1Editedby FunkooPark, Young-burn Park, Gordon Betcherma.n,Amit Dar A THE WORLD BANK I KLIKOREALABOR INSTITUTE Labor Market ReformsinKorea: Policy Optionsfor theFuture Edited by Funkoo Park Young-bum Park Gordon Betcherman Amit Dar The World Bank KoreaLabor Institute copyright0200I The World Bank and Korea Labor Institute All rights reserbed, Except for the brief quotations in a review, this book, or parts therefor, must not be reproduced in any form vvithout pennission in writing for the publisher. For information, address Korea Labor Institute, 16- 2 Yoido, Youngdungpogu, Seoul, Korea (www@,kli.re.kr) First published 2001 by the World Bank and Korea Labor Institute Park. Funkoo. Young-bum Park. Gordon Betcherman and Amit Dar, editors ISBN 89-7356-326-2 ThisBook is dedicated to the memory of Nicholas Prescott(1953-2QQQ) Foreword In the wake of the 1997 financial crisis, Korea experienced unprecedented economic and social disruptions. In the course of combating the crisis, institutionalizing labor market flexibiliry and building a stronger and more durable safety net stood out as challenging policy issues. The government, employers, labor leaders, and academics devoted themselves to resolving these issues in order to mitigate the painful costs of the crisis and to move forward afterwards. In this context, the World Bank and the Korean Ministry of Labor launched a project on labor market reforms in Korea. This project was financed through a grant from the ASEM-EU Financial Crisis Fund, with the Korea Labor Institute as the major implementing agency. The project aimed to assess the progress of the Korean labor market reforms after the 1997 financial crisis and provide policy directions for the future so as to respond to challenges created by changing economic and social circumstances. This book brings together the wide range of studies that were commi- ssioned through this project. It covers income support programs, labor law and labor market regulations, active labor market programs, labor market and gender, and the Korean experience in an international comparative context. Many prominent local and international experts from academia, government and international organizations including the World Bank, OECD and ILO participated in the preparation of these studies. An international conference was held in Seoul in May 2000 to discuss the preliminary outcomes of the project. At this conference, experts from other Asian crisis countries also shared the experiences and lessons of their countries from the crisis. The ASEM grant facility has played a major role in ensuring the success of this project and 1 would like this opportunity to thank the EU for their substantial and continuous support for this exercise. Many people also contributed to the successful implementation of the project and the publication of the book. The main initiator of the project was Nicholas Prescott (World Bank) who tragically passed away before this project could be completed. This project would not have been undertaken without his enthusiasm, knowledge, and efforts and we have chosen to dedicate this book to his memory. I would like to thank Funkoo Park (former President of the Korea Labor Institute), Young-bum Park, and Sri-Ram Aiyer (former Director of Korea Country Ofice, the World Bank) who were closeIy involved with this project right from the very beginning. 1 would also like to gratefully acknowledge the assistance and contribution of many colleagues at the World Bank. Specifically, I am heavily indebted to Tamar Manuelyan Atinc, Gordon Betcherman, Amit Dar and Dalhoe Coue (former staff of Korea Country Office) for their continued collaboration during this project. I would like to express my deepest appreciation for the support provided by the staff of the Korea Labor Institute, including Mr. Chanyoung Park, chief of the publication division. I sincerely hope this book will contribute to a wider understanding of labor market issues and serve to stimulate a productive dialogue as policymakers map options for the future. Wonduk Lee President Korea Labor Institute Contents Foreword v 1.Introduction 1 I . Overview 3 Funkoo Park, YoungbzfmPark, Gordon Betchermanand Amif Dar 11. Income Support Programs 23 2. Expanding the Coverage of Korea's Unemployment Insurance 25 Jai-Joon Hur 3. The Potential Role of an Employinent Guarantee Scheme in 46 Korea's Social Safety Net Gary S. Fields 4. Feasibility of Introducing a Non-contributory Cash Benefits 53 System for the Unemployed in Korea Young-bum Park 5. Unemployment Assistance: Policy Implications from the Crisis of 1997 77 Alan Abrahart 6. Assessing the Coverage and Effectiveness of the Wage Claim 99 Guarantee System in Korea Tai-gi Kim, Kung-Shik Choi and Joyup Ahn 111.Labor Law and Labor Market Regulations 123 7. Assessment of Labor Market Response to the Labor Law 125 Changes Introduced in 1998 Duck Jay Park, Jonghee Park and Gyu-Chang Yu 8. Labor Market Flexibilityand Employment Agencies in the UK, Japan 151 and Korea Barry Wilkinson IV.Active Labor Market Programs 173 9. Income Assistance and Employment Creation through Public Works 175 in Korea Joohee Lee 10. Public Works Programs in Korea: A Comparison to Active Labor 196 Market Policies and Workfare in Europe and the US Ivar Lsu'emel and Espen Llahl 11. Review of the Effectiveness of Public Employment Services 215 Jaeho Keum 12.Evaluating the Training for the Unemployed in Korea 248 Soon-Hie Kang and Byung Hee Lee 13. Publicly Funded Training for Unemployed Adults: Germany, 265 the UK and Korea Paul Ryan V. Labor Market and Gender 293 14. Were Korean Women More Affected by the Asian Crisis? 295 Kye Woo Lee and Kisuk Cho VI. The Korean Experience in an International Context 325 15. Labor Adjustment, Non-Standard Work,and Employment Programs: 327 Korea in an OECD Context Gordon Betcherman, Amit Dar, Amy Luinstra and Makoto Ogalva 16. Korean Labor Market and Social Safety-Net Reforms: Challenges 358 and Policy Requirements John P. Martin and Raymond Torres 17. The Challenge of Labor Reform in Korea: A Review of Contrasting 386 Approaches to Market Enhancement and Experiences from Chile and Deninark Lozlise Haagh Bibliography 420 List of Tables Table 1.1 : Selected labor indicators in Korea, 1995-2000 Table 1.2: Expenditures of social protection for the unemployed in Korea, 1997-2000 Table 2.1: Number of insured employees and estimated number of employees to be covered by the Korean UIS, December 1995 to December 1999 Table 2.2: The number of regular and temporary employees and actual number of insured employees of the Korean UIS by firm size, December 1999 Table 2.3: Number of regular and temporary employees and actual number of insured employees of the Korean UIS by industry, December 1999 Table 2.4: Ratio of UB recipients with respect to each category of the unemployed, December 1999 Table 2.5: Employees and unemployed employees by employment status: December 1999 Table 2.6: Tenure period, employed period, employment period and unemployment period for workers by employment status, June 1999 Table 2.7: Temporary and daily employees by firm size in Korea, December 1999 Table 2.8: Temporary and daily employees by industry in Korea, December 1999 Table 2.9: Tenure and employed period by employment status in Korea, August 1998 to July 1999 Table 4.1 : Unemployment benefits claimantsin Korea. 1997-1999 Table 4.2: Comparison between Livelihood Protection Act and National Basic Livelihood GuaranteeAct in Korea Table 4.3: Poverty and economic activities of survey respondent Table 5.1: Months of entitlement to unemployment benefit in Korea Table 5.2: Working age population in Korea Table 5.3: Unemploymentin Korea, July 1998 Table 5.4: Employment decl.ines(joblosses) in Korea, December 1997 to December 1998 Table 5.5: Labor force in Korea, 1999 Table 5.6: Government budget allocations on social protection in Korea, 1999 Table 5.7: Social protection in Korea, 1999 Table 5.8: Number of recipients of income support in Australia, 1965-1998 Table 6.1 : Amount and accumulated records of subrogated payments in Korea,1998 and 1999 Table 6.2: Subrogated payments made by type (wages and retirementallowances) Table 6.3: Subrogated payments by gender Table 6.4: Subrogated payments by age group Table 6.5: Subrogated payments by type of insolvency Table 6.6: Subrogated payments by firm size (confirmationof insolvency) Table 6.7: Characteristicdata of workersand their households Table 6.8: Deferred payment per person Table 6.9: Determinantsfor the size of deferred wages: 2SLS Method Table 6.10: Severity for the confinnation of insolvency: logit analysis Table 7.1: Impacts oftheenactment of act concerningcollective dismissals on regular employees Table 7.2: Main reason of employment adjustment Table 7.3: Change of regular employees of com~nercialbanking sector in Korea; 1997-1998 Table 7.4: Characteristicsof the firms that carried out dismissal Table 7.5: Characteristicsof firms that carried out 'honorable' retirement Table 7.6: Characteristicsof f m s that carried out spin-offor early retirement Table 7.7: Payment of compensation for 'honorable' retirement Table 7.8: Number of firms having new recruitment among the firms that carried out dismissal or 'honorable' retirement Table 7.9: Recent trends in temporary work agencies in Korea, 1998 Q4-1999 Q2 Table 7.10: Trends in the average number of workers per TWA, 1998 Q4-1999 Q2 Table 7.1 1: Companies using dispatched workers Table 7.12: Trends in number of dispatched workers used per firm Table 7.13: Future plan to use dispatched workers Table 7.14: Companies using dispatched workers by industry Table 7.15: Finns regarding using dispatched workers by firm size Table 7.16: Reasons why firms use dispatched workers Table 7.17: Employers' opinions regarding the impact of the act concerning temporary work agencies Table 7.18: Employers' opinions regarding the amendment of the act concerning the temporary work agency Table 7.19: Unions' opinions regarding the impact of the act concerning temporary work agencies Table 7.20: Unions' opinions regarding amendment of the act concerning temporary work agency Table 7.2 1: Firms that had a labor-management dispute in the process of employment adjustment Table 7.22: Most difficult thing to firms in processof employment adjustnient Table 7.23: Union's positionsto the attempt of firm's employment adjustment Table 7.24: F&S concerning redundancy package Table 7.25: Reasons for paying redundancy package Table 7.26: Firms that acted for outplace~nentand morale enhancement Table 8.1 Temporary staff on agency payrolls in the U.K., 1997-1998 Table 9.1: Changes in wage rates for public works projects in Korea Table 9.2: Characteristics of the public works project participants in Korea, 1998-1999 Table 9.3: Descriptive statistics for survey respondents' characteristics (totaI=1505) Table 9.4: Economic status of survey respondents Table 9.5: Economic status of survey respondents by sex Table 9.6: Types of benefited unemployment policies (totak248) Table 9.7: Employment structure of those employed before and after PWP participation (total=before:750; after:399) Table 9.8: Logistic regression coefficients for regression ofjob attainment after PWP participation on selected independent variables Table 9.9: Job search methods of survey respondents Table 9.10: Five items measuring respondents' evaluation of public works Projects Table 9.1 1: OLS coefficients for the regression of PWP evaluation on selected independent variables (total number=1505) Table 10.1: Distinguishing characteristics of three forms of active policies targeted at unemployed people Table 11.1: Number of employment service agencies in Korea, February 1997 to December 1999 Table 11.2: Records of publicemployment offices in Korea, 1994- 1999 Table 11.3: Job search methods of the unemployed Table 11.4: Job search method of respondents Table 11.5: Difficulties in job-search activity Table 11.6: Reason for visiting PES agencies Table 11.7: Type of desired service and degree of satisfaction Table 11.8: Improvement directions of the PES-job searchers Table 11.9: Improvement directions of the PES-counselors Table 11.10: Characteristics of counselors Table 11.11: Counselor's duties and working time per week Table 11.12: Counselor's duties to be expanded Table 11.13: Duties of counselors to be curtailed Table 11.14: Reason for ineffectiveness of the PES Table 11.15: Job satisfaction of counselors Table 11.16: Public employment service in selected countries Table 12.1: Training for the unemployed in Korea, 1998 and 1999 Table 12.2: Comparison of personal characteristicsof re-employ- ment training for the unemployed participants and non-participants Table 12.3: Re-employment experience and current employment status of subjects depending on re-employment training for the unemployed participation Table 12.4: End point of training and beginning of non-employment depending on re-employment training for the unemployed participation Table 12.5: Employment trend depending on re-einployment training for the unemployed participation by spell length of non-employment duration Table 12.6: Trend of economically active status depending on re- employment training for the unemployed participation Table 12.7: Length of period from job loss or end of training to re- employment depending of re-employment training for the unemployed participation Table 12.8: Employment status upon re-employment depending on re-employment training for the unemployed participation Table 12.9: Firm size upon re-employment Table 12.10: Varizbles in re-employment hazard Table 12.11: Estimation of re-employment hazard Table 12.12: Current economically active status of re-employment training for the unemployed participants and non- participants Table 12.13: Current economically active status of those who are not re-employed by training category Table 13. I: Shares of training categories in publicly funded training for adult workers, old and new Bundeslander in Germany, 1993 Table 13.2: Attributesof public training programs for adult workers, old and new Bundeslander in Germany, 1993 Table 13.3: Attributes of public adult training programs in UK, 1986-1998 Table 13.4: Vocational qualifications gained by leavers from adult programs in the UK, 1986, 1994 and 1997 Table 13.5: Evaluations of adult training in Germany Table 13.6: Evaluations of training programs for adult workers in UK Table 13.7: Training activity by program in Korea, 1997 and 1998 Table 13.8: Attributes of training participants and of unemployed workers in Korea, 1998 Table 14.1: Economically active population, employment and unemployment in Korea, 1996-1998 Table 14.2: IJnemployment (including hidden unemployment) in Korea, 1996-1998 Table 14.3: Estimated unemployinent spell based on employment rates alone in Korea, 1996-1998 Table 14.4: Estimated une~nploymentspell (including transition to economicallyinactive population) in Korea, 1996-1998 Table 14.5: Employed persons by industry in Korea, 1997-1998 Table 14.6: Employed persons by occupation in Korea, 1997-1998 Table 14.7: Employment status in Korea, December 1997 and December 1998 Table 14.8: Involuntary disengagement in Korea, 1997-1998 Table 14.9: Average labor earnings by gender in Korea, 1997-1998 Table 14.10: Starting earnings level by industry in Korea Table 14.11: Starting earnings level by size of firms in Korea Table 14.12: Occupational concentration of female workers and gender earnings gap in Korea Table 14.13: Recruitment by the firms with/without trade unions Table 15.1: International comparison of GDP growth rates Table 15.2: Summary of economic trends in selected countries, 1980-1998 Table 15.3: Selected labor force characteristics, crisis countries, 1995-1997 Table 15.4: Summary of unemployment, underemployment, and employment trends in selected countries Table 15.5: Real wage trends in selected countries Table 15.6: Proportion of non-standard employment in Korea, 1990and1999 Table 15.7: Percentage change in employment by status in Korea, 1990-1999 Table 15.8: Employmentserviceagencies in Korea, 1997 and 1999 Table 15.9: Effectiveness of public employment service agencies in Korea, 1994-1998 Table 15.10: Training programs for the unemployed in Korea, 1998-1999 Table 16.1 : International comparison of employment/population ratios, 1999 Table 16.2: International comparison of employment by statusof workers, 1999 Table 16.3: Job tenure in Korea and selected OECD countries Table 16.4: Coverage of the employment insurance system(E1S) in Korea, July 1995 to February 2000 Table 16.5: Characteristics of unemployment benefit recipients in Korea. first semester of 1999 Table 16.6: Monthly benefit levels per household, under the 374 livelihood protection program in Korea, 2000 Table 16.7: International comparisons of multi-factor productivity 376 growth, 1980-1998 List of Figures Figure 2. 1: Structure of construction labor Figure 2.2: Four important aspects of the RERS Figure 5.1: Replacement and contribution rates for different unemployment scenarios in Australia (loo?' takeup) Figure 5.2: Proportions of working-age population (15-64) who were jobless, receiving income support andlor unemployed in Australia, 1965-1998 Figure 5.3: Net replacement rates of income support versus minimum wage, various non-renter income unit types in Australia, 1972-1997 Figure 9.1: Budget for public works projects in Korea, 1998-2000 Figure 9.2: Future direction of PWP Figure 11.1: Evaluation of the PES-job searchers, counselors and the survey of 1998 Figure 11.2: Job satisfaction of counselors and wage workers (1998 KLIPS) Figure 11.3: Educational mismatch between job seekers and vacancies Figure 1 I.4: Mismatch in occupational type between job seekers and vacancies Figure 13.1: Unemployment and labour market slack in Germany, 1991-1997 Figure 13.2: Rate of participation in publicly funcied adult training in Germany, 1990-1997 Figure 15.1: Wage and non-wage employment growth in Korea, 1980, 1990, 1997 and 1999 Figure 15.2: Percentage change in employment by status in Korea, 1990-1999 Figure 15.3: Employment change by gender and status in Korea, 1990-1999 Figure 15.4: Standard and non-standard employment trends in Korea, 1989-1999 Figure 16.1: Composition of employment by status of work in Korea, 1990-1999 Figure 16.2: Wage differentials (ratios) in Korea, 1980-1999 ~igure16.3: ~a&ci~ationof unemployed workers in different labour market programs in Korea, 1999 Figure 16.4: Financial incentives to work for social assistance clients in Korea, 2000 Figure 17.1: Linkages in an individual-centered system of labor flexibility Figure 17.2: Linkages between individual mobility and efficient flexibility, and policy implications xviii *pJf+iŁ>r1 r I INTRODUCTION flOverview & Funkoo Park* Young-bum Park" Gordon Betcherman"' Amit Dar"' Korea's financial crisis was as dramatic as it was unexpected. Between October and December 1997, Korea was reduced from the world's I lth largest economy to an economy surviving on overnight loans from international financial institutions. Consequently, Korea experienced a devastating economic decline. In 1998 real GDP fell by 5.8 % and the number of unemployed rose by over one million. The Korean government implemented wide-ranging labor market reforms to accommodate needed structural changes, and was determined to build a stronger and more durable social safety net so as to mitigate the painful social costs of structural adjustment. In order to promote labor market flexibility, in February 1998 the government repealed the two-year suspension in the application of Article 3 of the Labor Standards Act thereby permitting employment adjustment for economic reasons. At the same time, in order to facilitate quick redeployment of labor, it enacted legislation to allow the operation of private manpower leasing services, and relaxed restrictions on the private provision ofjob placement services. These developments were controversial. Confrontation between unions and employers concerning the flexibility measures increased as unemploy- ment continued to rise, reaching a record high of 1.8 million in February 1999. Union leaders and workers insisted that layoffs should be averted through wage cuts and work-sharing while management argued that layoffs were necessary for companies to overcome the economic difficulties. The measures to enhance flexibility were accompanied by interventions to improve labor market functioning. For example, the Korean government ' FunkooPark is with the Kim and Chang Law Ofice, Seoul, Korea. *' Young-bum Park is with Hansung University,Seoul, Korea. "' Gordon Betcherman and Amit Dar are with the World Bank, Washington D.C.. 4 Labor h4a1-ketReforms in Korea. Policy Options,for the Future increased the number and quality of public employment offices while making sure that they would complement rather than crowd out private employment offices. The coverage of labor market information systems was extended to monitor the effects of adjustment on less privileged workers who were employed in establishments with less than five employees. The government underpinned these labor market reforms with measures to cushion the shock of layoffs as well as to protect vulnerable groups by expanding workers' legal rights. It spent trillions of Korean won on projects to help the unemployed. The coverage of protective articles of the Labor Standards Act was extended to cover employees in establishments with fewer than five workers. Social protection for workers was strengthened by extending coverage of the Employ~nentInsurance Scheme (EIS) to all establishments in October 1998. Against this background, a grant from the ASEM-EU Financial Crisis Fund was provided to study the labor market impacts of the crisis; to evaluate the experience of interventions designed to alleviate the impact of the crisis; to learn from international experience on the design and implementation of labor policies and interventions; and, based on these diverse set of experiences, to develop policy recommendations for the Korean labor market in the medium and long-term. National and intemational experts were commissioned to undertake a wide range of studies on these topics. These studies were presented and discussed at an international conference in Seoul in May 2000. Following the conference, they were revised for this volume. The remainder of this introductory chapter proceeds as follows. First, Korea's labor market developments after the financial crisis are reviewed. Employment, wage and other major labor market indicators as well as industrial relations are discussed. The labor market situation before the crisis is also compared to the post-crisis situation. This is followed by brief summaries of the studies that comprise this volume. These are organized into five areas: income support programs, labor law and labor market regulations, active labor market policies, the labor market and gender, and Korean experiences in a comparative context. Critical assessments of the findings of the papers are also discussed briefly. These assessments are mainly based on comments and discussions made at the conference in Seoul. The chapter concludes with a discussion about emerging policy concerns and areas for further research. Labor Market Developments after the Financial Crisis 1987 was a turning point for labor in Korea, with the government initiating its withdrawal from its authoritarian approach to become a conciliator within OVERVIEW 5 a legal framework.' A revival of the labor movement followed, manifested in dramatic increases in the number of unions, union membership, and disputes. But only minor amendments were made to labor laws, which continued to recognize only the apex union body, the Federation of the Korean Trade Unions (FKTU), and to lock out 'third parties' (i.e. union officials not employed by the enterprise concerned) from collective bargaining. Korean labor laws were not amended substantially until March 1997 when most of the restrictions on union activities were lifted as a result of these amendments. Measures were also taken to increase labor market flexibility. One of the most controversial aspects of this law was the introduction of collective redundancy dismissals. Under the new law, beginning in March 1999, companies would be able to dismiss workers if there were urgent economic reasons. However, the crisis overtook all these proposed changes. With the IMF's offer of a financing package in 1998 to help alleviate the impacts of the crisis, the Korean govemment and the IMF agreed that further steps should be taken to improve labor market flexibility. Recognizing that this issue could not be taken solely by the government, on January 15, 1998 President- elect Daejoong Kim established a tripartite committee for labor relations. This included representatives from unions, employers, the ruling political parties, Cabinet members, as well as public organizations. From the start, the committee faced severe opposition from labor over the issue of redundancies. Nonetheless, in February 1998 the committee agreed to allow dismissals in cases where they were absolutely unavoidable or as a consequence of mergers and acquisitions.' In addition, under the new agreement, dismissals for collective redundancy were allowed earlier (i.e. starting from April 1998) than had been originally scheduled (March 1999). Additional agreements reached by the tripartite committee were as follows: 0 The government should establish a five trillion won unemployment fund. 0 Public servants should be allowed to form a labor consultation body in 1999. Teachers should be allowed to forin a labor union in July 1999. O Labor unions should be allowed to engage in political activities as of the first half of 1998. 0 A dispatched worker system should be a~lowed.~ I Protests in the 1980s led by students and union activ~stsagainst the military government came to a head in June 1987 when the presidential candidate of the ruling party presaged political liberalization. See Park and Leggen for details (1998). In the earlicr 1997 amendment to the labor law, mergers and acquisitions were not considered a justifiable reason for collective dismissals. This system refers to hiring arrangements whereby an employer employs workers through temporary work agencies instead of hiring them directly. Labor Market Refornzs in Korea: Policy Optionsfor- the Future After this landmark agreement was reached, some members of the more radical Korean Federation of Trade Unions (KCTU) voted against the deal despite their representatives' endorsement of the package. The dissenters claimcd the agreement did not fully reflect the interests of labor. However, they did not go ahead with major strikes. Most people hailed the agreement saying that it would contribute greatly to restoring international confidence in Korea. Following the financial crisis, the Korean economy experienced many difficulties. In real terms, GDP decreased by 5.8% in 1998 (Table 1 .I). Onc of the consequences of the depressed economy was a rapid increase in unemployme~t.Job losses accelerated through 1998 and the unernploy~nent rate reached 8.6 % in February 1999, the highest it has been since official statistics have been published. This figure represented 1.8 million unemployed Korean workers. Before the financial crisis, less than 500,000 were unemployed. Job losses were concentrated in certain sectors of the economy. Manufacturing experienced the largest losses, with employment in that sector decreasing by 473,000 between the last quarter of 1997 and the first quarter of 1998. Job losses were also large in the construction sector. On the other hand, employinent in the agriculture1 forestrylfishery sectors increased during the financial crisis as unemployed workers went back to the rural areas to seek a livelihood. With mounting job losses, workers and their unions called for guarantees of job security, suspension of unilateral moves by employers to lay off workers in the name of structural refonn, and a stronger safety net. Workers felt that government measures did not guarantee a miniinum livelihood for the unemployed. As of August 1998, only regular employees working for establishments with five or inore employee were entitled to unemployment allowances. Moreover, according to a survey by the Korea Labor Institute (March 1998), two out of five jobless persons had not received severance pay, which is the alternative means of social assistance for the unemployed. At the workplace level, management was having a hard timeintroducing labor market flexibility measures. This was especially true in large firms with strong unions. Opposition from workers hardened as there were few re- employment prospects for those who were dismissed. In non-unionized sectors or in smaller firms, on the other hand, downsizing measures were introduced more easily. Thus, a short-tenn outcome of these developments has been a further polarization of working conditions between large and small To some extent, hnwever, institutionalizing a system of collective dislnissals for managerial reasons was much easier in the context of the IMF package. Unionr as well as workers have liad little alternative but to accept the principle of employment adjustment which is required by the IMF and internationalinvestors. Table 1.1: Selected labor indicators in Korea, 1995-2000 Q2 Q2 GDP change rate 8.9 5.0 -5.8 10.7) 10.8 9.6 (%) Economically 62 62.2 60.7 61.5 61.0 61.3 active population rate (%) Employed(000s) 20,377 21,048 19.926 21,634 20.362 21,268 Unemployed 419 566 1,463 1,353 1,435 840 (000s) Unemployment 2.0 2.6 6.8 6.3 6.6 3.81 rate (%) Nominal wage 11.2 7.0 2.4 12.1 10.6 8.7 change rate (YO) Real wage 6.4 -2.5 -9.3 11.1 10.0 7.2 change rate (%) Workinghoursper 207.0 203.0 199.2 208.1 month Union membership 1,615 1,484 1,402 (000s) " Number of industrial 88 78 129 198 disputes2) Number of foreign 128,906 267,546 157,689 227,384 workersin Korea3' Notes: 1) Union membershippeaked to 1,932,000in 1989. 2)7,238of strike took place from 1987-1989. 3) As of December 31 of each year. 4) From January to September1999. Source: KoreaLaborInstitute (VariousIssues),QuarterfyLaborReview. Ofticeof Statistcs.Korea (VariousIssues),SurveyReporton EconcrnicallyActivePopulation. Industrial disputes increased sharply as job losses mounted. In 1998, 129 strikes were recorded, a 65% increase over 78 recorded in the previous year. Workers who took part in the strikes numbered 146,000, three times as many as the 44,000 who participated a year earlier. The number of lost production days reached 1,452,000 in 1998, compared with 444,000 days a year earlier. Tension between unions and the government concerning economic restructuring peaked in April 1999. The union representing employees of the Seoul Subway System staged a strike in order to protest the scheduled restructuring plan which would inevitably have led to massive dismissals. However, other unions did not join in a general strike called by the KCTU and the dispute ended unsucccssfully for the union. The aftermath of the crisis also affected workers' earning through widespread wage cuts or freezes. In the Labor Ministry's survey of wage 8 Labor .Market Reforms in Korea: Policy Options for the Future deals in June 1998, 87% of the 2,408 firms polled nationwide reported that they had frozen or reduced salaries during the year. After annual wage gains typically greater than 6% until 1996, wage increases were only 0.9% in the fourth quarter of 1997. In 1998, they decreased by 9.3%. Cuts in bonus payments were particularly substantial. There has been some improvement in the situation over the past two years as the economy has started to grow quite rapidly again. Since February 1999, the unemployment rate has been decreasing - by late 1999, the unemploy- ment rate has dropped to below 5% from a high of close to 7% in 1998. The unempjoyment rates of the first and second quarter of 2000 were 5.1% and 3.8%, respectively.5 The improved labor market conditions mirror the marked improvements in Korea's economic conditions since mid-1999. This also confirms that the primary problems concerning unemployment during the financial crisis was deficient aggregate demand (Fields, 1999). Reduced unemployment rates were also partially attributed to income support measures to cushion the shock of layoffs put in place by the Korean government after the financial crisis began. The negative impacts (in terms ofjobs) of the flexibility measures introduced during the crisis were to some extent mitigated by these income support programs. Youth unemployment was a labor market policy concern even before the crisis. In November 1997, the unemployment rate among the 15-24 age cohort was 7.7% (5.7% for the 15-29 age cohort) when the economy-wide unemployment rate was just 2.6%. With the crisis, these youth rates jumped to 19.7% and 14.6% in December 1998 for the 15-24 and 15-29 age cohorts, respectively. While the youth unemployment rate also decreased with the improved economic situation, it is still very high.6 However, the improvements in the overall unemployment picture do not mean that the social protection for workers is no longer a problem. The employment situation of the most marginal unemployed workers who do not qualify for unemployment benefits has not improved. While Korea's Employment Insurance System, which was set up in 1995 to cover establishments with 30 workers or more, and then progressively expanded during the crisis (finally covering all enterprises regardless of size by October 1998) provides assistance through three components - employment subsidies, vocational training, and unernploy~nentbenefits - only a small proportion of workers remain insured and benefits are very modest by international standards. Furthermore, young people tend not to receive unemployment benefits since many of them have just entered the labor force. Or, if they do receive benefits, their eligibility period is shorter. Another problem is that the proportion of long-term unemployed (those unemployed for more than six months) is high. It was about 20 % of total unemployed in 1999. However, it had decreased to 13.6 %inthe first quarterof 2000. "he unemployment rate of the 15 to 19 age cohort was 13.9 % and 10.0 % in 1999 and November 2000, respectively. OVERVIEW 9 Since the crisis, there seems to have been a shift towards temporary employment and away from regular forms of work. For example, the number of daily workers, who are not covered by the unemployment insurance scheme, has increased substantially. The proportion of daily employees with less than a one-month employment contract period reached over 18.0% of total employment in 2000, an increase from 13.8% in 1996. Korean labor market conditions had been very tight up until the financial crisis, with many industries suffering from a shortage of unskilled workers. The rate of unfilled vacancies for unskilled workers was 13.5% in 1996. This situation had drawn foreign workers to Korea in increasing numbers, growing from a few thousand in the early 1980s to almost 300,000 in late 1997 (Table 1.1). While, current immigration laws do not allow unskilled foreign labor to enter Korea, they are, however, allowed to enter as 'trainees'. Nevertheless, a large number of unskilled foreign workers entered Korea for the purpose of employment and undocumented foreign workers comprised 55.3% of the total foreign labor force in Korea in December 1997.' Although the popular view called for the repatriation of foreign workers as the impact of the financial crisis unfolded, many Koreans came to realize that the country still needed unskilled foreign workers. Despite rapidly growing unemployment, not many Korean workers wanted to work in "three-D (difficult, dangerous and demanding) jobs". Small firms in labor- intensive manufacturing sector are still having difficulties in locating Korean nationals who want to work for them. The poor success of government measures to replace foreign labor also confirmed that unemployed Koreans were reluctant to take up 3-D jobs, even under the difficult circum~tances.~ Income Support Programs When the financial crisis began, Korea was not prepared for such high levels of unemployment. An unemployment insurance scheme had been introduced only in 1995 with limited coverage. Only 3% of the total population were protected under social assistance programs. In 1997, the livelihood protection program accounted for the major share of government spending for social safety nets. With mounting unemployment, the Korean government continuously ' After the financial crisis started, the Korean government offered two amnesty programs to undocumented foreign workers in 1997 and 1999. Some 55,000 undocumented foreign workers left the country under these programs. The Korean government offered wage subsidies to establishments that replaced their foreign workers with Korean nationals. However, only some con~paniesapplied for the subsidy. For example, until September 8, 1999, 16.6 %of the targeted replacements were accomplished. As a result, the government has doubled the wage subsidy amounts. 10 Labor Mnrket Reforms in Korea: Policy Options for the Future expanded the coverage of the Employment Insurance System and benefits. The total amount of the unemployment allowances increased from 79 billion won ($64 million) in 1997 to 850 billion won ($708 million) in 1998, then to 1.5 trillion won ($1.3 billion) in 1999 (Table l . ~ ) With employment . ~ prospects improving, overall unemployment payments decreased in 2000. During the first six months of 2000, unemployment allowances of 227 billion won ($189 million) were paid, which represented a decrease of 31.3% as compared to the same period in 1999. The Korean government has also put greater resources into livelihood protection programs. Temporary public assistance programs have been introduced. Around 5 trillion won were (and will be) spent in order to create short-term jobs, mostly through public works projects. Special programs targeting particular groups such as the homeless were launched. In addition, various kinds of training programs were offered for the unemployed. Government expenditures on social assistance have more than doubled since 1997. Dcspite all these efforts, only a limited number of the unemployed has benefited from these programs. Just 20% of the unemployed received some kind of social protection from the government in 1998 and 1999. Clearly, income support for unemployed workers remains a critical policy area and the first section of this volume includes five chapters on this issue. In Chapter two (Expanding the Coverage of Korea's Unemployment Insurance System), Jai-joon Hur (Korean Labor Institute) proposes a measure to expand Korea's unemployment insurance system. In June 1999, only 10% of the unemployed workers enjo ed access to unemployment benefits even after the system had expanded I . Recognizing the incomplete ;Y coverage of Korea's Unemployment In~uranceScheme, Hur proposes that the system be expanded to include daily workers whose einployment contract period is shorter than one month. About 20% of paid employees are daily workers. Another 30% are temporary workers and the distinction between the daily and temporary workers is often blurred. The author proposes a Refined Employment Record-keeping System (RERS). Under the new system, employers will be obliged to report the affiliation of all of their employees to the labor authorities. The introduction of an Employment Record Form (ERF) or Sinart Cards is also suggested in order to support employees in keeping track of their elnployment records. The ERF is a written form of employment record which employers have a responsibility to fill out and give to employees, as well as to submit to the authorities. Although the RERS proposal addresses one obvious concern with the coverage of the existing unemployment system, various issues need to be The exchange rate of Korean wo11with respect to the US dollar has flxctuated between 900 and 2,000in the aftermathof the crisis.This paper uses 1,200 Korean won to one US dollar which is close to the rate after the Korean won had moreor less stabilized. '" This ratio even decreased to 7.9% during the first half of 2000. Table 1.2: Expenditures of social protection for the unemployed in Korea1', 1997-2000 -- ----- - 1997 1998 1999 2000 2, ---- (won in billions) Unemployment allowances 78.7 850 1,522.7 1,013.6 Livelihood protection program 1,121.0 1,634.9 2,314.7 2,313.6 Regularprogram 1,379.1 1,488.4 1,709.0 Temporaryprogram 255.8 826.3 604.6 Supplementary social safety net 0.0 2,262.7 4,132.3 2,553.0 Short-term job creation 1,170.4 2,671.8 1,210 Public loans for the unemployed 772 1,179.2 515.2 Provision of lunches andschool fee 107.4 281.6 82.8 to middle andhigh school students Payments for wages in arrears 190.0 0.0 0.0 Training for the unemployed 754.1 583.2 350.9 Employment stabilization scheme 122.4 483.2 366.3 Job placement services 66.0 103.6 79.6 Total 1,199.7 5,667.2 9,140.0 5,922.0 Notes: 1) The figures of 1998 are from the Prime Mlnister Office and the \Ministryof Labor (1999). The figures of 1999 and 2000 are from \Ministry of Flnance and Economy and other government ministries(2000). 2) The figures of 2000 are as of January 11.2000. Source. Prime IVinisterOfficeand the Ministry of Labor. Korea (1999). Ministryof Financeand Economy and Other Government Ministries. Korea (2000). resolved before such a program can be implemented. One issue involves the potential financing burden, especially because of the seasonal nature of much daily work. Depending on the numbers of daily workers included in the Unemployment Insurance Scheme, funding requirements might need to be substantial to cover the high levels of unemployment among daily workers during the winter. Moreover, the RERS would likely only be applicable to daily workers who have a sort of guaranteed employment period as found in construction. Hence, many daily workers In manu- facturing and services would still be excluded from unemployment insurance, even with the RERS. Finally, the new system wouId likely encounter enforcement problems. In Chapter three (The Potential Role of an Employment Guarantee Scheme in Korea's Social Safety Net System), Gary Fields (Cornell University) proposes a second pillar of the Employment Insurance System which he labels as the Employment Guarantee Scheme (EGS). The EGS would consist of public works jobs, offered on a daily basis to all who wanted them. The EGS would differ from the existing public works syste~n in several respects. First, it would be open to everyone except fir persons currently receiving U1 benefits. Second, it would be a guarantee scheme 12 Labor Market Reforms in Korea: Policy Optionsf o the Fzrtttre ~ which offers social assistance to those who need help. Third, jobs would be offered for a period as short as a day. Finally, the benefits would be below the minimum wage level. However, designing social usefully public works jobs on a large scale raises several questions. Particularly, given the EGS proposal that partici- pants receive wages below the minimum wage level, and duration of partici- pation can be as little as a day, is it reasonable to expect that meaningful tasks will be performed? Furthermore, financing the system might be a problem particularly if the country runs a deficit in national financing. In Chapter four (The Feasibility of Introducing Noncontributoy Cash Benefit Systzms for the Unemployed in Korea), Young-bum Park (Hansung University) considers the feasibility of introducing a non-contributory cash benefit system for the unemployed with special reference to the one put forward by the government under the National Basic Livelihood Guarantee Act. Park recognizes the need to enhance social assistance for the unemployed. However, he argues that efficiently and fairly implementing a non-contributory cash benefit system, such as the one proposed by the government, would be difficult until conditions for means and work tests for the unemployed reach a reasonable level. Therefore, Park proposes that a public works scheme should be maintained as a short-term social safety net even after the financial crisis is over. Park also argues against the extension of the UI system to small employers unless an appropriate system to solve enforcement problems is developed. This view is quite different from the one proposed in chapter two and, of course, runs counter to what the government has done. The author does suggest, however, that the coverage of UI should be expanded by providing benefits to voluntary job quitters after a certain period. The chapter argues that the complete disqualification of voluntary job quitters is relatively harsh compared to international standards. While some countries do disqualify voluntary quitters, others do not. In Chapter five (Unemployment Assistance: Policy Implications~omthe Crisis of 1997), Alan Abrahart examines Korea's experience during the financial crisis in operating the UI system and compares it with Australia which offers only non-contributory unemployment assistance. The author draws a number of observations about Korea's UI system during the crisis, concluding that it offered little protection to those workers who were most vulnerable. First, its short-term viability was ensured through generating contributions from employees who were unlikely to make significant claims for unemployment compensation. The people who actually contributed to unemployment insurance were less than 25% of the labor force in 1999, and were those who were least likely to be unemployed. This implies that a relatively s~nallproportion of the labor force is allowed to get access to the system compared to other countries. Second, daily workers were excluded from the system, limiting its scope. Third, the risk of unemployment continues to be highest among those who are uninsured. Finally, and OVERVIEW 13 relatedly, the greatest problem during the financial crisis turned out to be providing social protection to those who were not excluded from the unemployment insurance scheme. The Livelihood Protection Program (LPP) and public works, the two significant instruments for assisting these workers, fell far short of the potential level of demand, leaving Inany people to rely on family and community support while unemployed. Recognizing that the government has chosen to supplement the UI system with a non-contributory scheme of assistance, Abrahart argues that Korea should decide the extent to which the UI system and the social assistance scheme are to be considered in one administrative framework. Australian experience suggests that welfare dependency can emerge among people who would not have UI entitlements in a typical contributory system. These are precisely the sort of marginal workers, often working in the informal sector, who suffered most during Korea's financ~alcrisis. The author claims that a crucial issue to consider is the level of assistance to be provided under the two schemes. In Chapter six (Assessing .the Coverage and EfJiciency of the Wage Guarantee System in Korea), Taigi Kim (Dankook University), Kang-shik Choi (Myungji University), and Joyup Ahn (Korean Labor Institute) describe the Wage Claim Guarantee System which was introduced after the crisis for insuring payment of severance allowance and wage arrears. Based on data collected from the beneficiaries of the system, their study shows that 90% of the surveyed employees found the system helpful in maintaining their livelihood. The authors suggest that the benefits and contributions should be differentiated depending on company size as well as type of industry. They also argue that a standardized manual for judging company insolvency should be developed and that there should be efforts to publicize the Wage Claim Guarantee System. Some reservations have been expressed in the study's conclusion that that the Wage Claim Guarantee System has been implemented successfully and effectively. These are mainly based on the fact that the survey only covered people who had received benefits. More research needs to be done to include people who qualified for the benefits, but did not receive them (presumably because they were not aware of the system). The effectiveness of the system in terms of costs and benefits should be also reviewed. Labor Law and Labor Market Regulations As we have described, Korea's labor laws have undergone revisions to allow for more flexible layoff provisions and for temporary worker agencies (dispatch workers). While it is too early to judge the impact of these regulations, there has been considerable discussion a-bout their impact on labor market flexibility and how these regulatory changes should be linked 14 I.nbor Market Refornzs in Korea: Policy Optionsfor.the Future to the safety net system. Section 11 of this volume includcs two chapters addressing these questions. In Chapter seven (Assessment of Labor Market Responses to the Labor Law Changes Introduced it? 1998), Duck Jay Park (Korea National Open University), Jong Hee Park (Korea University), and Gyu-Chang Yu (Sookmyung Women's University) analyze the effects of the legislation legalizing collective dismissals for managerial reasons. Their study, based on an enterprise survey, concludes that even after the Labor Standards Act was amended, the major mode of employment adjustment was not dismissal. Most separations were 'honorable retirements', spin-offs, early retirement, or attrition. This has led employers to claim that greater reliance on these forms of separation, particularly honorary retirement which is a costly mode of employ~nentadjustment, proved that Korea's labor market still lacks flexibility despite the concerned labor law amendment. However, it cannot be denied that the labor law change helped troubled companies reduce their workforces. It has contributed to workers' become more accepting of employment adjustment in such companies. However, the authors expect that collective dismissals for managerial reasons are not going to be the dominant mode of employment adjustment in the near future considering that a sentiment of strong egalitarianism exists in the Korean society. The authors argue that Korea's social safety net needs to be expanded to enhance labor market flexibility and they feel that this will be necessaly for workers to accept dismissal more easily. They also suggest that, in addition, policies to upgrade workforce skills need to be developed and government- private sector training partnerships should be established. In Chapter eight (Labor Market Flexibility and Emnployment Agencies in the UK, Japut7 and Korea), Barry Wilkinson (University of Bath) addresses key policy issues relating to the new worker dispatching law. His analysis considers evidence from the U.K. which has longer and more extensive experience with employment agencies. Wilkinson argues that Korea needs to further study the potential costs and benefits of the new worker dispatching law and makes the point that the implementation of the new law ,wasaccelerated under crisis conditions. Based on evidence from the U.K., the chapter raises some issues about the use of temporary worker agencies. First, the contribution of labor market flexibility, including temporary work agencies, to long term employ~nent creation remains unknown. Second, there is a negative relationship between employment agency activity and skills formation except for a small minority of highly qualified agency labor. Third, employment status is highly ambiguous and many temporary agency workers are unclear even as to who their employer is. The author stresses that policy makers have to find a way to achieve a balance between employer and employee needs. He states that this is urgent in Korea since many employment agency activities are illegal despite the recent labor law change and work done by dispatched workers is much wider than the allowed scope of temporary work agency activities. OVERVIEW Active Labor Market Programs (ALMPs) Unlike most OECD countries, Korea does not have a strong tradition of Active Labor Market Programs (ALMPs). One of the main reasons has been the low rates of unemployment prior to the crisis. However, the crisis has led to a significant expansion of these programs as the increases in unemploy- ment, the difficulties facing many laid-off workers in reentering the ranks of the employed, and the obstacles facing young people trying to enter the labor force have all underscored the need for policy-makers to consider new options. Section I11 of this volume includes five chapters on ALMPs. In Chapter nine (Income Assistance and Employment Creation through Public Works in Korea), Johee Lee (Korean Labor Institute) evaluates the impact of the public works programs during the crisis using the results of surveys of program participants and administrators." The author argues that, although these programs were officially classified as an active labor market measure by the government, the major role of the public works scheme was as a source of complementary income. The program especially benefited women, older, and less skilled workers who were mostly excluded from other income support measures. According to the survey results, participants were mostly satisfied with the scheme and wanted it to continue. Lee also claims that deadweight losses were relatively small as a large majority of the participants were atypical workers who had difficulty in finding jobs during the financial crisis. This claim is arguable. Even during the financial crisis, many small businesses experienced labor shortages. It seems likely that many indivi- duals who would otherwise have looked for a job participated in the public works especially given that wage rates were much higher than the minimum wage. Hence, critics contend that such schemes deepened the welfare depen- dency of the participants. On the other hand, supporters of these schemes, including the author, argue that these schemes were critical during the crisis as they assisted the poor to maintain incomes during a difficult period. Lee makes several recommendations regarding the future direction of Korea's public works programs. First, rather than drastically reducing the projects, he suggests that they should be retooled to cater to the demands of the disadvantaged groups in the private sector which may not be able to benefit from social assistance but can avail of public works. Second, close interactions between public works projects and other unemployment measures including social welfare programs should be established. In particular, with the development of the new social assistance scheme under " Conclusions drawn IYoiti this study are tentative since the study does not include a comparison group of individuals (who wzre unemployed but did not participate in public works project). 16 I.ahor Market Reforms in Korea: Policy Optionsfor the Future the Basic Livelihood Guarantee System, the role of public works should be redefined. Lee suggests that public works projects could be used as a work test, which would allow many of the unemployed poor to remain attached to the labor force. Third, Lee argues that the wage rates of public works participants should not be lowered considering that they are usually the major breadwinners of their families. In Chapter ten (Public Works Program in Korea: A Comparison to Active Lubor Market Policies and Worwre in Europe and the US.), Ivar Lodemel and Espen Dahl compare Korea's public works programs with workfare and other active labor market programs in Western countries. They note that the scale and the nature of the problems experienced in Korea during the financial crisis are more similar to the experiences in the West in the 1930s rather than under present circumstances. They point out that while the Korean scheme offered an effective buffer against the immediate effects of large scale unemployment, it fell short in developing the skills of the work force needed for a changing economy. According to them, Korea's challenge is to transform public works into a modern active labor market program and to reform the livelihood protection program into a modern social assistance program. The enactment of the Basic Livelihood Guarantee Law to replace the livelihood protection program suggests that Korea may be now ready to move in this direction. The authors observes that culturally Korea appears to have more in common with Western European countries rather than the U.S., even though there is a dominance of U.S. style policies which are based on a individualized view of people, their behavior and their aspirations. They caution that when options for reform are considered, careful consideration should be given to the fact that Korea is a cohesive and close knit society, rather than an individualistic one. In Chapter eleven (Review of the Effectiveness of Korean Public Employment Services), Jaeho Keum (Korean Labor Institute) looks at the Korean public employment service (PES) which expanded substantially during the financial crisis. The share of the unemployed who registered with public employment offices increased during the crisis partly because they were able to claim their unen~ploymentbenefits as well as receive job search assistance at the PES offices. The author argues that there still remains much room for the expansion of the PES in Korea as the number of public employment agencies, counselors, and beneficiaries assisted by each agency or counselor fall far short of the norms established in OECD countries. However, Korea has yet to reach a consensus about the extent to which the PES should be expanded and what the role of private employment agencies should be. Some argue that the role of the PES should be minimal - such as providing services to the most vulnerable - and that private employment agencies should play a greater role than at present. However, this does not seem to reflect the current view of the Korean Labor Ministry. Some also observe that the Korean PES enjoys monopoly power in the einployment services market which is not the case any longer in most OECD countries. Based on his survey of PES beneficiaries as well as service providers, the author argues that the Korean public e~nploymentservice still needs to improve the quality of its service and suggests some possible means to do so. First. the PES should be restructured to strengthen areas such as in-depth counseling, career guidance and in-depth occupational information as its job matching services will be increasingly carried out via Work-Net on the internet. Second, the PES should strengthen services to vulnerable groups such as elderly or female household heads, and low-skilled workers. Currently, PES customers mostly comprise of young or highly educated workers. Third, there should be measures to increase the competence and morale of counselors. The survey also showed that counselors had complaints about their status as contract employees, about their pay and promotion prospects, and about working conditions. The next two chapters look at evidence from evaluations of training programs. In Chapter twelve (Evaluating the Trainingfor the Unemployed in Korea), Soon-Hie Kang and Byung Hee Lee (Korean Labor Institute) present results from an evaluation of Korea's training programs for the unemployed. Training has played an important role in Korea's active labor market policy during the crisis, with participation expanding eight-fold between 1997 and 1998. However, the evaluation shows that re-employment rates after participation in training have been no higher than employment rates for the non-participating unemployed. This has been widely interpreted in Korea as proof of the failure of the training programs. The authors' study is Korea's first scientific evaluation of training programs. It compares employment-related outcomes nine to fifteen months afier training among 1,000 unemployed workers who had participated in a training course targeted at the insured unemployed workers in 1998. The comparison group is those individuals who had separated from a job during the same period but who did not undertake training during 1998. The analysis show that in terms of gross outcomes, the Korean training programs have not helped their participants: about 40% of both the trainees and the control group were employed at the end of the observation period. However, when the different observed attributes of the two groups are controlled for, differences emerge. The probability of moving into employment after training was significantly higher for female trainees than for non- participants (though the impact for males reinains insignificant). Although some methodological issues remain to be addressed, this study underlines the importance of conducting scientific evaluations when evaluating active labor programs. Chapter thirteen (Publicly Funded Trainzng for Unemployed Adults: Germany, the UK and Korea) by Paul Ryan (Cambridge University) surveys the evaluation evidence available on public training programs for (East) Germany and the U.I<.. The results are mixed. The Eastern German training programs do not appear to have generated significant Izbor market benefits for their participants. The British programs have, in contrast, helped 18 Labor Market Reforms in Korea: Policy Options for the Future participants find work and enhance their earnings. The author observes that Korean training programs during the financial crisis resembled their East German counterparts of the early 1990s in terms of the rapidity of expansion, their initial emphasis on breadth of coverage at the expense of depth of learning, their overwhelming orientation toward classroom training, and the wide range of participants and training provided. However, the implications of the U.K. and German experiences for Korean policy are limited as the impacts vary significantly based on economic conditions and type of training. Labor Market and Gender Gender discrimination was a significant issue in Korea even prior to the crisis, and this issue was further highlighted during the crisis period. In Chapter fourteen (Were Korean Women Idore Afiected by the Asian Crisis), Kye Woo Lee and Kisuk Cho (Ewha Woman's University) analyze the effect of the financial crisis on women. The authors find it difficult to draw an unequivocal conclusion on whether female workers suffered more during the financial crisis since both sexes were adversely affected. However, they conclude that male and female workers were treated differently in the labor market during the crisis, especially in terms of earnings, involuntary disengagement, and rehiring status. The average earnings of female workers were lower than those of male workers. The incidence of involuntary unemployment was higher among female than male workers. When female unemployed workers were re-employed, they tended to get non-regular or daily jobs more often than male workers. Ilowever, their data do not allow them to analyze whether labor market discrimination against females increased during the crisis. The authors try to identify which gender discrimination theory best explains these differences in the effects of the financial crisis. While there are some methodological problems with their data and test approaches, their survey data do not conclusively support any of the discrimination theories. Nonetheless, they argue that the data render more support for the cultural gender-type theory than the statistical discrimination or labor market risk theory. The Korean Experience in an International Context As the government instituted and expanded a range of labor market interventions to alleviate the impact of the crisis, this provided an ideal opportunity for Korean policymakers to' learn and benefit from the experiences of more advanced OECD countries. The final three chapters of the volume look at the Korean experience within the international context. Chapter fifteen (Labor Adjustment, Non-Standard Work and Employment Program: Korea in an OECD Context) by Gordon Betcherman, Amit Dar, Amy Luinstra, and Makoto Ogawa (World Bank) focuses on Korea's adjustment to the crisis and two labor market issues of growing importance for the country - ALMPs and non-standard employment. They show that the structure of the Korean workforce is rapidly becoming more similar to other OECD countries, with low and shrinking agricultural employment and growing employment in the service sector. Further, the educational attainment of the labor force now compares with that of high-income OECD countries. Their review highlights the flexibility of Korean labor market - both in terms of prices and quantities - during the crisis. This raises questions for social policy and labor market segmentation. Their analysis reveals that Korea has an underlying propensity toward non-standard employment. From an international perspective, Korea has a high share of non-standard employees. While this has been key to the country's labor market flexibility, they view this as an issue that rightly concerns policy makers. There is a clear segmentation between relatively well paid regular workers with access to many kinds of social protection and non-standard workers who are in an insecure position. This is at least partly due to labor policies and practices which create incentives for non-standard jobs. Whether recent reform to extend the coverage of labor standards and unemployment insurance to individuals in these forms of employment will alter these incentives or not will have to been seen. The crisis has led to a significant expansion of ALMPs. The authors do not draw a definitive conclusion as to whether such an expansion has been justified on economic or social grounds although they view the limited evidence on the effectiveness of these programs in Korea as discouraging. However, they strongly suggest that policy-makers in Korea should not avoid developing ALMPs as the country emerges from the financial crisis. But this should be done cautiously and large scale application of these programs should be avoided without piloting and carefully evaluating them. Furthermore, they argue that Korea's transition toward a knowledge-based economy will have implications for active labor programs in terms of institutional capacity building, delivery techniques, increasing reliance on technology, innovations in financing, and monitoring and evaluation of these programs. Korea can benefit from the experiences of some OECD countries where these transformations are already evident at the cutting edge of active labor market programs. In Chapter sixteen (Korean Labor Market and Social Safety-Net