- U - t - H I .-Az~ ~ ~ ~~~~ - F _ * _ r t t ,, , _ _ _ b _ _ IS _ 4: >- ---- g '=5i -- S - ~~~~ I : ' g 5-r ~ w1 - a x ~ - Other EDI Development Studies Does Privatization Deliver? Highlights fiom a World Bank Conference Edited by Ahmed Galal and Mary Shirley ISBN 0-8213-2589-2 7The Adaptive Economy: Adjustment Policies in Small, Low-Income Countries Tony Killick ISBN 0-8213-2125-0 Financial Regulation: Changing the Rules of the Game Edited by Dimitri Vittas ISBN 0-8213-21234 The Distribution of Income and Wealth in Korea Danny Leipziger and others ISBN 0-8213-2124-2 Public Enterprise Reform: The Lessons of Experience Mary Shirley and John Nellis ISBN 0-8213-1811-X Also available in French and Spanish Privatization and Control of State-Owned Enterprises Edited by Ravi Ramamurti and Raymond Vernon ISBN 0-8213-1863-2 Finance at the Fronder: Debt Capacity and the Role of Credit in the Private Economy J D. Von Pischke ISBN 0-8213-1818-7 EDI DEVELOPMENT STUDEES Labor Markets in an Era of Adjustme'nt Volume 2 Case Studies Edited by Susan Horton Ravi Kanbur Dipak azumdar The World Bardc Washington, D. C. Copyright 0 1994 The International Bank for Reconstruction and Development / THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. AU rights reserved Manufactured in the United States of America First printing July 1994 TheEconomic Development Institute (EDI) was establishedby theWorldBank in 1955 to train officials concemed with development planning policymaking, investment analysis, and project implementation in member developing countries. At presentthe substance of the EDI's work cmphasizes macroeconomic and sectoral economic policy analysis. Through a variety ofcourses,seninars, andworkshops,mostofwhich aregiven overseasin cooperation with local institutions, the EDI seeks to sharpen analytical skdlls used in policy analysis and to broaden understanding of the experience of individual countries with economic develop- menL Although the EDI's publications are designed to support its training activities, many are of interest to a much broader audience. EDI materials, including any findings, interpre- tations, and conclusions, are entirely those of the authors andshiould not be attnbuted in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Diectors or the countries they represenL Because ofthe informality of thisseries and to makethe publicationavailable with the least possible delay, the manuscripthas not been edited as fullv as would be the case with a more formal document, and the World Bank accepts no responsibility for errs. Some sources cited in this book may be informal documents that are not readily available. The material in this publication is copyrightedL Requests for permission to reproduce portionsofitshouldbesenttotheOfficeofthe Publisheratthe addressshown inthecopyright notice above. The World Bank encourages dissemination of its work and will normally give permission promptly and, when the reproduction is for noncommercial purposes, without asling a fee. Permission to copy portions forclassroom use is granted thmugh the Copyright Clearance Centerlnc, Suite 910, RosewoodDdve,Danvers,MassachusettsOl923, U.S. A. The bacdist of publications by the World Bank is shown in the annual Inder of ublca- dions, which is available from Distnbution Unit, Office of the Publisher, The World Bankl 1818 H Stret, N.W., Washington, D.C. 20433, US.A, or from Publications, Banque mondiale, 66, avenue d'Idna, 75116 Paris, France- Susan Horton is an associate professor of cconomics at the University of Toronto; Ravi KanburistheWorldBank'sresidentrepresernative in Ghana;andDipakMazumdaris alabor markets specialist in the World Bank's Office of the Chief Economist, Africa Library of Congress Cataloging i-Publication Data Horton, Susan. Labor markets in an era of adjustment / Susan Horton, Ravi Kanbur, Dipak Mazumdar. p. cam-(EDI development studies) Includes bibliographical references. Contents: v. 1. Issues papers-v. 2. Case studies. ISBN 0-8213-2680-5 (v-l).-ISBN 08213-2681-3 (v. 2) 1. Labor market-Developing countries-Congresses. 2. Structural adjustment (Economic policy)-Dteloping countries-Congresses- 3 Labormarket-Developingcountries-Casestudies-Congresss 4. Strucural adjustrnent (Economic policy)-Developing countries- Case studies-Congresses. I.Kanbur, Skl Ravi U. Mazumdar, Dipak, 1932- .HIl ritle. IV. Series. 1D5852 331.12f09172'6-dc2O 93-34978 CIP CONTENTS VOLUME 2: CASE STUDIES Foreword v Acknowledgments vii Preface ix Abbreviations and Acronyms iii 1. Labor Markets in an Era of Adjustment An Overview 1 SusanHorton, RaviKanbur,an DipakMazumdar 2. Argentina 61 Luds A. Riveros and Carlos E. Sdnchez 3. Bolivia 99 Susan Hof-on 4. Brazil 143 MA Louise Fox, Edward Amadeo, and Jose Marcto Camargo 5. Chile 169 Luis A. Riveros 6. CostaRica 217 T. H. Gindling and Albert Berry 7. COte d'Ivoire 259 Richard BtundeL4 Christopher Heady, and Rohizton Medhora 8. Egypt 317 Ragui Assaad and Simon Commander 9. Ghana 357 P. Beaudry and N. K. Sowa 10. Kenya 405 William J. Mime and Monica Neiz,err Hii iv Contents 11. Malaysia 459 DipakMazumdar 12. The Republic of Korea 535 DipakMammdar 13. Thailand 585 Chalongphob Sussangkarn Index 613 VOLUME 1. ISSUES PAPERS Labor Markets in an Era of Adjustment An Overview Susan Horton, Ri Kanbur, and DipakYMazumdar 1. Recent Developments in the Developed-Country Literature on Labor Markets and the Implications for Developing Countries .ea i-PazdAzam 2. Labor Market Distortions and Structural Adjustment in Developing Countries AI4andra CaoEdnds and Sebastan Edwards 3. The Poverty Effects of Adjustment with Labor Market Imperfections Tony Addison and Lionel Demery 4. Wage Indexation, Adjustment, and Inlaton Micheelif. Devereux 5. The Long-Run Consequences of Short-Run Stabilization Policy Edwwrd F B 5 ,1 10a .011 -.0001 -.268 region, sector, formal, .278 tenure c 5 .085a .027 -.0003 .272 urban, others .281 Kenyab 1977-78 0073 .0759c -.0O08c -.1188 occupation, city, .413 (.0033S2) age, education 19-6 .0222 .0784c *,0008c -.1587 .537 (.0057S2) Malayslac 1970 Malay men ,142 .093 -.0012 n.a. n.a. .451 Malay women .147 .071 -.0011 n.a. n.a. A21 Chinesc mcn .139 .110 .001 n,a. n.a. .521 Chinese women .133 .680 -.0007 n.a. n.a. .437 1987 Malay men .171 I11 -.0014 n.a. n.a. .439 Malay wormien .196 .110 -.0016 n.a. n.a. .421 Chinese men .153 .098 -0102 n,a. n.a. .437 Chlnesc womnn .152 .076 -.0009 n.a. n.a. .326 n.a. = not available n. Secondary school dummy. b. Urban only, c. Age. Sources: Country studies, except Cdtc d'[voiro source is van der Gaiag anid Vijvorberg (1989). Labor Markets in an Era ofAdjustment: An Overview 41 employment, in contrast to the results for Asia and Latin America, where structural adjustment often adversely affected earnings and unemployment for the educated. Unemployment could be relatively persistent: of those seeking employment in 1985, 81 percent were still unemployed in 1986, although 42 percent of the original group had stopped looking. Finally, the study had somne interesting results on the effects of crop price indexes on work behavior in rural areas. Increases in these indexes had a positive effect on work supply both for those who were working and in school in the first of the two survey years, but a negative effect on work supply for those in fuill-time education in the first year. In other words, crop price increases could increase effort, but not at the expense of interrupting human capital acquisition, an interesting finding. The use of earnings functions is obviously a useful direction for firther work on labor markets and adjustment, and in this respect the technique of dividing the sample (as used in the Ghana study) seems a promising wrt of teasing out trends from a single cross-section of data, which zi"At be particularly useful for African countriea Labor Market Institutions Two issues papers deal with labor market institutions, Devereux's on wage indexation (vol. 1, chapter 4) and Nelson's on political economy issues (vol. 1, chapter 7). The latter paper focuses on the effect of - unionism, both private and public, on labor market flexibility. It also discusses economic and political factors that affect how militant or cooperative labor movements are likely to be. Nelson argues the existence of theoretical reasons that explain why unions in developing countries might be more militant than in developed ones. The relationship between union organization and militancy is an inverted U-shape: weak unions exhibit a low level of militancy, and very strong centralized unions are also less militant as they can no longer consider only sectoral gains. Developing country unions fall in the middle, with some strongly organiZed sectors, but no strong central union body. Unions in most developing countries do not fit the corporatist model, where wage gains are traded off for better employment security and where labor may take account of the macro impact of sectoral wage demands. Another feature of unions in 42 Susan Horton, RaviKanbur, andDipakMazumdar developing countries is the greater role of public sector unions due to the greater share of public sector employment in total formal employment. A feature of the public sector is the greater difficulty experienced in laying off workers and the large severance payments offered. Economic factors may