycle. Concertation with unions and neocorporatist arrangements are difficult to achieve in such conditions. While this “valley of transition” is being crossed (that is, until the conditions for a social democratic supply-side strategy are created), the electoral demands of the support coalition of the government must not be at odds with this strategy. Otherwise, a government interested in its own survival, myopic or not, will abandon its course. When the PSOE won the general elections of 1982, 2.24 million people were unemployed. Jobs had disappeared at an annual rate of 2.0% between 1975 and 1982. The unemployment rate had gone up from 4.5 to 16.3% in that period. The existing economic and institutional conditions were very unfavorable for the social democratic strategy that we have examined. The technological backwardness of the economy made a competitive strategy of high added value hardly possible in the short term.2 And the stock of human capital was low, although the flow of educated and trained individuals had increased: thus, while nearly three out of every four people in the 20–24 age group had entered into secondary education, the proportion was only 15% for the 45–64 group (Fundación Argentaria 1995: 132). The labor market was extremely rigid, as under Francoism job protection was used to win workers’ passivity in the absence of democratic rights. Real wages had risen an annual average of 5.8% between 1961 and 1982, compared with 3.9% for Western Europe as a whole (Commission Européenne 1995: 162–63). The economy also had to provide jobs to an active population that was growing rapidly: it increased 18.1% over fifteen years, which meant an additional 2.36 million people over 16 years of age. Foreign migrations, which had taken over 2.4 million Spanish workers to other European labor markets between 1960 and 1975, had come to an 2

The following are a few examples of this technological backwardness of the Spanish economy. In 1974, total expenditure in R&D amounted to only 0.34% of GDP, and in 1982 it rose to only 0.47% (the European Community average was 1.2%). Business enterprise expenditure in R&D stood at 0.26% of GDP in 1984, against an EC average of 0.81%. In 1982, the number of national patent applications per 10,000 population was 2.69, while the EC average was 9.57. And the coverage ratio of the technological balance of payments was only 0.21 (OECD 1990: 17, 25, 47, 49).

296

The Politics of Unemployment

end. The participation rate of women in the labor market, which had been very low, was increasing: while it amounted to 28.7% in 1977, it went up to 37.2% in 1993, which meant an additional 1.99 million were either working or in search of a job. And the cohorts entering the labor market were growing: the 16–25 age group was 27% larger than the 26–35 group. To sum up, the conditions included an economy in deep crisis in the area of competitiveness, demographic trends that contributed to the deterioration of employment, technological backwardness, increasing requirements of capital per employed worker,3 and a workforce that lacked the education and training needed to ensure the short-term success of a highadded-value strategy. The institutional conditions for concertation were also absent. Two trade unions (the UGT and the CC.OO.) competed for the representation of the working class. Trade union affiliation, which had risen rapidly with democracy, was now falling. Yet, despite their organizational weakness, unions were still able to attract over 50% of the vote in elections to works councils, and 94% of representatives on such councils were elected on trade union lists. Workplaces that mobilized when unions called a strike employed nearly two-thirds of workers (Escobar 1995: 159–60). This type of mobilizational strength and a strategy of intermediate centralization of bargaining (at the sectoral/provincial level) was more conducive to conflict and strong wage demands than to concertation and wage moderation (Soskice 1990; Jimeno and Toharia 1994). In this institutional context, the management of the economy was basically very orthodox. Inflation was brought down from an annual rate of 14.6% to 4.9% between 1982 and 1995. The average annual variation of monetary circulation went down from 18.3% under the period of the Suárez government to 12.3% under the socialists. Exports of goods and services went up from 18.4% of GDP to 24.1% in the same period. The terms of trade, if we take 1980 as 100, improved to 116 in 1995. Foreign currency reserves measured in écus multiplied by four times in the decade from 1982 to 1991. The supply-side structural reforms tried to open the economy, transform uncompetitive sectors of production, and deregulate markets. The increasing international integration of the Spanish economy is reflected in the growing share of GDP of the sum of exports and imports of goods and services: they had amounted to only 26% in 1970, and to 29% in 1980, but reached 61% in 1995 (Consejo Económico y Social 3

Between 1974 and 1995 such capital requirements per employed worker doubled: they went from 6.23 million pesetas to 12.95 million (Espina 1997).

297

José María Maravall and Marta Fraile

1996). The reconversion of noncompetitive firms extended to one-third of Spanish industry, at a cost of $9.2 billion from public funds. The liberalization of markets had different purposes: to introduce greater competition in an overprotected financial system, to limit inflationary pricing in the service sector as a whole, to facilitate adjustments to external shocks, and to stimulate the creation of jobs in times of economic expansion. Thus, a first reform of the labor market in 1984 suppressed administrative restrictions to hiring and dismissing workers, and introduced more flexible contracts for temporary or part-time jobs. In some other crucial respects the government departed from neoliberal advice. First, as unemployment soared between 1975 and 1985, some economists argued that only if wages were cut by 30% would jobs be created in sufficient numbers. Rather than accept this supposed trade-off between wages and jobs, the government only sought to moderate the evolution of real wages, if possible by concertation with the unions. Real wages grew by an average 0.8% a year from 1983 to 1995, a slightly lower rate than that of Western Europe as a whole (1.0%). Second, its fiscal policies were expansionary, contrary to neoliberal recommendations, although the government did slow down the fast increase experienced by the public budget under the new democracy since the mid-1970s. Public revenues and expenditures went up: the first by 7.1 points of GDP, the second by 8.3 points. These increases were considerable by comparative standards: the corresponding figures for Western Europe as a whole were 1.3 and 2.7 (Commission Européenne 1995: 160–61, 212–15). This fiscal expansion served to finance the “social democratic” supply-side strategy cum redistribution. Public sector investment rose from 2.9 to 5.2% of GDP between 1982 and 1991; the budget for education increased from 2.8 to 4.5% of GDP in the same period; the number of people covered by vocational training programs went up from 67,965 in 1984 to 290,869 in 1992, while the share of GDP of active labor market policies rose from 0.22 to 1.03% between 1982 and 1991; expenditure on research and development as a proportion of GDP doubled, from 0.47 in 1982 to 0.90 in 1990 (Maravall 1997: 88–89; Ministerio de Trabajo y Seguridad Social: annual series). As for public programs of social protection, their budgets increased from 17.5 to 23.4% of GDP between 1980 and 1993: if we look at some of these programs, expenditures on health went up from 4.7 to 6.0%; pensions from 5.4 to 7.1%; unemployment subsidies from 2.7 to 4.8% (Fundación Argentaria 1995: 326–29). Thus, a rough calculation shows that the social democratic strategy added 11.2 points of GDP to public expenditure, of 298

The Politics of Unemployment

which 5.3 corresponded to state intervention on the supply side and 5.9 to social policies. This increase led to some restructuring of the public budget, which grew 2.9 points of GDP below the cost of this sum of social and supply-side policies. But, because structural reforms of public expenditure were limited and tough choices avoided, such expansive trends could only be preserved if fiscal revenues continued to grow and if economic conditions were favorable. The combination of a European economic expansion, cheaper oil prices, and the effects of economic policies led to rapid economic growth in the second half of the 1980s: it reached an annual average of 4.5%, compared with 3.3% in Western Europe as a whole. The number of jobs also increased by 1,468,300: the average annual rate of job creation in this period was 3.2%, which was far higher than the European annual average of 1.3%. But, owing to the reforms of the labor market, most of these jobs were temporary: their proportion increased from 15.8% of all contracts in 1987 to 33.5% in 1992. So, both greater labor market flexibility and growth over a seven-year period increased dualistic tendencies within the labor force, reinforced by the persistence of highly protective legislation for the other two-thirds of jobs.4 And, at the end of this period of economic growth, as we have already noted, unemployment still stood at 16.2%, against a Western European average of 7.6%. Table 10.1 provides information on the evolution of unemployment and the active population from the first election of the new democracy, won in 1977 by a center-right party (UCD), to the last electoral victory of the socialists (PSOE) in 1993, when growth had sharply come to an end, falling to 0.7% in 1992 and -1.1% in 1993, and unemployment had climbed again by 4.4 points in two years only. The table shows first the impact of demographic trends: between 1977 and 1993, the active population grew more than unemployment. Second, unemployment grew more under the five years of UCD rule (1977–82) than under the eleven years (1983–93) of social democratic government (1.51 million in the first period, against 1.42 million in the second), although the active population expanded much more in the second period (by 2.1 million under the PSOE, against 0.3 million under the UCD). Third, the incidence of unemployment over the whole period was particularly large among young people of both sexes (an increase of 875,200) and women (1.51 million). Fourth, this incidence varied considerably between the two political 4

Particularly due to the definition of “unfair” dismissals, and the cost of indemnity payments (Bover, Arellano, and Bentolila 1996).

299

300

Table 10.1. Active Population and Unemployment Trends, 1977–1993

Source: Instituto Nacional de Estadística, Encuesta de Población Activa, annual series.

The Politics of Unemployment

periods: under the UCD, young people suffered the cost to a much larger extent than under the PSOE (in the first case, they represented 54.9% of the increase in unemployment, but only 3.3% in the second). On the contrary, the number of women without jobs rose more under the socialist than under the conservative governments (they contributed 66.2% and 37.7%, respectively, to the total increase in unemployment). This was associated with the much faster growth of the participation of women in the active population under the PSOE government: while their numbers increased by 0.25 million under the conservatives, they went up by 1.74 million under the socialists. The new European economic crisis in the first half of the 1990s had a devastating effect on the labor market. Unemployment went up again, reaching 24.1% in 1994. It also increased in Europe as a whole, but only by 3.5 percentage points (it went from 7.6% in 1990 to 11.1% in 1994), or less than half the Spanish case. The standard interpretation was that the expansionary fiscal policies, the evolution of wages, and the inefficiency of the labor market were to blame. While the economic cycle turned downward, the budget continued to go up by an additional 3.3 points of GDP due to the postponement of reforms on public expenditures and to a further expansion of social policies. And as unemployment went up, with a segmented labor market and a trade union movement with no institutional incentives to restrain wage demands, these also went up by an annual average of 1.8% in real terms. In response, the government resorted to a tight monetary policy, which aggravated the recession, punished productive investment, and increased unemployment. When fiscal austerity was finally imposed, of the social democratic panoply the supply-side policies suffered much more from the cuts than the social policies. In hard times, the government tried to preserve the still-precarious welfare state, a “caring” and “redistributive” political image, and its core electoral support. But the whole social democratic strategy faced widespread criticisms: on the one hand, its supply-side policies needed too much time to be effective, as the country had only limited stocks of physical and human capital and an underdeveloped R & D system. On the other hand, its social policies had swelled the budget, and their design stimulated unemployment. So the dominant policy advice was (1) to balance the budget with structural reforms of public expenditures and with one-shot revenues from the privatization of public firms; (2) to liberalize sectors where protection and the absence of competition fueled inflationary pricing, and where wage inelasticity regarding unemployment was high; 301

José María Maravall and Marta Fraile

(3) to carry further the flexibility of the labor market; (4) to introduce drastic reforms in the welfare state. Some of these recommendations were adopted by the social democratic government of the PSOE in its last couple of years. It cut public expenditure by 2.7 points of GDP, real wage increases were limited to 0.4% per year, and the economy was further liberalized. And a new reform of the labor market was introduced in 1994, which decentralized collective bargaining, made indefinite contracts more flexible, and limited the use of temporary job offers. When the economy started to grow again, job creation was much more sensitive than in the economic recovery of the earlier decade: in 1985, a 2.3% annual rate of GDP growth still destroyed 0.9% of total employment; in 1995 a 3.0% rate of growth produced a 2.7% increase in jobs. Between 1994 and 1996, 362,000 jobs with indefinite contracts were created, whereas 45,000 temporary jobs disappeared (Espina 1997). But it was too late for the government to benefit from the economic recovery: after fourteen years in power, the socialists lost the 1996 general elections. To put it rather crudely, social democrats, after two decades of frustration over high unemployment, appeared to believe that easier dismissals facilitated the creation of jobs and became more orthodox in their fiscal policies. Over a long time, social demands from their support coalition had led to a rapid increase in public expenditure; because of institutional conditions, the government had been unable to attract trade union complicity; social democratic supply-side policies appeared to require a long time to produce effects, as the initial conditions were unfavorable. The result had eventually been a particular mix of social policies cum monetarism; both unemployment and welfare had expanded. As in pre-Keynesian times, social democrats ended up with no economic policies of their own: they adopted orthodoxy and combined it with social “remedialism.”

Unemployment and Welfare Dilemmas An active fiscal policy, without an incomes policy and under unfavorable conditions, ended up with an unintended outcome: monetary austerity and high unemployment. We now turn to the impact of such unemployment on income inequality and on the electoral support for the government. We have already mentioned the cost of unemployment benefits, which reached 4.8% of GDP and represented a considerable constraint on the expansion of other expenditure programs. 302

The Politics of Unemployment

At the beginning of the 1980s income distribution in Spain was considerably unequal by Western European standards. The share of total household disposable income among the bottom decile of the population was 2.7%. This share was the lowest in the Southern European countries, where the average was 3.0%,5 and lower than the average of 3.4% in fourteen Western European countries (Atkinson 1995: 53). Over time, the inequalities increased almost everywhere. Table 10.2 shows the trends in the countries of the European Union, as well as in the United States and Japan. Income inequality went up most sharply in Great Britain and Denmark. By contrast, inequalities decreased in Italy, Spain, Portugal, and Ireland. The trends in income distribution do not appear to have been related to unemployment: for instance, Spain and Ireland had greater equality with high unemployment, whereas Portugal had greater equality with low unemployment. Welfare state reforms were related to such trends. In the case of Spain, we have already examined the substantial increase in social expenditure. Half of the total amount had a redistributive content; that is, it deviated from the distribution of disposable income to the benefit of the lowerincome groups. This was the case with noncontributive unemployment subsidies, noncontributive pensions, grants and scholarships, public housing, health, social services, and primary and secondary education, which amounted to 48.5% of all social expenditure (Bandrés 1996: 135–46). As a consequence, while in 1980 this “indirect wage” represented 119% of the pretransfer monetary consumption of the poorest decile of households, by 1990 this figure had risen to 204% (Gimeno 1993: 105–7). Table 10.3 provides detailed data on the distribution of the major public expenditure programs by deciles of income in 1980 and 1990. It shows the percentage of the average household expenditure in each interval that these programs represented. Over the decade, public expenditure grew in each of them, with the exception of “economic services.” A consequence was that distribution through public expenditure (excluding the effect of taxation) consisted basically of a nonzero sum game: if we look at the two columns furthest to the right, everybody won between 1980 and 1990, 5

The national figures were 3.1% for Italy (1986) and Portugal (1980), 3.0% for France (1979), and 2.7% for Spain (1980). In the other European countries the shares were 4.5% in Finland (1987), 4.4% in the Netherlands (1983), 4.3% in Luxembourg (1985), 4.2% in Belgium (1985), 4.1% in Norway (1979), 4.0% in Sweden (1981) and Germany (1984), 3.5% in the United Kingdom (1979), 2.8% in Switzerland (1982), and 2.3% in Ireland (1987).