Reforms: Challenges and Policy Requirements), John Martin and Raymond Torres (OECD) argue that Korea is still facing key challenges with respect to industrial relations, labor markets and social protection. They set out reform 20 Labor Market Reforms in Korea: Policy Options for the Future proposals to tackle these challenges. The authors state that, although the recovery from the financial crisis is well underway, Korea should not reduce the momentum of reforms since further progress may be more difficult than is often thought. Indeed, future improvements in living standards may have to rest on using resources more efficiently, rather than more extensively as was the case before the 1997 crisis. They suggest a more qualitative approach to the development strategy encompassing an increased emphasis on enterprise training, reducing labor-market duality, and the establishment of a well-functioning social safety-net that covers all individuals. Martin and Torres also argue that these measures need to be adopted in a context of Pi111 respect for basic workers' rights and, in this sense, resolving remaining industrial relations' issues should rank high on the policy agenda. They argue that in a modem and productive economy, it is important that all social actors be involved in the reform process in a climate of mutual respect and trust. In Chapter seventeen (The Challenge of Labor Reform in Korea: A Review of Contrasting Approach to Market Enhancement and Experience from Chile and Denmark), Louise Haagh (Oxford University) points to the importance of an individual-centered approach to labor flexibility issues. Haagh argues that issues of labor market flexibility and social sector reform should not be regarded as separate concerns in an environment where human capital has become central to economic growth. She emphasizes an approach focused on individual welfare to labor market flexibility issues and a strong link between labor institutions and welfare reform. According to the Haagh, Danish labor reforms in the past decade have succeeded because they have followed such an approach. Reforms have been geared toward supporting the mobility of the individual workers and of the development of human capital. According to Fiaagh, the Danish experience shows that it is possible to enhance the labor market in the sense of making it easier for individual workers to move in labor markets without necessarily reducing the flexibility of firms. Haagh argues that this can be seen in the high levels of intra and inter firm labor mobility that characterized the labor market throughout the 1990s. Emerging Policy Concerns and Areas for Further Research During the financial crisis, the Korean government implemented many temporary measures in order to provide income support to displaced workers and their families, in addition to expanding the coverage of the Employment Insurance Scheme. However, only a small portion of the Korean labor force actually benefited from these programs. Many Koreans in hardship had to fall back on informal safety nets especially family support. Although these traditional forms of support should not be discounted, Korea's public safety net needs to be further developed in order to assist the most disadvantaged. One of the fundamental problems for the expansion of Korea's social safety net is the dualistic nature of labor market. Korea has one of the most flexible labor markets in the OECD and this flexibility has been demonstrated throughout the crisis. However, this flexibility is derived at least in part by the large share of non-standard (including temporary) employees who are entitled to very limited unemployment insurance or social protection benefits. The Korean labor market seems to have a propensity toward the creation of these forms of employment that can not be fully explained by the crisis nor by the country's stage of development. Recent government policy measures to extend social insurance and labor regulations to all firms and more categories of workers might decrease the incentives toward non-standard employment. However, the government's capacity to enforce these measures effectively is open to question. On the other hand, the loosening of restrictions on temporary worker agencies is likely to lead to increases in non-standard jobs. In this respect, there is a view that strengthening labor and social protection for workers in non-standard jobs will decrease labor market flexibility. Whether it will or not deserves further research. What is clear, however, is that a challenge emerging for Korea is to improve social protection for all workers, to create incentives for access to secure work, and to support workers in transition without constraining job creation. The impact of legislative changes concerning dismissals for economic reasons during the crisis, although not insignificant, was not as significant as expected. There is a school of thought that argues that this proves the lack of flexibility in the labor market. This may be true for some part of the Korean labor market - especially for workers in large sectors protected by strong unions - but it is most likely not true of workers in small enterprises. Hence, a short-term outcome of the crisis and the implementation of the reform package have been the further polarization of working conditions between large and small firms. On the other hand, it is possible that enlarging the social safety net could promote labor market flexibility as workers might be expected to accept dismissals more easily. This is clearly an area that will have to be studied in greater detail in the future. Labor force participation rates among women are low by OECD standards. Further, the incidence of non-standard work is high among females. During the financial crisis, women were more often subject to dismissals than were men. All these facts imply that policy initiatives should be developed to boost women's participation and upgrade their status in the Korean labor market. This is also an area that needs to receive further attention. Many of the unemployed was not insured by the unemployment insurance scheme at the beginning of the crisis. While the scheme was gradually extended to smaller private sector companies over the crisis period, less than 20% of the unemployed have benefited from the scheme. This is mainly due 22 Labor Marker Reforms in Korea: Policy Optionsfor the Fzlt~lre to the fact that those who contributed to the scheme were the least llkely to become unemployed and that temporary workers were excluded from participating in the scheme. Policymakers need to examine in greater detail how the coverage of the system can be extended while maintaining its financial sustainability. While the public works scheme contributed to reducing the hardship of the unemployed, especially those who were not entitled to unemployment benefits, critics argue that this scheme was poorly designed as the wages offered were significantly higher than the minimum wage which may have hindered employment generation in the private sector. A challenge for policymakers in Korea will be to design public works programs that are targeted to the most disadvantaged. The government has recently introduced a more extensive livelihood scheme under the National Basic Livelihood Guarantee Act. However, this new livelihood program is likely to face significant enforcement problems, especially given the dualistic nature of the labor market. Ensuring that the livelihood program is effective in meeting the needs of the target groups is an area that needs to be explored in greater detail. The relative scarcity of rigorous evidence on the effectiveness of ALMPs in Korea does not allow a definitive conclusion as to whether such interventions have been justified on economic or social grounds. However, it ISlikely that the importance of ALMPs will increase in the future as Korea's labor market structure more and Inore resembles that of other OECD countries. It is important that further scientific evaluations be conducted in order to identify the impacts of the active labor market programs that were implemented during the financial crisis. As Korea continues the transition towards a knowledge-based economy, it can learn from other OECD countries, especially in relation to the design of ALMPs and their implementation. Korea's economic progress over the past three decades and its apparent rapid resurgence from the economic crisis have been quite remarkable. This pace of development and change has been accompanied by pressures in the labor market - e.g. high levels of non-standard employment, inadequate levels of social protection, gender discrimination - creating formidable challenges for policy makers. IIowever, Korea's responsiveness to problems over the last few decades gives us confidence about its future prospects as it begins to tackle these labor market challenges. ? \ II INCOME SUPPORT PROGRAMS 2Expanding the Coverage of Korea's Unemployment Insurance Jai-Joon Hur* Problems with the Existing Unemployment Insurance System Objectives On July 1 1995, before the financial crisis, Korea had put into place an ambitious employment insurance system. The three components of this system are traditional unemployment insurance, job training, and employ- ment maintenance/promotion subsidies. Other elements of Korea's social safety net system include a livelihood protection program for the poor, a livelihood stabilization fund for the unemployed whose assets fall below a certain level, a public works program for the unemployed and for the spouses of low-wage workers, and many other smaller programs. The Korean Employment Insurance System (EIS) is of a compulsory social insurance type. Thus all the employers and employees in the covered firms should be insured and they are entitled to receive grants or unemployment benefits from the Employment Insurance Fund. As elsewhere, the unemployment insurance (UI) component of Korea's EIS protects against the risk of income loss due to unemployment. The UIS provides people with a temporary income after they have been completely unemployed for two weeks. When the UIS started in 1995, only employees in large firms (hiring more than 30 workers) were entitled to the U1 benefit. After the financial crisis. the coverage was rapidly expanded to all firms regardless of their size. The compliance rate, however, is low and UIS covers only 65% of the employees who are supposed to be covered. The objective of this paper is to assess why the coverage of Korea's UIS is incomplete and how the coverage * Korea Labor Institute. 26 Labor Market Reforms in Korea: Policy Optionsfor the Future can be extended. . Before proceeding, it is important to adopt a guiding framework with which to assess the LTIS and around which to make proposals for change. One possible model is social insurance. Under this model, in so far as possi- ble, every employee in society contributes and all of the insured that become unemployed can collect. Under the social insurance model, the financial resources of the system are funded based on pay-as-you-go principle. A second model is the individual account model. Individual accounts can be publicly-run or publicly-mandated. In either case, the individual builds up the balance in his or her own account while working. In the event of unemployment, the account balance can be drawn upon to provide an income to the unemployed person. Otherwise, the money belongs to the individual, to be used in some other way, e.g., for spending in their old age, home-buying, or medical expenses. Under this model, UI is a type of compulsory savings with self-insurance against unemployment. The third model is social assistance. Here, the nation taxes itself in order to provide benefits for the needy members of society. Under this model, benefits are means-tested - that is, in order to collect UI benefits beneficiaries must prove not just that they are unemployed but also that they are poor. Such means-testing is not necessarily carried out through declarations or administrative verification of low income or assets; self- declaration through self-targeting is also possible, as is the case with the public works program. Korea's system has been set up on the social insurance model. Our recommendation will be set forth in the context that this model is continued. The Labor Market Context and Inconzplete Coverage o f the Existing System Beginning in November 1997, the Republic of Korea underwent a devastating economic crisis. Declining macroeconomic conditions brought about major labor market disruptions in 1998: a quadrupling of unemploy- ment, a fall of 9% in real wages, informalization of the remaining jobs, increased job insecurity, and rising poverty and inequality. Disadvantaged groups suffered a disproportionate impact. The result was not only economic misery but also social pain: increased homelessness, rising crime, heightened school dropouts, an accelerating divorce rate, and an over- whelming sense of social malaise. 1999 marked a major turnaround for Korea. GDP grew by more than 10.796,and real wages increased apace. The unemployment rate is now less than half of its peak level. Women are again returning to the labor force and are finding employment. A sense of guarded optimism is returning to Korean society. Nonetheless, important labor market an&social problerns remain. The unemployment rate is now twice as high as it was before the crisis, EXPANDING THE COVERAGE O F KOREA'S UI 27 and temporary and daily workers are much more vulnerable to unemploy- ment than before. They account for 82.1% of unemployed wage workers, while only 52.9% of the total number of employees. They experience unemployment recurrently to the extent that 77.5% of those who experience unemployment more than twice during given 18 months are temporary and daily workers (Lee, 2000). In sum, they are principal group of workers suffering unemployment. Even if Korea has expanded coverage of UIS to all firms regardless of their size from Oct. 1998, the UIS fails to cover most of temporary and daily workers who are supposed to be covered. Therefore, one major issue for the UIS in Korea is that as coverage is iricomplete what measures can be taken to expand coverage so that more wage workers are insured against income losses due to unemployment. Before proceeding, we should point out that financing is apparently not an issue. The UIS is running a large surplus - that is, it has a sound financial footing even after the Fina:icial Crisis (Hur, 2000). This allows, without any further increase of contribution rate, room for extending the current UI structure by adding temporary and daily workers who frequently experience unemployment. Expanding the Coverageofthe UnemploymentInsuranceSystem Coverage of UTIS In 1999, there were 12.5 million wage workers in the Korean labor market. Of these, only 6.1 million employees were working in insured employment - that is, in jobs that would qualify them for U1 benefits in the event of unemployment. The gap in coverage comes from two sources: employees who are supposed to be covered but are not, and employees who are not now meant to be covered. As of Dec. 1999, the estimated number of employees to be covered amounts to 9.2 million, but only 6.1 million are registered at public employment office. These and earlier figures are presented in Table 2.1. About 35 % of the employees that should be covered by the UIS are not actually registered on the EI database. Even if all the employees, except in some designated cases', are now officially entitled to UIS coverage by law, many employees are still not protected by UIS. Table 2.2 contrasts the number of regular and temporary employees, which can be regarded as a proxy for the number of employees who are supposed to be covered with actual number of insured employees by firm size. The uncovered employees are concentrated among small firms ' Among them are employees working less than a month in a firm,government employees and private school teachen, etc. See Yoo (1999)for more details. 28 Labor Market Refornzs in Korea: Policy Optionsfol. the Future Table 2.1: Number of Insuredemployees and estimatednumber of employees to be covered by the Korean UIS,December 1995 to December 1999 Dec. 31. 1995 Jun. 30.1998 Dec. 31. 1999 Scope of coverage 30 employees or 5 employees or 1 employee or (firm size) more more more3) Estimated number of 4,280 5,710 9,243 employees to be coveredl'(A) Actual number of insured 4,204 4,797 6,054 employees2' (6) Coverage ratio (=EVA)(%) 98.2 84.0 65.5 Notes: 1) Estimated numbersof employeesto be coveredas of Dec.1995 and Jun. 1998 are cited from Yoo (1999) and that of Dec. 1999 was estimated based on National Statistics Office's Economically Active Population Survey database, Government Employees Pension data and Private School Teachers'Pension data. 2) Actual number of registered employees on the Central Employmentlnformation Office's El database. 3) Including part-timeemployees working 80 hours or more in a month. Table 2.2: The number of regular and temporaryemployees and actual number of insured employeesof the Korean UIS by firm size, December 1999 1-9 10-49 50-299 300- Total Regular and temporary 3,442 3,385 2,249 1,523 10,598 employees (A) Actual number of insured 1,329 1,356 1,498 1,871 6,054 employees (B) B/A(%) 38.6 40.1 66.6 122.8 57.1 Source: NationalStatistical Office,Korea,EconomicallyActivePopulation Survey database. CentralEmploymentlnformation Center, Korea, Employment Insurancedatabase. with less than 50 employees.' Although it is the employers' duty to declare their employees to the public employment office and pay U1 premiums, many small business employers neglect to do this. Informal characteristics in the employment relationship, the loose attitude of economic agents vis-a-vis legal regulations, together with the insufficient capacity of the EI administration underlie the For employees belonging to firms with equal to or more than 300, the actual number of insured employees exceeds the number of regular and temporary employees. The discrepancy seems to result from the inherent limit of employment statistics by firm size in the EAPS which is a householdsurvey.However, we continue the comparisonas thcrc is no better alternativeand his does not seriouslyharm our discussion incomplete coverage of the UIS in small business firms. Therefore, how the labor administration can effectively identify small business firms and their employees has been an important issue in implementing the UIS since coverage was expanded to smaller firms. The coverage ratio by industry is given in Table 2.3. Coverage rates are very low in the primary sectors, the wholesale and retail trade sector, the hotel and restaurant sector, and community, social and personal services sector.' These are industries with large proportions of temporary and daily employees. Table 2.3: Number of regular and temporary employees and actual number of insured employees of the Korean UIS by industry, December 1999 Regular and temporay A c t u ~of~ ~BIA ~ r ~ e m ~ l O ~ e(A)s e employees (B) (%) Agriculture, hunting and forestry 24,706 10,689 43.3 Fishing 24,869 7,314 29.4 Mining 16,729 17,856 106.7 Manufacturing 2,998,105 2,470,361 82.4 Electricity,gas and water supply 58,620 55,170 94.1 Construction 539,388 467,226 86.6 Wholesale & retailtrade, repairing 1,559,951 637,669 40.9 consumer goods Hotel and restaurant 639,730 150,405 23.5 Transport,storage and 814,038 552,243 67.8 communications Bank, insuranceand other financial 660,306 385,299 58.4 intermediation Real estate, rentingand corporate 910,868 751,283 82.5 services Public administration,defense and 632,839 0 n.d. compulsorysocial security Education 847,896 96,408 11.4 Health and social works 314,610 247,929 78.8 Other community, socialand personal 436,493 182.980 41.9 services Private households with employed 104.178 0 n.d. persons Extra-territorialorganization and 15,097 0 n.d. bodies Others in El database n.d. 21,647 n.d. Whole industry 10,598,423 6,054.479 57.1 Note : n.d.means not defined. Sources: NationalStatisticalOffice, Korea, EconomicallyActivePopulationSurveydatabase. CentralEmploymentInformationCenter, Korea, Em~loymentInsurance database. The coverage ratio in Education is low as privale school teachers are not entitled to UIS coverage. 30 Labor Market Reforms in Korea: Policy Optionsfor the Futu1.e As of Dec 1999, the 6.1 million registered employees account for 46.5% of total employees. This means that the UIS coverage is too low to be considered as a primary safety net for employees. Also, this is one of the reasons why only 10.7% of the unemployed, on average, received UB in 1999. In Dec. 1999, UB recipients accounted for 9.8% ofthe total number of unemployed and 15.5% of unemployed wage workers (Table 2.4). Table 2.4: Ratio of UB recipients with respect to each category of the unemployed, December 1999 Proportion of UB recipients with Person respect to each category of the ('000s) unemployed (%) UB recipients 102 Unemployed 1,040 9.8 Unemployed who had work experience 947 10.8 Unemployed who lost their job within 770 13.2 a year Unemployed employees who lost their 660 15.5 job within a year Sources : National Statistical Office. Korea, Econom~callyActive Population Survey database. Central Employment Information Center, Korea, Employment Insurance database. When we look at the composition of employees and that of unemployed employees, temporary and daily employees turn out to be principal group experiencing unemployment (Table 2.5). Daily employees account for 18.6% of total number of employees while daily employees account for 46.4% of the unemployed. Table 2.6 shows that unemployment span of daily employees amounts, on average, to 1.2 month a year, suggesting that they are more in need of a safety net even though the EIS fails to cover them now. Table 2.5: Employees and unemployed employees by employment status, December 1999 - -- - Employees Unemployed employees ('000s) ('000s) Total 13,024 (100.0%) 660 (100.0%) Regular 6,131 (47.1%) 118 (17.9%) Temporary 4,468 (34.3%) 236 (35.8%) Daily 2,426 (18.6%) 306 (46.4%) Note : The unemployed employees are those who lost their job within a year. Source : National Statistical Office, Korea, Economically Active Population Survey database. EXPANDING THE COVERAGE O F KOREA'S UI 31 Table 2.6: Tenure period, employed period, employment period and unemployment period for workers by employment status, June 1999 Non- Tenure Employed Employment Unemployment employment 3 period" periodz) period ' period4) (months) (months) (months) (months) period5' (months) Regular 10.5 11.6 11.7 0.1 0.3 Temporary 7.5 9.7 10.1 0.6 1.9 Daily 4.3 7.4 8.3 1.2 3.7 Notes : 1) Tenureperiod= Employed period as wage and salary workers in a firm 2) Employedperiod = Employed period as wage and salary workers 3) Employment period = Employed period +Employed period as self-employed or non- paid family workers 4) Unemploymentperiod= Unemployedperiod while searchinga job 5) Nonemploymentperiod= Unemploymentperiod+Period as out-of-labor-force Source : Panel constituted from Economically Acbve Population Survey database for the period of June 1998 to May 1999. Table 2.7 shows that daily employees are intensively distributed in firms with less than 10 employees. Among industrial sectors, manufacturing, construction, wholesale and retail trade, and hotel & restaurant are the four dominant sectors with large numbers of daily employees (Table 2.8). In general, where the proportion of temporary employees is important, so is the proportion of daily employees. The construction sector, however, is an exception. In the construction industry the proportion of daily employees is much more important than that of temporary employees. The distinction between "temporary employee" and "daily employee" in Korea depends on whether employment is scheduled to last more than one Table 2.7: Temporary and daily employees by firm size in Korea, December 1999 All sizes 1-9 10-49 50-299 300- Number of employees ('000s) -- -- - -- Temporary and Daily 6,893 4,154 1,938 603 198 (100.0%) (60.3%) (28.1%) (8.8%) (2.9%) Temporary 4,468 2,565 1,319 451 132 (100.0%) (57.4%) (29.5%) (10.1%) (3.0%) Daily 2:426 1,589 619 152 65 (100.0%) (65.5%) (25.5%) (6.3%) (2.7%) Source : National Statistical Office,Korea, EconomicallyActivePopulation Surveydatabase. 32 Labor Market Reforms in Korea: Policy Options for the Future Table 2.8: Temporary and daily employees by industry in Korea, December 1999 Temporaryand Temporary Daily employees daily employees employees (persons) Wholeindustry 6,893,371 (100.0%) 4,467,646 (100.0%) 2,425,725 (100.0%) Agriculture, huntingand forestry 100,981 (1.5%) 12,285 (0.3%) Fishing 30,076 (0.4%) 17,500 (0.4%) Mining 3,075 (0.0%) 1,550 (0.0%) Manufacturing 1,547,494 (22.4%) 1,045,992 (23.4%) Electricity,gas andwatersupply 17,744 (0.3%) 11,374 (0.3%) Construction 904,686 (13.1%) 226.975 (5.1%) Wholesale& retailtrade, 1,247,805 (18.1%) 988,374 (22.1%) repairingconsumergoods Hotel and restaurant 885,063 (12.8%) 563.456 (12.6%) Transport,storage and 230,848 (3.3%) 176,259 (3.9%) communications Bank,insurance and other 268,460 (3.9%) 244,289 (5.5%) 24,171 (1.O%) financialintermediation Real estate, rentingand 471,640 (6.8%) 390,394 (8.7%) 81,246 (3.3%) corporate services Publicadministration, defense 292,205 (4.2%) 86.016 (1.9%) 206,189 (8.5%) and compulsory social security Education 333,816 (4.8%) 293,843 (6.6%) 39,973 (1.6%) Health and socialworks 92,618 (1.3%) 82,420 (1.8%) 10,198 (0.4%) Other community, social and 302,533 (4.4%) 221,002 (4.9%) 81,531 (3.4%) personalservices Privatehouseholds with 161,689 (2.3%) 103,279 (2.3%) 58,410 (2.4%) employed persons Extra-territorialorganizationand 2,638 (0.0%) 2,638 (0.1%) 0 (0.0%) bodies Source: National Statistical Oftice,Korea. Economically Active Populafion Survey database. month ("temporary") or less than one month ("dail~").~For policy purposes, the distinction between temporary and daily employees is not important, as both are largely unprotected in the labor market. For this reason, it is often more convenient to integrate them into one labor market category.' For moredetails, seeBox 2. 1. j In this paper, we will use the term non-regularemployees in the same meaning as temporary & daily employees. EXPANDING THE COVERAGE O F KOREA'S UI 33 I Box 2. 1: Classification of employees by status It is a common tradition in the analysis of the Korean labor market that employees are classified into three categories: regular employees, temporary employees and daily employees. Statistically, regular employees are defined for operational purposes as employees whose employment contract spans, implicitly or explicitly, more than a year or is without any fixed term. Temporary employees are defined as employees who have an employment contract lasting between a month and a year. Daily employees are those employees whose contract period is less than a month. In reality, however, the usage of those terms has nuances other than just demarcation by contract period. Temporary employees can be fixed term contract workers or project based employees. It is not rare that their employment period exceeds one year or is without any fixed term; as is the case with most of part-time employees and dispatched employees. They are highly substitutable with other workers as theirjobs do not require professional skills. Employers can freely dismiss them without offering any severance pay. The employer, regardless of regulations in law, does not contribute a social insurance premium for them as their wage is considered to be inclusive of severance pay, social insurance premium, etc. In many cases, the employer does not keep their employment records so they cannot prove their employment career. Temporary employees are, in the usual sense, equivalent to atypical or non-regular employees. Daily employees in practical usage can be understood as a subset of temporary employees. They are paid on a daily basis and their employment spans a well-defined project. The employment contract is terminated automatically when the project is completed, even if they remain under the same employer or Shipjntig.They frequently move from one work place to another. Sometimes it is hard to distinguish them from the self-employed. They conclude a contract as an independent contractor as if they are a self-employed when they have opportunity of getting independent jobs, such as repairing or petty construction. Regular employees are those whose employment contract period is not fixed. Their employment is stable. Their wage increases on the basis of seniority and they have a predictable promotion path. They fully enjoy the social insurance advantage as regulated by law. The National Statistics Office takes into consideration these common sense meanings during surveys. When it is difficult to distinguish if an employee is regular or temporary, whether or not he/she is expected to receive severance pay serves as an important criterion during the survey. If an employee is not expected to receive retirement pay from the employer, even if the employment contract is without fixed term, helshe is classified as a temporary employee. If an employee is paid on daily basis and hisher employment contract is terminated automatically when the project is completed, helshe is classified as a daily employee even when helshe is expected to work more than a month in a workplace. Vhipjang is an informal organizerof the manpower pool in the construction industry.He is not an officially permitted employer or business owner but concludes a contract with the construction company as an independent contractor to pertom a process of construction. Traditionallq: any experienced worker who is able to organize a group of craftsmen and workers in the construction industry has played a role as a Shrpjangwhen he finds asmall construction project. 34 Labor Market Kefornls in Korea: Policy Optionsfor the Altur+e Under the present law, employers are expected to submit an affiliation form to the El administration of those employees who have worked at least one month. An indispensable condition for affiliating an employee to the El is that the employers should declare himlher to the El administration. In the non-regular employees' market where labor turnover is frequent, many employers do not declare their employees due to the exclusion rule of employment for "less than a month". Employers' negligence is not checked afterwards either by employees' claim or inspection by the El administration. Table 2.9 shows that, between August1998 and July 1999, 78.2% of daily employees worked more than a month in the same workplace, and 60.1% of them worked more than 6 months as wage workers. But declared daily employees are rare. For example, in the construction industry, the average number of declared daily employees is 3.6 persons per site (Hur and Shim, 1999). Table 2.9: Tenure and employed period by employment status in Korea, August 1998 to July 1999 Tenureperiod Employed period (%I (%I Months Regular Temporary Daily Regular Temporary Daily Total employees employees employees Total employees employees employees 1 1.9 9.4 21.8 8.0 0.3 2.3 6.4 2.1 2 4.5 12.0 15.8 9.0 0.5 3.0 5.7 2.2 3 1.8 5.2 10.9 4.6 0.5 2.9 6.5 2.4 4 1.8 5.0 12.0 4.8 0.5 3.3 6.8 2.6 5 2.0 6.8 10.6 5.1 1.0 4.3 7.8 3.3 6 1.4 3.2 3.3 2.4 0.8 3.2 6.9 2.7 7 1.9 4.7 4.1 3.2 0.7 3.7 6.1 2.7 8 2.0 4.7 2.6 3.0 0.9 4.1 6.9 3.1 9 2.0 4.0 1.9 2.6 1.2 5.0 8.0 3.7 10 2.1 3.6 2.3 2.6 1.4 6.3 8.1 4.2 11 2.4 4.3 2.9 3.1 2.3 7.6 7.9 5.0 12 76.3 37.1 11.9 51.6 90.0 54.5 23.1 66.0 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Note: The employed period is the period during which the workers were employed as wage and salary workers. Source: Panel constituted from Economically Acttve Population Survey database for the period of August 1998 to July 1999. Cited from Hwang (2000). The duty of declaration is left only to employers and non-regular employees remain without any employment certificate. It is usual in the Korean labor market that employees, whether they are regular or non-regular, move from one workplace to another without any employment certificate or EXPANDING THE COVERAGE O F KOREA'S Ul 35 reference. Employees are not accustomed to asking for an employment certificate when they leave a workplace. Non-regular employees lie, in general, outside of the employers' personnel management system. Those non-regular employees who did not request affiliation do not think of retroactive affiliation afterwards, either because they do not have any employment certificate or because they think of it as irksome task. Non-regular employees who want to be covered by UIS have to request affiliation to the LTlS during their tenure at the workplace. However, these employees do not know about the details of the EI regulations, such as who is excluded from affiliation and who has to be declared. If they do know, they do not have much incentive to request affiliation to an employer as they can hardly satisfy the eligibility condition. This is because on the one hand employers do not declare them properly, and on the other hand they are not entitled to UIS if they work less than a month in a given workplace. Therefore, even if they are affiliated in one workplace it will be practically impossible for them to satisfy the eligibility condition for UI as many work days are not registered in the EI database. Employers have not felt any obligation to keep employment records, and therefore the El administration is not capable of judging if an employee in a firm has worked more than a month or not. As a matter of fact, this is one of the reasons why the non-regular employees' market has been, implicitly or explicitly, considered to be an informal sector and the labor administration has until recently placed them outside of the policy area. Therefore, the remedy for improving the low coverage of employees is to establish an employment reporting system. Such a system will also resolve the irony that the most suffering group is now excluded from the UIS. Informal Employment Relations:the Case of the Construction Industry After the financial crisis in Nov. 1997, when how to protect non-regular employees became a critical issue, the term used in the debate was not 'non- regular employees' but 'daily employees'. The reason is that the most conspicuous landmark group who needed a safety net was considered to be former daily employees, in particular those who worked in the construction industry. Diagnosing the possibility and proposing a possible way of covering construction daily employees was the most pressing issue for the UIS in Korea. Therefore, it seems reasonable to concentrate our discussion around the construction daily employees' market.' The main obstacle for covering daily employees in Korea's social insurance system is that there does not exist any employment career ' Even if we substitute 'daily employees' with 'non-regular employees', it does not make any ditierence in the following discussion. The terminology 'daily employees' can be substituted in many places with 'temporary & daily employees' or 'nun-regular employees'. 36 Labor Market Refor,ms in Korea: Policy Options for the Fziture certifying mechanism for daily employees in markets such as construction. Construction daily employees are difficult to cover under an employment insurance system for two reasons. They are hired on a daily basis, generally for cash, and their employers are not accustomed to keeping records. Many of them are elnployed under vague employment relationships. Elnployees employed by Shipjangs in construction are examples of this category of workers. Figure 2. 1 shows in a simplified form how these Shipjangs fit into the employment structure in the construction industry. Shipjangs have not had legal status and their identity is unstable. Thus, construction daily employees have not been able to prove their work experience, while regular employees could provide their employment record when it was necessary by submitting the 'employment certificate' issued by the employer. Figure 2.1: Structure of construction labor employed . daily employees employees temporary 8 ,, -...-... .-.. ( DT:Shipjang ) .*. ( Dn: ) =Independentcontract -Directlycontrolled.Employer-employee relationship .-..Formally ~ndependentcontract. partially Y: Temporary 8 employment-employee relationship dally employees 1 Together with the above reasons, UI coverage of daily workers in the construction industry, where most workers are non-regular, has been hindered by the UIS's current exclusion rule. Many regulations of the current UIS are regular-employee-oriented. According to the current rules, employers are expected to submit an affiliation form within 14 days when they employ workers and also hand in a disafiliation form within 14 days when their employees quit. Employers don't have to declare to the insurance administration those employees who have worked less than a month. In the construction industry where labor turnover is frequent and where the predominant number of employees are non-regular, many employers have not declared their employees due to the exclusion rule. A large number of workers come into and out of a construction site following construction process. Many specialty contractor employers who maintain a traditional EXPANDING THE COVERAGE OF KOREA'S Ui 37 management style and share informality with the Shipjang have not conformed to the regulations. Frequent employee turnover and employment relation via Shipjangs have meant that labor administration control has been difficult. Box 2.2: Some informal characteristics of the daily employees' market O It is hard to distinguish the employer-employment relationship from an independent contract to the extent that employment is extremely unstable. Daily employees are freely dismissed by simple notice or the employment contract ends automatically when an ongoing project is completed. Not only employers but-also daily employees tend to take this for granted. O Dealing in cash is universal behveen employees and employers. They do not conclude contracts with written documents. They think that requesting any written contract is a sign of distrust and dishonor. O Employees think that the cost of litigation is vcry high (including indirect and implicit costs such as reputation damage, lost time, mental damage, etc.), or that the expected return for litigation is very small. UPeople can violate regulations just out of sympathy without any deep consideration about legal responsibility. Many do not distinguish redundant regulations from necessary ones and tend to reject regulations as a whole. Cheating or the false declaration of employment information is, thus, possible and its probability is not negligible. O Most of employees' time preference rate is so high that they count much present income and count little future income. They prefer not to contribute premiums even if the expected rate of return from their contribution is higher than the market return. Neglect of social insurance such as National Pension and Employment Insurance can prevail becausc of the high rate of time preference. O There are still many people of traditional mind who do not have any idea that government can run a contribution-based cash benefit program for them. Therefore, employers 'negligence, active or passive, of their duty is not rare and employees fail to ask of the requirements of social insurance. Non-regular employees have been beyond the scope of the employers' personnel management. Thus, it has not been possible for a non-regular employee who has resigned to request an employment certificate. And, if it is not impossible, there are many daily employees who think that it is a irksome task to request and accumulate employment records for the future possibility of receiving unemployment benefits, given those restrictions on eligibility and low probability of proper record-keeping in most of workplaces. If Korea can find any reasonable way of enhancing the coverage of the UIS for non-regular employees in construction, such a system can be extended without too many difficulties to non-regular employees in other industries. 38 Labor Market Reforms in Korea: Policy Options for.the Ftitzire Proposal for a Refined Employment Record-keeping System (RERS) Under the current EI administration process the local labor office collects insurance premiums from the original contractors. They are responsible for paying the insurance premium for the whole workforce involved in the construction project. They submit affiliatioddisaffiliationforms when they employ/layoff an employee. Specialty contractor companies are expected to submit affiliation/disaffiliationforms to the labor office and to transmit the withheld insurance premium to the original contractor company. But this process is not undertaken in actuality and neither do temporary & daily employees pay the insurance premium contribution. Four aspects of the current system can be improved upon. These are summarized in Figure 2.2. Figure 2.2: Four important aspects of the RERS Ins. premium exclusionwle RecwdYorm paymentand ernploymant dedaration 4 mlum and decla Exclusion Every employee should be affiliated as soon as he starts to work and the exclusion should be judged by the EI administration after a certain period of time. Thus, employees who work less than a month in a firm should be declared. According to the current rule, employers don't have to declare to the EI administration those employees who have worked less than a month. Employees working less than a month in a firm are excluded from coverage. EXPANDING THE COVERAGE OF KOREA'S UI 39 The El administration, however, is not capable of judging whether a daily employee at a workplace has worked more than a monrh or not and construction companies declare only a few of temporary or daily workers at their discretion. The consequence is the average number of declared daily employees is 3.6 persons per site. The pre-exclusion rule has in fact been adopted for the accommodation of the labor administration. The labor administration has considered that exclusion ex post is too costly. However, this rule encourages the moral hazard of construction companies in declaring employment infonnation to the El administration. Naturally a plausible idea to improve the low coverage of the ElS is to apply the exclusion rule differently from the way it is now. Every employee except for self-evident cases (for example, part-time student workers) should be, in principle, affiliated as soon as he starts to work and the exclusion should be judged by the EI administration after a certain period of time, for example one year after. If the total number of days worked by an employee is less than one month during a year, the insurance premium shall be paid back to the employer and the employee. Declarin~Work Records At present, maintaining work records is the responsibility of the employer only. Another modification necessary in the current EI administration is to encourage employees to declare when work has been performed. In general, in the Korean labor market employees move among workplaces without any employment certificate. Naturally, employees are not accustomed to asking for an employment certificate when they leave a workplace. As for non-regular employees, no records are left in the firm. Non-regular employees, therefore, have to be able to request affiliation to the UIS during their tenure at a workplace or, at the 'latest, at the moment of quitting. Thus, it is desirable that employees have work records when they leave a workplace. The EI administration should allow employers to declare employment information for short-term contract workers in a simpler way than now. According to the current rule, employers are expected to submit an affiliation form within 14 days of the day when they employ a worker and also hand in disaffiIiation form within 14 days of the time when their employees quit. The firms' paper workload (in particular, that of small business firms) is not negligible in employee administration due to frequent employment turnover. As a matter of fact even now there are many firms submitting affiliationidisaffiliation form far later than 14 days after the recruitlquit period, respectively. 40 Labor Market Reforms in Korea: Policy Options for the Future The labor administration can introduce an Employment Record Form (ERF) so that employers can declare at one time and, if possible, instan- taneously short-term contract employees. An ERF is a written form for employment record keeping and declaration for employers who employ workers where there is a high turnover rate, such as in the case of daily employees. An ERF is composed of two parts: one part is written at the beginning of the employment contract and a duplicate is given to the employee. The other part is filled in at the end of employment contract. The first part of ERF delivered at the beginning of employment substitutes for an employment contract and eases inspection as inspectors check whether or not a daily employee and the employer have filled in the ERF. An ERF is composed of 3 sheets. Those three sheets are shared among the employer, employee and EI administration. Employers complete- the reporting process by sending a sheet to the EI office. ~ m ~ l o ~have e r sthe responsibility for keeping the record for a certain period, for example 3 years. In sum, when an employee starts working, the employer writes the basic conditions of their contract such as the daily wage. When the job is terminated, the employer and the employee fill in an ERF and submit a copy to the labor office. The information written on the sheet is put into the database of the EI office. Employers could save on paper work by instantaneously declaring employees' record at one time while e~nployeescould declare their own employment information on their own initiative. The declaration process could be further simplified by using tools such as a reader-cum-card in order that employers' instantaneous declaration can be possible and the initiative of employees can be systematic. Construction Company's Responsibility The responsibility of paying the insurance premium and submitting the affiliationldisaffiliation form (under the RERS, it will be ERF) for e~nployeesshould be given to each company. At the moment the responsibility for paying the insurance premium for employees is imposed on the original contractors, while the specialty contractor companies have to submit affiliationldisaffiliationforms to the labor office. Specialty contractors may as well pay the insurance premium when reporting their employees. As for Y-type employees (in Figure 2.1), the responsibility for their affiliation should be imposed on the specialty contractor companies. Specialty contractors can claim that Shipjangs conclude independent contracts with them and that the labor administration should request the employment information of Y-type employees to Shipjangs. But as the Shipjang is an informal employer and Shbjang-ship is unstable, it"seems reasonable for the labor administration to impose on specialty contractor companies the duty of declaring the employment information for the EXPANDING THE COVERAGE OF KOREA'S UI workers employed by Shipjangs. Eligibility,Benefit Durationand Other Aspects of the RERS The EI, as well as other social insurance systems, can run properly only with a proper employment record-keeping system. Social insurance systems such as Industrial Accident Compensation Insurance can be run simply with employment records. The application of the El, however, requires some rules in the Employment Insurance Act such as an eligibility condition, criteria for the approval of unemployment, etc. Thus, apart from the record- keeping system, the question of how non-regular workers would be covered by an expanded UI system should be discussed. In principle, it will be desirable for the EI administration to have a single rule that applies to all worker types rather than multiple rules applying to different categories of workers. Finally there are other aspects of the RERS to be discussed. Benefit Eligibility and BenefitDuration For how long would daily employees have to pay into the system before becoming eligible for benefits? The same rule can be adopted as is applied to other employees. The benefit claimant has to be insured for at least 180 days over last 18 months. For how long could they collect benefits? The same existing benefit duration matrix can be applied to daily employees. The benefit duration ranges from 90 days to 240 days, depending upon the claimant's age and insured employment period. However, another important aspect should be discussed because daily employees frequently go through a sequence of employment and unemploy- ment spells.' It will be important for the Ell administration to carefully set rules determining when daily employees are entitled to Ul benefits. It does not seem to be consistent with the purpose of the U1 to pay unemployment benefits during a short spell of unemployment. In fact, the spirit of the waiting period rule is to prevent the payment, and thus the abuse, of benefits during a short spell of unemployment. The livelihood of the daily employees will not be threatened during a few days of bad weather or during the short interval when daily employees move from one construction site to another. Some have asserted that the system can be easily abused and the expenditure can be excessive if daily employees are entitled to UI as daily employees frequently repeat employment and unemployment spells. An- other difficulty arises from frequent turnover. During the benefit duration after a daily employee started to receive UB, he can soon find a job and can be unemployed again after several days. The rules determining eligibility have to be matched with the contribution formulas so that daily workers do "ccording to the Table 2.9, 74.4% of daily employees work less than 6 months and 21.8%of them work less than a month in a workplace. 42 Labor Market Reforms in Korea: Policy Optionsfor the Future not become an unreasonable drain on the UI system. One possible way of coping with this problem is taking fuller advantage of the job search reporting process. According to the current regulations a beneficiary must report every two weeks hisiher job search activities. The El administration can prove a daily worker is unemployed if he/she worked, for example, less than 7 days and apply a sliding scale benefit rule following the number of days worked during the previous two weeks. Someone who has worked equal to or more than 7 days out of two weeks will be regarded as employed. When would they become eligible for benefits a second time? This raises a second issue. As is shown in the Table 2.6, yearly employment span of non- regular employees ranges from 8.3 to 10.1 months a year. One simple way is imposing the same eligibility condition on claimants independently of their frequency of benefit collection. As can be expected from the employment span of non-regular employees, many of them can then claim unemployment benefits every year. To prevent excessive benefit claims, an experience rating can be introduced. Or alternatively, the EI administration can put a certain limit on the total length of the benefit span during a certain period (for example, during 3 years), Those unemployed daily workers who are not eligible for UB will be able to participate in the Public Works Program. Smart Cards When the RERS is launched, it would be an ideal opportunity to introduce Smart Cards. All of the processes done by the ERF can be done with reader- cum-card system. The EI administration can distribute mobile card readers to construction companies and Smart Cards to daily employees. At the moment of quitting, or everyday a worker comes to the construction site, he/she inserts hisiher card into the reader and records their employment information. This would enable the EI administration to track employment records person by person. Smart Cards are small cards like credit cards or bank automatic teller machine cards that are embedded with computer chips on which information can be stored. These are used, for example, throughout France as debit cards and throughout much of Europe as telephone cards, where they complement and more often replace coin telephones. Recording employment information and making wage payments to employees by Smart Card would have a number of advantages. It would get Korean workers accustomed to dealing with Smart Cards. Workers would not be eligible for unemployment benefits unless they could prove through their electronic accounts that contributions had been made and that the EXPANDING THE COVERAGE OF KOREA'S LII 43 maximum benefit period had not been exceeded. It would ensure that workers who receive unemployment benefits would not simultaneously receive benefits under other programs such as the Basic Livelihood Protection program, which will be implemented from Oct. 2000. Many have said that the RERS will not function at all because of 'informal' behavior of economic agents in the construction market. If the process is to work, at least either employers or employees should try to conform to the regulations. Ilowever, according to them, not only employers but also daily employees will not declare their employment information. Employers will resist by negligence. Daily employees will not be interested in UB at all as the benefit is too trivial and they may run the risk of paying income tax, etc. This implies that the tax administration should cooperate with the El administration when the ERF is introduced. Construction employers should withhold the income tax of daily employees to pay it directly to the tax office, while daily employees should properly settle the accounts of their annual incgme by a declaration process once a year. This directive has been impossible to impose because of false declarations are made possible by implicit or explicit collusion of employers and employees. The RERS proposed above can provide a ~nonitoringdevice for the tax administration. Conversely the tax administration can help the smooth and efficient settlement of the RERS by applying general income tax regulations to daily employees. In this way employers will be induced to be faithful to employment record-keeping and daily employees will be interested in their record-keeping. One advantage of the proposed system is that it can prevent frauds: daily workers who have become unemployed becoining eligible to collect UI benefits, but then surreptitiously returning to work. However, it will not be difficult to work without declaration at least before the RERS is completely settled because of the informalities of the construction labor market. Daily employees can be tempted to receive UB when it is possible to work without declaration. One way of preventing frauds is to operate inspection-cum-penalty rule so that the expected income of a false claim is negative. But, for the moment it is not certain to what extent this can be prevented by the inspection system. The Advantage of the RERS for Other Social Programs Covering construction daily employees with the El is a kind of bridgehead 44 Labor Market Refol-nzsin Korea: Policy Options for the Futl~re for other social insurance systems such as the National Health Insurance (NHI) and the National Pension (NP). If construction daily employees are not covered by EI, the NHI and the NP will not be able to protect them as wage workers either. The RERS can serve not only the EIS but also other social insurance systems as well. Principal Recommendations In this paper, we have identified one of major problems with Korea's unemployment insurance system. Despite the expansion in October 1998 to all firms, coverage remains quite incomplete. The compliance rate is low and the UIS covers only 65% of the employees who are supposed to be covered. The gap in coverage is mainly due to temporary and daily workers. The main obstacle to covering temporary and daily workers with the UIS is that there does not exist any employment career certifying mechanism for them. In order to overcome this, we have suggested a new Refined Employment Record-keeping System (RERS). Four aspects of the current system can be improved upon. O Under current law, employers have to declare those employees having worked more than a month in their establishment. We recommend that every employee except for self-evident cases (for example, part-time student workers) should be, in principle, affiliated as soon as helshe starts to work and the exclusion should be judged by the EI administration after a certain period of time, for example one year after. If the total number of days worked by an employee is less than one month during a year, the insurance premium shall be paid back to the employer and the employee. O At present, maintaining work records is the responsibility of the employer only. It is desirable to encourage employees to keep and declare their employment records before their employment contract ends. Temporary and daily employees may not be interested in declaring their employment information and receiving UB at all as the benefit is trivial and they may run the risk of paying income tax. This implies that the tax administration should cooperate with the El administration by applying general income tax regulations to temporary and daily employees. O In order to allow employees to maintain their work records and employers to declare the employment information of short-term contract workers in a simpler way than now, we suggest that the labor administration introduce an Employment Record Form (ERF) or Smart Cards. Ci Specialty contractors nlay as well pay the insurance premium when reporting their employees so that daily workers in the construction EXPANDING THE OVERAGE OF KOREA'S UI -45 industry are correctly declared to the labor office. Declared workers should include those workers employed byShipjangs. The proposed RERS will make it possible to extend EIS coverage to temporary and daily workers, which would make major contributions to Korea's UIS. It will provide substantial benefits to the most disadvantaged group in the labor market who face serious risks of unemployment and are not now covered by the EIS. The Potential Role of an Employment Guarantee Scheme in Korea's Social Safety ~ e t * Gary S. Fields'" The Current Employment Insurance System Beginning in November 1997, the Republic of Korea underwent a devasta- ting economic crisis. Declining macroeconomic conditions brought about major labor market disruptions in 1998: a quadrupling of unemployment, a fall of 9% in real wages, informalizationof the remainingjobs, increased job insecurity, and rising poverty and inequality. Disadvantaged groups suffered a disproportionate impact. The result was not only economic misery but also social pain: increased homelessness, rising crime, heightened school dropouts, an accelerating divorce rate, and an overwhelming sense of social malaise. 1999 marked a major turnaround for Korea. GDP grew by 10.7%,and real wages are growing apace. The unemployment rate is now less than half of its peak level. Women are again returning to the labor force and are finding employment. A sense of guarded optimism is returning to the Korean society. Nonetheless, important labor market problems remain (and social problems as well). The unemployment rate now is twice as high as it was before the crisis. But high unemploymentis the tip of the proverbial iceberg. Income inequality is 15% higher and poverty is twice as high as before. This report is part of a larger study looking at Korea's social safety net and ways of improving it. One component of that larger study is an assessment of how the unemployment insurance (UI) component of the nation's Employment Insurance System (EIS) should be redesigned or supplemented. Oficial recommendations are contained in a Korea Labor Institute report "Supplementing Korea's Unemployment Insurance System." What follows are additional recommendations reflecting my own personal views and not those of KLI or any other institution. " Cornell University,USA. THE POTENTIAL ROLE OF AN EMPLOYMENT GUARANTEE ... 47 What Korea has is not just an "unemployment problem" but a more general "employment problem." (Fields, 1999) The unemployed are not the only ones who are suffering. Those who lost jobs and have become re- employed in less good ones; those who have kept their old jobs and whose real earnings have declined; those who are employed as daily workers and temporary workers and who have fewer work days than before; and those who face precarious employment and earnings prospects - all these are part of Korea's employment problem too. Before the crisis, Korea had put into place an ambitious employment insurance system (EIS). The three components of this system are traditional unemployment insurance (UI), job training, and a job information service. Other elements of Korea's social safety net system include a Livelihood Protection Program for the poor, a Life Stabilization Fund for those unemployed whose assets fall below a certain level, public works jobs for the unemployed and for the spouses of low-wage workers, employment maintenance subsidies, support for venture enterprises, and many other smaller programs. Despite all that Korea has done, the nation's social protection system still has more holes than net. In 1999, there were 35.8 rnillion Koreans aged 15 and over. The average number of unemployed in any given month was 1.4 million. Yet, in a typical month, just 0.1 million received unemployment benefits and 0.3 million participated in public works programs. 1.7 million Koreans received Livelihood Protection benefits at some time during 1999. Taken together, about 0.5 million unemplcyed had been covered by those assistance programs. It is important to be clear about the risk being faced. Unemployment insurance in Korea, as elsewhere, protects against the risk of income loss because of unemployment. To date, the system has protected against the risk of losing all of one's income because of the complete loss of job. The EIS provides people with a temporary income after they have been completely unemployed for two weeks. The broader risk, not yet covered in Korea, is unemployment because of a less-than-full work week. UI is not intended to be an income supplementation program for people who are working but earning little; rather, it provides those who are not working with a temporary income while they are searching for employment. Thus, the millions of Koreans who have suffered income losses while remaining in a job are excluded from the EIS. Economic growth and other programs are needed to take care of their problems. It is also important to adopt a guiding framework for the EIS. One possible model is socialinsurance. Under this mcdel, in so far as possible, everyone in society pays in and all of the insured who become unemployed can collect. Under the social insurance model, benefits are limited only by the overall financial solvency of the system. A second model is the private account model. Private accounts can be publicly-run or publicly-mandated. In either case, the individual has his or 48 Labor Market Refornzs in Korea: Policy Option for the Future her own account. In the event of unemployment, the account balance can be drawn down to provide an income to the unemployed. Otherwise, the money belongs to the individual, to be used in some other way, e.g., for old age spending, home-buying, or medical expenses. Under this model, UI is a type of compulsory savings with self-insurance against unemployment. The third model is social assistance. Here, the nation taxes itself in order to provide benefits for the needy members of society. Under this model, benefits are means-tested -that is, beneficiaries must prove not just that they are unemployed but that they are poor in order to collect UI benefits. Korea's system has been set up on the social insurance model. 