affect labor's intransigence: they tend to show more concem for wages during upswings and more concem for employment protection during downswings, although unions foreseeing times getting worse may try to grab what they can early in the downswing. Political factors also matter. authoritarian regimes tend to use coercion more than democratic regimes, with some exceptions on both sides. The stage of the electoral cycle matters, as does labor's role in the political and party process. Labor may be attached to one party in a polarized system, or have access to more than one party in a more open system, or be largely excluded from the political arena. Likewise the regime's degree of stability matters, with new democracies in particular being susceptible to the revolution of rising expectations. Nelson makes the important point that successful adjustment in the long run not only requires investor confidence in the government's long-run ability to fulfill its promises, but also the confidence of the labor movement. The degree of equity in a society may be an important ingredient in sustaining such confidence. The Latin American country studies dwell at length on labor market institutions: unions, indexation, minimum wages, legislation on benefits and job security, and segmentation. For the African countries these institutions receive less coverage in the country studies, altho-gh they do exist. As the Kenya study shows, however, it is one thing for the institutions to exist, and another for them to be effective, and their force tends to be weakened by the highly elastic labor supply to urban areas in Africa. It is also likely that the much lower proportion of urbanization and of for, nal sector employment makes a difference. The Asian countries have some similar institutions (two-year wage contracts in Malaysia and the same kind of long-term contract/temporary labor division in Korea as in Brazil). However, the role of unions in Asia is dearly very different from their role in Latin America and Africa. Labor Markets in on Era ofAdjunent: An Overview 43 The five Latin American country studies provide an interesting contrast in terms of the alleged effect of labor market institutions in causing rigidities in the labor market. In three of the countries (Argentina, Brazil, and Costa Rica) the institutions remain strong despite the economic crisis, whereas in the other two (Bolivia and Chile) they have been suostantially weakened and/or dismantled. Some of the country authors criticize these institutions. For Argentina and Brazil they argue that they impeded adjustment and labor market mobility, and in Chile they receive partial blame for the painful nature of the recession and ensuing high unemployment. At the same time the Bolivian and Costa Rican cases are interesting counterpoints. In Costa Rica labor institutions survived relatively unscathed, for example, over 500 minimum wages are legislated, and are generally enforced, which did not prevent moderate adjustnent. In Bolivia much labor legislation was dismantled and large-scale. labor shedding occurred without as yet strong recovery. To some extent it sterns that labor market institutions are often a symptom of underlying political and economic difficulties, which make adjustment difficult, and the institutions are unfairly blamed for causing problems. The Brazil study describes labor market institutions in some detail. Unions are very strong (in the form in which they reemerged during the democratization period from the late 1970s onward), and are linked to political parties along the lines of the corporatist state discussed by Nelson. They combine strong plant-level organization with a previously legislated strong centralized structure, which allows them to transmit bargains struck at the best organized plants to national level. Wage indexation is perhaps the most sophisticated in Latin America, with monthly adjustments. Job security legislation used to be an important hindrance to mobility, but the setting up in 1964 of a fund (to which employers contribute) to provide severance pay has eased the problem. Tradables predominantly hire formal sector (that is, signed contract), unionized workers, whereas nontradables hire all types of workers, formal and informal, unionized and nonunionizedL Argentina has many of the same institutions. The author -links union strength to inward-oriented economic policy, since the oligopolistic nature of employers demands an equally centralized representation for labor. The author also mentions a compulsory wage 44 Susan Horton, RaviKanbur, andDipakMazumdar policy, whereby bargains struck by the unions are obligatory for all firms, which he argues harmed small finns. One difference from the Brazil case is that the main exportable in Argentina is food, and unionization is therefore concentrated in nontradables or importables. This arguably has been a major hindrance in changing the relative price of tradables and nontradables. One difference in Costa Rica is that although legislation is equally strong, unions are relatively weak, having been broken in an unsuccessful face-off with Standard Fruit in the 1970s. Wage indexation in Costa Rica, far from being an impediment to desirable relative price changes, is given much of the credit for allowing a real wage decline at a critical point following devaluation. Since indexation was imperfect, real wages fell, but by an apparently impartial mechanism. This tactic, however, can only be used infrequently, and Brazfl, for example, is no longer able to make such gains from unanticipated inflation. Two Latin American countries undertook major labor market reforms. Chile between 1973 and 1975 eliminated unions and job security and removed much of the force from minimum wages, benefits, and wage indexation mechanisms (the government actually cheated on the price index used for wage indexation). However, the author argues that lack of labor legislation during 1973-79 was detrimental to growth because employers feared that the law, once reinstated, would be unduly favorable to labor. Bolivia, the other Latin American severe adjustment case, likewise removed similar institutions, with the exception that wage indexation had never been particularly important and had not survived the hyperinflation as an institution. Job tenure was ended and job security reduced, thus allowing labor shedding- The government stepped out of previously centralized wage bargaining. In both Bolivia and Chile the public sector shed a substantial amount of labor, equal to 25 percent of Bolivia's public sector labor force and 3 percent of Chile's total labor force (the author does not specify as to whether total urban or total urban plus rural is meant). Comparisons between the Latin American countries in terms of the success of adjustment are instructive. Contrasting, for example, the relatively successful adjustment in Costa Rica and the problematic one Labor Markets in an Era ofAdjument An Overview 45 in Bolivia, evidently dismantling labor institutions is neither necessary (Costa Rica) nor sufficient (Bolivia) for successful adjustment. Another interesting comparison is between Brazil and Costa Rica. In Brazil large political-economic tensions exist, such that consensus over the division of output is lacking, which causes continual inflationary tendencies (tensions that similarly pushed Bolivia over the brink into hyperinflation). Although wage indexation has sometimes been blamed for perpetuating Brazil's inflation, it is more a symptom of the defensive ability of one of the groups engaged in underlying conflict In Costa Rica, by contrast, a higher degree of social consensus allowed a union-backed president to undertake some of the painful initial steps toward successful adjustment, in which wage indexation actually helped the process. The Asian countries also have institutional structures in the labor market The Korean govemment has followed a highly interventionist policy with respect to unions. The right to strike was banned in 1971 and only recently reinstated, and unions need government permission to undertake collective bargaining. The author argues that wage and productivity trends and their consequent effect on unit costs has been crucial in Korea's export success. In this respect the govermnent was heavily involved in ensuring that 'wages did not get ahead of productivity, and at the same time that workers did share in the frits of higher productivity. Increasing union autonomy and increasing strikes in the late 1980s may herald a change in the so far virtuous productivity and wage nexus in Korea. In Malaysia union power is similarly limited. The level of unionization is low, less than 25 percent in manufacturing, and unions are banned in some sectors. Paradoxically unions are strongest in the plantation sector, where wages stagnated in the 1980s. Malaysia has relatively long (three-year) wage contracts, which may have hindered adjustment. Unions in Thailand are also weak except in the public sector. In both Malaysia and Korea the importance of bonuses in earnings (around 30 percent of pay in Korea and 15 percent in Malaysia) has been argued to cause flexibility, since earnings and profits are related- Latin American countries also have bonuses, but less related to productivity and profits than