21904 i February 2001 EAST ASIAN LABOR MARKETS AND THE ECONOMIC CRISIS I MPACTS RESPONSES & LESSONS Edited by Gordon Betcherman and Rizwanul Islam EAST ASIAN LABOR MARKETS AND THE ECONOMIC CRISIS IMPACTS RESPONSES & LESSONS Edited by: Gordon Betcherman Rizwanul Islam rAwl, 0j4 The World Bank, Washington, D.C. The International Labour Office, Geneva Copyright © 2001 The International Bank for Reconstruction and Development / THE WORLD BANK 1818 H Street, NW. Washington, D.C. 20433, USA and the International Labour Organization All rights reserved Manufactured in the United States of America First printing February 2001 1 234030201 The findings, interpretations, and conclusions expressed in this book are entirely those of the authors and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. Publication of this work does not constitute an endorsement by the International Labour Office of the opinions expressed therein. The World Bank does not guarantee the accuracy of the data included in this publication and accepts no responsibility for any consequence of their use. The boundaries, colors, denominations, and other information shown on any map in this volume do not imply on the part of the World Bank Group any judgment on the legal status of any territory or the endorsement or acceptance of such boundaries. The material in this publication is copyrighted. The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. Permission to photocopy items for internal or personal use, for the internal or personal use of specific clients, or for educational classroom use is granted by the World Bank, provided that the appropriate fee is paid directly to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; telephone 978-750-8400, fax 978-750-4470. Please contact the Copyright Clearance Center before photocopying items. For permission to reprint individual articles or chapters, please fax a request with complete information to the Republication Department, Copyright Clearance Center, fax 978-750-4470. All other queries on rights and licenses should be addressed to the Office of the Publisher, World Bank, at the address above or faxed to 202-522-2422. Cover design by Design Studio Grafik, Herndon, VA. Library of Congress Cataloging-in-Publication Data East Asian labor markets and the economic crisis: impacts, responses, and lessons / edited by Gordon Betcherman, Rizwanul Islam. p. cm. Includes index. ISBN 0-8213-4478-1 ISBN 92-2 112274-3 (ILO) 1. Labor market-East Asia. 2. Labor market-Asia, Southeastern. 3. East Asia- Economic conditions. 4. Asia, Southeastern-Economic conditions. I. Betcherman, Gordon. II. Islam, Rizwanul. HD5826.A6 E37 2000 331.12'095-dc2l 00-066782 Contents Foreword xi Acknowledgments xiii SECTION I: LABOR MARKET EXPERIENCES IN THE CRISIS COUNTRIES 1 East Asian Labor Markets and the Economic Crisis: An Overview 3 Gordon Betcherman and Rizwanul Islam Chairman's Summary 38 Kenji Tunekawa 2 The Economic Crisis: Labor Market Challenges and Policies in Indonesia 43 Rizwanul Islam, Gopal Bhattacharya, Shafiq Dhanani, Max Iacono, Farhad Mehran, Swapna Mukhopadhyay, and Phan Thuy 3 Korea: Labor Market Outcomes and Policy Responses after the Crisis 97 Soon-Hie Kang, Jaeho Keum, Dong-Heon Kim, and Donggyun Shin III IV CONTENTS 4 Malaysia: Protecting Workers and Fostering Growth 141 Norma Mansor, Tan Eu Chye, Ali Boehanoeddin, Fatimah Said, and Saad Mohd Said 5 The Philippines: Labor Market Trends and Government Interventions Following the East Asian Financial Crisis 195 Jude Esguerra, Arsenio Balisacan, and Nieves Confessor 6 The Labor Market and Labor Policy in a Macroeconomic Context: Growth, Crisis, and Competitiveness in Thailand 245 Moazam Mahmood and Gosah Aryah SECTION II: LABOR MARKET ISSUES AND POLICIES IN THE REGION 293 7 Active Labor Market Policies: Issues for East Asia 295 Gordon Betcherman, Amit Dar, Amy Luinstra, and Makoto Ogawa 8 The Economics of Employment Protection and Unemployment Insurance Schemes: Policy Options for Indonesia, Malaysia, the Philippines, and Thailand 345 Alejandra Cox Edwards and Chris Manning 9 Vulnerable Groups and the Labor Market: The Aftermath of the Asian Financial Crisis 379 Susan Horton and Dipak Mazumdar 10 Social Dialogue and Labor Market Adjustment in East Asia after the Crisis 423 Duncan Campbell Japan's Experience with Employment Policy 467 Naoyuki Kameyama FIGURES 1.1 Quarterly GDP Trends, Crisis Countries 7 3.1 Unemployment Rates 100 CONTENTS V 3.2 Labor Force Participation Rate 101 3.3 Unemployment Rate by Educational Attainment 102 3.4 Long-Term Unemployment Rate 103 4.1 Manpower Planning Machinery 178 5.1 Visible and Invisible Underemployment Rates, 1996-99 203 5.2 Employment Creation in 1998 204 5.3 Unemployment and Underemployment Rates, January 1996-April 1999 205 5.4 Employment Creation (January 1997-April 1999) 207 5.5 Output Growth in the Major Sectors 207 5.6 Technical and Vocational Education in the Philippines 226 7.1 OECD Average Active Labor Market Program Expenditures 308 7.2 OECD Spending on Active Labor Market Policies 309 11.1 Changes in Employment Policy 487 TABLES 1.1 Summary of Economic Trends, Five Crisis Countries, 1980-96 5 1.2 Summary of Labor Market Institutions, late 1990s, Crisis Countries 10 1.3 Summary of Unemployment, Underemployment, and Employment Trends, Crisis Countries, 1996-98 13 1.4 Employment Change by Industrial Sector, Crisis Countries, 1997-98 14 1.5 Real Wage Trends, Crisis Countries, 1996-98 14 1.6 Labor Force Participation Rates, Crisis Countries, 1996-98 15 1.7 Employment Status Trends, Crisis Countries, 1997-99 16 1.8 Summary of Labor Adjustment in the Crisis Countries 18 1.9 Job Creation through Public Works Programs 22 1.10 Job Creation through Support of SMEs and Self-employment 25 1.11 Interventions in Skill Training 27 1.12 Response through Employment Services 28 2.1 Aggregate Changes in the Labor Force 47 2.2 Characteristics of the Unemployed, 1994-98 49 2.3 Trends in Underemployment, 1994-98 51 2.4 Underemployment in Selected Sectors, 1994-98 52 VI CONTENTS 2.5 Job Losses and Absorption in Agriculture, 1997-98 53 2.6 Structural Changes in Employment, 1994-98 54 2.7 Changes in Employment Status, 1994-98 56 2.8 Trends in Monthly Nominal and Real Earnings, 1994-99 60 2.9 Microcredit Providers 72 2.10 Vocational and Technical Secondary School Enrollments (1996-97) 82 2.11 Trainees in Skill Training Programs (1994-95 to 1998-99) 84 2.12 Productivity Levels of Asian-Pacific Countries 86 2.13(a) Expatriate Workers by Main Industry, 1995-97 87 2.13(b)Expatriate Workers by Main Occupation, 1995-97 87 2.14 Formal Education of Public Vocational and Skill Training Instructors, 1996 88 3.1 Public Expenditures on Unemployment Measures, 1998 108 3.2 Workers Receiving Unemployment Benefits, 1996-99 112 3.3 Unemployment Benefit Coverage 113 3.4 Participation in Employment Subsidy Programs, 1998 114 3.5 Public Works Project Participants, 1998-99 117 3.6 Reasons for Job Separation 122 3.7 Employment Service Agencies 123 3.8 Outcomes for Public Employment Service Agencies 124 3.9 Evaluation of Public Employment Services 125 3.10 Public Employment Service Agencies and Counselors, 1997 126 3.11 Vocational Training Institutes 127 3.12 Vocational Training by Program in 1997 129 3.13 Legislation on Vocational Training in Korea 131 3.14 Training for the Unemployed, 1998 132 4.1 Selected Macroeconomic Indicators 142 4.2 Industry Shares in GDP, 1986-98 143 4.3 Registered Job Seekers by Sex 146 4.4 Labor Force Participation Rates by Gender 147 4.5 Trade Unions and Membership, by Sector, 1992-96 150 4.6 Trade Unions by Membership, 1992-96 151 4.7 Collective Agreements, 1992-96 152 4.8 Net Job Creation by Sector, 1996-00 154 4.9 Funds Allocated to Distributive Trade and Supporting Services, 1996-2000 158 CONTENTS VII 4.10 Employer Actions during Recession, Peninsular Malaysia, 1998-99 160 4.11 Malaysia: Registration and Placement by Private Employment Agencies, 1997-99- 164 4.12 Registration and Placement of Local Retrenched Workers, 1998-99 166 4.13 Malaysia: New Registrants, Vacancies, and Placement in the Labor Market, 1992-99 169 4.14 Malaysia: Vacancies by Sector, 1996-99 170 4.15 Malaysia: Active Registrants, by Employment Status, 1997-99 171 4.16 Statutory Benefits Paid to Retrenched Workers, 1998 173 4.17 Malaysia: Trade Disputes, 1994-98 174 4.18 Intake of Public Training Institutions 1995-2000, by course 179 4.19 Output of Public and Private Training Institutions, 1995-2000, by course 181 4.20 Major Public Vocational Institutions 182 4.21 Expected Demand for Skilled and Semi-skilled Workers, 1990-98 184 5.1 Educational Attainment and Unemployment 196 5.2 Employment in Agriculture and Services, 1990-98 198 5.3 Unemployment and Underemployment, April 1999 206 5.4 Impact of the Crisis on Households 208 5.5 Households' Response to the Crisis 209 5.6 Government Livelihood Programs 217 5.7 Public and Private Training and Vocational Education Institutions (December 1997) 229 5.8 How Public and Private Institutions Use Their Facilities 232 6.1 Growth Rate of Real GDP 246 6.2 Macroeconomic and Social Indicators in Thailand 248 6.3 Structure of Output 250 6.4 Structure of the Labor Force, 1998 251 6.5 Sectoral Employment across Urban and Rural Areas, 1998 252 6.6 Employment by Status, February 1996-99 253 6.7 Manufacturing Wage Rates 254 6.8 Structure of Demand 255 6.9 Integration with the Global Economy 256 6.10 Structure of Manufactured Exports 257 VIII CONTENTS 6.11 Recent Thai Manufactured Export Performance 259 6.12 Unit Labor Costs in Manufacturing, Selected Asian Economies, 1990 260 6.13 Labor Productivity 262 6.14 Educational Enrollment, 1999 (or most recent years) 263 6.15 Savings, Investment, and Growth 264 6.16 Total Factor Productivity Growth Rates 265 6.17 Labor Force Status, 1994-99 (February) 268 6.18 Overseas Job Placement 1998 283 6.19 Educational Levels of Thai Labor Force 288 6.20 Estimated Labor Force Requirements According to Educational Level 289 6.21 Shortage of Science and Technology Manpower, by Level of Education 290 7.1 Tailoring Programs to Objectives 298 7.2 Active and Passive Labor Market Programs: Some Key Features 299 7.3 Expenditures on Labor Market Programs, Selected OECD Countries 310 7.4 Evaluation Results for Active Labor Market Programs 314 7.5 Losses from Wage Subsidy Programs 319 7.6 Annual Cost of Job Creation in Public Works 321 7.7 Failure Rates for Businesses Receiving Microenterprise Development Assistance 322 7.8 Employment Services 324 7.9 Labor Market Training 329 7.10 Job Creation Activities 333 8.1 Basic Indicators, 1998 350 8.2 Distribution of Employment by Major Sector, 1985-97 351 8.3 Basic Labor Market Indicators 352 8.4 Private Transfers among Households 353 9.1 Per Capita GDP Growth in Five Asian Countries 380 9.2 Selected Labor Market Variables for Five Asian Countries 381 9.3 Participation Rates before and after the Crisis 382 9.4 Estimated Magnitudes of Labor Market Response to the Crisis 385 9.5 Unemployment Rates before and after the Crisis 387 9.6 Sectoral Employment before and after the Crisis 391 9.7 Per Capita Income Measures, Thailand 396 CONTENTS IX 9.8 Labor Market Institutions Prior to Crisis 397 9.9 Changes in Labor Market Institutions Following the Crisis 401 10.1 Estimated Union Densities 430 10.2 Trends in Trade Union Membership in Thailand 430 11.1 Employment by Sector 473 11.2 Why Unemployed Workers Had Left Their Jobs 485 BoxEs 5.1 Loss of Competitiveness and Declining Labor Costs 199 5.2 TESDA's Responsibilities 227 8.1 Experience Rating as a Means of Financing Unemployment Insurance 348 8.2 Provident Funds and Worker Security 355 10.1 Social Dialogue and Macroeconomic Performance 428 10.2 One Enterprise Responds to the Crisis 437 10.3 Alternatives to Retrenchment 438 10.4 Labor Market Adjustment in Thailand 439 10.5 Tripartism in Labor Affairs: Problems from a Labor Perspective 443 10.6 Social Dialogue, the Enterprise, and Adjusting to the Economic Downturn: Some Examples 447 10.7 The Philippine Airlines Dispute 452 10.8 Highlights of the 90-point Agreement of the First Tripartite Commission 457 10.9 Highlights of the Agreement in the Hyundai Motor Company Dispute 460 10.10 Trade-offs Resulting from Social Dialogue at the National Level 461 11.1 Assisting Displaced Coal Mine Workers 477 11.2 Employment Measures for Older Workers 480 Foreword AFTER YEARS of remarkable progress in the region, the financial crisis in East and Southeast Asia had sudden and often painful consequences for millions of workers. With disappearing employment and earning opportunities in the formal sector, affected employees and their families had little option but to survive on reduced incomes, informal ac- tivities, meager social assistance, and support from others. As the effects of the crisis worsened in late 1997 and 1998, the World Bank and the Inter- national Labour Office (ILO) began to explore ways in which they could collaborate with the region's governments, employers, and workers in their response to the difficult social and economic situation. The two organizations agreed that one important contribution they could jointly make would be to create an opportunity for the affected countries to share their experiences regarding the crisis, its labor market impacts, and the policy lessons that have emerged. With substantial fi- nancial support from the Asia-European Meeting (ASEM) fund, the World Bank and the ILO sponsored a regional labor market project cov- ering the Republic of Korea, Indonesia, Malaysia, the Philippines, and Thailand. This included the preparation of five country papers analyzing the labor market aspects of the crisis. The difficult events of the late 1 990s also raised important questions for these countries regarding needed labor policy reforms for the longer run. Accordingly, the World Bank and the ILO also sponsored a series of papers by international experts on policy options in key areas including unemployment benefits, active labor mar- ket programs, support for vulnerable groups, and social dialogue. XI XII FOREWORD In October 1999, these papers provided the basis for a workshop in Tokyo organized and hosted by the Japanese Ministry of Labor and the Japan Institute of Labor. This workshop involved tripartite delegations from the five Asian countries as well as Japan, academic specialists, and of- ficials from international organizations including the Organisation for Economic Co-operation and Development and the European Union, as well as the World Bank and the ILO. This book includes the country reports and the international policy pa- pers prepared for the Tokyo workshop, revised to reflect the discussions at that seminar. It describes how the labor markets in the region were af- fected by the financial crisis and how governments and communities re- sponded. It also looks forward in setting out the labor policy options for the future, based on international experience. We hope that these chapters will be of interest for the countries in East and Southeast Asia in their quest to understand the crisis and to move forward from it. We also hope the book will be a valuable resource more broadly in a world where eco- nomic crises seem to occur too frequently and where developing countries are grappling with a range of labor market challenges. We are pleased that the World Bank and the ILO, with ongoing sup- port from the Japanese government, will continue to collaborate with gov- ernments in the region in exploring labor policy options. A follow-up proj- ect now under way focuses on the application of active and passive labor programs. With the cooperation of the Philippine government, we will discuss these issues at a regional seminar in Manila in early 2001. Robert Holzmann Goran Hultin DIRECTOR EXECUTIVE DIRECTOR SOCIAL PROTECTION UNIT EMPLOYMENT SECTOR WORLD BANK INTERNATIONAL LABOUR OFFICE Acknowledgments THIS PROJECT could not have been successfully under- taken without a number of partner organizations and the contributions of many individuals. The Asia-European Meeting (ASEM) fund provided substantial financial support for the research, the Tokyo workshop, and the publication of this volume. We also want to acknowledge the major contribution of the Japanese Ministry of Labor UMOL) and the Japan In- stitute of Labor UIL) in organizing and co-hosting the Tokyo workshop. We would especially like to thank Kimie Iwata UMOL) and Kunihiko Saito (JIL) in this regard. Robert Holzmann of the World Bank's Social Protection Unit was instrumental in the initiation of the project and con- tinued support throughout. From the ILO, Katherine Hagen provided encouragement in launching the project; Samir Radwan and Mitsuko Ho- riuchi provided active and ongoing support throughout the project; and Stanley Taylor coordinated ILO involvement. In addition to their sub- stantive contributions, Amit Dar, Makoto Ogawa, and Amy Luinstra of the World Bank flawlessly assumed responsibility for virtually all aspects of the project's coordination. Mr. Ogawa deserves special mention for ini- tiating the involvement of the government of Japan and coordinating the workshop arrangements with the Japanese Ministry of Labor and the Japan Institute of Labor. Melvina Clarke provided excellent administrative sup- port throughout the project. Emily Chalmers expertly edited the volume in an efficient manner. Studio Grafik designed the volume cover. Finally, we are grateful to Nicola Marrian, Sheila Queano-Colletta, and Carlos Rossel of the World Bank's Office of the Publisher for their wise advice and for guiding us through the entire publication process. XIII SECTION Labor Market I Experiences in the Crisis Countries 1 East Asian Labor Markets and the Economic Crisis: An Overview GORDON BETCHERMAN AND RIZWANUL ISLAM THE EAST ASIAN crisis has had serious consequences for workers throughout the region. Unemployment has risen, earnings have fallen, and workers' rights have been endangered. The resulting poverty, declines in living standards, and social tensions have threatened the gains made during the unparalleled growth of the 1980s and most of the 1990s. The negative effects of the crisis have also raised difficult questions about development in East and Southeast Asia and, in particular, about the so- cial protections it has afforded. More than three years after the onset of the crisis, evidence suggests that macroeconomic stabilization and recov- ery are beginning to take hold. But for millions of workers, social and eco- nomic hardship continues. Important challenges remain for governments as they work to alleviate the hardship the crisis has caused and to build a sustainable path toward future prosperity. In 1998 the World Bank and the International Labour Organization (ILO) initiated a project on labor markets and the East Asian crisis. This project was one element in a larger collaboration between these two insti- tutions aimed at providing support as the region responds to the social di- mensions of the crisis. The labor market project covers the five countries that received the brunt of the impact-Indonesia, the Republic of Korea, Malaysia, the Philippines, and Thailand-and is designed to stimulate analysis and policy dialogue in the context of international experience. The project is concerned with both the immediate crisis and long-term policy issues. The crisis has generated a pressing need for an accurate, em- 3 4 EAST ASIAN LABOR MARKETS pirically based picture of the impact on employment. How are the labor markets adjusting to the economic contraction? What groups are espe- cially vulnerable to unemployment, underemployment, and wage cuts? And what actions have governments taken to mitigate the hardships work- ers face and to help those affected cope with the downturn? Understand- ing the immediate labor market and labor policy as it relates to the crisis is essential for the countries discussed here. But such a study also has wider relevance. Serious economic shocks are occurring with increasing fre- quency. Policymakers everywhere are eager to learn from the Asian expe- rience in order to improve their understanding of the ways shocks are transmitted to the labor market and to identify measures governments can take to minimize the impacts. The Asian countries need to begin building on their experiences with the crisis in order to develop long-term labor market policies. With the ex- ception of Korea, the region has seen little policy development in this area. The high rates of economic growth of the 1980s and early 1990s meant that unemployment was not a serious problem, and workers' incomes were rising. Even before the crisis, however, questions were being raised about the adequacy of existing arrangements in areas such as support for unemployed workers, vocational education and training, employment ser- vices, and industrial relations. The crisis has underscored the need for careful analysis and consideration of labor policy options. These issues-some immediate, others long-term-have been at the center of the Asian labor market project. To inform the dialogue, analysts prepared papers on recent labor market developments in each of the five countries. A second series of papers placed the developments in the con- text of international experience and identified best practices in labor pol- icy. Together these papers formed the basis for a three-day workshop sponsored by the Japanese Ministry of Labor and the Japan Institute of Labor. The workshop, held in Tokyo in October 1999, included tripartite delegations from Indonesia, Korea, Malaysia, the Philippines, and Thai- land, as well as specialists from international organizations and the re- search community.1 This volume is based on the papers and discussions that took place in Tokyo. Part I includes chapters on each of the five countries under dis- cussion. Each chapter presents an empirical overview of recent labor mar- ket trends in the country and then reviews current policies in employment creation and maintenance, income support for unemployed workers, em- ployment services, and vocational education and training. The chapters in part II place the experiences of these countries in an international context EAST ASIAN LABOR MARKETS AND THE ECONOMIC CRISIS: AN OVERVIEW 5 by drawing on the experiences of other countries. The chapters cover ac- tive labor market programs, unemployment benefits, vulnerable popula- tions, the relationship between social dialogue and labor market adjust- ment, and the Japanese experience with employment policy. The Impact of the Crisis on Labor Markets In the 15 years before the crisis, Southeast Asia enjoyed dramatic growth. Basic indicators for the five countries under discussion show the regions remarkable growth path (table 1.1). Except for the Philippines, annual GDP (gross domestic product) growth in 1980-95 was more than 6 per- cent. A key element in this success was the region's openness to interna- tional trade and investment. Export growth reached as high as 16 and 22 percent (for Korea and Thailand, respectively) in 1980-96. Net capital in- flows were more than 9 percent in Malaysia, the Philippines, and Thailand in the mid-i 990s. By 1997 the region, one of the world's poorest a gen- eration earlier, had risen to middle-income level. Korea was moving to- ward the level of development seen in countries of the Organization for Economic Cooperation and Development (OECD). This rapid development led to significant changes in the region's labor markets. For the most part, these changes brought improved outcomes for workers. In the 1990s real wages increased significantly in all of the coun- tries except the Philippines.2 At the same time strong job growth kept unemployment at very low levels (typically well below 5 percent) in all countries but the Philippines. The structure of employment changed as well, with major shifts out of agriculture and into industry and service employment in urban areas. By the onset of the crisis in 1997, at least half of all jobs in each of the five countries were in nonagricultural sectors. Important developments also occurred in the region's labor force: educa- tional levels rose, and female participation increased in every country but Thailand. The financial crisis began in July 1997 in Thailand and quickly spread throughout the region. The timing and magnitude of the economic con- traction varied across countries (figure 1.1). GDP began declining in Thailand in the third quarter of 1997. In the other countries economic growth slowed but remained positive until the beginning of 1998. GDP decreased in 1998 by 0.4 percent in the Philippines, 5.8 percent in Korea, 7.5 percent in Malaysia, 10.0 percent in Thailand, and 13.7 percent in In- donesia. The decline in each country bottomed out sometime between the second and fourth quarters of 1998, and all five countries were growing TABLE 1.1 Summary of Economic Trends, Five Crisis Countries, 1980-96 GDP growth GDPper capita Industrial structure (Average annual % change) ($) % distribution, 1996 (% change since 1980) 1980-95 1996 1997 Agriculture Industry Services Indonesia 6.8 8.0 1,110 44.0 (-21.0) 18.1 (37.1) 37.9 (25.5) Korea, Rep. of 8.7 7.1 10,550 11.6 (-65.9) 32.5 (29.0) 55.9 (37.0) Malaysia 7.9 8.6 4,530 19.4 (-47.8) 32.2 (33.6) 48.4 (25.1) Philippines 2.1 5.7 1,200 41.7 (-19.5) 16.6 (7.8) 41.6 (26.8) Thailand 6.7 6.4 2,740 50.0 (-29.4) 20.8 (101.9) 29.1 (54.0) Sources: World Bank, World Development Indicators; International Monetary Fund, World Economic Outlook; International Labour Organization, Key Indicators of the Labor Market. EAST ASIAN LABOR MARKETS AND THE ECONOMIC CRISIS: AN OVERVIEW 7 FIGURE 1.1 Quarterly GDP Trends, Crisis Countries 15.0- 10.0 5.0- -5.0- Indonesia -15.0- A Malaysiab-/ -200 0 - Philippines - + Thailand -25.0- 1997 Ql 1997 Q2 1997 Q3 1997 Q4 1998 Ql 1998 Q2 1998 Q3 1998 Q4 1999 Ql 1999 Q2 1999 Q3 Source: World Bank 2000b. again by the second quarter of 1999 (figure 1.1). However, the strength of the recovery has varied widely throughout the region. For example, real GDP growth for 1999 has been estimated at 8.0 percent in Korea but at 0.0 percent in Indonesia (World Bank 2000b). How did the labor markets in the region adjust to the economic crisis? Since the countries affected were at different stages of development and had different labor market policy regimes, their responses also differed. The variations in their adjustment efforts provide an opportunity to assess the role of some potentially key variables. Such an assessment can offer impor- tant evidence on the impacts of major shocks, their transmission to different markets, and the distribution of their social and economic consequences.3 How Labor Markets Adjust Reduced aggregate demand can be transmitted to the labor market in a number of different ways. These include supply as well as demand re- sponses.4 On the demand side the key question is whether a decrease in the 8 EAST ASIAN LABOR MARKETS demand for labor results in lower employment levels or lower wages. Un- der perfectly competitive market conditions, adjustment occurs in prices (wages) but not in quantity (employment). But a range of behavioral and institutional factors can generate both types of responses. In developing countries that are undergoing structural adjustment, economic contrac- tion, or both, adjustment occurs primarily through wage declines, which are often greater than declines in GDP (Horton, Kanbur, and Mazumdar 1994; Fallon and Lucas 2000). However the studies also show evidence of adjustment through decreasing employment and rising unemployment.5 Important adjustments also occur on the supply side. Labor supply de- cisions made when demand for labor is weak can lead to either higher or lower participation rates, depending on the relative importance of the dis- couraged worker and added worker effects. The case studies reported in Horton, Kanbur, and Mazumdar (1994) show that both effects can be im- portant. Similarly, workers faced with a weak labor market may decide to relocate. In Southeast Asia both internal and international migration are important supply-side responses. Workers move between rural and urban areas in response to changes in economic opportunities. Workers also mi- grate across borders, moving in and out of countries, although such move- ments are often constrained by policy. A final aspect of adjustment involves the interaction among segmented labor markets. When labor demand decreases in the formal market, dis- placed workers have the option of moving to the informal sector, which often expands during economic contractions (Horton, Kanbur, and Maz- umdar 1994). This sector includes both rural agricultural employment and informal urban activities. Involuntary open unemployment is essen- tially nonexistent in informal sectors. But when the labor supply expands because formal-sector workers are displaced, we can expect to observe downward pressure on earnings. Labor market adjustment to shocks such as the Asian financial crisis, then, involves a number of dimensions: open unemployment, underem- ployment, declining wages and earnings, changes in labor force participa- tion and migration, and informalization. The importance of each of these aspects depends on various factors, including the magnitude of the eco- nomic shock, the structure of the economy, the composition of the labor force, and labor market policies and institutions. THE MAGNITUDE OF THE SHOCK. The severity of a macroeconomic cri- sis affects both the scale of labor adjustment and the forms adjustment takes. As we have noted the magnitude of the shock differed across coun- EAST ASLAN LABOR MARKETS AND THE ECONOMIC CRISIS: AN OVERVIEW 9 tries in Southeast Asia. Indonesia experienced the largest contraction and the Philippines the smallest, with the other three countries falling some- where between these extremes. THE ECONOMIC STRUCTURE. Reduced demand is most likely to be transmitted to the labor market through quantity rather than price ad- justments in countries with a relatively high proportion of employment in nonagricultural industries and the formal urban sector. Over the past two decades, major structural shifts away from agriculture and into industry and services occurred throughout the region. But in the late 1990s the five crisis countries were at different stages of development and had different economic structures. According to indicators such as industrial compo- sition and urbanization, Korea was the most developed, followed by Malaysia. Indonesia, the Philippines, and Thailand still had large agricul- tural sectors, and Indonesia and Thailand had low rates of urbanization.6 THE LABOR FORCE. The composition of the labor force can also affect responses to reduced labor demand. Characteristics such as gender, age, and level of education are likely to affect a number of variables, including workers' reservation wage and willingness to move to the informal sector. We have already noted the potential role of migration.7 LABOR INSTITUTIONS. Because they affect the decisions of both em- ployers and workers, labor institutions and policies affect labor adjustment in many ways.8 In Southeast Asia wage setting is generally decentralized and collective bargaining is relatively uncommon, except in the Philip- pines and Korea (table 1.2). There are certain restrictions on hiring and firing in some countries, although enforcement is often weak. Social pro- tection for unemployed workers is minimal. Only Korea has an unem- ployment insurance scheme (introduced in 1995), but coverage is low. In the other countries some unemployed workers (usually older workers) have certain drawing rights on provident funds, but these funds tend to offer only minimal support. Severance exists in all countries, although again enforcement is weak. And none of the existing institutions offers protection directly to the informal sector. Evidencejfrom Southeast Asia The chapters that constitute the first part of this volume describe the labor market impacts of the economic shock in each of the five crisis countries. TABLE 1.2 Summary of Labor Market Institutions, late 1990s, Crisis Countries Indonesia Korea Malaysia Philippines Thailand Minimum wage Yes Yes No Yes Yes Unionization rate Around 5% 12.6% (1998) 8.4% (1996) 27.0% (1997) 1.7% Level of wage Company level Company level Company level Company level Company level determination Restriction on Fixed term contract Fixed term contract No restriction No restriction No restriction hiring approved in certain more than one year cases prohibited Restriction on Dismissals require Dismissal article will be Advance notice Advance notice Advance notice dismissal tripartite enforced from 1999 required required required committee approval Dismissal article was enacted on Feb. 1998 Severance Mandatory severance Mandatory severance Mandatory severance Mandatory severance Mandatory severance payment payment payment payment (83% payment payment paid in 1998) (compliance is noted to be low) Social protection Employees with more UI introduced in 1995. Employees age 50+ Emergency loan for Employees have for than 5 year service Covering more than can withdraw part displaced workers some drawing unemployed can withdraw 30 employees' firms. of provident fund. (introduced in rights from accumulated fund Coverage of UI was March 1998) provident fund from expanded to all since JAMSOSTEK. Jul. 1998 Public 1.5 million job 2.1 million job 1.2 million job 0.4 million job 0.7 million job Employment applicants (37% applicants (140% of applicants (30% applicants (10% applicants (40% Service of unemployed) unemployed) of unemployed) of unemployed) of unemployed) registered with PES registered with PES registered with registered with registered with (1997) (1998) PES (1998) PES (1998) PES (1998) Source: Based on country papers (this volume). I2 EAST ASLAN LABOR MARKETS These chapters focus on the first full year of the crisis. Cross-country em- pirical analysis tends to be complicated by data comparability issues, and the discussions therefore take into account definitional differences and se- ries breaks. We look here at the major points covered in the chapters. ADJUSTMENTS IN THE DEMAND FOR LABOR. Quantity-related adjust- ments varied across the region but were most pronounced in Korea and Thailand (table 1.3). Between 1997 and 1998 open unemployment more than doubled in these two countries, although previous levels had been relatively low. Increases in unemployment were relatively small in Indone- sia and the Philippines. The modest increase in the Philippines reflects the equally modest impact of the crisis there. But unemployment rose only 0.7 percentage points in 1998 in Indonesia, despite the fact that GDP de- clined by 13.7 percent. Similar results emerge for underemployment.9 Employment trends repeat this pattern. The largest absolute declines in 1998 occurred in Korea, Thailand, and Malaysia (in that order). Indone- sia and the Philippines recorded employment gains, despite the effects of El Nifio and despite Indonesia's substantial decline in output. In Indo- nesia the increase was the result of a gain of 5.4 percent in agricultural employment (table 1.4). In contrast the other countries in the region ex- perienced either no growth or declines in agricultural employment. All countries, including Indonesia, experienced employment losses in indus- try (manufacturing and mining). Losses in this sector were particularly large in Korea, the most industrialized country in the region. The crisis also resulted in real wage decreases in all countries (table 1.5).1o The decline was particularly dramatic in Indonesia, where real wages fell an estimated 41 percent, reflecting the intensity of the impact there. But in comparison with the other countries, Indonesia's labor mar- ket response involved mainly prices rather than quantity. Downward pres- sure on wages was also significant in Korea (9.3 percent) and Thailand (7.4 percent). ADJUSTMENTS IN THE SUPPLY OF LABOR. Labor force participation rates vary for the five countries we discuss (table 1.6). In three countries (Korea, Malaysia and to a small extent Thailand), the participation rate fell be- tween 1997 and 1998, suggesting that many workers became discouraged and stopped looking for work. In Indonesia and the Philippines, partici- pation increased as more workers joined the labor force (the added worker effect). The labor supply response was generally greater among women than men. In countries where aggregate participation rates fell (Korea and TABLE 1.3 Summary of Unemployment, Underemployment, and Employment Trends, Crisis Countries, 1 99%...98a Unemployment Underemploymentb Employment % Change % Change % Change 1996 1997 1998 1997-98 1997 1998 1997-98 1997-98 Indonesia 4.9 4.7 5.4 14.9 35.8 39.1 9.2 2.7 Korea 2.0 2.6 6.8 161.5 7.3 9.3 29.2 -5.3 Malaysia 2.6 2.6 4.0 53.8 7.3 7.9 8.2 -2.7 Philippines 8.6 8.7 10.1 16.1 11.3 11.9 5.3 5.8 Thailandc 2.0 2.2 5.2 136.4 0.9 1.2 33.3 -2.8 a. Thailand figures for February; Indonesia for August; Korea and the Philippines are averages of quarterly estimates. b. Underemployment (as % of employed population) defined as: Malaysia, less than 30 hours per week; Korea, 35 hours or less per week; In- donesia, less than 35 hours per week; Thailand and the Philippines, less than 40 hours per week and available for more hours. c. Working age population in Thailand defined as 13 years of age and older. Sources: Country papers with updates (this volume); Kakwani (1998); World Bank, Thailand Economic Monitor, January 2000. I4 EAST ASIAN LABOR MARKETS TABLE 1.4 Employment Change by Industrial Sector, Crisis Countries, 1997-98 % Change in employment, 1997-98a Agriculture Industry Servicesb Indonesia 5.4 -1.5 1.3 Korea 0.0 -13.2 -4.4 Malaysia -5.4 -2.9 -1.9 Philippines -8.3 -8.6 21.4 Thailand -1.8 -1.9 n/a a. Philippines figures for January 1998-99; Korea for 4th quarter 1997-98. b. Includes construction. Sources: Country papers (this volume); World Bank (2000b). Malaysia), declines were primarily the result of falling female participation rates. But participation rose in Indonesia as more women joined the labor force. Migration, both internal and external, has been an important response to the crisis. In Indonesia and Thailand in particular, urban-rural migra- tion has been an important element in the adjustment process." In Thai- land labor flows from rural areas to Bangkok had been large before the cri- sis, but these were significantly curtailed with the onset of the crisis. Ultimately Thailand began to experience reverse migration that placed downward pressure on rural incomes (World Bank 2000b). The number of nationals going overseas to work increased in both Indonesia and the TABLE 1.5 Real Wage Trends, Crisis Countries, 1996-98 % Change in real wages 1996 1997 1998 Indonesia 10.8 8.7 -41.0 Korea 6.7 2.4 -9.3 Malaysia 2.1 3.3 -1.1 Philippines -2.0 -1.1 -2.0 Thailand 2.3 1.4 -7.4 Sources: Based on country papers with updates (this volume) and Economist Intelli- gence Unit. EAST ASIAN LABOR MARKETS AND THE ECONOMIC CRISIS: AN OVERVIEW I5 TABLE 1.6 Labor Force Participation Rates, Crisis Countries, 1996-98 Participation rates Absolute change 1996 1997 1998 1997-98 Indonesia 66.9 66.3 66.9 0.6 Korea 62.0 62.2 60.7 -1.5 Malaysia 66.7 66.6 64.3 -2.3 Philippinesa n/a 65.5 66.1 0.6 Thailandb 53.4 53.0 52.7 -0.3 a. Philippines figures are for October. b. Thailand working age population defined as 13 years of age and older. Source: Country papers and updates (this volume). Philippines. The number of foreign workers employed in Malaysia fell as working conditions worsened, in part because the government stopped is- suing work permits and raised the costs of permits for workers who al- ready had them. THE CRISIS AND INFORMAL EMPLOYMENT. Data availability is a limiting factor in assessing how much adjustment has occurred through the infor- malization of employment. One indicator is the share of workers in dif- ferent forms of employment. Employees include wage earners, who typi- cally work in the formal sector. Self-employed and unpaid family workers are often (though not always) in the informal sector, so an increase in their relative shares suggests that many formal workers are moving to the infor- mal sector. Wage employment declined substantially in Indonesia, Korea, and Thailand, and more modestly in Malaysia (table 1.7).12 In Korea the number of those engaged in nonwage employment (employers, the self- employed, and unpaid family workers) decreased slightly; in Malaysia, they increased. But in Thailand the number of workers in the employer and self-employment categories increased, and unpaid family work de- clined. In Indonesia the informal sector showed strong absolute growth in the first year of the crisis, accounting for the country's overall increase in employment. THE IMPACT ON VULNERABLE GROUPS. In chapter 9 Horton and Mazumdar examine how the economic crisis affected vulnerable members of the labor force, including women, youth, workers with few skills, and TABLE 1.7 Employment Status Trends, Crisis Countries, 1997-99 Indonesia Korea Malaysia Philippines Thailand Distribution Change Distribution Change Distribution Change Distribution Change Distribution Change 1997 1998 1997-98 1997a 1998a 1997-98 1997 1998 1997-98 1997 1998 1997-98 1997b 1998b 1997 98 Employer 1.7 1.7 3.4 2.6 2.9 13.5 2.5 2.7 1.3 Self- employed 43.7 45.2 6.3 28.1 29.2 -2.3 16.9 17.7 5.0 37.5 37.6 2.5 30.7 32.2 1.9 Unpaid family worker 19.6 20.6 8.1 8.8 9.5 1.2 6.0 6.0 0.3 14.8 13.4 -7.4 20.3 19.7 -5.5 Employee 35.0 32.4 -4.9 63.1 61.3 -8.3 74.5 73.4 -1.1 47.7 49.0 5.0 46.4 45.4 -5.0 Total 100 100 100 100 100 100 100 100 a. Figures are for fourth quarter. b. Figures are for February. Sources: Country papers and updates and Horton and Mazumdar (this volume). EAST ASIAN LABOR MARKETS AND THE ECONOMIC CRISIS: AN OVERVIEW I7 migrant laborers. Data for the latter two groups are particularly difficult to obtain. Even for women and youth information is generally available for unemployment and labor force participation, but not for other im- portant indicators such as wages and hours. Have female workers been especially hard hit by the crisis in Asia?13 The available evidence offers only a partial picture. Differentials in unem- ployment rates between men and women are not strong. In Korea the un- employment rate for men increased more than the unemployment rate for women, but the other countries show no significant differences in male and female unemployment rates. And as we have already noted, female participation rates rose in Indonesia. Young workers tend to be particularly vulnerable during an economic crisis, not only because of their lack of experience and low productivity, but also because of hiring slowdowns and seniority practices. The crisis did not substantially affect the traditionally high ratio of youth unem- ployment to aggregate unemployment. But with the increasingly slack labor market, youth unemployment rose to very high levels. For young males (15-24 years old), 1998 unemployment rates ranged from 11.2 per- cent in Thailand to 20.8 percent in Korea. For females the range was 8.6 percent in Thailand to 22.1 percent in the Philippines. Participation rates for youth decreased substantially in Korea and Thailand and slightly in In- donesia, but they increased somewhat in the Philippines. Interpreting the Evidence In the first year of the crisis, the Korean, Malaysian, and Thai labor mar- kets seem to have responded in broadly comparable ways (table 1.8). All experienced downward pressure on both employment and wages, declin- ing labor force participation, and a shift in employment from the formal to the informal sector. But the magnitude of these adjustments was much more severe in Korea, despite the fact that the scope of the economic shock was slightly smaller there than it was in Malaysia or Thailand. Korea had the region's biggest increase in the region in unemployment (4.2 per- centage points) and the largest decline in employment (5.3 percent). Sim- ilarly employment declines in the industrial sector (13.2 percent) and in wage employment (8.3 percent) were considerably larger than in the other four countries. Real wages also declined substantially (9.3 percent). The overall pattern was much the same in Malaysia and Thailand, al- though adjustments were more moderate, especially in Malaysia. In each country unemployment rose in the first year of the crisis (3.0 percentage TABLE 1.8 Summary of Labor Adjustment in the Crisis Countries Panel A Panel B Severity of Industrial crisis structure (% change (% of Labor market institutions Labor in GDP employment in Unemployment Quantity Price supply Informal 1998) agriculture) Regulation benefits adjustment adjustment adjustment adjustment Indonesia -13.7 44 Medium No Low Very high Small Large increase Korea -5.8 12 High Yes High Medium Decrease Medium Malaysia -7.5 20 Low No Medium Low Decrease n/a Philippines -0.4 42 Medium No Low Low Small Small increase Thailand -10.0 50 Low No High Medium Small Medium decrease Source. Assessments by the authors based on country papers. EAST ASIAN LABOR MARKETS AND THE ECONOMIC CRISIS: AN OVERVIEW Iv points in Thailand and 1.4 percentage points in Malaysia), and employ- ment fell (2.8 percent in Thailand and 2.7 percent in Malaysia). In both countries wage employment declined (5.0 percent in Thailand and 1.1 percent in Malaysia). Migration was also an important adjustment mech- anism in Thailand, which saw extensive urban-to-rural flows, and in Malaysia, where the number of foreign workers fell. The major difference in the Thai and Malaysia experiences was the larger decline in real wages in Thailand-7.4 percent, compared with 1.1 percent.14 Compared with the other countries, labor adjustment in the Philip- pines was moderate in all ways. This moderation reflects the much smaller output shock (a 0.4 percent decline in GDP in 1998) caused by the crisis. The labor market experience in Indonesia was also unique, with huge downward wage adjustments and a significant shift in employment to the informal sector. Real wages declined by an astounding 41 percent, with all sectors experiencing similar downward pressures. Workers displaced from the wage sector and new entrants went into informal employment in large numbers. In 1998 agricultural employment increased by over 5 percent, and self-employment and unpaid family work grew by 6.3 and 8.1 per- cent, respectively. Despite the fact that the output shock was greatest in Indonesia, however, the unemployment rate rose less than 1 percentage point, and employment actually increased by almost 3 percent in the first year of the crisis. The labor force participation rate also rose slightly. Many aspects of the experience in Southeast Asia-the decline in real wages, labor displacement in the industrial sector, and a shift from wage to nonwage employment-are common across countries in the region. But the way the labor markets have adjusted also differ in important ways. As we have seen Korea saw significant reductions in employment, while Indonesia saw informal sector expansion and extremely high wage reduc- tions. These patterns are not surprising, given the differences between the two countries in level of development, labor regulation, and formal social protection and in the severity of the economic shock.15 In the final analysis, evaluating how differences in the adjustment process affected the welfare of workers will be of the utmost importance. Economists often argue that adjusting to reduced labor demand through downward wages is preferable to quantity adjustments that result in open unemployment. But as Horton, Kanbur, and Mazumdar (1994) point out, the net welfare effects of these different adjustment paths are not so clear. Unfortunately assembling the evidence necessary to compare out- comes is not straightforward. Poverty rates are one potentially important indicator, and these rates rose significantly in both Korea and Indonesia. 20 EAST ASIAN LABOR MARKETS In Korea the urban poverty head count went from 9 percent in 1997 to 19 percent in 1998 (World Bank 2000a).16 In Indonesia the rate rose from 11 percent in 1996 to 20 percent in late 1998 (World Bank 2000b). But these data are spotty and not generally comparable across countries, and their links to labor market experiences are difficult to establish. More de- tailed analysis will be required to identify the full labor market impacts of the Asian crisis. Government Responses to the Crisis Governments in Southeast Asia have responded to the crisis with two types of interventions: active labor market programs, and measures de- signed to support incomes. Before the crisis experience with active labor market interventions (especially the kind found in industrial economies) in the region was limited, except in Korea. Since the crisis Indonesia, the Philippines, and Thailand have instituted direct employment creation schemes, including labor-intensive infrastructure construction programs and credit schemes to promote self-employment and enterprise develop- ment.17 All five countries have functioning institutions that provide voca- tional training and employment services, and their responses to the crisis have included interventions in skill training and job search services specif- ically for displaced workers. Job Creation Measures All five countries discussed in this volume have introduced or revived pub- lic works programs, in part because introducing such schemes as an emer- gency measure was a natural response to the crisis. In addition to these programs, and as a longer-term measure, governments have put in place or stepped up programs that support (mainly in the form of credit) self- employment and enterprise development. PUBLIC WORKS PROGRAMS. Governments introduced a variety of public works programs in Southeast Asia (table 1.9). Indonesia reintroduced the labor-intensive infrastructure construction programs (padat karya) that had been phased out in the early 1990s as the labor market tightened. The programs started in 1997 with four-month projects and were followed in 1998 by large-scale programs covering the entire country. The programs were expected to generate 300 million person-days of work. In Korea pub- lic works programs generated 440,000 jobs in 1998 and nearly 1.2 million EAST ASIAN LABOR MARKETS AND THE ECONOMIC CRISIS: AN OVERVIEW 2I jobs in 1999, providing work for around 70 percent of the country's 1.7 million unemployed in 1999. The Thai government also launched massive programs using both government and donor funds. In the absence of unemployment insurance schemes (except in Korea), public works programs provided unemployed workers with much-needed income. These schemes have also proved more cost-effective than other types of income support, such as subsidies. The Indonesian public works programs, for example, spent less than $4 for each $1 that was transferred to the poorest 15 percent of the population. The cost of rice subsidies was twice as high-$8.20 for each $1 transferred (World Bank 1998). Despite their positive aspects, all the programs suffered from a number of problems. First, they were not well targeted. In general the best type of targeting for public works programs is self-targeting, but in Southeast Asia this method did not work because wages were set too high. Second, many of the programs were poorly designed, as they were rushed into operation when the crisis hit. And in countries like Indonesia and Thailand, where public works programs had already been phased out, planning and imple- mentation capacity was severely limited. Third, the programs suffered from poor coordination and a lack of proper monitoring. In Thailand, for example, several programs were launched at once without proper coordination, and the projects were un- able to recruit enough workers. Fourth, female participation was low, largely because women often do not participate in construction projects and no programs were targeted specifically to women. Finally, in Indone- sia the wage bill as a percentage of total project costs was too low, so that fewer jobs than anticipated were created. Even with their shortcomings, public works programs in Southeast Asia have demonstrated that carefully designed and implemented schemes can succeed. In Indonesia schemes that involved local communities have per- formed much better than those planned and implemented without such participation. In Korea women accounted for nearly half of the partici- pants in public works programs in 1999. PROGRAMS IN SUPPORT OF SELF-EMPLOYMENT AND ENTERPRISE DEVEL- OPMENT. Many Southeast Asian countries had programs supporting self- employment and the development of small enterprises in place before the crisis. These programs focus mainly on credit, although they also include a variety of technical support services. High interest rates and the credit crunch during the crisis posed problems for some of these initiatives, but TABLE 1.9 Job Creation through Public Works Programs Brief description of the Country public works program Problems identified Observations Indonesia A four-month, Rp. 42 * Poor program design Projects with community billion program targeted - Poor targeting participation in design and retrenched workers from * Too low a wage bill implementation did better construction and compared with material manufacturing (December costs 1997). -Low incidence of female Massive infrastructure participation construction project with -Significant leakages 16 sub-programs covering all provinces (April 1998). Korea * Public works programs * High wage and poor * High female participation created 440,000 jobs in targeting 1998 and 1,190,000 jobs in * Poor project design 1999. Malaysia * Public works program extended Philippines * Public works programs * No geographical targeting * Department of Labor and created 476,000 jobs * No targeting of vulnerable Employment (DOLE) (1997) groups in some projects projects did target * Rural works program * Administrative delays vulnerable groups generated 4,837 jobs * High wages preclude self- targeting Thailand * Ministry of Interior * Lack of information about projects targeted to create location of the jobs 788,799 jobs (1999) * Lack of coordination of * Various donor-funded projects created labor projects targeted another 11 shortage million persons-months of * Low quality of projects jobs 24 EAST ASIAN LABOR MARKETS several countries implemented new programs aimed at creating jobs through enterprise development (table 1.10). In 1998 Korea initiated two programs. This first was designed to pro- vide jobs for unemployed professionals through new ventures and self- employment; the second combined training and start-up loans for small businesses. In 1998 Malaysia launched a fund for small-scale entrepre- neurs aimed at creating self-employment opportunities and a second fund for small and medium-size industries. The Philippines already had over 100 "livelihood programs" covering nearly 90,000 beneficiaries when the crisis struck in 1997, but high interest rates and reduced funding limited these programs during the crisis. Thailand also launched several programs to promote self-employment, but they were small in both size and cover- age. The government and some commercial banks introduced credit pro- grams for small and medium-size enterprises (SMEs) in 1999. The Indonesian government requires commercial banks to set aside a percentage of their loans for small borrowers. Cooperatives offer subsi- dized credit, and at least 24 government programs offer credit to micro- entrepreneurs and SMEs, particularly farmers, transmigrants, and women. Despite the plethora of programs, most require collateral and thus are not available to borrowers without assets. Only a few offer collateral-free credit-far fewer than are needed-and many have been constrained dur- ing the crisis by high interest rates and the credit crunch. Skill Training Interventions All five countries discussed here had national systems of vocational train- ing in place when the crisis hit, but the governments made special efforts to respond to crisis-related training needs (table 1.11). In Indonesia and the Philippines, the skill training programs undertaken in response to the crisis were small compared with the number of displaced workers. Indo- nesia's programs also suffered from design and targeting problems. Korea's training program, however, covered nearly a quarter of the unemployed in 1998, and in 1999 the government increased the training budget by 19 percent. Preliminary assessments indicate that the program in Korea did help increase employability, despite inadequate labor market information and management problems. Malaysia also launched several programs, including a retraining scheme for retrenched workers and a graduate en- trepreneur scheme. Thailand initiated several government- and donor- funded skill training programs, but they have not yet been assessed. EAST ASIAN LABOR MARKETS AND THE ECONOMIC CRISIS: AN OVERVIEW 25 TABLE 1.10 Job Creation through Support of SMEs and Self-employment Support for self- Country employment and SMEs Observations Indonesia * Government required * Most credit programs percentage of require collateral commercial lending to small borrowers * Subsidized credit for * Too few collateral- cooperatives free microcredit programs * 24 government * Very high interest micro and SME rates during the credit programs crisis targeted groups such as farmers, transmigrants, and women Korea * Government * The program allocated 700 billion targeted unemployed won and targeted professionals and 50,000 persons to managers support new ventures and self-employment (1998) * Training and start-up loans allocated 300 billion won and targeted 10,000 persons (1998) Malaysia * Small-scale * Aimed at creating Entrepreneur Fund self-employment allocated 390 million opportunities RM (1998) * SMI Fund (RM * For small and 1,123.4 million medium industries in through first quarter manufacturing of 1999) (Table continues on the following page) 26 EAST ASIAN LABOR MARKETS TABLE 1.10 (continued) Job Creation through Support of SMEs and Self-employment Support for self- Country employment and SMEs Observations Philippines * Over 100 "livelihood * High interest rates programs" covering and reduced nearly 30,000 availability of credit beneficiaries (1997) adversely affected * Multiple credit and small borrowers guarantee facilities for SMEs Thailand * Loan scheme of * Small in scale 10,000 baht and a group-based micro grant scheme * Various SME credit programs, public and private While the public sector remains the primary provider of training in all the countries mentioned, private training institutions are gradually as- suming an important role in this field. The Philippines and Thailand re- port that private providers are gaining in importance, and in Indonesia the private sector has become a significant source of skill training. A number of innovations have also been introduced, especially in financing. In 1998 Korea introduced a voucher-based pilot program that allows trainees to choose their providers. The Philippines has also introduced a voucher- style scholarship program, and Malaysia has set up a fund for retraining and skill upgrading financed with taxes on employers. Employment Services All five countries have public employment services that gather informa- tion on job seekers and vacancies and provide job-matching services (with varying degrees of success). The services offered differ markedly from those available in industrial countries, largely because of the lack of un- employment insurance, established labor market programs, and labor market information in the region. EAST ASIAN LABOR MARKETS AND THE ECONOMIC CRISIS: AN OVERVIEW 27 TABLE 1.11 Interventions in Skill Training Interventions in skill training Country in response to the crisis Observations Indonesia * Padat karya, P3T * Targeting problem. provided retraining * Design deficiency. for wage employment and self-employment. Korea * Training programs for * Increased the the unemployed employability of the covered one-quarter worker. of the unemployed * Provided mental (1998). security to the * A pilot program of workers. "training vouchers." - Inadequacy of labor market information and problems of management. Malaysia * Retraining scheme for retrenched workers (1998). - Graduate entrepreneur scheme. * Special apprenticeship programs. Philippines * Training grant scheme * Small portion of of DOLE covered displaced workers 1,545 displaced reached. workers in 1998. Public employment services across the region have responded to the economic crisis in a variety of ways (table 1.12). Indonesia and Thailand have introduced no new initiatives, while Korea, Malaysia, and the Phil- ippines have instituted a number of measures. An increasing number of Koreans are accessing the public employment service. To accommodate them the government has increased the number of offices and staff and introduced an Internet-based labor exchange system. The Philippines has 28 EAST ASIAN LABOR MARKETS TABLE 1.12 Response through Employment Services Country Response to the crisis Observations Indonesia * Regional employment service offices registered about 8 million laid- off workers; but no special measures undertaken. Korea * Increased number of public * The number of job employment services seekers using the PES agencies and counselors. increased. * Information technology * The effectiveness of utilized. PES is still a * One-stop "Employment question. Security Centers" set up. Malaysia * Registration and placement * New programs of retrenched workers achieved some introduced in January success. 1998. * "Job fair"/"job link" programs for retrenched workers. Philippines * 146 new Public Employment Service Offices (PESOs) were established and 118 were reactivated (1998). Expansion continued in 1999. * Philippine Job Exchange Network launched in 1998. * Jobs/Livelihood fairs conducted. Thailand * No significant new initiatives. EAST ASIAN LABOR MARKETS AND THE ECONOMIC CRISIS: AN OVERVIEW 29 also expanded its public employment service system and launched an Internet-based job-matching system. Malaysia has introduced a program to register and place retrenched workers, and both Malaysia and the Philippines have organized job fairs for retrenched workers. Malaysia has reported success with both its initiatives. Wage and Employment Subsidies Incentives for employers-often in the form of wage subsidies-are an- other active labor market measure that helps maintain existing jobs and create new ones. Such incentives have not been a significant policy tool in the Asian countries affected by the crisis, except for Korea and Malaysia. (The Philippines has two small wage subsidy programs targeting young people.) In Malaysia job-maintenance programs are designed to encourage the private sector to choose pay cuts, temporary layoffs, retraining, flexi- ble work hours, and part-time employment over retrenchment. Korea has introduced an effective program to provide subsidies to firms that agree to maintain their current workforce. Income-Support Measures Income support for workers in Southeast Asia is limited (see chapter 8). The only country in the Asian region that had unemployment insurance when the economic crisis hit was Korea, and none of the other countries has instituted an unemployment insurance scheme since the crisis. The Employment Insurance System established in Korea in 1995 was initially limited to workers in firms with more than 30 employees. By October 1998 coverage had been extended to firms with fewer than 5 workers, temporary workers employed at least 1 month, and part-time workers working more than 18 hours a week. The length of time workers can col- lect benefits was also increased. Despite these reforms only 12 percent of unemployed workers were receiving benefits in mid-1999. As we saw earlier severance pay is widely used in Southeast Asia, but en- forcement is weak. Severance pay varies from two to six months' salary in the countries affected by the crisis. But many bankrupt companies have not met their obligations, and they have not been forced to do so. Special funds have been set up in Korea and Thailand to guarantee that severance is paid (Lee 1999). A final form of income support available to retrenched workers is money they have saved in the state-run provident funds designed mainly 30 EAST ASIAN LABOR MARKETS to provide retirement benefits. The proportion of employed workers able to access this benefit varies considerably across countries but is only 12 percent in Indonesia and 16 percent in Thailand. The average balance per worker in the national provident fund of Indonesia in 1997 was the equiv- alent of $22 and thus could not be regarded as a substantial means of in- come support. Emerging Labor Policy Issues in Southeast Asia18 For the most part the impact of the Asian economic crisis on labor mar- kets has been similar to the impact of other serious downturns. Workers throughout the region have suffered declines in real earnings, industrial sectors have contracted, and employment has shifted from formal to in- formal activities. These adjustments have been most dramatic in Indone- sia, where the effects of the crisis have been most severe. In parts of the region, most notably Korea, open unemployment has increased as well. With most countries now enjoying some degree of stabilization and re- covery, employment is beginning to pick up. However experience from earlier crises suggests that the upswing in labor markets during recoveries progresses more slowly than the downswing during contractions. The experience of the crisis has been an important one for policymak- ers, both in terms of responding to major economic downturns and in terms of building a sustainable development path for the future. During the era of rapid development in East and Southeast Asia, labor market pol- icy was put on the back burner. Growth, after all, is the best way to in- crease employment and raise incomes. When the crisis hit, none of the af- fected countries (except to a small extent Korea) had policies in place to help workers cope. Governments throughout the region must now work to develop policy instruments to manage labor market risks. As they do so they must address several issues. First, they must find the right mix of poli- cies to promote job growth, improve skills, increase labor productivity, and protect workers against both structural and cyclical shocks. Second, given that traditionally the family and community have served as support structures, governments must find a balance between formal and informal mechanisms. And policymakers in each country must identify the kinds of industrial relation institutions that are appropriate to both the country's culture and its social and economic stage of development. The Southeast Asian countries will cope with these complex issues in different ways. There is no simple solution to the question of income EAST ASIAN LABOR MARKETS AND THE ECONOMIC CRISIS: AN OVERVIEW 3I support for unemployed workers, for instance. As mentioned earlier, un- employment insurance had a limited coverage in precrisis East Asia. Lee (1999), however, makes a strong case for introducing unemployment in- surance. While refuting the arguments against this form of income sup- port, he addresses the technical issues of affordability and feasibility from the point of view of a number of countries and concludes that "an average required contribution rate of between 0.3 and 0.4 per cent of payroll from 1991 to 2000 would have made a significant contribution to cushioning the harsh impact of the crisis on modern sector workers"(p. 83).19 There is also the view that many countries in the region have relatively few resources and a small formal sector, and in these economies unem- ployment insurance may not be the most effective means of providing support for the unemployed. Chapter 8 raises a series of questions about unemployment insurance and suggests alternative approaches that merit careful consideration from governments in the region. Labor market interventions by and large have focused on direct job cre- ation rather than on skill training and job search assistance. One reason for this focus is that in the absence of income-support measures such as unemployment insurance, governments saw job-creation schemes as a quick means of generating much-needed incomes for those in distress.20 Chapter 7 notes that the effectiveness of skill training as an active labor market measure in industrial countries has been called into question be- cause of the low rates of return associated with training programs. But the social benefits of skill training, especially for disadvantaged groups, also need to be examined, along with cost-cutting measures. And given the fact that job search assistance programs have been found to be relatively cost- effective, countries may want to use them to broaden the range of active labor market policies during the recovery. Public Works A number of issues relating to public works have emerged from the expe- rience of the Southeast Asian countries that implemented these programs in response to the crisis. First, there is the question of objectives. While it is possible to think of job creation and income generation as the sole ob- jectives of such programs, experience shows that pursuing other goals- for instance, the provision of much-needed infrastructure and social capital-is possible and indeed desirable. If labor-intensive infrastructure projects are carefully selected and implemented, they can meet the crite- 32 EAST ASIAN LABOR MARKETS rion of economic efficiency in low-wage situations. And the main devel- opment justification for such programs is their ability to serve a dual func- tion as a social safety net in crisis situations and as a long-term means of creating assets. But the objectives need to be made explicit so that they can be taken into account in designing and evaluating the programs. Institutional capacity for designing, implementing, and monitoring job creation programs is a second important issue. Countries where labor- intensive public works programs were phased out or had never existed had limited capacity to institute them. Such countries could not realistically be expected to launch projects that would meet the criteria for sound design and quality. The programs they rushed to implement were certainly open to criticism. In some cases low-priority projects that had been discarded earlier were resumed. And the administrative machinery of the govern- ments was also caught unprepared, resulting in inefficient implementation and monitoring. Policymakers must now consider how best to prepare for a crisis situation and then set up the necessary organizational machinery. A third important issue for direct job creation programs is targeting. While the poor constitute a category that is obviously in need of as- sistance, the crisis affected vulnerable groups differently, as chapter 9 dis- cusses, and its impact also differed across geographic regions. These dif- ferential effects need to be considered in planning and implementing job creation programs. In addition, mechanisms need to be devised that will attract target groups. Self-targeting through wage setting is an effective mechanism for ensuring that public works projects benefit the poor. Wages that are set too far above market levels-especially in the absence of means testing-often attract those with high reservation wages and other nontarget groups. Alternative targeting mechanisms, such as involv- ing communities in the selection of beneficiaries, also need to be explored. Se/f-employment and Entrepreneurship Support Support for self-employment and entrepreneurship, especially programs aimed at microentrepreneurs and small enterprises, can be a promising means of job creation. These programs also involve issues of targeting, as well as the problem of collateral for credit. But two additional issues have emerged from the experience in Southeast Asia. First, in a major economic crisis with soaring interest rates and a severe credit squeeze, promoting and sustaining credit-based programs may be difficult until interest rates fall to reasonable levels. Second, in a deep recession, contractions in demand may add to the challenges microentrepreneurs and small enterprises face. EAST ASIAN LABOR MARKETS AND THE ECONOMIC CRISIS: AN OVERVIEW 33 Vocational Education and Training The economic crisis has also changed the overall context of skill training. In many countries the crisis has highlighted the lack of institutional ca- pacity to respond to changing situations. The importance of preparing training systems to respond to structural reforms has become increasingly apparent. Even in noncrisis situations growing and dynamic economies undergo structural reforms, and governments must provide the training necessary to meet the ensuing challenges. Doing so will require improving coordination among providers and incorporating a stronger demand-side orientation into the training systems themselves. Employment Services A number of issues have emerged in the area of employment services. The key question concerns the effectiveness of publicly provided job search as- sistance during a period of recession. The crisis has called into question the institutional capacity of Southeast Asian economies to provide effective public employment services. Governments must find ways to extend the geographic reach of employment services and the range of services pro- vided, to ensure that the services offered are in demand, to update tech- nology (particularly computer networks), and to provide training for staff. By offering comprehensive services, including counseling, labor market information, and referrals to skill training, public employment services can make themselves more relevant. And as the economies recover, em- ployment services can play a valuable role in efficiently matching workers with new job opportunities. Governance and Dialogue The Asian economic crisis has highlighted the importance of transparency and accountability in governance and the need for dialogue that will in- volve civil society in the process of change. Even before the crisis many Southeast Asian countries were looking at political reforms that would fos- ter an environment of participation and dialogue. As chapter 10 points out, dialogue has economic as well as social benefits. And Lee (1999) ar- gues that "the strengthening of democratic institutions is central to the post-crisis economic model that is required" (p. 64). Indeed, the economic crisis created situations and opportunities for closer adherence to basic in- ternational labor standards and further consolidation of appropriate insti- 34 EAST ASIAN LABOR MARKETS tutions of social dialogue at the macro level. The experience has not only demonstrated the negative consequences of a relative neglect of labor rights and social protection in the precrisis period, but also the potential as well as actual role of labor standards and social dialogue in ensuring smooth ad- justment to structural change and in coping with economic crisis.21 But meaningful and effective dialogue in the area of labor policy re- quires strong labor market institutions as well as an enabling political and social environment, and the economic crisis has weakened many of these institutions. Large-scale retrenchments have resulted in declining union membership. And the national-level tripartite initiatives some countries have instituted are not effective substitutes for established institutions. Governments need to continue working to institutionalize a culture of co- operation and dialogue. The legislative reforms that have been imple- mented so far represent important steps toward creating such a culture. These efforts must not be abandoned as economies find themselves on the road to recovery. Notes 1. The project received substantial funding from the Asia-Europe Meeting Trust Fund, which is sponsored by the European Union. The World Bank and the International Labour Organization greatly appreciate the contributions of the European Union, the Japanese Ministry of Labor, and the Japanese Institute of Labor. 2. Real wage growth was 52 percent in Korea (1990-97), 25 percent in Malaysia (1990-96) and Thailand (1990-97), and 13 percent in Indonesia (1990-96) (ILO 1999a). The decline in the Philippines was 5 percent between 1990 and 1995. 3. A caveat should be made concerning the role of economic policy. The par- ticular macroeconomic responses to the financial crisis have not been uniform across the region. For example Korea and Thailand largely adhered to the ortho- dox prescriptions of the International Monetary Fund, while Malaysia took a dra- matically different approach that included instituting capital controls. These dif- ferences in economic policy undoubtedly have affected how the crisis has been transmitted to the labor market, but we do not explicitly consider them here. 4. For a conceptual discussion see Layard, Nickell, and Jackman (1994). Also see Horton, Kanbur, and Mazumdar (1994) for a review of these issues based on case studies in developing countries. 5. In terms of quantity adjustments, reduced labor demand can also be man- ifested in underemployment, or reduced working hours. Unfortunately there is little evidence of such adjustment. 6. The GDP per capita figures shown in table 1.1 support this relative ranking. EAST ASIAN LABOR MARKETS AND THE ECONOMIC CRISIS: AN OVERVIEW 35 7. The participation rates for Thailand in table 1.1 are based on a working- age population defined as 15 and older (ILO 1999b). These rates differ from those cited later in the chapter, which come directly from the Thai Labor Force Survey and are based on a population 13 years and older. 8. The extensive literature on this issue examines the impact of minimum wages, wage determination processes, employment protection laws, active labor market programming, and social insurance on labor market performance. See, for example, OECD (1999); Elmeskov, Martin, and Scarpetta (1998); Nickell and Layard (1997). On labor market adjustment to shocks, see Blanchard and Wolfers (1999). This literature deals largely with labor markets in industrial countries. Much less analysis has been done on developing countries. See, however, Freeman (1992) and Lindauer (1999). 9. The number of hours that constitute underemployment varies across coun- tries (see table 1.5), but in all cases the underemployment rate represents the pro- portion of the employed labor force that is working less than a full-time standard. 10. Wage earners represent only between one-third and two-thirds of the labor forces in the countries discussed here. Data on labor income for nonwage earners are scarce, so that estimating price adjustments for this group is difficult. Kakwani (1998) has computed labor incomes for different classes of Thai workers. Ac- cording to these calculations, average wages decreased by 7.1 percent between 1997 and 1998 (February surveys), and average total labor income (both business and farm) decreased by 10.1 percent. 11. ILO (1 999a) reviews migration patterns in Indonesia during the crisis. 12. Wage employment also shifted from regular and temporary forms to more informal daily work. 13. This volume is directly concerned with the position of women in the labor market. Second-order effects may also exist (such as the intrafamily distribution of income and declines in women's economic position before the crisis, especially for single mothers) that have had negative welfare impacts. 14. These figures relate only to formal (wage sector) employment. Possible inaccuracies exist in nominal wage data, and price deflators are often uncertain. The figures based on the country reports are not always consistent with real wage trends reported elsewhere. 15. The differences across countries are also attributable to the state of the labor markets at the onset of the crisis. For example the chapters on Thailand and Korea argue that even before the crisis, labor competitiveness in these countries had been eroding. These observations are supported by unpublished World Bank estimates, which find that unit labor costs in Korea and, to a lesser degree, in Thailand increased in the late 1980s and early 1990s. But unit labor costs in In- donesia declined during this period. The quantity adjustments observed in Korea and Thailand, but not in Indonesia, reflect trends in competitiveness that predate the crisis. The macroeconomic shock simply provided a stimulus for overdue labor rationalization. 36 EAST ASIAN LABOR MARKETS 16. This rate seems to be dropping rapidly as the recovery takes hold in Korea. The figure for the first quarter of 1999 was 15 percent (World Bank 2000b). 17. Such schemes are better known by their generic name, public works, al- though this term no longer reflects the nature of such schemes. In many instances the projects are run by private contractors. 18. The Chairman's Summary from the seminar on Economic Crisis, Em- ployment, and the Labor Market in East and Southeast Asia, which is appended at the end of this chapter, provides a good overview of the labor issues that emerged from the recent economic crisis in Asia as well as ideas for possible re- sponses to such issues. This section should, therefore, be read along with the Chairman's Summary appended to the present chapter. 19. Indeed, the ILO Governing Body symposium on the Social Impact of the Asian Financial Crisis held in Geneva on 19-20 March 1999 called for the assignment of high priority to the design and implementation of efficient social insurance systems. 20. In Korea, where some unemployment insurance was available, the gov- ernment made considerable efforts in the areas of skill training and job search assistance. 21. One of the conclusions of the ILO Governing Body Symposium on the Social Impact of the Asian Financial Crisis referred to earlier statements that so- cial dialogue confers substantial economic and social benefits in terms of its con- tribution to non-conflictual economic restructuring and to the weathering of economic crisis. References Blanchard, Olivier, and Justin Wolfers. 1999. "The Role of Shocks and Institu- tions in the Rise of European Unemployment: The Aggregate Evidence." Cambridge, Mass: National Bureau of Economic Research. Working Paper 7282. Elmeskov, Jorgen, John Martin, and Stefano Scarpetta. 1998. "Key Lessons for Labor Market Reform: Evidence from OECD Country Experiences." Swedish Economic Policy Review 5(2) pp. 205-58. Fallon, Peter R., and Robert E. B. Lucas. 2000. "The Impact of Financial Crises on Labor Markets, Household Incomes and Poverty: A Review of Evidence." Washington, D.C.:World Bank. Unpublished paper. Freeman, Richard. 1992. "Labor Market Institutions and Policies: Help or Hin- drance to Economic Development?" Proceedings of the World Bank Annual Conference on Development Economics. Washington, D.C.: World Bank, pp. 117-44. Horton, Susan, Ravi Kanbur, and Dipak Mazumdar. 1994. "Vulnerable Groups and the Labor Market: The Aftermath of the Asian Financial Crisis." Paper prepared for the seminar "Economic Crisis, Employment, and the Labor Mar- EAST ASIAN LABOR MARKETS AND THE ECONOMIC CRISIS: AN OVERVIEW 37 ket in East and South-east Asia," sponsored by the World Bank, International Labor Organization, and Japanese Ministry of Labor-Japan Institute of Labor, Tokyo. ILO (International Labour Office). 1999a. Indonesia: StrategiesforEmployment-Led Re- covery and Reconstruction. Geneva. ____. 1999b. Key Indicators of the Labor Market. Geneva. Kakwani, Nanak. 1998. "Impact of the Economic Crisis on Employment, Un- employment, and Real Income." Bangkok: Thailand National Economic and Social Development Board. Conference Paper. Layard, Richard, Stephen Nickell, and Richard Jackman. 1994. The Unemploy- ment Crisis. New York: Oxford University Press. Lee, Eddy. 1999. The Asian Financial Crisis: The Challenge for Social Policy. Geneva: International Labor Office. Lindauer, David. 1999. "Labor Market Reforms and the Poor." Department of Economics Wellesley College. Wellesley, Mass. Unpublished paper. Nickell, Stephen, and Richard Layard. 1997. "The Labor Market Consequences of Technical and Structural Change." Discussion Paper Series 23. Oxford: Centre for Economic Performance. OECD (Organization for Economic Cooperation and Development). 1994. The Jobs Study. Paris. ____. 1999. Employment Outlook. Paris. World Bank. 1998. Indonesia in Crisis: A Macroeconomic Update. Washington, D.C. . 2000a. Global Economic Prospects and the Developing Countries 2000. Washington, D.C. _____. 2000b. East Asia Quarterly Brief January. Washington, D.C. Chairman's Summary KENJI TUNEKAWA THE SEMINAR ON THE Economic Crisis, Employment, and the Labor Market in East and Southeast Asia was held in Tokyo under the sponsorship of the World Bank, the International Labour Organiza- tion, the Japanese Ministry of Labor, and the Japan Institute of Labor on October 13-15, 1999. Tripartite delegations from Indonesia, the Republic of Korea, Malaysia, the Philippines, and Thailand; officials of international organizations; the Japanese government; and academic resource people met to discuss recent labor market trends related to the financial crisis; examine the roles of gov- ernments, trade unions, and employers' associations in each nation; and consider policy options both for mitigating the impacts of the crisis and for building strong labor policies in the region. The discussions were very constructive because of the interaction of the tripartite delegations, the well-prepared documents on which the pro- ceedings were based, and the sharing of experiences across countries. This was the first time for the World Bank and the ILO to hold such a seminar regarding the Asian financial crisis. The seminar demonstrated the usefulness of collaboration between the World Bank and the ILO in supporting a social development agenda. The following conclusions emerged from the discussions: 1. Generally, the macroeconomic and labor market conditions in the five Asian countries are improving, although the overall situation remains severe. 38 CHAIRMAN S SUMMARY 39 2. Wide diversity exists throughout the region in the stage of economic development, the impacts of the financial crisis, and the current status of the recovery. Therefore, countermeasures need to be designed in ac- cordance with each country's situation. 3. The financial crisis has created an opportunity for strengthening de- mocratic mechanisms, particularly social dialogue, in each country. This can enhance the search for solutions to the negative socioeco- nomic impacts of the crisis, and facilitate recovery by building consen- sus and securing commitment to decisions made. Therefore, it is im- portant to create an effective tripartite framework to enhance the exchange of ideas and policymaking. This must be accompanied by ef- forts to strengthen the capacity of trade unions and employers' associ- ations to engage in constructive social dialogue. 4. A strong labor policy framework includes active labor market pro- grams, human resource development, a social safety net for workers, and appropriate labor laws and standards. These components need to be well balanced and integrated. Countries need to consider both growth and social protection objectives and the formal and informal sectors. However, there is no uniform solution: an individual country's strategy will depend on its stage of development and its institutions. More detailed conclusions include: (a) Active labor market programs (ALMPs) include employment ser- vices such as job search assistance and placement services, training, and employment creation, including wage subsidies and public works. These programs have the potential to reduce unemploy- ment, enhance human resource development, increase the employ- ability of disadvantaged workers, and improve the functioning of the labor market. Evaluations of these programs, however, mainly in OECD countries, show that economic conditions, worker char- acteristics, and program design have significant impacts on their effectiveness. An effective active labor market program is essential as formal labor markets grow, and a skilled work force becomes in- creasingly important. East and Southeast Asian countries now need to think about the roles of the public and private sectors and build- ing capacity to design, implement, and evaluate programs. An im- mediate priority is to develop a strong employment service, since this is the cornerstone of the overall strategy. (b) Programs of labor-intensive infrastructure construction, if care- fully targeted and properly designed and implemented, can not 40 EAST ASIAN LABOR MARKETS only provide a valuable safety net for the poor and the unem- ployed, but also contribute to further economic recovery and de- velopment. Efforts should be made, therefore, to overcome the ob- served deficiencies in designing and implementing such programs. Increased involvement of local communities and private sector participation can contribute to their improvement. (c) Small and medium enterprises (SMEs) that utilize local resources, are less dependent on imported inputs, and are capable of exploit- ing export markets have demonstrated a degree of resilience during the current crisis as well as their potential for growth and employ- ment creation during the recovery period. Policies need to be geared toward creating an enabling environment and providing necessary support services for boosting their development. (d) The fulfillment of employment service systems is one of the most inexpensive and effective policies, but introducing this system dur- ing a recession might limit its effect. The development of public employment services during recovery is necessary not only for this crisis, but also for future economic development. (e) Human resource development is essential for national competi- tiveness. This includes strong universal basic skills, workers with advanced technological abilities, and personnel with the capa- city for creativity and innovation. While human resource develop- ment policy is more effective in economic recovery, it is now important to formulate an effective strategy and measures to meet the challenge. (f) The crisis has underlined the vulnerability of certain groups such as women, children (especially child laborers), the disabled, and foreign workers. Generally, protection for these groups has not been adequate compared with the hardship they have experienced. Appropriate policy responses will differ according to the group and its needs. In any event, policies need to reflect the fact that many vulnerable workers are in the informal sector. (g) The importance of a safety net for workers, including income sup- port for the unemployed, has been underlined by the impacts of the financial crisis. While it might be difficult in some countries to introduce a well-designed system that would have a significant impact during the current crisis, evaluating the design of new, or improving existing, systems should begin immediately. For for- mal sector workers, options include unemployment insurance, unemployment assistance, and strengthened severance systems. CHAIRMAN'S SUMMARY 4I There are relative strengths and weaknesses to each approach. Well-designed systems, however, can both limit the social costs of an economic downturn and help to speed recovery. Consideration of options includes financial and political sustainability, adminis- trative capacity, the degree of support offered to the unemployed, and the implications for labor market efficiency and economic re- structuring. Policymakers also must recognize the large numbers of informal risk management strategies already in place and how to extend and improve these. (h) Labor laws affecting hiring, deployment, and dismissal are needed to achieve the twin objectives of economic efficiency and social protection. Both of these objectives need to be considered jointly. Labor standards are also critical for protecting the rights of work- ers. The core labor standards included in the ILO Declaration on the Fundamental Principles and Rights at Work should be en- couraged in each country. The principles embodied in these stan- dards should be promoted, and the ILO Conventions to which they refer should be reviewed by governments and the social part- ners with a view to their ratification. 5. It is important to strengthen technical cooperation across countries and between countries and the international organizations to cope with both the impacts of the crisis and the need for policies for the long run. The collaboration between the ILO and the Bretton Woods institutions is important in order to implement this cooperation ef- fectively. The seminar has been an important event in promoting this cooperation. 6. The sharing of the experiences in each country and the international experience should enhance policy dialogue throughout the region. Ultimately, each country must examine its options to address labor is- sues. Further tripartite dialogue will be important. The ILO, the World Bank, and the Japanese Ministry of Labor look forward to following up the seminar with technical assistance and further dialogue. 2 The Economic Crisis: Labor Market Challenges and Policies in Indonesia RIZWANUL IsLAM, GOPAL BHATTACHARYA, SHAFIQ DHANANI, MAX IACONO, FARHAD MEHRAN, SWAPNA MUKHOPADHYAY, AND PHAN THUY UNTIL THE EAST AsIAN crisis Indonesia had enjoyed re- markable success with its economic development strategies. Annual GDP growth throughout the 1980s stood at 6.1 percent, rising to 7.6 percent in the first half of the 1990s and to 7.8 percent in 1996. The economy diversified as it grew. The manufacturing sector in particular turned in an impressive performance, growing at a rate of 10 percent yearly between 1985 and 1995 and ultimately accounting for a quarter of the nation's GDP Indonesia exported a wide range of manufactures, including textiles and apparel, wood products, and petrochemicals. Merchandise exports grew at an average of nearly 15 percent a year in 1986-93 but then began to decline; in 1996 the growth rate was 8.8 percent (World Bank 1997a; ADB 1997; ILO 1996). Indonesia's gross investment rate went from 24 percent of GDP in 1980 to 32 percent in 1996. Domestic savings also grew, and the economy con- tinued to finance about 90 percent of its investment domestically. In 1996 the domestic savings rate was 31 percent of GDP (World Bank 1997a). Indonesia's performance on the employment front was less impressive in the years leading up to the crisis. Employment increased more slowly than the size of the labor force (Islam 1998). Between 1988 and 1995 em- 43 44 EAST ASIAN LABOR MARKETS ployment grew at 2.3 percent annually, while the labor force grew at 3.1 percent. Annual employment growth actually declined from 2.8 percent in 1985-90 to 1.8 percent in 1990-95. The rate of open unemployment jumped from 3.2 percent in 1990 to 7.0 percent in 1995.1 In 1996 (im- mediately before the crisis) approximately 4.4 million people, or nearly 5 percent of the labor force, were already looking for jobs. The problem of unemployment was most serious in urban areas, where the rate rose from 6.1 percent in 1990 to 10.9 percent in 1995. Unemployment also seri- ously affected those with a university education: for this group, the rate jumped from 6.7 percent in 1990 to 11.5 percent in 1995. The 1996 Labor Force Survey showed that nearly a third of all workers were under- employed (that is, working fewer than 35 hours per week) in 1996. The informal sector accounted for nearly two-thirds of total employment. The labor force continued to expand annually by around 2.2-2.3 million, de- pending on the source of the data.2 Despite the increase in unemployment, Indonesia made progress in re- ducing poverty. In the two decades between 1976 and 1996, the incidence of absolute poverty (the percentage of the population living below the poverty line) declined from 40.08 percent to 11.34 percent.3 The benefits of growth appear to have been distributed fairly widely. All provinces ex- perienced increased income, consumption, and employment levels, although at different rates (and beginning at different stages of development). Indi- cators for life expectancy, infant mortality, literacy, and education also im- proved (World Bank 1997b). Indonesia's economy began to show signs of strain in 1996, when export growth slowed to less than 9 percent from its average of nearly 15 percent for 1986-93. GDP growth began slowing in 1997, falling from nearly 8 percent in 1996 to less than 7 percent in the first quarter of 1997 and to less than 6 percent in the second.4 The growth rate for 1997 as a whole was 4.9 percent. A severe drought during the year adversely affected agriculture, and the production of rice (the staple food in Indonesia) was an estimated 4 per- cent below 1996 levels. As a result the country had to import food grains for the first time in many years. The worst forest fires in more than a decade exacerbated the agricultural problems and devastated the tourist trade. In effect the financial crisis hit at a time when the Indonesian economy was already facing dwindling exports, a slowdown in overall growth, sharply reduced rice production, and a sharp increase in food prices. This chapter examines the effects of the crisis on the already strained labor market. In particular it explores the government's policy response to the crisis in three areas: employment creation, labor market services, and vocational educa- tion and training. LABOR MARKET CHALLENGES AND POLICIES IN INDONESIA 45 Recent Macroeconomic Trends In 1998 Indonesia's GDP contracted by 13.7 percent. Electricity, gas and water, oil and natural gas, and agriculture were the only sectors to record positive growth in that year. A drought induced by El Nifio affected agri- culture, and the growth rate declined from 3.1 percent in 1996 to 0.7 per- cent in 1997 and 0.2 percent in 1998. The construction sector was hit hardest by the crisis, experiencing a contraction of nearly 40 percent in 1998. Nonoil manufacturing and services also experienced very sharp de- clines (14.5 and 16.6 percent, respectively). Inflation, as measured by the consumer price index (CPI) stood at about 78 percent in 1998, contrast- ing sharply with the price stability of the precrisis period. All components of the CPI increased dramatically, especially food and clothing prices. Despite the massive depreciation of the rupiah, export revenues stag- nated in dollar terms in 1998-99. But the flat export revenues masked ro- bust growth in the export volumes of agricultural commodities such as coca, coffee, palm oil, spices, and tea. But low international prices for many of these commodities in 1998 meant that overall export revenues remained unchanged or declined. Imports declined by 11 percent in 1998-99 over the previous fiscal year. This compression, which was caused by the mas- sive depreciation of the rupiah and the collapse of aggregate demand, in- cluded declines in all categories of imports-consumer goods, raw mate- rials, and capital goods. The sharp decline in imports combined with flat export growth to generate a sizeable current account surplus for 1998-99 of $1.4 billion, or 1.1 percent of GDP, in contrast to the 1997-98 deficit of $1.7 billion. This development cannot be regarded as wholly favorable, however, because lower imports of capital goods and raw materials reflect declines in investment and manufacturing output. The overall macroeconomic picture began improving in 1999. GDP rose by 1.4 percent in the first quarter from the last quarter of 1998. And in the second quarter of 1999, GDP registered positive growth compared with the corresponding period in 1998. The economy appeared to be bot- toming out sooner than predicted. For 1999 as a whole, GDP rose by 0.31 percent, and during the first quarter of 2000, the growth rate rose to 5.01 percent. The risk of hyperinflation was contained, and forecasts for infla- tion were adjusted downward to just 10 percent for 1999. Declining in- flation and the relative stability of the currency in late 1999 allowed the central bank to lower interest rates to just over 20 percent-down from the very high rates (30 to 50 percent) that prevailed in 1998 and the first half of 1999. 46 EAST ASIAN LABOR MARKETS Labor Force Participation Rates Indonesia's working-age population grew by 3.5 million persons (2.6 per- cent) between 1997 and 1998, continuing earlier trends. A stable employ- ment rate of just over 63 percent of this group translates into more than 2 million additional labor market entrants. These new entrants consisted of teenagers just turning 15 and nearly 0.7 million persons from outside the labor force (homemakers and others), who were absorbed primarily into agriculture. Labor force participation among Indonesians of working age (15 years and older) remained essentially unchanged (at about 67 percent) in 1998. The slight upward trend of the female rate was largely offset by the slight downward trend of the male rate (table 2.1). Participation among adults (25 years and older) increased slightly in the aftermath of the financial crisis, ris- ing from 72.5 percent in 1997 to 73.5 percent in 1998. The decrease in the rate of participation among youth (those ages 15-24) was smaller than an- ticipated. The rate fell a statistically significant 2.7 percent in 1994-98 (from 53.0 percent to 50.3 percent). Similar standard error calculations for urban and rural trends also show no significant trends. Regional differences in labor force participation rates in Indonesia are relatively small (the spread between the highest and lowest regional participation rate is 18 percentage points).5 In 1997 children ages 10-14 accounted for 1.6 million workers, or less than 2 percent of total employment. The vast majority of them were engaged part-time in unpaid agricultural work for their families. Around 1.0 million of these children, or just under two-thirds of the total, attended school. Employment among these children did not increase sig- nificantly in the first year of the crisis, nor did the number actively look- ing for work. The number of child workers rose by just 80,000, or 5 per- cent (mosdy boys), and they were relatively evenly divided between urban and rural areas. Most of the children in this age group (over 85 percent) continued to attend school full time in both 1997 and 1998, although the rate for male rural boys dropped slightly. The proportion of teenagers ages 15-19 who were attending school increased slightly. Unemployment and Underemployment The economic crisis did not lead to massive open unemployment, as had been feared. In 1997-98 overall employment increased by nearly 3 percent (2.5 million) compared with 1.5 percent in 1996-97.6 The labor force ex- LABOR MARKET CHALLENGES AND POLICIES IN INDONESIA 47 TABLE 2.1 Aggregate Changes in the Labor Forcea Annual increase 1994 1996 1997 1998 94-96 96-97 97-98 (millions) Population (10+) 125.3 131.9 135.1 138.6 2.59 2.42 2.58 Labor force 83.7 88.2 89.6 92.7 2.65 1.61 3.50 Unemployed 3.7 4.3 4.2 5.1 8.31 -2.08 20.61 Employed 80.0 83.9 85.4 87.7 2.38 1.79 2.65 Nonparticipating 41.6 43.7 45.5 45.8 2.48 4.07 0.78 Students 9.8 10.8 10.8 11.3 4.87 0.18 4.25 Homemakers 22.1 24.0 25.9 25.2 4.08 7.95 -2.43 Other 9.7 8.9 8.8 9.3 -3.88 -1.70 5.98 (percent) Unemploymentb 4.36 4.89 4.68 5.44 0.26 -0.21 0.76 Labor forcec 66.80 66.87 66.34 66.93 0.04 -0.53 0.59 Unemployment 2.92 3.25 3.11 3.65 0.17 -0.14 0.55 Employment 63.88 63.62 63.23 63.28 -0.13 -0.39 0.05 Nonparticipating 33.20 33.13 33.66 33.07 -0.04 0.53 -0.59 Students 7.83 8.19 8.01 8.14 0.18 -0.18 0.13 Homemakers 17.67 18.19 19.17 18.24 0.26 0.98 -0.94 Other 7.70 6.76 6.48 6.70 -0.47 -0.27 0.21 Note: 1994-97 figures exclude 10-14 age group. a. Defined as those aged 15 and above. b. Percent of labor force. c. These and following figures are percent of population. Source: National Labor Force surveys; Central Bureau of Statistics. panded by 3.5 percent between 1997 and 1998, compared with just 1.6 percent in the previous year-a rate of increase 1 percent higher than the growth rate of the working-age population. Though employment increased in almost all age categories, adults over 30 accounted for most additional employment (90 percent). The open unemployment rate increased by less than 1 percent between August 1997 and August 1998, the first year of the crisis. The increase re- mained small because displaced workers from sectors such as manufactur- 48 EAST ASIAN LABOR MARKETS ing and construction were simply too poor not to work. Thus they were forced to find income-generating occupations in the agricultural and in- formal sector. For this same reason the open unemployment rate remained low in 1999.7 In the first year of the crisis, the number of Indonesians not working and actively looking for work increased by nearly 1 million, or 20 percent, climbing from 4.2 to 5.1 million. This increase followed a decline of 2 per- cent in 1996, so that the open unemployment rate actually increased from 4.7 percent to 5.4 percent of the labor force (table 2.1). But the crisis also drew nearly 2.5 million new participants into the labor force, about 1.0 mil- lion of them from urban households.8 This number is smaller than the cor- responding number for rural areas because the number of urbanites who left or remained outside the labor force actually increased by about 1.2 million persons, half of them students or homemakers. The situation was reversed in rural areas, where over 1.0 million women who had been homemakers and 0.5 million former students of both genders entered the workforce. A closer look at the statistics for the working-age population outside the labor force shows that the number of people under "nonparticipating" who are neither homemakers nor students increased by 0.5 million (table 2.1). Since this category includes unemployed people who are not actively looking for work (especially in urban areas), the overall increase in both open and disguised unemployment was probably on the order of 1.4 mil- lion persons, or a third more than before the crisis. Unemployment affected men more than women because two of the hardest-hit sectors-construction and manufacturing-employed mostly men. In the first year of the crisis the number of unemployed men rose by 0.6 million, or 26 percent, over the previous year. The number of unem- ployed women grew at just half this rate. Urban and rural areas were af- fected equally, although the overall unemployment rate remained much lower in rural areas (around 3.3 percent) compared with urban areas (9.3 percent) (table 2.2). Job seekers in 1998 differed markedly from those seeking work in 1997. Before the crisis most job seekers were young high school and col- lege graduates looking for their first job.9 Once the crisis hit, however, many job seekers were older and had prior work experience. The number of those who had worked nearly tripled, rising from just under 0.8 million to nearly 2.0 million. The share of those ages 30 and above, many of them with just a junior secondary school certificate, also tripled, climbing from 11 to 30 percent of the total. Unemployment also increased significantly among those with secondary and some university education. But it de- LABOR MARKET CHALLENGES AND POLICIES IN INDONESIA 49 TABLE 2.2 Characteristics of the Unemployed, 1994-98 Number Distribution 1994 1996 1997 1998 1994 1996 1997 1998 (millions) (percent) Total unemployed 3.65 4.29 4.20 5.06 100.0 100.0 100.0 100.0 Male 1.98 2.29 2.26 2.86 54.2 53.3 53.8 56.5 Female 1.67 2.00 1.94 2.20 45.8 46.7 46.2 43.5 Urban 2.14 2.49 2.57 3.10 57.9 57.9 61.2 61.3 Rural 1.51 1.80 1.63 1.96 42.1 42.1 38.8 38.7 Worked before 1.05 0.78 0.77 1.96 28.7 18.1 18.3 38.7 Never worked 2.60 3.51 3.43 3.10 71.3 81.9 81.7 61.3 15-19 1.07 1.39 1.39 1.44 29.2 32.4 33.1 28.5 20-24 1.55 1.65 1.60 1.19 42.4 38.4 38.1 23.6 25-29 0.65 0.77 0.75 0.93 17.7 17.9 17.9 18.3 30+ 0.39 0.48 0.46 1.50 10.7 11.3 11.0 29.6 Primary educa- tion or less 0.91 1.05 0.98 1.17 24.8 24.4 23.3 23.1 Junior secondary 0.63 0.79 0.74 0.98 17.1 18.1 17.5 19.4 Senior secondary 1.81 2.06 2.11 2.48 49.6 48.2 50.2 49.0 Diploma/ university 0.31 0.40 0.38 0.43 8.5 9.3 9.0 8.5 Source: National Labor Force Surveys. creased for those at the top of the educational ladder (including university graduates). Since open unemployment increased only marginally in the first year of the crisis, the real impact was felt in less work and lower real incomes. Un- deremployment (working fewer than 35 hours per week) increased in the first year of the crisis, while opportunities to work overtime and earn extra income declined. The number of people working fewer than 35 hours a week rose by 3 percent of the workforce, mainly because of the substan- tial increase in the number of agricultural workers, who often work shorter hours (especially unpaid family workers). Both urban and rural areas were 50 EAST ASIAN LABOR MARKETS affected, and men were more affected than women. In addition the pro- portion of those working more than 45 hours per week declined by 4 per- cent (from 39 to 36 percent) in both urban and rural areas. Again men were more likely than women to bear the burden of reduced hours; in fact women saw their share of overtime increase by 2 percent. Fewer hours and reduced opportunities to work overtime meant a reduction in take-home pay for most workers and contributed to the decline in living standards and increasing poverty (tables 2.3 and 2.4). Before the crisis underemployment had been falling steadily, in line with the declining importance of agriculture and nonwage employment. In 1997-98, however, the number of people working fewer than 35 hours per week rose from 37 to 40 percent. Underemployment rose more rapidly in urban areas (4 percent) than in rural areas (3 percent). It also rose more rapidly for men (3 percent) than for women (2 percent). Most of this in- crease reflected growth in the number of agricultural workers (nearly 5.0 million), particularly in the number of unpaid family workers. Underem- ployment rose even in this sector, however, rising by around 2 percent, just as it did in nonagricultural sectors as a whole. In fact, underemployment rose by nearly 3 percent in manufacturing, compared with 1.0-1.5 per- cent in trade and services. Job Losses Around 2.5 million workers, or 3 percent of the total work force, were dis- placed by the crisis in 1997-98. Job losses came from all sectors of the economy except agriculture and the small transportation and communi- cation sectors. Most displaced workers were wage employees, and men ac- counted for three-quarters of all job losses. The manufacturing sector was easily the largest loser, accounting for nearly half of all job losses, followed by construction and mining and trade and services. About three-quarters of the job losses in these sectors were in rural areas. In urban areas many workers displaced from the manufacturing and construction sectors en- tered the trade and other service sectors, keeping net job losses in the cities minimal. Displaced workers in rural areas did not have this option. Rural workers thus formed the bulk of all displaced nonagricultural workers, and many were forced into agricultural labor. Job losses in the manufacturing sector exceeded 1.0 million persons, or nearly half the total job losses, with men accounting for around 60 per- cent (table 2.5). Around three-quarters of all retrenchments in this sector occurred in rural areas. The decline in manufacturing employment was LABOR MARKET CHALLENGES AND POLICIES IN INDONESIA 5I TABLE 2.3 Trends in Underemployment, 1994-98 1994 1996 1997 1998 (millions) All employees 80.04 83.90 85.40 87.67 <35 hours/week 30.60 31.86 30.56 34.30 35-45 hours/week 20.12 21.36 21.25 21.87 >45 hours/week 29.32 30.68 33.59 31.50 (percent) All employees 100.0 100.0 100.0 100.0 <35 hours/week 39.3 38.9 36.6 40.0 35-45 hours/week 24.7 25.1 24.6 24.6 >45 hours/week 36.0 36.0 38.8 35.4 Male 100.0 100.0 100.0 100.0 <35 hours/week 29.6 28.4 26.5 29.6 35-45 hours/week 26.7 27.7 26.6 27.6 >45 hours/week 43.6 43.8 47.0 42.8 Female 100.0 100.0 100.0 100.0 <35 hours/week 55.5 56.6 53.8 54.4 35-45 hours/week 24.1 23.0 23.3 20.7 >45 hours/week 20.4 20.3 22.9 24.9 Urban 100.0 100.0 100.0 100.0 <35 hours/week 21.9 20.9 20.9 24.1 35-45 hours/week 26.4 28.0 26.5 27.1 >45 hours/week 51.7 51.1 52.7 48.7 Rural 100.0 100.0 100.0 100.0 <35 hours/week 46.9 47.8 44.9 47.0 35-45 hours/week 25.5 25.0 24.8 23.8 >45 hours/week 27.6 27.2 30.3 29.2 Note: Data are for the general population 15 years and older. Source: National Labor Force Survey. not limited to the most industrialized provinces, since 60 percent of work- ers in the manufacturing sector were self-employed or family workers. Many of these small-scale and household enterprises relied on sales to the general population, whose purchasing power was eroded by inflation. 52 EAST ASIAN LABOR MARKETS TABLE 2.4 Underemployment in Selected Sectors, 1994-98 1994 1996 1997 1998 Agricultural 100.0 100.0 100.0 100.0 <35 hours/week 56.7 58.6 56.7 58.5 35-45 hours/week 23.7 23.8 23.5 23.3 >45 hours/week 19.5 17.6 19.8 18.3 Nonagricultural 100.0 100.0 100.0 100.0 <35 hours/week 24.3 23.4 22.6 24.6 35-45 hours/week 25.6 26.1 25.3 25.7 >45 hours/week 50.1 50.5 52.1 49.6 Manufacturing 100.0 100.0 100.0 100.0 <35 hours/week 24.3 24.9 23.4 26.0 35-45 hours/week 26.6 26.1 25.2 25.2 >45 hours/week 49.1 49.0 51.3 48.8 Trade 100.0 100.0 100.0 100.0 <35 hours/week 27.2 26.5 26.4 27.9 35-45 hours/week 21.7 21.6 20.9 20.8 >45 hours/week 51.2 51.8 52.7 51.3 Services 100.0 100.0 100.0 100.0 <35 hours/week 28.8 26.1 25.5 26.8 35-45 hours/week 33.6 35.5 35.2 35.4 >45 hours/week 37.6 38.3 39.2 37.9 Note: Data are for the general population ages 10 and older. Data for the 10-14 age group are included for 1994-97. Source: National Labor Force Surveys. The construction sector suffered job losses of nearly 0.7 million primar- ily male workers, or nearly a third of all job losses. As with manufacturing, three-quarters of these job losses occurred in rural areas. The other most seriously affected sectors-mining and trade and services-experienced nearly 0.2 million job losses each. Women accounted for two-thirds of all job losses in trade, while men accounted for virtually all job losses in the service sector. In urban areas some displaced workers from the manufac- turing and construction sectors found refuge in the informal trade and ser- vice sectors, thus increasing urban employment in these sectors by about 0.2 million. However, this increase was insufficient to compensate for the 0.4 million jobs shed by the rural trade sector. LABOR MARKET CHALLENGES AND POLICIES IN INDONESIA 53 TABLE 2.5 Job Losses and Absorption in Agriculture, 1997-98 Changes in employment By location By gender By status millions % Urban Rural Male Female Wage Other All employment +2.267 ... 0.951 1.316 0.894 1.373 -1.684 3.951 Distribution (%) 100.0 . . . 42 58 39 61 -74 174 (percentage) Agricultural +4.625 100 23 77 57 43 9 91 Nonagricultural -2.358 100 4 96 73 27 88 12 Mining -0.201 9 (7) 107 61 39 36 64 Manufacturing -1.075 46 29 71 58 42 66 34 Utilities -0.085 4 91 9 97 3 100 0 Construction -0.663 28 27 73 98 2 109 (9) Trade and hotels -0.139 6 (166) 266 37 63 201 (101) Transportation 0.028 -1 207 (107) 40 60 (284) 384 Finance -0.039 2 38 62 90 10 95 5 Other services -0.184 8 (104) 204 97 3 51 49 ... not applicable Note: Figures in parentheses denote job gains in nonagricultural sectors. Source: National Labor Force Surveys. Agricultural employment increased by nearly 5 million (3 million men and 2 million women), or 6 percent of the total workforce, and was mostly supply driven. Workers from other sectors who had been displaced by the crisis accounted for nearly half this growth. The other half consisted of new entrants to the labor force, primarily young workers and women seek- ing work to cope with the crisis. Rural households accounted for most (nearly 80 percent) of the new employment in agriculture. Men accounted for nearly 60 percent of the new agricultural employment, which was pre- dominandy nonwage (self-employment and unpaid family work). The increase in agricultural employment was not evenly distributed across the country. The large province of Java and the provinces domi- nated by smallholder agriculture (Maluku, Southern Sumatra, and South Sulawesi) expanded their work force by 1.0-1.6 percent. But the provinces of Aceh, Central Kalimantan, North Sumatra, Riau, and West Kaliman- tan experienced a contraction in employment. So did the commercial and tourist centers of Bali, Jakarta, North Sulawesi, and Yogyakarta, where the agricultural sectors were (and remain) too small to absorb many workers. 54 EAST ASIAN LABOR MARKETS Changes in the Employment Structure The crisis produced shifts in the country's employment structure. The share of agriculture in overall employment declined between 1996 and 1997, falling from 44 to 41 percent. With the onset of the crisis, however, agri- cultural employment began climbing again, reaching 45 percent in 1998 (table 2.6). The share of all nonagricultural sectors, which had been in- creasing, began declining correspondingly. Manufacturing employment, which had reached 13 percent of total employment in 1997, dropped back 2 percent, and employment in trade and services fell by 1 percent each (to 19 and 14 percent, respectively). The gender ratios did not change, how- TABLE 2.6 Structural Changes in Employment, 1994-98 Workers Distribution 1994 1996 1997 1998 1994 1996 1997 1998 (millions) (percent) Total employment 80.04 83.90 85.40 87.67 100.0 100.0 100.0 100.0 Agriculture 36.51 36.50 34.79 39.42 45.6 43.5 40.7 45.0 Nonagricultural 43.53 47.40 50.62 48.26 54.4 56.5 59.3 55.0 Mining 0.72 0.75 0.87 0.67 0.9 0.9 1.0 0.8 Manufacturing 10.59 10.57 11.01 9.93 13.2 12.6 12.9 11.3 Utilities 0.18 0.16 0.23 0.15 0.2 0.2 0.3 0.2 Construction 3.54 3.78 4.18 3.52 4.4 4.5 4.9 4.0 Trade and hotels 13.72 15.84 16.95 16.81 17.1 18.9 19.9 19.2 Transportation 3.36 3.94 4.12 4.15 4.2 4.7 4.8 4.7 Finance 0.62 0.69 0.66 0.62 0.8 0.8 0.8 0.7 Other services 10.79 11.67 12.58 12.39 13.5 13.9 14.7 14.1 Male 49.14 51.91 53.00 53.90 61.4 61.9 62.1 61.5 Female 30.90 31.99 32.40 33.77 38.6 38.1 37.9 38.5 Urban 24.04 27.43 29.35 30.30 30.0 32.7 34.4 34.6 Rural 56.02 56.47 56.05 57.37 70.0 67.3 65.6 65.4 Wage 26.79 28.70 30.28 28.80 33.5 34.2 35.5 32.9 Nonwage 53.25 55.20 55.13 58.87 66.5 65.8 64.5 67.1 Source: National Labor Force Survey. LABOR MARKET CHALLENGES AND POLICIES IN INDONESIA 55 ever, because while men accounted for most job losses in nonagricultural sectors, many of them moved to employment in agriculture. Similarly, the urban-rural shares of employment did not change, because most of the ad- ditional employment was located in rural areas, compensating for the job losses there. Increases in agricultural employment accounted for much of the rise in total employment in provinces that saw employment expand in 1997-98. These provinces saw the share of agriculture rise by about 10 percent, ex- cept for the larger Java provinces, where the increase was 4-6 percent. The average increase in agricultural employment for the country as a whole was 4 percent, in part because it remained flat or declined in East Timor, North Sumatra, South and Southeast Sulawesi, Yogyakarta, and West and Central Kalimantan. WAGE EMPLOYMENT. The crisis also affected wage employment. Before the crisis the share of wage employment had risen steadily, climbing from 33 to 35 percent in 1994-97, in line with the growth of manufacturing, construction, and other sectors with large numbers of employees. In 1998, however, wage employment declined by nearly 1.5 million, mainly in man- ufacturing and construction, but also in all other sectors. In contrast non- wage employment increased by nearly 4.0 million, primarily in agriculture but also in the urban informal sector, particularly trade and services. As a result, the overall share of wage employment fell back sharply to 32 per- cent in 1998 (table 2.7). Both wage and nonwage employment increased in agriculture. Wage employment rose by around 0.5 million, most of them relocated urban workers who became agricultural laborers, and was divided equally be- tween male and female workers. But nonwage employment increased even faster, mostly in the self-employment category (including hired and fam- ily laborers), so that the share of wage employment in total employment actually fell slightly, dropping from 13.4 to 13.0 percent. Wage employ- ment outside the agricultural sector fell by nearly 2.0 million. A third of these losses came from manufacturing and construction, while another tenth originated in the trade sector. Nearly two-thirds of all losses in wage employment originated in rural areas, primarily in manufacturing, con- struction, and the service sector. The service sector alone accounted for a quarter of the losses in wage employment. Even nonwage employment outside of agriculture fell (by around 0.5 million) in the first year of the crisis. In fact nonwage employment in- creased significantly only in agriculture (by nearly 4.5 million). In the ag- 56 EAST ASIAN LABOR MARKETS TABLE 2.7 Changes in Employment Status, 1994-98 Workers Distribution 1994 1996 1997 1998 1994 1996 1997 1998 (millions) (percent) All employment 82.04 85.70 87.05 89.40 100.0 100.0 100.0 100.0 Self-employment 16.60 18.35 19.98 20.63 20.2 21.4 23.0 23.1 Self-empl.a 18.92 21.29 18.05 19.79 23.1 24.8 20.7 22.1 Employer 0.78 1.20 1.48 1.53 1.0 1.4 1.7 1.7 Employee 27.06 28.95 30.49 29.00 33.0 33.8 35.0 32.4 Family workers 18.67 15.91 17.06 18.45 22.8 18.6 19.6 20.6 Nonwage 54.98 56.75 56.56 60.39 67.0 66.2 65.0 67.6 Agriculture 37.86 37.72 35.85 40.61 100.0 100.0 100.0 100.0 Self-employment 5.78 6.23 6.77 7.25 15.3 16.5 18.9 17.8 Self-empl.a 12.80 14.48 11.64 13.86 33.8 38.4 32.5 34.1 Employer 0.20 0.31 0.44 0.46 0.5 0.8 1.2 1.1 Employee 4.85 4.94 4.81 5.27 12.8 13.1 13.4 13.0 Family workers 14.23 11.75 12.18 13.77 37.6 31.2 34.0 33.9. Nonwage 33.01 32.78 31.04 35.34 87.2 86.9 86.6 87.0 Nonagricultural 44.18 47.98 51.20 48.79 100.0 100.0 100.0 100.0 Self-employment 10.82 12.12 13.21 13.38 24.5 25.3 25.8 27.4 Self-empl. a 6.12 6.81 6.40 5.93 13.9 14.2 12.5 12.2 Employer 0.59 0.88 1.04 1.06 1.3 1.8 2.0 2.2 Employee 22.21 24.01 25.68 23.74 50.3 50.0 50.1 48.7 Family workers 4.44 4.16 4.87 4.67 10.0 8.7 9.5 9.6 Nonwage 21.97 23.97 25.52 25.05 49.7 50.0 49.9 51.3 a. Self-employment assisted by family members and temporary employees. Source: National Labor Force Surveys. gregate, therefore, nonwage employment (both self-employment and un- paid family work) increased by nearly 4.0 million. It absorbed not only many of those losing wage employment but also most new entrants into the labor force. Nearly half the additions were self-employed workers as- sisted by hired labor or family members, while a fifth were unassisted self- LABOR MARKET CHALLENGES AND POLICIES IN INDONESIA 57 employed workers. The remaining third consisted of unpaid family work- ers. This last category was most important for women, accounting for nearly 60 percent of all new female employment, compared with only a quarter of all new male employment. In manufacturing the majority (60 percent) of workers who lost their jobs were employees; the rest were self-employed and unpaid family work- ers. This fact suggests important linkages between the formal and informal manufacturing sectors in Indonesia, particularly for female workers. Non- wage employment losses for women in manufacturing reached more than half (54 percent) of total losses. Likewise job losses in trade and services did not affect only those in the employee category but also self-employed work- ers assisted by hired labor and family members. Only the number of self- employed persons working on their own expanded in these two sectors. The exception was the service sector, which added female employees but not unassisted self-employed workers. Wage employment declined in almost all provinces, in line with employ- ment losses in the manufacturing, construction, and trade and service sec- tors across the country. The only significant exception was North Sumatra, where wage employment expanded a little for women in manufacturing and for both men and women in the service sectors. Definitive information on how women workers have been affected by the crisis is limited. Labor Force Survey data on open unemployment sug- gest that women have been less severely affected than men. The unem- ployment rate for male workers increased 22 percent in 1997-98, reach- ing 5.04 percent. The rate for females, however, rose less than 9 percent, although the 1998 rate was higher (6.12 percent). The degree of under- employment also increased more for men than for women. These and other such figures can be interpreted in different ways, however. One in- terpretation suggests that women worked longer hours in order to com- pensate for lost household income as men were retrenched. In urban areas female workers' real income declined sharply (by 29 percent) compared with the decline for males. In rural areas, however, the reverse was true. THE INFORMAL SECTOR. Before the crisis the informal sector played an important role in Indonesia's employment structure. But the share of in- formal employment in total employment began increasing once the crisis hit. Only indirect indicators of this trend are available, as no time-series data exist for the sector. The share of wage employment in total employ- ment is one such indicator. It increased in 1994-97, indicating a concom- itant rise in the share of formal sector employment (table 2.6). In 1998 58 EAST ASIAN LABOR MARKETS nonwage employment increased sharply, not only in agriculture but also in sectors such as trade and services, while the share of wage employment fell from its 1997 level of 35 percent to 32 percent. Self-employment statistics provide a second indirect measure of the size of the informal sector. Presumably self-employed workers (with and with- out temporary hired labor or family help) form the urban informal sector in some industries. Calculations based on this assumption for the whole- sale and retail trade sector show an increase of at least 300,000 informal sector units in 1997-98. This finding should be taken as a lower-bound estimate because of the presence in the 1997 data of self-employed work- ers ages 10-14 who cannot be harmonized with the 1998 data, which do not cover this age group Clearly, then, arriving at concrete quantitative estimates about informal sector employment is at best difficult. But the precrisis trend toward a de- cline in the informal sector's share in total employment appears to have re- versed itself afterward. Informal sector employment clearly has grown in urban areas, especially in trade, transportation, and services. MIGRATION. Migration has been an important strategy for the Indone- sian labor force in coping with the economic crisis. Reverse internal mi- gration (from urban to rural areas) is in part responsible for the increase in agricultural workers. In 1998 around 1.0 million urban workers took up agricultural work in rural areas while their families stayed in urban areas. Official and unofficial international migration has accelerated, par- ticularly to Malaysia (for men) and the Middle East (for women). A sig- nificant proportion of Indonesians-well in excess of 2 million-now work overseas. Some 70 percent of them are women. The number of contract workers migrating with official permission in- creased by 75 percent between 1998 and 1999. Contract workers also mi- grated without such permission, and this group (particularly males enter- ing Indonesia) is estimated to be several times larger than the group of documented migrants (Hugo 1999). Migrant workers did not return from Malaysia in large numbers, as had been feared, for two reasons. First, Malaysia weathered the crisis much better than had been anticipated. And second, redundant construction workers were easily absorbed into planta- tions, the informal sector, and the trade and service sectors. Real Wages and Living Standards Nominal earnings of employees increased by less than 20 percent in the first year of the crisis. Meanwhile consumer prices for workers rose by 100 LABOR MARKET CHALLENGES AND POLICIES IN INDONESIA 59 percent between August 1997 and August 1998, for a drop in real earn- ings of 40 percent for the period. Subsequent wage surveys of the manu- facturing and hotel sectors indicate that employers may have frozen pay between September 1998 and December 1998 and that this practice may be ongoing, considering the relatively low inflation rate in early to mid- 1999. Real earnings of workers were probably still 40 percent lower in mid- 1999 than they were in mid- 1997. Nominal wages in rural areas rose somewhat more than in urban areas owing to a higher-than-average in- crease in pay for agricultural laborers. Male and female workers were equally affected in the aggregate. While female workers in rural areas did not benefit from the same pay increases as their male counterparts, they did obtain higher pay raises in urban areas. According to the National Labor Force Survey, nominal wages increased by 17 percent on average in the first year of the crisis, rising from Rp. 241,000 to Rp. 282,000 (around $35) per month between August 1997 and August 1998. Nominal earnings increased by 14 percent in urban areas and by 22 percent in rural areas. Though on average male and female earnings kept pace with each other, in rural areas male earnings increased 23 percent, while female earnings rose just 19 percent. The reverse hap- pened in urban areas, where female earnings outpaced male (16 percent and 13 percent, respectively) (table 2.8). The statistical authorities do not publish official time series for real wages. An appropriate deflator reflecting increases in the prices of goods and services consumed by workers could be used to derive such a series, but such a deflator is not available. An average consumer price index is available for over 44 cities but not for rural areas. Even if a composite index were available, a consumer price index would have to be developed that reflects a basket of goods for workers earning the low daily average (around $1 per day). For the present purposes we have estimated an ap- proximate worker's consumer price index by weighing the price indices of food and other expenditure categories of the official urban consumer price index against the proportion of expenditures that urban and rural worker households devote to these categories. SECTORAL EARNINGS. Wage increases varied considerably across eco- nomic sectors. National Labor Force Survey data show that agricultural earnings increased almost twice as rapidly (more than 30 percent per year) than average earnings in other sectors (17 percent). The subsistence-level earnings that traditionally prevailed in agriculture ($0.65 per day in 1998) may in part explain the dramatic increase in earnings in this sector, since these initial earnings were only half the average earnings in sectors such as 6o EAST ASIAN LABOR MARKETS TABLE 2.8 Trends in Monthly Nominal and Real Earnings, 1994-99 Index (1997= 100) US$/month August US$/mst) 1998 1999 1998 (August) 1994 1996 1997 Aug Dec Mar Rp. o0Oa 1997 1998 Nominal Index Labor Force Survey 65 86 100 117 - - 282.3 96 35 Male 66 86 100 118 314.3 107 39 Female 62 84 100 118 215.5 73 27 Urban 68 86 100 114 328.0 115 41 Rural 65 87 100 122 227.2 28 75 Agriculture 70 90 100 131 155.3 48 19 Manufacturing 62 84 100 111 253.3 91 32 Construction 68 87 100 115 279.3 97 35 Trade and hotels 70 89 100 120 279.2 93 35 Services 65 84 100 118 333.6 113 42 Farmers' Surveyb 70 92 100 133 145 155 146.0 43 18 Quarterly Wage Survey Manufacturing 66 86 100 122 120 271.3 89 34 Mining (median) 99 97 100 154 216 700.5 182 88 Hotels 80 89 100 140 134 350.9 100 44 Real Index Labor Force Survey 83 92 100 59 - - 135.0 Male 84 92 100 59 150.4 Female 79 90 100 59 103.1 Urban 85 91 100 58 160.8 Rural 82 93 100 60 107.0 Agriculture 89 96 100 65 74.3 Manufacturing 79 90 100 56 121.2 Construction 86 92 100 58 133.6 Trade and hotels 88 95 100 60 133.6 Services 82 90 100 59 159.6 LABOR MARKET CHALLENGES AND POLICIES IN INDONESIA 6i TABLE 2.8 (continued) Index (1997 = 100) US$lmonth 1998 1999 1998 (August) 1994 1996 1997 Aug Dec Mar Rp. 000o 1997 1998 Farmers' Survey 89 98 100 65 66 68 71.9 Quarterly Wage Survey Manufacturing 85 93 100 59 57 123.0 Mining (median) 127 106 100 74 103 316.9 Hotels 102 96 100 67 64 158.7 Deflator (U+R)c 79 94 100 200 213 225 Urban 77 94 100 195 208 218 Rural 79 94 100 203 218 229 - not available. a. $1 = Rp. 2,500 in Aug 1997 and Rp. 8,000 in mid-1998. b. Weighed ploughing wage Java, 40 percent; Sumatra, 20 percent; other, 20 percent). c. Deflator: consumer price index for workers. Source: National Labor Force Surveys; Farmers' Terms of Trade Surveys; Quarterly Wage Surveys. manufacturing and construction. Rapid inflation in food prices meant that agricultural workers had to substantially increase their earnings or go without the food they and their households needed to survive. But the in- creased earnings are also in part the result of the fact that some laborers re- ceive their wages in kind, monetizing the value of food. While agricultural earnings increased, however, manufacturing and construction saw below- average increases (11 and 15 percent, respectively). Increases in service sec- tor earnings were slightly above average at 20 percent per annum for males but not for females. Other wage surveys confirm the above changes. The Farmers' Terms of Trade Survey indicates that the earnings of hired agricultural laborers rose by 33 percent during the first year of the crisis. They rose by another 22 per- cent in the subsequent nine-month period ending in March 1999. Quar- terly wage survey data show increases in gross earnings in manufacturing (22 percent), mining (54 percent), and hotel work (40 percent) between September 1997 and September 1998. Preliminary data for December 1998 indicate that earnings remained stable in manufacturing and hotels over the next three months while continuing to rise in mining. 