1 recommend that this model be continued. The next question is: Keeping the social insurance framework, what are the issues in providing the best unemployment insurance possible for Korea at the present time? Three concerns are central. The first is incomplete coverage. Only half of wage workers are insured by the UIS. 60% of employed people are wage workers. Therefore, only about 30% of working Koreans are insured by UIS. The first issue, then, is how to expand coverage so that more Koreans are insured against income losses due to unemployment. The paper prepared for this conference by Dr. Hur (2000) details a plan for doing this. The second concern is inadequacy of benefits. The problem is that in Korea, UI benefits replace only a fraction of workers' prior earnings, and the benefit duration is shorter than the typical spell of unemployment. The issue, then, is how to redesign the benefit formulas in order to provide better social insurance. The KLI report already mentioned (Korea Labor Institute, 2000) recommends doing this by extending the maximum benefit duration to 49 weeks. The third concern is weaving a social safety net. The problem is that the employment insurance system, the livelihood protection program, and the other programs in Korea, taken together, still leave major gaps. Gaping holes would remain even if the Refined Employment Record-Keeping System recommended by Dr. Hur were to be enacted, permitting UI benefits to be paid to non-regular workers in the construction sector. Accordingly, the next section of this report proposes that the gaps be filled further by creating an Employment Guarantee Scheme as a second pillar of the unemployment insurance component of the EIS. Before proceeding, let it be noted that financing is apparently not an issue. KIA studies conclude that the EIS is on sound financial footing, even after the IMF Crisis. Alternative financing arrangements are not likely to be enacted any time soon, and so that issue is not addressed in this report. A Proposed Employment Guarantee Scheme for Korea Key Features In order for the nation to provide all workers with social insurance against the risk of income loss due to unemployment, I propose a second pillar of the Employment Insurance System. This new pillar is an Employment Guarantee Scheme for Korea (EGS). This new scheme would be a formal part of the EIS and would work alongside traditional unemployment insurance as follows. As its name implies, an Employment Guarantee Scheme would consist of public works jobs, offered on a daily basis to all who want them. Under this proposal, anybody who is unemployed and wants to receive a day's benefit in exchange for doing a day's work would be eligible for an EGS job. There would be no requirement of having worked previously, no prior contributory period, and no limitation on the duration of benefits. All who are unemployed - including daily workers, temporary workers, new entrants, re- entrants and anybody else who might be interested -would be eligible, with one important exception: because the EGS would be Pillar 2 of the EIS, no worker currently receiving UI benefits under Pillar 1 would be eligible for EGS benefits under Pillar 2. Under such a scheme, establishing eligibility is not an issue. Doing a day's work would qualify the participant to receive a benefit at the end of the day. In other words, the work performed is itself prima facie evidence that the worker is deserving of benefits. What would be an appropriate wage to set? International experience has shown that guaranteed jobs cannot be too remunerative, otherwise workers will prefer the EGS jobs over private sector jobs. High EGS wages would result either in a very large and expensive system, or else the employment guarantee feature of the system would have to be sacrificed in order to maintain financial solvency. The solution that has been found to work in other countries is to set the wage for the EGS at below-market-levels - in the case of Korea, the national minimum wage might be an appropriate amount -sothat only those who cannot find more attractive jobs elsewhere in the economy would opt for the guaranteed jobs. In this way, the EGS would be self-targeting, so that those with poor market opportunities, and only those people, would participate in the EGS. The jobs themselves should be designed so that EGS participants will perform socially useful tasks. The work sites would have to be dispersed throughout the country, otherwise the guarantee aspect of the job does not have much meaning. Still, in the best of circumstances, it may be necessary 50 Labor Market Reforms in Korea: Policy Option for !he Future for workers to have to travel considerable distances in order to get to where the jobs are. The EGS would replace Korea's current system of public works. Public works were the largest single expenditure item for social protection for the unemployed in Korea in 1999, covering an average of 300,000 participants per month. The new EGS would differ from the existing public works system in several respects. First, it would be open to everyone except for persons currently receiving UI benefits; the present system, by contrast, has a detailed selection process. Second, it would be a guarantee scheme; the present system is not. Third, jobs would be offered for a period as short as a day; the present system provides continuing, ongoing jobs lasting as long as several months. And finally, the benefits would be below the minimum wage level; the current system pays benefits at or above the minimum wage. (This last feature would have to be carefully explained as a way of enabling the country to afford to offer guaranteed employment on even a daily basis.) Because the EGS will be a central part of the nation's system of social insurance and will have to be regarded as such, the benefits will need to be paid for by the larger society. Participants would make their contributions in kind by contributing a day's work in exchange for a day's wage, which will have to come from other sources. One possibility would be to levy a tax on "regular"employers; the justification for this would be that is their failure to create full employment that is creating the need for an EGS in the first place. Another possibility is to pay for the benefits out of general revenues, on the grounds that it is society's responsibility to assure earning opportunities to all of its citizens. Lessom)orn International Experience with Public Works Progralns As the details are worked out, the Korean authorities will be able to draw on much relevant international experience. Such public works programs to create employment for the poor have been established in the Maharashtra state of India, the principal city of which is Mumbai (formerly Bombay), as well as in Argentina, Bangladesh, Bolivia, Chile, Honduras, Cape Verde, and Botswana. Lipton (1998) has conducted a thorough review of these programs and recommended the following design features: design employment for low opportunity cost individuals; seek alternatives to direct targeting; use scheme rules and conditions to discriminate for the poor; allow for poor workers' frequent physical difficulties; minimize poor participants' transactions costs; reduce covariate stresses on public works resources; use retailer, employer, and public works competition 'for the poor'; before starting, check that low demand for labor causes poverty; subsidize coverage, THE POTENTIAL ROLE OF AN EMPLOYMENT GUARANTEE ... 51 sustainability, and graduation; encourage grassroots pressure groups to improve the scheme; seek complementarities among employment schemes; build up capacity of schemes and workers before works begin; and use performance incentives for officials and participants. Another source of guidance is the World Bank's forthcoming World Development Report 2000/2001, which reaches the following conclusions. 0 The wage rate should be determined by the local market wage for unskilled labor, not by the program's budget. 0 Usingadditionaleligibilitycriteriaforscreeningistobeavoided. O Wage schedules should be gender neutral. flLabor intensity should be higher than the local norm for similar projects. 0 Communities should be involved in project selection to derive maximum benefit from the infrastructurecreated. 0To maximize risk mitigationeffects, the program should be availableat all times,expandingautomaticallyduring crises,as demand increases. A final aspect, not touched upon in either of these reviews, is a technological one. When this new system is created, it would be an ideal opportunity to introduce Smart Cards. Smart Cards are small cards like credit cards or bank automatic teller machine cards which are embedded with computer chips that can be written on. These are used, for example, throughout France as debit cards and throughout much of Europe as telephone cards, where they complement and more often replace coin telephones. The bus card used in the Seoul metropolitan area is similar to a Smart Card. Making EGS payments to Korean workers by Smart Card would have a number of advantages. It would get Korean workers accustomed to dealing with Smart Cards. It would assure that workers who receive benefits under Pillar 2 would not simultaneously receive benefits under Pillar 1 - these benefits would also be paid by Smart Card. It would enable the Ministry of Labor to track usage of EGS jobs person by person. And it would reduce the scope for misappropriation of funds by local EGS administrators. Diagnosis and Principal Recommendations In my view, Korea's unemployment insurance system has three major problems. First, despite recent extensions, coverage remains quite incom- plete. Second, even if all wage workers were to be brought into the system, a large proportion of Koreans would remain without social protection. And third, the maximum duration of benefits is shorter than the typical spell of unemployment. 52 Labor Market Reforms in Korea: Policy Option for tile Ftrture This diagnosis leads to three major policy suggestions. First, because temporary and daily workers face serious risks of unemploymentand are not now covered by the current UIS, coverage should be extended to this group. In order to be able to do this, a new Refined Employment Record-Keeping System (RERS) has been recommended. Second, even if the UIS were extended to include temporary and daily workers, there would still be important groups who would be left out of the social protection system. These include some regular workers (government workers and teachers), the self-employed, new entrants and reentrants to the labor force, and those out of the labor force. In order to include those now excluded, i: has been recommended that Korea create an Employment Guarantee Scheme (EGS) that offers an assured job to all who wish to work. And third, because UI benefits do not last long enough to cover a typical spell of unemployment, benefit duration should be extended. It has been estimated that with current financing rates and coverage ratios, the system could afford benefits lasting as long as 49 weeks (Korea Labor Institute, 2000). However, I would not recommend that benefits be extended to this length for two reasons. First, the EIS needs to protect itself against a serious future adverse shock, such as occurred in the 1997198 financial crisis. Second, part of the current surplus would be better used, I think, to extend benefits to those who do not now receive them rather than lengthening benefits for those now covered. Accordingly, I recommend that the surplus funds in the EIS be used in part to extend the maximum benefit duration and in part to finance an Employment Guarantee Scheme. This combination of measures - extension of EIS coverage to temporary and daily workers, a second pillar guaranteeing public works jobs to all who want them, and longer benefit duration for the insured unemployed - would make major contributions to Korea's social insurance system. In addition to providing more substantial benefits to workers now insured who become involuntarily unemployed, the expanded UIS would also cover daily and temporary workers, the self-employed, new entrants and re-entrants to the labor force, and the re-unemployed. Still left out would be the old, the sick, and the infirm. For them, what is needed is an expanded livelihood protection program, guaranteeing an escape from poverty. Such a program has already been promised for October 2000 and is now being implemented. If this proposal were enacted, Korea would have a social insurance scheme with three major components: the current unemployment insurance system, extended to cover more wage and salary workers; an employment guarantee scheme for all who are capable of working; and an anti-poverty program for all, whether capable of working or not. The result would be a true social safety net for the first time in Korea's history. This would be a great national achievement. 4Feasibility of Introducing A I a Non-contributory Cash Benefits System for the Unemployed in Korea Young-bum Park* Introduction Korea's financial crisis put great strains on employment as Korea underwent a painful structural adjustment. The unemployment rate jumped from 2.1 % in October 1997 to 8.6 % in 1999. This left 1.8 million people without jobs. Korea was not prepared for such a high unemployment. In order to mitigate the painful social costs of the structural adjustment, the Korean government has built a stronger and durable social safety net. It expanded the coverage of the unemployment insurance scheme to all sizes of establishments and it also increased the benefits amount. Also, temporary pubIic assistance programs have been introduced. Trillions of (millions of US dollars) have been spent to create short-term jobs, mostly through public works projects. Special programs targeting particular groups such as the homeless and various training schemes were provided to the unemployed. However, despite all of the government programs, it turned out that only a small fraction of the unemployed workers actually benefited. Since mid 1999, economic conditions have improved markedly. For instance, the GDP growth rate recovered to 9.0 % for the 3rdquarter of 1999. Also, the unemployment rate fell below 5% beginning in September 1999. This confirms that the primary problem concerning unemployment was deficient aggregate demand (Fields, 1999, p.15). This also proves that the government's position relying more on temporary or supplementary schemes of social safety net during the financial crisis was right (Y.-b. Park, 1999). However, despite reduced unemployment with improved economic conditions, providing proper social protection for the unemployed is still a Hansung University, Korea. 54 Labor Market Reforms in Korea: Policy Option for the Fzltzrre serious policy concern. For instance, the unemployment rate is still twice as high as before the crisis (4.6 % in October 1999 versus 2.1 % in October 1997). The unemployment rate of young workers is still high even though it has been decreasing slowly (13.2 % in December 1999 for 15 to 24 age group). The proportion of the long-term unemployed in the total of unemployed also remains high (19.4 % in September 1999 for those unemployed more than six months). Furthermore, regular jobs decreased to less than 50%. And, inequality of urban household income increased after crisis (the Gini coefficient rose from 0.301 in the 3'* quarter of 1999 to 0.322 a year later). An increasing number of people now lives under the poverty (KLI and KIHASA, 1999; D.-s. Hwang et al., 2000). Recognizing these problems, the government plans to introduce a new social assistance scheme in October 2000. Under the new National Basic Livelihood Guarantee Act, without contributions, unemployed workers will be entitled to cash benefits if they pass means test and work test. This paper studies the feasibility of introducing a non-contributory cash benefits system for the unemployed with special reference to the one put forward by the government under the National Basic Livelihood Guarantee Act. Particularly, this report focuses on the institutional arrangement for accommodating the new scheme. This report also investigates to what extent social protection for the unemployed can be strengthened through reforms of Korea's current unemployment insurance scheme. Current Social Protection Schemes for Unemployed Workers: Contents and Limitations Coverage of the Social Protection Programs for the Unemployed With mounting unemployment, the Korean government continuously expanded the coverage of the unemployment insurance scheme and strengthened the benefits amount. The total amounts of the unemployment insurance allowances increased from 79 billion (US$ 64,000,000) in 1997 to 850 billion won (US$ 708,000,000) in 1998, then 1.5 trillion won (US$ 1,300,000,000) in 1999'). With employment improving, the unemployment allowance payment is expected to decrease in 2000. The Korean government has also put increasing resources on livelihood protection programs. Temporary public assistance programs have been introduced. In two years the government expenditures on social assistance more than doubled. Since the new social assistance scheme will be introduced in late 2000, the budget I The exchange rate of Won with respect to US dollar fluctuated between 900 and 2,000. This paper used 1,200 Won versus one US dollar which was closed to the rate after the Wan became more or less stabilized. of the social assistance of the year 2000 increased. Around 5 trillions of won were spent (and will be) in order to create short-term jobs mostly through public works projects. The government's outlay for the public works of 2000 will decrease more than 50 % since unemployment is on decline2. Various kinds of training programs have also been offered for the unemployed3. Despite all the government efforts, however, only a limited number of the unemployed persons turned out to benefit from the government programs. KIHASA-KLI (1999)4 showed that only 18.7 % of the unemployed persons benefited from government social protection programs; 16.2 % participated in only one program; 2.3 % received two kinds of benefits; and 0.2 % received three kinds of benefits. A detailed breakdown showed that 7.5 % of the unemployed received unemployment allowances; 7.7 % participated in public works projects; 4.2 % received training for the unemployed; 1.6 % received a loan for the unemployed; and 0.8 % benefited from a temporary public assistance program. Government anti-unemploy- ment measures helped 6.9 % of current wage earners that once experienced unemployment. The same study also showed that the role played by social support for unemployed households was small. The main sources of their earnings included wage earnings of other household members (54.5 %), savings (33.7 %) and borrowings (21.9 %: 21.9 % (private borrowing) + 2.3 % (government loan)). D.-s. Hwang et al. (2000) which was based on a survey taken from during the second half of 1999 also showed similar results. 22.6 % of the unemployed households benefited from the government anti-unemployment program including job placement services5. On the other hand, 16.6 %of the The unemployment rate of January 2000 turned out to surpass 5 %. The government is likely to spend more money than it planned. In that case, more government budget will be allocated to the public works projects. ' Other social protection programs are as follows. 1) special public loan for the low-income unemployed; 2) agriculture start-up loans for urban people that wish to go to rural areas to start farming; 3) provision of lunch to middle and high school students; 4) free shelters and meals for the homeless; 5) public compensation fund for guaranteeing severance payment and wages of the employees of the bankrupt firms. The Korea Labor Institute and the Korea Institute of Health and Social Affairs jointly conducted a survey to investigate how the lives of unemployed workers and their families changed after being unemployed and what they wanted the government to do. The survey was conducted and analyzed based on households as well as individuals. A total of 4,339 households as well as a total of 5,357 individuals were survcyed in depth from September 14, 1998 to October 3, 1998 The unemployed households numbered 2,771 with the unemployed persons of 3,230. Meanwhile, the number of non-unemplo)~edhouseholds turned out to be 1,561.See KLI and KlHASA (1999) for details. It includes the job placement services, but excludes the temporary livelihood protection 56 Labor Market Reforms in Korea: Policy Option for the Future households that once experienced unemployment since 1988 had help from the government programs. Unemployment insurance scheme as a social safety net An insured unemployed worker can claim the job seeking allowance or the employment promotion benefits if he or she meets the eligibility require- ments for the benefits. However, only a small portion of the unemployed workers enjoys access to the unemployment benefits. In June 1999, only 144,773 unemployed persons (10.6 % of the unemployed persons of the labor force survey) received the benefits, while the number of the unemployed workers was 1,356,000. The limited coverage of Korea's unemployment benefit scheme is attributed to the following factors. First, as of the end of June 1999, only 46.3 % (5,844,018) of the total paid employees is insured. This proportion dropped to 28.3 % if the insured are compared to total employment. The coverage of Korea's unemployment benefits scheme is limited compared to other countries (OECD, 1991, Table 7.1). Only some countries do not cover established civil servants since generally they do not run risk of redundancy (Austria, Ireland, Japan, Netherlands and Spain). There is also regional and occupational variation in unemployment insurance scheme in some countries (Belgium, Canada, Denmark, Finland, Sweden and Untied States), which is due to that earlier systems in many countries which were run by unions on a voluntary and commercial basis. However, in Korea, the excluded employment types and/or workers are excluded because the nature of the concerned jobs is temporary or seasonal or it is difficult to obtain correct information on employment record, except civil servants and private school teachers, The problem of temporary worker, part-time workers and daily workers pose big challenge for policy makers. Their number is not negligible and it has been increasing since the financial crisis took place, as shown above. The exclusion of the daily worker also contributes to not including many temporary workers since the distinction between the daily workers and temporary workers (whose employment contract period is between one month and one year) is often blurred. Second, small employers often neglect to register at the public employment office and pay their premium. But, the public employment offices find it difficult to enforce the law on small employers. As of May 31, 1999, 8,342,000 employees in 664,000 establishments were supposed to be covered by the law according to estimations based on a survey data of the National Bureau of Statistics and enterprise data of the National Tax Office (K.-s. Yoo, 1999, p.24). However, only 70 % of the concerned employees in 84 % of the concerned establishments paid the insurance premiums. As of December 31, 1995, when establishments with more than or equal to 30 employees were required to register, 98 % of the supposed-to-be-covered employees were insured. The Employment Insurance Act is the only labor law applied to all establishments6. Third, all insured workers do not receive the unemployment benefits. The law says that only qualified, insured workers can receive the job-seeking allowance (JSA) that is the main type of e unemployment benefits7. During the first half of 1999, 1,776,637 insured workers were separated from their jobs. Out of these workers only 30.5 % met the qualification (4), i.e. they were separated against their will (Table 4.1). However, only 189,582 (52.8% ofthe involuntary separators) applied for the benefits. Many ofthem (29.1% of the involuntary separators) did not meet the insured period requirement Table 4.1: Unemploymentbenefits claimantsin Korea, 1997-1999 Disqualifiedemployeesfrom Claimant the insurancescheme BIA3 Total Prop 1'' Prop 11'' Total (%) Recognized Unrecognized number(A) (%) (%) number (B) Jan.1999 167,702 36.6 28.4 41,290(100.0%) 41,054(99.4%) 236(0.6%) 24.6 May1999 192,221 25.6 19.4 28,377(100.0%) 28.102(99.0%) 27q1.0%) 14.8 June1999 227,665 35.1 23.4 25.788(100.0%) 25,582(99.2%) 206(0.8%) 11.3 Notes: 1) Proportionof the involuntarydismissed 2) Proportion of (the sum of the involuntary dismissed and ones who paid enough insurance premiums to qualify for the unemploymentbenefits) 3) This ratio needs to be understood with caution. There is a two-week of waiting period and the claimantscan choose when they receive the benefits since they can be eligible for the benefits during the next ten months afler being unemployed. Source: Ministry of Labor, Korea (calculatedfrom the unpublisheddata). "he Labor Standards Act was enacted in 1953, however, most clauses in the law are still applied only to establishments with five or more than employees. ' During the first half o f 1999, 97.5 % o f the unemployment benefits were job-seeking allowance. Employment promotion benefits and sickness benefits constituted the rest. 58 Labor Market Refirms in Korea: Policy Option for the Ftrttrre or some of them found a job before the waiting period (14 days after separation) expired or they did not know of the benefits scheme. As a result, many of the insured separated workers do not received the benefits. During the 1" half of 1999, only 16.1 % of the insured separated workers received their benefits8. This ratio has dropped substantially since March 1999, reflecting the coverage of the UIS expanded to all sizes of the establishments in October 1998. Nearly all of the insured unemployed workers who applied for the benefits received the benefits. Which means that the activity test is not rigorously enforced. Fourth, the duration of JSA varies depending on the insured employment period and the age of the claimant, ranging between a minimum of 90 days and a maximum of 240 days9. The average duration of JSA during the first half of 1999 was 118.2 days. It was 98.1 days in 1998 The increased duration of JSA reflects the 60 days' special benefits that was effective from July 15, 1998 to December 31, 1999. For a majority of the U1 beneficiaries, their entitlement period expires without their gaining employment. Among the UI beneficiaries whose unemployment was first recognized during the second half of 1998, only 24.6 % left the UI benefits with a job. Another 7.4 %lofound a job after the entitlement period expired". Finally, the rise in the number of daily workers and the high incidence of unemployment among the long-term unemployed and young people are also contributing factors to Korea's limited UI coverage, as discussed above. Social Assistance System for the Unemployed Korea has a national social assistance system with clear legal codification (which is implemented under the Livelihood Protection Law) and an unemployed person is eligible for the low-income care if he or she meets certain requirements (Table 4.2). Since financial crisis took place, a temporary social assistance program has been implemented. Under the temporary scheme, as the assets test requirement has been lowered more low-income people have benefited. In This ratio needs to be understood with caution. There is a two-week of waiting period and the claimants can choose when they receive the benefits since they can be eligible for the benefits for the next ten months after being unemployed. ' It was a minimum of 60 days and a maximum of 210 days until 1999. Since the unemploy- ment insurance benefits scheme was only introduced on July 1, 1995, the actual duration of the JSA is 90 to 180 days until June 30,2000. loThis only included worker who had a job in establishments that were covered by the unemployment insurance scheme. " The insured workers are supposed to claim their benefits within the next ten months (twelve months from April 2000) after their separation from the jobs. Some unemployed workers become not-entitled to benefits with some unclaimed benefits due to this requirement. FEASIBILITY OF INTRODUCING A NON-CONTRIBUTORY ... 59 1999, a cash livelihood benefit was also given to the low-income care recipients during the winter period and about 50 % of them received the benefits. Korea's social assistance scheme has been criticized for its limited coverage andlor its inappropriate benefit levels. 1,920,000 people, who are 4.2% of the total population, received social assistance under the livelihood protection scheme in 1999. However, among the livelihood protection program recipients, only 60.4% received the regular protection. This means that only 2.5 %of the total population received a regular social assistance. In terms of the beneficiary coverage, this ratio is very low compared to some OECD countries (in 1992)". Among the compared 23 countries, Table 4.2: Comparisonbetween LivelihoodProtection Act and National Basic Livelihood Guarantee Act in Korea LivelihoodProtectionAct (in 1999) NationalBasic Homecare recipient LOW-incomecarerecipient GuaranteeAct Regular Temporary Regular Temporary - Eligibility Age etc. Same as the Assets Same as the Recipient's criteria Older than 64 or regularone Less than 28 regular one income Younger than 18 or except millionwon per except appropriation Pregnantor assets household & assets amount Lost abilltyto work criteria(less Incomes criteria(less shouldbe less due than 44 Less than than 44 than minimum Illnessoracddentor million won 23, 000won millionwon living Peopleliving with per per monthper per expenses; thosemeeting the household) person B household) Until 2002, the abovecriteriaand No persons asset and caringthem B legally incomes of the Assets responsiblefor livelihood Less than 28 million supportingthe protection wonper household8 claimant schemewill be incomes used. Less than 23,000 won per monthper person 8 No personslegally responsiblefor supportingthe claimant Type of Livelihood;education; maternity; Education;maternity;medical; Housing aid aids medical;funeral self-supporting and emergent Regular livelihoodaid amounts Cash benefits equivalentto paymentadded arelargerthan temporaryone thoseof 50% of low-income care recipientsfor six months Work Yes test l 2Only the Livelihood Aid is considered a cash benefit. 60 Labor Market Reforms in Korea: Policy Option for the Fzlture Korea's ratio is ranked almost the bottom with Japan (Eardley et al. 1996a; Table 2.6). The coverage of Korea's social assistance is very limited, considering that the livelihood protection scheme is the only public assistance program available to low-income individuals, the national pension is not yet available for most people, and the national medical insurance scheme is also a contribution-based program. Public Works Projects The public works scheme is the program on which the government spent the largest amount of money during the high unemployment period of 1998 and 1999. The government spent about 3.8 trillion won (US$ 3,100,000,000) in two years. In 1998 and 1999, according to the government estimation a monthly average of 175,000 and 390,000 jobs were created, respectively. Without the public works, the unemployment rate of 1998 and 1999 would have gone up to a maximum of 7.6 %and 8.1 %, re~pectively'~. A survey (J.-h. Lee, 2000) shows that the participants seem to be pleased with the role of the public works as a social safety net. However, public works scheme was not much help for them to get a job. Only 31% answered that the public works helped them to get a job. On the other hand, the government officials involved with the public works had the negative evaluation about the public works scheme (J.-h. Lee, 2000)14.Except for 'helping living of the low-income people', they did not appreciate the contribution that the public works made. They believed that the public works are poorly related to employment opportunities. 90 % of the government officials answered that the public works should be abolished (58.8%) or reduced (29.2%). Training Programs for the Unemployed If an unemployed person undergoes a training program offered or authorized by the government, the trainees could have the maximum of 350,000 won (US$ 292) per month as the allowance. The types of allowance available to the unemployed tell us that the training program for the unemployed has some role as a form of social assistance. Low-income people receive a higher training allowance. However, B.-h. Lee, (2000) showed that the trainees found the allowances l 3We assume that all of the public works were created as planned and all of the participants would had looked for a job without the public works. Thus, these estimated rates are the ones with the possible maximum impact. I' The study is based on a sample survey on 585 government officials involved with the adrninistration of the public works scheme. The survey was taken from May 31 to June 5, 1999. FEASIBILITY OF INTRODUCING A NON-CONTRIBUTORY. .. 61 not that helpful for their living. The trainees from the UIS-covered establishments received the monthly average of 228,000 won (US$ 183) and 45.4 % ('very much' for 8.6%; 'some' for 36.8%) found that the allowances were helpful for their living. On the other hand, the UIS-not-covered trainees received only the monthly average of 58,000 won (US$48) and only 19.6 % ('very much' for 2.8%; 'some' for 16.8%) found that the allowances were helpful for their living. Promoting Social Protection for the Unemployed through Reforms of the Unemployment Insurance Scheme The proportion of the unemployed1 Korean persons who receive the unemployment insurance benefits is about 10% as shown above. This figure is small compared to other OECD countries such as Japan (36% in 1990) or the USA (34% in 1990)(OECD, 1994, Table 8.4)15. Like Korea, Japan and the USA have a relatively short duration of unemployment insurance benefits payment and disqualify or postpone the benefits for a long period following a worker voluntary leaving a job. Consequently, there is need to consider expanding the UIS beneficiaries, through expanding the benefit, or the scope and coverage for which, and to whom, the benefit is paid, as a way to promote the social protection of the unemployed workers. Expandingthe Coverage of the UIS to DaiIy Workers When Korea introduced the UIS, the government excluded some sectors. The government gave the following reasons for the limited coverage of the UIS (Fields, 1993). It was very difficult to collect contributions from small business or some type of workers. And, other social insurance benefits and labor laws also excluded the small business sector. Other countries also introduced the system gradually. It was also argued that the gradual implementation would enable the Korean authorities to learn from the experience. The Korean government had expanded the system to include firms of all size of firms during the financial crisis. However, some workers are still excluded even though they are those worse-off. It has been argued that daily workers are those who need to be included in the system. The number of daily workers is large and many of them are to some extent those who should be protected, more so than some others who currently received unemployment insurance benefits (J-h. Lee and J.-y. Jang, 1999). 'jThe low ratios of these countries reflect short duration of unemployrnent insurance benefits, disqualification or postponernent of benefits following voluntary job leaving and limited entitlement to assistance benefits (OECD 1994; 189). 62 Labor Market Reforms in Korea: Policy Option for the Future The exclusion of daily workers from the unemployment insurance scheme is closely related to the enforcement problem in the small business sector. 69.3 % of the daily workers are employed in establishments with less than 10 employees. Also, 63.4% of temporary employees work for establish- ments with less than 10 employees. The distinction between these two types of employment contracts is generally difficult, particularly for the small business sector. Only 20.2 % of the workers in establishments with less than 10 employees have a regular employment status. There is no doubt that the UIS should be expanded to include daily workers. However, saying that is much easier than implementing it. This is well reflected in that after the UIS was expanded to small firms, a majority of whose workers are temporary, only 70 % of the supposed-to-be-covered workers were actually covered. Many, including Fields (1993, p.12), have proposed a contribution book system to overcome this administrative difficulty. Under this system, workers are provided with contribution books, into which the employer is required to put stamps indicating that the required UIS contribution has been paid for that particular worker in the particular period. Currently, the construction industry, more than 50 % of whose employees are daily workers, has a contribution book system for severance payments on construction projects above a certain size. The employers are required to buy stamps equivalent to the wages and employment period of a particular worker and put the stamps on the contribution book of that particular worker. About 6 % of the daily workers in coostruction are covered by this scheme (H. Kim and K.-b. Shim, 1999). However, this contribution book system does not work well. The employers who are required to buy stamps are different from the ones who actually employ workers. In construction, the primary contractors hire subcontractors and the subcontractors at the bottom of the subcontracting chain are unauthorized small enterprises, actually employ the workers. Contribution books are not properly managed on an individual worker basis since the primary contractors usually give stamps to the subcontractors and the stamps are traded as a commodity in a sort of black market. The introduction of a contribution book system for the purpose of the UIS is expected to face a similar problem. Instead, J.-j. Hur and K.-b. Shim (1999) strongly argued for the introduction of a centralized employment-record keeping system. When employers hire and dismiss a daily worker, they are required to submit an employment certificate for that particular employee to the employment authorities. Under this system, basic subcontractors have the responsibility of reporting the employment certificate. However, this system will also encounter enforcement problem due to the primary-sub contract structure of the construction industry. In addition, practically this employment-record keeping system is only applicable to daily workers whose employment with FEASIBILITY OF INTRODUCING A NON-CONTRIBUTORY. .. 63 an employer is longer than a certain period, i.e., to the ones who have a sort of guaranteed employment period as seen in construction. Hiring practices of many daily workers is another problem encountered in the inclusion of the daily workers. Many daily workers get a job through private employment agencies. These private employment agencies often give advance payment to the daily workers who want to be paid on a daily basis, while employers still want to pay daily workers on a weekly basis (J.- h. Kuem, 1999). Under this circumstance, it is even more difficult to collect insurance contribution from employees as well as employers. The current income tax system also poses a problem. Employers do not have to collect income taxes if the daily employee's wages are less than 50,000 won (US$ 42) per day. This also contributes to the authorities' having difficulties in collecting information on the daily workers' true incomes. Consequently: There is no doubt that daily workers should be covered by the UIS. However, it is difficult to introduce an appropriate system under the current circumstances. Alternative means for strengthening social protection of the daily workers should be introduced. Including Family Workers or Small Employers under the UIS In 1999, employers and family workers constituted 28.8% and 9.5 %of total employment, respectively. Expanding the UIS coverage to these workers may be one way to boost their social protection. In Korea, it is well known that small employers do not report true incomes in order to evade taxes. Even medical doctors and lawyers do not reveal their true incomes. This is reflected in the harsh criticism government encountered when it expanded the coverage of the National Pension scheme in 1999, since the government did not have correct information on the incomes of the better-off small employers. If the UIS is expanded to include the employers and family workers, many of the worse-off employees will subsidize the small number of the better-off small employers. Therefore, these workers should not be included. However, this does not mean that there is no need for the social protection for the small employers and/or family workers. On the contrary, they are more like to belong to the poverty group, which will be discussed. Concurrently: The coverage of UIS should not be expanded to cover the small employers andlor family workers Entitlement Conditions for the UnemploymentBene$ts Basically, the insured unemployed worker should meet the following three conditions in order to be eligible for the benefits. (1) That person should look for a job actively and be available for the job. (2) He or she should pay 64 Labor Market Reforms in Korea: Policy Option for the Fzltrlre the insurance contribution for a certain period. (3) And, he or she should be separated with neither by the person's own will nor by his or her own fault. The first entitlement condition is rarely applied in Korea. The second condition was recently relaxed.Therefore, the issue is whether voluntary job quitters should be provided with unemployment benefits. During the 1" half of 1999, 34.0 5% of the insured separated workers who paid enough contributions for qualifying for the benefits were deprived of the benefits due to the nature of their job termination reasons. We do not know what proportion of the separated workers were those unemployed. Around 11% of the qualified insured workers who became voluntarily separated from the their jobs during the first half of 1999 became economically inactive (due to studies, marriage, housework or military service). If unemployment benefits were paid to voluntary job quitters, a substantial portion of the unemployed workers would be entitled to the benefits. Considering the expansion of the UIS coverage has limitations,especially concerning daily workers, voluntary job quitters should be provided with unemployment benefits. Korea is one of the few countries that disqualify voluntary job quitters completely from receiving unemployment insurance benefits(OECD, 1994, Table 7.5). On the other hand, with the relaxed entitlement conditions, the unemploy- ment duration is expected to be longer. Empirical evidence of other countries generally confirms the positive relationship between the generosity of benefits, in terms of level and eligibility, to the length of period unemployed although this effect is not always large (Tzannatos and Roddis, 1998, p.3). A Korean study (H.-m. Pang, D.-u. Kim and Y.-s. Ha, 1999) also revealed that unemployment insurance benefits had the adverse effect on the reemployment of unemployed workers. Then, there is a need to develop active labor market programs to reduce the possible adverse effects of the relaxed entitlement condition. Concurrently: The complete disqualification of voluntary job quitters is too harsh compared to the international standards. Moreover, relaxation of this entitlement condition seems to be only the practical way to expand the coverage of the UIS beneficiariesunder the current circumstances. Therefore, changing the entitlement condition in this direction should be seriously considered. Benefits Duration of Job Seeking AlIowance The benefits duration will be extended for another 30 days for all types of workers from April 1, 2000. However, this benefits duration period is still low compared to many of the OECD countries. There is no urgent need to extend the benefits duration period further. The Korean workers need to pay only six months during the previous eighteen months in order to qualify for the insurance benefits, in addition to that Korea has just extended the benefits duration period. Furthermore, the benefits duration can be extended temporarily under the current system, just as in the case of the crisis years of 1998 and 1999. As more jobs are being created with improved economic conditions, there is a need to wait and see how the new scheme of benefits duration period works under the improved economic conditions before the benefits duration period is extended further. In addition, under the newly introduced social assistance scheme, qualified long-term unemployed workers will be entitled to another type of unemployment benefits. Concurrently: The benefits duration should not be extended further. Need for A Non-Contributory Cash Benefit System for the Unemployed According to a joint survey of the Korea Institute of Health and Social Affairs and the Korea Labor Institute (KLI and KIHASA, 1999)the earnings of many beneficiaries of government unemployment programs turned out to be much lower than minimum living costs. 39.4 %, 65 % and 54 % were the proportion of the beneficiaries whose earnings were below the minimum living cost among the beneficiaries of unemployment allowances, loans for the unemployed and public works projects, respectively. D.-s. Hwang et al. (2000) estimated poverty groups who required social assistance by using survey results. The 1999 criteria of selecting temporary low-income care recipients (44 million won (US$ 37,000) per household for assets; 230 thousand won (US$ 192) per month per person for income) were used as the poverty line. 24.5 % of the households and 23.1 % of the individual persons belonged to the poverty group, respectively. More of urban areas and unemployed household belong to the poverty group. However, as they argued, it should be recognized that the estimation of poverty based on the survey results such as this one tend to overestimate the size of the poverty groupI6. The same survey also showed that more of the persons belonging to the poverty group had unstable jobs or were unemployed (Table 4.3). Hence, the government needed to help at least 16.4 % of them (the unemployed of 8.7 %; the disappointed unemployed of 1.8%; economically inactive, but wanted to have a job to have a job of 5.9%) to find ajob. Korea is also one of the few OECD countries that does not have a cash benefits social assistance for the unemployed. For many of the OECD countries, the ratio of the number of unemployment beneficiaries to the total l6In further analysis, D-.s. Hwang et al. (2000) assumed that 12 % of the households and 10 % of individual persons belonged to poverty group. 66 Labor Market Reforms in Korea: Policy Optionfor the Future Table 4.3: Poverty and economic activities of survey respondent Below poverty line Above poverty line (%) (%) Regular employees Temporary employees Daily employees 5.6 2.8 Employer 14.5 13.0 Family workers Unemployed I ') Unemployed II 2, Once looked for a job, now Economically Inactive Now economically inactive, but want to have a job Economically inactive 42.7 40.4 Total 100.0 100.0 Notes: 1) Unemployed workerswho looked for a job for the lastone week 2) Unemployedworkerswho looked for a job for the last two weeks Source: D-s. Hwang etal. (2000),Table 5.2. number of the unemployed is close to 100 % if the unemployment assistance benefits are included. Even low-benefit-coverage countries such as Spain, Greece and Portugal have a ratio of more than 30% (OECD, 1994, Table 8.4). Furthermore, Enlarging the coverage of Korea's unemployment insurance beneficiaries by including uncovered sectors such as the daily worker is not practical, as seen above. Thus, Korea needs to consider a non-contributory cash benefit system for the unemployed, and the Korean government plans to introduce a new social assistance system under National Basic Livelihood Guarantee Act, which will be in effect from October 2000. Under the new social assistance scheme, unemployed workers whose means are below a certain level are paid cash benefits". And, the recipients are supposed to make full use of their personal abilities to become self-supporting, which means that they should actively look for a job andlor they should try to upgrade their skills, if necessary, like the beneficiaries of the UIS are required. Therefore, a task ahead Korea is to have the new social assistance scheme to fit with the needs of the unemployed workers. Concurrently: There is need to consider a non-contributory cash benefit scheme for the unemployed and the government plans to introduce a scheme from October ''Many of the small employers andlor family workers are also expected to receive the social assistance under the new system. 52.7% of the employers and 59.0 % of the family workers fell into the 'poverty group 1', respectively, while only 26.5 % o f the employees with regular status fell below the 'poverty groupI' (D.-s. Hwang etal. (2000),Table 5.2). F E A S ~ B I L I TOF IN T R OD UC IN GA NON-CONTRIBUTORY. Y .. 67 2000. Hence, a major take ahead Korea is to make the new social assistance scheme fit with the needs of unemployed workers. Proposed Non-ContributoryCash Benefit Scheme of National Basic Livelihood Guarantee Act In October 2000, the Livelihood Protection Act will be replaced with the National Basic Livelihood Guarantee Act. Under the new framework social assistance will be available to all citizens who can not maintain the minimum standard of living subject to a means test and the beneficiaries are obliged to seek work unless they have no ability to work (Table 4.2). This means that the former low-income care recipients will be entitled to the cash benefits which would remedy one of the most serious deficiencies of the old scheme. Under the livelihood protection scheme criteria of the assets and income tests between the two recipient types are inadequate as their differences are too small (only one million won (US$ 833) for the assets and ten thousand won (US$ 8) for the income), while the difference in the benefits between them is substantial. In addition, many of the low- income care recipients turned out to have no ability to work. M.-k. Kim (1998) showed that 47.5 O/o of the low-income care recipients (in 1996) did not have any individual with ability to work in their households. Under the new scheme, double requirements of incomes and assets will be integrated into one criterion which is called 'the income appropriation amount'. There will be seven kinds of allowances; Livelihood Allowance, Education Allowance, Maternity Allowance, Medical Allowance, Self- supporting Allowance, Funeral Allowance and Housing Allowance. The Livelihood Allowances will be calculated by taking the minimum living expenses and subtracting the Medical Allowance, Housing Allowance, Education Allowance, income appropriation amount and other social payments. The minimum living expenses will be calculated by a similar method to that used under the current livelihood protection program. In assessing the income appropriation amount, the incomes and assets will be considered together. However, until 2002, the current means test of the livelihood protection scheme will be used instead of the income appro- priation amount. Housing Allowance will cover key money18for housing rental and rent for renters up to a predetermined maximum period. Housing repair costs will ''Key money is the lump sum amount of the bond money that renters pay. If a renter pays the key money, he or she does not pay periodic rents and the key money is returned to the renter when the rental contract expires. 68 Labor .Market Refirms in Korea: Policy Option jor the Future also be given to owner occupiers. The entitlement conditions for other allowances are the same as under the current livelihood protection scheme Recipients will be required to make full use of their personal abilities to become self-supporting. Those with ability to work (generally ones whose age is between 17 and 61) who are called conditional recipients will be paid the Livelihood Allowance under the condition that they will participate in programs to promote their self-supporting capability. Urgent payments will be also introduced for exceptional or urgent needs falling outside which is expected to be met from regular allowances. This allowance will be made at the discretion of local social welfare officers. Criterion Concerning Resource Unit of Claimants Under the new scheme, the social assistance claimants will be still expected to seek support from those legally responsible for supporting him or her in addition to their spouses, which is rare among OECD countries (Eardley et al., 1996a, p.65). This criterion has been criticized for being too tight (M.-k. Kim, 1999). Claimants seek support from siblings belonging to different households and these siblings often refuse to support them. The social welfare workers are supposed not to apply this criterion if the legally responsible family members, who do not belong to the same household, refuse to support the claimants. However, it is known that the social welfare workers often do not exercise their discretionary power since they are afraid of their decisions being questioned during the audit and inspection. Discussed below is the often-mentioned criticism that the government lacks the capacity to assess the resources of claimants. Particularly, local governments do not have the proper capacity to check the supporting capability of the legally responsible family members of the claimants who do not belong to the same household. Currently, local authorities do not apply this criterion to low-income care recipients. This means that this problem will be more serious with the new scheme since former low-income care recipients will also receive cash benefits under the new scheme. J.-h. Lee (2000) showed that among the public work program participants who received the livelihood protection only 35 % got help from the legally responsible siblings. It is reasonable that the resource units go beyond the duty of spouses or of parents and children when Korea's strong tradition of family responsibility and obligation is taken into account. This strong family bond also proved to be an important informal social safety net for many Koreans during the financial crisis. But, as Korean society moves toward individualism and as government lacks the capacity to assess the resources (which will be discussed below in detail), social welfare officers tend to apply the resource unit criterion very strictly. It has been suggested that-they should be encouraged to exercise their discretionary power of approving more of the claimants who have legally responsible family members, but can not get support from them. The government should also consider narrowing the scope of family members legally responsible for social assistance claimants as Korea becomes a society based more on the nuclear family. Concurrently: There is no need to narrow the scope of resource unit at the present time. However, social workers should be encouraged to use their granted discretionary power for favoring claimants when necessary. In the longer terms, narrowing the scope of family members when assessing the resource unit also should be considered as Korea moves further toward individualism. Resources of Clainzantsfor Bene$ts In calculating the income appropriation amount, virtually all kinds of income and resources of all members of the household are taken into account. The same household members include persons registered as the same household with the citizen register of local governments. The spouse and not-married children of the household head as well as the spouses of the children and grandchildren of the household heads (if the children and grandchildren are registered as the same household) are also considered as the same household members. All assets, including fixed assets, will be counted as resources in the means test. They will be considered to generate income that is assessed by using a formula. All liabilities and a basic amount of assets (that is deter- mined by the Minister of Health and Social Welfare) will be ~ubtracted'~. The recipients will be allowed to retain some of their earnings. Higher disregards will apply to earnings made through rehabilitation centers for handicapped persons as well as student earningt0. Despite that the Korean system has the features necessary for preventing the possible negative effect of the social assistance scheme, such as earning regards, to some extent it is meaningless to discuss the institutional features of the Korean system concerning the treatment of resource. As seen above, the size of the informal labor market is substantial. Small independent businesses in general do not report their true incomes. Many claimants are likely to claim a payment while still working in the cash economy. However, the government does not have the proper administrative capacity to check the reported income1assets of the low-income claimants. In other words, due l9Korea is one of the few OECD countries that take assets into consideration in assessing resources. Korea also has wider family base as the resource unit as seen above. Therefore, adult children might be expected to sell their properties before their parents have recourse to social assistance, which seems unreasonable. Less strict rules have to be applied to assets in assessing the resources of the legally responsible family members who are beyond the nuclear family. 2"It was also proposed that higher earning disregards should be applied to earning made through public works and participation in group/community cooperative enterprises. However, this proposal was abandoned due to the opposition of the budget otfice. 70 Labor Market Reforms in Korea: Policy Option for the Future to the same reason why Korea can not include daily workers and/or small employers in the coverage of the unemployment insurance scheme, Korea's system of assessing the resources will not work well. The fact that the government decided to postpone the implementation of the income appropriation amount method until 2002 itself proves that the government is not ready for using the new resourcetreatment measure. Currently, in selecting the livelihood protection recipients, the govern- ment uses an 'estimated income method' in order to remedy this problem. Based on the discretion of the social welfare workers concerning possible working days and daily incomes, the recipients are given an estimated income. However, the estimation income method has many deficiencies. It is often pointed out that social welfare officers have too much discretion in assessing the estimated income and their power is not exercised properly. Often, an estimated income is allocated to persons who are not capable to work or do not have a job. Not-married children who are registered as the same household, but do not live with the claimant, are also given an estimated income. Under the new scheme, it is very likely that social welfare workers will also have same discretion power to determine the estimated incomes used for assessing the amount of the Livelihood Allowance. Particularly, concerning conditional recipients who seem to have some incomes, the authorities will decide their income based on the average income of relevant occupations or industries and with the assumption that they work two days a week. Many suggest that the estimated income method should be abandoned and more scientific ways to assess the claimants' resource should be developed (D.-s. Hwang et al., 2000; M.-k. Kim, 1999). However, due to the peculiar structure of the Korean labor market such as the large size of informal sector and poor income tax collection system as discussed above, the scientific method alone does not guarantee the effective and fair assessment of resources of claimants. This problem will also be more serious under the new basic livelihood guarantee system since about one million of former low-income care recipients have cash benefits based on estimated incomes. Developing a data-matching system, which matches data held by the local government authorities with data held by other agencies such as the National Tax Office, the National Pension Authorities and the National Medical Insurance Authorities to identify possible discrepancies is the most prominent mechanism to reduce possible incorrect payments. However, in Korea, establishing a data-matching network of this sort takes some time considering that even a nationwide data network connecting different local governments has not been established. The Korean government plans to finish the installation of the nationwide public administration data network of local governments and a data- matching scheme of different government ministrieslagencies by September 2000 (Ministry of Health and Social Affairs, 1999, p.114). But there is still some doubt about the government capacity in finishing these installations on time. Concurrently: A data-matching network for controlling incorrect payment and fraud should be established. However, establishing this system alone does not guarantee the effective and fair assessment of resources since the govern- ment's lack of capacity to control cash economy is evident. Government's Capacity to Refer Proper Activities to Promote Self-Supporting Ability of Conditional Recipients Under the current livelihood protection scheme, generally claimants initiate their claims in person by visiting their local government offices. Then, social welfare officers review their claims and decide their eligibility and the amounts of aids. Claims should be reviewed at least once a year. In principle, the allowances are paid monthly and the payment should be made directly to claimants' bank account This procedure of making a claim and receiving the payment will be the same under the new scheme. But, as the new system is introduced. a thorough examination of claims based on the new entitlement conditions will be conducted during May to July of 2000. Once the recipients are determined, in case of conditional recipients their claims should be reviewed every three months. The government is expecting substantial extra burdens to be placed on social welfare officers, as the-recipients of cash benefits will rise from 540,000 in 1999 to 1,540,000 in 2000. In order to encounter this situation, staff numbers will increase gradually from the current number of 4,200 to 7,200 by 2001 (government estimation; Ministry of Health and Social Affairs, 1999a). In 2000, 600 more of social workers will be hired. The government claims that even with 7,200 of staff the ratio of the social welfare officers to all recipient households will be still low compared to those of advanced countries (Ministry of Health and Social Affairs, 2000). However, increasing the number of social welfare staff is not sufficient. Under the new scheme conditional recipients are supposed to participate in activities to promote their self-supporting capabilities and social welfare officers are supposed to refer recipients to self-supporting programs that fit a particular conditional recipient to a particular program. -programs for promoting a recipient's self-supporting capability include educatiodtraining, job placement services, public works and support programs for self- employment. But, social welfare officers do not have experiences of or expertise in matching recipients to appropriate programs and many social workers will be newly employed. Since registration with public employment ofices is considered an activity to promote the self-supporting capabilities, some sort of partnership is to be established between local governments and public employment 72 Labor Market Reforms in Korea: Policy Optionfor the Future offices, the details of which will have to be worked out among concerned government ministries. But, it is often pointed out that Korea's public employment services are still incompetent (J.-h. Keum, 1999; S.-h. Kim, 1999; Korea Labor Institute, 1999a). The number of job seekers registered with public employment services increased substantially during the crisis years. This is due, in part, to being mandatory for job seekers to register with the public employment services in order to participate in government unemployment programs. But, the penetration rate (the fraction of job placements that are filled through the public employment offices) is still low (9.1 % according to KLI and KIHASA, 1999). Seeking information is the main reason for using pubic employment services for 76.9 % of the unemployed who used public employment services. The same study showed that most customers of the public employment services are not satisfied with the services provided by the public employment offices. Another study done a year later also showed that the quality of the services of the public employment offices did not improve much (J.-h. Keum, 1999). The majority of the staff in public employment offices have little on the job experiences and their turnover rate is very high (Korea Labor Institute, 1999a, p.30). Some (Korea Labor Institute, 1999a; J.-h. Keum, 1999) argue that the public employment services should be enlarged. The size of the public employment services is still very low compared to other OECD countries such as Germany and USA (S.-h. Kim, 1999, Table 4.3). However, expanding the public employment service needs more thorough considera- tion. Fields (1993) argued that it would be justified if job seekers had poor information regarding job vacancies and had difficulty becoming employed due to lack of information about where the jobs were. He claimed that this was not the case in Korea before the crisis. He also argued in another study (1999, p.16) that job placement services was criticized in Korea during the crisis because Korea's unemployment problem was not frictional, hence placement services did not work due to that they focused on the wrong problem. Public works need to be a big component of the programs that conditional recipients are required to participate in. Public works is a responsibility of the Ministry of Home Affairs and local governments, while social assistance is that of the Ministry of Health and Social Affairs. However, coordination among relevant government agencies seems to be poor. In 2000 the government plans to spend 64.5 % of the public works budget in the first three months based on the assumption that unemployment rate will be high during this period. There will be little of the budget left for public works related to the new social scheme, which will be implemented in October of this year, unless the government allocates more budget later this year. Also, training will not absorb a large number of the conditional recipients. In 1999 about 60,000 unemployed low-income people went through a government-supported training program. This is very small compared to the number of the conditional recipients that are expected to be around one million people. The government plans to encourage the development of group/community cooperative enterprises. The concerned budget will increase from 1.2 billion won (US$ 1,000,000) in 1999 to 2.7 billion won (US$ 2,300,000) in 2000, and the designated enterprises will also rise from 20 to 70. S.