to Christmas, seniority, and so on. 46 Susan Horton, Rav Kanbur andDipakMazrsndar Although studies of Latin American countries frequently blame labor market segmentation (formallinformal) as a problem, some kinds of segmentation also exist in the Asian countries. In Korea labor is divided into permanent, temporary, and casual, and much labor market adjustment falls upon the casual and temporary workers, particularly women- Another type of segmentation between large and small firms is also quite marked in Korea, and small firms tend to pick up the slack during recessionary periods. Segmentation also seems to persist over time, although taking the form of a widening gap in the human capital levels of large as compared to small firms, rather than a widening of wage differentials. Fimaly public sector employment and adjustment is a topic worthy of separate study in its own right The growth of public sector employment as an initial response to economic crisis is mentioned in many of the studies (all of the Latin American studies, Egypt, and Malaysia). The eventual need to shed public sector labor was a difficult undertakdn& Bolivia, Chile, Costa Rica, and Ghana have bitten the bullet, Argentina has been unable to; and in Egypt, Kenya, and Malaysia adjustment took the form of a substantial slowdown in government hiring. In the latter three countries one consequence discussed was a rise in educated unemployment, particularly of women in Egypt and Malaysia, where educated women have few private sector alternatives. The relative decline in public sector wages observed in almost all the countries reflects the greater difficulty of adjusting labor quantity in the public than the private sector. Consequences of Labor Market Adjustment Labor market adjustment has consequences for income distribution and poverty, and on long-mn growth. The country study authors were asked to consider these, paying particular attention to the role of women in labor markets. Income Distribution As Addison and Demery show (vol. 1, chapter 3), theoretical discussion of the effects of adjustment on poverty yields ambiguous predictions. Their paper begins Nwith the standard Salter-Swan account of expenditure reduction and expenditure switching, and works out LaborMarerr in an Era ofAdjusznnmeAn Overview 47 wage and employment effects, assuming competitive labor markets. These wage and employment effects are then fed through a poverty index, but yield ambiguous predictions. The rest of the paper examines how these effects are modified by the introduction of different labor market imperfections. The first case is where there exists an economywide "quantity rationing" framework, that is, unemployment can persist. In this case the discussion of poverty becomes more complicated, since one must consider poverty among those employed in tradables, those employed in nontradables, and those unemployed. In this case although a devaluation may increase poverty because it shifts workers to the tradable sector, where greater poverty is assumed, and because it lowers the real wage, it will decrease poverty because of the unemployment reduction. Thus, ambiguity in predictions persists, but of a different type thn before The paper then moves on to discuss partial labor market imperfections, dividing the labor market into a formal and an informal sector. The analysis is similar to that by Edwards (1988) and Edwards and Edwards (vol. 1, chapter 2). The authors consider different types of wage inflexibility and trace out the consequences for sectoral employment, wages, and unemployment These are again fed through a poverty index Ambiguity is again the order of the day, although the analysis does illuminate the different components. A third variant is where barriers exist to entry into the formal labor market. Here Addison and Demery (voL 1, chapter 3) argue that an expenditure switching policy is quite likely to reduce poverty if barriers to entry into nontradables or tradables exist The fourth and final case is where labor market imperfections exist in both sectors, and the authors distinguish between unemployment and employment in informal tradables, formal tradables, informal nontradables, and formal nontradables. They follow through the real wage and labor allocation consequences of expenditure switching, and again feed them through the poverty index. They conclude that the effects of switching under these assumptions seem to be the most promising as far as poverty reduction is concemed. Tracing the effect of adjustment on poverty and income distnrbution empirically is no easier than doing so theoretically. 48 SusanHorton, RaviYCanbur, and DipakLMazumdar Asking the counterfactual question as to what happened during adjustment as compared to what would have happened otherwise is difficult, as many countries were on unsustainable courses. The data available also affect the conclusions that one can reach. It is usually more difficult to obtain information on overall economywide changes in income distribution from nationwide income-expenditure surveys than to obtain results on the urban distribution of earned income from labor force surveys. However, if real wages fall by more than GDP and urban-rural differentials change, then the latter data only tell part of the story. We focus here on relative earnings distnrbution. Several studies also document increases in poverty, unsurpising as a consequence of economic crisis. For Africa almost no time series data exist with which to make comparisons. The Kenya study does cite 7NICEF's finding that the share of the bottom 10 percent declined. For Egypt no distnrbution data are available after 1981/82. Changes in urban-rural income differentials are of great interest in the case of Africa and are the focus of studies elsewhere (Jamal and Weeks 1987), but country- studies here lacked the data to examine the issue. In Latin America income distnbution is a key issue related to the political economy of the economic growth process, and all the studies provided data. Brazil's income distnbution has long been of interest given that inequality increased during the long boom "economic miracle" period between 1967 and 1974, when there was a type of structural adjustment as the economy became more open. Some improvement in income distribution is evident between 1974 and 1981, with a worsening duiing the recession and stabilization (1981- 85), and since then a slight recovery- One interesting finding is that interregional equality increased during structural adjustment, which hit harder at the more affluent urbanized south than the more rural northeast For Chile the pattern was somewhat similar, but more exaggerated, with a sharp increase in 1974-76 accompanying the start of adjustment, the Gini remaining constant during 1976-79, increasing again in 1979-84, and since then decreasing slightly, but to a level much higher than at any time during 1960-74. It is not surprising that distribution worsened so much, given the massive cuts in real wages LaborMarkets in an Era ofAdjustrment: An Overview 49 and the very high unemployment levels. The measured changes may be offset somewhat by changes in social expenditures. In Argentina income distribution also worsened during the stop-go cycles (although the only data available are for income earners in Buenos Aires during 1974-88). The top two deciles gained at the expense of all others. For Bolivia and Costa Rica data are more scanty and knowing exactly what happened is harder. In Costa Rica inequality may have increased between 1971 and 1983 (before adjustment), but after the onset of adjustment different data sources give conflicting trends.. For Bolivia the data are also not very good, but suggest a possible improvement between 1982-85, when informal sector wages rose relatively during the hyperinflation, but by 1988 distribution had reverted back- to 1982 levels. In Asia, income distribution may have improved in Malaysia and worsened in both Thailand and Korea In Malaysia resources were put into agriculture, including food agriculture, whereas in Korea policy focused for at least some of the period on heavy industry, and in Thailand little was done about. the problem of urban primacy (concentration in Bangkok)- Women and Labor Market Adjustment Much of the literature on women and structural adjustment has concentrated on the effects of structural adjustment on women. Collier and others (vol. 1, chapter 6), using evidence from Africa, examine the opposite issue, namely, how women's economic mobility may affect the success of adjustment. They argue that women face constraints not only in the labor market and in access to education, but also in credit markets, which may affect adjustment. In particuIar, women in Africa are frequently concentrated in food production. The authors present three possible cases relevant to adjustment. Food may be a tradable, in which case its output should expand with adjustment; it may be a nontradable, in which case output should contract; or it might be nontradable in rural areas but tradable in urban areas, in which case food marketing (again frequently a female preserve, at least in West Africa) would need to expand. If food crops are to contract, this requires a reallocation of women's labor into other activities, and if they are to