62 EAST ASIAN LABOR MARKETS In the manufacturing sector, wage increases were modest at first but began rising in earnest after the first year of the crisis, for two reasons. Many firms began dismissing the lowest-paid workers and casual and con- tract laborers first, while keeping their better-paid regular staff. At the end of the first year, firms began to implement early retirement plans and other measures to reduce their regular work force. In both cases average reported earnings increased for the relatively well-paid staff that remained. For this reason the average figures in the wage surveys may overstate the actual in- crease in nominal wages. REGIONAL DIFFERENCES. The labor force survey data show that increases in nominal earnings were well above average in Maluku, all the Sulawesi provinces, and most of the Sumatra provinces except Aceh and North Su- matra (where they were below average). The increases were slightly above average in Bali, the Kalimantan provinces, and Nusa Tengara but remained average in most Javanese provinces. The exceptions were West Java and Yo- gyakarta, which saw increases that were well below average between August 1997 and August 1998. The provinces that did relatively well in Sumatra (Kalimantan, Maluku, and Sulawesi) have a strong base of smallholder agri- culture. The Sumatran provinces that experienced below-average wage in- creases are dominated by large-scale plantations. Java also experienced sub- stantial layoffs in manufacturing and construction, putting downward pressure on wages in these and other sectors, including agriculture. Only limited information is available on the income of self-employed and other nonwage workers, despite the fact that these workers account for over two-thirds of the work force. The monthly Farmers' Terms of Trade Survey provides some idea of the changes in the income of land- owning farmers in 10 provinces. But only the annual national socioeco- nomic survey (Susenas) shows overall changes in the living standards of households. This survey is of limited use, since the income data are never fully tabulated or published. Expenditures are tabulated and published, but without any corresponding information on the economic sector(s) generating household income. THE NUMBER OF PEOPLE IN POVERTY. The number of Indonesians who have been pushed into poverty by the economic crisis is a subject of con- troversy. Projections made in mid-1998 indicated that by the end of 1998, 48 percent of the population (around 90 million people) would fall below the poverty line. However, this estimate was based on the very high infla- tion rates prevailing at that time and on an assumption of stagnant wages LABOR MARKET CHALLENGES AND POLICIES IN INDONESIA 63 and incomes in nominal terms. Indonesia's Central Bureau of Statistics projected a similar increase in the incidence of poverty-39 percent in June 1998. But a study by Poppele and Pritchett (1999) presents a very different view, suggesting an incidence of poverty of only 14 percent in 1998 (compared with 11 percent in the precrisis period). An independent evaluation of these and other studies challenges this assertion, arguing that more than one-third of all Indonesians (41 percent in rural areas and 21 percent in urban areas) were propelled into poverty in 1998 (Islam 1999). A recent World Bank Poverty update puts the incidence of poverty in In- donesia at 19.9 percent in 1998-also a very sharp increase over the pre- crisis period. Employment Creation and Maintenance In the wake of the economic crisis, the government of Indonesia instituted a number of measures designed to address unemployment and underem- ployment. These measures create jobs directly and promote other employ- ment opportunities, including self-employment and small and medium- size enterprises (SMEs). Direct Job Creation Padat Karya, or labor-intensive job creation programs designed to allevi- ate poverty in the short term by generating employment, were in vogue in Indonesia in the 1970s and 1980s. Much of the social capital in rural In- donesia, including village roads, schools, and secondary and tertiary irriga- tion channels, were built during the 1970s under these programs. Over the years, however, the government gradually replaced Padat Karya with other poverty alleviation strategies and in 1994 discontinued the programs altogether. In the wake of the drought conditions, economic crisis, and political tur- moil, the government revived Padat Karya. Padat Karya 1 (PK1) had the sole objective of providing emergency income for displaced workers through direct job creation. It targeted primarily retrenched construction and man- ufacturing workers in urban Java, which had taken the initial brunt of the crisis. The government quickly disbursed Rp. 42 billion to support the pro- gram, which began in December 1997 and lasted four months. The next generation of Padat Karya programs, PK2, was initiated in April 1998 during a period of widespread unrest. PK2, which was much larger in scope (its 16 subprograms were implemented by different govern- 64 EAST ASIAN LABOR MARKETS ment agencies) and had multiple objectives and a variety of target groups. The budgets for the various programs ranged from Rp. 95 billion to Rp. 1.6 trillion.10 The ultimate goal of PK2 was to create a total of 226 mil- lion person-days of employment. PK2 was designed not only to provide emergency income but also to create social capital. The target beneficiaries of PK2 included the newly poor and recently retrenched workers as well as the long-term poor and unemployed. The subprograms covered all provinces and targeted specific regions, sectors, and population groups. Some were designed and imple- mented as top-down programs, while others sought to involve community groups and nongovernmental organizations (NGOs), at least in principle. PROBLEMS AND CHALLENGES. PK2 suffered from a number of problems, including inadequate coverage and significant leakage of benefits to the nonpoor. A lack of clear objectives, weak program design, insufficient com- munity involvement, and the absence of a functioning decentralized labor market information system made targeting beneficiaries difficult. The best option would have been to allow for self-selection (especially in the case of unskilled labor) by offering below-market wages. Most successful labor- intensive public works programs, like India's Maharastra Employment Guar- antee Scheme, keep wage rates low in order to reach the truly needy and keep out those with high reservation wages. PK2 suffered from two other problems: low wage bills and low female participation rates. The wage bills of a large number of projects were too small compared with expenditures for materials and other costs. Even in highly capital-intensive infrastructure projects, a large untapped potential may exist for increasing labor intensity without compromising quality. Identifying additional essential (if sometimes peripheral) activities that re- quire large quantities of labor can help projects realize this potential. The second problem-the lack of female participation-stemmed largely from the fact that women in most parts of Indonesia (except perhaps Bali) are generally not involved in construction activities. Project planners need to identify and design a range of activities that will attract women. POLICY IMPLICATIONS. The numerous problems inherent in the second stage of PK2 resulted in substantial cutbacks in the Padat Karya budget for 1999-2000 (to around 50 percent of the previous year's levels). The cut- backs are unfortunate, for the fault lay not in the programs themselves, but in the manner in which they were designed and implemented. Plan- ners did not pay adequate attention to the need for a disaggregated data- base and supporting community-level structures to ensure proper targeting LABOR MARKET CHALLENGES AND POLICIES IN INDONESIA 65 and sustainability. A centrally designed program targeting the poor and unemployed with virtually no community participation in needs assess- ment, selection, implementation, or monitoring is unlikely to succeed. The current situation had a number of important policy implications: * Substantial reverse migration, the influx of urban workers into farm- based activities, and a substantial drop in real incomes have created an urgent needfor additional income opportunities in the rural sector. The 1990s saw very little investment in rural infrastructure, for example. Ample opportunities exist for labor-intensive work programs to con- struct, renovate, and simply maintain school buildings, health centers, and public irrigation facilities. Activities such as forestry and soil con- servation offer additional possibilities. Such programs will contribute to the long-term growth of the economy. Careful targeting and sound design, implementation, and monitoring are essential to their success. * Effectively managing such programs requires a high degree of coordi- nation among line ministries at every level of government. Such co- ordination has been lacking, and the ministries have been duplicat- ing programs in some areas while providing none of the activities that are needed in others. Activities must be rationalized and policies as well as programs coordinated. Provincial and district-level admin- istrations need technical assistance as well as financial resources. * Localgovernments must make sure that they develop projects appropriate to community needs and targeted to those in need. A necessary condi- tion for ensuring fair and adequate project selection and targeting is community involvement in assessing and implementing projects and in monitoring them for greater transparency. Decentralization must begin at the top and be accompanied by a bottom-up process of in- stitutional strengthening. * The process of institutional strengthening requires that local institutions receive adequate funding and technical expertise. Introducing special quotas for vulnerable segments of the populations (such as the poor and landless, women, and minorities) can make these institutions more representative. Representatives of civil society can also be in- cluded in village-level forums-another change that will facilitate the targeting and monitoring of programs and enhance transparency. Strong village institutions are also well placed to identify surplus la- bor at local levels, and workers' organizations can serve not only as custodians of worker rights but also as important agents of change. Such organizations help identify areas needing special attention and emerging training needs. 66 EAST ASIAN LABOR MARKETS Self-Employment Generation and the Role of Microfinance In Indonesia self-employment on a very small scale-and at the bottom of the economic ladder-has long been the only alternative for displaced agricultural workers unable to enter the fast-growing industrial sector. Yet the informal sector's role in absorbing surplus labor during the country's economic transition went largely unrecognized. Instead the government systematically implemented a range of policy interventions that worked against the interests of the self-employed who operate at the lowest eco- nomic levels, including regulatory monopolies that limited production trade for a large range of agricultural commodities. Faulty pricing and fis- cal policies are virtually decimating small producers and traders in some markets, such as sugar cane. Indiscriminate taxes and levies imposed by local governments, ad hoc regulations on local trade and transport, and a range of statutory licensing requirements and restrictions on the produc- tion and marketing of products pose serious problems for small producers and businesses. The drought, the economic crisis, and historically high in- flation only made matters worse. Despite the severity of these problems, this group of workers has long faced a much more serious constraint: lack of access to working capital. Most credit instruments are not accessible to the assetless poor, since most require collateral or are geared toward those with regular incomes. Even many successful microfinance programs, such as Badan Kredit Desa (which has operated since 1898), cater to the relatively affluent segments of the rural economy. As a result the self-employed at the lower rungs of the in- come ladder have always depended heavily on informal credit sources. In the late 1990s the microfinance sector began expanding, although its share in total credit disbursed remains insignificant. The new programs provide some alternatives to very small entrepreneurs needing credit. One such program is the Bank Kredit Kecamatan, which is owned by the pro- vincial governments and has branches in subdistrict capitals. Field staff from these offices travel to the surrounding villages to transact business. The provincial governments also own the Lumburg Kredit Pedesaan, a sys- tem of semi-formal financial institutions. Bank Indonesia and the German Agency for Technical Cooperation (GTZ) sponsor the Program Hubun- gan Bank dan KSM, which operates in 10 provinces. It is designed to help low-income, self-employed workers develop skills in managing and imple- menting group lending strategies. Finally, the Pembinaan Peningkatan Pendapatan Petani-Nelayan Kecil operates in 18 provinces as a group- based microenterprise credit and development program for the poor. LABOR MARKET CHALLENGES AND POLICIES IN INDONESIA 67 All these microcredit programs carry high interest rates-far higher than the rates of some of the more formal programs, such as the highly ac- claimed Unit Desa network of Bank Rakyat Indonesia. Yet they have been largely successful, with low default rates. Two of these groups-Lumburg Kredit and the Peningkatan program-make only group loans of the Gra- meen Bank variety. The loans are small (usually less than $100), require no collateral, and have short maturities (from 3-18 months). Some loans require savings deposits. POLICY CONSIDERATIONS. Most of these programs received initial subsi- dies from the government or from donors. But with proper management they began breaking even within a few years and now provide a valuable service to low-income households and especially women, who make up most of the borrowers. Their success is in part attributable to sound de- sign that includes measures such as incentives for staff and clients. With the help of a contingent of field staff, programs such as Bank Kredit can reach clients in remote and relatively inaccessible areas. These lenders pro- vide women with the working capital they need to establish themselves (primarily in petty trading activities). The rates are high but affordable- much lower than the implicit rates charged in informal credit markets. These programs demonstrate that repayment rates can be kept high even with high interest rates. In fact such programs show that contrary to popular belief, microfinance institutions do not need subsidies to turn in strong financial performance. Subsidized credit, which is often provided for political reasons, may in fact undermine the healthy functioning of these programs. Government ownership can also undermine their financial per- formance through undue political pressures. Bank Kredit and Lumburg Kredit, for instance, face pressures to underwrite large numbers of loans during election times. Government control may do damage in other ways as well. Government regulations can seriously hamper the growth of mi- crofinance institutions. For instance, in Indonesia all microfinance institu- tions that are too small to become rural banks and abide by the 1992 Bank- ing Act must either become cooperatives or stop taking deposits. And the Badan Kredit Desa, the country's oldest source of microfinancing, was not allowed to expand because of the presence of another network, the Kope- rasi Unit Desa village cooperatives. This group has not generally followed sound financial principles, and Badan Kredit provided too much compe- tition. The Indonesian government needs to rationalize the existing pro- grams and develop a comprehensive national microfinance policy. 68 EAST ASIAN LABOR MARKETS Employment Funds Employment funds can be used to complement targeted employment generation-job creation programs of the Padat Karya type and micro- credit programs for the self-employed. In Indonesia an employment fund needs to have two components. The first would provide retraining and re- deployment assistance for retrenched workers, supplementing locally im- plemented programs that target hard-hit regions and sectors. Resources for this component would come from proportional contributions from regions (under the planned decentralized financial system) and govern- ment allocations. Workers' organizations can be included in the manage- ment of these funds to improve targeting and transparency. The fund would focus on four areas: * Targeting. The fund would operate in those geographic areas with the most severe retrenchment and unemployment problems, the highest poverty rates, and the most severely hit sectors and industries. * Design. Program design must reflect the needs of the community and provide for effective links to local industry. * Female participation. The fund needs to incorporate measures spe- cifically targeted at increasing female participation-for instance in- creasing awareness of women's needs and, where necessary, taking proactive steps such as reserving a proportion of training spots for women. The fund could also promote work opportunities for women in nontraditional fields and new occupations that emerge as the coun- try develops. * Management. Procedures for selecting organizations to manage the programs would have to be carefully designed, and ongoing support would need to be provided to the agencies administering the pro- grams. Competitive bids could be used to choose the best candidate among several suitable agencies. The second component of the fund would be a microcredit program to foster self-employment among the poor, especially in regions hit hard by the crisis. Policymakers would need to draw up a sensible microcredit pol- icy to help bolster self-employment for the poor, as the existing schemes are scattered and ad hoc and need to be rationalized. Such a policy must be developed in tandem with a clear policy on subsidized rural credit schemes, in part since subsidized credit is not necessarily the best way of making credit available to the poor. LABOR MARKET CHALLENGES AND POLICIES IN INDONESIA 69 A self-employment program of this type would be able to target poor women effectively. A number of small women's organizations cater to the needs of self-employed poor women. These organizations could be pro- moted and encouraged to participate in public forums (such as village coun- cil meetings) in order to bring forward their priorities and requirements. Small and Medium-Size Enterprises In 1994 SMEs, including household enterprises, made up 99.8 percent of all businesses and accounted for 39 percent of the country's GDP. The total number of SMEs grew from 34.2 million in 1994 (33.5 million of them household-based enterprises) to an estimated 39 million in 1999. Outside of agriculture SMEs are concentrated in food, beverages and to- bacco, textiles and garments, wood and wood products, and nonmetallic mineral production. An important feature of the Indonesia's SME sector is the existence of cooperatives, or loosely knit groups of SMEs that share some common interest. Around 52,000 of these cooperatives function throughout the country. Two opposing forces are at work in Indonesia regarding policy for SMEs. Donors such as the World Bank, Asian Development Bank (ADB), and In- ternational Monetary Fund (IMF), as well as many government officials, support an approach that offers technical assistance to microentrepreneurs. Proponents of the People's Economy Movement, which emphasizes the need to move away from a conglomerate-driven economy to a more diver- sified one, see the cooperatives as the main vehicle for redistributing assets and economic control to the people. Proponents are using at least 17 dif- ferent credit schemes in an attempt to pump $1.5 billion of subsidized credit into cooperatives. Early in March 1999 the government set up a state financing company for SMEs with an initial budget of Rp. 1.5 trillion. This politically motivated agenda may not result in effective long-term economic or social policy and could worsen short-term economic distor- tions. Pumping credit into cooperatives has caused bank deposits with these institutions to rise rapidly. The cooperatives have been further urged to borrow at subsidized rates of 16-18 percent but can readily deposit their money at more than 30 percent interest. Under these circumstances cooperatives' bank deposits will continue to rise with little real increase in economic activity. The People's Economy is a political reality and must be considered in designing any broad policy initiative. Accepting and articulating that two different agendas are driving policy formation can have specific benefits. 70 EAST ASIAN LABOR MARKETS First, policymakers can identify any policies that are best phased out when the political climate stabilizes. Second, acknowledging the situation can help restore confidence in the economy. Investors, both domestic and for- eign, are more likely to accept visible economic distortions if the govern- ment states clearly that these distortions are the result of short-term polit- ical or social considerations and not part of a long-term economic strategy. Policymakers themselves need to be able to distinguish between short- term programs that are inspired largely by political considerations and long- term programs that will produce sustainable economic growth. These latter programs are more likely to produce genuine social equity. And common policy initiatives can be developed in areas where the two types of programs overlap. The policy debate aside, Indonesia has a plethora of programs designed to support SMEs.11 The technical programs focus on marketing, inputs, manufacturing and processing, the regulatory environment, and linkages to other firms. They are designed to provide two types of assistance: ex- tension services and partnerships. Extension services have been provided in a variety of forms, including the World Bank's Dapati scheme, the clus- ter development program supported by the United Nations, the Indone- sian Business Advisory Network (supported in part by GTZ), and a num- ber of government programs. Several government ministries and agencies have put forward mandatory partnerships using a range of different laws and regulations. These partnerships are based on the premise that SMEs are a weak economic group and require government support. Government policies and programs frequently overlap or duplicate each other. Duplication wastes scarce resources and leads to interagency ineffi- ciencies that are particularly onerous when they take the form of fixed costs. SMEs generally pay a large proportion of their income in fixed reg- ulatory costs; in fact the burden of taxation and licensing requirements in all forms falls most heavily on these enterprises. And over the last 30 years a complex legal and regulatory system has developed that increases the burden on SMEs. Indonesia's overlapping laws, rules, regulations, licenses, and fees seriously constrain SME development. Some of these schemes (like the World Bank's) are relatively successful, but many (including most of the government's major programs) are not. ACCESS TO CREDIT. Since the onset of the economic crisis, lending to SMEs has decreased in both absolute and relative terms. This decline has generated some concern among policymakers and international agencies seeking to assist Indonesia. For example various IMF programs have em- phasized the need to ensure that adequate amounts of credit continue to LABOR MARKET CHALLENGES AND POLICIES IN INDONESIA 7I flow to SMEs. The lack of in-depth analyses of existing programs makes assessments of this issue particularly difficult. But the general picture of SMEs that is emerging shows that these businesses are reducing their debt burden in the face of high interest rates and increasing savings in response to high deposit interest rates. Obviously many SMEs have been hurt by the increase in interest rates and the consequent high costs of credit. And many are undoubtedly not in a position to take advantage of interest earn- ings through deposits. But in the aggregate it appears that many SMEs are making rational economic adjustments to the prevailing commercial cli- mate and will continue doing so as long as they enjoy access to a rational financial system that permits it. At least 24 microfinance and SME credit programs currently operate in Indonesia. These programs differ widely in orientation, methods of opera- tion, and structure. Some are directed primarily at special groups, such as farmers or transmigrants. The range of major commercial and industrial SME programs includes lending by commercial banks that are required to de- vote at least a fixed proportion of their lending to small borrowers, including SMEs. It also includes subsidized credit targeted to cooperatives and their members, specialized rural banks, and donor-supported SME programs. Three types of financing are available for SMEs: formal commercial, semi-formal and informal, and directed (table 2.9). The formal institutions epitomize what might be described as the ruling paradigm in small-scale financing and perhaps in financial intermediation in general in Indonesia. They are depositor based and self-sustaining, and they move considerable volumes of credit. A large share of total private finance, perhaps as much as a quarter of the total loans in the banking sector in 1997, flowed to SMEs (approximately Rp 83.3). Undoubtedly some of this credit (perhaps a substantial share) would have been extended to SMEs without any spe- cial programs. However, it seems clear that the regulation requiring that 22.5-25 percent of commercial banks' loans be directed to small borrow- ers has had an influence on the pattern of credit. Semi-formal and informal credit sources include private moneylenders, rotating savings accounts, credit associations, suppliers, and pawn shops. While most of these financing sources are difficult to examine, anecdotal evidence from people familiar with the market indicates that these arrange- ments provide significant services. The existence of this robust informal market raises two important points. The first is that these sources may well be essential to the survival of SMEs lacking access to other sources of credit. The second is that these sources of finance are themselves thriving small commercial enterprises. Their role in meeting the credit needs of SMEs has been largely overlooked and merits further study. 72 EAST ASIAN LABOR MARKETS TABLE 2.9 Microcredit Providers Types of Number of Total amount Number of Average institution branches of loans borrowers loan size Savings BRI Unit Desa 3,701 Rp 511 2.5 million Rp 500,000 Rp 16 billion in (14.6%) trillion new lending Rp 500,000- RP 32.3 trillion 1 million cumulative (30.3%) Rp 1-2 million (31.8%) BPR 2,056 Rp 2 trillion 2 million Rp 1 million Savings and loan cooperatives 37,595 Rp 4 trillion Rp 500,000 Rp 1.3 trillion Commercial banks 221 Rp 45 trillion Maximum Rp in small loans 350 million Source: Bishnu P. Shreshta, Rural and Micro Finance in Indonesia, Jakarta: APRACA Consulting Services, April 1999 p. 73, 75. Kebijaksanaan Pemerintah Dalam Mengem- bangkan Usaha Simpan Pinjam, Ir. Deswamdhy Agusman Jakarta, 5 May, 1999. STRATEGIES FOR PROMOTING THE GROWTH OF SMEs. Promoting SMEs requires more than targeted programs. It requires emphasizing macroeco- nomic and sector-specific policy reforms that affect the trade regime and the fiscal, monetary, and regulatory frameworks. Programs that favor SMEs must not be used to compensate for policy distortions; instead the policies causing the distortions need to be reformed. This approach is in keeping with increased reliance on market mechanisms. But it is also in line with the principal of putting in place complementary sets of public policy ini- tiatives to improve the functioning of these mechanisms while also trying to compensate for any market failures. A second strategy involves establishing appropriate institutional pack- ages that focus on four areas essential to the growth of SMEs. Designing and implementing such packages requires: Developing policy, legal, and regulatory environments in each sector that are conducive to competition and trade; LABOR MARKET CHALLENGES AND POLICIES IN INDONESIA 73 * Improving access to credit and financial services; * Providing access to general, sectoral, and locality-specific business ser- vices; and * Establishing innovative links between urban and rural markets, as well as relevant intersectoral linkages and backward and forward pro- duction and trade linkages, in order to stimulate demand. The pro- cess should include large and small domestic firms as well as interna- tional firms. In implementing these approaches, policymakers need to differentiate among emerging, recovering, transforming, and shrinking sectors and subsectors. SMEs themselves need to become more proactive and to engage in ac- tivities that will help them grow and develop. For instance, they need to prepare for increased competition under a free-trade regime. SMEs need to increase their efficiency by introducing modern management styles, im- proved technology, and best organizational practices. They need to coop- erate with each other to overcome their size constraints and improve col- lective efficiency. And they must integrate themselves in supplier and technological networks, both domestic and international. Large companies can play an important role in this process by rationaliz- ing their production systems and investing in more efficient supplier net- works. A strong business consulting sector is needed to support adjustment and restructuring processes within and among SMEs and between SMEs and large enterprises, as SMEs do not generally have easy, reasonably priced ac- cess to professional business advice. The government needs to offer initiatives to facilitate networking among SMEs, create a new balance between public and private service providers, and move from a supply-oriented to a demand- oriented approach to SME promotion in fields such as management advice, human resources development, technology upgrading, and marketing. Initiatives such as those included in the Asia Foundation's SME pro- gram have been particularly helpful in bringing entrepreneurs together and in strengthening links with government officials and NGOs, both lo- cally and nationally. Regional fora have brought together SMEs at the pro- vincial level, raising awareness of the role these enterprises play in the economy and winning them a hearing at all stages of the policymaking process. This kind of work can benefit from broad-based joint efforts that include donors and government and that are widely replicated. IMPROVING ACCESS TO CREDIT. A recent study examines changes in the availability and terms of credit for SMEs during the economic crisis and 74 EAST ASIAN LABOR MARKETS assesses possible short-term policy responses (Musa 1998). The study con- cludes that subsidized credit is not the solution to the problems of SMEs and that these enterprises are more concerned with business opportunities, access to information on markets, and a nondiscriminatory business envi- ronment. More than access to credit, SMEs need lower transaction costs, freedom from unnecessary bureaucratic interventions, simplified licensing procedures, and fewer taxes at the regional level. Based on this and simi- lar studies (see, in particular, Wieland 1998), we have identified a num- ber of policy initiatives that will improve the flow of credit to SMEs. First and most important is an environment conducive to the efficient devel- opment and operation of the commercial financial institutions that serve SMEs. Encouraging commercial lending is clearly the most cost-effective means of ensuring credit with minimal intermediation costs. Reducing the costs of commercial lending will increase the volumes of credit that flow through commercial institutions. To reduce these costs the government needs to encourage efficiency-enhancing practices within lending institu- tions and to improve the regulatory and policy environment for the finan- cial sector. Second, the government can continue requiring banks to set aside a certain proportion of loans for small borrowers. In general this approach increases the volume of credit flowing to SMEs, though it can also have negative consequences. Regulations mandating such set-asides can act as a tax on the financial system by reducing total loan volumes. Further, the regulations require banks to engage in an area of business with which they may have little or no experience. Some banks do well in these markets, while others do not. Third, the government needs to consider that private credit is more ap- propriate than subsidized credit when it comes to SME financing and to foster the growth of private credit programs for SMEs. Subsidized credit carries high fiscal costs and can impose significant administrative costs on participating institutions. These additional costs undermine repayment discipline and create unfair competition for commercial lending opera- tions. Several of the subsidized credit programs have demonstrated poor re- payment performance over the years. When the government forgives these bad debts, borrowers are left with a transfer of wealth and the perception that it is not necessary to repay loans. The result could well be a "culture of nonperformance" that further undermines repayment performance, jeop- ardizes commercially sustainable SME operations, and offers rent-seeking opportunities for target clienteles. Given that successful SME lending programs already operate in Indone- sia, the focus of any efforts to expand access to credit should be replicating the most successful aspects of those programs. Lending institutions wish- LABOR MARKET CHALLENGES AND POLICIES IN INDONESIA 75 ing to develop an SME portfolio can adopt innovations that reduce trans- action costs, improve repayment performance, and increase operational ef- ficiency. Moreover, given the difference of opinion among policymakers on how best to support improved access to financing for SMEs, standards for evaluating SME financing operations must be agreed to quickly. Standards for SME financing programs will need to address levels of outreach, cost- effectiveness, and the distribution of benefits. Technical assistance needs to focus on identifying the expected performance of SME financial operations in line with these three factors. Labor Market Services and Regulations Indonesia's Public Employment Service is within the Ministry of Man- power, under the Directorate General of Manpower Development and Placement. The service focuses on four areas, each with its own directorate: labor market information and support services, self-employment and job ex- pansion, domestic employment, and overseas employment. Regional and district offices provide outreach and collect information. The Employment Service has registered more than 1 million workers in 27 provinces each year since 1989. In 1996 and 1997 the number increased sharply, jumping from 1,198,281 in 1995 to 1,497,159 and 1,542,522 re- spectively. The number of job vacancies registered in these provinces in- creased from 227,539 in 1989 to 629,464 in 1996 but fell to 593,153 in 1997. Overall placement rates (the ratio of registered placements to regis- tered job vacancies) were 83.90 percent in 1996 and 83.07 percent in 1997. The highest ratios are found in the electricity, gas, and water sector and the mining sector. Labor Market Information The Department of Manpower is not only one of the primary sources of labor market information in Indonesia but also one of the major users of this information. The department is responsible for regularly collecting representative data on manpower and employment throughout the coun- try. Three of the directorates are particularly important suppliers of such data: * The Directorate of Overseas Employment Placement maintains and an- alyzes records of legal overseas workers. * The overseas directorates, where Indonesian workers register before they are placed by private employment agencies, collect data on the arrival and departure of overseas workers 76 EAST ASIAN LABOR MARKETS * The Directorate of Labor Market Information and Support Services to Placement maintains and analyses administrative records on job seek- ers, job vacancies, and placements. Employment service personnel in regional and district offices collect and process the data. * The Manpower Planning and Information Section collects aggregate labor market information. This agency is in charge of annual publi- cations on the manpower and employment situation in Indonesia. The dissemination of timely, accurate labor market information is hindered by several constraints. First, Employment Service officers do very little job canvassing, perhaps because they lack transportation. Second, labor mar- ket information received from the regional and district offices arrives late and is not sufficiently disaggregated by job, gender, skills, or age. The of- fices also do not supply information on layoffs; rather, it is collected by an- other agency that does not communicate it to the appropriate offices. In addition, job seekers are not required to register with local Employment Service offices. Those workers who register voluntarily are mainly from the formal sector. Further, decisions on where to locate district offices are based on the cost of land, rental prices, and the location of local authori- ties rather than on job seekers' needs and the location of enterprises. And finally, most data is still processed by hand because of a lack of technical expertise in modern data collection and retrieval systems. Job Matching at District Offices The aggregate placement figures cannot serve as a performance indicator for the employment service in job matching at the district level. Accord- ing to the 1998 National Labor Force Survey, only 4 percent of the un- employed regarded registering with their district office as a primary job- search strategy. Computerized systems have been installed in 15 district offices (9 in Jakarta and West Java and 6 in East Java) to register job seek- ers and job vacancies. However these pilot systems are neither compre- hensive nor fully operational and cannot be considered reliable labor ex- change systems, for several reasons. They are not linked to a local area network (LAN) that would allow district offices to access all jobs within a region. In addition the official data on job seekers and job vacancies is nei- ther comprehensive nor accurate, since most job seekers and employers (especially in the informal sector) do not register with their local offices. Likewise the information on job seekers that is collected does not include details of their employment history or training. And finally only basic in- LABOR MARKET CHALLENGES AND POLICIES IN INDONESIA 77 formation on job vacancies is registered-the name and address of the company, job title, and proposed salary. No job descriptions are available. For these reasons, district-level job-matching services tend to be ineffi- cient and ineffective. The problem is compounded by the Employment Service officers, for several reasons. First, the officers lack the skills neces- sary for job canvassing, analyzing job vacancies and workers' skills and re- quirements, and registering data. In addition they have no reference guides, books, or other material that could help them. They also face major problems in transferring laid-off workers from urban areas and in- dustrial sectors to rural areas and agricultural sectors, as Indonesia lacks a coherent redeployment strategy. And finally, Employment Service officers do not hook up job seekers with labor market adjustment programs for youth or women or with social safety net programs in rural areas. Efforts to place workers are also hindered by economic factors, including: * A decrease in job vacancies, both domestically and overseas; * Difficulties in identifying job vacancies in the informal sector and in companies that recruit workers directly; * Problems shifting job seekers from the formal to the informal sector or to Social Safety Net programs; and * Difficulties applying regulations on placing disadvantaged groups, such as the disabled. In February 1999 the Employers' Association of Indonesia established the National Tripartite Coordinating Body for Empowering Manpower. The group has two objectives: to set up a database on layoffs, new job seek- ers, and vacancies abroad and to do on-line job matching. This body could be a potential competitor for the Public Employment Service. Private employment agencies, which place workers both domestically and overseas, are strictly regulated in Indonesia. In 1998 the Department of Manpower issued 204 licenses to private agencies. Of these agencies 178 worked with overseas employment, 16 with domestic employment, and 10 with seafaring jobs. Around 8 million laid-off workers have registered with the Employment Service as companies and factories in the major contracting sectors close. If the Department of Manpower receives notice of mass layoffs in time, the appropriate agencies provide labor standards advice and mediation (if the unions and firm cannot agree on compensation). If mediation fails, the case goes to the Labor Court. The "early warning" system often did not work well during the crisis. Consequently, no planning took place before mass 78 EAST ASIAN LABOR MARKETS layoffs, and retrenched and dislocated workers did not receive immediate, proactive assistance from government agencies. It was left to the initiative of individual companies to respond to the crisis. In some big multinational companies (over 3,000 employees), the human resources department issued guidelines for mass layoffs, including programs such as early retirement. No retrenchment management program was available to other firms, and no as- sistance was available to alleviate the hardship of dislocated workers. Terminating Employment 12 Several ways of terminating an employment contract are recognized in In- donesia. Termination may occur at the initiative of either the employer or employee or with the agreement of both parties. Employment may also be terminated by a worker's retirement or the expiration of a fixed-term con- tract. However Indonesian law generally does not support termination at the employer's initiative. This policy is embodied in the law governing ter- mination of employment, which stipulates that employers must do their best to prevent terminating employees. This general prohibition is given legal force in the procedures set up to regulate dismissals. The legislation requires employers to obtain government authorization before terminat- ing employees. Employers may not simply give notice, as they do in other countries, notably the United States. Employers must have authorization in all but a very few cases. These include terminations made during probationary periods, which last for a maximum of three months; terminations resulting from the expiration of fixed-term contracts; resignations; and retirements that follow company regulations or collective agreements. Dismissal permits are issued in cases of "grave mistakes" such as theft, fraud, and criminal acts. Despite the general prohibition against dismissal, the law gives employ- ers some discretion in terminating workers. When an employer believes that termination is unavoidable, the law requires that discussions take place with the workers' organization or even with the worker. Only when the discussions have been held and produced no agreement can the employer ask the Regional Committee for the Settlement of Labor Disputes for per- mission to terminate. Without an agreement the employer may dismiss the worker only after obtaining this permission. In determining whether to permit the termination, the Regional Committee considers labor mar- ket conditions and the interests of both the worker and employer. Parties to a deliberation concerning a dismissal may also apply to the Ministry of Manpower for mediation. The ministry must deal with the matter within 30 days. When mediation fails the matter is forwarded to the LABOR MARKET CHALLENGES AND POLICIES IN INDONESIA 79 regional or central committee. Employees may not be suspended or their employment affected in any way while a request for permission to termi- nate employment is pending. Employees may appeal a regional or national committee's decision to permit termination within 14 days by appealing to the Central Committee for Settling Labor Disputes. COLLECTIVE DISMISSALS AND DISMISSALS FOR ECONOMIC REASONS. In- donesian law regulates collective or mass dismissals, which are defined as 10 or more dismissals at one enterprise in one month or a series of termi- nations that demonstrate an intention to carry out a mass dismissal. The number 10 is somewhat arbitrary; the important consideration is the em- ployer's intention to carry out mass layoffs. Except where such dismissals are effected with consent, employers wishing to carry out mass layoffs must obtain prior authorization in the form of a permit from the National Committee for the Settlement of Labor Disputes. The committee consid- ers the condition of the labor market and the interests of both employers and workers. The committee also holds consultations with the affected workers. When mass dismissals occur as a result of government measures or activities, the government must try to move the affected workers to other areas or to public works projects. SEVERANCE PAY. Regional and national committees have the discretion to award severance pay and other compensation to workers after granting an employer's request to terminate employment. Severance pay, which is set by ministerial regulations, is paid at a base rate of one month's pay for each year of service, up to a maximum of five months' wages. So-called service monies are also available for long-term employees (those with more than five years' tenure). Severance is payable at the base rate for company closures and changes of location but must be paid at twice the base rate for redundancies for economic reasons and dismissals for reasons other than grave mistakes. THE REGULATORY IMPACT. Assessing the actual impact of existing legis- lation and regulations on the functioning of the labor market in general and on job losses in particular is a difficult task because of the shortage of accurate data. Did the manufacturing firms that laid off some 1 million workers in the first year of the crisis actually follow the prescribed proce- dures? According to Department of Manpower records, some 1,125 enter- prises presented cases in 1998 concerning 182,442 workers. If these fig- ures are any indication, then mass layoffs took place that did not adhere to the procedures required by law. 8o EAST ASIAN LABOR MARKETS Whether alternatives to layoffs have been the subject of dialogue or con- sultations with workers and their representatives is also a matter of debate. What is known is that in early 1998, the Federation of All-Indonesia Trade Unions, which until mid-1998 was the only officially recognized trade union federation, set up a team to consult with and assist companies con- templating mass dismissals. The team's main objective was to promote al- ternatives to layoffs such as reducing the number of hours worked, elimi- nating overtime, cutting production costs, and sending workers for "home rest." The Indonesian Employers' Association also provided a range of similar advisory services to its members. The government is currently considering reforms to important compo- nents of the country's labor legislation. A degree of uncertainty will pre- vail within the regulatory framework until the reform process is complete. The Manpower Act, passed in September 1998, was suspended before it could come into force and is being revised to create a number of separate laws. The provisions on termination of employment, for instance, have been noted for possible inclusion in legislation on dispute settlement laws. But these reforms are unlikely to have any immediate impact on regula- tions governing termination of employment, as the new legislation specif- ically retains the terms of the previous legislation in this area. However, the new legislation does envisage a series of future implementing regula- tions that may affect certain provisions. Vocational Education and Training The recent economic and financial crisis in the Southeast Asian economies has brought the issue of the relationship between education and training and employment to center stage. The most difficult issue for Indonesia is identifying the precise nature of the training interventions that will pro- vide the relief the country needs as it follows the path of economic recov- ery. The process of recovery will require a degree of restructuring among as well as within various sectors. For these reasons educating and retrain- ing retrenched workers for redeployment must be regarded as both long- and short-term issues. Education and Training in Indonesia The national education and training systems in Indonesia were under re- view throughout the 1990s in an effort to enhance their efficiency, effec- tiveness, and relevance. Although the review should have enabled the gov- ernment to map out a rapid response to the challenges arising out of the LABOR MARKET CHALLENGES AND POLICIES IN INDONESIA 8I crisis, it did not. In addition the financial stringency resulting from the crisis could have a negative impact on education and training, particularly skills training programs for the formal sector. Budgetary allocations for the education sector as a proportion of GDP have in fact declined steadily since 1986-87 (ILO forthcoming). Further underinvestment in new equipment and raw materials and inadequate funding for upgrading the skills of the teaching staff will lower the quality of training. Education in Indonesia had been expanding rapidly before the crisis. Net primary enrollment reached 92 percent in 1992. In 1996-97 primary school enrollment reached 29.24 million, increasing to 29.27 million in 1998. Junior secondary school enrollment rose from 9.28 to 9.69 million during the same period. The dropout rate had also been declining, falling to less than 3 percent for primary schools and 3.6 percent for junior secondary schools in 1996-97. In 1998-99, however, enrollments declined and the dropout rate increased. The available data show that in 1998-99, the drop- out rate at the primary level rose to 5.7 percent. At the junior secondary school level it increased dramatically, reaching 11.5 percent (ILO 1999). Vocational Education The Ministry of Education and Culture provides vocational and technical education at the secondary level through the Directorate of Secondary Vo- cational Education (formal programs) and the Directorate of Mass Edu- cation (nonformal programs). In the 1996-97 school year, some 712 gov- ernment and more than 3,000 private vocational secondary schools were in operation. The contribution made by the private sector in promoting vocational education can be gauged from the fact that in 1996-97, about 83 percent of students in vocational schools were enrolled in private insti- tutions (table 2.10). The number of government schools increased to 751 in 1997-98. The two largest groups of government-run vocational senior secondary schools focus on economics (340 schools) and technical trades (154 schools). The technical schools offer formal education at the senior secondary level to students ages 16-19. The three-year courses are designed to develop broad- based technical skills in preparation for employment as skilled industrial workers and low-level technicians. In 1996-97 economics schools accounted for 48 percent of the total enrollment of senior vocational secondary schools (1.77 million students), technical schools for 41 percent, and other voca- tional institutions for approximately 11 percent. In 1994 the Ministry of Education and Culture introduced the concept of the dual education system (Pendidikan Sistim Ganda) to implement its 8Z EAST ASIAN LABOR MARKETS TABLE 2.10 Vocational and Technical Secondary School Enrollments (1996-97) Public Private Percent Percent School types Number of total Number of total Economic 245,923 28.8 606,880 71.2 Home economics 45,373 78.9 12,168 21.1 Other vocational 25,483 39.9 38,339 60.1 Social work 7,447 59.9 4,987 40.1 Handicraft industry 11,125 76.8 3,365 23.2 Traditional music 2,537 75.8 824 24.2 Music 503 87.3 73 12.7 Arts 3,012 84.7 543 15.3 Tourism 859 2.9 28,547 97.1 Technical 157,647 22 560,109 78 Agricultural 24,880 66.4 12,597 33.6 Other technical 14,646 38.8 23,116 6.2 Graphics 1,812 46.1 2,117 53.9 Marine engineering, ship building and maintenance 1,181 31.6 2,559 68.4 Chemistry 1,635 50.1 1,627 49.9 Flying 1,128 21.3 4,169 78.7 Development 8,890 100 0 0 Textile 0 0 1,224 100 Maritime 0 0 1,1420 100 Total 513,952 29.1 1,253,209 70.9 Source: Ministry of Education and Culture 1998. "link and match" policy, which is intended to connect vocational and tech- nical education and training directly to the labor market. The new dual system is modeled after the German system. So far all public and 20 per- cent of the private vocational and technical schools have participated in this program, along with some 11,000 SMEs. Skill Training Skill training in Indonesia is provided through a number of different mech- anisms. These include the formal educational system, special training cen- ters, public and private training providers, and some informal training LABOR MARKET CHALLENGES AND POLICIES IN INDONESIA 83 venues. Some 19 government departments have established and regulate 815 training programs (MOM 1997). But primary responsibility for occupational training in technical, man- agerial, and entrepreneurial fields, for skill certification, and for establish- ing skill standards lies with the Ministry of Manpower. 13 The ministry's tar- get group is primarily job seekers, who receive training at the 156 public institutions throughout the country. Other public agencies offering skill training include the Ministries of Trade, Agriculture, Health, and Tourism. The World Bank (1997b) has identified a number of weaknesses in the Ministry of Manpower's training programs. Among them are low internal and external efficiency, high unit costs, low capacity utilization, weak link- ages with employees, and uncertain labor market outcomes for graduates. Despite these weaknesses enrollment in the training centers grew in al- most every field between 1994 and 1998 (table 2.11). Enrollments did de- cline in 1998-99, mostly because of reductions in budgetary support dur- ing the economic crisis. Reduced budgetary support for skill training is not a new phenome- non. Similar reductions in 1984-87 resulted in serious underutilization of training facilities and equipment. In response the Ministry of Manpower initiated "third-party utilization" of its facilities, opening them to private training providers, employment agencies, businesses, and other govern- ment agencies. In 1996-97 nearly 31,000 workers attended third-party courses, but that number fell sharply to less than 12,000 in 1998-99 with the onset of the economic crisis. The ministry needs to revisit third-party utilization. Information on enterprise-based training is scarce. The World Bank (1997b) estimates that some 2,500 companies operate their own training centers for around 230,000 trainees. Prior to the crisis some enterprises were making significant investments in training, especially foreign firms and joint ventures producing high-value goods. Although private providers make an important contribution to the na- tional training effort, they have not been effectively integrated into the planning process for the national vocational training system-a situation that needs to be rectified immediately. The government can create an en- vironment that fosters efficiency among private training providers by keeping regulation to a minimum. At the same time, however, regulatory polices must ensure that the excesses evident in the private training sector in many countries do not develop in Indonesia. The Ministry of Manpower has been providing facilities for upgrading the skills of private trainers. However, participants do not rate these pro- 84 EAST ASIAN LABOR MARKETS TABLE 2.11 Trainees in Skill Training Programs (1994-95 to 1998-99) Fields 1994-95 1995-96 1996-97 1997-98 1998-99 Total Regular Engineering 8,321 10,176 11,176 12,968 9,008 51,649 Electricity 7,955 10,707 12,959 14,269 9,364 55,524 Automotive 8,929 10,162 10,849 12,383 8,608 50,931 Construction 6,736 7,977 9,139 9,568 6,224 39,644 Marketing 4,393 5,872 6,018 7,486 5,280 29,049 Other vocations 11,494 13,882 16,650 19,189 10,992 72,207 Agriculture 11,477 10,231 14,979 19,846 20,592 77,125 Total 59,305 69,007 81,770 95,709 70,068 375,859 Third-party training Engineering 4,812 5,123 6,726 6,037 2,848 25,546 Electricity 2,900 3,940 5,304 5,387 2,340 19,871 Automotive 5,441 8,604 9,232 5,008 4,115 32,400 Construction 1,320 942 1,182 797 505 4,746 Marketing 2,074 4,161 5,532 1,191 1,066 14,024 Other vocations 1,503 4,441 2,449 3,790 326 12,509 Agriculture 980 1,688 402 260 441 3,771 Total 19,030 28,899 30,827 22,470 11,641 112,867 grams highly. Participants want less of the theoretical training these pro- grams provide and more hands-on experience in public training centers. This suggestion merits serious consideration, especially in view of the fact that public training facilities are presently underutilized. In addition, the government needs to foster private sector participation in the national training system. The (sectoral) industry training boards can play an important role in this respect, and the government needs to facil- itate their creation. These boards can serve as catalysts in the process of de- veloping links between skill training and industries. The benefits of such linkages would accrue to both partners. Retraining Retrenched Workers The Indonesian government recognizes the importance of providing train- ing for the large number of workers retrenched because of the crisis. But the lack of data on retrenched workers, especially educational and skill pro- files, hinders efforts to plan and organize retraining programs. To be ef- LABOR MARKET CHALLENGES AND POLICIES IN INDONESIA 85 fective retraining programs must take into account differences in individ- ual workers, in the goals of retraining, and in the systems that deliver it. Indonesia has four types of retraining programs for retrenched workers, most of them focused on developing income-generating opportunities. These programs include entrepreneurship development and self-employment, micro- and small enterprise development, community-based training (sup- ported by the International Labour Organization [ILO]), and the develop- ment of cooperative enterprises. One of the most successful retraining programs is the Adjustment Train- ing Program (Pelatihan Penyesuaian), which is conducted by the Ministry of Manpower. This program provides retrenched workers with the training necessary to secure employment in the domestic market as well as overseas. The government has also used public works to provide income support for the unemployed, including two public works programs specifically target- ing retrenched workers. In the absence of an adequate database providing information on re- trenched workers, identifying the specific training methodologies that would be most suitable for Indonesia is difficult. However experiences from other countries (both industrial and developing) suggest some potentially effec- tive approaches. These entail creating learning conditions tailored to spe- cific needs, problems, and potential. For instance modular materials can be used to create an individualized, task-based, learner-paced environment. The modules can be dosely matched with individual skill profiles to make training more effective and to minimize the time and costs involved. The methodology developed by the ILO to generate employment and income sources in rural areas may be the most useful alternative for re- trenched workers. Community-based training recognizes the importance of forging effective links between skill training and employment opportunities in rural areas where the market (or the potential market) for newly acquired skills is generally unknown. The methodology involves identifying income- generating opportunities and corresponding training needs, designing and delivering appropriate training programs, and providing follow-up support with a view to ensuring the sustainability of the activities trainees under- take. One of the key elements in the success of this methodology is access to credit. In Indonesia this issue can be appropriately addressed by linking the program with an employment fund like those described earlier. Maintaining a Skilled Workforce The Indonesian government is trying to improve the skill profile of the Indonesian workforce in the medium term in order to meet the challenges 86 EAST ASIAN LABOR MARKETS of globalization and international competitiveness. Meeting this challenge requires improving the effectiveness and efficiency of the training system. Two factors add to the urgency of making these changes. The first is the country's productivity level, which is one of the lowest in Southeast Asia (table 2.12). The second is the large number of expatriate workers in the country (table 2.13a). Indonesia needed close to 17,000 technicians and operators alone in 1996, many of whom were imported (table 2.13b). The fact that this figure represents a downward trend from 23,000 in the pre- vious year is of little comfort given the scope of the problem. The government needs to take action on several fronts to improve skill training and increase productivity. First, it must increase its efforts to train workers for those industries with the most growth potential over the next two to three years, including agroprocessing, tourism, and certain types of manufacturing (such as rattan furniture) that produce goods for export and do not rely on imported materials. Second, the government must re- view the functioning of public vocational training centers. Their focus needs to be reoriented so that they assume different roles as functions are de- volved, shed some of the training activities the private sector can provide, and offer advanced training courses. One particularly important step in- volves offering skill upgrading to increase productivity in the agricultural sector, including in agroprocessing. Third, the government needs to consider ways to improve the quality of teaching in public training institutes. Until 1996 most of the instructors (60 percent) had only a senior secondary school education; 19.8 percent TABLE 2.12 Productivity Levels of Asian-Pacific Countries (thousand US$)a Country 1990 1991 1992 1993 1994 Japan 275,689 280,796 280,972 278,570 280,351 Singapore 177,788 182,757 187,286 204,246 217,051 China 116,044 122,503 127,863 134,184 139,821 Malaysia 64,839 68,319 71,518 74,344 78,848 Republic of Korea 83,881 89,205 91,984 95,462 100,954 Thailand 20,691 22,207 22,998 25,109 - Philippines 17,190 16,757 16,154 16,145 16,352 Indonesia 15,593 16,310 17,021 18,098 19,256 - not available a. Output per worker per year. Source: Asian Productivity Organization, 1996. LABOR MARKET CHALLENGES AND POLICIES IN INDONESIA 87 TABLE 2.13(a) Expatriate Workers by Main Industry, 1995-97 Agriculture, forestry, hunting, and fishing 1,963 4,243 1,431 Mining and quarrying 6,358 6,946 2,997 Manufacturing 36,429 22,299 19,574 Electricity, gas, and water 948 930 332 Construction 2,196 311 3,086 Wholesale and retail trades, restaurants, and hotels 5,015 525 4,049 Transportation, storage, and communication 792 971 590 Financing, insurance, real estate, and business services 771 882 1,045 Community, social, and personal services 2,485 4,027 4,088 Total 56,957 41,134 37,192 had the Diploma II, and just 7.5 percent had some tertiary education (table 2.14). To compensate for these relatively low educational levels, instructors were exposed to further training. Between 1989 and 1993 some 1,994 in- structors attended training within the country, and 393 were sent overseas (MOM 1997). The government needs to continue monitoring the quality of instruction, taking remedial action when necessary. Fourth, the government can step up its efforts to provide skill certifica- tion for workers going overseas. The overseas market has always been an important source of jobs for Indonesian workers and is even more so now. Indonesia expects to export 2.8 million workers in 1999-2003. Skill certi- fication plays an important role in wage setting because it helps ensure fair working conditions for workers. Testing the skills of migrant workers and TABLE 2.13(b) Expatriate Workers by Main Occupation, 1995-97 Occupation 1995 1996 1997 Manager 13,624 12,663 8,762 Professional 11,874 11,163 12,969 Supervisor 8,254 8,281 5,409 Technician or/operator 23,407 16,551 10,052 Total 57,159 48,658 37,192 Source: Ministry of Manpower. 88 EAST ASIAN LABOR MARKETS TABLE 2.14 Formal Education of Public Vocational and Skill Training Instructors, 1996 No. Vocation SLTA DI DII DIII SI Total I Automotive 216 0 206 0 59 461 2 Mechanical/Technology 427 0 288 289 47 1,051 3 Electricity/Electronics 392 40 106 46 40 624 4 Construction 347 0 106 0 23 476 5 Business 222 20 0 0 45 287 6 Various vocations 244 20 0 0 22 286 7 Agriculture 308 20 0 0 32 360 Total 2,156 100 706 335 268 3,565 Note: SLTA, senior secondary school; DI, diploma I; DII, diploma 11; DIII, diploma III; SI, tertiary school. Source: Ministry of Manpower 1997. providing certification of their competency before they leave should there- fore be a priority. Finally, the Ministry of Manpower needs to step up its overall skill test- ing initiatives. It has made considerable efforts to establish skill standards and tests for a range of occupations. It has put in place the infrastructure required to support the system and the expertise needed to develop stan- dards, test design, and certification arrangements. But the program suffers from inadequate staffing and funding. The ministry has initiated steps to review the current system with a view to improving its efficiency and ulti- mately to devolving skill testing and certification functions to the industries themselves. This important initiative needs to be followed up vigorously. Looking to the Future Training policies have an essential role to play in promoting employment, especially during economic and financial crises. In terms of the Asian fi- nancial crisis, training and human resources development initiatives will be measured by their success in creating and sustaining the country's com- parative advantage and meeting the demands of industry for skilled work- ers, irrespective of the health of the economy. In the aftermath of the cri- sis in Indonesia, where low-skilled workers are the most deeply affected, retraining programs will also be important to ensure that retrenched work- ers become (and remain) employable. LABOR MARKET CHALLENGES AND POLICIES IN INDONESIA 89 The government of Indonesia has initiated a number of important mea- sures designed to develop human resources. It has made nine years of basic education mandatory for all Indonesians and instituted reforms of the technical and vocational education and training systems. The government must now build on these initiatives. Some short- and medium-term con- cerns include: * Establishing a structured coordinating mechanism with a clearly de- fined mandate and sufficient autonomy and authority; * Evaluating public training centers in order to reorient their focus and enhance internal and external efficiencies; * Targeting training programs to the needs of industries with the most growth potential and to the most vulnerable groups; * Increasing the involvement of private training providers in national training efforts and facilitating the establishment of industry training boards; * Improving the skill testing and certification system; * Enhancing women's access to educational opportunities and training; * Creating a viable system for financing training; and * Collaborating with employer and worker organizations on improv- ing the national training system. Many of these points were set out in a 1995 report by the Task Force on Technical and Vocational Education, which was appointed by the Min- ister of Education and Culture.14 The task force recommendations are an important attempt to look at the vocational educational system in a holis- tic manner. Two important issues were inadequately addressed, however. The first is funding-a more important issue now than in 1995. In the aftermath of the crisis, efficient utilization of available resources is one of the government's major concerns. The second issue involves linking the vocational education program (run under the school system) and the vo- cational training program for those who have left school. If any effort to improve the national vocational training system is to succeed, it must ef- fectively integrate all subsystems. The task force also recognized the need for coordination at the national level and recommended establishing a new body called the National Voca- tional Education and Training Council (NVETC). Establishing such a body would lead to a duplication of efforts, since another body with a similar mandate (the National Training Council established by the Ministry of Manpower) already exists. The task force's recommendation is reflective of 90 EAST ASIAN LABOR MARKETS the ongoing "turf wars" among the various major players in the field of tech- nical and vocational education and training. Any coordinating body will have little effect as long as the ministries fail to work together-especially since they are technically not bound by each other's administrative fiats. Such a body, then, should be set up under a presidential decree. It should be headed by a representative of the private sector and should include, among others, private employers (from both the formal and informal sec- tors), ministries involved in training, workers' organizations, and private training providers.'5 An effective labor market information system is es- sential to the smooth functioning of such a body. This information system must include an inventory of public and private training providers, with attendant details such as the location of training centers; the type and length of the training offered; intake capacities; fees payable (where applicable); and entry requirements. Arrangements must be made to disseminate in- formation, including through the employment services. The government also faces challenges to the educational system in gen- eral. At both the national and provincial levels, the educational system faces a number of issues, including access, gender equity, a shortage of fa- cilities, and quality. Some of these problems are specific to individual prov- inces. Educational quality and opportunities vary widely, indicating that some provinces are experiencing serious funding problems. The issue of access manifests itself in high dropout rates and low educational levels throughout the population, showing that the educational system (partic- ularly at the upper levels) is out of the reach of many Indonesians. The im- pact of the crisis may drive more people away from the educational sys- tem, particularly from the vocational stream, which is perceived as more expensive. Given these problems and the fact that the depressed economic situa- tion is likely continue for some time, what is the outlook for vocational and technical education? General education may weather the storm be- cause it is less expensive to implement than vocational and technical edu- cation, which require expensive inputs in the form of equipment and fa- cilities. In general, however, while the population and workforce continue to expand, the funding available for education continues to decline. This scenario raises two questions. First, how will those with low levels of edu- cation secure productive, well-paying employment? And second, how will those with low levels of education manage to send their children to school and keep them there? These are the challenges policymakers and program managers must meet. Unless the government takes steps to make the best use of all available resources, both general and vocational and technical ed- ucation are likely to continue to suffer. LABOR MARKET CHALLENGES AND POLICIES IN INDONESIA 91 The government must pay more attention to the availability of educa- tional opportunities and to the quality of the education offered. Merely expanding the school system will not suffice; enhancing access is equally important. Areas where school-age children are not in school must be iden- tified and the causes of their nonparticipation addressed if educational poli- cies are to succeed. More measures need to be initiated to increase enroll- ments at all levels, for two reasons. First, a sound educational system means a more skilled, more trainable workforce. And second, increasing the num- ber of educated workers will decrease pressure on the labor market in the short to medium term. Some Final Remarks Although Indonesia has not yet recovered from the economic crisis that has gripped its economy since the middle of 1997, growth resumed in 1999. But the economy is unlikely to attain the growth levels of the last two decades within the next few years. Current projections indicate that GDP growth for 2000 will remain at around 3 percent. The dilemma faced by local industries because of the crisis is one reason growth will remain low. The domestic market is limited by the ongoing recession, and the quality of local products is not up to the standards of competitive foreign markets. At the moment industries are surviving from year to year, using up savings rather than productive money while waiting for the results of new elections-and ultimately for economic recovery. The crisis has been a major setback for the Indonesian economy in other ways. Before the crisis Indonesia had been steadily progressing toward its goal of achieving full employment and eliminating underemployment. The crisis has seriously impeded this progress. Unless the government de- votes special attention to the problem, unemployment and underemploy- ment are likely to grow worse. Durable solutions to the problems of un- employment and underemployment can be found only in the context of sustained economic growth. For this reason policymakers must not only restart economic growth-a daunting challenge-but also get the econ- omy back on the road to full employment.16 The government must begin thinking in terms of a two-pronged inte- grated strategy of employment-led recovery and reconstruction. This strat- egy would include policies and measures designed to make the growth pro- cess more employment friendly and activities aimed at creating additional jobs through direct employment programs. An employment-friendly, effi- cient growth process should make it possible to achieve output growth with a lower rate of investment than would otherwise be required.17 And 92 EAST ASIAN LABOR MARKETS such a strategy would be consistent with Indonesia's quest to attain pre- crisis growth rates without the same investment rates. Since estimated out- put growth in the short and medium term is unlikely to generate sufficient employment for the entire labor force, the government also needs to focus on the second element of the two-pronged strategy-additional job cre- ation through direct employment programs. While macroeconomic stability is essential in order to return to a sus- tainable growth path, making that growth employment friendly will require macroeconomic trade, exchange rate, fiscal, monetary, and sectoral poli- cies conducive to such a process. Such policies will have to be carefully de- signed to have the maximum positive impact on labor-intensive sectors, subsectors, and technologies. Sectoral employment issues will have to be a priority. The government needs to examine its strategies and policies (and design new ones) for employment in the agricultural, manufacturing, and informal sector. SMEs are expected to play a major role in growth and em- ployment generation in the medium term. The government can also introduce appropriate measures to minimize job losses and facilitate the process of restructuring the economy and re- deploying workers. It can begin this process with restructuring at the en- terprise level, appropriate reforms in labor market policies and institutions, measures to strengthen employment services, and training programs aimed at redeploying workers.18 Programs are needed that equip employers and managers with the practical concepts and tools necessary to undertake so- cially sensitive and economically effective enterprise restructuring. Most enterprises can benefit from specific assistance in: * Designing and implementing appropriate measures for repositioning financing and debt, production capacity, human resources, research and development, technology, and marketing; * Refocusing overall business strategies from both a short- and medium- term perspective; X Developing strategies to meet various types of international stan- dards and norms; * Increasing competitiveness and productive efficiency; * Installing modern human resources management, corporate citizen- ship, and social responsibility policies; and * Becoming involved in the newly emerging democratic environment by advocating policies that are of particular benefit to specific industries. LABOR MARKET CHALLENGES AND POLICIES IN INDONESIA 93 Reforms in labor market institutions and legislation need to use tripar- tite consultation to achieve an appropriate balance between efficiency in the workplace and effective worker protections. Given the economic re- structuring that must take place in response to the changed realities fol- lowing the East Asian crisis, retraining workers in general and retrenched workers in particular is a priority. Retraining programs must be carefully formulated, keeping in view the changing demands for skills and differ- ences in regional requirements. Labor market information will play a cru- cial role in formulating such programs and identifying target groups of workers. Self-targeting mechanisms offer the best means of selecting work- ers for retraining. Employment services also need reforming, in part by linking the assistance they provide to retraining programs. Incorporating "rapid response" capabilities will allow employment services to provide more effective assistance to retrenched and dislocated workers, as well as to com- panies and communities. Notes 1. The data presented above are from censuses and intercensal surveys. The Labor Force Survey for 1996 shows a much lower rate of unemployment (4.89 percent of the total economically active population in 1996). But even that fig- ure represents a sharp increase in the rate of unemployment from 2.6 percent in 1986 to nearly 5 percent in 1996. 2. The Labor Force Surveys yield higher figures for the total workforce than the census and intercensal data. The reason(s) for this difference are difficult to understand. Godfrey (1993) provides an overview of the difficulties of compar- ing figures obtained from various sources. 3. Data compiled by the Central Bureau of Statistics, based on the National Socio-Economic Survey of Consumption. The poverty line is calculated based on the daily minimum nutritional requirement of 2,100 calories per capita, plus a minimum amount of nonfood items such as clothing, education, transportation, and other basic household and individual needs. 4. Based on data available in Central Bureau of Statistics, Statistical Indica- tors, December 1997, GDP growth during the first two quarters of 1997 over corresponding periods in 1996 was 6.94 and 5.95 percent respectively. 5. Dki Jakarta has the lowest rate (58.2 percent) and Bali the highest (76.8 percent). The spread and the highest and lowest rankings did not change between 1997 and 1998. 6. This increase occurred because the job losses in nonagricultural sectors (2.5 million) were offset by the increases in agricultural employment (5 million). 94 EAST ASIAN LABOR MARKETS 7. The open unemployment rate was highest in the relatively industrialized and commercialized provinces of Bali, Java, North Sumatra, and South Sulawesi. Most other provinces experienced lower increases of 0.3-1.0 percent. 8. Some 56 million persons, or 40 percent of the working-age population of 139 million, lived in urban areas in 1998, compared with 53 million, or 39 per- cent, in 1997. 9. Even before the economic crisis, the government had expressed concern about the increase in graduate unemployment in Indonesia. To address this con- cern it implemented the Young Professional Entrepreneurship Development Pro- gram to help young unemployed graduates create self-employment opportunities. 10. Some of the PK2 projects were conceived as part of the Social Safety Net program induced by the crisis. But only US$1 billion of the Social Safety Net budget of US$15.7 billion was allocated to these projects, and a large proportion of even this amount apparently suffered leakages. 11. The most prominent government-funded programs are managed by the Ministry of Industry and Trade and the Ministry of Cooperatives and Small-Scale Industry. Donor-funded programs include those supported by the World Bank, U.S. Agency for International Development (USAID), Asia Foundation, German Agency for Technical Cooperation, United Nations International Development Organization (UNIDO), and Swisscontact. 12. The description of legislation in this section is based on ILO forthcoming. 13. The ministry is also responsible for training that enhances productivity and promotes self-employment and for programs targeting overseas workers and apprentices. 14. The task force argued for technical and vocational education that is driven by industry and guided by market signals. Training should meet nationally recog- nized standards, provide multiple entry and exit points and flexible delivery, and recognize competencies wherever and whenever obtained. Importantly it should cater to the informal as well as the formal sector. And finally, it should be oriented toward (and able to facilitate) the integration of education and training and sup- ported by a decentralized system of management. 15. The functions of such a body are discussed in ILO (forthcoming). 16. The economy's observed employment elasticity stood at 0.2936 from 1985 to 1995. The government's rule of thumb regarding employment elasticity is that a 1 percent increase in GDP will generate 400,000 jobs. Using this measure Indo- nesia needs a growth rate of over 5 percent annually just to absorb the 2.2 mil- lion workers who join the labor force each year-and this statistic does not take into account the backlog of the unemployed and underemployed. The 1.2 mil- lion additional jobs that should accompany the projected growth rate of 3 percent for 2000 are clearly insufficient. 17. Such a strategy would not create a trade-off in terms of productivity and incomes. As Indonesia still faces the problems of unemployment and underem- LABOR MARKET CHALLENGES AND POLICIES IN INDONESIA 95 ployment and wages are still relatively low (and also flexible), it should continue to enjoy a comparative advantage in producing and exporting labor-intensive items. It is of course important to ensure that factor prices do not get distorted because of government policies. 18. For more details, see ILO (forthcoming). References ADB (Asian Development Bank). 1997. Asian Development Outlook 1997. Manila. . 1998. Asian Development Outlook 1998. Manila. . 1999. Country Economic Review, Indonesia. Manila. APO (Asian Productivity Organization). 1996. Productivity Statistics: Productiv- ity Indexes and Levels in APO Member Countries. Tokyo. Godfrey, Martin. 1993. Labor Market Monitoring and Employment Policy in a De- veloping Economy: A Study of Indonesia. New Delhi: International Labour Of- fice and Asian Regional Team for Employment Promotion. Hugo, G. J. 1999. "The Impact of the Crisis on International Population Move- ment in Indonesia." Paper prepared for the International Labour Organization Employment Strategy Mission. ILO (International Labour Office). 1996. World Employment 199697. Geneva. ____. 1999 (forthcoming). International Digest on Termination of Employment Laws. Geneva. . Forthcoming. Indonesia: Strategies for Employment-led Recovery and Re- construction. Report of an Employment Strategy Mission. Geneva. Islam, Iyanatul. 1999. "Poverty, Inequality and the Indonesian Crisis: What Do We Know?" Jakarta: United Nations Support Facility for Indonesian Recovery. Islam, Rizwanul. 1998. Indonesia: Economic Crisis, Adjustment, Employment and Poverty. Issues in Development Discussion Paper 23. Geneva: International Labour Office. MOEC (Ministry of Education and Culture). 1998. Education Statistics in Brief 1996-97. Center for Informatics, Board of Research and Development. Jakarta. MOM (Ministry of Manpower). 1997. Profil Balai Latihan Kedja Menyongsong Kemandirian. Directorate General for Training and Productivity Develop- ment, Jakarta. Musa, Agustina. 1998. "Access to Credit for SMEs in Indonesia Before and Dur- ing the Economic Crisis (1997-98)." Jakarta: Asia Foundation and U.S. Agency for International Development. Working paper. Poppele, J., and Lance Pritchett. 1999. "Social Impact of the Indonesian Crisis: New Data and Policy Implications." Social Monitoring and Early Response Unit, Semarang. Widarti, D. 1999. "Skills training in Indonesia." Paper prepared for the Interna- tional Labour Organization, Jakarta. Unpublished. 96 EAST ASIAN LABOR MARKETS Wieland, Robert. 1988. "SME Finance in Indonesia: An Assessment." Washing- ton, D.C.: Institute for International Economics. World Bank. 1997a. World Development Report 1997. Washington, D.C. _____. 1 997b. Training and the Labor Market in Indonesia: Productivity Gains and Employment Growth. Washington, D.C. . 1998. Education in Indonesia-fiom Crisis to Recovery. Washington, D.C. 3 Korea: Labor Market Outcomes and Policy Responses after the Crisis SOON-HIE KANG, JAEHO KEUM, DONG-HEON KIM, AND DONGGYUN SHIN THE THREE DECADES leading up to 1997 were a time of phenomenal economic growth for Korea. But the financial crisis at the end of 1997 and the subsequent structural reforms and macroeconomic stabilization programs constricted economic activities on a large scale. In 1998 real GDP growth was -5.8 percent, and inflation rose to 7.5 percent. Korea's precrisis growth sharply increased the demand for labor. The unemployment rate hovered around 2 percent from 1989 to 1997, and as 1997 approached the country faced a severe labor shortage. Analysts ar- gued that the unemployment rate had fallen too far below the nonaccel- erating wage rate of unemployment. Wage increases soon began to outstrip productivity increases, discour- aging potential investors. Until the mid-1980s Korea had enjoyed cheap labor costs compared with competing countries such as Hong Kong, Sin- gapore, and Taiwan. However, the rapid rise in wages after 1987 increased unit labor costs, and Korea could no longer count on cheap labor to give the country an edge in international competition. After the democratiza- tion of 1987, trade unions with relatively advantageous collective bargain- ing contracts were often successful in wage negotiations with employers. Nominal wages doubled within just four years. Real wages came close to doubling between 1987 and 1997, while the growth rate of nominal wages continued to climb, rising by approximately 13 percent in 1990-97. 97 98 EAST ASIAN LABOR MARKETS One profound effect of the crisis, then, was to accelerate the restructuring of the labor market to achieve labor market flexibility. Jobs-which in Korea had often been perceived as having a lifetime guarantee-are no longer pro- tected. Labor market policies are now deigned to protect overall employ- ment. This change is not without costs. The Korean labor market experi- enced record levels of unemployment along with very strong labor mobility. The official unemployment rate for the first quarter of 1999 reached 8.4 percent, with the number of unemployed exceeding 1.7 million. Most of the unemployed had been employed in the construction, trade, and manufacturing sectors. They were primarily nonprofessionals, including technical personnel and laborers (53.6 percent) and workers in services and sales positions (22.1 percent). The proportion of regular workers in the work force continued to decrease, while the proportion of temporary and daily workers rose. The crisis also increased procyclical nominal and real wage movements. The growth rate of nominal wages decreased from 12 percent in 1996 to 5 percent in 1997. The results of wage bargaining through July 1998 show that wage increases declined an average of 2.4 percent. Of 3,337 workplaces 2,259 agreed to a wage freeze, and another 559 agreed to wage cuts. In 1998, as the inflation rate rose to 7.5 percent, real wages dropped by almost 10 percent. As growth in wages fell, so did the rate of unionization. Most collective bargaining in Korea takes place at the enterprise level, as it does in Japan and the United States. Union membership had been decreasing since the late 1980s and with the crisis declined to a little over 12 percent of the total work force-a smaller percentage than in the United States, where the rate of unionization is only 14 percent. As a result of the painful economic restructuring, the Korean economy began recovering from its deep recession. Growth resumed as private con- sumption increased, and private investment continued its slow (but siz- able) recovery. Real GDP grew by 4.6 percent in the first quarter of 1999, and inflation stood at only 0.7 percent. Although the unemployment rate stood at 8.4 percent in the first quarter, this high rate was the result mainly of lags in the effect of business cycles on the labor market. The unem- ployment rate started to decrease significantly in April, and as of Septem- ber 1999 stood at 4.8 percent, with 1.07 million people unemployed. Employment and Unemployment The East Asian financial crisis resulted in massive job destruction in Korea. Total employment, which had stood at 21.1 million in the fourth quarter of 1997, was reduced to 19.8 million by the fourth quarter of 1998 and to KOREA: LABOR MARKET OUTCOMES AND POLICY RESPONSES 99 19.0 million in the first quarter of 1999. The average quarter-to-quarter employment reduction for January 1998-March 1999 amounted to 0.4 million. The reduction in total employment was close to 20 percent. This fall in total employment is attributable primarily to job losses in the manufacturing and construction industries. From the fourth quarter of 1997 until the fourth quarter of 1998, employment fell by 583,000 in manufacturing and 533,000 in construction, for a total decline of 1.1 mil- lion-more than 90 percent of total job losses. Employment in wholesale and retail also fell significantly. The community, social, and personal ser- vices industry grew by 17,000, even during the economic downturn, re- flecting the massive public works program implemented as part of the government's unemployment policy. In keeping with expectations of countercyclical behavior, employment in agriculture increased sharply for the first three quarters of 1998 but fell dramatically in the first quarter of 1999, reflecting the recent economic recovery. Across industries the hardest-hit jobs were those in manual production and clerical work. The number of managers and professional workers ac- tually increased by 2.5 percent, rising from 3,722,000 in the fourth quar- ter of 1997 to 3,815,00 in the fourth quarter of 1998. This increase sug- gests that job losses during the economic crisis in Korea were not evenly distributed across different occupations, with workers at the lower eche- lons bearing most of the burden. The downturn negatively affected the structure of the Korean labor market. The number of regular workers decreased by 733,000 between the fourth quarter of 1997 and the same quarter of 1998, accounting for ap- proximately 60 percent of the total reduction in employment during this period. Temporary workers accounted for the next largest proportion. However, the number of daily workers, which had dropped sharply in the first quarter of 1998, began increasing steadily over the year, rising from 1,933,000 in the fourth quarter of 1997 to 1,961,000 in the same quar- ter of 1998.1 The number of unpaid family workers also increased by a small amount for the same period. According to a recent World Bank survey (Hallward-Driemeier and others 1999), 60 percent of Korean manufacturing firms employed fewer workers after the crisis than they had before. This difference is the highest among the five countries under discussion (Indonesia, Korea, Malaysia, the Philippines, and Thailand). Korea and Thailand had the largest pro- portions of redundancies, with over a quarter of the firms reporting a de- cline in employment of more than 25 percent. Perhaps the single most important indicator of labor market perfor- mance is the unemployment rate. As we have seen the recent economic 100 EAST ASIAN LABOR MARKETS downturn destroyed a large number of jobs, and the unemployment rate is high by historical standards. The annual average unemployment rate for 1989-97 was 2.2 percent, with little yearly variation. The recent financial crisis and resulting cyclical and structural adjustments sharply raised the unemployment rate, which climbed to 6.8 percent in 1998 and then to 8.4 percent noted for the first quarter of 1999 (figure 3.1). Who Is Unemployed? Breaking the unemployment rate down by gender helps us to understand the recent increase in overall unemployment. There is little difference between genders in terms of the number of unemployed. But in Korea men tend to remain in the labor force when they lose their jobs, while women tend to leave the labor market. As a result the difference in male and female unem- ployment rates increased from 0.5 percentage points in 1997 to 2 percentage points in 1998, rising to 2.3 percentage points in the first quarter of 1999. Since the unemployment rate measures the percentage of the labor force that is actively looking for work, it may underestimate the actual number of job losses if some jobless workers become discouraged and stop seeking work. FIGURE 3.1 Unemployment Rates Percent 10.0 9.0 -+ total 8.0 -*- female 7.0 - 6.0 - 5.0- 4.0- 3.0- 2.0- 1.0 0.0 _. I I 1~~~~~~~~~~~I I 89 91 93 95 97 99 1/4 Source: National Statistical Office. KOREA: LABOR MARKET OUTCOMES AND POLICY RESPONSES 101 FIGURE 3.2 Labor Force Participation Rate 80.0 Percent 75.0. 70.0 + total - male 65.0 m female 60.0 - 55.0 50.0 45.0 40.0 . _l_l_l_l___l_l_l_l_l 89 91 93 95 97 99 1/4 Source: National Statistical Office. Korea's overall labor force participation rate had been increasing until 1997, reaching 62.2 percent in that year. It decreased sharply to 60.7 per- cent in 1998 and to 58.6 percent in the first quarter of 1999, largely be- cause of the huge exodus of women from the labor force (figure 3.2). In 1997-98, the number of women in the labor force decreased by 330,000, while the number of men increased by 120,000, for a reduction of 210,000 in absolute terms. If these women had not left the labor force, the unemployment rate would have been 7.8 percent instead of 6.8 percent in 1998.2 Traditionally the unemployment rate in Korea is higher for workers with a college-level education than for those with a high school education or less. However, the unemployment gap between the two groups nar- rowed over time, and the crisis reversed the situation of these two groups of workers (figure 3.3). What lies behind this shift in unemployment trends? Empirical evi- dence shows that the recent structural adjustment process is associated with significant downward mobility for less educated workers as their jobs are filled by those with more education (NSO various issues). Often these less skilled workers remain unemployed. Another explanation for the nar- I02 EAST ASIAN LABOR MARKETS FIGURE 3.3 Unemployment Rate by Educational Attainment Percent 10.0 9.0 - | high school and under 8.0 - o--0 college and over 7.0 - 6.0- 5.0- 4.0 3.0- 2.0- 1.0 0.0 - I 89 91 93 95 97 99 1/4 Source: National Statistical Office. rowing of the unemployment gap between highly educated workers and those with less education lies in the growing demand for college graduates relative to the demand for high school graduates. The unemployment rate in the first quarter of 1999 was 9.1 percent for workers with a high school education or less and 6.4 percent for those with a college-level education. According to a recent study (Shin 1999), the monthly averages for Jan- uary 1998-February 1999 show that 567,000 people joined the ranks of the unemployed, while 496,000 left the unemployment pool. Net growth in unemployment therefore amounted to 71,000 people. Among the 567,000 newly unemployed workers, 58 percent had had jobs in the pre- vious month and 42 percent had not. Among the 496,000 people who left the unemployment pool, 64 percent found jobs, and the other 36 percent left the labor force. The monthly hazard rate from the unemployment pool was 31.8 percent. To put it another way, for every 10 unemployed people in a month, roughly 1 person left the labor market, 2 found em- ployment, and the other 7 remained in the unemployment pool for the following month. This 1:2:7 relationship remained stable after the crisis, supporting the notion that labor mobility was very strong. KOREA: LABOR MARKET OUTCOMES AND POLICY RESPONSES 103 If we apply the 31.8 percent average monthly exodus rate from the un- employment pool to the 934,000 unemployed in January 1998, we find that fewer than 10,000 should remain in the unemployment pool after 12 months. However, the actual number of unemployed persons who re- mained in the unemployment pool continuously between January 1998 and January 1999 was around 40,000-almost 4 times the expected fig- ure. These calculations show that people who experienced long-term un- employment had a smaller probability of escaping from unemployment than people who were unemployed for only a short period. This fact in turn suggests that the overall duration of individual periods of unemploy- ment in Korea was relatively long. One of the most important and least visible consequences of a deep re- cession or structural change is in fact a large and expanding pool of peo- ple who have been unemployed for a long period of time. The term long- term unemployment rate (LTUR) is traditionally used to represent the proportion of people who are continuously unemployed for 12 or more months. Figure 3.4 illustrates how the LTUR changed after 1998. A sys- tematic pattern tends to emerge when the LTUR series is contrasted with FIGURE 3.4 Long-Term Unemployment Rate Percent 9.0 - 8.0 -| unemployment rate 7.0 | LTUR 6.0 - 5.0- 4.0- 3.0- 2.0- 1.0 0.0. 88 89 90 91 92 93 94 95 96 97 98 991/4 Source: National Statistical Office. 104 EAST ASIAN LABOR MARKETS the usual unemployment rate series. The LTUR tends to fall with sharp increases in unemployment, when large numbers of newly unemployed workers join the pool of jobless workers. As the newly unemployed work- ers find jobs and the overall unemployment rate decreases, the LTUR tends to rise. Korea experienced this pattern in 1993-94, after a mild re- cession. In 1993, the year when the unemployment rate reached its local peak, the LTUR fell slightly. In 1994, however, it went up by 3 percent- age points, rising from the 1993 levels of 2.5 percent to 5.5 percent. The recent-and much more severe-downturn produced a similar pattern in the LTUR. As the unemployment rate jumped by 4.4 percent- age points in 1998, the LTUR dropped by 1 percentage point. The LTUR in the first quarter of 1999 was 5.8 percent. Although the LTUR fell in 1998, the absolute number of long-term unemployed almost doubled. Traditionally, the LTUR is higher for men than for women and is posi- tively correlated with education level. These patterns still prevail in Korea. The rising unemployment rate only partly reflects the consequences of the crisis in terms of employment. The number of hours worked also de- creased significantly during the downturn. The average number of hours worked per week fell from 47.3 in 1996 to 46.7 in 1997, dropping to 45.9 in 1998. The increase in the proportion of part-time workers (those work- ing fewer than 36 hours a week) in the total labor force partly explains this decline. The proportion of part-time workers rose from 7.2 percent in 1997 to 9.3 percent in 1998. Industrial Relations When it comes to the development of Korea's industrial relations, the 1987 Declaration of Democratization is considered a turning point. The decla- ration amended the constitution to endorse the principle of freedom of as- sociation and autonomy for labor and management. The pre-Declaration period is classified as the period of employer-dominated industrial rela- tions, and the post-1987 period, which was marked by massive industrial disputes, as the period of confrontation and conflict. During this period employers and employees strenuously sought to establish a "power equilib- rium." Before 1987 the number of labor disputes averaged 200 a year. In 1987, however, the number totaled 3,749. As they went through massive labor disputes, trade unions strengthened their organizing capacities. The number of unionized workers increased from 948,000 in June 1987 to 1,270,000 at the end of the year and to 1,930,000 in 1989. The number of labor disputes peaked in 1987 at 3,749 cases. But in 1989 the government began to take strong action against illegal strikes, and KOREA: LABOR MARKET OUTCOMES AND POLICY RESPONSES 105 the number of disputes dropped remarkably, falling to 1,616. It fell again in 1992 to just 234. In 1995 there were only 88 labor disputes, and in 1996 only 85. The number of unionized workers also continued to fall, decreas- ing from 1,930,000 in 1989 to 1,610,000 in 1995. However, these devel- opments did not mean that industrial relations in general stabilized. They indicated only that the time of turbulence that followed 1987 began to sta- bilize somewhat after 1990. But labor and management still had not quite shed their age-old tendency toward confrontation and conflicts. Evidence shows that the total number of work days lost due to labor disputes in- creased from 392,581 in 1995 to 892,987 in 1996 (KLI 1998). In 1997, recognizing that an economy with inefficient labor-manage- ment relations cannot survive in an age of borderless competition, the government, labor, and management reached an agreement. Korea needed a new system of harmonious industrial relations in order to achieve sus- tainable development. Korea's labor laws, which had not been significantly changed since they were enacted in 1953, were thoroughly revised. In par- ticular, the Act Concerning the Promotion of Worker Participation and Cooperation, promulgated in March 1997, required all enterprises with 30 or more workers to organize a labor-management council. Each council would be composed of an equal number of labor and management repre- sentatives and would hold quarterly meetings. The labor-management councils discuss, among other things, ways of improving productivity and workers' welfare, preventing labor disputes, and addressing environmental issues. They have the authority to introduce and manage education and training plans and welfare facilities in the workplace. The Central Labor-Management-Government Council is the highest national tripartite body and is composed of labor, management, and public interest groups. It deliberates on labor policies and their rela- tionship to industrial, economic, and social policies. And it considers any other matters that may have a bearing on industrial peace. As the mood of industrial relations changes from one of confrontation and conflict to one of participation and cooperation, labor and manage- ment are seeking solutions through dialogue and negotiation. Nevertheless, Korea needs a dispute resolution system. In principle the Trade Union and Labor Relations Adjustment Act prescribed that whenever disputes occur that involve working conditions, labor and management must work to re- solve the dispute autonomously. The new law also narrowed the scope of and imposed stricter conditions on compulsory arbitration. Only essential public services such as water, electricity, gas, oil, telecommunications, rail- roads, hospitals, inner-city bus services, and banking services would be sub- ject to compulsory arbitration. The Labor Relations Commission could io6 EAST ASIAN LABOR MARKETS conduct compulsory arbitration if the Special Mediation Committee, which is composed of three public interest representatives, recommended it. Problems in the Labor Market The question of whether the rise in the unemployment rate was cyclical or structural is difficult to answer, as dramatic cyclical changes induce struc- tural adjustments. Nor is it simple to compare the degree of labor market flexibility among economies. However, the above statistics show that most labor market variables responded very quickly and procyclically during the recession, suggesting that the Korean labor market was not excessively rigid. Moreover, Korea made efforts to improve labor market flexibility during the adjustment process. Newly introduced measures allowed for a system of flexible work hours, dismissals for economic reasons, and a worker dispatch system.3 The primary problems facing the Korean labor market after the Asian fi- nancial crisis were wage rigidity, a worsening employment structure, long- term unemployment, youth unemployment, and family unemployment (unemployment among heads of households with no other wage earners). WAGE RIGIDITY. Although labor market variables show very procyclical movements, some are functionally very rigid. In particular individual wages do not appear to reflect productivity differentials, creating functional wage rigidity. According to the 1996 Survey on Collective Bargaining on Wages conducted for employers and workers at 262 unionized workplaces in the manufacturing sector, the most common wage determination system is based on seniority. As of 1996 the cost of the labor needed to produce one unit of output was approximately 30 percent higher in Korea than in France and Japan, 15 percent higher than in the United States, and 10 percent higher than in Taiwan (KLI 1997). This fact explains why the international com- petitiveness of the Korean labor market gradually declined, even when the economy was growing rapidly. The wage system can be rationalized, however, through the promotion of an annual salary system based on ability and per- formance rather than on seniority. Functional wage flexibility should be the primary policy goal if the labor market is to be healthy and competitive. THE EMPLOYMENT STRUCTURE. As previously discussed the proportion of well-paying jobs and "regular" employment decreased dramatically dur- ing and after the crisis. But the trend toward increasing numbers of non- regular workers also appeared in many industrial countries. The most sen- sible way to address such a phenomenon is by asking how these workers KOREA: LABOR MARKET OUTCOMES AND POLICY RESPONSES I07 can best be protected and to what extent they should be. These questions concern not only basic labor standards but publicly administered social in- surance programs. LONG-TERM UNEMPLOYMENT. The LTUR reached 3.8 percent in 1999- still low relative to the rate in industrial countries. But with Korea's social safety net in the early stages of development, concern about the long-term unemployed in Korea could be justified. By 1999 this class of unemployed workers numbered around 51,400, and these workers seemed destined to form a new underclass. YOUTH UNEMPLOYMENT. As of 1998 the highest proportion of the un- employed (approximately 36 percent) belonged to the 20-29 age group. The next-highest proportion came from the 30-39 age group. Workers ages 15-19 accounted for a smaller proportion of the total unemployed but faced an unemployment rate of 20.8 percent in 1998. (The unem- ployment rate stood at 11.4 percent for the 20-29 age group.) When the crisis began, the practice of hiring new graduates from high schools and colleges virtually halted, generating a large pool of unemployed new grad- uates who would have trouble finding employment even when the econ- omy started its recovery. The government initiated a sizable number of ready-made public jobs and internship programs for this group, but these publicly funded programs were not expected to last indefinitely. FAMILY UNEMPLOYMENT. The problem of family unemployment in Korea is becoming increasingly serious. As of 1998 some 668,000 of the 1,463,000 unemployed, or 45.7 percent, were family heads, compared with 34.5 percent in 1997. Moreover, about 19.1 percent of this group were heads of families with no other income earners. The majority were low- paid workers even before becoming unemployed. Statistics show that 60,000 were temporary workers, 90,000 daily workers, and 39,000 unpaid family workers, new entrants to the labor force, or those with work expe- rience who had been jobless for more than one year (NSO various issues). Employment Creation and Maintenance In response to rapidly rising unemployment, in March 1998 the Korean government began implementing comprehensive measures to address a number of labor market issues. These issues included job protection and creation, vocational training, job placement, and social protection (Lee and Choi 1998; Uh 1999). 1o8 EAST ASIAN LABOR MARKETS The job-protection program was intended to minimize unemployment in two ways. First, it would help keep viable profit-making firms from col- lapsing. Second, it would provide firms experiencing short-term financial difficulties with subsidies to help them retain employees. The Employ- ment Insurance System (EIS) would provide workers with financial sup- port and assistance. The job-creation program aimed to provide new jobs through public investment projects, public works projects, and support for business startups. The vocational training and job placement programs were designed to enhance the employability of the jobless and to expand the networking system for job matching. The social protection program would provide income support to unemployed workers and their families. Included among the social protection programs were unemployment ben- efits (through the EIS), public works, temporary livelihood protection, and loan projects for the unemployed. To implement these unemployment policy measures, the government spent almost 10 trillion won in 1998 (table 3.1). Public expenditures on labor market programs were almost negligible in Korea before the crisis TABLE 3.1 Public Expenditures on Unemployment Measures, 1998 Budget No. ofpersons Unemployment measures (billion won) covered Job creation 3,817 Expansion of SOC investment 3,295 Support for venture enterprises 522 Job protection 2,117 781,000 Support for efforts to retain employees 122 781,000 Support for managerial stabilization of SMEs 1,995 Vocational training and job placement 901 363,000 Social protection 3,235 1,843,000 Public works projects 1,044 438,000 Unemployment benefits 850 441,000 Loan projects for the unemployed 750 109,000 Temporary livelihood protection, etc. 591 855,000 Total 10,070 5,611 - not available. Source: Ministry of Labor, White Book on Labor, 1999. KOREA: LABOR MARKET OUTCOMES AND POLICY RESPONSES IO9 (Park and Lee 1999). For example, Korea spent 0.1 percent of GDP on active labor market policies in 1996 (Martin 1998). However, the picture changed significantly after the crisis, and total labor market-related ex- penditures expanded to 2.2 percent of GDP in 1998.4 In 1999 a total budget of 16 trillion won was allocated for unemployment policies (in- cluding a supplementary budget): 6.5 trillion won to create jobs in the pri- vate sector, 2.5 trillion won to create public works jobs for 300,000 peo- ple, 0.5 trillion won to support firms that retain employees, 1.1 trillion won to support vocational training and job placement, and 5.4 trillion won to support the social safety net (OECD 1999). Although the basic framework of measures to address unemployment had not changed since the programs were enacted, the government's pri- ority shifted from assisting the unemployed to creating jobs. Job creation was the top priority among unemployment policies in 1999. The govern- ment's efforts centered on creating a favorable environment for the private sector to create jobs (KOILAF 1999; MOL 1999). These efforts included: * Providing necessary infrastructure (such as business management consulting resources and start-up information); * Increasing the number of small business support centers and business "incubators" to help support the start-up of small and medium-size enterprises (SMEs) and venture enterprises with high potential for employment; * Supporting service industries such as tourism that have high job- creation potential; and * Increasing investment in social overhead capital and extending hous- ing construction projects. Income Support for the Poor and Unemployed As we have seen the government made a number of provisions to help sup- port the vulnerable and unemployed. These social safety nets include the EIS, public works projects, and temporary livelihood protection. In addi- tion, the government offers special loans for the unemployed and a wage claims guarantee system for workers from bankrupt firms. The Employment Insurance System5 In the early 1990s the government began discussing an unemployment insurance scheme that would be an institutional device offering compre- IIO EAST ASIAN LABOR MARKETS hensive employment-related services (KOILAF 1998). There was some concern that such a system would reduce workers' motivation to seek em- ployment and thus prolong the period of unemployment. Finally, how- ever, a social consensus was reached to adopt the EIS as part of an active and comprehensive manpower policy. The Employment Insurance Act mandating the EIS was enacted in December 1993 and took effect on July 1, 1995. The EIS was intended to serve two purposes. First, it would provide unemployed workers with unemployment benefits, and second, it would offer workers the chance to upgrade their job skills and thus enhance their chances of finding and retaining stable employment. The program com- bined traditional unemployment benefits and active labor market pro- grams such as skills training and job matching. For this reason the system was called the employment insurance system rather than the unemploy- ment insurance system. The EIS has three major components: unemployment benefits, an em- ployment stabilization scheme, and a vocational competency development scheme. Unemployment benefits aim not only to stabilize unemployed workers' living conditions but also to promote reemployment through job-seeking and employment promotion allowances. The employment stabilization scheme aims to prevent unemployment from massive layoffs and to stimulate reemployment at those times when changes in the in- dustrial structure or technology result in extensive corporate restructuring. The vocational competency development scheme seeks to develop the vo- cational skills of workers through life-long vocational training. Each of the three schemes has its own premium. The premium for un- employment benefits is 1.0 percent (0.5 percent each from employers and employees) of total payroll.6 The premium for the employment stabiliza- tion scheme, which is financed entirely by employers, is 0.3 percent of total payroll. The vocational competency development scheme is also fi- nanced by employers and carries a premium ranging from 0.1 percent to 0.7 percent of total payroll, depending on the size of the firm. Compared with other contingencies covered by social insurance pro- grams, unemployment is subject to a much greater degree of uncertainty (Yoo 1 999a). Economic fluctuations affect the level of unemployment and thus the financial situation of the employment insurance fund. The Ko- rean government maintains a reserve for the employment insurance fund to cover any unexpected demand. When the economic crisis began at the end of 1997, Korea had a sufficient amount of reserves in the fund-2 trillion won, the equivalent of almost 18 months' expenditure in an envi- KOREA: LABOR MARKET OUTCOMES AND POLICY RESPONSES III ronment of high unemployment. These reserves allowed Korea to cope with high unemployment after the crisis. UNEMPLOYMENT BENEFITS. Unemployment benefits include job-seeking and employment promotion allowances. To receive unemployment bene- fits, claimants must have been insured for 6 months or more during the 12-month working period prior to the date of termination. Workers who are voluntarily unemployed without good cause or who are terminated for cause are ineligible for unemployment benefits. Claimants must register at their local labor office as job seekers. They must be ready and able to work and actively searching for a job. The duration of unemployment benefits is set at between 60 and 210 days, depending on age and length of enroll- ment in the EIS. Unemployment benefits are paid to eligible claimants every two weeks following a two-week waiting period. Thus even workers who register im- mediately following a layoff must wait four weeks for the first payment. The amount of the benefits workers receive equals 50 percent of their average wage rate during the three months immediately preceding layoff. Because unemployment benefits are nontaxable, the level of income re- placement is often higher than it is in other countries. Unemployment benefits were initially limited to workers in firms with more than 30 employees, but coverage was extended rapidly after the fi- nancial crisis (table 3.1). Coverage was extended to workplaces with more than 10 employees in January 1998 and to workplaces with more than 5 employees in March 1998. It was further extended in October 1998 to employees in firms with fewer than 5 workers, temporary workers em- ployed at least one month, and part-time employees working more than 18 hours a week. In addition a special 60-day extension was granted to un- employed workers whose benefits expired between July 1998 and June 1999. The extension was renewed in the second half of 1999. As of July 1999 the government had granted a combined 1.5 trillion won to 718,000 people in the three years since July 1, 1996, when un- employment benefits first began (table 3.2).7 Claims for unemployment benefits had increased sharply by the end of 1997 but showed a downward trend as the economy gradually picked up. The average duration of un- employment benefits was 85 days in 1997, 91 days in 1998, and 126 days in July 1999. Over the three-year period, men accounted for 70 percent of recipients. As of August 1999 the ratio of unemployment benefit recipients to the total number of unemployed stood at 12.3 percent-considerably lower 112 EAST ASIAN LABOR MARKETS TABLE 3.2 Workers Receiving Unemployment Benefits, 1996-99 1996 1997 1998 As ofJune 1999 Males 8,017 37,960 305,280 128,192 (79.1) (74.4) (69.6) (67.6) 13,057 133,185 61,390 (25.6) (30.4) (32.4) Females 2,116 5,563 119,550 56,739 (20.9) (10.9) (27.3) (29.9) 12,463 133,808 52,920 (24.4) (30.5) (27.9) 13,942 95,664 38,354 (27.3) (21.8) (20.2) Under 30 854 16,163 79,820 35,988 (8.4) (31.7) (18.2) (19.0) 30-39 2,264 2,886 9,623 5,581 (22.3) (5.7) (2.2) (3.0) 40-49 3,335 (32.9) 50-59 3,087 (30.5) Over 60 593 (5.9) Total 10,133 51,017 438,465 189,582 (100) (100) (100) (100) Note: Numbers in parentheses are percent of total participants. Source: Ministry of Labor. than in industrial countries (table 3.3). The government aims to increase the figure to 20 percent by 2002. In addition the government expects 7.6 mil- lion workers, or 80 percent of the workers eligible for employment insurance (about 9.5 million), to have insurance coverage by that year (MOL 1999). EMPLOYMENT STABILIZATION. The employment stabilization scheme has two parts: aid for employment adjustment and aid for employment pro- motion. Aid for employment adjustment consists of employment mainte- nance and hiring subsidy programs. Hiring subsidy programs assist em- ployers who contribute to the stabilization of the labor market by hiring KOREA: LABOR MARKET OUTCOMES AND POLICY RESPONSES II3 TABLE 3.3 Unemployment Benefit Coverage (millions of employees) July January March October June 1996 1998 1998 1998 1999 30 10 5 Scope of employees employees employees All All coverage or more or more or more workplaces workplaces No. of employees to be covered (A) 4.31 5.67 6.11 8.50 8.60 No. of insured employees (B) 4.3 4.30 4.69 4.90 6.00 B/A (%) 99.8 75.8 76.8 57.6 69.8 Source: Korea Labor Institute. laid-off workers from restructuring enterprises. Employment maintenance subsidy programs offer wage subsidies to firms that retain redundant work- ers during periods of short-term financial difficulty (OECD 1999; Yoo 1999b). The programs provide subsidies of one-half to two-thirds of em- ployees' wages (depending on the size of the firm) for six months to firms that provide their employees with at least one month's paid leave, transfer workers to affiliates, or switch to a new line of business while retaining at least 60 percent of their employees.' Other subsidies cover: * Temporary shutdown. Firms that dose for at least two days a month can receive one-half to two-thirds of the temporary shutdown al- lowance they must pay employees. * Reduced working hours. Firms that reduce regular working hours by more than 10 percent can receive up to one-tenth of the wage bill from before the reduction. * Training to retain employees. Firms that provide training for their workers receive one-half to two-thirds of wages for trainees and are reimbursed for training expenses. In 1998 a total of 1,805 firms received employment maintenance sub- sidies that covered nearly 660,000 workers. Some 91 firms received hiring subsidies for employing 5,210 displaced workers (table 3.4). Contrary to initial expectations, these employment subsidy programs were used by less 114 EAST ASIAN LABOR MARKETS TABLE 3.4 Participation in Employment Subsidy Programs, 1998a Number of employees Total 5-49 50-99 100-299 300-499 500-999 1000+ All firms 199,272 179,902 11,100 6,477 911 580 302 Firms participated 1,896 755 366 445 118 103 109 Participation rate (%) 0.95 0.41 3.29 6.87 12.95 17.75 36.09 a. Employment maintenance and hiring subsidies only. Source: Korea Labor Institute. than 1 percent of private employers. A strong association exists between the size of the firm and the likelihood of utilizing the subsidy programs, suggesting that small firms did not utilize the subsidy programs because they simply did not know the programs existed or could not cope with the complex rules. The employment maintenance subsidy program was expanded in 1999, allowing firms to receive more than one type of subsidy instead of just one at a time. In addition, the duration of the subsidies was tem- porarily increased from 6 to 8 months for the first half of 1999. A recent study investigated the effects of these subsidy programs on employment maintenance, using employer surveys as well as case studies (Kim and Park 1999). Samples were drawn from 533 firms that received the subsidies be- tween January 1, 1998, and March 20, 1999. Fay (1996) posits that, theoretically, the net employment effect of a subsidy can be computed as follows: net effect = gross effect - deadweight loss - substitution effect - displacement effect. By definition the substitution effect does not occur in wage subsidy pro- grams designed to maintain employment. Therefore we can compute the net employment effect by estimating precisely the magnitude of the dead- weight loss and displacement effect-an often difficult task. Thus the study estimated the net employment effect of the employment mainte- nance subsidy using the percentage of layoffs that might have occurred among subsidized employees in the absence of the subsidy. The result was 22 percent on average, suggesting a deadweight loss in the 70 percent range. Although judging the robustness of the result is also difficult be- KOREA: LABOR MARKET OUTCOMES AND POLICY RESPONSES 115 cause the study was based solely on interviews with employers, this result roughly coincides with the experiences of OECD countries (Fay 1996). Studies of the wage subsidies in Europe in the late 1980s have found that most such programs yielded small net employment effects. According to an OECD report, the estimated net employment effect of wage subsidy programs in Australia, Belgium, Ireland, and the Netherlands was a mere 10 percent (Martin 1998). The results by type of subsidy program are as follows: paid leave sub- sidy, 29 percent; unpaid leave subsidy, 26 percent; and temporary shut- down, reduction of working hours, and training, around 20 percent each. Analyzing these results by industry, we find that in nonmanufacturing in- dustries, 28 percent of the subsidized employees would have been laid off in the absence of the subsidy program-higher than the 21 percent in the manufacturing industries. Looking at the results by firm size (number of employees), we find that firms employing fewer than 10 workers would have eliminated a third of their subsidized employees in the absence of the subsidy program. Firms employing 100 workers or more would have re- duced 16-19 percent of their subsidized employees. These results suggest that the smaller the size of the firm, the higher the employment mainte- nance effect of the subsidy program. THE VOCATIONAL COMPETENCY DEVELOPMENT SCHEME. The vocational competency development scheme under the EIS is an incentive system that promotes voluntary in-house training by providing financial support to employers and employees from the EIS fund. The scheme offers two types of financial support. Employers that implement training programs receive support for in-plant and other types of education and training and paid leave for employees enrolling in training. Employers can also apply for a loan or subsidy from the EIS to establish training facilities, purchase training equipment, or both. To promote training in SMEs, the EIS helps arrange-and provide financial support for group vocational training. Support for employees covers education and training and includes training incentives for the elderly and tuition loans. Unemployed workers can engage in any training activity regardless of their eligibility for unem- ployment benefits. This scheme is based on the notion that receiving training with the goal of becoming reemployed in a secure job is prefer- able to the passive protection of unemployment benefits. In 1997 approximately 13,888 firms and 3,422,444 individual workers were entitled to use the vocational competency development programs. Of the workers 48 percent were in the manufacturing industry and 51 ii6 EAST ASIAN LABOR MARKETS percent in the service sector (19 percent in finance, insurance, and real es- tate and 13 percent in transportation, warehousing, communication, and related industries). These figures indicate that the focus of vocational training is shifting from the manufacturing to the service sector. Only 1 percent of workers were in the agricultural sector. In 1999 EIS funds supported a total of 199,880 people. Of these 81,324 (41 percent) received support for in-plant training, 102,683 (51 percent) received support for alternative types of education and training, 5,559 (3 percent) took paid leave for education and training, 1,949 (1 percent) undertook training during a period of unemployment, and 8,365 (4 percent) received tuition loans. Total financial support for the programs was about 59 billion won. Public Works Projects Public works programs have two policy objectives: to create temporary job opportunities and to reinforce social safety nets for the unemployed (Yoo 1999a). By providing temporary job opportunities for the unemployed in the public sector, these programs guarantee the basic livelihood of unem- ployed families. In principle those who are eligible for public works proj- ects must be between 18 and 60 years old at the time of application. They must be unemployed or daily workers without regular income. Homeless individuals whose status can be verified by administrative agencies or ap- proved organizations are also eligible. Those receiving unemployment benefits are not eligible, regardless of the amount of benefits they receive. However, spouses of those who receive less than 300,000 won in unem- ployment benefits are eligible for jobs with public works projects. Public works are divided into central and local government projects. Selection criteria for workers may differ for central government projects. Screening decisions for local government projects are made based on scores in nine categories: age, status in household (head or otherwise), number of dependents, property status, gender (for household heads), handicap, duration of unemployment, public works experience (including in previous stages of the same project), and household income. The budget for public works projects amounted to 1 trillion won in 1998 and 2.5 trillion won in 1999. Public works projects are carried out in stages. The first two stages take place over three months, with a 10-day break so workers can look for jobs. Those who participate in three stages consecutively are excluded from participating in the next stage. Although designed primarily for male heads of household, the projects also attract KOREA: LABOR MARKET OUTCOMES AND POLICY RESPONSES II7 TABLE 3.5 Public Works Project Participants, 1998-99 (percent) 1998 Stage 1 Stage 2 Male 67.7 57.0 Female 32.3 43.0 Under 20 1.2 0.7 20-29 12.0 8.8 30-39 20.7 19.6 40-49 26.1 26.9 50-59 27.6 30.6 60-65 12.4 13.4 1999 Stage 1 Stage 2 Stage 3 Male 58.7 50.5 45.6 Female 41.3 49.5 54.4 Primary school or less 37.4 36.0 34.7 Middle school or dropouts 22.3 23.3 21.3 High school or dropouts 27.2 27.3 27.1 Two-year college or dropouts 5.0 4.9 6.1 University dropouts or higher 8.0 8.5 10.7 Under 20 1.0 0.9 1.3 20-29 11.6 13.5 16.0 30-39 22.7 21.4 18.8 40-49 32.4 30.4 27.2 50-59 27.4 22.5 29.0 60-65 5.0 11.2 7.8 Source: Ministry of Labor. many female workers who have been excluded from other job opportuni- ties (Park and Lee 1999) (table 3.5). In 1998 some 41 percent of the par- ticipants were female. By 1999 the female participation rate had risen to an average of 49 percent during the first three project stages. The daily wage rate for participants depends on the type and difficulty of the work. In 1998 the wage rate for stage 2 projects ranged from II8 EAST ASIAN LABOR MARKETS 22,000-35,000 won. However, the wage rate was cut by 3,000 won in Oc- tober 1998 and by a further 3,000 won in 1999. Currently the daily wage rate ranges from 19,000-29,000 won, higher than the minimum wage.10 Short-term job creation projects target special groups. For example in 1998 the government began sponsoring an internship program in re- sponse to the problem of unemployment among young people. The pro- gram was originally introduced to provide new college graduates with short-term jobs and was extended to unemployed high school graduates in 1999. The internship program is a kind of wage subsidy program. Par- ticipating firms are offered 500,000 won per month for 6 months if they hire new college (high school) graduates who cannot find a job. The in- ternship program, which in 1999 had a budget of 209 billion won, bene- fited 58,000 unemployed college graduates and 10,000 unemployed high school graduates. The public works program was often criticized. Some public works projects are considered to be unproductive or even wasteful (Yoo 1999a). Early in the program some well-off people were able to participate and many poor unemployed people were excluded, largely because local gov- ernments often selected participants on a first-come-first-served basis re- gardless of the selection criteria. The high wage rate for public works was also criticized as having the potential to distort the labor market. In re- sponse the government cut the wage rate by 6,000 won. Temporary Livelihood Protection The Livelihood Protection Program, created by the Livelihood Protection Act in 1961, provides support to those who are unable to work, including the handicapped, the elderly, and children. It is a means-tested program designed not for the unemployed but for the poor in general (Yoo 1999a). In 1998 the program protected 1.16 million people-approximately 2.5 percent of the total population. If no one in a poor household is able to work, the government subsidizes living, educational, medical, maternity, and funeral expenses. If anyone in the household is able to work, the gov- ernment subsidizes educational, medical, maternity, and funeral expenses and supplies free vocational training and education but not living ex- penses. Instead, the government offers a long-term loan with a low inter- est rate to help set up and operate a business. If an unemployed person is unable to work, the program supplies direct income support. The government expanded the program and introduced the Temporary Livelihood Protection Program for the unemployed in March 1998 in KOREA: LABOR MARKET OUTCOMES AND POLICY RESPONSES II9 order to protect poor unemployed people who do not qualify for unem- ployment benefits. The original program had stiff eligibility criteria: in- comes of less than 230,000 won monthly and property valued at less than 29 million won. Some of the strict eligibility criteria were later eased to allow participants to own property worth up to 44 million won. This program offers long-term, low-interest loans for livelihood ex- penses, housing costs, and business operations and supplies medical care and education. If an unemployed person is unable to work, the program offers direct income support. The program protected 310,000 unemployed low-income people in 1998 and 760,000 in 1999. However, it is still too limited in its coverage and generosity. For example, a household of four members can receive 250,000 won per month under the program. But in 1999 the minimum living expenses for such a household were estimated at 880,000 won. The government plans to reform the social safety infrastructure with the Act on Ensuring People's Basic Living Standards (MOL 1999). With this reform the government aims to ensure that people in the lowest in- come brackets have the basic necessities, including clothing, food, medical services, and education. The act will take effect in October 2000, and the Temporary Livelihood Protection Program, together with the Livelihood Protection Program, will be integrated into the Program for Ensuring Peo- ple's Basic Living Standards. Other Social Safety Nets An unemployed person needing direct income support or money for ur- gent medical, wedding, or funeral expenses, children's school fees, housing costs, or business-related expenses may apply for a long-term loan. This special loan project was introduced by the tripartite agreement of Febru- ary 1998. In order to get the loan, the applicant must register at the local labor office as an unemployed person prior to applying for the loan and must be actively seeking work. The applicant must also satisfy certain cri- teria in terms of income, housing, and number of dependents. The amount of a loan ranges from 3 million won for livelihood maintenance, urgent medical, wedding, or funeral expenses, or school tuition to 30 mil- lion won for operating a business. As of June 1999, 596 billion won remained of the original 1.85 trillion won in the loan fund. The original amount was funded with the sale of employment security bonds. The loan project was expected to continue until the fund was completely paid out (MOL 1999). 120 EAST ASIAN LABOR MARKETS The government operates other loan funds. Some of these programs have highly beneficial effects, especially the loans to jobless female house- hold heads and funding for the long-term unemployed to start businesses. These programs will be considered for permanent implementation. A large number of firms went bankrupt after the crisis (Park 1999).11 Many of the bankrupt firms could not make severance payments or even pay wages to their employees. The government established the Wage Claims Guarantee System for such employees in July 1998. Under this system these workers receive wage arrears for up to three months and sev- erance payment equal to their wages for the last three years at the bank- rupt business. In 1998 the system was financed by a levy on firms of 0.2 percent of the total payroll and a budget outlay of 190 billion won (OECD 1999). The levy was raised to 0.3 percent in 1999. As of June 1999, 39.5 billion won had been paid to 12,210 workers at 278 work- places. About 72 percent of the recipients were male. The Financial Crisis and Labor Market Services The impact of the crisis on workers was particularly severe because Korea had only minimal social safety nets. Families broke apart; violent and petty crime alike increased sharply; and many small children were left at orphanages and other institutions. As a result the government undertook a number of programs designed to mitigate soaring unemployment. In order to minimize job losses, the Korean government encouraged the private sector to adopt more flexible management practices, allowing part- time employment, work-sharing, and flexible work hours.12 Wage flexibil- ity was also an important part of the government's response to the crisis. Employers and workers are encouraged to agree to a performance-based wage system and to consider pay cuts instead of retrenchment. Dismissals Responding to criticism that labor hoarding within firms was one of the causes of the financial crisis, the Central Labor-Management-Government Council initiated an amendment to the Labor Standards Act. The amend- ment gives firms the legal right to dismiss workers in what are termed "managerially urgent" cases in order to lower labor costs and build com- petitiveness. Previously firms had been unable to fire workers without run- ning the risk of lawsuits, and it was generally understood that job security was guaranteed unless a worker committed a serious crime. Before the new KOREA: LABOR MARKET OUTCOMES AND POLICY RESPONSES 12I legislation, firms (especially large ones) forced workers to quit "voluntar- ily" by removing their desks or transferring employees to a distant area or a lower-level position. The Tripartite Committee was established in January 1998 as a forum for organized labor, business, and government to discuss appropriate labor market policy responses to the crisis. The committee considered the rec- ommendations of the Central Labor-Management-Government Council to amend the Labor Standards Act and allow dismissals for economic rea- sons. The agreement reached by the committee amended the act to allow dismissals, providing that every effort to avoid retrenchment had been ex- plored, that rational and fair standards had been used to select employees for dismissals, that employees had been given advance notice, and that sincere consultations had been held with trade unions or worker repre- sentatives. Large firms were also obliged to notify the Ministry of Labor in advance of any layoffs. Conditions for dismissals are still relatively strict. Moreover, a firm is required to make efforts to rehire the dismissed workers if it intends to re- cruit workers at all within two years of the day of dismissal. In 1998 only 115 firms notified the Ministry of Labor of their intention to lay off 12,000 workers. The EIS reported that dismissals made according to the legislation totaled just 7.1 percent of all job separations, or roughly 140,840 workers (table 3.6). However 14.2 percent of the unemployed who had job experience had been dismissed from their previous jobs (KLI and KIHASA 1998). This figure is far higher than the figures reported by the EIS and Ministry of Labor, suggesting that most dismissals did not meet the required conditions and violated the necessary procedures. Many firms, especially SMEs, where labor unions have little influence and man- agement has the upper hand, laid off their employees arbitrarily, without due process. Moreover, the legalization of dismissals "for managerial needs" encouraged this phenomenon, since many firms regard this legislation as permission to dismiss workers unconditionally. Act for Dispatched Workers The Tripartite Committee also tried to enhance employment flexibility by legalizing the use of dispatched workers. Even though many temporary work agencies lease workers such as secretaries, maintenance personnel, and drivers, the law had prohibited dispatching workers. The Act Protect- ing Dispatched Workers, passed in February 1998, changed the legal cli- mate in this regard. 