-h. Kim (1999) argued that the budget for group/community cooperative enterprises should be at least 500 billion won (US$ 416,000,000) in 2001. However, this much of the sudden rise of the budget can not be expected. All of these discussions mean that the government is not ready for implementing a work test scheme through which claimants are required to meet the new entitlement conditions. The government should consider a way of operating the work test that requires only some competence of the authorities until a further institutional framework is developed to accommo- date the new entitlement conditions to reasonable level. Co~lcurrently: Korea is not ready for implementing a thorough and reasonable work test. Thus, the government should operate a work test scheme that requires only some competence of the authorities until a further institutional framework is developed to accommodate the new entitlement conditions more thoroughly. Benefit Level of Allowuncesfor Conditional Recipients In 2000, a home care recipient of a regular program receives a monthly average of 188,000 won (US$ 156) of aids (135,000 won of livelihood aid; 6,000 won of education aid; 49,000 won of medical aid) (Ministry of Health and Social Affairs, 1999a, p.21). The actual amount of aids he or she receives is dependent on the amount of the incomes of his or her household and the number of household members. A home-care recipient of single- member household on a temporary program receives a maximum 79,000 won of livelihood aid (US$ 66) in a month. The actual amount of livelihood aid of temporary program also depends on the recipient's income and the number of household members. A low-income care recipient of single- member household on temporary program receives a maximum 79,000 won (US$ 66) in a month, but the aid is paid only for three months. It is difficult to estimate the exact amount of additional cash benefits for each recipient with the new scheme. But, it may be inferred from the allocation of the government budget for the years 2000 and 2001. Additional benefits will not be that large for former home-care recipients. According to the government's budget plan for the year 2000, a monthly average allowance of home care recipients of the current livelihood program will be increased by 9.0 % from October 2000 (Ministry of Health and Social Affairs, 1999a, p.113). According to the Ministry of Health and Social Affairs, the additional budget that it will need for the year 2001 is a maximum 900 billion won 74 Labor Market Reforms in Korea: Policy Optionfor the Ftrture (US$ 750,000,000) or a minimum 500 billion won (US$ 417,000,000) (Ministry of Health and Social Affairs, 1999b) ''. If the government decides to spend 500 billion won, and this additional 500 billion won is spent only for the cash benefits of additional recipients, each additional recipient will have additional 500 thousand won for the whole year. The annual additional amount of 500 thousand won is divided by nine months since the additional cash benefits recipients already receive the livelihood allowances for the three months in 2000. This means that amount of their monthly allowance for the first nine months will be a maximum 56,000 won (US$ 47) per person22.Or, if the government decides to spend 900 billion won, the benefit amount of a month per person will be 100,000 won (US$83). Three-member household, which consists of one adult and two children, for livelihood aid, would receive a maximum 300,000 won (US$ 250) to a minimum 168,000 won (US$ 140) per month. If education allowance of 6,000 won and medical allowance of 49,000 won are included, this household will receive a maximum 465, 000 won (US$ 387) to a minimum 333,000 won (US$ 278). This is not low compared to Korea's other social benefits. The monthly minimum wage of between September 1999 and August 2000 is 361,600 won (US$ 301). The monthly allowance of this household will be a minimum 92.2% of the minimum wage. The minimum amount of 333,000 won is 77.4 % of the minimum monthly allowance of 430,000 won (19,000 won times 24 days) that a participant of public works would receive. Also, it is not low compared to the benefits levels of social assistance of other OECD countries (Eardley et aL, 1996a, Table 5.1 & Table 5.2). This means that the government has to make every effort for conditional recipients to be referred to appropriate activities to promote their self-supporting activities2'. Concurrently: Conditional recipients of the new basic livelihood guarantee scheme will have reasonable amounts of allowances. This makes the government's need to deliver the new social assistance system with a proper arrangement of work and means tests more compelling. 2LIt also includes budget for hiring more social welfare workers and the amounts equal to providing 9% increase of allowances to former home care recipients for the first nine months compared to year of 2000. 22It is also assumed that the number of recipients and the amounts of the cash benefits per recipient remain the same. 23If a reasonable amount of budget is not allocated to the new scheme, it is very likely that many conditional recipients will be only those with distinct labor market handicaps. It is due to social workers will have much discretion with the government lack of capacity in work and means tests. It can be inferred from types of households receiving social assistance in Japan, the system of which Korea is now mimicking. In Japan, the largest proportion of social assistance recipients is older, sick or disabled people (Eardley et al. (I996b), Table 14.4). PubIic WorbScheme:Alternative Means of Social Protection for the Unemployed Among active labor market programs such as training, retraining or support for self-employment, which the authorities are supposed to refer to condi- tional recipients, public works often act as a temporary safety net by providing current cash benefits to unemployed people. In Korea, public works scheme has been extensively used since the crisis began. It took the largest fraction (25.9%) of government expenditures concerning unemployment during the crisis years of 1998 and 1999. It will be phased out as the country overcomes the crisis. In 2000, public works budget decreased by 54.7 %. It was 49.7% of the total cut of unemployment program budget in 2000. But, maintaining public works scheme to a certain level, even after the crisis is over, should continue to be a major alternative social protection means to a public assistance such as the basic livelihood guarantee scheme or an unemployment insurance system in the delivery of social welfare. It is even more so since Korea is not yet ready for imposing reasonable work and means tests to many of unemployed workers who will receive social assistance, as seen above. Public works should be maintained as an a major means of social protection of permanent nature in the sense of maintaining the scheme until necessary conditions for imposing work and means tests fairly and efficiently are met to a satisfactory level without regard to how long it will take. Public works, unemployment assistance (cash benefits without contri- bution) and unemployment insurance schemes should be managed as an integrated system as, in Korea, the informal labor market is large and flexible. Many of unemployed workers who need social assistance will have to have the option of choosing public aids or pubic works. However, the current features of public works scheme (and the perception of government officials about the public works) will have to be changed somewhat. Like other public works schemes, Korea's current program is to provide temporary employment, in addition to its other role as a short-term social safety net. As seen above, government officials as well as the participants were positive about the role of public works program as a social safety net, while the government officials were skeptical about its role of helping re- integrating or increasing labor force attachment of the unemployed. These contracting evaluations between the participants and government officials are expected as the experiences of other countries have shown similar results (Dar and Tzannatos 1999, p.18). As a part of the integrated system of social protection for unemployed workers, public works' role of a social safety networlc for unemployed persons, who otherwise need social assistance, should be emphasized more than the role of an active labor market program. And, the system needs to be 76 Labor Market Reforms in Korea: Policy Option for the Future perceived and managed that way by all the concerned parties including the government officials. In this context, relationship between the basic liveli- hood protection scheme and public works programs including disregards concerning public works earningz4 in the assessment of resources of social assistance claimants should be re-examined. More detailed proposals of the public works scheme go beyond the scope ofthis study. Concurrently: Public works scheme will have to be a major means of social protection for the unemployed, even after the crisis is over, until necessary conditions for reasonable means and work tests are achieved to satisfactory level. Also, by providing public works to a large number of conditional recipients, the authorities will be able to manage the work test scheme only with some competence. But, the relationship between the basic livelihood protections scheme and public works program should be re-examined. Public works' role of a social safety net should be emphasized more. Concluding Remarks There is a need to enhance social assistance for the unemployed. However, implementing a non-contributory cash benefit system efficiently and fairly for them will be difficult until conditions for means and work tests for the unemployed recipients of the cash benefits reach a reasonable level. In the meantime, therefore public works should continue to be maintained as a short-term social safety net even after the financial crisis is over. 24Under the basic livelihood protection scheme, recipients will be allowed to retain 50 % of earning made through public works. 5Unemployment Assistance: Policy Implications from the Crisis of 1997 Alan Abrahart* In 1997, Korea, which had set about introducing an unemployment insurance scheme (UIS) at a measured pace, was suddenny faced with the need to adjust the scheme in a very short time. Action was necessary to ensure the viability of the scheme in the short-term and the government needed to re-think its views about the longer-term sustainability of the scheme. Questions arose that touched on options for supporting the most vulnerable groups in the economy, especially given the correlation between the lack of employment and poverty, and the role that unemployment insurance might or might not be expected to play. The events of this time offer a case study into the operal.ionof unemploy- ment insurance schemes during a crisis. They have raised issues about the extent to which such schemes need to be supplemented schemes to assist those who are not covered by insurance, that is, by forms of non- contributory unemployment assistance, particularly. This paper looks at the experience of Korea and compares it with circumstances in Australia, a country that offers only non-contributory unemployment assistance. The paper then draws some implications for the future of unemployment assistance in Korea. The experience of Korea also offers lessons for developing countries that may have reached a point where, with a growing formal sector of the economy, they, too, need to consider the type of social protection to be offered those who become unemployed. The Impact of the Unemployment Insurance Scheme during the Crisis The Korean Unemployment Insurance Scheme (UIS) was instituted in 1995, * The World Bank. 78 Labor Market Reforms in Korea: Policy Options for the Future at a time of rapid economic growth. For a short while after that, the growth continued, the unemployment rate remaining at only 2% or less. As to be expected, greater job opportunities meant declining poverty, the poverty rate falling from about 40% to less than 10% in the decade before the crisis. The UIS was initially introduced to protect workers in larger private sector enterprises, that is, those with at least 100 employees. Insurance premiums were paid 50:50 by employers and employees at an overall rate of 0.6% of wages. Compensation was set at a wage replacement rate of 50%. The duration of entitlement, shown in the following table, was related to age and years of contribution. Table 5.1: Months of entitlementto unemploymentbenefitin Korea Years of Age in years contribution Less than 30 30 to 50 50 and over 1-3 years 2 months 3 months 4 months 3-5 years 3 4 5 5-10 years 4 5 6 10 vears or more 5 6 7 Given the state of the labour market at the time, the different durations of entitlement seemed more than adequate; unemployment was low and periods of unemployment were relatively short. Prior to the crisis, less than 1 in 10 of the relatively few unemployed had been looking for work for over 6 months. The scheme was therefore consistent with a steady state unemploy- ment rate of no more than 2% and with a prognosisof continuing growth. The impact of the crisis, however, put these design parameters under severe test. In the first 12 months of the crisis, employment fell by over 5%. Taking into account the total turnaround, the effective loss of jobs amounted to more like 7%. This loss was comprised of increased unemployment of about one million and by labour force withdrawal of about another half million. Almost 70% of the additional unemployed was for males and by mid-1998, two thirds of unemployment was for males, one third for females. Table 5.2: Working age populationin Korea October 1997 October 1998 (persons in millions) Population 34.90 35.37 Labor force 21.8 (62.5%)') 21.7 (61.4%)') Employed 21.3 20.2 Unemployed 0.5 (2.1%)') 1.5 (7.1%)') Notes: 1) The numbersin parenthesisare labor force participationrates 2) The numbersin parenthesisare unemploymentrates. Source: KoreaLabor Institute(January1999). Following the onset of the crisis, the government came under pressure on four fronts: 1) The premiums could no longer cover the insurance payouts; 2) The duration of entitlements were considered too short given the increasing periods of unemployment and given that the scheme had been in operation too briefly for anyone to have built up long entitlements; 3) The levels of benefit were perceived to be too low; and 4) Many of the unemployed had no insurance cover. To address the first of these issues, the government increased premiums to 1% of wages, still equally shared by employers and employees. To address the second concern, it increased all of the respective durations of entitlement shown in Table 5.1 by two months. To address the third concern, it increased the wage replacement ratio to at least 70% of the national minimum wage. To address the last of the concerns, UIS was gradually extended to smaller private sector companies: to those with more than 30 employees in December 1997; with 10-30 employees in January 1998; with 5-10 employees in March 1998; and with less than 5 employees in October 1998. Although the government intended to extend the scheme to part-time and temporary workers this has so far proved impractical (as of March 2000). The extension of UIS to smaller companies, increased its nominal coverage to about 8.5 million, about 45% of the 19.5 million workers in the private sector. By the end of 1998, however, there were actually only 4.9 million contributors. The MOL estimated that compliance among small employers (5 employees or less) was only about 25%. The relatively fast extension of the UIS was hardly in accordance with the government's original intentions and it took place under economic, social and political stress. But it turned out to be a critical move in ensuring at least the short-term viability of the scheme. The low rate of compliance in the newly covered sectors was not the issue. The key to bridging the crisis period, which turned out to be briefer than most commentators had expected, was the increased revenue brought about by the wider coverage, despite the overall fall in compliance. At the same time, expenditure was constrained because of the low or zero level of entitle- ments that had accrued to most of those who had become unemployed. The following tables show the extent to which unemployment was heavily biased towards people with either no insurance cover or limited entitlements. The data show that people who had been unemployed for less than year, in effect, those displaced or unable to find first jobs during the first stage of the crisis, were mostly workers in Korea's secondary labor market: people who were uninsured and whose work was marginal. Less than a quarter of them had been in permanent jobs, and perhaps less than 1 in 10 had been in companies with more than 100 employees. 80 Labor Market Reforms in Korea: Policy Optionsfor the Future Table 5.3: Unemployment in Korea, July 1998 Number Category All unemployed Non- family heads Family heads Family heads -no employed members Education Middle school education or lower High school graduates College educationor higher People unemployed for less than I year Previouslya temporary worker Daily workers Permanent workers Self-employed Unpaid family workers Disengaged from firm with: 10 employees -= 10-19 employees 20-49 employees 50-299 employees 300 or more em~lovees Source: Korea Labor Institute (December 1998). Based on data from the Monthly Review on the EconomicallyPopulation. In addition, Table 5.4 shows that the impact of the decline in employment during 1998 was concentrated on either younger or older workers. Older workers may have had the option of early retirement, while the young, many seeking their first job, would have had few if any UI entitlements. In neither case would the call on unemployment insurance compensation have been as high as it would have been among employees aged 30 to 50 years, few of whom lost their jobs. As well, the major declines occurred among lesser skilled workers, in manufacturing and in construction, that is, among the lower paid. On the other a hand, employment actually increased among professional and managerial employees and in the service sector. During 1999, less than 25% of the labour force was actually contributing to unemployment insurance and only about 10% of the unemployed (or 1% of the labour force) were receiving benefits.In broad terms,throughout1999, the distributionof the working age population was as shown in Table 5.5. Table 5.4: Employment declines (job losses) in Korea, December 1997 to December 1998 Age classification('000s) 15-19 20-29 30-39 40-49 50-59 60 plus Total 1997 347 4,671 5,890 4,856 3,101 1,818 20,683 1998 321 3,956 5,856 4,811 2,835 1.741 19,520 Loss 26 714 34 45 266 77 1,162 LOSS (%) 7.5% 15.3% 0.6% 0.9% 8.6% 4.2% 5.6% Proportion 2.2% 61.5% 2.9% 3.9% 22.9% 6.6% 100.0% lndustrvclassification ('000s) Agriculture Manufgl Constr'n Retail1 Utilities Services Total /fishing mining wholesale 1997 1.917 4,433 1,966 5,863 2,030 4,474 20,683 1998 1,882 3,878 1.428 5,637 1,985 4,711 19.521 LOSS 35 555 538 226 45 -237 1,162 LOSS (%) 1.8% 12.5% 27.4% 3.9% 2.2% -5.3% 5.6% Prop'n (ex 2.5% 39.7% 38.5% 16.2% 3.2% 100% services) (Proportionsare expressed as excludingservices) Occupational classification('000s) Professional Clerical Service/ Skilled Unskilled1 Total /managerial sales agriculture prod'n workers LOSS -115 268 177 36 796 1,162 LOSS (%) -3.1% 10.5% 3.6% 2.0% 10.4% 5.6% Prop'n (ex - 21.O% 13.9% 2.8% 62.3% 100% proflman'l (Proportions are expressed as excludingprofessional and managerial occupations) Source: Korea LaborInstitute (December1998). Table 5.5: Labor force in Korea, 1999 Number % Proportions of: Classification ('000s) Population Labor force Working age population 35,000 100 Labor force 21,500 61.4 100 Employed 19,850 56.7 92.3 Nominally covered by UI 8,800 25.1 40.9 Actually contributing to UI 5,300 15.1 24.7 Unemployed 1,650 4.7 7.7 Receiving benefits 175 0.5 0.8 Source: Author'scalculations basedon governmentbudget figures(January1999). 82 Labor Market Reforms in Korea: Policy Options for the Future The ratio of beneficiaries to unemployed was thus about 1:lO. (a) Many of the unemployed were not insured to begin with. The Government expects effective coverage to increase with time to reach closer to the nominal target. But plans to include daily workers (largely building and construction workers) and part-time workers have stalled. However, any longer-term industry restructuring that takes place in Korea should increase the proportion of the unemployed that does come from the covered sector. (b) Others had exhausted their limited entitlements. It is difficult to be precise from the data available here, but perhaps two thirds of those who became unemployed were aged less than 30 years and therefore probably had no more than 3 months entitlement (subsequently increased by two months). In any case, the scheme had been in operation for only about 3 years at the onset of the crisis and consequently the maximum entitlement (see Table 5.1) that anyone could have accrued was only four months (subsequently increased by legislation to six months). (c) Still others would not have accumulated any entitlements. This would have include all those employees who were newly covered by the extensions to the UIS but who were required to contribute for at least six months before becoming entitled to any compensation in the event of becoming unemployed. In effect, the scheme was able to maintain its reserves during 1998 and 1999 only because the total impact of the full benefit structure had not really been felt. In the meantime, the GOK expected econolnic growth to resume, which it has done. However, there is still some expectation that unemployment will not come down to its former levels; anymore than it did in many OECD countries in earlier decades. Figure 5.1: Replacementand contribution rates for different unemploymentscenarios in Australia (100% takeup) i o./.: ~. - . . ,-.~ 1 , - ~ - -- .~ ..~.- ,I - - r - ~ ~ ~ 0.6% 0.6% 1.0% 1.2% 1.4% 1.646 1 8 % 2.0% 2.2% 2.4% 2.6% 2.6% 3.0%, - . -. ~ ~ ~ ~ x-axis is contribution rate; y-axis is wage replacement rate The question therefore arises as to the longer-term adequacy and sustainability of the scheme. In the absence of administrative complexities, there is a relatively simple relationship balancing a given replacement rate of UI compensation and the level of contribution needed to sustain it. The following graphs demonstrate this. For example: 0 to achieve average wage replacement of 50% under a steady unemployment level of 2%, would require a 1%contribution level. 0 on the other hand, if unemployment were to be 5%, a 1% contribution would enable a wage replacement rate of less than 20%; 0 to continue paying a wage replacement rate of 50% in the face of 8% unemployment would require the contribution to rise to about 4.3%. The simplistic approach assumes that all contributors would be entitled to receive benefits, which is never the case in any insurance-based scheme. All schemes set limits on entitlements. These may be limits on the duration of compensation, which imposes a brake on expenditure at times of higher than usual unemployment when the length of unemployment tends to increase. Waiting periods can also be introduced, as they are in Korea, the impact of this being noticeable no matter the level of unemployment. As we have seen in the case of Korea, the ratio of beneficiaries to unemployed was about 1:lO. It is not possible to say just what sort of beneficiary to unemployed ratio might be considered more appropriate, or even whether there is such a thing as an "appropriate" level. It is all a matter of the design parameters. But a ratio of 1:10 is surely low and perhaps a figure of 1:3 would be more typical. In the case of Korea, the nominal coverage of the scheme is now far wider and compliance rates may well improve with greater enforcement of the legislation. The likelihood of an unemployed person having some entitlement to UI compensation should therefore increase significantly. Under a scenario in which the beneficiary to unemployed ratio was 1:3, the current contribution rate of 1% of wages could support a wage replacement of 50% for unemployment levels up to 5%. The current UIS parameters would therefore be adequate provided unemployment does stay around 5% or less; but if the government wishes to continue with a wage replacement level of 70% the premium will need to rise closer to 1.5 or even 2%. In summary, then, a number of conclusions about the UIS can be drawn from the crisis. 0 Thesystem wasimmatureatthe onsetofthecrisis -No insurance business could be expected to handle such an event so soon after its introduction. Plans for its development over time and for the building up of reserves were seriously disrupted. 0 The short-term viability of the UIS was ensured only by an expansion of revenues through generating contributions in respect of employees who were unlikely to make significant claims for unemployment 84 Labor Market Reforms in Korea: Policy Options for the Future compensation - Although clams for compensation did exceed contributions during 1998199, the growth in contributions through expansion of the scheme was sufficient for the UIS to retain positive reserves. 0 The UIS was limited in scope to people who were unlikely, for a variety of reasons includingthe impact of labour legislation, to become unemployed. - Labour legislation restricted hiring and firing practices among the very enterprisesalso covered by the UIS; and -Changes to labour legislation are addressing this but questions concerningthe portability of insurance contribution will remain. O The risk of unemployment continues to be higher among those who were uninsured -Korea still has a significant informal sector and it has so far proved impracticable to devise a system to cover workers in marginal jobs; and - Youth unemployment presents an especiallydifficult case. 0The governmentchanged the rules of the schemethe moment it was called on to respond to the crisis -These changes were beneficial only to those who were insured to begin with and had no impact on the majority of unemployed. O Compliance with the UIS legislation is lower among smaller enterprises where the likelihood of labour turnover, and therefore intermittent unemployment, is highest -The total wage-based levies under the so-called the Employment Insurance Scheme (the EIS, of which the UIS is just one component) is now 1.5% of payroll for small firms and yet they are probably less likely to receive assistance under the various components of the scheme (see the next section). The UIS, of course, offers a guarantee of compensation under certain conditions.Contributors are entitled to compensation, under UIS guidelines, irrespective of their circumstances. It was not, and is not, a poverty-related measure. Government decisions to extend the duration of compensation, for example, had to be made universally, without consideration of contributors' individual circumstances. The greater problem during the financial crisis therefore turned out to be providingsocial protection to those who were not insured, particularly those who had fallen into poverty. Other Form of Social Protection During the Crisis The weaknesses that the UIS exhibited during the crisis will certainly be eased with time but they will not be eliminated entirely. Many people will continue to be uninsured. Others, although nominally covered, will still lack entitlements or have exhausted what little the had. This will especially be the case for young people at a time when youth unemployment is likely to be greater than before. And within the sectors that are supposedly covered by insurance, compliance with the UIS legislation cannot be totally enforceable. UI reserves stood at only a little over W 1trillion by the end of 1997. Table 5. 6 below shows that by 1999 the government budget for unemploy- ment insurance compensation alone was W 15 trillion, with the total budget for social protection being almost W 77 trillion. The budget for LI1 compensation was thus only 20% of the budget for social protection. The budget can be considered in two parts: expenditure on labour market programs; and expenditure on social safety nets, including UI compensation. The budget allocation for labour market programs (excluding UI compen- sation) was about W 20 trillion. 40% of this was spent on vocational training through a program known as the Vocational Ability Development (VAD) scheme. VAD was financed by a further levy on employers only, averaging about 0.4% of wages, although it varied according to the size of the enterprise. Subsidies were also made available to employers to avoid dismissals. They were applied through a program called the Employment Support (ES) scheme, financed by third levy on employers, equivalent to 0.7% of wages. Table 5.6: Government budget allocations o n social protection in Korea, 1999 - - -- Budget allocation Size of program (won in millions) (persons) TotalLabour MarketPrograms 2,069,300 Vocationaltraining 795,700 320,000 Loans to the unemployed 638,200 70,000 Support to avoid dismissal 491,500 1,020,000 Expansion of placementnetwork 63,600 Support for highly educated unemployed 60,000 25,000 Trainingljobplacementfor women 20,300 6,400 Total Social SafetyNet 5,621,800 LivelihoodProgram (long-term and 1,950,400 1,730,000 Temporary) Public works 1,600,000 370,000 Unemploymentinsurance compensation 1,501,200 530,000 Other expenses (supportto children, 570,200 Living costs,rural start ups, reserves) Total Budget 7,691,100 Source:Ministryof Finance(January1999). 86 Labor Market Refornzs in Korea: Policy Optionsfor the Future Both of these schemes, VAD and ES are components, together with UIS, of the Employment Insurance Scheme (EIS). The total levy on wages to finance all elements of EIS therefore averaged about 1.7%, rising to 2.1% during 1998. The sizes of the respective programs shown in Table 5. 6 represent the totalflow of people who were be assisted in 1999. A typical stock situation during the year could be expected to look something like the following table. Table 5.7: Social protection in Korea, 1999 Unemploymentassistance Number of unemployedreceiving: Training and job placement 160,000 22% Unemploymentcompensation 170,000 23% Loans to unemployed 70,000 10% Total 400,000 55% Poverty alleviation Number of people benefitingfrom: Public works LivelihoodProgram Total Total 725.000 100% Source: Authois calculationsbased on KLI (December1998) This stocktaking is against a background of an underlying stock of 1.6 million unemployed. In addition, another 1 million people were to be supported during the year by subsidies to employers designed to avoid dismissals. Since, the budget allocation for that was equivalent to about one third of the allocation for UI compensation, the monthly stock Figure 5. would have presumably been in the order of another 100,000 workers. The table shows that the social safety net budget contained three big-ticket items, which between them accounted for 90% of the social safety net budget: 0 Cash assistance through the Livelihood Protection Program (LPP), providing both long-term and temporary protection, and costing W 20 trillion for an estimated 940,000 participants 0A public works program costing W 16trillion for an estimated 370,000 people. 0 Unemployment insurance compensation costing W 15trillion for an estimated 530,000 people. The two significant areas of unemployment assistance other than unemployment insurance were thus LPP and public works. The LPP was designed originally to provide long-term assistance to the chronically, rather than to the intermittently or occasionally poor or to the unemployed as such. The Program was designed as a safety net for what was seen as a diminishing number of poor families, diminishing because economic growth was bringing about continuing reductions in the poverty rate. The Ministry of Health and Welfare reported that 87% of Livelihood registrants before the crisis were located outside Seoul. They attributed this to the impact of the assets test which value property at much higher levels in Seoul, disqualifying many people in the city. However, the number of Livelihood recipients did not markedly increase during the crisis. It essentially remained a program to deal with chronic poverty. The greater crisis response came from the expansion of public works programs. Public works had already been in place but they were driven largely at the local government level using local government finance. Because of this, the main take up of places was in Seoul, the only local government area with sufficient financial backing to put in place a reasonable scheme. Some 80% of places under the existing scheme were within Seoul, virtually the mirror image of the LPP. Work was said to be cleaning, laboring, housekeeping, house maintenance, and so on. Under the impetus of the crisis, the expansion of the program was nationwide, providing public works for the first time in many rural areas. Even so, as the 1999 budget showed, the expansion was still limited. In addition, the function of public works was not always clear and it became somewhat contentious in the legislature. Many considered that public works should provide long-term work opportunities. Public works funds were provided, for example, to the Ministry of Information and Communication for the IT projects such as work associated with the Y2K problem. In addition, the extended public works program during 1999 established priorities for selection as follows: (a) "highly" skilled workers; (b) people aged 30-55; and (c) heads of households. The first category seems unexpected and undeserved in a poverty-related program. The end result of all of this was that the Livelihood Protection Program and Public Works fell far short of the potential level of demand, leaving many people to rely on family and community support while unemployed. In discussion with the World Bank leading towards structural adjustment loans, mainly during 1998, Korean authorities made clear their reluctance to undertake any sweeping expansion of safety net programs so as to avoid any growing welfare dependency. However, as Professor YB Park's paper in this conference discusses, the government does now intend to expand the Livelihood Protection Program. In October 2000, the National Basic Livelihood Act will introduce a program to provide assistance based on the administration of work tests and means tests. In effect, the government intends to expand non-contributory assistance to the unemployed. 88 Labor Market Reforms in Korea: Policy Optionsfor the Ftlttrre Observationsfrom Another System The development of an enhanced Livelihood Protection Program will add to the array of government instruments an expanded poverty-related program that complements the system of entitlements under UIS. But a program of this type has its own complexities. Benefits are provided on the basis of need rather than entitlement and that, of course, raises one fundamental issue: how to assess need. Two methods may be used: categorical testing and financial assessments. Whether these two can be easily mixed is open to question. We may consider the example of Australia', which was looked at more closely in the context of examining the feasibility of introducing a non- contributory scheme of unemployment assistance. Australia is one of only a handful of countries that has no system of unemployment insurance, and consequently, no system of entitlementsfor the unemployed. The Australian system is designed to provide an adequate level of income for all Australians irrespective of any circumstances, other than their financial situation. The essential feature is therefore meant to be an income and assets test (a "means' test). 7 Spending on welfare payments generally increased from around 3% of GDP in the 1960s to over 7% at present, although OECD data indicate that spending on social security is only about 64% of the OECD average. Even so, the Australian Department of Family and Community Services (FaCS) reports that during the 1990s, growing welfare dependency has been a problem for Australia. The proportion of children living in families without any members in work was high in comparison to other OECD members and a significant problem of inter-generational welfare dependency has arisen. FaCS has reported that "young people from income support recipient families were more likely to leave school early, to experience unemploy- ment and long-term unemployment, to have children by the age of 19 years, to receive income support themselves andlor to be classified as homeless for income support purposes"2. The extent of welfare dependency within Australia can be seen from the following graph. ' The following draws substantially on the Technical and Other Appendices to the Interim Report of the Reference Group on Welfare Reform: Participation Support for a More Equitable Society, Department of Family and Community Services, March 2000. Copies of the Technical and Other Appendices can be obtained through the Australia Government web site or directly from http://www,facs.gov.au/ ' Australian Department of Family and Community Services (1999) Figure 5.2: Proportions of working-agepopulation(15-64)who were jobless, receivingincome support andlorunemployedin Australia, 1965 - 1998 % I n c o m e support +Unemployed The term 'jobless' is used here to refer to all people without work, whether or not they are in the labour force. One in four people of working age are now in receipt of some form of income support, although it should be understood that this includes support to students in education. As the graph shows, joblessness has actually fallen over the years, even as recorded unemployment may have grown. Part of the increases in the proportion of people receiving income support comes about because of the changing distribution of work among families. There have been increases in the proportion of couples with no paid work, at the same time as there have been increases in the proportion of couples with two incomes. The proportion of couple-families in which there is only one income has declined. In effect, there is a tendency for husbands and wives to be either jobless together, or both earning incomes. In addition, there has been a sustained increase in the rate of single-parent families, which are much more likely than couple- families to be receiving income support. Another part of the increase in the numbers receiving support comes about because of the way the income support schemes actually operate. To reduce any tendency towards welfare dependency, the system encourages recipients to earn income (including from investments) while still retaining some income support payment. The level of income support provided depends on family structure and the income earned. Naturally, income support is reduced as other income is earned. Nevertheless, almost 20% of working age income support recipients now receive only part-payments through income support because they have separate earnings, even though the total earnings and income support payments are taxable. The level of income support provided depends on the exact structure of the means test, which is meant to ensure that payments are targeted to those most in need. The test has two components: an assets test and an income test. 90 Labor Market Reforms in Korea: Policy Optionsfor the Futzlre The assets test takes into account assets such as investments, motor vehicles, bank accounts and so on, but excludes the primary (family) home. The income test is comprised of: (7 afree area -the amount of private income a person can receive from earnings or other sources before their income support payment is reduced; 0a taper rate-the rate at which income support payments are withdrawn or 'tapered away' over the range of private income between the free area and the final cut-out point; and 0a cut-outpoint - the amount of private income at which a person ceases to receive any income support payment. A tight income test (some combination of a low free area, a steep taper and a low cut-out point) ensures more accurate targeting of assistance to those with little or no income. But it can reduce the financial rewards from increasing private income (such as through earnings) through the interaction of the income test and increased tax liability. On the other hand, a generous income test (some combination of a high free area, a low taper rate and high cut-out point) will reduce the extent to which assistance is targeted to those with the lowest incomes while increasing financial rewards from working. The best measure of the proportion of income that is lost through the dual application of taxation and an income test is the effective marginal tax rate (EMTR). High EMTRs reduce the incentive to work since people get to keep only a small amount of the additional income they earn. In some cases, the EMTR may become greater than 100 %, offering little or no incentive to undertake any work. An increase in the EMTR may bring about two opposite effects: 0An'income effect': wherebyincreasesinthetax ratereducethe amount of disposable income from a given amount of work and therefore act to encourage greater work effort to make up for this loss; and 0 A 'substitution effect': whereby the return for giving up an extra hour of 'leisure' in order to work has been reduced by the increase in the EMTR and therefore encourages people to work fewer hours. The net effect on work incentives will depend on individual's preferences and their circumstances. Judging from what the FaCS has written, it appears inclined to consider that mothers who are only second income earners in a family are more influenced by the substitution effect; while primary earners are more influenced by the income effect. EMTRs are not easy to calculate and although individuals may be aware of the general impact that any change to the system will have on the level of their income support payments, they are unlikely to know accurately what EMTR they face. However, a household survey by Australian Bureau of Statistics in I998 showed that the way schemes are constructed and the effect that earnings may have on pensions and benefits is not necessarily the individual's major consideration in deciding whether or not to participate in the labour force. Non-financial factors are aIso important, such as self- UNEMPLOYMENT ASSISTANCE: POLICY IMPLICATIONS... 91 esteem, social contacts and work ethic. In addition, income support recipients may be influenced by the availability of other benefits. For example, the prospect of losing a concession card, which allow recipients access to cheaper public transport or health services: may be a greater disincentive to go off income support entirely than a high EMTR. Although non-contributory schemes do not have specified wage replace- ment rates as an insurance scheme would have, it is still possible to assess the generosity of a scheme by using effective replacement rates, based, say, on average earnings as a proxy measure. Until the 1970s, income support for those of workingage in Australia was generally paid at a relatively low rate and there were few incentives to stay on unemployment benefits: During the 1970s,there were large increases in payment levels effectively pushing replacement rates much higher. Some of the increases took place in the early part of the decade when it was thought that the economy could afford to be more helpful to the poor. A national poverty inquiry in 1973 assessed the benchmark income required to support the basic needs of a family of two adults and two dependent children. Equivalence scales were devised for other types of family and government policy was designed to increase income support levels gradually, to the point where family income at least reached these levels. Because of supplements payable for spouses and children, the higher replacement rates that began operating in the 1970s had their greatest impact on families that might have otherwise had only a single income earner. For larger families in these circumstances, total income support could often exceed the likely earnings from that person. In addition, supplementary benefits available to people who were renting privately could push the effective replacement rates even higher. The introduction of means-tested family payments increased the attrac- tiveness of lower-paid work to those with children although at the same time real income support payment rates continued to rise gradually, which offset the effect of the new family payments on replacement rates for families. As the following figure shows, typical replacement rates for families with children did rise during the 198Os, due largely to increases for people who were renting privately. Although the increases were more discemable for single people, again especially for those renting privately, they were from a low base. It is evident from the figure that replacement rates are considerably higher for couples and families with children than for single people. This is inevitable given that the system is based on family need. However, it also means that welfare dependency is more likely to occur among larger families as they may face little net return from one low-paid full-time job and, if the costs of working are high (for example, transport or clothing costs), they may be no betteroff at all. 92 Labor,Market Reforms in Korea: Policy Options for the Future Figure 5.3: Net replacementrates of income support versus minimumwage, various non-renter income unit types in Australia,1972 - 1997 (%) 100: 10 0 L 1972 1977 1982 1987 1992 1997 Year Despite all of this, FaCS reports that Australia's net replacement rates (taking account of both taxation and other benefits) is actually comparatively low among OECD countries. For short spells of unemployment, Australia had the fourth lowest gross replacement rate for single people out of twenty- one OECD countries (1996) while replacement rates for couples were consistent with the OECD average. FaCS states that the "income floor set by income support payments in Australia is sufficiently low, relative to wages, to have had little impact on the supply of full-time workers (compared to other countries)". Finally, it should be pointed out that although the Australian system is essentially needs-based, it still retains many features of a categorical system. The three most important categories to be considered here are the unemployed, the disabled and youth. As with any system of unemployment insurance or unemployment assistance, the unemployed must face activity tests, that is, they must be available for work and, under various conditions, be willing to accept work that is offered to them. Immediate problems arise. Should a person be required to take work that is distant from their home and how is the term 'distant' to be defined? Should a person be required to undertake work that is not related to their experience o r is at a low wage rate? Should they be expected to take up a 3-D jobs (difficult, dangerous, demanding)? The issue becomes especially relevant when considering how to deal with public works programs. In Australia, public works programs have come and gone in fashion (but are generally very poorly regarded) but they have never been treated as poverty programs per se. Wages have been set according to market standards for the work being undertaken. This has meant that public works programs could be used to administer activity tests. It would not be possible to administer a pure poverty-related public works program in this way. Activity tests in Australia have been increasingly tightened over the years and are now increasingly enshrined in the phrase "mutual obligation". Recently (March 2000) the Australian announced its intention to further tighten the tests, making it obligatory for unemployment benefi- ciaries to accept any jobs offered to them. This new hardening of policy is taking place against a background of relatively strong economic growth. However, during periods of economic weakness and of slackness in the labour market, administering activity tests has special problems. The number of unemployed registered for work may be many multiples of the number of jobs notified to the national employment service. It is impossible under these circumstances to administer a tight test. The end result is that activity tests become most rigorously applied only at the same time as unemployment falls anyway as a result of economic growth. In effect, they mop up the least job ready, those who are probably the most welfare dependent. The requirements faced by applicants for disability pensions involve medical tests designed to measure a person's impairment for work. Their conceptual basis is therefore quite different to that of activity tests. The following table shows the relative movements in the numbers of disability pensioners and unemployment beneficiariesover the years. Table 5.8: Number of recipients of income support in Australia, 1965-1998 1965 1970 1975 1980 1985 1990 1995 1997 1998 Disability pensioners Number 107,500 134.500 171,500 236,800 271.500 328,200 464,400 527,500 553,300 Annualrateof change 4.6% 5.0% 6.7% 2.8% 3.9% 7.2% 6.6% 4.9% Unemplovmentbeneficiaries Number 12,700 13,000 160,700 311,200 561,400 419,800 795,500 801,800 790,300 Annualrateof change 0.5% 65.4% 14.1% 12.5% -5.6% 13.6% 0.4% -1.4% Ratioof DPlUB 8.46 10.35 1.07 0.76 0.48 0.78 0.58 0.66 0.70 The impact of the recession in the mid-1970s on unemployment beneficiaries is clear. Prior to that, unemployment was low, perhaps 1% or less, and the number of people with disabilities far outnumbered unemploy- ment beneficiaries. Since then, the number of unemployment beneficiaries has fluctuated somewhat, partly to do with economic cycles and partly to do with changing rules on eligibility. But the number of people on disability pension has steadily increased, by a factor of 4.3 in the 30 years to 1995. Disability pensioners are heavily biased, virtually 2:1, in favour of males. The most common conditions of disability are musculo-skeletal and psychological/psychiatric,which between them make up just over 50% of all disabilities. 38% of all new approvals are in respect of people who are transferring from unemployment benefit. 94 Labor Market Reforms in Korea: Policy Optionsfor the Firture On the face of it, then, the evidence suggests that the growth in disability pensions is one example of increasing welfare dependency, and one that is related to the underlying growth of unemployment in the last quarter century. FaCS now estimates that by 2006 over 750,000 people will be in receipt of a disability pension and that they will once again outnumber unemployment beneficiaries. Unlike unemployment benefits, however, changes to the regulations on disability pensions are not so easily implemented. Appendix 3 to the Interim Report on Welfare Reform in Australia highlights two patterns af income support receipt. One pattern "is found among (mainly older) people who come onto payment and remain there until age pension or death. Payments that show this pattern includes Disability Support Pension...There is little turnover in the populations on these payments". Another pattern "is found among people on (unemployment assistance). These payments have high population turnover, implying signficant proportions of short-term receipt. It is common for people on these payments to intersperse periods on inconze support with periods of self-reliance". The last category to be considered is young people. The impact of various income support measures on the activities of young people has long been debated. Because unemployment benefits were not insurance-based, people as young as 16 were eligible for benefits immediately they left school. This did not become a problem until youth unemployment rose as a result of circumstances in the mid-1970s but thereafter the policy came under heavy criticism for being a significant contributor to unemployment. Leaving aside the merits of the long debate that ensued, the consequence in the 1980s was a sharp improvement in allowances for students and a reduction in allowances for unemployed young people, especially those under the age of 18 years. In the 1990s, youth unemployment rates remained high although the absolute number of young unemployed did decline because of dramatic increases in student retention. Youth have therefore been separated out in terms of income support. Student allowances and unemployment benefits have been brought together under a Youth Allowance, applying to people under the age of 21 years. The Youth Allowance is subject to an activity test, which requires beneficiaries to undergo a work test if they are unemployed or maintain full-time attendance in studies if they are students. In addition, the Allowance is subject to a parental means test as well as a personal means test. Exceptions apply, of course, for young people who must be considered as independent, such as those who are married or those, including single parents, who have dependents of their own. Issues for Unemployment Assistance in Korea The main policy issue for Korea, unlike many other OECD member U N E M P L O Y M E N T ASSISTANCE: POLICY IMPLICATIONS ... 95 countries, is not how to overcome the well-established problems that have developed in social protection systems for the unemployed but how to avoid them in the first place. Those well-established problems faced in varying degrees by OECD members are: O Persistent long-term unemployment; OIncreased reliance on income support; OThe polarisation of households into work-rich and work-poor; OHigh levels of welfare expenditure; 0Financial disincentives to work; O The impact of increasing part-time and temporary work in the labour market; and 0 Changes to family structures, especially the growth in single parent families. In effect, development in Korea and reform elsewhere should end up at similar points. Korea has already chosen to continue developing its unemployment insurance scheme, UIS, and to supplement it with a non- contributory scheme of assistance based on its existing Livelihood Protec- tion Program. It has not chosen to avoid welfare dependency, clearly one of its major concerns, by avoiding or minimising welfare altogether; but it will still be necessary to proceed with caution. The future of the UIS is perhaps more certain but even here issues of administration and management need to be considered. The first question to consider the extent to which the UIS and LPP are to be considered in the one administrative framework. The UIS, notwithstanding its 'insurance' tag, is really a first tier of social protection, a system of entitlements generated from contributions that enable a person to claim benefits not available to others. Controls are placed over the way that entitlements are established, much as they are now, for example by being limited in size and duration. In addition, there may therefore need to be tighter controls over the number and type of occasion on which compensation may be claimed; especially as Korea is increasing its labour market flexibility, one consequence of which could be higher labour turnover among the insured labour force. This will become important if the effective coverage of UIS does increase among smaller firms and if ways are found to extend UIS to day workers or part- timers. There would need to be restrictions on how many times compen- sation can be claimed, perhaps over some specified timeframe, or longer waiting periods may be needed. It is also important to decide on the extent to which UIS should be self- contained and self-financing. Total monthly payouts exceeded total monthly contributions for a time during the crisis, but the reserves remained in surplus. They were, nevertheless, run down and had the crisis continued, or even if the level of unemployment had risen again, it would probably have proved impossible to protect them. If the UIS were to be regarded solely as an insurance scheme, it would not be possible to run an operational deficit, 96 Labor Market Reforms in Korea: Policy Options for the Future or even allow the reserves to be exhausted, for anything other than short periods. Strictly speaking, allowing long-term deficits to develop runs the risk of undermining the whole insurance of the scheme. However, unemployment insurance does not have the same operational basis as, say, health insurance where the risks are more easily spread over time. It is probably inevitable with unemployment insurance that a government would want to provide financial support to the system during economic downturns; and even more obviously during periods of crisis or sharp economic adjustments. Neverthe- less, it is still important, especially when government comes to consider changes in guidelines, to know how-self-contained the UIS should be. Short- term changes that are inconsistent with the longer-term sustainability of the system would not seem desirable. If the system is not to be self-contained, there will be a disconnection between the guidelines governing contributions and the guidelines govern- ing compensation, the one not really governing the financial viability of the other. In this case, the contributions, no matter how separately accounted for, are really another form of taxation and could just as easily be paid into consolidated government revenues, with the system operating more closely akin to a partly-funded system of unemployment assistance. This raises questions about the conjunction of the two schemes, UIS and LPP. The Australian experience, and probably the wider OECD experience as well, suggests that welfare dependency becomes generated among people who would not have UIS entitlements to begin with. They are precisely the sort of marginal workers, often working in the informal sector, who suffered most during the 1997-99 crisis in Korea. The threshold issue to consider is the level of assistance to be provided under the two schemes. In 1998, the estimated monthly cost of a minimum consumption basket for an average household was about W 800,000. This was about 50% of average earnings, which were about W 1.5 million during that year. The average payment to UI beneficiaries throughout the year was W 578,000 per month; about 40% of average wages. However, there were substantial differences in earnings across different employer groups: 0 Monthly earnings in larger companies were W 1.7 million; 115% of average wages; 3Monthly earnings in smaller companies (10-29 employees) they were W 1.2 million; 80% of average wages; 0Forcompanies with fewer than 5 employees, theaverage monthly wage was only W 600,000; 40% of average earnings. DThe national public works program introduced during the crisis allowed 20 days work a month at 23,000 won per day giving a maximum of 460,000 won a month; 30% of average wages. DThe original (pre-crisis, local government-based) public works program restricted work to 15 days a month at a going rate of about 17,000 won per day giving a maximum of 205,000 won a month; 15% of average wages; These figures are meant to give broad comparisons only and can hardly be considered precise. But they suggest some sort of floor to the level of assistance that may be provided. Ideally, that might be one of the possible measures of poverty that could be derived but whether that would be affordable, let alone desirable, is doubtful. An expansion of the LPP at an effective replacement rate of 25% of average earnings would have a noticeable budgetary impact and yet still provide support at a relatively low level compared to the minimum consumption basket. At present, the priority would seem to be an extension in the coverage of the scheme rather than an improvement in benefits to those already receiving assistance. As we noted previously, the basic LPP is biased towards areas outside Seoul, while the established public works program is biased towards Seoul. The essential difference here is not the nature of the two schemes but their different sources of financing, national and local government. Local governments do have a role in the delivery of social protection, perhaps even more so than national agencies, but they cannot finance it to any significant extent. An extension of national policies and the financing of national programs would seem to be called for. The relative levels of earnings and benefits also suggest that replacement rates of 50% for UIS are reasonable and affordable provided the 1% contribution rate remains unchanged. But they are probably as high as they should be. At these replacement rates, higher income earners, for example, would be receiving amounts sufficiently in excess of minimum consumption costs; and even lower income earners could be receiving significant, though not necessarily adequate, contributions to household incomes during periods of unemployment. Even if there were an effective replacement rate for LLP of 25% of average earnings, there would therefore be a significant difference between the two, a forceful incentive to take out insurance; in effect an incentive to join the formal sector of the economy. Finally, we should consider the use of categories, or target groups, to define eligibility for assistance. Even when governments establish poverty- related programs, they usually find it expedient to define target groups rather than rely solely on means test to determine the most needy. In fact, avoiding the use of categories entirely is really impossible. The question becomes whether the categories do identify groups more likely than others to be in need. Even when they do, of course, there will still be a leakage of assistance to those who are not in need. The example of the 1999 guidelines for the extended public works program, in which the highly skilled were established as the preferred target group, is a good example of a misplaced target group in this respect. Whether there is a substantial political or social significance to be attached to a target group is, of course, another question. Of the categories that could be of concern youth may be the most important. This has not been the case so far in Korea where the emphasis has 98 Labor Market Reforms in Korea: Policy Options for the Futzlre been on supporting skilled workers first and adult heads of households second; and whether youth unemployment does become a problem in Korea remains to be seen. But there is little doubt from international experience that the incentives between work and education and the relative roles of the individual and parents are of great importance. Unless this experience is to be entirely dismissed, youth issues will grow in importance in the coming years. We should conclude by observing that the relevance to Korea of any international experience has to be considered very closely indeed. Develop- ments in Korea are not moving in the same direction as much of the OECD. This not because it is pursuing different objectives but because it is coming at them froin a different starting point. Fortunately for Korea, its starting point in policy terms is better placed. 6Assessing the Coverage and Effectiveness of the Wage Claim Guarantee System in Korea Tai-gi Kim'' Kang-Shik Choi" Joyup Ahn"' Introduction The financial crisis that emerged in late 1997 directly led to the rapid deterioration of the labor market in Korea. During the ensuing period of turmoil, a substantial number of workers were laid off without receiving wage arrears and retirement allowances, as their employers suddenly found themselves bankrupt and unable to pay. The need for a social mechanism that protects the livelihood of laid-off workers and their families has, in fact, been long-standing in Korea. Prior to 1998, employees were not able to receive wage arrears and retirement allowances in cases where companies were found insolvent and financially incapable of paying their debts, even though there is a procedure to sell certain assets possessed by bankrupt companies through various channels. For this reason, on July 1, 1998, the Korean government introduced the wage claim guarantee system and set up the Public Compensation Fund for insuring payment of retirement allowances and wage arrears. The major objective of the wage claim guarantee system is to provide basic livelihood stability for workers laid off from bankrupt companies. Under the system, workers laid off without pay can claim wage arrears and retirement allowances within a fixed limit from the government, which acts on behalf of financially insolvent employers. The purpose of this study is to assess the effectiveness of the wage claim guarantee system and to propose directions for further improvements by examining any implementation problems. This paper is organized as follows. 'DankookUniversity,Korea I*Myungji University, Korea. "'KoreaLaborInstitute. 100 Labor Market Reforms in Korea: Policy Optionsfor the Future The next section describes the background of the introduction of the wage claim guarantee system by briefly reviewing the preferential reimbursement for wage claims in the Labor Standards Act as well as its content, nature, and relationship with the employment insurance system. Then, the paper evaluates the wage claim guarantee system by analyzing the data on employees who have received wage arrears and retirement allowances through the system. The final section summarizes the results and proposes policy directions. The Wage Claim Guarantee System in Korea Background of the Wage Claim Guarantee Fund There are various reasons for the preferential treatment of workers' claims when an enterprise becomes insolvent.' First, workers' claims are likely to be the only source of family income for laid-off workers. Second, workers' claims can often lack an institutional shield because workers are in a less favorable position than the other external creditors. And finally, there is no possibility of shared liability for workers' claims since the employer is the only debtor for these claims. The preferential or privileged treatment scheme ranks wage claims and other claims ahead of the claims of certain other creditors in the insolvency proceedings according to general preference, special preference, and super- preference. The wage guarantee institution scheme, which seeks to offset any shortcomings of the preferential treatment scheme, guarantees workers' claims by operating a guarantee institution according to the methods applied in social insurance schemes. In Korea, workers' claims generally had been protected by a Labor Standards Act provision, which stipulatesthat "wages, retirement allowances, accident compensation and other claims arising from employment shall be paid in preference to other claims as to the total property of an employer except for pledges, mortgages, taxes, and public levies" (Presidential Emergency Measure No. 3, Article 19; on January 14, 1974). On December 24, 1974, the Labor Standards Act was revised with the insertion of Article 30-2, stipulating that "workers with wage claims and other labor-related claims have prior claim to employers' assets. However, these claims do not precede the right of pledges, mortgages, taxes, public levies and other equivalent claims." The special preference of workers' claims was amended further on four later occasions (Dec. 31, 1980; Nov. 28, 1987; Mar. 29, ' Edward Yemin and Arturo S. Bronstein (eds.), The Protection of Workers' Claims in the Event of the Employer's Insolvency, Labour-Management Relations Series, International Labour Office, Geneva, 1991. 