expand, this requires women's access to credit. In either 50 Susan Horton, Ravd Kanbur, and DipakMazumdar case, constraints on women's flexibility will hinder the success of structural adjustment. Collier and others therefore urge that government policies should focus on relaxing constraints to women's economic activities. AMother reason cited in favor of this strategy is that it also improves household income security if higher women's incomes offset the loss of men's jobs in the formal or government sectors during adjustment, although they do not consider the potential costs involved, such as women's responsibilities for children. The paper by Collier and others also discusses women in South Asia, again focusing on women as participants in, rather than victims of, structural adjustment It deals with both rural and urban activities of women, and draws somewhat on the earlier experiences of women in export-oriented industries in East -Asia Bardhan sees structural adjustment as potentially altering the existing U-shaped pattern of female labor force participation with education: in South Asia women tend to participate either with very low education in menial and low- productivity activities, or in high-skilled, high-education activities. The author argues that adjustment may increase the demand for labor- intensive industry output, requiring women workers with medium education, with resulting beneficial effects on reduced fertility and increased incentives for female education. Adjustment may also involve costs for women, such as those where the male family members or the whole family migrate,. and the costs imposed particularly on women's time when social infrastructure deteriorates- Like Collier, she sees a role'for government in relaxing the constraints on women's activity. Labor market legislation aimed at protecting women has ended up tending to exclude them from the formal sector. Bardhan foresees benefits to women in selective deregulation of some sectors in India, such as electronics. Another aspect of the paper by Collier and others focuses rather more on the effects of structural adjustment on women. In Latin America, studies on women seem to focus mainly on labor force participation, and little information is available on trends in relative earings. Women's labor.force participation has been increasing, partly due to sectoral shifts, in particular, increased employment in the service sector, but largely due to higher participation within sectors. Labor Markets Ui an Era ofAdjusiment An Overview Sl The participation increases vary somewhat across countries. The authors undertake econometric analysis for Chile, which suggests that unemployment that accompanies structural adjustment does not have differential effects on discouraging female and male labor force participation. One interesting avenue they suggest for future work is to examine how increased female participation fits in with the trend in much of Latin America toward increased informalization of the labor force. The country studies concentrate more on the effects of structural adjustment on women. As Collier and others argue, the effects are likely to depend on the preceding sectoral distribution of women workers and on the effect on participation rates. However, the likelihood exists that women workers' more tenuous attachment to the labor force means that they are more likely to lose jobs during periods of labor shedding. The countiy studies do not give a single story, although there seems to be a lot of evidence of adverse impacts, but the data are not very complete. Even for the United States, where data are available, understanding how male/female wages, for example, had changed over time due to changes in female labor force participation was difficult. For the deve!oping countries female labor force participation has exhibited trend changes plus cyclical responses due to crisis. Tracing the effects on women's welfare is even harder if most women live in households with men. Although the effects on female- headed households are less ambiguous to interpret from the data, this was a topic well beyond the scope of the country studies. The Ghana study documents that women suffered rather more from structural adjustment than men as they were concentrated in the informal sector, which tended to absorb excess labor. Women are also predominantly in food crop agriculture, whereas resources have gone instead to cash crops. In C6te d'Ivoire, insofar as education had a positive effect on the probability of remaining in employment or of entering employment, and women tend to have less education, they are likely to have faced disadvantages. The Egypt study documents an adverse effect on women due to the lengthening queue for government employment, and the more limited private sector alternatives available to women. 52 Susan.fIorton, RaiKanbur, andDipakMazmdar In Bolivia the male/female differential fell between 1981 and 1987 as measured from earnings functions, although aggregate data suggest the opposite (the difference is perhaps explained by changes in participation rates). Although anecdotal evidence suggested that labor shedding from the, formal sector-was to the detriment of women, who are more costly workers in terms of benefits, this may have been offset by much of the employment loss being focused in mining, a male- dominated sector. In Costa Rica the male/female earnings differential increased during the crisis and decreased thereafter, which the authors attnrbute to rising female participation during the crisis (added worker effect), where the female entrants were less well quaified. In both Bolivia and Chile the emergency employment schemes explicitly targeted male workers, at least initially, and in Chile public sector hring in the early part of the crisis also favored men. In Malaysia some evidence suggests that women last ground during the recession due to the firing of labor in a weaker position in the labor market; however, a trend increase in female wages is evident over the 1970s and 1980s. The relative earnings of Malay women in particular increased between 1970 and 1984, and the returns to female education and experience rose absolutely and relative to the same returns for men. However, these gains were all reversed in the 1984- 87 recession. Nevertheless, Malaysia differs from some of the other countries studied in that women are a higher proportion of wage employment th;an of self-employment, and are concentrated in some export industries, such as electronics. In Korea women are at a disadvantage, crowded into low paying, white collar sectors, and providing a disproportionately high share of family workers, the most disadvantaged group in the labor force. Female participation rates are also surprisingly low in Korea compared to other East and Southeast Asian countries. Women also tended to lose out in the recession. Whereas male employment shifted continuously toward the permanent category, this proportion declined for women during the recession. Effects on Long-Rum Growth Most of the issues papers focus on demand side effects of adjustment and the labor market. Buffie's (vol. 1, chapter 5), by contrast, highlights the supply side consequences of fiscal contraction, LaborMarkets in an Era of Adjustment An Overvew 53 and hence the impacts on long-run growth. Demand side complications are abstracted from by assuming that the economy is small and open. Two traded goods, agricultural exports and manufactures, are produced using labor and capital. Manufacturing also requires an intermediate input, which is supplied by the public sector. Labor employed in the public sector and in manufacturing is paid a higher than competitive wage, and the rest of the labor is underemployed in agriculture. Buffle assumes a fixed wage differential between the modem and the informal/agricultural sectors. Capital accumulation dynamics are also mcdeled. Human capital is modeled by distinguishing between skilled and unskmilled labor. Skilled labor growth is determined by human capital investment by the government. If factors are complementay, then the productivity of unskldled labor declines when investment is cut, as does the productivity of capital. Overall, Buffie shows that disinvestment in human capital leads to capital decumulation. The paper suggests two broad policy lessons. First, productive government investments in human capital should be protected, which requires broadening the tax base. Second, a more gradual approach to adjustment is likely to entail fewer adverse impacts on productive investments vital for long-mn growth. To some extent the topic of adjustment and long-rn growth is a difficult one to study empirically, since many countries are still grappling with short- and medium-term issues, but some of the studies provide information on investment, in particular, human capital investment, as discussed by Buffie. The Kenya and COte d'Ivoire studies discuss falling investment, but do not blame labor markets. The Argentina study throws the blame for stop-go cycles onto the labor market's inability to allow prices of tradables to rise relatively in a srstained way, thereby harming long-run growth. Similarly, in Chile a lack of labor legislation and fears of a return to previous laws that favored labor are assigned the