122 EAST ASIAN LABOR MARKETS TABLE 3.6 Reasons for Job Separation (percentage of sample) Survey on Unemployment and Welfare Reason EIS data Needs Voluntary resignation (to change jobs) 50.3 41.5 Voluntary resignation (to marry, care for children, or related reason) 2.3 2.6 Resignation (health reasons, age) 1.2 4.2 Dismissal (for managerial needs) 7.1 14.2 Bankruptcy or shutdown of firm 6.6 24.0 Termination of contract 2.6 1.2 Early or honorable retirement (usually with allowances) 15.9 2.0 Regular retirement 0.8 0.8 Others 10.0 9.5 Total 100.0 100.0 Source: EIS data: Phang (1999); Survey: KLI and KIHASA (1998). Under the act dispatched workers can be used for up to one year in 26 occupations deemed to require specialized knowledge, skills, and experi- ence. The act permits one-year extensions if both parties agree. Employers must make potential workers aware of the working conditions. And to protect dispatched workers, the act stipulates that employers must not dis- criminate between them and regular workers. Their contracts cannot be terminated because of religion, gender, or social status. By the end of 1998, some 789 agencies had been established that employed a total of 42,000 dispatched workers. Employment Services and Labor Market Information The Korean government expanded and reinforced public employment ser- vice agencies and created a nationwide network for job information.13 The number of public employment service agencies managed by the central government increased from 52 in February 1997 to 134 in March 1999 (table 3.7). The number of public workers specializing in job placement ser- vices also increased, climbing from 141 to 2,684 during the same period.14 KOREA: LABOR MARKET OUTCOMES AND POLICY RESPONSES 123 TABLE 3.7 Employment Service Agencies February 1997 March 1999. Central government Local labor office 46 - Manpower bank 3 20 Employment security center 3 99 Employment center for daily workers - 15 Subtotal 52 134 No. of counselors 141 2,684 Local government Employment center 251 253 Information center 34 28 Subtotal 285 281 Private agencies 1,432 1,756 Total 1,769 2,156 - not available. Source: Ministry of Labor. In addition to dramatically increasing the number of public employ- ment service agencies and counselors, the government launched several projects designed to provide easy access to labor market information. The government formally launched an electronic labor exchange system, Work- Net, in May 1999, using Canada's WorklnfoNet as a standard. Work-Net, which workers can access via the Internet, provides information on job vacancies, vocational training, career guidance, employment policies, em- ployment insurance, labor market statistics, and labor laws. In July 1999 the government published the Occupational Outlook Handbook, along with various career guidance books and a CD-ROM version of the Dictionary of Occupational Titles.15 It is currently developing an occupational prefer- ence test and in 1998 piloted two other tests, the General Aptitude Test Battery and the Occupational Interest Test Battery, which were adminis- tered nationwide to 300,000 people and 150,000 people, respectively. To improve the quality of employment services and make them more user friendly, the government combined the employment insurance and employment security sections of local labor offices into employment secu- rity centers. These centers, which are based on the concept of one-stop ser- vice, are designed to provide job seekers with all the information and ser- vices they need for their job search, from job vacancy information to vocational training. The primary advantage of these centers is that unem- I24 EAST ASIAN LABOR MARKETS ployed workers can receive unemployment benefits and job search assis- tance in the same place. The central and provincial governments run a joint manpower bank that specializes in job-matching services and an em- ployment center for daily workers. The government also eased regulations on and increased its support for private employment agencies. In 1998 some 1,902 private agencies helped 1,537,857 workers find jobs.16 The government also supports trade unions and employers' organizations that provide job placement services. In June 1998 the requirement that employers report all vacancies to the government was eased, and unnecessary requirements (such as regulations governing office space) were eliminated. For-profit enterprises that pro- vide labor market information such as job vacancies through newspapers, periodicals, and the Internet must register with the Ministry of Labor. By the end of 1998, 47 such enterprises had registered with the government. Finally, the government initiated the development of a set of databases of unemployed workers and introduced the Worker Profiling System, a statistical model based on the reemployment experiences of previously un- employed workers with similar characteristics.17 The purpose of profiling is to identify hard-to-employ workers early on and refer them to services that can help them find jobs quickly. The Effectiveness of Employment Services As a result of the government's efforts to expand public employment ser- vices, the number of job seekers using the public employment system jumped from 243,467 in 1997 to 2,130,687 in 1998 (table 3.8). The number of job vacancies posted with public services also increased dra- matically. A survey conducted in September 1998 reported that the ratio TABLE 3.8 Outcomes for Public Employment Service Agencies Vacancies Job seekers Employed (A/B) X (C/B) X Year (A) (B) Placements (C) 100 100 1994 225,652 148,597 180,248 28,141 152 18.9 1995 196,319 116,147 175,416 20,938 169 18.0 1996 215,925 150,668 306,501 20,939 143 13.9 1997 245,223 243,467 644,295 36,425 101 15.0 1998 410,005 2,130,687 2,356,881 157,442 19 7.4 Source: Ministry of Labor. KOREA: LABOR MARKET OUTCOMES AND POLICY RESPONSES 125 of unemployed registered with public employment service agencies rose from 4.8 percent in 1996 to 22.5 percent in 1998 (KLI and KIHASA 1998).18 Only 9.1 percent of the unemployed used private employment services, suggesting that public employment services played an increas- ingly important role in job matching in 1998. However, the effectiveness of these services is still in question. Even though the absolute number of job seekers finding jobs through the pub- lic system more than quadrupled, the employment rate ([employedljob seekers] X 100) decreased only by half from 1997 (15.0 percent) to 1998 (7.4 percent). In part this disparity reflected a tight labor market still feel- ing the effects of the economic downturn. But other statistics show that only 5.8 percent of the unemployed found jobs through the public em- ployment services. According to the September 1998 survey, unemployed workers used an average of 2.36 job search methods. These findings sug- gest that public employment services are relatively inefficient compared with other job search methods such as friends and relatives and direct contacts. 19 Many of the unemployed visited public employment service agencies such as local labor offices and manpower banks. But the general feeling was that the current system did not provide customers with timely and re- liable information on the supply of and demand for labor (table 3.9). TABLE 3.9 Evaluation of Public Employment Services (percentage of respondents) Agree Agree Disagree Disagree completely somewhat somewhat completely Staff is kind and friendly 18.8 54.0 19.0 4.5 Staff know their duties very well 14.7 53.8 19.4 3.4 I can find plenty of job openings 2.5 10.8 47.7 32.1 Services are provided quickly 2.1 18.9 41.2 30.2 Job openings suggested are adequate for me 0.9 8.7 28.3 54.2 I received in-depth counseling on career development 3.1 10.9 32.8 42.0 Source: KLI and KIHASA (1998). I26 EAST ASIAN LABOR MARKETS Most job seekers believed that public employment services failed to pro- vide adequate job opportunities and counseling on career development. Most of the employees working in public employment service agencies had less than one year of experience in this field. Furthermore, both the quality and quantity of the information the system supplies were seriously limited. Before the financial crisis, information on job openings was usu- ally sufficient for job seekers. However, in an environment of high unem- ployment and economic contraction, job seekers want information not only on available job opportunities, but also on vocational training, career development, the labor market situation, and occupational prospects. The government's failure to provide much-needed information becomes clearer when we compare Work-Net with labor market information sys- tems in other countries, including Canada's WorklnfoNet and the America Labor Market Information System (ALMIS) in the United States. Work- Net did not supply occupation-specific information, local labor informa- tion, or occupational prospects. These services are essential to helping work- ers plan their careers and fully utilize Korea's limited human resources. Additionally, the number of public employment agencies and coun- selors was insufficient. Even after the initial drastic increase in the number of public employment service agencies and counselors, the number of workers per counselor stood at 3,901 in 1997, compared with 364 in Ger- many, 325 in Sweden, and 745 in the United Kingdom (table 3.10). Japan's figure was closer to Korea's, but Japan's unemployment rate was still relatively low. In any case the number of unemployed (as opposed to workers) per counselor was 150 in Japan but 664 in Korea. Vocational Education and Training In 1967 the government established a national training system with the enactment of the Vocational Training Law. In 1973 the National Techni- TABLE 3.10 Public Employment Service Agencies and Counselors, 1997 Country Agencies Counselors Workers per counselor Japan 619 15,320 3,401 Germany 842 93,000 364 Sweden 570 11,000 325 United Kingdom 1,159 34,000 745 Koreaa 119 2,684 3,901 a. Data are for March 1999. Source: Ministry of Labor. KOREA: LABOR MARKET OUTCOMES AND POLICY RESPONSES 127 cal Qualification Law put in place a system of skill certification. Laws en- acted in 1976 introduced and formalized mandatory in-plant vocational training and established a Vocational Training Promotion Fund financed by a training fee levied on firms that did not meet the stipulated training requirements. In 1982 the Ministry of Labor founded the Korea Man- power Agency (KOMA) as its training arm. In 1989 the Korea Institute for Technology and Education was established to provide teacher training programs for the vocational training centers. By 1998 approximately 500 vocational training institutes were provid- ing various levels of training for a broad range of occupations (table 3.11). KOMA is in charge of 40 public vocational training institutes aimed at providing the skills industries need most but cannot teach in in-plant training. The remaining institutions are in-plant training centers (237) and authorized training centers (178) run by nonprofit organizations and private sector firms. Between 1967 and 1998 approximately 3.3 million people attended training courses at these institutions. Government support for vocational and technical education and train- ing in Korea was important in developing the skills that have helped fuel Korea's rapid growth. Having concentrated on expanding primary educa- tion during the 1950s, Korea focused on vocational secondary schools dur- ing the 1960s and 1970s. In the late 1960s and 1970s, the government stepped up efforts to develop skilled and technical workers outside the for- mal education system and established many vocational training centers. In the mid-1970s the government began encouraging firms to offer in-service training. During the 1980s enrollment in vocational secondary schools and in-plant training began to decline, although enrollment in junior vo- cational colleges rose. In 1991, in response to perceived shortages of skilled workers, the government once again began emphasizing vocational edu- cation in secondary schools. In 1994 the government initiated a review of vocational education and training policies as part of a comprehensive ex- TABLE 3.11 Vocational Training Institutes Public Central Local Total KOMA government government In-plant Authorized 1990 316 37 38 7 122 112 1998 509 40 46 8 237 178 Source: Ministry of Labor. 128 EAST ASIAN LABOR MARKETS amination of human resource development policies by the Presidential Commission on Educational Reform. The most important development in technical education at the upper secondary level and in public vocational training programs occurred in the 1 970s, however. Government policy specifically addressed these areas, and appropriate financial support was made available to provide a first-class teaching environment in these programs. Also important in developing these programs was foreign assistance-both financing and technical ser- vices-especially in establishing public training centers. Ongoing efforts were made to adapt the vocational education and train- ing system to changes in the economic environment. When the construc- tion boom in the Middle East in the late 1970s offered opportunities for skilled construction workers, special attention was paid to training work- ers in those trades. In the early 1970s, when the rural electrification proj- ect was implemented, rural training centers were established in each prov- ince. These centers provided short-term training in electrical trades so that locally trained workers would be available to help implement projects. Vocational training programs are classified as initial training, upgrade training, retraining, and job transfer training. But there are no clear dis- tinctions among the last three categories, and in general further training or in-plant training covers all three.20 Initial training, as recognized by law, is for trainees who are recruited specifically for vocational training and new employees who have been on the job less than one month. Initial training programs include three things: i General education, which is coordinated with practical training; i Basic training in the knowledge and skills necessary to the occupa- tion, provided by a training institution or in a plant, either on or off the job; and * Specialized training to improve employability. In 1997 the number of people completing initial training programs ac- counted for 27 percent of all trainees completing craftsmanship training. Of the people who completed in-plant training, which focuses primarily on more advanced training, 84 percent were enrolled in further training and only 16 percent in initial training (table 3.12). Recent Reforms in Vocational Training The Korean government attempted to correct market failures in the area of training by imposing levies on firms that do not provide training, imple- KOREA: LABOR MARKET OUTCOMES AND POLICY RESPONSES I29 TABLE 3.12 Vocational Training by Program in 1997 (number of trainees) Initial Further training Institution Total training Upgrade Retraining Job transfer Total 245,044 66,015 156,067 22,199 763 Public 49,257 15,644 33,107 506 - In-plant 173,686 28,312 122,918 21,693 763 Authorized 22,101 22,059 42 - - - not available. Source: Ministry of Labor, Yearbook of Vocational Training, 1997. menting a national technical qualification system, and providing public vo- cational training. Until the 1980s these systems served the needs of eco- nomic development well. During the 1990s, however, the levy system failed in its goal of encouraging firms to undertake and upgrade training pro- grams voluntarily. The most visible evidence of this failure was the lack of upgrade programs in enterprises. Training programs for blue-collar workers (including new entrants to the workforce) concentrated on basic skills to enhance employability and did not improve the technical knowledge and skills of existing employees. This emphasis on basic skills training can be traced to a policy designed to channel youths who were not in school into the industrial sector dur- ing the economic expansion of the 1970s. This compulsory training sys- tem, like the levy system, contributed to an initial expansion of in-plant training. But it did not meet the training demands of private firms because it did not respond to changes in the Korean economic environment. It also did not provide employers with the incentives and assistance they needed to invest in the further education and training of their workers, largely owing to the regulations governing the use of training funds, the qualifications of trainers, and training materials and equipment. Early in the 1990s training provided by companies on a voluntary basis increased sharply. At the same time the number of youths undertaking vo- cational training decreased markedly owing to the declining proportion of young people in the overall population and the growing number of youths entering universities. Some of the hottest issues in designing the EIS were determining how to encourage the private sector to provide more training voluntarily, boost upgrade training and retraining for in-service workers, and offer job transfer training to the unemployed. I30 EAST ASIAN LABOR MARKETS The government decided that it was premature to abolish the compul- sory system, which had existed for almost two decades. For the time being, it was decided to apply the system to companies of 1,000 or more that still needed a large number of trained workers. As a result, enterprises with fewer than 1,000 employees were exempted from the compulsory voca- tional training system after July 1995. These companies then came under the jurisdiction of the EIS, which leaves the decision to provide training up to the employers and provides financial support based on actual costs. Promoting Vocational Trainingfor Workers As discussed previously, the Basic Vocational Training Act, including the compulsory training system, helped workers acquire certain skills but did not meet the changing demand for industrial manpower. In addition, the law limited the ability of enterprises to provide upgrade training to their employees. In response to this situation, in 1997 the government passed the Act Promoting Workers' Vocational Training. The act establishes a sys- tem of vocational competency development and encourages enterprises to provide further training for their employees on a voluntary basis. The act went into effect in January 1999, replacing the Basic Vocational Training Act (table 3.13). Under the new legislation, the Ministry of Labor can offer financial sup- port to employers implementing vocational competency development pro- grams. The Ministry can also offer support to employees seeking to develop their vocational competency by enrolling in vocational competency devel- opment training programs, programs designated in the Education Law, or certification programs. The enactment of the law provided momentum for another transfor- mation in vocational training. The act removes restrictions on in-plant training, encouraging voluntary training, demand-oriented training, and job competency training for the employed. The government will play a supportive assisting role. Vocational Trainingfor the Unemployed The severe economic downturn, recent bankruptcies, and the downsizing of many companies resulted in rapid unemployment growth in Korea. The government responded by providing training programs for the jobless in order to increase employability. In 1998 vocational training programs for the unemployed were provided to 362,941 persons, approximately 8 times as many as in the preceding year. Almost a quarter of all the unem- KOREA: LABOR MARKET OUTCOMES AND POLICY RESPONSES I3I TABLE 3.13 Legislation on Vocational Training in Korea Basic Vocational TrainingAct Act Promoting Workers' (1976-98) Vocational Training (1999) Premise Compulsory system of Companies encouraged to training. provide voluntary training. Target group Most training is initial Training includes vocational training for graduates. competency development for the employed, the unemployed, and graduates. Training area Training concentrates on Training is expanded to production work in the include clerical, managerial, manufacturing industry. and service workers. Classification Training may be public, in Standard training is based of training plant, or authorized. on standards set up by the Programs include initial, Ministry of Labor. Vocational upgrade, and job transfer competency development training as well as retraining. includes initial, upgrade, and job transfer training. Training Training materials compiled Any training materials can materials or approved by the be used. government. In-plant Enterprises with 1,000 or The government provides training more employees in certain financial support to all industries are required to companies for vocational provide training or pay a competency development training levy. programs. Source: Ministry of Labor. ployed received government-sponsored training. The total budget for vo- cational training in 1998 amounted to 738 billion won, approximately 13 percent of which was spent on programs for the unemployed (table 3.14). About half of those who received training once worked at enterprises covered by the EIS. These workers are therefore eligible for reemployment training lasting from one month to one year. They may take training up to three times until they find a new job. However, training allowances are cut in half for the second period of training and are reduced to zero the third TABLE 3.14 Training for the Unemployed, 1998 Number Number of Number of Budget in training participants (B) participants Type of training (billion won) plan (A) B/A (1997) Total 737.7 320,000 362,941 113.2 42,182 Reemployment training for the unemployed 327.5 289,000 333,541 115.4 27,738 Reemployment training for the unemployed from enterprises covered by the EIS (trainees receive a training allowance of up to 90 percent of the minimum wage, except those receiving unemployment benefits) 178.5 141,000 170,096 120.6 1,949 Training for employment promotion (for the unemployed from enterprises not covered by the EIS, especially economically disadvantaged groups and women) 96.9 110,000 101,709 92.5 25,789 Training to return to agriculture 1.2 2,000 5,126 256.3 - Training for the newly unemployed who are not entitled to unemployment benefits, such as new graduates (especially from polytechnic colleges and universities) 46.1 26,000 43,012 165.4 Training to start a business, especially for the white-collar unemployed 4.8 10,000 13,598 136.0 Manpower development training 69.6 31,000 29,400 94.8 14,444 Initial training to be a craftsman (to promote employment for workers with no skills who do not go on to a higher education) 9.6 11,000 14,515 132.0 5,900 Training for the 3-D (difficult, dirty, dangerous) fields 35.0 10,000 11,000 110.0 2,985 Paid leave training (for workers from enterprises covered by the EIS) 25.0 10,000 3,885 38.9 5,559 Other 340.6 - - - - - not available Source: Ministry of Labor. KOREA: LABOR MARKET OUTCOMES AND POLICY RESPONSES I33 time. Trainees who do not qualify for unemployment insurance receive a monthly benefit of 200,000 to 300,000 won (60-90 percent of the mini- mum wage) during the training period, including allowances for trans- portation, family support, and child care. Trainees learning skills used in the "3-D" (dirty, difficult, and dangerous) industries, which still face a la- bor shortage, receive an additional bonus. Similar training opportunities and benefits are offered to the unemployed who are not covered by the EIS. A pilot program of training vouchers was launched in 1998. The ob- jectives of the voucher system are to provide the unemployed with more choices for receiving training and to promote competition among institu- tions providing it. Colleges, universities, and other educational institu- tions became involved in training activities. In addition the program makes monitoring registration and course attendance easier. Despite some confusion in the program early on, in late 1999 the voucher initiative ex- panded from just three cities to all of Chungchong province (in central South Korea). An evaluation at the end of the year will help determine whether the initiative should be expanded nationwide. Vocational training for the unemployed in 1998 had a number of pos- itive effects, increasing employability, stabilizing livelihoods, and provid- ing a degree of mental security for the unemployed. Despite their success, however, these programs were not without their problems. Some training institutions participated in the programs simply to receive government money and did not provide appropriate training. And some trainees were interested in receiving the training allowances but not in attending the programs. Two other difficulties associated with the programs were the underde- veloped labor market information system and the lack of expert managers. Training programs were selected based on what existing training institu- tions had to offer (for instance, in equipment and facilities) with little thought for either labor market demands or the needs of the unemployed. Thus training institutions often provided the same training programs they were offering before the financial crisis. The placement figures for those completing the training reflect these problems. As of 1998 the employ- ment rate for individuals completing training programs for the unem- ployed was 19.8 percent. New Measures in 1999 In 1999 the Korean Ministry of Labor began implementing plans to pro- vide training for 341,000 jobless workers (200,000 of them under the I34 EAST ASIAN LABOR MARKETS EIS), and the training budget rose almost 19 percent to 880 billion won. Government policy for vocational training in 1999 placed top priority on enhancing employability and increasing the efficiency of the training it- self. To improve the relevancy and efficiency of vocational training for the unemployed, the government implemented the following initiatives: * Transforming the vocational training system to meet trainees' needs. Training courses would be evaluated and redesigned to respond to de- mand. Regulations governing such things as instructors and curricula would be adjusted to meet training needs, and training providers would have more autonomy. * Expanding the kinds of training offered to include more topics and to cover promising job fields. Training programs would give priority to such high value-added fields as information technology. Training programs for currently employed workers would be upgraded. The examination criteria of the National Technical Qualification Test would be revised to meet the needs of knowledge-based industries. e Making training more efficient. As a means of controlling the quality of training programs, the government began developing an effective evaluation system. To increase the placement rate for trainees, a "tailor-made" training system was introduced in March 1999. The system assists trainees by securing employment contracts with com- panies and then providing any necessary training. The companies re- ceive 10 percent of the monthly training costs as an incentive to pro- vide courses that result in placement rates of 50 percent or more within the 3 months following the course. Comprehensive evalua- tions are to be performed of the job placement rate of training in- stitutions, the number of trainees obtaining qualifications, trainees' satisfaction, and the management of trainees after training. The in- dividual evaluation system was designed to increase the probability of job placement. i Enhancing information systems. The government sought to improve the labor market information system in order to supply industries with workers that have relevant skills, thereby fulfilling demand and at the same time ensuring employment. To this end the government be- gan constructing a vocational training information system, enhancing counseling for trainees using individual profiles, and improving the available information on training, especially on employer demand. * Increasing private participation in training. The Act Promoting Work- ers' Vocational Training, as we have seen, was designed to encourage KOREA: LABOR MARKET OUTCOMES AND POLICY RESPONSES I35 the private sector to become more involved in vocational training. The act helps adjust training and subsidies to the needs of customers and reduces regulations on training. The government committed it- self to conducting a comprehensive review of the current training subsidies with an eye to further adjustment. The government targeted specific groups and firms in its effort to boost private involvement in training, including trade unions and small to medium-size enterprises. In general trade unions in Korea had not been involved in vocational training. The government encouraged trade unions and employers to consult about matters related to training under the aus- pices of the labor-management councils at the firm level. Furthermore, the government encouraged SMEs to install training facilities and equipment with a view to boosting the firms' participation in training. It also began examining the possibility of granting subsidies to large enterprises (and their associate companies) that develop training programs. Finally, the government sought to identify companies and training institutes with out- standing performances in order to disseminate their experiences as "best practices" to others involved in vocational training. Looking to the Future The financial crisis that began at the end of 1997 had severe consequences for the Korean labor market. In the wake of the crisis, the government re- sponded by providing an arsenal of both active and passive labor market pro- grams. The EIS is the most important social safety net for the unemployed in Korea (Yoo 1999b). However the EIS covers only about 12 percent of all the unemployed. To expand the number of unemployment benefit recipi- ents, every effort must be made to include in the EIS all workers in small workplaces and temporary workers. In addition, the government needs to carefully review the possibility of applying the EIS to daily workers. Since unemployment benefits cover only a limited group, the govern- ment created a large number of public works jobs to assist the unem- ployed. But these programs have their problems. Some ineligible parties received benefits, and others drew double benefits; meanwhile those most in need of assistance were left out of the projects. Alleviating this situation will require establishing a database of participants in public works projects in order to ensure an unbiased selection process and reduce the possibility that participants draw double benefits. Efforts to monitor the projects will help make implementation more efficient. In the long run public works 136 EAST ASIAN LABOR MARKETS projects need to target unemployed groups and establish links with voca- tional training providers and job search agencies (Park and Lee 1999). Because high unemployment is expected to persist in Korea some years into the 21st century, vocational training to enhance workers' employa- bility and job search assistance through the public employment service are increasingly important. Job search assistance is usually the most inexpen- sive active labor market program, and evaluations from other countries show consistently positive outcomes for this type of activity. To speed up the reemployment process of unemployed workers and enhance labor market efficiency, the public employment service system requires ongoing improvement. The government plans to increase the ratio of job seekers utilizing the public service to 30 percent within a few years. Meeting this goal will require establishing a comprehensive labor market information system and opening more public employment agencies staffed with skilled counselors.21 Finally, even though many of the key components of the in- frastructure needed to build a comprehensive program of public employ- ment services are in place, more can (and should) be done, including: * Revising the current occupational classification system, which is in- adequate, to meet labor market needs; * Improving Work-Net, the integrated labor market information system; • Developing career guidance information; • Improving research on occupational prospects; * Enhancing the connection between private and public employment services; * Improving internal coordination within the Ministry of Labor and with local governmental employment service centers; * Emphasizing data development, improving the quality of data, and expanding dissemination of labor market information; • Developing standard job manuals for counselors; and i Enhancing the expertise and morale of staff working in employment service agencies. Owing to a faster-than-expected economic recovery, the unemploy- ment rate began declining after reaching a record high of 8.6 percent (1.78 million) in February 1999, dropping to 4.8 percent (1.07 million) in Sep- tember 1999. At the same time, however, the employment structure wors- ened, and the proportion of long-term unemployed increased rapidly. Fur- ther, the share of nonregular work in the labor market is expected to persist. To address these issues effectively, the government needs to care- fully evaluate its policy responses.22 KOREA: LABOR MARKET OUTCOMES AND POLICY RESPONSES I37 Although Korea developed various social safety nets for the unem- ployed after the crisis, their absolute level of social protection is still much lower than in industrial countries. The government should make every ef- fort to reinforce the role of the EIS as the primary social safety net and ini- tiate other income support activities to prevent unemployed workers and their families from slipping into poverty. In addition to the income sup- port activities, more active labor market measures are needed. What is equally crucial, however, is the evaluation of these measures to ensure that they are being implemented effectively. Notes 1. For statistical purposes regular employees are defined as workers with an implicit or explicit employment contract of more than one year. Temporary work- ers have an employment contract of one month to one year, and daily workers have a contract of less than one month. 2. Recent empirical findings suggest that the overall labor force participation rate is procyclical (NSO, various issues.) 3. A worker dispatch system allows an employer to lease its employees to an- other firm. The dispatching firm retains an employment relationship with its leased employees. 4. Korea's GDP was 449.5 trillion won in 1998. Park and Lee (1999) exclude expenditures on job creation and support for the managerial stabilization of small and medium-size enterprises, perceiving these as expenditures on indirect unem- ployment measures. Thus their calculations reflect only direct public expenditures on unemployment measures (table 3.1). 5. The following description of the EIS draws heavily on Yoo (1999a, 1999b). 6. The premium had been 0.6 percent until 1998. It was lowered in January 1999, when Korea found itself facing high unemployment. 7. Korea Herald, August 31, 1999. 8. Subsidies for manpower reassignment are paid for one year following the completion of the reassignment. 9. Most evaluation studies show that subsidies to private sector employers have both large deadweight losses and substitution effects (Fay 1996; Martin 1998). Deadweight losses occur when the outcome of a wage subsidy program is no different from the outcome in the absence of the subsidy (that is, employers use the subsidy to hire workers they would have hired anyway). The substitution effect occurs when a worker hired by a firm in a subsidized job is substituted for an unsubsidized worker who would have been hired. Typically, the displacement effect refers to displacement in the product market. 10. Specifically 19,000 won or less for positions such as simple office and out- door work; 24,000 won or less for work that requires certain skills or qualifica- tions or has a high labor intensity; and 29,000 won or less for professional and I38 EAST ASIAN LABOR MARKETS skilled work. The minimum wage rate applicable from September 1, 1999, to August 31, 2000, is 1,600 won per hour (361,600 won per month). 11. More than 15,000 firms went bankrupt between December 1997 and April 1998. 12. For an in-depth discussion on the issue, see KOILAF (1998). 13. The government neglected employment services until the financial crisis. Until the onset of the crisis, the government's share of job matching was roughly 2 percent. 14. Most of the newly hired counselors have civilian status. And these coun- selors are involved not only in job-matching services but also in other areas, such as employment insurance. Thus only a portion of the new counselors are involved full time in job-matching services, a factor that tends to lower their morale. 15. Unlike similar handbooks in other countries, Korea's handbook is not based on a rigorous econometric model. Such a model still needs to be developed. 16. However, there is no reliable data on the proportion of private agencies in- volved in job matching. 17. The worker profiling system is similar to the one developed by the Aus- tralian government, which is used mainly as an advisory tool in the field. For de- tail, see Eberts and O'Leary (1996) and Deety (1998). 18. However, a large number of unemployed workers register in order to par- ticipate in programs such as vocational training and public works rather than to seek jobs. 19. Since 22.5 percent of unemployed workers registered with the public em- ployment service agencies and unemployed workers used an average of 2.36 job search methods, 9.5 percent of the unemployed had to succeed in finding jobs through the public employment services if the services are as efficient as other methods. But the data show otherwise. 20. Further training is generally regulated under the EIS as part of the voca- tional competency development training component. 21. In particular, the government must concern itself with counselors' morale and expertise. 22. As previously discussed, although the EIS has been extended to all em- ployers, the system does not protect a large number of workers classified as daily workers. Daily workers make up an estimated 70 percent of construction work- ers, and the unemployment rate of the daily construction workers is the highest of all occupational groups. A comprehensive employment stability plan for daily construction workers is urgently needed. References Deety, A. 1998. Job Seeker Classification Instrument. Canberra, Australia. Eberts, R. W, and C. J. O'Leary. 1996. Profiling Unemployment Insurance Bene- ficiaries. Kalamazoo, MI: WE. Upjohn Institute. KOREA: LABOR MARKET OUTCOMES AND POLICY RESPONSES 139 Fay, Robert G. 1996. "Enhancing the Effectiveness of Active Labour Market Poli- cies: Evidence from Programme Evaluations in OECD Countries." OECD Labour Market and Social Policy Occasional Papers 18. Paris: OECD. Hallward-Driemeier, M., and others. 1999. "Asian Corporate Recovery: A Firm- level Analysis." Paper presented at the conference organized by the World Bank and the Korea Economic Research Institute, May 27. Kim, Dong-Heon and Eui-Kyoung Park. 1999. "Evaluating the Efficiency of Em- ployment Maintenance Subsidies." Seoul: Korea Labor Institute. KOILAF (Korea International Labor Foundation). 1998. Labor Reform in Korea Toward the 21st Century. Seoul. ______. 1999. Unemployment Measures: From Desperation to Reconstruction. Seoul. KLI (Korea Labor Institute). 1997. 1996 Overseas Labor Statistics. Seoul. ______. 1998. Overseas Labor Statistics. Seoul. KLI and Korea Institute for Health and Social Affairs. 1998. Survey on Unem- ployment and Welfare Needs. Seoul: Ministry of Labor. Lee, Won-Duck, and Kang-Shik Choi. 1998. Labor Market and Industrial Rela- tions in Korea. Seoul: Korea Labor Institute. Martin, John P. 1998. What Works amongActive Labour Market Policies: Evidence from OECD Countries'Experiences. OECD Labour Market and Social Policy Occasional Paper 35. Paris: OECD. MOL (Ministry of Labor). 1999. Midterm Unemployment Policy. Seoul. NSO (National Statistical Office). Various issues. Annual Report on the Economi- cally Active Population Survey. Seoul. OECD. 1999. OECD Economic Surveys: Korea, 1998-99. Paris: OECD. Park, Funkoo, and Joohee Lee. 1999. "The Social Impact of the Financial Crisis: Labor Market Outcomes and Policy Responses in Korea." Seoul: Korea Labor Institute. Unpublished Paper. Park, Young-bum. 1999. "Social Safety Nets for the Unemployed: The Korean Case." Paper presented at the Institute for Social Sciences International Con- ference on Unemployment, Job Creation, and Employment Relations, Octo- ber 29-30. Phang, Hanam. 1999. Characteristics of the Unemployed Who Received Unemploy- ment Benefits in 1998. Seoul: Korea Labor Institute. Shin, D. 1999. Unemployment Structure in Korea. Seoul: Korea Labor Institute. Uh, Soo-Bong. 1999. Employment: Structure, Trends and New Issues. In Labor Relations in Korea, edited by KOILAF. Seoul. Yoo, Kil-Sang. 1999a. Employment Insurance and Social Safety Nets for the Un- employed. In Labor Relations in Korea, edited by KOILAF. Seoul. ____. 1999b. The Employment Insurance System in Korea. Seoul. 4 Malaysia: Protecting Workers and Fostering Growth NoRMA MANSOR, TAN Eu CHYE, ALI BOEHANOEDDIN, FATIMAH SAID, SAAD MOHD SAID FROM 1987 UNTIL the onset of the East Asian financial crisis in July 1997, the Malaysian economy had been one of the fastest- growing economies in Southeast Asia, with an average annual growth rate of 8.5 percent (table 4.1). The crisis trimmed Malaysia's growth rate to 7.7 percent in 1997 and in 1998 plunged the country into a severe recession, reducing the growth rate to -7.5 percent. Prior to the crisis, Malaysia's growth had been driven largely by private investment in productive ca- pacities with increasingly sophisticated technology.' The country had also generally been enjoying a merchandise trade surplus.2 The surplus has be- come even more significant in light of the growth in exports and the dra- matic slowdown in imports amid the recessionary circumstances. For almost 10 years the manufacturing sector rather than the agricul- ture sector was the mainstay of the economy (table 4.2). In 1987 the two sectors contributed around the same amount (22 percent) to Malaysia's GDP. But the manufacturing sector recorded average annual growth of 13.8 percent in 1987-97, while the agriculture sector averaged just 2.9 percent. In 1997 the agriculture sector accounted for only 11.9 percent of Malaysia's real GDP, but the manufacturing sector weighed in at 35.7 per- cent. The construction sector's real GDP contribution rose only margin- ally in 1987-97 (from 3.4 percent to 4.8 percent in 1997). The mining and quarrying sector appears to be playing an increasingly small role in the Malaysian economy, given the contraction in the sector's share in real I4' 142 EAST ASIAN LABOR MARKETS TABLE 4.1 Selected Macroeconomic Indicators GNP per capita Real GDP Infation Unemployment Population at current market Year growth (%) (%) rate (%) (millions) prices (US$) 1985 -1.1 0.4 6.9 15.68 1,896 1986 1.2 0.6 8.3 16.11 1,595 1987 5.4 0.8 8.2 16.52 1,814 1988 8.9 2.5 7.2 16.94 1,869 1989 9.2 2.8 6.3 17.4 2,064 1990 9.7 3.1 5.1 17.8 2,311 1991 8.7 4.4 4.3 18.2 2,474 1992 7.8 4.7 3.7 18.6 2,882 1993 8.3 3.6 3.0 19.6 2,970 1994 9.3 3.7 2.9 20.1 3,515 1995 9.48 3.4 2.8 20.7 3,960 1996 8.6 3.5 2.6 21.2 4,463 1997 7.7 2.7 2.6 21.7 4,284 1998 -7.5 5.2 4.0 22.2 3,013 Source: Treasury, Economic Report, various issues. GDP from 10.5 percent in 1987 to 6.7 percent in 1997. The share of the services sector has remained relatively stable at around 41 percent in 1987-97. During the high-growth era of 1987-96, Malaysia experienced relative price stability. Inflation never exceeded 5 percent, largely because of pru- dent monetary policy and government intervention. The government im- posed direct price controls (especially on consumer essentials), improved distribution and marketing channels, and curbed profiteering and other unethical business practices. Despite the threat of stagflation following the abrupt depreciation of the ringgit and the danger of imported inflation (owing to the economy's heavy dependence on imported intermediate and capital goods), the government succeeded in containing inflation at 5.2 percent in 1998. Malaysia has been experiencing a population growth rate of 2 percent a year. The population rose from 16.52 million in 1987 to 22.2 million in 1998 (table 4.1). The government's relatively prudent macroeconomic management has resulted in a remarkable 66.1 percent rise in per capita GNP, which climbed steadily from US$1,814 in 1987 to US$4,284 in TABLE 4.2 Industry Shares in GDP, 1986-98a (RM million) 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 Agriculture, livestock, 11,854 12,348 13,216 13,933 14,768 14,827 14,828 15,531 16,205 16,047 16,231 16,584 16,804 15,813 forestry and fishing (2.5) (4.2) (7.0) (5.4) (6.0) (0.4) (0.0) (4.7) (4.3) (-1.0) (1.1) (2.2) (1.3) (-5.9) Mining and quarrying 5,985 6,368 6,409 6,803 7,383 7,757 7,944 8,075 8,039 8,241 8,979 9,381 9,475 9,399 (-1.4) (6.4) (0.6) (6.2) (8.5) (5.1) (2.4) (1.6) (-0.4) (2.5) (9.0) (4.5) (1.0) (-0.8) Manufacturing 11,263 12,111 13,734 16,151 18,444 21,340 24,307 26,859 30,324 34,842 39,790 44,684 50,270 47,354 (-3.8) (7.5) (13.4) (17.6) (14.2) (15.7) (13.9) (10.5) (12.9) (14.9) (14.2) (12.2) (12.5) (-5.8) Construction 2,738 2,354 2,077 2,133 2,380 2,832 3,240 3,619 4,023 4,589 5,385 6,150 6,732 5,439 (-8.4) (-14.0) (-11.8) (2.7) (11.6) (19.0) (14.4) (11.7) (11.2) (14.1) (17.3) (14.2) (9.5) (-19.2) Electricity, gas, 948 1,027 1,109 1,211 1,344 1,526 1,697 1,931 2,176 2,474 2,797 3,134 3,543 3,791 and water (6.5) (8.3) (8.0) (9.2) (11.0) (13.5) (11.2) (13.8) (12.7) (13.7) (13.0) (12.1) (13.0) (7.0) Transport, storage, 3,630 3,851 4,055 4,412 4,839 5,487 6,079 6,481 6,921 7,776 8,852 9,711 10,530 10,835 and communication (4.8) (6.1) (5.3) (8.8) (9.7) (13.4) (10.8) (6.6) (6.8) (12.4) (13.8) (9.7) (8.4) (2.9) Wholesale and retail trade, 6,911 6,147 6,423 6,988 7,687 8,807 10,068 11,190 12,428 13,427 14,781 16,163 17,290 17,047 hotels and restaurants (-2.8) (-11.1) (4.5) (8.8) (10.0) (14.6) (14.3) (11.1) (11.1) (8.0) (10.1) (9.4) (7.0) (-1.4) Finance, insurance, real estate and 5,094 5,071 5,482 6,088 6,771 7,759 8,733 9,644 10,650 11,713 12,938 14,825 16,240 16,911 business services (4.1) (-0.5) (8.1) (11.0) (11.2) (14.6) (12.6) (10.4) (10.4) (10.0) (10.5) (14.6) (9.5) (4.1) Government services 6,957 7,253 7,543 7,819 8,185 8,579 8,964 9,201 10,073 11,022 11,454 11,931 12,654 12,958 (2.1) (4.3) (4.0) (3.7) (4.7) (4.8) (4.5) (4.9) (9.5) (9.4) (3.9) (4.2) (6.1) (2.4) Other services 1,301 1,353 1,400 1,454 1,522 1,678 1,831 1,983 2,146 2,298 2,478 2,687 2,880 2,923 (4.1) (4.0) (3.5) (3.9) (4.7) (10.2) (9.1) (8.3) (8.2) (7.1) (7.8) (8.4) (7.2) (1.5) Less: Imputed bank 1,834 1,891 2,235 2,820 3,356 4,076 4,804 5,376 6,411 7,381 8,503 10,032 11,498 12,280 service charges (15.0) (3.1) (18.2) (26.2) (19.0) (21.5) (17.9) (11.9) (19.3) (15.1) (15.2) (18.0) (14.6) (6.8) Add: Import duties 2,245 1,759 1,650 2,131 2,442 2,947 3,458 3,728 4,043 4,927 5,090 5,402 5,765 3,747 (-11.0) (-21.6) (-6.2) (29.2) (14.6) (20.7) (17.3) (7.8) (8.4) (21.9) (3.3) (6.1) (6.7) (-35.0) GDP at market 57,093 57,751 60,863 66,303 72,409 79,463 86,345 92,866 100,617 109,976 120,272 130,621 140,684 133,939 prices (1.2) (5.4) (8.9) (9.2) (9.7) (8.7) (7.8) (8.3) (9.13) (9.4) (8.6) (7.7) (-4.8) a. In 1978 prices. Note: Figures in parentheses are annual percentage changes. Source: Treasury, Economic Report, various issues. I44 EAST ASIAN LABOR MARKETS 1997. But the steep depreciation of the ringgit following the de facto de- valuation of the Thai baht has severely eroded Malaysians' living stan- dards, scaling down per capita GNP to USUS$3,013 in 1998-a hefty 29.7 percent decline. In fact, the per capita income target of RM 14,788 set for the year 2000 has become impossible to meet. Instead, the per capita income (both in nominal and real terms) for 2000 is projected to be lower than in 1997. Labor Market Trends The manufacturing sector has become not only an increasingly important source of national income but also a prime generator of national employ- ment. Manufacturing employment accounted for 27.1 percent of total na- tional employment in 1998, up from 15.7 percent in 1987. The share of the agriculture sector in national employment fell considerably, dropping from 31.9 percent in 1987 to 16.5 percent in 1998. This decrease can be traced in part to the adoption of mechanized production processes and in- tensive farm management practices, and in part to the growing emphasis placed on developing other sectors. The share of the mining and quarry- ing sector has been relatively stable at around 0.005 percent. Both the construction sector and the services sector registered a moderate expansion in their shares. Construction saw its share of employment expand from 6 percent in 1987 to 8.8 percent in 1998, and employment in the services sector grew from 45.8 percent to 47.1 percent. The national unemployment rate declined steadily, falling from 8.2 per- cent in 1987 to 2.6 percent in 1997, primarily because of the government's progrowth policies (table 4.1). In fact, Malaysia had already achieved full employment in 1993 with an unemployment rate of 3 percent, and the labor market tightened further in the years that followed. Employment The deflationary effect of the financial market turmoil led to a surge in the unemployment rate, which rose to 4.0 percent in 1998. Despite this in- crease, the crisis has had a limited effect on job opportunities in Malaysia. Workers have been retrenched as firms restructure their operations, but many of those retrenched have been the foreign workers Malaysian indus- tries had come to depend on in the overheated precrisis economy. Despite the retrenchments, then, Malaysia continues to face labor shortages in cer- tain sectors of the economy owing to a lack of intersectoral mobility and the MALAYSIA: PROTECTING WORKERS AND FOSTERING GROWTH 145 strong preference among locals for jobs that are not perceived as menial. Since February 1998 the government has required employers to inform the Department of Labor at least one month in advance of any retrenchments, wage cuts, or temporary layoffs in order to monitor workers' welfare. THE INFORMAL SECTOR. Malaysia has no unemployment insurance. Some of those who have lost their employment in the formal sector have ventured into the informal sector, becoming petty traders and street ven- dors or operating from their residences, performing odd jobs and engag- ing in other income-generating activities. Hawker stalls have proliferated in cities. While no studies have been made of the informal sector per se, it is safe to assume that in Malaysia, as in many developing countries, the sector serves as an important source of employment for the poor and un- employed, especially during shocks. The absence of a social safety net makes informal ventures more com- pelling. In order to promote self-employment and self-reliance, in 1996 the government implemented a revolving fund of RM 300 million (the Amanah Ikhtiar Malaysia). This program to aid the poorest in setting up small businesses has disbursed a total of RM 289.8 million in interest-free loans to 54,856 participants, and its success is reflected in the 100 percent repayment rate. The number of participants has increased because of the economic downturn and is projected to reach 4,260 participants in 2000. Another program designed to assist petty traders and informal sector workers is the Yayasan Tekun Nasiona4 which has disbursed loans totaling RM 8.7 million to 2,890 entrepreneurs. While these programs help some retrenched workers maintain an income, the informal sector cannot be ex- pected to provide full insulation from shocks such as the Asian crisis. Busi- nesses in the informal sector are also likely to suffer as the formal sector falters. UNEMPLOYMENT. During the high-growth period before the crisis, Malaysia recorded an absolute decline in the number of job seekers, espe- cially among workers in the 15-19 and 25-29 age groups. In what may be a reflection of Malaysia's success in providing greater access to tertiary education, those in the 20-24 age group constituted the bulk (44 percent) of total job seekers in 1997. Women appear to be having more trouble finding jobs than men. While males constituted 63.2 percent of the total number of registered job seekers in 1987, they accounted for about 52.1 percent in 1997 (table 4.3). The number of female job seekers, however, rose from 36.8 percent in 146 EAST ASIAN LABOR MARKETS TABLE 4.3 Registered Job Seekers by Sex (end of period) End of Total job seekers Males Females 1985 80,681 49,279 31,402 1986 86,896 55,857 31,039 1987 78,553 49,642 28,911 (63.2) (36.8) 1988 78,200 49,478 28,722 1989 72,127 46,130 25,997 1990 54,387 33,915 20,472 1991 50,159 30,050 20,109 1992 42,344 25,287 17,057 1993 31,617 18,594 13,023 1994 26,445 15,095 11,350 1995 25,546 13,935 11,611 1996 21,747 11,863 9,884 (54.5) (45-4) 1997 23,762 12,386 11,376 (52.1) (47.8) July 34,514 18,423 16,091 1998 (53.4) (46.6) Note: Figures in parentheses refer to percentage shares in the total number of job seekers. Source: Treasury, Economic Report, various issues. 1987 to 47.8 percent in 1997. In part, this phenomenon is attributable to the increasing entry of females into the labor force as Malaysia modernizes (table 4.4). The overall labor force participation and male labor force par- ticipation rates are relatively stable, but the female participation rate in- creased from 44.9 percent in 1987 to 47.3 percent in 1997. This increase in women seeking work may also mean that women are feeling the conse- quences of the East Asian crisis slightly more than their male counterparts. Full employment has also contributed to the rising incidence of job switching-workers who are already employed seeking alternate employ- ment. The share of the employed registering as job seekers expanded from 9 percent in 1987 to 33.8 percent in 1997. The economic recession brought forth by the financial market turbulence has helped reduce the percentage of employed job seekers, however, and their number fell to 20.7 percent of total job seekers in 1998. Prior to the crisis, the labor mar- ket seemed to favor those with less education. The percentage of those MALAYSIA: PROTECTING WORKERS AND FOSTERING GROWTH 147 TABLE 4.4 Labor Force Participation Rates by Gender Labor force Labor force participation rates Year (millions) Totala Maleb Femalec 1985 6,039.1 65.8 87.4 44.3 1986 6,222.2 65.8 87.5 44.2 1987 6,408.9 65.9 86.9 44.9 1988 6,658.0 66.1 85.8 46.5 1989 6,850.0 66.3 85.8 46.9 1990 7,042.0 66.5 85.6 47.3 1991 7,204.0 66.6 85.7 47.5 1992 7,370.0 66.7 85.7 47.6 1993 7,627.0 66.8 87.0 46.1 1994 7,834.0 66.8 87.1 46.5 1995 8,256.8 66.9 86.8 47.1 1996 8,641.4 66.7 86.6 47.2 1997 9,038.2 66.6 86.9 47.3 1998 9,006.5 64.3 86.7 46.8 a. Total number of people economically active as a percentage of total working age pop- ulation (15-64 years). b. Total number of males economically active as a percentage of total number of males in the working age population. c. Total number of females economically active as a percentage of total number of females in the working age population. Source: Treasury, Economic Report, various issues. with little education registered as job seekers contracted from 27.2 percent in 1987 to 14.1 percent in 1997. The percentage of job seekers with some higher education increased, while their share in the labor force fell. Labor Productivity and Wages The economic slowdown of 1985-86 had already forced establishments to make productivity improvements as part of the adjustment process. Rising unemployment and retrenchments exerted pressure on employees to boost productivity in order to retain their jobs. For the economy as a whole, labor productivity, or value added per worker, increased from RM 10,160 in 1985 to RM 10,867 in 1988, an increase of 2.2 percent an- nually. The manufacturing sector, which underwent severe adjustments (including retrenchments), recorded an increase of 6.5 percent per year 148 EAST ASIAN LABOR MARKETS in value added per worker. In the agriculture sector labor productivity in- creased by 2.5 percent annually during 1985-88. The economic condi- tions that prevailed during 1986-88 did not warrant the rapid wage in- creases of the first half of the 1980s, and wages were moderated because of the need for economic adjustment. In the manufacturing sector, aver- age nominal earnings of workers increased by 2.2 percent per year, while the increase in real earnings was close to 1 percent. Overall, average wage increases in collective agreements negotiated in 1986 was 4 percent for 3 years. For the manufacturing sector, the rate was 4.3 percent. In 1987 the average rate dropped to 3.2 percent and 2.5 percent, respectively. Labor productivity continued to increase in the years before the finan- cial crisis. Despite the strong demand for labor in 1991-95, labor pro- ductivity rose across all sectors. For the economy as a whole, labor pro- ductivity measured as GDP per worker in constant 1978 prices increased by 5.1 percent a year, rising from RM 11,870 in 1990 to an estimated RM 15,200 in 1995. (This rate of growth exceeded the overall labor produc- tivity growth of 3.3 percent during 1986-90.) Output per worker in the manufacturing sector, which stood at RM 19,410 in 1995, was higher than for the economy as a whole. Overall economic growth averaged 3.9 percent annually between 1990 and 1995. Labor productivity growth in the agriculture and services sectors was 6.1 percent and 5.9 percent, re- spectively, during this period. At 5.9 percent a year, growth in GDP per worker in the services sector was particularly impressive. A growth rate of 1.8 percent in 1986-90 indicates increased efficiency in the use of labor resources and the widespread use of information technology. In 1991-95, however, the tight labor market resulted in upward pres- sure on wages. Average nominal manufacturing wages rose by about 27 percent between 1990 and 1994, or 6.2 percent annually. With pro- ductivity growth lagging behind wage growth in the sector, pressure on unit labor costs increased. This phenomenon was particularly marked in 1990-92, when high wage growth unaccompanied by productivity growth resulted in rising unit labor costs. Productivity growth in the manufactur- ing sector began rising at the end of 1992, however, climbing 4.6 percent and 6.8 percent in 1993 and 1994, respectively. Consequently, unit labor costs in the manufacturing sector began declining in the mid-1990s. The tight labor market situation since 1995 has exerted pressure on wages, with wage increases outstripping productivity growth. Nominal wage rates grew by 5.7 percent in 1996 and 6.1 percent in 1997 but in- creased at just 4 percent in 1998. However, labor productivity grew at a slower rate during those years, rising to only 3.5 percent in 1996 before MALAYSIA: PROTECTING WORKERS AND FOSTERING GROWTH I49 dipping slightly to 3 percent in 1997. Moreover, productivity growth reg- istered a negative 3.8 percent owing to the underutilization of existing capacity-the result of the 1998 economic slowdown. In the manufactur- ing sector, average nominal wages increased by 10 percent a year during 1996-97, but productivity rose by a mere 1 percent. The result has been pressure on unit labor costs, particularly in 1996 and 1997, when they rose by 4.8 percent and 2 percent before declining by 4.5 percent during the 1998 economic slowdown. Although the economy appears to have recovered, many firms are still paying wages that are lower than they were before the crisis. However, this situation can be viewed as a technical correction, as the country had sus- tained a loss of competitiveness in the precrisis years, with wage growth outstripping productivity growth. Bank restructuring and mergers, which are gathering momentum in the country, and the gloomy outlook for the electronics industry argue for at least a cap on further wage increases. Industrial Relations The number of trade unions increased from 479 in 1992 to 516 in 1996 (table 4.5). The government favors in-house rather than national unions. Many of the unions (316 out of 516 in 1996) are for workers in the man- ufacturing and services sectors. The number of trade unions increased in four sectors-agriculture, forestry, and fishing; manufacturing; electricity, gas, and water; and commerce-but consolidation and rationalization in other sectors prevented the unions from expanding. Inclusive of employ- ers' trade unions, the number of trade union memberships increased steadily from 680,647 in 1992 to 728,246 in 1996 (table 4.5). Private- sector employees accounted for slightly more than half the memberships (table 4.6). These developments suggest that the government allows work- ers to organize. Although recently the Malaysian Trade Union Congress has pushed for a minimumn wage law, no such law has been enacted. Collective agree- ments appear to be an significant means of determining wages. The num- ber of collective agreements concluded increased from 334 in 1992 to 398 in 1996 (table 4.7). In 1996 over half of the agreements (210 out of 398) were in the manufacturing sector. The tendency to link wages to workers' productivity is growing and may help to contain inflation and enhance the country's international competitiveness. In 1994 about 60 percent of the collective agreements in force contained productivity- or performance- related incentives. TABLE 4.5 Trade Unions and Membership, by Sector, 1992-96a 1992 1993 1994 1995 1996 Sector Number Membership Number Membership Number Membership Number Membership Number Membership Agriculture, forestry 35 84,415 37 82,742 36 83,943 39 86,421 43 91,983 and fishing (12.4) (11.9) (12.0) (12.2) (12.6) Mining 5 3,110 5 2,921 5 2,186 4 1,531 3 1,491 (0.5) (0.4) (0.3) (0.2) (0.2) Manufacturing 102 149,569 106 147,487 117 147,359 119 141,021 119 144,814 (22.0) (21.3) (21.1) (20.0) (16.9) Construction 7 3,402 7 4,588 6 4,661 6 2,849 6 4,032 (0.5) (0.7) (0.7) (0.4) (0.6) Electricity 26 33,585 26 36,625 24 32,725 27 51,436 30 54,843 (4.9) (4.7) (4.7) (7.3) (7-5) Commerce 30 465,01 32 53,035 31 54,574 45 102,803 45 100,480 (6.8) (7.7) (7.8) (14.6) (13.8) Transport and 73 60,123 78 55,413 78 55,908 75 46,018 73 42,963 communication (8.8) (8.0) (8.0) (6.5) (5.9) Services 201 299,302 205 314,770 204 318,017 189 274,174 197 287,640 (44.0) (45.4) (45.5) (38.8) (39.5) Total 479 680,007 496 693,581 501 699,373 504 706,253 516 728,246 (100.0) (100.0) (100.0) (100.0) (100.0) a. Does not include Employers' Trade Union and Federation of Trade Unions. Note: Figures in parentheses are percentages of total numbers. Source: Department of Trade Union Affairs, Ministry of Human Resources. TABLE 4.6 Trade Unions by Membership, 1992-96 1992 1993 1994 1995 1996 Category Number Membership Number Membership Number Membership Number Membership Number Membership Employees' trade unions Private-sector 258 373,288 276 384,867 281 384,867 281 396,663 292 407,303 employees (54.8) (55.4) (55.0) (56.1) (55.9) Government 131 215,304 132 219,579 133 225,897 135 226,823 136 241,411 employees (31.6) (31.6) (32.3) (32.1) (33.1) Statutory body and local authority 90 91,415 88 89,402 87 88,609 88 82,767 88 79,532 employees (13.4) (12.9) (12.7) (11.7) (10.9) Employers' trade 18 640 16 616 16 616 13 572 13 528 unions (0.1) (0.1) (0.1) (0.1) (0.1) Total 497 680,647 512 694,197 517 699,989 517 706,253 529 728,246 (100.0) (100.0) (100.0) (100.0) (100.0) Note: Figures in parentheses are percentages of total numbers. Source: Department of Trade Union Affairs, Ministry of Human Resources. I52 EAST ASIAN LABOR MARKETS TABLE 4.7 Collective Agreements, 1992-96 1992 1993 No Wage No Wage of Workers increase of Workers increase Sector C/A involved (%) C/A involved (%) Agriculture and estates 12 6,515 8 15 27,916 5 Mining and quarrying 4 1,784 7 3 2,439 5 Manufacturing 191 51,042 11 199 54,940 11 Construction - - - - - - Electricity, gas, and water - - - - - - Commerce 46 3,404 8 44 6,727 10 Transportation, storage and communication 40 14,277 8 31 3,633 8 Service 34 32,177 10 39 10,001 9 Others 7 614 6 1 2 5 Total 334 109,813 332 105,658 - data not available. Note: C/A: Collective agreement. Source: Industrial Relations Department, Ministry of Human Resources. Job Creation and Maintenance and Income-support Activities Rapid economic growth during the 1990s led to an expansion of employ- ment opportunities in Malaysia. Most of the new jobs were in the manu- facturing, construction, and services sectors, highlighting the importance of these areas (table 4.8). Private investment continues to be the primary source of new jobs, but the government has supported job creation di- rectly (through public work and privatization programs) and indirectly (through loan and grant programs). While it is still too early to assess the full effects of these initiatives, they have contributed to the net gain in job opportunities and have aided unemployed workers seeking to support themselves in the informal sector Private Direct Investment Private direct investment (including foreign direct investment), particu- larly in manufacturing, was the primary source of employment creation in MALAYSIA: PROTECTING WORKERS AND FOSTERING GROWTH I53 1994 1995 1996 No Wage No Wage No Wage of Workers increase of Workers increase of Workers increase C/A involved (%) C/A involved (%) C/A involved (%) 15 1,481 10 14 91,863 15 32 12,990 15 - - - - - - 6 1,342 9 196 53,468 9 216 58,656 28 210 47,674 11 - 162 - 1 1,237 10 - - - - 20,452 - - - - 9 29,713 5 53 10,080 11 95 15,116 12 42 4,866 13 31 3,093 15 35 32,399 12 55 9,918 10 46 214,77 9 38 8,458 11 38 6,364 9 6 25,51 9 5 232 7 6 466 9 348 112,764 404 207,961 398 113,333 1996-98. It is expected to be the primary source of jobs in 1999-2000 as well, creating a total of 129,900 jobs, or 39 percent, of all new jobs. The financial crisis has not seriously affected private investment in Malaysia. The latest figures on direct investment projects prove that Malaysia continues to attract both local and foreign capital. In the first quarter of 1999 (anuary-March), Malaysia received 160 applications to set up manufacturing projects worth RM 2.67 billion. Of these, 78 were for new projects and 82 for expanding and diversifying existing invest- ments. A total of 145 manufacturing projects were approved, with total capital investments totaling RM 1.81 billion. Of this amount, RM 1.32 billion was foreign direct investment. Of the 145 approved projects, 75 were new and 70 expansion and diversification. Clearly investors recognize that Malaysia is an attractive and viable manufacturing and export base. Direct Job Creation Since 1996, the government has sought to increase employment opportu- nities through direct job creation. To this end it has funded public works projects and undertaken privatization projects with the private sector.3 The major public work programs include infrastructure projects, agricul- I54 EAST ASIAN LABOR MARKETS TABLE 4.8 Net Job Creation by Sector, 1996-00 Average annual Net job creation growth rate (percent) (thousands) Sector 1996-98 1999-00 1996-98 1999-00 Agriculture, forestry, livestock and fishing -2.0 -0.3 -90.6 -7.9 Mining and quarrying 1.1 0.5 1.2 0.4 Manufacturing 3.8 2.8 243.9 129.9 Construction 0.7 2.1 15.6 30.3 Electricity, gas, and water 5.7 3.1 11.8 4.9 Transport, storage, and transportation 3.2 2.2 39.2 19.8 Wholesale and retail trade, hotels, and restaurants 3.1 1.8 124.2 53.4 Finance, insurance, real estate, and business services 3.6 3.9 41.9 32.7 Government services 0.2 0.2 5.5 4.0 Other services 5.7 4.1 121.1 65.4 Total 2.1 1.9 513.8 332.9 Source: Government of Malaysia, Midterm Review of Seventh Malaysia Plan, 1996-2000. tural development projects, and rural development projects. Public works projects had a generally positive impact on job creation in 1996-98 (table 4.8).4 The exceptions were in the agriculture, forestry, livestock and fish- ing sector, which lost jobs during the period. Privatization occurred under the government's privatization program from 1996-98. Both the public sector and privatization programs were directed toward expanding capac- ity to support the private sector and meet the needs of Malaysia's mod- ernizing economy. The government's allocation of development resources shows the extent of the public sector projects. In the Seventh Malaysia Plan for 1996-00, economic and social programs accounted for the major part of develop- ment expenditure. Of RM 89.5 billion in total expenditure, almost 80 percent is allocated to economic and social programs. Some RM 43 bil- lion, or 48.2 percent, is allocated to economic sectors and RM 27.6 bil- lion, or 30.9 percent, to social sectors. In the economic sector, the largest portion (22.1 percent) is reserved for transportation and communication, with another 9.6 percent going to commerce and industry, 9.3 percent to MALAYSIA: PROTECTING WORKERS AND FOSTERING GROWTH I55 agriculture and rural development, and 4.3 percent to water resources. The breakdown of social sector spending is 17.8 percent for education and training, 3.9 percent for housing, 3.8 percent for health and population, and 2.0 percent for local authorities and welfare services. INFRASTRUCTURE PROJECTS. In its quest to become an industrial econ- omy by the year 2020, between 1996 and 1998 Malaysia devoted a large share of its development expenditure to the transportation and communi- cation subsectors (around RM 13.5 billion, or 28 percent). Most of the funds were used to develop railroads, ports, highways, and the new gov- ernment administrative center at Putrajaya, as well as to support sewerage, waste disposal management, and water supply projects. Among the large infrastructure projects completed in 1996-98 were the North-South Highway, the West Port in Selangor, the Light Rail Transit (LRT) in Klang Valley, and the telecommunication and television satellite. During the period of economic slowdown, the government reviewed the expenditures of both economic and social programs. As a result, it redirected policy to give more weight to helping those most severely af- fected by the crisis, particularly low-income groups and the poor. Specific steps to redirect policy to meet this goal include: C Continuing to develop infrastructure projects. For the remainder of the Seventh Malaysia Plan, the government will continue to develop in- frastructure projects essential to an industrial economy. The govern- ment has provided RM 5.243 billion for developing infrastructure and public utilities and will give priority to projects that are regarded as catalysts for economic growth, including ongoing projects and those nearing completion. - Reducing operating and maintenance cost through the privatization. Since 1993, a total of 434 projects have been completed under the Privatization Program. From 1996 to 1998 alone, the total costs of 68 projects amounted to RM 667 trillion. The program is said to be relieving the government's financial and administrative burden and improving the efficiency and productivity of the economy. The pro- gram is believed to be facilitating economic growth while helping to meet planned targets in pursuing growth and equity, including en- hancing the participation of Malaysia's indigenous people (Bumipu- tra) in the corporate sector. During the crisis, the privatization program has continued to play an important role in facilitating economic recovery and growth. To 156 EAST ASIAN LABOR MARKETS ensure the effectiveness of the programs, the government is taking steps to streamline procedures, provide effective coordination and monitoring, and strengthen the regulatory aspects. Priority is given to those projects with the greatest multiplier effect. Providing funding for injrastructure projects. To help finance infra- structure projects and projects involving large public facilities, the government has established the Infrastructure Development Fund with an initial allocation of RM 5 billion. The fund aids companies undertaking infrastructure projects that require large capital outlays, particularly the strategically important privatized road projects. The government provides financing in the form of soft loans through the newly restructured Development and Infrastructure Bank Malaysia Berhad. The fund not only helps the companies undertaking such projects but also indirectly creates and maintains employment. AGRICULTURAL DEVELOPMENT. The government's agriculture develop- ment projects are designed to improve productivity through commercial- ization and private-sector participation. The government has two ap- proaches to furthering its goals: the In-Situ Development Program and the New Land Development Program. The In-situ Development Program focuses on integrated agricultural de- velopment, drainage, irrigation, replanting, and land consolidation for smallholders. In 1996-98 it developed a total of 156,000 hectares. The program helps smallholders group themselves into larger production units, creating economies of scale by increasing productivity and alleviating labor shortages. The program is projected to develop another 120,000 hec- tares of land in 1999-2000 and will include rehabilitation programs. The New Land Development Program developed a total of 65,000 hec- tares of new land for agricultural purposes in 1996-98. In line with the policy of limiting the government's role in this area, the private sector de- veloped another 50,000 hectares. The program will develop a projected 72,000 additional hectares in 1999-2000. The private sector will be re- sponsible for 55.5 percent of this figure. RURAL DEVELOPMENT. The government's rural development policy em- phasizes improvements in the living standards of the rural sector. The new rural development philosophy places equal emphasis on human and phys- ical development. The government has introduced several programs, in- cluding the Rural Vision Movement, One Village One Product, Home Rehabilitation, Uplifting the Village Economy, and Entrepreneurial De- velopment, among others. MALAYSIA: PROTECTING WORKERS AND FOSTERING GROWTH I57 In 1996-98, the government undertook a total of 3,000 rural water sup- ply projects, 270 road projects, 10,000 projects under the Home Rehabilita- tion Program, and 6,000 projects under the Uplifting the Village Economy Program. In 1999-2000, 1,800 villages will be developed under the Rural Vision Movement and 1,400 under the Community Enhancement Pro- gram. In addition, the government will increase its actions to promote en- trepreneurship, develop infrastructure, promote organized farming activities, implement the One Village One Product program, establish rural industries, and foster agro- and ecotourism ventures. These measures are expected to generate employment as well as to expand development in rural areas. Indirect Job Creation During the economic slowdown that has followed the financial crisis, the government has actively supported workers seeking to become entrepre- neurs. It offers a variety of schemes providing loans and grants for both in- dividuals and small to medium-scale industries (SMIs). Funds for SMIs are intended to promote new productive capacity in targeted sectors. MFASURES FOR THE SELF-EMPLOYED. In an effort both to ensure that the self-employed are able to continue their activities during the crisis and to create new self-employment opportunities, the government has allo- cated RM 200 million for petty traders and another RM 214 million for vendor development programs. It has also introduced a venture capital financing scheme and a "graduate entrepreneurs" scheme aimed at devel- oping entrepreneurs. In particular, the programs provide support for self- employment in the distributive trades, including petty trading (vegetable cultivation and livestock rearing, for instance), hawking, farming, and running small businesses. Allocations from the development fund alloca- tion for such activities during the Seventh Malaysia Plan increased from RM 249.4 million to RM 638.7 million (table 4.9). Other financial schemes available to entrepreneurs are the Islamic Fi- nancing Scheme, the Incubator Financing Scheme, and the General Fi- nancing Scheme (run by the Bank Pembangunan Malaysia). Funds are also available through Tabung Amanah Ikhtiar Malaysia, the Small-Scale En- trepreneur Fund, and the Economic Business Group Fund (which includes assistance for women entrepreneurs). The funds are expected to help around 12,000 petty traders and small businessmen set up or expand their businesses. Institutions such as the Board of People's Trustees (MARA), the State Economic Development Corporation, Perbadanan Usahawan Na- I58 EAST ASIAN LABOR MARKETS TABLE 4.9 Funds Allocated to Distributive Trade and Supporting Services, 1996-2000 (millions of RM) Seventh Malaysia Plan Estimated allocation expenditure Balance Program Original Revised 1996-98 1999-2000 Fund for petty trades - 200.0 - 200.0 Upgrading and modernization of trade services 28.8 30.8 10.1 20.7 Business premises 212.7 400.0 130.1 269.9 Trading and constancy services 7.9 7.9 7.9 5.8 Total 7.9 638.7 142.3 496.4 - not available. Source: Government of Malaysia, Midterm Review of Seventh Malaysia Plan, 1996-2000. sional Berhad and Perbadanan Nasional Berhad (PNS), and the Urban De- velopment Authority have similar programs. During the economic slowdown, the government further assists fran- chisers and franchisees with the Franchise Financing Scheme and the Small- Scale Entrepreneur Financing Fund.5 The Franchise Development Assis- tance Scheme assists entrepreneurs in starting a business and developing a product. Measures for Small and Medium-Scale Industries In Malaysia firms with paid-up capital amounting to RM 500,000-1.5 million are considered SMIs. SMIs make up a significant portion of the manufacturing sector. At present, however, their output of intermediate goods is limited, and they operate at a low level of technology. Further, the industrial linkages between SMIs and the manufacturing sector are inadequate. To address these problems, the government implemented eight devel- opment programs in 1996-98 through the Small and Medium-scale In- dustries Development Corporation. These programs include initiatives aimed at creating industrial linkages, developing and acquiring technol- ogy, developing enterprises and markets, and providing financial support MALAYSIA: PROTECTING WORKERS AND FOSTERING GROWTH I59 and skills training. More than 400 SMIs received grants totaling RM 25.1 million for industrial linkage and technology-related programs. Another 527 enterprises received a total of RM 5.0 million for marketing develop- ment programs. SMIs suffered a setback during the crisis because of insufficient demand, increases in production costs, and a credit squeeze. On January 2, 1998, the government established the RM 1.5 billion SMI Fund to promote new pro- ductive capacity and improve the utilization of existing capacity in manu- facturing, agriculture-based, and supporting services industries. In 1998 participating financial institutions approved disbursements of RM 882.23 million for 873 applicants. The fund will run until December 2005 and will provide between RM 50,000 and RM 5 million per customer. All Malaysian-owned SMIs in the target sectors are eligible to apply to the fund. In response to the crisis, on November 23, 1998, the government es- tablished the Rehabilitation Fund for Small and Medium-scale Industries. The fund provides financial assistance during the period of crisis to viable SMIs with nonperforming loans and temporary cash flow problems. The fund has a budget of RM 750 million and is available to all SMIs in the manufacturing, agriculture-based, and services industries. While the SMI Fund and the Rehabilitation Fund are meant to be used by existing busi- nesses, prospective entrepreneurs can use the New Entrepreneur Fund to start a new business. This fund has been in existence since 1989 and has been used by over 2,000 entrepreneurs. Employment Maintenance Neither work sharing nor wage support is practiced in Malaysia, since the layoff rate for skilled workers is not alarming. Despite the absence of a wage support system (either before or after the crisis), workers in certain essential professions in both the public and private sectors enjoy some kind of wage support. Generally these workers receive additional allowances on top of their monthly pay as an incentive to stay in their professions. For example, public-sector employees in certain professions, such as medicine, receive a Critical Service Allowance as well as their regular monthly salary. The Malaysian government has changed its policy on retrenchment be- cause of the economic crisis. In August 1998 the government introduced measures that encourage companies to maintain their workers. These mea- sures include support for retraining workers; recommendations for pay cuts, temporary layoffs, and voluntary separation schemes; and flexible working hours and part-time employment. i6o EAST ASIAN LABOR MARKETS To encourage companies to send their workers for training during the crisis, the government has agreed to reimburse registered employers for the full costs of training (up from the previous reimbursement rate of 75-80 percent). At the end of July 1999, the government had approved 181,032 training centers at a total cost of RM 74.4 million. During the same pe- riod, a total of 806 participants received approval to follow courses at cer- tificate and diploma levels to upgrade their existing skills and knowledge. The government also encourages employers to maintain their workers by reducing salaries. Since August 1998 the Ministry of Entrepreneurship Development has been collecting the data on companies that have fol- lowed this recommendation. Employers appear to prefer voluntary sepa- ration to temporary layoffs. Between August and December 1998 only 4.4 percent of employers opted for temporary layoffs. Another 28.4 percent opted for voluntary separation, with the remaining 67.2 percent choosing pay cuts. Most of the firms opting for voluntary separation were in the manufacturing sector, particularly in areas with an oversupply of women workers. On August 1, 1998, the government introduced flexible working hours and formalized part-time work by amending the Employment Act of 1955. The amendment is aimed at reducing the country's dependence on foreign workers by encouraging more women to participate in the labor force. The government is also contemplating the introduction of "career breaks" that would allow women to leave and reenter the workforce with- out losing their statutory benefits. Table 4.10 shows the actions employers took in 1998-99 when faced with the need to cut labor costs. The most popular course of action among TABLE 4.10 Employer Actions during Recession, Peninsular Malaysia, 1998-99 UJanuary-June) Number of employers Number of employees Type of action taken 1998 1999 1998 1999 Retrenchment 4,485 1,342 75,432 20,292 Salary reduction 795 369 22,516 13,246 Layoffs 52 26 6,342 5,277 Voluntary separation scheme 336 255 5,982 8,152 Total 5,668 1,992 110,272 46,967 Source: Labor Department, Ministry of Human Resources. MALAYSIA: PROTECTING WORKERS AND FOSTERING GROWTH i6i the 4,485 employers was retrenchment, followed by salary reductions and layoffs. Retrenchments seemed to slow down somewhat in the first half of 1999, with 1,342 employers retrenching a total of 20,292 workers. In 1998 some 41,877, or 55.5 percent, of the retrenched workers were male. Of this group 9,415, or 22.6 percent, were professionals and tech- nical workers. But almost half of the female workers retrenched in 1998 were skilled and semi-skilled production workers. The same pattern ap- peared in 1999. The only data available on retrenchment and related activities are com- piled from the mandatory reports filed by employers. These reports cover only manual workers (regardless of wage level) and other workers with monthly wages of not more than RM 1,500 (the definition of employee given in the First Schedule of the Employment Act of 1955). Employers retrenching other categories of workers are not compelled to report to the Labor Department. Income-support Activities As has been noted, unemployed Malaysian workers do not receive any form of income or in-kind support. However, employed workers are protected against old age and incapacitation and their survivors cared for in case of death (from any cause) through a compulsory saving scheme, the Employ- ees Provident Fund (EPF). The EPF is financed directly by monthly con- tributions from both employers and employees (12 and 11 percent, re- spectively, of the worker's monthly wage). Employees can withdraw their outstanding deposits with the fund ei- ther in full or in part, depending on the circumstances. Employees who are incapacitated, intend to emigrate, or have reached the age of 55 can withdraw the full amount, as can relatives of an employee who dies. Em- ployees may also withdraw part of their deposit when they reach 50 and when they need money to build or purchase a house, repay a housing loan, invest in funds managed by approved fund management institutions, or pay for medical treatment for specified illnesses. Employees are also protected against the contingencies of industrial ac- cidents, occupational diseases, disability, and death from any cause through two social security schemes, the Invalidity Pension Scheme and the Em- ployment Injury Insurance Scheme. These schemes apply to all private- sector employees earning less than RM 2,000 per month. The Workmen's Compensation Act of 1952 requires private-sector employers to insure legally employed foreign employees against the same contingencies. I62 EAST ASIAN LABOR MARKETS In-kind Support Through collective agreements, some employees enjoy benefits such as private health insurance, car and housing loans, creches, prayer rooms, and all kinds of leave, including leave for pilgrimages, prolonged illnesses, and drug rehabilitation and leave in consideration of long service. With the exception of private health insurance, public-sector employees enjoy the same benefits. Severance Pay Severance pay in Malaysia is known as termination benefits and is regulated under the Employment Act of 1955. Employers are required to pay this benefit to any employee who has been employed under a continuous con- tract for a period of not less than twelve months. The amount of the ter- mination benefit the departing employee receives varies from 10 to 20 days of wages for each year of service. During the crisis termination ben- efits have been extended to cover employees who leave their jobs under a voluntary separation scheme. Labor Market Services For more than two decades Malaysia has had employment services that match workers' qualifications with employers' specific requirements. The job-matching process is intended to reduce the length of time it takes workers to find jobs, contributing to the efficient functioning of the labor market while making the best use of available human capital. The De- partment of Manpower of the Ministry of Human Resources is responsi- ble for employment services, including registration and placement, coun- seling, occupational guidance, and technical training for workers. The department also registers and regulates private employment agencies and collects and disseminates employment information. It maintains regional and state Manpower Offices throughout the country and has its own web- site where vacancies are posted. Public Employment Services The Department of Manpower's Public Employment Services Section provides free employment assistance to domestic job seekers and employ- ers. It is assisted by four units: the Operational Unit, the Private Employ- ment Agencies Licensing Unit, the Enforcement Unit, and the Occupa- MALAYSIA: PROTECTING WORKERS AND FOSTERING GROWTH I63 tional Guidance Unit. The section not only registers and places job seek- ers but also provides employers with lists of qualified candidates (usually within two working days of a request). Private Employment Agencies Private employment agencies cater to both domestic and foreign job seek- ers. These agencies complement the registration and placement activities of the Public Employment Service. Based on the services they provide, private employment agencies in Malaysia can be grouped into three categories: * Regular private employment agencies, which cater to white-collar and blue-collar workers. They provide placement services for local and overseas job seekers and charge a registration fee; * Consultant management agencies, which cater to top managerial and professional job seekers; and * Labor contracting agencies, which cater to certain industries, usually domestic and construction work. As of the end of 1998, a total of 898 licenses had been issued under the Private Employment Agencies Act. Some 324 of the registered agencies were active, and 574 were considered inactive. Table 4.11 shows the reg- istration and placement activities of the private employment agencies based on the monthly reports received by the Manpower Department. Prior to 1998 the government required private employment agencies to submit reports on registration on a quarterly basis and reports on place- ment on a monthly basis. But since the crisis these agencies have been re- quired to submit monthly reports for both activities. Demand for the ser- vices of private agencies has increased markedly and was particularly high in 1998, when 29,135 job seekers registered-an increase of more than 100 percent over the previous year. But in the first half of 1999, the number of job seekers using the agencies declined to 6,773 (compared with 19,398 in the first half of 1998). This decrease in demand could reflect either of two scenarios. First, some job seekers may have become discouraged and with- drawn from the labor force as the probability of finding a job fell and the expected value of wages declined to less than the reservation wage. Married women and teenagers, in particular, may have given up the job search alto- gether in favor of nonmarket activities such as looking after children, pur- suing further studies, or just consuming leisure. Second, the decline reflects Malaysia's nascent economic recovery, which was evident in a positive eco- nomic growth rate of 4.1 percent for the second quarter of 1999. 164 EAST ASIAN LABOR MARKETS TABLE 4.11 Malaysia: Registration and Placement by Private Employment Agencies, 1997-99 Uanuary-June) Registration Placement 1997 1998 1999 1997 1998 1999 January 911 1,319 71 790 493 February 0 816 61 847 206 March 1,047 9,595 2,382 131 880 372 April 412 1,552 84 730 268 May 345 196 90 809 53 June 1,016 8,135 508 87 769 578 July 142 148 911 August 1,589 201 627 September 2,356 6,493 295 596 October 1513 565 610 November - 834 - December 9,587 - 915 - Total 14,006 29,135 6,773 3,482 7,569 1,970 - not available. Source: Ministry of Human Resources, Manpower Department. The placement rate of private employment agencies does not appear to be high. Of 29,135 job seekers in 1998, a total of only 7,569 job seekers, or 26 percent, were reported as successfully placed. The actual perfor- mance of private employment agencies could be evaluated by calculating the overall penetration rate. But the lack of both data on vacancies and a detailed breakdown of registrants, vacancies, and placement makes such an assessment impossible. Services for Retrenched Workers The government provides a number of services for retrenched workers, in- cluding registration and placement through the Public Employment Ser- vice. It also provides a job link program, occupational guidance, and job counseling. REGISTRATION AND PLACEMENT. Since the government introduced reg- istration and placement programs for retrenched workers in January 1998, a total of 17,622 workers, or 29 percent of all those retrenched, have reg- MALAYSIA: PROTECTING WORKERS AND FOSTERING GROWTH I65 istered with the Public Employment Service. Of these, 12,062, or 68.4 percent, were successfully placed in 1998 (table 4.12). The other 53,333 retrenched workers-71 percent-cannot be traced. Presumably they have been absorbed by other firms or have obtained jobs through other channels, such as advertising or personal contacts. About 16 percent of the retrenched workers were high-level professional workers, and 15,042, or 19.9 percent, were skilled production workers. These groups are more likely than others to find new positions without having to turn to the Pub- lic Employment Service for help. The performance of the Public Employment Service in registering and placing retrenched workers improved significantly in the first half of 1999. The percentage of untraceable workers has fallen from 83 percent in the first half of 1998 to 58 percent. In the same period, the service successfully placed 83 percent of the registered retrenched workers, up from 71 percent in 1998. Publicity campaigns may be partly responsible for this improvement. JOB LINK PROGRAM. When the Public Employment Service receives no- tification of an impending retrenchment, it informs registered employers searching for workers about the potential supply of labor. This program has had some success. For example, in July 1998 some 713 of the 4,128 workers retrenched by a company in Penang were rehired immediately. A total of 2,203 of those workers have been reemployed by other companies, while the rest have turned to self-employment or withdrawn from the labor market. OCCUPATIONAL GUIDANCE AND JOB COUNSELING. Since the crisis the Public Employment Service has continued to play its traditional role in providing occupational guidance and job counseling services for job seek- ers and retrenched workers. However, its emphasis has shifted. As well as providing job seekers with information about wages, it offers information about the current labor market situation. Job seekers are briefed on em- ployers' requirements for new hires, industries producing the most new jobs, and available training. To improve the capabilities of those providing guidance and counsel- ing, the Manpower Department organized a number of training seminars and courses for them. In 1998 a total of 830 officials and staff from the Manpower Department were trained overseas and locally. As of December 1998, the Manpower Department had conducted a total of 103 occupational guidance talks throughout the country with 55,318 participants and had held 9 job fairs. The activities took place at headquarters and at sites throughout Malaysia. TABLE 4.12 Registration and Placement of Local Retrenched Workers, 1998-99 U anuary-June) 1998 1999 Alternative Alternative Month Total employment Untraced Registrants Placement Total employment Untraced Registrants Placement January 1,654 1,653 1 2,613 104 1,820 689 376 February 6,590 40 6,498 52 29 4,565 277 2,387 1,901 1,679 March 9,945 2 9,665 278 247 2,496 355 889 1,252 1,174 April 6,028 2 5,959 67 17 1,092 136 639 317 207 May 5,591 20 5,504 67 18 2,135 91 1,396 648 500 June 5,391 355 4,071 965 720 1,022 28 777 217 213 July 10,488 2,307 4,504 3,677 2,608 1,047 21 797 229 207 August 6,427 112 5,006 1,309 830 September 6,099 246 1,040 4,813 3,749 October 6,434 541 2,397 3,496 2,064 November 5,083 120 3,372 1,591 678 December 5,169 199 3,664 1,306 1,102 Total 74,899 3,944 53,333 17,622 12,062 14,970 1,012 8,705 5,253 4,356 Source: Manpower Department, Annual Report, 1998. MALAYSIA: PROTECTING WORKERS AND FOSTERING GROWTH I67 Labor Market Information In Malaysia a number of agencies are responsible for collecting and dissem- inating labor market information and formulating labor market policies. These agencies include the Ministry of Human Resources, the Department of Statistics, the Ministry of Education, the Public Service Department, the Economic Planning Unit, and the Malaysian Administrative Moderniza- tion Manpower Planning Unit. The economic recession following the financial crisis has created addi- tional demand for new sources of data that are essential to implementing successful remedial policies. The labor market information system during the recession must be related to the policy framework for employment cre- ation for target groups such as the unemployed, the poor, and retrenched workers. Successful planning, implementation, and monitoring depend on having detailed information on the characteristics of labor and the structure of labor markets. To meet this requirement, the Labor Depart- ment of the Ministry of Human Resources is producing a statistical report on retrenchment to facilitate policy formulation during periods of eco- nomic recession. The National Recovery Plan provides the most comprehensive labor market information and has been one of the main public policy responses to the economic downturn. It is the result of collaboration and coordina- tion between those who produce labor market information and those who use it. The government put together the plan with organizations represent- ing universities, employers, workers, and consumers. Among other things, the plan contains information on the causes and effects of the economic cri- sis, the policy response and its objectives, and actions that will be taken. Labor Market Regulation Employers in Malaysia are restricted in the actions they can take against employees. Since the financial crisis, laws have been amended to cover in- creasing numbers of retrenchments. First, however, it is important to make a distinction between dismissal and retrenchment, two common modes of terminating a contract of service between employers and employees. Em- ployers cannot terminate a contract of service without proper cause, as pro- vided by Section 5(2)(a) of the Industrial Relations Act of 1967. * Dismissal refers to a disciplinary action taken against an employee who is found guilty of industrial misconduct. Section 14(1)(a) of the i68 EAST ASIAN LABOR MARKETS Employment Act, 1955 states that "an employer may, on the grounds of misconduct . . . [and], after due inquiry, dismiss the employee without notice." Industrial misconduct refers to actions related to duty, discipline, and morality. In cases of industrial misconduct an employee's contract of service can be terminated immediately, without a prior notice from the employer. TERMINATION BY RETRENCHMENT. Under the Industrial Law, retrench- ments, redundancies, and reorganizations all fall into the same category. Companies retrench workers for various reasons, including a decline in aggregate demand during a recession, closures, mergers and takeovers, and technological changes that introduce labor-saving techniques of produc- tion. Employers wishing to retrench employees are required to serve them with a notice of termination. The termination takes effect on the expira- tion date of the notice. The Employment Termination and Layoff Benefit Regulations of 1980 (under the Employment Act of 1955) govern benefits for terminated em- ployees. The amount of a termination benefit is determined by the terms of the contract of service. However, it should not be less than the amount pro- vided for in the regulations. The Employment Act of 1955 was amended on August 1, 1998, to prevent inappropriate retrenchments and unlawful dismissals and to promote fair labor practices. REGULATIONS GOVERNING LABOR MOBILITY. Malaysia has no regulations on occupational and internal geographical mobility. The government pro- motes the migration of workers from west Malaysia to the state of Sabah. Both the Public Employment Service and the Malaysia Migration Fund Board (the statutory body within the Ministry of Human Resources re- sponsible for promoting geographical mobility) foster worker mobility. The Migration Fund Board oversees transfer schemes that give grants and allowances to workers migrating to Sabah. This program has gained in im- portance because the migrant workers from Peninsular Malaysia who once worked in Sabah are no longer willing to make the move. Working condi- tions on the plantations of Peninsular Malaysia are favorable, and the cost of living in Sabah is high. As the number of Malaysians emigrating to Sabah decreases, the number of workers from Indonesia and the Philip- pines willing to travel there increases. The government's program is de- signed to make the area more attractive to Malaysians. MALAYSIA: PROTECTING WORKERS AND FOSTERING GROWTH I69 Malaysian law classifies international labor mobility into two cate- gories: Malaysian workers seeking foreign employment, and foreign work- ers seeking employment in Malaysia. Private employment agencies play an active role in recruiting Malaysians for foreign employment, and their ac- tivities in this regard are regulated by the Private Employment Agencies Act of 1981. Foreigners seeking employment in Malaysia are placed in one of two categories: semi-skilled and unskilled foreign workers, and expatri- ates. They are governed by Malaysia's Immigration Guidelines for Foreign Workers (1995). Assessing the Public Employment Service The Public Employment Service in Malaysia has been viewed as one of the instruments for achieving and maintaining the goal of full employment set out in the National Labor Policy. The service mobilizes the country's man- power and improves human resource utilization through its traditional function as intermediary between employers and employees in job search and placement activities. Since the crisis, demand for the programs the service offers has risen dramatically (table 4.13). The number of new reg- istrants increased by 32.7 percent between 1997 and 1998, although it fell TABLE 4.13 Malaysia: New Registrants, Vacancies, and Placement in the Labor Market, 1992-99 Uanuary-July) Penetration ratea Year New registrants Vacancies Placement (percent) 1992 102,644 83,300 27,060 32.5 1993 102,624 77,031 25,901 33.6 1994 94,045 57,410 23,425 40.8 1995 85,768 58,412 22,552 38.6 1996 84,736 57,539 25,134 43.7 1997 91,491 64,463 26,233 40.7 1998 121,396 74,610 21,170 28.4 1999 76,074 57,214 12,513 21.9 a. The penetration rate is the total number of placements expressed as a percentage of total hires. Source: Manpower Department, Ministry of Human Resources, Malaysia: Labor and Human Resources Statistics, 1992-96. 170 EAST ASIAN LABOR MARKETS TABLE 4.14 Malaysia: Vacancies by Sector, 1996-99 (January-June) Sector 1996 1997 1998 1999 Agriculture, forestry, hunting, and fishing 1,993 1,855 5,231 10,991 Mining and quarrying 122 109 188 64 Manufacturing 36,107 40,645 52,159 25,645 Electricity, gas, and water 178 200 122 13 Construction 2,755 3,785 2,156 1,450 Wholesale and retail trade, restaurants, and hotels 6,159 6,420 5,281 2,354 Transport, storage, and communication 2,245 2,555 1,066 800 Finance, insurance, real estate, and business services 3,264 3,868 3,070 1,628 Other services 4,716 5,026 5,337 2,549 Total 57,539 64,463 74,610 45,494 Source: Manpower Department, Ministry of Human Resources. again in the face of the nascent recovery in 1999. This large increase in the number of job seekers signals a manpower surplus and a loose labor market. It also reflects the "added-worker effect" that occurred as the re- cession encouraged secondary workers to join the labor market to main- tain household income. Despite the increase in unemployment, the number of vacancies also rose (from 64,463 in 1997 to 74,610 in 1998, or 16 percent) because cer- tain sectors, such as agriculture, are experiencing labor shortages. Most of the new vacancies have been in the agricultural sector, which has seen its need for workers double each year since 1997 (table 4.14). The increase in vacancies in the sector is the result of the opening up of oil palm estates in the districts of Miri, Bintulu, and Sibu in Sarawak. The manufacturing sector also recorded an increase in new job vacancies of 28.3 percent in 1998. This increase may originate with export-oriented companies that are responding to Malaysia's improved export performance. Despite the increases in the number of registrants and vacancies re- ported in 1998, the number of placements has decreased. The percentage of total vacancies filled by the Public Employment Service, or the overall penetration rate that measures the effectiveness of placement services, is MALAYSIA: PROTECTING WORKERS AND FOSTERING GROWTH 17I low. It stood at just 28.4 percent in 1998 and fell to 21.9 percent in the first half of 1999, compared with 40.7 percent in 1997. The penetration rate by occupational categories also varies from a high of 70.7 percent for clerical and related workers to a low of 3.3 percent in agriculture. The limited number of vacancies available relative to the number of job seek- ers makes the Public Employment Service's job more difficult, contribut- ing to the low overall penetration rate. The extremely low penetration rate in the agricultural sector can be traced to the unwillingness of workers to take agricultural jobs, particu- larly on the plantations. Relatively low wages in the sector result in a high incidence of rural-urban migration, leading to labor supply constraints in the plantation sector. The low percentage of vacancies filled in other job categories is the result largely of the mismatch between job seekers' aspi- rations and characteristics and employers' requirements. When unemploy- ment is relatively high, employers tend to look for workers with more experience and skills than would usually be required, and employment ser- vices can face difficulties in meeting demand. Perhaps for this reason, re- trenched workers have a high placement rate. In 1998 the Public Em- ployment Service successfully placed 68 percent of all retrenched workers. Most active registrants seeking employment with the Public Employ- ment Service are already unemployed. In 1998, 80.2 percent of the active registrants were unemployed, compared with 69.9 percent in 1997. In 1999 that figure increased by 10.3 percent. Further, in 1998 only 6.7 per- cent of the active unemployed in the country registered with the Public Employment Service, so that calculating the unemployment rate in terms of all Malaysians is virtually impossible. TABLE 4.15 Malaysia: Active Registrants, by Employment Status, 1997-99 U(anuary-June) Employment status 1997 1998 1999 Unemployed 194,071 298,250 197,187 Employed 70,287 55,441 35,847 Self-employed and family workers 13,251 18,203 14,966 Total unemployed in the economy 233,100 443,200 Unemployed active registrants (% of total unemployed) 8.3 6.7 - not available. Source: Manpower Department Bulletin, various issues; Treasury Report 1998/99. 172 EAST ASIAN LABOR MARKETS Assessing the Impact of Regulations on Retrenchment The Malaysian government has passed new regulations or revised existing statutes to protect workers in the aftermath of the Asian financial crisis. While it is difficult to assess the exact effects of these changes, statistics show that termination benefits are being paid. Some employers, however, are taking advantage of the economic slowdown to replace local workers with cheaper foreign labor. TERMINATION BENEFITS. According to the regulations, in 1998 a total of RM 301.5 million should have been paid to 56,292 retrenched work- ers (table 4.16). Of this amount, 3,061 employers disbursed RM 249 mil- lion, or 83 percent of the required sum. Even though the law states that the benefit must be paid in full within 7 days of the contract termination date, in some cases the benefits have not been paid 10 months after work- ers have been retrenched. However, the figures show a high degree of com- pliance among employers undertaking retrenchments. UNFAIR DISMISSAL. Workers may be dismissed unfairly during the re- cession for one of two reasons. First, employers may wish to replace local workers with cheaper foreign labor. And second, employers may simply neglect to comply with the law, dismissing workers without prior notice or without complying with the principle of "last in, first out." The first reason is more commonly given for dismissing workers. Some employers in the manufacturing sector have seen the recent economic downturn as an opportunity to substitute foreign for local workers. Since foreign work- ers are willing to accept lower wages than local workers, these firms are able to reduce their labor costs. COLLECTIVE BARGAINING. In 1998, 442 trade disputes involving 93,530 workers were reported, a decrease of 3 percent from 1997 (table 4.17). Of these disputes 175 (40 percent) concerned mainly the terms and condi- tions of collective agreements. Although deadlocks in collective bargain- ing accounted for only 112 (29 percent) of the total disputes reported, they affected almost 55 percent of total workers. Disputes involving un- fair labor practices represented only 5 percent of the total disputes in 1998, but this figure represents a 600 percent increase over 1997 levels. Many of the 1998 disputes (216, or 49 percent) occurred in the manu- facturing sector-the only sector to experience an increased number of disputes during the recession. TABLE 4.16 Statutory Benefits Paid to Retrenched Workers, 1998 (Peninsular Malaysia, by state) Number of Number of Amount required Amount settled State employers employees (RM) (RM) Federal Territory of Kuala Lumpur 532 4,740 34,223,114.69 (100) 25,797,200.19 (75) Johore 391 5,327 17,101,673.84 (100) 14,135,585.51 (83) Kelantan 48 982 1,416,289.51 (100) 1,358,518.67 (96) Malacca 84 1,035 3,783,855.09 (100) 3,345,709.68 (88) Negeri Sembilan 144 1,709 6,520,227.52 (100) 5,839,149.32 (90) Pahang 83 2,830 5,593,475.95 (100) 4,629,672.56 (83) Perlis/Kedah 167 4,474 24,671,691.82 (100) 23,684,456.04 (96) Penang 275 15,284 78,967,250.17 (100) 76,021,118.22 (96) Perak 386 4,714 15,516,975.23 (100) 15,045,190.73 (97) Selangor 904 12,449 109,588,602.40 (100) 75,749,915.34 (69) Terengganu 47 2,748 4,154,177.74 (100) 3,298,858.77 (79) Total 3,061 56,292 301,537,333.50 (100) 248,905,374.50 (83) Note: Figures in parentheses denote percentage of total. Source: Labor Department, Ministry of Human Resources. TABLE 4.17 Malaysia: Trade Disputes, 1994-98 Number of disputes Number of workers involved Reasons for disputes 1994 1995 1996 1997 1998 1994 1995 1996 1997 1998 Refusal to enter into collective bargaining 25 25 0 12 19 3,924 2,106 6,356 4,003 2,644 Deadlock in collective bargaining 119 131 137 112 126 53,716 25,437 21,367 39,370 51,209 Terms and conditions in collective agreement and other service contracts 236 254 211 223 175 22,638 12,660 18,181 70,861 30,393 Retrenchment and layoff 7 10 17 4 9 833 634 1,368 151 971 Promotion, reallocation of duties, transfer, and other management prerogatives 28 8 1 17 24 259 55 18 13,426 1,282 Demotion, suspension, warning letter, and other disciplinary actions 23 16 3 17 37 93 60 22 2,242 1,103 Noncompliance with labor standards 15 10 8 7 2 929 2,418 180 565 97 Infringements of workers' rights and unfair labor practices 14 9 22 3 21 1,751 1,225 1,784 9 4160 Other 36 48 57 63 29 3,332 4,062 3,062 8,090 1,671 Total 503 511 476 458 442 87,575 48,657 52,338 138,717 93,530 Source: Industrial Relations Department, Ministry of Human Resources. MALAYSIA: PROTECTING WORKERS AND FOSTERING GROWTH I75 Limitations of Current Labor Market Information As we have seen, Malaysia has several sources of labor market information. However, the collection and dissemination of the data need to be improved. Some of the problems with obtaining the labor market information are: * Delays in the dissemination of information. Delays hinder the effec- tiveness of using data to monitor changes in the labor market. For example, the latest published issue of the Manpower Department's Labor and Human Resource Statistics is for 1992-96. The Labor Force Survey is conducted only once a year, and publishing the data takes at least a year. * Inadequate statistical coverage. Major gaps in statistical coverage of the labor market exist in Malaysia. First, little information is available on the labor market at the local level and on the informal sector. Cur- rently, only information on Peninsular Malaysia is available through the Labor Force Survey and the population census. Second, there is a dearth of information on unemployment and underemployment by season, age, sex, level of education, skills, and family obligations. An- other gap is lack of published information on the characteristics of those involved in the informal sector (sex, economic activity, level of education, skills, and income, for instance). And finally, nothing has been published on the number of workers holding more than one job. * Shortage of wage rate and earnings data. Surveys and reports on wages are not conducted and published at regular intervals. The Occupa- tional Wage Survey, which provides information on salary and earn- ings in selected industries, was initially a biannual survey but has been conducted annually since 1991 (the latest data are from 1993). The Annual Survey of Manufacturing Industries also provides wage data by employment category but at a highly aggregative level. The recent publication of the Fringe Benefit Survey for Executives and Nonexecutives by the Malaysian Employers Federation has helped fill the gap in information on wages and earnings. Vocational Education and Training Education and training have always been an important component of Malaysian economic development. However, vocational training started to receive greater government attention only in the mid-1980s under the Fifth Malaysia Plan (1985-90). The focus on vocational education and 176 EAST ASIAN LABOR MARKETS training (VET) supports Malaysia's efforts to become a more industrial and technology-based economy. Since 1985 VET has focused on high-tech jobs that involve labor- saving technology and robotics work production and on skills related to information technology. This focus supports the government's efforts to create a knowledge-based economy, as outlined in the Seventh Malaysia Plan. Private-sector participation in training has increased, with compa- nies not only providing input but also forming partnerships with public- sector organizations to offer training. More private training institutions have also emerged since 1985, although VET continues to be an activity led by the public sector. The Ministry of Human Resources is responsible for overall human re- source development issues. Its objectives are to contribute efficiently and effectively to the government's efforts to produce a skilled, competent, dis- ciplined, innovative, and efficient workforce with positive values, consis- tent with the nation's industrial needs, technological changes, and eco- nomic growth. To meet these objectives, the ministry has developed several strategies. It aims to: * Establish strategic alliances through consultations and cooperation with employers, workers' organizations, and enterprises to produce a highly skilled and productive workforce to meet the government's planning goals; * Hold dialogues with industrial organizations and employers' associa- tions to encourage greater private-sector participation in skills devel- opment and upgrading; * Expand the capacity and improve the quality of training programs at industrial training institutes; and * Adopt innovative and progressive strategies to market customized courses at the industrial training institutes. The Policy Research and Planning Division of the Ministry of Human Resources monitors policies and strategies within the ministry. The divi- sion has several important functions: * Assessing and identifying training needs for industry and analyzing both public and private training institutes; * Implementing government initiatives related to human resource de- velopment, such as the Cabinet Committee on Training and the In- dustrial Technology Development Action Plan, among others; and MALAYSIA: PROTECTING WORKERS AND FOSTERING GROWTH 177 * Representing the ministry at international and national forums re- lated to human resource planning and development. Along with the Ministry of Human Resources, the Ministry of Educa- tion makes important contributions to VET. A key part of its mandate is to ensure that school curricula are in line with the qualifications workers need in a high-tech labor force. Institutional Framework A number of key players are involved in manpower planning in Malaysia (figure 4.1): * Central planning agencies like the Economic Planning Unit, the Ministry of Human Resources, and the Public Service Department; * Supporting principal agencies like the Ministry of Education, the Ministry of International Trade and Industry, and the Statistics De- partment; and * Agencies directly involved in producing skilled labor, such as the Min- istry of Education, MARA, and the Ministry of Youth and Sports. Other agencies that play a role in human resource development include the National Vocational Training Council (NVTC), which is tasked with establishing a coordinated and comprehensive system of vocational and industrial training in line with Malaysia's economic development goals and technological development needs. At the end of 1998, some 158 in- stitutions had registered with the NVTC. The Manpower Planning Sub- committee of the National Development Planning Committee is respon- sible for the overall manpower policy and determines national education and training policies. These policies are then translated into specific pro- grams by the supporting and implementing agencies. The Human Re- source Development Council (HRDC) has a special role. It encourages re- training and skills upgrading for employees of firms registered with the council. In 1997 a total of 1,181 employers from the manufacturing sec- tor and 405 employers from the service sector registered with the HRDC. Public Vocational Training In 1995-98, the intake of public training institutions increased from 15,428 to 22,630 trainees (table 4.18). Over the same time period, their output (the number of trainees successfully completing courses) increased I78 EAST ASIAN LABOR MARKETS FIGURE 4.1 Manpower Planning Machinery | NDPC-MPS MDB; Manpower Policy Tripartite Proposals Board EPU Overall National Manpower Planning Skilled Labor Principal PSD HR Producing Agencies Support Agencies Manpower Manpower MOE Planning Planning MARA MOE Public Sector Private Sector MOYSMT DOSl Public Sector Agents Determination of Requirements in each Agency PC-MPS: Manpower Planning Subcommittee of the National Development Planning Committee MDB Manpower Development Board EPU Economic Planning Unit PSD Public Service Department MHR Ministry of Human Resources MOE Ministry of Education MOYS Ministry of Youth and Sports MTI Ministry of Trade and Industry DOS Department of Statistics Source: General Circular No. 2, 1986, Government of Malaysia. TABLE 4.18 Intake of Public Training Institutions 1995-2000, by course Average annual 1995 1998 2000 growth rate (%) Course Intake Percent Intake Percent Intake Percent 1996-98 1999-2000 Engineering trades 10,574 68.5 14,333 63.8 16,605 58.5 10.8 7.5 Mechanicala 5,991 38.8 8,508 37.6 9,769 34.4 12.1 7.5 Electricalb 4,323 28.0 5,406 23.9 6,306 22.2 7.7 8.0 Civilc 260 1.7 519 2.3 530 1.9 25.9 1.1 Building tradesd 1,660 10.8 2,312 10.2 3,063 10.8 11.7 15.1 Printing tradese 154 1.0 244 1.1 210 0.7 16.6 7.2 Othersf 2,457 15.9 3,541 15.6 6,338 22.3 13.0 33.8 Skill upgrading 583 3.8 2,100 9.3 2,166 7.6 53.3 1.6 Total 15,428 100.0 22,630 100.0 28,382 100.0 13.5 12.1 Notes: alncludes general mechanics, general machining, tool and die making, motor vehicle mechanics, welding, and sheet metal works and fabrication. blncludes electrical installation and maintenance, radio and television servicing, refrigeration and air conditioning, electrical fitting, and armature winding and engineering. 