1989; and Dec. 24, 1997). Finally, the workers' claims that fall under the special preference category are wages for the final three months of employment, retirement allowances for the final three years of service, and accident compensation. These claims take precedence over any obligations, taxes, public levies and other claims secured by pledges and mortgages. As a measure to cope with the economic crisis that erupted at the end of 1997, the Tripartite Com~nissionwas set up on January 15, 1998. The Commission adopted the "Tripartite Joint Statement on Fair Burden Sharing in the Process of Overcoming the Economic Crisis" on January 20, 1998. It also announced "The Social Agreement for Overcoming the Economic Crisis" on February 6, in which the wage claim guarantee system was conceived. The process for legally institutionalizing the wage claim guarantee system transpired as follows. Enactment of the Wage Claim Guarantee Act (Law No. 5,513) occurred on February 20, 1998, and the Enforcement Decree of the Act (Presidential Decree No. 15,804) was promulgated on May 26 of the same year. The Enforcement Regulations of the Act (Ministry of Labor Regulation No. 131) was enacted on June 15, and the wage claim guarantee system entered force on July 1, 1998. Characteristics and Contentsof the Wage ClaimGuarantee System2 The wage claim guarantee system, based on the Wage Claim Guarantee Act, is a subrogated payment system. In the case of bankruptcy of a business or insolvency of an employer, a worker's deferred wages and retirement allowance are reimbursed up to a fixed ceiling amount by the Wage Claim Guarantee Fund in lieu of payment by the employer in order "to contribute to the stabilization of workers' livelihood by seeking measures which guarantee the wage payment for the workers who retired with their wages unpaid because the company is not able to continue its business or its management is unstable due to fluctuations of the economy and changes in industrial structure, etc. (Article I of the Act)." There are two types of insolvency subject to subrogated payments (Article 6, Paragraph 1 of the Act): judicial insolvency as per a court decision according to the pertinent law and actual insolvency due to closure following deterioration of business activities. Judicial insolvency providing legitimate cause for subrogated payments refers to: (1) an adjudicated bankruptcy according to the Bankruptcy Act; (2) a decision on the start of composition according to the Composition Act; and (3) a decision on the start of reorganization according to the Company Reorganization Act. As such, the reasons for subrogated payments cease to exist when the procedures are completed, i.e., (1) when a judgment of composition is finalized (Article 58 of the Composition Act) and (2) when the procedure of This section refers to Lee, Jeongjo, The Introduction of the Wage Claim Glrorantee System, Jooangkyungje, Seoul, 1999 (in Korean). 102 Labor Market Reforms in Korea: Policy Optionsfor the Future reorganization is completed (Article 271 of the Company Reorganization Act). Actual insolvency, which is usually not judicial insolvency but rather virtual insolvency depending on the state of insolvency in practice, is confirmed by the head of a local labor administration. Its real conditions (Article 8 of the Enforcement Decree of the Act) are that there is no ongoing business activity, there is no prospect for the resumption of business, and the employer has little or no financial capability of paying wages or retirement allowances. The government takes full charge of the wage claim guarantee system, but administrative work is entrusted to the local labor administration and the Korea Labor Welfare Corporation (Article 24 of the Enforcement Decree of the Wage Claim Guarantee Act). The local labor administration is responsible for issuing the certification of bankruptcy, examining whether those companies applying for bankruptcy status meet the requirements for subrogated payments, receiving the necessary documents submitted by the company owners, and supervising these companies. The Korea Labor Welfare Corporation (1) collects contributions from employers, (2) provides subrogated payments for workers who satisfy the conditions for deferred payments, (3) exercises the right to claim wages and retirement allowances from employers within the limit of subrogated payments already made to the workers, and (4) reclaims any payments made to those who submit false applications by ordering repayment. In general, the payment of debts by a third party as prescribed by Article 469 of the Civil Code depends on the types and the nature of the debts and, when the debtor expresses an objection, the third party may not pay the debt on behalf of the debtor. The government's subrogated payment to workers is based on a political measure to protect the livelihood of workers. Article 6 of the Wage Claim Guarantee Act authorizes the government to pay a fixed amount of wage claims regardless of the opinion of the employer, overriding Article 469 of the Civil Code. When the government pays deferred payments to workers, the employer is released from hislher duty to pay workers and instead the government obtains the right of indemnity against the employer. Thus, the government then has claims against the employer within the amount of subrogated payments made to workers. The Wage Claim Guarantee Act applies to the same category of businesses or business establishments as prescribed in Article 5 of the Industrial Accident Compensation Insurance Act (Article 3 of the Wage Claim Guarantee Act). Article 2 of the Wage Claim Guarantee Act defines workers as prescribed in Article 14 of the Labor Standards Act, which stipulates, "... a person engaged in whatever occupation offering work to a business or workplace for the purpose of earning wages." The determination and administrative interpretation of the definition of a worker in the Labor Standards Act depends on whether the labor is actually provided to employers at businesses or workplaces in exchange for wages, and whether ASSESSING THE COVERAGE AND EFFECTIVENESS. .. 103 the work is conducted under a subordinate relationship, regardless of the form of the employment contract. In other words, determination does not look to whether it was an employment contract or a work contract under the Civil Code. Conditions for subrogated payment are that a worker should have retired or should retire from the pertinent company within one month before or within one year after the bankruptcy of the company and that the petition for the confirmation of insolvency was filed (Article 7 of the Enforcement Decree of the Act). The range of workers' claims protected under the guarantee system has been a matter of legislative policymaking. The amount of wage claims guaranteed under the Act is identical to the amount guaranteed in the Labor Standards Act, i.e., "the amount of wages for the last three months and the retirement allowance for the last three years7'(Article37, Paragraph 2). As in most countries, Article 6, Paragraph 1 of the Enforcement Decree defines the ceilings of the subrogated payment both for one month's wage and for one year's retirement allowance according to age group^.^ Article 15 of the Act allows the Minister of Labor to establish the Wage Claim Guarantee Fund to fill up the subrogated payment as prescribed in Article 6. Article 16, Paragraph 1 of the Act defines the financing of the Wage Claim Guarantee Fund.4Also, Article 4 of the Act defines the defrayal from the State Treasury as follows: "that the State shall, within the budget of every fiscal year, provide some parts of the expenses, which are required in enforcing the activities for guaranteeing the wage claims under the Act, from general accounting." According to Article 8 of the Act, the Minister of Labor collects defrayal from the employer equivalent to two-thousandths of the total wage in order to compensate for the expenses required in paying the deferred wages and retirement allowances as prescribed in Article 6. The Act enforces the defrayal by the employer of the concerned business without any compensation to the employer, which makes it different from insurance premiums. In other words, the defrayal is collected froin employers at a fixed rate regardless of the possibility of insolvency, wlhile the insurance premium is calculated at varying rates depending on the extent of risk- bearing. 800,000 Korean won for retired workers less than 30-years of age, 1 million won for retired workers 30-years of age or older but less than 45, and 1.2 million won for retired workers 45- years of age or older. The money redeemed by employers as provided by Article 7 (acquisition of the request right to deferred wages); charges paid by employers as provided by Article 8 (defrayal by employer); loans from special financial accounting, financial organizations or other funds on the security of the Wage Claim Guarantee Fund provided by Paragraph 2 of Article 16; revenue accrued from operation of the fund; and other revenues. 104 Labor Market Reforms in Korea: Policy Optionsfor the Future Article 9 of this Act states that the Minister of Labor may lighten the charges, as stipulated in Article 8, for an employer under retirement insurance, etc., under Article 34, Paragraph 4 of the Labor Standards Act. The purpose is to promote equality with regards to the defrayal by the employer considering the possible and stable payment of the retirement allowance through the financial system when the employer participates in retirement insurance, etc. It is a measure to emphasize the role of the government in promoting the operation and development of the wage claim guarantee system. The employers who implemented a policy of retirement insurance, etc. for all employees as beneficiaries as of the end of the previous year, as prescribed in the Labor Standards Act, qualify for a reduction in defrayal. This provision aims to provide balanced consideration for employers, for in this case external financial institutions pay retirement allowances to their employees on their behalf. Other Systems Related to the Wage Claim Guarantee System The wage claim guarantee system mainly focuses on the social protection of workers' livelihood. The retirement allowance system, the employment insurance system, and the national pension fund system have a similar purpose. The retirement allowance system maintains the stability of retired workers, the employment insurance system guarantees the security of the unemployed, and the national pension fund system secures workers' livelihood during the post-retirement period. The retirement allowance and employment insurance systems are of direct relevance to the wage claim guarantee system. Retirement allowances are monetary payments made to a worker by an employer when a practical labor relationship is terminated. The payments cover not only deferred wages but also social security and rewards for meritorious service. The retirement allowance system was introduced through a provision in the Labor Standards Act of 1953 in order to guarantee the income level of retiring workers. Article 34 of the Labor Standards Act stipulates that "an employer shall establish a retirement allowance system whereby an average wage of more than 30 days shall be paid to a retired worker as a retirement allowance for each year of the number of consecutive years employed; however, if the worker was employed for less than one year, this shall not apply." The retirement allowance system in Korea has unique features that are not prominent in other countries. It has the characteristics not only of a private welfare system but also of a public welfare system. The Korean system seeks to meet the need levels of household expenditures at each stage of the lifecycle beyond a simple guarantee of income security during the post- retirement period. Nonetheless, it bears an aspect of inequity in the sense that some workers benefit from the progressive accumulation of retirement allowances while many others are left out of the system. There have been arguments that the legal retirement allowance system should be changed into a voluntary system when the national pension fund system and the employment insurance system were introduced. In practice, the retirement allowance system and the national pension fund system are closely related to each other. Revisions will inevitably need to be made to the current retirement allowance system as some deposits of the retirement allowance are transferred to the national pension fund system. Another role of the retirement allowance system to ensure livelihood of the unemployed using a mid-term settlement scheme of retirement allowance was undertaken by the employment insurance system. The first government-level discussion on the introduction of employment insurance took place when the unemployment rate began to rise in late 1979 and 1980 due to political instability and economic recession. In the process of preparing for the 5IhFive-Year Economic and Social Development Plan in the early 1980s, the unemployment insurance system underwent the beginnings of an official examination, but this was put off indefinitely because of the possibility of negative side effects. The introduction of employment insurance went through a full discussion during the 7IhFive-Year Economic and Social Development Plan in the early 1990s. The Employment Insurance Act was enacted in I993 and went into force on July 1, 1995. It was amended in December 1998 to increase the minimum amount and the minimum benefit period for unemployment benefits. Further revision took place in September 1999 in order to extend the coverage of employment insurance. At the time of the first revision, it applied to business establishments with 30 employees or more, but it began to cover all workplaces from October 1998. The employment insurance system in Korea consists of the Employment Stabilization Program, Job Skills Development Program and Unemployment Benefits. It is a preventive and active policy scheme for the stabilization and restructuring of the labor market and the enhancement of vocational training. Recent economic crisis and high unemployment rates have proven its importance as an active labor market policy. Despite its role as an ALMP me~asure,the employment insurance fund is raised and maintained with contributions from employers and employees. The coverage of the employment insurance system in Korea has expanded rapidly. In particular, coverage was extended to cope with the ever-rising mass of unemployment after the economic recession and nationwide restruc- turing. The number of work places covered was increased from 43 in 1995, the first year it was introduced, to 664 in October 1998. Also, the number of workers covered increased from 4,280,000 to 8,342,000. 89 % of companies were covered by employment insurance as of September 30, 1999. Of the entire qualifying workforce, 71 % is currently covered by employment insurance. Despite the rapid expansion of the employment insurance system, the number of beneficiaries of unemployment benefits stood at 147,000 as of June 1999, a mere 12.8 % out of 1,356,000 unemployed. The re-employment 106 Labor Market Reforms in Korea: Policy Options for the Future rate for the recipients of unemployment benefits is lower than that of non- recipients. Empirical Analysis Basic Statistics5 Between the time the Wage Claim Guarantee Act was promulgated (July 1, 1998) and December 1999, a total of 54.9 billion won was paid out to 17,727 workers laid off from 460 bankrupt companie~.~On average, laid- off workers received 3.2 million won each, whereas the average work place received 119.4 million won (Table 6.1). For the year 1999, 12,588 workers displaced from 360 insolvent companies received wage arrears and retirement allowances totaling 38.8 billion won. On average, this translated into 107.8 million won per work place and 3.1 million won per worker. Table 6.1: Amount and accumulated records of subrogated payments in Korea, 1998 and 1999 Number of Number of Amount Period beneficiaries workplaces surveyed (wonin millions) (~ersons) 1998 3rdqtr 15 (3%) 378 (2%) 1,148 (2%) 41hqtr Subtotal 1999 ls'qtr 2" qtr 3rdqtr 4thqtr Subtotal Total Source: Korea LaborWelfare Corporation. - - The population used in the analysis includes all workers who benefited from the wage claim guarantee system during the period between the time of implemel~tationand December 1999. Korea Labor Welfare Corporation, "Performance Report of the Wage Claim Guarantee Act, 1999" (Press Release), January 2000. Subrogated payment is a legal term that refers to the third-party payment of debts on the condition that the debtor will make reimbursement at a hture date. In the case of the wage claim guarantee system, the government pays a portion of deferred payments (a maximum of 3 months wages and 3 years retirement allowances) to workers in advance, which later will be collected from the employer. ASSESSING THE COVERAGE AND EFFECTIVENESS. .. 107 According to this data on about 14,031 workers who received deferred wages through the system (the Korea Labor Welfare Corporation)', the total amount of deferred payments was 53.6 billion won, or 3.8 Inillion won per worker (Table 6.2).9 Deferred payments are largely divided into two categories: for deferred wages, a total of 22.2 billion won (41.3 % of the total amount paid) were distributed to 11,787 workers, or 1.9 million won per worker; for deferred retirement allowances, 10,553 workers received a total of 31.5 billion won (3.0 million won per worker). Table 6.2: Subrogated payments made by type (wages and retirement allowances) Number of workers Total amount Average amount (persons) (won in millions) (wonin thousands) Subrogated payment 14,031 53,614 (100%) 3,821 Wages 11,787 22,156 (41%) 1,880 Retirement allowances 10,553 31,457 (59%) 2.981 Source: Calculation of the data from Korea Labor Welfare Corporation The workers that benefited from the system break down to 10,033 male workers (71.5 % of the total) and 3,998 female workers (Table 6.3). In terms of amount, male workers received 4.4 million won on average, accounting for 81.9 % of the total payment, while each female worker received 2.4 million won. This difference is mostly due to the fact that male workers on average served longer years and thus received higher wages than their female counterparts under the same conditions, and also, the average age of the former is higher than that of the latter. Table 6.3: Subrogatedpayments by gender Number of workers Total amount Average amount Gender (persons) (won in millions) (won in thousands) Male 10,033 43,919 4,378 Female 3,998 9,694 2,425 Total 14,031 53,613 3,821 - Source: Calculation of the data from Korea Labor Welfare Corporation. ' The analysis results in this paper based on the personal information data differ from the data published by the Korea Labor Welfare Corporation, mostly due to differences in reference periods for data collection. Other reasons for data discrepancies may be that duplicate observations were eliminated and that the provided data did not cover all the workers that benefited from the system. However, these discrepancies are not considered critical in the analysis of workers' characteristics. ' his is a significant amount in comparison to 2,278,000 won, the average amount of unemployment benefits paid to the workers (approx. 306,000 workers) whose unemployment benefits ended in 1998. 108 Labor Market Reforms in Korea: Policy Optionsfor the Future The 24 and under age bracket accounted for 11.0 % of the total beneficiaries while those between 25 and 29-years of age accounted for 17.8 %, proportionately higher than the other age groups. (Table 6.4) The average amount of deferred payment received increases with each age bracket up to 55, after which it drops. This is due to the fact that wages (and retirement allowances) are in proportion to age. The drop in the deferred payment received by the over 55 age group also agrees with the typical age- earnings profile. Workers up to the age of 55 received wage arrears and retirement allowances in a proportion reasonable to their age. Table 6.4: Subrogated payments by age group Age Nuniber of workers Total amount Average amount group (persons) (wonin millions) (wonin thousands) 15-19 180 227 20-24 1,367 3,024 25-29 2,500 6,845 30-54 8,700 39,499 Over 55 1,284 4,019 Total 14,031 53,614 Source: Calculationof the data from Korea Labor Welfare Corporation By type of insolvency, 6,317 workers (or 45 % of the total) were able to claim their deferred pay on the basis of judicial insolvency, while the remaining 7,714 received it on the basis of the confirmation of insolvency from relevant authorities (Table 6.5). In terms of the amount of deferred payment, judicial insolvency cases accounted for 51.6 % of the total. Therefore, the average amount of deferred payment (4,379,000 won) was higher than that of other cases (3,361,000 won). The main reason for this may be that the latter case applied mostly to small- and medium-sized companies, where wages are relatively low compared to large corporations. Table 6.5: Subrogated payments by type of insolvency Type of insolvency Number of workers Total amount Average amount (person) (wonin millions) (wonin thousands) Judicial insolvency 6,317 27,663 4,379 (45%) (51.6%) Confirmationof insolvency 7,714 25,951 3,364 (55%) (48.4%) Source: Calculation of thedata from Korea Labor Welfare Corporation. In cases where laid-off workers received deferred payments based on the confirmation of insolvency and others, their personal data also provide the type and size of their employer companies. As this confirmation system was designed for small- and medium-sized companies, thus inapplicable to large ASSESSING T HE COVERAGE A N D EFFECTIVENESS ... 109 corporations, most companies studied had 300 employees or less. According to Table 6.6, the largest proportion of companies (40.1 %) had 30-99 employees (accounting for 39.7 % in terms of the amount of deferred payments), followed by those with 100-299 employees (21.6 %). Companies with 4 employees or less accounted for 9.0 %. The average amount of deferred payment received does not show any evident pattern depending on company size. The highest average amount was 3.5 million won for those with 10-29 employees. p Table 6.6: Subrogated payments by firm size (confirmation of insolvency) - - Number of workers Total amount Average amount Firm size (oersons) (won in thousands) 4 or less 694 (9.0%) 2,154 (8.3%) 3,105 5-9 286 (3.7%) 996 (3.8%) 3,481 10-29 1,963(25.4%) 6,947(26.8%) 3.539 30-99 3,090(40.1%) 10,309(39.7%) 3,336 100-299 1,665(21.6%) 5,504(21.2%) 3,305 300-499 16 (0.2%) 41 (0.2%) 2,551 Note: Numbers in parenthesis are % of the total. Source: Calculation of the data from Korea Labor Welfare Corporation. Descriptive Analysis of the Survey In order to further study the effectiveness of the wage claim guarantee system, a survey was given to 423 workers from 47 business companies who had benefited from the system.1° The survey includes 288 male workers (68.1 %) and 135 females, which is quite comparable to,that of the population who received deferred payments through the wage claim guarantee system (71.5 % male). By age, those between 15 and 19-years old accounted for 3.3 % of the survey group; those between 20 and 24-years old, 8.7 %; and those between 25 and 55-years old, 85.1 %. This distribution breakdown is also quite similar to the age distribution of the entire group of workers who benefited from the wage claim guarantee system. Respondents to this survey consist of 83 workers (19.6 %) with elementary or middle school education, 162 (38.3 %) with high school education, and 165 (39 %) l oThe survey took place between Feb. 14 through Feb. 25, 2000. The survey questionnaire for workers includes: information on workers, their families and employers; the size of deferred wages and retirement allowances: the actual amount of wage arrears and retirement allowances received through the system; difficulties experienced in the course of claiming deferred payments through the system; the general assessment of the wage claim guarantee system; and others. The survey questionnaire for employers includes: the details of the company; main causes for insolvency and the type of insolvency; the amount of deferred wages and retirement allowances; efforts made to sustain business; whether a retirement allowance fund was set up; and the assessment of the current wage claim guarantee system. 110 Labor Market Reforms in Korea: Policy Optionsjor the Future with junior college or higher education." The proportion of those with junior college or higher education is greater than that of the educational distribution of the economically active population of Korea in 1999. This indicates that the majority of the beneficiaries of the wage claim guarantee system were not workers with lower-level education. By occupation, executives and professionals accounted for 11.8 %; technicians and semi-professionals, 26.7 %; office and service workers, 27.9 %; skilled workers, 10.6 %; and machine operators and simple manual workers, 16.8 %. This indicates that a larger proportion of workers with higher education, such as executives, professionals and technicians, have benefited from the wage claim guarantee system compared to the occupation distribution of the entire economically active population. By type of employment status, 373 (88.2 %) were regular workers while temporary and daily workers accounted for a very small number. Since more than half of paid workers in Korea are temporary or daily workers, these figures also represent a much higher proportion of regular workers.12 The sample in the survey can be broken down by job tenure. Those with tenure of less than a year accounted for 25.1 %; those with between 1 year and 3 years of tenure, 32.2 %; and those with tenure of between 3 and 5 years, 22.0 %. Approximately 80 % of the workers worked for less than 5 years. However, those who worked 10 years or longer accounted for a considerable 6.6 %. In terms of work hours, those who worked 18 hours or less per week accounted for 12.8 %. The majority (86.1 %) of the survey group worked 36 hours or more per week. This is another indicator that the majority of the survey group worked full time, as shown in the case of previous work types, and hence it appears that the number of temporary workers is very small. As one of the main objectives of the wage claim guarantee system is to protect the livelihood of low-wage workers, it is necessary to look at information about their household as well. First, 70.2 % of the survey group were married and living with a spouse, while 29.8 % were either married but not living with a spouse, or single. Also, 18.0 % of the workers in the survey group had four persons or more to support; 31.0 %, three persons; and 18.7 %, two persons. These figures all illustrate the need to take a closer look at workers' households when studying how the wage claim guarantee system affects worker livelihood. The survey did investigate whether the worker was the head of household or not; and whether any other member of the household was engaged in economic activities and if so, how much they were paid. For the first - - " The remaining3.1%answered"Others"tothisquestion or didnot respond. '* One reason for this might be found in the common practice whereby an employment relationship is terminated in the case of temporary or daily work if the wage is deferred continuously. ASSESSING THE COVERAGE A N D EFFECTIVENESS ... 111 question, 60.0 % of the workers in the survey group were the head of their household while the remaining 40.0 % were not. (Table 6.7) Table 6.7: Characteristicdata of workers and their households - -- - Total Head of Non-head of household household Total 423(100.0%) 254 (60.0%) 169 (40.0%) Individual income:over 1.5 millionwon 42 (9.9%) 36 (8.5%) 6 (1.4%) lncome of other family members Over 1.5 million won 4 (0.9%) 1 (0.2%) 3 (0.7%) Less than 1.5 million 7 (1.7%) 5 (1.2%) 2 (0.5%) No income 31 (7.3%) 30 (7.1%) 1 (0.2%) Individual income:lessthan1.5 million 156 (36.9%) 95 (22.5%) 61 (14.4%) lncome of other family members Over 1.5 millionwon 14 (3.3%) 3 (0.7%) 11 (2.6%) Less than 1.5million 47 (11.1%) 23 (5.4%) 24 (5.7%) No income 95 (22.5%) 69 (16.3%) 26 (6.1%) Individual income: None 225 (53.2%) 123 (29.1%) 102 (24.1%) lncome of other family members Over 1.5 millionwon 46 (10.9%) 13 (3.1%) 33 (7.8%) Less than 1.5 million 70 (16.5%) 31 (7.3%) 39 (9.2%) No income 109 (25.8%) 79 (18.7%) 30 (7.1%) - In the following part of this section, the surveyed workers will be divided into several categories depending on the employment of the workers and that of their household members as well as the amount of monthly income. The distribution of these categories will also be examined. First, those who were the head of household and were currently employed with a new empIoyer accounted for 31.0 % of the survey group. The remainder of the survey group were either still unemployed or economically inactive. However, out of this 31.1 %, a mere 8.5 % were employed with monthly pay exceeding 1.5 million won. The remaining 22.5 % received less than 1.5 million won per month." This is not regarded as a sufficient amount to sustain the normal living of a household. Although they received less than 1.5 million won, 6.1 % of the survey group (among the 22.5 % with monthly pay of less than 1.5 million won) had other income sources, since other members of their household were employed. However, the remaining 16.3 %, out of the 22.5 % had no other l3 The caveat against the income report is that it is highly likely that the amount of wages reported through the household survey tends to be underreported. 112 Labor Market Reforms in Korea: Policy Optionsfor the Futzu-e income source. In this case, their income was the sole income source for the entire household and clearly insufficient to maintain the livelihood of the household. Second, those who were heads of household and unemployed accounted for 29.1 % of the survey group. This breaks down to unemployed heads of household with no employed family members (18.7 %) and unemployed heads of household with other employed family members (10.4 96). However, even in the latter case, 7.3 % of the survey group fell into the category where the collective income of other family members was short of 1.5 million won. Third, those who were not heads of household and who were currently employed accounted for 15.8 % of the survey group. However, 6.2 % of the survey group were those with monthly wages of less than 1.5 million won and with no other employed family members. So these also constitute cases where there is insufficient income to maintain the livelihood of the household. Fourth, those who were not heads of household and unemployed accounted for 24.1 % ofthe survey group. Specifically, those workers whose family members were also unemployed accounted for 7.1 %. Even when other family members were employed, 9.2 % of the survey group lived on a monthly income of less than 1.5 million won. In conclusion, households with no income source accounted for 25.8 %, and households with one employed family member (either the interviewed worker or another member), but with a monthly income of less than 1.5 million won, accounted for 39.0 %. These households, which collectively amount to 64.8 % of the survey group, clearly fall below the subsistence level. A significant proportion of the beneficiaries of the wage claim guarantee system are professionals or skilled workers with relatively higher education in comparison to the distribution of the entire economically active population. Flowever, the income of these households was very low (at the time of survey), and thus in urgent need of other measures or other sources of income. Deferred payment refers to the total sum of wage arrears and retirement allowances that has not been provided on time. For 21.8 % of the survey group, the total amount in arrears was less than I million won; for 24.2 %, it was less than 2 million won (but more than 1 million won); and for 15.9 %, it was less than 3 million won. However, those who had not received deferred wages exceeding 5 million won accounted for 15.3 % of the survey group, indicating that a substantial number of workers were laid off without receiving a sizeable amount of wage arrears. In terms of the number of months during which wages are deferred, the highest proportion, 33.1 % of the survey group, had three months of wages in arrears; followed by 24.5 % with one month of wages in arrears; and 10.8 % with two months in arrears. Also, those with four months of wages in arrears accounted for 31.7 %. Since the maximum amount of deferred wages that is provided under the wage claim guarantee system is 3 months of wages, these people cannot expect to receive the full amount. In terms of the amount of deferred retirement allowances, those with less than 1 million won accounted for 14.5 % of the survey group; less than 2 million won (but more than 1 miIlion won), 20.2 %; and less than 3 million won, 22.2 %. Furthermore, 12.5 % of the surveyed workers had deferred retirement allowances exceeding 5 million won. The total amount of deferred payments, which is the sum of deferred wages and deferred retirement allowances, varied between less than 1 million won (13.9 %), less than 2 million won (13.9 %), and less than 3 million won (13.2 %). In particular, surveyed workers with a total deferred payment exceeding 5 million won amounted to 38.3 % of the survey group. The average per capita amount of the combined deferred payments was 4.4 million won (based on a sample group of 423 workers). The average amount of deferred wages was 2.6 million (a sample group of 372), while the average amount of deferred retirement allowances was 3.0 million (a sample group of 297). The number of workers who had not received retirement allowances was smaller than that of those who had not received wage arrears, but the average amount of deferred wages was higher than that of deferred retirement allowances. (Table 6.8) Table 6.8: Deferred payment per person Number Of Total amount Average amount workers (persons) (won) Total Deferred wages Deferred retirement allowances 297 880,575 2,965 Moreover, the Wage Claim Guarantee Act sets the ceiling of the monthly amount of deferred payments depending on the age of the beneficiaries. This is in addition to limiting the amount of deferred wages (up to 3 months) and deferred retirement allowances (up to 3 year^).'^ Because of these limiting conditions, 200 (53.8 %) out of 372 workers with deferred wages did not receive the full amount. On average, 1,735,000 won remained unpaid. The workers who have benefited from the wage claim guarantee system, yet did not receive the full amount, can be divided into three groups. The average unpaid amount is 1.3 million won for workers under the age of 30 (25.0 %), 1.8 million won for workers aged between 30 and 45 (54.0 %), and 2.0 million won for workers aged 45 and older (21.0 %). In the case of deferred retirement allowances, 188 (63.3 %) of 297 workers who qualified to receive retirement allowances under the wage claim guarantee system failed to receive 1.7 million won, on average, due to the ceiling amount set by the Act. I'Refer to the description of the wage claim guarantee system in the previous section 114 Labor Market Reforms in Korea: Policy Optionsfor the Future The average amount unpaid is 0.8 million won for workers under the age of 30 (25.5 %), 2.1 million won for workers aged between 30 and 45 (32.7 %), and 2.0 million won for workers aged 45 and older (14.5 %). In order to assess how the wage claim guarantee system has benefited workers, those surveyed were asked how useful the wage claim guarantee system was. Overall, an absolute majority or 89.4 % of the sample survey group considered the wage claim guarantee system to be helpful.-~owevei, 5 % answered that it made no difference to them, and 3.3 % responded that it was not helpful. This result comes as no surprise given that the survey group did receive deferred payments through the system. To make a fair judgement, a comparison group made up of those who did not receive deferred payments is needed. The comparison between the survey group (target group) and the comparison group could reveal the true usefulness of the wage claim guarantee system. However, it takes enormous time and expense to construct a comparison group in a household survey. Therefore, in lieu of creating a comparison group, the surveyed workers were asked how the subrogated payment they had received was spent. Out of the respondents, 84.4 % answered that the money was spent on maintaining the livelihood of their famiIy after their retirement. (See Figure 3.22.) 5 % of them spent the payment on the startup of their own business, and 3.3 % spent it for vocational training. There were few respondents who placed the money into savings or other financial investments. This result indirectly suggests that the wage claim guarantee system has been of great help to laid- off workers for maintaining their livelihood, which is one of the objectives of the system. In order to investigate problems arising from the implementation of the wage claim guarantee system, the surveyed workers were asked how long it took between the time they left their job and the time they received subrogated payments. For deferred wages (deferred retirement allowances), it took less than 1 month for 6.2 % (8.4 %); longer than 1 month but less than 2 months for 17.5 % (17.2 %); and less than 3 months for 33.6 % (35.7 %). Approximately 87.9 % (84.2 %) waited for less than 6 months while 4.8 % (6.4 %) waited for more than one year. In order to receive subrogated payments based on the confirmation of insolvency, one must obtain-confirmation from the local labor authorities. To the question of "whether it is difficult to meet requirements for the confirmation of insolvency," 11.1 % replied "very difficult" and 32.6 % responded "difficult." In total, 43.7 % of the sample group wanted the requirements to be moderated. In contrast, 8.5 % replied "not so difficult" and 1.7 % answered "not difficult at all." To the question of when they became aware of the wage claim guarantee system, most workers answered around or after their retirement. The survey results showed that 20.6 % of the surveyed workers were aware of this system before they left the company, 22.9 % at the time of leaving, and 46.6 % after leaving. Moreover, 46.6 % of the surveyed workers learned of ASSESSING THE COVERAGE AND EFFECTIVENESS ... 115 the existence of the system through colleagues. 26.2 % heard about the system through "government-related organizations, local labor offices or labor inspectors," and 11.3 % through mass media. In contrast, a mere 2.4 % were informed of the system by their employer company. Therefore, it is necessary to educate employers about the system, and furthermore, to make it compulsory for employers to inform their employees of the system in case they become unable to pay wages in time. The survey for employers showed a rather poor respolnse rate as it was diff~cultto get in contact with them. Thus, one must bear in mind that the analysis of the survey results in the following section is somewhat lacking with regard to rigorous statistical significance. The surveyed employers were asked "whether they had applied for the subsidy for sustaining employment provided under the Employment Insurance Act or applied for other corporate loans." This question sought to investigate how much of an effort they had made to save themselves before going insolvent. Only 10.9 % of the surveyed employers (5 companies) gave an affirmative answer. The remaining 89.1 % answered no, indicating that few e~nployersmake use of the programs designed to help them. Out of the five companies which applied for any form of financial support, two applied for corporate loans, two applied for loans for living expenses, and one applied for the subsidy for sustaining employment. In Korea, companies are obliged to set aside part of their total employee retirement allowance as a contingency in case the employees leave the company all at once. Interestingly, most companies did not reply to the cluestion of whether they had a sufficient find for employee retirement allowances. Only 11 companies responded to the question about the reserve retirement allowance fund. Of these, 63.6 % chose an internal reserve system and the others, an external reserve system. To the question as to whether the retirement allowance fund was ever used as operating money for the company. 3 replied "yes" and 10 replied "no." Four companies used the fund as collateral for bank loans. Of the four, two had paid back the loans in full. These results suggest that business companies are quite reluctant to reserve a retirement allowance fund, and even when they do, they often use the fund as operating capital. The wage claim guarantee fund is financed with the employers' defrayal, the reimbursement from employers, and loans and profits from the fund operation. The amount of the employers' defrayal is determined by multiplying the total wage with a fixed rate. Currently, this rate is fixed regardless of business type and size. As shown in the previous sections, the amount and extent of deferred wages vary across industries. Those who answered that "the current fixed rate is fine" accounted for 21.3 % of the surveyed employers; those who answered that "the rate should be re- adjusted for different business sectors" amounted to 27.7 %; and furthermore, those who answered that "the rate should be readjusted not only for different business sectors but also according to the credit rating of the 116 Labor Market Reforms in Korea: Policy Optionsfor the Fzltzlre company or employer" accounted for 10.6 %. Taken together, almost 60 % of the employers expressed a negative opinion on the fixed rate for contributions. Most of the surveyed employers, that is, 89.4 % of the survey group, considered the wage claim guarantee system to be helpful. Only 4.3 % (2 companies) replied that it was not very helpful. Thus, the majority of employers responded positively to the wage claim guarantee system. On the other hand, 42.6 % of the surveyed employers considered it unfair that the government exercises the right of indemnity against them, even though they pay a defrayal for the fund. A considerable 29.8 % also answered that the current system is acceptable. Multiple answers were allowed to the question about any future improvements to the system. The most frequent answer was the abolition of criminal punishment (33.3 %), followed by the long-term or installment repayment (24.4 %) of the subrogated payment when the government exercises the right of indemnity. Those who proposed an elimination of the employers' dual obligations, along with their contributions to the fund and repayments of the subrogated payment, accounted for 20.0 %. Lastly, 15.6 % answered that a grace period be allowed for employers to repay the subrogated payment made by the government. Determinants of Deferred Payments and Severity of the System The size of deferred wages and retirement allowances depends on the average wage of the concerned worker and the length of deferment.'' The average wage depends on the level of human capital as well as the nature of work and the characteristics of the business establishment. Although the deferment length is generally determined by the characteristics of the business establishment, its current situation, and the overall economic situation, it is also affected by an employer's deliberate deferment. In other words, employers may intentionally abuse the wage claim guarantee system by deferring payment even when they have sufficient financial means to pay wages. In response to the survey question regarding the presumed presence of such intention, 20.3 % of the surveyed workers answered that "the employers did not pay wages and retirement allowances deliberately despite their financial capability." Therefore, it is important to examine those factors determining the size of deferred payments using a regression analysis with the total amount of deferred wages, i.e., the worker's wage multiplied by the length of deferment, as a dependent variable. The explanatory variables can be divided into two: the worker's wage and other factors. The other factors 'jThe Labor Standards Act in Korea stipulates that employers of business establishments with 5 employees or more should pay retirement allowances equivalent to a minimum of 30 days of pay for each year of service to retiring employees whose tenure exceeds I year. ASSESSING THE COVERAGE AND EFFECTIVENESS ... . 117 include those describing the characteristics of the business establishment, such as industrial classification, company size, location, presence of a labor union, company type, existence of subcontractors, and type of insolvency. Also included is the response of the surveyed employees regarding the moral hazard of employers. Most independent variables that reflect the characteristics of the business establishment influence not only the worker's wage but also the size of the deferred wage. This brings up a consistency issue with estimation in the sense that the wage variable is an endogenous variable and is correlated with the error term in the regression equation. For this reason, the two stage least- square method was used in the estimation. At the first stage, the worker's wage was estimated using only the exogenous variables and then. at the second stage, the estimated wage instead of the actual wage was used as an explanatory variable in estimation of the deferred wage. The typical Mincer type earnings equation was estimated first." The explanatory variables used in the earnings equation include the properties of human capital, such as educational background, gender and age, character- istics of the occupation, and characteristics of the concerned business establishment. The results of the second regression, which estimates the size of the deferred wage using the estimated wage and other variables as explanatory variables, are shown in Table 6.9. One can see that the worker's wage is a significant factor in determining the size of deferred wages (the coefficient of the (estimated) wage variable is 0.487). Another factor that affects the size of deferred wages is the size of the business establishment. The coefficient of the dummy variable representing business establishments with 500 employees or more is not significantly different from zero and, thus, these establishments and those with 4 employees or less appeared to register a larger amount of deferred wages.'' Roughly speaking, medium-sized companies tend not to defer wages in comparison to small co~npanies(4 employees or less). In addition, different industries showed varying trends in deferred wages. In the service sector, the deferment of wages was less observed than in the manufacturing sector. The coefficients of the dummy variable for the service sectors were all negative and significantly different from zero. In compari- son, the coefficients were positive for the construction industry and the agriculture, forestry and fishing industries, but they are statistically insignifi- cant. Thus, the size of deferred wages in the service industry was smaller than in the manufacturing and construction industries. 'The sizes of the deferred wages did not significantly vary across different types of business establishments. '" results of thc first regression analysis (estimation of the earnings equation) are available he iiorn the authors. "The reason why the dummy coefficientfor large establishmcntswith 500 employeesor more is ~nsignificantmay be that these establishments accounted for only 1.9 %of thc sample group. 118 Labor Market Reforms in Korea: Policy Optionsfor the Future Table 6.9: Determinants of the size of deferred wages: ZSLS method - Var~ables Coefficients Standard error Intercept 5.433 (1,000) =* Log wage (estimated) 0.487 (0.162) ** Seoul (=I) 0.483 (0.311) Big cities excluding Seoul (=I) 0.303 (0.300) Establishment size (compared to less than 5 workers) 5-9 workers -0.966 (0.830) 10-49workers -1.267 (0.782) 50-99 workers -1.071 (0.793) 100-299 workers -2.788 (0.835)** 300-500 workers -0.622 (0.919) 500 or more workers Union (=I) Industry (compared to manufacturing) Agriculture 1.317 (1.477) Construction 0.051 (0.329) Electricity, gas, wholesale, retail, restaurant, etc. -1.075 (0.556)* Financial, social and personal sewices -0.833 (0.483)* Type of company (compared to corporation) Non-corporation 0.328 (0.285) Others 0.613 (2.109) Subcontractor (=I) -0.082 (0.279) Judicial insolvency (=I) -0.034 (0.336) Moral hazard (=I) -0.378 (0.297) Cause for Insolvency (compared to transitory financial problem) Chronic financial problem -0.557 (0.313)* Decrease in sales -0.275 (0.299) Insolvency of the contractor -0.443 (0.455) Number of obse~ations 371 R2 (Adjusted RZ) 0.117 (0.0609) F-Statistics 2.09** Note: "*, "and *stand for 0.01. 0.05 and 0.10significancelevels, respectively. ASSESSING T HE COVERAGE AND EFFECTIVENESS ... 119 The amount of deferred wages varied little with the type of insolvency, but it did vary with the cause of insolvency. For insolvency due to a transitory financial strain, the amount of deferred wages was smaller than that due to chronic financial strains. In the former case, the dummy variable has a negative value and was significantly different from zero. When the employer was declared insolvent due to sluggish sales or due to the insolvency of their contractor who placed a contract order, the amount of deferred wages was small, but the coefficients were not significantly different from zero. Table 6.10: Severity for the confirmation of insolvency: logit analysis Standard Variable Coefficient error Constant -0.443 (0.380) Seoul -0.227 (0.163) Firmsize (compared toless than 5) 5-9 -0.501 10-49 -0.070 50-99 -0.338 100-299 -0.385 300-500 -1.142 Totalamount of deferredpayments 0.000 Period to receive deferredpayments 0.043 CauseofInsolvency(comparedtotemporaryprobleminfinance) Perpetual problemin finance 0.200 Decrease in sales -0.061 Insolvencyof the contractor -0.575 Union (=I) 0.070 Industry(compared tomanufacturing) Construction Electricity,gas, wholesale,retail,restaurant, etc. -0.055 Financial, social and personalservices 0.010 Occupation (compared to clericalworkers) Manager,specialist,technician 0.229 Service 1.438 Operator 0.037 Unskilledlabor -0.044 Number of observations 415 'Eligibility is stringent' 185 'Eligibility is not so stringent' 230 - (loglikelihood) 255.90 Note:"', "and 'standfor 0.01,0.05,0.10significancelevels, respectively 120 Labor Market Reforms in Korea: Policy Optionsfor.the Arture For the presence of moral hazard, there is no reliable evidence that supports the surveyed employees' affirmative response with regard to companies having failed to pay a larger amount of deferred wages. The coefficient of the dummy variable related to this question was not significantly different from zero. The labor inspector at the local labor office with jurisdiction over the concerned company determines whether or not to confirm the insolvency. At this stage, the labor inspector's personal judgement may come into play, and thus, depending on the inspector, the standards and requirements may vary significantly. The results in Table 6.10 show that compared to respondents in other cases (e.g., transitory financial strains or chronic financial strains), the respondents from companies that became insolvent following the insolvency of their contractor company were more likely to find the requirements as not so difficult to meet. From the standpoint of the labor inspector, it is easier not only to determine the insolvent state of contractor companies, which are generally large corporations, but also to secure wage claims for exercising the right of indemnity in the future. Thus, such cases give rise to fewer complications with regard to obtaining the confirmation of insolvency. However, in the remaining cases, it is not so easy to determine whether the concerned company is really insolvent or not, and it is even less easy to secure wage claims. For these reasons, workers froin companies that become insolvent due to the insolvency of the contractor company find it easier to obtain confirmation of insolvency. Also, in the case of service sector companies, the labor supervisor finds it rather difficult to confirm insolvency. Naturally, workers from these companies will consider it complicated to obtain confirmation of insolvency. Summary and Policy Suggestions On the surface, workers who benefited from the wage claim guarantee system do not appear to be the underprivileged in the Korean labor market. Their educational background was on average higher than that of the entire economically active population. Also, a larger proportion of them, as compared to the national average, were positioned at a managerial post before leaving the job. However, although a relatively larger proportion of workers benefiting from the wage claim guarantee system had higher education and held professional and skilled positions, their household income was so low at the time the survey was taken that an urgent call for additional policy measures is needed. The result of the survey indicates that both employees and employers find the wage claim guarantee system to be positively effective. Although those who benefited from the system were not A S S E S S I N G T H E COVERAGE A N D EFFECTIVENESS ... 121 considered as vulnerable groups in the Korean labor market, they found it difficult to maintain their livelihood after their employer became insolvent. There still are several improvements to be made when the wage claim guarantee system goes through a redesign process in the near future. First, it is necessary to differentiate the benefits of the workers as well as the rate of the employers' defrayal depending on the size of the company and the type of industry. Among the results found in this paper with regard to differentiation, of particular note is the parallel decrease in the remaining amount of deferred payments (= the total amount of deferred payments - the subrogated payments from the wage claim guarantee fund - any additional amount provided by the employer at a later date) with the company size. In light of this, more consideration should be given to the needs of employees at small-sized companies when improvements are made to the system. In the case of retirement allowances, consideration should be given to the trend whereby workers in the manufacturing and construction sectors show a larger amount of deferred retirement allowances in arrears than those in the service sector. Furthermore, different regulations should apply to different industries. Second, it is imperative that a standard manual for the confirmation of insolvency be developed. At present, labor inspectors at the relevant local labor office examine and determine the state of insolvency. Any decisions made regarding insolvency may reflect their personal opinion, and thus vary on a case-by-case basis. Moreover, labor inspectors also face intrinsic limitations in judgement as they have insufficient information regarding the concerned business establishment. Therefore, a manual that provides objective yardsticks to determine the state of insolvency is urgently required in order to prevent arbitrary judgements by labor inspectors. Third, the wage claim guarantee system is not widely known yet. Thus, employers should be obliged to inform their employees of the wage claim guarantee system, especially when they become insolvent and cannot pay wages and retirement allowances. Fourth, employers' contributions (defrayal) to the wage claim guarantee fund should be readjusted. When the government provides subrogated payments from the fund to laid-off employees, the government later exercises the right of indemnity against employers, even though they already paid their contributions to the fund. So employers are facing dual burdens. In order to alleviate the problem of duplicate burdens, the amount of reimbursement employers are to pay the government should be adjusted depending on the amount of contributions they already made to the fund. Also, further considerations in the form of long-term or installment repayments should be given to employers when the government exercises the right of indemnity. Lastly, it is necessary to improve the current retirement allowance system. The average amount of deferred wages received through the wage claim system is less than that of deferred retirement allowances. Nonetheless, 122 Labor Market ReJorms in Korea: Policy Options jor the Future business establishments are reluctant to reserve a retirement allowance fund, or if they have set aside a reserve, they are likely to use the fund for other purposes. Under these circumstances where companies themselves make very little effort, the subrogated payment of retirement allowances would impose serious financial strains on the operation of the fund in the long run. Also, the payment of retirement allowances might be overlapped by other systems, such as the national pension or the unemployment benefits system. However, this is not a matter that requires improvements within the wage claim guarantee system itself and is beyond the scope of this paper. Therefore, a separate study is needed to improve the legal retirement allowance system in Korea. LABOR LAW AND LABOR MARKET REGULATIONS 7AssessmentofLaborMarket Response to the Labor Law Changes Introduced in 1998* Duck Jay Park" Jonghee Park'" Gyu-Chang Yu"" Introduction In early 1998, Korea changed its labor law in the aftermath of the severe economic crisis that occurred late 1997. There are two main changes: legal permission is granted for dismissal for 'managerial reasons', and legal permission is granted to temporary work and temporary work agencies. The aims of these changes were to help firms overcome their predicament and to boost Korea's economy by making the labor market in Klorea more flexible. The main purpose of this paper is to analyze the effects of the legislative changes on the employment adjustment system and labor market flexibility in Korea. We had three main research objectives. First, we tried to analyze the impacts of the employment adjustment system and temporary work services on the labor market and human resource management. Second, we tried to identify the causes of labor-management conflict that occurred in the process of employment adjustment and the adoption of temporary work services. Third, we tried to find out policy implications for the employment adjustment system and the temporary work services for the future. For the study, we employed multiple methodological approaches. First, we conducted in-depth interviews with human resource managers and union officials in eleven firms and the representatives from two temporary work agencies. Second, we conducted a survey on enterprises. The survey was conducted during November and December 1999 by a professional survey team. Two We are very grateful to Professor Barry Wilkinson for his comments and help in English expression. However, the views and any errors are ours. ** Korea Nalional Open Univcrsity, Korea. **. Korea Univcrsity, Korea. .*.* Sookmyung Women's University, Korea. 126 Labor A4arket Refornzs in Korea: Policy Optionsfor the Futzire sets of questionnaires were developed: one for management, and the other for trade unions. The items included in the questionnaires were the general situations of firms, opinions about dismissal provisions, opinions about legislation on temporary work services, experience of e~nployrnentadjust- ment and related human resource management practices, and experience of using temporary work services. The population of the survey was 6,691 business firms with more than 150 employees, which are obliged to be inspected by external audit institution. Small firms with less than 150 employees were excluded as most of them are non-unionized enterprises in Korea. We used a simple random sampling method. A final sample universe was 638. The number of firms that responded to the survey was 483 (response rate of 76 %). The distribution of the firms by industry was: 226 in heavy and chemical manufacturing, 104 in light manufacturing, 47 in construction, 29 in wholesale & retail, restaurants & hotels, 21 in transport, storage & communication, 25 in finance and insurance, 31 in business, personal, public service & others. Among the 483 firms, 173 were unionized firms and the rest were non-unionized. This chapter is organized as follows. In the following section, we analyze the effects of the legalization on dismissal on the einployment adjustment practices of firms. Then, we analyze the effects of the legislation on temporary work and the temporary work agencies, and we look at the causes of the labor-management disputes in the process of employment adjustment. In the last section, we conclude with policy proposals.' The Effects of Legal Permission to Dismiss In Korea, lifetime employment used to be considered as a normal system. This was mainly because, during the past era of rapid economic growth, firms expanded their organizations continuously, and therefore, employment relations could continue throughout the lifetime of workers. In this era, this system seemed efficient for employers. As companies have traditionally hired new employees without job experience and developed most of them through firm-specific education and training, long-term employment relations had been regarded rational. In these circumstances, the concept of employment in Korea was regarded as a more concrete relation than that based on a contract between boss and worker. Lifetime employment also fitted well with the Korean sentiments of egalitarianism and communitarian consciousness. These sentiments are well ' Owing to the limit of space, we omitted the theoretical aspect of employment adjustment and temporary work. Some articles explaining thc theoretical aspect are presented in the references. expressed by the employee's phrase 'our company', meaning the company he works in. However, the economic crisis that occurred at the end of 1997 brought about a drastic change in the employment practices of Korean companies. During the time of crisis, many big companies collapsed or were downsized. In this situation, the practice of lifetime employment initiated by big companies also began to diminish. It was in 1989 that a precedent first presented detailed conditions for justified dismissal in Korea. The decision of the Supreme Court requested fulfillment of the following four conditions to justify dismissal of employees. First, there had to be urgent managerial needs. Second, all possible alternatives had to have been studied. Third, even if dismissal was unavoidable, employees subject to dismissal had to be selected by reasonable and just standards. Fourth, the employer had to consult with the labor union or employees on the measures for avoiding dismissal and setting fair standards for selecting employees to be dismissed.- In late 1996, the precedents governing dismissal were first legislated. However, as it was opposed by trade unions and was immediately substituted with amended law on March 13, 1997. Then the provisions for dismissal set forth in the Labor Standards Act revised in March 1997 were suspended for two years in accordance with addenda. The laws of 1998 were made to fulfill the condition that called for greater flexibility of the labor market in return for receiving emergency relief funds from the IMF. Current regulations were legislated at the temporary meeting of the National Assembly on February 20, 1998, based on the direction agreed at the Tripartite Commission composed with trade unions, business, &d the government. The ~rovisionsfor dismissal in the Labor Standards Act were in line with the precedents. In other words, the law demands urgent managerial needs, employer's efforts to avoid dismissal, and reasonable and just standards to select employees for dismissal as the preconditions for dismissals. In this section, we will analyze the effects of the change of law on the employment adjustment system of Korea. Employment Adjustment afier the Enactment of the Law concerning Collective Dismissals Change of E m p l o y ~ e a What was the effect of the change of law on employment adjustment in Korean firms? Table 7.1 shows the impact of law amendment on regular em- ployees after the change of law.2This reveals that dismissal was carried out - Employment adjustment was actively carried out mainly in 1998. Korea's rate of GDP growth was 8.9%during the early three quarters 1999, drastically increased from -5.8 % in 128 Labor Market Reforms in Korea: Policy Optionsfor.the Futtire only by 15 %of firms, 75 out of 483, during 10 months just after the change - of law in 1998. The average number of dismissals per firm was about 46. Table 7.1: Impacts of the enactment of act concerningcollective dismissals on regular employees Number of 'Honorable' Dismissal Others4' year employees retirement p -- Less than 10 57 17 3.5 employees 18 3.7 22 4.6 30-39 40 8.3 0 0.0 2 0.4 5 1.0 1998'' 40-49 33 6.8 15 2.9 13 2.7 7 1.4 50 or more 142 27.1 101 20.9 17 3.5 25 5.2 None 103 21.3 312 64.6 408 84.5 419 86.7 Total 483 100 483 100 483 100 483 100 -. Average3' 103 - 231 - 46 - 252 - Lessthan 89 18.4 20 4.1 13 2.7 employee 25 5.2 50 or more 84 17.4 22 4.6 3 0.6 15 3.1 None 148 30.6 399 82.6 439 90.9 428 88.6 Total 483 - 483 - 483 - 483 - Average - - - 3, 64 - 95 15 107 Notes: 1) 1998 meansthe period from March to Decemberof 1998; 2)1999meansthe periodfrom January tothe endof July; 3) Average means averagenumber of the employeesper firms that carriedout the program; 4) Others mean spin-off, early retirement,etc. The percent of firms that used 'honorable' retirement as a means of shedding labor was 35 %, much higher than that of firms using dismissals. The average number of workers per firm who quit by this method was 231, also much higher than the numbers being dismissed. Others, the last column in Table 7.1, include spin-off3, sub-contracting, early retirement, etc. These were used by about 13 % of the firms. The average number of workers per firm affected by these methods was 252. 