blame for lack of investment. As regards human capital investments, the Costa Rica study documents a sharp drop in school enrollment during the crisis, especially at the secondary and technical levels, with likely adverse effects on growth and distribution. By contrast, no such effect was predicted from cross-section regressions for the CBte d'Ivoire. In Asia 54 SusanHorton,RaviKanbur, andDipakMazundar where short-run problems of adjustment have been largely solved, the studies had more room to focus on long-run issues. The Malaysia and Korea studies examine changing returns to education, and the Thai study examines potential labor market skill mismatch issues. Conclusion This overview has summarized theoretical predictions and country study experience on two important topics related to labor markets and adjustment. First, how well have labor markets functioned, and have they assisted or impeded macro adjustment efforts? Second, what were the effects of some of these adjustments on the labor market? With respect to the issue of labor market functioning, labor markets have at least three allocative functions: they match workers to employment in such a way that overall unemployment levels and real wages matter, they allocate workers between sectors, and match worker skills to job requirements so that relative wages and employment matter, both for economic sectors and for skill categories; and they provide incentives for intertemporal allocation of resources, specifically for human capital accumulation in edlucation and firm- specific training. Applying these three criteria to the often descriptive country studies to assess how well or how badly labor markets performed is not easy. By and large individual country authors argue that the labor markets performed well; although authors of the studies for the big three Latin American countries, Argentina, Brazil, and Chile, were more critical. Theory suggests that labor market rigidities are only one of three possible reasons for unemployment. With the exception of Chile, the countries have not had prolonzged unemployment despite severe recession, however, cyclical increases have occurred. This fits with the presumption that in developing countries without unemployment insurance schemes, unemployment is not an option for primary household earners unless the household is unusually wealthy. The evidence on real wages casts considerable doubt on theoretical concerns about aggregate real wage rigidity and labor market inflexibility as a hindrance to adjustment. Real wage declines have been dramatic, and often far greater than the fall in GDP. For some Labor Markets in an Era ofAdjuslment: An Overview 55 countries the real wage declines may have been excessively large and led to a fall in domestic demand, which inhibited recovery. With regard to the sectoral employment shifts, these have generally been in the desired direction, that is, toward tradables, although this has generally meant that agricultural employment has increased relatively and manufacturing employment declined in all but the most successful countries. Shifts of employment into services and comnerce are, however, indicative of weak GDP growth, and hence growth of labor demand. Sectoral wage changes have aiso been largely in the appropriate direction, although little information is available on agricultural wages. The decline in relative government wages is one factor causing relative wages in nontradables to decline. Finally, on the intertemporal aspect, the evidence is a little more mixed. In Costa Rica the evidence showed that the recession had induced decreases in school enrollment, whereas in C6te d'Ivoire econometric results suggested that increases in crop prices, which would help adjustment, would not lead to parents pulling their children out of school. Earnings fimctions for Bolivia. Costa Rica, and Malaysia showed that returns to all formal sector characteristics including education and experience declined during adjustment, and in that government relative wages declined universally, and government tends predominantly to hire the more educated, this would decrease the incentives to acquire schooling. The country studies did not discuss another human capital issue, namely intemational migration, although for at least three of the countries- Cote d'Ivoire, Egypt, and Ghana-this was important. The country. studies also explicitly discussed labor market institutions, thought to be a source of rigidity. One possible interpretation is that where these institutions lack binding force, whether because of elastic labor supply (Africal or weak unions (Asia and perhaps Costa Rica), they were not perceived as obstacles to adjustment. Nevertheless, dismantling of the institutions and weakening of the unions as in Bolivia does not seem to be sufficient to ensure recovery, in that country imperfections in the functioning of the capital market seem to bear at least part of the responsibility for poor growth. The authors also argued here that labor market institutions in Latin America often receive the blame, whereas they are 56 Susan Horton, Ravi Kanbur, and Dipak Mazumdar only the symptoms of underlying political economy problems detrimental to growth. Turning now to the second broad topic, the outcomes of labor market adjustment, the authors had some difficulties in separating how far outcomes were due to structural adjustment, how far due to recession, and how far due to pre-existing trends. Severe adjustment, as in the case of Chile with high unemployment and sharp falls in real wages in an economy where urban employment predominates, can be very adverse to income distribution. Perhaps Brazil's worsening during the 1964-79 structural change period has some parallels, as does Korea's heavy industry phase. That is, unless countries make explicit provision for poorer groups, for example, the emphasis on food crop agriculture in Malaysia, structural change can worsen income distribution, although some of the changes, such as improved rural-urban relative income and possible improvement in informal/formal relative income, night militate in the opposite direction. Country-specific factors-success of indexation, wage and price freezes-also affect distnbution. The effects on distribution also depend on the level of the GDP. No data are available for Africa to test this hypothesis, but it seems plausible that improving rural-urban terms of trade and abolishing rents from prce distortions as part of adjustment programs could improve income distnbution nationally. The effects on women might be somewhat country specific, depending whether women were in tradables or notL but women are likely to face adverse effects of the employment shrinkages in some sectors due to their weaker attachment to the labor market The country studies generally confirmed this. Finally, the effects on long- term growth were adverse, but not directly attributable to labor market malfunctioning. Where should one go from here? One issue is that the apparently. benign conclusion that labor shifted into tradables masks that in response to structural adjustment, labor has moved in the direction opposite to that usually associated with economic development. Labor has shifted back into agrculture, out of manufacturing, and out of the public sector, although one might argue that this later sector was too large given the level of development reached. Recession plus adjustment has also resulted in an increase, in informalization, Labor Markets in an Era ofAdjustment: An Overview 57 increased use of casual labor, decreased worker benefits, and declines in skill and possibly education differentials. These trends are observed even in the most successful adjustment cases in Asia. Developing countries have long resisted being relegated to the role of primary producers in the internai3onal economic order, and it is unlikely that structural adjustment entailing further shifts of labor into agriculture would be highly sustainable. As regards possible further research, country study and some issues paper authors pointed the finger of blame for adjustment problems onto the capital market and possible. price rigidities in the output market. Another possibly fruitful topic is that of the role of labor market institutions, unions, and the political economy; something worth examining before launching into a wholesale advocacy of dismantling such institutions- Finally, as in all empirical research, better data are needed. One useful step would be to improve international collation of labor force statistics, clearly separating the results from household surveys from those of establishment surveys. Another would be to encourage further analysis of, and increased accessibility to, labor force surveys, which tend to be more expensive to analyze, but arguably yield more reliable results. References Dutt, A. K. 1984. "Stagnation, Income Distribution and Monopoly Power." Cambridge Journal of Economics 8(1): 25-40. Edwards, S. 1988. "Terms of Trade, Tariffs and the Labor Market Adjustment in Developing Countries." World Bank Economic Review 2(2): 165-185. Fallon, P. R., and L. A- Riveros. 1988. 