303

304

305 Table 10.2. Income Distribution and Unemployment, 1979–1992

306 Table 10.3. Distribution of Public Expenditure by Deciles of Income (Households), 1980–1990

The Politics of Unemployment

although the relative increases were greater as one goes down the income scale. Note that the redistributive impact of public health, pensions, and “other social expenditures” was vastly superior to that of public expenditure in education and training, which mostly favored the middle income groups,6 although important differences existed within educational subprograms (for instance, grants and scholarships had a considerably egalitarian effect, while the public financing of universities had not). Therefore, proposals to turn the welfare state into a gigantic educational program must be handled with care: they can have a devastating effect both on equality and on the material conditions of life of the poorest groups. The trends in unemployment and inequality were thus paradoxical. As unemployment rose dramatically, income distribution became more egalitarian, notwithstanding the limits that we have pointed out. In fact, unemployment seldom led to poverty. The key to this paradox lies in the compensatory effect of social policies and the support provided by families. Unemployment was suffered mainly by members of the family who were not the main provider. These were mostly the spouses, sons and daughters who had entered the labor market in the course of the decade, swelling the volume of the active population. This incidence of unemployment can be appreciated in Table 10.4. In two-thirds of the Spanish unemployed population between 1977 and 1994, other family members contributed at least one wage to the household budget. In most other cases, one member of the family received a pension or unemployment benefits. From 1975 to 1985, 25% of the unemployed lived in households with no income earner, but by 1994 this figure had fallen to under 10%. The average amount of transfers was above the poverty threshold: it was 111% higher in 1980, 21% in 1990.7 And the mean per capita income of the unemployed was the equivalent of almost two-thirds of that of the population as a whole (Toharia 1995). As Table 10.5 shows, 59% of the unemployed belonged to households in the lower half of the income scale; only 27% belonged to households in the two lowest deciles. Unemployment was not just concentrated in the poorest households, and it was not necessarily related to dramatic 6

7

This distributive impact of education was, however, largely the product of the changed social composition of the lower decile, which consisted mostly of aged people who did not use educational services. The figures in current pesetas were 253,400 in 1980, with a poverty threshold of 120,000; 421,960 in 1990, with a poverty threshold of 350,000 (Ayala, Martínez, and Ruiz Huerta 1996: 385).

307

José María Maravall and Marta Fraile

Table 10.4. Rate of Unemployment According to Position within the Household, 1980 and 1990

Table 10.5. Distribution of Unemployment and Unemployment Benefits by Deciles of Household Income

308

The Politics of Unemployment

material scarcity. Table 10.5 also shows, however, that, in general, people in the higher deciles were more protected by unemployment benefits when they lost their jobs. Paradoxically, however, 15% of those who received such benefits had these as their sole income, while still belonging to the richest 30% of the population. Internal differences among the unemployed were thus important, and for many, the experience of not having a job did not take them to the bottom of the income scale. This may be due to the availability of other sources of personal or family income. For the unemployed as a whole, family and welfare were the main safeguard against need. However, welfare provision became a major target of criticism. Wages and job security, as well as welfare remedialism, were presented as trade-offs with jobs. Neoliberal conceptions of such trade-offs between employment, wages, and welfare policies defended both a drastic deregulation of the labor markets and a rollback of the welfare states if jobs were to increase. These proposals reflected an emerging “policy paradigm” regarding the welfare state. Though this paradigm had influenced a panoply of OECD policy recommendations (Dell’Aringa and Ludovici 1997), greater equality and the protection from need provided by families and social policies appear to have prevented Spain from suffering the corrosive social and political consequences traditionally expected from high unemployment even though it had the highest unemployment rate of OECD countries. We now turn to a more careful analysis of these political effects.

The Political Consequences of Unemployment We have argued that the incentives for governments interested in their own survival to introduce reforms that modify the unemployment/inequality status quo will depend on the political reactions of their key constituencies, including the unemployed. Such incentives will be great if the negative reaction to high unemployment is strong, and that to increasing inequality is weak. On the contrary, incentives will be low if the political reaction to unemployment is mitigated by social policies, and that to a growing inequality is strong. Spain provides an ideal opportunity to examine such reactions to unemployment, inequality, and social policies. To quote Goldthorpe (1987: 382), “Although the attempt may be made to present such a situation as one that is beyond government’s control, and thus to induce a fatalistic response, high levels of unemployment 309

José María Maravall and Marta Fraile

must still be reckoned as in general rendering governments electorally vulnerable.” Hibbs (1987: 203–23, 229–37, 254–57) and Cheibub and Przeworski (1999) have verified this vulnerability in cross-national work, and individual national studies also show that unemployment damaged the electoral support of the executives: this has been the case in the United States (Meltzer and Vellrath 1975; Fair 1978), Mexico (Buendía 1996), Peru (Stokes 1996), Poland (Przeworski 1996), the former German Democratic Republic (Anderson, forthcoming), and Great Britain (Price and Sanders 1995). Among the unemployed voters in Spain, support for the government also declined, but it was still always greater than that provided to any of the opposition parties. Boix (1998: 145–55) has shown that, while in their first electoral victory of 1982 the socialists won the vote of 55% of the unemployed, in their last one in 1993 they attracted the vote of 35% of those who had lost their job and only 18% of those who were unsuccessfully looking for their first job. Evidence for the period 1989–95, examined by González (n.d.), also reveals a substantial erosion of electoral support for the government among the unemployed. Yet when the PSOE eventually lost the elections in 1996, a majority of the unemployed, 38%, still voted for the socialists, against 30% who chose the PP, the party that won that election. And while 40% of the unemployed declared that in no circumstance would they vote for the PP, only 15% expressed a similar rejection of the PSOE.8 Maravall (1997: 95–98) concluded, after studying survey data for 1993, that while a negative evaluation of general economic conditions detracted votes from the government, the experience of being unemployed did not. Mancha (1993: 143–51, 165–71), examining the period from 1977 to 1989, found no effect of unemployment on the vote at election time; however, evidence from surveys carried out between 1979 and 1990 revealed that the “popularity” of (that is, the vote intention for) the government was affected by the unemployment rates. Two explanations have been provided for such limited political effects. First, hardship from unemployment was mitigated by the protection offered by families and social policies, which, in the case of the latter, was attributed to both the regime and the government (Maravall 1997: 95–101; Boix 1998: 150). Second, a large section of the population (20% of the unemployed) 8

The data are from survey no. 2,207 of the Centro de Investigaciones Sociológicas, carried out in March 1996.

310

The Politics of Unemployment

believed that no other party would improve the performance of the government on this issue (Fraile 1996). We thus have inconclusive evidence on the electoral consequences of unemployment. Subsamples of unemployed people studied in surveys have generally been too small, the kinds of questions posed have been inconsistent, and the statistical techniques and the specification of the models have varied among studies. We must therefore examine more carefully additional empirical evidence, turning to the analysis of individual, rather than aggregate, data in order to contrast the effects on vote intention of being unemployed; of having a particular level of income; and of holding specific perceptions of inequality, ideological positions, and evaluations of economic conditions and of social policies. The survey data that we use were collected in April 1995.9 The economic context was then as follows: the unemployment rate stood at 22.9% of the active population; following the economic crisis of 1993–94, monetary and fiscal policies remained restrictive, although the government was trying to preserve social expenditure; the 1994 reform of the labor market was largely unpopular. And although the economy was growing again, the socialists were only a few months away from their defeat in the 1996 general election, after fourteen years in government. Thus our dependent variable is the vote intention for the PSOE government in April 1995.10 As for our independent variables, we assumed that being unemployed as well as being critical of the evolution of social inequalities reduced the probability of voting for the government, a fortiori for a social democratic one, normally expected to have employment and equality as policy priorities. We expected however that other factors mitigated this effect: a leftist ideology and a position in the lower income levels should reinforce partisan loyalties, while positive evaluations of the general economic conditions and on social policies should also diminish 9

10

The survey was carried out by the Centro de Investigaciones Sociológicas, to a representative nationwide sample of 4,000 adults. The questionnaire and the research design were drawn by José María Maravall. Voting intention was defined as 1 for the incumbent and 0 for the other parties. For obvious reasons, we excluded from the analysis those respondents who declared that they would abstain or were undecided as to whether to vote or not, or about which party to vote for. Such cases do not usually enter the final models since they tend not to provide opinions that were considered here as independent variables. We experimented with a model that included them and found that the number of cases was almost the same.

311

José María Maravall and Marta Fraile

hostility toward the government.11 We ran different models of logistic regression, and the best specification and goodness of fit that we found is shown in Table 10.6.12 We tried alternative models that included other social policies, such as health, pensions, and noncontributive social benefits. The evaluation of such policies made no difference on the vote intention. We thus kept unemployment benefits and education because they provided the best model from a statistical point of view. Besides, these two policies represent well the two traditional dimensions of income transfers and public services.13 We found that, although the unemployed were somewhat more critical in their evaluation of these social policies, the differences were not great. It was regarding unemployment benefits that the unemployed had substantially more negative views than the rest of the people. The level of education and the attribution of responsibilities to the government for reducing social inequalities were never significant, and we excluded them from the model. Table 10.6 shows that being unemployed had a significant effect on support for the government: the propensity to vote for the socialists decreased if somebody was unemployed.14 Perceptions of social inequali11

12

13

14

The codes of the independent variables included in the model can be found in the statistical appendix. The goodness-of-fit test (Hosmer and Lemeshow 1989) shows the following: c2 (8): 12.48, p > c2 = 0.1310; Pearson c2 (804) with 812 covariate patterns = 870.75; p > c2 = 0.058. From the residual analysis, we conclude that the model specification is right. Out of the 1,391 observations, eight problematic cases exist, four with Pearson residuals bigger than 4.9 and four with the highest value of the dBeta, Pregibon’s db (and therefore the most influential cases in the calculation of the coefficients of the model). The model was run without seven of the eight mentioned cases (the other one presented the highest values in both residual and dBeta) and the coefficients remained practically the same, except for the unemployment variable, which became more significant: going from 92% to a level of 95%. Finally, according to the test carried out, there was no multicollinearity between the variables included in the model. All these tests and information are available for the interested reader. Policy evaluations among citizens in general were better for social services than for income transfers: the contrasts between positive and critical views were 75% versus 15% for education, 76% versus 13% for social services for the poor, 69% versus 18% for health. As for income transfers, the variations were greater: 58% versus 24% for pensions, 36% versus 50% for unemployment benefits. Positive views among the unemployed were as follows: 75% for education and for social services, 65% for health, 54% for pensions. They were more critical regarding unemployment benefits: only 27% thought that they had improved, while 62% believed that they had worsened. We also tried alternative specifications to the contrast between the unemployed voters and the rest: in particular, we ran a model that distinguished between (1) voters who were unemployed, (2) voters who were employed, and (3) voters who were outside the labor

312

The Politics of Unemployment

Table 10.6. Unemployment and Vote for the Government Independent variables

Coefficients

Odds-ratio

INEQUAL UNEMPLOY SOCIOTROP INCOME IDEOLOGY BETUNEMPL BETEDU Constant

-0.51 (0.13)** -0.36 (0.20)* -0.58 (0.08)** -0.41 (0.05)** -0.32 (0.03)** 0.62 (0.13)** 1.09 (0.19)** 3.34 (0.43)**

0.60 0.70 0.56 0.66 0.72 1.86 3.00 —

Number of cases c2 (7) Pseudo R2 % correctly predicted

1,391 409.56** 0.23 77%

Note: Entries are maximum-likelihood estimates with Logit, and their associated standard errors are in parentheses. * Significant at the level of 95%. ** Significant at the level of 99%.

ties also had an impact on electoral support: when somebody thought that the former had increased, the probability of voting for the government would decrease. Thus, the experience of being unemployed and having critical views on social inequality had an impact on electoral intentions. However, other factors influenced the vote, particularly, ideology, income, as well as considerations about the state of the economy and about social policies. The more to the right an individual’s ideological position, the less likely that person was to vote for the socialist government. The higher the levels of income, the lower the probability of such a vote. When we looked for statistically relevant differences in the distribution of income between the employed and unemployed groups, we found only a slightly greater concentration of the latter group in the lower income level (less than 50,000 pesetas).15 So, besides being unemployed and having highly

15

market. As a categorical variable, we included it in the model as dummies, taking as reference those voters who were outside the labor market. Support for the government was somewhat higher among the latter; the difference was more marked compared with the unemployed voters. But the results were very similar to those of our final model, and the specification of the variable posed considerable sociodemographic problems. For this category, the Adjusted Residual (AR) is positive: +5. The AR for the third category (between 100,001 and 150,000 pesetas) is also positive: +3.5. The rest of the residuals are nonsignificant. Therefore, 10% of the unemployed were placed in the poorest category of income, against 6% of the total population.

313

José María Maravall and Marta Fraile

critical views on social inequalities, vote intention for the socialists declined as the voters’ ideology moved to the right and their income levels went up. We expected unemployment to reinforce egocentric economic judgments: not having a job should increase the concern for one’s personal economic conditions and diminish the relevance of sociotropic considerations. We also believed that such unemployed voters, owing to what Stokes (1996: 13) has called the “social comparison principle,” would not be likely to vote for the government. Yet egocentric views turned out to be of no statistical significance, and we excluded them from the model. Instead, sociotropic ones were relevant, so that the probabilities of voting for the government decreased as the voters’ retrospective economic assessments became more negative. But the incidence of sociotropic views among the unemployed is a potential source of exonerative voting: as it has often been argued (Fiorina 1981; Feldman 1982; Kiewiet 1983), sociotropic voting means that people do not attribute the responsibility for their own personal circumstances to the government. Table 10.6 also indicates that the evaluation of social policies had an important and significant impact on the vote for the government. When somebody thought that unemployment benefits had improved over the past ten years, the probability of voting for the PSOE increased. And the same applies to positive views regarding education. All of the preceding statements refer to the coefficients of the logit model. It is well known that the logit model translates into individual probabilities in a nonlinear, nonadditive manner (see Long 1997). Let us now move one step further and examine variations in real probabilities by means of simulations, in order to better assess the political consequences of being unemployed and of holding critical views on social inequality. Hence, we transform the coefficient estimates of Table 10.6 into probabilities, attributing hypothetical values to the independent variables included in the model (simulations). Suppose, for instance, that a voter was moderate (Voter 1) if he or she believed that social inequalities had not increased, held a centrist ideological position (5 over 10) and received a middle-level of income (4 over 10), was neither enthusiastic nor very critical of economic conditions, and was critical of social policies (that is, thought that education and unemployment benefits had worsened or, at least, not improved). In these circumstances, if Voter 1 was unemployed, the probability of voting for the PSOE was 12%; if he or she was not, this probability went up to 17%. Suppose 314

The Politics of Unemployment

now that a voter shared the same traits but believed that social inequalities had increased (Voter 2): if he or she was unemployed, the likelihood of voting for the government was 8%; if he or she was not, it rose to 11%. To what extent did ideology and views on social policies modify the combined effect of being unemployed and being critical of social inequalities? Imagine now a new voter (Voter 3) who shared all the traits of Voter 2, but who stood more to the left (in position 3): if he or she was unemployed, the probability of voting for the PSOE government was 14%; if he or she was not, it reached 19%. Consider finally another voter (Voter 4) who shared all the traits of Voter 2 (including ideology) but thought that unemployment benefits had improved over the past ten years: if he or she was unemployed, the probability of supporting the socialists was 20%; if he or she was not, the probability went up to 27%. If the same voter also had a positive view about education, the probability of voting for the government was 43% if he or she was unemployed, and reached 52% if she or he was not. Thus, although important differences remained between unemployed voters and the rest of the voters, ideology and views on social policies had a strong effect on the vote intention, which was similar in both cases when the rest of the variables take the values that we have attributed to them. We have explored further the effect of different combinations of these variables both for the unemployed and for the rest of voters. Figures 10.1–10.5 represent such combinations. We start with the combination of the effects of ideology and social policies. As Figure 10.1 indicates, the impact of ideology on the vote for the socialists varied when views on social policies (unemployment benefits and education) changed. That is, the effect of ideology was reinforced by views on social policies. The combined effect of leftism and positive opinions on unemployment benefits and education reduced the political consequences of unemployment on the vote for the government.16 The second combination consists of income levels and social policies. As Figure 10.2 reveals, both for voters who were unemployed and those who were not, the effect of income on the probability of voting for the government changed when views on social policies varied. In both groups 16

BETUNEMPL and BETEDUC are not related to IDEOLOGY. We ran a logistic regression for the first two variables, and ideology as an independent variable: ideology was not significant in the model. The correlation of the variables is also very low: IDEOLOGY and BETUNEMPL -0.05; IDEOLOGY and BETEDUC -0.18. Again, all these statistical results are available for the interested reader.