1998. So the need for employment adjustment decreased drastically in 1999 Spin-off means aprocess of a corporaterestructuringwhereby the capitalstock of adivision or subsidiary of a corporation is transferred to a separate firm, which generally i s established by employees buy-out or managementbuy-out of the employees (management) of the division or subsidiary Tn this case, mother corporation gcncrally guarantees work by subcontracting. Attrition was used most widely in reducing employment. About 80 % and 70 % of the firms reduced employment by way of attrition in 1998 and 1999, respectively. The average numbers of employees per firm reduced by attrition was 103 and 64 in those years, respectively. At the time of the survey, the average numbers of employees of the 380 and 335 firms that reduced elnployees by attrition were 1,368 and 1,627, respectively. Hence, the annual attrition rates were very high at 9 % and 8 % respectively. Why did firms carry out employment adjustment? As shown in Table 7.2, as many as 57 % of the firms replied 'overcoming the crisis of the firm' as the main reason for employment adjustment. This indicates that many firms were in deep crisis when they were trying to adjust employment. Table 7.2: Main reason o f employment adjustment (firms) (%) To make vacancy for promotion 7 2.9 To reduce excess of employees generated by automation 13 5.4 Corporate restructuring and downsizing 39 16.3 To reduce wage costs 31 12.9 To overcome crisis of the company 136 56.7 Merger and acquisition with other company 3 1.3 Others 11 4.6 Total 240 100.0 The fact that so many firms were in deep crisis gives us a general picture of employment adjustment at the time. Confronted with a severe business downturn, with decreasing orders and slack work, and some firms on the brink of insolvency, employees could not help but quit their job. By doing this, they thought they could, at least, retrieve some retirement allowances in addition to the severance pay. In these circumstances, legal permission for dismissal by the change of labor law gave employers the upper hand in the process of persuasion to quit. Many workers were worried that, however strongly they opposed dismissal programs, employers would be able to dismiss them anyway, paying only little or no compensation. In this context, a large number of voluntary or 'honorable' retirees surged. We can see this in an airline company. At first, the employer had planned 'honorable' retirement of about 200 employees in 1998. After the firm's announcement of 'honorable' retirement program, more than 700 out of 16,000 employees applied, and in the end, 695 employees retired by the program. Seeing rampant vacant seat owing to the decrease of passengers and frequent cancellation of flights, many employees were worried about the possibility of bankruptcy of the firm. That was the major reason why so many of them rushed to apply for 'honorable' retirement. We can see the same phenomenon in the banking industry, too. In this sector nine major banks cut huge numbers of staff by October 1998 through 130 Labor Market Reforms in Korea: Policy Options for the Ftltzu-e 'honorable' retirement. Even though 9,000 staff cuts were planned at first, an extra 1,000 took 'honorable' retirement as they worried they would lose any chance of retrieving a retirement allowance if they stayed. Chohung bank was to cut 2,222, but an extra 200 took the 'honorable' retirement. Those to go were primarily young females in low-level jobs. After that, banks rehired some retirees on a temporary basis (The Chosunilbo, October 9, 1998). As a result of this kind of employment adjustment, there were large-scale job cuts in the commercial banking sector. After the end of 1997, the banking sector cut the number of regular employees drastically. As shown in Table 7.3, the total number of employees in this sector was 113,994 at the end of 1997. At the end of 1998, however, 75,677 remained, a 34 % cut. At the end of June 1999, this number was further reduced to 74,851.4 Table 7.3: Change of regular employees of commercial banking sector in Korea, 1997-1998 End of 1997 End of 1998 End of June 1999 (persons) (%) (persons) (%) (Persons) (%) National banks 94,065 100.0 64,830 68.9 65,187 69.3 Regional banks 19,929 100.0 10,847 54.4 9,664 48.5 Total 113,994 100.0 75,677 66.4 74,851 65.7 Source: Financial Supervisory Commission. In 1999, firms that cut regular employees by attrition or any other intentional devices decreased compared with 1998. As will be explained shortly, some of this reduction may be the result of the speedy recovery of Korea's economy in 1999. In 1999, however, the order of employment adjustment devices used by firms and the number of employees affected by them were the same as in 1998. Detailed Analysis of Employment Adjustment What kind of firms carried out which employment adjustment programs? Let's look at the firms which carried out dismissals. Table 7.4 shows that, dismissal was most common in small firms with employment of less than 300.5 Big firms with employment of 1,000 or more were far less likely to In terms of employment adjustment, the banking sector was in a peculiar position after the economic crisis. While the major reason for Korea's economic crisis was the bankruptcy of several heavily indebted business groups, their problems left the banking system in tatters, necessitating a large-scale rescue financing by the government. The relative briskness of dismissal in small firms is also ascertained in other survey of the j Korea Labor Institute. This shows that 87 out of 355 firms that carried out dismissals from April to October 1999, 32 firms (37 %) were firms with less than 100 employees (Korea Labor Institute, 1999b. p.28). carry out dismissal. Non-unionized firms were more active in dismissal compared with unionized firms. The low percentage of dismissals in big or unionized firms might be related to the opposition from trade unions or large number of employees to dismissals (see Table 7.4). Table 7.4: Characteristics of the firms that carried out dismissal Industry Number of employees Union More Mfgl') Mfg22) Others3) Less than 300 999 than Uq Nu" 299 999 1999 17 13 14 24 13 7 22 22 Co. (7.5%) (12.5%) (9.2%) (14.5%) (7.4%) (4.9%) (7.1%) (12.7%) Workers 6 17 27 6 35 7 11 21 No. 226 104 153 166 175 142 310 173 Notes: 1) Manufacturingsectors such as metal and mechanical products, transport equipment (including cars), chemicals,medicines,plants, etc. (so-called heavy) and chemical industry; 2) Manufacturing sectors such as food products, textile and garments, printing and publishing. medicalequipment, etc. (so-calledlight industry); 3) All industriesexcept manufacturing; 4) Unionized firms; 5) Non-unionized firms; 6) Number of firmsthat carried out dismissal; 7) Average number of workers dismissedper firm that carried out dismissal; 8) Number of firms that responded to the question; Numbers in the parenthesis are the % of the total respondents Table 7.5 shows that firms in heavy and chemical industry and non- manufacturing industry rather than firms in light industry carried out 'honorable' retirement more often. Big firms with 1,000 or more employees and unionized firms were more active in carrying out 'honorable' retirement. Unionized firms were a bit more active in 'honorable' retirement compared with those without a union. 'Honorable retir'ement' seemed to function as a substitute for dismissal for the big unionized firms. In the case of spin-off or early retirement (others), firms in heavy and chemical industries were more active compared to those in light industry or those of non-manufacturing industries. Big firms with 1,000 or more employees were also more active rather than small-and-medium size firms in adjusting employment by these programs. Non-unionized firms were more active in utilizing these programs compared with unionized firms (see Table 7.6). 132 Labor Market Reforms in Korea: Policy Options for the Future Table 7.5: Characteristics of firms that carried out 'honorable'retirement Industry Number of employees Union More Mfgl lJ MfgZ2' OLhers3' Less than 300 999 than U" Nus) 300 999 Workers7J No. 1999 Co. Workers No. Note: same as Table 7.1 and Table 7.4 Table 7.6: Characteristics of firms that carried out spin-off or early retirement Industry Number of employees Union More Mfgl l) MfgZ2' Othecs3' than Less 300 999 than ~4 Nu" 300 999 1998 34 11 19 18 20 26 37 27 CO.~' (15.0%) (10.6%) (12.4%) (10.8%) (11.4%) (18.3%) (11.9%) (15.6%) Workers 162 54 57 9 31 188 141 46 NO. 226 104 153 166 175 142 310 173 Note: same as Table 7.5. Lastly, firms that reduced employees by attrition were prevalent regard- less of industry, number of employment, unionized or not. 'Honorable' Retirement: Characteristic Features of Employment Adjustment System in Korea 'Honorable' retirement appears to be a system of employment adjustment ASSESSMENT OF LABOR MARKET RESPONSE TO THE LABOR ... 133 peculiar to Korea. Here, we will analyze it in more detail. In general, its process proceeds as follows. When an employer finds out that he has an excess number of employees in relation to current and anticipated demand, he sets a target of job cuts, in accordance with the business strategy. When the workplace is unionized, he consults with the union and seeks its consent. Then, he invites employees to apply for 'honorable' retirement, promising to pay some compensation to the retiree until the number of applicants reaches the target. If the number does not reach the target, they sometimes persuade, induce, or coerce certain employees to apply. The expression of 'honorable' here is meant to give impression of praising applicants for the sacrifices they make for the sake of colleagues and the company. In general, they have a target group who they wish to release. The specific features of the target group depend on the employment structure of the firm. They are generally employees of low ranking in performance appraisals, old workers anticipating retirement in the near future and female or young workers who do not have a large family to support. As it is based on the application of individual employee, there is a risk of adverse selection. That is, competent employees, who the employer did not wish to lose, may apply. Especially in circumstances of bleak business prospects, the risk escalates. In a normal firm, however, the employer is able to reduce the size of the workforce through persuading incompetent employees to retire, paying some compensation. Employers paid more money in compensation to the workers who took 'honorable' retirement than to those to be dismissed. The amount of compensation for individual applicants depends on factors such as seniority, remaining work year, rank and position, etc. As Table 7.7 shows, 95 % of the firms paid retirement allowances. The average amount was about 7 months' wages of the retiree. Table 7.7: Payment of compensation for 'honorable' retirement (firms) (%) Paid Did not pay Total The average amount of ~avment Waqes of 7 months By adopting this method of employment adjustment system, which requires some compensation, employers in Korea could weaken the opposi- tion against employment adjustment from the workers and the union. In these firms, many workers appreciate 'honorable' retirees as martyrs sacrificing themselves for the company and the co-workers. So, it was helpful for improving employees' morale in times of hardship. Furthermore, it could be a convenient bypass for the leaders of trade 134 Labor Market Reforms in Korea: Policy Options for the Ftltzlr-e unions. They may understand the inevitability of employment adjustment of the firm in due course. However, it is taboo to attend a meeting for consultation on dismissal, least of all to agree to it, in Korea. 'They are liable to be discharged by a no-confidence resolution by the agitation of militant sections of the union. However, as 'honorable' retirement is carried out based on de jure application, although not necessarily a defacto one, of the employees, trade union leaders could be excused for not being militant enough to block the employment adjustment. An example of an 'honorable' retirement program of a TV maker, one of our case studies, will help in understanding the system. Because of the excess supply of TVs in the world market, its profits plummeted since 1991, with a huge loss in 1998. Accordingly, its employment decreased to 1,400 in 1997, mainly through attrition, from the peak of 2,500 in 1991. In 1998, the firm adopted three devices for further cuts. Firstly, by spin- off it reduced employment, mainly in sectors of automatic inserting and packing materials, by 130. Secondly, by subcontracting operations previously done by its own employees in sectors like transport and logistics, it reduced employment by 70. Lastly, by 'honorable' retirement program it reduced einployment by 94. Any employees who were in service inore than 5 years could apply for it. The amount of compensation promised by the firm for the 'honorable' retirees was about 1,600,000 - 1,800,000 won (US$ 14,000-15,600) per person depending on seniority and the rank of the retiree. The first 'honorable' retirement program was in February 1998. At this time there were 46 retirees, all of them de facto applicants and most of them manual workers. As the first cuts were small compared to the firm's plan, there was a second 'honorable' retirement program in August 1998. At this time the employer eagerly tried to invite mainly non-manual workers to apply for the program. In particular, the employees in the bottom order of the performance appraisal were the main target of the persuasion. At this time, the firm cut by 48. As trade union leaders and employees had been watching slack orders and idle factory lines for years, they did not oppose to the program. As the employer coerced the non-manual workers to apply, union leaders were not so eager to block it, either. The main target of the employer's invitation was non-manual workers, who were not members of the trade union. Instead of opposing it: the union leaders demanded increased compensation, to which the firm agreed. In addition to the comparatively generous compensation, the support program for the retirees which helped retirees find new jobs probably made it proceed more smoothly. As employers utilized devices like 'honorable' retirement, they had more discretion in recruiting new employees. We could find that many firms recruited new employees in the process of the adjustment. Table 7.8 shows that about half of the firms that carried out dismissals or 'honorable' retirement recruited regular employees in 1998. In the same period, about 34 % of them recruited non-regular employees, too. In 1999, the number of firms that recruited regular or non-regular employees increased.The fact that so many firms were recruiting new employees in the midst of employment adjustment seems to indicate that Korea's employers have strong authority in recruitment, with no strong intervention from the union, unlike in the U S 6 According to Table 7.8, the percentage of the firms with newly recruited employees is higher in non-unionized firms than unionized firms. The gap, however, is not so large. Moreover, the average number of recruitment of regular employees in unionized firms is much larger than that of the non- unionized firms, both in 1998 and 1999. Under the seniority wage system in Korea, old worker's wages are higher than those of the young in the same job. Most of the programs for 'honorable' Table 7.8: Number of firms having new recruitment among the firms that carried out dismissal or 'honorable' retirement Regular workers Non-regularworkers Number of workers year Ci:i recruited union Total Union Total Less than 10 workers 28 11 39 24 14 38 10-19 13 8 21 16 2 18 20-29 9 5 14 2 2 4 30-39 7 8 15 4 0 4 4049 2 2 4 1 0 1 998 50or more 21 6 27 6 8 14 Did not recruit 89 27 116 115 41 156 Total 169 67 236 168 67 235 Averagenumberofwcrkers 79 46 68 34 8 14 recruitedinihefirmsthatrecruited Less than 10 employees 10-19 20-29 30-39 4049 999 50 or more Did not recruit Total Averqenumberofwakers recruitedinthefirmsthatrecruited Notes: 1) Respondents are employers of the firms that conducted dismissal or 'honorable' retirement programs. 2) Average means average number of the newly recruitedemployees per firm that respondedto haverecruited In the US, most collective agreements provide for recall of employees after layoff. These provisions usually specify that employees be rehired in order of seniority (Holley and Jennings, 1997, p. 459) 136 Labor Market Reforms in Korea: Policy Optionsfor the Future 'honorable' retirement, therefore, target primarily old workers. If there are not enough applicants to reduce wage costs, employers sometimes coerce old employees to apply. By replacing young workers for old ones, they could succeed in reducingsubstantive wage costs. Temporary Work Services Introduction After the financial crisis in late 1997, increasing flexibility of the labor market was one of the important conditions necessary to recover from the economic crisis in Korea. The Tripartite Commission of the Republic of Korea, established on January 15, 1998 to overcome the economic crisis, issued its first joint statement on January 20, 1998, that labor, management and the government would each bear the burden of responsibility for the crisis equally. As one of the means to achieve labor market flexibility, the Tripartite Commission made a critical compromise regarding temporary work agencies. On February 20, 1998, the Act Relating to Protection, etc., for Dispatched Worker was enacted. The "temporary work" defined in the Temporary Work Act7 means that the temporary work agency hires employees (dispatched workers),and while maintaining employment relations, leases these employees to a third party (user firms) to work under the service recipient's supervision and direction in accordance with the employment leasingagreement. The Temporary Work Act places restrictions on the types of business for which temporary work is allowed. These are divided into two categories. One is temporary work in areas that require special knowledge, technique, or experience, other than manufacturing, falling under the 26 categories set forth in the Presidential Decree. The term for such temporary work shall not exceed one year. The purpose behind such restrictions is to uphold the original purpose of temporary work business, which is to meet the temporary demand for workforce and the demand for specialists who cannot be employed at all times. In other words, by making sure employees are leased for just the appropriate amount of time, the service recipient may not make a common use of dispatched workers for a long-term to replace regular employees. However, if an agreement is reached between the temporary work agency, the company that uses these employees, and the employees dispatched, the term may be extended by one year, limited to one extension.' ' We use "Temporary Work Act" here instead of its full name of "Act relating to Protect~on etc., for D~spatchedWorkers" just for convenience. "Temporary work" is more commonly used term in journals written in English in Korea. If a company uses dispatched workers for an uninterrupted period of more than two years, these workers shall be regarded as regular employees of the company after the two year The other is dispatched workers to fill a temporary vacancy. When a company needs dispatched workers to fill a temporary vacancy or meet increased need (not for the purpose of replacing regular workers), it may lease workers for a limited time regardless of the type of work, as long as the work is not listed in the Decree. However, even in such a case, the employer must consult in good faith, prior to such leasing, with the labor union or the representative of a majority of workers if there is no such union. The term of lease is restricted even in cases of leasing employees for such temporary and intermittent need. When it is clear that e~nployeesare needed to fill a vacancy or a temporary need, the term of lease may be such as to solve the problem. For instance, to fill a vacancy for maternity or sick leave, employees may be leased for the term of such leave. In other cases, the term may not exceed three months, but if an agreement is reached between the three parties, the term may be extended for not more than a three-month period.' The Current Situationof TemporaryWork Services in Korea Although temporary work agencies had been illegal in Korea, quite a large number of manpower leasing agencies had actually operated even before the passage of the law under the implicit allowance of the government (Jeong & Yoon, 1993; Jeong, 1999). The precise number of temporary work agencies was not known. However, it was estimated that at least hundreds of agencies operated. After the passage of the law, we can estimate the number and activities of temporary work agencies since the Ministry of Labor of Korea (MOL) announces key statistics for them. The most recent data available at this time are presented in Table 7.9. In the 4thquarter of 1998 (about a half year after the passage of the law), 789 temporary work agencies had obtained legal permission from MOL. The number increased to 968 during the 1" quarter of 1999 and again to 1,095 during the 2ndquarter of 1999. That is, during a period of 6 months, the number of legally permitted temporary work agencies increased by almost 40 %. However, the number is somewhat exaggerated because not all permitted agencies actually provide services. According to MOL, only 784 firms out of 1,095 are currently in the temporary work business, which is term expires, provided that the dispatched employees do not clearly demonstrate a wish to the contrary. " There are cases in which temporary work is prohibited even in temporary or intermittent need. These may be categorized as the following: construction work, unloading work related to transportation or circulation business, ship work under the Seamen's law, work harmful to public morals or sanitation, work harmkl or dangerous under the Industry Safety and Health Act, driving in the transportation business, and certain work under laws governing the medical business 138 Labor Market Refornzs in Korea: Policy OptionsJor the Futzire Table 7.9: Recent trends in temporary work agencies in Korea, 1998 Q4-1999 Q2 Number of temporary work agencies (TWAs),permittedby MOL Number of W A S ,currentlyin actual 564 728 (29.1%) 784 (7.7%) operation Number of user firms,received services 4,302 4,178 (-2.9%) 4.701 (12.5%) from TWAs Number of dispatchedworkers 41,545 44,665 (7.5%) 46,407 (3.9%) Average number of workers per agency 73.7 61.4 59.2 Note:The numbers in parenthesis are the rates of increase compared to previousyear 71 % of the total agencies. As of the 2ndquarter of 1999, 4,701 employers used dispatched workers, which represents 9 % increase after the passage of the law. On the other hand, over the same period, the number of dispatched workers employed by TWAs increased by 12 %, from 41,545 in 1998 to 46,407 in 1999. Although the increasing rate of dispatched workers is very high compared to that of regular workers, the proportion of dispatched workers out of total employees in Korea is still 0.2 %, which is very small cornpared to other OECD countries (OECD, 1999). The average number of user firms per TWA is currently 6 as of the 2nd quarter of 1999, a slight increase from 5.7 in the first quarter, but a decrease from 7.6 in 1998. Also, the average number of dispatched workers is currently 59 per TWA, a decrease from 61 in the second quarter of 1999 and 74 in 1998. This indicates that temporary work agency activity is rapidly expanding, but their size is very small, and getting smaller. Table 7.10 shows more detailed figures of the average number of workers per TWA. As indicated, 311 agencies (28 %) are not currently operating businesses." Also, most agencies (87 % of acting agencies) have less than 100 dispatched workers. Only 101 agencies have more than 100 dispatched workers and 2 % of all have more than 300 dispatched workers. While this figure confirms that most TWAs in Korea are very small, there is also the phenomenon that a few agencies are getting bigger. The president of the Association of TWAs in Korea told us in an interview that a few leading agencies are preparing M&A in response to both the demand of the market and the entry of large international TWAs. "' The reason they are not operating after receiving permission for business from MOL is not clear. We hypothesize that they overestimated the market situation in the temporary work industry, but it has turned out to be too competitive for many small agencies to compete profitably ASSESSMENT OF LABOR MARKET RESPONSE TO THE LABOR ... 139 Table 7.10: Trends in the averagenumber of workers per TWPI, 1998 Q4-1999Q2 Number of 1998 Q4 1999 Q1 1999 Q2 workers (persons) (%) (persons) (%) (persons) (%) 0 225 28.5 240 24.8 311 28.4 1- 49 407 51.6 548 56.6 583 53.2 50-99 71 8.9 81 8.4 100 9.1 100 -299 64 8.1 74 7.6 77 7.0 More than 300 22 2.9 25 2.6 24 2.2 Total 789 ,100 968 100 1,095 100 User Firms In our survey, as well as asking questions on dismissal procedures, we also sought to identify the extent to which firms used dispatched workers (See Table 7.11). Thirty-seven % of firms reported that they currently use dispatched workers. During 1998, after passage of the law, 27 % of firms used dispatched workers. However, the usage rate of dispatched workers employed by firms rapidly increased in 1999 from 27 %(129 firms) to 37 % (177 firms). The survey also showed that user firms tend to employ more dispatched workers than in the past. Firms that employed more than 50 dispatched workers increased from 16 % in 1998 to 23 % in 1999 (See Table 7.12). Similar to the worldwide tendency of an expansion of dispatched work, we found strong signs that firms increasingly rely more on TWAs for their human resource management practices in Korea. We asked firms whether they would expand or discontinue use of dispatched workers if they currently used them or whether they planned to employ dispatched workers in the near future if not currently used by them. Only 33 % of respondents said they would neither expand, nor plan to use, dispatched workers (See Table 7.13). Table 7.11: Companies using dispatchedworkers Use Do not use Total (firms) (%) (firms) (%) (firms) ("4 178 37.0 305 63.0 483 100.0 Table 7.12: Trends in number of dispatched workers used per firm March 1998- .January1999- December 1998 November 1999 Less than 10 dispatched workers 59 (45.7%) 76 (42.9%) 10- 19 19 (14.7%) 25 (14.1%) 20-29 12 (9.3%) 21 (11.9%) 30-49 19 (14.7%) 14 (7.9%) More than 50 20 (15.5%) 41 (23.2%) Total 129 (100.0%) 177 (100.0%) 140 Labor Market Reforms in Korea: Policy Optionsfor the Ftlttlre Table 7.13: Future plan to use dispatched workers (firms) (%I Expand or newly plan to use 211 43.9 Neither expand nor planto use 161 33.3 Don't know 111 23.0 Total 483 100.0 There were differences in industries regarding the extent to which firms use dispatched workers (See Table 7.14). It was shown that firms in the service sector, such as retail, banking and other services industries, tended to use dispatched workers more frequently than manufacturing firms. Around half of service firms among survey respondents responded to use dispatched workers. We also analyzed size differences regarding the extent to which firms use dispatched workers (See Table 7.15). Larger firms tend to use TWAs more frequently than smaller firms. However, there was little difference regarding future plans to use dispatched workers, in terms of both industry and firm size. Table 7.14: Companies using dispatched workers by industry Heavy Light ~onstnrc- Transpor- Other manu- manu- Retail tion tation Banking selvice faduring faduring Use (%) 38.9 30.8 17.0 44.8 23.8 64.0 51.6 Donot(%) 61.1 69.2 83.0 55.2 76.2 36.0 48.4 Total(%) 100 100 100 100 100 100 100 Totalnumber 226 104 47 29 21 25 31 offirms Table 7.15: Firms regarding using dispatchedworkersby firm size Number of 1 299 - 300- 1000 More than 1000 employees (firms) (%) (firms) (%) (firms) (%) Use 39 22.0 66 37.3 72 40.7 Do not use 127 42.2 109 36.2 65 21.6 Total 166 34.7 175 36.6 137 28.7 User firms were asked why they resort to dispatched workers. The most frequent answer was to fulfill peripheral work (26 %). Next to that were to obtain employment flexibility (23 %), to reduce labor costs (22 %), to supplement a temporary labor shortage (11 %), due to a temporary high workload (6 %), and due to jobs which require special knowledge and skills not available within the firm (6 %) in that order. ASSESSMENT OF LABOR MARKET RESPONSE TOTHE LABOR ... 141 Table 7.16: Reasons why firms use dispatched workers - lst reason 2" reason 3rdreason (firms) To supplementa temporarylabor shortage 20 12 22 To fulfilla short-termdemand increase 11 15 17 To fulfillperipheralworks 46 31 23 Dueto jobswhichrequirespecialknowledgeandskills 10 6 5 To supplementworkers for specific time period 8 8 8 To reduce labor costs 38 35 21 To obtainemploymentflexibility 41 49 23 To preventlabor-managementconflict 2 1 5 Others 0 1 1 Total 176 158 125 Impact of the TenzporaryWork Act on Labor Market in Korea We asked questions regarding the impact of the passage of the Act in 1998 on labor markets and the interests of workers to both employers and union officials (Table 7.17). First of all, employers regarded the passage of the Act to have had a positive effect on the growth of temporary work services and also a positive effect on labor market flexibility. About 40 % of employers believed that the Act helped the activation of temporary work services in Korea, while only 11 % of employers disagreed. About 50 % of employers agreed that the Act helped enhance labor market flexibility. Employers did not think that the contents of the law were inappropriate in terms of the protection of dispatched workers. However, e~nployerswere concerned about types of jobs permitted and length of employment periods. About 60 % of employers thought that the types ofjobs permitted in the law are too limited, whereas 58 % of employers complained that periods of temporary work services are too restrictive. In interviews, employers argued that the maximum two years of contract are too short to fully utilize dispatched workers in terms of the skills and experiences obtained within the firms. Likewise, when we asked an open question regarding improvement of the law in the hture, the most frequent answers were types of jobs and lengths of employment periods (Table 7.18). However, opinions of union officials were very different from those of employers (See Table 7.19). Seventy-seven % of union officials did not agree that that the Act includes regulations for appropriate protection of dispatched workers. Also, 61 % of union officials thought that the types of jobs were rather too comprehensive, and 63 % of them disagreed that periods of services were appropriate. Union officials were mostly concerned about protection of dispatched workers. In an open question regarding improvement of the Act, the most frequent answers were, "Need more 142 Lnbor Market ReJornzs in Korea: Policy Optionsjbr the Fz~tzlre protection for workers regarding working conditions"and "Need methods to prevent wage appropriation by temporary work agencies". Some of the union officials even argued that the Act must be abolished because of the possibility of abuse by TWAs and user firms. Table 7.17: Employers' opinions regarding the impact of the act concerning temporary work agencies Neither agree Agree Strongly Don't Strong'y Disagree Total disagree nor disagree agree know Helped the growth temporarywork services Helped enhance labor market flexibility Due to too much protectionof workers, made it more difficult to use dispatched workers Included appropriate protection of workers Types of jobs are too limited Periods of temporary work services are too restrictive Table 7.18: Employers' opinions regarding the amendment-ofthe act concerning the temporary work agency (firms) Lifl the limitation of types of jobs or diversifyfurther 115 Lift the limitationof periods or expand longer 100 Total 225 Table 7.19: Unions' opinions regardingthe impact of the act concerning temporary work agencies Strongly Disagree Neither Agree Agree Strongly Don't Total Disagree nor Disagree Agree Know Included appropriate 66 81 23 12 1 8 191 protection of workers (34.6%) (42.4%) (12.0%) (6.3%) (0.5%) (4.2%) (100.0%) Types of jobs are too 11 33 15 95 21 16 191 comprehensive (5.8%) (17.3%) (7.9%) (50.0%) (11.0%) (8.4%) (100.0%) Periods of temporary 38 83 19 21 1 29 191 work servicesare (19.9%) (43.5%) (9.9%) (1.0%) (0.5%) (15.2%) (100.0%) appropriate Table 7.20: Unions' opinionsregardingamendmentof the act concerning temporary work agency (frequencies) Must provide ways for preventingwage appropriation Types of jobs must be restricted Must abolish the law because of abuse 12 Need more protection for workers regarding working conditions 16 Total 43 Labor-Management Relations in the Process of Employment Adjustment Wherever they were, workers and trade unions were never insensitive to employment adjustment. Even in the US, where an 'employment at will' doctrine predominates, large non-union firms are reluctant to lay off employees for fear of unionization. In the case of unionized firms in the US, management has obtained union acceptance of layoffs in times of downturns, in return for management's acceptance of seniority as the basis of layoffs (Mills, 1994, p. 487). As there is not such a systematic compromise between the two sides, the employers in Korea have reasonable grounds to fear employees' opposition to labor adjustment. This is another important reason why so many employers choose 'honorable' retirement rather than dismissal. By adopting 'honorable' retirement, they hope to evade industrial conflict. As explained above, employers must fulfill 4 preconditions if they are going to dismiss employees due to managerial reasons. During the process of fulfilling these conditions, they are liable to collide with trade unions or representatives of employees. Trade unions representing employees of big corporations were especially militant against the employers' attempts at dismissal. The militancy against employment adjustment is partly connected to the unions' tough bargaining tactics which are used in anticipation of a better pay off. This was well illustrated by industrial disputes in some of the Korean firms. For instance, in 1998, Hyundai Motor Company was going to dismiss 1,448 employees after domestic car sales plunged by 50 %. Before that, the firm cut 4,858 jobs from a total of 46,000 through 5 separate 'honorable' retirement programs. However, trade union leaders insisted that all dismissals be revoked in exchange for wage cuts and paid leave. As the employer could not accept that, the union went on strike, although it was illegal, for about four weeks. After an acrimonious showdown between the union and the company, the union finally agreed to accept dismissal. The deal between the two sides, with state intervention, appeared to achieve the 144 Labor Market Reforms in Korea: Policy Options for the Futzrre union's goal. As the unionists insisted on dismissal of a minimum number, it was reduced to a token 277 from the originally targeted 1,538. Workers who lost jobs received compensation of up to nine months' wages. Another 1,261 workers were to go on 18 months' unpaid leave, with the last six months spent in job retraining schemes (The Korea Herald, August 25, 1998). There is another example of an industrial dispute on the dismissal issue. The trade union in Mando Machinery Corporation led a strike in August 1998 protesting against the company's plan to dismiss 1,090 workers out of 4,500. After a 17-day-long strike, the police raided the six factories of the company. The police raid, although it ended the strike, touched off strong protest by labor sympathizers nationally, and the Korean Confederation of Trade Unions (KCTU) threatened to stage an anti-government campaign and a sympathy strike (The Korea Herald, September 4, 1998). In the case of trade unions in small-and-medium firms and those in firms verging on collapse, the opposition against employment adjustment was not so strong as to lead to strikes. Table 7.21 shows this. Even among unionized firms, only 5 % of the firms (9 out of 171) replied there were labor- management disputes. Among non-unionized firms, there were labor- management disputes in 3 % of the firms. Table 7.21: Firms that had a labor-management dispute in the process of employmentadjustment Had Did not have Total - Unionized 9 (5.3%) 162 (94.7%) 171 (100.0%) Not unionized 2 (2.8%) 69 (97.2%) 71 (100.0%) Even in the firms that had no labor-management disputes, curtailing employment was not easy. What was most difficult in the process of employment adjustment for the employers? Table 7.22 shows that about half the employers replied that it was to make employees understand the necessity of employment adjustment. Next came 'consultation about the criteria to select who are to lose their jobs', 'consultation about the compensatory package for the workers to be adjusted', and 'consultation about the number to be adjusted', in that order. A higher percentage of employers at unionized workplaces than those of the non-unionized workplaces answered that making employees understand the necessity of employment adjustment was a difficult task. We can also see the difficulties faced by employers through trade union's responses to the attempt to enact dismissals or 'honorable' retirement. About 45 % stated they were strongly opposed to dismissal, and 32 % were strongly opposed to 'honorable' retirement (see Table 7.23). According to the opinion of employers, paying a redundancy package and allowances for 'honorable' retirement is a way of reducing opposition from workers (Table 7.24). In the case of dismissal, 73 % of the firms paid ASSESSMENT OF LABOR MARKET RESPONSE TO THE LABOR ... 145 redundancy packages in addition to the severance pay required by the law. The average amount was as much as 2.6 months' wages. In the case of 'honorable' retirement, employers generally promised to pay a higher amount of retirement allowances than in the case of dismissal. The difference of the average payments between the two seems to be related to two facts. One is that mainly small-and-medium firms carried out dismissals. The other is that in the case of dismissals, employers cannot soothe workers resistance by the payment of allowances. Table 7.22: Most difficult thing to firms in process of employmentadjustment Unionized Non-unionized Total Make understandthe necessityof 73 (51.8%) 16 (39.0%) 89 (50.9%) employmentadjustment Consultationabout the number to 16 (11.3%) 5 (12.2%) 21 (12.0%) be adjusted Consultationabout the criteriato 28 (19.9%) 14 (34.1%) 42 (24.0%) select to be adjusted Consultationabout the 18 (12.8%) 4 (9.8%) 22 (12.6%) compensatorypackage Others 6 (4.3%) 2 (4.9%) 1 (0.6%) Total 141 (100.0%) 41 (100.0%) 175 (100.0%) Table 7.23: Union's positions to the attempt of firm's employmentadjustment 'Honorable'retirement Dismissal Support for stable employmentof 20 (18.2%) 22 (30.1%) majority of the employees Reluctantlyacquiesce Strongly oppose Total Table 7.24: Firms behavior concerning redundancypackage Paid 62 (72.9%) Did not pay 23 (27.1%) Total 85 (100%) Amount of payment 2.6 months' wages Note: Amount means the average of the firms that respondedto have paid. We can see this by the answers of employers about the reasons of their paying the redundancy package. Table 7.25 shows that most employers thought it reasonable and humane to pay some compensation and show their sympathy to the workers who were going to retire against their wish. On this matter, there was not much difference in the percentage between the firms with trade unions and without unions. 146 Labor Market Refirnu in Korea: Policy Options for the Futzlre Table 7.25: Reasons for paying redundancy package Unionized Non-unionized Total Smoothing resistance of retirees 12 (26.7%) 6 (23.1%) 18 (25.4%) Sympathy, goodwill to the retirees 33 (73.3%) 17 (65.4%) 50 (70.4%) Others'] 0 (0.0%) 3 (11.5%) 3 (4.2%) Total 45 (100%) 26 (100%) 71 (100.0%) -- Note: 1) Others include methods for improving morale of remaining employees, donation of the rema~ning employees, etc. This employer attribute may be related to their needs to uphold a community consciousness among their employees. Table 7.26 shows that among the firms that carried out dismissal or 'honorable' retirement, 53 % carried out some activities for enhancing the morale of remaining employees after adjustment. About 33 % of them undertook activities to help outgoing workers find jobs. These activities include providing information on the labor market, supporting business start-ups, providing training for getting jobs, etc. A high proportion of firms was active in helping not only the remaining employees but also the outgoing employees. This indicates that the employer's relation with the employees does not end completely through dismissal or retirement. Table 7.26: Firms that acted for outplacement and moraleenhancement Did Didnot Total Outplacement 80 (32.9%) 163 (67.1%) 243 (100.0%) Morale enhancing 128 (52.9%) 114 (47.1%) 242 (100.0%) From the employees' point of view, compensation for 'honorable' retirement allows them to maintain their living standards for the time being. Korea currently has a poorly developed welfare system and has a meager social safety net. Considering this, paying some amount of inoney in addition to the mandatory severance pay to the retirees must be helpful for smoothing opposition from the unions and boosting the morale of the remaining workers as well. Policy Suggestions and Conclusion There was active employment adjustment in Korea after the change of the Labor Standards Act in 1998, though the major method for employment adjustment after the passage the law was not dismissal but 'honorable' retirement, spin-off, early retirement, and attrition. However, the overall employment adjustment undoubtedly helped by the change of law permitting dismissal due to managerial reasons. As we have seen in above, ASSESSMENT OF LABOR MARKET RESPONSE TO THE LABOR ... 147 the number of temporary work agencies, dispatched workers and user firms has also rapidly expanded in the period after the passage of the Act. Employment adjustment in 1998 and early 1999 was facilitated not only by legislative change but also by the economic crisis. As so many firms were in deep trouble many employees chose to leave jobs for fear of losing severance pay, or on behalf of the firm they have always worked for and identified with. Furthermore, the unions were weakened by a sharp increase in unemployment and by public opinion unsympathetic to their militancy. However, in future the need for employment adjustment will not come from business downturns alone. As globalization of the economy progresses, as advanced technology becomes prevalent, and as the lifecycle of products shrink, more flexibility in the labor market is likely to be called for, whatever type it may be. Furthermore, the conditions underpinning long- term employment - strong corporate profitability, easy bank lending terms and a steady supply of graduates committed to one company - are more likely to change as the chaebols are driven by the government towards reform. Recently, with the success of new firms utilizing advanced technology, there has been a collapse of long-term employment in the labor market of university graduates. Nowadays, talented people are less likely to choose long-term employment in search of higher earnings. They change jobs, arrange flexible working hours, and even work from homel which were all unthinkable a few years ago. In this era of rapid change, big corporations, which are not necessarily sufficiently flexible to adapt, are losing hundreds of talented brains. These trends could exert a great influence on employment practices in the 'old' industries. Nevertheless, considering the sentiments of strong egalitarianism and communitarianism in Korea, employment adjustment by dismissal is not expected to be dominant in the near future. Rather than dismissal, 'honorable' retirement and functional flexibility through job rotation will still be more popular. Further, if firms rely on a harsh approach of downsizing and drastic job cuts, they must take a risk of tearing up their loyalty contracts with their workers, the majority of whom are not of the 'new graduate' category with easily transferable scarce skills. There are differences of opinion on labor market flexibility. The position taken in this paper is that it is inevitable that employers vvill be allowed to dismiss employees due to managerial reasons, at least for a considerable time. However, labor flexibility must be pursued without destroying workers' employment stability and living conditions. This is not an easy task in any country, least of all in Korea which is in a vulnerable position with a bailout from the IMF. So, what should government policy concerning labor market flexibility be? First of all, the government should enlarge the social safety net so that unemployed people can live without a drastic d.egradation of living conditions. For instance, the employment insurance system in Korea was 148 Labor Market Reforms in Korea: Policy Optionsfor the Future initially relevant only to the better-off employees, excluding the daily workers and part-time workers. Because of this, when the crisis broke out and the workers lost jobs en masses, it could not adequately cope with the situation. It could provide benefits to only about 99,000 (5.8%) persons among the 1.7 million unemployed. This gave rise to a widespread sense of discontent (Pak et al.,1999, p. 44). By improving the employment insurance system and social safety net the government can enlarge the protection for the unemployed and unskilled workers. In view of the growing income gap in Korea, these policies should have priority." In an upcoming era of the 'digital economy', the inequality between the permanent skilled workers and the temporary unskilled ones is likely to increase further. Second, there should be a policy to upgrade workforce skills to keep up with changing technology. Up to the present, firms were the most important actors in providing training for workers. As workers' mobility increases and more workers are employed on a temporary basis, the benefits for firms to invest in the training of under-skilled employees diminish. As global competition increases, firms can no longer afford to wait for their employees to develop over many years. In these circumstances, government-private sector 'training partnerships', as have been considered in the UK, could be invaluable. Third, there should be a legal stipulation which helps identify the legitimate representative of employees. According to the present law, the employer who intends to dismiss employees due to managerial reasons should consult with the representative who stands for more than half of all employees. The trade union representing more than half of the employees in the firm is deemed to be a legitimate representative. However, there can be a problem of misrepresentation if the targets of dismissal are employees in higher positions, because most of the unions in Korea limit their membership to employees below a certain position. As a matter of fact, there was a court verdict in February 2000 judging it illegal to dismiss a worker in a senior position, on grounds that the union, which was the consulting partner then, could not be a legitimate representative. Therefore there should be a stipulation to select the representative of all employees who will consult with the employer on matters such as dismissal or 'honorable' retirement. Fourth, if we accept the logic that a surge in merger and acquisition (M&A) activity is crucial to Korea's economic prospects, there may be a need for a detailed stipulation or precedent making employment adjustment easier in these cases. In article 31, section 1 of the Labor Standards Act of " Recently, the relative earnings of low-skilled workers deteriorate in Korea. According to the National Statistical Office, income inequalities between the rich and poor worsened in 1999. despite the overall income increase and a government package of measures to support the middle class and the poor. The Gini coefficient of concentration, a measure of inequality among income brackets, amounted to 0.3204, the highest since 1979 (National Statistical Office, 2000). ASSESSMENT O F LABOR MARKET RESPONSE TO THE LABOR ... 149 Korea, there is a description concerning dismissal in the case of M&A. It says, "the case of M&A for overcoming business crisis is considered to have an 'urgent managerial need' for dismissal". There is no denying that this is a progress in terms of labor market flexibility if we compare the description to the previous precedent that made the employer responsible for maintaining employment in times of business acquisition. However, the meaning of this description is so equivocal that it is likely to cause conflicts between the dismissed workers of acquired firms and the employer of acquirer after the M&A deal. Furthermore, the need for employment adjustment in the case of M&A does not arise by a crisis of the firm alone. As we can see in the recent takeover of the German telecommunications company Mannesmann by Vodafone AirTouch, the British telephone company, deregulation and globalization of financial markets are promoting M&A deals even of the prosperous firms. By jettisoning peripheral or overlapping activities and focusing on a few core businesses, successful M&A accelerate business restructuring and boost profits of the firm. To make this process smooth, there should be a stipulation governing employment adjustment in the case of M&A. In particular, the employer should have some leeway for employment adjustment when there is overlapping activities and redundant workers as the result of M&A. Fifth, there should be policy for improving the profitability of TWAs. As most temporary work agencies are small and their management abilities are poor, they have far weaker bargaining positions than user firms. Therefore, they often play the role of a placement service firm rather than a temporary work agency. This lack of bargaining power creates several negative effects on temporary work services in the labor market. It leads to excessive competition on price among agencies, which then imposes lower wages on workers. It also leads to a lack of proper protection, such as social insurance and industrial safety, which must be provided to the workers under the law. In the long run this eventually leads to negative effects on user firms as temporary agencies cannot reinvest their profits and provide user firms with quality services such as the training of workers and adequate customer services. Sixth, there should be policy to reduce the possibility of the misuse of temporary work services by user firms. As we have seen, user firms tend to use dispatched workers for dealing with simple, monotonous, repetitive work and reducing labor costs rather than as temporary substitution due to seasonal and short-term labor shortages. The unions argue that user firms are using dispatched workers to replace regular workers. If this is true, although the introduction of temporary work services in Korea may help expand labor market flexibility, it may also distort the labor market for regular workers. We do not know how many regular workers have been replaced with dispatched workers. Such a misuse of temporary work services may lead dispatched workers to remain as "Temps" in the longer term during their career. A "temporary" work position could be a "permanent" job for 150 Labor Market Reforms in Korea: Policy Options for the Future dispatched workers, not an opportunity for getting new skills and job experience. It may hurt the job security of regular workers as well. It could lead to poor productivity and poor morale among workers, which may have a negative effect on firm performance and on the national economy in the long run. Seventh, there should be a deregulation of the law in terms of number of jobs and the length of temporary work services permitted. Tight regulation can restrict (good-minded) user firms in the expansion of the use of temporary work services. This may lead user firms to disguise their use of dispatched workers as subcontractors. Many firms we interviewed said that although they would prefer to use dispatched workers in many areas, they could not do so because those jobs are not permitted under the current law. For example, human resource officials in one manufacturing company argue that they would like to transform jobs currently performed by subcontractors to temporary work services. They prefer temporary work services as they can directly control work performance, which is difficult when they use subcontractors. To work properly, the society requires a balance of openness and inclusion. Up to 1997, Korea's labor market indicated the characteristics of excessive inclusion. Under this system, new employees could be con- tinuously recruited, which was helped by rapid economic growth. However, in the forthcoming era of slower growth, which Korea is entering, too much inclusion in the labor market could negatively impact on the profitability of firms and reduce the employment opportunities of new participants in the market. In this era, we need more openness in the labor market. However, we must be cautious not to swing to the opposite position by eliminating the employment stability of most workers. In the system of employment adjustment, we should try to maintain the balance of power between employers and the unions. &?Labor Market Flexibility and Employment Agencies in the UK, Japan and ~ o r e a * Barry Wilkinson" Labor Market Flexibility in the UK, Japan and Korea It seems ironic that as recently as the 1980s, when the concerns of many economists and sociologists were focussed on flexibility of labor deployment, Japanese and other Asian (and within Europe German and Italian) models were held up as exemplars for the rest of the world (Kern and Schumann, 1987; Kumazawa and Yamada, 1989; Oliver and Wilkinson, 1988; Piore and Sabel, 1984; Piore, 1986). At the turn of the century, in the context of a long period of high unemployment and slowed growth in many European countries, and severe economic problems in Japan and elsewhere in East Asia, it is the US which has come to the forefront as the model for labor flexibility the world over, and within Europe it is frequently the UK which is held to provide the model of a flexible labor force. This shift is clearly related to the apparent economic success of the US and UK in recent years, which has underpinned a resurgence of neo-liberal economic ideology. This section focuses attention on labor market flexibility at the national level in the UK, Japan and Korea, though developments in the strategies and structures of employing organizations are included as important factors. For each of the three societies, there is discussion of labor market institutions and mechanisms in their economic and political/legislative contexts. Labor Market Flexibility in the UK The election of a Conservative government under the leadership Mrs This chapter benefits from descriptive and quantitative data provided by Professor Park Duck Jay of the Korea National Open University. The author, however, takes sole responsibility for the arguments put forward. "UniversityofBath,UK. 152 Labor Market Reforms in Korea: Policy Options for the Future Thatcher in 1979 heralded a major change in approach to the management of the UK economy. Neo-liberal rhetoric was used to justify the reduction of subsidies and other forms of assistance to industry, the wholesale privati- zation of the publicly owned utilities, and the introduction of market mechanisms in the public sector. The assertion of the primacy of competition and capital markets had implications for labor markets, which were seen by the government as rigid and distorted due to excessive trades union power, and over the next 15 years or so a whole raft of legislation aimed at curbing that power was introduced (see Brown and Wadhwani 1990, and Brown et al. 1997 for detailed discussion). Also important were competition policy, the increased exposure of UK (or UK-based) businesses to international competition, and shifting firm strategies. These forces combined with legislation to produce a decentralization of collective bargaining activity, a reduction in the size of the (traditionally highly organized) public sector, and a restructuring of the economy which decimated some of the industries (for instance coal and steel) that were traditional trades union strongholds. Such developments helped pave the way for the extension of the contingent labor market. The reduced size of the UK's public sector was associated with government encouragement of competition within the remaining public sector through 'competitive tendering'. In-house activities were reduced through the contracting out of services that could be provided more cheaply and (it was argued) effectively by private sector organizations, many of which would be less likely to use unionized labor on permanent terms. In cases where the outsourcing company takes over the existing labor force the Transfer of Undertakings (Protection of Employment) (TUPE) Regulations 1981 comes into play. These specify that the workforce is entitled to the same terms and conditions as offered by the previous employer, though it is possible for the new employer to 'whittle away' any terms and conditions it considers undesirable by making new appointments on new terms and inducing existing employees to voluntarily accept such (see 'Privatizing Personnel', IRS Employment Review, no.648, 1999 for a detailed example.) International exposure to competition in product markets has also had an effect, albeit immeasurable. Manufactured imports increased as a proportion of home demand from 26 % in 1980 to 45 % in 1995 (Brown et al., 1995) forcing business restructuring and cost cutting, which often includes changing employment relations. The increase in foreign investment activity has also had an impact, with companies setting up on green-field sites able to establish new employment practices (often with the cooperation of single unions desperate for membership) which include mechanisms for flexible usage of labor (Bassett, 1986; Oliver and Wilkinson, 1988). Employing organizations have, then, shifted their employment practices in ways that increase the size of the contingent labor force. These shifts are undoubtedly related to the shift in the political climate, reduced trades union power, competition policy, and increased exposure to international com- LABOR MARKET FLEXIBILITY AND EMPLOYMENT AGENCIES. .. 153 petition. A whole industry of 'strategic human resource management' has grown out of these changes, with 'flexibility models' distinguishing between 'core' and 'peripheral' workers who contribute in different ways to firm competitiveness. Whether HRM in these organizations really is 'strategic', and whether the use of contingent labor really does contribute to firm competitiveness is, however, open to question. According to Purcell (1996, p.19): In practice this rational, ordered planning and analysisis rarely elnbarkcd on. What evidence we have is that firms adopting peripheral policies do so for short-term cost reasons and rarely link this development to core strategies. Contingent working seen in temporary work and the growth of insourcing and outsourcing has been more a reflectionof cost reduction exercises in the context of labor market deregulation and continuing high levels of unemployment. This observation raises serious questions about the real economic and social consequences of extensive use of contingent labor which will be addressed in section three. We should also comment on what the future might hold. In 1997 a Labor government was elected following 18 years of Conservative rule. The new government has vigorously asserted a continuing commitment to a flexible labor market policy, but has attempted to combine this (or for cynics to legitimize this) with notions of 'fairness at work'. Indeed it was the Fairness at Work White Paper that led to the 1999 Employment Act. This includes 'family friendly' - in particular more generous rules for parental leave on birth or adoption of a child. Overall the Act leaves employer prerogatives over flexible labor deployment intact. However, in the longer term it is possible, particularly since the UK recenily signed up to the Social Chapter, that pressures from the EU might force the UK to adopt stronger measures that circumscribe employer prerogatives. On the other hand if convergence across Europe is to occur it could be in the form of the rest of Europe moving to the UK model: German employers for instance have been suggesting they might be given freedoms similar to those in the UK, and Murray (1999) suggests a recent ideological shift in the European Commission which is more supportive of extensive use of temporary workers. There is not room in this paper to dwell on this issue, but lively discussions of the prospects can be found in Guest (1997) and Barnard and Deakin (1999). Significant grey and black elements in the contingent labor market mean there can be no definitive statistics on its size. Different methods of calculation cause problems too. For instance the Labor Force Survey relies on self-definitions of worker respondents, while the Workpluce Industrial Survey (see Cully et al., 1999) relies on responses from employing organizations. Even so, it is possible to get a feel for the characteristics and 154 Labor Market Reforms in Korea: Policy Optionsfor the Future types of contingent labor. A preliminary point is that there is evidence of ethnic and sex biases: it has been reported that eight % of women in the UK are temporary workers compared with six % of men, and 10 % of ethnic minorities are temporary workers compared with seven % of whites. Given what we know about sex and race discrimination in employment generally in the UK such a bias may not be surprising. (In the US too, women and ethnic minorities are over- represented among 'non-traditional' workers (Houseman, 1999; US Depart- ment of Labor, 1999.) The major types of contingent labor used in the UK include home workers (common in manufacturing), labor-only sub-contractors (popular in construction), casual workers (agriculture, hotels, catering, dock work), fixed term contract workers (especially important in the public sector), and agency supplied temporary workers (commonly found in an increasing number of sectors). For discussions see: Beatson (1995); Labor Market Trends (1997); Ministry of Agriculture (1999); Local Government Manage- ment Board (1998); IRS Industrial Relations Law Bulletin (no.534, 1995); Casey et al. (1997); FRES (199718). Labor Market Trends (1997) estimated that the number of temporary workers in the UK grew from 5.2 % of the working population in the mid 1980s to 7.1 % in 1996 amodest expansion. - However, according to the Labor Force Strwe); the numbers of economically active people in the UK defining themselves as 'self employed' almost doubled between 1979 and 1990 to represent 13 %of the employed workforce. A large, but unquantified, number of these self- defined 'self-employed' are in reality contingent workers in all but name. Hence the real numbers of temporary workers in the UK are almost certainly far greater, and have grown far more rapidly, than official figures suggest. Labor Market Flexibility in Japan Japan has taken a different route to labor market flexibility: There is a flexibility trade off. Concern with labor market flexibility - especially managers' ability to hire and fire at will-is strengthened in the Anglo-Saxon economies by the inflexibilityof the financial marketsthey face. Japanese firms, being more insulated from the short-term demands of shareholders, have hitherto been able to afford more 'rigid' employtnent systems from which they gain the advantage of employee commitment and cooperativeand flexibleattitudesto work. (Dore,1996) The sovereignty of capital markets and the power of shareholders in the UK and US is today being pushed as a model for the way forward for Japan (and elsewhere in East Asia). The pressure comes from international (especially US) sources but also from within Japan, which is searching for a way out of a recession that has persisted for nearly ten years. The difficulty LABOR MARKET FLEXIBILITY AND EMPLOYMENT AGENCIES ... 155 in adopting less 'rigid' employment systems and establishing an ability to 'hire and fire at will', as Dore puts it, appears to be not so much a function of trades union power (as arguably was the case in the past in the UK) but of corporate decision making structures. 'Corporate governance' has been a key focus of critics, with many commentators finding problems in 'too' close relationships between the state and business, between businesses inter- locked through shared ownership and shared senior directors, and between businesses and their employees. The proposed solution, well rehearsed in the pages of Far Eastern Economic Review and other popular Dow Jones publications, is to loosen the ties that bond businesses to a broader communitv of interests. and to focus their attention on shareholder demands (usually short-term profits and rising share prices). Boards of directors should be restructured in ways that increase 'transparency' and 'accoun- tability' to shareholders, hence giving control over the direction of companies to capital markets. Such a transformation of corporate governance in Japan could be as radical as that which was forced on Japan during the post-war US occupation, and those scholars who see Japanese employment practices as rooted in hundreds of years of culture and tradition (Whitley 1992) might rule it out as a possibility. However, there are at least two additional reasons, found in more recent history, for reluctance to change radically. Firstly, Japan has been highly successful for most of the second part of the twentieth century, such that the recession may need to get even deeper and scandals even more outrageous to precipitate a sea change. Second, established obligations - to inter-locked businesses and local communities as well as w employees - are not easily broken. In those cases where .they have been in recent years, it has generally been a last resort rather than a choice. Obligations to regular employees do not derive, as in Germany for instance, from formal employee representation on boards of directors. Rather, Japanese directors are usually selected from middle management ranks, and are expected to act as leaders of employees who monitor their affairs. Well established life-time employment expectations and 'bottom-up' ringgi decision making, combined in some companies with employee stock ownership plans, also contribute to the difficulty of 'downsizing' at will (Yamakawa, 1999; Yasui, 1999). The mechanisms supporting labor market flexibility in Japan were until recently studied by many foreign observers for the lessons which might be gained. One focus of attention was the internal labor market flexibility of 'core' employees enjoying life time employment and seniority wages within the large corporations: the common observation was that employer bene- volence fosters employee loyalty and a willingness to work flexibly in terms of job function and working hours (OECD, 1977; Dore, 1987). Employment security is offered in return for wage restraint, and competitiveness is achieved through skills development and investment rather than treating workers as disposable resources (Sako, 1997). 