'Macroeconomc Adjustment and Labor Market Response: A Review of the Recent Experience in LDCs." Washington, D.C.: World Bank. Draft, processed. Fields, G. 1990. "Labor Market Policy and Structural Adjustment in Cote d'Ivoire." Ithaca, New Yorlc Cornell University. Draft CGhai, D. 1987. Economic Growth, Structural Change and Labor Absorption in Africa: 1960-85. Discussion Paper No. 1. 58 Susan Horton RaviKanbur, andDipakMazwndar Geneva: United Nations Research Institute for Social Development. Horton, S., R. Kanbur, and D. Mazumdar. 1988. "Labor Markets in an Era of Adjustment: A Project Proposal." Washington, D.C.: World Bank, Economic Development Institute. Processed. ILO (International Labour Organisation). 1987. World Recession and Global Interdependence: Effects on Employmen4 Poverty and Policy Formation in Developing Countries. Geneva: 0LO World Employment Program. Various years. Yearbook of Labour Statistics. Geneva: ILO. . 1989. Yearbook of Labour Statistica Geneva: ILCO. Jamal, V., and J. Weeks. 1987. Rural-Urban Income Trends in Sub- Saharan Africa. World Employment Programme Labor Market Analysis and Employment Plamning Working Paper No. 18 (WEP 2-431WP.18). Geneva: ILO JS'JPA (Jobs and Skills Program for Africa). 1988- Africa Employment Report 1988. Addis Ababa: ILO. Johnson, 0. E. G. 1986. "Labor Markets, Extemal Developmegts, and Unemployment in Developing Countries." Washington, D.C.: IM Staff Studies for the World Economic Outlook Lavy, V., and J. Newman. 1989. "Wage Rigidity: Micro Evidence on Labor Market Adjustment in the Modem Sector." World Bank Economic Review 1(1): 97-111X. Lindauer, D. L., 0. A. Meesook, and P. Suebsae.ng. 1988. "Government Wage Policy in Africa: Some Findings and Policy Issues." World Bank Research Observer 3(1): 1-26. Ramos, J. R. 1980. "The Economics of Hyperstagflation: Stabilization Policy in Post 1973 Chile.' Journal of Development Economics 7(4): 467-88. Riveros, L 1989. "Recession, Adjustment and the Performance of Urban Labor Markets in Latin America." Washington, D.C.: World Bank. Processed. LaborMarkets in an Era ofAdjustment.-An Overview 59 Taylor, L. 1988. Varieties of Stabilization Experiences Towards Sensible Macroeconomics in the Third World. Oxford, UK.: Clarendon Press. Tokman, V. E. 1984. "The Employment Crisis in Latin America?" International Labor Review 123. van der Gaag and Vijverberg. 1989. "Wage Determinants in C6te d'Ivoire: Experience, Credentials, and Human Capital." Economic Development and Cultural Change 37: 371-381. Wong, P. K 1985. "Economic Development and Labor Market Changes in Peninsular Malaysia." Working Paper No. 12. Kuala Lumpur and Canberra. ASEAN-Australia Joint Research Project 2 ARGENTINA LuisA. Riveros Carlos E. Sdnchez Poor growth and macroeconomic imbalances have characterized Argentina's recent history. In combination with long-mun stagnation, the country has suffered chronic higb inflation and deep cyclical fluctuations compounded by intractable balance of payment crises. Attempts to stabilize the economy and achieve a structural adjustment to restore sustained growth and basic balances have failed due to both a fragile political climate and inconsistent policies. The lack of ad- justment has also been partly due to anticipated negative short-run la- bor market outcomes. In turn, the existence of persistent economic imvalances over the long run has negatively affected wages, employ- ment, and income distnbution. This chapter analyzes the performance of Argentina's labor mar- kets in recent years. It points out both how poor economic perfor- mance has affected labor market outcomes and how anticipated short- run costs have hindered reform efforts. Long-Term Economic Trends and Short-Term Adjustment Policies Global economic trends in Argentina have produced contradictory quantitative results and forced growing state intervention in the labor market. The authors gratefu2lly acknowledge comments on earlier drafts by M. Faig, R. Newfarmer, S. Horton, R. Paredes and seminar participants at the World Bank, the University of Chile, and the University of Warwick, as well as the efficient research assistance of 0. Giordano. 6C 62 LUis A. Riveros and Carlos E. Sdnchez Macroeconomic Policies, Economic Organization, and Growth An understanding of the trade and macroeconomic policies Argentina foUlowed after the Great Depression helps explain the country's poor economic performance. During 1860-1929, the gov- ernment pursued an export-led growth strategy, which included almost free trade and appropriation of the benefits of trade according to the country's comparative advantages. After the Great Depression, the government adopted an import substitution strategy. After 1945, gov- ernment policies aimed at expanding domestic markets through over- valued exchange rates and high import tariffs, which distorted re- source allocation and thwarted exports over the long-run. The eco- nomic results of the inward-ornented policies were poorer than those under the export-led strategy (table 2.1). Macroeconomic policies that affect variables such as the share of government consumption, public debt, and the money stock in total income, as well as commercial (export taxes and import controls) and exchange rate policies, were mostly responsible during the 1950s for the increase in the effective exchange rate for imports relative to that for exports. The resulting price increase of import goods made import substitution activities relatively more attractive for investment d6ci- sions, thereby prompting an inefficient specialization of production. Macroeconomic policies also affected the relative prices of productive Table 2.1 Strategies and Growth (real growth in per capita G1DP) Strategy Years Growth (percent) Export-led strategy 1900-29 1.5 Inward-oriented strategy 1929-58 0.9 1958-87 0-7 (Entire period) 1900-87 1h0 Source: IERAL data base. Argentina 63 factors, resulting in a distorted capital/labor mix in production that affected resource allocation at the sectoral and regional levels (see Cavallo -1986; Cavallo and Cottani 1986; Cavallo and Domenech 1988; Cavalo and Mundlak 1982; Nogues 1981; Sanchez 1987). The Argentinian economy can be divided into three sectors: a rural sector and two urban sectors (Llach and Sanchez 1984). The rural sector is a net exporter of wage-goods, mainly agriculturaL Therefore, real wages are inversely related to the incentives to produce exportable goods for a given exchange rate. The two urban sectors are net im- porters. They consist of an import substituting sector (which under the prevailing inward-oriented strategy is both a marginal exporter and a net importer of inputs and capital goods), and a sector producing nontradables. This economic structure led to a tradeoff between the trade balance situation and the prospects for domestic growth- Long-term economic growth requires a stable, high real exchange rate, but a high exchange rate implies higher food prices, and thus lower urban real wages. Since the level of -urban real wages have traditionally been a key political variable, the government has tended to hold down incentives to export. Thus, due to a deliberate policy of overvalued exchange rates (that is, a policy mix aimed at yielding high real wages in combination with low real exchange rates), relative prices attained two simultaneous roles: a mechanism for resource allocation and a distnrbutive device. This led to contradictory economic targets in a long-term context: mn- efficient organization of production or improved well-being of urban workers; a conflict generally resolved in favor of the latter - A deIiberate policy of overvalued exchange rates was the usual mechanism the government used to enlarge domestic markets, and therefore to increase real GDP and real wages in the short run. As soon as the growth in foreign terms of trade began to decline, a bal- ance of payments crisis arose, which made a devaluation and a decline in wages unavoidable. Thus, in the context of a deliberate policy of overvaluation, FTT can be seen as a determinant of the real exchange rate (figure 2.1).1 As a result of this policy, Argentina's economic 1. Cavallo and Domenech (1988) have modeled the behavior of the real exchange rate depending on the foreign terms of trade, taxes on imports and exports, the income leveL and macroeconomic policy. 64 Luis A. Riveros and Carlos E. Sdnchez Figure 2.1 Real Exchange Rate, Foreign Terms of Trade, 1962-87 (index, 1970 = 100) 160 140 820 . . \ ~~~~~~~~~~~~/ 100 -~ ~ ~ ~ ~~ V 40 20 _ m ,_ __ arrr-r 1963 1965 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 1987 Year -Rcal echangemate Foreign tmsoftade Source: Institute of Eccrnomic Studies of Argentina (IEERAL). history in recent decades can be described as a series of redistrbutive periods (domestic currency overvaluation as well as increased activity and real wages) followed by periods of stabilization (devaluation and reductions in the levels of activity and real wages). Poltical and Economic Developments of the 1970s Argentina experienced crucial political and economic changes during the second half of the 1970s. The years 1963-73 had seen un- usually high growth rates-of per capita GDP, which averaged 3-9 per- cent per annum. However, this growth was not a result of specific do- mestic policies, but of favorable foreign terms of trade, especially during 1964-66 and 1971-73. Nonetheless, this economic growth allowed for policies aimed at overvaluing the real exchange rate and at increasing both the domestic absorption of goods and real wages (table 2.2, figures. 