315

316

Figure 10.1 Effect of Ideology on the Probability of Voting for the Incumbent When Evaluation of Social Policies Changes (holding the rest of variables constant at their means for sociotropic and income, and unequal = 1). Circles indicate positive visions of social policies; triangles indicate negative visions of social policies.

317

Figure 10.2 Effect of Income on the Probability of Voting for the Incumbent When Evaluation of Social Policies Changes (holding the rest of variables constant at their mean values for sociotropic and ideology, and unequal = 1). Circles indicate critical visions of social policies; triangles indicate positive visions of social policies.

José María Maravall and Marta Fraile

of voters, the inverse relationship between levels of income and socialist vote increased when views on social policies are taken into account. An individual in the lower income groups was much more likely to vote for the socialists, but positive opinions on social policies increased this probability very dramatically both for those who were unemployed and for those who were not. Our next combination consists of ideology and income. Figure 10.3 shows that the two variables went in the same direction and reinforced their effect on the vote for the socialist government. This happened both for voters who were unemployed and for those who were not. The influence of ideology on the probability of such a vote varied with different levels of income, holding the remaining variables constant. Thus, such influence was greater when the levels of income were lower. We also examined the effects of sociotropic evaluations of the economy combined with social policies. As Figure 10.4 shows, opinions on social policies influenced the relationship between evaluations of the economy and the probability of supporting the government. When such opinions were favorable among unemployed voters, they increased the support for the government, reinforcing the effect of sociotropic economic considerations. The same happened for the rest of the voters. Finally, we looked at the combined effect of ideology, social policies, and the impact that views on social inequality had on the probability of voting for the government, again both for unemployed voters and for the rest. In Figure 10.5 we consider voters that changed such views: they could be more or less critical about what had happened with social inequality over the past ten years. We assumed that voters had a positive opinion on social policies. By contrast, their income levels and evaluations of the economy were indifferent regarding support to the government. Under these conditions, did the effect of ideology on the vote for the government change? As Figure 10.5 shows, benevolent views on social inequalities increased the probability of support for the socialists somewhat, but the shape of the curve that represents the effect of ideology remains the same. Thus, the negative impact that critical views of social inequalities had on the likelihood of voting for the government appears to have been mitigated by ideology and by positive opinions on social policies. And this pattern existed both for voters who were unemployed and for those who were not. So far, we have looked at the linear and additive effects of each independent variable on our latent dependent variable (let us call it Z), and we 318

319

Figure 10.3 Effect of Ideology on the Probability of Voting for the Incumbent When Level of Income Changes (holding the rest of variables constant at their means for sociotropic, and unequal, betunem, and betedu = 1). Circles indicate lowincome population (second category); triangles indicate high-income population (eighth category).

320

Figure 10.4 Effect of Sociotropic Evaluations of the Economy on the Probability of Voting for the Incumbent When Evaluation of Social Policies Changes (holding the rest of variables constant at their means, for ideology and income, and unequal = 1). Circles indicate positive visions of social policies; triangles indicate negative visions of social policies.

321

Figure 10.5 Effect of Ideology on the Probability of Voting for the Incumbent When Visions about Level of Social Inequalities Change (holding the rest of variables constant at their means for sociotropic and income, and positive visions on social policies). Circles indicate social inequalities have increased; triangles indicate social inequalities have decreased or remained the same.

José María Maravall and Marta Fraile

transformed these effects into the probability of voting for the government.17 We shall now test the hypothesis that being unemployed conditioned the effect of the rest of independent variables on the probability of voting for the PSOE. For this purpose, we tried different interactions: did being unemployed change the direct effect of ideology, income, or sociotropic evaluations on the probability of voting for the incumbent? None of the interactions turned out to be significant except that of income. But this effect must be taken into account in order to understand the nature of the relationship between being unemployed and the vote for the government. When we examined the relationship between the unemployment rate and the general levels of inequality, we noted that higher levels of unemployment did not translate automatically into higher levels of inequality (Table 10.5). Social policies mediated such a relationship. The effect of the level of social protection is shown in Table 10.7. The interaction between unemployment and income is indicated in the coefficient of UNE*INCO: a one-unit increase in the level of income of the unemployed increased the propensity to vote for the socialists, whereas a one-unit increase in the level of income of the rest of voters decreased the probabilities of voting for the government. Hence, the negative effects of having a high income on the vote for the socialist government were lower for those unemployed than for the rest of the population.18 Note also that 17

18

As the following equation shows, the logit model is linear in the logit: lnW (x) = xß. Nevertheless, the model is linear in terms of odds, but a constant factor change or amount of change in the odds does not correspond to a constant factor change or amount of change in the probability (see Long 1997). Note that we are testing the following hypothesis: the negative effect of being unemployed on Z (transformed into probabilities of voting for the PSOE) is lower when the income of the unemployed is higher. Hence the specification of the interaction takes this form: Z = b0 + g1X1 + b2X2 + . . . . + ei where g1 = b1 + b3X1 * X2. The inclusion of multiplicative terms in multiple regression equations has been criticized on the grounds that the results are difficult to interpret, unreliable (as a result of multicollinearity between the multiplicative term and its constituent variables), and even meaningless. Such criticism is wrong. Coefficients in models with multiplicative terms must be understood as the conversion of a general statement of relationship into a conditional one. Therefore, high standard errors are related to problems of statistical efficiency (because what is being calculated are coefficents for small groups of cases), not to problems of multicollinearity (see Friedrich 1982; Jacard, Turrisi, and Wan 1990).

322

The Politics of Unemployment

Table 10.7. Unemployment Protection and Vote for the Government Independent variables

Coefficients

Odds-ratio

INEQUAL UNEMPLOY SOCIOTROP INCOME IDEOLOGY BETUNEMPL BETEDU UNE*INCO Constant

-0.53 (0.13)** -1.06 (0.36)** -0.57 (0.08)** -0.46 (0.05)** -0.32 (0.03)** 0.62 (0.13)** 1.12 (0.19)** 0.40 (0.16)** 3.46 (0.43)**

0.58 0.23 0.56 0.63 0.72 1.86 3.06 1.49 —

Number of cases c2 (8) Pseudo R2 % correctly predicted

1,391 415.8** 0.24 78%

Note: Entries are maximum-likelihood estimates with Logit, and their associated standard errors are in parentheses. * Significant at the level of 95%. ** Significant at the level of 99%.

when the impact of those unemployed with the highest income is isolated in the interaction, the effect of being unemployed, for those having the lowest incomes, becomes statistically more significant.19 This is further evidence that when the income of the jobless was protected, the political effects of unemployment were mitigated. The government was punished only by the unemployed in the lowest deciles of the income scale. To sum up, this analysis shows that being unemployed as well as holding critical views on the evolution of social inequalities reduced the probability of voting for the socialist government. It also reveals that, to a considerable extent, this impact was mitigated by the influence of ideology, income levels, and evaluations of the general economic conditions and of 19

Note that the coefficient for being unemployed was significant in the first model of Table 10.6 with a level of 92%, whereas in this new model the level reaches 99.9%. As for the coefficient for those having the second lowest income, the effect is also significant at the level of 99% (results are also available).

323

José María Maravall and Marta Fraile

social policies. The evidence seems to indicate that trade-offs between social policies, inequality, and unemployment were hardly perceived by citizens. When they thought that inequalities had increased or were unemployed, the likelihood that they would vote for the socialist government went down sharply. However, when views on social policies were positive, the government extracted important electoral benefits, both among voters in general and among those who were unemployed. Moreover, when those who were unemployed had their incomes protected (either through social policies or other incomes in the family), the odds that they would vote for the socialist governments increased. Therefore, these mechanisms limited the negative political consequences of unemployment and of critical views on the evolution of social inequalities.

Conclusions The Spanish experience provides an excellent case of the new challenges faced by European social democracy as well as the growing policy dilemmas about how best to combine competitiveness and growth, wages, jobs, and welfare. The government tried to combine an orthodox macroeconomic management with active supply-side policies. But we have argued that two factors made the success of this strategy difficult: one was the set of economic constraints that made a “high-value strategy” almost impossible; the other was a trade union movement unwilling to make wage sacrifices. For a long time, a high rate of economic growth facilitated the postponement of difficult choices regarding costs. But as soon as economic conditions (and fiscal resources) deteriorated, the government had difficulties in preserving the distinctive elements of the social democratic supply-side strategy. Policy recommendations of greater labor market flexibility, lower labor costs, fiscal austerity, and strict monetary management became increasingly influential. When the crunch came, the social democratic government chose to preserve social expenditure and redistributive policies. This choice was institutionally framed: it was largely due to the failure to reach agreements with unions regarding wages and labor market reforms; it led to a tight monetary policy within the European Monetary System and with an independent central bank. Monetary orthodoxy and expanding social welfare were the eventual, unintended policy mix of the socialists. This outcome was at the cost of jobs. Yet, on the one hand, even under vast unemployment, income inequalities decreased; on the other, social policies helped the gov324

The Politics of Unemployment

ernment to survive electorally. While support for the government was lower when a voter was unemployed or was highly critical of social inequalities, positive views of social policies, as well as comparatively higher levels of income among those unemployed due to welfare protection, mitigated or even reversed these effects. If this conclusion is correct, initiatives that would increase social inequalities or dismantle popular social policies would make a government electorally vulnerable, at least one of a social democratic persuasion. As a consequence, incentives for policy reform will be weak and unemployment will remain high unless unions are willing to sacrifice wages and agree on greater labor market flexibility in order to achieve a different combination of redistributive social expenditure and jobs.

Statistical Appendix Codification of the Variables Included in the Model 1. INEQUAL is a dummy variable with a value of 1 for those who thought that in the past ten years social inequalities had increased and 0 for those thinking that they had diminished or remained the same. 2. UNEMPLOY is also a dummy variable with a value of 1 for those being unemployed and 0 for the rest of the population. 3. SOCIOTROP expresses an evaluation of the economic situation of the country that may go from 1 (very good) to 5 (very bad). 4. IDEOLOGY refers to self-placement on a continuum that goes from 01 (extreme left) to 10 (extreme right). 5. INCOME represents the household monthly income, going from 00 (less than 50,000 pesetas) to 09 (more than 1 million pesetas). The minimum wage in Spain in January 1995 was 62,700 pesetas (Anuario El País, 1997). 6. BETUNEMPL is a dummy variable that has a value of 1 for those considering that unemployment benefits had improved during the past ten years (under the socialist government) and 0 for those thinking that they had worsened or remained the same. 7. BETEDUC is another dummy with a value of 1 for those who believed that education had improved during the past ten years and 0 for those who thought that it had worsened or remained the same. 325

José María Maravall and Marta Fraile

Descriptive Statistics of the Variables Included in the Model Variable

Number of observations

Mean

Standard Deviation

Minimum

Maximum

IVPSOE INEQUAL UNEMPLOY SOCIOTROP INCOME IDEOLOGY BETUNEMPL BETEDU

2,439 3,575 3,972 3,868 2,901 3,133 3,470 3,627

0.32 0.50 0.13 3.83 2.18 4.74 0.36 0.76

0.47 0.50 0.33 0.85 1.46 2.00 0.48 0.43

0 0 0 1 0 1 0 0

1 1 1 5 9 10 1 1

References Anderson, C. J. Forthcoming. The Economy and Public Opinion in East Germany, 1991–95. In Public Opinion and Economic Reforms in New Democracies, ed. Susan C. Stokes. New York: Cambridge University Press. Atkinson, Anthony B. 1995. Incomes and the Welfare State. Cambridge: Cambridge University Press. Anuario El País. 1997. Madrid: PRISA. Ayala, Luis, Rosa Martínez, and Jesús Ruiz Huerta. 1996. La Distribución de la Renta en España desde una Perspectiva Internacional. In Fundación Argentaria, La Desigualdad de Recursos. Madrid: Visor. Bandrés, Eduardo. 1996. ¿A Quién Beneficia el Estado de Bienestar? In Fundación Argentaria, Dilemas del Estado de Bienestar. Madrid: Visor. Boix, Carles. 1998. Political Parties, Growth, and Equality. New York: Cambridge University Press. Bover, Olympia, Manuel Arellano, and Samuel Bentolila. 1996. Duración del Desempleo, Duración de las Prestaciones y Ciclo Económico. Estudios Económicos 57, Servicio de Estudios, Banco de España. Buendía, Jorge. 1996. Economic Reform, Public Opinion, and Presidential Approval in Mexico, 1988–1993. Comparative Political Studies 29, 5. Cameron, David. 1984. Social Democracy, Corporatism, Labour Quiescence, and the Representation of Economic Interests in Advanced Capitalist Society. In J. H. Goldthorpe, ed., Order and Conflict in Contemporary Capitalism. Oxford: Oxford University Press. Cheibub, José Antonio, and Adam Przeworski. 1999. Democracy, Elections, and Accountability for Economic Outcomes. In Adam Przeworski, Bernard Manin, and Susan C. Stokes, eds., Democracy, Accountability, and Representation. New York: Cambridge University Press. Commission Européenne. 1995. Economie Européenne 60. 326