'Includes construction. dIncludes carpentry and joinery, woodwork machining, bricklaying, and plumbing. eIncludes hand composing, machine composing, offset printing, and bookbinding and letterpress. fIncludes surveying, architectural draftsmanship, photography, laboratory science, dispensing optics, computer programming, information pro- cessing, heavy plant operation, architecture, quantity surveying, hotel and catering and home economics. Source: Government of Malaysia, Midterm Review of the Seventh Malaysia Plan, 1996-2000. i8o EAST ASIAN LABOR MARKETS from 15,844 to 20,898 (table 4.19). Malaysia has four basic training schemes.6 THE SBL SCHEME (TRAINING SUPPORT SCHEME). This scheme allows employers to develop their own training programs relevant to their busi- ness goals. It also permits various modes of training, namely: * In-house training, either on or off the job or both, using internal, ex- ternal, and even foreign resources; * Institution-based training, with public and private training institu- tions offering evening and weekend programs; * Training by employers' or industry associations such as the Federa- tion of Manufacturers Malaysia; * Training by industry-managed training centers such as the Penang Skills Development Center; * Cooperative training, with large companies offering their programs to smaller employers; and * Overseas training when domestic offerings do not meet employers' needs. The PROLUS SCHEME (APPROVED TRAINING PROGRAM SCHEME). This scheme aims to facilitate the retraining of workers through preap- proved programs, sparing employers the burden of applying for approval themselves.Training providers, including government agencies, private or- ganizations, and consultants, are encouraged to register with the HRDC. All training providers must satisfy certain conditions in order to win Ap- proved Training Program status. A committee with majority employer rep- resentation assesses the trainers. PLT SCHEME (ANNUAL TRAINING PLAN SCHEME). Under this scheme employers are encouraged to submit annual training plans covering the next year. Smaller employers may engage training consultants for this pur- pose and have the HRDC reimburse their costs. The PERLA SCHEME (TRAINING PROVIDER AGREEMENT SCHEME). Un- like the above three schemes, which provide reimbursement only after the training has been completed (and fees incurred), PERLA provides the em- ployers with the opportunity to participate in training programs while pay- ing an approved minimum in fees. The HRDC provides the trainer with the balance of the fees, using a fund supported by employer contributions. TABLE 4.19 Output of Public and Private Training Institutions, 1995-2000, by course Average annual growth rate (%) 1995 1998 2000 1996-1998 1999-2000 Course Public Private Total Public Private Total Public Private Total Public Private Public Private Engineering trades 10,758 7,496 18,254 12,857 9,823 22,680 15,014 11,615 26,629 6.1 9.4 5.3 8.7 Mechanical 6,804 1,679 8,483 7,958 2,100 10,058 8,932 2,416 11,348 5.4 7.7 3.9 7.3 Electrical 3,734 5,743 9,477 4,471 7,628 12,099 5,616 9,087 14,703 6.2 9.9 7.9 9.1 Civil 220 74 294 428 95 523 466 112 578 24.8 8.7 2.9 8.6 Building trades 1,792 205 1,997 2,122 255 2,377 2,780 290 3,070 5.8 7.5 9.4 6.6 Printing trades 108 25 133 213 31 244 210 37 247 25.4 7.4 -0.5 9.2 Others 2,623 576 3,199 3,676 718 4,394 3,780 828 4,608 11.9 7.6 0.9 7.4 Skills upgrading 563 - 563 2,030 - 2,030 2,116 - 2,116 53.3 - 1.4 - Total 15,844 8,302 24,146 20,898 10,827 31,725 23,900 12,770 36,670 9.7 9.3 - 8.6 - not available. Note: Does not include output from courses such as commerce, agriculture, home science, and other "soft" skills. Source: Midterm Review of the Seventh Malaysia Plan, 1996-2000. I82 EAST ASIAN LABOR MARKETS TABLE 4.20 Major Public Vocational Institutions Ministry Institution Human Resources Industrial Training Institute (CIAST) The Center For Instructor and Advanced Skills Training (CAST) Education Polytechnics, Technical Secondary School, Vocational School Entrepreneur Development Institut Kemahiran MARA (IKM) Youth and Sports Institut Kemahiran Belia Negara (IKBN) The Ministry of Human Resources and HRDC introduced the Re- trenched Workers Retraining Scheme as a specialized occupational mobil- ity program in May 1998. It is designed to make workers more employ- able by upgrading their skills and to provide employers with highly skilled workers (especially when the economy has recovered). The government has also established a number of vocational institu- tions, some of them under various ministries (table 4.20). Two advanced skill-training institutes have also been set up. The British-Malaysian Institute offers courses in electrical and electronic engineering, medical electronic engineering, communication engineering, computer system en- gineering, and information technology. The Japan-Malaysia Technical In- stitute specializes in computer engineering technology and industrial elec- tronic technology. In addition, France and Germany have signed bilateral agreements with Malaysia to provide VET in highly advanced technical programs. The HRDC has also introduced an 18-month special apprenticeship program designed particularly for graduates and retrenched workers. The Hotel Industry Apprenticeship Scheme prepares workers for the service industry, and the Mechatronics Apprenticeship Scheme provides training in electronics, electrical, mechanical, and other technical fields. The En- trepreneur Development Ministry has set up another apprenticeship pro- gram, the Graduate Entrepreneur Scheme, which trains graduates to be- come entrepreneurs. Private Vocational Training In 1994 around 40 private vocational institutions trained 6,278 workers in 39 different areas. Most of the trainees were in five areas: electrical engi- MALAYSIA: PROTECTING WORKERS AND FOSTERING GROWTH I83 neering (2,193 trainees, or 34.9 percent); microcomputers (785 trainees, or 12.5 percent); mechanical engineering (573 trainees, or 9.1 percent); radio and television (393 trainees, or 6.3 percent); and computer graphics design (287 trainees, or 4.6 percent). Of the trainees, 50.6 percent completed their courses with a diploma. The other 49.4 percent received certificates. The government has developed a scheme that could benefit private vo- cational institutes. The Human Resource Development Fund (HRDF), established under the Human Resource Development Act of 1992, offers employers incentives (in the form of grants) to establish and accelerate sys- tematic training programs. The fund is supported by contributions from employers equal to 1 percent of the monthly payroll. Companies can ei- ther hire trainers or send employees for training at private vocational in- stitutes. The HRDF focuses on three types of training: * Skill areas such as computers, crafts, technology, and professional ed- ucation, as well as specialized training for trainers, research and de- velopment, and basic education; * Management and supervisory training; and * Companywide productivity and quality improvement. To make it easier for employers to train their workers, the HRDF offers distance learning, joint training schemes, and cluster training schemes. The HRDF has pumped RM 15 million into an apprenticeship scheme in machining, mechatronics, tool and die manufacturing, and hotel work, using both on- and off-the-job training. By the end of 1998, a total of 694 trainees had been trained under the apprenticeship program. A Special Skilled Apprenticeship Program has also been developed. Lon-term Perspectives Public and private vocational institutions are producing a growing num- ber of skilled and semi-skilled workers to meet the country's needs (table 4.21). But Malaysia still suffers from imbalances in the labor market, with surpluses in some areas and shortages in others, for two reasons. First, the number of semi-skilled workers is increasing faster than the number of skilled workers. The overall number of skilled and semi-skilled workers increased in 1993-97 in a number of major industries (automotive, electronics, electrical, carpentry, and printing, for example). However, the increase in semi-skilled workers has been larger than the increase in skilled workers, leaving a shortage of highly skilled labor in these industries. I84 EAST ASIAN LABOR MARKETS TABLE 4.21 Expected Demand for Skilled and Semi-skilled Workers, 1990-98 (by industry) Basic Computer- Semi- Industry skills aided design skilled Total Percentage Food 1,271 436 6,160 7,867 1.84 Drinks and tobacco 53 - 520 573 0.13 Textile 1,660 3,831 31,391 36,882 8.62 Wood products and furniture 5,829 3,291 48,884 58,004 13.56 Paper and printing 1,716 180 6,536 8,432 1.97 Chemicals 1,389 1,207 10,880 13,467 3.15 Coal and petroleum 1,173 124 734 2,031 0.47 Rubber products and plastic 2,897 2,056 25,913 30,866 7.22 Nonmineral material 1,558 980 10,531 13,069 3.05 Basic minerals 7,421 207 8,996 16,624 3.88 Manmade minerals 4,153 360 4,956 9,469 2.21 Electrical and electronic 7,811 5,973 174,300 188,084 43.96 Engineering and vehicle tools 8,238 531 18,242 27,011 6.32 Measurement tools 151 42 1,840 2,033 0.47 Other 1,518 773 11,184 13,475 3.15 Total 46,838 19,991 361,067 427,896 100.0 - not available. Source: Ministry of Human Resources, Research and Planning Division, Prospective Em- ployment in the Manufacturing Sector 1990-98. Second, demand for labor and labor absorption of both skilled and semi-skilled workers) are highest in a handful of industries (table 4.21). These include the automotive, electrical, electronics, metal-based, rubber- and plastic-based, textile, and wood-based industries. These industries are the country's primary source of growth. Demand for both skilled and semi-skilled workers in these areas is expected to con- tinue as the industries expand. Based on its projections of future demands for labor, the Ministry of Hu- man Resources has identified the training areas that need to be emphasized: * Electrical and electronics industry group. Computers, mechatronics, and advanced manufacturing technology such as micromechanics, robotics, digital imaging, micro- and nanofabrication. MALAYSIA: PROTECTING WORKERS AND FOSTERING GROWTH I85 * Automotive and transportation industry group. Industrial and engi- neering design, automobile engineering, automotive ergonomics, au- tomotive testing analysis, tools, mold and die, general machining, and machine operation. * Textile and apparel industries. Textile science and technology, fashion de- sign and marketing, engineering design, and manufacturing processes. * Rubber-based industries. Computerized manufacturing, robotics for SMIs, advanced processing techniques, measurement and sensor tech- nology, computer modeling (in compounding), computer simulation in rapid prototyping for design work on rubber engineering applica- tions, precision machining, and precision tool-and-die fabrication. * Wood-based industries. Specialized industrial training in advanced wood-based methods and sciences. Metal-based industries. Metallurgy and metallurgical engineering. Evaluating Public Training Despite the relative success of public institutions in training and retrain- ing workers, the public VET delivery system suffers from several serious shortcomings. Predominant among these is a lack of coordination and haphazard planning that leads to duplications in curricula. This situation is the result of each institution's tendency to promote its own strategic in- terests and the lack of effective overall coordination at the national level. To address this problem, the NVTC acts as a focal point for vocational and skills training. Unfortunately, it does not have the necessary jurisdic- tion to implement a coordinated VET delivery program in public and pri- vate training institutions, although the government envisages giving the council the necessary mandate. The government has also set up a standing committee, the Coordinating Committee for Public Training Agencies, to deliberate on common issues faced by the public skills training provider. But this committee has also proved inadequate to the task of providing an effective delivery system. As a result most training centers have their own separate curricula and types of certification, all of which differ from one another. Most of the public institutions offer similar courses in a few areas (such as heavy trans- portation mechanics, electrical wiring, air conditioning and refrigeration, industrial electronics, and welding) and cover a limited range of skills. This duplication suggests that public training institutions are unable to identify market trends and offer courses relevant to them. Contributing to this situation is a problem discussed earlier: a lack of timely and relevant information. The public agencies involved in planning VET are poorly in- i86 EAST ASIAN LABOR MARKETS formed about the human resource requirements of industries and the state of the labor market. Another serious problem is the quality of the students who attend vo- cational institutions. Most have not succeeded in other educational pro- grams and opt for vocational training as a last resort. Of those opting for VET, those taken in by public vocational institutions are generally less aca- demically qualified than those who attend private vocational institutions. In addition, the students themselves tend to perceive vocational training as the least-best option. VET lacks the status of more traditional educa- tional programs in the Malaysian schools. And public servants tend to earn less than their private-sector counterparts and often have little op- portunity for promotion. Experienced trainers often move to the manu- facturing sector, which offers higher remuneration and benefits, leaving public training institutions without enough qualified trainers. For all these reasons, industries have questioned the quality of the train- ing the public institutions provide. The trainees are not up to industry standards, so that much of the effort that has gone into training them is wasted. Many firms are unwilling to hire graduates of these public voca- tional institutes as skilled workers, taking them on as semi-skilled work- ers. Manufacturers complain that graduates of public training institutions are unable to fit into a high-tech occupational environment. In other words, public training providers are not market driven. Collaboration between public-sector training institutions and industry is imperative in order to correct this problem. Public training institutions need to enhance their ability to respond to rapid technological changes. The tasks of identifying the competencies that are actually required in the manufacturing sector, assessing the competency of trainers and institu- tions, and overseeing certification lie with the NVTC. It has instituted a new five-level skill qualification framework for vocational training and certification to help provide workers with the skills they need and create a systematic career path for skilled craftsmen. The new qualification frame- work will encourage trainees to seek advanced training after completing lower-level courses. The Government Response The government has embarked on a few encouraging initiatives to improve both public and private VET. Among these are the RM 5 million retrain- ing program for retrenched workers and an exemption from HRDC con- tributions for employers in sectors facing severe difficulties. This last mea- MALAYSIA: PROTECTING WORKERS AND FOSTERING GROWTH I87 sure is intended to ease employers' financial burden and thus moderate the level of retrenchment. The NVTC has developed the National Occupational Skills Standards in an effort to standardize training curricula and requirements in certain areas. The accreditation system is based on a continuous assessment of the trainees and has been fully implemented in all public training institutions offering courses covered by the standards. Implementing the accreditation system is essential not only to producing highly skilled manpower but also to developing a career path for skilled workers that runs parallel and is seen as equal to academic-based career development. As part of its efforts to encourage greater private-sector participation in industrial and skill training, the government has raised the industrial building allowance from 2 to 10 percent a year. This allowance helps off- set the cost of constructing buildings that will be used for VET by allow- ing businesses to write off the costs within 10 rather than 45 years. Fur- ther, the government has provided financial support to skill development centers managed by the private sector in the amount of RM 35.7 million. Policy Implications and Recommendations The Malaysian government has responded to the effects of the financial crisis with a number of labor market reforms and initiatives that address unemployment problems. The reforms speak to both the short-term con- sequences of the economic slowdown and the long-term need for an ade- quate supply of labor in all sectors. They also seek to ensure the insensi- tivity of unemployment in the event of future declines in economic activity. Despite these efforts, more can be done in the areas of employ- ment creation and maintenance and income-support activities, labor mar- ket services, and vocational education and training. Employment Creation and Maintenance and Income-support Activities Both the government and private sector must assume responsibility for creating and maintaining employment. The government has two impor- tant roles to play. First, it needs to continue to actively create employment through pub- lic works projects, especially when faced with economic slowdowns and market failures (when the private sector is unlikely to take on certain responsibilities). I88 EAST ASIAN LABOR MARKETS * Second, the government must ensure that the private sector observes all rules and regulations regarding employment, particularly during the pe- riod of crisis. The private sector meanwhile should play a positive role in employ- ment creation during periods of economic boom. During economic slow- downs, private firms should actively seek alternatives to retrenchment in order to maintain employment levels. One way the private sector can fulfill this responsibility is by adhering to government regulations concerning re- trenchment, which offer alternatives to eliminating jobs. The government of Malaysia has repeatedly stressed that it does not need an income-support policy for the unemployed. Besides not wanting to incur the financial burden, the government feels that such a policy would discourage the unemployed from seeking jobs, prolonging the pe- riod of unemployment. As an alternative to unemployment schemes, the government has elected to provide training in skills required by industries and to work to ensure that enough job opportunities are available in the country. However, the government should seriously consider implementing an unemployment insurance scheme to provide financial support for retrenched workers. Such a scheme would minimize the financial burden involuntar- ily retrenched workers face, particularly in the period of crisis. The gov- ernment can design the scheme so that it is financed by worker and em- ployer contributions, particularly from those sectors that tend to be most affected by economic slowdowns or that have high turnover rates among workers. The scheme would also provide employers with another reason to seek alternatives to retrenchment, since firms would be partly responsi- ble for financing the cost of their unemployed workers. The increase in the unemployment rate in Malaysia during the crisis can be considered a temporary phenomenon. Employment should in- crease as the economy improves. In the short run, then, the best policy is to let the market make the necessary adjustments. However, the govern- ment can introduce a series of measures to speed up the adjustment pro- cess. Among these are: * Creating labor market information centers with rigorous employment search services and improving coordination between the government and private sector. * Promoting part-time work and temporary employment in the medium and long term. The objective of such measures is to enhance flexibility in the job market, thus reducing dependency on foreign workers in MALAYSIA: PROTECTING WORKERS AND FOSTERING GROWTH I89 sectors facing labor shortages, such as manufacturing, agriculture, and services. Legislation expanding opportunities for part-time employ- ment has already been introduced. Expanded part-time opportunities will increase women's participation in the labor market and help ensure an adequate supply of workers, particularly in service-related sectors. * Offering part-time workers some of the benefits available to full-time workers will further encourage women to join the labor force. Pro- viding day-care centers for children of employees at the workplace should encourage mothers to reenter the work force and discourage women who become pregnant from quitting their jobs. * Introducing temporary employment can help eliminate the high turnover rate in some sectors.7 Temporary employment needs to be properly monitored to ensure that employers do not abuse this form of hiring. Government guidelines on the hiring of part-time and temporary workers need to be particularly clear, for instance. Experi- ence suggests that temporary workers who are employed for a speci- fied period are reemployed to do the same job. Establishing more temporary employment agencies throughout the country can also help promote both part-time and temporary employment. * Instituting tax and subsidy incentives designed to create andpreserve em- ployment in the long term. Income subsidies can be an incentive to en- courage employment in sectors that are facing labor shortages. Sub- sidies can also encourage employers to hire individuals they would not normally take on, such as young people and the elderly. Labor Market Services The labor market services currently available in Malaysia need to be more effective. As discussed earlier, a number of shortcomings prevent the Public Employment Service from providing the best services to those seeking em- ployment, including a lack of coordination and a shortage of timely infor- mation. The government and private sector can take several steps to im- prove the available services. Increasing the effectiveness of labor market intermediaries would ensure better matches between job seekers and vacan- cies, reducing the time and cost of the job search for both employers and workers. For instance: * Establishing regulations governing the Public Employment Service and private agencies. In Malaysia, the Public Employment Service is the main instrument for organizing the labor market. It issues licenses I90 EAST ASIAN LABOR MARKETS and implements rules regulating the placement activities of private employment agencies. It can promulgate regulations that safeguard the interests of both job seekers and employers and ensure that in- formation is compiled, analyzed, and disseminated. * Improving coordination between public and private employment agen- cies. To expedite the placement of unemployed job seekers in times of recession, public and private employment agencies need to co- operate closely and exchange information. The Public Employment Service is within the Ministry of Human Resources, which has direct access to information on labor market conditions and policies. Pri- vate agencies need to have close contact with the service in order to obtain timely labor market information that will help improve their placement activities. * Enhancing the quality of the Public Employment Service. The effective- ness of the placement service must be judged in qualitative as well as quantitative terms. Quality placement from the employers' side in- volves satisfying their demands for qualified workers. On the employ- ees' side it involves providing well-paying, stable job opportunities. * Requiring businesses to register vacancies with the Public Employment Service. Especially when the number of job seekers considerably ex- ceeds the number of vacancies, compulsory listings can facilitate the job search. These listings also make it easier for the Public Employ- ment Service to assess employment opportunities and match workers with jobs. Such data also provide essential information in terms of available jobs and retrenchment activities. * Placing more emphasis on vocational and employment counseling, espe- cially in periods of recession. When unemployment rates are high, the Public Employment Service needs to encourage unemployed workers to undertake entrepreneurial activities or to become self-employed. Counselors can identify those job seekers with the initiative to ven- ture into business and provide them with financial and management assistance. An employment guidance program can help change job seekers' perception of the job market and shorten the time unem- ployed workers spend seeking work. Such a program helps reduce frictional and structural unemployment. * Reassessing and increasing retrenchment benefits. Retrenchment creates major financial problems for workers, especially those who are sole breadwinners. For this reason retrenchment benefits set out in the Employment Act of 1955 (which was revised in 1983) can be reex- amined and increased. Such a revision will take into account the de- MALAYSIA: PROTECTING WORKERS AND FOSTERING GROWTH I9I terioration in the real value of benefits from inflation. Employees' unions are one of the best ways to ensure that workers get the opti- mum amount of retrenchment benefits (over what is provided in the Employment Act of 1955). Employers are unlikely to agree to such a recommendation. Al- though the government requires them to inform the Labor Depart- ment of intended retrenchments, many apparently do not. Estimating the actual percentage of employers that do report their retrenchment exercises is difficult. One of the reasons employers are reluctant to pro- vide such information is that companies often operate at a loss during the recession and are unable to afford the retrenchment benefits and other payments the law requires. * Increasing the efforts of the Manpower Department and the HRDC to disseminate information on registration, placement, and retraining ser- vices. Only 24 percent of retrenched workers register at employment offices. HRDC figures indicate that less than 1 percent of retrenched workers seek retraining under the Retrenched Workers Retraining Scheme. These low percentages may reflect the fact that many work- ers are not aware that such services exist. Publishing vacancies regu- larly and establishing regular contact with employers, employees, and government agencies to coordinate and disseminate labor market in- formation will also improve labor mobility. * Providing adequate training for staff entrusted with generating labor market information. Successful training efforts require carefully de- signed training programs with appropriate course offerings and ef- fective trainers and course material. Courses such as data analysis are relevant in this regard. Vocational Education and Training The list of recommendations for improving vocational training is long. The government, private sector, and training institutions themselves need to take steps to improve the VET delivery system. These steps include: * Having an independent central agency with the tools and mandate it needs to oversee and guide all public and private-sector VET The NVTC, for instance, cannot function effectively unless it includes not only government and private-sector representatives but also rep- resentatives of trade and workers' unions and manufacturers' groups. This group can draft appropriate programs and requirements for I92 EAST ASIAN LABOR MARKETS VET, coordinate and monitor programs as they develop, and make recommendations to the government. The recent economic crisis has shown that those involved in VET have not studied the impact of the crisis on human resource development in Malaysia. The only organi- zation known to have evaluated its VET is MARA, but only in its branches. * Fully implementing national standards for curricula and certification at both public and private training institutions. A national policy must be in place that requires all vocational institutions to offer courses in line with national objectives. The government has already begun taking such steps with the new skill standards. More should be done, including developing a national assessment system and a national certification system. A National Reference Structure for Vo- cational Qualifications can provide a basis for comparing the various types of diplomas and certificates offered, helping students choose appropriate courses and employers find the types of workers they need. * Encouraging employers to provide training. The government needs a firm directive and guidelines on training for employers in the manu- facturing sectors. These employers need to provide opportunities for trainees to undergo practical training with their companies. * Improving the collection and dissemination of information on labor market trends andfuture demandfor workers. Institutions must be able to base their course offerings on demand-that is, they must design their curricula to meet the needs of the labor market. * Upgrading the quality of the students who enter public (andprivate) vo- cational institutions. Since Malaysia is moving into high-tech, capital- intensive industries, the quality of students selected to undergo skills training must be upgraded accordingly. The task is to set a higher standard of admission requirements and to give VET the same status as other forms of education. The Ministry of Education should try to change the curriculum at the secondary level by giving equal status and significance to vocational subjects. * Encouraging the participation offoreign expatriates and vocational in- stitutions from industrial countries in VET The government should be more flexible in allowing foreign expatriates to work with both pub- lic and private vocational institutions, especially in areas where local expertise is scarce. At the same time, the government must encourage vocational institutions from other countries to operate in Malaysia by providing attractive incentives and infrastructure support. MALAYSIA: PROTECTING WORKERS AND FOSTERING GROWTH I93 * Offsetting costs for both trainees and employers. Costs for training in the most-needed specialties must not be too high, and scholarships, in- frastructure facilities, equipment, and trained instructors must be made available. Constraints on space, staff, budgets, and equipment have often been responsible for the overabundance of courses that promise few prospects in the job market. * Setting up centers of excellence for specific industries. Such centers pro- vide state-of-the-art, industry-specific training that will help develop a work force suited to the demands of the high-tech workplace. These centers will also make the delivery system for VET more pro- fessional and acceptable to employers. * Improving coordination betwveen training institutions and local organi- zations. Each training institution must establish a close rapport with employers in its area in order to gather up-to-date and relevant data on labor market demands. Such data will help in designing courses that meet employers' demands. * Amending course offerings to include nontechnical skills that improve job performance. Areas such as English language competency, creativity, initiative, entrepreneurial know-how, and teamwork are often ne- glected. However, these skills can improve a candidate's chance of ob- taining employment. * Ensuring the competency of training providers. Training providers need to meet certain standards. They need to keep abreast of the latest technological developments in the workplace. They must be effective presenters able to conduct and evaluate the material they are provid- ing. The pace of change in technology and in labor market needs is such that the VET system must itself be capable of rapid change. Modular courses are one way to increase flexibility by allowing stu- dents and employers to select only the modules they need. Incorpo- rating a variety of delivery methods, including on-the-job and dis- tance learning, can also keep training flexible, as can ensuring that decisions are made at the local level so that training is responsive to the needs of local industries. Notes 1. In 1987 and 1997, however, exports fueled the country's growth, and in 1992 public investment played the leading role, recording the fastest growth rate of all the demand components of GDP. 2. The exceptions were 1991 and 1994-97. 194 EAST ASIAN LABOR MARKETS 3. It is difficult to calculate the exact number of jobs created by public work projects, since the private sector is also directly involved in developing these proj- ects. The government administers and funds the projects, and private companies that have been awarded the tender do the actual construction. 4. Public works projects were clustered primarily in four sectors: construction; electricity, gas and water; transportation and storage; and government services. 5. The fund was launched in June 1998 with an initial capital of RM 90 mil- lion. The funds have been disbursed to more than 8,500 applicants. Demand for disbursements has been substantial enough that the government is providing an additional RM 300 million to augment the fund. 6. Total output exceeds input because some courses take longer than one year to complete. Some students may also have been "repeaters." 7. Temporary employment in this context does not include the probation pe- riod for employment, although employees are classified as temporary during this time. 5; The Philippines: Labor Market Trends and Government Interventions Following the East Asian Financial Crisis1 JUDE ESGUERRA, ARSENIO BALISACAN AND NIEVES CONFESSOR IN THE LAST half of the 1990s, the Philippines has expe- rienced both high unemployment rates and low productivity growth. In the wake of the East Asian financial crisis, unemployment and underem- ployment have increased, and the impact on households has been pro- found. This chapter examines the effects of the crisis on low-income groups and the success of the government programs designed to support unemployed and underemployed workers. These programs include in- come support, job creation, loans for small enterprises, and labor market services. The evidence suggests that the majority of these programs need to be strengthened and that targeting needs to be improved. The training and vocational education system, which is also examined here, is likewise in need of reform, and some measures have already been drafted. These improvements and reforms will provide workers in the Philippines with improved security during future crises and shocks. The Situation before the Crisis Unemployment was already high before the crisis, particularly among youth. Productivity was low, in part because of the high reservation wages I95 i96 EAST ASIAN LABOR MARKETS of the unemployed. The crisis only made matters worse: more unskilled workers became unemployed or underemployed, and the country's output became less and less competitive. Firms and workers adjusted by shorten- ing hours, but the impact remained severe. High Unemployment Rates Unemployment rates in the Philippines were already high before the fi- nancial crisis struck in 1997. During most of the 1990s they hovered be- tween a little below 8 percent and 9.5 percent. The El Nifio phenomenon and the financial crisis brought these rates up even more. In Europe high unemployment rates are blamed in part on overgenerous unemployment and social insurance schemes that workers can resort to as an alternative to work. A similar phenomenon may be at work in the Philip- pines, even though the rudimentary programs that are found in the coun- try cannot be compared to the mature social security systems in Europe. On the face of it, high unemployment rates in the country are largely a youth-related phenomenon. Theoretically, reducing unemployment among young people in the 15-24 age group-especially those with some high school or college education-will largely solve the country's unemploy- ment problem (Canlas 1997). But that observation does not yet constitute a diagnosis. In January 1997, before the crisis, only those with a high school education and those who had some undergraduate college credits registered rates of unemployment significantly higher than the overall un- employment rate of 7.75 percent (table 5.1). Unemployment among col- lege graduates was also above the overall rate, but only by less than a per- TABLE 5.1 Educational Attainment and Unemployment Unemployment rates (percent) Jan-97 Jan-98 Jan-99 Overall unemployment 7.75 8.42 8.98 No grade completed 5.94 7.34 7.45 Elementary 5.63 6.13 6.04 High school 10.42 11.14 12.12 College undergraduate 12.78 13.47 14.80 College graduate 8.63 10.22 12.07 Source: Bureau of Labor and Employment Statistics, Current Labor Force Statistics 1999. THE PHILIPPINES 197 centage point. As the crisis progressed, the unemployment rate rose among all workers regardless of their educational attainment but especially among those with college degrees. It would be wrong, however, to conclude that crisis-related job losses were highest among well-educated workers. One hypothesis that deserves closer study posits that the reservation wage of educated young people in search of employment is high because of overseas income from family members and relatives.2 The amount of this type of income compares favorably with the local value added of Philippine manufactured exports and allows young people, especially those from rela- tively well-off families, to endure long periods of unemployment as they look for work. An overly generous social security system would have simi- lar effects. This hypothesis is further supported by the fact that the preva- lence of households that rely on foreign remittances as their main source of income rises noticeably between the lowest and highest income deciles. In 1991, for example, for household heads in the first and second deciles, only 0.7 and 1 percent relied on foreign remittances as their main source of in- come. The percentages for the 9th and 10th deciles, however, were as high as 13.2 percent and 19.8 percent, respectively. The proportion of unem- ployed heads of households who are dependent on remittances increases as income increases. Only 16.8 percent of unemployed households heads in the 1st decile receive remittances,3 but more than half (53 percent) in the 10th decile do.4 In contrast to Europe, however, there is no easy response to the situation such remittances create. Social security benefits can (at least in theory) be reduced or redesigned to ameliorate the perverse effect on unemployment in industrial countries. But the same cannot be said of the Philippines. These remittances from abroad are, after all, private transfers. Given this situation we are able to say that the observed unemployment in the coun- try before the crisis was not really entirely the result of distress-that is, it was not entirely involuntary. A significant portion of it was likely the effect of unemployment among young educated people who had high reservation wages because their families could afford to support them for long periods during the job search. In the long term, when overall productivity and wages have caught up with the reservation wages of this segment of educated youth, unemploy- ment will go down. Productivity growth is the crucial factor that will allow the economy to catch up to the high expectations and reservation wages of these young people and their families. Indirectly we are therefore also say- ing that productivity growth can be linked to unemployment. The ques- tion then becomes one of how to raise productivity.5 198 EAST ASIAN LABOR MARKETS TABLE 5.2 Employment in Agriculture and Services, 1990-98 (percent) Industry shares (percent) 1990 1991 1992 1993 1994 1995 1996 1997 1998 Agriculture 44.97 44.93 45.30 45.71 45.10 43.43 42.84 40.83 39.2 Industry 15.42 15.85 16.15 15.61 15.78 16.13 16.30 16.71 16.4 Services 39.62 39.22 38.55 38.68 39.12 40.44 40.86 42.46 44.4 Source: Current Labor Statistics (DOLE). Productivity and the Reservation Wages of the Unemployed The Philippine labor force was around 30.3 million in 1997 and 31.1 mil- lion in 1998. Among those older than 15, labor force participation aver- aged 65.8 percent from 1987 to 1997. In 1998 it was 66.1 percent, or slightly lower relative to 1996 and 1997. Where did these people work? Table 5.2 tells the story of an agricultural sector that is contracting, an in- dustrial sector that has been stagnant, and a services sector that has be- come the last resort for people who cannot find work elsewhere. In the past productivity growth has stalled because of stagnation in agriculture and the failure of manufacturing to create job openings. The contribution of the manufacturing sector to the country's GDP inched down from more than 27 percent in 1980 to 24.9 percent in 1998. As a result, industry's contribution to employment remained at around 16 per- cent throughout the 1990s. Industry has failed to employ labor that cannot be employed in agri- culture. This fact has many causes, including an inhospitable macroeco- nomic environment that in the past has led to the intensive use of capital rather than labor. Failures in the exchange rate policy of recent years are central to this development. The loss of competitiveness in the tradables sector, in addition to factors such as deficient infrastructure, is the result of the real (at times nominal) appreciation of the peso during the first half of the 1990s (box 5.1). While there was some economic growth before the crisis, much of it occurred in construction, public utilities, real estate, and banking-sectors that are not subject to competition from imported sub- stitutes made cheap by an overvalued currency and the progressive reduc- tion of import tariffs and restrictions. Investments moved toward these sectors because of the loss of profitability in tradables (manufacturing and agriculture). THE PHILIPPINES 199 BOX 5.1 Loss of Competitiveness and Declining Labor Costs Unit labor costs fell in the years before the crisis, but the decline was generated by falling wages rather than by rising productivity. Productivity had in fact been down in a number of key sectors. Rising productivity in the tradables sector was the result of labor shedding (World Bank 1997). It is possible to classify the dif- ferent sectors according to whether they are open to competition or not, and anal- ysis has shown that those most open to competition have lost share in total employment even while experiencing increased productivity (Lim 1997). The wage-setting system the government introduced in the 1990s did not cause rigidities in the labor market serious enough to threaten competitiveness (Reyes 1996a). Compensation indexes had been declining in the years before the crisis, and ironically, in the run-up to the crisis, Philippine wages in foreign currency terms were high and rising. They were in fact on a par with those of Thailand, which had a per capita income almost twice that of the Philippines, and much higher than wages in Indonesia, which has a similar per capita income (World Bank 1997). Since overall compensation in real peso terms has been relatively stagnant in the 1990s, the nearly 40 percent real exchange rate appreciation between 1990 and mid-1997 must bear much of the responsibility for the increased remunera- tion, measured in foreign currency terms.6 Higher wages may in turn have con- tributed to the relative stagnation of traditional manufacturing industries (such as garments and footwear) in recent years. This trend is probably premature, since the Philippines has a large pool of surplus labor and a rapidly growing labor force relative to its major competitors. Productivity also slowed because periodic cycles of boom and bust pre- vented firms and workers from accumulating the know-how and master- ing the technologies that would have allowed them to sell in bigger and more demanding markets and that would have spurred job creation. The economy's tenuous foothold in the demanding export market-the result in no small part of an overvalued currency prior to the crisis-also dimin- ished the demand for product innovation and efficiency enhancement among firms. During the recovery, which is already under way, this initial condition also means that increasing demand in major Asian commodity and component importers such as Japan and the Republic of Korea will have a limited impact on prospects for the Philippines. 200 EAST ASIAN LABOR MARKETS Agriculture, which continues to employ the biggest chunk of workers in the Philippines, has also been suffering from the effects of very low in- vestment rates owing to the prolonged uncertainty over land ownership and the land reform program that was started in 1988. The land reform program was supposed to have been completed in 1998 but at that point had been implemented on only 40 percent of target lands. The country's fiscal problems and the fact that the lands to be distributed in the coming years are relatively expensive make it difficult to be optimistic about the resolution of this problem. When the crisis brought down the overvalued peso, it also inadvertently addressed one of the most important reasons for the economys weak ca- pacity to absorb labor into high-productivity sectors while maintaining its competitiveness. The financial crisis may therefore have marked an impor- tant turning point for the Philippines, even if (on the surface at least) eco- nomic policy has exhibited more continuity than change. Labor Market Flexibility Labor regulations can cause high unemployment rates. Philippine labor regulations, however, including wage setting and retrenchment, are not in general related to the high precrisis unemployment rates. The possible ex- ceptions are regulations on nonregular work-specifically certain prohibi- tions pertaining to labor-only contracting that existed before 1997. Fur- ther, during the crisis both workers and employers found flexible means of keeping firms from shutting down altogether. In the formal sector these methods resulted in underemployment among workers who had not been retrenched. WAGE SETTING. Wage fLxing in the Philippines has been decentralized since 1990 to allow for more sensitivity to regional economic circumstances. Regional Tripartite Wage and Productivity Boards are responsible for set- ting wage levels. The law itself allows for many exceptions, although this fact does not mean that setting a minimum wage will not influence the lev- els at which other wages are set. One study reveals that a significant pro- portion of entry-level employees in the years after the regionalization of wage setting received the minimum wage (Reyes 1996a). The same study showed that in the period after wage setting was regionalized (1989-95), wage adjustments did not contribute to inflationary pressures. The mini- mum wage increases granted by the regional wage boards since the crisis have been below inflation rates. THE PHILIPPINES 201 SEPARATION PAY. The authorities have recently strengthened the en- forcement of separation pay provisions in the Labor Code.7 There is no ceiling on the amount of separation pay an employee may receive, but the Labor Code mandates that permanent workers who are laid off must be given the equivalent of one month's salary for every year of service. Little is known about the consequences of this provision on the behavior of managers, but it did not prevent the Philippines from having the highest level of layoffs among older workers of any country hit by the crisis (Lam- berte, Guerrero, and Orbeta 1999). In the absence of unemployment in- surance in the country, tightening the legal mandate on separation pay was a logical step for the government to take during the crisis. NONREGULAR WORK. In the years before the crisis, prohibitions on "labor- only contracting" created an institutional disincentive to the emergence of a flexible labor market and negatively affected small and medium-size en- terprises (SMEs) and their potential to create jobs. To protect workers' rights, the Labor Code expressly prohibited the contracting out of labor. The regulations also prohibited subcontracting, preventing employment creation through that channel. In 1997, however, the Labor Department revised these regulations.8 The new rules declare that labor contracting and subcontracting arrange- ments are legal as long as the contractor or subcontractor has sufficient capitalization and contractual employees are assured of their rights. The principal employer is jointly liable for any violations of labor standards by the subcontractor. The new rules specify which services can be contracted out or subcontracted: • Work or services that are needed to meet unusual production vol- umes and that the regular work force cannot handle; * Occasional work or services requiring experts or highly technical individuals; • Services for the promotion of new products, but only for the dura- tion of the promotional period; * Services not directly related to the main business or operation of the principal, including janitorial and security work, landscaping, mes- senger work, and tasks not related to manufacturing; * Merchandising or publicly displaying a manufacturer's products, but not selling them or performing any work that involves issuing a re- ceipt or invoice; 202 EAST ASIAN LABOR MARKETS * Specialized work involving unusual or specific skills, tools, or equip- ment; and * Temporary services that substitute for those of a regular employee during a period of absence, including during suspensions but not during legal strikes. The practices of hiring temporary in-house workers through agencies and subcontracting parts of production to outsiders are prevalent in the Philippines despite the legal prohibitions.9 The use of nonregular workers differs considerably across industry groups, however. They are a significant presence in construction, in some segments of manufacturing, and to some extent in the financing, community services, sales, and utility industries. They are used the least in the mining and quarrying sector. The Effect of the Crisis on Labor Market Flexibility Many firms and their workers were able to be flexible in adjusting to the weakness in demand caused by the crisis, although low demand is the main reason for retrenchments and factory closures. The adjustments-in par- ticular reduced hours-ameliorated the impact of the crisis on unemploy- ment, but at the cost of greater underemployment, even in the formal sec- tor. In January 1998 the rate of visible underemployment began to increase over previous years (figure 5.1). The rate had not gone down as of Octo- ber 1998 as it normally does when agricultural underemployment falls. By October 1998 the full impact of the crisis was being felt (indepen- dent of the effects of the drought on agriculture). Both wage and salary workers adjusted by working shorter hours based on various forms of job rationing (de Dios 1999). Daily work hours were reduced, work weeks cut, and workers put on job rotation or forced vacation leave. Because vis- ible underemployment in the formal sector translated into lower pay, firms were able to achieve a flexible downward adjustment in their wage bill corresponding to the business cycle. Industrial relations did not seem to worsen, however. In fact, indicators of strike activity declined through- out the crisis period. Involuntary Unemployment and Underemployment during the Crisis For the poor underemployment is normally a more viable option than outright unemployment. For the whole of 1997, employment continued to rise in all three major industrial sectors-agriculture, manufacturing, THE PHILIPPINES 203 FIGURE 5.1 Visible and Invisible Underemployment Rates, 1996-99 (percent of total employed) 16 16 - |-- Invisible 14 --i Visible | 12 8 6 4 c N. _ u 00 0c 00 Nc Y ¢ O cd ¢ \ Q\ ¢ OO , m' - ~ON N 0 Source: National Statistics Office, Labor Force Surveys. and services-although at a declining rate. This slowing down in employ- ment was one of the first effects of the crisis that began in July of that year. Even in 1998, with GDP declining after the second quarter, the econ- omy still managed to generate slightly more employment than in the pre- vious year. But the number of new entrants to the labor force was greater than the number of new jobs created. Labor Force Surveys for 1998 show that aggregate employment began to decline only in April 1998, at the height of the El Nifio phenomenon. Employment growth picked up after that point, even though GDP for the rest of 1998 declined at an increas- ingly rapid pace before posting a modest recovery during the first quarter of 1999 (figure 5.2). Unemployment and underemployment affected workers most severely in October 1998, more than one year after the crisis struck. The rise in unemployment in April 1998 (to more than 13 percent) was primarily the result of the effects of the El Nifio phenomenon and not of the financial crisis. Throughout 1998 the various subsectors of the economy were still creating net positive employment, though not at a pace sufficient to em- ploy all new entrants to the labor force (figure 5.3). By October 1998 there were 6.7 million underemployed workers in the Philippines, an increase of 894,000 (or 15.4 percent) over the number in the previous year. The October unemployment rate rose to 9.6 percent in 1998, compared with 7.9 percent in 1997, and underemployment was 2.2 204 EAST ASIAN LABOR MARKETS FIGURE 5.2 Employment Creation in 1998 6 GDP growth --Employment growth 2 0. -2 1.97 11.97 111.97 IV.97 1.98 11.98 111.98 IV.98 1.99 -4 Source: National Statistics Office, Labor Force Surveys. percent higher than it had been the previous October. Generally unem- ployment takes a seasonal downturn during this month, reaching its low- est point for the year. In 1998, however, this downturn did not occur- another reason for our contention that October 1998 was the worst month for workers in terms of unemployment. This finding contrasts with the analysis of the Department of Labor and Employment (DOLE), which dates the most severe impact of the crisis to the first quarter of 1998. DOLE relies heavily on administrative reports on closures and re- trenchments in formal sector establishments. UNDEREMPLOYMENT: AN INDICATOR OF DISTRESS AMONG THE POOR. The social impact of the crisis did not necessarily begin to abate once the unem- ployment rates began to go down, as they did in January and April 1999 (compared with the same periods in 1998). In the Philippines underem- ployment remains a serious consequence of the crisis. The country's most de- prived sectors are found among the ranks of the underemployed. We define underemployment as employment that does not pay enough to meet basic needs, so that workers (often unskilled) must search for additional sources of income. Balisacan (1994) shows that self-employed subsistence farmers and fishers, seasonal workers, and informal sector workers-regarded as among the poorest groups-are also among those who are most frequently under- employed. Results obtained by Alba and Esguerra (1999) also suggest that the average worker " would prefer to be fully employed until his unemploy- THE PHILIPPINES 205 FIGURE 5.3 Unemployment and Underemployment Rates, January 1996-April 1999 (in percent) 25 23 19 L 21.5 21.1 28 21.6 21 20.8 17 194 15 -- 7 ot\l \.7os o om c e 7 4 7 9 0 ~~~~. ~~~~~~~~0 c 00 c() Underemployment - Unemployment Source: National Statistics Office, Labor Force Surveys. ment income exceeds 9,000 pesos. Only then does unemployment begin to dominate all the other forms of labor force participation as the worker's preferred mode" (de Dios 1999, p. 21; italics added). Outright unemploy- ment is something that the poor would avoid completely if the choice were up to them. For this group the period of unemployment is thus bound to be short, even if keeping it short means settling for a job that pays poorly. The income deficit that may result is manifested in the search for additional work, even for those who are already employed full time. The phenomenon of full-time workers seeking additional income sources in order to make ends meet is known as invisible underemployment. Underemployed workers are not always poor, although a rise in under- employment often signals difficulty among the very lowest echelons. How- ever, the invisibly underemployed-those working at least 40 hours a week and still in search of more employment-tend to be worse off than the vis- ibly underemployed. The visibly underemployed seek work because they are working part time and would prefer to work more. Unlike the poor, workers who have had some college education are more likely to be unemployed than underemployed. Alba and Esguerra (1999) conclude that additional education possibly raises a worker's reser- vation wage and that the job search tends to be longer for individuals with 2o6 EAST ASIAN LABOR MARKETS TABLE 5.3 Unemployment and Underemployment, April 1999 Number Percent Increase from (millions) of laborforce previous year's levels Unemployed 3.95 11.8 -324,000 Underemployed 6.69 22.7 857,000 TOTAL 11.54 34.5 535,000 Source: Labor Force Surveys. higher reservation wages. Using panel data from the four quarterly Labor Force Surveys for 1998 (5,890 households), Balisacan and Edillion (1999) show that the period of unemployment increases with precrisis income, re- gardless of the sector in which the individual is employed. This result squares with the findings of Alba and Esguerra (1999) and supports the notion that unemployment of a household head is not a good predictor of distress. In this regard, although unemployment had already gone down by April 1999, the number of the underemployed still rose by 857,000 dur- ing that month over April 1998. This scenario indicates that the effects of the crisis on poorer households were persisting and may have worsened (table 5.3). Half the underemployed were in the agricultural sector, al- though output and growth and employment creation in the sector had al- ready resumed (figures 5.4 and 5.5). The Impact of the Crisis on Households The Ministry of Labor's 1998 Annual Poverty Indicator Survey includes two questions pertaining to the crisis. The first asks whether the house- hold has been affected by price increases, job loss, reduced wages, or El Nino. Table 5.4 shows the percentage of households affected by each event, distributed by income deciles that are based on the 1997 Family In- come and Expenditure Survey, which is conducted by the National Statis- tics Office. The number of households affected by price increases and El Nifno varies across income deciles. Price increases affected poor households more than relatively wealthy ones. Job loss within the country and reduced wages affected middle-income deciles the most, while overseas job loss affected more upper-income households than households from other deciles. THE PHILIPPINES 207 FIGURE 5.4 Employment Creation (January 1997-April 1999) 20 - | Agriculture 15 - Industry --Services/ 10- 5 -5- -9 i 4 Ju98Ot. A.99 0 -10 -15 - Source: National Statistics Office, Labor Force Surveys. FIGURE 5.5 Output Growth in the Major Sectors 10 5 0 97.1 97.2 97.3 97.4 \41 98.2 . -5 Agriculture Industry\ / iO- -- Services \ -15 Source: National Statistics Office and Coordination Board. How Households Responded to the Crisis Around half of all households (especially among the poorest deciles) re- sponded to the crisis by changing their eating patterns (table 5.5). House- holds (mostly those in the lower-income deciles) also increased their work- 2o8 EAST ASIAN LABOR MARKETS TABLE 5.4 Impact of the Crisis on Households (percent) Event Job loss of Income decilea Price Job loss in migrant Reduced (poorest) increases country worker wages El Nifio 1 93.5 17.0 3.8 15.4 78.6 2 91.5 16.6 3.2 13.9 72.7 3 90.9 18.3 2.9 15.5 68.3 4 91.7 18.5 4.1 17.1 64.5 5 90.0 21.5 4.5 17.1 61.7 6 90.2 20.5 3.8 16.8 55.0 7 89.7 20.7 4.7 17.1 51.4 8 89.6 19.4 4.8 15.2 45.2 9 88.3 18.3 5.1 14.2 43.5 10 84.7 14.7 4.8 11.2 37.8 Overall 90.0 18.5 4.2 15.3 57.9 a. The sample consisted of 2,315 households per decile, or 23,150 households. These households represent the overlap between the sampling frame of the 1998 Annual Poverty Indicator Survey and the 1997 Family Income and Expenditure Survey. Source: 1997 Family Income and Expenditure Survey, National Statistics Office. ing hours and took their children out of school. Administrative reports and labor force surveys show that high school students were the most likely to be removed from school. The proportion of households receiving assis- tance from friends and relatives was consistently higher than the propor- tion of households receiving assistance from government agencies. More poor households than middle- and upper-income households depended on government assistance, but households in all the income deciles (including the wealthiest) received assistance from relatives and friends. Income Support, Job Maintenance and Creation, and Employment Services The Philippine government responded to the crisis with a number of poli- cies aimed at maintaining workers' income and jobs, creating new jobs, and easing the job search. Specifically, it provided emergency loans, subsi- dized rice, and established public works programs for displaced workers. THE PHILIPPINES 209 TABLE 5.5 Households' Response to the Crisis (percent) Response Received Income Took Migrated assistance Received decile Changed children to city from assistance Increased (1997 FIES) eating out of or other friends! from working (poorest) patterns school countries relatives government hours 1 56.7 12.4 7.8 16.5 10.7 37.5 2 52.3 9.3 5.4 17.1 8.8 36.8 3 50.7 7.3 5.4 16.3 8.4 33.6 4 51.0 8.7 5.2 17.0 6.8 33.1 5 47.8 7.1 4.5 17.2 5.9 29.4 6 48.3 5.6 3.8 16.4 5.7 27.0 7 47.0 5.0 3.7 15.0 4.5 26.1 8 44.1 3.5 3.4 12.5 2.9 22.3 9 41.4 3.2 3.1 13.8 3.9 23.1 10 33.3 1.2 3.5 12.0 2.6 18.2 Total 47.5 6.4 4.6 15.4 6.1 28.9 Source: Panel data constructed from the overlap between the 1997 Family Income and Expenditures Survey and the 1998 Annual Poverty Indicator Survey (21,602 households). See Balisacan and Edillion (1999). The Emergency Loan Facilit for Displaced Workers The Philippines has no unemployment insurance system. However, during the crisis, the Employees Compensation Commission initiated the Emer- gency Loans Program. The program supported the Emergency Loan Facil- ity for Displaced Workers in the amount of 500 million pesos and the Emergency Loan Facility for Displaced Sugar Workers in the amount of 100 million pesos. Ordinary workers who did not usually have access to loans from the banking sector were able to use emergency loans for con- sumption smoothing. The program started on March 25, 1998, under the Social Security Sys- tem. From March 1998 up to May 1999, a total of 433.4 million pesos in loans was released from the displaced workers' fund to 40,491 persons, and 36.6 million pesos in loans was released to 3,481 displaced sugar workers. Displaced workers were able to apply for loans of twice their monthly salaries (up to the 12,500 pesos maximum) at the subsidized 210 EAST ASIAN LABOR MARKETS interest rate of 6 percent per year. The loans were payable within two years, with a one-year grace period. Workers wishing to borrow from the emergency loan program must have been up to date in their payments on any other loans before their separation from their job. Similar funds had been set up for workers enrolled in the Social Security System who were affected by the eruption of Mount Pinatubo in Central Luzon and the electricity shortages of 1991-92. Thus the loan funds were a fairly routine institutional response to a crisis situation. Such schemes are not, however, a replacement for a full-fledged unemployment insurance sys- tem for the private sector that responds to the needs of workers displaced as a result of the normal dynamics of a highly open market economy. As of January 2000 the Social Security System does not even have the flexibility to allow unemployed workers to benefit from past contributions. Such a mechanism would allow workers to borrow from their own future retirement benefits, which they would replenish after finding employment. Under the system's rules, workers who lose their jobs also temporarily lose their access to loans from their future retirement benefits. Because em- ployers must approve loans that use salaries as guarantees, workers who lost their jobs as a result of the financial crisis had no other recourse but the emergency loan facility. Because the DOLE and Emergency Loans Program databases are not linked, identifying the characteristics of displaced workers who took out loans is difficult. What is immediately apparent, however, is that at least 60 percent of those who did are from Manila and nearby provinces. This percentage roughly corresponds to the regional distribution of workers in formal establishments who were affected by temporary and permanent layoffs, job rotation, and other adjustment measures. The loan facilities were not designed to respond to the record unemployment and underem- ployment in the agricultural sector. Thus the profile of beneficiaries cor- responds to those retrenched from formal sector establishments because these workers are monitored by the agencies of the Labor Department. Also, for a worker to gain access to a loan, the DOLE must certify that the potential borrower is an employee of a firm listed in the department's ros- ter of formal establishments that are encountering difficulties. The pattern of formal establishment difficulties does not correspond to the general pattern of distress, however. The regional and sectoral distribution of un- employment and underemployment as reflected in the Labor Force Sur- veys, rather than the administrative reports, is the more reliable basis for assessing social distress. We do not know if contractual workers have been able to access these benefits. Even if some of these workers are enrolled in the Social Security THE PHILIPPINES 2II System, the frequency with which informal workers experience unemploy- ment makes it unlikely that their contributions to the system would be current. In 1998 some 19.25 million of the 32 million workers in the labor force were registered with either the regular Social Security System or the system for government employees. Workers' contributions to the system are automatically deducted from their pay, and employers are sup- posed to remit these payments together with the employer contribution. Many employers fail to do so, however. The compliance rate among em- ployers in 1993 was estimated to be anywhere from 35 to 55 percent (World Bank 1995a). Self-employed workers and workers in family-based enterprises are not usually enrolled. Subsidized Rice The National Food Authority (NFA) administers a subsidy scheme that al- lows it to purchase unmilled rice (palay) at a high price and sell milled rice at a low price. For the last six years, the NFA has been able not only to pro- cure at least 3 percent of the country's total palay production but also to store, process, and even distribute the rice. The government provides the NFA with a direct subsidy of roughly 1 billion pesos annually. But the major portion of the NF's budget actually comes from domestic bank credits guaranteed by the government. The NFA's subsidy scheme is basically a general food subsidy program. Around 90 percent of its price support funds go to rice rather than palay, so the scheme fails as a subsidy program for producers. There are three rea- sons for this failure: i The NFA is unable to influence farm-gate prices. The agency's paddy procurement support prices for 1995-98 were below actual palay prices. The support price was fixed at 8 pesos per kilo in 1996, while farm-gate prices averaged 8.20 pesos per kilo until 1997. * Buying levels have been low and declining. From 1975 to 1996 the NFA procured a yearly average of just 5 percent of total palay production. It was able to exceed its target procurement level only in 1990. * There have been persistent delays in procurement. The NFA usually buys one or two months after the peak season, so only relatively wealthy farmers and traders-those with storage capabilities or access to postharvest facilities-benefit from the agency's program.'0 As a food subsidy program, the NFA's untargeted scheme is also defi- cient. Balisacan (1994) shows that a general food price subsidy such as the 2I2 EAST ASIAN LABOR MARKETS one commonly employed by the NFA in its poverty alleviation efforts is not cost-effective. The price subsidy for rice exhibits substantial leakage of benefits to the nonpoor, who account for about one-half of total rice con- sumption. Arguably the NFA's operations need to be rationalized, and the agency itself has recently moved in the direction of improving the target- ing of the rice subsidy program. Joseph Estrada's government has implemented a food safety net pro- gram specifically for the poor. This new program represents a shift from general subsidies that allow anyone to buy discounted NFA rice to subsi- dies targeting families living below the poverty line. In collaboration with the Department of Social Welfare and Development (DSWD) and the Department of the Interior and Local Government, the NFA in 1999 pi- loted targeted rice subsidy programs in 12 municipalities in the provinces of Antique, Iloilo, Sorsogon, and Surigao. If the program moves beyond the pilot stage, it will require a subsidy larger than the 1.19 billion pesos the NFA received in 1999. Assuming no leakages and no additional ad- ministrative costs, the program would require a budget of around 5 billion pesos (calculated using a 5 pesos subsidy, multiplying it by a per capita consumption of 100 kilos, and multiplying the result by around 10 mil- lion people living below the poverty line)." The NFA maintains a number of programs tailored specifically to the poor, including the Enhanced Retail Access for the Poor (ERAP) program. This program sells cheap basic commodities such as canned goods, cook- ing oil, milk, noodles, rice, sardines, and sugar. By the middle of 1999, 2,784 privately owned stores and 434 market stalls had been set up under the program. ERAP also fielded 384 rolling stores for those in remote and depressed areas. The Philippine government is in no position to implement a general rice subsidy, and in fact one is not needed. The key to increasing the im- pact of the NFA's operations is effective targeting of its interventions. Ad- ministrative costs for targeting, however, can be anywhere from 10 to 30 percent of delivery costs. The most reasonable starting point would be to increase the volume of the NFA's operations in the poorest provinces. Ad- ministratively undemanding schemes such as the provision of low-quality rice may also be needed to reduce leakages to the nonpoor. Finely tuned targeting mechanisms may be considered after these adjustments have been introduced. While rural households are net buyers of rice, this fact should not pre- vent the government from realizing that government interventions in the rice price market will negatively affect farming households or from identi- THE PHILIPPINES 2I3 fying areas that will suffer. For this reason the government may want to consider shifting to food stamp programs that deliver the subsidy to house- holds as a direct income support. Food stamp programs can also reduce the significant leakages that tend to occur when subsidies are implemented in urban areas with high numbers of displaced workers. Direct Job Creation When the crisis unfolded, the Philippines had no sizable public employ- ment programs that could respond appropriately to large-scale economic shocks. Past government infrastructure programs-notably the Commu- nity Employment Development Program (CEDP) during the administra- tion of Corazon Aquino and Kabuhayan 2000 under the administration of Fidel Ramos-had focused on rural poverty. Kabuhayan 2000 was, however, incoherent under the Ramos administration. Subbarao, Ahmed, and Teklu (1996) observe that the program never managed to ensure em- ployment among the poorest groups. In fact, the bulk of the infrastructure program benefited nonpoor areas. Evidence shows that of the two pro- grams, the CEDP was the more successful in terms of reaching poor households in poor regions (Quibria 1996). The CEDP program was intended to hasten the economy's recovery from the deep recession of the mid-1980s. The program aimed to create employment opportunities for an estimated 1.2 million workers by the end of 1988. The emphasis was on small-scale infrastructure projects such as small, often dirt roads, communal irrigation systems, and school build- ings. These projects accounted for nearly 70 percent of the total allocation of 9.1 billion pesos, roughly 90 percent of the approximately 53,000 of the projects supported by the program, and close to 80 percent of the total employment target. Nongovernmental organizations (NGOs) were active in the delivery of these programs. The CEDP, however, often had to delay project implementation-primarily because of slow funding and dis- bursement procedures-and political pressures tended to distort priorities in project selection. Cases of graft and corruption during project imple- mentation were also reported (Quibria 1996). The CEDP, however, achieved results that were favorable compared with the outcomes of previous programs. A 1988 National Economic De- velopment Authority survey covering 11,086 workers showed that a large portion (54 percent) of the workers hired under the program were unem- ployed before the CEDP. In addition 54 percent also had monthly in- comes below the minimum food consumption threshold, 82 percent had 214 EAST ASIAN LABOR MARKETS monthly incomes below the poverty threshold, and 86 percent had in- comes below the minimum wage. The program also aimed to utilize work- ers from the project areas. In this respect the CEDP was again successful, because 78 percent of the workers did come from the project areas. The CEDP was also exceptional among public infrastructure programs in its efforts to evaluate outcomes, and its evaluations showed similar results in almost all the regions. PROBLEMS WITH PUBLIC WORKS PROJECTS. A major problem in employment-creating infrastructure programs in the country (including past food-for-work programs) is the government's policy of compensating beneficiaries with at least the minimum wage. The experience of other countries has shown that a compensation level lower than the prevailing market wage is an important feature of public works schemes, because low wages discourage the nonpoor from participating and preempting slots in- tended for distressed workers. The CEDP was successful under the Aquino administration despite this constraint. The Estrada administration has yet to formulate its own public works strategy. The present pattern of infrastructure provision remains largely an inheritance from the Ramos administration, so that the jobs created in the course of infrastructure provision are largely incidental. Even programs that trickle down to poor agricultural areas, such as the construction of farm- to-market roads, are delayed by complications in budgetary release pro- cesses. The delays significantly reduce the ability of the already limited rural infrastructure programs to provide temporary employment from March to April, the period of highest unemployment and underemploy- ment in the agricultural sector. The "thinness" of rural industrialization in the Philippines means that off-farm employment opportunities are highly covariant with peak farming and harvesting seasons. In other words, the seasonal scarcity of income opportunities on farms is not offset by off- farm income opportunities. Under the new Estrada administration, the only public works initiative as of this writing has been the formulation of a task force. An executive order signed in March 1999 stipulates that the Department of Public Works and Highways (DPWH) and the DOLE will head an interagency task force designed to formulate and recommend steps to increase the labor content of the government's infrastructure programs.'2 Despite official intentions to enhance the ability of infrastructure pro- grams to employ labor, no significant public works programs are available to workers on demand. The initiative that comes closest to an employment- THE PHILIPPINES 2I5 creation program designed to reduce the risk of falling into poverty be- cause of seasonal joblessness is the DOLE's 14 million pesos pilot Rural Works Program (originally the Emergency Employment Program). This program offers jobs in school repair and maintenance projects; drainage and waterworks systems; health and day care centers; roads, bridges, and irrigation networks; and reforestation, cleanliness, and beautification pro- grams. Workers are paid the minimum wage. The DOLE pays 60 percent of the workers' wages, and local government units and NGOs shoulder the rest of the burden. The government also sponsors a food-for-work pro- gram that is spread over 13 cities and has an annual budget of 15 million pesos. The program was started in 1995 but is already in the process of winding down its operations. LABOR-BASED TECHNIQUES. The promotion of labor-based techniques is in many ways an appropriate response to the problem of unemployment and underemployment in rural areas that have suffered from the lingering effects of the Asian financial crisis and the El Nifio and La Nifia weather disturbances. Importantly, the labor-based public works approach would not require new expenditures. This approach is of course not entirely new. The DPWH itself has a "labor-based advisory and training team." In 1994 the DPWH stated that 14 percent of infrastructure programs for that year under Kabuhayan 2000 would use labor-based equipment-supported methods and began providing training in using this approach. These projects, amounting to 1.9 billion pesos, included the construction and maintenance of secondary roads, water supply projects, and flood control structures. Between 1987 and 1993, the Second Rural Roads Improvement Project of the DPWH (funded in part by the World Bank) had a labor-based component that received technical as- sistance from the International Labour Organization (ILO). This compo- nent produced 250 kilometers of small roads between 1987 and 1993 and generated around 1 million person-days of employment in 1988-93. The wage rate adopted for the project was 25-30 percent above the minimum wage. The project was not seasonally targeted, so it may have competed with agricultural actlvltes. Such projects face a number of roadblocks. The first is the perennial delay and uncertainty of fund releases from the Department of Budget and Management. These delays prevent agencies oriented toward rural areas, like the Department of Agrarian Reform (DAR), from initiating all of their infrastructure projects on time during the hot first half of the year. If and when the DAR learns the funds are available, it is usually too close 2I6 EAST ASIAN LABOR MARKETS to the planting season (which is ushered in by the rains) to initiate proj- ects. Labor becomes scarce, and the race against the oncoming rains makes the choice of mechanized methods that require the least amount of labor a very practical one. The second issue is the problem of specifying and in- troducing incentives and regulations that will induce private contractors (for example, in low-cost housing programs) to employ more labor. Sig- nificant trade-offs may exist between efficiency and job creation that need to be studied, especially for categories of public works like housing. Exist- ing guidelines to encourage labor-based production have not been effec- tive because they have not addressed these difficulties. Reforms that succeed in reducing the uncertainty concerning the avail- ability of funds and the timing of budget releases will go a long way to- ward putting rural infrastructure programs in a better position to respond to the seasonal employment needs of the rural poor (see, for instance, World Bank 1995b). Such reforms may also provide the agencies with greater scope for employing labor-based technologies, even if these are not always as efficient or as fast as technologies that require more equipment support. There are other pressing issues-for instance, the skewed distributions of infrastructure projects toward areas that do not necessarily have an ur- gent need for them. The Kabuhayan 2000 experience indicates that as- sessments of the depth of poverty in different parts of the country do not seem to play a defining role in legislative and executive decisions to allo- cate infrastructure across and within regions. This deficiency in the pub- lic investment program persists. For instance, the proposed national bud- get for the year 2000 allocates one-half of the 2.5 billion pesos School Building Program of the Department of Education, Culture, and Sports (DECS) and the DPWH on the basis of student populations in congres- sional districts rather than on the basis of assessed shortages of facilities. A public works act that allots 30 million pesos in infrastructure funds to every congressional district, even those that are highly urbanized, has been receiving wide support in the Congress. The solution to the problems of public works schemes, however, may have to go beyond training more en- gineers in LB-ES methods. It may need to include a change in the politi- cal incentives that underpin the system of intergovernmental transfers and lead to the irrational allocation of infrastructure funds. Livelihood Programs The government has been able to generate employment indirectly through various livelihood and income-generating programs and projects under THE PHILIPPINES 217 the DAR, DSWD, Department of Trade and Industry (DTI), and the Labor Department. Incomplete reports submitted by these agencies show that in 1997 a total of 89,780 individuals benefited from the government's various livelihood and income-generating programs. These livelihood programs, however, have reflected a supply-led credit policy. Consequently they have high default rates and excessive adminis- trative costs. They have also failed to identify and serve the poor and are not sustainable, in part because of the way the loans are structured. The loans are from government agencies and not financial institutions, so bor- rowers have tended to see them as one-time grant opportunities, and re- payment has been low. One inventory showed that only 13 out of 111 programs in existence in 1995 were directed to the ultra-poor (table 5.6). The DSWD is the government agency that has had the most success pro- viding credit for the poor through its Self-Employment Assistance (SEA) program, a livelihood scheme aimed at developing the capacity of the poor- est segments of society for self-reliance and improving their socioeconomic status. The program provides no-interest, collateral-free loans payable in two years, social development assistance, capital skills development assistance, and legal administrative support. Its various livelihood projects benefited some 31,043 persons in 1997. The SEA organizes and supports kaunlaran associations, small local groups designed to provide community-based, self- TABLE 5.6 Government Livelihood Programs Borrowers targeted Small and Salaried medium- Agricultural and self- scale credit Ultrapoor employed enterprises Total Number of programs 39 13 21 38 111 Number of implementing agencies 19 4 12 9 Estimated funds available (billions of pesos) >P11.06 - >P3.2 >P16.7 >P30.9 0-6 to Annual interest market rate (percent) 6-14 rate 3-18 7-24 - not available. Source: Ledgerwood (1995). 2I8 EAST ASIAN LABOR MARKETS managed credit facilities and access to community social services. The associ- ations have around 25 members each and receive technical assistance and seed capital for microenterprises that averages 150,000 pesos (at 4,000- 5,000 pesos per member). Starting with a revolving fund of 65 million pesos in 1993, the SEA Project has provided technical and capital assistance to 3,492 successful associations all over the country and 86,890 member fami- lies that are engaged in various microenterprises, for a total of 337 million pesos. In 1997 alone, the DSWD extended seed capital of 73.7 million pesos to 633 associations with 16,542 member families. DOLE runs a small self-employment scheme, PRESEED, in rural areas. PRESEED, which delivers a package of interlinked services for new entrepreneurs, has been subjected to better-than-average monitoring. The program is directed at landless rural workers who want to learn new en- trepreneurial skills and are capable of providing equity of at least 10 per- cent of the total project cost. Participants qualify only if their incomes fall below the poverty threshold. PRESEED released a total of 117 billion pesos from 1989 to June 1996, funding a total of 11,869 projects and benefiting 27,738 workers. In 1993-94, however, only 7.7 percent of the beneficiaries had been unemployed prior to the program. Some 60 percent had existing livelihood projects or businesses even before they applied for the loans, 19 percent were employed in various firms and agencies, and others were working on personal or family endeavors-an indication that PRESEED could have tapped the wrong target beneficiaries. Support for Small and Medium-Size Enterprises Assessments of the situation before the crisis estimate that deregulating the financial sector made it much easier for SMEs to gain access to formal credit markets (see, for instance, Llanto and Sanches 1997). According to Llanto and Sanches, the 10 new foreign banks granted licenses to operate since 1995 cultivated a client base among SMEs. The study says that this devel- opment was aided by the abolition of the antiusury law, which prescribed limits on the interest rate lenders can charge. The law had the perverse ef- fect of making it difficult for lenders to recoup the higher transactions costs that accompanied lending to small borrowers. With deregulation, even the progressively rising mandatory lending requirements on commercial banks for SMEs (under the country's Magna Carta for SMEs) were exceeded.13 These changes did not mean that any eligible SME that wanted to take out a loan could do so, and SMEs continued to be rationed out of the credit market. Increased access to credit markets was just the start of a process that was bound to take a long time, even if there were no reversals (World Bank 1998). In spite of the progress that had been made, SMEs THE PHILIPPINES 2I9 lost access to credit during the crisis, once again highlighting the problem they have obtaining adequate, affordable credit. The Philippine experi- ence suggests some positive lessons. Using a firm-level survey of 300 firms as well as a control group, the World Bank (1998) showed that between 1985 and 1995, firms receiving on-lent credit from a project in support of SMEs generated more jobs more quickly than similar firms without such credit. Beneficiary firms in the sam- ple created 4,821 new jobs over a decade, or 27 new jobs per firm, while nonbeneficiaries lost 1,400 jobs, resulting in an average of 10 fewer jobs per firm. Moreover the employment generated by beneficiary firms grew more than twice as fast as the Philippine labor force over this period (5 percent annually, compared with the economywide rate of 2.4 percent). The on- lending activities to SMEs were so successful that they can be seen as add- ing, within limits, to the social safety net for workers. Such activities can, for instance, help address high urban unemployment rates (World Bank 1998). The Small Business Guarantee Fund The lesson from the World Bank evaluation cited above did not include the claim that guarantees were effective in preserving and creating jobs. Guar- antees were rarely used in the World Bank on-lending programs in the Philippines, because in cases of default, commercial banks that lacked suf- ficient collateral for their loans were required to go through a long and un- certain bureaucratic process before they could utilize the World Bank guar- antee. Actual payments on claims were made only after two to seven years. The Small Business Guarantee Fund is an offshoot of the Magna Carta for SMEs. It aims to increase the flow of funds from formal lending insti- tutions to SMEs, especially those without collateral. The fund was origi- nally expected to provide large-scale guarantees for SMEs by encouraging financial institutions to lend to these enterprises. It offers a guarantee of up to a maximum of 90 percent on loans to qualified entrepreneurs. The guarantee works primarily as a collateral substitute or supplement. But since its inception in 1993, the fund has had an average volume of guar- anteed loans of only around 400 million pesos. Because the Philippines has many credit and guarantee facilities for SMEs, and efforts to rational- ize these programs are ongoing, guarantee programs such as the fund may not be the most appropriate source of support for these enterprises. Wage Subsidies In some countries wage subsidies have been used to induce firms to train and employ workers who would not otherwise be employed. In Canada for 220 EAST ASIAN LABOR MARKETS instance, wage subsidies coupled with reimbursements on training costs have been used to encourage firms to create slots for the long-term unem- ployed. In the Philippines, the government provides wage subsidies through the Special Program for the Employment of Students and the Work Appre- ciation Program. The program for students helps children of the long-term unemployed pursue their education by providing employment during summers and Christmas vacations. The program places students in public and private es- tablishments, with the employer paying 60 percent of the prevailing mini- mum wage and the DOLE footing the other 40 percent in the form of vouchers. The DOLE spent 174 million pesos on this project in 1997, ben- efiting a total of 86,018 of the 113,939 students registered with the DOLE regional offices and the Public Employment Service. In 1998 the program received 110.4 million pesos and placed a total of 96,557 beneficiaries out of the 118,265 registered. Since the onset of the Asian crisis, children have been receiving priority over displaced workers. This program therefore becomes something of a direct transfer scheme for displaced workers. It does not just displace costs that the private sector might have been willing to incur on its own. Rather, it results in the employment of young people who may not be the most qualified workers that firms could have recruited. The Work Appreciation Program aims to develop a respect for work and ethics in young people by exposing them to actual work situations. The program teaches them work procedures, the use of machines and tools, and other practical knowledge relevant to specific occupations. This pro- gram facilitated the placement of 14,126 youths in different private estab- lishments nationwide in 1998. Passive Public Labor Market Services 14 The Philippines has a number of passive labor market programs in place. They include the Public Employment Service and an automated system that provides applicants with information on jobs. These services are aimed at shortening the job search for the unemployed by providing timely labor market information. PUBLIC EMPLOYMENT SERVICE. In establishing a Public Employment Service Office in every province, key city, and highly urbanized municipal- ity, the government has several goals. It aims to facilitate the exchange of labor market information between job seekers and employers, establish a national manpower skill registry, and serve as a referral and information THE PHILIPPINES 22I center for employment placement. The offices also serve as clearinghouses for the various employment promotion programs and services offered by both government and private agencies. To date, there are 1,825 offices situated in local government agencies, NGOs, and schools nationwide. From 1993 to the first quarter of 1999, a total of 1,685,282 job applicants-or about 62 percent of the 2,722,858 who registered-were placed or otherwise assisted. In 1998 the nationwide network of offices registered a total of 418,617 job applicants; referred 347,606 applicants for wage employment, self-employment, and training; provided employment counseling to 209,641; and tested 23,143. Further linking these offices is a priority on the DOLE's legislative agenda. House Bill 7127, otherwise known as the Public Employment Service Office Act of 1999, is intended to be the enabling law that will ensure a national facil- itation service network in every province and key city in the Philippines.15 PHILIPPINE JOB EXCHANGE NETWORK (PHIL-JOBNET). The Phil-Jobnet, launched in November 1998, is an automated job and applicant matching system that aims to facilitate job seekers' search for work and employers' search for human resources. The Phil-Jobnet maintains information on ex- isting vacancies and the corresponding qualifications required by employers and can be accessed via the Internet. Applicants can open a file containing information such as their age, educational attainment, skills, and other qualifications. Workstations accessible to the public have been installed in 43 DOLE regional offices and local government units nationwide. JOB AND LIVELIHOOD FAIRS. Job and livelihood fairs allow job seekers, employers, and overseas recruitment agencies to meet in one venue on a specific date in order to reduce cost, time, and effort, particularly for ap- plicants. Applicants can ask about positions and receive an on-the-spot in- terview and are sometimes hired immediately. Government agencies are also invited to provide information on self-employment and training as- sistance. In 1998 some 246 job and livelihood fairs drew 47,828 registered job applicants, and 6,442 were hired. In the first quarter of 1999, 67 job fairs were conducted, drawing 13,666 registered applicants and generating 2,694 placements. AssISTANCE FOR DISPLACED ovERsEAs woRKERs. The Placement Assis- tance Program provides employment referral services to returning workers. The program matches workers' skills and qualifications with the require- ments of local employers to facilitate access to employment for this group. 222 EAST ASIAN LABOR MARKETS Private Labor Market Service Initiatives The private sector offers services for job seekers and employers. Traditional private employment agencies recruit for both domestic and overseas posi- tions. In some cases the private sector is trying out innovative approaches. One such approach is a job-matching radio program that provides infor- mation on openings at local firms. MATCHING JOBS ON THE AIR. An excellent example of a private sector initiative in employment facilitation is a public service program offered by 702 DZAS, a Far East Broadcasting Company AM radio station. Since 1986 the program, Action Line, has aired job openings, profiled job seek- ers, and referred workers to positions that suit their qualifications. The radio sponsor tied up with DOLE to provide a format for airing impor- tant public announcements and discussions on labor issues on a regular basis. The program, considered a first in the history of Philippine radio broadcasting, is featured twice each weekday and receives an average of 30 calls daily from job seekers and companies in need of personnel. RECRUITMENT AGENCIES. Recruitment agencies provide employment facilitation services for local and overseas workers for a fee (the fee is charged to the job seeker or employer or is shared by both). The network of agencies and services is an important link between companies and job seekers and reports a significant number of placements. In 1998 alone local agencies recruited about 60,345 job seekers, of which 39,895 were placed. In the first quarter of 1999, these agencies placed 11,397 of 13,927 job seekers. The government monitors and regulates the operations of these agencies. In 1998 it issued 77 licenses to newly established local re- cruitment agencies and renewed 101 licenses for existing agencies. In the first quarter of 1999, 33 new licenses were issued and 11 were renewed. Recruitment agencies for overseas employment numbered about 1,053 as of May 1999. Of these 705 specialize in land-based employment and 348 in sea-based jobs. While no system for monitoring placements made by these agencies is in place, their contribution is clearly significant. The scope of their services can be deduced from the large number of workers who leave for overseas employment via formal or authorized channels. Labor Department Services during the Crisis16 The labor market services available in the Philippines during the Asian fi- nancial crisis may be likened to a half-finished house that suddenly had to THE PHILIPPINES 223 be used for shelter from a storm. In general terms, the available govern- ment labor market services (provided by the DOLE) sought to prevent job losses wherever possible and to make an effective and efficient match be- tween workers and companies seeking them-that is, to help job seekers find employment in the shortest possible time and at the least possible cost. One reason for the temporary nature of the DOLE's safety net pro- grams is that the department has no direct mandate to provide social pro- tection services. Within its limits, however, the DOLE has provided a number of useful labor market services that formed an important, if not entirely effective, safety net for some displaced workers. The most significant set of policies is directed at workers displaced ei- ther as a result of the economic crisis or because of adjustments to the opening up of the economy via trade liberalization. These measures make up a package of assistance that reaches mainly workers in formal sector es- tablishments.'7 Local labor offices try to identify firms in distress early on, before labor underutilization, permanent or temporary layoffs, or a per- manent closure of the whole establishment. In certain instances the field offices are able to help prevent firm closures and layoffs by assisting in ne- gotiations (social accords) that can be crucial in transforming potential solvency problems into more manageable temporary liquidity problems. In most instances these negotiations reduce labor costs through techniques such as job rotation and shortened work weeks, relieving firms of the bur- den of having to provide full payment for partially employed or under- utilized labor. Since the early 1990s the DOLE has had specific guidelines that make these flexible responses to depressed market demand not only possible but also in compliance with the country's labor laws. The DOLE is the third party in conflicts that could result in mutually disadvantageous outcomes for both the other parties. The department provides information about possible modes of adjustment that the two parties acting by themselves may not have considered and, by assuring la- bor that proposals are fair, increases the likelihood that the agreement will be mutually beneficial. And the DOLE acts as a witness to agreements that might not be self-enforcing in the absence of a third party. In the event that layoffs become inevitable, the DOLE provides work- ers with four essential services: * It ensures that workers get the benefits due them by law. * It sees that workers receive counseling and information about job prospects. 224 EAST ASIAN LABOR MARKETS * It provides information on ways to help smooth consumption, help- ing workers gain access to emergency loans and income-augmenting possibilities for children who need to continue schooling. * It helps workers gain access to livelihood possibilities such as tempo- rary employment through public works programs, self-employment through livelihood programs, and retraining. PROBLEMS WITH THE DELIVERY OF LABOR MARKET SERVICES. In imple- menting its social protection and active labor market policies, the DOLE inadvertently concentrates mainly on workers in large formal sector estab- lishments. Many of these workers are among the most highly skilled seg- ment of the labor force and are not usually the poorest workers. Further, they are usually among the most capable of finding new employment. The DOLE is not intentionally biased, but because the department's original mandate centers on industrial relations and not on employment or social protection, the focus on formal sector establishments was inevitable. Agri- culture, for instance, is not within its purview. It also works mostly with large firms that have the capacity to comply with labor standards and vol- untary reporting requirements. Smaller firms have little incentive to seek the DOLE's assistance, not only because they lack the capability to com- ply with labor standards (such as minimum wages), but also because they are unlikely to be able to offer laid-off workers the requisite separation pay. For these reasons displaced workers from smaller firms are not usually in the DOLE's roster of establishments and thus have limited access to DOLE's programs. The same is true for agricultural workers. Aside from assistance for sugar workers, the DOLE has only small livelihood programs and rural works programs to offer. These programs are too limited to make a significant impact in times of large shocks, and no other agency has pro- grams for agricultural workers. The country's existing agricultural pro- grams concentrate mainly on perennial poverty and low productivity, but not transient poverty. Covariant shocks-such as those caused by the El Nifio weather phenomenon and the recent financial crisis-require re- sponses that differ systematically from the present programs for rural areas. The DOLE's programs for displaced workers exclude virtually no one, however, because the eligibility criteria are so general. The department can say that its services are available to everyone, not just to workers affected by economic crises. Job seekers in general may avail themselves of job fairs, job-matching services, and some degree of job counseling; skills acquisi- tion and upgrading (for formal sector workers); employment and liveli- hood assistance (for informal sector workers, including assistance to help them enter formal sector employment); services for returning or displaced THE PHILIPPINES 225 overseas contract workers; and assistance to improve employability (for new entrants to the labor force), including some assistance in making the transition from school to work. Yet because the most significant services are delivered on site to workers in formal sector establishments, an infor- mal screening mechanism is in fact operative. LABOR MARKET INFORMATION AND ITS EFFECT ON SERVICES. The DOLE's programs do not make sufficient use of the labor market information that already exists. The Family Income and Expenditures Survey, the Quarterly Labor Force Surveys, and the Annual Poverty Indicator Survey can be very useful instruments in rationalizing the allocation of funds within the DOLE and other agencies with active labor market policies. The DOLE's most significant programs rely mainly on administrative reports prepared by the Regional Layoffs Monitoring System, which tend to overrepresent large formal sector establishments. At present DOLE officials believe that these field reports provide the best statistics of the impact of the crisis on employment. The department disputes this study's claim, for instance, that the impact of the crisis on unemployment was most marked in October 1998. This judgment is based on administrative reports from the monitor- ing system showing that the lowest number of firm closures for 1998 oc- curred during the fourth quarter. The administrative reports suggest that the worst effects of the crisis were evident sometime in the first quarter of 1998. After October 1998 no major assessments of the impact took place because of the DOLE's confidence in the regional monitoring system. It felt that the quarterly Labor Force Surveys were not timely enough and did not have the right focus for the DOLE's information needs.18 If the DOLE begins paying attention to the most complete sources of labor market information, the department will begin to see that its pro- grams are inadequate relative to the magnitude of the problem and that they usually fail to address the needs of the most distressed workers. This realization should call attention to the need to redesign and scale up pro- grams such as the public works schemes. According to the Institute for Labor Studies, "if given more funds," the DOLE "would want to expand the Phil-Job Net and the rural public works component of the adjustment program." 19 Training and Vocational Education Programs Until the establishment of the Technical Education and Skills Develop- ment Authority (TESDA) in 1994, responsibility for vocational and tech- nical training was divided among several agencies. This subsector overlaps zz6 EAST ASIAN LABOR MARKETS with basic and higher education, as technical and vocational education programs are conducted at both the secondary and postsecondary levels. In 1994 the education department was split into three departments-the DECS, which is in charge of primary and secondary education; the Com- mission on Higher Education (CHED), which is in charge of tertiary and higher education; and TESDA. At that time 207 high schools were trans- ferred to TESDA, of which 163 were subsequently transferred back to the formal high school authority because they did not meet the criteria for vo- cational and technical high schools. The problem of the overlap of technical and tertiary education has ex- isted for many years. CHED-supervised institutions and state universities and colleges account for an estimated half of the spending on the technical and vocational subsector (TESDA 1999). But TESDA does not oversee and regulate these institutions; rather, they are supervised by the CHED, which is not particularly qualified to do so (figure 5.6). TESDA, which is primarily a regulatory agency, does not have a man- date to run training centers, much less schools, on a day-to-day basis (box 5.2). The Director General believes that it is possible to improve the sit- uation of the CHED-supervised institutions without necessarily ceding them to TESDA by placing only those programs that offer technical, FIGURE 5.6 Technical and Vocational Education in the Philippines PRIVATE PUBLIC Enrollment 80 percent 20 percent Type Nonformal' Formalb Formal Funding ^ & Almost exclusive private £ Government subsidies to public sources (commercial establishments, NGOs, schools (high schools and regional nonprofit organizations, and reli- and provincial training centers) from gious groups), mainly through tu- TESDA's regular annual budget, ition fees and endowment income. with contributions (often in kind) 5 Minor exceptions, such as Pri- from some local governments. vate Education Student Financial dh Enrollment charges and subsi- Assistance, which provides public dies for programs under the purview scholarships that benefit 3 percent of state colleges and universities, of students enrolled in private which have their own charters and schools. annual budgets. a Organized programs that are part of the school system. b Organized classes outside the school system. THE PHILIPPINES 227 BOX 5.2 TESDA's Responsibilities Mandated by law Inherited tasks (largely regulatory and facilitative) (provided directly by TESDA) 1. Overall regulation and direction of 1. Direct operation of over 50 high the technical and vocational educa- schools with postsecondary techni- tion system cal enrollment * School-based training 2. Direct operation of network of * Center-based training 59 regional and provincial train- * Enterprise-based training ing centers 2. Regulatory functions 3. In the future: regulation and * Development of industry-based operation of technical programs standards and certification in state colleges and universities * Registration and accreditation of that are at present formally under training establishments the Commission on Higher * Training of trainers Education * Research on the subsector * Promulgation of information about supply and demand for midlevel skills and advocacy for clientele, including enterprises vocational, and middle-level skills under the agency's supervision. Given the fact that state universities and colleges have their own charters and are therefore autonomous, the agency is unlikely to be able to extend its au- thority further. Gaining responsibility for these programs is one of the most urgent reforms on TESDA's advocacy agenda. Another key initiative already listed in the National Technical Education and Skills Develop- ment Plan would set up a system for gathering data so that schools and TESDA itself could assess the effectiveness of training. In 1999 TESDA administered only 2 percent of the education sector's spending. Even when state universities and colleges are included, the tech- nical and vocational subsector in the Philippines is much smaller and less mature than it is in other developing countries in Asia (Maglen and Man- asan 1996). TESDA's main task is not the direct administration of train- ing institutions. TESDA's mandate enjoins it to take on a facilitative, reg- 2Z8 EAST ASIAN LABOR MARKETS ulatory, and standards-setting role (figure 5.6). However, at present, TESDA operates most of the public training institutions outside of the universities and colleges in 14 regional and 45 provincial centers. The vocational schools and centers use up two-thirds of TESDA's budget. Aside from ask- ing for a larger allocation from the national government, TESDA is cau- tiously testing schemes to shed itself of the financial burden of running these institutions. Private versus Public Funding Funding for vocational and technical education is inadequate. TESDA re- ceived only a modest budget of 22 billion pesos in 1997-99. The pro- posed budget for 1999 of 2.2 billion pesos did not include any adjust- ments for inflation; nominally, it was the same as the 1998 budget. The budget for tertiary training is around eight times the budget for vocational education. However, it is unclear how much money universities spend on vocational courses and similar training. Aside from this lack of funds, there are other very good reasons for en- ticing the private sector to take on more responsibility for running the country's technical and vocational schools. Highly subsidized public insti- tutions tend to squeeze out private institutions, which are known to be more cost-effective in delivering services. They have higher utilization rates for both facilities and teachers and employ more teaching than nonteach- ing staff. The higher utilization rates for facilities and faculty may well re- flect the efficiency that derives from the strong link between private pro- viders and the labor market that they serve-in other words, their courses sell better. The emerging consensus appears to be that the scope for gov- ernment financing of training remains, but that the government's com- parative advantage does not appear to be in setting up or maintaining training institutions (see table 5.7). The government's role in financing technical and vocational education is in part the result of the failure of financial markets to operate properly. If banks lend at all, they demand a high premium from borrowers. The private supply of training outside enterprises is constrained by the unwill- ingness of financial institutions to fund training organizations that do not possess adequate collateral. One way to circumvent this problem is to con- sider subsidizing private training facilities directly. The draft National Plan provides for the adoption of a rational scheme of resource allocation with "institutional capability" as the first criteria. The plan involves establish- ing a system of competitive tendering (bidding) for training provision. Al- THE PHILIPPINES 229 TABLE 5.7 Public and Private Training and Vocational Education Institutions (December 1997) Tertiary Technical- institutions, Industry- NGO-based vocational nondegree based training training schools programs centers centers Others Total Private 825 374 64 92 28 1,383 Public 240 182 57 137 107 723 Source: National Technical Education and Skills Development Plan. though the details are not yet in operational form, presumably both pri- vate and public training institutions may bid through this system. Private training and vocational education are financed almost exclu- sively by tuition fees and endowments. The government does not channel financial resources directly through private institutions, with one excep- tion. It operates the Government Assistance to Students and Teachers in Private Education through the voucher system of the Private Education Student Financial Assistance program. The Private Education Student Fi- nancial Assistance scheme is one mode of financing that continues to ad- dress the government's equity concerns, and there is certainly room to ex- pand it. It provides performance-based scholarships for a maximum of three years. Qualifying exams help screen the applicants. These scholar- ships function like vouchers, so that students may choose to enroll in ei- ther private or public institutions. More than 7,600 scholarships were awarded in 1997 and 1998, and a budget for 10,300 scholarships was al- lotted for 1999. The scheme serves two purposes. First, it is a scholarship fund that tar- gets poor students. To qualify, the student's family must have an income below the official poverty line. Vouchers for each region are allocated ac- cording to a quota system, so that regions with a greater share of people living below the poverty line receive more vouchers.20 Regions that have a higher share of people in the labor force who reached high school but not college also receive more vouchers. The vouchers are an important resource that help keep private training institutions liquid by supplying them with students whose fees are paid for by the government. The vouchers are there- fore an indirect means of supporting the private sector. A "containment" strategy requires students to use vouchers within the local region, assuring that the program benefits private providers in all regions. Despite this re- 230 EAST ASIAN LABOR MARKETS striction, however, the system creates incentives for schools to compete with other private providers and to offer courses the students demand.' Outcomes of the Training and Vocational Education System The collection of data concerning the performance of technical and voca- tional education graduates is spotty, to say the least. The Congressional Commission on Education (1992) reported that these graduates find ob- taining employment difficult. Employment rates were 31 percent for grad- uates of postsecondary nondegree courses. Graduates of short-term, gen- erally nonformal courses had slightly more success, with an employment rate of 54 percent. Godfrey and Hunting (1996) show that graduating from vocational or technical training does not reduce unemployment and that the unem- ployment rate for graduates of these programs is in fact higher than the prevailing unemployment rate for nongraduates in similar age brackets.22 However, the same study argues that training increases labor force partic- ipation rates. This finding may reflect the fact that people who have un- dergone training are less likely to give up on the search for work. Training also appears to open up opportunities in paid employment (rather than in entrepreneurial or family-based modes), and private training (as opposed to government programs) affects how much workers earn. The private training institutions surveyed had slightly higher unit costs per graduate for nonformal training than their government counterparts, but nonfor- mal welding, business studies, automotive, and electrical technology courses boosted trainees' employment prospects. Training and Vocational Education as a Response to the Crisis TESDA was able to train only about 1,500 displaced workers in 1998. The displaced workers in the formal sector that were monitored by the DOLE for 1998 numbered 155,000. Several factors might explain the weakness of TESDA's response thus far. First, the effects of the crisis were not apparent until well into 1998. Consequently, there was little sense of alarm. The national budget for 1999 was prepared by line agencies and made ready for passage in Con- gress by July 1998, when the full impact of the crisis was not yet appar- ent. Second, a degree of complacency may have been bred by the constant public pronouncements that the Philippines was the least affected of all the countries affected by the crisis. The fact that national output growth THE PHILIPPINES 231 was far healthier than it was in other East Asian economies (which expe- rienced free fall earlier on) doubtless contributed to this complacency. Third, there was a substantial fiscal constraint. A precautionary program with the International Monetary Fund (IMF) entered into by the Ramos government in early 1998 involved reducing the appropriated budgets of all agencies by 25 percent (the level of mandatory reserves). The Ramos government, however, was also involved in a window-dressing operation of its fiscal situation. While it reported a slight fiscal surplus at the end of 1997, the accounts subsequently showed that the administration owed gov- ernment contractors more than 100 billion pesos. The Estrada administra- tion inherited the need for drastic fiscal adjustment just as the impact of the crisis was worsening. Finally, the national elections and the new administration-which was formed around a new party-may have prevented a rapid transition from taking place. Because of the winner-take-all character of Philippine presi- dential elections, a new president always has a chilling effect on govern- ment activity, as middle- to upper-level public officials are customarily obliged to hand in their resignations as a courtesy to the new government. Against this backdrop, TESDA was in its first full year of operation as an integrated structure. Private and Public Resource Use23 By 1997 there were 723 publicly funded vocational-technical institutions and centers in the Philippines and 1,383 private vocational and technical institutions and centers. Private institutions account for 80 percent of en- rollment in technical and vocational programs. Both public and private training providers have dropout rates of 10-15 percent. However, the two systems differ sharply in the effectiveness with which they use resources. The costs per graduate of private formal training centers are only 45 to 70 percent of those for graduates of public training institutions (Godfrey and Hunting 1996; Johanson 1998) (table 5.8). The low staff-to-student ratios in public training institutes is consistent with findings that show widespread underenrollment in some courses. However, this generalization requires a caveat, because public training in- stitutions are heterogeneous. The regional training centers generally have very high utilization ratios, in part because they have the best facilities. The expansion of private vocational training is hindered by one im- portant factor: the credit constraint faced by small private providers. These small enterprises generally have no collateral, a factor that prevents them 232 EAST ASIAN LABOR MARKETS TABLE 5.8 How Public and Private Institutions Use Their Facilities (percent) Public Private Utilization rate for laboratories 74 44 Utilization rate for workshops 80 55 Providers with staff-to-student ratios of more than 1:12 90 66 Source: TESDA (1999). from accessing credit markets. If they do gain access to credit, the lender charges very high interest rates to compensate for the risk. Public credit windows are available, but transaction costs are high and processing an ap- plication can take up to two years. Some observers theorize that a second constraint to the expansion of private training provision exists in the form of unfair competition from highly subsidized public training institutions, which often do not charge fees Uohanson 1998). Given the large market share of private institutions, however, any unfair competition is likely to occur in areas that have a disproportionate share of public institutions. If the phenomenon of unfair competition is widespread, a voucher scheme that allows students to choose their institution is one solution. Subsidies to public training institutions can also be reduced using a schol- arship scheme similar to the Private Education Student Financial Assis- tance program, which allows students to choose either public or private providers.24 A complementary approach would involve allowing public in- stitutions to charge fees and use some of the funds generated to compen- sate for reduced government subsidies. Still another solution would in- volve giving private institutions access to government subsidies-but this notion is somewhat unrealistic, given TESDA's miniscule budget. Vocational Training within Firms A recent survey of firms in the Philippines found that more than 40 per- cent had made significant expenditures on training for their employees (ohanson 1998). More than one-half (55 percent) of firms with more than 500 employees and around one-third of firms with 50 to 499 em- ployees had developed training units and human resource development plans. Nearly half of these medium-size and large firms also had staff members specifically for training, and slightly more than 10 percent had their own training centers. All the firms were keenly aware of the incen- tives the government provides for private training, THE PHILIPPINES 233 Small firms are less able to institute their own training programs. There is a clear market failure here, because small firms invest less in training and are particularly vulnerable to the free-rider problem (the poaching of staff trained by other firms). Other countries have used various methods to overcome this market failure, including direct subsidies for employers that provide training. Subsidizing employers' expenses is one promising ap- proach, as Singapore's experience shows. In Singapore such subsidies have encouraged small employers to provide training. The Philippines, how- ever, has no single strategy to encourage small firms to provide training. The Philippines has two apprenticeship programs. First, the appren- ticeship law governs training provided by the enterprises themselves. In 1997 some 20,000 apprentices were enrolled in such training, a figure that represents less than 0.1 percent of all wage and salary employees in the country. Several standing proposals (including one from TESDA) argue for revising the law. The draft National Technical Education and Skills Development Plan (TESDA 1999) states that employers are using the ap- prenticeship law to exploit workers by treating apprentices as a source of cheap labor. In any case the law is ineffective, because it does not require firms to draw up training schemes for approval. The law also restricts training to a maximum of six months-too short a time for trainees to ac- quire many skills. And TESDA is unable to monitor the apprentices, as they are not considered workers. The dual training system, which is modeled on the German apprentice- ship system, is the second kind of enterprise-based training. Trainees spend 70 percent of their time working in a firm and the rest of the time in des- ignated learning centers. Trainees receive three-fourths of the minimum wage but retain only 30 percent, with the rest going to the training provider. The system brings together on-the-job experience and theoretical instruc- tion. Firms are not obliged to hire trainees after the 30-36 month training period, and they do not spend as much as German firms, which shoulder most of the cost. However, this type of training can be expensive. While the annual cost per trainee of the dual system in Germany is about the same as the country's per capita GNP, simulations show that this ratio is much greater in other countries that have adopted the system. It is more than 2:1 in Korea, more than 3:1 in Indonesia, and more than 4:1 in Egypt.25 The program is in its very early stages of implementation and outside of Manila is constrained by the lack of trained personnel. Monitoring mechanisms remain weak, and despite recent gains the adoption rate has been slow. In 1997 the number of graduates was only 235, down from the 273 of the previous year. The weakest link in the dual training system 234 EAST ASIAN LABOR MARKETS appears to be the rate at which training institutions are accredited. Twenty- seven training institutions had been accredited as of 1998, but the target was 48.26 Making Training and Vocational Education Responsive to Market Signals A major step in TESDA's medium-term planning involved an exercise to elicit from its institutional constituency a roster of skills and industries that should be given priority in 1999-2004. Included in this exercise were staff in the regional offices, schools and training centers, chambers of com- merce, and labor unions. What emerged from this exercise was a long list of "sectoral priorities" and "critical skills" for each industry and region. Unfortunately, the exercise included no quantification, so the results can- not be used as a tool to decide exactly how TESDA should use its limited resources. A crude first approximation might, for instance, have involved identifying professions for which compensation indices have been rising faster than average in the past few years. The country's exports of skilled labor could have provided another measure. It makes sense to think twice about investing skills in sectors with a surplus, as indicated by the dispro- portionately large human resource exports in recent years. The human resource planning exercise also failed to identify the profes- sions and skills that enterprises and private training institutes are most likely to underprovide or the reasons for not emphasizing them. The gov- ernment's role in training provision will vary depending on the cause of these market failures. TESDA can begin providing the needed training in these areas. Subsidies (in the form of directed credit and teacher training assistance for private sector providers) can help correct policy distortions (such as excessively high wages for trainees) that cause enterprises to un- derprovide training. Career counseling can address the lack of information. The requirements of human resource planning can be daunting, and attempts at such planning elsewhere have not proved successful. Others have arrived at the conclusion that manpower planning does not work (Schweitzer 1994). In this case it becomes very important to make sure that both private and public training providers have the capacity and in- centives to respond to market signals rather than being guided by man- power plans. Several means of ensuring this response are available. TRAINING VOUCHERS. Vouchers such as those available under the Private Education Student Financial Assistance program allow trainees to choose a training provider. The National Technical Education and Skills Develop- ment Program includes a plan to expand the number of vouchers under the THE PHILIPPINES 235 current program, which is a more efficient mode of providing vocational education than direct government provision. A 200 million peso fund for vouchers benefits 10,000 trainees, while a similar 200 million peso fund benefits only 5,000 student under TESDA.27 A VOUCHER SYSTEM FOR THE PUBLIC SECTOR. Voucher systems can also be applied to public schools and training centers. At present TESDA pro- vides block grants for its schools and training centers, and some local gov- ernment units participate by making school buildings and agricultural lands available. Enrollment is highly subsidized, and there is no means testing in the allocation of the slots. Because grants routinely increase by a certain percentage each year, public schools and centers have no incen- tive to improve their productivity, resource use, or the quality of their course offerings. A voucher system can create a link between the amount of public resources a school can attract and the number of students it draws. As of the middle of 1999, however, there were no plans to set up such a system. DEVOLUTION AND THE RATIONAL ALLOCATION OF RESOURCES. A voucher system for the public sector is only one possibility for using resource allo- cation to create incentives for schools and training centers to perform bet- ter. However, in some settings the existing capabilities of a provincial or a regional training center depend on either past investment or past neglect. In these situations the number of voucher holders a school is able to at- tract may not give an accurate picture of the efforts the institution has made. An incentive system based mainly on vouchers could reward those training outfits that perform well primarily because they have been fa- vored with capital investments in the past or because they are located in highly urbanized centers. The absence of a voucher system for public providers therefore does not mean that reforms are not being planned. One specific strategy being con- sidered is the establishment of a competitive bidding system for training providers (TESDA 1999). While the formulation may seem general at this point, the language of the provision allows for a wide range of possibili- ties. People at the policy planing staff do not believe that this reform would force public providers to compete with private providers for con- tracts. However, if TESDA's training facilities are devolved to local gov- ernment units, they will undoubtedly have to compete. In deciding which regions or provinces will receive the larger share of training resources, other criteria over and above institutional capability, program equity, and pro- gram relevance may be considered, such as a system of matching grants. 236 EAST ASIAN LABOR MARKETS This system would allocate more resources to provinces (regions do not have funds) that show a willingness to invest some of their own resources in training. The proposal to devolve training functions to local government units will encounter many difficulties. Having seen the experience of the devo- lution of hospitals, TESDA regional staff members are naturally wary. The devolved hospital functions were matched with compensating increases in the Internal Revenue Allotment of the receiving local government units.28 But it proved difficult to compel local governments to use the additional funds to maintain facilities and pay personnel. The concern of trainers is an especially valid one because the local government unit has the legal au- thority to declare their positions redundant. The concerns arising from proposals to devolve functions raise almost the same concerns as the pro- posal to shed the central administrative functions of public training facil- ities through management contracts with NGOs or the private sector. RESOLVING THE PROBLEM OF SUPERVISION. Improving responsiveness is particularly difficult in schools run by the CHED and by chartered uni- versities, because it is often difficult to make rapid changes in the curricu- lum. School-based training has been known to succeed only when the link between the schools and enterprises is strong. Labor market responsiveness also improves when employers are linked directly to the school-for ex- ample, by direct employer financing or involvement in school governance. Because more than half of the budget for training and vocational educa- tion is being spent by state universities and colleges and institutions super- vised by the CHED, creating a channel for TESDA to introduce broad su- pervision will potentially have a tremendous impact. Some Recommendations for Future Policy The government of the Philippines has taken a number of steps to address the problems arising from the East Asian financial crisis. However, much remains to be done. The government can take a number of steps in the near to medium term that will build on the efforts already made. UNDERTAKE FURTHER RESEARCH. Such research will help policymakers determine the depth of needed labor market interventions. The initial condition of high unemployment in the Philippines-a situation that sets the country apart from its neighbors-deserves further consideration. If unemployment is once more between 7 and 9 percent after the crisis, re- THE PHILIPPINES 237 search can resolve the issue of whether or not the country needs extensive safety nets, active and passive labor market policies, and training. If labor regulations or structural mismatches between the supply and demand for skills are causing market rigidities, safety nets and labor market policies can improve social welfare. We did not find any obvious instances of labor market inflexibility in wage setting or the decision to hire and fire. Even the political power that unions are normally able to wield appears to have been overwhelmed by the overriding need to preserve jobs in the face of a precipitous decline in aggregate economic demand. STRENGTHEN LABOR MARKET POLICIES. Labor market policies can be strengthened to assist displaced workers. Falling aggregate demand did in- deed result in significant job losses and involuntary unemployment. In ad- dition, and possibly to an even greater extent, job losses since the middle of 1997 had another cause. They were also the result of the belief that the drastic shift in relative prices would mean a recovery of aggregate demand insufficient to restore the profitability of many of the enterprises that thrived when the overvalued peso was at its highest. A period of adjust- ment and economic restructuring, not just of growth recovery, is most likely already under way. Labor market policies, especially training for dis- placed workers, will be useful as labor is shed and shifted into temporary unemployment or long-term redundancy. IMPROVE PRODUCTIVITY. Productivity needs to be improved to reduce unemployment. One novel hypothesis proposed here is that a significant portion of the unemployment in the Philippines, even during the crisis, is voluntary and requires a long-term process of sustained productivity growth rather than a high dose of labor market policies. Economywide produc- tivity improvement may be a necessary condition for reducing the high level of unemployment. Many of those who are unemployed simply have unusually high reservation wages because they receive support from their families (which receive significant income from foreign remittances) for extended periods during the job search. USE INFORMATION TO IMPROVE TARGETING. Labor market information can improve targeting. The phenomenon of underemployment, especially during periods of economic shocks, provides a good first approximation of the location of distress. Quarterly labor force surveys, in conjunction with other economywide surveys like the Annual Poverty Indicator Sur- veys and Family Income and Expenditure Surveys, can provide insight into 238 EAST ASIAN LABOR MARKETS the kinds of policies needed. The failure of agencies like the DOLE and the Social Security System to launch adequate poverty-targeted responses can be traced to their reliance on monitoring schemes that are able to cap- ture only the most obvious manifestations of the crisis. STRENGTHEN SOCIAL SAFETY NETS. Stronger safety nets will help sup- port income during crises. As in the past the Social Security System and the Employees Compensation Commission were able to provide a welcome safety net in the form of emergency loans. But the number of people af- fected by the East Asian crisis was far greater than in previous calamities. The Social Security System's response had clearly been inadequate, in sev- eral senses. First, 13 million people working in the informal sector do not fall within the ambit of the system. A systematic inventory of the govern- ment's possible role in encouraging formal and informal safety net mech- anisms for these workers is essential. Policymakers must also understand the adequacy and limits of safety net mechanisms such as transfers from relatives in order to establish effective policies. Second, emergency loans are a cumbersome replacement for drawdowns that members of the Social Security System should be able to get routinely from their accumulated savings. The system needs to explore ways to allow members to access funds they have accumulated in the pension fund through the years. Fi- nally, in the medium term the Social Security System should further in- crease its capacity to monitor and punish employers who do not remit contributions for their employees. REFORM PUBLIC WORKS PROJECTS. Public works programs need to be made more effective. Previous programs such as the CEDP offer some pos- itive lessons. Any serious effort to use rural works programs to confront co- variant risks in the farm and nonfarm sectors of the rural Philippines must begin by reforming the system of intergovernmental transfers that under- mines poverty-focused infrastructure allocations for the provinces. The na- tional government's system of budget planning and releases also needs to be reformed to provide the predictability that agencies implementing labor- intensive infrastructure programs require. INCREASE PROVISION OF LABOR MARKET SERVICES. The DOLE is an agency in transition. Its original mandate was to oversee industrial rela- tions, but in recent years that mandate has changed in two ways. The agency has been involved in disaster relief through job fairs and employ- ment creation in areas affected by volcanic eruptions and earthquakes. It has also been called on to concern itself with problems of employment THE PHILIPPINES 239 even in sectors where there are no strict employer-employee relationships. For example, it has stepped up efforts to monitor the enforcement of key labor standards among subcontracting firms. In these situations it strives to provide job-matching services, counseling, retraining, rural works, and livelihood programs and to facilitate access to other government programs such as emergency loans and basic labor standards enforcement, especially in distressed firms. The DOLE can usefully explore these directions further. In particular it can help to fulfill the need for a lead agency in the creation of social pro- tection institutions, since it is strategically placed to take on this role. The DOLE can move in this direction by: * Strengthening and scaling up the delivery of its establishment-based ser- vices while looking beyond them. At present these services are simply inadequate. The DOLE can go beyond formal sector establishments by using economywide databases more intensively. These databases include the Annual Poverty Indicator Survey, the quarterly Labor Force Surveys, and the Family Income and Expenditure Surveys. These databases can be enhanced further and may be used to gauge the adequacy of existing programs and target beneficiaries and to cal- ibrate the timing and extent of needed labor market interventions. The depth of agricultural unemployment and underemployment even in early 1999, for instance, was not reflected in the DOLE's ad- ministrative reporting system. If it had been, the agency might have seen the potential contribution scaled-up versions of its own public works programs could have made in rural areas. In addition, it could have implemented an interagency initiative on labor-based infra- structure programs. Again, economywide surveys in the fall of 1998 would have allowed the DOLE to see that unemployment and under- employment were rising in many industries, particularly manufac- turing and construction. Using only establishment reports on clo- sures and layoffs led the DOLE to believe that the worst effects of the crisis had been felt during the first quarter of 1998. * Influencing other agencies with a key role in social protection. With a more comprehensive grasp of the crisis (beyond what the adminis- trative reports indicate), the DOLE could undertake important ad- vocacy. For instance, during the East Asian crisis it could have argued for relaxing the eligibility requirements for emergency loan programs for distressed workers and increasing the scope for on-lending to SMEs. 240 EAST ASIAN LABOR MARKETS * Designing its own programs and influencing those of other agencies to pre- vent leakage to the nonpoor. Because of the fiscal crisis, services should increasingly be targeted to poor areas (the rice subsidy) and means- tested where necessary (retraining). Workers in formal sector estab- lishments are not normally the most vulnerable segments of the work force, even though they are also affected by unemployment and job rationing. Because resources are always in scarce supply, the DOLE needs to identify early on a set of selection mechanisms for beneficia- ries that will allocate limited resources (the limited slots for retraining, for instance) to those who need these services but cannot afford them. These may include older workers, people with disabilities, people with low educational levels, minorities, and others. Moving away from an almost exclusive reliance on establishment-centered administrative re- ports will go a long way toward improving the impact and equity of the government's programs. It will also allow the DOLE to assume a larger role in the provision of services and to develop an economywide constituency for its programs. * Reforming training and vocational education programs to improve delivery and efficiency. Reforms are being contemplated in a number of areas. Possible reforms include: * Reducing the number of public training centers in areas where they offer the private sector too much competition; * Privatizing maintenance for training facilities by transferring re- sponsibility for it from the central agency to local governments; * Implementing user charges; and * Auctioning training funds to the private sector or enhancing scholarships for use in the private sector. Notes 1. The authors would like to thank Marites Cruz, Mina Pangandaman, Aleli Rosario, and Mariflor Liwanag for their able assistance. We also gratefully ac- knowledge the assistance of the senior staff of the Philippine Department of Labor's Institute of Labor Studies, the Bureau of Labor and Employment Statis- tics, and the National Wages and Productivity Board. 2. Balisacan and Edillion (1999) reveal that 12-16 percent of the more pros- perous half of Philippine households receive transfers from friends and relatives, although the proportion of families in the poorer half that receive transfers from relatives is higher still. 3. The term unemployed as it is used here is not comparable to the official definition of unemployment. Household heads who are not seeking work-and are therefore not in the labor force-are included here. THE PHILIPPINES 24I 4. Based on an examination of the 1991 Family Income and Expenditure Survey (Reyes 1996b) 5. Note that high firm-level productivity will not translate into higher real wages throughout the economy so long as there are many involuntarily unem- ployed workers at the factory gates who are qualified replacements for the em- ployed workers who want to bid up wages. Sustained job creation is a necessary condition for bridging the gap between reservation wages and prevailing real wages. 6. Capital inflows outpaced the increases in the current account deficit and resulted in the real appreciation and stabilization of the currency as international reserves grew. The real (and sometimes nominal) appreciation of the peso in a pe- riod of heavy and expanding trade and growing current account deficits led to the overvaluation of the currency. For a discussion of the relationship between the ap- preciation of the peso and the onset of the crisis see de Dios and others (1997) 7. Personal communication with staff of the Labor Department's Institute for Labor Studies (1999). 8. Order Number 10, Series of 1997, which amended the rules implement- ing Books 3 and 4 of the Philippine Labor Code. 9. For a statistical overview see BLES (1998). 10. See BleedingAgency, Margarita Debuque's three-part article in the Philip- pine Daily Inquirer, August 23-25, 1999. 11. Bleeding Agency, Margarita Debuque's three-part article in the Philippine Daily Inquirer, August 23-25, 1999. 12. The order is titled Establishing the Policy Direction and Institutional Frame- work to Implement Labor-based Equipment Supported Infrastructure Program. 13. The Magna Carta for Small Enterprises (Republic Act 6977) aims to pro- mote, develop, and assist small and medium-size enterprises through the creation of a council and the rationalization of government assistance, programs, and agencies concerned with the development of these firms. 14. Information for this section was put together by the DOLE's Institute for Labor Studies 15. President Estrada signed this bill into law in January 2000. 16. These programs are formally enumerated in Department Order No. 6, Se- ries of 1998 of the Department of Labor and Employment. 17. The Special Program for the Employment of Students and training schol- arships at the Technical Education and Skills Development Authority give prior- ity to displaced workers. These programs are the most generously funded and the easiest to access. The 500 million peso Emergency Loans Program under the Social Security Administration is also accessible mainly to workers displaced from formal sector establishments. PRESEED and the pilot Rural Works Program of the Bureau of Rural Workers, which are the key programs designed for rural workers, are very small. 18. A more detailed presentation of the DOLE's position can be found in BLES (undated). 19. Personal communication with Teresa Soriano, Executive Director of the Institute for Labor Studies. 242 EAST ASIAN LABOR MARKETS 20. We should note that no regions have been designated as poor, although some provinces and municipalities have. A formula sensitive to poverty incidence within regions would improve targeting. The Family Income and Expenditure Surveys can be used to derive a ranking of provinces according to poverty crite- ria, for instance. 21. Caveat: around half of the vouchers are spent on computer-related courses. No assessment of the program has yet probed the possibility that students' deci- sions do not maximize their chances of becoming employed. There are, however, doubts within TESDA's planning office as to the wisdom of enrolling in com- puter classes instead of agriculture or fisheries courses. 22. We cannot discount the self-selection and screening problems involved here, however. If employers think or know that those who ordinarily qualify for college slots go to college instead of vocational and technical schools, then they tend to believe that those who have "only" technical or vocational credentials are not bright enough to be accepted to college. 23. The data from this section may also be found in NTDSD 1999 (May draft). 24. Current recipients of Private Education Student Financial Assistance scholarships do not attend public institutions. 25. An evaluation of the costs of the scheme in the Philippines is presently being conducted. 26. Personal communication with Guy Adrias, Office of Apprenticeship, TESDA. 27. Meeting with TESDA Director-General Edicio dela Torre (1999). 28. Intergovernmental transfers from the national government comprised a man- datory 40 percent of the national government's internal revenue of two years ago. References Alba, M., and E. Esguerra. 1999. "Estimating the Modes of Labor Force Partici- pation in the Philippines." Discussion Paper 9903, School of Economics, Uni- versity of the Philippines. Annual Poverty Indicator Surveys. 1998 and 1999. National Statistics Office, In- come and Employment Statistics Division, Manila, Republic of the Philippines. Balisacan, Arsenio. 1994. "Design of a Poverty-Targeted Food subsidy Program in the Philippines." Policy paper for the Agribusiness Assistance Program, School of Economics, University of the Philippines. Balisacan, Arsenio, and Rose Edillion. 1999. "Human face of the Asian Crisis: What Do Nationwide Panel Data Philippine Households Show?" Paper pre- sented at the national conference of the political Science Association, Univer- sity of the Philippines, Diliman, Quezon City. THE PHILIPPINES 243 BLES (Bureau of Labor and Employment Statistics). 1998. "Trends and Patterns in Nonregular Employment in the Philippines." Labstat Updates 2(2) Depart- ment of Labor and Employment, Manila. ____. 1999. "Statistics on DOLE Assistance Provided to Displaced Workers Affected by the Economic Crisis" Labstat Updates 3(7). Department of Labor and Employment, Manila. ____. Undated. "An Alternative Approach to Monitoring Labor Market Re- sponse." Manila. Canlas, Dante. 1997. "Unemployment and Monetary Policy in the Philippines." In Esguerra Emmanuel and Kazuhisa Ito, eds., Employment, Human Capital and Job Security: Recent Perspectives on Philippine Labor. Tokyo: Institute for Developing Economies. Congressional Commission on Education. 1992. Post-Secondary Education. Vol. 2 of Book 2 of Making Education Work. Manila. de Dios, Emmanuel. 1999. "The Economic Crisis and Its Impact on Labor." Philippine Center for Policy Studies. Unpublished working paper. de Dios Emmanuel, Benjamin Diokno, Raul Fabella, Felipe Medalla and Solita Monsod. 1997. "Exchange Rate Policy: Recent Failures and Future Tasks." Public Policy October/December. Family Income and Expenditure Surveys. 1997. National Statistics Office, Income and Employment Statistics Division, Manila, Republic of the Philippines. Godfrey, Martin, and Gordon Hunting. 1996. "Evaluation of the Technical education and Skills Deveopment System." Washington, D.C.: World Bank. Unpublished discussion paper. Johanson, Richard. 1998. "Philippines Technical-Vocational Education and Train- ing (TVET)." Background paper for Education Sector Study, Asian Develop- ment Bank and World Bank. Manila. Labor Force Surveys. Bureau of Labor and Employment Statistics. Manila, Re- public of the Philippines. Lamberte, Mario Cesar Cororoton, Marge Guerrero, and Aniceto Orbeta. March 1999. "Results of the Survey of the Philippine Industry and the Financial Cri- sis." Unpublished discussion papeL Lim, Joseph. 1997. "Macro Labor Statistics and Sectoral Employment." In Es- guerra Emmanuel and Kazuhisa Ito, eds., Employment, Human Capital andJob Security: Recent Perspectives on Philippine Labor. Tokyo: Institute for Develop- ing Economies. ____. 1999. "The East Asian Crisis and Child Labour in the Philippines." Tokyo: Institute for Developing Economies. Draft. Llanto, Guilbert, and Teresa Sanches. 1997. Impact of Financial Liberalization on Micro and Small-scale Manufacturing Industries. Manila: Philippine Institute for Development Studies. 244 EAST ASIAN LABOR MARKETS Maglen, A., and Manasan, Rosario. 