22 and 2.3). The new (Peronist) administration that took over in 1973 inherited the combination of a satisfactory eco- Argentina 65 nomic performance with persistent imbalances associated with the economy's structural organization. Table 2.2 Economic Indicators, 1963-87 (index, 1970 = 100) Foreign terms Real Per capita Real Year of trade exchange rate GDP wages 1962 94 125 78 83 1963 101 120 75 81 1964 115 100 81 92 1965 114 95 87 98 1966 112 90 87 98 1967 107 112 88 96 1968 106 106 90 91 1969 101 103 96 96 1970 100 100 100 100 1971 114 94 103 104 1972 125 109 105 98 1973 142 S8 109 107 1974 114 82 114 120 1975 95 82 111 115 1976 90 117 109 74 1977 94 105 114 69 1978 92 80 107 72 1979 100 58 113 80 1980 117 46 113 93 1981 125 57 104 85 1982 105 100 97 74 1983 102 122 98 90 1984 111 107 99 99 1985 95 128 93 81 1986 81 102 96 76 1987 70 105 97 66 1988 82 107 93 63 1989 88 123 87 50 Source IEERAL data base. 66 LuisA. Riveros and Carlos E. Sdnchez Figure 2.2 Real Wages and Per Capita GDP, 196247 (index, 1970 = 100) 120 110 90 60 50. 1963 196S 1967 1969 1971 1973 197519'77 19U79 1981 1983 1985 1987 Year ---- - RcalwSgs R per otpia GDF Source IEERAL Figure 2.3 Foreign Terms of Trade and Real Wages, 1962-87 (index, 1970 = 100) 160 140- 120 60. 1963 196S 1967 1969 1971 1973 1975 1977 1979 1981 193 1985 1967 Rcal wages Foxeiptaus oftru Source: IEERAL. Argentina 67 The foreign terms of trade were still changing favorably in 1973- 74, and the Peronist administration continued to use redistributionist policies. Achievement of increasing real per capita GDP and real wages at the cost of deteriorating real exchange rates was still feasible. However, the increase in oil prices and the ensuing global recession revealed the fragility of this policy. Imported inflation and a sharp decline in foreign terms of trade in 1975 made it virually impossible to maintain an overvalued domestic currency without creating a sharp external deficit When the military overthrew the Peronist government in 1976, it inherited an acute balance of payments crisis and a huge fiscal deficit of more than 13 percent of GDP. The military government's objective was to reduce inflation drasti- cally and to initiate a longer-term strategy that would encourage sus- tained growth and full employment A central part of that strategy was a two-stage trade reform in 1976-81. During the first stage (1976- 78), the govermment introduced an unannounced tariff cut, eliminated export taxes, and replaced quotas with tariffs for all commodities ex- cept steel and aluminum. During the second stage (1979-81), the government implemented a preannounced schedule of quarterly tariff reductions. Even though the program was supposed to have continued until January 1984, trade reforms were reversed in 1981 due to a sharp balance of payments crisis. Completion of the trade liberalization was jeopardized not only by the excessive gradualism used- in the second stage, but also because a uniform tariff was not the final objective. Stikingly, the final result on overall nominal protection-including tariff and nontariff barriers- was contradictory: in 1979-80 the unweighted average nominal tariff rate decreased slightly from 51.9 to 49.2 percent, but its dispersion in- creased greatly (Cavallo and Cottani 1986). Another problem with the trade liberalization program was the use of an accompanying inconsistent mix of fiscalfmonetary and ex- change rate policies. In 1979 and 1980, the government's use of an active crawl reduction scheme based on a preannounced schedule of future devaluations resulted in a severe overvaluation. Although tax collection increased, the fiscal policy produced a large budget deficit because of the increase in both current spending and public invest- ment. Thus, the overall deficit rose from 13 percent of ODP in 1976 08 LuisA. Riveros and CarlosE. Sdchez to 16 percent in 1981 and 17 percent in 1932. Since a basic aim of the macro policy was to curb inflation, the government reduced the monetary financing of the deficit and began to rely heavily on do- mestic and external borrowing, thereby pushing up market interest rates and significantly increasing the public external debt. The persis- tence of a large fiscal imbalance in combination with a severe overval- uation produced disequilibrium in the balance of payments accom- panied by high real interest rates and low employment, production, and investment. The economic policy of the late 1970s did not improve the prevail- ing domestic imbalances. Although stabilization reduced annual infla- tion of the consumer price index from a peak of 441 percent in 1976 to 101 percent in 1980, inflation remained high: in 1981 it was still 104 percent per annum. Likewise, the active crawl reduction scheme implemented after 1978 and the external financing of the governmen- t's excess demand caused the real exchange rate to appreciate: taldng 1976 as a base year, its level was 39 in 1980 and 49 in 1981. The financial policies used after 1978 also resulted in volatile real interest rates. In January 1979, the government implemented its previ- ously announced nominal devaluations and eliminated most restric- tions on capital mobility. During the first eight months of 1979, when the policy still had some credibility, real interest rates were low, but reached negative values at times. Later on, uncertainty increased and risk premiums became high, wnich raised real interest rates from 2 to 6 percent per month. The increased uncertainty observed in 1980-81 was closely related to both the existing gap between inflation and the rate of devaluation (in 1980 inflation was 17 percent while the rate of devaluation was only 6 percent) and to observed changes in external accounts (table 23). The long-term structural adjustment program was barely imple- mented, and its final result was a shift in incentives in favor of non- tradable activities. The short-term stabilization program failed: in the presence of a persistent budget deficit, financing via capital markets (which replaced simple money creation) produced a crowding out ef- fect and was strongly deflationary (Mann and S6nchez 1984, 1985). Table 2.2 also shows the significant variability in real output between Argentina 69 Table 2.3 Selected External Accounts, 1980-84 (US$ millions) Account 1980 1981 1982 1983 1984 Exports 8,021 9,143 7,623 7,838 8,100 Imports 10,540 9,430 5,336 4,505 4,600 Trade balance -2,519 -287 2,286 3,331 3,500 Interest payments 956 2,925 4,400 4,983 5,273 Current account balance -4,769 -4,714 -2,357 -2,461 -2,492 Source: Cavallo (1986, table 1). 1975 and 1980, before the economy moved into another recession in 1981. The Crisis of the 1980s In 1981, new economic authorities had to address the external and internal imbalances that had resulted from overvaluation and the fis- cal/monetary mismagement. The authorities instituted a drastic pro- gram of exchange rate devaluations to deal with the most urgent pol- icy problem. From 1981 to 1983, the real exchange rate depreciated by 115 percent In 1981 and 1982, years of macroeconomic adjust- ment, the real exchange rate depreciated sharply and both the real per capita GDP and real wages experienced large reductions (table 2.2)_ The current account deficit in 1982 was substantially smaller than that ob:trved in 1980 and 1981 (table 2.3), thus many policymakers probably believed that further adjustment was not necessary, and opted for a new shift in policies during 1983 and 1984. During 1983, the last year of the military government, policies were aimed at recovering real wages, thereby reinstating the deliberate pol- 70 LuisA. Riveros and Carlos E. Sdndzez icy of overvalued exchange rates. As a result, real wages rose 22 per- cent in 1983, while the wage/exchange rate ratio increased over 40 percent. The overvaluation "'as accompanied by active fiscal policies and a resurgence of inflation. Long-term adjustment was abandoned and traditional populist policies returned to guide policymaking. A civilian administration (the Radical Party) took office in December 1983. This government inherited a very weak economic situation and vast public expectations of improved social welfare re- sulting from the restoration of democratic institutions. Activity levels and wages continued to grow in 1984 accompanied by a high fiscal deficit, growing inflation, low public utility rates, and exchange rate overvaluation (table 2.2 shows that the real exchange rate declined 13 percent between 1983 and 1984, while real wages increased 10 per- Cent). At the end of 1984, the government signed an agreement with the Inernational Monetary Fund (IMF) that initiated an external sector adjustment based on demand reduction. During the last quarter of 1984 and the first half of 1985, the economy suffered a drop in real wages and activity levels, a depreciation of the exchange rate, and ris- ing inflation. In June 1985, after sharp increases in public utility rates and a drastic devaluation, the government introduced the Austral Plan, which contained both heterodox and orthodox measures io curb in- flation. The former included a wage and price freeze and a deindexa- lion of debt. The latter inc-luded long-term measures, such as a high