The Politics of Unemployment Consejo Económico y Social. 1996. Economía, Trabajo y Sociedad. Madrid. Deininger, Klaus, and Lyn Squire. 1996. A New Data Set Measuring Income Inequality. World Bank Economic Review 10, 3. Dell’Aringa, Carlo, and Manuela Ludovici. 1997. Policies for the Unemployed and Social Shock Absorbers. The Italian Experience. In MIRE-Florence, Comparing Social Welfare Systems in Southern Europe. Paris. Escobar, Modesto. 1995. Spain: Works Councils or Unions? In Joel Rogers and Wolfgang Streeck, eds., Works Councils. Chicago: Chicago University Press. Espina, Alvaro. 1997. El Sistema de Relaciones Industriales y la Concertación Neocorporativista en España: 1977–1997. Manuscript. Fair, Ray C. 1978. The Effect of Economic Events on Votes for President. Review of Economics and Statistics 60. Feldman, Stanley. 1982. Economic Self-Interest and Political Behavior. American Journal of Political Science 26, 3. Fiorina, Morris P. 1981. Retrospective Voting in American National Elections. New Haven, Conn.: Yale University Press. Fraile, Marta. 1996. La Introducción de “Temas Relevantes” en el Modelo de Explicación del Voto al PSOE en las Elecciones de 1982 y 1993. Manuscript. Juan March Institute. Friedrich, Robert. 1982. In Defense of Multiplicative Terms in Multiple Regression Equations. American Journal of Political Science 26, 4. Fundación Argentaria. 1995. Las Desigualdades en España. Madrid: Visor. Gimeno, Juan Antonio. 1993. Incidencia del Gasto Público por Niveles de Renta. In Fundación Argentaria, Sector Público y Redistribución, I Symposium sobre Igualdad y Distribución de la Renta y la Riqueza, Madrid. ———. 1996. Los Beneficios del Gasto Público. In Fundación Argentaria, Las Políticas Redistributivas. Madrid: Visor. Golden, Miriam, Peter Lange, and Michael Wallerstein. 1995. The End of Corporatism? In Sanford Jacoby, ed., The Workers of Nations. Industrial Relations in a Global Economy. New York: Oxford University Press. ———. 1997. Unions, Employers’ Associations and Wage-Setting Institutions in North and Central Europe, 1950–1992. Industrial and Labor Relations Review 50, 3. Goldthorpe, John H. 1987. Problems of Political Economy after the Post-War Period. In Charles S. Maier, ed., Changing Boundaries of the Political. New York: Cambridge University Press. González, Juan Jesús. N.d. Algunas Observaciones sobre el Declive Electoral del PSOE. Manuscript. Universidad Nacional de Educación a Distancia/Centro de Investigaciones Sociológicas. Hibbs, Douglas A. 1987. The Political Economy of Industrial Democracies. Cambridge, Mass.: Harvard University Press. Hosmer, David W., and Stanley Lemeshow. 1989. Applied Logistic Regression. New York: John Wiley & Sons. Instituto Nacional de Estadística. Encuesta de Población Activa. Annual series. Jaccard, James, Robert Turrisi, and Choi K. Wan. 1990. Interaction Effects in Multiple Regression. London: Sage. 327

José María Maravall and Marta Fraile Jimeno, Juan Francisco, and Luis Toharia. 1994. Unemployment and Labor Market Flexibility: Spain. Geneva: International Labor Organization. Kiewiet, Roderick. 1983. Macro-Economics and Micro-Politics: the Electoral Effects of Economic Issues. Chicago: University of Chicago Press. Lange, Peter, and Geoffrey Garrett. 1985. The Politics of Growth. Journal of Politics 47, 3. Long, Scott. 1997. Regression Models for Categorical and Limited Dependent Variables. Advanced Quantitative Techniques in the Social Sciences Series, 7. Beverly Hills: Sage Publications. Mancha, Tomás. 1993. Economía y Votos en España. Madrid: Instituto de Estudios Económicos. Maravall, José María. 1997. Regimes, Politics, and Markets. Oxford: Oxford University Press. Meltzer, Allan, and Marc Vellrath. 1975. The Effects of Economic Policies on Votes for the Presidency: Some Evidence from Recent Elections. Journal of Law and Economics 18. Ministerio de Trabajo y Seguridad Social. Anuario de Estadísticas Laborales, 1984 to 1992. Annual series. Madrid. Organisation for Economic Co-operation and Development (OECD). 1990. Main Science and Technology Indicators. Paris. Price, Simon, and David Sanders. 1995. Economic Expectations and Voting Intentions in the UK, 1979–87: A Pooled Cross-Section Approach. Political Studies 43. Przeworski, Adam. 1996. Intertemporal Politics: The Support for Economic Reforms in Poland. Comparative Political Studies 29, 5. Regini, Marino. 1995. Uncertain Boundaries. Cambridge: Cambridge University Press. Saint-Paul, Gilles. 1993. On the Political Economy of Labour Market Flexibility. Discussion Paper 803. London: Centre for Economic Policy Research. ———. 1996. Exploring Labor Market Institutions: A Political Economy Analysis. Document 96-03. Paris: DELTA and CERAS. Scharpf, Fritz W. 1987. The Political Calculus of Inflation and Unemployment in Western Europe. Center for European Studies Working Paper 6. Cambridge. ———. 1991. Crisis and Choice in European Social Democracy. Ithaca, N.Y.: Cornell University Press. Schmitter, Philippe C. 1979. Still the Century of Corporatism? In Philippe C. Schmitter and Gerhard Lehmbruch, eds., Trends Toward Corporatist Intermediation. London: Sage. Soskice, David. 1990. Wage Determination: The Changing Role of Institutions in Advanced Industrial Countries. Oxford Review of Economic Policy 6, 4. Stokes, Susan C. 1996. Public Opinion and Market Reforms: The Limits of Economic Voting. Comparative Political Studies 29, 5. Toharia, Luis. 1995. Unemployment in Spain: How Can It Be So High? Manuscript. Universidad de Alcalá. Wallerstein, Michael. 1990. Centralised Bargaining and Wage Restraint. American Journal of Political Science 34. 328

11 Conclusion UNEMPLOYMENT, THE NEW EUROPE, AND THE OLD INEQUALITIES

Nancy Bermeo

Unification has the nasty habit of creating division, and opinions are certainly divided over the merits of a united and liberalized European market. Where some fear that the “surf of market forces” will turn the New Europe into “post-modern flatlands” (Gowan and Anderson 1997: xiv), others are convinced that marketization is the only means of ensuring economic buoyancy. The essays in this collection are reactions to both these perspectives. They are also reactions to the coincidence of high unemployment with the high-gear drive toward monetary and economic union. Though this research project began with no political agenda, the sum of its parts yields a message with important implications for employment policy: the organized managed economies of Europe should be reformed but not replaced. A united Europe should be wary of modeling itself on the United States. I defend this argument by integrating evidence from the collection’s essays with a closer and more critical look at the workings of the paradigmatic example of the liberal model in the United States. I begin with a review of what the consequences of high unemployment are supposed to be and then illustrate the extent to which these negative consequences have been avoided in Western Europe. I then move on to link the mitigation of unemployment’s worst effects to the bedrock institutions of the organized managed economy. I conclude by setting the European experience in comparative perspective and arguing that the adoption of the U.S.-style liberal model is both unlikely and unwarranted.

329

Nancy Bermeo

The Consequences of Unemployment Unemployment is associated with a broad range of negative consequences – poverty being only the most obvious. Britain’s William Beveridge saw it as a source of moral “evil” and the breeding ground for “hatred and fear” (Allen and Waton 1986: 16). Many others have shared this view and drawn empirical connections between unemployment and trouble of all sorts. These include personal consequences that lie beyond the purview of this essay, such as psychological distress, marital breakup, and even attempted suicide (Gallie 1994; Whelan 1994; Allen and Waton 1986: 14), but they also include a wide range of consequences that bear directly on the workings and quality of political life. Unemployment is frequently associated with an affinity for extremist political parties of both the antidemocratic left (Evans 1987: 16–18; McElligott 1987: 253) and the antidemocratic right (Betz 1994: 189; Braun 1997: 6; Givens 1997; Kagedan 1997: 130). Even when the unemployed are not active supporters of extremist parties, they have been found to be a “receptive” constituency for extremist mobilization (Kitschelt 1995: 278), and to be overrepresented in xenophobic protest movements (Lipset and Rabb 1970: 156, 177; Verba and Scholzman 1997: 295). Though the unemployed have been found to be overly active on the extremes of the political spectrum, they have been found to be excessively passive in democratic political life more generally. Unemployment is thought to bring on “a withdrawal from community life” (Verba and Scholzman 1997: 294), a decline in sociability (Gallie 1994), and a decline in sense of efficacy (Cohn 1978). Together these factors are often associated with a decline in voting, and with low rates of political participation more generally (Milbrath and Goel 1977; Rosenstone 1982). The negative effects of unemployment are not confined to the unemployed themselves. A number of studies have suggested that citizens vote according to their evaluation of how incumbents have affected “collective” economic well-being (Lewis-Beck 1988: 155). Comparative research has suggested that the unemployment rate is “highly predictive” of “opposition vote shares,” (Lewis-Beck 1998: 9), that people will reject both the government and the policies they associate with joblessness (Hibbs 1987; Przeworski 1996: 540–41), and that “unemployment is such a catastrophic event that when people think the probability of losing their job is high, they interpret this as unambiguously bad and hold the government responsible” (Stokes 1998: 20). At least one researcher has found that “even short330

Conclusion

term evaluations of government [economic] performance” in the realm of unemployment and elsewhere “hold considerable sway over levels of political alienation” (Lockerbie 1993: 291). If these findings can be generalized, they suggest that persistent unemployment should be associated with high governmental turnover, high levels of alienation, low levels of participation, and high levels of xenophobia and extremism. Happily, contemporary Western Europe seems to be avoiding at least some of these negative consequences. As Anderson’s research suggests, the unemployed are no more given to xenophobia and extremism than the employed. In Western Europe as a whole, the unemployed were slightly less interested in politics and slightly less likely to vote, but in specific countries, the opposite pattern prevailed. These observations have been confirmed by other research (Amoretti 2000). Maravall and Fraile show, similarly, that the consequences of high and persistent unemployment in Spain were not nearly as negative as the literature would lead us to expect. Neither the extreme right nor the extreme left reaped dramatic benefits from high unemployment and as a consequence, support for Spain’s relatively new democratic regime did not decline (Maravall 1997). Though the unemployed were less likely to vote Socialist at the end of the party’s fourteen years in power than at their beginning, a majority of the unemployed were still voting Socialist in 1996 when the party lost the general election. People seemed to “exonerate the government from responsibility” (Chapter 10). The chapters on trade unions in this volume show that the effects of unemployment on at least one form of associational life are generally negative but not uniformly bad. On the negative side, three chapters (4, 6, 7) show how precarious employment has significantly lowered levels of involvement with unions in a variety of states. Yet, looking at the question of density specifically, Scruggs and Lange (Chapter 5) paint a more mixed and positive picture. Despite the onset of rising unemployment, union density across sixteen countries was still higher in 1989 than it had been in 1970. Rising unemployment had no effect on trade union membership where union density was already high; and in certain institutional contexts, unions actually expanded when unemployment rose. This occurred in the cases of Belgium, Denmark, Finland, and Sweden. Western and Healy (Chapter 5) provide a reminder that even where unemployment has had a deleterious effect on union density, huge sectors of the workforce are still covered by collective bargains anyway. In every 331

Nancy Bermeo

country but Switzerland, over three-quarters of the workforce has wages that are covered by collective agreements regardless of union density (Conover and Feldman 1986: 53). Unemployment’s effects on both wages and representation are highly variable. Evaluated as a whole, the preceding essays illustrate that the consequences of unemployment vary greatly across different institutional contexts. The unreformed institutions of the organized managed economies have certainly created inefficiencies, but they have also had salutary effects. The welfare state, the institutions of corporatist bargaining, and the public sector more generally have helped to mitigate the worst consequences of high unemployment. Yet, as Cameron (Chapter 1) and Hall (Chapter 2) both make clear, each of these institutional configurations is currently being threatened by the drive toward economic liberalization and European Monetary Union. As policy makers evaluate the costs of maintaining these institutional configurations, they should also evaluate their benefits. If not, as Hall warns, they may lose more than they gain. The benefits of using the welfare state to mitigate the worst consequences of unemployment are illustrated by the Spanish case. Despite its extraordinary scope and duration, unemployment in Spain has “seldom led to poverty.” In fact, “income distribution” actually “became more egalitarian” as unemployment rose (Chapter 10). This was not the result of a process of leveling down. In fact, average annual GDP per capita rose 2.6% in the decade of Spain’s highest unemployment. The avoidance of poverty was partly attributable to the welfare function of the Spanish family (Bermeo 2000), but both the avoidance of poverty and the avoidance of greater inequality owed a great deal to investment in the Spanish welfare state. Determined to make the Spanish economy more competitive in preparation for the many challenges of economic integration, the Socialist Party deliberately allowed unemployment to rise, but it attempted to compensate the unemployed in a variety of ways. The ruling party expanded existing unemployment benefit programs: replacement values rose, the length of time for insurance was extended, and the percentage of registered unemployed receiving unemployment compensation went from under 40% to over 80% in the party’s first decade in office (Toharia 1997: 158). The party also expanded programs for the training and placement of the unemployed. In the Socialist’s first decade in power, the number of people in vocational training programs more than quadrupled, the share

332

Conclusion

of GDP in active labor market policies increased nearly 500%, and spending on education (as a percentage of GDP) rose from 2.8 to 4.5%. The expansion of the Spanish welfare state had redistributive (i.e., inequality-reducing) effects in that fully half of spending expansion went to lower income groups. In the decade beginning in 1980, the “indirect wage” received by the poorest tenth of Spanish households through various welfare programs rose by over 70% (Chapter 10). The fact that Spain managed to mitigate the inequality and disaffection that might have been the by-products of high unemployment had positive effects on the consolidation of Spanish democracy. Christopher Anderson’s research (Chapter 9) shows that the size of the welfare state has significant effects on the beliefs of the unemployed. The jobless who live in the more advanced welfare states are much more likely to be satisfied with their democracy and with their lives more generally (Amoretti 2000). The connection between state activity, unemployment, and democratic consolidation is illustrated in another way by the Portuguese case. The 1974–75 revolution left Portugal with a huge public enterprise sector. Worker mobilization and a brief period of communist rule had led to the nationalization of a broad range of enterprises, including all domestic banks, insurance firms, and basic industries. As the first freely elected government in over forty years began its negotiations for entrance into the European Community, an estimated 53% of the country’s fixed industrial investment was under state control. Despite strong external pressures (and three IMF-inspired austerity programs), Portugal’s policy makers did not restructure public enterprises until 1990 (Torres 1994). The pace of restructuring was delayed primarily by Article 83 of the Portuguese Constitution, which prohibited privatization. Its amendment required a twothirds majority vote in the legislature, and this was not secured until late in 1988. The delay in privatization (and in coping with overmanning more generally) meant that a substantial sector of Portugal’s working class enjoyed a safe haven from the ravages of unemployment. Their protection had high economic costs, but the political costs of an earlier restructuring might have pushed Portugal’s new democracy beyond its limits. As Stoleroff (Chapter 6) reminds us, Portuguese trade unions once enjoyed formidable mobilizational power, and the public enterprise sector was the center of the most radical trade union activity. Exposing its workforce to