1996. Education Sector Study. Manila: Asian Development Bank and World Bank. Quibria, M.G. 1996. Indonesia, Republic of Korea, Philippines and Thailand. Vol. 2 of Rural Poverty in Developing Asia. Manila: Asian Development Bank. Reyes, Celia. 1996a. "An Assessment of the System of Minimum Wage-Setting." Policy and Development Foundation, Inc. Available at the National Wages and Productivity Commission, Department of Labor and Employment. Manila. . 1996b. "Monitoring Systems for Poverty Tracking." Micro Impacts of Macro Adjustment Policies Project Research Paper Series 30. Legaspi Village Makati: Policy and Development Foundation, Inc. Schweitzer, Julian. 1994. "Vocational Education and Training: the Role of the Public Sector in a Market Economy." Human Resources Development and Operations Policy Working Paper 16. Washington, D.C.: World Bank Subbarao, Kalanidhi, Akhter Ahmed, and Tesfaye Teklu. 1996. "Selected Social Safety Net Programs in the Philippines: Targeting, Cost Effectiveness and Op- tions for Reform." Discussion paper 317. Washington D.C.: World Bank. TESDA (Technical Education and Skills Development Authority). 1999. The Na- tional Technical Education and Skills Development Plan, 1999-2004, "A Vi- sion and Strategy for the Development of Middle-Level Manpower." Manila. World Bank. 1995a. "The Philippines: An Agenda for the Reform of the Social Security Institutions." Sector Report 3400. Washington, D.C. World Bank. 1995b. "The Philippines: Public Expenditure Management for Sus- tained and Equitable Growth." Sector Report 14680. Washington, D.C. World Bank. 1996. "A Strategy to Fight Poverty: Philippines." Working Paper 16118 Country Operations Division, East Asia and the Pacific. Washington, D.C. . 1997. Philippines: Managing Global Integration. 2 vols. Poverty Reduc- tion and Economic Management Sector Unit Report 7024-PH. East Asia and the Pacific Regional Office. Washington, D.C. . 1998. World Bank Supportfor Small and Medium Industries in the Phi- lippines: An Impact Evaluation. Operations Evaluation Department Report 18041. Washington, D.C. 6 The Labor Market and Labor Policy in a Macroeconomic Context: Growth, Crisis, and Competitiveness in Thailand MOAZAM MAHMOOD AND GOSAH ARYAH THIS APPRAISAL OF Thailand's labor market and labor policy comes at a critical juncture for the Thai economy.' The crisis in growth and employment the country is experiencing at the beginning of the 21st century is unparalleled in Thailand's recent history. The country's recovery-especially to the high growth and full employment that char- acterized the 1 980s and much of the 1 990s-is uncertain, largely because of the variability in labor market policy. Past policies may have delivered high trend growth in output and employment, but they also engendered structural weaknesses in the economy that helped bring about the crisis. These problems aside, such policies are unlikely to be effective in the changed global environment. This chapter attempts to fill the current pol- icy "gap" in Thailand by analyzing the related problems of the labor mar- ket and the economy. It argues that new policies are needed to foster growth in both output and employment. The Current Labor Market Situation In the fall of 1999, the Thai economy was embroiled in the Asian finan- cial and economic crisis. The economy had contracted by 9.4 percent 245 246 EAST ASIAN LABOR MARKETS TABLE 6.1 Growth Rate of Real GDP (percent per year) Year 1980-90 1990-96 1995 1996 1997Q1 1997Q2 1997Q3 1997Q4 1998 Ql GDP 7.6 8.3 8.8 5.5 3.8 5.1 -3.2 -7.5 -13.4 since 1998 (table 6.1). But the rate of contraction slowed in the last quar- ter of the year, falling from its earlier high of 13 percent to just 4 percent. Estimates of economic growth for 1999 projected an increase to 3 percent. Trend growth was expected to remain low, ranging between 3 and 5 per- cent, until 2003. The situation is made more dramatic by the high growth that preceded it. Trend growth in Thailand in 1980-96 topped 8 percent and in 1990-95 neared 9 percent (table 6.1). It declined in 1996 and the first half of 1997, ranging between 4 and 6 percent. The financial crisis hit Thai- land in the third quarter of 1997, causing the economy to contract by 3 percent and then to shrink by 8 percent in the last quarter alone. Growth is expected to remain low-in the area of 3 to 5 percent-in 2000-2003. The Thai economy had attained a per capita GNP of around $3,000 in 1996. The crisis is broadly estimated to have reduced real per capita in- come by about 20 percent (Kakwani 1998). In 1996 the unemployment rate stood at 2 percent-virtually full employment (table 6.2). By 1997, however, the rate had doubled and continued to climb to around 7 per- cent. By one estimate the crisis generated approximately 90,000 redun- dancies, raising the level of unemployment to 2.2 million as of June 1999.2 Real wages declined following the precrisis tightening of the labor market. Real wage growth, which stood at over 2 percent per year in 1996, reversed itself. The real wage fell by more than 7 percent in 1998 and by 1.5 percent in 1999. The large debt overhang underscores the difficulty of generating a strong financial and economic recovery. Almost one-half of all debt now consists of nonperforming loans. Private investment is virtually nonexis- tent, and capacity utilization has collapsed (table 6.2). The exchange rate has also plummeted. The Thai baht was pegged to the dollar at 25:1 in 1996, but the exchange rate is expected to remain around 40:1, up from 55:1 in July 1997. Despite the depreciated baht, exports have dropped from their 1997 level of $57 billion to $53 billion for 1998. The depreci- ation of the baht has led to a severe compression of the import market, with imports falling from $64 billion in 1996 to $37 billion in 1998. The GROWTH, CRISIS, AND COMPETITIVENESS IN THAILAND 247 1998 Q2 1998 Q3 1998 Q4 1999 Ql 1999 Q2 1999 Q3 1999 Q4 2000 2001 2002 2003 -12.8 -9.5 -3.7 -9.4 4.0 -9.4 3.3 4.8 4.6 4.8 5.1 high precrisis growth of GDP and exports was in part financed by large capital inflows (5 percent of GDP in 1996). Thailand had a significant current account deficit of 8 percent of GDP in 1996, but the currency cri- sis (which reversed the inflows) and the fall in imports resulted in a large current account surplus-13 percent in 1998. The major impact of the reduction in growth, employment, wages, asset prices, and income has been a degradation of human development indicators (table 6.2). Between 1992 and 1996 Thai poverty levels had vir- tually halved, with only 11 percent of the population falling below the poverty line. The crisis has raised the poverty level from 11 to 13 percent, pushing 1 million more people into poverty. Macroeconomic policies designed to combat the crisis have been based largely on financial sector reform. At the onset of the crisis, the Thai gov- ernment committed itself to an economic reform package that was a con- dition of a $17 billion loan from the International Monetary Fund (IMF). Until the year 2000, fiscal and monetary policy will be shaped largely by the terms of this commitment. The IMF program is based on several key measures. First, the indefensible pegged exchange rate has been floated and the depreciating baht propped up by interest rate increases to attract capital inflows. Second, fiscal contraction and a cap on the budget deficit are being used to curtail the threat of inflation. Third, the ailing financial sector is being revived through bank closures, restructuring, and recapital- ization. However, the current state of the economy shows that the eco- nomic sectors have not recovered. Devising policies to revive the economic sectors, generate employment, and improve labor conditions requires understanding both the roots of the high growth that preceded the crisis and the reasons it collapsed. For in- stance financial sector recovery may be necessary to increase employment levels, but it may not be sufficient. If the factors that caused the financial and economic crisis were purely financial, then financial reform will spur economic recovery. But if the causes are a more complex combination of financial and economic sector weaknesses, then economic sector reform will be required as well. 248 EAST ASIAN LABOR MARKETS TABLE 6.2 Macroeconomic and Social Indicators in Thailand 1996 1997 1998 1999 2000 GDP growth 5.5 -0.4 -9.4 0.9 2.9 Inflation (%) 5.6 5.7 8.1 2.4 4.7 Budget balance (% of GDP) 2.3 -0.9 -2.3 -3.5 -3.4 Current account balance -8.1 -1.9 11.5 8.7 5.1 (% of GDP) 54.4 56.7 52.9 54.6 58.3 Exports (US$ bn.) 63.9 55.1 36.5 42.0 49.0 Imports (US$ bn.) 27.0 29.5 32.34 Gross reserves (US$ bn) 33.1 28.1 38.0 31.3 28.8 Savings ratio (% of GDP) 40.8 32.3 26.7 24.4 25.2 Gross fixed investment (% of GDP) -23.5 -16.6 Private investment index (Dec.) (May) 65.0 52.1 57.6 Capacity utilization (%) (May) 13.4 13.7 15.2 12.0 11.0 Interest rate (%) B25.3 B31.4 B41.4 B38.2 B39.5 Exchange rate against US$ (baht) 32.0 32.5 33.0 33.4 33.7 Labor force (million) 2.0 3.7 5.2 7.5 6.8 Unemployment rate 2.2 (% labor force) (June) Unemployment level (millions) 2.3 1.4 -7.4 -1.6 2.4 728 911 Real wages 11.4 12.9 Poverty line (baht/person/ month) 6.9 7.9 Population below (%) No. (million) Source: Finance Ministry, Labor Ministry, and the National Economic and Social De- velopment Board. Economic Intelligence Unit Forecast (3rd and 4th quarters, 1999). We argue that the economic sectors of the country exhibit fundamen- tal weaknesses that helped cause the crisis. These weaknesses exacerbated the problems caused by financial sector weaknesses, which have been widely acknowledged as key factors in the crisis. Our analysis of Thailand's high precrisis growth levels reveals that production efficiency had declined for both labor and capital, especially in the 1990s. This decline in com- GROWTH, CRISIS, AND COMPETITIVENESS IN THAILAND 249 petitiveness has affected exports, the major engine of the country's growth. Macroeconomic policy to revive growth and employment has concen- trated so far on eliminating the financial inefficiencies. But the weak state of the economy argues for additional macroeconomic policies directed specifically at redressing the inefficiencies in production. We look at the policy actions the government has taken to combat the crisis-including employment-creation and income-support measures, measures to improve and intensify public employment services, and training programs-and assess their ability to deal with the current economic situation. The Economy and the Labor Market before the Crisis Before the crisis Thailand enjoyed high growth in both output and em- ployment. The growth in output caused a shift in the relative importance of various sectors, with agriculture losing share in GDP and manufactur- ing gaining. Participation in the labor force was also high, especially for women. Despite these positive factors, the informal sector continued to flourish, and underemployment persisted. Output Growth Thailand had very high output growth in the 15 years leading up to the crisis. The rate of growth throughout the 1980s was 8 percent annually (table 6.1)-a rate exceeded in the region only by the Republic of Korea (with 9 percent) and China (with 10 percent). From 1990 until 1996 only China's growth rate (12 percent) was higher. Industry, and especially man- ufacturing, led this high growth, expanding at 10 percent. The services sector grew at 7 percent, while agriculture expanded at only 4 percent. The high growth in industry and stagnation in agriculture changed the relative shares of these two sectors. In 1980 agriculture accounted for 23 percent of GDP, while industry had a 29 percent share and services a 48 percent share (table 6.3). By 1996 agriculture's share had shrunk to 11 percent, industry's had expanded to 40 percent (owing primarily to manufacturing), and services had remained relatively constant. The Structure of the Labor Market The high output growth and sectoral change that characterized the 1980s and 1990s had a major impact on the labor market in Thailand. Thai- land's population grew at a rate of 1.6 percent over this period to an esti- 250 EAST ASIAN LABOR MARKETS TABLE 6.3 Structure of Output Thailand 1980 1996 Value of GDP (US$ billion) 32.4 185.0 GDP share Total 100.0 100.0 Agriculture 23.0 11.0 Industry 29.0 40.0 (Manufacturing) 22.0 29.0 Services 48.0 50.0 1980a 1989h All manufacturing 100.0 100.0 Food, beverages, tobacco 55.0 21.3 Textiles, apparel, leather 8.0 15.2 Wood, paper, printing - 21.0 Chemical, rubber - 17.0 Minerals - 6.3 Metals - 5.5 Machinery, transportation 9.0 14.2 Jewelry, others 21.0 1.4 - Not available a. World Bank 1998. b. Report of the 1989 Industrial Survey. mated 61.6 million in 1999.3 In 1998 some 53 percent of the population over the age of 13 was economically active (including the employed, the unemployed, and the seasonally inactive), making up a labor force of 32 million (table 6.4). Those older than 13 who were not in the labor force accounted for another 15 million, including students (6 million), women doing housework (4 million), and the very young and elderly (4 million). Thailand's 77 percent mean labor force participation rate (LFPR) was high for the region, and its female LFPR the region's highest (59 percent). Thailand's LFPR declined in the 1990s, falling from nearly 80 percent to its 1998 level. The male LFPR declined from 85 percent to 77, while the female LFPR declined more, from 74 percent to 59 percent. As a re- sult, while the labor force grew at 3 percent in the second half of the 1980s, its growth rate dropped to 0.4 percent in the 1990s. Falling popu- lation growth rates and increased school enrollment ratios (the result of an increase in compulsory education levels) are responsible for these declines. GROWTH, CRISIS, AND COMPETITIVENESS IN THAILAND 251 TABLE 6.4 Structure of the Labor Force, 1998 Total Males Females (million) % % (million) % (million) % Total population 61.10 100 30.48 30.62 > 13 years 47.15 77.2 100.0 23.60 100.0 23.73 100.0 Economically active 32.17 52.7 68.22 18.07 76.6 14.09 59.4 Employed 28.55 60.55 16.57 70.2 11.98 50.5 Unemployed 1.61 3.41 0.86 3.6 0.75 3.2 Seasonally inactive 2.00 4.24 0.64 2.7 1.36 5.7 Not in labor force 14.98 31.77 5.34 22.6 9.64 40.6 Household work 4.27 9.06 0.08 0.3 4.19 17.7 Studies 6.05 12.83 3.06 13.0 2.99 12.6 Too young/old 3.84 8.14 1.68 7.1 2.16 9.1 Others 0.82 1.74 0.53 2.2 0.29 1.2 Source: Labor Force Survey (1998). Even after the sectoral changes of the 1980s and 1990s, half the Thai labor force was still in agriculture in 1998, but almost all of these workers lived in rural areas (table 6.5). After agriculture the three sectors with the largest shares of employment were manufacturing, sales, and services, all with approximately 13 percent. Construction and transportation had much smaller shares (6 and 3 percent, respectively). In urban areas sales and ser- vices occupied nearly a third of the labor force, and manufacturing ac- counted for less than a fourth. In rural areas agriculture had just under a two-thirds share of the rural labor force, followed by manufacturing and sales with a 10 percent share each and services with a 9 percent share. The high concentration of women in certain sectors has resulted in high LFPRs for females in these areas. By 1998 there was virtually no gen- der gap in manufacturing, commerce, or services, while agriculture had a small gender gap of 6 percent. The major gender gaps remained in con- struction, utilities, and transportation. According to an Asian Institute of Technology study (AIT and ILO 1999), this reduction in the gender gap is the result of the feminization of particular export industries. High growth in exports propelled growth in manufacturing, drawing women into assembly lines as production workers. Before the crisis women con- stituted an estimated 65 percent of the labor force in the six leading export industries. Women had a 65 percent share in the food sector, a 61 percent share in textiles, a 74 percent share in garments and leather, a 252 EAST ASIAN LABOR MARKETS TABLE 6.5 Sectoral Employment across Urban and Rural Areas, 1998 (percent) Female (percent Sector Total Urban Rural of total) 100.0 100.0 100.0 0 Agriculture 50.3 1.9 62.6 0 Nonagriculture 49.7 98.1 37.4 1 12.9 22.9 10.4 0 6.1 5.2 6.3 4 Manufacturing 13.9 30.7 9.6 5 3.0 7.5 1.8 4 Construction 13.1 30.3 8.8 0 Sales 7.0 1.5 0.5 5 Transport Services Other Source: Labor Force Survey, 1998. 75 percent share in electronic products, a 64 percent share in jewelry, and a 45 percent share in plastic products. In 1996, just before the crisis, employees made up the largest occupa- tional category, accounting for 14 million workers, or just under half the labor force of 30 million (table 6.6). Self-employed workers were the next largest category, with just under 10 million (one-third of the labor force), fol- lowed by unpaid family workers with 6 million (one-fifth of the labor force). Women are still underrepresented in the wage employee category (25 percent, compared with 30 percent for men) and the self-employed cate- gory (18 percent, compared with 41 percent for men) (ILO 1999). And in fact women are strongly concentrated in the family worker category. At 54 percent their share is twice that of men's. While leading export indus- tries may have been strongly feminized, women constituted a high pro- portion of unskilled labor in these industries, with correspondingly low proportions in engineering and management (AIT and ILO 1999). High output growth resulted in a tight labor market before the crisis, and the unemployment rate stood at just 2 percent of the labor force in 1998, or a little more than half a million unemployed (table 6.2). Given the high proportion of seasonal unemployment in total unemployment, a 2 percent unemployment rate verged on full employment. (Seasonal fluc- GROWTH, CRISIS, AND COMPETITIVENESS IN THAILAND 253 TABLE 6.6 Employment by Status, February 1996-99 (in millions) 1996 1997 1998 1999 Employers 0.84 0.77 0.78 0.94 Government employees 2.31 2.42 2.67 2.64 Private employees 11.68 11.64 10.68 10.22 Self-employed 9.37 9.30 9.48 10.09 Unpaid family workers 5.90 6.14 5.80 6.13 Total labor force 30.10 30.27 29.41 30.02 Source: Labor Force Survey, ILO, CEPR. tuations, which render a portion of the agricultural workforce inactive each year, may have accounted for up to 2 million unemployed in the wake of the crisis, or more than half the total unemployment then.) The tight labor market resulted in real wage growth of approximately 3.5 percent per year between 1987 and 1997, or approximately the same as real wage growth in Malaysia (but only around half that in Korea) (table 6.7). Women's wages as a proportion of men's wages are estimated to have increased from 64 percent in 1991 to 71 percent in 1994. This proportion dropped down to 68 percent in 1995, however (ILO 1999). It is possible that the increased supply of women workers became a source of cheap labor, raising the wage differentials. Despite the precrisis tightening of the labor market, two of its charac- teristics endured: the dualism inherent in the existence of a burgeoning in- formal sector alongside the formal sector, and underemployment (ILO 1999). Data from the Labor Force Surveys show that the informal labor market has not changed significantly over the last decade (ILO 1999). In fact in absolute terms informal employment increased from 1.8 million to 2.4 million in 1988-96. Underemployment, defined here as working fewer than 35 hours per week, stood at an estimated 2.43 million in 1997, or a significant 7.6 percent of the labor force (NSO 1999). Sources of Output Growth and Employment Two primary sources of growth and job creation are important to Thai- land's future growth. The first is the export sector, which has increased in importance throughout the 1990s. The second (and related) source is gross domestic investment, aided by capital inflows. TABLE 6.7 Manufacturing Wage Rates Thailand (baht/month) Korea (won/month) Philippines (pesos/month) Malaysia (ringgit/month) Nominal Real Nominal Real Nominal Real Nominal Real wage wage wage wage wage wage wage wage Year rate CPI rate rate CPI rate rate CPI rate rate CPI rate 1988 - - - 393.1 100.0 393.1 2995 100.0 2995 620 100.0 620 1989 2996 100.0 2996 491.6 105.8 464.7 3441 112.1 3070 640 102.8 623 1990 3357 105.9 3170 590.8 114.7 515.1 4263 128.0 3330 660 106.0 623 1991 3688 111.9 3296 690.3 125.4 550.5 4831 151.9 3180 719 110.6 650 1992 3986 116.5 3421 798.5 133.3 599.1 5386 165.5 3250 794 115.9 685 1993 4138 120.4 3437 885.4 139.7 633.8 5584 178.0 3133 848 119.9 707 1994 4229 126.6 3340 1002.5 148.3 689.5 6272 194.2 3230 928 124.4 746 1995 4994 133.9 3730 1123.9 155.0 725.1 6654 209.8 3172 1002 128.7 779 1996 5502 141.7 3883 1261.2 162.6 775.2 1997 5935 149.5 3970 1326.2 169.9 780.6 Annual growth, 1989-97 (percent) ... ... 3.5 ... ... 7.6 ... ... 0.8 ... ... 3.3 - not available. ... not applicable Source: ILO (1999a). GROWTH, CRISIS, AND COMPETITIVENESS IN THAILAND 255 THE EXPORT SECTOR. Sources of growth are indicated by the compet- ing demand for resources within an economy. In Thailand the largest in- crease in demand for resources has come from the export sector, which ex- panded by 15 percent between 1980 and 1996, raising its share of GDP from 24 to 39 percent (table 6.8). Between 1986 and 1996, indicators of Thailand's integration into the world economy more than doubled (table 6.9). External trade rose from 15 percent of GDP (in terms of purchasing power parity) to 31 percent. Of the other countries in the region, only Malaysia and the Philippines matched this achievement. Increased demand in the export sector resulted in a very high growth rate for exports. Thailand's export growth rate in 1980-90 was 14 percent, and it increased further in 1990-95 to 21 percent, for an average growth rate for 1980-96 of 17 percent. In each period Thailand's export growth rate was the highest in the region. In value terms Thai exports rose from $6 billion in 1980 to $56 billion in 1996. Manufactured exports showed the strongest growth. The share of food and raw materials in exports fell from almost two-thirds to one-third, while the share of manufactures in exports rose from one-quarter to three-quarters. Lall (1999) argues that this high export growth was based on a shift in the structure of the export sector, where complex activities began to re- place simple production. The sector encompassed four types of technolo- gies: resource-based (food processing), low (textiles, leather, and plastics), medium (the automotive industry), and high (complex electronic and electrical products). Between 1985 and 1996, the share in exports of prod- ucts manufactured using medium and high technology rose from among the lowest in the region (20 percent) to the highest (50 percent) (table TABLE 6.8 Structure of Demand (percent of GDP) 1980 1996 Private consumption 65.0 55.0 Government consumption 12.0 10.0 Gross domestic investment 29.0 41.0 Exports 24.0 39.0 Imports 30.0 44.0 Gross domestic savings 23.0 35.0 Note: Total GDP = private consumption + government consumption + investment + exports - imports. Source: World Bank 1998. 256 EAST ASIAN LABOR MARKETS TABLE 6.9 Integration with the Global Economy Trade Gross private (as percent of PPP GDP) capitalflows 1986 1996 1986 1996 Thailand 14.7 31.3 1.6 5.0 Korea 33.6 46.7 3.5 11.1 Indonesia 10.7 13.6 2.0 2.1 Malaysia 33.6 70.2 2.8 4.6 Philippines 8.0 21.3 2.3 4.8 Source: World Bank 1998. 6.10). Thailand's medium- and high-technology product export shares exceeded those of China, Hong Kong (China), and Indonesia, although those of Korea, Singapore, and Taiwan (China) were higher. INVESTMENT AND INFLOWS. Gross domestic investment rose by 12 per- cent of GDP between 1980 and 1996, climbing from 29 to 41 percent (table 6.8). Its largest component, gross domestic savings, rose by 12 percent over this period, bringing its share of GDP up from 23 to 35 percent. These very high levels of investment and savings helped fuel growth in Thailand. Capital inflows filled the gap between the domestic savings rate and the domestic investment rate. In the 1980s capital inflows doubled, rising to $4.5 billion per year, and between 1990 and 1996 they tripled to $14 bil- lion per year. Thailand's private capital flows (5 percent of GDP) com- peted well with those of Indonesia, Malaysia, and the Philippines. Foreign direct investment as a proportion of inflows fell gradually over time from about two-thirds in 1988 to 17 percent by 1996. Portfolio investment and bank and trade lending became the major form of capital inflows. The in- crease in the volume of capital inflows raised the current account deficit. In 1980-85 the gap between domestic savings and domestic investment was filled by a current account deficit of 6 percent of GDP. This current account deficit had risen to approximately 8 percent by 1996, by one es- timate (table 6.11). Weaknesses in the Labor Market and the Economy Even as Thailand's growth peaked in the mid 1990s, the sources of that growth were creating hidden weaknesses in the economy. The competi- GROWTH, CRISIS, AND COMPETITIVENESS IN THAILAND 257 TABLE 6.10 Structure of Manufactured Exports (percent) 1985 1996 Economy RB LT MT HT RB LT MT HT Argentina 67.5 15.6 11.8 5.1 49.1 18.8 28.8 3.3 Brazil 32.6 33.3 27.1 7.1 25.6 31.8 34.0 8.6 China 11.7 57.1 21.8 9.4 9.8 56.3 13.4 20.6 Hong Kong (China) 2.1 64.3 14.2 19.3 4.4 52.7 14.0 28.9 India 40.3 46.1 10.6 3.0 31.1 52.3 13.1 4.4 Indonesia 72.2 19.2 5.9 2.8 34.9 41.9 8.5 14.7 Korea 7.8 59.9 12.2 20.1 9.4 28.4 26.6 35.7 Malaysia 53.7 9.7 5.5 31.0 17.8 13.1 8.7 60.4 Mexico 20.2 15.0 29.2 35.6 7.1 20.9 35.2 36.9 Singapore 42.3 10.8 14.6 32.3 12.7 7.9 14.0 65.4 Taiwan (China) 8.7 57.3 13.3 20.7 5.1 33.9 20.2 40.9 Thailand 42.1 38.2 6.6 13.1 14.5 35.6 13.5 36.3 Turkey 22.0 62.3 13.4 2.3 17.5 63.9 12.8 5.7 Note: China's export structure for 1985 is based on 1990 figures. RB = resource-based. LT = low-technology. MT = medium-technology. HT = high-technology. Source: Lall (1999). tiveness in manufacturing that had fueled high export growth throughout the 1980s and early 1990s began to erode. The high levels of inflows and investment may even have helped disguise the decline in competitiveness. But when the currency crisis reversed the capital inflows, the decline in competitiveness became clear. Thailand's exports fell during the global trade slump and did not recover. The very high investment levels and cap- ital inflows may also have led to inefficiencies in the use of capital, con- tributing to a decline in capital productivity throughout the 1990s. LABOR PRODUCTIVITY AND THE EROSION OF COMPETITIVENESS. Manu- factured exports grew at 23 percent in 1993-94 and at 26 percent in 1994-95. The value of manufactured exports was $46 billion in 1995. Of this amount medium- and high-technology products had the largest share ($24 billion) (table 6.11). These exports are considered critical to the transformation of the manufacturing base from simple to complex tech- 258 EAST ASIAN LABOR MARKETS nology. The slump in world trade in 1995-96, from trend growth of 9 percent to 2 percent, reduced Thai manufactured exports to negative growth (Lall 1999). They have not recovered from that jolt. In 1996-97 manufactured exports grew by 5 percent, but in 1997-98 growth of these exports once again was negative. What are the winners and losers among Thailand's exports? Thailand's medium- and high-technology manufactured exports appear to be more competitive than low-technology and resource-based products. Despite the global slump medium- and high-technology manufactured exports have grown on trend by 12 percent, rising from $24 billion in 1995 to $27 bil- lion in 1998. In contrast low-technology exports declined from $12 billion in 1995 to $8 billion in 1998, and resource-based products have stagnated at $6 billion. Within medium- and high-technology exports, two impor- tant product groups-hard-disk drives and semiconductors-have contin- ued growing throughout the crisis, albeit at a lower rate (Lall 1999). If labor- and resource-intensive manufactured exports have been falling since the mid-i 990s, then the competitiveness of Thailand's labor-based products has declined. 'While competitiveness has many dimensions- including costs, product types, and quality-essentially the problem of competitiveness is one of labor productivity. One important indicator of competitiveness (in this case, of labor-based products) is unit labor cost (ULC). In 1990 Thailand's ULCs were higher than those for China, Indo- nesia, Korea, Malaysia, or the Philippines (table 6.12). China and Indonesia had lower ULCs in virtually all product categories. ULCs for the Philippines were lower in all product categories except textiles, wood, and fabricated metals. And ULCs in Malaysia were lower for chemicals and metals. Three major components determine the ULC. They can be written as follows: ULC = f (wage rate, labor productivity, exchange rate) The ULC increases as the real wage and exchange rate rise and decreases as labor productivity rises (Mahmood 1998). Thailand's ULCs, which were not particularly competitive for labor- intensive products even at the end of the 1990s, were the result of wage increases that outstripped labor productivity increases throughout the decade (TDRI 1998). The real wage rate increased at the rate of about 3.5 percent annually throughout the 1990s (table 6.7). In comparison Korea had much higher wage growth, Malaysia about the same level of growth, and the Philippines lower wage growth. Labor productivity TABLE 6.11 Recent Thai Manufactured Export Performance Values (US$million) 1993 1994 1995 1996 1997 1998 Low-technology 9,048.1 10.669.2 12,123.4 10,201.4 9,518.7 8,422.6 Medium-/ high-technology 15,067.0 17,830.5 23,680.4 24,573.9 26,611.9 26,557.5 Resource-based 3,970.2 4,668.2 5,758.8 6,351.0 6,632.7 5,692.3 Other manufactures 1,636.8 3,523.5 4,640.1 4,310.4 4,704.2 4,040.0 Total manufactures 29,722.2 36,691.5 46,202.7 45,436.7 47,467.5 44,712.4 Growth rates (percent) 1993-94 1994-95 1995-96 1996-97 1997-98 1993-98 Low-technology 17.9 13.6 -15.9 -6.7 -11.5 -1.4 Medium-/ high-technology 18.3 32.8 3.8 8.3 -0.2 12.0 Resource-based 17.6 23.4 10.3 4.4 -14.2 7.5 Other manufactures 115.3 31.7 -7.1 9.1 -14.1 19.8 Total manufactures 23.4 25.9 -1.7 4.5 -5.8 8.5 Source: Lall (1999). TABLE 6.12 Unit Labor Costs in Manufacturing, Selected Asian Economies, 1990 Chemicals Fabricated Paper (including metal Total Textiles, Wood products, petroleum, Pottery, Iron, steel, products, Other manufac- Food, apparel, products, publishing, rubber, china, glass nonferrous machinery, manufac- Economy turing beverages footwvear furniture printing plastics) products metals equipment turing East Asia Chinaa 0.57 0.74 0.32 0.37 0.51 0.58 0.51 0.44 0.44 0.41 Hong Kong (China) 1.94 1.59 1.85 1.86 1.51 1.64 1.43 1.11 1.46 2.46 Korea, Rep. of 0.99 0.81 1.20 1.20 0.90 0.85 0.95 0.61 0.87 1.55 Singapore 1.12 1.58 1.57 1.57 1.04 0.99 0.94 0.73 1.07 1.80 Taiwan (China) 1.45 1.23 2.31 2.31 1.37 1.50 1.25 0.70 1.43 1.16 Southeast Asia Indonesia 0.71 1.15 0.79 0.84 0.80 1.01 0.83 0.16 0.68 1.15 Malaysia 0.95 1.04 1.35 1.31 1.03 0.69 0.80 0.65 0.79 1.41 Philippines 0.85 0.77 1.44 1.38 0.88 0.75 0.84 0.28 1.06 1.96 Thailand 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 South Asia Bangladesh 1.16 0.88 0.92 1.20 1.23 0.67 0.73 0.89 1.06 0.68 India 1.52 1.99 1.50 1.88 1.66 0.97 1.36 0.81 1.30 1.71 Pakistan 0.76 0.65 0.93 1.04 1.28 0.68 0.74 1.22 0.97 1.14 Sri Lanka 0.62 0.42 0.86 1.46 0.81 0.62 1.06 0.59 0.91 0.85 Note: Thailand= 1. a. Data are for 1986. Source: United Nations Industrial Development Organization industrial statistics database. GROWTH, CRISIS, AND COMPETITIVENESS IN THAILAND 26i in manufacturing, which in the second half of the 1980s had risen by 40 percent, increased by much less in the second half of the 90s-only 15 percent (table 6.13). Why did wage increases outstrip productivity increases in Thailand? The answer is that the economy turned the Lewisian corner in 1990 and ran straight into labor shortages (Warr 1998). In the 1980s agriculture had already experienced low labor productivity and a labor surplus that could be transferred to manufacturing, raising average productivity without a proportional increase in the wage rate. As the supply of surplus agricul- tural labor ran out in the early 1990s, peak season shortages developed and agricultural labor productivity increased. Soon the tighter labor mar- ket pushed up wages. Labor productivity may also be constrained by a relative lack of skilled workers in Thailand. In 1995 Thailand had the second-lowest secondary school enrollment ratios in the region (only Indonesia's were lower) (table 6.14). Thailand's tertiary enrollment ratios were higher only than those in China, Indonesia, and Malaysia. Similarly tertiary enrollment in technical subjects like mathematics was the lowest in the region, and enrollments in engineering and technology were among the lowest.4 Another factor contributing to the rise in ULCs is real exchange rate ap- preciation. The real exchange rate (the rate at which commodities are traded) is strongly affected by movements in the nominal exchange rate. The standard argument is that the Thai baht appreciated because it was pegged to the U.S. dollar, which had appreciated against the yen. But this scenario occurred only after 1995. Between 1990 and 1995, the dollar did not appreciate against the yen. Evidence is now emerging that the real ex- change rate appreciated because of the phenomenon known as Dutch Dis- ease (Warr 1998). With a fixed exchange rate like Thailand's, the price of traded goods is determined by the international market price, even when demand increases. But the price of domestic nontraded goods is bid up when demand increases, reducing average competitiveness in the economy. INEFFICIENCY IN THE USE OF CAPITAL. Investment peaked at 41 percent of GDP in 1996, domestic savings at 33 percent, and capital inflows at 8 percent. But weaknesses were emerging in the way capital was used. One measure of efficiency in the use of capital is the incremental capital out- put ratio (ICOR), which shows how much capital is needed to produce a unit of output. The smaller the amount needed, the lower the ICOR. A lower ICOR means increased efficiency in both the use of capital and the economy. Thailand's ICOR was 5 in 1980-85, but it dropped to 3 in TABLE 6.13 Labor Productivity Sector 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 Average 88.6 93.5 100.0 110.8 112.1 130.1 135.8 143.3 161.0 169.2 178.0 Agriculture *97.8 100.4 100.0 111.9 92.0 124.9 128.0 123.7 148.2 151.0 165.0 Mining 86.1 98.9 100.0 144.0 136.7 135.2 148.8 166.8 203.0 193.1 227.6 Manufacturing 83.0 80.8 100.0 97.6 119.9 103.0 107.5 115.9 121.8 127.8 134.5 Construction 96.3 87.8 100.0 109.5 123.4 92.0 85.4 101.3 85.4 89.3 80.0 Utilities 87.1 88.2 100.0 121.6 144.7 142.4 153.8 143.4 131.2 146.4 168.1 Transport 88.9 84.8 100.0 104.8 114.0 104.5 114.7 119.8 129.4 134.9 147.7 Trade 85.8 84.6 100.0 111.1 133.1 Public admin. and others 100.0 Source: ILO (1998). TABLE 6.14 Educational Enrollment, 1999 (or most recent years) Enrollments Tertiary level technical enrollments (percent) Natural science Math/computing Engineering Total technology Country Secondary Tertiary Number % Number % Number % Number % Argentina 67 36 69,727 0.21 ... ... 96,205 0.29 165,932 0.49 Brazil 46 11 46,322 0.03 92,701 0.06 149,660 0.10 288,693 0.19 China 55 4 95,492 0.01 174,862 0.02 1,156,735 0.10 1,427,089 0.13 France 106 50 304,093 0.53 ... ... 50,845 0.09 354,938 0.62 Germany 101 36 310,435 0.39 ... ... 389,182 0.49 699,617 0.88 Hong Kong (China) n.a. 21 5,503 0.09 6,441 0.11 14,788 0.25 26,732 0.46 India 49 6 869,119 0.10 ... ... 216,837 0.02 1,085,956 0.12 Indonesia 45 10 22,394 0.01 13,117 0.01 205,086 0.11 240,597 0.13 Japan 98 29 59,030 0.05 20,891 0.02 488,699 0.39 568,620 0.46 Korea, Rep. of 99 55 81,222 0.18 171,147 0.38 437,537 0.98 689,906 1.55 Malaysia 61 10 8,776 0.05 4,557 0.02 12,693 0.07 26,026 0.14 Mexico 58 14 42,457 0.05 97,575 0.01 221,867 0.27 361,899 0.45 Philippines 79 27 27,200 0.04 121,000 0.18 225,700 0.33 373,900 0.55 Singaporea 68 46 1,281 0.05 1,420 0.05 13,029 0.47 15,730 0.56 Taiwan (China) 88 38 16,823 0.08 32,757 0.16 179,094 0.86 228,674 1.09 Thailand 49 21 77,098 0.14 1,292 0.00 105,149 0.19 183,539 0.32 UK 94 41 105,983 0.18 76,430 0.13 219,078 0.38 401,491 0.69 USA 97 81 496,415 0.19 525,067 0.20 801,126 0.31 1,822,608 0.70 Source: UNESCO, Statistical Yearbook, various years; Government of Taiwan Statistical Yearbook 1994; and Lall (1999). a. Singapore's tertiary enrollment figures include polytechnics, which enroll 27 of the age group. n.a. not available. ... not applicable (included in another discipline). 264 EAST ASIAN LABOR MARKETS 1985-90, rose again to 5 in 1990-95, and then peaked at 6 in 1996 (table 6.15). In other words, the Thai ICOR doubled during the 1990s, halving the efficiency of capital use and of the economy. In 1985-90 Thailand had the lowest ICOR in the region and the high- est GDP growth (10 percent annually). But in the latter half of the 1990s the situation reversed itself, and the Thai ICOR rose to 5-the highest in the region. Countries with lower levels of investment, including China, Malaysia, and Singapore, achieved higher or equivalent growth rates. The very high levels of investment and capital inflows in Thailand, which peaked in the 1990s, appear to be associated with this drop in the effi- ciency of capital use. The high ICORs and inefficient use of capital reflect a substantial re- duction in the rate of technical change in the economy. Technical change can be measured by total factor productivity (TFP), which shows the change in output for one unit of input. TFPs for Thailand follow the same pattern across time as ICORs (table 6.16). In the first half of the 1980s, when the ICOR was 5, the TFP stood at a low 0.5. In other words tech- nical change was negative over this period, when a unit of capital and labor produced less than a unit of output. In the second half of the 1980s, when the ICOR fell to 3, the TFP was high (over 3). And from 1990 to 1995, when the ICOR doubled to 5 and then rose to 6, the TFP fell to near zero. The highly negative TFP of -4 in industry and manufacturing was largely responsible for the near-zero aggregate TFP. The high level of capital inflows are also partly responsible for the high ICORs and falling TFPs. Paradoxically macroeconomic policies designed to deal with the inflows may have heightened their impact (Warr 1998). Macroeconomic policy deals with inflows by sterilizing them through the sale of central bank bonds. Sterilization offsets the impact of capital TABLE 6.15 Savings, Investment, and Growth 1980-85 1985-90 1990-95 1996 Output growth 5.4 10.3 8.0 6.4 ICOR 5.3 3.1 5.1 6.4 Investment/GDP 28.8 31.8 41.1 41.0 Domestic savings (% of GDP) 22.9 25.5 33.6 33.0 Current account deficit (% of GDP) 5.9 6.3 7.5 8.0 Source: Dwor-Frecaut (1998). GROWTH, CRISIS, AND COMPETITIVENESS IN THAILAND z65 TABLE 6.16 Total Factor Productivity Growth Rates GDP Land Capital Labor TFP Total 1981-85 5.45 0.16 3.39 1.37 0.53 1986-90 10.34 -0.02 4.92 2.20 3.24 1991-95 8.56 -0.04 6.73 1.84 0.03 1981-95 8.12 0.03 5.01 1.80 1.27 Industry 1981-85 6.47 ... 5.58 2.76 -1.87 1986-90 14.42 ... 7.07 3.83 3.52 1991-95 10.62 ... 8.97 5.30 -3.65 1981-95 10.50 ... 7.20 3.97 -0.67 Manufacturing 1981-85 4.99 ... 3.41 2.85 -1.27 1986-90 15.10 ... 7.19 3.93 3.98 1991-95 10.95 ... 8.28 5.74 -3.07 1981-95 10.35 ... 6.29 4.18 -0.12 Services 1981-85 5.33 ... 3.99 2.79 -1.45 1986-90 10.01 ... 5.22 3.26 1.53 1991-95 8.15 ... 6.71 2.73 -1.29 1981-95 7.83 ... 5.30 2.93 -0.40 ... not applicable. Source: TDRI 1998. inflows and the resulting current account deficit on the money supply. As the public buys bonds, the money supply shrinks and interest rates rise. But another irony is at work: as rates of return rise, more portfolio capital flows into the economy. The added inflows increase the money supply, nullifying the sterilization and increasing demand (the Mundell- Flemming effect). Increased demand bids up the price of nontraded goods (Dutch Disease), attracting more resources to this sector than to the traded goods sector and decreasing the efficiency of the economy. Certain macroeconomic conditions can help economies avoid the Dutch Disease dilemma. A current account deficit is more sustainable if capital inflows are invested in exports (Dwor-Frecaut 1998). Net export growth should be greater than the interest rate on foreign savings, or the high interest rate will simply attract more portfolio capital inflows to the nontradables sector (bypassing the tradables sector). In Thailand these macroeconomic conditions may have been violated, leading to major losses in efficiency and competitiveness. The interest rate spread-the dif- ference between the lending and deposit rates, or the lending and London 266 EAST ASIAN LABOR MARKETS Interbank Offer Rates (LIBOR)-is the incentive for foreign savings to flow into an economy. Thai spreads rose in 1990-96 from 4 to 6 percent. In comparison spreads in Korea ranged from 0-1 percent. Similarly Thai spreads in terms of the LIBOR also rose, climbing from 8 to 10 percent, while Korean spreads ranged between 2 and 3 percent. These high spreads competed with exports, increasing capital inflows to 8 percent of GDP. But these inflows were increasingly and predominantly portfolio capital, so that large amounts of both foreign and domestic savings were chan- neled into nontradables, especially real estate, resulting in a concomitant decline in capital efficiency, technical change, and competitiveness. A macroeconomic policy that maintains high spreads and therefore even higher interest rates has a major impact on productivity and profitability through yet another channel. A survey of 1,227 Thai firms shows that the top 30 percent have a high TFP (approximately 175) while the bottom 70 percent have a very low TFP (approximately 30) (Dollar and Hallward- Driemeirer 1998). Small and medium-sized firms (SMEs) have low levels of productivity and technical change because of the high interest rates they face and are acknowledged to be starved for credit (Colaco 1998). Finally, a survey of firms shows that their profitability had declined considerably prior to the crisis (Dollar and Hallword-Driemeirer 1998). Net profits relative to capital and retained earnings had dropped from 17 to 4 percent. The high real exchange rate was largely responsible for this de- cline, along with the high cost of capital. The tradable sector is unable to bid up prices in a fixed exchange rate regime in an open economy that al- lows international markets to set prices. The Impact of the Crisis on the Labor Market There are several ways to estimate the impact of the crisis on the labor market, but the methods vary according to the rigor of the estimation. We look at unadjusted estimates, including raw employment levels, and esti- mates adjusted for such factors as seasonal unemployment. In 1997 em- ployment stood at 30.3 million, but it fell to 29.4 million in 1998 (table 6.17). Although the employment level rose again in 1999, it did not reach the 1997 level. Unadjusted Estimates Unemployment jumped from 0.7 million in 1997 to 1.7 million in 1999. Official statistics show that recorded layoffs increased from 5,000 in 1996 to 42,000 in 1997 and 52,000 in 1998 (ILO 1999). Women constituted GROWTH, CRISIS, AND COMPETITIVENESS IN THAILAND 267 a somewhat higher proportion of those laid off (53 percent). Because of the preponderance of women workers in the export sector and the decline in exports, women were severely hit in particular industries. Women as a proportion of all those retrenched equaled 95 percent in garments, 88 per- cent in toys, 80 percent in knitting and electrical appliances, 73 percent in jewelry, 71 percent in plastic products, and 68 percent in shoes and leather products. The crisis has generated underemployment as well as unemployment. Employers wishing to avoid severance payments have an incentive to re- duce working hours rather than laying workers off altogether. Labor Force Survey data show that underemployment increased from 2.43 million (7.6 percent of the labor force) in 1997 to 4.41 million (13.7 percent of the labor force) in 1998 (NSO 1999). The unemployment and underemployment levels do not show the full impact of the crisis, however. Declines in the wage rate are an important indicator of the impact on income, because very low-income workers fac- ing total loss of income will take jobs at even lower rates of pay. Labor Force Survey data show that between February 1997 and February 1998, the nominal wage rate dropped by 6 percent, while the real wage rate dropped by 4 percent (ILO 1999). The decline in urban areas (8.3 per- cent) was almost double the decline in rural areas (4.7 percent). At the ag- gregate level men and women seem to have experienced the same levels of decline. But in urban areas women's real wages eroded by 10.5 percent, more than the decline in men's wages. In rural areas, however, the drop in men's real wage rate was higher than it was for women's. In urban sectors the steepest declines in real wage rates were in manufacturing (13 percent) and construction (24 percent). The unemployed had recourse not only to work at lower wages but also to work in the informal sector, self-employment, and migration to rural areas. Table 6.6 reflects the shift in the occupational structure prompted by the crisis. Of an employed labor force of 30 million in 1997, private sector employees constituted the largest group (11.6 million). By 1999 this group had shrunk by 1.4 million, but the number of self-employed workers had increased by 0.8 million. This shift out of wage employment into self-employment represents a shift primarily to the informal sector in urban areas and to the agricultural sector in rural areas. Adjusted Estimates Estimates made using data from the Labor Force Survey are based on one particular indicator (such as employment) at one point in time. Even TABLE 6.17 Labor Force Status, 1994-99 (February) 1994 1995 1996 1997 1998 1999 Total population 59034.9 59112.9 59750.4 60350.7 60949.0 61551.2 Total labor force 31049.9 31347.9 31898.4 32000.2 32143.1 32810.2 Employed 28233.5 29055.1 30099.2 30266.4 29412.9 30024.5 Unemployed 1244.4 723.5 641.3 697.9 1479.3 1715.7 Looking for work 200.1 167.8 119.6 179.6 402.8 475.8 Available for work 1044.2 555.7 521.7 518.3 1076.5 1239.8 Seasonally inactive 1571.9 1569.2 1157.8 1035.9 1250.8 1069.8 Unemployment rate (%) 4.0 2.3 2.0 2.2 4.6 5.2 Open unemployment rate (%) 0.6 0.5 0.4 0.6 1.3 1.5 Source: Labor Force Survey, ILO, CEPR. GROWTH, CRISIS, AND COMPETITIVENESS IN THAILAND 269 comparisons of a single indicator at two points in time do not necessarily reflect a specific event such as the crisis. The change in the indicator over time must be adjusted for the internal dynamic of the economy. In Thai- land's case two important factors integral to this internal dynamic that need to be considered are seasonality and the trend growth of the economy. The impact of the seasons on labor market variables.is high in Thai- land. There are two distinct seasons-dry (winter and spring) and wet (summer and autumn). Kakwani (1998) estimates the seasonal impact on employment at a highly significant 10 percent-that is, employment in the dry season drops by 3 million (out of a labor force of 30 million). Women are the most seriously affected (2.2 million). The second factor, the trend growth of the economy, shows us what would have happened to an indicator like employment had the crisis not occurred. Kakwani takes Labor Force Survey data for six wet and six dry seasons for the years 1992-98 and adjusts them for seasonality and trend growth. He then creates a "crisis index" for 1997-98 that estimates the impact of the crisis on the indicator. The estimates show that the trend growth of employment in Thailand resulted in 0.22 million jobs per year before the crisis. Calculating the immediate impact of the crisis involves comparing wet season data for 1997 to dry season data for 1998. The un- adjusted data show that employment dropped by 3.8 million (from 33.2 million to 29.4 million). But adjusting for seasonality and trend growth, we find that employment dropped by only 0.7 million (from 31.7 million to 30.9 million). Trend growth in unemployment was negative before the crisis, with the number of unemployed falling by 0.08 percent per year. The gender gap in unemployment rates was narrowing, with women's unemployment rates falling faster than men's (-0.18 percent compared with -0.1 per- cent). Seasonality accounted for a loss of 0.52 million jobs in each dry sea- son. Given these variables the adjusted crisis index shows that the crisis generated unemployment of 2.6 percent (2.9 percent for men but only 2.2 percent for women). The result is an adjusted unemployment level for the first quarter of 1998 of 3.9 percent. Adjusted estimates of formal sec- tor unemployment place the level higher, at 5.4 percent, twice as high as informal sector unemployment of 2.6 percent. Growth in underemployment was also negative before the crisis, with underemployment falling at an annual rate of-0.0 1 percent. The adjusted estimates show that the crisis increased underemployment by 0.11 mil- lion, raising the underemployment level to 0.36 million (0.19 million men and 0.16 million women) in the first quarter of 1998. 270 EAST ASIAN LABOR MARKETS Kakwani (1998) estimates that real wage growth was experiencing an upward trend before the crisis, rising at an annual average rate of TB 188 per month, with the gender gap decreasing over time. He also estimates that the adjusted average real wage dropped by TB 506 per month because of the crisis (at the basic rate) and by TB 629 per month (with overtime reductions). The impact on men's wages was more pronounced than the impact on women's wages. Kakwani also estimates a standard-of-living index by adjusting for re- gional variation in prices. According to his estimates the standard of liv- ing was increasing at a rate of 6.8 percent annually before the crisis and dropped 21 percent after the crisis began. Disaggregating this figure shows that the largest contribution is a decline in income per earner of 19 per- cent. A less significant contribution (2.7 percent) is attributable to a drop in employment, and the least significant contribution (0.4 percent) can be traced to a drop in the LFPR. The crisis caused poverty to increase in 1998 by 2 percent, raising it to 13 percent (table 6.2). The occupational impact of the crisis was especially severe on wage employees. Wage and salary employment dropped by an estimated 1.2 mil- lion, while farm employment picked up by 0.6 million. Wage income over- all dropped by an estimated TB 768 per month. The impact was particu- larly profound in construction, manufacturing, trade, and banking, with the adjusted real wage in manufacturing dropping by an estimated TB 857 per month. The crisis did not significantly affect farm income, however, confirming that an occupational shift took place. Migration to rural areas became a significant strategy for maintaining income among the urban unemployed. The scope of internal migration is difficult to estimate, not the least because Thailand's population has always had migratory habits. Of the 47 million Thais older than 13 in 1998, approximately 2.3 million had migrated in the last year and 5.2 million in the last 5 years. Recent migrants make up one of two groups that have been particularly vulnerable during the crisis. The unemployment rate for recent migrants has increased 13 percent since the crisis began, and their income has dropped from an average of TB 3,765 per month to TB 3,294. The other particularly vulnerable group is children. Kakwani (1998) estimates the incidence of child labor in Thailand at 15 percent in 1998. But even this high level represents a decline in the trend growth of child labor of over 5 percent per year. The adjusted crisis index shows that child labor increased by 0.4 million because of the crisis. GROWTH, CRISIS, AND COMPETITIVENESS IN THAILAND 271 The Policy Response to the Crisis: A Matter of Causality In essence there are four explanations for the Asian financial crisis, each with its own policy implications. We offer a fifth explanation based on a consensus that offers more comprehensive options for designing policy instruments. The first position argues that the crisis was induced by currency policy (IMF 1998). An overvalued pegged exchange rate reduces export compet- itiveness, creates a current account deficit, and depletes foreign exchange reserves. In this situation tight monetary and fiscal policies, especially higher interest rates and tight credit controls, can help to prevent further exchange rate depreciation and inflation. A second position characterizes the Asian crisis as a financial panic. Thai banks especially borrowed short-term, unhedged capital from international markets and lent it to the corporate sector on a long-term basis against in- flated assets. The banks and the corporate sector were highly leveraged. When portfolio investors saw that the exchange rate regime was unsus- tainable, they panicked, reversing the huge inflows. The panic became a self-fulfilling prophecy created by investors withdrawing capital from in- debted but solvent borrowers in a situation that offered no lenders of last resort to bail the borrowers out. The macroeconomic policy prescription in these cases is to protect the economy through lenders of last resort. A third position holds that the Asian financial crisis was caused by moral hazard (Lee 1998; Krugman 1998). Moral hazard occurs when banks borrowing with public guarantees use the funds for overly risky ven- tures. These banks and their risky debtors are prone to failure, triggering downturns in economic activity. Since moral hazard is based on collusion among the government, banks, and the corporate sector, and government intervention has gone wrong, the policy position is to reduce government intervention. The fourth explanation maintains that a lack of government inter- vention caused the crisis. The governments of the region had a solid 20- year track record of generating high growth and lowering poverty through judicious regulation and intervention. The problems underlying the crisis arose when the governments deviated from their successful policies, prin- cipally through poorly managed financial liberalization. The policy posi- tion suggested by this viewpoint, then, is a return to the regulatory poli- cies of the past. 272 EAST ASIAN LABOR MARKETS Our consensus-based explanation of the Asian crisis is grounded in the recognition that all three major markets-currency, capital, and labor- were experiencing problems simultaneously. All three had major imper- fections. While the imperfections in the currency and capital markets have been widely discussed the problems of the labor market have not been stressed and therefore have been assigned a low priority in the process of policy reform. THE CURRENCY AND CAPITAL MARKETS. The currency market suffered as the government attempted to maintain a pegged exchange rate with an open capital account. There were a number of capital market imperfections- for instance, the problem of moral hazard highlighted by Lee (1998) and Krugman (1998) and, perhaps most important, the inefficient patterns of investment. As we have seen, to a large extent the declining efficiency of capital was the result of a combination of the Mundell-Flemming effect and the Dutch Disease syndrome. Macroeconomic policy designed to san- itize inflows into Thailand reduced the money supply, increasing interest rates and paradoxically leading to higher inflows. As demand increased the prices of nontradables such as real estate rose, and more and more re- sources were channelled to the sector. The result was not only inefficient capital allocation but also a decline in competitiveness. Further, high interest rates and spreads raised the opportunity cost of capital, reducing productivity, profitability, and technological develop- ment among SMEs, which account for 70 percent of all employment in Thailand. High spreads increased the inflows but at the same time created a disincentive against hedging. When the capital flows were suddenly re- versed, the banks-which had borrowed short and lent long (in unhedged funds)-faced a liquidity crisis that flowed on to domestic firms, and both banks and firms faced closure. THE LABOR MARKET. As we have argued, the labor market suffered from declining competitiveness in all categories of manufacturing, but espe- cially in labor-intensive industries. Thailand's ULCs were not competitive in the region in the 1990s, in part because of growth in wages that out- stripped growth in productivity. This weakness is also in part attributable to Dutch Disease syndrome during the first half of the 1990s. As an in- creasing volume of resources flowed into nontradables, aggregate produc- tivity fell and the real exchange rate (and therefore ULCs) rose. And in the mid-1990s the nominal exchange rate, which was pegged to the U.S. dol- lar, appreciated against the yen, raising ULCs even more. GROWTH, CRISIS, AND COMPETITIVENESS IN THAILAND 273 Because of these weaknesses, Thai exports never really recovered from the export recession in East Asia that began in 1995. Declining competi- tiveness in both traditional and technology-intensive industries left Thai- land at a disadvantage. When the regional currencies depreciated in mid- 1997 and capital flows reversed, the liquidity problems of Thai banks and firms brought economic growth to an immediate halt. Thailand's growth was unsustainable because its fundamentals were wrong. It was based on simple input increases rather than on increased productivity and techno- logical advancement. At the time the crisis struck, the economy needed very high investment levels (financed through inflows) in order to con- tinue growing. The Debate over Policy Instruments The financial crisis (and arguably the current reform package) has severely compressed demand. Liquidity problems have kept Thai firms from main- taining past levels of production. Without the goods to export, the com- petitive edge provided by the depreciation in the real exchange rate means little. Thus import compression rather than export expansion has reduced the current account deficits. The crisis may have been deepened and prolonged by some of the do- mestic policies the Thai authorities followed. Monetary tightening may have accounted for as much as a quarter of the contraction in GDP be- tween 1997 and 1998 (IMF 1999). Monetary tightening may also have ex- acerbated liquidity problems for firms, especially SMEs. But the key prob- lem was the effectiveness of interest rates, the chosen policy instrument. The most recent studies show that the relationship between interest rates and the exchange rate (the variable that interest rates were supposed to in- fluence) was not the predicted positive but instead was highly ambiguous. In other words raising interest rates did not raise the exchange rate. Policy measures such as closing banks and raising interest rates could have increased the panic rather than stemming it. The authorities believed that tight money would attract capital. However the Kindleberger effect (which was operative here) introduced another element-the elasticity of investors' expectations. If interest rates rise and investors' expectations are not elastic (that is, investors do not expect a crisis), the higher interest rates will attract capital. But if interest rates rise and investors' expectations are elastic (that is, investors expect a change in prices or an exchange rate de- preciation), capital will flee. 274 EAST ASIAN LABOR MARKETS A New Macroeconomic Policy A new macroeconomic policy designed to increase the growth of output and employment needs to be based on trends in both demand and supply. The Thai economy is experiencing a severe compression of demand. Re- versing this situation requires reflation through expansionary fiscal and monetary policy and higher fiscal deficits to expand consumption. In order to improve the supply of goods and services, policies must be designed to increase competitiveness, particularly of SMEs. A more ex- pansionary monetary policy based on lower interest rates will help increase their survival rate, competitiveness, and employment capacity. Other pol- icies must address training and education, research and development R&D, and labor productivity. Weaknesses in all these areas prevent Thai- land from regaining its competitive standing. Given the credit bias against SMEs, a much higher proportion of credit needs to be directed to them. Improving the competitiveness of inefficient SMEs is more cost-effective than fostering marginal improvements in the competitiveness of larger enterprises that already have relatively high productivity levels. One institutional possibility would be to set up a trust fund to bail out the most vulnerable SMEs. In order to regain its competitive edge, Thailand also needs to improve skills training, vocational education, and R&D. The lack of much-needed skills contributes to the country's low rates of productivity and techno- logical change and low value added in medium- and high-technology manufacturing. Thailand has low enrollments in secondary and tertiary institutions as well as in technical areas and insufficient investment in R&D. Policies need to be designed to increase enrollments and technical training. To increase investment in R&D, the government must invest in capacity and provide incentives for private firms to increase funding for innovation. Increasing labor productivity requires a number of enabling conditions. First, it requires a forum for cooperative dialogue with employers in order to prevent conflict. Next, it requires effective training for management as well as labor. It requires incentives for firms to provide training, including financial support. And finally, because employment and welfare are inex- tricably linked to productivity, it requires attention to the conditions of employment. The benefits of productivity increases can be reaped only if firms retain their skilled labor by providing job security and a safety net that includes unemployment insurance, accidental death and injury in- surance, and a retirement system. GROWTH, CRISIS, AND COMPETITIVENESS IN THAILAND 275 The Response to the Crisis The government's actions in response to the crisis fall into three areas: em- ployment generation, employment services, and vocational education and training. In 1998, as the recession worsened, the government launched a social safety net program. The program aimed to: * Generate employment and stimulate the economy, especially through productive investment and labor-intensive job creation; * Mitigate the short-run impact of the crisis by providing income support and social services for the unemployed, the vulnerable, and the poor; * Develop the skills workers need to stay abreast of technological changes; and * Implement the structural reforms necessary to increase economic competitiveness. The government allocated a budget of approximately TB 20,000 mil- lion for the program. For the 1999 fiscal year, the allocation was TB 157,800 million, or 21 percent of the expenditure budget. Additional financial support from the World Bank, the Asian Development Bank (ADB), the Overseas Economic Cooperation Fund, and the Japanese Export-Import Bank has also been secured under various programs. These include a Social Sector Program ($500 million), a Social Investment Pro- gram ($462 million), and the Miyazawa Initiative ($1.5 billion).5 Job Creation and Employment Maintenance To create jobs and encourage reverse migration from urban to rural areas, in 1999 the government allocated TB 51 billion from its regular budget to provide direct and indirect employment in rural areas. Of this amount TB 18 billion was to be used to employ 800,000 people. Direct. employ- ment initiatives include 68 rural job-creation projects involving public works. These projects rehabilitate and construct economic infrastructure (roads, drainage, dredging, weirs, bridges, dams, airports, tourism facili- ties, utilities, and offices), construct public health centers, build low-cost housing, and help conserve forests. Indirect job-creation projects involve vocational development (such as handicrafts, vocational education, and rubber tapping) for youth and women, agricultural irrigation projects, and livestock projects. 276 EAST ASIAN LABOR MARKETS The Social Investment Program supports 5 main projects that generate 557,000 person-months of employment plus 16,375 road construction jobs. The work includes constructing village feeder roads, constructing and dredging weirs, rehabilitating and expanding small-scale irrigation networks, rehabilitating and constructing tourist facilities, rehabilitating and building schools, and upgrading civic amenities in the Bangkok urban area. The Social Structure Program Loan funds a project that provides em- ployment at village welfare centers for 13,000 unemployed and laid-off workers and new university graduates. Project participants assist welfare workers in improving services to villages. Of the TB 852 million budget approved for the project in December 1998, however, less than a quarter had been disbursed by March 1999. Of the Miyazawa Initiative budget of TB 53 billion, just under half has been used to create direct employment for unskilled and educated work- ers. The unskilled workers upgrade rural irrigation systems and construct small infrastructure projects, while some 89,000 educated workers assist with government services. Ministry of Interior projects received approxi- mately 35 percent of the total Miyazawa fund to generate employment for 24,000 educated workers and provide 50 million person-days of employ- ment for unskilled workers. Income-support Measures Income-support measures have included severance pay, tax and utility price cuts, a price support program for rice, and extension of health care and educational subsidies. As Thailand has no unemployment insurance scheme, the first (and primary) income support workers receive is sever- ance pay. The new Labor Protection Law (in effect since September 1997) raises the maximum severance pay rate to the equivalent of 10 months of the employee's salary. However, compliance is apparently low. Although official statistics on severance pay have not been available since the cri- sis, casual evidence from small enterprises indicates that none of the re- trenched had been given any severance pay. Some reported that they were not even given their last paycheck. Most people do not consider the long court procedure entailed in recovering this money worthwhile, given the amounts involved. But the Ministry of Labor and Social Welfare closely monitors enterprises that look likely to shut down and warns management about its legal obligations to employees. The government announced tax reductions in March 1999. The most sig- nificant reduction is in the value-added tax (VAT), which was reduced from GROWTH, CRISIS, AND COMPETITIVENESS IN THAILAND 277 10 to 7 percent for two years. Income tax was waived on the first TB 50,000 of income, as was a corporate tax of 1.5 percent of sales levied on small-scale enterprises earning less than TB 1.2 million per year. And the excise tax on fuel oils was slashed from 17.5 percent to 5 percent. The government esti- mates that these tax cuts will cost TB 57 billion a year in lost revenue. The government also reduced electricity tariffs as of March 1999 by an average of TB 0.26 per unit. This reduction will cost TB 19 billion per year in lost revenue. At the same time the government reduced the price of cooking gas by 10.5 percent, a move that will save consumers an esti- mated TB 229 million per month. To help rice farmers affected by the crisis and by falling world paddy prices, in February 1999 the government initiated a price support pro- gram for rice at a level above the prevailing market price. TB 3,500 mil- lion will be used to buy 250,000 tons of rice at fixed prices, but the rice will be exported in order not to affect domestic prices. Thailand has two public health insurance systems: a health-card scheme run by the Ministry of Public Health and a social security scheme run by the Social Security Office of the Department of Labor and Social Welfare. The health-card scheme is open to the general public. Thais pay a fee of TB 500 to obtain a health card for one year, with the government contributing another TB 500. The health card entitles the holder to health services at a designated health center and to emergency treatment at any public health center. In 1999 the number of subscribers was raised from 1.2 million to 2.4 million, and the government's contribution rose to TB 1,000 under the Social Safety Net Program. The social security scheme is a form of health insurance, but it also includes compensation for hospitalized workers and a death benefit. Par- ticipation is mandatory for all workers in enterprises with a staff of 10 or more. The government, employers, and employees all contribute 1 percent of the employee's salary as the insurance premium. Workers receive free medical treatment at a designated hospital and, in case of prolonged treat- ment that causes them to miss work, 50 percent of their salary. Because of the crisis the government extended coverage to laid-off subscribers, who can receive hospitalization benefits for a year without having to pay monthly contributions. The retrenched can continue subscribing to the scheme after a year but must make monthly payments. To improve educational opportunities for the underprivileged, the gov- ernment gives education loans to households with incomes of less than TB 150,000 per year. The loans, which can be used up to the bachelor's degree level, enable borrowers to meet school fees and some living expenses. Re- payment begins two years after students complete their education and can 278 EAST ASIAN LABOR MARKETS be extended for up to 15 years at an interest rate of 1 percent. As of the end of 1998, TB 13 billion had been loaned to about 638,000 students. Additionally, the Miyazawa Initiative has provided TB 869 million to expand the school lunch and school uniform programs for poor students. And a TB 35 million loan from the ADB under the Social Structure Pro- gram provides milk and lunch to 15,000 disadvantaged children in urban areas and in child development centers. A loan of TB 480 million from the Miyazawa Initiative provides subsidies for 400,000 elderly poor and for 50,000 poor families needing to supplement their income. Skill Development and Vocational Training To upgrade the skills of the unemployed, in 1999 the Ministry of Labor and Social Welfare launched a skill development program targeting 123,000 people over three years. The program provides training in a range of skills, including truck driving, equipment maintenance, appliance re- pair, and computer programming. Approximately 10,000 unemployed workers are scheduled to undergo training in programming and system design and analysis. The ministry also provides training in managing home-based production networks for leaders of subcontracting groups in rural areas. Approximately 7,500 project leaders are expected to receive this training. The Ministry of Industry provides occupational and vocational train- ing for rural workers, and the Bangkok Metropolitan Administration ser- vices those residing in the capital. While rural workers receive training pri- marily in handicrafts, the Bangkok Metropolitan Administration aims at self-employment in the service industry. Under the Miyazawa Initiative, Thammasat University will train 10,000 unemployed university graduates to become computer programmers. Capacity-building Schemes In addition to skill development, the government offers programs to im- prove productivity and competitiveness in manufacturing. The Ministry of Industry, for instance, will train 4,500 people to act as advisors to some 2,000 export businesses on adopting the ISO 9000 standard.6 In collabo- ration with the National Productivity Institute, the Thai Japanese Tech- nological Promotion Association, and the Small Industrial Financial Cor- poration, the ministry has also set up a TB 72 million project to evaluate enterprises. The program aims to improve the competitiveness of SMEs by GROWTH, CRISIS, AND COMPETITIVENESS IN THAILAND 279 enhancing their management capabilities. Evaluators trained through the program will use an "enterprise diagnosis system" with an evaluating index to assess enterprises. By the end of 2000, the program will have certified an estimated 100 evaluators and assessed 2,000 SMEs. SMEs can also receive assistance in creating jobs and improving pro- ductivity. The Department of Industrial Promotion in the Ministry of Industry, in collaboration with the Krung Thai Bank, has set up a loan facility for SMEs. The facility has funds totaling TB 10 billion and will lend to a projected 2,700 SMEs in 1999-2000. The program, which tar- gets primarily rural enterprises, is expected to generate direct and indirect employment for about 120,000 persons. The Krung Thai Bank has also allocated TB 3 billion in loans for approximately 200 SMEs in 1999, and the Government Savings Bank has set aside TB 1 billion to support SMEs. In addition the Bank for Agriculture and Cooperatives is propos- ing to amend its operations to include lending to farmers in the off season to support income-generating activities. Once the change is approved, the Bank is expected to provide TB 2,000 million to support 10,000 microenterprises. The Ministry of Industry is also providing venture capital for SMEs (approximately TB 3-5 billion) from the government budget, the Over- seas Economic Cooperation Fund, the ADB, and private investors. These funds will be used to buy shares of less than 50 percent in targeted indus- tries, including electrical appliances and vehicle parts and industries using local technologies and inventions. In addition, credit guarantee facilities for SMEs are being planned to assist enterprises that have good prospects but lack collateral. Assessing Program Performance The programs the government launched in response to the crisis have en- joyed only limited success. The job creation and maintenance programs suffered from delays in implementation and are only stopgap measures. Income-support measures such as the severance pay scheme bypass a large number of workers released by failing firms that lack the funds to make the payments. Securing the severance pay system is especially important in light of the fact that Thailand has no unemployment insurance system. JOB CREATION AND MAINTENANCE PROGRAMS. Each of the employment generation projects creates a large number of jobs. The Ministry of the Interior alone expects to provide some 790,000 jobs using the regular gov- 280 EAST ASIAN LABOR MARKETS ernment budget and approximately 2 million person-months of work under the loan from the Miyazawa Initiative and the World Bank (through the Social Investment Program). Under the Social Investment Program, the Ministry of Agriculture and Cooperatives is generating em- ployment equivalent to 9 million person-months. Despite these encour- aging numbers, the programs have a number of problems. First, despite the fact that the programs may have tightened up the labor market to some extent, numerous projects have been delayed. For instance only 58 village feeder road projects out of 883 had been carried out under the Social Investment Program as of May 1999, although all the funds were scheduled for disbursal that year. The weir projects under the same program have had similar problems. By May 1999 only 662 of the planned 1,635 weirs had been dredged and only 11 were under construction. Second, many of the projects-especially under the Miyazawa Initiative- require a large number of workers with secondary and college education. Ministry of Interior projects alone will generate employment for 24,338 educated workers. At the same time other agencies are also either hiring or training these workers. The National Election Committee plans to recruit 20,000 graduates to create greater awareness of the new constitution, and Thammasat University expects to train 10,000 as programmers. These projects are unlikely to be easy to implement in view of the small number of qualified workers, many of whom may not be willing to relocate. Third, the jobs that are being created are in general temporary. The pay for educated workers is below the market rate, creating a disincentive for workers to stay and jeopardizing the projects. Because the work is tem- porary, it affords no job security-another disincentive to stick with a project through completion. The temporary nature of the jobs also re- duces their multiplier effect. While the measures are meant to meet more than one objective (for instance, stimulating economic growth while pro- viding assistance to the unemployed), the general outcome appears to be various forms of work sharing. The emphasis thus far has been on direct job creation rather than self- employment. The Social Investment Fund, which provides assistance to communities, the Ministry of the Interior (under the Miyazawa Initia- tive), and the Ministry of Labor and Social Welfare have plans to encour- age self-employment by providing funds for start-up investment. But with the credit crunch and high interest rates, unemployed workers need more assistance in training and equipping themselves to become self-employed. INCOME-SUPPORT ACTtVITIES. As we have seen, the government's income- support activities take a variety of forms. Its activities include direct trans- GROWTH, CRISIS, AND COMPETITIVENESS IN THAILAND 281 fers, subsidies for school lunches and uniforms, price supports for rice and utilities, health insurance (through the health-card scheme and social security), tax reductions on income and excise taxes and VAT, and educa- tion and business loans. The school lunch and uniform projects were in place before the crisis. The crisis has made it difficult to maintain these projects, especially since the budget deficit must be cut to meet the conditions of the IMF loan. Support from international donors allows the government to continue the projects and to step up the projects to accommodate more children. Price support programs (such as the program for rice) have been invoked from time to time to lessen the consumption burden of the poor. To make these programs effective, however, the price support level must be above the market price, and the funding for the programs must be substantial. Other- wise the programs cannot be sustained and should be avoided. Since the crisis the government has extended health insurance coverage to more people through both the health-card scheme and the social secu- rity program. While more people now have coverage, the composition of demand for services has changed. As incomes have fallen, public health centers rather than private health clinics and hospitals are increasingly in demand. The government has not addressed the problem of increased de- mand at public clinics, where long lines of people await care. The absence of a nationwide unemployment insurance scheme leaves many workers with no income protection when they lose their jobs. The severance pay scheme and a provident fund program are limited to em- ployees in the formal sector, leaving workers in the informal sector un- protected. As we have seen the severance pay system is often ineffective in times of crisis, as failing firms (both large and small) often lack the money to make the mandated payments. A severance pay guarantee scheme may be the only method of ensuring that workers have some form of social pro- tection when they lose their jobs. One alternative scheme that has been under consideration is a public compensation fund for unpaid severance payments financed with fines on firms that violate the Labor Protection Act. But this scheme may be most feasible during times of relative nor- malcy. During a crisis of any magnitude, the funds generated by the fines would not be enough to pay the large number of unemployed workers. For this reason businesses may also need to contribute to such a fund. An unemployment insurance scheme could be instituted to ensure that work- ers are not left without some form of support. The benefits of tax reductions should not be overstressed. Tax reduc- tions (especially in the VAT) are often effective in increasing purchasing power. But the VAT does not apply to essential goods such as foods, limit- 282 EAST ASIAN LABOR MARKETS ing the impact of such a reduction. Moreover, an economy realizes the full impact of tax reductions only if the government maintains its expenditure levels and accepts a larger deficit. If it maintains the deficit at the existing level, the tax reduction represents expenditure switching (from the govern- ment to consumers). The effect of an increase in government expenditure on the economy is greater than the effect produced by a tax reduction. Labor Market Services Increased unemployment and underemployment have created a significant demand for labor market services. Both public and private market services are available in Thailand. The government provides services that include job matching, career counseling, and testing, but its effectiveness is limited by a weak labor market information system. Private employment agencies assist workers in finding employment domestically and abroad. Public Employment Services The Department of Employment provides services for both employers and job seekers through its 9 Bangkok and 76 provincial branches. Job seekers and employers needing workers are listed in a database for one year and matched on request. Periodic job fairs (over 800 in 1998 alone) bring prospective employers face to face with job seekers. The department pro- vides job counseling and career guidance at job fairs, at educational insti- tutions for graduates, and in villages, where people may seek subcontract- ing work from home. Some 894,000 members of the general public and 814,000 villagers received job counseling and career guidance in 1998. Of the 1.5 million Thai workers who were unemployed in 1998, some 700,000 applied for 550,000 jobs through public placement services. However, only half the vacancies were filled. This low success rate under- scores the importance of private placement services. One method of boosting employment has been to reduce the number of foreign migrant workers in the country by increasing security at bor- ders. While attempts to push out illegal migrant workers have been rela- tively successful, firms used to the cheaper wage rates of migrant labor have experienced problems as a result. A second method of promoting employment has been to send workers abroad. This approach also creates problems, as the contagion effect of the crisis has reduced regional demand for labor. Overseas employment is ex- pected to peak at just under 0.5 million in 1999, with new placements ta- GROWTH, CRISIS, AND COMPETITIVENESS IN THAILAND 283 pering off sharply from 200,000 per year to a quarter of that number. The Department of Employment provides some overseas placement services. Through its missions abroad the department identifies job vacancies, an- nounces them, and processes applications for prospective employers. The department also provides skill testing to screen applicants and orientation sessions for newly hired workers, all free of charge. (In 1998 a quarter of a million workers were tested and over 100,000 workers going abroad re- ceived orientation.) But overall the department's placement rate for over- seas jobs is low (table 6.18). Private Employment Services Private placement agencies, which operate temporary employment and permanent placement services, are under the supervision of the Ministry of Labor and Social Welfare through the Job Search and Worker Protec- tion Act of 1985. The act allows the ministry to monitor the agencies to protect workers and to set fee levels. Statistics on private agencies are not available. The Department of Em- ployment's data on number of placements, applicants, and vacancies offer the best estimate of the magnitude of private employment services. These statistics provide only a residual, but the implication is significant-public employment services accounted for only half of job placements in 1998. Private agencies also accounted for almost 46 percent of foreign placements.7 The Department of Employment licenses private agencies as well as supervises them. The operating conditions include a deposit of TB 5 mil- lion with the department and registered minimum capital of TB 1 mil- TABLE 6.18 Overseas Job Placement, 1998 (number of workers) Self- Private arranged DOE Employers agencies Total Middle East 5,914 2 623 11,244 17,783 Africa 209 - 136 - 345 Asia 81,428 1,181 9,298 76,091 167,998 Others 4,661 87 537 324 5,609 Total 92,212 1,270 10,594 87,659 191,735 - not available. Source: Department of Employment. 284 EAST ASIAN LABOR MARKETS lion. Private agencies must submit work contracts to the department for approval and repatriate and reimburse job seekers whose contracts are withdrawn. In addition placement firms and overseas employers contrib- ute to the Overseas Workers Aid Fund, which provides welfare and orien- tation training and skill testing for job seekers. Despite these safeguards the agencies are not trouble free. Some unscrupulous agencies charge ex- orbitant fees, and in some cases workers have been stranded abroad when promised jobs fell through. Labor Market Support Services The crisis has resulted in numerous business failures and the subsequent termination of employees. The Department of Labor Protection and Wel- fare provides assistance to the retrenched, especially in getting severance pay. The department inspects enterprises that are likely to close down and coordinates with other government agencies in determining the financial status of these businesses. When businesses close and the employees are terminated, the department examines documents such as wage and pay- roll accounts and employment agreements. The department also discusses with managers their obligations to their employees. When severance pay is not forthcoming, the department helps employees prepare legal docu- ments (such as petitions) and joins in requests to sequester assets. It also coordinates with the Revenue Department to stop tax refunds to enter- prises that shut down and leave their workers with no severance pay. The Effectiveness of Labor Market Services The Public Employment Service accounts for only a small share of job placements in Thailand. The service has trouble matching workers to jobs in part because it has limited contact with job seekers. In addition, Thai- land has a weak labor market information system. This system is relatively new, and nationwide data are available, but the processing capacity is in- adequate. The timeliness of the data is also an issue, as job opportunities change continually. The department has relied mainly on employers to provide information on vacancies, although it also initiates some contact with firms. It needs a more proactive strategy (based on ongoing contact with employers) in order to gather all the necessary information. A more dynamic information system will also provide the data needed to improve job counseling and career guidance services. GROWTH, CRISIS, AND COMPETITIVENESS IN THAILAND 285 Vocational Education and Training Both private and public vocational education and training are available in Thailand. Vocational education involves some years of formal schooling, while vocational training is generally relatively short term and skill spe- cific. The institutions responsible for vocational education include the De- partment of Vocational Education of the Ministry of Education and vari- ous universities offering professional curricula as well as liberal arts. The major public institutions that provide vocational training are the Depart- ment of Skill Development (of the Ministry of Labor and Social Welfare) and the Bangkok Metropolitan Training Center (of the Bangkok Metro- politan Administration). Vocational Education The Department of Vocational Education offers a number of options ranging from short training courses to diplomas in five subject areas (in- dustrial mechanics, handicrafts, home economics, commerce, and agricul- ture). The department has 413 schools and training centers around the country. The curricula include both formal education, which requires 2-3 years of study, and informal education, which requires anywhere from 45 to 225 hours. Enrollment in vocational education courses in 1997 was ap- proximately 0.5 million, of which 54 percent was in industrial mechanics, 32 percent in commerce, 6 percent in agriculture, 5 percent in home eco- nomics, and 2 percent in handicrafts. In July 1998 the department started a new project to provide vocational and technical education to retrain unemployed workers. In the first three months the 27,000 workers received training at 54 centers. Some 1,000 assistant trainers were also recruited from laid-off workers. The govern- ment urges workers who have completed the training to return to their hometowns-a strategy aimed at generating local income and reversing rural-urban migration. The project was deemed successful, with demand exceeding supply, and was repeated in 1999 with an additional budget of TB 35 million and 700 assistant trainers. A project is also under way to upgrade the department's 413 training institutions. Under this project 2,200 new graduates or retrenched work- ers are setting up a database of enterprises and their occupational require- ments. The government will use the database to estimate demand for vo- cational training. These workers will also execute an in-school job creation project designed to give students on-the-job training by subcontracting 286 EAST ASIAN LABOR MARKETS work from the private sector. The program will also supplement the in- come of both the school and the students. Skill Development The main role of the Department of Skill Development is to provide skill training to both new labor market entrants and employed workers from both the public and private sectors wanting to upgrade their skills. The service is also offered to the economically vulnerable, particularly in the agricultural sector, and the urban poor. The department has established skill standards and testing and provides training courses for trainers. The department has sought greater private sector participation in skill development through a number of measures. In 1994 it set up a Skill Development Fund under the Vocational Training Promotion Act with initial capital of TB 400 million. Businesses can apply for loans from the fund at below-market interest rates to train their employees. Employers providing training for employees in the workplace can also deduct train- ing costs from their taxes. The department supports the private sector in establishing Skill Standard Testing Centers. Training organizations regis- tered under the Vocational Training Promotion Act are exempted from some labor regulations and statutes governing private schools. The de- partment is also revising the Vocational Training Act of 1994 to further encourage private involvement in skill training. In response to the economic crisis, the department has initiated a num- ber of projects. The Thais Help Thais project provides vocational skills to retrenched and unemployed workers. Low-cost funding is available to un- employed workers to upgrade skills and prepare for self-employment. The Urban/Rural Employment project provides training in rural areas, partic- ularly in subcontracting work such as rubber tapping and other forms of self-employment that can be carried out in villages. Finally, the Employ- ment in the Industrial Sector project provides skill development to the general public and training for trainers. It also establishes test standards and encourages enterprises to train their own workers. Metropolitan Vocational Training The Bangkok Metropolitan Vocational Training Department of the Bangkok Metropolitan Administration offers short vocational and occu- pational training courses to the general public, but particularly to low- GROWTH, CRISIS, AND COMPETITIVENESS IN THAILAND 287 income groups and the vulnerable. Participants are trained in skills they can use in wage labor and in self-employment. The department has 54 training centers in the Bangkok area. Courses offered include mechanics, electronics, small appliance repair (radio, tele- vision, computer, and air conditioner), stone cutting, dressmaking and tailoring, barbering, food preparation, typing, handicrafts, and music. The training takes from 48 to 600 hours, depending on the subject and the level, and carries a small tuition fee. Private Vocational Training Private businesses have made only limited efforts to provide training for employees beyond what is needed to perform job functions. Several disincentives-including training costs and the possibility of losing newly skilled workers-keep firms from expanding training possibilities. Crisis- induced declines in profits have also had a negative effect on workplace training. Only a small number of firms have taken advantage of the Skill Development Fund and the tax incentives. The Vocational Training Promotion Act that created these incentives also allows private companies to set up training organizations as business enterprises. In 1998 Thailand had 268 such companies with 2,329 courses and 190,358 trainees, up from 2 companies with 2 courses and 132 trainees in 1996. In the first half of 1999, there were 125 training com- panies with 1,584 courses and over 150,000 trainees. Six skill-testing cen- ters for overseas workers had also been established and had tested 14,543 workers, and 8 new licenses had been granted. Meeting the Demandfor Skills in the Labor Market The Thai labor force comprises mainly unskilled workers with elementary- level schooling. In 1999 an estimated 73 percent of all workers had six years or less of education, and most of these had completed only four years of schooling (table 6.19). About 18 percent of the labor force has a secondary-level education and about 9 percent a university degree. The proportion of workers with a university degree is projected to increase to 14 percent of the total labor force by the year 2006, and the proportion with a secondary-level education is expected to rise to 24 percent (table 6.20). The percentage of the work force with only a primary-school edu- cation will shrink to 62 percent. 288 EAST ASIAN LABOR MARKETS TABLE 6.19 Educational Levels of Thai Labor Force (in millions) Lower Upper College Primary secondary secondary education Year education education education and above Total 1997 25,258.8 3,359.3 2,074.9 2,530.2 33,223.3 1998 24,960.0 3,543.5 2,206.1 2,772.0 33,481.6 1999 24,624.7 3,747.7 2,343.2 3,014.3 33,730.0 2000 24,264.7 3,960.2 2,485.5 3,256.5 33,966.9 2001 23,854.7 4,180.5 2,636.1 3,503.7 34,174.9 2002 23,428.5 4,393.2 2,794.9 3,757.2 34,373.8 2003 23,007.0 4,589.2 2,959.8 4,046.2 34,572.2 2004 22,592.3 4,776.9 3,133.2 4,288.9 34,971.3 2005 22,191.6 4,954.4 3,310.3 4,569.6 35,025.9 2006 21,791.6 5,119.7 3,490.7 4,860.4 35,253.7 Source: Master Plan for Manpower Development in Thai Manufacturing and Service Sector: 1998-2206, Thailand Development Research Institute, September 1998. The relatively low educational levels in Thailand mean that the coun- try has a surplus of unskilled labor. The country has more workers than it has jobs, despite a labor shortage at the upper end of the educational scale. A surplus in the science and technology sector is projected to disappear by 2006, leaving a labor shortage in this area (table 6.21). The labor surplus is also shrinking rapidly in skill areas addressed by vocational education and in manufacturing engineering, raising the possibility of labor short- ages in these areas by 2006. There may also be a shortage in managerial and entrepreneurial occupations. Some Thoughts for the Future The Thai economy is facing a policy dilemma. Prior to the crisis the econ- omy had strong export-based growth, with medium- and high-technology products approximating a third of the total. But sluggish labor and capi- tal productivities weakened the country's competitiveness in the 1990s. The government must now decide how to shift from crisis-oriented policies to policies designed to restore competitiveness and economic growth. The Thai economy needs to raise the share of medium- and high- technology products in exports from one-third to the two-thirds share that is the average in East Asian economies. The increase in labor and capital TABLE 6.20 Estimated Labor Force Requirements According to Educational Level (in millions) Lower Upper College College Post- Primary secondary secondary Vocational liberal professional graduate Year education education education education education education education Total 1997 24,026.7 3,371.1 1,061.3 856.2 1,736.2 658.0 120.0 31,829.5 1998 22,975.5 3,385.4 1,111.8 857.6 1,812.5 700.0 126.1 30,969.0 1999 22,377.0 3,481.6 1,190.1 880.3 1,915.6 762.5 134.1 30,741.3 2000 21,938.9 3,620.4 1,287.5 916.3 2,039.1 840.5 143.6 30,786.2 2001 21,507.4 3,769.1 1,394.8 955.7 2,172.2 928.1 153.8 30,881.1 2002 21,101.6 3,927.8 1,512.3 997.9 2,315.5 1,026.0 164.6 31,045.7 2003 20,682.1 4,087.1 1,636.3 1,040.0 2,465.1 1,132.7 175.8 31,218.9 2004 20,227.7 4,240.8 1,766.0 1,081.5 2,622.7 1,248.4 187.7 31,374.7 2005 19,763.4 4,393.1 1,901.3 1,122.3 2,783.7 1,372.9 199.6 31,536.4 2006 19,291.3 4,545.3 2,044.2 1,163.8 2,951.9 1,508.5 212.0 31,716.9 Source: Master Plan for Manpower Development in Thai Manufacturing and Service Sector: 1998 - 2206, Thailand Development Research Insti- tute, September 1998. TABLE 6.21 Shortage of Science and Technology Manpower, by Level of Education 1998 1999 2000 2001 2002 2003 2004 2005 2006 Lower-level vocational education -36,469 -19,711 -14,886 -11,010 -9,609 -6,667 -5,967 -4,749 -1,950 Upper-level vocational education -51,835 -34,248 -25,513 -21,896 -19,104 -16,429 -15,596 -13,872 -10,761 University -39,247 -21,800 -15,655 -15,051 -17,117 -16,594 -17,193 -16,585 -14,415 Engineering -24,750 -11,176 -6,813 -5,414 4,234 -3,420 -3,489 -2,969 -1,632 Manufactoring -7,595 -4,518 -2,517 -1,879 -1,379 -1,043 -1,200 -987 -410 Electrical and electronics -8,743 4,903 -3,553 -3,119 -2,756 -2,594 -2,649 -2,654 -2,287 Infrastructure -7,635 -1,656 -728 496 -226 51 243 508 820 Material science -777 -199 -15 80 128 164 116 164 244 Science -9,896 -6,799 -4,955 -5,518 -8,381 -8,078 -8,566 -8,315 -7,662 Engineering -2,147 -1,476 -1,260 -1,243 -1,404 -1,329 -1,309 -1,332 -1,177 Science -2,454 -2,349 -2,628 -2,877 -3,098 -3,766 -3,830 -3,969 -3,944 Source: Master Plan for Manpower Development in Thai Manufacturing and Service Sector: 1998-2206, Thailand Development Research Institute, September 1998. GROWTH, CRISIS, AND COMPETITIVENESS IN THAILAND 29I productivities required to make this change entail increased investment in human capital and R&D. It also requires increased efficiency in the use of capital, in the form of significant restructuring of the corporate and bank- ing sectors. Thailand, with its history of strong economic growth, can make this shift. Notes 1. Like all markets the labor market is an abstraction. But it is a particularly difficult abstraction given the difficulty of separating the labor market from the macroeconomy. In classical models of the economy, the labor market is the core of the economy, and for some prototypes, like that of Lewis (1954), it is the econ- omy. This chapter assumes that the labor market is inextricably embedded in the macroeconomy, so that the two must be viewed together. 2. Labor Protection and Welfare Department, Government of Thailand. Per- sonal communication. 3. National Statistical Office, Labor Force Survey, Round 1, 1999. 4. Lall (1999) ranks 120 countries according to a skills index (tertiary plus technical enrollment multiplied by a factor of 10). On this scale, Thailand ranks 55 out of 120. But for the region it ranks above only three other countries- China, Indonesia, and Malaysia. 5. The Miyazawa Initiative refers to Japanese bilateral aid offered to East Asian economies affected by the crisis. 6. ISO 9000 is a set of quality management guidelines used by companies throughout the world to establish systems of continuous improvement (e.g., Total Quality Management). 7. The largest category (self-arranged placements) was made up of workers re- turning to jobs after an absence. Most of the workers private placement agencies send abroad and most returning workers are unskilled laborers and low-level tech- nicians. Most of those recruited by employers, however, are white-collar workers. References The word processed describes informally reproduced works that may not be com- monly available through libraries. Asian Institute of Technology and ILO (International Labour Organization). 1999. "Gender Policy and the Economic Crisis." Working paper, processed. Bangkok, Thailand. Colaco, E X. 1998. "Thailand and International Competitiveness: A Framework for Increasing Productivity." Working paper, processed. Washington, D.C.: World Bank. 292 EAST ASIAN LABOR MARKETS Dollar D., and M. Hallward-Driemeirer. 1998. "Competitiveness and the Indus- trial Sector." Working paper, processed. Washington, D.C.: World Bank. Dwor-Frecaut, D. 1998. "Thailand's Balance of Payments and Financial Crisis: Export Competitiveness, Investment Efficiency, and Financial Fragility." Work- ing paper, processed. Washington, D.C.: World Bank. ILO (International Labour Organization). 1999a. Key Indicators of Labor Markets. Geneva. . 1999b. "Country Employment Policy Review for Thailand." Geneva. Processed. IMF (International Monetary Fund). 1998. World Economic Outlook. Washington, D.C. . 1999. World Economic Outlook. Washington, D.C. Kakwani, N. 1998. Impact of the Economic Crisis on Employment, Unemployment, and Real Income. National Economic and Social Development Board and Asian Development Bank. Krugman, Paul. (1998) "Saving Asia: It's Time to Get Radical Part 1. Asia: What Went Wrong" Fortune Investor September (available online at http://www. fortune.com/fortune/investor/ 1998/980907/sol l .html. Lall, S. 1999. Raising Competitiveness in the Thai Economy. Geneva: International Labor Office. Working paper. Processed. Lee, Eddy. 1998. The Asian Financial Crisis: The Challenge for Social Policy. Geneva: International Labor Office. Lewis, W A. 1954. Economic Development with an Unlimited Supply of Labour. The Manchester School of Economic and Social Studies. Vol. 22, pp. 139-191. Mahmood, M. 1998. Establishing Some Determinants of Competitiveness in Thai Manufacturing. Bangkok: International Labour Organization, Regional Office for Asia and the Pacific NSO (National Statistical Office). 1999. Labor Force Survey, Round 1. Bangkok TDRI (Thai Development Research Institute). 1998. Processed. Warr, PG. 1998. Thailand-What Went Wrong?Washington, D.C: World Bank. World Bank. 1998. World Development Indicators. Washington, D.C. . 1999. World Development Indicators. Washington, D.C. SECTION Labor Market I I Issues and Policies in the Region 7 Active Labor Market Policies: Issues for East Asia GORDON BETCHERMAN, AMIT DAR, AMY LuINSTRA, AND MAKOTO OGAWA AFTER 1960 ACTIVE labor market policies emerged as an important employment policy tool, particularly in industrial countries. This policy envelope includes a wide range of measures intended to increase the quality of the labor supply through activities such as retraining, raise the de- mand for labor with direct job creation and related programs, and improve job matching for both workers and employers by expanding job search as- sistance. The objective of these measures is primarily economic. They are in- tended to increase the probability that the unemployed will find jobs and that the underemployed will increase their productivity and earnings.' More recently the case for active labor market policies has emphasized the poten- tial social benefits (inclusion and participation) of productive employment. The debate that surrounds these labor market policies is often formu- lated in terms of the relative value of active versus passive measures in com- bating unemployment. So-called passive programs, such as unemployment insurance or social transfers, mitigate the financial needs of the unem- ployed but are not designed to improve employability. Active programs are meant to directly increase unemployed workers' access to the workplace. As the disincentives and dependencies inherent in passive programs re- ceive more emphasis, active policies become increasingly attractive. The now-familiar safety net and trampoline metaphors for the passive and active approaches illustrate the growing interest in active measures. Not only are trampolines more politically acceptable, but they also have a theoretical ra- 295 296 EAST ASIAN LABOR MARKETS tionale in models of the labor market that incorporate asymmetric infor- mation and market failures associated with investments in human capital. Active programs (specifically retraining) will become increasingly important as technological change raises skill requirements and quickens the pace of obsolescence. But as the experiences of the late 1980s and 1990s show, ac- tually implementing active labor market policies poses many challenges. The immediate question is whether these programs do any good. Eval- uations of their impact are mixed, with many programs assessed as having little or no effect on the employability or earnings of participants. Even policymakers who judge the evidence favorably or feel compelled to intro- duce active policies for political reasons must address a host of complex de- sign and implementation issues in order to maximize the probability of success. These issues include the complementarity of public and private roles, optimal resource allocation, targeting, delivery, monitoring and eval- uation, and feedback. Further, effective and efficient active labor market programs require considerable capacity. For this reason most of what is known about them is based on experience in industrial countries. Clearly, however, the role and nature of active programming varies at different stages of development. And as the experiences of industrial countries show, culture and institutions matter a great deal as well. This chapter reviews international experiences with active labor market policies and discusses their applicability to five East Asian countries: In- donesia, Korea, Malaysia, the Philippines, and Thailand. While several of these countries have some active policies, no country makes active mea- sures an important policy instrument. But East Asian policymakers need to consider these policies carefully as these economies respond to the cri- sis and the long-term requirements of development. Active Labor Market Programs: An Overview Active labor market programs, including job creation (public works, sup- port for self-employment, and wage subsidies), training, and employment services can affect both the supply of and the demand for labor, as well as the way the labor market matches the two.2 The overall objective of these interventions is to increase employment and incomes. They can serve eq- uity objectives as well, most obviously when they target vulnerable groups. Active policies increase employment and incomes in various ways. Di- rect employment creation-for instance, temporary jobs through public works-has a stabilizing effect. Training and wage subsidies move supply and demand curves outward. Training, mobility incentives, and other em- ployment services reduce structural imbalances by improving the match be- ACTIVE LABOR MARKET POLICIES: ISSUES FOR EAST ASIA 297 tween workers and jobs. By decreasing the number of vacancies (and thus reducing both upward wage pressure and labor bottlenecks), these policies also increase employment. The employment and income effects are further transmitted through increases in skill levels and productivity. Even though the net employment effects may not be significant, some active labor mar- ket programs increase the attachment of the long-term unemployed to the labor force and decrease their dependence on unemployment benefits. By assisting the most disadvantaged workers, these policies break down the po- tentially negative consequences associated with "outsider" phenomena. Active labor market programs can positively affect employment and in- comes in many ways, then. But substitution, deadweight, and displace- ment effects can dissipate or even eliminate their potential benefits.3 To minimize these losses policymakers must address various issues in design- ing and implementing active programs, including the overall strategy as well as specific details of program design. The overall strategy for active labor market programs involves identifying clear objectives, determining the composition of programs, targeting priorities, and establishing an ef- fective balance with passive policies. As we have noted active programs can serve several objectives, and policymakers need to determine which of these are the most important. An active strategy can be designed to moderate cyclical downturns, reduce structural imbalances, or otherwise improve the functioning of the labor market. It may also aim to increase productivity, support disadvantaged or at-risk workers, assist at-risk employers or indus- tries, or achieve more than one of these goals. These objectives have differ- ent client populations and call for a variety of policies (table 7.1). The relationship between active and passive policies is an important strategic issue. As a general rule countries with active programs also have unemployment insurance or some other form of passive support. But for the most part coordination between the two types of policies is weak. A few countries, including Austria, Germany, Japan, Norway, and Spain, have integrated systems. Others, such as Canada, are moving toward integra- tion by coordinating active program options with unemployment insur- ance. This trend is likely to continue, for two reasons. First, attempting to reintegrate unemployed workers into the labor market is more politically attractive than simply providing income support or insurance. Second, in- tegrated systems may have positive economic outcomes. After a period with little labor market intervention in the United Kingdom, the public employment service began interviewing unemployment insurance clai- mants. The agency found that even minimal contact reduced the number of claims (OECD 1994). Furthermore, coordinating benefits distribution with job search assistance can save on administrative costs. Z98 EAST ASIAN LABOR MARKETS TABLE 7.1 Tailoring Programs to Objectives Objective Program orientation Targeting orientation Moderate cyclical * Direct job creation (for * Vulnerable groups (those downturns instance, public works) with the least resiliency) * Wage subsidies * Hard-hit regions and * Training (subsidies or industries grants to workers or employers) * Self-employment support Reduce structural * Employment services * Proximate regions, imbalances (information and search industries, or and mobility assistance, occupations among others) * Training * Wage subsidies Improve general * Employment services * All labor market * Training (such as functioning apprenticeships) Enhance skills * Training and retraining * At risk or disadvantaged and productivity (including in-service and worker categories apprenticeships) (especially for retraining) Support * Employment services * At-risk or disadvantaged disadvantaged (counseling and job worker categories or at-risk workers search assistance) * Training (for instance, grants and subsidies) * Wage subsidies Types of Active Labor Market Programs Policymakers involved in planning and implementing active labor mar- ket programs confront a variety of issues specific to the type of program (table 7.2). We discuss some of the key issues involved in the main types of labor market programs: employment services and job search assistance, training, job creation, microenterprise development and support for self- employment, and public works. TABLE 7.2 Active and Passive Labor Market Programs: Some Key Features Program Objective and activities Possible pros Possible cons Key issues Job search assistance To match jobs and job 1. Helps reduce 1. Crowds out private 1. Determining the roles and employment seekers through services the length of services. of public and private services such as initial inter- unemployment. 2. Results in deadweight job search agencies. views at employment 2. Generally is reasonably losses. 2. Providing integrated offices, in-depth coun- inexpensive. 3. Benefits only a frac- services. seling, and job clubs. 3. Prescreens participants tion of job seekers. 3. Using monitoring and who may be receiving evaluation to improve assistance from other effectiveness. programs. Training and retraining To help new and rede- 1. Increases productivity 1. When poorly targeted 1. Defining the roles of ployed workers (the and enhances workers' results in deadweight the government and long-term unemployed skills. losses. private sector. or those laid off during 2. When well targeted 2. Produces poor out- 2. Improving linkages restructuring) accum- may benefit groups comes when the with the labor market. ulate employability such as the disadvan- economy is flat. 3. Resolving costs. skills. taged and women. 3. Is one of the most costly programs (but is extremely popular). (Table continues on thefollowing page) TABLE 7.2 (continued) Program Objective and activities Possible pros Possible cons Key issues Wage subsidies To encourage employers 1. May lead to perma- 1. Can result in dead- 1. Determining the to maintain employees nent employment by weight losses. duration of the and hire disadvantaged helping individuals 2. Can displace workers subsidies. groups such as the develop work-related and cause a substitu- 2. Determining the ideal long-term unemployed skills. tion effect if employers level of the subsidies. and youth by paying 2. Helps individuals replace unsubsidized part of workers' salary maintain contact with employees with subsi- for a period of time. the labor market. dized workers. 3. May be viewed by employers as a source of cheap labor, with workers laid off once the subsidies end. Microenterprise To create and promote 1. Can assist in creating 1. May have high 1. Identifying the kind of development and self- small-scale businesses entrepreneurial spirit. deadweight losses. support to provide employment support and self-employment 2. May displace other (financial, technical, or activities with technical small businesses that other). assistance, credit, and do not get this 2. Improving targeting to other support. assistance. minimize deadweight 3. Has low take-up losses. rate among the 3. Measuring the impact unemployed. of programs, including the proportion of un- employed that use them, the proportion of businesses that sur- vive, and the number of additional jobs created. Public works and To address poverty and 1. May assist disadvan- 1. May crowd out pri- 1. Setting the wage level public service nutrition objectives taged groups to regain vate sector jobs, espe- and determining the employment and create temporary labor market contact. cially if targeting is proportion of wages in employment by 2. Leads to production ineffective. total program costs. generating income of public goods 2. May be perceived neg- 2. Determining whether through low-wage and develops atively and thus not these programs gener- temporary work. infrastructure. increase participants' ate employment and 3. Can be self-targeting if employability. higher wages and are wages are set cost-effective. effectively. 3. Choosing whether to hire private or public contractors to imple- ment the projects. 302 EAST ASIAN LABOR MARKETS Employment Services Employment services serve brokerage functions, matching jobs with job seekers. This assistance includes many different types of activities-for ex- ample, initial interviews at employment offices, in-depth counseling dur- ing the unemployment spell, and job clubs. In Hungary and Poland in the mid-1990s, workers had access to job referrals, job counseling, skills as- sessment, job search training, resume preparation, and job clubs (O'Leary 1998a, 1998b). In New Zealand job seekers attended job screening inter- views, workshops, and follow-up interviews and received personal case management (New Zealand Department of Labor 1995). Australia of- fered training in resume writing and interviewing techniques. Such services are relatively inexpensive and can help shorten spells of unemployment by providing job seekers with up-to-date information on jobs. On the negative side these interventions usually have deadweight losses, since individuals who use them are generally more qualified than most job seekers and could probably find jobs on their own.4 Policymak- ers designing such interventions must address three key issues: public- private sector complementarity, the integration of active and passive ser- vices, and monitoring and evaluation. PUBLIC-PRIVATE SECTOR COMPLEMENTARITY. Increasingly, public and private services coexist in many countries. Public employment services are justified on the grounds that they benefit the disadvantaged, including the poor and the long-term unemployed (Fretwell and Goldberg 1994). Private fee-charging agencies typically provide labor exchange services to more fa- vored segments of the labor force, such as the employed, skilled, and white- collar workers. Some countries ban or restrict private agencies, so that pub- lic employment services operate under near-monopoly conditions. While governments need to provide services to certain segments of the popula- tion, public employment offices should not be viewed as substitutes for pri- vate agencies. Private agencies can enhance the operation of the labor mar- ket, especially when they are appropriately regulated to ensure quality. INTEGRATED SERVICE PROVISION. Optimally employment services are integrated with other active as well as passive programs. This integration is beneficial to the extent that it helps the unemployed acquire the skills and knowledge they need to fill available job vacancies. The benefits must be weighed against the administrative requirements such integration en- tails, however. ACTIVE LABOR MARKET POLICIES: ISSUES FOR EAST ASIA 303 MONITORING AND EVALUATION. As with all active labor market policies, monitoring and evaluating the impact of this intervention are essential. Countries of the Organization for Economic Cooperation and Develop- ment (OECD) use a variety of methods to enhance the effectiveness of public employment services. Some countries use administrative data such as the number of registered job vacancies to set targets that measure the ef- fectiveness of these services and then allocate budgets accordingly. For ex- ample in Sweden and Finland funds allocated to the employment service are disbursed to the regional and local levels based on the authorities' suc- cess in meeting performance targets (OECD 1997). Labor Market Training Public support for training can take the form of direct provision through public training institutes; financial support for trainees in the form of funds for training costs, subsidies for trainees, or both; and "infrastructure" services such as labor market information, licensing, monitoring, and cre- dentialing. Most countries focus on three types of training programs: i Retraining aimed at the long-term unemployed (12 months or longer); * Retraining for displaced workers, especially those laid off as a result of enterprise or industrial restructuring; and * Training programs targeting young people, often with special atten- tion to school dropouts. While these types of training programs can help increase productivity and employability, they have a number of limitations. First, they are rela- tively costly. Second, they often have little impact when the economy is not performing well and job opportunities are limited. Finally, training programs can also result in deadweight losses, as those participants who benefit the most may have more skills to begin with and could have found jobs even without training. In designing effective training programs, pol- icymakers must consider three central issues: the government's role in training provision, the role of private providers, and the link between training and the labor market itself. THE GOVERNMENT'S ROLE IN TRAINING PROVISION. Governments have a number of potentially important roles to play in the provision of labor market training. Governments may provide training directly, act in a reg- ulatory capacity, collect and disseminate information, set standards, and 304 EAST ASIAN LABOR MARKETS provide financing. Many governments are moving away from the role of direct provider to focus on market failures in information and financing, leaving delivery in the hands of private providers. This approach may be the most successful way for governments to foster the development of a relevant and cost-effective training system. THE ROLE OF PRIVATE PROVIDERS. To encourage private delivery of labor market training, governments must create a set of enabling conditions. First, they must ensure that the laws governing private provision are clear and do not discriminate against private providers. Second, they need to exercise caution in providing training directly in order to avoid crowding out private suppliers. Finally, they must let employment growth lead the demand for training. Countries that meet these requirements (including Australia and Indonesia) are rewarded with a growing, vibrant, and competitive private training sector. In these countries the public sector focuses on providing ser- vices to.the most vulnerable groups (Gill, Fluitman, and Dar 2000). LINKING TRAINING WITH THE LABOR MARKET. Creating strong linkages between the training system and the labor market requires governments to examine their own internal structures and operations. In some countries where training reform succeeded, governments developed strong institu- tional links with employers, making training institutions relatively flexi- ble. In Chile, for example, groups comprising representatives of employers, workers, and the government govern vocational training institutes. This tripartism strengthens accountability while offering the institutes the au- tonomy necessary to respond to the needs of employers. Job Creation These programs are intended to maintain existing jobs as well as to sup- port the creation of new ones. Three types of programs fall into this cate- gory: subsidies, public works, and self-employment support. Subsidies en- courage employers to hire new workers or keep employees who might otherwise be laid off. These programs may offer direct wage subsidies (for either the employer or worker) or social security payment offsets and are always targeted to a particular category of worker or employer. Public works and related programs provide temporary jobs in the public or nonprofit sector. Typically these programs cover the costs of hiring previously unem- ployed workers. Support for self-employment often takes the form of as- sistance for unemployed workers willing to start their own enterprises. ACTIVE LABOR MARKET POLICIES: ISSUES FOR EAST ASIA 305 This support can involve microfinancing for business start-ups or operat- ing costs, extended unemployment benefits while claimants start their own business, grants, and business support services. WAGE AND EMPLOYMENT SUBSIDIES. These subsidies support the long- term unemployed, those coming from severely disadvantaged areas (such as sectors with high unemployment), and special groups of workers (such as youth). The subsidies were instituted under varying economic condi- tions, though most often during slack periods. The programs often have a social objective, in that they encourage employment and thus the inclusion of disadvantaged individuals in productive activities. Detractors argue that designing subsidies that actually meet the goal of creating jobs in a cost- efficient manner is a difficult task and that in any case subsidies are also often associated with deadweight losses. Subsidies can also can have unin- tended effects-for example, firms may hire subsidized workers to replace unsubsidized staff or take on subsidized workers only to lay them off once the subsidy period ends. Good design, effective targeting, and careful mon- itoring can reduce these negative impacts. Subsidies may target individu- als to promote hiring in industries with excess demand (or by individual firms). Monitoring employers' behavior minimizes program abuse. DURATION AND LEVEL OF SUBSIDIES. Wage subsidy programs are most often payments to firms in the form of a wage offset-that is, they are an inducement to hire program participants. The level and duration of such subsidies vary significantly across programs and countries, and the opti- mum levels depend on specific conditions. For example under the U.S. Targeted Job Tax Credit, firms are paid 50 percent of the individual's wages for a period of up to two years. The job subsidy program in the United Kingdom provides up to 100 percent of wages for a period of six months. Careful monitoring and evaluation allows policymakers to arrive at an informed decision about the length and levels of subsidies. Public Works Some governments attempt to alleviate unemployment by creating jobs and hiring the unemployed directly; others contract with nonprofit orga- nizations or private businesses to provide jobs. Most programs target the displaced and the long-term unemployed (that is, the workers who are hardest to place). Some programs target youth, providing a way to intro- duce young workers to the labor market. 306 EAST ASIAN LABOR MARKETS The idea behind these programs is generally to help the unemployed reestablish contact with the labor market, minimizing the possibility that these workers will be stigmatized by long periods of unemployment and obsolete skills. Such programs can also produce public goods and develop basic infrastructure, and in many cases public works have exactly these goals and are not aimed primarily at creating jobs.5 Another advantage of these programs is that with appropriate wage levels they can be self- targeting, attracting only those most in need. In some countries public works jobs have a major drawback, however: the jobs carry a stigma and may decrease participants' employability in the long term. Policymakers designing public works programs need to consider two important issues, then: targeting and management. TARGETING. When the objective is to reduce poverty, the most success- ful form of targeting is an appropriate wage. The wage offered should be no higher than the prevailing market wage for unskilled manual labor in the area where the scheme is introduced. High wages can attract less dis- advantaged workers, in effect crowding out private sector employment. Restrictions on eligibility are best avoided. Ideally the fact that workers are willing to accept low wages should be sufficient as an eligibility require- ment. When the demand for jobs exceeds the budget, the projects can be targeted to poor areas using a credible "poverty map" and can focus on the assets that will be of most benefit to local residents. Flexibility in future budget allocations will allow the government to respond to demand from a variety of locales (Ravallion 1998). MANAGEMENT. Tendering public employment activities through pri- vate contractors and nonprofit organizations enhances the effectiveness of public job creation schemes. An evaluation of public works programs in Hungary showed that those operated by private contractors tended to be the most cost-effective (Fretwell, Benus, and O'Leary 1999). Another les- son from experience is that programs need not be managed at national lev- els. Decentralization often increases administrative efficiency and facili- tates appropriate targeting. Microenterprise Development and Self-employment Assistance Technical assistance, credit, and other support contribute to the creation and promotion of new small-scale businesses and self-employment. In coun- tries with an embryonic financial infrastructure, private banks are often un- ACTIVE LABOR MARKET POLICIES: ISSUES FOR EAST ASIA 307 able to conduct the comprehensive risk assessments necessary to offer credit to unemployed workers who want to create their own business. Public pro- grams to support small business loans can help eliminate this distortion. Microenterprise development assistance is offered on a universal basis to particular groups under varied economic conditions. These groups in- clude the newly unemployed (in the U.S. state of Massachusetts in the early 1990s), the long-term unemployed (in Denmark in the 1980s), and displaced workers (in Hungary and Poland in the 1990s). The programs vary in design. Participants may receive a lump-sum payment or periodic allowances to set up their businesses. A screening mechanism often sub- jects potential beneficiaries to a rigorous assessment that evaluates the like- lihood of success (for example, in Germany). But in other countries, such as the United States, screening is more cursory (Wilson and Adams 1994). In most cases participants also receive business advice and counseling. Few among the unemployed (usually less than 5 percent) typically take up opportunities for self-employment (Wilson and Adams 1994). One ex- planation for this low figure may be that individuals are generally risk averse and, given a choice, will opt for unemployment benefits. These pro- grams also have potential displacement effects, since small businesses that do not receive assistance are disadvantaged relative to those that do. The most important issues facing policymakers in designing these programs are determining the appropriate level of support, targeting, and screening. How MUCH SUPPORT? The experience of successful microcredit schemes such as Bangladesh's Grameen Bank shows that successful credit programs have several common characteristics.6 First, they offer small initial amounts of credit, with subsequent loans contingent on a good repayment record. Second, they charge market interest rates. They also use group lending with community guarantees rather than formal collateral, and they offer flexible repayment schedules. Finally, microenterprise credit programs start on a very small scale and grow gradually, allowing both the agency and the community to learn by doing and ensuring that supervision and training activities keep pace with the lending activities. SCREENING PARTICIPANTS. Screening is especially important in micro- enterprise development assistance and self-employment support programs. Instruments that improve screening include information sessions, detailed application forms, interviews, preentry business advisory services, training, and developing business plans. Using these instruments can sharply reduce deadweight losses and greatly enhance a project's chance of success. 308 EAST ASIAN LABOR MARKETS FIGURE 7.1 OECD Average Active Labor Market Program Expenditures Percentage of GDP 1.20 1.00 - 0.80 - 0.60 - 0.40 - 0.20 - 0.00- l l l l I l l l l l I Year Source: OECD, Employment Outlook, various years. The Cost ofActive Labor Market Programs The OECD collected statistics on its member countries' expenditures on active labor market programs from the mid- 1980s until the late 1990s and compared them with expenditures on passive programs such as unem- ployment insurance (figure 7.1). Such comparisons provide a rough mea- sure of the relative importance of the two types of labor market policies over time and across countries. Relative spending levels on active labor market programs increased in the early 1990s and continued at that level throughout the decade (figure 7.2). This increase likely reflected both the increasing preference for active programs and high unemployment rates in most OECD countries com- pared with the 1980s. An OECD (1993) analysis confirms that spending on active programs increases when unemployment rises. In 1990, for ex- ample, a 1 percent increase in the unemployment rate was associated with a 0.6 percent increase in the share of expenditures in GDP allocated for these programs. A second point emerging from the OECD expenditure data is that countries generally see active and passive programs as complements to rather than substitutes for each other. When spending for active programs is relatively high, it is also likely to be high for passive programs. In 1990 the correlation coefficient between national spending on active and pas- sive programs was .60 (OECD 1993). After diverging in the early 1990s, ACTIVE LABOR MARKET POLICIES: ISSUES FOR EAST ASIA 309 FIGURE 7.2 OECD Spending on Active Labor Market Policies 3.50 - 3.00 - 2.50 - 2.00 - __ _ - _ 1.50 - 1.00- --------- -- - -- -------…---- 0.50 - 0.00 1991 1992 1993 1994 1995 1996 1997 1998 ------- Passive Labor Market Programs -- - Active Labor Market Programs Total Labor Market Spending Source: OECD, Employment Outlook, various years. when income support jumped to accommodate workers laid off in the recession, the correlation grew strong again in 1993. Spending on active programs then increased slightly relative to spending on passive programs (figure 7.1). Nonetheless spending on passive programs remains roughly 50 percent greater than spending on active programs in the OECD re- gion. The level and composition of spending on active labor market pro- grams also vary widely across OECD countries (table 7.3). Evaluating the Impact of Active Labor Market Policies 7 In spite of relatively large public expenditures on active labor market pro- grams, rigorous evaluations of their impact are limited. Policymakers are increasingly realizing the importance of evaluation to improving program design, however. They want to know what programs accomplish, what they cost, and how they can be made cost-effective. Impact Evaluation Techniques Techniques for evaluating the effectiveness of programs can be either sci- entific or nonscientific. Scientific evaluations are of two types: experi- TABLE 7.3 Expenditures on Labor Market Programs, Selected OECD Countries (percent of GDP) Australia Denmark France Germany Italy Japan Spain Sweden United States Program (1997-98) (1998) (1997) (1998) (1996) (1997-98) (1998) (1998) (1997-98) Public employment services and administration 0.21 0.14 0.16 0.23 0.04 0.03 0.07 0.30 0.06 Labor market training 0.07 1.07 0.35 0.34 0.01 0.03 0.21 0.48 0.04 Training unemployed adults and those at risk 0.06 0.73 0.31 0.35 - 0.03 0.10 0.47 0.04 Training employed adults - 0.34 0.04 - 0.01 - 0.11 0.01 Youth measures 0.06 0.08 0.26 0.07 0.42 - 0.07 0.03 0.03 Measures for unemployed and disadvantaged youth - 0.08 0.07 0.06 0.04 - 0.07 0.03 0.03 Apprenticeship and related forms of general youth training 0.05 - 0.19 0.01 0.38 - - - Subsidized employment 0.13 0.30 0.52 0.39 0.61 0.02 0.35 0.58 0.01 Subsidies to employment in the private sector 0.04 0.03 0.32 0.03 0.56 0.02 0.24 0.15 Support of unemployed persons start- ing enterprises 0.02 0.04 0.03 - - 0.03 0.09 Direct job creation (public or nonprofit) 0.07 0.23 0.20 0.32 0.04 - 0.07 0.35 0.01 TABLE 7.3 (continued) Australia Denmark France Germany Italy Japan Spain Sweden United States Program (1997-98) (1998) (1997) (1998) (1996) (1997-98) (1998) (1998) (1997-98) Measures for the disabled 0.06 0.30 0.08 0.25 - - 0.02 0.62 0.04 Vocational rehabilitation 0.02 0.30 0.02 0.10 - - - 0.04 0.04 Work for the disabled 0.04 - 0.06 0.15 - - 0.02 0.58 - Unemployment compensation 1.17 1.86 4.50 2.29 0.68 0.43 1.64 1.91 0.25 Early retirement for labor market reasons - 1.88 0.35 - 0.20 - - - TOTAL 1.69 5.63 3.22 3.56 1.96 0.52 2.36 3.93 0.43 Active measures 0.52 1.89 1.37 1.27 1.08 0.09 0.72 2.01 0.18 Passive measures 1.17 3.74 1.85 2.29 0.88 0.43 1.64 1.91 0.25 - Not available. Source: OECD Employment Outlook, various years. 