exchange rate and fiscal restraint. Inflation declined rapidly in 1985 due to price and wage controls. HoNvever, since the fiscal problem remained unsolved, monetary policy continued to play an active role. At first, the demand for money in- creased substantially, but then inflation returned because of percep- tions that the program was -unsustainable. In August 1986 and February 1987, the government made two other attempts to reduce infiation by means of a tight monetary policy and control on wages and prices. However, inflation remained high because its primary source-lack of fiscal discipline-was not eliminated. In addition, in August 1986 a period of overvaluation began: The real exchange rate averaged 113 (index, December 1976 = 100) during the first seven Argentina 71 months of 1986 and then dropped to 106 during August 1986 to August 1987. A Frustrated Process of Adjustment (1987-88) During September and October 1987 the government began to implement some new policies that were much more in line with a pro- gram of structural adjustment. However, no positive result has been yet observed and no structural adjustment has taken place. The relevant question is why a government politically committed to structural ad- justment ended up with quite different results. The period 1987-88 can be divided into two phases. During the first one, from September 1987 to July 1988, the government implemented a devaluation fol- lowed by a crawling peg adjustment of the exchange rate. During the second one, after July 1988, the feal exchange rate again appreciated significantly. During the first phase, the wholesale prices of nonagricultural (essentially tradable) goods experienced significant increases with re- spect to private services and construction (49 and 15 percent, respec- tively). Cereal and oilseed prices increased 84 percent with respect to private services with the help of the increase in intemational grain prices. These figures give some idea of the improvement in the relative. price of tradables versus nontradables. During this phase, the ex- change rate policy provided substantial incentives to the tradable sec- tor (table 2.4). As a result, export activities expanded and a realloca- tion of resources towards export-oriented activities began. The effects of this policy on the volume of exports and the trade surplus were significant: in 1988, exports increased 43.6 percent and the trade sur- plus increased 607.0 percent. In the specific case of manufacturing-a potentially exportable sector in Argentina-the change in relative prices (wages, exchange rate, and domestic terms of trade) led to a pattern of increasing profits and remarkable export growth. Inflation accelerated, mainly due to the failure to reduce the fiscaI deficit. After falling from 25 percent per month. in October 1987 to 3 percent per month in December, the inflation rate climbed to over 20 percent per month in July 1988. At the time it imposed price and ex- change controls, the government announced long-term measures aimed at shifting resources to tradable activities, improving the x-efEi- 72 LuttA. River,s and CarlosE. Sdnchez Table 2.4 Changes in Relative Prices, June 1985-October 1988 (percent) June 1985- Aug 1986- SepL 1987- July 1988- Ratios Aug. 2987 Aug. 1987 July 1988 Oct 1988 Versus private services: Exchange rate -40.1 -5.2 25.0 -12.7 Nonagricultural WPI -44.7 -8.9 49.0 -0.5 Cereals-oilseeds -37.6 i 1.0 84.0 -21.8 Versus construction cost: Nonagricultural WPI -7.2 1.0 15.1 -4.2 Note: WPI = wholesale price index. Source: Institute of Economic Studies on Argentina (IEERAL). ciency of the public sector, and freeing up rigidities in factor and oAt- put markets. During the second phase-after July 1988-real wages in manufacturing increased and profits began to decline (table 25 and figure 2.4). Tabh!? 2.4 also indicates that nonagricultural wholesale prices (that is, mainly manufacturing prices) deteriorated compared to private services (-0.5 percent), and constuction (-4.2 percent). In other words, the domestic terms of trade turned in favor of nontradables. Since the increase in wages and in the price of nontradables was accompanied by overvaluation and elimination of import restrictions, the manufacturing sector faced faling profit margins and lower domestic market shares. The poor timing and lack of coordination between short-team and long-term economic policies hindered the achievement of structural adjustment. Argentina 73 Table 2.5 Manufacturing Wage, Exchange Rate, and Product Wage, 1987 and 1988 (index; 1988 3rd quarter = 100) _ Wa gel Product- Wagel Product- exchange wagel exchange wagel Year Ouarter ratea " insb Year Ouarner rarea rainsb 1987 1 S * 1989 1 116 115 2 0 2 43 78 3 121 125 3 47 56 4 111 116 4 74 86 1988 1 114 125 1990 1 70 -81 2 109 109 2 111 101 3 100 100 3 177 111 4 102 116 4 207 133 * not calculated. for this study a. Manufacturing nominal hourly wage divided into the exchange rate at which imports are traded. b. Ratio of wage cost (ratio between the nominal hourly wage paid in manufacturing and the corresponding wholesale product price) to productivity (output per manhour). Source: IEERAL data base. Figure 2.4 Real Wage and Product Wage in Manufacturing (index, 1988 3rd quarter = 100) 250 * 200 150 - 100 Ns X - 50 - 0 1 2' 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1987 l9S8 1989 1990 Ycurs Wage/changc mae --- -Prod wagewPtucdan gns Source: IEERAL 74 Lui A. Riveros and Carios E. Sdnclhez The Structure and Trends of Labor Markets The performance of labor markets has reflected overall economic tiends, which have produced low wage growth over the long run, as well as reduced employment growth in the private sector. Observed Trends in Real Wages Observed labor market results are paramount in analyzing Argentina's import substitution policy. The driving force behind labor markets was a wage setting mechanism based on a deliberate policy of overvalued exchange rates, restricted by the trade balance/domestic growth tradeoff. The basic policy tool was government -intervention supported by urban-based unions and political groups. The main observed outcomes were a slight growth in real wages over the long run, significant short-run economic fluctuations, and distorted relative wages among productive sectors- Between 1940 and 1985, real wages neither rose nor fell for more then three consecutive years. In all but one case (1969-71), periods of growth in real wages were followed by periods of sustained decline. By 1985, real wages were only 61 percent higher than in 1940, implying an average yearly growth rate slightly higher than 1 percent (Riveros 1989; Sanchez 1987). If the shorter period 1962-87 is considered, the evidence more than confirms the wage deterioration over time (see table 2.2 and figure 2.2). After growth during 1962- 74, real wages declined much more than real per capita GDP. In addition, the magnitude of short-term. fluctuations increased. For instance, in 1962-74 real wages rose at a yearly rate of 3.1 percent only to drop subsequently at an even higher rate (-4.4 percent). During the whole period 1962-87, real wages fell by an average of 0.8 percent per year. The observed trend in relative wages between tradable and nontradable sectors is an outcome of the inward-oriented growth strategy. Domestic market-oriented growth required relative prices favorable to urban activities and high purchasing power for wage earners. Most of Argentina's population is concentrated in a few urban centers, while services, construction, and import substituting industries produce and sell most of their output in these markets. Argentina 75 Thus, the evolution of relative wages from the 1940s to the 1980s has clearly favored labor in nontradable activities (Riveros 1989; Sanchez 1987). Moreover, observed wage changes did not reflect changes in labor productivity in nontradable activities (Sdnchez 1987), and did so only mildly in manufacturing. During 1962-87, however, relative wages in tradables and nontradables remained relatively stable (see table 2.6 and figure 2.5). However, this stability probably reflects government intervention more than relatively stable relative labor productivity. To avoid problems of interpretation associated with the peculiar behavior of public sector wages after 1985, namely, sharp wage cuts due to stabilization policies, they are not included in the group of nontradables. Table 2.6 The Evolution of Relative Wages, 1962-86 Wages in Wages in Wv'ages in potenialy tradable Wages in potentaly tradabk tradable industes tradable industieui indnriesAvagesin wages in undunriesivagesin wages in nonradable nontradable -intbi nontradabek Year industries industries Year industries indurui&s 1962 0.83 0.99 1976 0.93 1.04 1963 0.79 0.95 1977 0.94 1.04 1964 0.82 0.96 1978 0.91 1.01 1965 0.(8 1.01 1979 0.93 1.04 1966 0.90 1.03 1980 0.89 0.99 1967 0.89 1.05 1981 0.91 1.03 1968 0.88 1.03 1982 0.91 1.03 1969 0.88 1.04 1983 095 1.05 1970 0.89 1.04 1984 0.96 1.09 1971 0.91 1.04 1985 0.96 1.08 1972 0.91 1.03 1986 0.95 1.10 1973 0.88 1.00 1987 1.02 1.16 1974 0.93 1.05 1988 1.05 1.21 1975 0.92 1.02 1989 0.94 1O10 D __________________________ I990 0.95 1.06 Source. ISERAL data base. 76 Luis A. Riveros and Carlos E. Sdnchez Figure 2.5 The Evolution of Relative Wages, 1962-89 1.3. 1.1 . ,-