333

Nancy Bermeo

a liberalized market earlier in the democratization process might well have led to destabilizing extremism – especially since the budgetary legacies of the revolution made the Spanish option of expanding the welfare state untenable. Instead, the delayed reform of public enterprise in Portugal had serendipitous results. Privatization and restructuring began after accession to the European Union, after the economy was attracting increased investment, and after the private sector was able to absorb many of the workers released from restructured firms. As result, Portuguese unemployment has consistently been below the EU average throughout the 1990s. Indeed, for much of this period, it compares quite favorably with that of the United States. The absorption of excess labor in the public sector certainly had economic costs (and they were often paid by the workers themselves), but it appears that the long-term functioning of the Portuguese labor market was not compromised. The Dutch case is very different from the Spanish and the Portuguese, but it also exemplifies how the bedrock institutions of the organized managed economy perform positive functions. As Steven Wolinetz makes clear (Chapter 8), the Dutch engineered a dramatic reversal of skyrocketing unemployment. Though it was dubbed “the most spectacular employment failure in the advanced capitalist world” in the mid-1980s (Therborn 1986: 152), the Netherlands began the 1990s with an unemployment rate of only around 6% – slightly lower in fact than that of the United States. The Dutch “job machine” created new positions at a rate of 1.6% annually at a time when the average for the European Union as a whole was only 0.4% (Visser and Hemerijck 1997). Though part-time and flexible jobs account for over 70% of job growth, there is widespread popular support for the “one and a half jobs” per family model. Indeed, life satisfaction in the Netherlands is consistently among the highest in all of Europe (European Commission 1998: 81). How was joblessness reversed? The Dutch miracle, as it came to be called, rested on the reform and not the replacement of the organized managed economy. As Wolinetz illustrates, a renovated form of corporatist bargaining was fundamental to what went on. Unions decided to bargain in 1982 after a decade of militancy had yielded only stalemate, high unemployment, and a drop in union membership. Entrepreneurs decided to bargain because the Dutch government expressed its willingness to relinquish a long-standing power to set wages independently. With the prospect of bipartite autonomy in sight, the social partners traded wage 334

Conclusion

moderation for a reduction of work hours in a first set of bargains and then traded wage moderation for lower payroll taxes and social charges in a second set. As more jobs were created, unions were rewarded further with growing membership. Over time, unions became supportive of part-time and flexible employment, and the terms of the bargains between the social partners changed: by 1996, for example, unions were trading the liberalization of dismissal procedures for flexible workers’ pension and social security rights. The Dutch experience is a paradigmatic example of Martin Rhodes’s argument (Chapter 3) about the evolution of competitive corporatism. Dutch policy makers recognized that economic competitiveness (and jobs) would not be restored without wage moderation. By providing the incentives for a deeply rooted but moribund corporatist system to reemerge, they helped ensure that wage increases would indeed be held down. Wage moderation, according to the Central Planning Board, accounted for 50% of job growth in fifteen years (Visser and Hemerijck 1997: 113). In keeping with Rhodes’s argument, entry into the EMU did not hamper the restitution of corporatist bargaining. On the contrary, the Dutch Social and Economic Council Report on the EMU argued that “European integration requires a revitalization of corporatist adjustment strategies” (Visser and Hemerijck 1997: 111). Just as the Dutch revitalized the corporatist component of the organized managed economy (OME), they reformed its welfare component. Reforms involved cost containment at first. For example, the maximum benefit for social security payments was reduced from 80 to 70%, and eligibility criteria for disability payments were tightened after the number of people drawing disability payments exceeded 16% of the employed labor force (Visser and Hemerijck 1997: 18). These and other austerity initiatives were highly controversial but were modest in their effects and should not be mistaken for an attempt to dismantle the welfare state or the organized managed economy. On the contrary, the Dutch state has mostly rechanneled its resources, becoming more rather than less activist in key areas such as labor market policy. In a deliberate move toward becoming what Therborn terms a “strong” rather than a “soft” welfare state, the Dutch have “made a U-turn” embracing the “Scandinavian preoccupation with maximizing labor force participation,” and recognizing that the state must actively promote paid work for men and women. In the early years 335

Nancy Bermeo

of the correction process, the Dutch were criticized for reversing unemployment by channeling people into other programs or out of the labor force altogether (Iverson and Wren 1998). This policy has changed as governments adopted an increasingly activist stance. The number of people in subsidized employment has jumped 136% since 1980 and 88% since 1990. The number of people enrolled in some sort of state-sponsored job program grew 55% between 1990 and 1996 (OECD 1996: 41; Visser and Hemerijck 1997). Despite changes in the ways state funds are being allocated, the Netherlands has not created jobs at the expense of equality. Overall earnings differentials have not changed dramatically (Iverson and Wren 1998: 543). In 1995, long after the state reforms had begun, the adult minimum wage in the Netherlands was still double the wage in the United States. The case materials in this volume illustrate the theme that unites the opening chapters: there is more than one viable economic model. The organized and mixed economies of Europe may be able to cope with and even correct their unemployment problems while keeping the basic institutions of their current economic models intact.

Reassessing Unemployment A closer look at the alternative neoliberal model may make the organized and mixed models seem even more desirable. The workings and the consequences of the U.S. job machine deserve closer scrutiny. The United States – and the liberal model on which it rests – is widely revered for its capacity to generate employment. The fact that the movement out of joblessness is ten times faster in the United States than in Europe is especially impressive (Freeman 1996: 34). Yet we should be wary of exaggerating the scope and meaning of the U.S. success. The movement into joblessness is ten times as fast too (Freeman 1996: 34), and the U.S. unemployment statistics have to be contextualized to be fully understood. For example, Bruce Western and Katherine Beckett have recently demonstrated that the U.S. male unemployment rate rises by some 2% points if prison inmates are counted among the unemployed (Western and Beckett 1999: 55). Including inmates among the unemployed in European countries raises the jobless rate by only a few tenths of a percentage point. Showing that U.S. incarceration rates are only loosely connected to the level of criminal activity, Western and Beckett argue that “incarceration [has] generated a sizeable 336

Conclusion

nonmarket reallocation of labor” in the United States. Adding inmates to the official unemployment statistics in both the United States and Europe, they find that labor utilization is actually lower in the United States than in Europe for fifteen of the nineteen years between 1976 and 1994. The U.S. employment situation seems significantly better only after 1994 (Western and Beckett 1999: 56–58). Figure 11.1 shows how the general observations of Western and Beckett are replicated if women as well as men are considered. The effects of including the incarcerated in the ranks of the unemployed are not as dramatic if both sexes are included, but the jobless rate still rises one or more percentage points for most of the twelve years considered here. If we include people who feel “too discouraged” even to look for work in the ranks of the officially unemployed, the U.S. rate rises substantially higher – high enough to meet or approach the EU average in 1990 and 1991. If we include people who “want work” but do not seek it because of “family responsibilities,” “handicaps,” and the like, the rate rises even higher. These groups are included in Figure 11.1 as well.

Figure 11.1 Reassessing Joblessness in the United States, 1986–1997. Source: U.S. Department of Labour, Bureau of Labor Statistics, Employment Earnings; Western and Beckett (1999); Unpublished data from Bureau of Labor Statistics, compiled by Allan Beck and Darrell Gilliard (1997); OECD Labour Force Statistics 1977–1997. 337

Nancy Bermeo

The size of both these groups – that is, those who are so discouraged that they do not look for work, and those who want to work but feel unable to do so – derives in no small measure from the nature of the U.S. welfare state and the liberal model of the economy. Universal day care, more extensive elder care, state-sponsored programs for the handicapped, better public transportation, and more activist jobs programs would presumably deplete the numbers in both groups. How does the United States compare with Europe on these measures? Comparative data on the number of Europeans who want to work but feel unable to do so are not available. Comparative data on discouraged workers suggests that the United States is worse off than at least eight European states (see Figure 11.2). Comparative data on those who want full-time work but find only part-time work are displayed in Figure 11.3. On this measure too, many European states score higher than the United States. In fact, only Sweden and the Netherlands are worse off. The size and age of the U.S. student population affects the comparative assessment of unemployment as well. As Cameron illustrates (Chapter 1), joblessness in Europe affects the young disproportionately. Youth unemployment in the United States is also problematic, but the youth

Figure 11.2 Discouraged Workers as a Percentage of the Labor Force, 1993: The United States and Europe. Source: OECD (1995: 76). 338

Conclusion

Figure 11.3 Involuntary Part-Time Workers as a Percentage of the Labor Force, 1993: The United States and Europe. Source: OECD (1995: 76–77). Table 11.1. Proportion of Secondary School Graduates Going on to Higher Education, 1980–1994

unemployment rate is brought down considerably by the fact that in the United States, the vast majority of people aged eighteen to twenty-two who finish secondary school go on to some form of “third-level” education. The sharp contrast between the United States and selected European countries is displayed in Table 11.1. Because of dramatic variations in educational policy, the percentage of the population that can be classified as “in school” rather than unemployed 339

Nancy Bermeo

is higher in the United States than in Europe, despite the fact that large sectors of the same age cohort are without work and still heavily dependent on parental assistance in both societies. The point here is not to argue that U.S. employment performance has been poor. This is clearly not the case. But the success of the U.S. job machine is often exaggerated. Even without reassessing who should be considered unemployed, the simple disaggregation of European unemployment proves the point. Figure 11.4 shows the number of European countries which have either matched or exceeded the U.S. employment record during the past decade. The wide range of countries that compare favorably with the United States illustrates the argument presented by Cameron, Hall, and Rhodes (Chapters 1, 2, and 3): the liberal model is not the only viable model available. Just as the success of the liberal model needs to be put in comparative perspective, so do its costs. The liberal model of job creation in the United States has been extremely costly in terms of wage inequality. The vast majority of jobs that have been created have been service jobs with low

Figure 11.4 Official Standardized Unemployment Rate: United States versus Selected European States, 1990–1998. Source: OECD (1999: 224). 340

Conclusion

wages. As a consequence, real earnings for the less skilled have plummeted since the 1980s (Stokes 1998: 23). For all skill levels, workers who find employment after being jobless typically take a 23% pay cut (Freeman 1996: 34). In 1994, the U.S. Council of Economic Advisors reported that the mean family income of the lowest earnings quintile had actually dropped 14% since 1973 (Freeman 1996: 38). Of all the developed countries, the United States has the most inegalitarian earnings distribution, with the lowest decile of the U.S. workforce earning 38% of median earnings while their peers in Western Europe earn 68% (Freeman 1996: 32). The portion of full-time workers whose earnings are actually below the poverty line skyrocketed from 12% in 1979 to 18% in 1990 (Krugman 1994: 23). These inequalities nourish the growth of a permanent underclass where many find crime more attractive than a job with low pay. In one out of every seven census districts in major U.S. cities, a full two-thirds of all adults are jobless (Wilson 1996: 14), and children often grow up unaware of paid employment “as a central experience of adult life”(Wilson 1996: 52). The costs of extreme inequality are highest for the poor, but they are borne by whole polities. As Yale economist James Tobin reminds us, “there are limits to the . . . inequality that a democracy can tolerate” for “materialism, hardheartedness and incivility” are its “by-products” (Tobin 1996: 179). Where the European labor market creates “outsiders,” the U.S. labor market creates “marginals” instead. The former are the long-term unemployed who are unable to secure paid employment. The latter are the never or nearly never employed – that is, individuals who are either unable or unwilling to work for the low wages that their low skill level affords them. The number of men who have permanently dropped out of the U.S. labor force has more than doubled since the 1960s ( Wilson 1996: 23). These individuals do not show up in conventional unemployment data, but they are jobless nonetheless. Clearly, the realities of the U.S. liberal model are more problematic than a first glance at official unemployment figures suggests.

Is the Move toward the Liberal Model Inevitable? Regardless of the weaknesses discussed above, many see Europe’s eventual embrace of the liberal model as almost inevitable (Moody 1997: 126). Two factors suggest that it might be. First, the real and imagined success of the 341

Nancy Bermeo

liberal model in the United States has given the LME a hegemonic position in key international policy-making institutions. Though there is often great debate about its individual elements, no other economic management model is as widely revered by economists nor as widely purveyed by international financial institutions (Williamson 1990; Stallings 1992). Second, and relatedly, many argue that the structure of the European Monetary Union will force a new economic model on European states regardless of their political preferences. In fulfilling its statutory goal of price stability (defined as an inflation rate of no more than 2%), the European Central Bank inevitably restricts the autonomy of national political elites. Berman and McNamara compare the current dynamic to a “secular version of Lent” in which governments prove their “faith” by giving up “the ability to control their countries’ economic destinies” (Berman and McNamara 1999: 3). The EU’s power to extract sacrifice from the believing was made manifest in 1995–97 as countries scrambled to meet the EMU convergence criteria. Yet this is a Lent with no Easter, for the loss of autonomy is permanent. Though it is not unreasonable to assume that permanent restrictions on budget deficits and government debt mean a retraction of state intervention and an embrace of liberalism, Europe’s policy-making terrain still looks vastly different from the policy-making landscape of the U.S. For the most part, budgets and debts were not brought into line with the Maastricht criteria by radically cutting social programs but by raising taxes instead (Cottrell 1999). Thus, if EMU has brought a retraction of the state in some realms, it has brought an expansion of the state in others. The current European emulation of the Internal Revenue Service constitutes an interest in a decidedly nonmarket institution. Despite a myriad of individual policy changes in different countries, Europe seems not to be converging on the neoliberal model thus far, despite the fact that twenty years have passed since the neoliberal revolution began in Britain. On the contrary, Franz Traxler’s recent survey of state/industrial relations policies in twenty advanced industrial nations found “no evidence” of a “convergence toward neoliberal regulation” (Traxler 1999: 55). What he found instead looks much like the findings of this collection: “both neoliberalism and corporatism increased” (Traxler 1999: 69, emphasis added). New policy approaches are consistently combined with traditional ones. Continuity is related to an intrinsic interconnectedness wherein certain structures and production regimes “ ‘fit’ each other” (Huber and Stephens 1998: 8) and constrain change. As a result, 342

Conclusion

the state’s role in industrial relations “has remained rather stable” (Traxler 1999: 79). Research on individual countries, such as Portugal and Spain, has drawn similar conclusions (Glatzer 2000). Is the eventual adoption of the liberal model a necessary condition for the solution of Europe’s employment problems? Cameron, Hall, and Rhodes (Chapters 1, 2, and 3) argue that it is not, and further evidence suggests that they are correct. In several European countries, unemployment seems to be dropping without the dismantling of the OME. Table 11.2 tells the story. The cases are sorted on two dimensions. Vertically, they are divided according to the severity of the unemployment problem: with high unemployment defined as above 10%. Horizontally, they are divided according to how much an individual country has decreased unemployment from whatever peak it reached in the period running from 1984 to 1998. A high correction score means that unemployment has dropped over 40%. A medium correction score means a drop between 40 and 20%. The “No Correction” column is reserved for states in which unemployment is on the rise. The “No Correction Needed” column lists those states in which unemployment has never exceeded 5.5% – the rate considered full employment by most economists.