156 Labor Market Reforms in Korea: Policy Options for the Future A second focus was on 'peripheral' workers who work on temporary bases. Official figures show over five million 'temporary' plus over one million 'daily' workers in Japan in 1997. Additionally there were nearly eight million 'self-employed', an unknown number of whom might be temporaries in all but name. Peripheral workers are less likely to be unionised, are paid lower wages, and tend to bear the brunt of the impact of any economic downturn (Oliver and Wilkinson, 1988; Berggren and Nomura, 1997). Japan was considered to benefit from a 'dual' labor market, with core workers providing 'internal' flexibility, and peripheral workers 'external' flexibility. However, the prolonged recession finally began to take its toll towards the end of the 1990s, when the big firm sector saw some major casualties, and many more big firms are looking for downsizing on a scale which could cut into the core workforce. The main means to cope with over-staffing in most big companies for most of the 1990s have been reducing temporary worker headcounts, hiring freezes, reducing the extent of overtime worked, and extending the use of shukko and tenseki transfers (Benson and Debroux, 1997; Berggren and Nomura, 1997). Shukko refers to temporary transfers of employeei to other companies, and tenseki refers to permanent transfer to a new employer. This practice typically occurs between companies that are related by capital or business transactions, and often means larger firms sending employees to smaller firms. Several purposes have traditionally been served, including transferring technology and managerial expertise, but increasingly the practice has been used to 'shuffle off older employees to jobs in other firms with lower status and less favourable terms and conditions (Sato, 1996). However, firms are still reluctant to lay off permanent workers - dismissals remain rare as a means of employment adjustment (Genda, 1999). The famous Yamaichi Securities case may be illustrative of the obligations to permanent employees. Yamaichi had no choice but to dismiss its 7,694 workers soon after its collapse. But the firm set up an in-house working party that managed to obtain 28,000 job offers from 4,000 companies for its dismissed workers. Over 3,000 employees were successfully placed very quickly - older employees were more difficult to place. Yamaichi even helped 241 out of 279 graduates who were due to take employment at Yamaichi to secure positions at other companies (Japan Labor Bulletin, 1 May 1998). Many firms in Japan still appear to be attached to long-term employment relations, and the reluctance to abandon the Japanese employment system is shared by the Japanese state. The Ministry of Labor's (1999) White Paper on Labor is ambiguous about the wisdom of abandoning long term employment practices, expressing a range of fears including economic de-stabilization, rising unemployment benefit payments, reduced company training provision, industrial relations conflict, and loss of employee morale. The outcome is that while government and employers have been searching for mechanisms that can help labor force adjustments to changing economic structures and LABOR MARKET FLEXIBILITY AND EMPLOYMENT AGENCIES ... 157 circumstances, so far changes in both corporate,approacl?es and in labor legislation have been limited (Yamakawa, 1998; 1999). Labor Market Flexibility in Korea For the bulk of the period from the 1960s, when the c o u n t ~embarked on its j ambitious program of export-oriented industrialization, to the latter part of the 1990s, Korea enjoyed a situation of full employment. The giant conglomerates, or chaebol, which have led the economy, recruited most of their employees without job experience, and developed their human resources through company based education and training. Long- term employment relations were regarded as rational as the focus was on securing a sufficiently large workforce to meet rapid growth prospects. The ownership and control of chaebol differs from the Japanese keiretsu - in Korea family owners keep control over their vast conglomerates and have typically been described at the authoritarian end of the paternalist spectrum (Rhee, 1985; Eckert, 1990; Park, 1992), whereas in Japan, as described above, ownership is corporati~edand control is more diffuse. However, as in Japan, long-term employment systems have combined with seniority wages and enterprise based trades unions to create a situation where workers' interests became closely aligned to those of the company, making labor adjustments during Korea'stlate 1990s crisis both painful and controversial (Park et al., 2000). A great deal has been written and said about the causes of Korea's crisis and many have offered prescriptions to cure the ills. Most importantly of course the IMF has laid down loan conditions focused on corporate governance mechanisms which seek to increase the transparency and responsiveness of chaebol to capital markets. (See 11 Chong Nam et al., (1999) for detailed characterization of corporate governance in Korea and discussion of measures taken by the Korean government.) The focus here is specifically on the measures taken to ease labor market adjustments with a view to enabling smoother and quicker corporate restructuring. Following the onset of Korea's economic crisis and under pressure from the IMF, the government of Korea convened a Tripartite Commission consisting of government, employer and labor representatives in February 1998. The Commission signed an agreement on 'Social Compromise to Overcome the Economic Crisis in Korea; which included an article on 'Enhancement of Labor Market Flexibility'. Changes to the Labor Standards Act followed making dismissal for economic reasons much easier, and the National Assembly passed the 'Act Relating to Protection, etc. for Dispatched Workers'. This Act legalized temporary employment agencies and became effective in July 1998 (Park et al., 2000). It is difficult to separate out the effects of these legislative changes from shifts in companies' employment strategies, though at the very least they served an important symbolic function in signaling that en~ploymentsecurity could no 158 Lnbor Market Refornw in Koi-en: Policy Options for the Future longer be taken for granted. IJnemployment rose dramatically from around two % in 1997 to 7.4 % by the end of 1998, but Park et al's (2000) research, based on a large sample of companies across a range of sectors, suggests that dismissals were relatively rarely used by firms in their employment adjustment during the 10 months after the change of law. Only 15 %of firms in the sample dismissed workers, while attrition was used to reduce headcounts in 80 %, and sub-contracting, early retirement and outsourcing was used in 13 %. Those firms that did dismiss workers laid off an average of 51 using this method during 1998, whilst averages of over 200 per firm were removed from employment by the other mechanisms. However, 34.2 % of firms used 'honorable retirement' as an adjustment mechanism, and these firms on average shed over 50 staff through this method. It seems that 'honorable retirement' is induced in employees through a mix of retirement incentives, moral persuasion, and sometimes coercion. While the numbers of workers losing their jobs in 1999 was a little less than in 1998, the popularity of the different mechanisms remained basically the same. Korean employers may have tended to avoid straightforward dismissals on economic grounds because of felt obligations to employees, though it is difficult to distinguish this as a factor from the fear of militant trades union action - several heated industrial disputes over the size and procedures of dismissals occurred during 1998. Either way, while simple dismissals are not expected to become the predominant mechanism for labor adjustment in the near future, employment security can no longer be taken for granted by Korean workers. While corporate governance structures in Korea are quite different to those in Japan, there are some strong similarities with regard to employment practices, and both societies are going through the painful process of adjusting such practices in the face of economic crisis, global competitive pressure, and international political demands. The UK (and the US) offer a well established alternative labor market paradigm, but so far the evidence suggests there has probably been more lip service than action to implement the espoused mechanisms. We now turn attention specifically to the mechanism of temporary employment agencies, which have been offered a wide role in Japan and Korea only very recently, before focussing in on the potential costs and benefits of labor market transformation in the following section. Employment Agencies: Context and Characteristics While employment agencies have a long history in the UK, their activities have grown enormously during the 1990s. In Japan and Korea, employment agency activities were severely restricted until the last couple of years of the 1990s. This section characterizes employment agencies and agency workers in the three countries - and sets them in their economic and institutional contexts. Employment Agencies in the UK While employment agencies have been significant in the UK for at least a century, it is only in recent years (with a few exceptions) that serious attempts have been made to document the size and characteristics of the market. The figures available today, however, remain fraught with difficulties, partly related to methods of calculation, and also to what we define as an employment agency. Labor Force Survey (LFS) data, frequently used by labor economists, may underestimate the significance of employment agencies: for Spring 1998 it found there were 222,000 agency supplied temporary workers - less than one % of the employed population. Figures from a survey conducted by the Federation of Recruitment and Employment Services (FRES, the industry association) on the other hand estimate 879,000 temporary staff on agency payrolls for one sample week in November of the same year (FRES, 1997198). If accurate this would represent close to four % of the working population. Labor Force Survey data may grossly underestimate the size of the agency labor force because they are based on individuals' self-reporting of their status: many may mis-categorize themselves as employed by the client, or as self-employed. Further, agencies often treat temporary workers as self-employed when the reality may be rather different (Burchell et al., 1999). The definition of employment agency also makes for difficulty in estimating the size of the agency workforce. For instance as mentioned in section one, labor-only sub-contracting is prevalent in the construction industry, where the 'gang boss' may make the arrangement for his workers to be paid on completion of a job. We might legitimately wish to define the gang boss as an employment agent, yet construction workers hardly figure in data on temporary agency workers. The extent to which inclusion of gang bosses and other intermediaries (often in grey or black market sectors) would affect our calculations of the size of the temporary agency labor force is not known, but it would be to calculate upwards, and it would be significant (Rothwell, 1990). Exactly where the boundary is drawn around what counts as an 'employment agency' has to be a problematic, if not random affair, but a convenient one here is that used by the industry association FRES, by the market intelligence agency MINTEL, and by the British government in its 1999 consultation document (described below). Since 1995 there has been no requirement for employment agencies to be licensed by the government, so there is no definitive register. However, FRES uses Business Pages listings to make annual estimates of numbers of employment agencies, and 160 Labor Market Refornzs in Korea: Policy Options for the Fzltzrre the government's business register is used to gather detailed information on industry structure, turnover, etc. Here the industry is defined by the SIC code 74500, 'Labor Recruitment and the Provision of Personnel'. FRES data (FRES, 1998) estimates a growth in the number of employment agencies operating in the UK from 7,745 in 1995 to 10,570 in 1998. Those agencies with FRES membership have also grown, from 2,824 in 1994 to over 4,700 in 1998.In 1998 the industry directly employed over 63,000 staff, handIing nearly 900,000 temporary placements and nearly 400,000 permanent placements. An industry turnover growth of 325 % in real terms between 1993 and 1998 is recorded. Focussing on the temporary staff on agency payrolls, the extent of business derived from different industrial sectors is indicated in Table 8. 1. Table 8.1: Temporary staff on agency payrolls in the U.K.,1997-1998 - - -- Numbers ('000) (%) Secretariallclerical Blue collar Technicallengineering Nursinglrnedical/care Drivers ComputingIlT Hotellcatering Financial Professionallmanagerial Other Source: FRES Suwey (1997-1998) The high profile of certain client industry sectors in these figures is no great surprise, but it is worth noting the absence of some sectors where we know temporary working is extensive - in local government, education, and construction sectors for instance. This is explained by the predominance of alternative temporary arrangements such as short-term contracts in the public sector and 'lump labor' in the construction industry. However, it is possible that employment agencies may increase their presence in more sectors in the years to come as new opportunities arise associated with, for instance, changing labor market circumstances and changing human resource strategies of client companies, and as agencies themselves work to increase their presence in new markets. Further, while the UK's Einploy- ment Service is likely to remain the biggest job broker through its Job Centers for the foreseeable future, public-private partnerships being pushed by the present government could lead to further opportunities for employment agencies. Limited partnership activity between the Employ- ment Service and selected private sector agencies began in 1998, and could expand. The structure of the UK employment agency industry is characterized by quite extreme fragmentation due in part to low entry barriers and low start up costs. This contrasts with most EU countries where industry concent- ration is now well established, partly because EU social regulation is much tighter, making it difficult for small players (MINTEL, 1998). The largest player in the UK, Manpower, accounts for 3.7 % of total industry turnover, and only another eight agencies account for more than one % of turnover. However, acquisition activity by (especially US) foreign multinationals has been strong in recent years -they now own a third of the top twenty agencies, and most are seeking to expand their markets. Three additional forces which could encourage industry concentration may be: (i) changes to UK legislation and the adoption of EU legislation which might add to the employment responsibilities of the agencies and therefore raise barriers to entry; (ii) changes in the requirements of many large client companies who are looking to outsource whole 'non-core' functions, or to find 'suprana- tional solutions' to their human resources problems; and (iii) technological advances which are allowing employment agencies to drastically reduce response-to-client times through the use of call centers and computer software which uses large databases to automatically match the C.V.'s of workers with client company specifications. All of these developments could put the bigger employment agencies in a better position to meet client company expectations. In the UK, employment agencies are set minimum standards of conduct via The Employment Agencies Act 1973. The Act makes the distinction between an employment agency, which finds employment for workers, or provides client employers with workers, for employment by them, and an employment business, which supplies its own employees to act for, or be under the control of, others. The term 'employment agency' is typically used to cover both types of agency: most agencies in any case undertake both types of activity. Regulations made under the Act in 1976 required employment businesses to issue written terms and conditions to recruits; failure to do so could lead to a revoked license. But licensing requirements were repealed in 1994 (IRS Industrial Relations Law Bulletin, 534, December 1995). The 1975 Finance Act obliges all employment agencies to deduct tax and national insurance contributions from the temporary workers they place, which can make the agency the employer for tax purposes but not necessarily for any other purpose. As well being among the most free in Europe, UK regulations are also among the most ambiguous, giving scope for agencies andlor client businesses to enter into relationships with workers on a variety of bases - a favored one is to define the worker as 'independent contractor', which means no employment rights, and may help account for the dramatic rise in the numbers of 'self-employed'. The DTI does take complaints directly and is willing where it deems necessary to investigate and take action. Since 1996 over 10,000 enquiries per year led to 1,300 formal investigations into actual or alleged breaches of 162 Labor Market Reforms in Korea: Policy Optionsfor the Future the standards set out in 1976 regulations (DTI, 1999). Over the years disputes over employment rights (in relation to home workers, lump labor, trainees and others as well as agency workers) have given rise to four tests which are used in law to determine when a worker might be defined as an 'employee' rather than 'independent contractor' (IRS Industrial Relations Law Bulletin, 533, November 1995). The control test examines the extent of the employer's control over how, where and when work is done. The integratiovl test asks to what extent the worker or job is an Integral part of the organization. The economic realiv rest questions whether the worker is in business on his or her own account. And the multiple test takes all the circumstances of the case into account before making a judgement. Agency workers, however, are often particularly troublesome to define because of the tripartite nature of the relationship - agency, worker, and client business. (For a detailed discussion of the problematic nature of the relationship between the three parties, see Fenton-Jones, 1999). Two more pieces of legislation of relevance are the recently introduced Minimum Wages Legislation, and the European Working Time Directive which came into force in the UK in October 1988. Minimum wages legislation applies to agency workers in the same way as to all employees - the impact is not clear, though MINTEL (1998) claims most agency workers were already paid above the minimum. The Working Time Directive specifies three weeks annual holiday after a three-month qualification period, and working hours limited to 48 hours per week (though there is provision for employee opt-out). There is also provision on rest breaks and night working. The Directive applies to temporary and contract staff working through agencies, many of who have had to renegotiate contracts with their temporary workers and clients (and indeed their own direct employees!) (MINTEL, 1998). The biggest legislative impact on the employment agency industry could arise out of the adoption of the European Social Chapter, to which the present government is committed to implementing after signing up in June 1999. This puts pressure on the UK to bring its employment agencies in line with mainland European practice by making them liable for the full costs of employment protection, including redundancy. However, before the govem- ment signed the agreement, the European Commission had shifted from its previous position that temporary work was a low quality form of employment which should be limited to certain strictly specified circum- stances, to one which asserts a 'symmetry of need' for temporary work between employers and workers (Murray, 1999). In other words, EU demands on the UK (and some other European countries) could have been far more extensive and radical in the absence of an ideological shift regarding temporary work. Whether such a shift will be maintained remains to be seen. Perhaps partly in anticipation of EU pressure, one month before signing the Social chapter the government's Department of Trade and Industry LABOR MARKET FLEXIBILITY A N D EMPLOYMENT AGENCIES ... 163 published a consultation document on The Regulation of the Private Recruitment Industvy that recognizes legal ambiguities and the scope for abuse of agency workers (and indeed client businesses). This document, which came out ofthe Fairness at Work White Paper, will form the basis for the introduction in early summer 2000 of a new set of regulations to govern the industry as part of The Employment Relations Act 1999 (the White Paper and the 1999 Act are described briefly in Section one ofthis report). Like the Fairness at Work White Paperj the consultation documentation on the regulation of employment agencies begins from the premise that 'fairness in the workplace' must go hand in hand with the continued pursuit of labor market flexibility. The document states (p.10) If the industry is to continue to expand, it needs a modem legislative framework that ensures that the high standards set by reputable bureaux are not undermined by cowboy operators interested only in cutting comers. Four objectives intended to inform the regulations are laid out as follows: (i) increasing clarity: in particular, client businesses should be given unambiguous terms which specify who is engaging whom on what basis, and work seekers should be given clear information about their employment rights; (ii) promoting labor market flexibility: outdated regulations which unnecessarily restrict employment agencies' activities should be removed, and agencies should not issue contracts to workers which restrict their flexibility; (iii) proper proteclion for clients and the general public: agencies should themselves ensure the suitability and qualifications of workers - especially for pilots, doctors, care workers, and other groups who have significant social responsibilities; also for workers being placed in jobs which could be hazardous without the correct training and experience; (iv) curbing payment abuses: the regulations must ensure worlcers are paid fully and on time for their work. In addition, the document makes it clear that under normal circumstances temporary workers supplied to a client by an agency should be treated as employed by the agency. The final regulations are still to be seen, but it seems unlikely the government will stray from the principles established in the consultation document. How far they might go towards curbing malpractices remains to be seen, but it appears that the employment agency indrrjtry will have to make significant changes. If 'fairness at work' really could be established at the same time as labor market rigidities are reduced even further, this would be considered quite a feat. Enzployment Agencies in Jupan In Japan employment agencies are known as worker dispatching agencies (WDAs). Because of the harsh exploitation of dispatched workers before the Pacific war, WDAs were illegal right up until 1985, when they were 164 Labor Market Reforms in Korea: Policy Options for the Futtlre legalized for 26 allowable occupational categories. Legalization followed strong political controversy (Araki, 1994). The number of dispatched workers then grew from 145,000 in 1986 to over 700,000 in 1996, a little over one % of the Japanese working population. The categories allowed meant that most of these dispatched workers were female white collar. In 1999 restrictions on allowable occupational categories (there were only 26) were abolished through an amendment to the Worker's Dispatching Law - suddenly only a few categories where serious exploitation was feared (notably including construction, security work and port work) were prohibited (Araki, 1999). The extension of allowable WDA activity was the result of government attempts to activate an external labor market which was seen as impeding the flow of workers between businesses during economic restructuring. WDAs were seen as a key means of providing the rising ranks of unemployed in the mid to late 1990s with employment opportunities and career development choices. (Around the same time private employment placement services were given legal permission to extend their activities, with the same object of improving labor market. flexibility in mind.) As in the UK, the WDA is expected to be the employer of the dispatched worker while they are under the direction of the client company,-though (also as in the UK) there are cases where the client company has been held up in law to be the employer (Yamakawa, 1995). The revisions to Worker Dispatching Law in 1999 were preceded by heated debates in the tripartite Employment Security Council and in the Japanese Diet, the outcome being a compromise whereby WDAs were given the go-ahead to extend their activities at the same time as regulations governing their operation were tightened. Fears that client firms might use dispatched workers to replace permanent workers were addressed through a one-year limit on their use in the same post at a client company. Further, under the new law client companies are legally obliged to offer dispatched workers the same working environment as their own employees, and dispute resolution and personal information protection mechanisms are made tighter. The growth of WDA activity is likely to be substantial, but tightened regulations associated with legalization - such as restrictions on renewals of dispatching contracts - could conceivably limit the emergence of a 'hire and fire mentality'. While we await evidence of the impact ofthe new legislation and the response of client firms to the new employment possibilities, we can only speculate on the likely consequences for the Japanese employment system. Employment Agencies in Korea Prior to the legalization of temporary work agencies (TWAs) in Korea in 1998, the government had turned a blind eye to grey market TWA activities. Surveys by the Korea Labor Institute prior to legislative change suggested that while employers claimed positive effects on labor market flexibility, LABOR MARKET FLEXIBILITY AND EMPLOYMENT AGENCIES ... 165 dispatched workers were provided with far poorer wages, social protection and employment terms and conditions than their permanent counterparts. (Park et a/.,2000). The number of agencies was not known. The Act Relating to Protection, etc. for Dispatched Workers became effective in July 1998. This rules out ambiguity over employment status by specifying that the TWA, not the client company, is the legal employer. The TWA must gain approval to operate from the Ministry of Labor, and this approval has to be renewed after three years. There are restrictions on the type of work dispatched workers may be used for, and on the length of contracts (one year renewable for a maximum of one further year), and there is requirement for consultation with the trades union prior to engagement of temporaries under some circumstances. Various iegal safeguards to prevent abuse of dispatched workers are included in the Act. and at the same time the Act provides guards against the use of dispatched workers to replace permanent employees. By the last quarter of 1998, a few months after passage of the Act, 789 TWAs had gained legal permission to operate from the Ministry of Labor, increasing to 1,095 by the second quarter of 2000. (A little over a quarter of these were not currently operating.) The nurnbers of dispatched workers accounted for by registered agencies grew from 41,545 in the last quarter of 1998 to 46,407 in the second quarter of 1999, representing 0.23 %of Korean employees. While the average size of TWAs is very small, a few agencies have been growing rapidly, and agency industry concentration via mergers and acquisitions is expected. Park et al.'s (2000) survey of 483 firms found 37 % of the sample using dispatched workers, with 43.9 % planning to introduce or extend their use. Larger firms in banking, retail and other services were most likely to use them, though they were commonly found across a range of sectors and firm sizes. Employers were generally positive about the effects of TWAs on labor market flexibility, but felt the periods of temporary work allowed and the limitations on job types were too restrictive. Union ofticials, on the other hand, felt the regulations provided insufficient protection for agency workers, and that permitted jobs were too comprehensive. While it is clearly too early to attempt to judge the impacts TWAs will have in Korea, it is clear that employers are likely to gradually extend their use. However, as in Japan (but not yet the UK) the adoption of tight regulations under pressure from the labor movelnent could limit their usefulness to firms wishing to practice 'hire and fire' policies. Flexible Labor Markets and Employment Agencies: Economic and Social Consequences Two popular arguments are used to legitimate elnployment agency activity, 166 Labor Market Reforms in KO?-earPolicy Optionsfor the Fzltzrre and the promotion of temporary working generally: 'more employment' and 'better employment' (Murray, 1999, p.273). 'More employment' and 'better employment' need to be broken down a little further, and in this section various evidence and argument is examined in order to pass comment on the potential economic and social consequences of the expansion of temporary employment agency activity. In particular we examine the consequences for: (i) overall employment levels and labor adjustment; (ii) skills development, innovation capability and quality; (iii) labor rights and representation; and (iv) worker well being. Evidence is taken primarily from extensive research undertaken in the UK, and this is used to point to the issues which may need to be addressed in Korea (and Japan) as temporary work agencies gradually extend their activities. Employmenf,Skills, Rights and Well-being It is frequently argued that labor market flexibility has beneficial effects on employment by improving the prospects for job creation (and also low inflation and productivity growth) (OECD, 1986). Indeed this was a central tenet of the Conservative administration in the UK which, despite its poor record in the preservation and creation of jobs for most of its tenure, could report a consistent reduction in unemployment whilst maintaining low inflation during its last three years in office in the mid-1990s. Further, it was claimed the UK emerged from recession before many European countries in the 1990s because employment legislation had made labor adjustment easier, so that employers were more willing to take on extra labor (Guest, 1997). Recently the European Community has begun to move from an active labor market intervention stance (Potts, 1999) towards an acceptance of a relationship between flexible labor deployment and the achievement of high employment (Barnard and Deakin, 1999). Employment agencies are seen to play a key role in encouraging and enhancing flexible deployment by matching workers to companies at the same time as handling employment obligations - effectively agencies help reduce the costs and risks of hiring and firing labor to the client firm. However, there is little direct evidence on the effects of flexibility on employment creation, and what little research has been done often uses dubious proxies such as presence of a trades union or presence of temporary workers as indicators of flexibility within sample firms (Haskel et al., 1997). Further, separating out the effects of flexible employment practices from other variables is a difficult task, such that results of studies, whether positive or negative, are always likely to be open to alternative interpretation. Haskel et al.'s own research suggests that labor input (in terms of hours worked and/or numbers employed) is more responsive to demand change in a flexible labor market. This means that 'whilst flexible firms are more likely to create jobs in upturns they are more likely to destroy them in downturns' (p.375). More flexible firms find it less costly to adjust the size LABOR MARKET FLEXIB~LITYAND EMPLOYMENT AGENCIES. .. 167 of the workforce and therefore are more responsive to business cycles, and this may help explain why e~nploymentpicked up much quicker following the early 1990s recession compared with the early 1980s recession (Beatson, 1995). This research does not tell us, however, about the long-term effects on overall employment levels, which remain subject to debate in the absence of systematic evidence, and the consequences for productivity remain unclear. While Beatson (1995) found some evidence of an association between the extension of flexible working and a better productivity performance in the UK, a European Comlnission Expert Working Group concluded that 'when employment relationships are more precarious, motivation and willingness to cooperate are likely to be lower, with negative consequences for firms' productivity' (Bosch, 1994, p.26). Nor does it tell us about other potential benefits or costs to national economies, to which we now turn. Evidence with regard to the relationship between labor force flexibility, skills formation and product and service quality in the UK is mostly negative. Arulampalam and Booth's (1998) study of flexible employrnent and training concluded that the expansion of flexible forms of employment was negatively related to work-related training. The Institute of Management and Manpower (1996) found that the use of flexible manpower was nlostly driven by cost considerations, and one of the costs that could be minimized was training. Casey et al. (1999, p.76), in a review of research findings in the area, found that in many cases managers have sought to employ flexible workers precisely in order to distance themselves and the employing organization fi-om the individual employees and to minimise the orgaliization's obligations and commitments towards that worker. It also appears to be the case in the UK that temporary work often goes hand in hand with low cost product strategies. Casey et al. (1999) use the examples of the hospitality sector and services contracted out from the public sector to make the point. In both .cases, temporary labor is chosen becazlse it helps cuts costs: once chosen, it is difficult to move to higher value offerings because it produces conditions 'that render the sophisticated, developmental people management systems the high performance model demands almost impossible to achieve' (p.76). Not dissimilar findings have applied to UK manufacturing too. A review of the evidence available by Ackroyd and Procter (1998) led them to conclude that large UK manu- facturing organizations were not using the high value, high trust flexible firin model as advocated by the EU and OECD. Nor were they developing product market strategies, employment practices and people management systems that would provide for the emergence of such a model. Relatedly, it has been argued that the removal of labor market rigidities in the long run have a detrimental effect on the product and process innovation capabilities 168 Labor Market Reforms in Korea: Policy Optionsfor the Ftlture of firms (Kleinknecht, 1998). There are a few exceptional areas where highly skilled and professional workers such as IT consultants, interim managers and research and development workers carry scarce skills between organizations, often via employment agencies. For these workers temporary work is often taken out of choice rather than necessity: they benefit the client organization with their specialist skills, and they gain a degree of control over their own working time and income. Flexible work practices do not have to be associated with low cost, low value strategies, but the evidence to date suggests this is the route most UK employers have taken. The present Labor government perhaps recognizes it must play some role in guiding employers along desired paths. Tregaskis (1997) is skeptical of the prospects for individual workers taking on responsibility for their own skills development, and suggests the state should extend its 'public-private partnership' ideas more firmly into workforce skills development activities, with 'training partnerships' between employers, individuals and the state. This might help improve individuals' future 'employability', eventually creating a virtuous circle whereby the organization will have to provide learning opportunities for highly skilled candidates. It is also feasible that the forthcoming Employment Agencies legislation could have some impact, at least for those employees who gain temporaiy jobs through the agencies, though developments here are difficult to foresee. However, in the absence of at least some government interven- tion, it seems likely that the flexibilization of the UK labor force will continue to 'undermine the national policy objective of securing the upskilling of the UK workforce' (Casey et al., p.76). Employee representation and contingent labor rights have been found to be problematic issues. Union membership among 'flexible' workers has been found surprisingly high at around 30 %, though problems with defining 'flexible workers' (see section one) might mean this figure is an over- estimate. Probable reasons why unionization is higher than might be expected are that many types of 'flexible worker' are most common in larger workplaces where union density is highest (Labor Market Trends, 1998) and that unions in the UK actually have a reasonable tradition and good experience of recruiting contingent labor - in hospitals, docks and taxi driving for instance (Croucher and Brewster, 1998). For many of the newer types of flexible worker, however, unions face particular difficulties - of recruitment, communication, and provision of cost effective services. Temporary agency workers who occasionally or frequently move clients are particularly difficult to deal with. Their dispersal means unions are rarely in a position to offer them collective bargaining services, and frequently it is not clear whom they should be bargaining with (Croucher and Brewster, 1998). The latter problem might be partially resolved with the changes to employment agency legislation due in summer 2000 (see section two). At the same time as trying to find ways to extend representation of the contingent labor force, unions have been vocal politically about the spread of temporary working. A motion supported by the General Council at the Trades Union Congress of 1994 noted that Agency workers...are continuously denied employment rights as they are moved on to different workplaces. In addition, agency workers are often denied all the pay and conditions benefits provided under collective agreements and opportunities for training provided and/or paid for by their employers. (TUC, 1994) The motion included a call for the reform of statutory employment law to provide full employment rights. It should be emphasized, however, that the trades union movement is not against temporary working per se. TUC Secretary General John Monks has recently commented that workers must be flexible in a 'twenty four hour society' with round the clock services. The point is that employees should have more choice and flexibility themselves to determine times worked: employers must keep their side of the psychological contract (IRS Employment Review, no.163, December 1999). In reviewing research and literature on the implications of contingent working for the psychological and social well-being of workers it quickly becomes apparent that the negative viewpoint is most strongly represented. However, there are those who point out significant potential benefits to workers. The two benefits most commonly cited are contingent working as a means to enhance lifestyle, and as a means to work whilst continuing family care responsibilities (Bronstein, 1991; Osawa and Kingston, 1996). The former benefit is largely restricted to those with scarce skills who can develop more than one 'income stream' which enables them to gain control of their working time (a relatively small but growing proportion of the UK workforce). The latter relates primarily to part-time work, which, it is almost universally agreed, does benefit parents with families by allowing both parents to undertake some paid employment. Of course much part-time work is also permanent, and is not necessarily of the precarious kind found in other categories. The government's view is that temporary working can and should be 'family fiendly7,and Wood (1999) has produced evidence that in certain types of organization 'family friendly management' is being institution- alized. These are large firms that place a high value on employee welfare and see a 'bottorn line' in adopting family friendly policies and gaining workforce commitment. The consensus, however, is that such firms are thin on the ground. An Institute of Personnel Development study (Stredwick and Ellis, 1998) describes flexible working as often a euphemism for chronic instability and exploitation: 'the threat of flexible working lies in the increased potential for employee exploitation, unregulated sweated labor and the end of the meaningful long term employment relationship'. And on the 170 Labor Marker Reforms in Korea: Policy Optionsfor the Futtire basis of its extensive of handling temporary workers' grievances the independent National Association of Citizens Advice Bureaux (1997) comments that: 1;or many (employers) flexibility means withdrawal from the responsibilities associated with the employment relationship. From this perspective, basic employment standards such as protection from unfair dismissal and decent rates of pay are seen as constraints on flexibility, and so damaging to the health of the economy. Purceli et a!. (1999) found positive benefits for employees where flexibility was 'sensibly managed' and not exploitative, but that the costs of temporary unstructured employment outweigh the benefits unless employees have scarce skills and can be involved in choosing when and where to work. In sum, if the published evidence provides a reasonably accurate portrayal of worker well being under conditions of temporary work, there is a way to go before 'symmetry of need' between employers and workers comes into existence. Policy Issues Among other forms of contingent working, temporary work agencies have been accepted as a legitimate means of enhancing labor market flexibility in furtherance of economic and societal objectives in Korea (and Japan). While it appears unlikely that the principle of temporary work agency activity will be challenged in the foreseeable future, policy issues related to their use will be carefully debated in the light of agency behavior and in the light of the experience of agency activities from the perspectives of the state, employers, labor and trades unions. Following discussion of the consequences of labor market flexibilization and temporary work agencies in the UK, the four key areas which will become the focus of concern are clear. Firstly, do temporary work agencfes help employment creation? Evidence from the UK and elsewhere is ambiguous. It is generally accepted that agencies can contribute to the efficient matching up of labor supply and demand. Park et al. (2000) found that Korean employers were positive about the effects of TWAs on labor flexibility, effects which they suggested would be even more positive if there were fewer restrictions on the use of dispatched workers. However, the contribution of labor market flexibility generally, let alone specific flexibility mechanisms, to long term employ- ment creation remains unknown as problems of measurement and methodology dog research. Until these problems are resolved, this key policy question will remain the subject of ideological debate in the absence of unambiguous evidence. L A B O R MARKET FLEXIBILITY A N D EMPLOYMENT AGENCIES ... 171 Second, what is the relationship between employnzent agency activity and skills formation? Here most of the evidence from the UK (and US) strongly suggests a negative relationship except for a small minority of highly qualified agency labor such as consultants. interim managers and research staffs, who develop and update their portfolio of skills as they move between assignments. So far in Korea, most agency labor has been employed in unskilled jobs which demand relatively little training (Park, 2000) such that the impact on skills formation is likely to be neutral for the employer and negative for the employee. The question then becomes how to establish the conditions whereby a positive relationship can occur. 'Training partnerships' between the state, employers andlor agencies, and labor have been mooted as a way forward in the UK, though we await developments on this front. In the meantime the objective of upskilling the UK workforce may continue to be undermined by short tenn employment practices and 'hire and fire' policies. Korea needs to avoid such a development. Third, what are the employment rights of contingent workers generally and temporary agency labor speczjically? In the UK employment status is highly ambiguous and many agency workers are unclear even as to who their employer is. The current government has recognized this as a serious problem and promises to address the issue in forthcoming legislation. The new legislation on agency work in Korea (and Japan) is rather less ambiguous on this front, though it will be important to monitor the situation carefully to prevent abuses by unscrupulous agencies or employers. Park et al. found that Korean employers felt the level of legal protection afforded agency workers was adequate, whereas Korean union representatives felt there was a need for tighter legislation on working conditions and a need for measures to prevent 'wage appropriation' by agencies. Employment rights for temporary workers will undoubtedly be a policy issue for Korea in the years to come. Fourth, how can a balance between employer and employee needs be achieved? Contingent working as a means to enhance lifestyle or combine work with family care responsibilities is frequently argued. But several studies have demonstrated that for the majority of contingent workers in the UK such benefits are imaginary. While 'family friendly' employment policies are now on the UK (and Japanese - see Sato 2000) political agenda, the whole rationale of employers in their use of temporary workers is to do with their own flexibility. (tt is worth noting here that Korean temporary worker well-being may be less likely to appear on the national political agenda than is the case in the UK because far fewer. temporaries are unionized and they have little by way of other means of representation.) Conceivably temporary work agencies could play a role in matching employees' needs for employment flexibility with those of employers, but the evidence suggests that at the moment it is the agency workers who have to make the adjustments. Industry concentration which could enable the emergence of larger and more professional employment agencies could be a 172 Labor Market Reforms in Korea: Policy Optionsfor the Future basis for changing this situation (Park et al., 2000), but evidence from the US, where large professional outfits already dominate the agency market, would suggest this development in itself would have little impact (Houseman, 1999). Conclusion In accepting IMF loans Korea has been pressured into accepting reforms of its systems of economic and corporate governance, entailing among other things significant changes in labor market adjustment mechanisms. In the space of less than two years employment security for core workers has been drastically reduced if not curtailed, and the national labor market has been sigiiificantly freed up. The speed of change is reflected in the fact that temporary work agency activity was liberalized sooner than in Japan, which had been debating extensions of the role of dispatched workers under recessionary conditions for several years previously. While consensus among the interested parties was gained for the principle of labor market flexibilization and the adoption of employment agencies through Korea's Tripartite Commission, accelerated implementation of new worker dispatch- ing laws under crisis conditions in Korea meant little time for considered internal debate over its potential costs and benefits, and how these might be dealt with. As employment agency activity grows, further consideration in the light of solid documentation needs to be given. The experience of the UK could be a good starting point. While the results of research are in some respects inconclusive, and while lessons from one society are not necessarily easily transferred to another, the areas of serious concern and therefore in need of careful monitoring are quite clear. It is hoped this paper has pointed clearly to these areas and thereby might contribute to an important policy research agenda. ACTIVE LABOR MARKET PROGRAM 9Income Assistance and Employment Creation through Public Works in Korea Joohee Lee' Instruction Public expenditure on labor market programs was almost negligible in Korea before the financial crisis in the winter of 1997. Throughout the 199Qs, Korea spent less than 0.1 % of GDP on these programs, one of the lowest rates in the OECD. The relatively small amountof spending reflects the low level of unemployment (OECD, 1998, p. 148). Compressed industrial trans- formation in Korea has constantly created new jobs in the nonagricultural sectors of the economy, and as a result Korea was accustomed to near full employment for several decades. This picture has significantly changed after the crisis. The growth in employment was radically reversed by the severe economic recession in 1998. Due to the recession the unemployment reached 6.8 % in 1998, and 6.2 % in 1999. Although still lower than the average OECD spending, Korean public expenditure on unemployment measures expanded to 1.26 %of GDP in 1998, and to 1.87 % in 1999 (Jeong et al., 2000). One notable feature of Korean public spending on unemploy~nent measures was that it struck a balance between passive income support and more active measures' that were designed to improve access to the labor market for the unemployed and to improve the employment situation. Except for a few countries, the share of spending on passive measures was far larger than that on active measures in most OECD countries. On the one hand, it is a promising sign that Korea is in line with the general principle of ' KoreaLabor Institute. ' For analytical and policy purposes, the OECD splits public spending on labor market programs into 'active' and 'passive' measures. Active measures comprise a wide range of policies aimed at improving labor market access for the unemployed and the employment situation, job-related skills and the functioning of the labor market (Martin 1998). 176 Labor Market Refornzs in Korea: Policy Optionsfor the Ftrt~rre shifting public resources from income support to active labor market policies, a shift that has been endorsed by many Labor Ministers in developed countries (Martin 1998). On the other hand, however, social assistance programs in Korea are in general underdeveloped, with numerous cut-off levels for eligibility and restrictions in entitlements. As Korea lacked broad based social assistance programs, the situation was quickly reached where Korea had to accommodate a number of the unemployed who were not eligible for unemployment and at the same time not sufficiently poor enough to be covered by livelihood protection programs. It is in this context that public works projects emerged as a crucial une~nploymentmeasure in the overall system of income support and social protection. In 1998, over 1 trillion won was poured into the projects, and in 1999, about 2.5 trillion won (Lee 1999). Although officially classified as an active measure, these schemes mostly played an important role as complementary income support in Korea. Even in developed countries it is a common phenomenon that many active measures tend to become inad- vertently 'passive' during periods of high and persistent unemployment as there is more emphasis on the social objectives of the projects than on their efficiency objectives (OECD 1998). This is the reason that Korean public works projects should not be solely evaluated based on the net effects of programs on aggregate employment and unemployment. This research therefore attempts to evaluate the distribution effects of the projects as well as employment effects. As some projects hastily made by the unprepared municipalities were not always effective or useful to the public, the private sector was increasingly critical of the projects, regarding them as "unnecessary make-work".2 Actually, most evaluations show that subsidized employment schemes are ineffective due to considerable dead-weight, displacement and substitution effects (BECD 1998). However, the system-wide effects ofthe public works projects, which were not easily measurable, could be more beneficial than expected. The projects boosted and created effective demand, and provided much needed services to the disabled, the elderly and children of low- income families. Thus whether the projects bear beneficial outcomes, along with unexpected externality effects, is also investigated in this paper. The remainder of this paper is organized as follows. The next section provides a detailed description of Korean public works projects. It is followed by sections that seek to advance an analytical discussion regarding the various dimensions of project evaluation, and explanations of the research methods utilized to conduct project evaluation. The majority of empirical work in the result section is devoted to the impact of project New Deal emergency relief projects were similarly criticized in the U.S. For more informa- tion on this issue, see Rose (1994). She argues that inefficiency was built into the public works projects because these projects were mandated to use a maximum of human labor and a minimum of machinery. EMPLOYMENT CREATION THROUGH PUBLIC WORKS 177 participation on individual workers, but it also briefly engages in a broader discussion of project effects on the national economy. In tlhe final Section, we conclude this study by recapitulating the findings and discussing future policy agenda, with special emphasis on their relationship with the Basic Livelihood Security System for the Nation to be enacted in October 2000. Project Profile Public works projects were implemented as a two-stage project in 1998 (Phase I: from May to August; Phase 11: from September to December), and as a four-stage project in 1999 (Phase I: from January to March; Phase 11: from April to June; Phase 111: from July to September; Phase IV: from October to December). As shown in Figure 9. 1, the projects consumed a substantial portion of the entire budget for unemployment measures. The public works projects generated about 400,000 jobs on average during the year 1999. Figure 9.1: Budget for public works projects in Korea, 1998-2000 (won in millions) --- -- 1130ther unemployment measures UPubl~cworks project - In general, the public works projects fell into four broad categories. Infrastructure works(kongkong saengsansung)projects involved forestation, new construction of small public facilities such as community parks, and repairing public property. Public service (kongkong service) projects provided temporary workers to public organizations and community welfare service centers. These public workers engaged in a variety of activities, which ranged from managing cultural assets in national museums to teaching children from low-income families in after-school classes. Maintenance (HwankytlngChtlnghwa) projects were mainly composed of such activities as garbage coIlection and lawn maintenance in national parks, 178 Labor Market Reforms in Korea: Policy Optionsfor the Fzztzlre snow removal, and street cleaning. Information Technology (Churigbohwa) projects, added to the public works programs in 1999, can be classified as professional or nonmanual projects. Relatively young and educated workers were included in this category, and they mostly constructed database, and provided assistance on resolving the year 2000 computer problems. In 1998, those who wanted to participate in the programs had to be une~nployedpersons between 15 to 65 years old. In Phase 11 of 1999, age restrictions were tightened to include only those between 18 to 60 years old, which brought about wide-spread complaints from the old unemployed about the selection procedure. At last, those between 61 to 65 were allowed to participate as long as they were within the 5 % of the total selected participants in Phase 111 of 1999. Selection criteria are different depending on the skills and credentials required by the particular projects. But priorities were given to the head of the household, those in their prime working age, i.e., 30 to 55 years old, and the poor unemployed. Since Phase 111 of 1999, screening decisions have been made based on the aggregated scores given to the following 9 factors: age, householder status, number of dependents, assets, female householder status, handicapped, duration of unemployment, the first-time participant status, and household income. Changes in wage rates for the public works projects are shown in Table 9. I . There were no differentials in wages by sex or by age. At the end of Phase 11 1998, wage rates were highest, paying about 35,000 won per day to those performing professional work, and 25,000 won to low-skilled manual workers. Determination of the wage rates was controversial process as the business sector argued for lower rates. This criticism has led to the reductions of wage rates, once in Phase I1 of 1998, and again in the beginning of the Phase I 1999. These rates were maintained through Phase I 2000. As wage rates had been reduced, the projects began to attract more and more disadvantaged workers in the regular labor market. Table 9. 2 demonstrates this point. Women, less educated, and older workers dominated the projects by the end ofthe Phase IV, 1999. Table 9.1: Changes in wage rates for public works projectsin Korea 1998 1998 Oct. 1, 1998 1999 Phase l Phase II (won) Low-skilled 20,000 22,000 22,000 or less 19,000or less office work Manual labor 25,000 25,000 22,000 or less 19,000or less Labor-intensive 30,000 27,000 or less 24,000 or less manualLabor Professional 35,000 32,000 or less 29,000 or less work E M P L O Y M E N T CREATION T H R O U GH P UBLIC WORKS 179 Table 9.2: Characteristics of the public works project participants in Korea, 1998-1999 -- .. . P 1998 1999 Stage 1 Stage 2 Stage 1 Stage 2 Stage3 Stage 4 lo/"\ (Oh\ Sex Male Female Education Primaryschool or less Middleschool or dropouts High schoolor dropouts 2-yearcollege or dropouts Universitydropoutsor higher Age Under20 1.2 0.7 1.O 0.9 1.3 1.2 20-29 12.0 8.8 11.6 13.5 16.0 13.0 30- 39 20.7 19.6 22.7 21.4 18.8 16.6 40-49 26.1 26.9 32.4 30.4 27.2 26.2 50-59 27.6 30.6 27.4 22.5 29.0 37.3 60-65 12.4 13.4 5.0 11.2 7.8 5.8 Principle Objectives of the Public Works Projects Public works projects in Korea have not simply been job creation programs. As in other debt-stricken developing countries, poverty as well as unemploy- ment became a much greater problem. This problem was exacerbated by the fact that Korea lacked a pure income maintenance program for the unemployed. The Korean government expanded the existing livelihood protection in order to provide temporary relief for the unemploy- ed that did not qualify for unemploylnent insurance. In order to be eligible for the original program, one should have income less than 230 thousand won per month, must pass an asset test, and must have no family that can support them. This prograin was rnostly for those who were unable to work, such as the handicapped, elderly, and children. The temporary livelihood protection program eased some of its strict means-test criteria to cover additional people, however, it is still too limited in its coverage and Korea introduced social assistance long before it introduced programs of universal social insurance. A case in point is the livelihood protection program. created by livelihood protection act in 1961. 180 Labor Market Refornzs in Korea: Policy Options for the Fzrtzli*e generosity to complement unemployment insurance and to substitute for unemployment assistance. Among those covered under the program, more than two-thirds were classified as the "self-supported," for whom cash benefits were not available except during winter. Thus the expanded public assistance scheme in Korea was not well equipped to function as an ultimate safety net of income support for the newly unemployed. This was the reason that the public works projects had to offer short-term income opportunities to those newly unemployed. Therefore whether the public works projects actually cover the major target groups with high unemployment risks: especially the low skilled, the long- term unemployed, and the poorly unemployed heads of the household is an important indicator that measures the distribution effect of the public works projects. It is also necessary to explore the economic status of women participants, as they were frequently suspected of being econolnically inactive people who were not eligible for the projects. If these participants turn out to be the actually unemployed, and non-beneficiaries of the other unemployment measures, the distribution effect of the program can be more positively evaluated. The major goal of the direct employment creation projects was to increase the available number ofjobs while satisfying the needs that were not met by the private sector. To put it otherwise, public sector job creation should not compete with or substitute for existing jobs. Although the direct effect on employment creation seemed significant, the literature on labor market policy and evaluation (Forslund and Krueger 1994; Calinfors 1994) has cautioned that the net impact of job creation, that is, gross impact minus deadweight, displacement and substitution effects, could be much less than expected. Furthermore, public sector job creation could produce indirect employme~~teffects that may or may not be favorable for labor market outcomes of project participants. In order to assess such net impact of public works projects on job creation, the following concerns must be fully explored. First of all, labor market outcomes for participants are not always positive. Participation in the projects may stigmatize them, reducing reemploy~nent opportunities and future earnings. Second, the net employ~nentgenerated by a public employ~nentprogram decreases in proportion to the incidence of Freedman (1990)explains why public employment programs often fail to meet the eligibility requirements as follows. First, those who suffer the most serious economic and educational disadvantages may be the rrlost dificult to reach for reasons of access to information about and personal attitudes towards the programs. Second, programs and project organizers are often concerned with producing rcsults that look good on paper, and may tend to favor participants who are likely to fare well in the programs. Third, where programs offer employers a subsidy to hire members of the target group, it is likcly that "creaming" will occur, i.e., employers will seek to minimize their risks by selecting those participants who appear better qualified. EMPLOYMENT CREATION T H R O U G H PUBLIC WORKS 181 deadweight losses, or fiscal substitution,' such as when a local administra- tion takes advantage of a federal subsidy to create a job that administration would have created otherwise (Erhel et al., 1996). Such an effect is only indirectly investigated in this study. Programs targeted on the disadvantagcd would be less vulnerable to fiscal substitution because these target group workers are seldom qualified to do a job that required regular employees. Third, the employment creation effects of the public works projects diminished if the projects made participants continuously dependent on the projects. This problem is called the locking-in effect. It is therefore necessary to investigate whether participants want to continue their partici- pation and how long the participation would last. Finally, subsidized employment schemes such as public works projects may have a number of objectives other than creating additional jobs. They may enhance effective labor supply by helping individuals to keep in contact with the world of work, thus maintaining their motivation and skills (OECD, 1998). Again, this effect cannot be directly measured, but it is necessary to investigate whether project participation enhances willingness to work by surveying the attitudes of the participants. Research Methods The empirical research reported in the following section was based on a survey of project participants, conducted during September-October 1999. The sample for this survey consists of adults who participated in the public works projects during 1998 and 1999. Respondents were sampled with a multi-stage cluster sampling procedure. The integrated list of those who participated in central government projects in 1998 was not available, so they are excluded. This exclusion sho~~ldnot lead to a serious bias in analyzing the data, as public works projects in general were most extensively implemented in 1999, and a majority of participants worked.for local government projects. There were also no compelling reasons to believe that those who participated in 1998 central government projects differ significantly from those participated in 1999 projects. Those who participated in 1998 local government projects were sampled based on local government participant lists, and those who participated in 1999 central and local government projects were sampled based on the lists obtained from the central employment information institution of the labor ministry. Among 1578 cases selected, a total of 1505 cases were collected Schmid (1996) argues that whereas efficiency is always impaired Iby deadweight effects, d~splacementand substitution might be intentional. Preferential treatment for one person at the cost of another -promoting women at the cost of mcn, long-term employed at the cost of short-term employed- can be an explicit social policy. 