312 EAST ASIAN LABOR MARKETS mental and quasi-experimental. Experimental evaluations require the se- lection of both treatment and control groups before the intervention. Treat- ment groups are made up of those receiving assistance; control groups in- clude only nonrecipients. Quasi-experimental studies select these groups after the intervention. Nonscientific techniques do not use control groups in evaluating the impact of interventions, relying instead on statistics compiled by program administrators. Since there is no counterfactual information detailing, for instance, what might have happened in the absence of the program, these evaluations are of little use in determining whether participants benefit. Nonscientific evaluations provide some information on deadweight losses and substitution and displacement effects. EXPERIMENTAL (CLASSICALLY DESIGNED) EVALUATIONS. If large samples are randomly assigned to treatment and control groups, on average neither the observable or unobservable characteristics of the two groups should differ, and any difference in outcomes can be attributed to participation in the program. The main appeal of experimental evaluations, then, lies in the simplicity of interpreting results. The program impact is the simple difference between the means of the treatment group and control group members on the outcome of interest. Although experimental evaluations have many virtues, they have pitfalls as well. First, they require careful plan- ning and design in advance of the experiment. In addition groups may not be assigned randomly owing to factors such as nepotism and the desire to exclude high-risk individuals in order to produce certain results. Partici- pation in experimental analyses may affect participants' behavior (the Hawthorne effect). Further, such experiments typically incur high costs and can raise ethical questions if some people are excluded. QUASI-EXPERIMENTAL TECHNIQUES. With these techniques the treat- ment and control groups are selected after the intervention. To isolate the effects of the program, econometric techniques correct for differences in the characteristics of the two groups. The main appeal of this technique lies in its relatively low cost and in the fact that evaluations can be done at any time after the intervention. The main drawbacks are statistical complexity and a failure to account for all differences in the two subsamples. Tech- niques for adjusting for differences in the observable attributes of a group (such as sex, education, and region) are relatively straightforward, although they are subject to specification errors. Correcting for unobservable char- acteristics (such as motivation) requires a complex procedure that can yield ACTIVE LABOR MARKET POLICIES: ISSUES FOR EAST ASIA 3I3 different results depending on the specification. Quasi-experimental evalu- ations are of three types: regression-adjusted for observables, selectivity- corrected, and matched pairs. • Evaluations that are regression-adjusted for observables assess the im- pact of participation in a program when the observable characteris- tics (sex, age, and education, for example) of the participant and comparison groups differ. They are used when these differences ex- plain variations in outcomes for the two groups. * Evaluations that are regression-adjusted for observed and unobserv- able variables (selectivity-corrected) are used when treatment and control groups have unobservable differences. Such differences arise when selection is not random and participation is thus based on both types of characteristics. Unobservable differences (such as innate abil- ity) can cause nonparticipants to respond differently than they would have if they been included in the program. In these cases evaluations that are regression adjusted for observables only are likely to be bi- ased. The Heckman selectivity method is the most commonly used method of controlling for unobservables. i Matched pairs are used to control for differences in the observable characteristics of the individuals in the control and treatment groups. Individual characteristics always differ, so that the groups are likely to have different success rates in finding employment even in the ab- sence of active labor market programs. To control for these spurious differences, a synthetic control group is constructed using matched pairs. The synthetic control group, a subset of the entire control group, is composed of individuals whose observable characteristics closely match those of the treatment group. Interpreting Evaluation Results We now turn to the evidence on program impacts. While we include some results based on evaluations in developing countries, this evidence focuses primarily on the experiences of OECD countries. Here we merely review the main conclusions emerging from the literature (table 7.4).8 An important caveat with these evaluations concerns their summary nature. The studies treat the programs as "black boxes" by not evaluating issues relating to program design and implementation, staffing, and the intensity and quality of services. These factors are obviously important and will have an impact on a program's likelihood of success. 3I4 EAST ASIAN LABOR MARKETS TABLE 7.4 Evaluation Results for Active Labor Market Programs Program Appear to help Comments Job-search assistance Unemployed adults Relatively more cost- and employment when economic effective than other labor services (19 conditions are market interventions (such evaluations) improving; women may as training) because of low benefit more costs. Possible deadweight losses if not effectively targeted. Training for Women and other No more effective than job- long-term disadvantaged groups search assistance in increas- unemployed ing reemployment probabili- (28 evaluations) ties and postintervention earnings; two to four times more costly. Retraining workers Little positive impact; No more effective than job- displaced positive results mainly search assistance and signifi- in mass layoffs when economy is cantly more expensive. Rate (12 evaluations) improving of return usually negative. Training for youth No positive impact Employment and earnings (7 evaluations) prospects not improved. Negative real rate of return to these programs when costs are taken into account. Employment and Long-term unemployed Benefits to treatment group wage subsidies not significant compared (22 evaluations) with control group. Some- times used by firms as a per- manent subsidy program. High deadweight and substi- tution effects. Public works Severely disadvantaged Long-term employment programs groups prospects not helped. Pro- (17 evaluations) gram participants less likely to be hired for permanent jobs than control group and ACTIVE LABOR MARKET POLICIES: ISSUES FOR EAST ASIA 315 TABLE 7.4 (continued) Program Appear to help Comments less likely to earn the same wage. Not cost-effective if objective is to get people into gainful employment. Microenterprise Relatively senior and Very low take-up rate development more educated groups among unemployed. programs Significant failure rate of (15 evaluations) small businesses. High deadweight and displace- ment effects. High costs- cost-benefit analyses rarely conducted but sometimes show higher costs than for control group. Source: Dar and Tzannatos 1999. JOB SEARCH ASSISTANCE AND EMPLOYMENT SERVICES. Expenditures on these programs range anywhere from 5 percent of the budget for active labor market programs (Denmark) to over 70 percent (the Czech Repub- lic). On average OECD countries spend about a quarter of the funds bud- geted for such programs on employment services, including job search as- sistance. These expenditures include not only the costs of financing job search assistance programs but also the administrative costs associated with operating the unemployment benefit system and other active labor market programs. Of the 19 evaluations we examined, all except 1 are scientific. Six of the scientific evaluations are experimental and 12 quasi-experimental. The evaluations suggest that job search assistance is one of the most successful active labor market programs. In general the programs cost little and are as effective as other more expensive active labor market initiatives. How- ever, much depends on whether the economy is growing and on the avail- ability of public funds (which can be scarce during a recession). Most results for job search assistance programs are positive, though not all. The least successful programs tend to operate during periods of reces- sion and rising unemployment. For example job search assistance to work- ers displaced during mass layoffs in Canada in the late 1980s did not raise 3I6 EAST ASIAN LABOR MARKETS the probability of employment or increase earnings compared with the con- trol group. In fact participants terminated during mass layoffs spent a sig- nificantly greater amount of time searching for jobs than nonparticipants (Fay 1996). At the time, however, unemployment was rising. The effective- ness of job search assistance seems to increase when economic conditions improve and when new jobs are being generated. When unemployment rates were falling in the Netherlands in the late 1980s, program participants were more likely to be employed than those in the control group (OECD 1993). Evaluations in Hungary and Poland show similar results (O'Leary 1998a, 1998b). Studies that examine both cost and effectiveness data generally con- clude that job search assistance is one of the most cost-effective of the active interventions. Leigh (1995) finds that training and retraining pro- grams are two to four times more expensive than measures designed to help workers find jobs and that the interventions are equally effective. Job search assistance is not a substitute for training, however, especially since people who use such assistance may well be more "employment-ready" than individuals who get training. However, if job search assistance and training programs cater to roughly the same clientele, policymakers may prefer the less expensive option. Overall, then, the evidence suggests that job search assistance can have some positive effects and is usually cost-effective relative to other active labor market interventions. A positive correlation exists between the like- lihood that these initiatives will succeed and local labor market condi- tions. Finally, like other interventions job search assistance does not seem to help all types of workers equally-for example, these programs have lit- tle impact on youth. TRAINING PROGRAMS. Training and retraining programs generally ac- count for a significant share of expenditures on active labor market initia- tives (between 40 and 60 percent in most countries, but over 75 percent in Denmark in the early 1990s). Training can reach many different groups. We concentrate on training for three groups: the long-term unemployed, those displaced by mass layoffs, and youth. Training programs for the long-term unemployed. We reviewed 28 studies (6 experimental, 18 quasi-experimental, and 4 nonscientific) of training programs for the long-term unemployed. A few of these studies are longi- tudinal and thus provide evidence on the long-term impact of the pro- grams. Scientific evaluations suggest that these programs can have a positive impact but that they sometimes do not. As we have seen in most cases train- ACTIVE LABOR MARKET POLICIES: ISSUES FOR EAST ASIA 317 ing programs are generally no more effective than job search assistance in increasing the probability of reemployment and postintervention earnings. As with job search assistance initiatives, the success of programs for the long-term unemployed tends to be heavily dependent on the business cycle. Programs perform better when they are instituted during times of economic expansion. In Hungary training outcomes improved as the economy grew (O'Leary 1995, 1998a). In general programs seem to be more effective for women (Puhani 1998; Friedlander, Greenberg, and Robins 1997; Goss, Gilroy, and Associates 1989). Longitudinal studies provide mixed results. While in some cases the positive effects of training dissipate within a year or two after the program ends, in some cases the impacts persist. For ex- ample in Sweden training for the unemployed raised earnings in the short term, but the long-term impact (more than two years) was somewhat neg- ative (Meager and Evans 1998). Conversely the long-term unemployed trained as part of the New Jersey Reemployment Demonstration Project in the mid-1980s were earning more than the control group 30 months after program ended (Anderson, Corson, and Decker 1991). Costs-when they are known-vary substantially. In most cases they are so high compared with the benefits of the program that social returns can remain negative (especially for males) even if the effects persist for 10 years (Friedlander, Greenberg, and Robins 1997).9 The U.S. Job Training Part- nership Act of 1982, which gave employers subsidies to hire workers for on-the-job training, offers a rare exception. Both male and female partici- pants did significantly better than the control group, and the training was relatively inexpensive. However, single mothers benefited more than men (Bloom 1990). In spite of the positive results, some evaluators caution that the aggregate effects are likely to be modest, both for the target population and for the labor force as a whole (Friedlander, Greenberg, and Robins 1997). One of the major implications emerging from our analysis is that train- ing should not be seen as a panacea for securing jobs for the long-term un- employed, especially when demand is slack. Evaluations also show that tightly targeted on-the-job training programs, which are typically aimed at women and other similarly disadvantaged groups, often offer the best re- turns. Rigorous cost-benefit analyses are seldom performed, but the lim- ited evidence that does exist shows that training programs are not usually cost-effective. Retraining programs for workers displaced during mass layoffi. We sum- marized the results of 12 studies of retraining programs for displaced workers: 5 quasi-experimental, 6 nonscientific, and 1 mixed (using both 3I8 EAST ASIAN LABOR MARKETS techniques). The studies relate to retraining programs for workers dis- placed by mass layoffs during major enterprise restructuring and plant clo- sures. Unfortunately none of the evaluations is longitudinal, so we lack in- sights into the long-term effects of these programs. Scientific quasi-experimental evaluations find that though some re- training programs may result in a modest increase in reemployment, the result is often statistically insignificant (Corson, Long, and Maynard 1985). The results for postprogram earnings are even more discouraging. Wage effects on participants (compared with control group workers) are generally negative (OECD 1991). Evaluations seldom report the full costs of retraining. But the available evidence shows that direct costs (usually measured as total recurrent program costs) vary from US$3,500 to US$25,000 per participant (Dar and Gill 1998). As noted earlier given that retraining and job search assistance have roughly similar impacts, job search assistance can be more cost-effective than retraining. Some re- searchers claim that they can find no evidence of an incremental effect for retraining programs that exceeds the effects of job search assistance (Cor- son, Long, and Maynard 1985). As with some other initiatives, these pro- grams may be serving different groups of unemployed workers and thus may not be direct substitutes for one another. The experience of OECD countries with retraining programs for work- ers displaced in mass layoffs may be useful in designing assistance programs in transition countries and other economies that experience large-scale labor redundancies. The evidence on the expense and lack of effectiveness of these programs suggests that they should not be the principal source of support for individuals seeking gainful employment. Instead they should be relatively small and targeted toward the subgroups that will benefit the most from them. Training programs foryouth. We examined seven evaluations, five experi- mental and two quasi-experimental. These evaluations are mostly discour- aging, even though the programs were often introduced during periods of relatively stable youth unemployment. We find that training rarely has an effect on the earnings or employment probabilities of program beneficiaries (Fay 1996). A more mixed (and more promising) picture arises from the evaluation of the Canadian Job Entry Program. A quasi-experimental eval- uation of the program shows that while youth who received only classroom training did no better than the control group, those who undertook enter- prise training did significantly better than the control group (OECD 1993). This positive effect was attributed to the fact that many of those trained in enterprises stayed on with the firms. Cost-benefit analyses of several of the ACTIVE LABOR MARKET POLICIES: ISSUES FOR EAST ASIA 3I9 youth training programs suggest that the social rates of return to these pro- grams are typically negative in the short as well as the long run (Friedlander, Greenberg, and Robins 1997). Correcting the failures of the educational system with what is generally short-term training is very difficult. WAGE AND EMPLOYMENT SUBSIDIES. We analyzed 22 evaluations, 6 of them experimental, 8 quasi-experimental, 7 nonscientific, and 1 mixed. Wage and employment subsidy programs are some of the most poorly funded active labor market initiatives in OECD countries, attracting less than 10 percent of expenditures on active labor market programs. Fund- ing for these programs is also negligible in the United States and the United Kingdom. Evaluations tend to agree that wage and employment subsidy programs have high deadweight losses and substitution effects. In the most extreme case, the losses from Ireland's wage subsidy program totaled over 95 percent, while the program's net incrementality was a meager 4 percent (OECD 1993). Evaluations of similar programs in Australia, The Netherlands, and the United Kingdom also show losses, though not as severe (table 7.5). Equally disappointing are evaluations that compare wage and employ- ment outcomes of participants with those of a control group. For example TABLE 7.5 Losses from Wage Subsidy Programs (in percent) Deadweight and Country substitution effects Additionality Australia, mid-1980s (Jobstart Program) 65 D 35 Belgium, early 1990s (recruitment subsidy) 89 (53 D, 36 S) 11 England, 1986-90 (training grant) 69 D 31 England, late 1980s (Workstart 1) 75 (45 D, 30 S) 25 England, mid-1 970s (employment subsidy) 70 D 30 England, early 1980s 73 (63 D, 10 S) 27 Germany, mid- 1970s (wage subsidy) 75 D 25 Ireland, 1980s (Employment Incentive Scheme) 91 (70 D, 21 S) 5-10 Netherlands, early 1980s (Vermeend- Moor Act) 75 (25 D, 50 S) 25 Netherlands, late 1980s (JOB scheme) 80 S 20 United States, mid-1980s (Targeted Job Tax Credit) 80 (70 D, 10 S) 20 Note: D = deadweight losses; S = substitution effect. Additionality is the employment effect after accounting for deadweight, displacement, and substitution effects. 320 EAST ASIAN LABOR MARKETS a longitudinal study of the U.S. Targeted Job Tax Credit program found that participants earned significantly more than individuals in the control group for the first year. But this effect declined in the second year and dis- appeared after that (OECD 1993). Similarly evaluation results for Hun- gary show that participants were significantly less likely to be employed and earned less (though not significantly so) than those in the control group (O'Leary 1998a). However, a few exceptions do exist. For example Australia's program experienced deadweight losses of around 30 percent, but the subsidies had a significant positive impact on postprogram employment. Participants were 15 percent more likely to be employed than members of the control group (Webster 1998). And as we have seen, the U.S. Job Training Part- nership Act had positive effects for participants (Bloom 1994; Friedlander, Greenberg, and Robins 1997). Overall, however, the high deadweight losses and substitution effects strongly suggest that wage and employment sub- sidies are unlikely to have positive social returns that economists can mea- sure, although these initiatives may help reduce social exclusion among older workers and single mothers. PUBLIC WORKS PROGRAMS AND PUBLIC SERVICE EMPLOYMENT. Public works programs are among the most heavily funded active labor market interventions in OECD countries. We summarized the results of 17 evalu- ations of public works programs, 13 of them quasi-experimental and 4 nonscientific. These evaluations suggest some general conclusions. First, nonexperimental evaluations show that these programs have some desirable short-run effects in the form of increases in employment and de- clines in unemployment. Second, some scientific evaluations suggest very high displacement effects (in Sweden they reached 100 percent) (Skedinger 1995). Third, workers who participate in these programs are less likely to find employment than their counterparts in the control group and tend to earn less. Finally, these programs do not significandy reduce long-term unemployment, and whatever small short-run impact they do have tends to diminish over time (Webster 1998). These conclusions are to a large extent predictable. Unlike other active labor market initiatives, public works provide current benefits and are only temporary escape routes from unemployment. Irrespective of the merits of such projects, an economist's first impression is that public works are generally expensive and are not an effective instrument if the objective is to get people into long-term gainful employment (table 7.6). ACTIVE LABOR MARKET POLICIES: ISSUES FOR EAST ASIA 32I TABLE 7.6 Annual Cost of Job Creation in Public Works Egypt Honduras Nicaragua Madagascar Bolivia Senegal Ghana Cost per job (US$) 1,401 2,120 2,580 786 2,700 5,445 2,122 Cost per job (PPP)Y 7,212 9,759 14,302 3,620 9,388 12,100 10,610 Per capita GDP (US$) 790 600 380 230 800 600 390 Ratio (1:3) 1.77 3.53 6.79 3.42 3.38 9.08 5.44 a. PPP = purchasing power parity. Source: Subbarao (1997); World Bank (1997). Public works can serve as a short-term antipoverty intervention. For this reason some developing countries use them extensively during periods of hardship. Keddeman (1998) suggests that labor-intensive infrastructure construction projects, if carefully targeted and properly designed and im- plemented, not only provide a valuable safety net but also contribute to further economic recovery and development. Increasing the involvement of local communities and the private sector in the design and implemen- tation phases can lead to improved outcomes. MICROENTERPRISE DEVELOPMENT AND SELF-EMPLOYMENT SCHEMES. We summarized the results of 15 evaluations of programs aimed at help- ing unemployed individuals start their own businesses. Two of these eval- uations are experimental, seven are quasi-experimental, five are nonscien- tific, and one is mixed. These programs are described in various ways-as microenterprise or self-employment schemes, for example-but we refer to them as microenterprise development assistance. These programs have high deadweight losses. Estimates of such losses range from about 30 percent (for the self-employment experiments in the U.S. states of Massachusetts and Washington) to over 50 percent (for Can- ada's self-employment assistance program and Denmarkls enterprise start- up grant scheme) (Fay 1996; Graves and Gauthier 1995; Balakrishna 1998). Evaluations show that businesses started with assistance from these schemes are short-lived. Typically one-third to one-half of them close down in the first year of operation (table 7.7) .0 Even when businesses survive, the multiplier effect is small. Most surviving businesses create, on average, half an additional job. In Hungary each surviving enterprise created 0.3 addi- tional jobs, in France 0.5, and in Australia (during a period of declining un- 322 EAST ASIAN LABOR MARKETS TABLE 7.7 Failure Rates for Businesses Receiving Microenterprise Development Assistance (percent) Program Failure rate of businesses Australia (New Enterprise Initiative, late 1980s) 58 (1 year), 71 (2 years) Canada (Self-Employment Assistance Program, early 1990s) 20 (1 year) Denmark (Enterprise Allowance Schemes, late 1980s) 60 (I year) France (Microenterprise Development, early 1980s) 50 (4.5 years) Hungary (Microenterprise Development Assistance, mid-1990s) 20 (15 months) Netherlands, early 1990s 50 (4 years) Poland (Microenterprise Development Assistance, mid-1990s) 15 (2 years) United States (state of Washington, Self-Employment Experiment, 1990) 37 (15 months) employment) 0.7 (OECD 1993; Wilson and Adams 1994). Businesses that receive mentoring and business counseling are more likely to succeed, and even though a majority of them may close, the entrepreneurs are likely to move to employment in another industry rather than into unemployment. Despite heavy deadweight losses and high rates of business failures, par- ticipants in these programs fare reasonably well in terms of employment outcomes compared with control groups. Scientific evaluations show that participants are more likely to be employed than individuals in the control group. However in this case employment does not necessarily translate into higher earnings. For example, in the Washington self-employment ex- periment, participants were more likely to find employment than the con- trol group but earned significantly less (Fay 1996). In Hungary and the Czech Republic, participants were more likely to be employed than indi- viduals in the control group but earned US$30 per month less. In Poland, however, participants were 25 percent more likely to be employed than the control group and earned significantly more. In the Hungarian and Polish programs, women and older workers generally had better outcomes than individuals in other subgroups (Fretwell, Benus, and O'Leary 1999). The cost-benefit issue is rarely addressed. The available data indicate that the cost of starting up a small business varies from $4,500 in France ACTIVE LABOR MARKET POLICIES: ISSUES FOR EAST ASIA 323 to $13,000-14,000 in Canada and Denmark. The Canadian evaluation concludes that the long-term cost-effectiveness of these programs is un- certain. Preliminary analyses from Poland and Hungary indicate a loss to the unemployment insurance system. In both countries the duration and amount of benefits are greater for participants than for the control group. But drawing any conclusions on the cost-effectiveness of these programs would be premature given the small amount of evidence. Overall the evaluations show that microenterprise development assis- tance programs work for only a small subset of the unemployed popula- tion and that they have high deadweight losses and displacement effects (as well as high rates of business failures). In the end, then, the net effects of these programs are small. As in the case of training, assistance targeted at particular groups, such as women and older individuals, shows the great- est likelihood of success. The East Asian Experience with Active Labor Market Programs East Asia does not have a strong tradition of active labor market programs. The one exception is the fairly widespread use of public works to create short-term earning opportunities. Other instruments, such as retraining and employment services, are not used on any significant scale to integrate unemployed workers into the labor force. In part this policy reflects the stage of development of most countries in the region and the prevailing ideologies regarding the role of public policy. It also reflects the low un- employment rates in most countries before 1997. The potential for using active labor market initiatives in the region changed with the crisis. The increases in unemployment, the difficulties many laid-off workers face in finding new employment and earning an ad- equate salary, and the obstacles confronting young people trying to enter the labor force all underscored the need for policymakers to consider new options. While the crisis increased the need for new programs, however, fiscal pressures soon restricted the spending capacity of governments. Even with these constraints governments in all countries are assessing how they can use active labor market interventions to help alleviate un- employment problems (table 7.8). These programs are particularly im- portant because unemployment insurance exists only in Korea. But the governments of the region must address many problems if such initiatives are to make an important contribution, not only to the crisis but to the development of a long-term labor policy framework. TABLE 7.8 Employment Services Services offered Public-private provision Labor market information Crisis-specific interventions Indonesia 1. Job search assistance, 1. Strict regulation of 1. Limited computer literacy None vacancy tracking, and private agencies among administrative placement clerks 2. Voluntary registration of 2. Interagency overlap and unemployed workers communication problems Korea, Rep. of 1. Registration and 1. Recent easing of 1. New Internet-based 1. Increased number of administration of restrictions on private career guidance and job public employment unemployment benefits agencies search system (Work-Net) service agencies. 2. Job search assistance, 2. New "Worker Profiling 2. New one-stop vacancy tracking, and System" "Employment Security placement 3. Major revisions in Centers" 3. Career guidance and occupational counseling classification Malaysia 1. Job search assistance, 1. Licensing requirements 1. Computer-based 1. Tracking and monitoring vacancy tracking, and for private agencies Employment Service of placements through placement Automatic Reporting task force 2. Career guidance and System counseling 3. Monitoring of labor mobility TABLE 7.8 (continued) Services offered Public-private provision Labor market information Crisis-specific interventions Philippines 1. Registration of 1. Innovations in private 1. Long-term project under 1. 146 new Public unemployed, national provision (radio station GATT Adjustment Employment Service skill registry airs job openings and Program known as Offices (PESO) in 1998 2. Referral for job takes calls from job Systematization of Labor 2. National facilitation placement locally and seekers and employers) Market Information and network in every overseas 2. Government regulation Employment Counseling province and key city 3. Career guidance, job of private agencies 2. Job-matching service fairs, and placement in 3. Growth of successful (Phil-Jobnet) other programs private agencies, especially for overseas placements Thailand 1. Job search assistance, 1. Monitoring and 1. Very little coordinated or 1. No significant new vacancy tracking, and supervision of private comprehensive effort to initiatives placement agencies (under Ministry collect data 2. Job fairs and counseling of Labor and Social Welfare) 326 EAST ASIAN LABOR MARKETS Employment Services All the East Asian countries have public employment services that have two responsibilities: to gather information on job seekers and job vacancies and to provide access to services that will match the two. These services are delivered through a network of employment bureaus in each country. The Philippines has an extensive network of public employment service offices (1,825 across the country), buttressed by new offices opened since the crisis. Such offices also seem plentiful in the other countries (for ex- ample, Korea has about 120, Malaysia about 40, and Thailand 85). But the size, population, and topography of most of the countries make pro- viding complete geographical coverage difficult. Another constraint stems from what public employment services can offer in terms of services for the unemployed. In most industrial countries the public employment service is driven by three activities: * Delivering unemployment insurance benefits (and validating job availability and other criteria); * Serving as a point of access to labor market programs like retraining and job search assistance; and * Disseminating labor market information to workers, employers, and service providers. The lack of unemployment insurance, limited number of active labor market initiatives, and shortage of labor market information (and the fact that when it is available it is inadequately used) mean that public employ- ment services in East Asia do not serve these functions. On the positive side, however, policymakers can take advantage of the experiences of effec- tive employment services in industrial economies. We know, for instance, the importance of coordinating and integrating services, of providing com- plementary public and private service delivery, and of using technology effectively. But ultimately the value of public employment services as an element of active labor market policy depends on how well the services are integrated with other aspects of labor policy (active and passive). East Asian countries are making efforts to foster such integration. Korea and the Philippines (and possibly Indonesia) are transforming traditional em- ployment offices into one-stop centers where job seekers can access un- employment benefits (in the case of Korea), job search assistance, and vo- cational training opportunities. Throughout the region committees, task forces, and government and tripartite bodies are examining how inte- grated and coordinated services can improve labor adjustment. ACTIVE LABOR MARKET POLICIES: ISSUES FOR EAST ASIA 327 All the countries also allow private companies to provide employment services alongside the public systems, although in some cases (for instance, Indonesia), some concerns exist about the efficiency of government regu- lation. Currently private service providers are facing difficulties because of depressed business conditions (especially in Malaysia), while public sys- tems are being forced to handle rapidly increasing loads. In Korea, for ex- ample, the number of job seekers using public employment services in- creased almost 10-fold between 1997 and 1998. East Asian countries vary in their use of information technology, and its apparent effectiveness also varies throughout the region. In Malaysia and Korea employment offices in different geographic areas can share in- formation about job seekers and vacancies on computerized networks and the Internet. In Malaysia the Employment Service Automatic Reporting System assists the Manpower Department in matching job seekers and job vacancies. In Korea the government has an electronic labor exchange sys- tem, "Work-Net," which provides information on job vacancies, voca- tional training, career guidance, employment policies, unemployment in- surance, and labor market statistics. In the Philippines the public can search for job vacancies and find application information at selected gov- ernment offices and on the Internet through "Phil-Jobnet," an electronic information system. Deficiencies in training, networking, and the computers themselves continue to constrain the effectiveness of information technology in much of the region. The Philippines network, for example, registered only 1,500 vacancies and 6,000 workers in 1998, its first year of operation. In Indo- nesia computerized data in district employment service offices are not available in other regions. And because many staff do not have appropri- ate training or experience, much of the data is still collected and analyzed manually. Some limited activity is taking place in the areas of counseling and place- ment services. Korea stepped up its placement and counseling services by rapidly increasing the number of counselors in the public employment service offices. In Thailand and the Philippines, the public employment service holds job fairs to bring prospective employers and job seekers to- gether. Counseling is provided at the job fairs, as well as at educational in- stitutions and other outreach venues. How is the effectiveness of employment services in the region evalu- ated? Placement rates, the most commonly used indicator, fell once the crisis began. In Korea, for example, a survey on unemployment and wel- fare needs found that only 5.8 percent of the unemployed succeeded in finding jobs through the public employment services. This finding sug- 328 EAST ASIAN LABOR MARKETS gests that unemployed workers in Korea find jobs through other more im- portant channels. Information from Thailand suggests various reasons for, the difficulties registered workers have finding jobs through the public employment service, for two reasons. First, a mismatch exists between workers' qualifications and job requirements (and between workers' ex- pectations and the available jobs). Second, employers continue to rely on other means of finding workers. Labor Market Training Training is an increasingly popular intervention in East Asia. The region is moving along a long-term development path toward industrial activities that require highly skilled workers. All the countries now have relatively complex systems of vocational training at the national level (table 7.9). For the most part these systems are funded by the public sector and operated by the government. Not surprisingly the crisis changed the context for training. In general the increase in unemployment created pressure to step up training activities designed to reintegrate laid-off workers into the job market and provide social relief. In Korea, for example, the number of people participating in vocational training programs for the unemployed rose eightfold between 1997 and 1998. In Thailand the Department of Employment launched a special program of three-month courses in 1998 to retrain workers laid off because of the financial crisis that drew more than 27,000 unemployed workers. Vocational training and education in Indonesia, however, de- creased once the crisis began because of the worsening fiscal situation. Despite the growing popularity of training programs in the wake of the crisis, governments need to address a number of important challenges in order to improve job training in the long term. Some of these challenges are common across East Asia, including enhanced coordination and a stronger demand-side orientation. Coordination among the various arms of gov- ernment responsible for vocational education and training is currently limited. For example, in the Philippines and Thailand three separate gov- ernment agencies have overlapping responsibility for vocational educa- tional and training activities. The situation in Indonesia is even more problematic: a recent report identified 19 different departments that were involved in 815 vocational training programs (see chapter 2). Virtually all the East Asian countries need to make their vocational training programs more responsive to labor demand. Specifically they need to generate more high-technology training, and some governments have IA1LtE /.9 Labor Market Training Vocational education system Public and private provision Training innovations Crisis-specific interventions Indonesia 1. 19 different government 1. Estimated capacity of 1. German-style dual 1. Public vocational departments run 815 private training centers education system education somewhat vocational training more than that of the operating in 11,000 threatened by fiscal programs public SMEs constraints of crisis 2. Ministry of Manpower 2. Heavy regulation of oversees 156 public private training and lack training institutions of public-private coordination Korea 1. National training system 1. Approximately 500 1. Pilot program of training 1. Eightfold increase in since 1967 vocational training vouchers since 1998 training programs for 2. Ministries of Labor and institutes, 237 in-plant the unemployed from Education responsible centers, and 178 centers 1997 to 1998 for programs run by nonprofit 3. 3.3 million in training organizations and private courses since 1967 firms Malaysia 1. Vocational training 1. Government promotion 1. Training fund for skill None policies overseen by of private training but upgrading financed by Ministry of Human Re- little coordination levies on firms sources and administered between public and through the National private Vocational Training 2. Many agencies not Council (public accredited training) and the 3. Dearth of public Human Resource Devel- instructors owing to low opment Council (private pay sector training) (Table continues on thefollowingpage) TABLE 7.9 (continued) Vocational education system Public and private provision Training innovations Crisis-specific interventions Philippines 1. Technical Education and 1. 723 public and 1,383 1. Government-funded 1. Expansion of scholarship Skill Development private vocational scholarships for students program and shift Authority responsible for technical centers in private training toward private provision programs since 1994 2. 80 percent of trainees in institutions of services 2. Programs smaller than in private institutions 2. Dual-training system other East Asian being tested countries Thailand 1. 413 training centers 1. Restrictions on private 1. Training institutions to 1. New 3-month training under the Department agencies relaxed be improved by hiring courses for laid-off of Vocational Education 2. Tax deductions for 2,155 new college workers (27,000 since (Ministry of Education) company training graduates to set up July 1998) 2. New entrants and expenses national database on 2. Donor-financed training current employees under 3. Small but growing training needs by for groups such as entre- the Department of Skill private training sector province, assist staff in preneurs, rubber-tappers, Department (Ministry of training and supervision, electronics repair Labor and Social and execute in-school persons, and tailors Welfare) job creation dual- training project for students ACTIVE LABOR MARKET POLICIES: ISSUES FOR EAST ASIA 33I introduced initiatives in this area. The Malaysian government, for instance, established three centers for high-technology skills training and negotiated bilateral agreements with France and Germany for additional facilities. Countries are also making attempts to increase the role of the private sec- tor in training. Private training institutions represent one vehicle for getting the private sector involved in training. The Philippines and Thailand report increasing numbers of private providers, but public institutions continue to predominate in the region. Even in the Philippines, which probably has the highest number of private trainers, the private sector faces constraints in terms of access to credit and competition from highly subsidized public in- stitutions. A more neutral policy environment or one that encourages the entry of private providers is likely to lead to a more efficient and responsive training sector."1 Creating such an environment requires governments to set standards and develop accreditation arrangements. Governments are introducing a number of innovations in financing for training. While there is no guarantee that all the innovations will be use- ful in the long run, policymakers need to try new ideas in this area, given the well-known financial market imperfections associated with human capital. In Korea the Employment Insurance System provides financial in- centives both for employers to provide training and for employees to take it. Employers can apply for money from a special fund to implement var- ious educational and training activities, including offering paid leave for workers who participate. Korea launched a pilot program for training vouchers in 1998 with the objective of providing more choice in the train- ing system. The Philippines also introduced a voucher-style scholarship program through the government-sponsored Private Education Student Finance Assistance program. In Malaysia the Human Resources Develop- ment Council encourages retraining and skill upgrading for current em- ployees in part by requiring firms registered with the council to contribute to a training fund. A number of countries in the region are now considering German-style vocational education systems that integrate vocational and technical edu- cation with work experience. In 1994 the Indonesian government intro- duced a "dual" system, and all the public institutions offering vocational education, plus a higher-than-expected number of small and medium-size enterprises (SMEs), are participating. The Philippines is testing an ap- prenticeship program modeled on the German system. The program lasts 30-36 months, with trainees spending 70 percent of their time in the firm and 30 percent at a training center. Trainees receive 75 percent of the min- imum wage but retain only 30 percent of their wages after paying the 332 EAST ASIAN LABOR MARKETS training centers.12 While such dual systems have strong features, they have been criticized for a lack of flexibility, and some firms are unwilling to in- volve themselves in these programs. Job-Creation Activities Traditionally job creation is the major active labor market policy in all the East Asian countries (table 7.10). Most job-creation activities take the form of public works, but these projects are generally motivated by social relief objectives rather than by employment development goals. Two other activities used in East Asia fall under the rubric of job creation programs. Both Malaysia and Korea have employment subsidies to firms to encour- age new hiring and maintain existing jobs. And all East Asian countries have some type of initiative to support the self-employed and SMEs. Public Works All five countries implemented substantial public works programs as a so- cial safety net provision in the wake of the 1997 financial crisis. While these countries had offered similar programs before, the extraordinary growth of the mid-1990s substantially reduced the need to create infrastructure and provide jobs. In Indonesia, for example, labor-intensive job creation pro- grams officially ended in 1994, only to be resurrected in 1997 following an extreme drought, the economic crisis, and political turmoil. Design flaws can limit the effectiveness of public works programs, and these flaws are issues in East Asian countries. Some are the result of the haste with which the projects were put together or expanded. In Indone- sia, Korea, and the Philippines, public works programs struggled to reach their target populations because wages were set too high. Other design flaws in Indonesia included a low wage bill (as a percentage of total proj- ect costs) and the absence of women in most public works projects. Proj- ects in the Philippines suffered from insufficient monitoring and a failure to identify those areas most in need of antipoverty initiatives. The rush to respond to the crisis before all the recent gains in eco- nomic, social, and human development were lost created numerous prob- lems in policy coordination and follow-through on public works projects. The Thai government, for example, hurried to design and implement new public works projects by soliciting input from various ministries. The re- sulting menagerie of programs included a number that had previously been dismissed or rated as low priorities but that were implemented any- TABLE 7.10 Job Creation Activities Public works Employment subsidies Self-employment and SME support Indonesia 1. A 4-month, RP 42 billion- None 1. Percentage of commercial lending program targeting retrenched set aside for small borrowers, construction and manufacturing subsidized credit for cooperatives workers (implemented December 2. At least 24 credit programs for 1997) groups such as farmers, trans- 2. Massive public works program migrants, and women (micro- with 16 subprograms targeting entrepreneurs and SMEs) various groups in all provinces (implemented April 1998) Korea 1. Major focus of postcrisis policy 1. Subsidies offered to firms that 1. Postcrisis support through new (400,000 participants in 1998) maintain employment by Korea Venture and Investment 2. Internship programs for high reducing work hours, offering Fund Cooperative, a public fund school and college graduates paid leave, training employees, of 100 billion won expected to create 10,000 and temporarily shutting down, or 2. "Business incubators" offering 57,000 jobs, respectively dispatching workers to weaker management training, technolo- affiliates gical capacity, and office space (Table continues on the following page) TABLE 7.10 (continued) Public works Employment subsidies Self-employment and SME support Malaysia 1. Public works programs in infra- 1. Primary labor market strategy that 1. Industrial linkages, technology structure and agricultural and encourages employers to use pay development and acquisition, rural development cuts, temporary layoffs, flexible enterprise development, financial work hours, and part-time work support, and skill upgrading rather than retrenchment provided by Small and Medium Industries Development Corporation Philippines 1. Public works projects (with 1. Wage subsidy program for 1. Over 100 "livelihood" programs multiple objectives) in students (children of displaced administered by government transportation, agriculture, workers) during school vacations 2. Multiple credit and guarantee environment, communications, facilities for SMEs and housing departments Thailand 1. Primary labor market policy None None response to crisis, with Ministry of Interior projects for 1999 creating an estimated 788,799 jobs 2. Additional 11 million person- months provided by donor- financed projects ACTIVE LABOR MARKET POLICIES: ISSUES FOR EAST ASIA 335 way in lieu of expanding existing high-priority public works projects. The numerous public works programs implemented in 1998 in Indonesia suf- fered from a lack of conceptual clarity about objectives and beneficiaries, in part because of inadequate labor market information. As in Thailand there were numerous problems of overlap and lack of coordination among implementing agencies and a virtual absence of program monitoring. Wage and Employment Subsidies Providing incentives to employers to maintain current employment and create new jobs is not a major policy tool among East Asian governments. The exceptions are Malaysia, where such subsidies are the primary labor market strategy, and Korea, which implemented subsidies in the immedi- ate aftermath of the crisis. The Philippines has two small wage subsidy programs targeted to young people. In Malaysia the government launched efforts to encourage the private sector to choose pay cuts, temporary layoffs, flexible work hours, and part- time employment for workers rather than retrenchment. Initial evalua- tions of government incentives report that these measures are having some success in preventing unemployment. In particular employers are choos- ing to cut pay rather than instituting voluntary retrenchment or simply re- trenching workers. In Korea employment stabilization programs immedi- ately following the crisis took several approaches. The government offered subsidies to firms that agreed to maintain their current workforce using any of the following measures: shutting down temporarily, reducing work- ing hours, offering in-firm training or paid leave, reassigning workers, and even taking up a new line of business while retaining at least 60 percent of the staff. The employment effects of this program appear positive. In a survey of 533 firms receiving government support, managers estimated that they would have had to layoff 22.3 percent more workers in the ab- sence of subsidies (see chapter 3). Supportfor the Self-employed and SMEs Many programs in the region support the self-employed and SMEs with a variety of technical services and, most prominently, financial assistance. A number of the programs are not driven by immediate labor market con- cerns but are part of the overall industrial policy envelope. Because gov- ernments now view them as instruments for encouraging employment, we include them here. 336 EAST ASIAN LABOR MARKETS Financing can be an obstacle for microentrepreneurs and small enter- prises, and governments everywhere have responded by introducing pro- grams designed to increase access to credit for these groups and in some cases to improve the terms of financing. After the East Asian crisis began, declines in product demand, increases in production costs, and a credit squeeze exacerbated the financing issue, raising demand for further gov- ernment intervention to protect SMEs and their employees. Most of the programs existed prior to the crisis but received no additional funding afterward. In the Philippines SMEs had access to credit through government pro- grams long before the crisis. According to a World Bank evaluation, par- ticipating firms perform better in many dimensions than other small firms, but the magnitude of this job-creation effort and its effectiveness as a poverty reduction tool are relatively minor (World Bank 1998). The Philippines also offers self-employment initiatives through various govern- ment departments. The Self-Employment Assistance livelihood program offers access to credit as well as social services delivered at the community level and targeted to poverty reduction goals. In Malaysia the government created a number of schemes for entrepre- neurs that are expected to assist about 12,000 traders and small businesses in setting up or expanding their businesses. Indonesia currently has at least 24 credit programs for microentrepreneurs and SMEs. Some target special groups such as farmers and transmigrants or subsidize credit for coopera- tives (typically groups of SMEs with common interests). The latter pro- grams have come under criticism for allowing cooperatives to profit at the government's expense by borrowing cheaply and investing the money at higher interest rates. A recent assessment of credit programs for SMEs in Indonesia argues that the best government intervention in support of SMEs is the development of an environment conducive to the efficient operation and development of the commercial financial institutions serv- ing them (Wieland 1998). Various technical assistance programs are also available in these five countries. The Korean government is developing "business incubators" to assist entrepreneurs with management training, technological capacity, and office space. The Malaysian government provides opportunities for self-employment not only through financing assistance but also through franchise development, various farming extension services, and basic busi- ness training for unemployed graduates. Indonesia provides technical as- sistance through different channels. For example prior to the crisis the government had established the Small and Medium Industries Develop- ment Corporation (SMIDEC) to assist industries with issues such as in- ACTIVE LABOR MARKET POLICIES: ISSUES FOR EAST ASIA 337 dustrial linkage, technology development, technology acquisition, market development, enterprise development, financial support, and skill devel- opment and upgrading. The Future of Active Labor Market Policies What can policymakers in East Asia conclude from our review of active labor market programs? First, the OECD experience illustrates the range of programs industrial countries have undertaken and the issues that are in- volved in this area of public policy. While various types of interventions exist that address different policy objectives, our review of the evidence from evaluations certainly sends out cautionary signals. Ultimately active labor market programs are judged by their performance in improving the em- ployability and earnings of workers, and the evaluations reviewed here sug- gest that these programs often have little or no impact on these outcomes. However in putting forward this evidence, we are not arguing that pol- icymakers in countries that have not made major investments in active labor market programs should avoid this area in the future. First, active labor market programs can serve social as well as economic objectives. Re- searchers have not addressed the question of the programs' social impacts, which may be more positive. Second, work-force development, the social and economic integration of marginalized and at-risk groups, and the sta- tus of unemployed workers are central concerns for policymakers. Active labor market programs are obvious instruments for addressing these issues. Third, the disappointing performance of these programs in the aggregate masks the fact that some program designs do seem to lead to positive out- comes for some types of workers. Moreover in the 1990s policymakers in some countries utilized novel and promising approaches to active labor market programs, including finding new roles for the private sector and instituting innovative tripartite arrangements. The challenge is to learn from experiences such as these and to design future programming along lines that appear to work. The evidence from the evaluations suggests that policymakers should exercise caution in two areas. First, policymakers should be realistic about what active labor market programs can do. Second, they should make only careful and modest investments in this area. As policymakers in the East Asian countries look toward the future, we recommend that they carefully consider the following issues related to active labor market policies. Priority setting. As we have noted, active labor market programs can have multiple policy objectives, including reducing unemployment dur- ing cyclical downturns, correcting structural imbalances, improving labor 338 EAST ASIAN LABOR MARKETS market functioning, and assisting disadvantaged groups of workers. In de- signing an overall strategy, policymakers need to identify which of these objectives will have top priority and therefore determine the choice and design of programs. We do advocate the use of private placement agencies. But because these businesses serve only relatively skilled white-collar workers, they are no substitute for public employment services. Regardless of the specific objectives, one immediate and high priority is to develop a strong employment service-the first link in the chain of active labor market policies. The roles of the public and private sectors. This consideration is key both to developing an overall strategy and to designing and implementing pro- grams. At one time OECD governments developed and delivered virtually all active labor market programs, but increasingly governments are recon- sidering the role of the private and nonprofit sectors in program delivery. In many countries these sectors now play important roles in the delivery of services. This development can lead to more diverse, innovative, and cost- efficient services and to programs that respond to demand. But even in countries where the scope for private sector involvement is considerable, governments still play the central role. They must remain responsible for the overall system, ensuring that it remains focused on public priorities. They must also address distributional issues (for instance, ensuring that all types of workers receive adequate services) and provide public goods. They also must be the catalysts in the process of harnessing private sector in- volvement in retraining and other active labor market programs.13 Partnerships and dialogue. The process of identifying priorities for active labor market policy and program choices can benefit from ongoing dia- logue among government, business, labor, and other relevant organiza- tions (such as private and nonprofit service providers). When this dialogue is conducted effectively, policymakers understand the needs of the labor market and can maximize support for active programs and policies. The dialogue needs to be carried out both at the national level, where priori- ties are set, and at the local level, where programs are delivered. Govern- ments typically must be the leaders and catalysts of this process. "Infrastructure"for the labor market. Infrastructure services are essential if active labor market programs are to be useful policy instruments. By infrastructure we mean labor market information, a viable and complete network of employment service offices, and certification and accreditation systems. These services are the cornerstones of an effective system, inform- ing the program choices that should be made. They provide the bridges among the labor market, service deliverers, workers, and employers. And ACTIVE LABOR MARKET POLICIES: ISSUES FOR EAST ASIA 339 they ensure quality throughout the system. As largely public goods, they are inevitably the responsibility of governments. In countries where the development of active labor market policies is at an early stage, these in- frastructure services should be the first priority. Coordination within government. In many countries (including those in East Asia), a number of government agencies share the responsibility for administering active labor market programs. In these cases overlapping re- sponsibilities and a lack of coordination complicate service delivery. Too often mechanisms to ensure that the various departments are working to- gether are not in place. In addition overall economic planning does not take into account the need to set priorities and design effective active labor mar- ket programs. Coordination within the government needs to be improved to ensure the relevance and efficiency of active labor market programming. Policy and administrative and operational capacity. Designing and im- plementing active labor market programs requires considerable capacity within the government itself. In many ways active programming is more complicated than passive income support programming, especially since capacity needs differ significantly by program. Public works can be rela- tively straightforward to design and implement-so much so that these programs are often the primary active labor market intervention in many developing countries. Employment services, however, require a network of facilities with extensive geographic coverage, skilled counselors, reliable connections with employers and the educational communities, and the technological resources and know-how to generate and disseminate accu- rate and timely labor market information. Training programs also require labor market information, training and occupational standards, monitor- ing and evaluation capabilities, and capacity (increasingly in the private or nonprofit sectors) to deliver good programs. Governments must recognize that capacity building is a slow but necessary process. Financing. The most important question in this area concerns the balance of public and private financing. Clearly the rationale for public spending is strong: market failures exist with respect to human capital in- vestments, and active labor market policies have some elements of public goods. But there are also private gains to employers and employees as a re- sult of training and other interventions, and governments need to think about how financing can reflect such gains. They can consider the applic- ability of innovative financing arrangements (such as income-contingent loans) that address market imperfections but reflect the fact that active labor market programs offer some private returns. In terms of public fi- nancing the essential choice for policymakers is drawing from general rev- 340 EAST ASIAN LABOR MARKETS enues or creating a fund financed by employer and (perhaps) employee contributions. Each option involves important considerations, including fungibility, assumptions about responsibility for labor programs, funding integrity, and incentives for formal employment creation. OECD coun- tries, which use a full range of funding arrangements, offer useful lessons for countries in East Asia. Monitoring and evaluation. In spite of large public expenditures on ac- tive labor market programs, OECD countries generally have not under- taken rigorous evaluations. In an effort to improve the targeting and effi- ciency of social programs, governments need to conduct sound impact evaluations. These evaluations can compare labor market outcomes for program participants with those of a control group. Data on program costs can be used to determine the impact on individuals, the net social gains, and the outcomes-that is, to determine whether the programs provide the optimum results for the money spent. Clearly many considerations are involved in developing a strong active labor market policy. The experience of the OECD countries and the ex- tensive resources such programs require suggest that East Asian countries should move slowly to build on what already exists. Over the long run building the capacity to implement active labor market programs will be important as formal labor markets grow and the need for a skilled work- force increases. Countries now need to think about priorities, the role of government, and a range of issues related to how these programs are best carried out. Given its significance, labor market "infrastructure" demands immediate attention. As the East Asian economies continue to develop, active labor market policies will need to become part of the overall policy tool kit. Notes 1. Objectives can focus on the needs of employers as well-for example, the need for an adequate supply of trained (and trainable) workers. These needs may take priority in times of rapid expansion, when vacant jobs rather than unem- ployed workers are the predominant form of labor market imbalance, as they were in the 1960s when active labor market programs were first introduced on a sig- nificant scale. 2. For an extensive listing of active labor market programs, see OECD (1993). 3. Displacement effects usually occur in the product market. A firm with subsidized workers increases output, displacing output from firms with no subsi- dized workers. Displacement effects also occur when some individuals receive support for self-employment activities and others do not. The substitution effect ACTIVE LABOR MARKET POLICIES: ISSUES FOR EAST ASIA 34I occurs when a subsidized worker replaces an unsubsidized worker who would otherwise have been hired, for a zero net employment effect. 4. Deadweight losses occur when program outcomes are no different from what would have happened in the absence of the program. In this case the workers receiving job search assistance would have found jobs on their own. Similarly, wage subsidies may go to support workers who would have been hired without them. 5. We count as active labor market policies only public works designed specif- ically to alleviate unemployment and poverty, not those routinely planned to con- struct infrastructure. 6. Grameen Bank, which has more than 2 million members (94 percent of them women) has become a successful mechanism for reducing poverty. Over time it has demonstrated its ability to operate with market resources and to re- duce its dependency on subsidized funds. Its loan recovery rates are consistently higher than 90 percent. 7. This section is based on Dar and Tzannatos (1999) and Dar and Gill (1998). 8. For a more detailed review, see Dar and Tzannatos (1999). 9. Social returns are based on a comparison of measurable economic costs and benefits that do not take into account either possible externalities associated with the reintegration of the long-term unemployed into the labor force or reduc- tions in high levels of unemployment in specific regions. 10. We do not have data on the failure rate of small businesses that do not take part in these programs. But anecdotal evidence suggests that these rates are usu- ally similar. While businesses receiving assistance, then, do no worse in terms of survival rates than businesses not receiving assistance, they do not appear to do any better. 11. Assessments in the region have identified problems with the quality and efficiency of public institutions. In Malaysia, for example, low salaries resulted in a shortage of qualified instructors. A World Bank evaluation of public training institutions in Indonesia found low levels of efficiency and capacity utilization (World Bank 1997b). 12. Another apprenticeship initiative in the Philippines allows firms to pay wages considerably below market levels to apprentices receiving on-the-job train- ing for up to six months. But the law that allows this exemption does not require firms to establish a specific training program. Thus a recent government review suggests that firms use the law as a means of exploiting workers (see chapter 5). 13. For discussions of these aspects of government roles, see ILO (1998) and Betcherman, McMullen, and Davidman (1998). References Anderson, P., and B. D. Meyer. 1994. "Unemployment Insurance Benefits and Take-up Rates." NBER Working Paper 4787. Cambridge, Mass.: National Bureau of Economic Research. 342 EAST ASIAN LABOR MARKETS Anderson, P., W. Corson, and P. Decker. 1991. "The New Jersey Unemployment Insurance Re-employment Demonstration Project Follow-up Report." Unem- ployment Insurance Occasional Paper 91 - 1, U.S. Department of Labor, Wash- ington, D.C. Balakrishna, R. 1998. "An International Review of Active Labor Market Policies and the Lessons for the GCC Countries." Discussion paper. London: London School of Economics. Betcherman, G., K. McMullen, and K. Davidman. 1998. Training for the New Economy. Ottawa: Canadian Policy Research Networks. Bloom, H. 1990. "Back to Work: Testing Re-employment Services for Displaced Workers." Working paper. Kalamazoo, Michigan: WE. Upjohn Institute for Employment Research. Corson, W, S. Long, and R. Maynard. 1985. "An Impact Evaluation of the Buf- falo Dislocated Worker Demonstration Program." Working Paper. Princeton: Mathematica Policy Research. Dar, A., and I. Gill. 1998. "Evaluating Retraining Programs in OECD Countries: Lessons Learned." World Bank Research Observer 13(1): 79-101. Dar, A., and Z. Tzannatos. 1999. "Active Labor Market Programs: A Review of the Evidence from Evaluations." Social Protection Discussion Paper Series 9901. Washington, D.C.: World Bank. Fay, R. 1996. "Enhancing the Effectiveness of Active Labour Market Policies: Evi- dence From Program Evaluations in OECD Countries." OECD Labor Mar- ket and Social Policy Occasional Paper 8. Paris. Fretwell, D., and S. Goldberg. 1994. "Developing Effective Employment Ser- vices" in "Equity, Efficiency and Adjustment in Labor Markets." Discussion Paper 208. Washington, D.C.: World Bank. Fretwell, D., J. Benus, and C. J. O'Leary. 1999. Evaluating the Impact of Active Labor Programs: Results of Cross Country Studies in Europe and CentralAsia. So- cial Protection Discussion Paper Series 9915. Washington, D.C.: World Bank. Friedlander, D., D. Greenberg, and P. Robins. 1997. "Evaluating Government Training Programs for the Economically Disadvantaged." Journal of Economic Literature 35(4). Gill, I., F. Fluitman, and A. Dar. 2000. Vocational Education and Training Reforms: Matching Skills to Markets and Budgets. New York: Oxford University Press. Goss, Gilroy, and Associates. 1989. "Evaluation of the Job Development Program: Final Report." Report prepared for the Program Evaluation Branch, Employ- ment and Immigration Office, Canada. Graves, E, and B. Gauthier. 1995. "Evaluation of the Self-Employment Assis- tance Program." HRDC Working Paper. Ottawa, Canada. Heckman, J. 1992. "Randomization and Social Policy Evaluation." In C. Manski and I. Garfinkel, eds., Evaluating Welfare and Training Programs. Cambridge, Mass.: Harvard University Press. ACTIVE LABOR MARKET POLICIES: ISSUES FOR EAST ASIA 343 ILO (International Labour Office). 1998. World Employment Report: Employabil- ity in the Global Economy: How Training Matters. Geneva. Keddeman, W 1998. "Of Nets and Assets: Effects and Impacts of Employment- Intensive Programmes-A Review of ILO Experience." Development Policies Department, International Labour Organization. Geneva: International Labour Office. Leigh, D. 1995. Assisting Workers Displaced by Structural Change: An International Perspective. Kalamazoo, Mich.: W E. Upjohn Institute for Employment Research. Mangan, J. 1988. "Wage Subsidies for the Disabled: A Discussion of Their Im- pact in Australia." International Journal of Manpower 11 (1). Meager, N., and C. Evans. 1998. "The Evaluation of Active Labor Market Mea- sures for the Long-Term Unemployed." Employment and Training Papers 16. Geneva: International Labor Office. New Zealand Department of Labor. 1995. "Evaluation of Job Action." New Zea- land Employment Service Report. Wellington, New Zealand. OECD (Organization for Economic Cooperation and Development). 1997. OECD Employment Outlook. Paris. . 1994. The OECD Jobs Study: Evidence and Explanations. Paris. . 1993. OECD Employment Outlook. Paris. . 1991. Evaluating Labor Market and Social Programs: The State ofa Com- plex Art. Paris. O'Leary, C. 1998a. "Evaluating the Effectiveness of Active Labor Programs in Hun- gary." Draft Working Paper. W. E. Upjohn Institute for Employment Research. _____. 1998b. Evaluating the Effectiveness of Active Labor Programs in Poland. Draft Working Paper. W. E. Upjohn Institute for Employment Research. ____. 1995. "An Impact Analysis of Employment Programs in Hungary." Re- search Working Paper 95-30. W E. Upjohn Institute for Employment Research. Puhani, P 1998. "Advantage Through Training? A Microeconometric Evaluation of the Employment Effects of Active Labor Market Programs in Poland." ZEW Discussion Paper 98-25. Warsaw, Poland. Ravallion, M. 1998. "Appraising Workfare Programs." Working Paper Series 1955. Washington, D.C.: The World Bank. Skedinger, P. 1995. "Employment Policies and Displacement in the Youth Labor Market." Swedish Economic Policy Review 2(1). Subbarao, K. 1997. "Public Works as an Anti-Poverty Program: An Overview of Cross-Country Experience." American Journal ofAgricultural Economics. Webster, E. 1998. "Microeconomic Evaluations of Australian Labor Market Pro- grams." Australian Economic Review 3(12): 189-201. Weiland, R. 1998. "SME Finance in Indonesia: An Assessment." Discussion paper prepared for the Trade and Industry Strategy project (BAPPENAS) and the Indonesian Ministry of Industry and Trade, Jakarta. 344 EAST ASIAN LABOR MARKETS Wilson, S., and A. V. Adams. 1994. "Self-Employment for the Unemployed: Ex- perience in OECD and Transitional Economies." Discussion Paper 263. Wash- ington, D.C.: World Bank. World Bank. 1997a. World Development Report: The State in a Changing World. Washington, D.C.: World Bank. . 1997b. Training and the Labor Market in Indonesia: Productivity Gains and Employment Growth. Washington, D.C. . 1998. "World Bank Support for Small and Medium Industries in the Phillipines: An Import Evaluation." Operations Evaluation Department Re- port 18041. Washington, D.C. 8 The Economics of Employment Protection and Unemployment Insurance Schemes: Policy Options for Indonesia, Malaysia, the Philippines, and Thailand ALEJANDRA Cox EDWARDS AND CHRIS MANNINGI THIS CHAPTER EXAMINES the current systems of income protection in place in Indonesia, Malaysia, the Philippines, and Thailand and looks at policy options for protecting workers in both the formal and informal sectors. While acknowledging the problems that have arisen be- cause of the Asian financial crisis, we take as our main focus future insti- tutions. We start from the assumption that employment stability, and more precisely income stability, is the preferred state for individuals. At the same time employment creation and destruction are normal and perma- nent features of market-oriented systems. It follows, then, that private employment contracts must deal with the finite nature of employment re- lations in some way. But it is also clear that legislation establishing rules for terminating employment or restricting fixed-term contracts affects contract negotiations between employers and workers. 345 346 EAST ASIAN LABOR MARKETS All economies offer a continuum of employment opportunities, from self-employment within the household to formal employment in regis- tered enterprises. Part of a household's strategy for allocating its labor time-that is, choosing activities from this continuum-involves dealing with the risks surrounding temporary and permanent declines in income. Some of the risks households face are crop failure, unemployment, ill health, and incapacity in old age. Managing these risks is often intimately tied up with decisions affecting labor supply. Households make decisions about the allocation of labor to different activities (working at home, in other towns, or abroad, for instance) based on whether the activities re- duce or increase potential risks.2 The expansion of formal labor and financial markets does not entirely replace traditional forms of dealing with risk, but these wider markets offer households alternative mechanisms for managing it. Knowing that workers are concerned with a variety of risks, firms interested in a stable and loyal labor force compete in the labor market using not only wages but also the overall labor contract, which includes important benefits such as health insurance. Similarly, well-developed financial and insurance mar- kets offer households the opportunity to earn a return on even small sav- ings and the option of spreading their risks more broadly, lowering the unit cost of insurance. Labor Market Policy in Developing Countries In developing countries informal and formal jobs together usually provide some form of employment for the entire labor force. A large number of rural workers are in the informal sector and may not be filly employed. Some work only part of the year, while others average fewer hours than they would like or find themselves performing low-productivity activities and earning meager wages. But the nature of production in agricultural economies is such that open unemployment is relatively rare. Poor households cannot afford to be without any income at all, and sharing low-productivity work is a common practice. The result is widespread underemployment. As developing countries seek to transform themselves from traditional to modern economies, open unemployment rises. The transition from widespread underemployment to open unemployment in these economies is in part an income effect. As countries grow and household incomes rise, jobless workers are able to endure periods without work while waiting for a job to open. Rising open unemployment is also in part a result of ur- banization and the shift to modern forms of production, since modern POLICY OPTIONS 347 economies organize work in ways that do not lend themselves as easily as agricultural arrangements to work sharing. Realistic policy interventions, then, must recognize the need for an approach that includes economic ac- tivities appropriate to both types of economies. Ideally such policies will lend themselves to a smooth transition between traditional and modern forms of production. One important policy issue is loss of income during the transformation. Two kinds of policy interventions address the problems associated with in- come loss. The first includes poverty-targeted emergency programs. Not necessarily focused on the unemployed (since unemployment is not the most important characteristic of the poor and may not be the most ap- propriate targeting mechanism), these programs are considered effective mechanisms for reaching households severely affected by income loss. The second set of policies, which we focus on here, includes all permanent labor market interventions that focus on dismissals and the unemployed. Labor market interventions can address the risk of income loss associ- ated with formal employment during the normal operations of a growing economy as well as in times of recession. Such labor policy interventions may help households deal with the risks associated with employment. But if labor policy overlooks the role of wages and working conditions as in- centives and market signals, it will end up closing the formal labor market to the influence of market forces and discouraging the formalization of labor contracts. Ideally policy interventions reduce the transaction costs associated with employer-employee negotiations and help workers manage job-to-job transitions. The central question is how employers and employees conduct themselves in the event of job termination and what to do in these situa- tions. Legislation can simply require that all employment contracts in- clude a clause on termination, potentially preventing disputes and reduc- ing transaction costs. Alternatively legislation can impose specific rules concerning job dismissals, establishing property rights and liabilities such as severance payments. Legislation can also create a program of transfers for the unemployed. But such transfers rely on some source of funding, usually a payroll tax, and are likely to alter some or all of the following: job creation, job search incentives, labor costs, and net wages. The existing evidence indicates that firing costs deter hiring, reduce labor demand, and hamper the economy's ability to deal with uncertainty and structural change (OECD 1995; Saint-Paul 1999). In general unem- ployment insurance lowers the incentives to look for work, but the rela- tive size of this effect is a function of the program's design and thus varies 348 EAST ASIAN LABOR MARKETS across countries. European unemployment insurance programs pay bene- fits according to certain preestablished rules and collect the same tax rate from all employers, generating large cross-subsidies from some taxpayers to program beneficiaries. In contrast the unemployment insurance pro- gram in the United States is financed by an experience-rated payroll tax (box 8.1). In this case the employers that dismiss the most workers pay the BOX 8.1 Experience Rating as a Means of Financing Unemployment Insurance Experience rating is unique to the U.S. unemployment insurance system. Em- ployers finance the system with payments that are essentially a state-level tax. Firms pay a percentage of the earnings of each employee, up to a relatively low maximum level that varies across states. (The maximum taxable wage was only $7,000 in most states in 1999.) In California, for example, new employers are as- signed a rate of 3.4 percent. After an employer has made payments for three years, the company's rate is adjusted according to its "reserve ratio"-the sum of all con- tributions minus the benefits paid to former employees, divided by the total base payroll. Based on the 1999 schedule of tax rates, an employer could see the tax rate fall as low as 0.7 percent or rise as high as 5.4 percent. The rate continues to be adjusted yearly as new information on the employer accumulates. Employers receive a yearly summary of their account and can contest items that are incon- sistent with their records. The experience-rating system can make the marginal cost of firing more than zero and is intended to force firms to internalize the cost of unemployment ben- efits for its employees. However, because the tax rates have both lower and upper limits, many firms are in effect not experience rated-that is, an additional layoff or week of unemployment by a former employee has no effect on the firm's un- employment insurance tax bill. Feldstein (1976) was the first to note the fre- quency with which laid-off workers were recalled, citing it as evidence that many workers do not search for alternative employment while receiving unemployment insurance but simply wait for their previous employer to recall them. Topel (1983) argued that most of the impact unemployment insurance systems have on layoffs is caused by the fact that current methods of experience rating are incom- plete. In other words, in spite of experience rating, in a large number of cases the marginal cost of layoffs for the firm is zero or negligible. Unemployment insur- ance systems can create an incentive for employers to rely on unemployment ben- efits to adjust cyclical labor costs. POLICY OPTIONS 349 highest tax rate, especially if these workers collect from the unemployment insurance system. When looking at the impact of job security legislation or unemploy- ment insurance programs, we must keep in mind the possible interaction between the two types of programs. European unemployment insurance programs coexist with job security legislation. Such legislation tends to re- duce the number of dismissals, but once employees are terminated they have a difficult time finding new jobs because of the "slowing" effect of job security on job creation. In the United States employment at will pre- vails, making the event of dismissal more likely but easing the process of reemployment. The employment-at-will doctrine allows employers to dis- miss workers for a just reason, an economic reason, or no reason at all, as long as the dismissal does not violate federal antidiscrimination legislation (such as the Civil Rights Act of 1964) or the National Labor Relations Act. Not surprisingly one of the key differences between the labor markets in Europe and the United States is that in Europe workers experience much longer periods of unemployment (Blanchard and Portugal 1998). Informal Employment Protection Arrangements The four countries discussed here differ not only in per capita incomes but also in the organizational features that characterize their economies (table 8.1). According to the World Bank's classification system, Indonesia is a low-income country, the Philippines and Thailand are lower-middle- income economies, and Malaysia is in the upper-middle-income category (World Bank 1999). At the beginning of the 21st century, these countries are undergoing a rapid transformation. The share of overall employment in agriculture fell on the order of 10 points between 1985 and 1997 (table 8.2). Traditional forms of production coexist with modern methods in economies that, in spite of vast regional differences, have well-integrated labor markets. Inter- nal migration and the ongoing reallocation of labor from the least produc- tive to the most productive activities have contributed to more than two decades of sustained growth and an impressive trend of real wage growth. These achievements are particularly striking compared with the situation in developing countries outside the East Asia region (World Bank 1995). The labor markets of these countries have three distinct features. The first is their size. The large numbers of active workers, especially in In- donesia, require decentralized programs in order to make administering labor market regulations a manageable task. Second, unemployment rates 350 EAST ASIAN LABOR MARKETS TABLE 8.1 Basic Indicators, 1998 Population Per capita Per capita Population density GNP GNP Country (millions) (per kin2) (US$) (PPP) Indonesia 204 112 680 2,790 Philippines 75 252 1,050 6,740 Thailand 61 120 2,200 5,840 Malaysia 22 68 3,600 6,990 World 5.9 45 4,890 6,200 Low income 3.5 85 520 2,130 Lower middle 908 25 1,710 4,080 Upper middle 588 27 4,860 7,830 Note: PPP = purchasing power parity. Source: World Bank (1999). have been modest, suggesting that focusing income security policy on the unemployed may not be appropriate (table 8.3). Third, with the excep- tion of Malaysia more than half the workers in these countries are self- employed. For this reason programs focused on wage employees miss a sig- nificant proportion of the labor force. Wiebe (1996) argues that income insecurity in Indonesia stems from a combination of low earnings and seasonal unemployment. Households have two types of protection against the possible consequences of these fluctuations-private strategies and government welfare programs. Private transfers flow among rural households and from urban to rural areas. Cox and Jimenez (1990) report that these types of transfers are also important in Malaysia and the Philippines (table 8.4). At the community level, In- donesian neighborhood associations (Rukun Tetangga) collect an informal tax to provide welfare assistance in emergencies. Employers sometimes provide a third form of private insurance, even in the informal sector. Studies of labor markets in developing countries find that implicit contracts capture a number of social and economic forces (Hopenhayn and Nicolini 1997). Employers typically put in place systems of incentives that link compensation to productivity and reduce monitoring costs. Employers also make widespread use of fringe benefits in an attempt to make jobs more attractive and to secure workers' loyalty. These benefits include meals, transportation, some form of medical cov- POLICY OPTIONS 35I TABLE 8.2 Distribution of Employment by Major Sector, 1985-97 (percent) Malaysia a Thailandb Philippinesc Indonesia d Agriculture 1985 30.5 63.5 49.6 54.7 1990 26.0 64.0 45.2 56.0 1997 17.3 50.3 40.4 41.2 Industry 1985 23.9 13.0 13.8 13.4 1990 27.5 14.0 15.0 13.8 1997 33.7 19.7 16.7 19.0 Manufacturing 1985 15.1 9.4 9.5 9.3 1990 19.9 10.2 9.7 10.2 1997 23.4 12.9 9.9 12.9 Services 1985 45.6 23.5 36.5 31.8 1990 46.5 22.0 39.7 30.3 1997 49.0 29.9 42.9 39.8 Note: Data are based on annual or quarterly labor force surveys (August round in Thai- land, October/fourth round in the Philippines). Data for 1985 for Indonesia are taken from the National Intercensal Survey. a. Ages 15-64 years. b. Ages 11 and above (1985); ages 13 and above (1990 and 1997), c. Ages 15 and above. d. Ages 10 years and above. Source: ILO (1998). erage, and some form of income smoothing or employment continuity. In agriculture, for example, wages often vary by task. Tasks are specific to seasons but do not directly reflect the value of marginal productivity in each season. Instead the most productive workers tend to be employed more steadily than less productive workers, who are likely to be hired dur- ing peak seasons only (Dreze and Mukherjee 1997; Hart 1984; Otsuka, Chuma, and Hayami 1992). A relatively large number of wage workers are on fixed-term contracts and move easily among jobs in search of the best working conditions. In TABLE 8.3 Basic Labor Market Indicators Labor Distribution of labor force by sector (percent) force Unemployment (millions) rate (percent) Agriculture Industry Services 1990- Other Other Other Country 1997 1996 1997 1998a Wage wage Wage wage Wage wage Indonesia 91.3 2.5 4.7 5.5 6.5 43.9 10.0 5.8 14.3 19.5 Philippines 30.3 8.1 7.9 - 9.6 35.7 12.3 3.7 23.6 15.1 Thailand 33.3 2.2 4.6 6.6 59.7 8.7 3.2 11.6 10.2 Malaysia 8.6 2.5 6.4 8.8 21.8 19.0 3.6 34.0 12.7 Note: Labor force statistics are based on quarterly or annual labor force surveys and cover those ages 15 and above, except for Indonesia (ages 10 and above), Thailand 1997 (ages 13 and above), and Malaysia (ages 15-64). a. Indonesia, August 1998; Thailand, February 1998; Malaysia, projected (Economic Planning Unit). Source: World Bank (1995); Economic Planning Unit, National Economic Recovery Plan 1998; Department of Statistics, Labor Force Survey 1996; NSO (1998); various press reports, January-March 1999. TABLE 8.4 Private Transfers among Households Percentage of households Average transfer as a receiving or sending percentage of average transfers income Per capita Receiving Sending Receiving Sending Country Year GNPa households households households households Indonesia Java) 1982 490 Rural 31 72 10 8 Urban 44 45 20 3 Malaysia 1977-78 1,830 19-30b 33-47' 11' Philippinesd 1978 560 47 - 9 a. In 1986 dollars. b. Average not available. Figures denote upper and lower bounds. c. The average transfer was 46 percent of income for households in the lowest income quintile. d. Cash gifts in a large informal housing area. Source: World Bank (1995). 354 EAST ASIAN LABOR MARKETS the Indonesian garment districts worker mobility is very high, reflecting instability in the placement of sales orders and the continuous entry and exit of workers to and from the labor market. This mobility can be inter- preted as a form of negotiation, where workers who find conditions unfavorable simply leave the job rather than voice a complaint. It also reflects the uncertain nature of production contracts in that sector of eco- nomic activity-an uncertainty that discourages employers from engaging workers on long-term contracts. When hiring costs become relatively high, employers are willing to pay for workers' "waiting time" in an effort to avoid high rotation rates. In fact in a very tight labor market employers must offer incentives to retain workers. For example in 1995 a growing number of Indonesian construc- tion workers in Medan were reportedly being hired on weekly contracts. These contracts set a daily wage and a bonus for workers who remained at the work site continuously for the week (Edwards 1996). Should the legal system enforce such implicit understandings? The key to contract enforcement is the expected penalty associated with a breach of contract, and the existence of penalties does not necessarily depend on a formal legal system. In traditional communities, for example, a person's reputation may be a more binding guarantee. In either a formal or an in- formal system, however, local authorities who are genuinely respected by the population are in the best position to deal with disputes over failures to comply with such contracts. A formal legal system based on written agreements is unlikely to be able to deal appropriately with informal con- tracts. But in order for the legal system to take part in the conflict resolu- tion process, contracts that are acceptable to employers and employees must also be recognized as legal agreements. Policymakers must pay at- tention to the reality of the traditional or informal sector and encourage the coexistence of traditional and modern forms of conflict resolution. Formal Employment Protection Arrangements In most countries existing legislation establishes rules and procedures governing employment termination. This legislation is particularly impor- tant in the four countries under review in light of the fact that none of them has an unemployment insurance system. (A few self-financed private schemes do exist, but they cover only a small proportion of highly paid employees.3) Provident funds (described in box 8.2) and severance pay are the two forms of income support available to workers dismissed from formal-sector jobs. Severance pay is the centerpiece of employment pro- tection legislation. POLICY OPTIONS 355 BOX 8.2 Provident Funds and Worker Security Along with severance pay, provident funds are available to workers in the for- mal sector in Indonesia, Malaysia, the Philippines, and Thailand. These funds provide pensions, disability, health, and severance payments based on years of ser- vice with a firm. Employers and employees contribute to the funds in relatively equal amounts. Workers generally withdraw the money as a lump sum at retire- ment (age 55). Total contributions in Thailand, which introduced the funds in 1990, amount to 4.5 percent of wages. In Indonesia, where the funds were put in place since 1992, contributions equal 5.7 percent of wages. Malaysia's funds began operating much earlier, and contributions amounted to a much higher 22 percent of monthly wages following reforms in 1993. Coverage extends to approximately 12-16 percent of the employed population in Indonesia and Thailand, 24 percent in the Philippines, and a much higher 48 percent in Malaysia (Lee 1998). Contributions are mandatory for all firms with 10 or more employees in Thailand and Indonesia, but in practice those firms en- rolled in the scheme tend to be large establishments (100 employees or more). In the mid-1990s in Indonesia, the average enterprise in the scheme had 165 em- ployees, although the average manufacturing enterprise had only 50. At the same time employer-financed contributions amounted to 0.52 percent of wages for ac- cident insurance and 36 percent (individual and family coverage, respectively) for health insurance. The Malaysian Employment Act of 1955 established a minimum level of severance pay. Workers with 1-2 years of tenure would get 10 days of pay per year of service. Workers with 2-5 years of tenure are entitled to 15 days and workers with more than 5 years of tenure to 20 days. The In- donesian law (12/1964) and its implementing regulations control termi- nation of employment. After discussing termination with the employee or the union, the employer must seek permission to dismiss the employee from a regional committee set up to resolve labor disputes, even if the em- ployee consents to the dismissal. Severance pay is set at a minimum level of one month's pay for less than one year of service, two months' pay for one to two years of service, three months' pay for two to three years of ser- vice, and four months' pay for more than three years of service. The same regulation sets merit allowances, which are added to severance pay, in ef- fect doubling it. 356 EAST ASIAN LABOR MARKETS Severance pay in Indonesia does not apply to fixed-term contracts. Fixed-term contracts make the contracting parties more aware of the finite nature of employment relationships and are a desirable aspect of the sys- tem. In countries where employers do not have the legal option to estab- lish fixed-term contracts, job security legislation becomes binding on all employment contracts. For example in Latin America (with the notable exception of Mexico) the law restricts the use of temporary contracts. For- mal contracts become indefinite after a probationary period, which can be as short as three months. Once contracts become indefinite under the law, they are subject to a minimum severance in case of dismissals. Similar con- ditions used to apply in many European countries, but during the 1990s fixed-term contracts became more available. Job security is traditionally an important component of labor legisla- tion in Africa, the Middle East, Latin America, and South Asia. In general laws in these regions establish limits on employers' freedom to hire work- ers on fixed-term contracts and to dismiss employees. These types of re- strictions are also common in European countries, but not in the United States, as we have seen. However in the United States the experience-rated payroll tax that finances the unemployment insurance program is de- signed to reflect the costs employers impose on the program (see box 8.1). Protection against dismissal (or termination) may take any of several forms: mandatory advance notice, mandatory severance pay, or the option to question what appears to be wrongful termination. In Indonesia the mandatory formal authorization is intended to prevent wrongful termina- tions. In Latin America the law establishes a set of legal causes for dis- missal, and workers have the right to appeal any firing. Employers unable to provide proof of due cause may be required to reemploy workers or to pay increased severance monies-either a predetermined sum or an amount set by the judge. The relative weights of severance payment rules and wrongful termina- tion hurdles are important features of employment protection systems. The reality of market forces tells us that job terminations are necessary. Legally mandated severance pay may coexist with employment at will and may well serve the function of unemployment insurance systems in devel- oping countries that lack them. But the emphasis on wrongful termina- tion distorts the functioning of the labor market. It delays access to sever- ance payments and makes voluntary job separations a rare occurrence, because workers who leave voluntarily do not receive it. Arguably these rules do exactly the opposite of what is intended, creating conflicts and in- creasing transaction costs. POLICY OPTIONS 357 Employment Protection and Severance Pay The simplified model proposed by Lazear (1990) considers a government- ordered severance payment equal to Q (a fixed amount). Assuming that newly hired workers are willing and able to make an up-front payment equal to Q, the rule will have no efficiency effect. Lazear concludes that in the absence of constraints on borrowing or lending, any state-mandated severance pay can be seen as a delayed payment scheme, with no efficiency effect. Lazear's conclusions extend to severance payments that are based on a defined-contribution plan. In this case, just as in the case of a fixed amount Q, we can think of the severance payment as a payoff based on the accumulated value of periodic employee contributions. Countries such as Brazil have a system that requires employers to contribute a frac- tion of workers' monthly salary to a fund that covers severance pay (Fondo de garantia do tempo de servicio). The state guarantees this fund, which is indexed and earns a minimum yield. It is portable, and workers may with- draw funds from it in case of involuntary separations. The nominal inci- dence of the contribution (employee or employer) does not determine its economic impact. Three sets of issues arise here. First, severance schemes are generally defined-benefit rather than defined-contribution plans-that is, they pay a multiple of the employee's salary at the time of separation. Second, leg- islation governing job security not only defines a minimum level of sever- ance pay but also makes the payments a function of two factors: the party that initiates the separation and the cause. Third, to say that a defined- contribution benefit has no efficiency effect is an oversimplification. As Lazear recognizes, lack of liquidity (imperfections in the capital market) or lack of confidence in the system's viability (the absence of guarantees on defined-contribution funds) may hurt workers required to pay the amount Q If capital markets are limited, poor workers do not have the option of making an advance payment for the minimum amount and are blocked from accepting permanent contracts. Furthermore, if there are no credible guarantees on the defined-contribution funds, the mandated severance payments have a nontrivial tax component and labor costs increase rela- tive to the initial equilibrium, inducing inefficiency. WHY IS THE BENEFITS FORMULA IMPORTANT? Mandated severance pay defined as a function of workers' final salary and number of years on the job has two effects. First, it causes the employer's liability to vary with changes in the prevailing wage. Second, it means that employers must 358 EAST ASIAN LABOR MARKETS make relatively high payments to employees with many years of service. A useful way to look at the effects of this type of formula on labor contracts is to consider a situation in which a contract worker pays a marginal con- tribution to a severance fund. We assume that all workers receive severance pay at the end of their contract and that they are willing to accept a re- duction in their current cash wages in return for a future payoff Given this assumption workers receive compensation composed of a cash wage (W ) and a contribution to a severance fund (a). The other variables are T, the worker's years on the job; L, the employer's liability for severance pay when the worker is dismissed (equal to a factor oc times the formal wage FW, times the years of tenure T); and A, the present value of the ac- cumulated contributions to the worker's severance pay fund. In one portion of the labor market, the cash economy, severance pay- ments are not required. In Indonesia this part of the labor market includes all workers on fixed-term contracts. The presence of this alternative sector affects competitiveness in the formal sector-that is, total compensation in the formal sector must be at least equal to the cash wage, since workers can easily move to the cash economy. Starting from a situation where L ex- actly equals A, we posit an increase in the level of wages in the cash econ- omy. Market competition then forces an increase in compensation in the formal sector. From the point of view of both employers and workers, the compensation increase has two components: an increase in the wage in the current period d(FWt), plus an increase in the liability dL, a multiple of d(FWt). Assuming that L = o(FWt)* Tand that initially A = ot(FWt)* T, then: d L = d(FWt)* TPa. But the marginal condition for labor employment is: d VALUE MARGINAL PRODUCT = dA MARGINAL C0ST, or: d VALUE MARGINAL PRODUCT = d (FW,) + dA. A financially sound policy that allows a firm to honor its liabilities when employees are terminated requires that dL = dA. For this relation- ship to be true, however, the following must also hold: d VALUE MARGINAL PRODUCT = d(FW,) + d(FW) *T*a = d(FW,)(1 + T*a). POLICY OPTIONS 359 This relationship implies that the formal wage adjustment required to maintain competitiveness is a function of a worker's years of experience and the factor a. If the firm's employees have relatively few years of ser- vice, the wage adjustment will be close to the change in cash wages. But if the firm's employees have more experience, the wage adjustment will be much smaller. Since most firms are likely to have a combination of newly hired and experienced workers, this financially sound policy poses a diffi- cult management challenge. How does the formal sector react to a reduction in real wages in the cash economy? Employers in the formal sector may choose to reduce cur- rent wages, simultaneously reducing the amount of severance they will have to pay the most experienced workers. Assuming that all workers are dismissed at some point and receive the severance payment, a current wage reduction represents a much larger reduction in permanent income. The more years an employee has spent on the job, the larger the reduction in permanent income will be relative to the reduction in current wages. Workers will oppose the wage reduction, but they will have a very weak case. First, the threat of quitting is not very credible, since it entails giving up the right to severance. Second, alternative sources of employment will force them to accept lower market wages. In short a wage reduction in the cash economy poses an enormous threat to formal sector workers with many years of experience. Employers can try to reduce the cost of severance payments by main- taining a very young work force with a high turnover rate or simply by try- ing never to dismiss workers, thus avoiding the issue altogether. These choices have further efficiency effects. In the first case the type of workers the firm employs affects its activities.4 In the second labor is treated as a fixed factor. Hiring decisions are delayed and unfavorable market condi- tions make severance payments realized losses for employers and a bonus for employees. Since employers prefer to avoid losses, this system of pro- tection against dismissals is likely to offer long-term employees a high de- gree of job security. Employers are also likely to be reluctant to offer wage incentives for loyalty and performance. Instead, they are likely to offer bonuses or other one-time awards that do not permanently affect wages in order to keep severance payments as low as possible. Pages and Montene- gro (1998) found that relaxing job security regulations in Chile increased hiring and encouraged firms to move away from the traditional policy of firing younger workers. Stricter employment security legislation, however, reduced the rate of labor market participation among the young. Severance payments based on a multiple of final wages conspire against good labor relationships, the use of long-term contracts, and competitive- 360 EAST ASIAN LABOR MARKETS ness.5 In particular, changes in market conditions that result in variations in current wages have an asymmetrical effect on firms and workers, de- pending on whether real wages increase or decline. Wage increases have a magnified effect on firms' costs as a function of workers' tenure, and wage reductions have a magnified effect on workers' permanent income. These problems can be avoided if severance payments required by law are de- fined in terms of explicit contributions to a fund-that is, if severance payments become a forced savings program. The exact amount individual workers receive at termination (voluntary and involuntary) equals the amount of funds in the account. SEVERANCE PAY AND WRONGFUL TERMINATION. Employers generally do not have to make severance payments when employees are dismissed for due cause. Due causes for dismissal are spelled out in labor codes and typ- ically include mutual agreement, voluntary retirement with minimum no- tice, death of the employee, termination of fixed-period contracts, and behavioral problems such as bad conduct, intoxication, and excessive ab- sences. Dismissals for economic reasons (financial problems of the enter- prise) are not always included in this list. Argentina and Chile are among the exceptions to this rule. The laws of both countries treat dismissals for economic reasons (redundancies), requiring employers to pay severance based on years with the firm, but subject to a cap. Such dismissals entitle workers to 50 percent of the allowable severance pay for unjust dismissals in Argentina and to severance pay equaling one month per year of service (to a maximum of 11 months) in Chile. How different is job security legislation in industrial countries? Grubb and Wells (1994) ranked the countries of the European Union according to the strictness of regulations protecting regular and fixed-term workers against dismissal. According to their assessment employment protection is relatively low in Denmark, Ireland, and the United Kingdom and high in Italy, Portugal, and Spain (see also OECD [1994]). The study shows that Finland, Norway, and Sweden also restrict fixed-term contracts and that the latter two restrict temporary work agencies significantly. The United States has no federal policy mandating severance payments. Employers that pay severance do so in response to collective bargaining agreements, individual contracts, or market incentives. Survey data indi- cate that of those companies offering severance, most provide payments to workers who are dismissed permanently for economic reasons. The most common payment schemes use functions of tenure to determine the ben- efit level, typically one week's salary per year of service. Schedules based on age and length of service are also common.6 POLICY OPTIONS 36I Despite the prevalence of the employment-at-will doctrine throughout the United States, in the 1980s various state court rulings weakened em- ployers' right to dismiss workers at will. By 1987, when Montana passed landmark legislation requiring firms to have "just reason" to fire a worker, nine other states had introduced similar legislation. These statutes include economic cause as just cause for dismissal. Kruger (1991) argues that such legislation has been introduced to limit employer liability, expedite dispute settlements, reduce legal costs, and clarify property rights. But a study by Dertouzos and Karoly (1992) provides evidence that making employers li- able for wrongful termination creates substantial costs beyond those di- rectly attributable to lawsuits. When such legislation in is in place, they argue, firms hire fewer workers, causing a decline in aggregate employ- ment of 2-5 percent. They also found that the decline in employment is greater in large firms, suggesting that costs vary on a per-employee basis. Why do some countries adopt wrongful termination doctrines? Der- touzos and Karoly look at supply and demand as they relate to the will- ingness of state governments in the United States to adopt such legisla- tion. They argue that declines in union membership increase the number of workers without employment protection under collective bargaining agreements, increasing the demand for common-law remedies to wrong- ful discharge. When the unemployment rate rises, the number of poten- tial litigants increases, leading courts to look for ways to lighten their workload. An increase in the lawyer-to-population ratio, as a proxy for liti- giousness, also increases the demand for wrongful termination legislation and coincidentally the likelihood that states will recognize a new doctrine. In the United States an increased awareness of workers' rights, particu- larly the advent of civil rights legislation at the federal and state levels, may have contributed to the general belief that workers can use the justice sys- tem to pursue claims of unfair treatment from their employers. Ideologi- cal factors may also explain the willingness of state judiciaries to provide increased job protection. All of the reasons that Dertouzos and Karoly offer could apply to developing countries, except for the fact that in most of these countries state or provincial governments do not have jurisdiction over labor legislation. Many labor laws in developing countries date to the 1950s and 1960s and the technical assistance provided by the Interna- tional Labour Organization (ILO) in response to individual governments' attempts to improve labor practices. Extreme job security was imposed in Zimbabwe and in India in 1980. Fallon and Lucas (1991) determine (after controlling for other factors) that these laws significantly reduced the demand for labor. The weighted- average estimate of the impact of this legislation on jobs, given output, is 362 EAST ASIAN LABOR MARKETS on the order of 20 percent. Moreover the data for India show an interest- ing pattern across industries. Those industries with the most public enter- prises were the least affected by the legislative change. The authors inter- pret this finding as a sign that job security was already high within the public sector. But the authors find no evidence that the legislative changes affected the speed with which employment adjusted to changing market conditions. Lazear (1990) found that mandatory severance payments reduced em- ployment across countries of the Organization for Economic Cooperation and Development (OECD), although the results were not robust. Alba- Ramirez (1991) reported that the introduction of fixed-term contracts in Spain in 1980 spurred employment growth. Marshall (1991) noted that temporary and part-time employment became more common in Lima than in Buenos Aires when Peru began encouraging temporary contracts in an effort to reduce unemployment (Argentina did not follow such a policy). Improving the Design of State-Mandated Severance Schemes Experience shows that different types of severance and pension schemes are successful when they are largely self-financed, when contributions and benefits correspond to each other, and when they are designed to mini- mize incentives for workers to leave the labor force or shift to informal or casual contracts. Severance payments defined as multiples of prevailing salaries at the moment of separation (as in East Asia, most of Latin Amer- ica, and Europe) are not defined-contribution programs. As we have seen workers who have been on the same job for different lengths of time need to be compensated differently. This fact creates personnel management conflicts and limits employers in their efforts to improve working condi- tions and induce cooperation. Given the long tradition of severance pay- ments in many countries, improving on the existing systems makes more political sense than adding an unemployment insurance plan. Alterna- tively a system can be designed that requires all workers (through their employers) to establish a "job termination account" that is accessible at the time of dismissal or voluntary separation. In some countries severance pay is entirely the responsibility of em- ployers. The advantage of this design is that it avoids the problems posed by employers unable (or unwilling) to pay severance benefits. If severance payments are financed through employer-paid premiums, firms need not be reinsured or granted public guarantees in the event of insolvency. In Bolivia, for example, workers have access to the same severance payments POLICY OPTIONS 363 regardless of the conditions surrounding their separation (which may be voluntary or involuntary) after five years on the job. Some countries, in- cluding Brazil, Colombia, and Peru, replaced the traditional mode of sev- erance pay (a specified amount, such as a month's pay, for each year of ser- vice) for a Time of Service Fund-the Fondo de Garantia do Tempo de Servicio (FGTS) in Brazil and Compensacion por Tiempo de Servicio (CTS) in Peru. Employers pay a percentage of workers' salaries (8 percent in Brazil, 8.33 percent in Peru) into the fund, which is available in case of justified layoffs or voluntary separation. Chilean law allows workers to choose this type of arrangement instead of the traditional severance pay- ment scheme after seven years on the job. Employers are required to de- posit the equivalent of one-twelfth of the worker's annual salary in a sav- ings account (Edwards 1993). This type of system is likely to operate smoothly in an economy that has introduced a social security system based on individual accounts, al- lowing the same institutions that manage pension funds to manage the funds. Such systems can also be indexed, so that all deposits are shielded from the inflation tax. But contribution rates, like those in Brazil and Chile, can be too high relative to what is needed for the program. Feld- stein and Altman (1998) provide a full calculation of the contributions necessary to finance a system with the level of benefits that the current sys- tem in the United States offers. They argue that 4 percent of salaries, up to a maximum, is sufficient. In the cases described above, employees will receive severance payments some time in the future. In the meantime employers treat the amounts they deposit as an expense for tax purposes. At the time of separation workers receive severance payments equal to the amount that has accu- mulated in the fund (in real terms plus interest). This type of system has many advantages over the traditional method. First, workers receive sever- ance pay equal to what has accumulated in their funds in all cases of ter- mination-whether they quit, are fired, or are laid off. Second, the funds are portable, so workers do not lose the accumulated amounts when they move to a new job. Third, these funds help firms avoid the liquidity prob- lems that making severance payments can create. Fourth, the system is very transparent. Employers see it as part of the current cost of labor. For employees it is a deferred payment, since the funds are portable and pay- ment is not conditional on the circumstances of termination. The specifics of the benefits program may vary. For example, benefits can be a multiple of individual salaries. The benefit period and replace- ment ratio can also be defined in advance. Since workers will use employ- 364 EAST ASIAN LABOR MARKETS ment transition funds to look for new jobs, and the typical job search lasts six months or less, the mandated contributions can be subject to a maxi- mum relative to the income level of the individual worker. Thus, as earn- ings increase, the maximum contribution employers must pay into the fund will increase as well. The fact that employers cannot use the funds as working capital is the most important aspect of this system. But this change is the price of mov- ing to a system that guarantees employees some security if they find them- selves unemployed. The alternative is a system of compulsory insurance, which is administratively complex and in any case is unlikely to be suffi- cient during an event such as the 1997 East Asian crisis. In such cases the government is left to make up the deficit (in the Republic of Korea the au- thorities created a special fund for this purpose in 1997). Since those laid off from the formal sector in a crisis are likely to be better off than those outside the sector, governments are likely to want to avoid relying on a public guarantee. Unemployment Insurance Systems Is an unemployment insurance system financed by both employees and employers a superior method of providing income security to terminated employees? What are the advantages and disadvantages of the U.S. system, which is financed by taxes on employers? What other alternatives (such as subsidized savings plans and welfare-for-work schemes) should be consid- ered in order to improve the income security of employees in the formal sector? Unemployment insurance programs are intended to protect workers from income losses associated with involuntary and unanticipated inter- ruptions of employment. These programs are managed by the public sec- tor, because two characteristics of unemployment risk discourage private insurance. First, unemployment is associated with business cycles. As a re- sult some types of workers regularly experience unemployment at certain periods in the business cycle. Second, unemployment insurance systems require significant institutional capacity, much like income tax systems, in order to operate properly. In fact taxpayer numbers (such as Social Secu- rity numbers in the United States) can be used to identify employers and employees and to maintain records of contributions. The Problem of Moral Hazard The frequency and duration of periods of unemployment are to a degree the result of employment decisions made by both firms and workers. The POLICY OPTIONS 365 existence of unemployment insurance makes retrenchment more attrac- tive to employers. To the extent that employment contracts do not per- fectly internalize the costs of unemployment insurance, the incidence of unemployment increases. Becker (1972) points out that the method of fi- nancing an unemployment insurance system is of key importance, because it affects firms' decisions to keep or lay off employees. For workers unemployment insurance makes longer spells of jobless- ness more attractive, creating problems of moral hazard. Unemployment insurance affects search strategies by raising reservation wages and there- fore increasing the average period of unemployment (McCall 1970; Mortensen 1970). Holen (1977) and Ehrenberg and Oaxaca (1971) find that relatively liberal unemployment insurance payments result in ex- tended periods of unemployment for those eligible for compensation. To eliminate the problem of moral hazard, unemployment insurance plans must be designed so that the insured share some of the costs. A self- financing program of social insurance, for instance, can offer partial in- surance and benefits that decline over the period of unemployment. Coverage and Funding Self-financed unemployment insurance systems pay benefits only to work- ers who have made contributions to the system either directly or indirectly (through their employers). Such systems generally exclude informal-sector workers. In the United States employers make mandatory payments that are essentially a tax (see box 8.1). Because employers fund the system, eli- gibility for benefits depends on a worker's having a substantial attachment to the labor force. One of the methods used to measure this attachment is a minimum earnings test that requires a certain level of earnings in a 12-month period.7 Obviously, informal-sector earnings do not help a worker meet the minimum earnings test, since employers are not making contributions on those earnings. All self-financing unemployment insurance systems must be experience rated in an aggregate sense. Tax rates must be high enough so that inflows equal outflows over a period of time long enough to meet demand during recessions as well as ordinary business cycles. Most European countries use a flat tax rate to finance unemployment insurance programs and other policy instruments to discourage firings. But average unemployment rates are very different across industries, so with this system firms with relatively few terminations subsidize the benefit payments of firms with more lay- offs and firings. In the United States firing practices are considered liberal compared with other countries, and experience rating is intended to force employers to subsidize the employees they terminate. 366 EAST ASIAN LABOR MARKETS What are the effects of the unemployment insurance payroll tax? The U.S. state of Washington provides a natural experiment that tells us much about the impact of methods of funding unemployment insurance sys- tems. From 1972 to 1984 unemployment insurance taxes were not expe- rience rated in Washington state. Instead all firms paid the same payroll tax (either 3 or 3.3 percent, depending on the year). Changes in federal legislation prompted the state to introduce experience rating in 1985. Be- tween 1984 and 1985 some firms saw their tax rate increase from 3.0 to 5.4 percent, while others saw it fall from 3.0 to 2.5 percent. Anderson and Meyer (1998) compared wages and labor market outcomes before and after the change and were able to draw conclusions regarding the impact of experience rating. Overall the results suggest that experience rating re- duces claims for unemployment insurance and stabilizes employment. But the experiment sheds light on the behavioral changes experience rating in- duces. In particular Anderson and Meyer find a significant decrease in the level and seasonality of claims and an increase in the number of benefit de- nials because firms have an incentive to "police the system" and contest in- valid claims. Anderson and Meyer are also able to tackle the question of who bears the cost of the unemployment insurance tax. They separate the payroll tax into two parts: a base tax that applies to a "market" or industry, and the remainder, which they descibe as "firm based." Their hypothesis is that the market part of the tax is internalized in the wage determination process and reappears in the form of lower market wages. But the firm absorbs the firm-based portion of the tax in the form of higher costs. Although the au- thors are limited to imprecise estimates, they present two emerging pat- terns from the data that are consistent with the theory. Firms seem to be able to pass on a substantial part of the market part of the payroll tax but none of the firm-specific portion. Perfect experience rating means that each firm pays (over time) a tax rate proportional to the differential between its rate of insured layoffs and the industry average. The firm-based part of the premium is thus analo- gous to the extra costs incurred by a firm that is required to make sever- ance payments and that lays off more employees than the average, raising labor costs. The main difference, of course, is that while unemployment insurance entitlements usually increase with duration of employment, they are rarely proportional. Entitlements are generally greater than the value of contributions for those with a short employment record and less than accumulated contributions for long-term employees. These systems tend to support those with short-term employment tenure at the cost of POLICY OPTIONS 367 premiums paid on behalf of long-term employees. Employers and em- ployees are therefore encouraged to make their contracts shorter than they would in the absence of unemployment insurance. This bias is systematic in such unemployment insurance arrangements. Experience rating will re- duce the opportunistic use of the system to subsidize short-term turnover but will not eliminate the general bias toward shorter contracts. Unemployment insurance systems therefore have the same bias as sev- erance payment systems based on workers' final salaries. Both systems create an incentive for employers to move to short-term contracts. The distortions such systems impose lead to administrative controls, such as denying eligibility for voluntary separations and requiring those receiving benefits to actively look for work, so that benefits are not used to subsi- dize leisure or other pursuits. The same reasoning that argues for a more "neutral" system of severance payments has led to recent proposals for in- come security systems that eliminate these short-term biases. Alternatives to Unemployment Insurance Is an improved severance payment system a viable alternative to unem- ployment insurance? Taking into consideration conditions in developing countries, this option may be the most reasonable approach to providing income insecurity. As we have seen unemployment insurance systems re- quire an administrative capacity that is often lacking in these economies. Existing severance pay systems can be the basis for more functional arrangements, such as the defined-contribution system, that help workers who are between jobs and also make formal contracts more transparent. Employers must recognize that a formal contract involves obligations that employees value, such as severance pay (or, in industrial countries, unem- ployment insurance). A system of severance payments based on accumu- lated contributions to individual funds is equivalent to an unemployment insurance system in which benefits are proportional to contributions. However, unemployment insurance systems do not pay benefits in cases of voluntary separation, while defined-contribution systems do. UNEMPLOYMENT INSURANCE SAVINGS ACCOUNTS. The key feature of these accounts is that they are defined-contribution funds (Feldstein and Altman 1998; Cortazar 1994). Each individual, or the individual's em- ployer, is required to contribute 4 percent of the employee's wage income to an unemployment insurance savings account (UISA).8 Unemployed workers receive benefits according to the U.S. system-that is, equal to 50 368 EAST ASIAN LABOR MARKETS percent of the worker's salary, or a replacement rate of 50 percent, up to a maximum amount for six months. As individuals withdraw benefits, their accounts are debited. Individuals whose account balances are insufficient to pay the benefits to which they are entitled can borrow from the gov- ernment at the same rate they earn on their account. The wages on which contributions are based are limited to a maximum level that may be slightly above the median wage. This level is roughly equal to the level of wages on which unemployment insurance benefits in the United States are based. An individual or employer is exempted from making UISA contributions after the employee's fund reaches a certain amount. Since benefits are 50 percent of wages (up to the ceiling) and last for no more than six months at a time, the maximum benefit the employee can draw during a single period of unemployment is one-quarter of annual wage income from the previous year. Most periods of unemployment are substantially shorter than six months, the median being less than 10 weeks in almost all years. (These episodes would be even shorter with the change in incentives the savings accounts would provide.) Contributions stop, then, when the accumulated balance in an account reaches 50 percent of the individual's wage income in the previous year or 50 percent of the ceil- ing, whichever is smaller. Like employer payments to the U.S. unemployment system, the funds deposited in UISAs are from pretax income and accumulate tax free. If a worker withdraws funds, the money would be considered taxable income, again like unemployment benefits in the United States. The tax would also apply to funds withdrawn in retirement or by heirs. Alternatively the funds deposited in UISAs can come from after-tax income, so that subse- quent withdrawals are not taxed.9 Does such a system in the formal sector raise labor costs and drive some employers into the informal sector, or do employees absorb the costs in the form of lower wage rates? The impact on wages and employment is a function of the program design. A tight link between contributions and benefits at the individual level creates a kind of forced savings program, and the impact on wages is insignificant. But for low-wage workers, a forced savings program may be equivalent to a tax. Thus the self-employed can be exempted, providing room for choosing between formal and infor- mal methods of income security. Evidence on the coexistence of formal systems of income security and informal employment is already available from almost two decades of experience with social security reform in Chile (Edwards 1999). The Chilean law requires wage employees to make contributions to their per- sonal retirement accounts. The self-employed may make voluntary contri- POLICY OPTIONS 369 butions to the pension system. In urban areas, for example, employees ac- count for around 72 percent of working men and 60 percent of working women. Within this group, 82 percent of males and 85 percent of females have fixed-term or permanent contracts, and among those with contracts, 95 percent of both men and women make contributions. The decision to become an employee with a contract is not independent of the decision to contribute to social security. Up to 25 percent of individuals not required to contribute (the self-employed and employees without contracts) do make contributions. This proportion increases with age and schooling and does not vary much with salary level or sector. Enforcement and UISA Accounts Even if an individual holds informal jobs while drawing UISA benefits, el- igibility is limited by the worker's attachment to the formal sector, and the level of benefits is a function of the individual history of contributions. Payments continue until the recipient finds a new job in the formal sector (that is, one that pays premiums into the system) or for up to the maxi- mum benefit period. Because the benefits stream consists of withdrawals from workers' own funds, the system offers no incentive for workers to continue receiving funds. Workers want to find a job quickly so they can stop withdrawing funds from their account. Workers with a negative sav- ings balance understand the consequences of extending the job search. Administrative remedies such as requiring workers to visit employment of- fices weekly to document the job search are not necessary. The reality of the job search is complex. One worker may limit the search to newspaper advertisements. Another may prefer to obtain informal employment. Still others may work at casual jobs or attend a training pro- gram. These choices should not affect eligibility for benefits. The key actor is the unemployed worker, who is not receiving encouragement to stay un- employed. For the same reason unemployed workers do not need the in- centive of a bonus when they find a formal job (the practice in Japan and Korea). Individuals do not receive a free transfer under the UISA system. Government Programs and Income Security in the Informal Sector Instituting UISAs in lieu of severance pay systems would still leave a large proportion of the labor force with no income security coverage. What measures should be put in place (if any) to provide some income security for day laborers and other nonpermanent employees who are not eligible for severance payments under present legislation and would not be cov- 370 EAST ASIAN LABOR MARKETS ered by UISAs? Unemployment insurance would not help those affected by the East Asian financial crisis, making some form of direct assistance more appropriate. How should such a scheme be designed? If eligibility is based on family deprivation, how is it determined in societies with a large informal sector? First, complementary programs are best designed to alleviate poverty rather than to mitigate unemployment, since income- and wage-replace- ment programs are impossible to administer in the absence of effective verification procedures. A number of food-for-work programs or employ- ment guarantee schemes are in place that focus on the unemployed (Ravallion, Datt, and Chaudhuri 1993). But the targeting method is pri- marily wages (or their equivalent in food and other transfers). Without other verification methods, low wages become a self-targeting mechanism. Poverty alleviation programs, even when they are set up in response to a recession or major employment reduction, should not be funded from the same source as unemployment insurance or UISAs. Separate funding for these programs is an important condition if the unemployment insurance program is not to distort labor market incentives. EMERGENCY EMPLOYMENT PROGRAMS. Through the antipoverty pro- gram IDT (Impres Desa Tertinggal), Thailand and Indonesia adopted public works programs (Padat Karya) and revolving credit schemes. These programs are considered cost-effective given the limited resources avail- able. Latin American countries developed a few large-scale employment programs after the mid-1970s, including the Minimum Employment Plan and the Employment Program for Heads of Families in Chile and the Temporary Income Support Program in Peru. The Chilean programs cov- ered about 500,000 people in 1983 and the Peruvian program about 375,000 at its peak. Mexico developed an array of public works and com- munity employment programs that provide training, credit, emergency employment, and public services. Flexible approaches and a large degree of local autonomy in the dis- bursement of funds characterize emergency employment programs. Some impose stringent eligibility requirements. Others pay wages below the legal minimum, and others require participants to work. These restrictions are a function of the availability of funds relative to the number of poten- tial beneficiaries. For example, the proportion of the labor force employed in the Chilean emergency program was virtually constant from 1977 to 1981 in spite of the economic recovery and a decline in real compensa- tion. The explanation lies in increasingly flexible entry restrictions that al- lowed large numbers of potential workers to qualify. POLICY OPTIONS 371 Response to the Economic Crisis: Social Safety Nets in Indonesia and Thailand Thailand and Indonesia, the countries worst hit by the economic crisis in 1997-98, sought to dampen the effects on labor through public sector spending on social safety net schemes, including employment creation (public works) and education and health subsidies (see chapters 2 and 6). Both countries abandoned initial balanced budget policies in favor of a fiscal deficit, a significant proportion of which was allocated to social expenditures (Lee 1998).10 While unemployment rose as a result of the crisis, only 4-6 percent of the labor force in both countries had been af- fected by late 1998 (Manning 1999a). The government saw its main pri- ority as stimulating spending and economic activity in order to support jobs in the flexible informal sector, where many displaced workers had sought work. INDONESIA. Confronted with spiraling food prices, Indonesia allocated a substantial amount to rice subsidies in 1998 and 1999.11 It was also forced to retain subsidies to other goods and services, most notably kerosene and electricity. The rice subsidies (60-70 percent of the market price) were expected to benefit around one-third of the total population in the first half of 1999. The employment support schemes involved three sets of activities in budget year 1998-99: special labor-intensive programs, regular labor- intensive programs, and public works programs (through local gov- ernments). Outside of the rice subsidy the government allocated by far the largest part of the social safety net budget in 1998-99 to credit subsi- dies for small-scale industries and cooperatives at interest rates of 6-16 percent (market rates were above 50 percent). A budget of Rp 20 trillion (US$2.67 billion, or close to 10 percent of all public expenditure) was al- located for distribution by the Department of Cooperatives and Small En- terprises to small-scale firms and cooperatives throughout the country. The government sought to protect the education budget from the ef- fects of inflation by maintaining expenditures in real terms in 1998-99. The main concern was dropouts who failed to make the transition from primary to lower secondary school. Intervention involved scholarships for needy primary and junior secondary students in their final years (with the support of the World Bank and Asian Development Bank)."2 In the area of health, the main form of assistance was subsidies for imported inputs used in the manufacture of basic drugs. Emergency procedures were also put in place to combat shortages of major drugs at the district level. 