Table 11.2. Correcting Unemployment: United States versus Europe, 1984–1998

343

Nancy Bermeo

The United States is only one of several states that have been able to decrease their unemployment rate by over 40% in the period surveyed. The broad variety of states that have achieved this level of correction suggests that the adoption of the liberal model is not a necessary condition for unemployment relief. The heterogeneity of the successful case lists is a constant, whether one considers countries in which the peak of unemployment has been relatively high or relatively low. The fact that the performance of the United States is similar to that of Norway and Portugal makes the point. These three countries, as Hall (Chapter 1) reminds us, represent three different economic models (the LME, the OME, and the mixed model, respectively). The nature of the countries which have been able to avoid serious unemployment altogether should lead us to question the necessity of the liberal model even further. Neither Austria nor Luxembourg nor Switzerland can possibly be classified as LMEs, yet their performance has been consistently superior to that of any liberal economy in the sample. Norway’s organized managed economy has avoided an unemployment rate of over 5.5% for all but three of the last thirty years. This is a fourth case in which a highly successful employment record has been achieved outside the liberal paradigm. The fact that each of the “liberal” economies in the sample scores high on correction performance appears to legitimate arguments about the relative merits of the LME. Yet, a closer look at the cases indicates the need for caution. Britain’s unemployment rate averaged 9.1% in the years of conservative rule. This means that over a decade and a half of liberalization yielded an average unemployment rate just 0.4 percentage points below the EU average in the same period. Though unemployment was dropping in the last years of the Major administration, the U.K.’s correction score at the end of this period was only 29%. This is certainly a positive achievement but its relative merits should not be exaggerated. The recent correction of unemployment in Ireland has been more impressive. The “Emerald Tiger’s” economic growth between 1989 and 1997 produced a 23% rise in employment in just eight years (Tansey 1998: 35). From 1990 to 1997, total employment in the European Union dropped by 2.7% while total employment in Ireland rose by 17.3%. This is a dramatic difference, yet there are reasons to be cautious about interpreting this difference as a triumph of the liberal model. To begin with, total employment in the United Kingdom dropped in the same period by 4.7%, leaving the score for the liberal economies lower than it would be 344

Conclusion

if we simply considered the United States and Ireland (OECD 1997: Annex table 20). But more important than the British comparison are the many nonliberal elements in Ireland’s policy-making formula. Ireland’s economic model today is a mixture of liberal and organized elements, and the latter have played an important role in job development. Admittedly, Ireland fits the liberal model in several important ways. Policy makers cut employment in the public sector in the mid-1980s (Tansey 1998: 39). They implemented a “determined” program of fiscal consolidation that reduced the share of public expenditure by some ten percentage points at around the same time (de la Fuente and Vives 1997: 120). They instituted low corporation tax rates, which have been a key mechanism for investment growth (Sachs 1997: 60; Tansey 1998: 47), and they allowed some growth in income disparities: the ratio of top-to-bottom decile earnings rose from 4.2 in 1987 to 4.9 in 1994 (Gray 1997: xxv). This said, there are important differences between the Irish economic model and the British. In contrast to the “Thatcherite model of the 1980s,” Ireland “has not relied exclusively on private sector responses to tax-cuts for expansion. Public spending, broadly defined to include EU transfers, has been visibly important” (Tansey 1998: 49). State-owned enterprises “play a major role in the Irish economy” (Lane 1996: 121), and “social safety nets have been strengthened” rather than weakened (Tansey 1998: 247). Ireland, according to Michael Burda, simply “did not experience the Great Regime Change of 1979.” It opted instead for a “continental” solution, including “a top-down industrial strategy” as well as “an incomes policy in the form of social contracts signed by government, trade unions and employers” (Burda 1997: 86; Economist 1997: 24). In keeping with the argument advanced by Martin Rhodes (Chapter 3), cooperative pacts have been fundamental to ensuring the competitive nature of the Irish economy. Though the Irish wage system once looked much like the British, “centralized” wage bargaining has become the norm in recent years, with “a large proportion of wages being determined by national wage agreements” (O’Hagan 1996: 254). Neocorporatist agreements have been fundamental to keeping wages moderate and to boosting both the growth of the economy and the growth of employment opportunities (Krugman 1997: 51). Like their counterparts in the Netherlands, Irish unions are now trading wage moderation for individual tax cuts and more jobs. Multiyear agreements launched in 1987, 1991, 1994, and 1996 have all functioned to lower labor costs and thereby make Ireland a more 345

Nancy Bermeo

attractive place to invest. The emphasis on social partnership and social consensus rather than class conflict has yielded what at least one economist has called “a supply-side revolution with a social democratic face” (Tansey 1997: 49). More important for our concern with unemployment, Ireland’s emphasis on social partnership has yielded the rapid expansion of investment and jobs. In emphasizing the multiple means of correcting unemployment, the picture sketched here coincides not simply with the essays in this collection but with broader literatures too. The OECD itself has recently published two studies that raise important challenges to the liberal premises that inspired the organization’s policy guidelines in the past. One concerns job protection legislation. Another concerns the nature of collective bargaining. The widespread belief that job protection legislation leads to higher levels of unemployment seems to be undergoing challenges and refinement. Though the OECD’s 1994 “Jobs Strategy Recommendations” called explicitly for the “reform [of] employment security provisions that inhibit the expansion of employment in the private sector” (OECD from Conover and Feldman 1986: 6), more recent OECD research suggests that employment security regulations might not be as deleterious as once thought. A major 1999 OECD study supported the hypothesis that strict employment protection legislation, or EPL, “lowers some forms of labor market turnover,” but then concluded, decisively that “EPL strictness has little or no effect on overall unemployment” (OECD 1999: 88, emphasis added). Though the evidence linking EPL to the expansion of self-employment is “robust,” there was “no clear link” between EPL and the share of workers in temporary employment. Nor was there more than a “very weak and statistically insignificant” correlation between EPL and the unemployment of either youth or women (OECD 1999: 88). The questioning of the common wisdom regarding different levels of job protection is mirrored in recent research on the relative merits of different forms of collective bargaining. The OECD Jobs Study of 1994 argued implicitly for liberal bargaining systems in recommending the removal of “restrictions that prevent wages from reflecting local conditions and individual skill levels” (OECD Jobs Study of 1994 from Conover and Feldman 1986: 6). In 1997, however, the OECD published a major comparative project with rather different implications. The study used measures of trade union density, bargaining coverage, centralization, and coordination to assess 346

Conclusion

collective bargaining in nineteen economically advanced nations over a fifteen-year period. Contrary to the well-known argument advanced by Calmfors and Driffill (1988) about the relative inferiority of intermediate systems, they found “no robust evidence” that any particular form of collective bargaining was more associated with positive economic performance than any other. Though it was clear that more centralized/ coordinated economies had significantly less earnings inequality, the analysis did not find “statistically significant relationships” (OECD 1997: 83) between any measures of collective bargaining and unemployment. Ironically, the study “tentatively suggests” that the countries which followed the liberal common wisdom and “moved toward decentralization or less coordination” actually “experienced larger declines in the employment rate” than those that did not (OECD 1997: 83, emphasis added). The effects of moving away from centralization and coordination are illustrated in Figure 11.5. The relationship between types of collective bargaining and the ability to correct unemployment is illustrated in Table 11.3. Table 11.3. The Nature of Collective Bargaining and the Capacity to Corrected Unemployment, 1984–1998

Sources: OECD Employment Outlook (1997), p. 71; OECD Employment Outlook (1999), p.224; OECD Historical Statistics 1960-1994, p. 47; Sepheriades (2000); Tansey (1998).

347

Nancy Bermeo

(a)

(b)

Figure 11.5 Change in Centralization and Percentage Point Change in (a) Unemployment Rate and (b) Employment Rate. Source: OECD (1997: 81).

A comparative analysis of the cases shows that the capacity to correct unemployment altogether is not confined to a particular type of system. The two (liberal) systems with low levels of coordination score well, but two systems with much higher levels of coordination score even higher, and five systems with moderately higher levels of coordination score as well as the liberal systems.

348

Conclusion

Conclusion Both the 1997 and 1999 OECD employment reports reinforce the idea that no single set of institutional arrangements proves uniquely superior for the correction of unemployment. Neither the wholesale dismantling of employment protection legislation nor the destruction of Europe’s current collective bargaining institutions is supported by the evidence. A move toward a U.S.-style model seems unwarranted. What does seem warranted is more informed thinking about two questions. First, what policies must change for unemployment to drop? Second, what actors should be included in the crafting and implementation of these policies? The first question turns out to be much more difficult to answer than the second. Many highly respected economists readily admit that there is no consensus yet on even the relationship between unemployment and inflation (Alessina and Rosenthal 1995). Blanchard and Jimeno argue that “economists are still a long way from understanding movements in structural employment rates across countries and time” and that “humility is in order” (Blanchard and Jimeno 1995: 217). We are clearly not in a situation where the economics are uncontested and only the politics are problematic. On the contrary, the answer to the political question concerning how policy changes are best brought about seems relatively clear. Both coordinated and uncoordinated policy models seem to “work” in that both are associated with drops in unemployment. Yet, ultimately, coordinated policy models are preferable. Why? The studies in this collection and in the literature more broadly, show that the coordinated policy models, resting on some form of competitive corporatist bargaining arrangements, are not decidedly inferior to uncoordinated, liberal models in economic terms. Compared with their liberal counterparts, coordinated arrangements were associated with stronger rather than weaker employment records in the 1960s and 1970s (Cameron this volume) and they are increasingly associated with dramatic corrections of unemployment in the 1990s (Pérez 1999; Rhodes this volume; Wolinetz this volume). They are also associated with the strengthening of associational ties (Scruggs and Lange this volume). Contrary to the expectations of many proponents of the liberal model, neither highly coordinated nor moderately coordinated systems seem any worse than uncoordinated systems in the provision of employment (OECD 1997). Coordinated systems do seem to be

349

Nancy Bermeo

associated with higher rates of long-term unemployment, but this is counterbalanced by the clear association between coordinated systems and lower wage inequalities. Most important perhaps for the correction of unemployment is the fact that coordinated systems seem more efficient at controlling real unit labor costs. As Table 11.4 shows clearly, real unit labor costs in the 1990s dropped much more in high-coordination systems than in low ones. The economic case against coordination is thus rather weak. Yet, the political case in favor of coordination is strong. Coordination – in some variant of competitive corporatism – has at least two political advantages over the liberal model. One derives from the association between coordination and consensus. The other derives from the association between coordination and inclusion. If the New Europe is to conquer unemployment under the constraints of monetary union, states will have to reform a whole range of popular policies. The politicians who undertake these reforms will prove less vulnerable if they seek broad consensus involving capital and labor. As Paul Pierson explains it, “broader consensus legitimates the claim that policy change is necessary,” and that “policies can be sustained.” Powerful interests “are likely to place a high value on predictability and continuity” (Pierson 1998: 555), and this is especially true of the investors needed to create jobs. By “extending the range of actors with a stake in the reform outcome” (Pierson 1998: 555), coordinated policy making trumps the more adversarial liberal model. Coordination – in the form of competitive corporatism – trumps liberalism through inclusion as well as consensus. Competitive corporatism is an inclusive means of policy making and inclusion is a fundamental dimension of democracy. Indeed, as Robert Dahl has persuasively argued, the Table 11.4. Real Unit Labor Costs: High Coordination versus Low Coordination Systems, 1991–1998

350

Conclusion

more inclusive a political system becomes, the closer it comes to being truly democratic (Dahl 1971: 7). With so many Europeans excluded from paid employment, and with such important policy parameters set by the nonelected officials of the European Central Bank, the value of political inclusion in the New Europe is especially high. The essays in this collection show that the costs of inclusion are relatively low and probably well worth the price. References Alesina, Alberto, and Howard Rosenthal. 1995. Partisan Politics, Divided Government and the Economy. New York: Cambridge University Press. Allen, Sheila, and Alan Waton, eds. 1986. The Effects of Unemployment Experience and Response. In The Experience of Unemployment. Basingstoke, Hampshire, England: Macmillan. Amoretti, Ugo. 2000. Does Labour Under-utilisation Really Affect Political Attitudes? Southern Europe in Comparative Perspective. In Nancy Bermeo, ed., Unemployment in Southern Europe: Coping with the Consequences. London: Frank Cass. Berman, Sheri, and Kathleen R. McNamara. 1999. Bank on Democracy. Foreign Affairs 78 (1999): 2. Bermeo, Nancy. 2000. What’s Working in Southern Europe? In Nancy Bermeo, ed., Unemployment in Southern Europe: Coping with the Consequences. London: Frank Cass. Betz, Hans Georg. 1994. Radical Right-wing Populism in Western Europe. New York: St. Martin’s. Blanchard, Olivier, and Juan F. Jimeno. 1995. Structural Slumps and Persistent Unemployment. American Economics Association Papers and Proceedings 85, 2. Braun, Aurel. 1997. Introduction. In Aurel Braun and Stephen Scheinberg, eds., The Extreme Right. Boulder, Colo.: Westview. Burda, Michael. 1997. Persistently High Irish Unemployment: A Comparison with the UK. In Alan W. Gray, ed., International Perspectives on the Irish Economy. Dublin: Indecon. Calmfors, Largs, and John Driffill. 1988. Bargaining Structure, Corporatism and Macroeconomic Performance. Economic Policy 6, 14–61. Cohn, R. 1978. The Effect of Unemployment Status Change on Self-Attitudes. Social Psychology 41 (June 1978): 81–93. Conover, Pamela Johnston, and Stanley Feldman. 1986. Emotional Reactions to the Economy. American Journal of Political Science 30, 1. Cottrell, Robert. 1999. Europe: So Far, It Flies. New York Review of Books, April 8. Dahl, Robert. 1971. Polyarchy. New Haven, Conn.: Yale University Press. De la Fuente, Angel, and Xavier Vives. 1997. The Sources of Irish Growth. In Alan Gray, ed., International Perspectives on the Irish Economy. Dublin: Indecon. European Commission. 1998. Eurobarometer Report 49. Standard Eurobarometer, fig. 6.2. 351

Nancy Bermeo Evans, Richard J. 1987. Introduction: The Experience of Unemployment in the Weimar Republic. In Richard Evans and Dick Geary, eds., The German Unemployed. London: Croom Helm. Freeman, Richard B. 1996. Doing It Right? The US Labor Market Response to the 1980’s–1990’s. In Herbert Giersch, ed., Fighting Europe’s Unemployment in the 1990’s. Berlin: Springer. Gallie, Duncan. 1994. Social Consequences of Long-term Unemployment in Britain. In Odile Benoit-Guilbot and Duncan Gallie, eds., Long-Term Unemployment. London: Pinter. Givens, Terri. 1997. Unemployment, Immigration and the Radical Right in Western Europe. Paper presented at Southwestern Political Science Association Annual Meeting, March 26–29, New Orleans. Glatzer, Miguel. 2000. Rigidity and Flexibility: Patterns of Labour Market Policy Change in Portugal and Spain, 1981–1996. In Nancy Bermeo, ed., Unemployment in Southern Europe: Coping with the Consequences. London: Frank Cass. Gowan, Peter, and Perry Anderson. 1997. The Question of Europe. New York: Verso. Gray, Alan. 1997. Forward: Irish Economic Challenges and International Perspectives. In Alan Gray, ed., International Perspectives on the Irish Economy. Dublin: Indecon. Hibbs, D. 1987. The Political Economy of Industrial Democracies. Cambridge: Harvard University Press. Huber, Evelyne, and John Stephens. 1998. Welfare States and Production Regimes in the Era of Retrenchment. Paper prepared for the “New Politics of the Welfare State.” Harvard University. Iverson, T., and A. Wren. 1998. Equality, Employment and Budgetary Restraint: The Trilemma of the Service Economy. World Politics 50, 4: 507–46. Kagedan, Ian. 1997. Contemporary Right-Wing Extremism in Germany. In Aurel Braun and Stephen Scheinberg, eds., The Extreme Right. Boulder, Colo.: Westview. Kitschelt, Herbert. 1995. The Radical Right in Western Europe: A Comparative Analysis. Ann Arbor: University of Michigan Press. Krugman, Paul. 1994. Europe Jobless, America Penniless? Foreign Policy 9 (Summer): 19–34. ———. 1997. Good News from Ireland: A Geographical Perspective. In Alan W. Gray, ed., International Perspectives on the Irish Economy. Dublin: Indecon. Lane, Philip R. 1996. Government Intervention. In J. W. O’Hagan, ed., The Economy of Ireland: Policy and Performance of a Small European Country. New York: St. Martin’s Press. Lewis-Beck, Michael. 1988. Economics and Elections: The Major Western Democracies. Ann Arbor: University of Michigan Press. Lipset, Seymour Martin, and Earl Rabb. 1970. The Politics of Unreason. New York: Harper and Row. Lockerbie, Brad. 1993. Economic Dissatisfaction and Political Alienation in Western Europe. European Journal of Policy Research 23: 281–93.

352

Conclusion McElligott, Anthony. 1987. Mobilising the Unemployed. In Richard J. Evans and Dick Geary, eds., The German Unemployed. London: Croom Helm. Maravall, Jose Maria. 1997. Regimes, Politics and Markets : Democratization and Economic Change in Southern and Eastern Europe. Oxford: Oxford University Press. Milbrath, Lester, and M. L. Goel. 1977. Political Participation. Chicago: Rand McNally. Moody, Kim. 1997. Workers in a Lean World. London: Verso. O’Hagan, John. 1996. Employment and Unemployment. In Alan Gray, ed., International Perspectives on the Irish Economy. Dublin: Indecon. Organisation for Economic Co-operation and Development (OECD). 1995. Employment Outlook (June). Paris. ———. 1996. Economic Surveys: The Netherlands. Paris: OECD. ———. 1997. Employment Outlook ( June). Paris. ———. 1999. Employment Outlook ( June). Paris. Pérez, Sofia. 1999. The Resurgence of National Serial Bargaining in Europe; Explaining the Italian and Spanish Experiences. Working Paper 1999/130. CEACS. Juan March Institute. Pierson, Paul. 1998. Irresistible Forces, Immovable Objects: Post-Industrial Welfare States Confront Permanent Austerity. Journal of European Public Policy 5 (December): 539–60. Przeworski, Adam. 1996. Public Support for Economic Reforms in Poland. Comparative Political Studies 29, 5 (October): 520–43. Rosenstone, Steven. 1982. Economic Adversity and Voter Turnout. American Journal of Political Science 26 (1): 25–46. Sachs, Jeffrey. 1997. Ireland’s Growth Strategy: Lessons for Economic Development. In Alan Gray, ed., International Perspectives on the Irish Economy. Dublin: Indecon. Sepheriades, Seraphim. 2000. Unemployment, Informalization and Trade Union Decline in Greece. In Nancy Bermeo, ed., Unemployment in Southern Europe: Coping with the Consequences. London: Frank Cass. Stallings, Barbara. 1992. International Influence on Economic Policy: Debt Stabilization and Structural Reform. In Stephan Haggard and Robert Kaufman, eds., The Politics of Economic Adjustment. Princeton, N.J.: Princeton University Press. Stokes, Susan. 1998. Introduction: Public Opinion and Market Reforms: The Limits of Economic Voting. Comparative Political Studies 29, 5: 499–519. Tansey, Paul. 1998. Ireland and Work: Economic Growth and the Labour Market, 1987–1997. Dublin: Oak Tree Press. Therborn, Goran. 1986. Why Some People Are More Unemployed Than Others. London: Verso. Tobin, James. 1996. Full Employment and Growth: Further Keynesian Essays on Policy. Cheltenham, England: Edward Elgar. Toharia, Luis. 1997. Labor Market Policies – Spain. Luxembourg: European Commission.

353

Nancy Bermeo Torres, Francisco. 1994. Portugal. In John Williamson, ed., The Political Economy of Policy Reform. Washington, D.C.: Institute for International Economics. Traxler, Franz. 1999. The State in Industrial Relations: A Cross-National Analysis of Developments and Socioeconomic Effects. European Journal of Political Research 36: 55–85. Verba, Sidney, and Kay Lehman Scholzman. 1997. Unemployment, Class Consciousness and Radical Politics: What Didn’t Happen in the 30’s. Journal of Politics 39: 291–323. Visser, Jelle, and Anton Hemerijck. 1997. A Dutch Miracle. Amsterdam: Amsterdam University Press. Western, Bruce, and Katherine Beckett. 1999. How Unregulated Is the U.S. Labor Market? American Journal of Sociology 104, 4 (January). Whelan, C. 1994. Social Problems and Personal Troubles: Unemployment and Psychological Distress in Ireland. In Odile Benoit-Guilbot and Duncan Gallie, eds., Long-Term Unemployment. London: Pinter. Williamson, John. 1990. What Washington Means by Policy Reform. In John Williamson, ed., Latin American Adjustment. Washington, D.C.: Institute for International Economics. Wilson, William. 1996. When Work Disappears: The World of the New Urban Poor. New York: Alfred A. Knopf.

354

Index

agricultural labor, 186 Aliança Democrática government (Portugal), 196 Anderson, Christopher, 4, 331, 333 Austria collective bargaining, 34 competitive corporatism, 106, 112 employment rates, 21 impact of age on unemployment, 26 unemployment, long-term, 26, 27 unemployment rates, 15, 17, 30, 150 social welfare program, 60–1 Aznar, José María, 33 bargaining. See collective bargaining; corporatist bargaining; wage bargaining. Beckett, Katherine, 336–7 Belgium, 133 employment, part-time, 29 employment rates, 30 impact of age on unemployment, 26 political behavior and attitudes of unemployed, 281, 284, 286 unemployment, long-term, 26 unemployment rates, 15, 20 xenophobia, 282 Berman, Sheri, 342

Beveridge, William, 330 Blanchard, Olivier J., 12, 349 Bloco Central government (Portugal), 179, 180, 197 Britain, 281, 283, 285, 286. See also United Kingdom. Calmfors, Lars, 347 Cameron, David, 2, 332, 338, 339, 343 capital mobility globalization, 93–4, 95, 96 Casey, B., 111 CC.OO. See Comisiones Oberas. CEOE (Spainsh employers confederation), 33, 240 Cerdeira, M. C., 194, 195 class and voting behavior, 276 Clayton, R., 95 collective bargaining, 34, 46–7, 76, 77, 251. See also corporatist bargaining. and union influence, 127–8 and wage moderation in corporatist systems, 126 Comisiones Oberas (CC.OO), 33, 224, 225, 238–40 Confederação Geral dos Trabalhadores Portugueses (CGTP) (Portugese union federation), 178, 181, 182, 194 355

Index corporatism, 2, 4, 89, 197 competitive, 102–12, 335, 349–51 demise of, 101–3 literature on, 251–2 polder model, 256–8, 264 corporatist bargaining, 34, 332, 334 decline of, 137 Craypo, C., 199 Crouch, Colin, 133 Dahl, Robert, 350 deindustrialization, 20 de Koning, Jan, 260, 261 de la Hoz, J. M., 238 Denmark collective bargaining, 34 competitive corporatism, 106, 108, 109 corporatist bargaining, 129 deregulation, 78 employment, part-time, 29 employment rates, 21 impact of age on unemployment, 26 income distribution, 303 job creation rates, 46 political behavior and attitudes of unemployed, 278, 281, 284, 286 unemployment, long-term, 26, 27 unemployment rates, 16, 17, 30, 150 den Uyl cabinet (Netherlands), 258, 259 deregulation financial, 75–6 de Vries, Bert, 263, 264 Driffill, John, 347 ECB (European Central Bank), 9 economic growth and unemployment, 32–3, 39–40 economic reform difficulty of in context of high unemployment, 291 strategies, 295–6 356

economy. See liberal market economy; organized managed economy. education and training, 34, 103–4, 255, 295, 297, 298, 307, 332–3, 339–40 in competitive corporatism, 110 promoted by OME, 59 undermined by unemployment, 76–7 employers’ organizations in corporatist system, 103 employment and gender, 184, 191 insider-outsider labor market, 221–2 part-time, 28–9, 34, 72, 105, 207, 245, 250, 254, 255, 261, 334 political, economic and institutional sources of, 29–36 protection legislation, 346–7 temporary, 205, 207, 187, 190, 204, 205. See also fixed-term employment. variations among European countries, 20–3 of women, 207 EMU. See European Monetary Union. endogenous-growth theory, 70 Esping-Anderson, Gösta, 60, 72, 99 Estivill, J., 238 European Central Bank (ECB), 9 European Economic Community (EEC), 184 European Investment Bank, 36 monetary policy, 40–4 European Monetary Union (EMU), 10, 332 and competitive corporatism, 108, 109–10 and corporatist bargaining, 335 employment policy of, 44–6 impact on interest rates and investment, 38–40 monetary policy and European Central Bank, 40–4 structural effects of, 37–8, 47

Index support of for liberal market economy, 344 and wage bargaining, 64–5 European Union (EU), 1 job creation rates, 55 unemployment rates, 7–9 variations in unemployment and employment within, 15–26

political behavior and attitudes of unemployed, 281, 286 real wage growth, 123 unemployment rate, 15, 17, 20, 124, 150 Freeman, R. B., 91 Frieden, J. A., 93 Fulcher, J., 210

Federation of Dutch Trade Unions (FNV), 35, 253, 258, 261, 262 financial services sector impact of EMU on, 37–8 Finland competitive corporatism, 106 employment, part-time, 29 employment rates, 21, 23 impact of age on unemployment, 26 unemployment, long-term, 26–7 unemployment rates, 17, 20, 30, 150, 157 fixed-term employment. See also employment: temporary. correlation with class, 216–20 and creation of insider-outsider divisions, 220–1 and job turnover, 214–16 as source of two-tier system, 211–13 and unemployment, 213–14, 218, 220–2 and union participation, 230–7 FNV. See Federation of Dutch Trade Unions. Foundation of Labor ( STAR, Netherlands), 254, 255, 256, 262 Fraile, Marta, 4, 331 France economic policy under Mitterrand, 138 employment rate, 23 impact of age on unemployment, 26 interest in politics, 284 neoliberal shift in 1980s, 131

Ganghof, S., 96 Garrett, G., 145–6 gender. See also women. impact on unemployment, 24, 25, 26 and part-time employment, 29 General Agreement on Tariffs and Trade (GATT), 87 Genschel, P., 96 German unification as cause of European recession, 12 Germany corporatism, 104, 111, 129 deregulation, 78 employment, part-time, 29 employment rates, 21, 23 impact of age on unemployment, 26 political behavior and attitudes of unemployed, 278, 280, 281, 286 social welfare program, 60–1 unemployment, long-term, 26 unemployment rates, 15, 17, 30, 54, 124, 157 wage growth, 123 wage growth/labor decline model, 135, 138 xenophobia, 282 Ghent unemployment insurance system, 148, 151–2, 157, 162–4, 166, 174 globalization, 273 competitive corporatism as response to, 101–12 debates on impact of, 87–90 finance, 93–4 multinationals, 96–8 357

Index globalization (cont.) and organized market economy, 68–72 tax competition, 94–9 trade and competition, 90–3 as undermining unions, 146 welfare and employment reform, 98–112 Glyn, A., 94 Gobeyn, M. J., 102 Gold, M., 111 Golden, Miriam, 129 Goldthorp, John H., 309 Gorz, A., 209 Grahl, J., 101 Great Britain, 303 Greece employment, part-time, 29 employment rates, 21–2, 23, 286 unemployment, long-term, 26 unemployment rates, 15, 20 xenophobia, 282 Gutiérrez, Antonio, 240 Hall, E., 210 Hall, Peter A., 2, 105, 252, 332, 340, 343, 344 health care reform, 99 Healy, Kieran, 331 Hemerijck, Anton, 264 Hobsbawm, Eric, 208, 209 Hyman, R., 209 IG Metall (German union), 111 income distribution and inequality, 303–7, 332, 333 U.S., 341 and voting behavior, 311, 312–13, 314, 315, 318, 322–3 inflation, 136 wage growth and, 124–5 International Monetary Fund (IMF), 9, 89 Ireland collective bargaining, 34, 35 358

competitive corporatism, 101, 107, 108, 112 economic growth, 32, 33 employment rates, 30 income distribution, 303 job creation rate, 21, 46 multinational corporations, 97 political behavior of unemployed, 281, 286 unemployment, long-term, 26 unemployment rates, 16, 17, 30 Iriso, P. L., 229 Italy competitive corporatism, 101, 107, 108 employment, part-time, 29 employment rates, 21–2, 23 income distribution, 303 political behavior and attitudes of unemployed, 281, 284, 285, 286 unemployment, long-term, 26 unemployment rates, 15, 17, 20, 124, 148, 150 wage growth, 123, 128 xenophobia, 282 Iverson, Torben, 64 Japan job creation rate compared with EU and U.S., 13–15 unemployment compared with Europe and U.S., 10f1.1, 150 Jimeno, Juan F., 349 job creation rate, 10, 13–15, 21, 46, 55, 104–5, 107, 245–6, 299. See also specific countries. Jones, Erik, 265 Kok, Wim, 259, 263 Kurzer, P., 102 labor market, 33 and globalization, 90–1 reform under competitive corporatism, 109–11, 112–13 regulation, 104–5