372 EAST ASIAN LABOR MARKETS THAILAND. Thailand faced a similar sharp cutback in labor demand, al- though in several ways it was better placed to deal with the crisis than In- donesia (Somsak 1999; Sussangkarn, Flatters, and Kittiprapas 1999). The country had put several social support programs in place prior to the cri- sis in response to lagging performance in key social indicators. Prices re- mained relatively stable despite the large drop in the exchange rate. Al- though substantial the declines in the exchange rate and GDP were much smaller than in Indonesia, and by late 1999 the economy showed strong signs of recovery in two successive quarters. Major efforts were taken early on to help displaced workers. These in- cluded the creation of a center that offered assistance to displaced work- ers, including help in obtaining severance pay and social security, plus counseling and job-matching services. A National Committee on Em- ployment Alleviation introduced measures designed to increase rural em- ployment and help displaced workers find employment at home and abroad. These measures set up loan and training programs and provided support for employment-creation programs through Government Savings Bank branches throughout the country. As in Indonesia these efforts were supplemented by social sector loans through the World Bank and Asian Development Bank. In particular, the government program set up several years earlier to provide interest-free educational loans was extended (and much more widely accessed) after the crisis began. A health card system for the poor and a medical assistance program for the elderly were also quickly expanded and helped cushion the effects of the crisis on the poor. In addition the government sought to dampen the effect of layoffs by in- creasing severance pay from 6 to 10 months and extending social security and health coverage from 6 to 12 months for 100,000 retrenched workers who were contributing to the social security fund. Some Final Thoughts As the 21st century begins, Indonesia, Malaysia, the Philippines, and Thai- land are making the transition from traditional to modern economies. A realistic approach to policy interventions aimed at improving income se- curity must include both ends of the spectrum and help smooth the tran- sition between traditional and modern forms of production. Two types of policy interventions address the problems associated with income loss. Poverty-targeted emergency programs do not necessarily focus on the un- employed, since unemployment is not the most important characteristic of the poor and may not be the most appropriate targeting mechanism. POLICY OPTIONS 373 These programs are considered effective mechanisms for reaching house- holds severely affected by income loss. The second set of policies includes all permanent labor market interventions focused on dismissals and the unemployed. Labor market interventions can address the risk of income loss associated with formal employment not only in a recession but also during the normal operation of a growing economy. In most developing countries, including the four cases specifically ad- dressed here, legislation is in place that deals with procedures and respon- sibilities surrounding dismissals. The existing evidence indicates that fir- ing costs deter hiring, reduce labor demand, and hamper the economy's ability to deal with uncertainty and structural change (OECD 1994; Saint-Paul 1999). Unemployment insurance systems are not necessarily a better alternative. In general unemployment insurance reduces the incen- tive to seek work, but the relative size of this effect is a design function and varies from program to program. European unemployment insurance pro- grams pay benefits according to certain preestablished rules and collect the same tax rate from all employers, generating large cross-subsidies from some taxpayers to other program beneficiaries. In contrast the U.S. pro- gram is designed so that employers dismissing the most workers pay the highest tax rate, especially if the workers collect from the unemployment insurance system. Severance pay based on final salaries is an inefficient method of pro- viding income support and smoothing. Severance payments serve wel- fare and efficiency objectives best if they are designed around defined- contribution funds workers can access without any reference to the cause of separation. Similar unemployment savings accounts are an optimum form of unemployment insurance but require a well-developed institu- tional setting. At the very least the system must be able to generate num- bers for employers and employees that can be used to track contributions and benefit payments. Furthermore a system that relies on self-controls re- quires a very efficient and relatively up-to-date flow of information. The administrative capacity that ultimately makes these information flows pos- sible takes time to establish. But the type of information system needed to manage a defined-contribution system of unemployment benefits is much like the one needed to administer a defined-contribution system of old- age benefits. Both types of programs can be designed to cover part of the labor force, and coverage can be extended as the degree of formalization and their relative value increases. A program designed to help workers transition from job to job, be it unemployment insurance based on individual accounts or a severance pay 374 EAST ASIAN LABOR MARKETS system that replaces existing programs, still leaves out the large proportion of the labor force in the informal sector. The most appropriate focus of a complementary program is poverty alleviation rather than unemploy- ment, primarily because an alternative program of income or wage re- placement is impossible to administer in the absence of verification meth- ods. Flexible approaches and a large degree of local autonomy in the disbursement of funds characterize effective poverty alleviation and emer- gency employment programs. Some programs impose eligibility require- ments and other restrictions that are very much a function of the avail- ability of funds relative to the number of potential beneficiaries. Notes 1. Alejandra Cox Edwards is a professor in the Department of Economics at California State University, Long Beach. Chris Manning is Head, Indonesia Proj- ect, Economics Division, at the Australian National University, Canberra. The authors are grateful to Peter Scherer for his detailed comments and suggestions. 2. Despite these strategies even poor working households succeed in having much less variation in consumption than in income. They use a variety of mech- anisms to maintain their consumption level-financial savings, transfers from other members of the family or the community, and borrowing. 3. See, for example McLeod (1993), Nayyar (1995), Khoman (1998), Betcher- man and Ogawa (1999). Among the East Asian countries affected by the crisis, only Korea had an unemployment insurance scheme in 1997. 4. With a young and inexperienced work force and high turnover, assembly- type activities that require little or no training become relatively more attractive. Fine crafts and industrial activities that require a high degree of training and that incur high costs when employees leave become relatively less attractive, all other things remaining constant. 5. These issues are similar to the distortions that result from defined-benefit pension schemes based on workers' final salary. 6. See Bureau of National Affairs, Inc., Personnel Policies Forum (various sur- veys, 1985-90). 7. Workers who have not earned the minimum are not eligible for benefits, on the assumption that low earnings indicate a short or temporary attachment to the labor force. The 12-month base period comprises 4 quarters, and the quarter with the highest earnings determines the weekly benefit amount. 8. The nominal incidence of current unemployment insurance taxes is on em- ployers, but the economic incidence of the tax would presumably be the same if the nominal incidence were on employees. Similarly, gross wages are adjusted downward if employers rather than employees make the contributions, since these deposits are the property of individual workers and thus are similar to tax- deferred cash compensation. POLICY OPTIONS 375 9. Feldstein and Altman (1998) consider several options for managing UISA funds. 10. Thailand projected a budget deficit of 3 percent in 1998, and Indonesia projected a deficit of 6 percent for the financial year 1998-99. 11. See Manning (1999b) for a discussion of the main programs. 12. Block grants were also provided to the poorest 60 percent of schools. References Alba-Ramirez, A. 1991. "Fixed Term Contracts in Spain: Labor Market Flexibil- ity or Segmentation" Unpublished disucssion paper. Anderson, Patricia M., and Bruce D. Meyer. 1998. "Using a Natural Experiment to Estimate the Effects of the Unemployment Insurance Payroll Tax on Wages, Employment, and Denials." NBER Working Paper 6808. Cambridge, Mass.: National Bureau of Economic Research. Becker, Joseph. 1972. Experience Rating in Unemployment Insurance: An Experi- ment in Competitive Socialism. Baltimore and London: John Hopkins Univer- sity Press. Betcherman, Gordon, and Makoto Ogawa. 1999. "Labor Market Institutions and Structures in Indonesia: Reforms to Labor Legislation and to Social Secu- rity." Social Protection Department. Washington, D.C.: World Bank. Draft paper. Blanchard, Olivier, and Pedro Portugal. 1998. "What Hides Behind an Unem- ployment Rate: Comparing Portuguese and U.S. Unemployment." NBER Working Paper 6635. Cambridge, Mass.: National Bureau of Economic Research. Cortazar, Rene. 1994. "Labor Policy in a Democratic Chile." Paper presented at the International Forum on the Latin America Perspective, Paris, November 2-4. Cox, Donald, and Emanuel Jimenez. 1990. "Achieving Social Objectives Through Private Transfers: A Review." The World Bank Research Observer 5(2): 205-18. Cox, James, and Ronald Oaxaca. 1990. "Unemployment Insurance and the Job Search." Research in Labor Economics 11: 223-40. Dertouzos, James, and Lynn Karoly. 1992. Labor Market Responses to Employer Li- ability. Santa Monica: Rand Corporation, Institute for Civil Justice. Dreze, Jean, and Anindita Mukherjee. 1987. "Labor Contracts in Rural India: Theories and Evidence." Discussion Paper 7, Development Research Pro- gramme, London School of Economics. Edwards, Alejandra Cox. 1993. "Labor Market Legislation in Latin America and the Caribbean." Regional Studies Program Report 31, Latin America and the Caribbean Technical Department. Washington, D.C.: World Bank. ____. 1996. "Labor Markets in Indonesia." Washington, D.C.: World Bank. Unpublished paper. 376 EAST ASIAN LABOR MARKETS . 1999. "Are Social Security Reforms Gender Neutral? Evidence from Chile." Economics Department, California State University, Long Beach. Un- published manuscript. Ehrenberg, Ronald, and Ronald Oaxaca. 1971. "Unemployment Insurance, Du- ration of Unemployment, and Subsequent Wage Gain." American Economic Review 66(5). Fallon, Peter, and R.E.B. Lucas. 1991. "The Impact of Changes in Job Security Legislation in India and Zimbabwe." The World Bank Economic Review 5(3): 395-413. Feldstein, Martin. 1976. "Temporary Layoffs in the Theory of Unemployment," Journal of Political Economy 84(5):937-57. . 1978. "The Effect of Unemployment Insurance on Temporary Layoff Unemployment." American Economic Review 68(December): 834-46. Feldstein, Martin, and Daniel Altman. 1998. "Unemployment Insurance Savings Accounts." NBER Working Paper 6860. Cambridge, Mass.: National Bureau of Economic Research. Grubb, D., and W Wells. 1994. "Employment Regulation and Patterns of Work in EC Countries." OECD Economic Studies 21 (Winter): 7-58. Hamermesh, Daniel. 1991. "Unemployment Insurance: Goals, Structure, Eco- nomic Impacts, and Transferability to Developing Countries." Population and Human Resources Department. Washington, D.C.: World Bank. Unpub- lished working paper. Hart, W 1984. "Agrarian Labor Arrangements and Structural Change: Lessons from Java and Bangladesh." Working Paper 67, World Employment Pro- gramme Research. Geneva: International Labor Office. Holen, Arlene. 1977. "Effects of Unemployment Insurance Entitlement on Du- ration and Job Search Activity." Industrial and Labor Relations Review 30(4). Hopenhayn, Hugo, and Juan Pablo Nicolini. 1997. "Optimal Unemployment Insurance." Journal of Political Economy 105 (2): 412-38. Khoman, S. 1998. "Social Policy in Thailand: Issues and Policy Actions." In United Nations Development Program, Social Implications of the Asian Finan- cial Crisis. EDAP Joint Policy Series. Seoul: Korean Development Institute. Kruger, Alan. 1991. "The Evolution of Unjust-Dismissal Legislation in the United States," Industrial and Labor Relations Review V4(2): 123-45. Kuruvilla, Saroth. 1996. "Linkages between Industrialization Strategies and In- dustrial Relations/Human Resources Policies." Industrial and Labor Relations Review 49(4): 635-57. Lazear, Edward. 1990. "Job Security Provisions and Employment." Quarterly Journal of Economics 105(3):699-726. Lee, E. 1998. The Asian Financial Crisis: The Challenge of Social Policy. Geneva: International Labor Office. McCall, John. 1970. "Economics of Information and Job Search." Quarterly Journal of Economics 84(1). McLeod, R. 1993. "Workers' Social Security Legislation." In Chris Manning and Joan Hardjono, eds., Indonesia Assessment 1993-Labor: Sharing in Growth? POLICY OPTIONS 377 Political and Social Change Monograph 20, Australian National University, Canberra, 88-107. Manning, Chris. 1999a. "Labor Market Trends in the NIEs and the ASEAN-4." Asian-Pacific Economic Literature 13(1): 50-68. ____ . 1999b. "The Economic Crisis and Child Labor in Indonesia." Paper prepared for the International Labour Organization, Canberra. Manning, C., and S. Jayasuriya. 1996. "Survey of Recent Economic Develop- ments." Bulletin of Indonesian Economic Studies 32(2) 3-43. Marshall, A. 1991. "The Impact of Labor Law and Employment Practices: Tem- porary and Part-time Employment in Argentina and Peru." Labor Market Pro- gramme DC/38/1991. Geneva: International Labor Office. Miller, Paul W, and Paul A Volcker. 1985. "Unemployment Insurance Eligibil- ity Rights: Evidence from a Comparison of Australia and Canada. Journal of Macroeconomics 7(2): 223-35. Mortensen, Dale. 1970. "Job Search, the Duration of Unemployment, and the Phillips Curve." American Economic Review 60(5). ____. 1994. "Reducing Supply-side Disincentives to Job Creation." In Re- ducing Unemployment: Current Issues and Policy Options. Kansas City: Federal Reserve Bank. NSO (National Statistical Office). 1998. The Impact of the Economic Crisis on Employment, Unemployment and Labor Migration. Bangkok. Nayyar, R. 1995. "Indonesian Labor Legislation in Comparative Perspective: A Study of Six APEC Countries," Unpublished paper, World Bank, Washing- ton, D.C. Nickell, Stephen. 1997. "Unemployment and Labor Market Rigidities: Europe vs. North America." Journal of Economic Perspectives 11(3), 55-74. OECD (Organization for Economic Cooperation and Development). 1994. The OECD Jobs Study. Evidence and Explanations. Vol. II: The Adjustment Potential of the Labor Market. Paris. ____. 1995. Employment Outlook. Paris. Ofreneo, R. 1995. "Philippine Industrialization and Industrial Relations." In A. Verma, T. A. Kochan, and R. D. Lansbury, eds., Employment Relations in the Growing Asian Economies. Londen: Routledge. Otsuka, Keijiro, Hiroyuki Chuma, and Yujiro Hayami. 1992. "Land and Labor Contracts in Agrarian Economies." Journal of Economic Literature 30(4): 1965-2051. Pages, Carmen, and Claudio Montenegro. 1998. "Job Security, Tenure and Em- ployment Dynamics: Theory and Evidence from Chile." Paper presented at the 72nd Western Economic Association Conference, Lake Tahoe, July. Ravallion, Martin, Gauran Datt, and Shubham Chaudhuri. 1993. "Does Maha- rashtra's Employment Guarantee Scheme Guarantee Employment? Effects of the 1988 Wage Increase." Economic Development and Cultural Change 41(2). Saint-Paul, Gilles. 1999. "The Political Economy of Employment Protection." Economics WP No. 355, National Center for Research Science, Universitat Pompeu Fabra Delta. Barcelona, Spain. 378 EAST ASIAN LABOR MARKETS Somsak, Tambunlertchai. 1999. "The Social Impact of the Financial Crisis in Thailand and Policy Responses." In United Nations Development Program, Social Implications of the Asian Financial Crisis. EDAP Joint Policy Series. Seoul: Korean Development Institute. Sussangkarn, C., F. Flatters and S. Kittiprapas. 1999. "Comparative Social Im- pacts of the Asian Economic Crisis in Thailand, Indonesia, and Malaysia." TDRI Quarterly Review 14(1): 3-9. Topel, Robert. 1983. "On Layoffs and Unemployment Insurance." American Economic Review 73(4): 541-59. U.S. Department of Labor. Various years. Foreign Labor Trends. Washington D.C. Verma, Anil, Thomas A. Kochan, and R. D. Lansbury, eds. Employment Relations in the Growing Asian Economie. London: Routledge. Warr, P G. "Thailand." In R. H. McLeod and R. Garnaut, eds., EastAsia in Cri- sis: From Being a Miracle to Needing One. London: Routledge. Welch, Finis. 1977. "What Have We Learned from Empirical Studies of Unem- ployment Insurance?" Industrial and Labor Relations Review 4(july): 451-61. Wiebe, Frank. 1996. "Income Insecurity and Underemployment in the Indone- sian Informal Sector." Paper presented at the Indonesian Workers in the 21st Century Workshop for Economic Reforms and Labor Market Restructuring for Indonesia, Jakarta, April 2-4. World Bank. 1995. World Development Report. Washington, D.C. . 1998. East Asia: TheRoad to Recovery Washington, D.C. . 1999. World Development Report. Washington, D.C. 9 Vulnerable Groups and the Labor Market: The Aftermath of the Asian Financial Crisis SUSAN HORTON AND DIPAK MAZUMDAR1 THE AsIAN FINANCIAL crisis has had severe repercussions for labor markets in the most seriously affected countries of East and Southeast Asia (Indonesia, the Republic of Korea, Malaysia, the Philip- pines, and Thailand). This chapter traces some of the consequences of the crisis for what we call vulnerable groups-young workers, women, workers with the lowest educational and skill levels, and migrant laborers. The cri- sis has played out differently in each of the five countries, in part because the magnitude and timing of events differed and in part because labor markets operated differently in each country before the crisis. For this rea- son the groups that are most vulnerable also differ across countries. The affected countries implemented a variety of measures in response to the crisis. Labor market interventions included expanded severance pay, public works programs, labor market information, and enhanced skill and vocational training, among others. Because of the relative lack of govern- ment intervention in labor markets in East and Southeast Asia prior to the crisis, the countries of the region had been held up as models of success- ful laissez-faire strategies. The crisis showed that while these strategies pre- sented few problems when the economy was growing rapidly, they were not effective during the downturn that followed the crisis. Thus we argue that Asian countries may wish to use the experience of the crisis to strengthen aspects of their social safety nets such as labor market policies. This chapter presents experiences from other regions (primarily Latin 379 380 EAST ASIAN LABOR MARKETS America and South Asia, but also Eastern Europe) that may help guide such reforms. Labor Markets Prior to the Crisis We begin by examining labor market functioning in the five countries prior to the crisis, focusing particularly on vulnerable groups. The five countries were at different stages of development, with per capita income in the most developed (Korea) almost 10 times higher than in the least developed (Indonesia). Several issues are pertinent here: labor supply, in- equality, the structure of employment, and educational levels. The ex- haustion of the labor surplus in the traditional sector (agriculture) was particularly important to the situation of the labor markets in the 1990s. Both Korea and Malaysia passed this point before 1990 and from then on exhibited tight labor markets with very low unemployment. Thailand also reached this point in the 1990s, although unemployment had been creep- ing up for some time because of emerging structural problems. Indonesia and the Philippines had substantial levels of unemployment and under- employment before 1997. The level of inequality is also an important background factor. Accord- ing to the Gini coefficients, Malaysia, the Philippines, and Thailand had higher levels of inequality than the other two countries (table 9.1). In- equality had been increasing in Thailand but decreasing in Korea (Kak- wani and Son 1999). Wage differentials had been narrowing in Korea since the 1980s, and the country had experienced increasing income dis- persion during the period of heavy industrialization (Lee and You 1999). Employment structures also differed across the five countries prior to the crisis. Korea had made the most complete transition out of agriculture, TABLE 9.1 Per Capita GDP Growth in Five Asian Countries (percent) Year Indonesia Korea Malaysia Philippines Thailand 1996 8.0 7.1 8.6 5.7 5.5 1997 4.0 5.0 7.8 5.2 -0.5 1998 -13.6 -5.8 -6.1 -0.4 -9.9 1999 (estimated) -4.0 2.0 0.9 2.0 1.0 Gini coefficient 34.2 36.3 48.4 44.9 46.2 Source: Mahmoud (1999); Gini coefficient for Korea: Lee and Kim (1999). THE AFTERMATH OF THE ASIAN FINANCIAL CRISIS 38I TABLE 9.2 Selected Labor Market Variables for Five Asian Countries Variable Indonesia Korea Malaysia Philippines Thailand Real wage growth (%) 1996a 6.6 6.8 5.8 3.3 2.3 1997a 4.2 4.3 4.2 7.7 -1.4 1998a -37.8 -5.7 -1.1 -2.0 -7.4 % labor force in agriculture, lggob 57 18 27 64 45 % labor force in industry, Iggob 14 35 23 15 14 Foreign workers precrisis (millions)c few 0.2 1.7d few 0.8 Workers overseas precrisis (millions)c 1.1 few few 0.75 0.3 a. Mahmoud 1999. b. World Bank 1993. c. Estimates, from country studies (estimates only); for Thailand: personal communica- tion with Chris Manning. d. An estimated two-thirds were in the country legally. followed by Malaysia (table 9.2). Thailand had more of its GDP and labor force in agriculture than might have been expected given the country's per capita GDP. This fact has important implications, since countries where rural-urban migration occurred closest to the crisis were the most likely to experience reverse migration (urban-rural) during the crisis itself. These countries also had higher levels of remittances between urban and rural areas, something that tends to diversify risks. Employment structures are related to educational levels. As educational levels increase with development, participation rates fall for younger work- ers, and the vulnerable age for recent labor market entrants increases.2 We can observe the effects of educational enrollment on the labor market par- ticipation rates for young East Asian workers. The participation rate for children ages 10-14 was zero or virtually zero in Korea and Malaysia be- fore the crisis, 7-9 percent in Indonesia and the Philippines, and as high as 15 percent in Thailand (table 9.3). A similar pattern occurred in the 15-24 age group. Two-thirds of the young men in this age group were in the labor force in Indonesia, the Philippines, and Thailand, but only one- third were in the labor force in Korea. We therefore see workers as young 382 EAST ASIAN LABOR MARKETS TABLE 9.3 Participation Rates before and after the Crisis (percent) Indonesia Indonesia Korea 1996 1998 1994 Male, total 84.8 83.2 76.4 Male, youth (15-24) 61.4 59.8 31.0 Female total 51.3 51.2 47.9 Female, youth (15-24) 42.9 41.1 42.3 Source: World Bank 1999. as 10-14 as vulnerable groups (new labor market entrants) in Thailand, but this group extends into the 25-34 age cohort in Korea, where educa- tional levels are much higher. Levels and patterns of female participation differed across the five countries (Horton 1996). Women in the Philippines and Thailand were the most strongly attached to the labor force, with a pattern of participa- tion similar to men's in both urban and rural areas. This pattern, which is typical of many rural economies, was also found in Indonesia. In contrast Malaysia exhibited the highest participation rates prior to marriage and childbearing. Korea displayed the pattern characteristic of industrial coun- tries, with women leaving the labor market when their children are born and returning later. (In countries where labor market participation is in- terrupted by childbearing, women tend to receive less on-the-job training and smaller investments in skills training.) Wage differentials between genders also differed across the five countries before the crisis. Data on these differentials are hard to find. Sources such as the International Labour Organization (ILO) provide information only on manufacturing and are thus not sufficiently representative. Data from labor force surveys (Horton 1996) suggest that gender differentials were largest in Korea, followed by Indonesia, where women earned only 50-60 percent of what men earned. The smallest differentials were in Thailand, where women earned almost 90 percent of what men earned. These data must be used cautiously, since they include hourly, quarterly, and annual wages and some are for employees only, excluding the self-employed. How- ever the data do confirm that the two countries where women had the strongest attachment to the labor force (the Philippines and Thailand) THE AFTERMATH OF THE ASIAN FINANCIAL CRISIS 383 Korea Philippines Philippines Thailand Thailand 1998 1997 1998 1996 1998 75.2 82.4 82.9 82.5 71.5 26.4 60.6 61.9 64.4 58.9 46.9 48.9 48.7 66.9 63.8 35.8 35.9 37.1 56.6 50.1 tended to have the narrowest wage differentials between genders. In the Philippines the wage differential was further affected by the fact that women in the labor force had more education than their male counterparts. Migrant workers were (and remain) an important part of the labor mar- ket in all five countries (table 9.2). Indonesia and the Philippines, the two countries with a labor surplus, had a substantial number of overseas con- tract workers (an estimated 1.1 million from Indonesia and 0.75 million from the Philippines before the crisis). Korea and Malaysia had sizable populations of foreign workers (0.2 million and an estimated 1.7 million, respectively). Thailand is an intermediate case. About 0.3 million Thais worked abroad in relatively skilled jobs in 1996, and about 0.8 million foreign workers from low-income countries worked in Thailand. About one-third of the Philippine's overseas workers were in other Asian coun- tries in 1997 and were thus affected by the Asian crisis. Similarly, many foreign workers in Malaysia came from Indonesia and the Philippines, leading to the transmission of shocks within the region. Internal migration was also important, particularly in Indonesia and Thailand, which experi- enced significant rural-urban migration. The Labor Market Response to the Crisis How did the Asian crisis affect the labor markets, and in particular the most vulnerable groups? Our response to this question is restricted by data limitations. The most readily available data by age, sex, and educational and skill levels are for unemployment and participation rates. Unemploy- ment rates alone can be misleading, since women and secondary workers 384 EAST ASIAN LABOR MARKETS (young workers) are more likely than prime-age men to drop out of the labor force when labor demand is weak. Participation rates are useful, since they provide some guidance as to the extent of this discouraged worker effect or its opposite, the added worker effect. The discouraged worker effect occurs when workers give up looking for jobs and leave the work force. The added worker effect occurs when, in some crises, additional workers enter the labor force because the main household earners experi- ence a decline in income. Data on hours worked are also useful (estimates of underemployment often rely on hours worked) but are rarely available by age and sex. Real wage data are even more difficult to come by. The figures that are most readily available-real wages for workers in manufacturing or for employees (as opposed to the self-employed)-are misleading at a time when employment levels are changing rapidly, since the skill composition of the labor force is also changing. Typically the least-experienced and least-skilled workers are the first to be retrenched, making it difficult to interpret gender and age differentials. For example in Korea the gender gap in wages (for employees) appears to have nar- rowed following the crisis, almost certainly because those women who re- mained as employees were disproportionately skilled or experienced. One final and serious limitation is that data for Malaysia are not available for this discussion. The Malaysian Labor Force Survey data present informa- tion on the composition of the unemployed but not population-based rates, with breakdowns only for the registered unemployed, a small and unrepresentative proportion of the total (Mansor and others 1999). The estimates in table 9.4 on labor market responses to the crisis should be taken as orders of magnitude rather than precise estimates, since figures on (for example) job losses are hard to come by. The labor force in 1997 totaled 94 million in Indonesia, 22 million in Korea, 8 million in Malaysia, 30 million in the Philippines, and 35 million in Thailand (World Bank 1996). The data suggest that the Philippines was the least af- fected by the crisis and that in Malaysia the brunt of the adjustment was borne by migrant workers. Malaysia is unusual in its very large proportion of foreign workers (close to 20 percent of the labor force). That, plus the large fall in participation rates, left relatively little of the adjustment in Malaysia to open unemployment and real wage declines. Neither Malaysia nor the Philippines engaged in intensive efforts to cushion the adjustment, unlike Indonesia, Korea, and Thailand. The scale of labor market adjustment has been staggering in Indonesia and of course has been accompanied by tremendous social unrest. Even before the crisis Indonesia was having difficulty absorbing new labor mar- TABLE 9.4 Estimated Magnitudes of Labor Market Response to the Crisis Effect Indonesia Korea Malaysia Philippines Thailand Job losses (millions) 1997/98 2.5 0. a 0.1 n.a. 0.8 Usual labor force growth (annual) 3.5 0.4 0.4 n.a. 0.2 Unemploymentb increases (millions) 1997/98 0.9 0.8 0.2 n.a. 1.2 Underemployment increases (millions) 3.7 n.a. n.a. n.a. up I m (defined 1997/98 as < 20 hours/wk) Change in labor force participation (millions) +0.7 +0.1 (men) -0.2 n.a. -0.9 mc 1997/98 -0.3 (women) Change in migration (millions) overseas legal foreign overseas 1997/98 work workers work +0.8 n.a. -0.3 +0.8 n.a. Real wage growth (percent) 1996 6.6 6.8 5.8 3.3 2.3 1997 4.2 4.3 4.2 7.7 -1.4 1998 -37.8 -5.7 -1.1 -2.0 -7.4 n.a. not available. Source: Country studies; real wages from Mahmoud (1999). a. Data are for permanent job losses only; temporary job losses were much higher. b. Underemployment is defined as less than 35 hours per week for Indonesia and less than 20 hours in Thailand. c. First quarter to first quarter. 386 EAST ASIAN LABOR MARKETS ket entrants, and unemployment rates were relatively high (table 9.5). Even a large shift in labor to agriculture after the crisis-5 million work- ers-was not enough to prevent unemployment and underemployment from increasing or real wages from decreasing (by around 38 percent in one year). Participation increased during the crisis, largely because the number of women in the work force grew. As more women went to work, secondary school students were pulled out of school, an occurrence that is indicative of the desperate economic situation. These labor market re- sponses typify those of a low-income country where formal safety nets are completely inadequate to deal with the scale of a crisis. The response in Thailand was a larger proportionate increase in the open unemployment rate than in Indonesia and a significant decline in participation. The data used here compare August 1997 and November 1998 and overstate the decline somewhat, since there are seasonal effects, but the overall effects are still significant. Thailand also experienced sub- stantial internal migration that led to job losses in Bangkok and rising un- employment rates in rural areas, particularly the northeast. (Urban migra- tion in Thailand is a relatively recent phenomenon, and many individuals still retain ties to their original rural communities.) The response in Korea was more typical of an industrial country, with open unemployment increasing substantially and the participation rate of women in the labor force decreasing (the discouraged worker effect). Job losses affected casual rather than permanent workers. Many firms under- took multiple adjustments-shorter hours, temporary leave, and tempo- rary shutdowns (Shin 1999). Korea also exhibited the problems of indus- trial countries, with a growing proportion of long-term unemployed and difficulty reallocating the most senior workers, since firms customarily keep permanent workers whenever possible. The Crisis and Vulnerable Groups in the Labor Market The groups we have identified as vulnerable include young workers, who are vulnerable in all countries, women and their households, less educated workers, and migrants. Labor market concerns vary across these groups. With young workers the issue is generally not their immediate welfare (ex- cept for those young workers who do not live with their parents but are themselves household heads or live on the streets), but their long-term prospects for entering the labor market. Female-headed households in countries such as Indonesia, Korea, and Malaysia, where women receive low relative wages and tend to be seen as temporary workers, are especially TABLE 9.5 Unemployment Rates before and after the Crisis (percent) Indonesia Indonesia Korea Korea Philippines Philippines Thailand Thailand 1996 1998 1995 1998 1995 1998 1996 1998 Male, total 4.0 5.5 2.3 7.7 7.0 9.5 1.0 4.3 Male, 15-24 12.2 15.7 9.3 20.8 25.9 17.9 2.6 11.2 Male, 35-54 I.oa 2.3a 1.7 6.2 5.6b 5.2b 0.3 1.9 Male, 50+ n/a n/a 0.9 5.4 7.6b 6.60b .4 2.5 Female, total 3.3 5.0 1.6 5.6 8.2 9.8 1.1 4.6 Female, 15-24 15.0 19.1 6.0 12.8 33.9 22.1 2.3 8.6 Female, 35-54 1.8a 2.3a 0.7 4.5 6.7b 4.5b 0.5 3.2 Female, 50+ n/a n/a 0.3 2.4 6.9b 5.7b 0.8 2.7 a. Age 25+ for Indonesia. b. Age groups for Philippines are 35-54 and 55+. n/a not available. Source: BPS (1998); NSOP (1999); NSOT (1998); ILO (various years). 388 EAST ASIAN LABOR MARKETS vulnerable. Households where women are not well educated are also vul- nerable. Beegle, Frankenberg, and Thomas (1999) found that female edu- cation in Indonesia and Peru, respectively, was an important factor in ame- liorating the impact of crises on household welfare. Less educated workers, who are also at risk in all countries, are particularly vulnerable in areas where the option of going back to work in agriculture is not viable. Finally, migrant workers, such as low-skill foreigners in Malaysia and internal migrants in Thailand, are an important vulnerable group, as are internal migrants in Indonesia, although no data are available on these workers. Young Workers Young workers (new labor market entrants) have less labor market experi- ence, little firm-specific training, and (usually) low productivity rates. Un- employment rates for the young are higher than they are for prime-age workers even during normal times, and this group is one of the first af- fected during a crisis, when new hiring slows down or halts. Among coun- tries of the Organization for Economic Cooperationi and Development (OECD), unemployment rates for the young have climbed as high as 75 percent in particularly depressed regions. This level of unemployment is a recipe for social problems. Even if the majority of young workers still live with their parents and are not under imminent threat of starvation, there are negative welfare effects. "Scarring" is an issue, as these workers do not develop a long-run attachment to the labor force. And some young work- ers are household heads and may have children. Graduate unemployment, a special type of youth unemployment, has been a long-term problem in a number of developing countries. The prob- lem became acute in the 1980s in countries where a university degree had traditionally been the ticket to a government job. As government hiring slowed, some graduates opted to remain unemployed for several years and wait for their place in the job queue. The wait was often long-up to four years in Egypt, for example (Assad and Commander 1994). The Asian crisis had dramatic effects on youth unemployment (table 9.5). In Indonesia the unemployment rate for young workers (ages 15-24) is normally about 7-8 times as high as for older workers (25 and above) of the same sex. In Korea, the Philippines, and Thailand, it is around 4-5 times that of prime-age workers-that is, workers ages 35-49 in Korea and Thailand and ages 35-54 in the Philippines. In Korea, where tertiary ed- ucation is widespread, unemployment in the 25-34 age group is also about twice that of prime-age workers, as it is in the Philippines and Thailand.3 THE AFTERMATH OF THE ASIAN FINANCIAL CRISIS 389 Although the ratio of youth unemployment to that of older workers did not change much with the crisis, absolute rates went up to very high lev- els. Rates in subgroups, such as the 15-19 age group, were even higher. Participation rates for young workers declined by more than the aver- age in absolute terms. In all four countries participation rates decreased more in both absolute and relative terms for vulnerable groups (young workers, women, and older workers) than for prime-age males. The de- creases for young workers were 1-2 percent in Indonesia, 2-4 percent in Korea, and 4 percent in Thailand. Participation increased by just over 1 percent in the Philippines (table 9.3). Participation rates did not decrease because young workers were staying longer in school. In Indonesia, for in- stance, school participation for those ages 7-12 and 13-19 also decreased (Frankenberg, Thomas, and Beegle 1999), although falling school enroll- ments were not particularly noticeable in the other four countries. Pre- liminary results from Beegle, Frankenberg, and Thomas (1999) for In- donesia suggest that the presence of young women ages 15-19 in the household was one factor helping to maintain income. These girls may have assumed household responsibilities in order to free their mothers to participate in the labor market. Women A sizable literature exists on the question of whether structural adjustment in Latin America and Africa during the 1980s adversely affected women (Cox-Edwards and Roberts 1994; Bardan 1994). To some extent the out- comes were specific to country situations. In Bolivia, for example, the first large layoffs were in mining, a sector dominated by men. In other Latin American countries, cuts in public sector employment disproportionately affected women, who made up a smaller share of private formal sector em- ployment. In Africa relatively few women were in urban formal sector em- ployment (which shrank), and women working in food-crop agriculture could benefit from improved terms of trade from agriculture. But crises and recessions that slow (or reverse) the transition from the informal to the formal sector put women in all countries at a disadvantage, since women tend to shift relatively slowly out of unpaid family work into for- mal employee status.4 Do developments that disadvantage women in the labor market trans- late into disadvantages in terms of welfare? If most adult women live in households that also contain male adults, the effects of changes in total household income are of first-order importance. But there are second- 390 EAST ASIAN LABOR MARKETS order effects if women's earned incomes decrease relative to men's. The lit- erature on intrahousehold spending suggests that decreases in women's relative contributions to household income are associated with a decrease in household expenditures on both children and basic needs. Depending on the country, the share of female-headed households may also be so large that adverse welfare effects result when women are disadvantaged in the labor market. If women have to work longer hours in poorly paid jobs in the informal sector following a crisis, for instance, their ability to pro- vide childcare is reduced. The unemployment data do not suggest strong gender differentials in unemployment (table 9.5). In both the Philippines and Thailand, female unemployment is consistently higher than male unemployment, although the differential has been narrowing over time. The opposite is true in In- donesia and Korea, where male unemployment rose relative to female after the crisis. The most likely reason for the apparent increase is that female participation rates fell more than male participation rates in all countries except Indonesia. Participation rates for women are consistently lower than for men in virtually all age groups, with the exception of the youngest (15-24) in Korea, where men receive more education (table 9.3). As noted earlier, declines in the participation rates of women of all ages exceeded those for prime-age men after the crisis, with the sole exception of In- donesia, where participation rates for women over 25 increased. Data on hours worked by sex are not available for all the countries, but data from a smaller sample survey for Indonesia (the Indonesia Family Life Survey) suggest that the number of hours worked declined for both gen- ders following the crisis. The decline in hours worked was greater for women than for men, especially in the formal sector. The hours men in the formal sector worked dropped by 1.4 per week, while the hours worked by their female counterparts fell 2.9 per week.5 At the other ex- treme, the incidence of long hours (defined as more than 45 hours per week) for women increased, rising from 20 percent in 1994 to 22.9 per- cent in 1997 and 24.9 percent in 1998. For men it fell following the cri- sis, although about twice as many men as women work long hours (Islam and others 1999). This situation is consistent with the view that women entered the labor force in Indonesia to try to offset the drastic falls in household income. Jobs open to women are more frequently in the infor- mal sector, where the poorest work very long hours at jobs with very low returns. Data on sectoral employment are available by sex only for Korea (table 9.6). The main sectoral shift in Korea after the crisis was out of employ- TABLE 9.6 Sectoral Employment before and after the Crisis (percent) Indonesia Indonesia Ko rea Korea Philippines Philippines Thaikand Thailand 1996 1998 1994 1998 1997 1998 1996 1998 To al Employee 34.2 38.2 62.0 61.2 47.7 49.0 46.4 42.8 Self-employed 21.4 22.7 27.8 28.8 37.5 37.6 33.3 36.7 Unpaid family 44.4 39.2 10.2 10.0 14.8 13.4 20.3 20.4 Male Employee n.a. n.a. 64.2 63.1 n.a. n.a. n.a. n.a. Self-employed n.a. n.a. 34.0 35.1 n.a. n.a. n.a. n.a. Unpaid family n.a. n.a. 1.8 1.7 n.a. n.a. n.a. n.a. Female Employee n.a. n.a. 58.7 58.3 n.a. n.a. n.a. n.a. Self-employed n.a. n.a. 18.8 19.4 n.a. n.a. n.a. n.a. Unpaid family n.a. n.a. 22.6 22.3 n.a. n.a. n.a. n.a. na. not available. Source: NSOT (date) (key tables 1 996-98.) 392 EAST ASIAN LABOR MARKETS ment and into self-employment (with no other employees). The data also show that women are more likely than men to be unpaid family workers in Korea and less likely to be employees or self-employed. Data on relative wages by gender following the crisis are not readily available, especially household survey data, which would be more useful than establishment survey data. Workers with Low Educational and Skill Levels Workers with few skills and little education are vulnerable to economic downturns for two major reasons. First, many of them work in construc- tion, a highly cyclical industry, or agriculture, where they are susceptible to a variety of developments. (In Asia, for instance, the adverse effects of El Nifio reduced demand for agricultural labor.) Second, they have less training and less firm-specific capital and thus are the group most likely to be retrenched. In Asia wage differentials between skilled and unskilled workers have traditionally been high owing to the relative scarcity of skilled workers. However, the gap has been narrowing as educational levels rise. In Korea wages for middle school graduates have increased steadily, rising from 33 percent of those of university graduates in 1980 to 60 percent in 1997. Relative wages for high school graduates also rose from 46 percent of uni- versity graduates' to 68 percent over the same period. And the ratio of earnings for wage earners in the 90th percentile to earnings for workers in the 10th percentile fell from 4.7 to 3.7. These trends are unlike those in industrial countries, where wage differentials between skilled and un- skilled workers are on the rise.6 Information on the effects of the Asian crisis on relative wages broken down by skill level is not yet available for most countries. A little infor- mation can be gleaned from data on unemployment rates that are disag- gregated by skill level. In Korea relative wages fell for the least educated and for those in production jobs after the crisis, and distribution by quin- tiles also worsened. In fact the only wage gap that did not worsen was the differential between genders, which continued to narrow. Unemployment among those with a high school education or less had traditionally been lower than among college graduates, but after the crisis the unemploy- ment rate for those with less education began to exceed the rate for those with a college education (Shin 1999). In Thailand, where the average ed- ucational level is lower, the highest unemployment rates are usually in the group with either a lower or upper secondary education, and this trend re- THE AFTERMATH OF THE ASIAN FINANCIAL CRISIS 393 mained unchanged during the crisis (TDRI 1999). Indonesia experienced a small increase in the proportion of unemployed with a lower secondary education (Islam and others 1999). Thus Korea offers the clearest evidence of weakening demand for the unskilled. Migrants Recent migrants are among those most likely to lose their jobs in an eco- nomic downturn because they have less local labor market experience and often work as temporary laborers (a group that, as we have seen, is among the most vulnerable when the economy slackens). We do not have com- parable data on this vulnerable group for all five countries, but at least one source (Abubakar 1999) provides information about foreign migrant workers in Malaysia and data on internal migration for Thailand. The Philippines collects more extensive data on its overseas workers than most other countries, but the results of the 1998 survey of overseas workers were not available to the authors at the time of writing. FOREIGN MIGRANTS. Abubakar (1999) discusses the extent to which the situation of foreign migrants in Malaysia became increasingly desperate after the crisis. Employers began to skimp on housing costs by crowding more migrant workers into each housing unit. Bonuses for foreign work- ers disappeared, and hours of work increased. Abubakar suggests that in some cases domestic workers were retrenched because they were more ex- pensive than migrants and that migrants were forced to work longer hours. The government also increased the cost of renewing permits for foreign workers to 140 ringgit (RM) per month, a substantial sum for service workers who may only earn RM 350-400 monthly. The Malaysian gov- ernment stopped issuing new permits for foreign workers in the service sector and even considered trying to repatriate as many as a million for- eign workers whose work permits expired or who entered the country ille- gally. This scheme created so much concern that the government decided instead to redeploy foreign workers from services and construction to sec- tors unable to recruit enough nationals (plantations and parts of manu- facturing, for instance). But according to Abubakar, of the 1.7 million for- eign workers in Malaysia, only 1,000 had been formally redeployed as of 1998, and only 2,382 had applied to change sectors (table 9.2). INTERNAL MIGRANTS. In periods of economic downturn, when regional disparities tend to increase, we can expect internal migrants to bear the 394 EAST ASIAN LABOR MARKETS brunt of the recession. In developing countries generally, including those in Asia, labor markets are less integrated than those in industrial economies, in part because urbanization rates are lower. For this reason in- come dispersion and wage differentials among regions tend to be higher in developing countries. During periods of strong growth, a sizable num- ber of workers migrate from low- to high-income regions, but to a large degree this internal migration is temporary. The cost of this type of mi- gration tends to be relatively low, and temporary migrants from rural areas do not give up their stake in the land or family farm. Such temporary mi- grants are disproportionately represented in the lower echelons of workers in high-income regions (the informal sector, casual labor, and the non- tenured segment of the formal labor force).7 Typically such migrants re- turn to their regions of origin, where incomes are generally lower. Thailand offers a good example of the effects of the crisis on internal migration. Data from the National Migration Survey of Thailand from 1995 suggest that the country enjoys considerable interregional labor mo- bility, largely from richer regions to poorer ones.8 About a third of all in- ternal migrants have moved more than once in a two-year period, and about half of these were seasonal migrants who moved to urban areas in the dry season, returning to agricultural work (most often in the north and northeast) in the wet season. Short-term population flows are strongly affected by this trend. Net losses as a result of seasonal and other tempo- rary migration were nearly 150 per 1,000 for Bangkok in the 1995 wet season and 50 per 1,000 for the central region. At the receiving end, net gains of 33 and 68 people per 1,000 were recorded for the north and northeast, respectively. Rates of net permanent inmigration, however, are generally small, with gains ranging from 4.1 per 1,000 for Bangkok to 12.8 per 1,000 for the southern region over the two-year period. In the 1980s and early 1990s, Bangkok and the surrounding vicinity formed Thailand's major growth center. Because of this growth the region enjoyed some of the highest incomes in the country. In 1992-97, how- ever, incomes in the northeast grew at more than three times the rate of Bangkok and its surrounds. But the crisis interrupted this strong favorable trend toward increased income equality across regions. Financial crises and foreign investment flows arguably impact the modern sectors of an econ- omy much more than the traditional farming sectors-and in fact the rise in unemployment was proportionately higher in Bangkok than in other regions (NSOT 1998). But this hypothesis does not take into account the impact on regional economies of disrupted flows of labor migration. Tra- ditional labor flows from the north and northeast to Bangkok raised in- THE AFTERMATH OF THE ASIAN FINANCIAL CRISIS 395 comes through remittances. As unemployment in the Bangkok area in- creased, the flow of migrants slowed and incomes in the northern regions fell, reversing the trend toward more income equality across regions. We can say, then, that while the direct impact of the crisis in the poorer north- ern and northeastern regions of Thailand may have been smaller than it was in Bangkok, the net effect was greater. Kakwani calculated the trends in per capita incomes for different re- gions in 1992-98, with seasonal adjustments (NESDB 1998) (table 9.7). He then goes on to calculate the deviations from the trend line for in- come, which he terms "crisis indices." The crisis indices thus show not just actual declines in incomes in the postcrisis period but declines relative to the levels incomes would have reached if the crisis had not interrupted the trends. Labor Market Interventions Governments in East and Southeast Asia tend to intervene in labor mar- kets less than developing country governments in other regions. Despite the crisis, for instance, four of the countries studied here still have no un- employment insurance scheme (the exception is Korea; see table 9.8). The crisis has forced governments to intervene to some extent to protect work- ers, largely in the areas of income support and training. Labor Market Institutions before the Crisis Payroll taxes in the region traditionally tended to be somewhat lower than in Latin American countries, largely because individuals and households rather than the state provided social services such as health, education, and pensions (World Bank 1996).9 Those payroll taxes that were levied fre- quently support individual provident funds (in Malaysia and Singapore), individual health accounts (for instance, Medisave and Medishield in Sin- gapore), and severance pay (in Korea, Malaysia, the Philippines, and Thai- land). Contributions to general (rather than individual) pension funds were also less common than in regions such as Latin America. In Korea a national pension scheme began in 1988, but it covered only 58 percent of the working population (Lee and You 1999). Other labor market institu- tions were also weak before the crisis. PUBLIC WORKS. Public works had been phased out as an employment protection strategy in all five countries as a result of burgeoning growth. TABLE 9.7 Per Capita Income Measures, Thailand Per capita income Per capita income Q1 1998 Q3 1998 Long-term growth Kakwanis crisis Kakwanis crisis Region (bahtper month) (bahtper month) rate (percent) index, QJ 1998 index, Q3 1998 Central 2,791 2,791 6.8 -20.4 -26.1 Eastern 2,791 2,791 7.1 -29.6 -31.7 Western 2,569 2,518 5.1 -18.8 -22.1 Northern 2,096 1,896 6.0 -19.3 -26.0 Northeastern 1,488 1,166 10.2 132.4 -37.4 Southern 2,339 2,195 8.1 -22.7 -31.3 Bangkok and vicinity 5,439 5,147 2.9 -8.1 -14.5 Country 2,633 2,402 6.3 -19.2 -24.8 Source: NESDB (1998). TABLE 9.8 Labor Market Institutions Prior to Crisis Institution Indonesia Korea Malaysia Philippines Thailand Unemployment None Began 1995 for None None None (but can insurance firms >30 access some of employees; pension fund requires 6 when months of unemployed) contributions Severance pay No information Dismissal only for Those employed for 1 month per year of Yes (no details) cause prior to >12 months service 1997; severance receive 10-20 pay progressive days' pay per with experience; year of dismissal of experience senior workers costly Skills/vocational 712 public, 3,000 140 public, 178 No information TESDA began Public training training private schools private, 237 in- 1994: 80% of under Ministry plant providers; training private, of Labor, firms required to 20% public; Ministry of pay levy or PEFSA provides Industry and provide in-plant vouchers to Bangkok Metro training individuals Training Centre; few private institutions (Table continues on the following page) TABLE 9.8 Labor Market Institutions Prior to Crisis (continued) Institution Indonesia Korea Malaysia Philippines Thailand Labor market System not well Network of centers, Network of centers No information 7 branches in information developed; not heavily Bangkok, 76 in offices not utilized provinces, computerized, computerized and not linked and linked, but access is slow (hardware problems) Other Low minimum Minimum wages Minimum wages, wages relatively higher nonbinding than rest of Asia, but exemptions permitted Sources: Country studies; Abubakan (1999). THE AFTERMATH OF THE ASIAN FINANCIAL CRISIS 399 Indonesia was the last to phase out its program in 1994 (for more on labor market strategies before the crisis, see chapter 7). TRAINING AND VOCATIONAL EDUCATION. All five countries had training and vocational education systems before the crisis (table 9.8). These sys- tems utilized both private and public institutions. The Philippines, for ex- ample, instituted the Technical Education and Skill Development Au- thority (TESDA) in 1994. TESDA still operates as a regulatory agency and also runs most of the public training institutions (other than state uni- versities and colleges). The Philippines had a voucher system that permit- ted trainees to shop around for a suitable institution, either private or pub- lic. However outright competition between public and private providers was restricted, since public institutions continued to receive block grants, and the number of vouchers was fairly limited (around 10,000 in 1999). Prior to the crisis Korea had begun making the transition from a system of compulsory training that required firms to provide in-plant training or pay a levy to a voluntary system that would provide firms with incentives to offer training. UNEMPLOYMENT INSURANCE AND SEVERANCE PAY. Unemployment in- surance tends to exist only in high-income countries, as it requires not only extensive record keeping but also a reasonably large formal sector to support it. Korea began an unemployment insurance scheme in 1995 that covers firms with more than 30 employees. Employees must make contri- butions for six months in order to file a claim. None of the other coun- tries had an unemployment insurance scheme prior to the crisis. Severance pay has long been widespread in the formal sector in Asia. Both employers and employees contribute to a fund that guarantees dis- missed workers a certain amount of severance pay based on years of ser- vice. In Korea all firms with more than five employees are required to pro- vide severance pay. Dismissal (other than for cause) was not permissible for formal sector workers in Korea prior to 1997. These precrisis systems were not necessarily ideal. In Korea, as we have seen, severance pay accu- mulated with experience, making it almost prohibitively expensive to ter- minate senior workers while providing an incentive to dismiss workers in the middle of their careers. In other cases access to funds intended for sev- erance pay encouraged workers to leave their jobs. OTHER LABOR MARKET POLICIES. Traditionally other labor market poli- cies have been relatively noninterventionist in Asia. Minimum wages were 400 EAST ASIAN LABOR MARKETS not emphasized and in many cases were not binding in Asia before the cri- sis (with the possible exception of the Philippines). Korea, the Philippines, and Thailand all had minimum wages. In Thailand minimum wages rose 17 percent in real terms during the 1980s, but this increase meant that they declined sharply relative to average wages (Squire and Suthiwart- Narueput 1996). As of January 1, 1991, each of the 151 members of the ILO had rati- fied an average of 36 of its conventions. But Indonesia had ratified only 9, Malaysia (including component regions) 12, the Philippines 21, and Thailand 11. Korea was not even an ILO member at the time. In Asia only Sri Lanka and the Philippines ratified the 1990 ILO convention on the Protection of Rights of All Migrant Workers and Their Families. Enforce- ment of labor legislation was also an issue in the years before the crisis. The Philippines made fairly strenuous efforts to improve enforcement, dramatically increasing the number of staff available to conduct inspec- tions. But Thailand lagged in this respect, and Malaysia did not make do- mestic labor legislation binding in export processing zones.10 Changes in Labor Market Institutions Following the Crisis East Asian governments were forced to intervene in the labor markets fol- lowing the crisis. The most significant changes occurred in the three coun- tries with the greatest adjustment problems-Indonesia, Korea, and Thai- land (table 9.9). However in Indonesia these changes were concentrated in two areas: public works and self-employment. Thailand also made changes to its policies on severance pay and its vocational training schemes. Korea made the largest number of changes, expanding its programs in every cat- egory. The Philippines made no major changes, as it had been dealing with high unemployment for some time. Malaysia shifted much of the ad- justment onto the migrant workforce, blocking new recruitment in cer- tain sectors, increasing the penalties for citizens aiding illegal immigrants, undertaking a limited amount of redeployment, and raising some levies on migrants. The government also considered but did not implement repatriation. PUBLIC WORKS. Since 1997 Indonesia, Korea, and Thailand have all im- plemented large public works schemes that have created millions of person-days of employment. In Indonesia the Padat Karya programs were revived in Java for 4 months, followed by a second phase implemented as 16 subprograms by different ministries and institutions across the coun- TABLE 9.9 Changes in Labor Market Institutions Following the Crisis Institution Indonesia Korea Malaysia Philippines Thailand Public works Large public works Large public works No change No change Large public works schemes schemes; some schemes targeted to women Unemployment None Extended to firms None None None insurance >5 Mar 98, to all firms Oct 98, but only to those contributing >6 months; only covers 20-25% of job losers Severance pay No change Allow employees to No change No change Increased maxi- received mum benefits to severance for 10 months voluntary quits salary; set up public fund for workers whose firms go bankrupt (Table continues on the following page) TABLE 9.9 Changes in Labor Market Institutions Following the Crisis (continued) Institution Indonesia Korea Malaysia Philippines Thailand Skills/vocational No change 103 bn won Graduate TESDA trained Funds for skills training additional; entrepreneurial only 1545 development for shifting from training scheme, displaced new entrants, compulsory levy with business workers in 1998 unemployed; to incentives for loans considering voluntary making skills training development fund mandatory; loans to businesses for training Labor market No change Government set up No change No change No change information "Employment Security Centres"; large rise in usage; "Work-Net" begins Other Subsidized loans to 6.4 trn won in New immigrant Self-employment co-ops wage subsidies, recruitment loans some targeted to blocked in some youth and women; sectors; efforts to loans for venture redeploy existing business migrants; levies on migrants raised -,.. t;- T< -r l Kim 1q99: Ahbjbakar. 1999. THE AFTERMATH OF THE ASIAN FINANCIAL CRISIS 403 try. The second phase created around 225 million person-days of employ- ment between April 1998 and December 1999 (Islam and others 1999). The program suffers from some problems, however. It has a top-down management structure, so that the involvement of community organiza- tions is weak. Detailed background information on vulnerable groups is lacking, corruption and nepotism abound, and the scheme pays wages in excess of market levels. Thailand has several public works schemes. The Social Investment Pro- gram operates a Social Investment Fund for small projects originating or implemented at the community level. The Regional Urban Development Fund supports local governments seeking to expand urban infrastructure. The program, which started in September 1998 and is scheduled to run for at least 40 months, is expected to provide 557,000 person-months of employment. The projects experienced some initial difficulties. As of May 10, 1999, some seven months after the program commenced, only 2 per- cent of the funds under the Social Investment Fund had been dispersed, and only 4 percent (comprising two projects) under the Regional Urban Development Fund had been approved. Most of the proposed projects did not meet the program's approval requirements (Arya 1999). Thailand's Miyazawa program expects to fund newly created jobs for 86,000 educated workers and 400,000 unskilled workers in activities such as rural infrastructure. The program was implemented through local au- thorities in March 1999 and should provide 50 million person-days of work. In addition funds have been earmarked in the regular government budget to support the employment of almost 800,000 people in rural areas. In Korea public works projects benefited 440,000 workers in 1998 and were expected to benefit 1.2 million in 1999, peaking in the second quar- ter. Local governments use 10 percent of the general public works budget to implement projects that employ daily labor. The government has also allocated funds to help employ the highly educated by instituting an in- ternship program for graduates. The program is in such high demand that the government is introducing a similar one for high school graduates. Korea also has a venture capital fund, although the fund's restrictive con- ditions have prevented it from being used to create employment to any great extent. SELF-EMPLOYMENT. Indonesia, Korea, and Thailand have implemented schemes to encourage self-employment. The Indonesian government pro- vides subsidized loans to cooperatives. However, the cooperatives can get a safe return by depositing these funds at banks (which are offering higher 404 EAST ASIAN LABOR MARKETS interest rates), diminishing the effect of the loans on employment genera- tion. Thailand has also earmarked funds for self-employment loans, as has Korea with its venture capital loan scheme (for further discussion of these programs, see chapter 7). WAGE SUBSIDIES. Korea has implemented a large wage subsidy pro- gram. The program pays as much as two-thirds of a worker's wage if a firm can demonstrate its need to adjust by shutting down two or more days a month, cutting hours by more than 10 percent, placing employees on leave for a month or more, transferring workers to affiliates, or switching to a new line of business while retaining at least 60 percent of existing em- ployees. In 1998 the program covered 800,000 workers, and in 1999 the government expanded it and increased the duration of the subsidy from 6 to 8 months. A survey of enterprises suggests that without the subsidies, 22.3 percent of all jobs would have been lost (the figure is higher in the nonmanufacturing sector). TRAINING AND VOCATIONAL EDUCATION. In Thailand some funding was induded in the various programs for skills development for new job mar- ket entrants and the unemployed. Malaysia instituted a graduate entrepre- neurial training scheme as well as business loans. Korea increased its bud- get for training by 103 billion won. Lee and Kim (1999) argue that the change from a compulsory to a voluntary system "Mostly extends what has been proven ineffective" (p. 12). The changes do not offer training pro- viders incentives to meet market needs, and private providers continue to overemphasize skills for manufacturing, where demand has been declining. UNEMPLOYMENT INSURANCE AND SEVERANCE PAY. Korea broadened its unemployment insurance scheme following the crisis, extending it first to firms with more than five workers and subsequently to all firms. However the requirement that workers contribute for six months to establish eligi- bility limits coverage to only 20-25 percent of job losers. None of the othe-rsountries discussed here has instituted an unemployment insurance scheme since the crisis, and all still rely primarily on severance pay. Thai- land extended the maximum amount of severance pay to 10 months and set up a fund to make the legislated severance payments to workers whose firms go bankrupt (the fund is supported by fines on firms that break labor laws). Malaysia has also extended its severance pay scheme since the crisis to cover workers who quit voluntarily. THE AFTERMATH OF THE ASIAN FINANCIAL CRISIS 405 LABOR MARKET INFORMATION SYSTEMS. Labor market information sys- tems in the region tend to vary with the level of development and were ex- tremely limited before the crisis. Public employment offices in Indonesia, for instance, are seen as limited in their offerings because they are not computerized (and therefore not linked). Those in Thailand are comput- erized and linked, but some of the equipment is old, and access is slow. Malaysia and Korea have networks of public employment offices, but they are not widely used, and job seekers do not see them as a major strategy in the job search. Korea has made major changes in its labor market in- formation systems, setting up "employment security centers" that com- bine payment facilities for unemployment insurance with information on work applicants and vacancies. The government has also instituted a sys- tem (modeled after systems in the United States and Canada) that allows workers to check vacancies via the Internet. The number of people using the centers increased enormously in 1997-98, rising by a factor of nearly 10, and the number of vacancies posted doubled. The ratio of the unem- ployed registered with the centers went from 4.8 percent in 1996 to 22.5 percent in 1998. But only 5.8 percent of those who found a job found it through the public system (Keum 1999). The Effects of Labor Market Interventions on Vulnerable Groups Information about the effects of the interventions on vulnerable groups is limited, but we can make some observations based on program design. For instance, successful public works programs can be designed to include women. In Bangladesh food-for-work programs include projects aimed specifically at women and pay attention to issues such as childcare for participants. In Indonesia, however, the limited evidence available suggests that women have not participated heavily in public works projects, for three reasons. First, 70 percent of the factory workers (both male and female) surveyed by AKATIGA (1999) were reluctant to join the Padat Karya pro- gram. They felt the work provided was not appropriate for factory work- ers and preferred instead to start their own businesses. Women inter- viewed in Java further felt that the tasks involved in the public works projects were not appropriate for women and proposed instead work in- volving "women's skills" (including cleaning community facilities such as schools and parks, sewing, and cooking). A final factor deterring women in Indonesia was the location of the projects, which were implemented in areas far from people's homes." A local initiative in Surabaya did have 406 EAST ASIAN LABOR MARKETS some success in involving women. The local government negotiated with businesses and workers, encouraging local firms that had not been seri- ously affected by the crisis to hire unemployed factory workers as daily workers. Although the program was not targeted by gender, the main ben- eficiaries were women. In Korea one disincentive to incorporating women and young workers in public works is that compensation on the projects is 50 percent higher than the minimum wage. As a result the projects attract new participants from outside the labor force and draw workers away from small busi- nesses. In fact the higher the project wage offered, the less likely that women and young workers will obtain jobs. Some schemes have been tar- geted specifically to women and young workers, and women's participa- tion in these projects increased markedly between 1998 and 1999. Some schemes are specifically designed to provide women from poor households with short-term employment opportunities in the areas of social services and childcare. The government also specifically subsidizes new jobs (in existing firms) for women who are household heads, subsidizing wages by one-third to one-half. Another scheme provides internships for unem- ployed college graduates-for example setting up electronic network sys- tems in public institutions or working in private companies. However companies are permitted to hire new subsidized workers only if they are not laying off workers, a restriction that substantially limits the number of potential employers. In Thailand, where women's labor market participation is high, women also undertake heavy agricultural work. For this reason there may be fewer barriers to women's participation in public works projects. The majority of wage-support policies, including severance pay and unemployment insurance, are biased in favor of prime-age male workers. Severance pay, in particular, favors workers who have worked steadily in permanent jobs, those who have been employed in large firms (small firms can be exempt from the scheme), and those who have worked in the for- mal sector. Unemployment insurance tends to have similar effects. Link- ing unemployment insurance payment locations to job information cen- ters, as in Korea, is likely to make the centers more useful and available to males who have worked in the formal sector, since they are the ones eli- gible for payments. Minimum wages are also biased against vulnerable groups, unless separate lower wages are set for these groups. Both Canada and the United States, for example, have separate minimum wages for young workers. And finally wage subsidies of the kind implemented in Korea benefit mostly men. In Korea the smallest firms (which employ a THE AFTERMATH OF THE ASIAN FINANCIAL CRISIS 407 disproportionate number of women) are too financially risky to benefit from wage subsidies. Lessons from Other Regions We focus here on the experience of other developing countries, particu- larly those in South Asia and Latin America, with some reference to East European economies. There are also many lessons to be learned from the OECD experience that are more applicable to Korea than to the other Asian countries, given current incomes levels (Betcherman and others 1999). We focus more on income support policies than on labor market information and training, since more information is available on the effect of income support on vulnerable groups. Public Works Developing countries have used public works widely as income-support measures. The Maharashtra Employment Guarantee Scheme in India em- ployed roughly half a million workers, or around 2.5 percent of the labor force, each month between 1975 and 1989. The Bangladesh Food for Work schemes created 100 million person-days of employment per year in 1987-8 (Ravallion 1991). Latin American countries have implemented many public works schemes (Marquez undated). In Argentina, for instance, large public works projects employed up to 9 percent of the labor force. The majority of these workers participated in rwo large schemes, one operated by the social se- curity agency and state governments (Programa de Asistencia Solidaria), the other by local governments and nongovernmental organizations (NGOs) (Programa Trabajar). In Chile during the 1970s and 1980s, public works employed up to 10 percent of the labor force during the painful adjust- ment period, although these projects have now disappeared. In Mexico about 5 percent of the labor force has been included in two programs: the Program to Maintain Rural Roads [Programa de Conservacion de Caminos Rurales], financed by the federal government and implemented by state and local governments, and a similar but smaller program designed to improve social infrastructure (Marquez undated; Marquez and Martinez 1998). In Bolivia the Emergency Social Fund helped ease the plight of the poor during the adjustment that followed the adoption of the New Eco- nomic Policy in 1985 Uorgensen, Grosch, and Schacter 1991). At the end of 1988 about 2 percent of the labor force were participating in the fund. 408 EAST ASIAN LABOR MARKETS LESSONS FROM LATIN AMERICA. Discussing the Latin American experi- ence, Marquez (undated) notes that "labor-intensive public works have been the tools of choice to deal with economy-wide shocks" (p. 16). In fact Latin American governments spend more on public works than on other mechanisms for income support. For this reason observers have been able to draw a number of lessons about such projects. Marquez, for in- stance, cites three characteristics of good programs: * First, they are financed by the central government but implemented by a local agency-either a local government (as in Mexico and Ar- gentina) or local NGO (as in Bolivia). The relationship between the central government and the executing agency is well structured and appropriate to the political structure of the country.12 * Second, wage levels and selection criteria for beneficiaries are set cen- trally, but local agencies do the actual selecting. Community oversight ensures that there is no favoritism and is complemented with a non- trivial expenditure of central resources on monitoring. * Third, the wage level is lower than the prevailing market wage to en- sure self-targeting, even at the cost of stigmatizing participants. Marquez also counsels against the danger of allowing public works to become entrenched and lose their important countercyclical properties. In Bolivia the Emergency Social Fund was deliberately cut off after four years of operation, despite the expense. Generating a pipeline of adequate pro- posals takes about a year, so that programs may be cut off just as they are beginning to run smoothly. LESSONS FROM OTHER REGIONS. Ravallion (1999) makes several recom- mendations for successful projects. First, he maintains that work should not be rationed (the Maharashtra model) but notes that if rationing is necessary, it should be done geographically as well as temporally. He also argues that a project's labor intensity (the share of labor in total project cost) must be at least as great as the average for similar projects in the same setting. Finally he counsels that projects should be aimed at poor areas to ensure that the assets created are of maximum value to poor people. In the event that they are not, then local cofinancing is required. Ravallion further calculates benefit-cost ratios for public works in both a middle-income country (based loosely on Argentina) and a low-income country (based on South Asia). He suggests that under plausible assump- tions, $1 of government spending in the middle-income country leads to THE AFTERMATH OF THE ASIAN FINANCIAL CRISIS 409 a transfer of $0.20 to the poor. He assumes that the labor share of the cost of public works is 33 percent, that the wage is low enough to prevent leak- age to the nonpoor in employment, and that foregone income by partici- pants is $0.40 per $1.00. The same calculation for the low-income coun- try suggests that $1 of government expenditure leads to $0.28 for the poor. This calculation is based on a labor share in the project cost of 50 percent, a leakage of 25 percent of the employment created to the non- poor, and forgone earnings of $0.25 on $1.00. At first sight, then, public works are hardly more attractive than uni- form untargeted transfers (assuming that poverty rates are 20 percent and 50 percent in the two countries, respectively). Adding future returns to the assets created raises the benefits to the poor to around $0.40 per $1.00 of government spending in both low- and middle-income countries. This figure makes public works more attractive than untargeted transfers in the richer country but still not as attractive in the poorer country. One possi- ble reason for continuing to use public works in low-income countries may be that external funds cannot be mobilized for untargeted income transfers. Geographically targeted public works projects are also likely to be politically more acceptable than similarly targeted cash transfers. PUBLIC WORKS AND VULNERABLE GROUPS. Public works are generally successful at reaching unskilled workers but less so at reaching women. The effect on migrants is more difficult to gauge. Ravallion (1991) cites three studies of the Maharashtra scheme and one of the Bangladesh Food- for-Work scheme showing that these programs successfully reach the poor. One survey of Maharashtra participants found that 90 percent remained below the poverty line even when calculations of their income included project earnings (around 49 percent of the overall population lives below this level). Ravallion's estimates suggest that 60 percent of the participants in the Bangladesh project came from the poorest 25 percent of house- holds. Jorgensen, Grosh, and Schacter (1991) do a counterfactual analysis for Bolivia and find that 25 percent of the workers in the Emergency So- cial Fund would have been in the poorest decile without it. With it none were in this decile, because mean monthly earnings, while below average wages for construction work, were actually above the average for low- skilled construction workers with the same characteristics and only slightly below the average for all workers in the population. Public works may reach poor male workers, but experience suggests that unless projects are specifically designed for women and youth, these groups will not benefit. In Bolivia, for instance, 91 percent of the program 410 EAST ASIAN LABOR MARKETS participants were men, and 93 percent were heads of households. The Latin American projects tend to service prime-age men who are household heads, in part because these projects are designed to employ household heads during crises. A few schemes have been aimed specifically at women and youth. Two examples are the Mexican rural roads program for youth and a small program for female heads of household in Argentina, the Pro- gram for Community Service [Programa de Servicio Comunitario]. In Latin America programs aimed at women and youth have focused pri- marily on training and labor market insertion, including Chile Joven, Proyecto Joven (Argentina), and PROJOVEN (Peru). Programs in Chile and Costa Rica have also been designed to promote employment for those with handicaps. In general, though, public works leave out those who are poor but who have restrictions on their ability to work. The South Asian schemes have had comparatively high levels of female participation. In Maharashtra men and women have participated at around the same rate. In Bangladesh, where employment has been created primar- ily for men, some projects have specifically targeted women workers. The Maharashtra scheme has encouraged women by siting projects near vil- lages, not discriminating by gender in wages (piece rates are often the norm), and by providing childcare (usually by hiring women to provide it). Because few studies are available that provide poverty profiles or that examine the location of projects, information on migrants and public works projects is almost nonexistent. In Bolivia targeting by location was not based on formal data; rather, administrators would go out and look at sites to see if they were in poor neighborhoods. Given the prevalence of this type of targeting, the poorest and most remote locations-where mi- grants often live-are likely to be left out of public works projects. Wage Subsidies Wage subsidies have not been widely used in developing countries, partly because of the difficulties of monitoring and enforcement (Marquez un- dated). Governments can use subsidies to promote the employment of particular subgroups (women, youth, and former combatants, for in- stance) by creating promotional contracts that allow firms to hire workers without the usual social benefits. However, unions tend to oppose these contracts, which they see as promoting less desirable forms of employ- ment. Argentina adopted a voucher system that eased the tax burden on firms hiring unemployed workers in the target groups (workers turned the THE AFTERMATH OF THE ASIAN FINANCIAL CRISIS 411 vouchers over to employers, who submitted the paperwork to the govern- ment). Firms willing to hire particular groups of workers could also get tax rebates. Canadian employers using "coop" students (students in programs combining work and study) are eligible for tax rebates. Korea's experience with wage subsidies illustrates three potential pitfalls. First, as we have seen, only firms that were not laying off workers could hire subsidized workers, a condition that substantially reduced the number of eligible firms. Second, the wage subsidies created large displacement ef- fects. One survey suggests that 78 percent of jobs would have been main- tained even without the subsidies (chapter 3). Third, because it is politi- cally effective to tie wage subsidies to apprenticeship schemes and training programs, wage subsidies in Korea disproportionately benefited men. Severance Pay and Unemployment Insurance Severance pay is a much more widespread institution than unemployment insurance in developing countries, though not in the industrial world. Both schemes have their merits and their problems. Severance pay exists in much of Asia, Eastern Europe, and Latin America, although it is re- stricted to the formal sector. Severance schemes are fairly straightforward. Employers withhold a specific amount from employees' paychecks, de- positing it in a fund that is available to the worker upon dismissal. If not well designed, however, these schemes can generate undesirable incentives. In Brazil, for instance, the very high fluidity of the labor market (which inhibits firm-specific training) is in part the result of workers' ability to ac- cess their severance funds (Camargo 1997). 13 As we have seen, in Korea severance pay accumulates with seniority, making it difficult to dismiss se- nior employees. But in other countries (like Brazil) severance pay creates incentives for employees to force their own dismissal. Further, regulations regarding severance pay assume that firms (and the state) have the capa- bility to maintain records. And as we have noted, severance pay tends to benefit prime-age males more than other groups, since these men are more likely to work in large firms and maintain steady employment. Severance pay can be inferior to unemployment insurance in a more mature economy. Unemployment insurance covers a broader range of workers, though it also requires the state to keep detailed records and monitor firms. Unlike severance pay systems, which provide individual benefits, unemployment insurance typically involves some pooling of risk. In the United States unemployment insurance schemes use experience rat- 412 EAST ASIAN LABOR MARKETS ing, levying higher premiums on high-risk industries or firms. Canada of- fers more generous benefits (less stringent eligibility requirements and a longer benefit period) in regions of relatively high unemployment. This kind of fine tuning helps discourage migration to those regions with the best employment prospects. In Eastern Europe firms have switched from severance pay schemes to unemployment insurance but have not abolished severance pay for exist- ing workers. Dismissed workers can therefore access both types of bene- fits, although some workers (specifically in the Czech and Slovak Re- publics) exhaust their severance benefits before claiming unemployment insurance (Ham, Svejnar, and Terrell 1998). All the Central and East Eu- ropean countries had established unemployment insurance schemes by the end of 1991, with benefits that were initially generous compared with those in Western Europe. But these benefits were gradually reduced be- cause of budget considerations. In the Czech Republic, for instance, indi- viduals who had not worked but had in the last three years taken care of children or sick relatives, left the military, or spent time in prison were ini- tially eligible. In 1991 benefits lasted one year, but in 1992 that period was reduced to six months. Ham, Svejnar, and Terrell (1998) find that unemployment insurance has only a moderate disincentive effect on workers. A 10 percent increase in benefits raises the time the average worker is unemployed by only 0.61 weeks in the Czech Republic. This result indicates an elasticity of duration with respect to benefits of 0.34; the corresponding elasticity in Slovakia is 0.14. Even drastic experiments, such as switching individuals from recip- ient to nonrecipient status, increase duration only by 10-20 percent. The main factor in how long workers are unemployed is labor demand condi- tions. The authors conclude that there is latitude for governments to pro- vide social protection without unduly affecting the incentive to work. In Latin America experience with unemployment insurance is limited. The length of time unemployed workers can receive benefits is typically restricted to four months, and the number of beneficiaries is small because the schemes are restricted to those with full-time employment in the for- mal sector. Argentina has only 100,000-125,000 recipients at a time, and Brazil, which has the largest scheme, has only 300,000-400,000. Vene- zuela never implemented the legislation it enacted on unemployment in- surance, and Mexico permits workers only to draw against the pensions they have accumulated. In Latin America, as in Eastern Europe, unem- ployment insurance tends to be a complement to severance pay rather than a substitute for it (Marquez undated). THE AFTERMATH OF THE ASIAN FINANCIAL CRISIS 413 Retraining Schemes Betcherman and others (1999) provide a comprehensive overview of re- training schemes in OECD countries. They conclude that more scientific evaluations (those with controls) belie the overly optimistic nonscientific assessments of training programs for workers who have lost jobs in mass layoffs. They also conclude that retraining programs do little to raise ei- ther the long-term earnings or employment prospects of such workers. The authors conclude that it is more cost-effective to provide job search assistance for these workers, especially since the costs of such assistance may be as little as one-tenth of the costs of retraining. Training and job insertion programs may be more cost-effective for young workers, however. Chile Joven is one example of a successful reform of training programs for youth. The previous compulsory system of train- ing in Chile (with mandatory payroll taxes) had led to complaints that public sector training programs were not well suited to the needs of the labor force. Under Chile Joven governments provide scholarships for in- classroom training, with a three-month paid internship in a private firm (Marquez undated). The central government agency requests bids for projects, describing the content of courses and including a commitment form for firms willing to provide apprenticeships. This interaction with firms helps ensure that the training provided is relevant. Argentina and Peru have also followed this model. Brazil's scheme uses competitive bid- ding between the public and private sector, so that the national training institution must also bid. Some of the reforms of the Philippine training system (Esguerra, Balisacan, and Confessor 1999) have been made along similar lines, although the quantities of workers trained are much too low to be considered a significant response to the crisis. The success of the Mexican PROBECAT scheme (a job training and labor market insertion program) may also have been a liability. As partic- ipation rates rose, so did youth unemployment, raising concerns that the program was causing students to leave school in order to join (Marquez undated). Policy Considerations and Lessons from Experience From the information presented here, we draw three policy considerations and present some lessons regarding labor market interventions and vul- nerable groups. The policy considerations apply not just to vulnerable groups but to the labor markets of which these groups form an integral 4I4 EAST ASIAN LABOR MARKETS part. While it is true that some very vulnerable groups cannot be reached through the labor market interventions alone, effective targeting improves the chances that members of such groups will remain employable during -crisis periods. Policy Considerations Safety nets must be in place even in times of strong economic growth. At the onset of the financial crisis, the five Asian countries considered here were at different stages in the evolution of their labor markets and labor market institutions. One general characteristic was that public in- stitutions were not in place to provide social support during such a cri- sis, perhaps because these countries had experienced such extraordinary growth. With very little experience to back them up, the governments were then faced with the need to design and implement very rapid labor market interventions. Their responses tended to be short term, designed primarily to respond to the immediate needs of laid-off prime-age male workers who were household heads. Concerns about vulnerable groups (other than unskilled workers) were of secondary importance. These countries now have the opportunity to begin making some long- term changes in their labor market institutions. These institutions need to be able to meet the needs of an urbanized economy that faces the income variability associated with increased integration into the international economy. Marquez (undated) provides a useful agenda for reform in Latin America that is also applicable in Asia. * Up-to-date labor market information and the means to disseminate it are essential to effective policy responses during crises and beyond. During the Asian crisis the lack of up-to-date information on the dis- tribution of poverty hampered policy responses. Siting public works pro- grams in the neediest areas would have been easier if data on the distribu- tion of poverty had been available early in the crisis, for example. There is no substitute for timely analysis of such tools as labor force and income and expenditure surveys. In response to the crisis some governments un- dertook useful initiatives in this direction. For instance, the Economic and Social Development Board in Thailand began publishing a newsletter an- alyzing economic trends in the country. Indonesia initiated the Social Monitoring and Early Response Unit, which includes NGOs that post in- formation on an Internet website. THE AFTERMATH OF THE ASIAN FINANCIAL CRISIS 415 * In order to prevent inequalities from worsening during crises, social safety nets must make special provisions for vulnerable groups. The crisis in Asia worsened inequalities among economic groups. Al- though the crisis initially affected the formal sector (leading to employ- ment losses, including in exporting firms), it was transmitted quickly to the informal sector in at least three of the five countries studied here. In both Korea and Thailand the crisis exacerbated inequalities by reversing the trends toward narrower skill and age differentials (Korea) and more re- gional equality (Thailand). This result is consistent with what is known about crises in Latin America, but it is very unlike the situation in sub- Saharan Africa, where crises and structural adjustment in the 1980s nar- rowed urban-rural inequality Uamal and Weeks 1987). This result also suggests a hypothesis about the degree of market inte- gration that can benefit vulnerable groups. When these groups are not in- tegrated into the national market, they will not be as adversely affected by crises-but they will also not benefit as much from growth. However while vulnerable groups that are tied into the market can benefit dispro- portionately from growth, they suffer relatively more during crises. This fact lies behind our recommendation that policies concerning safety nets should take vulnerable groups particularly into account. Lessons from Experience Experience provides us with policy lessons to be learned about specific labor market interventions. We have discussed some of these experiences, which provide lessons in best practices for various activities. Public works need to be well designed, encouraging suitable proposals, set- ting appropriate wages, and targeting vulnerable groups such as women, young workers, and migrants. Public works have been a mainstay of income support in three of the countries studied here. Since none had recent experience with such pro- grams, there were inevitable difficulties in design. In Thailand for example, efforts to involve community organizations and NGOs in proposing proj- ects resulted in very few suitable proposals during the first year (Arya 1999). In such cases central agencies need to be more proactive, develop- ing "templates" for projects and actively encouraging suitable proposals. 4I6 EAST ASIAN LABOR MARKETS Public works programs in Indonesia were not successful in attracting women participants. To do so these programs must keep wages moderate in order not to attract men who are self-employed. Setting project wages below the minimum wage may not be appropriate, however, if minimum wages are already very low. Nevertheless, wages that are too high will dis- place workers rather than reaching those most in need of work. In addi- tion projects must provide work that is considered appropriate for women in the locale, and project sites must be accessible and childcare available. Finally, effective monitoring helps ensure that a project runs well. * Wage subsidies should be temporary and preferably directed to specific groups of workers. Of the countries discussed here, only Korea used wage subsidies to any extent. International experience (for example, in Argentina) with this measure suggests that temporary, targeted subsidies are best used in con- junction with training and labor market insertion programs. Otherwise deadweight losses will be high. - Because severance pay favors older male workers in permanent jobs, unem- ployment insurance schemes are an important addition to social safety nets, especially for vulnerable groups. The crisis exposed the weakness of severance pay schemes, which tend to cover few workers and are of little use when firms go bankrupt. Among the countries discussed here, only Korea has an unemployment insurance scheme in place. Although it had been expanded before the crisis, it had not been operating long enough to cover many workers when the crisis came. Reforming severance pay schemes or implementing unemployment insurance is a long-term effort, but it is one that will provide large bene- fits in future crises and recessions. Only the relatively high-income Asian countries are likely to follow the Latin American model and shift to unemployment insurance systems. Ex- perience with shifting out of severance pay schemes and into unemploy- ment insurance is limited. The East European and Latin American coun- tries, and even Canada and the United States (for those workers whose contracts include severance pay), have simply imposed one system on top of the other. Nevertheless, having only one mandatory system offers cer- tain advantages. (Although having two systems can provide workers with benefits for a longer period, it can also worsen income distribution.) A pe- THE AFTERMATH OF THE ASIAN FINANCIAL CRISIS 417 riod of economic recovery, with a young and growing labor force, can be used to make the shift from one system to the other. From the point of view of vulnerable groups, unemployment insurance is certainly prefer- able, as long as coverage is reasonably broad. Training programs are valuable for specific vulnerable groups, particu- larly women, andforyoung people needing assistance in labor market in- sertion, but most existing programs need to be reformed to be effective. Here Latin America offers useful experience (with similarities to the German dual model) of shifting to systems of training provision that in- volve competition among providers and encourage strong links with em- ployers. Government agencies still retain an important role in setting stan- dards, however. These again are the kinds of reforms that cannot be undertaken during a moment of crisis but are important to have in place for the future. Several countries brought in training programs as a com- ponent of crisis response (internships for high school and college gradu- ates in Korea, for instance, and skill development for the unemployed and new labor market entrants in Thailand). Training is not especially useful for mass layoffs, and alone it is not a cost-effective substitute for formal education for young workers. Finally we repeat one important point. Not all vulnerable groups can be reached by labor market interventions. Some groups, such as the very low-skilled working poor who will never earn enough to be able to save for retirement, single parents with young children, and those with chronic health problems that make it difficult for them to work, must be reached through other social measures. Notes 1. The authors work with the Centre for International Studies at the Univer- sity of Toronto. Their analysis in this chapter (especially the first two sections) draws heavily on a number of country studies prepared as background, namely Arya (1999), Esguerra and others (1999), Islam and others (I 999), Kang (I 999), Keum (1999), Kim (1999), Mahmood (1999), Mansor and others (1999), and Shin (1999). The authors would also like to thank Suying Hugh for excellent re- search assistance. 2. Again, Thailand had lower secondary enrollments than per capita income alone would predict. Observers frequently comment on this issue, as this fact is generally expected to make Thailand's transition from low-wage to high-skill, high-productivity manufacturing more difficult. 4I8 EAST ASIAN LABOR MARKETS 3. Data are not available for Indonesia. 4. For data on this transition in Asia, see Horton (1996). 5. Personal communication with K. Beegle. 6. The most commonly accepted explanation for the widening differentials in industrial countries is that changing technology has given skilled workers the ad- vantage in the labor market-an advantage that is buttressed by the skill intensity of the traded goods sector. Widening differentials would also be the norm in de- veloping countries if technology were the main factor driving wages. Trade effects that narrow these differentials would offset this trend, however. 7. In Asian formal sectors, large firms employ a substantial number of tempo- raries along with a tenured core of workers who enjoy job security and other ben- efits. In the Indian textile industry such laborers, or badlis, account for 25-30 per- cent of the labor force. This type of labor structure has also been a continuing part of the Japanese industrial scene. 8. In fact the gap in income levels among the country's regions is enormous. In 1998 per capita income in Bangkok and the surrounding area (the country's richest region) was four times the average for the northeast (the country's poorest region). Part of the reason for these income disparities is the relatively low ur- banization rate of poorer regions, since incomes tend to be higher in urban areas in all regions. 9. Some observers argue that the interventions of Latin American governments in labor markets has in part been responsible for the relatively large informal labor force there (World Bank 1996). 10. See Horton (1996) for a discussion of labor market legislation affecting women in Asian countries. 11. AKATIGA undertook a second study, which found that women are af- fected negatively when husbands are laid off. Women must continue working to supplement the household's income, but husbands do not increase their share of domestic chores or childcare. 12. Jorgensen, Grosh, and Schacter (1991) provide specific lessons for good program design from Bolivia, a country with limited local government capability. 13. But Brazil's very slow judicial mechanisms for enforcement often lead workers to accept less severance pay than is their right in order to get some money without a long wait. The slow judicial system also leads both employers and em- ployees to prefer informal rather than formal contracts, undermining statutory mandates for severance pay. References Abubakar, Syarisa Yanti. 1999. "Migrant Labor in Malaysia: Impacts and Im- plications of the Asian Economic Crisis. Malaysia Institute of Economic Re- search. Mimeo. AKATIGA. 1999. "Factory Unemployment: Gender Issues and Impacts." Jakarta. Mimeo. THE AFTERMATH OF THE ASIAN FINANCIAL CRISIS 4I9 Arya, Gosah. 1999. "Economic Crisis, Employment, and the Labor Market in Thailand." Paper prepared for the joint World Bank-International Labour Or- ganization seminar on Economic Crisis, Employment and Labor Markets in East and Southeast Asia, Tokyo, October. Japanese Ministry of Labor, Japan Institute of Labor. Assad, R., and S. Commander. 1994. "Egypt." In S. Horton, R. Kanbur, and D. Mazumdar, eds., Labor Markets in an Era of Adjustment, vol 2, Case Studies. Economic Development Institute, Development Studies series. Washington, D.C.: World Bank. BPS (Badan Pusat Statistik). 1998. The Labor Force Situation in Indonesia, August 1998. Jakarta. Bardan, K. 1994. "Gender and Labor Allocation in Structural Adjustment in South Asia." In S. Horton, R. Kanbur, and D. Mazumdar, eds., Labor Mar- kets in an Era ofAdjustment, Vol. 1. Washington, D.C.: World Bank. Beegle, K., E. Frankenberg, and D. Thomas. 1999. "Economy in Crisis: Labor Market Outcomes and Human Capital Investments in Indonesia." Santa Monica, CA: RAND. Mimeo. Betcherman, G., A. Dar, A. Luinstra, and M. Ogawa. 1999. "Active Labor Mar- ket Policies: Policy Issues for East Asia." Paper prepared for the joint World Bank-International Labour Organization seminar on Economic Crisis, Em- ployment and Labor Markets in East and Southeast Asia, Tokyo, October 10-13, 1999. Japanese Ministry of Labor, Japan Institute of Labor. Camargo, J. M. 1997. "Brazil: Labor Market Flexibility and Productivity, with Many Poor Jobs." In E. Amadeo and S. Horton, eds., Labor Productivity and Flexibility. Houndmills, UK: MacMillan. Collier, P. 1994. "A Theoretical Framework and the Africa Experience." In S. Horton, R. Kanbur, and D. Mazumdar, eds., Labor Markets in an Era of Adjustment, Vol. 1. Washington, D.C.: World Bank. Cox-Edwards, A., and J. Roberts. 1994. "The Effects of Structural Adjustment on Women in Latin America." In S. Horton, R. Kanbur, and D. Mazumdar, eds., Labor Markets in an Era of Adjustment, Vol. 1. Washington, D.C.: World Bank. Esguerra, J., A. Balisacan, and N. Confessor. 1999. "Philippines Country Paper." Paper prepared for the joint World Bank-International Labour Organization seminar on Economic Crisis, Employment and Labor Markets in East and Southeast Asia, Tokyo, October 23-25. Japanese Ministry of Labor, Japan In- stitute of Labor. Frankenberg, E., D. Thomas, and K. Beegle. 1999. "The Real Costs Of Indone- sia's Economic Crisis: Preliminary Findings from the Indonesia Family Life Surveys." Labor and Population Program Working Paper Series 99-04. Santa Monica, CA: RAND. Ham, J., J. Svejnar, and K. Terrell. 1998. "Unemployment and The Social Safety Net During Transitions to a Market Economy: Evidence from the Czech and Slovak Republics." American Economic Review 88:1117-42. 420 EAST ASIAN LABOR MARKETS Horton, S. 1996. Women and Industrialization in Asia. London: Routledge. ILO (International Labour Office). Various years. Yearbook of Labor Statistics. Geneva. Islam, R., with G. Bhattacharya, S. Dhanani, M. Iacono, F. Mehran, W. Mukhopadhyay, and P. Thuy. 1999. "Indonesia Country Paper." Paper pre- pared for the joint World Bank-International Labour Organization seminar on Economic Crisis, Employment and Labor Markets in East and Southeast Asia, Tokyo, October. Japanese Ministry of Labor, Japan Institute of Labor. Jamal, V. and J. Weeks. 1987. "Rural-Urban Income Trends in Sub-Saharan Africa." World Employment Program Labor Market Analysis and Employ- ment Planning Working Paper 18 (WEP2-43/WP. 18). Geneva: International Labor Office. Jorgensen, S., M. Grosh and M. Schacter. 1991. "Easing the Poor through Eco- nomic Crisis and Adjustment." Regional Studies Program Report 3, Latin America and the Caribbean Technical Department. Washington D.C.: World Bank Kakwani, N. and H.H. Son. 1999. "Economic Growth, Inequality, and Poverty: Korea and Thailand." Sydney, Australia: University of New South Wales. Mimeo. Kang, Soon-Hie. 1999. "Vocational Education and Training in Korea." Paper prepared for the joint World Bank-International Labour Organization seminar on Economic Crisis, Employment and Labor Markets in East and Southeast Asia, Tokyo, October. Japanese Ministry of Labor, Japan Institute of Labor. Keum, Jaeho. 1999. "The Financial Crisis and Labor Market Services in Korea." Paper prepared for the joint World Bank-International Labour Organization seminar on Economic Crisis, Employment and Labor Markets in East and Southeast Asia, Tokyo, October. Japanese Ministry of Labor, Japan Institute of Labor. Kim, Dong-Heo. 1999. "Employment Creation/Maintenance and Income Sup- port Activities in Korea." Paper prepared for the joint World Bank-Interna- tional Labour Organization seminar on Economic Crisis, Employment and Labor Markets in East and Southeast Asia, Tokyo, October. Japanese Ministry of Labor, Japan Institute of Labor. Lee, Ju-Ho, and Dae-Il Kim. 1999. "Crisis and Reforms in Koreds Labor Market." Seoul: KDI (Korean Development Institute) School of International Policy and Management. Mimeo. Lee, Ju-Ho, and Jong-Il You. 1999. "Globalization and Social Policy: the Case of South Korea." Paper prepared for Conference on Globalization and Social Pol- icy at the Centre for Economic Policy Analysis, New School for Social Re- search University in Seoul, October. Mahmood, Moazam. 1999. "Thailand Country Paper: A Macro Appraisal of the Economy and the Labor Market: Growth, Crisis, and Competitiveness." Paper prepared for the joint World Bank-International Labour Organization seminar THE AFTERMATH OF THE ASIAN FINANCIAL CRISIS 421 on Economic Crisis, Employment and Labor Markets in East and Southeast Asia, Tokyo, October. Japanese Ministry of Labor, Japan Institute of Labor. Mansor, N., T. E. Chye, A. Boerhanordin, E Said and Saad Said. 1999. "Malaysia Country Paper." Paper prepared for the joint World Bank-International Labour Organization seminar on Economic Crisis, Employment and Labor Markets in East and Southeast Asia, Tokyo, October. Japanese Ministry of Labor, Japan Institute of Labor. Marquez, G. Undated. "Labor Markets and Income Support: What Did We Learn from the Crises?" Washington, D.C.: Inter-American Development Bank. Mimeo. Marquez, G., and D. Martinez. 1998. Programas de empleo e ingreso en America Latinay el Caribe. Lima, Peru: Inter-American Development Bank and the In- ternational Labour Office. Mazumdar, D. 1999. "Trends in the Labor Market: The Last Quarter in the Per- spective of the 1998 Experience." University of Toronto, Center for Interna- tional Studies. Mimeo. NESDB (National Economic and Social Development Board). 1998. Indicators of Well-Being and Policy Analysis 2(4). Bangkok. ____. 1999. Indicators of Well-Being and Policy Analysis 3(1). Bangkok. NSOK (National Statistics Office, Korea). Various years. Annual Report on the Economically Active Population Survey. Seoul. NSOP (National Statistics Office, Philippines). 1999. Labor Force Survey. Income and Employment Statistics Division, Household Statistics Department, Manila. Available at http://www.census.gov.ph NSOT (National Statistics Office, Thailand). 1998. "Report of the Labor Force Survey, Whole Kingdom, Round 5." Bangkok: Office of the Prime Minister. Ravallion, M. 1991. "Reaching the Rural Poor through Public Employment: Ar- guments, Evidence and Lessons from South Asia." World Bank Research Ob- server, 6:153-175, 1991. ___ . 1998. "Appraising Workfare." WorldBank Research Observer 14:31-48. Scarpetta, S. 1998. "Labor Market Reforms and Unemployment: Lessons from the Experience of the OECD Countries." Working Paper 382. Washington, D.C.: Inter-American Development Bank, Office of the Chief Economist. Shin, Donggyun. 1999. "Statistical Overview of Current Labor Market Condi- tions and Trends." Paper prepared for the joint World Bank-International Labour Organization seminar on Economic Crisis, Employment and Labor Markets in East and Southeast Asia, Tokyo, October. Japanese Ministry of Labor, Japan Institute of Labor. Squire, L., and S. Suthiwart-Narueput. 1996. "The Impact of Labor Market Reg- ulations." World Bank Research Observer 11. TDRI (Thailand Development Research Institute). 1999. Social Impacts of the Asian Economic Crisis in Thailand, Indonesia, Malaysia and the Philippines. Bangkok. 422 EAST ASIAN LABOR MARKETS Thomas, D., E. Frankenberg, K. Beedle, and G. Teruel. 1999. "Household Bud- gets, Household Composition and the Crisis in Indonesia: Evidence from Longitudinal Household Survey Data." Paper presented at the Population As- sociation of America meetings, New York, March. World Bank. 1995. World Development Report 1995. Washington D.C. . 1996. World Development Report 1996 Washington, D.C. . 1998. World Development Report 1998/99. Washington, D.C. . 1999. Coping with the Crisis in Education and Health. Thailand Social Monitor, issue 2. Bangkok: World Bank Thailand Office. 1 O Social Dialogue and Labor Market Adjustment in East Asia after the Crisis DUNCAN CAMPBELL1 RELATIVELY UNDEVELOPED WORKER and employer orga- nizations and weak formal institutions for social dialogue in the labor market characterize much of East Asia. Both are a part of the legacy of ambivalence regarding freedom of association that persists in many coun- tries of the region. It is all the more noteworthy, then, that in the past two years a number of tripartite initiatives have emerged in East Asia. A more participatory approach is clearly evident in attempts to address the eco- nomic and social havoc wrought by the collapse of the baht in July 1997 and the contagious aftermath. In at least three ways, the emergence of this social dialogue represents a break with the past. First, the search for solutions has resulted in the creation of entirely new tripartite approaches in Cambodia, the Republic of Korea, Malaysia, the Philippines, Singapore, and Thailand. Second, some of the tri- partite initiatives have included government participation at the highest level. In Malaysia, for instance, the Prime Minister himself headed the tripartite mechanism set up to address the economic crisis.2 Third, in some instances, the agenda for dialogue has extended well beyond issues dealing solely with the labor market to include a range of macroeconomic policy choices. This chapter focuses on the most institutionalized modalities for dia- logue-those associated with a country's formal industrial relations sys- 423 424 EAST ASIAN LABOR MARKETS tem. We look at the East Asian countries most affected by the recent fi- nancial crisis: Korea, Malaysia, the Philippines, and Thailand. Good Governance Equals Participation-A New Equation? Why does the East Asian economic crisis seem to have increased not only interest in but the general willingness to experiment with participatory ap- proaches? The answer appears to lie in the political fallout from the crisis. The effect has been substantial, most obviously in Indonesia, where crony- ism and an autocratic polity fell victim to popular protest. In fact, how- ever, in all three of the countries most affected by the crisis, political trans- formation has been widespread. Popular resentment over the deterioration of the Korean economy galvanized support for the presidential candidacy of the country's best-known opposition leader and democracy advocate. In Thailand the crisis coincided with the culmination in August 1997 of a process of constitutional reform characterized by an unusually high degree of popular support and participation. The cornerstone of the Thai Con- stitution itself is an emphasis on openness, participation, and the consol- idation of democracy. Even prior to the baht's collapse, moreover, Thai- land's Eighth Development Plan had veered away from previous plans with its emphasis on community and "people-centered development." (NESDB 1997) Malaysia's unorthodox response to the crisis extends well beyond capi- tal controls and exchange rate policy and cannot be fully appreciated with- out reference to the rise of a credible political opposition. This opposition was personified by a former deputy prime minister and finance minister whose policy views were considerably more orthodox than the govern- ment's. While the press noted concerns over the resurgence of cronyism in the aftermath of the Philippine presidential election, there can be little doubt of the president's populist appeal, particularly among the poor. In both Malaysia and the Philippines, moreover, response to the crisis has raised the stature of the labor federations at the national level. In Malaysia this development is most evident in the inclusion of the country's largest worker organization, the Malaysian Trade Union Congress (MTUC), in decisionmaking at the national level. In the Philippines it is evident in the conclusion of the national tripartite social accords. And in Indonesia and Cambodia the ratification of the core standards of the International Labour Organization (ILO) on freedom of association signals the radically altered political climate. SOCIAL DIALOGUE AND LABOR MARKET ADJUSTMENT 425 To varying degrees, but consistently, political trends in the countries af- fected by the crisis underscore a rising demand for openness and popular participation in economic and social policy choices. While pressure for de- mocratic reform has long been a secular trend in the region, the crisis ap- pears to have given it new impetus. The increase in tripartite initiatives is one reflection of this greater momentum. The demand for greater popular participation is a reaction to the ori- gins of the crisis itself, since the weakness or absence of democratic mech- anisms is generally seen as one of the causative factors. Thus the crisis has been the catalyst for changes such as an ambitious project of legal reform in Thailand. The link between political institutions and economic and so- cial outcomes is worth probing more deeply, for it implies that economic and social costs may be associated with the absence of well-functioning democratic mechanisms. This point is debatable, of course. The Asian "miracle" occurred without strong democratic mechanisms. But a com- pelling link exists between democracy and economic performance over the long run, so that strengthening channels for dialogue may well have eco- nomic benefits. Here, then, is a view that fortifies what had been a purely normative trend, and justified as such (democratic participation as a hu- man right) with a utilitarian motivation. Beyond its moral justification, participation also appears to pay economic dividends. The Economics of Dialogue When commentators on the Asian financial crisis cite as one of its causes a lack of transparency or accountability, they are evoking underlying prob- lems, one of which is the inadequacy of information. For example, because international financial markets are dogged by the problem of asymmetri- cal information (that is, the extreme difficulty of obtaining accurate and timely information on the risks and rewards of distant investments), for- eign portfolio investors failed to see the crisis coming. Because informa- tion on business dealings was often closely held and shared only among a select, few-one of the defining features of "crony capitalism"-investors were further prevented from obtaining a true picture of just how risky some of the soon-to-be-nonperforming ventures were (Lee 1999). Markets, of course, run on information, and the lesson reinforced by the crisis is that information conveys economic benefits, whereas igno- rance has economic costs. This principle underpins the economic value of mechanisms for dialogue. In fact, when looked at in narrowly economic 426 EAST ASIAN LABOR MARKETS terms, dialogue or participation through "voice mechanisms" offers two main advantages. First, it is a tool for solving the problem of asymmetri- cal information. And second, it is a mechanism through which commit- ment is instilled and resistance minimized. Each of these advantages is briefly reviewed below. The Role of Information in the Quality of Decisionmaking One economic value of dialogue lies in the fact that dialogue generates in- formation, and information improves the quality of decisionmaking, in- creasing the likelihood that decisions will be successfully implemented. Both improved quality and better implementation are likely to increase efficiency, since bad decisions are costly. The problem of assymetrical in- formation means that these benefits are rarely available in the absence of well-functioning mechanisms for dialogue. Asymmetries arise from the probability that no one person or interest group is likely to possess all the relevant information needed to make complex decisions. More informa- tion is likely to become available when enough concerned, knowledgeable players are involved in the decisionmaking process.3 The field of industrial relations offers some of the best examples of just how useful (and also how difficult) information exchange can be. In mar- ket economies the interests of various stakeholders in the labor market are not identical. Industrial relations systems in market economies must ad- dress a common problem: while cooperation is essential to production, the interests of labor, management, and the state are not fully convergent. Many implications flow from this simple point. But as regards the first proposition-that information has economic value-we can draw one con- clusion: when interests are not thoroughly convergent, all parties are un- likely to share information equally. Nor are the parties likely to perceive the relevance of information identically. In consequence, some mechanism pro- moting dialogue needs to be devised for obtaining the information needed to improve a given decision-again, on the assumption that the greater the amount of relevant information, the better the decision is likely to be. Much thinking in industrial relations has always been directed to ques- tions of designing what we might call information- and commitment-gener- ating mechanisms. For example, one key question much in vogue in recent years is the construction of trust. Parties that do not trust each other have incentives to withhold information, and withholding information leads to bad decisions. A manager, for instance, may react to a high rate of unex- plained resignations by guessing that frequent unscheduled overtime is the SOCIAL DIALOGUE AND LABOR MARKET ADJUSTMENT 427 source of the problem. The manager accordingly curtails overtime use but observes no change in the rate of departures. Finally the manager learns from the employees themselves that a particularly abusive supervisor is the source of the problem. The moral of this simple tale is that an informed de- cision is better (for example, less costly) than a trial-and-error approach. The Role of Voice Mechanisms in Implementing Decisions4 The previous example illustrates a second economic advantage of partici- pation. Beyond getting the decision right is the matter of implementing it. Substantial empirical evidence shows that voice, or a well-functioning mechanism for information exchange, is a less costly alternative to other adjustment behaviors such as worker resignations. Evidence shows that the major advantage of unionized over nonunionized workplaces is that the voice mechanism available in unionized settings provides an effective means of channeling the facts required to make an informed decision. Ab- sent this channel, adjustment occurs in more costly ways-through em- ployee turnover, for example. But there are other potentially costly reac- tions to a decision in the form of noncompliance of various sorts, such as resistance to change and reductions in work effort, motivation, and com- mitment. Both theory and experience show that when the decisionmak- ing process internalizes the concerns of those who will be affected by the decisions, it results in more effective compliance. In other words partici- pation promotes compliance. These two main positive outcomes of the "economics of participation" have significance at both the macro- and microeconomic levels of social dialogue (box 10.1). At the societal level, in a world of increasing interde- pendence and rising external shocks, no one ministry or organization is likely to possess all the information required to make difficult decisions. The participation of specialized interest groups in the labor market has improved the decisionmaking process. Although employers and workers have separate interests, they also have separate information pools that can be invaluable to all parties. At the same time, despite episodes of social un- rest, it can be plausibly asserted that effective voice mechanisms have pre- vented disputes, as we will show. In the countries under review here, the recent financial crisis has increased communications and efforts at coop- eration. In the country most affected by the crisis, Indonesia, the collapse of nonparticipatory institutions and the inevitable institutional void their absence has created is related to the considerable social and political un- rest the crisis has occasioned. 428 EAST ASIAN LABOR MARKETS BOX 10.1 Social Dialogue and Macroeconomic Performance The ability to maintain macroeconomic stability in the face of turbulent external conditions is the single most important factor accounting for the diversity in post-1975 economic performance in the developing world. The countries that were unable to adjust their macroeconomic policies to the shocks of the late 1970s and early 1980s ended up experiencing a dramatic collapse in productivity growth. The countries that fell apart did so because their social and political insti- tutions were inadequate to bring about the bargains required for macroeconomic adjustment-they were societies with weak institutions of conflict management. In the absence of institutions that mediate conflict among social groups, the policy adjustments needed to reestablish macroeconomic balance are delayed, as labor, business, and other social groups block the implementation of fiscal and exchange rate policies . . . [To overcome policy paralysis and social divides], evidence shows that participatory political institutions, civil and political liberties, high-quality bureaucracies, the rule of law, and mechanisms of social insur- ance . . . can bridge these cleavages. Source: Rodrik 1999, pp. 2-3. Recent work by Rodrik (1999) supports the notion that democratic mechanisms of participation play an important role in improving eco- nomic performance and minimizing social instability. Democratic socie- ties perform better because they are able to resolve conflicts that otherwise impede adjustment to external shocks. Over the several years that globalization has been discussed, interest in labor-management cooperation has increased, largely because of its "fit" with the changing nature of competitive advantage. Two examples illus- trate this point. First, the most globally competitive firms in many indus- tries increasingly rely on the skills and decisionmaking abilities of their employees-that is, on the employees' knowledge and information.5 In fact, the greater the value placed on nonprice competitive assets such as speed, quality, and innovation, the greater the reliance on employees' ca- pabilities. Enterprises cannot hope to harness the skills of their employees, tap their knowledge, and institute a culture of commitment and adapt- ability to change without strongly developed participative mechanisms.6 Second, globalization has increased the rate at which enterprises need to SOCIAL DIALOGUE AND LABOR MARKET ADJUSTMENT 429 adapt to change in both technologies and product markets. This factor is what underlies corporate interest in flexibility. Businesses are unlikely to adapt to change, however (or at least are likely to resist it) in the absence of extensive mechanisms promoting communication and the exchange of information. Ultimately, however, it is design that matters to economic outcomes, not just the presence or absence of mechanisms for dialogue. Such mech- anisms affect more than distributional outcomes. They also affect measur- able indicators of performance-that is, variables such as the speed at which recovery occurs and the costs associated with reform. At present, there are grounds for guarded optimism about the awakening interest in social dialogue in the countries affected by the East Asian financial crisis. But these countries must still overcome a number of obstacles if social dia- logue is to play a more substantial role in the policy debate. The Institutional Framework for Social Dialogue In most Asian countries, weak trade unionism and underrepresentation of the labor force are the rule, for many reasons. Among these are the legacy of overt repression of labor organizations. All the countries under consider- ation have experienced periods of such repression, with the possible excep- tion of the Philippines (although efforts to curb labor organizations are not unknown even there). In all the countries under review, then, organized labor constitutes only a small minority of the total work force and an equally small minority of the modern, formal sector work force (table 10.1). The full brunt of the crisis, meanwhile, has of course fallen squarely on the individual-individual workers and their families and individual en- terprises. Absent collective agreements or social accords outside the enter- prise that commit it to a particular adjustment path, the individual bears the cost of adjustment. The simple fact in all countries is that social dia- logue over labor market adjustment during the crisis at the level at which that adjustment ultimately occurs has been largely absent-at least through any formal channels that would allow labor and management to interact. Weakly organized labor markets have become weaker still in Malaysia and Thailand as a result of the economic downturn. In Malaysia trade union membership in 1995 stood at 708,253, or approximately 9 percent of the work force. As a result of the financial crisis, union membership fell an estimated 1 percent (about 7,100 members), for a decline in union density of about .10 percent (Peetz and Todd 1999). Table 10.2 shows the trend in trade union membership for Thailand. 430 EAST ASIAN LABOR MARKETS TABLE 10.1 Estimated Union Densities (percent) Thailand 1.0 Philippines 11.0 Republic of Korea 11.2 Malaysia 9.0 Indonesia - not available. Source: ILO and author's estimate. Recent data for the Philippines show that trade union membership has continued to inch upward throughout the recession, despite the increase in registered unemployment. In 1996 membership stood at an estimated 3,611,000. By the end of the first quarter of 1998, that figure had grown to 3,670,000. In Indonesia, as well, new trade unions have emerged with the changed political climate. Coliective Bargaining Coverage Aggregate membership figures, of course, do not show the actual influence of trade unions on the economic and social choices made during adjust- ment. In all of the countries under review, however, collective bargaining takes place either solely or predominantly at the enterprise level. One mea- TABLE 10.2 Trends in Trade Union Membership in Thailand 1991 169,424 1992 190,142 1993 231,480 1994 242,730 1995 261,348 1996 280,963 1997 270,276 1998 265,982 Source: Ministry of Labor and Social Welfare. SOCIAL DIALOGUE AND LABOR MARKET ADJUSTMENT 431 sure of the extent of social dialogue at the microeconomic level, then, is the number of collective agreements. In Malaysia most trade unions are involved in collective bargaining, al- though as Peetz and Todd (1999) report, the number of workers covered by collective agreements varies from industry to industry. For example, well-established bargaining processes exist in the plantation industry at the enterprise level. But the construction industry registered only two collec- tive agreements covering just 1,400 workers in the 1994-96 period (most collective agreements have a duration of three years). Overall just 434,000 workers (6 percent of all employees) were covered by new collective agree- ments during the same period. Still, collective bargaining coverage in Malaysia is greater than union density, since the agreements also apply to nonunion workers who fall within their scope. Considerable disparity exists in the Philippines between estimates of trade union membership and the actual number of workers covered by collective bargaining agreements. In 1997 approximately 524,000 work- ers, a mere one-seventh of total trade union membership, were covered by collective agreements. As Kuruvilla and others (1999, p. 45) observe, "ab- sent the ability to expand collective bargaining beyond these small num- bers, the ability of unions to significandy voice the concerns of labor in the economic development process is seriously compromised." In Thailand collective bargaining and employee committees constitute the two main formal channels for labor-management dialogue at the en- terprise level. 