Index structural shifts in, 173–4 wage growth rate, 124–5 Lafontaine, Oskar, 43, 45 Lane, Christel, 71 Lange, Peter, 3, 98, 113, 129, 145–6, 331 Lash, S., 101 Levy, Jonah, 78 liberal market economy (LME), 2, 3, 52, 53 compared to organized market economy, 55, 56, 67 and future of Europe, 341–51 unemployment and union density, 154, 161, 164 U.S. compared to Europe, 336–41 Lisnave (Portugese company), 180–1 LME. See liberal market economy. Lockwood, D., 209 Lubbers, Ruud, 260, 261, 263 Lubbers cabinet (Netherlands), 31, 247, 259, 261 Lubbers-Kok cabinet (Netherlands), 249, 262, 263 Luxembourg collective bargaining, 34 employment rate, 30 job creation rate, 21 political behavior and attitudes of unemployed, 281, 285, 286 unemployment, impact of age on, 26 unemployment rates, 30 McCracken Report (OECD), 126 McNamara, Kathleen R., 342 Major government (U.K.) reduction of welfare state under, 99 Mancha, Tomás, 310 Maravall, José María, 4, 310, 331 Marshall, Gordon, 209 Martin, Andrew, 64, 111 Melkert jobs, 36, 254 Modern Capitalism (Shonfield), 7 monetary policy and unemployment, 64

Mosley, H., 112 multinational corporations (MNC), 87, 89, 94, 96–8 competitive corporatism, 109 Mundell-Fleming theorem, 93 Negrelli, S., 103–4 neocorporatism, 53. See also organized managed economy. neoliberalism, 89. See also liberal market economy economic reform, 295, 297–8 Netherlands, 4 collective bargaining, 34, 35, 77 competitive corporatism, 101, 106–7, 108, 109, 112, 245–53, 256–8 deregulation, 78, 105 employment, part-time, 29 employment rate, 28, 30 job creation rate, 21, 46, 246 polder model, 253–8, 262–5 political behavior and attitudes of unemployed, 281, 285, 286 social welfare programs, 60–1 unemployment rates, 16, 26, 30, 150, 248 unions and employer’s organizations, 253 wage restraint, 36, 167, 254 Wassenaar Agreement, 258–62 Nickell, S., 105 Norway corporatist bargaining, 106, 129 neoliberal shift in 1980s, 131 unemployment rates, 150 wage growth, 124 Notermans, T., 94 OECD, 9, 89 OME. See organized managed economy. Organization for Economic Cooperation and Development (OECD), 9, 89 359

Index Organization of Petroleum Exporting Countries (OPEC) oil price shock, 10, 12, 122, 128, 157 organized managed economy (OME), 2, 55, 329, 332, 334, 335 employment performance in, 58–61, 64 globalization, 68–72 institutional complementarities as limit on reform, 74–8 macroeconomic policy context for, 63–5 productivity, 61–3 relation of firms to, 56–7 service sector, 61–3, 72–4 social policy in, 63 strengths and weaknesses, 57–8 technological revolution, impact of, 65–8 unemployment and union density, 154, 161, 164, 166, 167 welfare programs, 335 Padilha, E., 194 Partido Socialista Obrero Español (PSOE, Spain), 203, 204, 205, 299, 332 economic philosophy in 1980s, 293 electoral support of by unemployed, 310, 311 government, 224, 239 and unemployment, 293 Pauly, L. W., 97 pensions, 99 Pierson, P., 90, 98, 99, 101 polder model, 248, 253–8, 262–5, 334–5. See also Netherlands. Pontusson, J., 95, 146 Portugal, 3 collective bargaining, 34 corporatism in, 101, 107, 108 employment rates, 21, 23, 30 income distribution, 303 job creation rate, 46 political behavior and attitudes of 360

unemployed, 281, 283, 284, 285, 286 unemployment, long-term, 26 unemployment rates, 16, 17, 30, 175–7 Portuguese Communist Party (PCP), 178n6, 179 privatization, 333 and decline of unions, 186–7, 193, 195 productivity wage growth and, 124 public sector employment in OMEs, 60, 61, 332, 333 and union density, 161–2, 164–5 public service employment, 26–7, 34 Ramaswamy, R., 90 recession impact on unions, 190–4 OPEC oil shock as cause of, 10, 12, 122, 128, 157 and unemployment, 12–13 Rehn-Meidner model (Sweden), 64 Reich, S., 97 retirement, 34 as workforce reduction, 187 Rhodes, Martin, 2, 252, 335, 340, 343, 345 Rijn Schelde Verolme Affair, 247 Rodrik, D., 94 Rose, David, 209 Rothstein, Bo, 76 Rowthorn, R., 90 Ruigrok, W., 97 salários em atraso (delayed payment of wages), 179, 181, 182 Scharpf, F. W., 73, 96, 291 Schlozman, Kay, 275–6, 283 Schlüter government (Denmark), 31 Schmid, G., 112 Schmitter, P. C., 101, 102, 145 Scruggs, Lyle, 3, 98, 113, 331 SER. See Social and Economic Council.

Index service sector employment, 20, 340 in organized market economy, 72–4 productivity growth compared to industrial sector, 61–3, 184 Shonfield, Andrew, 7 Silva, Cavaco, 31 Soares, Mario, 179 Social and Economic Council (SER, Netherlands), 254, 255–6, 262 social democratic governments decline, 130–1 globalization of finance, 94, 95 tax competition, 95 and wage moderation in corporatist systems, 126 Social Democratic Party (Portugal), 179, 181 social democrats corporatist bargaining, 294 economic reform strategy, 295–6, 297, 302 philosophical changes in 1980s, 293 and political consequences of high unemployment, 291 Socialist Party (Portugal), 179, 181 social market economies, 161, 165 social policy, 63 social welfare programs in OME, 60–1 Soskice, D., 70, 145, 161, 165 Spain, 4 competitive corporatism, 107, 108, 110, 112 economic reform, 203–6, 293–302 employment rates, 21–2, 30 income distribution, 303 labor market segmentation, 33, 211–22 political behavior and attitudes of unemployed, 281, 284, 286 political consequences of unemployment, 309–24 unemployment rates, 16, 17, 20, 30, 204, 293, 294, 296, 301 unions, 206–8, 222–38

welfare and unemployment, 302–9 working-class fragmentation, 208–11 Stability and Growth Pact, 36, 44 STAR. See Foundation of Labor. Stekelenburg, Johan, 265 Stokes, Susan C., 314 Stoleroff, Alan, 3, 333 Streeck, W., 101 strikes, 196–8, 257, 258 as measure of union militancy, 146 Suárez, Adolfo, 292 Summers, Lawrence H., 12 Swank, D., 95 Sweden, 76 competitive corporatism, 112 employment, part-time, 29 employment rates, 21, 24 social corporatism, collapse of, 106, 131 unemployment, long-term, 26, 27 unemployment rates, 15, 17, 53, 150, 157 wage bargaining, 77, 102 wage growth, 124 wage growth/labor decline model, 135 Swenson, Peter, 71 Switzerland, 133 unemployment rates, 150 Tanzi, V., 94 taxation, 94–6, 342 Teague, P., 101 technological revolution as challenge for OME, 65–8 labor market implications of, 91–2 Thatcher government (U.K.), 31 reduction of welfare state under, 99 Thelen, Kathleen, 71, 104 Therborn, Goran, 335 Tobin, James, 341 Toledo Pact (Spain), 107 trade competion globalization of, 90–3 361

Index training programs. See education and training. Traxler, Franz, 131–2, 342 UCD (center-right party, Spain), 299 UGT. See União Geral dos Trabalhadores (Portugal); Unión General de Trabajadores (Spain). unemployed, political behavior, 331 data and measures, 277–8 extremism and xenophobia, 272, 276, 280–3, 288 hypothses, 276–7 interest in politics, 283–4 life satisfaction, 278–9, 287 political satisfaction, 279–80, 287 research on, 272–6 voter turnout, 275–6, 284–5, 330 in welfare states, 277, 285–7, 288 unemployment, 15–17, 29–30, 46. See also unemployed, political behavior; and specific countries. by age and gender, 24–6 class, segmented by, 216–22 and collective bargaining, 347–8 in era of EMU, 36–46 European compared with U.S., 10–11, 336–41 and income distribution, 307–9, 322 long-term, 26–8, 260 mitigated by family, 307, 309, 310, 324, 332 political consequences of, 309–24 rates, 15–20, 135, 148–54, 190–1, 247, 271, 293, 296, 301 reduction, 74, 343–4 structural or equilibrium rate, 11–12, 193 system reform, 100–1 trade-offs among welfare, wages, and, 293–302 and union power in liberal market theory, 146–8 362

and wages, 121–2, 126, 128 of women and youth, 299, 301, 338–9 unemployment insurance union administered. See Ghent unemployment insurance system. União Geral dos Trabalhadores (UGT) (Portugese labor organization), 178n6, 181, 182, 183, 194, 224, 225, 238–40 Unión General de Trabajadores (UGT) (Spanish labor organization), 33, 204 unions and labor organizations, 2–3, 335 in corporatist system, 103 decline of, 128–31, 173, 194–200, 206–8 density and unemployment rates, 146–8, 150–60, 155–7, 162–4, 166–8, 174–7, 331–2 extension of agreements to nonmembers, 263–4 and Ghent unemployment insurance systems, 157–60, 162–5 growth trends, 194–9 and insider-outsider divisions, 207–8, 225–7 institutional strength, 139 labor market segmentation and participation in, 227–37 political influence, 145–6, 177–8 problems of in Spain, 222–5 and unemployment, 174–7 and wage increases, 198–9 and wage moderation, 59, 125–7, 137–8, 294, 297, 324 and women in the labor force, 206 and working class fragmentation, 208–11 United Kingdom employment, part-time, 29 employment rates, 21, 22, 23, 30 social inequality, 108

Index unemployment rates, 16, 17, 30, 148, 150 United States job creation rate compared with EU, 13–15 unemployment and political behavior, 275–6 unemployment compared with Europe, 10–11, 150, 336–41 Urry, J., 101 van Agt cabinet (Netherlands), 259 van Veen, Chris, 259 Van Tulder, R., 97 Verba, Sidney, 275–6, 283 Visser, Jelle, 264 VNO-NCW (Netherlands employers association), 36 vocational education. See education and training. wage bargaining, 64–5, 107, 111. See also bargaining; wage moderation. wage growth, 296 European compared with U.S., 121, 122, 123 institutional explanations, 125–8 and labor decline model, 132–9 market explanations, 124–5 stagnation post-1973, 123, 128–31 wage moderation, 36, 59, 125–7, 137–8, 223–4, 249, 254, 257, 294, 297, 298, 334–5 labor market institutions, 121–2 wages, 3, 259, 260 delayed payment, 179

trade-offs among unemployment, welfare, and, 293–302 Wagner, Gerrit, 247 Wallerstein, Michael, 129 Wassenaar Accord, 247, 258–62 Weiss, L., 97 welfare programs, 2, 4, 298–9 trade-offs among unemployment, wages, and, 293–302 welfare reform, 309 competitive corporatism, 109, 112 difficulty of in context of high unemployment, 291 welfare state, 332 challenges of globalization, 87–8, 92, 98–112 and political behavior and attitudes of unemployed, 285–7, 288 reduction of in Netherlands, 261 and unemployment, 292 unemployment mitigated by, 332, 333 Western, Bruce, 331, 336–7 Wolinetz, Steven, 4, 334 women employment rate, 184, 186, 297 labor unions, decline of, 195, 207 part-time employment, 29, 73, 245 unemployment rates, 24, 25, 26, 191, 338 Wood, A., 90 Wood, Stewart, 71 workforce flexibility, 187–90 World Bank, 89 xenophobia, 330

363

Other Books in the Series

Stefano Bartolini, The Political Mobilization of the European Left, 1860–1980: The Class Cleavage Carles Boix, Political Parties, Growth and Equality: Conservative and Social Democratic Economic Strategies in the World Economy Catherine Boone, Merchant Capital and the Roots of State Power in Senegal, 1930–1985 Michael Bratton and Nicolas van de Walle, Democratic Experiments in Africa: Regime Transitions in Comparative Perspective Valerie Bunce, Leaving Socialism and Leaving the State: The End of Yugoslavia, the Soviet Union and Czechoslovakia Ruth Berins Collier, Paths Toward Democracy: The Working Class and Elites in Western Europe and South America Donatella della Porta, Social Movements, Political Violence, and the State Gerald Easter, Reconstructing the State: Personal Networks and Elite Identity Roberto Franzosi, The Puzzle of Strikes: Class and State Strategies in Postwar Italy Geoffrey Garrett, Partisan Politics in the Global Economy Miriam Golden, Heroic Defeats: The Politics of Job Loss Merilee Serrill Grindle, Changing the State Frances Hagopian, Traditional Politics and Regime Change in Brazil J. Rogers Hollingsworth and Robert Boyer, eds., Contemporary Capitalism: The Embeddedness of Institutions Ellen Immergut, Health Politics: Interests and Institutions in Western Europe Torben Iversen, Contested Economic Institutions Torben Iversen, Jonas Pontusson, David Soskice, eds., Unions, Employers, and Central Banks: Macroeconomic Coordination and Institutional Change in Social Market Economies Thomas Janoski and Alexander M. Hicks, eds., The Comparative Political Economy of the Welfare State Robert O. Keohane and Helen B. Milner, eds., Internationalization and Domestic Politics Herbert Kitschelt, The Transformation of European Social Democracy Herbert Kitschelt, Peter Lange, Gary Marks, and John D. Stephens, eds., Continuity and Change in Contemporary Capitalism Herbert Kitschelt, Zdenka Mansfeldova, Radek Markowski, and Gabor Toka, Post-Communist Party Systems

David Knoke, Franz Urban Pappi, Jeffrey Broadbent, and Yutaka Tsujinaka, eds., Comparing Policy Networks Allan Kornberg and Harold D. Clarke, Citizens and Community: Political Support in a Representative Democracy David D. Laitin, Language Repertories and State Construction in Africa Mark Irving Lichbach and Alan S. Zuckerman, eds., Comparative Politics: Rationality, Culture, and Structure Doug McAdam, John McCarthy, and Mayer Zald, eds., Comparative Perspectives on Social Movements Scott Mainwaring and Matthew Soberg Shugart, eds., Presidentialism and Democracy in Latin America Anthony W. Marx, Making Race, Making Nations: A Comparison of South Africa, the United States and Brazil Joel S. Migdal, Atul Kohli, and Vivienne Shue, eds., State Power and Social Forces: Domination and Transformation in the Third World Wolfgang C. Muller and Kaare Strom, Policy, Office, or Votes? Ton Notermans, Money, Markets, and the State: Social Democratic Economic Policies since 1918 Paul Pierson, Dismantling the Welfare State?: Reagan, Thatcher and the Politics of Retrenchment Marino Regini, Uncertain Boundaries: The Social and Political Construction of European Economies Yossi Shain and Juan Linz, eds., Interim Governments and Democratic Transitions Theda Skocpol, Social Revolutions in the Modern World David Stark and László Bruszt, Postsocialist Pathways: Transforming Politics and Property in East Central Europe Sven Steinmo, Kathleen Thelan, and Frank Longstreth, eds., Structuring Politics: Historical Institutionalism in Comparative Analysis Sidney Tarrow, Power in Movement: Social Movements and Contentious Politics Ashutosh Varshney, Democracy, Development, and the Countryside Elisabeth Jean Wood, Forging Democracy from Below: Insurgent Transitions in South Africa and El Salvador