7 In Thailand the number of trade unions greatly exceeds the number of collective bargaining agreements, since collective bargaining occurs between a single employer or enterprise and a single industrial union. Since trade union membership covers only a small share of the total work force, collective bargaining coverage is even smaller (and with the crisis has become smaller still). The number of collective agreements fell sharply in the recession year of 1997, dropping from the 1996 level of 381 to 271 in December 1997. Data related to trade union demands include the num- ber of workers the demands will affect. Over the past three years, each de- mand for collective negotiation has affected an average of 450 workers (MOLSW, various years). Overall, then, just under 122,000 workers, or around 1.2 percent of the formal private sector work force, were actually covered by collective agreements in December 1997. The Labor Relations Act of 1975 also allows for (but does not man- date) the creation of employee committees in enterprises with more than 50 employees.8 Like trade unions, the number of employee committees grew over the 1990s, and their decline in 1997 may also be a reflection of 432 EAST ASIAN LABOR MARKETS the impact of the crisis. Thailand has 24,830 establishments employing more than 50 persons, and employee committees exist in only 2 percent of them. Moreover, the vast majority of employee committees exist in es- tablishments that also have trade unions (Piriyangsan 1998). The two par- ticipatory mechanisms are therefore not substitutes for each other. If they were, a larger share of the work force would be covered by at least one voice mechanism. A Changing Climate for Labor Organizations? The data above allow us to make the simple proposition that, with the possible exception of Malaysia, little social dialogue in the context of the crisis is likely to have occurred at the microeconomic level in the countries under discussion-at least through any formal channels. While it is a safe bet that weak labor market organization will continue to characterize these countries in the near future, some positive signs have emerged, for the most part in response to the crisis. In the three countries most affected by the crisis-Indonesia, Korea, and Thailand-the right to freedom of association has been affirmed in various ways. Indonesia, for example, ratified ILO Convention No. 87 on freedom of association in June 1998, with the strong encouragement of the International Monetary Fund (IMF). New freedoms of association are clearly evident in that country, where the number of trade unions and labor federations is growing. At the same time a new framework labor law and laws on trade unions and dispute resolution are well advanced in the legislative process. Similarly in Korea one outcome of the tripar- tite social accord process was the accelerated acceptance of the Korean Confederation of Trade Unions (KCTU) as a participant in collective bargaining not only at the national but also at the enterprise level. The KCTU's inclusion could help reverse the downward trend in labor mar- ket organization in that country. Freedom of association, formerly with- held from Korean teachers, has now been granted them: as of July 1999 Korean teachers also have the right to organize trade unions and bargain collectively. In Thailand the new constitution explicitly reaffirms the right to free- dom of association. Thailand's labor federations and major employer con- federations recognize that bipartite labor-management relations remain underdeveloped and need to be strengthened. The two main pieces of in- dustrial relations legislation in Thailand (covering state-owned enterprises and the private sector) are both well advanced in the legislative process. SOCIAL DIALOGUE AND LABOR MARKET ADJUSTMENT 433 Trade union membership had been cut in half in 1991 in the wake of a mil- itary coup and the subsequent banning of trade union membership in the state-owned enterprises. The country's ongoing legal reforms would reverse this ban. The current version of the bill would mandate social dialogue at the enterprise level in the private sector through compulsory labor- management councils. While many of these developments do not translate directly into strengthened institutions for social dialogue at the enterprise and other levels, they are all necessary if such strengthening is to occur. As argued at the outset, global economic integration brings with it an increasing degree of external influence on national economies and domes- tic institutional arrangements. Coping with integration requires strong, resilient, and well-designed institutions of domestic governance if coun- tries are to reap the advantages of globalization at minimal social cost. The need for effective labor market institutions is no less great. Reducing un- employment and equitably sharing the burdens requires national negotia- tions that create arrangements among firms, governments, and social ac- tors, along with strong organizations and channels for dialogue. For the countries under review the formal channels for social dialogue at the en- terprise level are, as we have seen, often rather weak and limited in cover- age. One consequence of this situation is that initiatives undertaken at the national level become all the more important for addressing the social and economic consequences of the crisis. Otherwise, the costs continue to be disproportionately borne by the individual. The Social Dialogue in Malaysia Malaysia has experienced far fewer retrenchments than might have been ex- pected during a crisis of such magnitude. From an impressive growth rate prior to the crisis, the Malaysian economy contracted in 1998, with the growth rate of GDP declining by 13 percent. Yet the Ministry of Human Resources estimates that the country experienced a total of only 80,000 re- trenchments between the onset of the crisis in late 1997 and February 1999. Of these workers, moreover, the overwhelming majority were placed in new jobs, although sometimes at lower wages.9 A Participatory Approach There can be little doubt that Malaysia's approach to "crisis management" has been characterized by a strengthening of tripartite dialogue at the na- tional level. When the crisis erupted in late 1997, the government responded 434 EAST ASIAN LABOR MARKETS by creating the National Economic Action Council (NEAC), which was charged with developing a National Economic Recovery Plan. The plan, re- leased in July 1998, had six main elements, including several unorthodox prescriptions for foreign-exchange transactions, capital controls, and fiscal stimulus. As a decisionmaking apparatus, the NEAC was an entirely new in- stitution, born of the crisis. Chaired by the prime minister, it comprises 12 ministers, representatives of various industries, and representatives from the MTUC and the Congress of Union of Employees in the Public Administra- tion and Civil Service. The new council marked a significant change in the sometimes uneasy relations between the government and the MTUC and a substantial elevation of the labor movement's stature in the economic and social policymaking apparatus. Two other indications of the MTUC's rising stature in terms of managing the crisis were the prime minister's well- publicized public recognition of trade union support for the recovery pro- cess and the appointment of the MTUC secretary-general to the senate. There are at least some indications that giving the MTUC a voice in the management of the crisis has been effective. First, as the crisis deep- ened, the government recommended a freeze on wages and a suspension of collective bargaining over wages. At the MTUC's urging, however, the government drew back from this approach and accepted the proposal that collective bargaining over wages should continue and that wage outcomes should be a function of ability to pay. Second, the absence of widespread systems of social protection, includ- ing unemployment insurance, exacerbated the costs of the crisis in the re- gion. In the summer of 1998, the MTUC proposed the creation of a "one dollar" retrenchment fund to be capitalized by contributions of one ring- git per worker from both employers and employees. The fund would pro- vide replacement income for retrenched workers not eligible for severance pay. Initially the idea met with a cold reception from both the Malaysian Employers' Federation as well as the government. Over the intervening months, however, the government's position changed, and the "dollar fund" concept remains under review. EncouragingAlternatives to Retrenchment Efforts to stem job loss have also been channeled through long-established mechanisms for dialogue. The National Labor Advisory Council, a tripar- tite consultative body dealing with a range of labor policy matters, drew up guidelines on retrenchment with a view not only to ensuring that pro- cedures are equitable but also to highlighting alternatives to retrenchment. SOCIAL DIALOGUE AND LABOR MARKET ADJUSTMENT 435 In their emphasis on layoffs as a last resort, the guidelines reflect the strong aversion most Malaysian employers have to retrenchments. While draw- ing firm empirical conclusions about the effect of the guidelines is diffi- cult, they are likely to have had some influence on the behavior of enter- prises, for several reasons. First, there can be little doubt that the regulatory framework for indus- trial relations in Malaysia is clearly harnessed to the state's objective of pro- moting industrialization and attracting foreign direct investment. Strong workplace unionism remains weakly developed in Malaysia because of the regulatory framework. For example, it is procedurally difficult to conduct a legal strike. The state is more likely to step in and resolve disputes, leav- ing negotiated compromise weakly developed. Second, in keeping with the government's "look east" emphasis of the 1 980s, policy has strongly fa- vored the development of in-house unions, which are prohibited from federating across industries. In export-oriented pioneer industries, unions are not allowed to use collective bargaining to improve on the terms and conditions of the employment laws. Finally, relative to other countries in Southeast Asia, Malaysian law is characterized by a distincdy narrow legal scope for bargaining. For example, hiring, promotions, transfers, and dis- missals are the prerogatives of management and remain outside the legal scope of bargaining. Despite the nominal constraints the law imposes, regulatory incentives encourage enterprises to avoid retrenchments. Malaysia's basic industrial relations law includes the Code of Conduct on Industrial Harmony. The code stipulates that enterprises seek alternatives to retrenchment before dismissing workers. It further states that sacrifice-for example, by lower- ing wages and salaries-must be equitable across the hierarchy of the en- terprise. The code is technically a voluntary prescription of good conduct, but when dismissals have been challenged through the Industrial Court system, the courts have routinely based their judgments on the enterprise's adherence to the code. As a result the legal constraints on the scope of bar- gaining have been somewhat eroded. In response to the crisis, the Ministry of Human Resources now requires enterprises to give one month's advance notice before resorting to collective dismissals. This requirement allows various reemployment initiatives to be put into place. It also opens a window for considering negotiated alterna- tives to retrenchment. Malaysia's severance pay requirement, which is rela- tively generous for the region, also affects the behavior of enterprises. Be- cause of the cost of retrenching workers, employers are more inclined to retain labor, particularly when they view a downturn as temporary. 436 EAST ASIAN LABOR MARKETS Innovation at the Sectoral and Enterprise Levels Labor market adjustment is affected by other factors in addition to the regulatory environment and national tripartite dialogue. Important initia- tives have taken place at the sectoral level, where substantial cooperation exists between trade unions and employers trying to lessen the effects of the crisis on jobs. Prior to the crisis, the Federation of Malaysian Manu- facturers became the first employer organization to establish a joint com- mittee with the MTUC. With the crisis, the joint committee has become a task force on retrenchments and vacancies, in essence functioning as a private employment service. It keeps a list of companies planning to re- trench staff and then liaises with other companies to see whether they can take on workers about to lose their jobs. The MTUC is also part of the in- formation conduit and seeks vacancies for the list of retrenched workers it maintains. While there are no firm indications that employee participation at the enterprise level has increased in response to the crisis, trade unions, individ- ual employers, and employer organizations all report an increase in cooper- ation and communication during the downturn. ILO-sponsored research on enterprise adjustment during the crisis, while neither comprehensive nor necessarily representative, reveals a pattern of labor retention beyond what might have been expected in the nadir of the downturn (Siengthai 1999a). Some of the alternatives to retrenchment that have been put in place are clearly innovative. One example from the automobile industry received sub- stantial national media attention, especially when the Minister of Human Resources visited the enterprise (box 10.2). Companies have put a variety of measures in place to avoid retrench- ments (box 10.3). In most instances these measures are outcomes negoti- ated with the union. One exception was the largely nonunion electronics industry, which avoided retrenchments for two reasons. First, many cor- porations had explicit no-layoff policies. Second, the multinational enter- prises in the industry had extensive financial resources to tide them over a downturn they viewed as temporary. Peetz and Todd (1999, p. 184) observe that: Retrenchments could not be avoided altogether. But retrenchments were much less likely to occur where the downturn in demand was not severe, where the firm was unionized, where the firm was in some way linked with the State (including as a supplier to government), or where the firm was from an Asian-based [multinational corporations]. SOCIAL DIALOGUE AND LABOR MARKET ADJUSTMENT 437 BOX 10.2 One Enterprise Responds to the Crisis Quality Assembly assembles, sells, and provides after-sales service for motor ve- hicles that are designed overseas but built locally. The company was seriously hit by the financial crisis. In 1997-98 sales of the Korean and Japanese vehicles it assembles virtually halted, while sales of its European vehicles fell by over 60 per- cent. In total, monthly sales fell by 70 percent between July 1997 and Novem- ber 1998. Since 1994 a new chief executive officer (CEO) has been trying to create a culture of commonality. The firm's new motto is "if we sink, we sink together; if we float, we float together." Union representatives find this culture very different from the culture promoted under earlier CEOs and endorse the new approach. How has the company stayed afloat? It has taken a number of carefully thought out steps: * In August 1997, when the first signs of a downturn appeared, hiring was frozen, employees were encouraged to conserve on utilities such as power and water, and incidental expenses were cut (for instance, the annual dinner and sports meet were cancelled). * In September 1997 overtime was reduced and assembly activities slowed. * In October 1997 overtime was abolished. Cuts were made in petrol allow- ances paid to employees with company cars and in expenditures on business travel and outstation allowances. * In November 1997 the company stopped importing some parts. - In December 1997, with still further reductions in assembly, excess employ- ees were redeployed to other departments or to maintenance. * On April 1, 1998, the assembly plant was closed for three months, and 435 employees were temporarily laid off at 75 percent of their base salary, with no reduction of other benefits in the collective bargaining agreement, in- cluding the contractual bonus and 100 percent holiday pay. This arrange- ment was less costly than retrenchment for employees with an average of three to four years of tenure with the enterprise. The three-month layoff was negotiated with the union, and employees were encouraged to accept tem- porary employment elsewhere (at no loss of layoff pay). The personnel di- rector took an active role in identifying vacancies. * At the end of the three-month period, all but 10 employees returned to the company. 438 EAST ASIAN LABOR MARKETS When retrenchments could not be avoided, job losses tended to be shared equally throughout enterprises and were not concentrated in the unionized work force. The Social Dialogue in Thailand Thailand has lost a substantial number of jobs-860,000 as of the end of 1998. But these losses are far fewer than might have been anticipated given the magnitude of the downturn. The Thai labor market has experienced significant downward wage flexibility, as well as numerical flexibility (mostly through reductions in hours worked). Numerical flexibility, which is de- fined as adjustments in the quantity of labor inputs, occurs primarily through two channels: adjustments in the number of persons working or in the number of hours worked. Enterprises appear to have hoarded labor, retaining workers but reducing the hours they work and their pay. Sub- stantial underemployment and loss of earnings have resulted from this practice (box 10.4). The Extent of the Dialogue The formal channels through which social dialogue can occur are relatively underdeveloped in Thailand. In fact, Thailand's labor market is the least or- BOX 10.3 Alternatives to Retrenchment 1. Insourcing of previously outsourced activities. 2. Increased employee training during the slack time. 3. Temporary transfer of redundant employees to other unaffiliated enterprises. 4. Temporary layoffs, during which employees are paid a proportion of their usual wages and even allowed to work elsewhere. 5. Adjustments in pay through overtime cuts or freezes and the deferral or non- payment of contingent bonuses. Infrequent: Cuts in contractual bonuses, es- pecially in the unionized sector, despite renogotiated payment schedules, and reductions in base pay (cuts in managerial salaries were the most significant of these-up to 50 percent in the hardest-hit firms). 6. Postponement of negotiations on new collective agreements on wages. 7. Reductions in profit and even losses. SOCIAL DIALOGUE AND LABOR MARKET ADJUSTMENT 439 BOX 10.4 Labor Market Adjustment in Thailand Methods of Macroeconomic level promoting * Measures to reduce excess labor supply, including the numerical flexibility repatriation of illegal migrant workers and promotion of overseas employment for Thai workers. * Measures to stimulate labor demand, including fiscal stimulus, targeted job creation schemes (some funded by the social sector loan programs of World Bank and Asian Development Bank), and employment of grad- uates on public stipends. * Measures to match workers with jobs by coordinating labor market information from provincial employment service centers and disseminating labor market infor- mation through the Mass Layoff Center of the Ministry of Labor and Social Work's (MOLSW) Employment Department. * The dissemination of various alternatives to retrench- ment through the MOLSW * Reform of the Labor Protection Act that increases the cost of dismissing workers with long tenure as a fur- ther incentive to avoid retrenchment. Microeconomic level v Reductions in number of hours employees work as an alternative to layoff. Macroeconomic level Measures to encourage wage * No adjustment in the minimum wage in 1998 (and flexibility possibly in 1999) by decision of the tripartite National Wage Committee. * Reform of the Labor Protection Act to stipulate grad- ual decentralization of minimum wage fixing to the (Box continues on the following page) 440 EAST ASIAN LABOR MARKETS BOX 10.4 (continued) Measures to provincial level. Ten pilot provinces in the process of encourage wage forming tripartite provincial wage committees. flexibility * Devaluation of the Thai baht and subsequent ex- (continued) change rate policy, leading to a de facto decline in real incomes and shoring up export competitiveness of Thai industry. Microeconomic level Reductions in number of hours worked, leading to pay decreases (and lower payroll costs), reductions in overtime, nominal declines in base pay, and the re- duction or elimination of annual bonuses. Methods of Macroeconomic level encouraging De facto market-driven (rather than policy-driven) flexibility functional flexibility through changes in the composi- tion of employment such as reverse (urban-to-rural) migration and growth of urban informal sector activ- ities that help absorb displaced formal sector work- ers. Policy has supported this recomposition through employment-generation schemes targeted to persons returning to rural areas. e A review of labor policy, including policy relating to wages and competitiveness (for example, productivity and skill development), spurred by the conditionalities of Asian Development Bank assistance. Microeconomic level * Possible rethinking of human resource basis of produc- tivity improvement; unknown whether labor market slack has been used for retraining or skills upgrading or if work reorganization to increase productivity has been a feature of the downturn. SOCIAL DIALOGUE AND LABOR MARKET ADJUSTMENT 44I BOX 10.4 (continued) Measures promoting Macroeconomic level flexibility in industrial * The seven-point program for the alleviation of unem- relations ployment under the chairmanship of the prime minis- ter (one outcome of the tripartite process, albeit one in which government representation was dominant). * Efforts to encourage greater direct, popular participa- tion in the Eighth Plan (preceding the onset of the cri- sis), especially in the focus on community-based orga- nizations as recipients of social sector loan grants and in a variety of electoral reform measures arising from the expansion of democratic rights in the August 1997 constitution. * Two pieces of industrial relations legislation, the Labor Relations Act and the State Enterprise Relations Act, that are well advanced in a parliamentary process of reform. * National promotion of bipartism through 1996 Code of Conduct on Labor Relations and 1998 Code of Practice on the Promotion of Labor Relations. Microeconomic level * In aggregate quantitative terms, some weakening of trade unions and of labor-management institutions (such as collective bargaining and employee commit- tees). e Despite crisis-related increase in the number of indi- vidual disputes (over wages, severance pay, and dis- missals), greater labor-management communication during the crisis. ganized of the four under review. Moreover, any dialogue on labor market adjustment at the enterprise level covers only a minority of Thai workers- those in the formal or modern sector. It covers an even smaller share of the work force when the rural and urban informal sectors are included. Exten- 442 EAST ASIAN LABOR MARKETS sive flexibility of the market-driven sort is the rule for the majority of work- ers in Thailand, where informal jobs and incomes are not even protected by law, let alone subject to social dialogue. And in the modern sector, the already low level of worker representation has fallen still further during the crisis. The first year in the decade that saw the number of trade unions de- cline was 1997, most likely because of the initial effects of the crisis, which began in July of that year. The greatest share of the decline occurred in the industrial province of Pathum Thani, where the number of registered unions decreased from 148 to 87 over the 12-month period leading up to December 1997. In addition, the relative shares of different types of trade unions have shifted as a result of the downturn and the business closures associated with it. Most of Thailand's unions are at the enterprise level and are thus more vulnerable to downturns than other types of labor unions. The share of industry unions in the total had been increasing gradually in the 1990s, rising from 43-44 percent of all unions in the first half of the decade to 46 percent in 1996 and 48 percent in 1997. The number of en- terprise unions declined by 9 percent in 1997, while the number of in- dustry unions held almost constant. Tripartism and the Crisis Chief among the weaknesses of tripartism in Thailand is the fact that it is only as strong as its constituent parts. Fragmented interest groups with low membership levels risk being unable either to command the credibil- ity or to develop the voice they need to become effective representatives of the country's work force. In practice this weakness is often reflected by the government's domination of tripartite consultations, as well as by the un- sound criteria that are used to appoint worker representatives to tripartite bodies. Both weaken the effectiveness of tripartism. The Thai trade unions' own evaluation of the weakness of tripartism is shown in box 10.5. As noted earlier, at least in quantitative terms the crisis seems to have weakened some of the building blocks on which social dialogue is founded, such as union membership. At the same time, however, the crisis has served to reinvigorate or add momentum to joint approaches to problem solv- ing. Two tripartite mechanisms-the National Wage Committee (NWC), which sets the minimum wage, and the National Labor Development Ad- visory Council, which provides policy advice on matters relating to the labor market-have the most influence on employment policy. SOCIAL DIALOGUE AND LABOR MARKET ADJUSTMENT 443 BOX 10.5 Tripartism in Labor Affairs: Problems from a Labor Perspective 1. Workers and trade union members continue to be unaware of their legal rights. 2. Worker representatives to tripartite bodies are elected regardless of their qual- ifications, resulting in a low level of competence and easy domination by em- ployers and government representatives. 3. Many tripartite mechanisms contain too many government appointees. Gov- ernment appointees should not be allowed to serve concurrently on more than two tripartite committees. 4. Some worker representatives are not honest. A transparent monitoring process needs to be developed to ensure that these individuals act in the interests of the members they represent. 5. Rivalry exists among worker representatives on tripartite bodies, weakening labor's ability to speak with one voice. 6. The number of votes cast is not proportionate to the size of the membership but to the number of unions, further fragmenting labor's voice in the tripar- tite machinery. Source: Labour Congress of Thailand seminar, March 18, 1999, Chaingrai, as reported in The Labor Chronicle, Series 272 (June 1999), p. 7. The Minimum Wage An important instance of the influence of social dialogue on employment policy in Thailand is the mechanism for determining wages. Wages in Thailand are guided by a nationally determined, three-tier minimum wage that is adjusted frequently. The NWC, which comprises equal num- bers of worker, employer, and government representatives, recommends these adjustments. The government has endorsed virtually every one of the NWC's recommendations. At the microeconomic level, wage setting in Thai enterprises can be thought of as the outcome of the ways market forces are mediated-that is, through individual bargaining power and unilateral employer decisionmaking. This interpretation is logical in light of the scant coverage provided by collective bargaining. But it is also mis- leading, as the existence of the NWC shows. 444 EAST ASIAN LABOR MARKETS The NWC is Thailand's primary vehicle for social dialogue related to macroeconomic policy choices. The crisis in Malaysia and Korea has led to national social dialogue over a broad range of macroeconomic policy variables, including fiscal and monetary policies. But social dialogue in Thailand does not embrace this full range of macroeconomic policy vari- ables and is instead confined to minimum wage adjustments. As a tool of employment policy, moreover, the NWC cannot be considered a full-blown incomes policy mechanism. By law a minimum wage adjustment must take into account inflationary effects and anticipate employment impacts, among other factors, but it is not within the NWC's mandate to set infla- tion targets or employment objectives. During the crisis period, Thailand has entered into various loan com- mitments from international financial institutions to be directed toward social programs. One of these is a loan from the Asian Development Bank (ADB) that is to be disbursed in tranches on fulfillment of certain con- ditions to which the Thai government has committed itself One such commitment is to review labor policy, including wage policy. To this end, the ADB commissioned a study by an independent consultant (Brooker Group 1999). Concurrently the NWC formed the Subcommittee on Wage Policy and commissioned its own study, in this case, from the ILO (ILO forthcoming). Both studies found that increases in the minimum wage during the 1990s may have been excessive and were one of the factors-or perhaps the only factor-behind the erosion of export competitiveness of labor- intensive industry in Thailand. 'O Three aspects of the minimum wage trend in the 1990s are of particular relevance here. First, the real minimum wage trend in the 1990s was generally in line with trend growth of the economy as a whole, especially as minimum wage growth lagged behind economic growth rates during the 1980s. To the extent that a problem existed, it lay in the rate of productivity growth. Thailand's high overall productivity growth rate reflected a shift in labor resources away from relatively low- productivity activities (such as rural agriculture) into more productive ones (such as manufacturing). Within sectors, however, the growth of labor productivity varied substantially. It was lowest in manufacturing. The second aspect concerns the effect of minimum wage adjustments on earnings distribution in the 1990s. The largest increases in real wages were concentrated at the low end of the earnings distribution ladder. In ef- fect minimum wage policy was compressing the wage structure. This com- pression was in turn reflected in a gradual increase in the ratio of the min- imum to the average wage. Because a high percentage of the Thai work SOCIAL DIALOGUE AND LABOR MARKET ADJUSTMENT 445 force is concentrated at or around the minimum wage rate, the adjustments affected a sizable share of the work force. Workers earning the minimum wage were affected directly, while workers at enterprises that adjusted the prevailing wage structure felt the indirect effects. Third, the minimum wage in Thailand clearly applies to far more people than just new entrants to the work force and unskilled laborers. Trade union officials complained that workers with both tenure and skill acquisition were stuck at the minimum wage. Employers complained that they could not reward employees through an internal, enterprise wage policy, because adjustments to minimum wage displaced the capacity of companies to offer enterprise-based wage increases. A substantial share of workers remains at the minimum wage despite three or four years of on-the-job experience. One unintended outcome may perhaps have been the displacement of enterprise-based mechanisms for wage determination. Wage setting at the national level may have supplanted a role that workers would otherwise seek through enterprise-based social dialogue via trade unions and collec- tive bargaining. This displacement could be undermining efforts to orga- nize labor by making unions less attractive. But as we have seen minimum wage adjustments have also partly displaced employers' efforts to develop more comprehensive, enterprise-based wage policies. Minimum wage pol- icy can also work against the development of performance-based person- nel policies that seek to build commitment and improve productivity, among other things. In response to the crisis, the minimum wage remained at its 1997 level in 1998 and 1999. The crisis also prompted a legal reform of the minimum wage-fixing mechanism with a view to decentralizing responsibility for set- ting the minimum wage to the provincial level. While the NWC retains its responsibility for setting a national basic minimum age as the lowest thresh- old or wage floor, the provinces themselves recommend, on the basis of sev- eral statutory criteria, a provincial minimum wage. The NWC accepts, amends, or rejects the provincial recommendation. Local minimum wage fixing is intended not only to help fine-tune adjustment, bringing costs in line with capacity to pay, but also to help attract investment capital. The Thai trade unions have voiced much of the concern over this de- centralization. The provincial wage committees would be, like the NWC, tripartite. But trade union organization is particularly thin outside the Bangkok metropolitan area and the contiguous five provinces.11 Finding legitimate, competent workers to sit on the provincial committees will be difficult.12 The new system is as yet untried, but decentralization will be piloted in ten provinces. While it is too early to claim that the trade unions' 446 EAST ASIAN LABOR MARKETS concerns will prove unjustified, the opposite scenario also remains a pos- sibility-that is, the growth of provincial wage committees may become an opportunity to extend and strengthen tripartism in Thailand. In addition to these ongoing mechanisms, other instances of tripartite consultation have occurred, often of some significance. One of these was the process of formulating the comprehensive government response to the crisis. The prime minister undertook this initiative in December 1997 by forming the Committee on the Alleviation of Unemployment. Irrespective of the substance of the seven-point plan that emerged, the process itself cannot be described as a truly robust example of tripartite social dialogue. Worker and employer representation on the committee was minimal, and it is unclear what contribution their presence made to the final shape of initiatives proposed. Each of these initiatives was government driven, with various ministries given responsibilities for each point. Program imple- mentation did not specify a role for the social actors, nor has the process of implementation been an explicitly tripartite or dialogic one. Bipartism and the Crisis The crisis has altered at least some of the underlying conditions that previ- ously hindered the development of sound industrial relations at the enter- prise level. In 1996 and again in 1998, Thailand's worker and employer or- ganizations committed themselves to two codes of conduct. The 1996 Code of Conduct on Labor Relations stipulated that the parties find ways to re- duce industrial conflict and set up mechanisms for bipartite dialogue. The 1998 Code of Practice to Promote Labor Relations outlined ways of reduc- ing workplace costs and committed parties to the principle that bipartite so- cial dialogue is a necessary prerequisite to any decision affecting employees' welfare in the workplace. The legal reforms under way for both the private and public sector labor relations frameworks are also a positive sign. One of Thailand's most important employer organizations, the Em- ployer Confederation of Thai Industries (ECONTHIAI), has noted that one effect of the crisis has been to improve the flow of information and communication in the workplace. Box 10.6 provides a few examples of such improvements. There is a growing awareness in Thailand of the need to improve social dialogue at the enterprise level, for three reasons. An argument can be made that the climate for industrial democracy is strongly shaped by atti- tudes toward democratic governance in society at large. The process of constitutional reform in Thailand enjoyed wide popular support and par- SOCIAL DIALOGUE AND LABOR MARKET ADJUSTMENT 447 BOX 10.6 Social Dialogue, the Enterprise, and Adjusting to the Economic Downturn: Some Examplesa Banking Background Unionized. 26,000 employees, 60 percent union members. Affected by crisis, but less severely than other major banks. Effect of the crisis on labor-management dialogue Marked increase in communication and information ex- change through joint consultations. High degree of coopera- tion reported (first institutionalized by an abrupt 1984 busi- ness downturn in the bank). Adjustment measures reached through dialogue * No involuntary layoffs. * Bankwide restructuring and negotiated internal re- deployment. * Reduction of bonus from 5-6 months to 3 months. * Early retirement package: Minimum age 45 + 15 years' tenure for employees Minimum age 50 + 20 years' tenure for supervisors * Reduction in salary increase from 8-10 to 3 percent. * Union withholds new demands for the present time. * Increased skills training for employees and labor law training (paid for by bank) for union officials. Automotive Background Unionized. 4,000 employees. 70 percent union members. Sharp decline in business during the crisis. Effects of crisis on labor-management dialogue Marked increase in communication and information ex- change through joint consultative committees. Adjustment measures reached through dialogue * Ongoing negotiations over wages and reductions in bonus. (Box continues on the following page) 448 EAST ASIAN LABOR MARKETS BOX 10.6 (continued) * 1996 no-strike declaration in return for greater company support for union. * No involuntary layoffs. * Increased job rotation. * Increased company-funded training in parent company. Automotive Background Unionized. 380 employees, 65 percent union members Sharp decline in business. (No new work from May 1997 until May 1999.) Effect of crisis on labor-management dialogue Marked increase in communication and information exchange through joint consultative committees. Adjustment measures reached through dialogue * No involuntary layoffs thus far. * 3 months' leave at statutory pay reduction (Nov. 1998-Jan. 1999) * Hiring freeze. * Elimination of bonus. * 10 percent wage reduction for those earning over 5,000 baht per month. Garment Background Unionized. 1,200 employees. Export oriented. Crisis has had no effect on business. Effect of crisis on labor-management dialogue Collective bargaining takes place each year. Because business has been stable throughout the crisis, the frequency of labor- management dialogue has not changed. Adjustment measures reached through dialogue * No layoffs. No additional measures. SOCIAL DIALOGUE AND LABOR MARKET ADJUSTMENT 449 BOX 10.6 (continued) Garment Background Nonunion. 700 employees. Export oriented. Crisis has had only slight impact on business. Effect of crisis on labor-management dialogue No effect reported. Company practices high degree of commu- nication through informal channels, including frequent man- agement meetings with employee groups and frequent meet- ings between supervisors and subordinates. Adjustment measures reached through dialogue * Current hiring freeze. * No involuntary layoffs. * Staff reduction through attrition. * Units asked to reduce production costs. * Efforts increased to find new product markets. a. Names of enterprises have been withheld by request. Source: Siengthai 1999b. ticipation, and the new Constitution of 1997 is unambiguous in its sup- port for freedom of association and democratic governance. If anything, the crisis has reinforced these trends. Social Dialogue in the Philippines Several features distinguish labor market adjustment in the Philippines. First, at least in terms of GDP growth during the crisis, the Philippines has fared better than the other countries. Growth prior to the crisis was pro- ceeding at an impressive rate but was still lower than elsewhere in East Asia, reaching an estimated 5-6 percent in 1995-96. The Philippine economy did not contract with the crisis like economies elsewhere in the region. Instead, it stagnated. The social impact has nevertheless been se- vere: the peso's devaluation resulted in a 25 percent decline in per capita 450 EAST ASIAN LABOR MARKETS income in dollar terms, and the Department of Labor and Employment (DOLE) recorded over 120,000 layoffs in the first 10 months of 1998. Second, as in Thailand, the number of unionized workers in the Philip- pines is extremely low-1 1 percent of all workers. Further, collective bar- gaining agreements cover only around one-seventh of this small group of unionized workers. "Real" union representation thus covers only about 1.5 percent of the total work force. The third distinctive feature of the Philippine labor market is the weak growth of manufacturing employment during the precrisis years. Despite steady increases in foreign direct investment, employment in manufactur- ing held constant in the years prior to the crisis and declined when the cri- sis struck. Among the factors at work here is the effect of trade and invest- ment liberalization on enterprise strategies. Downsizing (or rightsizin,) had become very much a part of the vocabulary of adjustment even before the crisis. While the fear of retrenchments has become generalized in many countries since the crisis, the fear-and the factual basis for it-were al- ready present in the Philippines. The focus of dialogue at the enterprise level (and the subject of most recent disputes) has therefore been on employers' interest in greater nu- merical flexibility. In addition to retrenchments, employers have sought numerical flexibility through increased use of nonregular employees. The share of nonregular employees in total employment rose from 20 percent in 1992 to 28 percent in 1997.13 Employers subcontract, hire casual labor (temporary or fixed-term contract workers), and in what has been a con- tentious matter throughout the 1990s, use labor-only contracting, leasing rather than hiring employees. Two issues are of particular concern with labor-only contracting. First, wages and other terms of employment are then left up to the contractor and thus are often less favorable to workers than they would be if the workers were employed directly by the leasing firm. Second, such workers are largely unorganized. Trade unions attached to enterprises generally cannot recruit them because leased workers are not employees. In effect the Philippines now has two tiers of workers: core and periph- ery. Core workers have permanent employment, receive training and other investments in human resource development, and are continually upgrad- ing their skills. But the number of periphery workers is increasing because of the financial crisis, and this group of casual, temporary workers is bear- ing the brunt of the adjustment effort. In the environment of increasing casualization and subcontracting that characterizes the Philippine labor market, labor laws are often violated (Kuruvilla and others 1999). SOCLAL DLALOGUE AND LABOR MARKET ADJUSTMENT 45I A fourth feature of industrial relations in the Philippines (but one that it arguably shares with Korea) is its adversarial tradition. Legalism per- vades labor-management relations and manifests itself in a variety of ways. Traditionally arbitration has been required to resolve disputes the parties themselves cannot conclude. Delays in dispute resolution impede the de- velopment of closer, more cooperative labor-management relations, and opposing parties simply become more firmly entrenched. Legalism, inher- ited in part from the U.S. model on which Philippine labor law is based, breeds mistrust. Finally, labor-management relations at the enterprise level are poorly developed in the Philippines. Like Thailand, the Philippines has a frag- mented trade union movement, with federations and individual unions vying with each other for members and thus unable to speak with a uni- fied voice. There are seven national labor centers, a number of unaffiliated independent unions, and one workers' mutual aid society. At the national political level, one indication of the disadvantages of such fragmentation is that trade unions have been unable to advance the candidacies of all but one of their choices in the recent legislative elections. At the enterprise level, interunion rivalry has often resulted in an inability even to elect a union to represent employees. Signs of Crisis-Induced Change in Philippine Industrial Relations The recession may be changing industrial relations in the Philippines, moving it toward a more cooperative model. Perhaps the most significant sign of this change lies in the growing recognition that existing practices are inadequate. In this regard, the well-publicized dispute surrounding Philippine Airlines (PAL) in 1998 is an excellent example of this new awareness (box 10.7). As noted, sensitivity to the negative effects of numerical flexibility ap- pears to have been increasing in the Philippines even prior to the crisis. Kuruvilla and others (1999) find that for a small sample of enterprises, the crisis has induced a change of attitudes among trade unions and that unions have become more accepting of the need for retrenchments-or at least more accepting of the inevitability of job loss. In some instances, the trade union bargaining agenda has shifted away from simple resistance to discussing severance pay and retraining opportunities. The pattern has not been uniform, however, and in some instances acceptance has not meant acquiescence. This lack of agreement is reflected in at least some workers' 452 EAST ASIAN LABOR MARKETS BOX 10.7 The Philippine Airlines Dispute Philippine Airlines' (PAL) service was disrupted by a major dispute in mid-1998. Both domestic and overseas flights largely ceased for three weeks in June during a strike by the pilots' union over the issue of retrenchment. PAL management sum- marily fired over 600 pilots when they failed to respond to a return-to-work order issued by the Secretary of Labor and Employment. The courts have not ruled de- finitively on the legality of either the pilots' strike or the management's sweeping layoffs, although the pilots' union has filed suit to have its pilots reinstated. PAL was able to hire enough replacement pilots to restore limited service. In June the company filed for protection from its creditors under procedures similar to Chap- ter 11 in the United States. When the 1998 PAL labor dispute widened in July to involve the ground crew and support staff, it became a landmark labor relations event. Newly elected President Joseph Estrada took a direct hand in dispute me- diation efforts in order to save the financially troubled carrier. Through his in- volvement, an unprecedented agreement was reached that required the union leaders to suspend their collective bargaining agreement for 10 years in exchange for employee stock shares and seats on the PAL board.... The union also agreed to a limited layoff plan (many fewer than management first announced). Al- though a strike was averted, a dissident faction within the ground crew union soon challenged the agreement in court as an illegal infringement of worker rights. Source: U.S.D.O.L. 1999, p. 10. disaffection with their organizations, as the increase in wildcat strike ac- tions shows. Evidence from the Kururvilla study suggests that both labor and man- agement have become more willing to negotiate and cooperate because of the crisis. In the small sample of enterprises in the study, management has often tried to improve communications and cooperation in the workplace. This change complements the unions' greater realism when it comes to job losses and competitiveness concerns. In particular, the crisis appears to have galvanized greater support for labor-management councils (LMCs), originally initiated under the administration of Corazon Aquino. LMCs have since become vehicles for improving labor-management relations with less conflict. Effective LMCs may even be able to defuse workplace tensions, quicken the pace of collective bargaining, and improve the qual- SOCIAL DIALOGUE AND LABOR MARKET ADJUSTMENT 453 ity of enterprise industrial relations. But LMCs also pose a danger: non- union enterprises can use them as a strategy to avoid allowing unions in the workplace. Efforts at the National Level The fragmentation of the Philippine labor movement among competing labor federations has impeded the development of a clear and compre- hensive labor strategy that addresses the effects of the East Asian financial crisis and other workplace issues. There have nonetheless been some pos- itive signs of change here, too. First, the DOLE joined in a national tripartite process involving em- ployers' organizations and two labor federations that in February 1998 produced a national social accord. The fact that the accord was initiated by the employers themselves rather than by the government is a particu- larly healthy sign. The accord committed trade unions to forgoing strikes during the economic downturn in return for an employer pledge to avoid retrenchments as often as possible. After a six-month trial period, the tri- partite process and accord were renewed, and the process was replicated at the regional level, resulting in a number of regional accords. While the accords mark an important instance of a shift in the national mood toward greater social cooperation, gauging their effect in more con- crete terms is difficult. The number of industrial disputes has decreased in the Philippines, but the trend has been secular and during the crisis more likely reflects a loss of bargaining power than anything else. Some indica- tions suggest that a number of small and medium-size enterprises see themselves as exempt from the national social accord. These firms have used layoffs, pay cuts, and the elimination of benefits as emergency mea- sures (U.S. Department of State 1999). In an effort to ensure that the accord is more than just a statement of good intentions, in late 1998 the DOLE established a monitoring program to evaluate and promote compliance. Other changes have occurred within existing tripartite mech- anisms. The administration of Joseph Estrada broadened worker repre- sentation on the national Tripartite Industrial Peace Council to include representatives from six labor federations instead of just two."4 The result could be greater cooperation among trade union federations. Efforts to overcome labor fragmentation are evident among trade unions as well. The Brotherhood of Union Presidents in the Philippines, which was formed in 1996 to promote unity among trade unions, groups together union presidents from both independent local unions and labor 454 EAST ASIAN LABOR MARKETS federations. Another group, SULOG, is a loose coalition of unions repre- senting workers in several industries. The group has developed a bargain- ing procedure for unions to use when faced with retrenchments and clo- sures and also lobbies at the national level for lower interest rates to protect local industries (Kuruvilla and others 1999). Responsibility for setting a minimum wage in the Philippines has been decentralized since 1999 and now rests with tripartite regional wage boards. As in Thailand, adjustments to the minimum wage can be influential, since the minimum wage approximates the average wage for a large share of the formal work force. When the crisis made itself felt in 1997, em- ployer and government pressure to keep wage adjustments low prompted the major labor federations to boycott the tripartite process. By 1998, however, the labor federations had not only rejoined the tripartite process but had agreed to delay demands for higher wages in view of the recession and rising number of layoffs. A final and hopeful indication of the changing mood in labor-manage- ment relations in the Philippines is the continuing trend toward finding alternatives to litigious, third-party dispute resolution. The DOLE has shown less eagerness to intervene directly in disputes on the grounds that such intervention crowds out the development of bipartite dispute resolu- tion. It has increasingly promoted voluntary mediation and conciliation through the National Mediation and Conciliation Board and other less le- galistic private and public alternatives to dispute resolution. Social Dialogue in the Republic of Korea The crisis in Korea has been characterized by two developments. First, there has been a marked increase in the number of industrial disputes, almost all of them over the issue of retrenchment. Second, the country has undertaken a historic and largely successful experiment in national tripartite social dia- logue on solving the economic and social problems the crisis has caused. Strikes increased from 78 in 1997 to 129 in 1998, and because the 1998 strikes were concentrated in large enterprises undertaking restruc- turing, the number of employees involved was substantially more-232 percent-than in 1997. In part in response to this increase, the Tripartite Commission and the Tripartite Social Agreement were established, and by early 1998 Korea had become the region's best example of the effectiveness of social dialogue in managing economic crises. Several factors lie behind this initiative. First, the election to the presidency of the country's fore- most democracy advocate signalled the beginning of broad social partici- SOCIAL DIALOGUE AND LABOR MARKET ADJUSTMENT 455 pation in national decisionmaking.. Second, the new administration knew that making choices would be difficult in the middle of the country's worst postwar economic crisis. A crippling general strike under the pre- vious administration in December 1996 had been largely triggered by the absence of dialogue (last-minute, late-night reforms were made to the labor law over the unions' objections). The strike was a pointed reminder of the consequences of excluding organized labor from discussions of reform. Third, an immediate catalyst for dialogue was a condition of the US$57 billion IMF loan that mandated improved labor market flexibility. This condition meant simply that layoffs should be made easier so that enter- prise restructuring could occur.15 In Korea, however, lifetime employment is the cultural norm and is strongly protected by labor laws. Anticipating that widespread social unrest would arise from any attempt to violate this cultural and legal norm, the IMF urged the Korean government to enter into dialogue with the trade unions. As of January 2000 the Korean economy has staged a rather stunning recovery. Despite a contraction in 1998, it grew at over 7 percent in 1999. By the end of 1999 unemployment had declined from its peak of 8.7 per- cent to just over 6 percent in May 1999. However, enterprise adjustment may not have gone far enough, and with growth resuming, the pressure to undertake further restructuring may have eased. The Tripartite Process The early accomplishments of the tripartite process in Korea are unique among the experiences we have seen, for several reasons. First, as in Malay- sia, the Korean tripartite discussions have involved private actors in the full range of macroeconomic and macrosocial decisionmaking. But the Korean Tripartite Commission, unlike tripartite Malaysian groups, is characterized by the balanced participation of workers, employers, and the government. On the worker side, the Tripartite Commission includes not only the well- established Federation of Korean Trade Unions (FKTU), but also the more militant Korean Confederation of Trade Unions (KCTU). Inclusion on the committee provided the KCTU with substantial support in its ongoing process of establishing itself in Korean political and economic life. Second, the tripartite mechanism, inaugurated by then President-elect Kim Dae-Jung in January 1998, explicitly staked Korea's chances for re- covery on the tripartite process itself. Recognizing both that substantial sacrifices would be required and that the various economic reforms envis- 456 EAST ASIAN LABOR MARKETS aged would be unlikely to succeed in the absence of cooperation, the com- mission initially produced the Tripartite Joint Statement on Fair Burden- Sharing in the Process of Overcoming the Economic Crisis. A third and related distinguishing feature of the process was the contentiousness sur- rounding the issue of layoffs. These would be required not only as an ad- justment response to the recession, but also in the context of what was an- ticipated to be widespread restructuring-all in a labor market unused to retrenchment. A fourth feature of the tripartite process was that it rapidly became an institution. The Tripartite Commission became a permanent feature of the industrial relations landscape, backed by legislation and strengthened by its own secretariat. It would involve government participation at high lev- els and would include the main economics and labor ministries. And fi- nally, the outcome of the dialogue was quickly and concretely legislated into action. The Social Agreement reached on February 7, 1998, resulted in no fewer than 90 recommended actions, many of which have been acted upon (box 10.8). Indicators of Success The first round of the accord process resulted in a trade-off, with each side giving up something to gain something more. The labor federations ac- ceded to legal reforms that make dismissals "on managerial grounds" easier and expand the use of temporary and fixed-term workers. In return the government expanded workers' rights to freedom of association (most no- tably to teachers) and undertook to increase the amount of unemployment benefits, the length of time workers can collect them, and the number of workers they cover. While employers gained the labor market flexibility they sought, they also committed themselves to greater managerial trans- parency, a reform of accounting practices, and to enterprise restructuring that would continue to use dismissals as a measure of last resort requiring consultations with employee representatives. The government played a cat- alytic role in the process of promoting one of the new administration's pri- ority objectives-the reform of industrial relations. In view of the magnitude of the economic downturn, and despite the increase in labor disputes, both the process and the outcome of the Tripar- tite Commission have played important roles in promoting economic re- form while maintaining relative social stability. At the very least, the par- ties themselves credit the initiative with having played such a role. The Korean Employers' Federation (KEF) observed in August 1999: SOCIAL DIALOGUE AND LABOR MARKET ADJUSTMENT 457 BOX 10.8 Highlights of the 90-point Agreement of the First Tripartite Commission Management transparency and business restructuring * Promote management transparency * Improve the corporate financial structure * Establish a responsible management system * Promote business competitiveness Price stabilization Employment stabilization and unemployment policy * Expand and improve the unemployment insurance system and extend coverage * Provide support for retired and unemployed workers * Strengthen employment services * Expand vocational training * Promote job creation * Foster employer efforts to stabilize employment * Reduce the number of migrant workers * Increase necessary financial resources Extension and consolidation of the social security system Wage stabilization and the enhancement of labor-management cooperation The crisis was a big shock to both labor and management. Over the last year, 23,000 enterprises have gone bankrupt because of high interest rates and stag- nation of sales. Workers were faced with huge wage reductions (a minus 2.5 percent wage increase) and severe unemployment. The unemployment rate of 2 percent soared to 8.7 percent and the number of unemployed amounted to 1.8 million. However, industrial relations remained relatively stable because of sustained social dialogue [emphasis added] (Kim 1999). If the process can be credited with having contributed to social stability during the crisis, a related question is whether it can also be credited with having improved the economic climate. Recent research has explored the possibility of an empirical relationship between indicators of successful di- 458 EAST ASIAN LABOR MARKETS alogue (such as the resumption of participation in the tripartite dialogue and the successful resolution of industrial disputes) and indicators of macroeconomic performance (such as stock market performance and for- eign direct investment) (Hayter 1999). Preliminary results show a convinc- ing positive correlation between the indicators of social stability and eco- nomic performance. Investors appear to respond positively to indications of social stability and therefore to the mechanisms that generate such stability. The tripartite discussions have included a number of procedural matters in industrial relations at the enterprise level, particularly the need to im- prove labor-management cooperation. One of the 90 specific points from the commission's first round involved the establishment of an organization designed to monitor labor disputes at enterprises, eliminate unfair labor practices, and promote cooperation between labor and management. Some Continuing Challenges The Korean dialogue has not been uniformly smooth or consistently pro- ductive. At various points over the past 19 months, one or both of the labor federations have withdrawn from the process, complaining that the pace of economic restructuring and retrenchments was proceeding without con- sultations with trade unions. Employers have also withdrawn, charging that the government was trying to make a deal with the trade unions that ex- cludes the employers. The government too has either berated the major enterprises for dragging their feet on restructuring or sought legal penalties against trade union officials for engaging in legal strikes. As of January 2000 the FKTU has made a commitment to continue the dialogue with the government and employers to promote recovery and avoid industrial disputes. The KCTU, however, remains outside of the process. Labor federations in Korea very clearly derive their power from the large industry- and enterprise-based unions that bargain at the enterprise level, where retrenchments occur. A very early consequence of the first round of tripartite dialogue-when the deal was struck that made layoffs easier- was that the rank and file of the KCTU, angered by the agreement, forced the top leadership to resign. The agreement was reached in the first half of 1998, when the number of retrenchments had reached its peak. The first real test case of the agreement occurred in the context of projected mass layoffs at one of Korea's chaebols, Hyundai. As with the PAL strike in the Philippines, the 1998 strike at Hyundai Motors was widely publicized. It became symbolic of the challenges of enterprise restructuring, a litmus test of the economic reforms of Kim Dae-Jung's government. In May of 1998 the company announced its plans to cut 4,830 jobs and sent layoff notices SOCIAL DIALOGUE AND LABOR MARKET ADJUSTMENT 459 to 1,538 workers. Sporadic strike action began in late May and culmi- nated in a plant occupation and 36-day strike in July and August. The KEF (1998b, p. 2) commented: The final agreement reached by the union and Hyundai stipulated that there would be only 277 retrenchments. On the face of it, the trade union appeared to have gained the advantage, as the majority of those affected were female cafeteria workers whose jobs had been secured by another com- pany. The Korean Employers federation took a dim view of the govern- ment's conduct, the agreement, and the trade union's right to strike in the first place. It condemned the government for having what it called a "dou- ble standard" on labor reform-that is, urging businesses to restructure to remain competitive while at the same time urging them to minimize lay- offs. The KEF saw the agreement as a dangerous precedent that threw into doubt the government's ability to make the labor market more flexible and that rewarded union militancy, especially since the employers believed that the strike had been illegal. For the government, however, the significance of the strike lay else- where. The trade union had officially acceded to the principle of layoffs (box 10.9). With the worst of the recession apparently behind it, Korea may find that economic growth will help to relieve some of the destabilizing ten- sions in the tripartite process. As a result of the first social agreement, re- placement in the event of unemployment now applies to about 40 percent of the Korean work force instead of 9 percent (the prevailing level when the crisis struck), a factor that may encourage labor market flexibility. Dis- cussions within the tripartite commission have clearly slowed, yet the agenda of reform measures to be discussed is lengthy and includes issues such as flexibility in hours of work and the activation of labor-manage- ment consultations at the regional and industry levels. While the mood in regard to increasing the momentum of dialogue is relatively optimistic, no clear steps have been taken in this regard. Plans to anchor the tripartite process firmly within Korean legislation will likely serve as a catalyst to sta- bilize and strengthen the social dialogue. Looking to the Future Interest in creating and maintaining mechanisms that allow popular par- ticipation in political and social institutions is clearly rising. These insti- tutions include those pertaining to labor market governance at both the 460 EAST ASIAN LABOR MARKETS BOX 10.9 Highlights of the Agreement in the Hyundai Motor Company Dispute 1. Of the 1,538 workers who had been officially notified that they would be dismissed, only 277 would be retrenched "for managerial reasons." 2. The company would give consolation bonuses to the dismissed workers according to the following schedule: workers employed for less than 5 years, 7 months' pay; workers employed from 5 to 10 years, 8 months' pay; and those with more than 10 years, 9 months' pay. 3. The 1,261 workers exempted from dismissal would take 18 months of unpaid vacation leave. After one year, they would receive retraining courses from an external organization. 4. The government would be asked to take measures to cover the living expenses, reemployment, and job training of the dismissed workers. The company would try its best to see that the dismissed workers were employed by other Hyundai subsidiaries. 5. Redundancy dismissals would not be made for two years. The union would prepare a Joint Declaration on Industrial Harmony and the Avoidance of Disputes that focused on industrial peace. national and enterprise levels and in some instances (Korea and Malaysia, for example) at the sectoral level. This interest is evident not only in the various examples of "crisis management" discussed here, but also in the important legal reforms that have either been completed (Korea and In- donesia) or are well under way (Thailand) (box 10.10). These signs are also evident in the emergence of entirely new dialogue mechanisms in Cambodia, Indonesia, Korea, Malaysia, the Philippines, and Thailand. It remains unclear whether every country can maintain the momentum of the shift toward greater political openness and dialogue. There is some concern that the momentum is inversely related to the pace of economic recovery-that with a return to growth, some of the pressure to keep chan- nels of dialogue open will diminish and the incentive to improve labor- management institutions will slow. This risk is perhaps greatest in those countries where efforts at social dialogue during the crisis have been one- time events rather than ongoing institutionalized mechanisms. The risk is SOCIAL DIALOGUE AND LABOR MARKET ADJUSTMENT 46I BOX 10.10 Trade-offs Resulting from Social Dialogue at the National Level Korea * The Tripartite Social Accord (February 1998) made layoffs easier to promote economic restructuring and at the same time improved social protection by expanding social security. * Layoffs are an adjustment measure of last resort and require negotiation. * Second Tripartite Commission is examining the possibility of making adjustments in working time as an alternative to retrenchment. Philippines * Social Accord on Industrial Harmony and Stability (February 1998) between Employers' Confederation of the Philippines and two major labor federations provided for six months' mutual restraint on layoffs and industrial disputes. * At regional level, the Department of Labor and Employment (DOLE) organized tripartite conferences on saving jobs and industries that culminated in regional social accords. Singapore * Tripartite Panel on Retrenched Workers (February 1998) pro- vides for the development of a network for information on po- tential retrenchments in order to match those being dismissed with job vacancies. Also explores alternatives such as redeploy- ment, adjustments in working time and wages, and training possibilities; access to advice on training opportunities. a The Skills Redevelopment Program (December 1996) is being used as an alternative to layoff. The program, spear- headed by the National Trade Union Congress (NTUC) and funded by the government, allocates training subsidies of up to 80 percent of course fees and 70 percent of the absentee payroll for training conducted during normal hours of work. Malaysia * Tripartite Committee on Retrenchment set up to monitor re- trenchment. Encourages reference to Malaysia' Code of Con- duct for Industrial Harmony-in particular, the passages re- lating to retrenchment that stipulate a range of alternatives. (Box continues on the following page) 462 EAST ASIAN LABOR MARKETS BOX 10.10 (continued) Malaysia * Government introduces legislation (following consultations (continued) with the National Labor Advisory Council) requiring em- ployers to notify the Labor Department of prospective layoffs one month in advance. New law also to expand married women's participation in the labor force by making hours of work more flexible. Malaysian Trades Union Congress (MTUC) and Malaysian manufacturing employers establish databank on job vacan- cies for retrenched workers. smallest in economies where reforms strengthening relevant social institu- tions have been legally sustained (such as Indonesia, Korea, and Thailand). Social dialogue and sound economic outcomes are positively related and thus move together. Arguably, the examples in this chapter have demon- strated that fact. Korea provides the most extensive and influential example of tripartite management of the effects of the East Asian financial crisis, and it is the Korean economy that has shown the strongest recovery. Yet as with most research topics, much remains to be learned about the causal rela- tionship between effective social dialogue and recovery from exogenous shocks such as the 1997-98 crisis. Finally, the institutions on which formal social dialogue are based in the countries under review are weak, sometimes profoundly so. Collective bargaining coverage is very slight-smaller even than trade union representation, which itself is often exceedingly weak. Trade unions themselves are inadequately financed, and in most cases their already small membership has been further weakened by the financial cri- sis. Labor movements are fragmented in the Philippines, Thailand, and in- creasingly in Indonesia. This weakness impedes the development of a co- herent voice for workers and their employers that could influence the social dialogue. Thus, while this chapter has maintained a tone of guarded opti- mism, it also makes clear that worker and employer organizations must be strengthened if social dialogue is to progress beyond the "crisis" stage. Notes 1. Duncan Campell is a Senior Industrial Relations Specialist with the Inter- national Labour Organization's East Asia Multidisciplinary Team in Bangkok. SOCIAL DLALOGUE AND LABOR MARKET ADJUSTMENT 463 2. So, too, in other countries. In the Republic of Korea, three-way dialogue is a primary government objective, and key economics ministries also participate. In the still-indeterminate case of Indonesia, the tripartite machinery is expected to extend beyond the Manpower Department to include government depart- ments in commercial sectors. The composition of Cambodia's new National Labor Advisory Committee has similarly broad government participation. 3. One can be concerned about (that is, affected by) a decision but thor- oughly ignorant of the information on which the decision can and ought to be made. For this reason, effective three-way dialogue relies on the competence and knowledge of the participants. 4. Much of the vocabulary and many of the concepts in this section are based on Hirschman (1970) and Freeman and Medoff (1984). 5. In the process, traditional notions of hierarchy are being eroded, since hi- erarchies can be barriers to the flow of information. See, for example, Hirschhorn (1998). 6. In the immediacy of the crisis, of course, the value of participation lies in the search for least-cost adjustment paths and ways to minimize the chance for social unrest. 7. Formal in the sense of provided for and regulated by law (unlike, for exam- ple, quality circles). 8. The Labor Protection Act of 1998 also mandates the establishment of em- ployee welfare committees and promotes tripartite mechanisms. 9. One factor that helped keep retrenchments low was the buffer provided by the 11-12 percent of the Malaysian work force that is foreign. Repatriating for- eign workers to safeguard the jobs of Malaysian citizens became explicit policy. Foreign workers were concentrated in particularly hard-hit sectors, such as con- struction. Job losses there, however, as well as for foreign workers generally, typi- cally resulted from unrenewed contracts rather than from retrenchment. One im- plication of this scenario is that the number of jobs lost in the recession was a multiple of the number of those actually retrenched. 10. The ILO study also finds that this explanation is not entirely satisfactory, for a number of reasons. For example, capital-intensive exports in which labor accounts for a small share of total costs also saw a decline in export performance. In addition, the United States, as a major market for garment exports, has been reducing its re- liance on Asian (not just Thai) sourcing in favor of closer-to-home, intraregional trade arrangements. And Thai exchange rate policy (rather than wage policy) greatly inflated real wage increases in Thailand when denominated in dollars. 11. The regional distribution of Thailand's trade unions suggests that they are geographically dispersed. In fact, however, when we look at Bangkok and the five industrial provinces surrounding it, we can see clearly that trade unions continue to be highly concentrated in the greater Bangkok metropolitan area. 12. Employers, on the other hand, have sometimes expressed concern that in the absence of effective worker voice on the provincial committees, the govern- 464 EAST ASIAN LABOR MARKETS ment will be compelled to adopt the workers' viewpoint on wage setting, to the detriment of employers' concerns about labor costs. 13. Perhaps in part because in 1997 the DOLE further eased restrictions on hiring short-term contractual workers. 14. Two representatives had been the norm under the two previous adminis- trations. 15. As might be expected, there is debate over the degree to which Korean chaebols are overstaffed. The Korean Employers' Federation (KEF) estimated in late 1998 (well after the legal reforms making layoffs easier had taken effect) that large enterprises continued to retain about 20 percent redundant labor (338,000 jobs), while the banking industry was hoarding 33 percent of employees it did not need (132,000). At the same time, the KEF estimated the labor shortages in small- and medium-sized enterprises at 120,000 (KEF 1998a). References Brooker Group. 1999. "Labor Policy in Thailand." Bangkok. Freeman, R., and J. Medoff 1984. What Do Unions Do?New York: Basic Books. Hayter, S. 1999. "Social Fundamentals and the Asian Financial Crisis." Geneva: International Labor Office. Hi rschhorn, L. 1998. ReworkingAuthority: Leading and Following in a Post-Mod- ern Organization. Cambridge: MIT Press. Hirschman, A. 1970. Exit, Voice, and Loyalty. New York: Oxford University Press. ILO (International Labour Organization). Forthcoming. "Minimum Wage Policy in Thailand." Bangkok. Working paper. . 1999. "Evolution of the Asian Financial Crisis and Determination of Policy Needs and Responses." Governing Body Document 274/4/2. Geneva. Kim, Jung-Tae. 1999. "Globalization and Industrial Relations: Employers' Per- spective." Note prepared for the Asian Regional Tripartite Seminar on Global- ization and Industrial Relations, Bangkok, August. KEF (Korean Employers' Federation). 1998a. Industrial Relations and the Labor Market. Seoul. . 1 998b. Quarterly Review XV(III). KLI (Korea Labor Institute). Various years. Quarterly Labor Review. Kuruvilla, S., Erickson, C. 1999. "Industrial Relations and Globalization in the Philippines." Bangkok: International Labour Organization, East Asia Multi- disciplinary Team. Unpublished manuscript. Lee, E. 1999. The Asian Financial Crisis: The Challenge for Social Policy, (Geneva: International Labour Organization. MOLSW (Ministry of Labor and Social Welfare). Various years. Yearbook of Labor Statistics. Bangkok. NESDB (National Economic and Social Development Board). 1997. The Eighth National Economic and Social Development Plan, 1997-2001. Available at www.nesdb.go.th/New_menu/plan 8e/content_page.html. SOCIAL DIALOGUE AND LABOR MARKET ADJUSTMENT 465 Peetz, D., and T. Todd. 1999. "Industrial Relations and Globalization in Malaysia." Bangkok: International Labour Organization, East Asia Multidisci- plinary Team. Unpublished manuscript. Piriyangsan, S. 1998. "Bipartite Systems: Workplace Relations in Thailand." Paper prepared for International Labour Organization, East Asia Multidisciplinary Team, Bangkok. Rodrik, D. 1999. The New Global Economy and Developing Countries: Making Openness Work. Washington, D.C.: Overseas Development Council. Siengthai, Sununta. 1999a. Employee Relations during the Crisis. Bangkok: Inter- national Labour Organization, East Asia Multidisciplinary Team. _____. 1999b. Industrial Relations and the Recession in Thailand. Bangkok: International Labour Organization, East Asia Multidisciplinary Team. Foreign Labor Trends, Philippines. FLT 99-4. U.S.D.O.L. (United States Department of Labor). 1999. Foreign Labor Trends, the Philippines. FLT 99-4. Washington, D.C. Japan's Experience with Employment Policy NAOYUKI KAMEYAMA THROUGHOUT THE 1990s the Asian economies were caught up in a whirlwind of change, moving from stagnation to rapid growth, collapse, and recovery. During this period virtually all the Asian governments sought to construct an effective employment policy. These efforts inform our analysis of Japanese employment policy since World War II. The East Asian economies are closely linked. But at the beginning of the 21st century they have reached separate stages of development and face very different employment (and unemployment) problems. Cultural differences and historical experiences have affected industrial relations in the region, making comparisons difficult. For this reason our analysis of the Japanese experience does not advocate best practices in employment policy. Instead we offer-in their historical context-lessons learned over the course of the last half-century. During the past 50 years Japan has ex- perienced a number of economic scenarios that have required the govern- ment to respond to a variety of employment problems. We hope that other Asian countries will be able to draw on this experience as they strug- gle to cope with their own employment issues.' In 2000 Japan is enduring a serious recession. The country has experi- enced periods of economic adjustment and stagnation before, even during the era of high economic growth era that lasted from the early 1960s to the mid-1970s. But the length and severity of the most recent recession, which began in 1992 with the collapse of the bubble economy, are unlike 467 468 EAST ASIAN LABOR MARKETS anything Japan has experienced before. The unemployment rate has reached nearly 5 percent. The collapse of lifetime employment, tradition- ally considered a fact of life for Japanese workers, is widely discussed. This chapter does not address the accuracy of such a claim, but the fact that it is being put forth at all is indicative of the degree to which the current re- cession has prompted new thinking about employment policies. Understanding the new direction Japanese labor policy must take re- quires us first to systematically reassess the employment policies of the last half-century. These policies are best examined as a group rather than indi- vidually. We emphasize the systemic aspects of employment policy, since developing effective policies in East Asia requires thinking systemically and making coordination and consistency the top priorities. Our ap- proach to post-World War II employment measures and policies also sup- ports the development of effective targeting, a key component for policy- makers to consider. We evaluate Japanese employment policy for each of the five distinct economic periods the country has experienced since World War II: * The immediate postwar recession (1945-60); * The period of high economic growth (1960-73); e The downturn induced by the oil crisis from the mid- 1970s until the mid-1980s; ' The "bubble economy" of the late 1980s;2 and i The Heisei recession, which began in 1992. The Postwar Recession (1945-60) World War II destroyed the industrial base of the Japanese economy. The economic and labor resources that had been devoted to the war industry were no longer needed, and strategic bombing by the Allied Forces de- stroyed a vast number of industrial facilities and disrupted urban func- tions near the end of the war. The Economic Stabilization Board estimated the economic damage Japan sustained as a result of the war at 65.3 billion yen (as of August 1945)-approximately one-fourth of the country's pro- duction capability. The Postwar Labor Market3 At the end of the war, Japan had a huge population of unemployed and underemployed workers, in part because of the vast number of people re- JAPAN'S EXPERIENCE 469 turning from overseas and the demobilization of the country's soldiers. This rapid increase in the productive population was noted in the 1950 census, which found that the number of people ages 15-59 had risen sharply during the previous decade, climbing from 40.7 million in 1940 to 44.7 million in 1947 and to 47.4 million in 1950. Further complicat- ing the situation was the exodus of workers from war-damaged cities to farming and fishing villages. This internal migration resulted in popula- tion overflows and stagnation in the rural sector. In addition a large num- ber of people were forced to stop working because of injuries suffered dur- ing the war or because they had been discharged from war industries. Statistics Bureau data from a survey conducted in April 1946 found that 1.6 million people who could and wanted to work could not find a job. Among those working, underemployment was a major concern. The number of employees, self-employed individuals, and family business workers who worked only 1-7 days per month was 1.96 million. Another 4.32 million worked only 8-19 days each month. These figures indicate that one out of every four Japanese laborers was unemployed or under- employed immediately following the end of the war. The employment situation continued to deteriorate in 1948, when the government adopted the Dodge Plan to counter inflation. This plan in- volved large-scale administrative readjustments and corporate consolida- tions that resulted in massive retrenchments. As many as 155,000 em- ployees in administrative organizations and 435,000 employees in private companies were discharged. The Japanese economy did not begin to grow substantially until June 1950 and the start of the Korean War. The war ef- fort meant huge orders for "special procurements" and an increase in ex- ports. The manufacturing production index, which had been assigned a value of 100 for June 1950, had expanded to 135 by December of that year. Private investment also gained momentum. Ultimately the Japanese economy would grow at more than 10 percent annually for the three years of the Korean War. By 1954 manufacturing production had expanded by 225 percent, and after 1954 Japan enjoyed a long period of sustained eco- nomic prosperity, interrupted by brief periods of economic adjustment. Unfortunately the renewed growth did not immediately eliminate the unemployment and underemployment problems, for a number of reasons. First, dramatic demographic changes in 1950-53 resulted in a growth rate of 10.9 percent in the labor force, which expanded from 65.5 to 68.3 per- cent of the total population. Second, although manufacturing production expanded, productivity also increased, holding down the demand for ad- ditional workers. Furthermore, those firms that did hire additional work- 470 EAST ASIAN LABOR MARKETS ers did so on a temporary basis because they feared that wartime procure- ments would not continue. Third, some underemployed people gradually became unemployed, raising the number of unemployed workers from 0.44 million in 1950 to 0.47 million in 1952. By March 1952 some 30 percent of the total workforce was underemployed.4 Postwar Employment Policy The common belief in every industrial nation after World War II was that the ultimate objective of a national employment policy should be to achieve full employment. Japan was no exception. The definition of full employment, however, was unclear, and policymakers soon narrowed the scope of employment policy to address the issue of underemployment. The government's policy had a number of objectives: * To reduce the need for underemployment, which exists to supple- ment inadequate household income; * To reduce the possibility that unemployment could be disguised as underemployment and thus remain invisible; * To reduce the scope for employer practices that encouraged under- employment through self-employment and family work; and * To absorb underemployed people as the economy developed. In addition to these efforts to address underemployment, the government created jobs directly (through public works programs), provided job-match- ing services, and offered training through public employment services. PUBLIC WORKS. Large-scale public works programs were initiated in late 1946 for the purpose of recovering national lands, developing social capi- tal, and absorbing the unemployed. The projects were not well targeted, however, and the projects absorbed only about half as many unemployed people as had originally been planned. The main problem was that the re- gions where public works were undertaken were not the areas where un- employment had increased dramatically. More than 70 percent of public works projects in those days were concentrated in farming villages and mountain areas in order to meet the urgent need to expand food produc- tion and improve the infrastructure around mountains and rivers. Al- though the top priority was eliminating unemployment, local farmers who were not among the unemployed were hired for these public works jobs. The deflationary Dodge Plan in 1948 cut government spending and shrank the public works programs. Tight economic policy and corporate JAPAN'S EXPERIENCE 471 restructuring led to considerable layoffs in the private sector and increased unemployment. The Emergency Unemployment Countermeasures Law was established in May 1949 to address this problem. These emergency measures created new work relief projects to absorb unemployment. This time, however, the unemployed were required to register at a Public Em- ployment Security Office (PESO), which then dispatched them to work sites. At their peak in the early 1960s, these public works projects em- ployed 350,000 people. Although they were much better than earlier programs at reaching truly unemployed workers, this generation of public works programs was not successful in providing only temporary employment. Those employed at work relief projects included large numbers of middle-aged and senior workers for whom reemployment was difficult, self-employed people who were having a hard time making a living, and women who had never worked before. Many of these workers tended to stay in public works jobs for a long time. Accordingly various changes were made to the projects, in- cluding establishing eligibility criteria that would screen out certain groups, enhancing work supervision, and setting wage differentials according to performance and skills, among others. These revisions met with limited success. As the workers got older, their capabilities declined, and the ten- dency to remain longer with the projects increased. Worker profiles col- lected in 1960 revealed the average age to be 50 and the length of service to be six years (MOL 1960). Another problem with the public works projects was that most of the jobs did not require many skills, and the wages were low. This problem was in part the result of the prevailing thinking on employment policy. One of the main objectives of postwar macroeconomic policy was resolv- ing what was called the "employment problem."5 The main purpose of the economic policies designed to meet this goal was to produce and expand overall demand for goods and services. As demand for goods and services increased, the demand for labor would increase along with it. The policies did not target certain groups among the unemployed or certain jobs or in- dustries, and the programs were not aimed at any particular category of unemployed worker. The goal of such general manpower policies was to provide a job for every worker, regardless of technical skills, knowledge, or physical or mental condition. JOB-MATCHING SERVICES. The government of Japan also sought to re- duce unemployment by matching job seekers and employers through the PESOs.6 These offices were also responsible for dispatching workers to the public works programs and were preparing to implement the newly es- 472 EAST ASIAN LABOR MARKETS tablished unemployment insurance system. Given the very high numbers of unemployed persons at this time and these additional responsibilities, Japan's public job-brokering services were not nearly as successful as had been envisioned. The Ministry of Labor provided guidelines and training for staff in these offices in order to improve operational techniques. The Period of Rapid Economic Growth (1960-73) Full-scale economic growth did not begin until the 1 960s, when the gov- ernment developed the National Income Doubling Plan. The goal of the plan was to improve living standards substantially and work toward full employment. To meet these goals, the plan identified five core needs: * To improve social capital; * To provide guidance in the establishment of an advanced industrial structure; * To promote trade and international cooperation; * To improve worker skill levels and promote science and technology; * To put an end to the "dual industry" economic structure and en- hance social stability.7 The second issue merits particular attention. The basic idea was to im- prove productivity at the industry or individual firm level and to revamp the industrial structure by emphasizing high-productivity rather than low- productivity sectors. Planning for a High-Growth Labor Market In order to create an advanced industrial structure, the labor force distri- bution across industries would have to be changed. The plan projected that the total number of workers would expand by 7.15 million during the 1960s. That estimate assumed the following shifts among industries: a de- cline in the number of workers in primary industries of 4.91 million; an expansion of workers in secondary industries of 5.62 million; and an ex- pansion in tertiary industries of 5.32 million workers. In addition, the plan forecast annual average real economic growth of 7.8 percent during the decade, resulting in annual average economic growth of 10.9 percent. The number of workers in primary industry declined almost in accor- dance with the plan, with an actual drop of 4.5 million (table 11-1). This decline had been expected to occur as older workers retired and a shift in TABLE 11.1 Employment by Sector, millionsa Employed persons Total (A) Primary industry Secondary industry Tertiary industry Employees (B) B/A 1950 36.0 (100) 17.5 (48.6) 7.8 (21.7) 10.7 (29.7) 1955 41.0 (100) 15.4 (37.6) 10.0 (24.4) 15.6 (38.4) 17.8 (43.4) 1960 44.3 (100) 13.4 (30.2) 12.4 (28.0) 18.5 (41.8) 23.7 (53.5) 1965 47.3 (100) 11.1 (23.5) 15.1 (31.9) 21.1 (44.6) 28.8 (60.9) 1970 50.9 (100) 8.9 (17.5) 17.9 (35.2) 24.1 (47.3) 33.1 (65.0) 1973 52.6 (100) 7.1 (13.5) 19.3 (36.7) 26.2 (49.8) 36.2 (68.8) 1975 52.1 (100) 6.6 (12.7) 18.4 (35.3) 27.1 (52.0) 36.5 (70.1) 1980 54.7 (100) 5.3 (9.7) 19.2 (35.1) 30.2 (55.2) 39.7 (72.6) 1985 57.8 (100) 5.1 (8.8) 19.9 (34.4) 32.8 (56.7) 43.1 (74.6) 1990 62.2 (100) 4.5 (7.2) 21.0 (33.8) 36.7 (59.0) 48.4 (77.8) 1991 63.3 (100) 4.2 (6.6) 21.6 (34.1) 37.5 (59.2) 50.0 (79.0) 1992 64.0 (100) 4.1 (6.4) 21.9 (34.2) 38.0 (59.4) 51.2 (80.0) 1993 64.3 (100) 3.8 (5.9) 21.7 (33.7) 38.8 (60.3) 52.0 (80.9) 1994 64.0 (100) 3.5 (5.5) 21.5 (33.6) 39.0 (60.9) 52.4 (81.9) 1995 64.1 (100) 3.4 (5.3) 21.3 (33.2) 39.4 (61.5) 52.6 (82.1) 1996 64.9 (100) 3.3 (5.1) 21.2 (32.7) 40.4 (62.2) 53.2 (82.0) 1997 65.2 (100) 3.5 (5.4) 21.3 (32.7) 40.4 (62.0) 53.9 (82.7) 1998 65.1 (100) 3.4 (5.2) 20.5 (31.5) 40.8 (62.7) 53.7 (82.5) aFigures in parentheses are percentages of total labor force. Source: Bureau of Statistics, Annual Report on the Labor Force Survey. 474 EAST ASIAN LABOR MARKETS labor occurred between the agricultural and industrial sectors. In the end, however, the decline was the result primarily of improvements in produc- tivity brought on by mechanization and a drop in off-farm jobs among rural households. In fact a large idle workforce was created in farming vil- lages, and these unemployed workers eventually drifted into manufacturing and construction industries as migrant workers, creating a new employ- ment problem. In secondary industries, the number of workers expanded by 5.62 million. Employment also rose in tertiary industries by 5.32 mil- lion, nearly in line with the plan. Importantly, no major social friction oc- curred during these shifts, thanks to the supply of new graduates during this period-over 20 million. This period was not without its employment problems, however. Facil- itating the movement of the workforce among regions became a priority, as did coping with the structural unemployment that developed as cer- tain industries declined. Furthermore, near the end of the period of rapid economic growth the supply of labor slowed. New employment policies were needed to address this problem, which had begun to hinder eco- nomic growth. Employment Policy during the High-Growth Period Rapid economic growth during this period meant a gradual reduction in unemployment and underemployment, but it soon created a new prob- lem-a labor shortage. The Employment Measures Law enacted in July 1966 addressed this issue by attempting to lift restrictions on the labor supply. Based on this law, the First Employment Measures Basic Plan was drawn up in March 1966 with the subtitle "Preparation for Full Employ- ment" UMOL 1966). The Basic Plan stated that: .... the objective of employment policy is to enable everyone to hold a job that is suitable to their aptitude and fully utilizes their capabilities, as well as to improve the economic and social position of these workers in order to contribute to the development of the national economy. Although it incorporated qualitative aspects into the objectives, the policy was primarily intended to facilitate economic growth-the "devel- opment of the national economy." The concept of full employment in Japanese labor policy continued to be expanded and modified. The Second Employment Measures Basic Plan January 1973), subtitled "For a Rewarding and Comfortable Working JAPAN'S EXPERIENCE 475 Life," recognized both a quantitative and qualitative dimension to the con- cept of full employment and began to incorporate qualitative objectives into employment policy: Amid high economic growth, employment conditions improved signifi- cantly in terms of volume. However, not all workers were given the oppor- tunity to develop their abilities throughout their lives, hold jobs suitable to their capabilities, or to find something they could live for at the workplace. The reality is a long way from true full employment, both in terms of qual- ity and quantity. Despite reduced unemployment, the public works programs initiated in the aftermath of World War II continued to employ significant num- bers of workers. In order to phase out this reliance on work relief projects, the Employment Security Law was revised in 1963. The government also shifted its employment policy priority from absorbing unemployment to promoting employment growth in other sectors. The PESOs focused on placing the unemployed in existing jobs whenever possible, falling back on public works jobs as a last resort. Thanks to these measures and the long-lasting economic prosperity that had begun in 1964, the number of workers participating in public works projects declined. In 1971 the Promotion of Employment of Mid- dle and Older Workers Special Measures Law had prohibited the absorp- tion of any additional unemployed people, marking the beginning of the end of public works projects.8 Although the existing projects did not take on any new workers, people who were already employed tended not to leave, resulting in a decline in capabilities and skills and more work-related accidents as the workers aged. The job-brokering role of the PESOs increased in importance as de- mand for labor grew. The offices were required to help reallocate the labor force among industries, jobs, and regions. In particular, demand for labor concentrated on new graduates, mainly junior high school graduates.9 The goal was to direct these new graduates to growth industries in accordance with their abilities, aptitudes, and desires. An amendment to Article 19 in the Employment Security Law allowing PESO staff to guide job seekers to positions in other regions was important in meeting the growing demand for labor.10 As labor shortages developed in some regions-either because of a lack of workers or because of a lack of workers with certain skills-companies diversified their recruiting activities to cover a wider area. Accordingly, in 476 EAST ASIAN LABOR MARKETS 1964 the government established the Labor Market Center. This center was intended to serve a number of purposes: * To integrate PESO functions nationwide; * To provide broad information on job vacancies and required skills in a timely and efficient manner; * To facilitate job placement outside local areas; and * To enhance clerical functions in areas such as the administration of unemployment insurance. The opening of the Labor Market Center made it possible to quickly iden- tify job seekers whose skills match specific job openings. The center also implemented a system to calculate the unemployment insurance eligibil- ity period for job seekers with more than one former employer. In the 1990s an on-line system was established to link employment organizations all over the country and facilitate data transfers. Employment programs of this era did not fail to target older workers. The economy was growing and the employment situation had improved substantially, but many older workers were still unemployed. The govern- ment introduced unique measures for this group instead of handling older workers as average job seekers. The Employment Security Law, which was substantially revised in 1963, contained new provisions for unemployed workers ages 35 and older. Five types of employment promotion measures were made available to these workers: long-term vocational training (six months to a year), short-term vocational training (about three months), occupational adjustment training (about six months), vocational courses (about two months), and job search courses (two to six months). One of the first selective manpower policies designed to target a partic- ular group of displaced workers was the 1959 Displaced Coal Mine Work- ers Temporary Measures Law (box 11-1). Other targeted initiatives were implemented during this period, including the 1958 Temporary Measures Law for Displaced Workers from Stationary Troops and the 1960 Law for Employment Promotion of the Disabled. Gradually selective policies began to replace general employment policies. The First Oil Crisis and the Period of Low Growth (1973-85) The first oil crisis at the end of 1973 pushed up the prices of commodities and raw materials (mainly oil), severely damaging oil-importing economies JAPAN'S EXPERIENCE 477 BOX 11.1 Assisting Displaced Coal Mine Workers As Japan has virtually no oil resources, coal was the main source of energy during industrialization. But in 1950 the country began to shift to oil as its primary fuel. Demand for coal fell rapidly, resulting in a chronic recession in the coal industry. From 1960 to 1970 the number of coal miners fell by 230,000, or around 80 per- cent (see table). Changes in the Number of Coal Miners Existing coal Number of miners workers Average age 1955 750 349,578 34.5 1960 622 302,768 36.1 1965 222 151,455 38.5 1970 74 70,565 41.2 1975 35 36,073 42.3 1980 25 30,070 42.4 1985 26 24,082 41.2 1990 21 7,977 42.4 1995 13 4,629 43.0 Source: Ministry of International Trade and Industry, Annual Re- port on Energy Production and Supply/Demand. In 1959 the Displaced Coal Mine Workers Temporary Measures Law was es- tablished to fundamentally rationalize the coal mining industry. This law was re- vised in 1962 to integrate and downsize the industry and assist coal miners in finding other jobs. The law had several key elements designed to help the dis- placed workers. First, all coal miners received job application "pocketbooks" designating the holders as displaced coal miners. Employers hiring pocketbook holders received incentives and housing allowances. Miners who started their own businesses re- ceived incentives or debt guarantees to help in borrowing start-up funds. This method of subsidizing the employment of a targeted group of workers became an important pillar of Japan's employment policy. Second, the government implemented a method of introducing jobs available in other areas to displaced coal miners. Under Japan's Employment Security Law, (Box continues on thefollowing page) 478 EAST ASIAN LABOR MARKETS BOX 11.1 (continued) job seekers were to be matched with local jobs. Finding employment opportuni- ties for displaced coal mine workers in the same locale was difficult, however. The result was a plan that provided workers with information about jobs in other areas in order to encourage geographic mobility and promote reemployment. Re- locating workers received funds for moving and housing. Third, the government provided various educational and training programs to make job transfers easier. Unemployed miners remained eligible for job-brokering services and vocational training longer than for other unemployment benefits. Fourth, public works projects were specifically designed for displaced miners. Initially, most displaced coal workers were absorbed into existing projects, in- cluding (among others) river development, erosion control, and land develop- ment. If 85 percent or more of employees in any one project were displaced coal miners, their wages were heavily subsidized (80 percent of a separately set unit price). These measures highlight the shift in Japanese employment policy from ab- sorbing unemployment by providing jobs directly to subsidizing employment in other sectors. This approach dominated employment policy in and after the 1 960s. Coal mining companies cofinanced these measures when miners were in- volved (55 yen per ton of coal produced), setting an important precedent for fu- ture policy. such as Japan. Conserving resources and changing the industrial structure became important policy issues in Japan as this scenario unfolded. The Labor Market Situation Rapid price increases attributable to the oil price shock dampened eco- nomic growth and employment expansion during this period. Energy- intensive, large-scale heavy industries such as steel and shipbuilding, which had previously taken the lead in the Japanese economy, faced a cri- sis. Companies responded by reducing their workforces. As the industries most seriously hit by the recession were geographically concentrated in company towns, local unemployment emerged as a key problem. Although Japan's unemployment rate had remained at the 1 percent level for 20 years, it reached 2 percent in 1976. This decline marked the beginning of permanent unemployment due to structural rather than JAPAN'S EXPERIENCE 479 merely cyclical influences. The unemployment rate has not dipped below 2 percent since this time. Another key labor market trend became visible during this period as well-the aging of the workforce.' 1 The growth rate of the senior population was perceived to require specifically targeted pro- grams (box 11-2). Employment Policies during the Low-Growth Period The rapid escalation of unemployment and underemployment forced the government to reassess the Second Employment Measures Basic Plan after only two years. In December 1975 the Ministry of Labor's Employment Policy Research Institute proposed an employment policy for the decade beginning in 1975 that focused on full employment amid slow economic growth. This proposal took note of the country's success in reaching lev- els close to full employment during the period of economic growth. This achievement allowed the government to join the governments of other in- dustrial economies in using monetary and fiscal policies to smooth fluc- tuations in the business cycle. The Third Employment Measures Basic Plan, drawn up in June 1976, was based on these ideas. This plan's basic concepts included employment measures appropriate to an environment of slow economic growth. These measures had two objectives. First, they were designed to promote changes in the structure of both the labor force supply and the country's industries. But the second objective-balancing supply and demand in the labor force-was the more important, addressing as it did the question of how to lower the unemployment rate. UNEMPLOYMENT INSURANCE AND ADJUSTMENT SUBSIDIES. A compre- hensive employment insurance scheme was implemented in 1975, replac- ing the 1947 system that guaranteed livelihoods during periods of un- employment. In addition to providing income support, the new system established a fund for "active" labor market programs such as vocational training and employment subsidies. Employers and employees would each pay 50 percent of the costs of unemployment insurance. The research group that designed the insurance scheme also created an employment adjustment subsidy scheme. The Ministry of Labor had es- tablished the group in May 1973 to develop an active manpower policy under the assumption that Japan's economic growth would continue. Early in its deliberations, however, the group realized that a national economic slowdown or recession might soon create an unemployment situation that 480 EAST ASIAN LABOR MARKETS BOX 11.2 Employment Measures for Older Workers Employment measures for older workers began in earnest in 1975 with discus- sions surrounding the extension of the mandatory retirement age. In the mid- 1970s the most common retirement age at large companies was 55, clearly too young in terms of working capabilities. By the mid-1980s the retirement age had risen to 60 (see table 1). The extension of the mandatory retirement age ignored individual differences among older workers. Depending on the situation and the choices available, older people have different ambitions, possibilities, and needs with regard to continuing employment. Extending the retirement age to 60 did not solve the employment issues sur- rounding older workers. In August 1979 the Fourth Employment Measures Basic Plan was drawn up with the goal of expanding employment opportunities for people in their early 60s who have a strong desire to work. The plan had three basic goals: * To extend employment by raising the retirement age again, with considera- tion for the conditions of individual companies; • To promote reemployment after formal retirement or extend the working period by providing employment subsidies and other benefits to companies willing to hire retired and older workers (see table 2); and * To promote self-employment, part-time employment, and various other forms of employment utilizing older workers' experience and abilities, with consideration for individual preferences. In 1986 the government enacted a law that prohibited companies from setting the retirement age lower than 60. But this prohibition did not automatically guar- antee employment until the age of 60. According to a 1992 government survey, BOX TABLE 1 Changes in the Mandatory Retirement Age (% of company surveyed) 55years oryounger 56 to 59 years 60years or older 1980 39.7 20.1 39.7 1985 27.1 17.4 53.3 1990 19.8 16.1 63.9 1995 7.6 6.5 85.9 1999 0.5 0.4 99.2 Source: Ministry of Labor, Employment Management Survey. BOX TABLE 2 Allowances Provided for Elderly Employment 1991 1992 1993 1994 1995 1996 1997 1998 Continual employment system implementation incentive (1,000 yen)a 8,330,400 9,939,200 10,421,200 11,437,200 11,371,600 22,377,200 5,601,000 14,910,280 (Number of companies) 1,591 1,894 2,008 2,133 2,226 4,371 4,542 11,279 Elderly employment environment development incentive (1,000 yen) 24,000 139,500 207,500 323,500 268,000 (Number of companies) 6 32 66 114 99 Special job seekers' employment development subsidy (1,000 yen) 31,331,327 30,500,582 29,937,076 40,946,946 73,869,654 48,913,556 40,083,306 (Number of people) 91,428 89,007 93,239 113,961 112,774 114,923 Elderly employment continuation subsidy (100 million yen) 120 370 570 770 (Number of people) 75,416 84,447 89,487 100,174 Source: Ministry of Labor Statistics. a Owing to a revision of the system in 1987, the numbers for continual employment system implementation incentives are not directly comparable to those prior to 1987. b Only the amount paid with regard to employment of the elderly has been included. (Box continues on the following page) 482 EAST ASIAN LABOR MARKETS BOX 11.2 (continued) people in the age group 60-64 who were employed at the age of 55 and who worked until the mandatory retirement age constituted only 37.7 percent of all workers that age. Employing older workers became even more urgent when the minimum age to begin receiving an old-age pension was raised from 60 to 65. The easiest way to way to maintain income for workers between the ages of 60 to 65 would be to secure continued employment at the same company. Maintaining older workers on the job would also be an effective way of utilizing job skills accumulated over a long career. Recognizing that skill levels are not always enhanced as workers age, however, the government chose to establish a system of subsidies for continued employment (or reemployment) rather than mandating another increase in the retirement age to 65. The plan established a "continual employment system" that would provide subsidies for five years to companies setting the retirement at age 60 or older that employed workers continuously until the age of 64. The subsidies amounted to between 1 million and 2.5 million yen each year, depending on the size of the company. In addition, eligible companies where workers ages 60 and older made up at least 10 percent of the workforce could receive 30,000-40,000 yen per month for each elderly worker for a maximum of five years. These subsidy systems have repeatedly been revised and expanded. Companies have used the subsidies as a vehicle to establish and improve a proper environment for retaining and hiring senior workers. In addition these subsidies have facilitated the development of measures to expand employment overall. These measures re- quire an established employment insurance with ample funds and therefore can- not be used as a model for every country. However, the idea behind the subsidy system may be applicable to active labor market programs in many countries. Japan is also taking a unique approach to elderly employment with its Elderly Employment Centers. Local governments in areas with more than 6,000 workers are eligible for a center, but they must prove that there is substantial need for one. As many as 1,417 centers had been set up as of the end of September 1999. The volunteer-run centers, which receive government subsidies to cover some man- agement expenses, offer temporary, short-term jobs in areas that support the local community, including professional services such as editing and translation, sim- ple clerical work, and supervisory services for local parks and parking lots. Work- ers at the centers may have specialized skills, or they may provide basic household services. Revenues are distributed among member workers. The centers not only help find jobs but also offer members something fulfilling to do during the tran- sition from work to full retirement. JAPAN'S EXPERIENCE 483 simple adjustment measures could not handle. Such a scenario would re- quire measures aimed at preventing unemployment as well as assistance for companies that needed help adjusting their employment levels. The result of this thinking was the Employment Adjustment Subsidy Scheme. It was designed to intervene before companies were forced to lay off workers rather than to support workers after they had been laid off. The program provides subsidies to companies that use temporary layoffs or retraining rather than permanent layoffs during times of economic cri- sis. Employers pay the costs of the program, and any subsidies the compa- nies receive cover a portion of workers' wages. In effect the system supports employment maintenance in the internal labor market by subsidizing a company's efforts to retain its workers. Recognizing the importance of the scheme, the Diet implemented it before the rest of the unemployment in- surance program in order to prevent large-scale retrenchments. Such sub- sidies soon became an integral part of Japan's employment policy. PUBLIC EMPLOYMENT SERVICES. In order to handle the deteriorating employment situation after the first oil crisis, the government enhanced the functions of the PESOs. In 1974 it began requiring companies con- templating substantial layoffs to notify the local employment office in advance. This regulation was designed both to provide the local office with timely and accurate information about the employment situation in its area and to facilitate ability to provide needed services. In addition the unemployment insurance payment system was automated and the PESOs reorganized. The staff's level of expertise was enhanced through training. Finally, job-matching services were revamped to include a clas- sification system that speeded up the process of referring job seekers to appropriate jobs. The Current Recession (1992-2000) Japan's "bubble economy" collapsed in 1990. Because this period lasted so long, the subsequent recession has been the most serious since the end of World War II. The proximate cause of the recession is generally viewed as the accumulation of too many bad debts by financial institutions. The sub- sequent tight financial market has led to stagnation in investment and in- dividual consumption. At the same time the Japanese economy has lost its relative advantages in mass production and certain "low-cost" industries. In 2000 Japan is still struggling to escape from the recession, and the employment and unemployment situations are more difficult than ever. 484 EAST ASIAN LABOR MARKETS The unemployment rate reached 4 percent in 1998 and 4.9 percent in June 1999-very close to the 5 percent level. The Current Labor Market Situation Several scenarios are being played out on the personnel front. Some work- ers have been discharged on a nonvoluntary basis from failing companies. Others have applied for early retirement and have left their jobs voluntar- ily, only to experience difficulties in finding new employment. The num- ber of workers who have left their jobs voluntarily is about the same as the number who have been discharged (table 11.2). Three groups in particu- lar have been unable to find employment: new graduates, those trying to enter the labor market to maintain their standard of living, and seniors reentering the market after a year or more. With so many challenges to address in the labor market, many ob- servers are beginning to realize that the conventional employment mea- sures taken in the past may not be sufficient in the current situation. The situation Japanese society faces is the result of structural change, and the view that a drastic change in employment policy is necessary is becoming dominant. Concern is growing that the current employment policy frame- work functions merely as a collection of individual employment policy measures and that what is required is a complete reassessment and reorga- nization of the country's employment policy at the macro level. As a result, although the eighth version of the Employment Measures Basic Plan was designed to cover six years (1995-2000), it has had to be reassessed dur- ing that period. A ninth Employment Measures Basic Plan has been de- veloped to cover the years from 1999 to the early part of the 21st century. Employment Policies The diverse populations of unemployed workers that characterize the cur- rent recession complicate employment policymaking. Conventional poli- cies for the unemployed, which primarily emphasize maintaining employ- ment in the in-house labor market, are being rethought. The idea of lifetime employment that has long permeated the Japanese labor market is one of the most important issues. It has long been believed that once people enter a company in Japan, they work for that company for their entire career.12 As a result, labor mobility among companies is very low. If the government develops policies that take this employment relationship as the norm, the primary emphasis is naturally on in-house measures. However, lifetime em- TABLE 11.2 Why Unemployed Workers Had Left Their Jobs (percent) 15-24years 25-34years 35-44years 45-54years 55 years and older Office closure, corporate bankruptcy, closure of a self-employed person's business 1.9 5.4 12.8 7.0 8.8 Dismissal, restructuring 5.7 8.1 15.4 25.6 19.1 Mandatory retirement, termination of the employment contract period 5.7 5.4 5.1 4.7 44.1 Potential future instability in the workplace 3.8 5.4 5.1 4.7 1.5 Deterioration of working conditions 13.2 9.5 12.8 7.0 4.4 Working conditions that differ from those expected prior to obtaining the job 11.3 9.5 5.1 4.7 0.0 Want to try some other job 18.9 12.2 7.7 4.7 1.5 Marriage, childbirth, childcare 3.8 9.5 5.1 2.3 0.0 Nursing care, housework, commuting, illness, etc. 5.7 9.5 7.7 9.3 4.4 Other 30.2 25.7 23.1 30.2 16.2 Source: Bureau of Statistics, Special Labor Force Survey, August 1999. 486 EAST ASIAN LABOR MARKETS ployment arrangements are changing. In 1998 Japanese companies hired 5.4 million workers, nearly 60 percent of them from other industries and companies (MOL 1998). This trend is not just a recent phenomenon but one that has developed slowly over the past 30 years. Clearly the current recession demands a different kind of thinking. The industrial structure needs substantial restructuring, and the labor force needs to be reallocated among different industries. New growth areas must be supported in an effort to create and expand employment opportunities. In addition, workers must be able to develop occupational skills that match the new employment opportunities. And finally, the government needs to develop measures to shorten the duration of unemployment. The latest employment policy envelope is depicted in figure 11-1. The new agenda reflects a policy shift from employment protection to job cre- ation and support for job transfers. Policy initiatives announced by the Ministry of Labor in January 2000 divide policies into three categories. The highest priority is employment development, which encompasses the following four initiatives: subsidies to promote employment creation among small and medium-scale enterprises (SMEs), a program targeting middle-age and senior workers, support for infant industries, and public works projects for local communities. The first initiative is the Special Subsidy System for Regional Employ- ment Creation by SMEs. The Ministry of Labor and prefectural govern- ments work together to choose promising companies from among SME start-ups to serve as models for the program. The program provides wage subsidies and allowances for employee benefits, human resource devel- opment, and the development of an employment management system. While past subsidies have been used to target particular groups of workers such as the elderly, this system targets particular companies that could be important employers. According to Ministry of Labor statistics, the pro- gram provided companies with subsidies for over 60,000 workers in 1999. The second measure, the Special Subsidies for the Emergency Creation of Employment, targets unemployed middle-aged people or seniors who leave their jobs on a nonvoluntary basis in regions where employment conditions have deteriorated badly (see table 2, box 11-2). A subsidy of 300,000 yen per person is available to companies that hire such workers. In 1999 this system was implemented in Okinawa, the Kinki district, and the southern Kanto district. To date the program has received subsidy ap- plications for 932 people. The third measure is the New Growth Area Employment Production Special Incentive, which provides financial support to companies in 15 FIGURE 11.1 Changes in Employment Policy Measures to support employers' efforts to maintain employment Employment Adjustment Subsidy (financial assistance to employers that retain employees through such methods as temporary leave, transfers, and training rather than dismissing workers). Measures to support labor shifts Measures to support the creation ofjobs Measures to support the upgrading without unemployment of human resources Support for business ventures, new busi- Support for shifts in labor among ness development (small and medium-sized Efforts to foster high-level and creative companies and industries brought about companies), and job creation in districts human resources to help companies by changes in the industrial structure. with severe employment conditions. develop new business areas and produce * Subsidies for companies that make * Assistance with personnel expenses for goods with higher value added. efforts to reassign workers, transfer workers engaged in establishing or * Promotion of the upgrading of public workers to other companies, introduce developing new business areas; assistance vocational training. places of reemployment, conduct for expenses and wages relating to in- * Support for the upgrading of training in training, and make capital investments. house and outside training; and wage private companies. * Implementation of consulting, seminars, subsidies for districts with severe * Promotion of vocational skills and so forth for companies transferring employment conditions. development for white-collar workers. or receiving workers. * Special matching system for venture * Promotion of individual vocational skills businesses, human resources, and development. consulting. 488 EAST ASIAN LABOR MARKETS growing industries such as information and communications. The Min- istry of Labor expects to create employment opportunities for 150,000 people through this measure. The fourth measure is the Emergency Local Employment Special Sub- sidy, under which each prefectural government provides a portion of the funds required to develop public works projects around the needs of local communities. Owing to the problems associated with public works proj- ects implemented just after World War II, the Japanese government has long held negative views about work relief projects designed to absorb the unemployed directly. However, as high unemployment continues, the government has decided to implement measures to absorb some unem- ployed workers (but within certain time limits). The Ministry of Labor expects to create employment opportunities for about 290,000 people through this measure. It is too soon to comment on the effectiveness of these measures, but the shift in focus this new policy represents-from employment mainte- nance to job creation-is significant. We can demonstrate this point even more clearly by assessing the employment adjustment subsidy that has been provided since October 1999. As stated earlier, supplementing and enhancing the employment maintenance function with employment ad- justment subsidies has been the main pillar of employment policy in Japan since the late 1 970s. But given the risk of hindering labor mobility by sub- sidizing stagnant industries, the government has revised its criteria for tar- geting industries. The primary focus is now on providing support for "temporary employment adjustments in response to economic fluctua- tions" (MOL 1999). By excluding industries that are stagnant for struc- tural reasons, the government sets the stage for transferring workers from these industries to growing industries. In addition to measures that address the specific concerns of the current recession, the government continues to rely on the services of the PESOs. The PESOs play an important role in adjusting the balance between sup- ply and demand in the labor market. As of 1999 Japan had 478 PESOs (26 main offices and 111 branch offices) with a total staff of 13,000 people. The labor force adjustment system for supply and demand is currently being restructured in order to determine which types of information and methods of job placement are most effective in the current employment situation. And in 1999 the Worker Dispatching Law and the Employment Security Law were revised to liberalize the types of jobs that employment agencies are allowed to handle. The revision has made it possible for pri- vate employment exchange services to broker jobs, and as a result private JAPAN'S EXPERIENCE 489 employment agencies are becoming increasingly active. It is too soon to judge the impact of either the restructuring or the new laws, however. Directions for the Future Japan is at a major turning point in terms of its employment policies. The problem is not just that previous policies have lost their effectiveness and that new policies need to be drawn up. Rather, the policy structure itself must be reconceptualized. Lessons from history are essential to this en- deavor. The situation requires analyzing the objectives of previous em- ployment measures, the effectiveness of the methods applied, and the tar- gets of such measures. To this end this chapter has analyzed employment policies through four phases of Japan's economy since World War II. During the initial postwar period, employment policy focused on di- rect job provision to absorb the masses of unemployed workers. The pro- grams functioned essentially as government transfer programs, providing primarily simple low-wage jobs to anyone in need of income. While such policies may be appropriate for short-term relief, they do little to stimu- late productive, long-term employment. In Japan workers remained in these jobs for long periods, and the government itself became responsible for having created employment opportunities with poor labor conditions. The key to job creation programs lies in appropriately targeting a pool of unemployed workers with certain attributes and providing jobs that allow these workers to display their abilities. Policies designed to alleviate short periods of unemployment resulting from economic downturns must build in incentives for workers to seek jobs outside of the public works projects as the economy recovers. As the Japanese experience proves, a system with- out such incentives tends to fall victim to social inertia and becomes ex- tremely difficult to discontinue. Employment policy during the period of economic growth in the 1960s, as initiated under the National Income Doubling Plan, sought to provide a continuous supply of labor for growing industries and a smooth transition for workers retrenched from sunset industries. The First Em- ployment Measures Basic Plan played an important role in promoting a balanced supply of labor to meet demand. The shift to introducing em- ployment opportunities from a wider geographic area in order to encour- age labor mobility promoted flexibility in the labor market and improved the ability of the labor force to respond to structural changes. The adapt- ability encouraged by such labor market policies supported the remarkable economic growth of this period. 490 EAST ASIAN LABOR MARKETS During the 1970s the focus of Japanese employment policy shifted to internal labor markets. The centerpiece was the employment adjustment subsidy scheme, which was based on the belief that unemployment could be prevented as long as employers received the right incentives and subsi- dies at the right time. This and similar assumptions underpinned many of the measures initiated at the time, including the extension of the retire- ment age and other initiatives directed at older workers. But such policies, which are based on encouraging employment maintenance, ultimately tend to hinder labor mobility. Subsidies may be workable when the in- dustrial structure is relatively stable and excesses in the labor supply are temporary. But employment policies based on the internal labor market and typified by employment adjustment subsidies are best introduced only after such conditions are verified. The serious long-term recession that has affected Japan since the disin- tegration of the bubble economy in 1992 proves the importance of con- sidering both the industrial structure and the macroeconomic conditions of an economy in developing employment policies. Currently, the gov- ernment is promoting new management strategies that jettison conven- tional management methods in an effort to speed recovery. Attention has turned to the promotion of businesses in nontraditional fields, and the growth of such companies in this sector has been striking. The govern- ment is also once again considering policies to facilitate the smooth real- location of labor. The success of these strategies remains to be seen. They draw on Japan's experience and are specifically geared toward meeting the country's current employment challenges. Policymakers from other Asian economies may find Japan's experience in developing employment policies useful in designing strategies to address their own employment issues. Notes 1. We do not offer encyclopedic information about Japanese employment policy, nor do we offer actual technical support for implementing employment policies. Policymakers who plan to apply the lessons from the Japanese experience must determine how best to use them. 2. In order to focus more clearly on the impact of employment policies dur- ing recession and recovery, we do not cover the bubble economy in depth. 3. There are no reference materials available that we can use to compare the employment situation prior to 1953 with the current situation. The most com- prehensive data available on Japanese employment conditions is the Labor Force Survey conducted every month by the Statistics Bureau of the Management and Coordination Agency. However, the size of the respondent pool and the methods JAPAN S EXPERIENCE 49I used to analyze the results have been revised a number of times. The revisions made in 1961 were particularly substantial, creating a considerable gap between the new results and those from earlier years. The survey group was able to recal- culate previous years' indicators only back to 1953. 4. According to the Bureau of Statistics, of the total workforce of 25.66 mil- lion, the number of self-employed people and their family workers earning annual incomes of 25,000 yen or less came to 5.14 million (20.0 percent). At the same time the number of employees (including those under the age of 20 earning monthly salaries of 3,000 yen or less and those ages 20 and older earning 4,000 yen or more per month) was 1.82 million (7.1 percent). The number of peo- ple, including the unemployed, whose income could not be determined came to 3.96 million. 5. The first Economic Recovery Plan, released in January 1946, was titled "Urgent Issues to Be Tackled to Resolve the Employment Problem." 6. In accordance with Article 19 of the 1947 Employment Security Law. 7. Under the dual structure, the labor force was divided into two groups: those who worked in large concerns, and those who worked in small- to medium- size enterprises. 8. In 1981, in order to help seniors and the infirm become independent and leave public jobs, the government offered a lump-sum payment of 1 million yen to those who withdrew from public works programs. This plan led 18,000 peo- ple to leave their jobs. In addition the government abolished or downsized a num- ber of extremely inefficient or uneconomical public works, and in March 1996 the Congress passed a law abolishing public works projects intended to absorb the unemployed. The difficulty of moving workers off the projects underscores the importance of learning from past mistakes. 9. Known as "Golden Eggs," these graduates were generally 15 or 16 years old. 10. Clause 2 of Article 19 in the Employment Security Law stipulates that ef- forts be made to direct job seekers to jobs that do not require them to move. However, in 1960 the law was amended to allow Public Employment Service staff to direct workers to other regions when appropriate work cannot be found locally. 11. Japan's senior population (65 years old and older) was only 9.0 percent in 1980, the lowest rate among advanced industrial nations. However, the rate was projected to rise to 16.4 percent in 2000, the same level seen in other advanced nations, and to reach 25.2 percent in 2020. This increase would make one of every four Japanese a senior citizen by 2025. 12. One of the first books on Japanese employment customs widely read out- side Japan was The Japanese Factory (Abegglen 1958), which offers this descrip- tion of lifetime employment: At whatever level of the organization in the Japanese factory, the worker commits himself on entrance to the company for the remainder of his 492 EAST ASIAN LABOR MARKETS working career. The company will not discharge him even temporarily ex- cept in the most extreme circumstances. He will not quit the company for industrial employment elsewhere. He is a member of the company in a way resembling that in which people are members of families, fraternal organi- zations, and other intimate and personal groups in the United States. References Abegglen, James C. 1958. The Japanese Factory. Cambridge, Mass.: MIT Press. Bureau of Statistics. Various years. Labor Force Survey. Tokyo. ILO (International Labour Organization). Action against Unemployment. Paris. MOL (Ministry of Labor). 1960. "Special Survey on Workers in the Work Relief Projects." Tokyo. . 1966. First Employment Measures Basic Plan (March 1966). . 1973. Second Employment Measures Basic Plan (anuary 1973). . 1998. Survey of Employment Trends. Tokyo. . 1999. "Emergency package of Measures to Increase Job Security and Strengthen Industrial Competitiveness." Tokyo. EAST ASIAN LABOR MARKETS AND THE ECONOMIC CRISIS The recent financial crisis in East and South East Asia had sudden and painful consequences for millions of workers and their families. Falling employment and earning opportunities in the formal sector forced them to survive on reduced incomes, informal activities, and meager social assistance. The World Bank and the International Labour Office (ILO) collaborated to explore ways in which they could support the region's governments, employers, and workers in their response to the difficult social and economic situation. Resulting from the collaboration, this joint study covers the Republic of Korea, Indonesia, Malaysia, the Philippines, and Thailand. It addresses the labor market impacts of the crisis, examines how the various countries responded to the crisis, and identifies needed labor policy reforms for the longer run, particularly in the areas of unemployment benefits, active labor market programs, support for vulnerable groups, and social dialogue. The book will be a valuable resource not just to policymakers and researchers but to everyone interested in a world where economic crises are a fact of life and where unfamiliar labor market conditions pose a new set of chailenges. The World Bank 1818 N Street N.W. Washington, D.C. 20433 USA Telephone: 202-477-1 234Il Bl Facsimile: 202-477-839 Intemet: www.worldbank.org E-mail: feedbackOworldbank.org ISBN 0-8213-4478-1 ISBN 92-2 112274-3 (ILO)