182 Lubor Murkel Refornzs in Korea: Policy Optionsfor the Fzltzlre via personal interviews conducted during the summer of 1999. Regional allocation of the samples are as follows: Seoul (274), Kyoungki (289), Inchon (52), Kangwon (84), Pusan (144), Ulsan (33), Southern Kyungsang (107), Taegu (105), Northern Kyungsang (1 I;), Taejon (50), Southern Chungchung (94), Northern Chungchung (45), Kwangju (48), Southern Cholla (SO), Northern Cholla (60). Since it was virtually i~npossibleto form a benchmark or control group of individuals who did not participate in the projects, we were not able to directly judge and compare earnings and employment outcomes of the participants against non-participants. Data Description Table 9.3 shows that overall, men accounted for 45 %, and women 55 %, of respondents, which closely reflects the characteristics of population reported in Table 9.2. However, younger participants were overrepresented, and older participants were underrepresented in the sample. The mean age of respondents was slightly over 42 years. Similarly, highly educated partici- pants were overrepresented, especially those who were college graduates. Those who held the legal head of the household status were 50 % of respondents. The mean household size was 3.5 persons. For those married only, about 56 % of respondents had unemployed spouses. Participation in public works projects has grown rapidly since the beginning of 1999. The projects were most extensively implemented during Phase I and Phase I1 of 1999. During that time, the number of unemployed, employed, and economically inactive respondents was subsequently reduced. A large number of respondents (66.8 %) participated in public works projects run by the local governments, and only 6.4 % of respondents participated in the central government information technology projects in 1999. About 27 % of the respondents participated in other central government projects in 1999. A majority of respondents (78.4%) participated in the projects to support a family. A majority of respondents participated in the projects just once or twice, and when they participated in the projects, they seldom quit the projects. Results Income Assistance: distribution eflects Although half of the respondents own houses, the average monthly household income of all respondents was just 950 thousand won (Table 9.4). Considering that the average household size was 3.5 persons, this amount was close to the poverty line. The minimum cost of living for a household with 4 persons was 930 thousand won in 1999. EMPLOYMENT CREATION THROUGH PUBLIC WORKS 183 The public works projects have received a great deal of criticism due to women participants not being considered eligible for the projects. The survey results in Table 9.5 demonstrate that these women came from economically disadvantaged households, where almost half of them (44.3 %) were major breadwinners and half of those married women's husbands were unemployed. Table 9.3: Descriptive statistics for survey respondents' characteristics (total =1505) Characteristics (%) Sex Male Female Age Age 20' Age 21 to 30 Age 31 to 40 Age 41 to 50 Age 51 to 60 Age 60' Mean age (SD) Schooling Elementary School Middle School High School College or university Family status Head of the household Dependent Mean size of household (SD) Marital status Married Never married Separated, divorced, widowed, etc Employment status of spouse (married'only) Regular worker Atypical worker Self-employed Non-paid family worker Unemployed 184 Labor Ma~.ketRefornu in Korea: Policy Optionsjor the Fltture Table 9.4: Economic status of survey respondents Characteristics (%) Housing Own house Chonsel) Monthly rent Other Major breadwinner Respondent Spouse Parents Child(ren) Other familymembers Household income(won) Monthly householdincome Monthly householdexpenses Indebted Yes No Loan amount and payment (indebted only)(won) Total amount 17,500,000 Monthly payment 170,000 Note: l ) A rent system in which the tenant pays a lump sum to the landlordand the tenant gets the money back when slheleaves Respondents spent 4.97 months on average participating in public works projects from January 1998 to July 1999. On average, they were employed for 4.04 months, unelnployed for 5.31 months, and economically inactive for 4.68 months. Those who were employed for less than a month amounted to 46.7 % of respondents, which reflects the harsh labor market conditions after the financial crisis. Those who were unemployed for less than a month amounted to 40.7 %, and economically inactivefor less than a month 54.3 %. These figures imply that a portion of participants might not be the unemployed, and the projects might have activated previously economically inactive population. The proportion of economically inactive population did not significantly differ by sex and residence. Only 15 % of respondents benefited from unemployment policies other than public works projects. The number of respondents who received NGO's relief activities was even smaller, being just 31 persons out of 1505 EMPLOYMENT CREATION THROUGH PUBLIC WORKS 185 Table 9.5: Economic status of survey respondentsby sex P - - Male Female Legal householder status Head of the household Dependent Major breadwinner Respondent Spouse Parents Child(ren) Other family members Employment status of spouse (Married only) Regular worker Atypical worker Self-employed/employer Non-paid family worker Unemployed Table 9.6: Types of benefited unemploymentpolicies (total=248) -- Characteristics (%) Livelihood protection program 54 Unemployment insurance 1 Vocational training 43 Loan schemes for the unemployed 1 Tax exemption 0 Others 1 respondents. A majority of the beneficiaries of government unemployment measures have participated in either livelihood protection programs or vocational training programs. Table 9.6 demonstrates that most public works project participants have not been benefited from other unemployment measures. In sum, the public works projects did appear to help those who needed social protection due to a sudden and massive upsurge of unemploy- ment and poverty. 186 Labor Market Refornzs in Korea: Policy Optionsfor the Future Job Creation: Labor Market Outcomes A large number of project participants were previously atypical workers, such as daily, temporary, or non-paid family workers. By occupation, they were also predominantly unskilled manual workers (Table 9.7). This table also illustrates that the employment structure of those 399 workers who found jobs after the project participation remained surprisingly similar to the employment structure shown before project participation. About 42 % of respondents answered that the compensation and working conditions of their current jobs were worse than their previous jobs, whereas only 18 % experienced improved compensation levels and working conditions in their current jobs. It suggests that the project participants had to face harsher labor market situations as a result of the financial crisis, and they took jobs that were inferior to their previous ones instead of lingering on the public works projects. Respondents' desired job characteristics, however, were significantly different from respondents' previous job characteristics and the current employment structure of those who were able to find jobs. Almost 80 % of respondents who are searching for employment prefer to have a regular job, whereas only 34 % of respondents who were previously employed had regular jobs. Their desired wage rate (96 thousand won) was much higher than the rate of those who found jobs after the project participation (70 thousand won). The respondents' strong desire to have a regular job is to a certain degree related to the fact that these projects were designed to be irnplernented for a relatively short period of time, i.e., two and a half months, in order to provide stronger incentives for job search activities. Logistic regression coefficients for regression of job attainment after public works project participation on several independent variables are reported in Table 9.8. The regression equation for the model reveals that male workers, younger workers and major breadwinners in the households were more likely to attain jobs after participating in the projects. Compared to the manual labor workers, those participated in the DB construction, mainly low-skilled white-collar workers were less likely to find jobs. Workers who were unsatisfied with working conditions, wage rates, labor intensity, and project duration were more likely to find jobs. The table also shows that previous job experience certainly helped most workers to find jobs after participation, with the exception ofthose workers who had worked in white-collar jobs. EMPLOYMENT CREATION THROUGH PUBLIC WORKS 187 Table 9.7: Employment structure of those employedbefore and after PWP participation (total=before:750;after:399) Before After Characteristics Employmentstatus I Regular worker Temporary worker Daily worker Self-employed Non-paid family worker Dispatched worker Employment status Il Full-time worker Part-time worker Occupation First job Same as previous job Different from previous job Compensation and working conditions Better than previous job Same as previous job Worse than previous job Industry Manufacturing Construction Service Hotels and restaurants Agriculture, forestry, etc. Other Type of worker Unskilled manual worker White-collar worker Skilled manual worker Service worker Other Monthly wage (won) (SD) Notes: 1) Wholesalelretailrepair service (6%),educationalservice (5%),housekeepingservice (4%), public administration (3%), transport distribution and telecommunications (3%), Finance and insurance(3%),social welfare (2%),fishery (1%), public utilities(1%),real estate (1%). 2) Technician (7%), skilled worker in agriculture and fishery (5%), expert (3%), machine operator (3%),high ranking manager (2%). 3) Educational service (5%), social welfare (5%), wholesalelretail repair service (5%), finance and insurance (3%), public administration (3%), transport, distribution and telecommuni- cations (3%),real estate (2%),fisheries (1%), mining (1%). 4) Skilled manual worker (?Ox), expert (6%),skilled worker in agriculture and fishery (6%), expert, machineoperator(2%) 188 Labor Market Reforms in Korea: Policy Optionsfor the Future Table 9.8: Logistic regression coefficients for regression of job attainment after PWP participation on selected independent variables Independentvariables Logisticcoefficients(SE) Gender (male=l) Age Education Project types (Manuallabor) DB construction Informationtechnology Administrativeassistant Educational activities Internship Welfare client assistant Degree of overall satisfactionon PWP Degree of satisfactionon projectmanagement Previousjob experience (no experience) white-collarworker Serviceworker Skilledmanual worker Unskilledmanual worker Other (Constant) ,317 (.462) -21og likelihood 1570.576 Chi-square 133,584"' Degree of freedom 17 Note: 1) Breadwinneris a dummy variablecoded 1 if the respondent is the major breadwinnerin the household. 'p<0.05. =*pn- employmen$ re-employment Table 12.6 shows the monthly economical activity change of the subjects since job loss or training in 1998. Since re-employment training for the unemployed participants spend an average of 4.8 months on participating training, their economically active status one month after training ends corresponds with that of the non-participant group six months after their job loss. For those whose training ended or whose non-employment began in January 1998, we could trace a maximum of 20 months of their economical activities since then until Sept. 1999; however, for convenience purposes we will only describe their economically active status up to the first ten months. It must however be noted that their training end point or non-employment beginning point vary, and therefore the actual months do not necessarily correspond with each other and thus requires more caution in interpreting their economically active status. Looking at the econo~nicallyactive status of cohorts by corresponding spell length of non-employment duration, employment rate of training participant group appears consistently higher than that of the non-participant group. Employment rate of participants at one month after their training ended was 26.8'36, which was higher than 24.4% of non-participant group at EVALUATING THE TRA~NING 257 Table 12.6: Trend of economically active status depending on re-employment training for the unenlployed participation After 1 After 2 After 3 After 4 A f r 5 After 6 After 7 After 8 After 9 After 10 mo, mo. mo. mo, mo. mo. mo. mo. mo. mo. Participants Employed 178 202 226 237 248 263 247 229 219 177 Unemployed 486 441 401 363 325 293 256 227 183 133 Nonecon.Activity 266 239 230 230 228 211 198 174 166 133 Total 930 882 857 830 801 767 701 630 568 443 Employmentrate 26.8 31.4 36.0 39.5 43.3 47.3 49.1 50.2 54.5 57.1 Participationrate 71.4 72.9 73.2 72.3 71.5 72.5 71.8 72.4 70.8 70.0 Non-participants Employed 52 69 84 94 115 132 143 152 166 182 Unemployed 499 485 471 456 434 410 391 377 353 328 Nonecon.Activw 246 243 242 247 248 254 262 268 278 278 Total 797 797 797 797 797 796 796 797 797 788 Employmentmte 9.4 12.5 15.1 17.1 20.9 24.4 26.8 28.7 32.0 35.7 Participationmte 69.1 69.5 69.6 69.0 68.9 68.1 67.1 66.4 65.1 64.7 six months after their job loss. This gap was expanded and by the fifth months after training ended, participants' employment rate was 43.3% whereas that of the non-participant group was only 35.7% at ten months after their job loss began. As previously described, the fact that non-participant group's re- employment experience rate or current employment rate is higher than those of the participant group has its root in their longer period of job search. That is to say, under the same job search period, participant's re-employ- rnent experience rate or current employment rate is higher than those of non- participants. Also, economically active participation rate of the re-employment training for the unemployed participants is also consistently higher than that of the non-participant group. In other words, training promotes re-employment as well as prevents withdrawal of participants from the labor market, increasing opportunities for active participation in economical activities. Period between Beginning of Non-employmentIEnd of Training and Re- employment Let us take a look at the period from job loss or end of training in 1998 to re- employment of those who were successful in finding jobs. Participants in re- employment training for the unemployed spend on the average about 4.3 months before re-employment. Non-participants on the other hand spend on average 8.6 months before finding jobs. These findings show that training participants find jobs twice as fast. Also, 49.6% of training participants have experienced re-employment, 32.3% who are re-employed found jobs within the first month out of training and more than half of the participants were re- 258 Labor Market Refornzs in Koruo: Policy Options for the Fntzlre employed within the first three months. On the other hand, 52.6% of non- participants have experienced re-employment. the most number of people are finding jobs after between six months and a year. More than half of this group is found to be re-employed six months afterjob loss. Table 12.7: Length of period from job loss or end of training to reemploy- ment depending of reemployment training for the unemployed participation Participants Non-participants Statistics (persons) (%) (persons) (%) 1month 116 32.3 31 8.9 1 - 3months 79 22.0 43 12.4 3-6 months 80 22.3 65 18.7 x2=128.954 6 months-1year 72 20.1 124 35.7 P=O.OOl Over 1 year 12 3.3 84 24.2 Total 359 100.0 347 100.0 Average length before 4.3 re-employment Employment Status upon Re-employment The quality of new jobs is as important as re-employment itself. Let us now look at the employment status upon re-employment. Employment status differed between participants and non-participants in re-employment train- ing for the unemployed, higher percentage of participants held regular status compared to non-participants whereas more number of non-participants held part-time positions, self-employed, and unpaid family worker. Among the Table 12.8: Employment status upon re-employmentdepending on re- employment training for the unemployedparticipation Participants Non-participants -- Statistics (persons) (%) (persons) (%) Wage worker Regular 218 52.5 210 46.7 Temporaw 94 22.7 92 20.4 Daily 43 10.4 61 13.6 Employer 17 4.1 15 3.3 x2=10.146p=O.O Self-employed 36 8.7 55 12.2 Unpaid family worker 7 1.7 17 3.8 Total 415 100.0 450 100.0 Wage worker (%) 85.5 80.7 t=-1.9181~ ~ 0 . 0 5 5 4 Regular(%) 61.4 57.9 t=-0.9705 p=0.3321 EVALUATING THE TRAINING 259 re-employed. higher percentage of re-employment training participants are wage workers than non-participants are. Firm Size upon Re-employment Companies to which the subjects are re-employed is significantly smaller in size when compared to their previous work places, showing that the quality of employment might be deteriorated. However, firm size does not differ by much depending on re-employment training for the une~nployedparticipation Table 12.9: Firm size upon re-employment participants Non-participants Statistics (persons) (%) (persons) (%) 500 - 38 11.0 35 10.3 Total 344 100.0 341 100.0 Dynamic Employment ESfect of Training Need for an Analysis of Dynamic Employment Effect of Training The analysis of employment effect of training defined by cohort by spell length of non-employment duration does not consider personal or previous employers' characteristics, which may have profound impact on the effect. As already identified, training participants tend to be younger and have higher education level than the non-participants, which may act to distort the employment effect of training and thus such influential characteristics need to be controlled. Also, as time goes on, chances of re-employment may decrease. As previously described, the longer the unemployment period is, the smaller the chances of being re-employed is; the analysis however shows that training leads to fast re-employment. Therefore, evaluating the employment effect of training at a fixed point of time rules out the time factor that certainly plays a major role in determining the accurate effect of training. To prevent this, the time spent between beginning of non-employment/end of training and re- employment needs to be considered for a more comprehensive analysis of the employment effect of training. The analysis of employment effect of training defined by cohort by spell 260 Labor Market Refor-n~sin Korea: Policy Optionsfor. the Futtrre length of non-employment duration has its limitation in that it also includes the incomplete spell length of non-employment duration. At the time of the survey, spell length of non-employment duration of non-employed is an incomplete spell, so an analysis based on the least square method using such data without controlling the right censoring loses statistical consistency. Therefore, we analyzed the employment effect of training, which determines the length of period from non-employment or end of training to re- employment, based on the hazard model that considers the censoring factor. Method of Analysis Based on the hazard model, which is widely used to analyze a movenient of the subjects from state to state, we defined a movement from non- employment to re-employment as hazard. From this point on, we will look into the factors that affect the trend of hazard change or the size of hazard. In order to analyze the quantitative effect of training on re-employment hazard, a Cox model which is equivalent to semi-parametric estimation by employing a proportional hazard model is used. In a proportional hazard model, the hazard at period of the observation with an independent variable x is expressed as follows: Here h, (t) represents a baseling hazard at t. Therefore, in a proportional hazard model, hazard change at every period is determined by the baseling hazard, and it is supposed that the independent variable (x) plays a role of scaling up or down the baseling hazard depending on its size andlor the value of j . A Cox model has the advantage in that, without knowing the shape of baseling hazard, it is able to derive the impact of each independent variable on hazard. Also, using the relative risk which is defined as the ratio of real hazard against baseling hazard, h(t)/ h,(t) , we are able to estimate the impact of independent variables. Natural logarithm of relative risk is expressed as Changes in relative risk from i change with all variables, other than independent variable i, fixed, is expressed as Therefore, relative risk of AX,, a change in X, is also expressed as ea '. Using the value of j and relative risk, we are able to estimate the size of independent variables' impact on hazard. If relative risk is bigger than 1, EVALUATING T H E T RA I N I N G 26 1 then transition probability increases; if relative risk is smaller than 1 , then transition probability decreases. Variables Because it is expected that independent variables and training participation's impact on re-employment hazard would certainly differ by sex, we estimated acoefficient for each case of male, female, and overall. As for personal characteristics, age seems only significant for women; especially at the age of 41, their re-employment possibility is the highest. In contrast, education level does not seem to influence re-employment possibility at all. By sex, men have 73% more chances of being re-employed than women. Marital status does not affect men; however for women, married women have just 33% chance of being re-employed of single women. Eliminating the marriage factor, position within a family as head of household works in favor of men, although not very significantly; however. it works as a disadvantage for women significantly. Table 12.10: Variables in reemployment hazard Variable factors Variables Mean Standarddeviation Spell length of non-employment (dependentvariable: month) Age AGE Sex MALE Marital status MARRIED Education SCH Head of family HEAD Employment lengthin previous job BTEN Previous job in manufacturing MANU Previous job in production PROD Firm with 300 or more SIZE Previous employment status REGULAR Involuntaryseparation SEPARATE Regionalunemployment rate REGURATE Training participation TRAIN -- As for the nature of previous employment, tenure very slightly decreases the chance of re-employment; although not as significant, this implies that tenure factor may work as an obstacle to re-employment. Manufacturing employment experience does not have any influence at all whereas males who were employed in production jobs have 27% higher chance of being re- employed than in other jobs. Also, the large size of previous firm acts as a disadvantage against the re-employment for men. Employment status and voluntariness of separation factors do not seem to affect the re-employment 262 Labor A4arket Refornzs in Korea: Policy Optionsfor the Fzltzlre possibility. Training participation factor, the main interest of this research. seems to increase re-employment possibility by 28%; this estimation differs by sex. For men, training participation does not seem to exert a significant influence on increasing re-employment possibility; however, for women, training participation increases re-employment possibility by 63% compared to non- participants under same conditions. Table 12.11: Estimationof re-employmenthazard -- All Male Female Coefficient Coefficient Coefficient AGE AGESQ MALE MARRIED SCH HEAD BTEN MANU PROD SIZE REGULAR SEPARATE REGULATE TRAIN -2 LOG L 7021.331 3882.054 2956.384 Chi-square 273.832 *** 121.347'** 146.539 *** N 1814 893 921 Note: ^*^, ** and stand for the significance levels of 0.01, 0.05 and 0.10, respectively Analysis of Labor Market Characteristics for those who are not re-employed Economically Active Status of Those Who Are Not Re-employed In order to find out if those who are not re-employed are actively seeking jobs, or if they are willing to work but are not seeking because there does not seem to be any jobs, or if they have retreated from the labor market completely, we categorized their economically active status into greater details. In accordance with [Economically Active Population Survey], unemployed persons are defined as "those who do not have any paid work EVALUATING T HE TRAINING 263 during the survey period, although are able to begin work immediately, and who are seeking jobs actively". We categorized economically inactive population as four groups by the subjects' length of job seeking activity, willingness to work, reasons for not seeking job, and job seeking activity during the past year. First group is categorized under "OECD standard unemployed" who have actively sought for jobs in the past four weeks; second group is so called "discouraged unemployed" who are willing to work, have sought jobs in the past year, but are not seeking jobs because there seems to be no jobs, that is discouraged by fzctors of the labor market; third group is categorized under "economically inactive population but willing to work" who are willing to work but have not sought jobs for household maintenance, raising children, education, and health reasons; fourth group is categorized under "purely econon~icallyinactive group" who have completely withdrawn from the labor market. Current Economically Active Status of Re-enzploynzent Training for the Unemployed Participants Who Are Not Re-employed Current economically active status showed difference between the re- employment training for the unemployed participants and non-participants at 10% significant level. 41.4% of those who are not re-employed after participating in re-employment training for the unemployed have actively sought for jobs in the past month; but 34.7% of non-participant group have actively sought for jobs during the same period. If we include those who are willing to work but are not seeking jobs because there does not seem to be enough jobs, so called the discourage unemployed, then 51.5% of training participants are found to be willing to work, which is higher than 42.6% of Table 12.12: Current economically active status of re-employment training for the unemployed participants and non-participants Participants Non-participants Statistics (persons) (%) (persons) (%) Unemployed 156 32.3 130 28.4 Economicallylnactivegroup OECD standard unemployed 44 9.1 29 6.3 Discouragedunemployed 49 10.1 36 7.9 x2=8.817 econ. inactive but willing to 73 15.1 74 16.2 p=o.o66 work purely econ. Inactive 161 33.3 188 41.1 sub-total 327 67.7 327 71.6 Total 483 100.0 457 100.0 264 Labor Market Refornzs in Korea: Policy Options for lhe Future the non-participant group. 33.3% of training participants who are not re- employed were purely economically inactive group who have completely withdrawn from the labor market whereas it was 41.1% for the non- participant group. These findings show that employment training maintain the subjects' desire to work, and contribute to relating them to the labor market. Current Economically Active Status of Those Who Are Not Re-enlployed by TrainingCategories4 Among those who received training but are not re-employed, there is difference of economically active status between participants and non- participants at 5% significant level. 34.5% of employment promotion training participants, who are not re-employed, have actively sought for jobs in the past month, showing a similar rate with the re-employment training for the unemployed non-participants. Also, 52.9% of re-employment training for the unemployed participants, including the discouraged unemployed group, were willing to actively seek jobs whereas the same for the employment promotion training participants was only 42.1%. Also, purely economically inactive group who have completely withdrawn from the labor market were higher from the employment promotion training participants who are not re-employed than from the re-employment training for the unemployed participants who are not re-employed. Table 12.13: Current economically active status of those who are not re- employed by training category Re-employment Employment training for the promotion training Statistics unemployed participants participants (persons) (%) (persons) (%) Unemployed 140 32.9 Economically inactive group OECD standardunemployed Discouragedunemployed Econ. Inactivebut willingto work Pumlyecon. Inactive Sub-total Total Re-employment training for the unemployed and employment promotion training for the unemployed are the two most important re-employment promotion programs in training. The former is for the unemployed from employment insurance applicable enterprises. and the latter is for the unemployed from employment insurance non-applicable enterprises. 43Publicly Funded Training for Unemployed Adults: Germany, the UK and ~ o r e a * Paul Ryan * a Introduction Training programs have bulked large in the Republic of Korea's response to the labor market difficulties precipitated by the 1997 financial crisis. The rapid increase in unemployment that accompanied the crisis was countered by an even faster expansion of publicly supported training. Participation in the two leading programs for the unemployed rose nine-fold between 1997 and 1998 (Kang et al., 1999). The rapid expansion oftraining in Korea has led to concerns about quality and effectiveness. Some providers appear to have taken advantage of a favorable opportunity to sell low quality programs. Participants appear not to have done well in the labor market: only 20 % found a job soon after training in 1998 (Cho and Ra, 1999). The Korean government has reacted to these difficulties in several ways: by stepping up the inspection of training providers and withdrawing funding from those found unsuitable; by basing funding of providers partly on the job finding rates of their ex-trainees (output-related funding); by funding trainees through training vouchers, in order to encourage them to pick better providers; by adopting a 50% target placement rate for 2000; and by favoring training programs tied directly to the needs of employers ('tailor made training'; Kang et al., 1999). Only limited evidence has been available to guide these reforms. Imme- diate job placement rates, on which much attention has focused, provide a The author would like to thank Joan Payne for comments and suggestions on previous related work, Bernd Fitzenberger, Reinhard Hujer, lngeborg Konig, Michael Lechner, Markus Pannenbeg and Viktor Steiner for assistance in locating German materials, the UK Department for Education and Employment for providing unpublished data, and participants in the Seoul workshop for helpful comments. All interpretations remain my own responsibility. " University ofCambridge, UK. 266 Labor Market Refornu in Korea: Policy Optionsfor the Future poor criterion of program success. The rate typically rises strongly during the first year after training. Other outcomes matter too, including the acquisition of vocational qualifications, and pay when employed. Moreover, these outcomes describe only the gross outputs of a program, not its value added - i.e., the change in outcomes relative to what they would have been in the absence of the program. Additional evidence can be sought from two sources. The first is Korea itself. A formal evaluation has recently been conducted for the leading training program for the unemployed in 1998. Its findings potentially change the co~~ventional'failure' perspective upon Korean training (Kang and Lee, 2000). The second source is other countries. The track record of adult training programs - particularly the remedial programs that have dominated in the US - is at best patchy (Dar and Gill, 1995; Martin, 1998). A better prospect of useful evidence are countries whose training programs have been evaluated in detail, as for the US, but in which public training services have undergone strains more similar to those experienced by Korea. Two national experiences that are potentially resonant from a Korean perspective are the East German and the British. German reunification in 1989-90 was followed, in what was formerly East Germany, by a sharp rise in unemployment, inducing the federal government to introduce training programs and expand them rapidly. Similarly, the large rise of unemployment in Britain during 1979-83 led during the mid-1980s to the expansion of public training and the recasting of its role in labor market policy. As the Korean economic recovery has gathered pace, so the pressure on public training programs has eased. The government has scaled back its plans for public training. The opportunity has arisen to reconsider the role of public training, in terms of its size, quality and role within labor market policy, less influenced by the distortions caused by labor market crisis. The role of public training in Germany and the UK is discussed in section 2, followed in section 3 by methodological issues in evaluation research. Section 4 considers evaluation findings for German and British programs, section 5 some relevant attributes of Korean training programs. Conclusions follow in section 6. Training Attributes: Germany and the UK In both Germany and the UK, economic dislocation has at times produced acute difficulties in the labor market, prompting governments to expand public training programs for the unemployed. This section considers both experiences. The primary emphasis is placed on East Germany, where the speed of developments after 1989 mirrors that of Korea after 1997, and where reunification provided a context similar to, if less acute than, that which may face the Republic of Korea in the future. The National Context: the Labor Market and Public Policy In mid-1990 the Federal Republic of Germany absorbed the former German Democratic Republic (East Germany). National reunification produced a boom in West Germany, resulting from the new demand unleashed in the East for the West's consumer and capital goods, but a slump in the East, as the demand for its output fell heavily. Eastern producers had to redirect their efforts froln the sheltered product markets of ex-communist east Europe to the exposed ones of the international market system. Their competitiveness proved low, as a result of poor products, low labor productivity and, given the 'one for one' terms of currency unification, high hourly labor costs. The demand for labor in the East collapsed. The result was an empIoyrnent crisis unprecedented in Germany since the 1930s. By 1992 the Eastern labor force had shrunk by one third. Despite that, the unemployment rate had nearly reached 20 %. One in every two members of the labor force received public support, whether unemployment benefit, training or another labor market measure. The labor market problem in the East eased only moderately thereafter, before intensifying again as the national economy faltered after 1995 (Figure 13.1). Figure 13.1: Unemployment and labor market slack in Germany, 1991-1997 Notes: 1) Un: registered unemployment 2) SI; 'Slack', registered unemployment plus participants in all labor market programs, including sholt-time working (full-timeequivalent) Source: Bach etal. (1998), Tables 1,4,11.12. 268 Labor lbfarketRefoornzs in Korea: Policy Optionsfor the Futzrre Labor market policy soon became a major plank in the federal govern- ment's response to the East's employment problems. Initially, subsidies for short-time worlting led the way, followed by early retirement subsidies, but both became unimportant after 1994. Job creation and labor market training were slower to come on stream, but by 1992 training alone accounted for more than 6 % of the labor force. Its coverage shrank to 3 % by 1997, but that was still more three times larger than in the West, where unetnployment had by then passed 10% (Figure 13.2). Figure 13.2: Rate of participation in publicly funded adult training in Germany, 1990-1997 0.0 1990 1391 1992 1393- 1994 1995 1996 1997 dEast 63 West Note. Training defined as in Table 13. 1 Source: Bach eta/. (1998). Tables 1,4,11,12 The UK context is the elnployinent crisis of the 1980s. Prompted by the 1979 oil price shock and the macroeconomic policies of the Thatcher Government, the unemployment rate rose during 1979-83 from 4 to 12 %, where it remained there until 1986. The policy reaction was less immediate and dramatic than in East Germany, and it concentrated during the first half of the 1980s on youth programs. By 1986, long-term unemploy~nenthad reached 4 % of the labor force, most of it adult. Training for the adult unemployed was expanded strongly in 1986-88 and its content reformed. Public Training Progranis: Content and Clientele Publicly supported training for adults is heterogeneous, in terms of services, intensity, duration, location and clientele in particular. The point must be underlined, as much of the evaluation literature (section 3, below) treats it as essentially homogeneous, as part of 'training' or human capital formation in the abstract. Germany Three main categories are typically distinguished in Germany' further training, retraining and familiarization training (Table 13.1). Familiarization training, aimed at new hires, is typically job-based, part-time and short-lived, and does not lead to a vocational qualification. At the other pole, retraining is typically 'classroom'-based, full-time, long-lasting, and it must lead to an occupational qualification. Inbetween, further training, aimed at the already qualified within an occupation, divides into adaptation training, which goes primarily to disadvantaged individuals and female re-entrants, and upgrade training, frequently aimed at advanced qualifications within an occupation. It contains some remedial training, which constitutes the dominant category in the US but which lacks a direct counterpart in Germany. Its content varies from short courses that are not linked to an additional qualification, to longer ones that culminate in the acquisition of Meister, Techniker and other qualifications (Blaschke, Plath and Nagel, 1992). All three types of training are eligible for public funding under the Employment Promotion Act of 1969, which was extended to the East after reunification. The key change was the temporary introduction in the East of ad hoc short courses, typically lasting for no more than one month, and accounting during 1991-92 for around one-fifth of all publicly supported trainees.' By 1993 training in the East had stabilized along lines similar to those in the West. Differences in emphasis between the two regions were still present, with training oriented in the East towards adaptation training and retraining, in the West towards upgrade training (Table 13.1). For the two main programs, adaptation training and retraining, training durations were markedly longer, and training more commonly conducted full-time in the East than in the West (Table 13.2). The East German training effort was strikingly intense and costly. Retraining was delivered overwhelmingly on a full-time basis, apart from employment, lasting on average (for completers) more than 18 months and containing extensive technical education. Adaptation training shared those attributes, in the East at least, where the great majority of training was full- time, and courses lasted nine months on average (Table 13.2). These attributes of adult training reflect its link in Germany to the occupa- tional categories defined for the national apprenticeship system. With the temporary exception of the short courses of 1991-2, publicly supported adult training is geared to the same occupations, and in the case of retraining ' Publicly supported adult training programs in Germany are typically grouped together in the category Fortbildung und Un?schul~~ng (FuU). Lechner (1996);Hu~jerand Wellner (2000), Figure 1. 270 Labor Market Refornzs in Koren: Policy Optionsfor the F~tlire Table 13.1: Shares of training categories in publicly funded training for adult workers, old and new Bundeslander in Germany, 1993(Q3com- pleters) Training category East West') All (%) (%) (%I Further Training: a. Adaptation 62.6 43.4 53.1 (Anpassungsfortbtbildung) b. Upgrade (berufliche Aufstieg) Retraining (Umschulung) Familiarization training 7.6 2.6 5.2 (Einarbeitung) All training (%) 100.0 100.0 100.0 (numbers) 59,600 58,200 117,800 Note: 1) the former West and East Germany constitute respectivelythe old and new Bundeslander Source: Blaschke and Nagel (1995),Table 3. requires the attainment of the same (craft plus) level of knowledge and skill as does an apprenticeship, albeit in as little as half the time. Adults undergoing retraining must also pass formal examinations, conducted by the same local bodies that examine apprentices (Johansen, 1994). The clientele for public training programs has traditionally been primarily the unemployed, but in East Germany the widespread need for skill enhancement after reunification led to an informal guarantee of access to all who demanded it. In practice, participation in training has been slanted towards the unemployed, females (except for familiarization training) and the more highly educated. The UK In Britain, adult training has evolved through three different programs since the mid-1980s (Table 13. 3).3The 'Old' Job Training Scheme (JTS)%f the The New Deal (1998- ) currently constitutes a fourth, evaluations of which have now begun to emerge. The descriptor 'Old' was subsequently appended to JTS in order to differentiate it from a short-lived replacement of 1987-88, which has been termed the 'New' JTS. The title is clearly ~~nofficial ;the marketing-oriented Government of the time would hardly have chosen such a title for anew program. Table 13.2: Attributes of public training programs for adult workers, old and new Bundeslander in Germany, 1993 (all 1993. Q3 completers) Training Share of females within Share of part-time trainees Average duration of training category category and region within category and region (months) West East West East West East Further Training: a. Adaptation 47.1 70.6 17.8 14.4 6.6 9.0 (Anpassungsfortbildung) b. Upgrade 25.6 45.9 44.1 29.3 n.a. n.a. (berufliche Aufstieg) Retraining (Umschulung) Familiarization training 37.7 31.5 8.8 4.0 n.a. n.a. (Einarbeitung) All Training 40.7 64.5 23.9 10.7 9.8') 12.1') Note: n.a,:not available 1) Adaptationtraining and retraining oniy Source: Blaschkeand Nagel (1995),Tables 3,5 Table 13.3: Attributes of public adult training programs in UK, 1986-1998 Program Periodof Eligibles Spell Services Participant Public Stock of operation duration income funding partidpants entitlement basis ('000) Age Status months (year) Job Training 1986-88 18+ Interested 5.0 Occupational No course fees Trainee Scheme individuals off-JT, public (96%); training starts and ('Old' JTS) provision allowance(83%) time spent in training Employment 1988-93 18+ Unemployed 4.9'' Trainingleading Unemployment Trainee Training (ET) 6+months to a VQ via benefit plus f10 starts and work, off-JT or per week weeks project TrainingFor Work 1993-98 18+ Unemployed 4.5 Training,work Unemployment Participant (TFW) 6+months experiencevia benefit plus f10 outcomes work, off-JT per week; (VQ, or project employee status permitted Notes: 1) author's estimate; participants who left within two months not included 2) author's estimate, based on 40.000 completions p.a.; dropout rate of 10 % assumed 3) from 1995. Sources:Payne (1990), Payne, eta/.(1996), Payne, etal. (1999). PUBLICLY FUNDED TRAINING FOR UNEMPLOYED ADULTS 273 mid-1980s provided occupational training, mostly retraining and upgrade training, to a clientele that included the unemployed, female labor market re- entrants and employees willing to quit their jobs in order to train full-time. The Employment Training (ET) and Training for Work (TFW) programs of 1988-98 offered retraining, remedial training and work experience, based variously at workplaces or in special projects, with eligibility focused on the long-term unemployed. The replacement of JTS by ETITFW saw several changes. Training was redirected from occupational training and retraining for a mixed clientele, delivered off-the-job in public training facilities, to job-related training and work experience primarily for the long-term unemployed delivered insofar as possible on the job. The scale of training rose seven-fold during 1986-92, but fell back thereafter. The average duration of training remained around four to five months, but the shift towards work-based training probably reduced training content. Finally, public payments to training providers were rebased in 1995 from inputs (starts plus progress) to outcomes (qualifications acquired and employment rates after training). The training content of British programs varied from extensive to negligible: from the certified occupati6nal skills that were prominent in JTS to the job search and work experience in which segments of TFW specialized - which should strictly be excluded from the category 'training program'. Training methods included various mixes of off-the-job ('classroom') training, as emphasized by JTS, and on-the-job training and work experience, as emphasized by ET. Comparison East German and British programs therefore resembled each other in terms of their rapid expansion, in the face of mounting unemployment (Germany) and long-term unemployment (the UK), their provision of varied training services according to participants' needs, and their emphasis on acquiring vocational qualifications. They differed in terms of a shorter duration and lower intensity of training in the UK, along with a greater orientation to remedial training, an increasing orientation towards work-based training, and the introduction of output-related f ~ n d i n g . ~ The contrast between the countries is sharpest for the amount and quality of training provided to individual participants. Short course durations, low trainee achievements and limited quality control in work-based training have The two countries also differ in the meaning of a training 'program'. In Germany, puhlic training services have evolved to meet changes in perceived needs, such as those in the East after 1989, but within the framework of the 1969 Act ( M G ) throughout. The concept of discrete programs with limited individual life-span is more appropriate for the UK, where Governments have frequently discarded an existing program, such as .ITS, in favor of a new one, such as ET. The e ~ n p h a s ~has s been placed in Britain on program innovation, in Germany on institution building. stability and adaptation. 274 Labor Market Refornzs in Korea: Polic-vOptions for the Flrtzrre caused concern in the UK. Only one-third of ETITFW participants gained a vocational qualification, and only a handful at craft, technician or higher levels (Table 13.4). These low attainment rates undoubtedly reflect the low prior skills of the long-term unemployed. They may also reflect employer use of trainee labor for job training and production work, at the expense of occupational preparation, when training is work-based and external quality control is weak (Ryan, 1994). Table 13.4: Vocational qualificationsgained by leavers from adult programs in the UK, 1986,1994 and 1997 JTS ETi) TFW1) 1986 Jan-Mar, 1994 Jul-Sep, 1997 (%) (%) (%) Full qualification 63') 34 Level 1(basic) n.a. 7 Level 2 (semi-skilled) n.a. 12 Level 3 (craft) n.a. 2 Level 4 (technician) n.a. 0 Other n.a. 13 Part Qualification n.a. 6 All 63 40 Notes:l) Englandonly. 2) includes completioncertificatesissued by training providers(17 % of trainees). Sources: Payne (1990), p.24. Department for Education and Employment, unpublished information, trainee database system Evaluation Methods The way to evaluate training is disputed, but the basics are widely agreed. Two issues may usefully be distinguished: the goals of training and the estimation of the counter-factual (Heckman, Lalonde and Smith, 1999; Grubb and Ryan, 1999). Program Goals and Evaluation Criteria The standard criterion is whether or not a program improves the labor market prospects of its participants. When unemployment is high, employment tends to be the priority, as measured by rates of employment, stable employment and unemployment after participation.' As the presumed " The dominance of employment-relatedoutcomes in the German evaluation literature is also consistent with the explicit goal of the Employment Promotion Act. viz. 'achieving and goals of goal of training (as opposed to, e.g., job creation) include skill enhancement, effects on pay (e.g., hourly wages) should also be considered. Pay effects are regularly considered alongside elnployment ones in the British literature, but only rarely in the German one. These two criteria are however not sufficient for effective evaluation. Their limitations become visible when the criteria of applied welfare economics, viz., efficiency and equity, come into consideration. An improvement in participants' subsequent elnployinent rates and pay implicitly suggests an efficiency gain. Two further adjustments are however required before such a conclusion can be reached. The first is to allow for displacement effects on non-participants. Higher subsequent employment amongst participants may come, partly or wholly, at the expense of lower employ~nentamongst non-participants. In that case, the sum of participant's benefits overestimates aggregate employment benefits. The second is to take account of the cost of the program. Cost-benefit analysis is widely used in US evaluation research, but neither it nor adjustment for displacement has featured in its German and British counterparts. Finally, the equity dimension of program effects is also itypically overlooked, even though many training programs are set up to help the disadvantaged, including the long-term unemployed in the UK and the uninsured unemployed in Korea. Adopting here an egalitarian social welfare function defined over economic outcomes, a program has merit from an equity standpoint when its clientele is economically disadvantaged and when that clientele is made better off than it would otherwise have been, during if not after participation.' Evaluation Methods: the Counter-Factual The counter-factual is defined as the situation of participants had they not taken part in training, which cannot be observed directly and has to be estimated. The effects of the program are then estimated as the difference between the actual and the counter-factual situations. Estimation of the counter-factual depends heavily in the US nowadays on highly experimental methods: controlled trials, in which eligible workers are assigned randomly to 'treatment' and 'control' groups, with the former receiving training, the latter (ideally) not receiving it. Social experiments are however rare outside the US. The acceptability of refusing training to a ' control group is especially low when unemployment is high, and when eligible workers are guaranteed access to, or required to undertake, training maintaining a high level of employment, constantly improving the structure of employment, thereby promoting economic growth' (Bach et al., 1998, p.15). ' The further, non-economic goals, such as reduced dependence on public income support, and cuts in cr~minalityand teenage pregnancy, that have often featured in the goals of training programs in the US (Grubb, 1996) have not done so in German and British programs. 276 Labor Market Rejbrms in KO/-eatPolicy 0ptionsjo1-the Future -asin boththe UKandEastGermany. Under such conditions, the counter-factual is typically estimated by quasi- experimental methods. Some mix of statistical selection and economic modeling is used to control for differences between the attributes of participants and of a 'comparison' group that is typically drawn from national surveys of households, unemployed workers, etc. The experiences of the latter are used to estimate the counterfactual. The difficulty is that the two groups may differ not just in observed (i.e., measured, and therefore statistically controllable) but also unobserved attributes that are relevant to both participation and labor market outcomes. For example, participants may be more able and more motivated than non-participants, and any superiority in post-program outcomes for participants over non-participants may then reflect the underlying difference in attributes rather than any effect of training. The threat of selection bias in quasi-experimental evaluations is widely recognized. The various ways of countering it divide here into two: matching and modeling. Both have been extensively used in research on East Germany. Matching methods pick for every participant one or more non-participants to minimize the difference between the two in observable attributes (e.g., schooling, age, sex). They use pre-participation period outcomes for the two to test for the presence of unobserved differences. If none are detected, program effects are estimated by the difference in outcomes in the two groups, adjusted for any differences in observable attributes that the matching procedure did not manage to eliminate (e.g., Lecher, 2000). Model-based methods, by contrast, apply an econometric model of the selection of eligible workers into participant and non-participant status. They seek to remove the selection biases caused by unobserved attributes with the use of such techniques as fixed effects models, instrumental variables and two-step estimators - all of which require potentially restrictive assumptions in order to generate estimates (e.g., Fitzenberger and Prey, 2000).' Hybrid approaches have also been used: e.g., the demands made on econometric modeling are reduced by combining it with individual matching, particularly by area of residence (Payne et al., 1999). The validity of quasi-experimental methods is however limited. The estimates that they generate have been shown to be sensitive to choice of both selection model and comparison group, in evidence for both the US and Germany, notwithstanding the expertise that has been devoted to refining ' For example. fixed effects models. which feature prominently in German evaluation research (Fitzenberger and Prey, 2000), implicitly assume that any unobservable factors that influence individuals' labor outcomes are constant, across both individuals and time. A program's effects can then be estimated as the difference between the two groups in the change in their members' labor market outcomes. To the extent, however, that shocks are specific to both individual and time (e.g., a particular person 'had a bad year'), the technique does not remove selection bias. them (Lalonde, 1986; Hiibler, 1997; Heckman, Lalonde and Smith, 1999). On the next rung down the ladder of evaluation technique stand weakly experimental methods. The counter-factual is suggested by the experiences of groups whose situation bears some similarity to that of participants. Possible choices for such a group include: participants in other programs, similar age groups, similar groups in other countries and even participants' own past situations: (beforelafter comparisons). Thus in Germany comparisons of participants' outcornes to those of both dropouts and participants in other programs were used before quasi-experimental methods became the norm (Hofbauer and Dadzio, 1984; Blaschke and Nagel, 1995). Uncontrolled differences in the unobserved attributes of the two groups mean however that the validity of such evidence is low. On the bottom rung, there are gross outcomes, such as the 20 % employ- ment rate of ex-trainees in Korea in 1998. Similarly, in the UK in 1997, 50 %of TFW ex-participants achieve a 'positive outcome', defined as employ- ment or entry to education or other training, as measured six months after leaving (DfEE, 1998). In Germany, the earliest evaluation research had also been limited to gross outcomes: e.g., the 59% employment rate 3 to 6 months after completing further training in the West in 1983 (Hofbauer and Dadzio, 1984). As gross outcomes ignore the counter-factual, however, this evidence is non-experimental, and its value for evaluation correspondingly low.9 Finally, there is implementation-oriented research: establishing how a program operated in practice, in relation to both its design specifications and programs in other times and places (e.g., section 2, above). How a particular program came to succeed or fail is likely to remain a mystery without such evidence. In sum, amongst the various evaluation methods in use, even the most sophisticated are imperfect. In addition, evaluation research often handicaps itself by taking a narrow view of program goals, by ignoring aggregation problems, and by emphasizing refinement of econometric technique rather than data quality. Training Evaluations for Germany and the UK The longer operational history of training programs for the unemployed in Germany and the UK than in Korea is associated in each case with a more extensive evaluation record. This section reviews the findings of the relevant national literatures. "ross outcomes, such as completion and employment rates, are widely used in the US for administrative purposes, e.g.. as indicators of the success of program delivery across localit~es or providers. Even there, however, selection biases undermine the valid~tyof the itiformation (Barnow. 1992). 278 Labor Market Refirms in Kor-ea: Policy Options for the Filtzl/-e Findings: Germany The evaluation studies for Germany and the UK discussed here (Tables 13. 5,6) all concentrate on subsequent labor market outcomes for participants. All are quasi-experimental, applying sophisticated econometric or statistical techniques to micro-data in order to estimate the counter-factual. Their limitations include small sample sizes (Germany only), and the neglect of program attributes (and their implications for equity), of aggregation problems and of operational attributes. Four recent evaluations of training programs in East Germany are considered, selected from the twelve covered by a recent survey (Fitzen- berger and Speckesser, 2000). All of the studies cover the two largest public training categories: further training and retraining (section 2, above). Two include familiarization training as well. All refer to part or all of 1990-94, i.e., the aftermath of reunification. All concentrate on employment-related outcomes for participants after leaving training, covering a maximum period of four years and, for most participants, no more than two years. Hazard rate models of transitions between labor market states are used in three studies. Methods of estimating the counter-factual divide evenly into statistical matching and econometric modelling (Table 13. 5). The results suggest that publicly supported training has failed. In particular, it has not improved participants' employment prospects. The estimated impact of training is for typically not significantly different from zero, statistically speaking. Two studies even find negative effects: Lechner for the first twelve months after training, and Kraus et al. for training undertaken during 1990-92. The only positive finding concerns the effects of training conducted during 1993-4, which Kraus el al. find helped participants to find work, relative to having remained unemployed. This pattern of results matches that in wider literature for East Germany. 'Most evaluation studies produce no significant effects' and 'the positive employment effects of job creation and training measures are in any case small' (Fitzenberger and Speckesser, 2000; 21,22). The same bleak picture applies also to publicly funded training programs in the West (ibid.,Table 3). It contrasts however with that forprivately funded adult training in the West, evaluations of which suggest favorable employment-related effects, particularly for programs lasting more than one year (Table 13.5; Fitzenberger and Prey, 2000)."' Various causes may be suggested for the absence of a favorable or even a stable pattern in estimates of employment effects. The first consideration is sample size. Positive effects may well exist, but, as only small sample sizes "' Again: the findings of various studies diverge. Pischke (1996). studying wage rather than employment effects. finds no general benefits from employee training in Germany. Table 13.5: Evaluations of adult training in Germany (quaxisxperimental methods only) Programeffects Territory Training Data Sample Size of sample Study Evaluationmethod Employment-related covered service5' source" frame [controlgroup] Variable Effect2' Public Lechner East GSOEP Entrantsaged 131 Matchingby observable lnudenceof unemployment - First 12 (2000) 1990-96 lessthan 53, [I2801 attributes and full-timeemployment, months 1990-93 subsequent4 years 0 sub sequent Hujerand East GSOEP, Entrantsaged 231 Matchingby observable Durationof employmentand 0 Wellner 1990-94 lessthan51, [1401] attributes unemployment.subsequent (2000) 1990-93 3years Kraus, East AMM, Entrants, 3.503 Correctedforobservables Transtionstoemployment - 1990-92 Puhaniand 1990-94 1989-94 [3095] (testedfor unobservables) andinactivity,subsequent Steiner(l999) Iyear + 1993.94" Fitzenberger East AMM, Entrantswo. 471 Observables:controls; Employmentrates, 0 0 andPrey 1990-94 othertraining, [4352] unobservables:Cxed subsequent0 4 years (real hoully (1999) 1989-94 effects pay) Non-public Fitzenberger West Allprivately GSOEP, Employeesaged 126 Matchingbyobse~ables, Employmentand 2%all andPrey fundedadult 1984-97 2 5 . 4 4 ~ [I1131 fixedeffectsfw unemploymentrates, 6% duration (2000) training 1984 unobservables subsequent3 years 12+mos Notes: 1) GSOEP: German Socioeconomic Panel (annual household survey); AMM: Arbeitsmarktmonitor (Labor Market Monitor: annual longitudinal survey of East German population of working age). 2) +,- indicate respectively statistically significant positive (i.e.,in desired direction) and negative (undesired direction) effects cant effects. 3) unpublished results, reported in paper. 4) off-the-job programs (1-2) only (for men). 5) training categories as in Table 13.2. 280 Labor Market Refornzs in KOI-ea:Policy Optionsfor file Ftrttr~-e are possible in both of the datasets available (GSOEP and AMM; Table 13. 5), statistical significance is difficult to establish. This interpretation is consistent with the finding by Fitzenberger and Prey (2000) of positive but statistically insignificant coefficients in almost all the models that they estimate." A second possible cause is methodology: the variety of assumptions involved in implementing the counterfactual may generate unstable results, much as was the case in the US until quasi-experimental methods gave way to highly experimental ones (Lalonde, 1986). For East Germany, Hiibler (1997) notes a similar sensitivity of results to choice of estimation method and dataset. This explanation suggests that little will be learned until social experiments are used in Europe." Thirdly, weak effects may reflect low quality training. The possibility is particularly relevant to 1990-92. Some research shows employment effects that change from negative to positive between 1990-92 and 1993-94 (Kraus et al.; Pannenberg, 1995, 1996). It is also consistent with the predominance of short courses and an underdeveloped training infrastructure in the East right after reunification. Yet over 1990-94 as a whole, East German programs were both intensive and long-lived by UK and US standards, which hardly suggests low training content. Evidence on the irnplernentation of training in the East, and on vocational qualifications obtained by trainees, is needed to assess the issue. A fourth explanation might be that not even high quality training can succeed in slack labor markets. In the post-unification East, even the markets for more skilled labor towards which trainees are being redirected are glutted, leaving few vacancies for ex-trainees to pursue. This view is certainiy consistent with the state of the Eastern labor market at the time. It sits uneasily with evidence from the West, however. Although labor market slack had risen there during the 1980s, it remained much lower than in the East and it fell during the reunification boom of the early 1990s. Nevertheless, eight evaluations of public training in the West, for various periods centred on 1984-93, have also produced a varied and contradictory pattern of employment-related effects, with similar numbers of 'significantly -- - " Kraus et a/.does however generate a much larger sample by aggregating across years. Small samples impose two further limitations on research on East Gennany. As many participants do not find work. pay effects have to be estimated from even smaller san~ples.and training effects cannot be broken down, e.g., by type of training. '*The problems of sample size and evaluation method interact unfavorably, particularly for evaluation methods that rely on the individual matching of participants and controls. Non- parametric matching techniques confined to the observable attributes of participant and comparison group members (e.g.. Lechner, 2000) require that there be no significant ditference in the pre-program labor market positions of the two groups. When sample sizes are small. however, the negative tinding of no statistically signiticant difference provides a partic~~larlyweak basis for the positive conclusion, that the two groups really are similar. The findings of the evaluation then remain subject to selection biases. caused by uncontrolled differences in unobserved group attributes. positive' and 'insignificant or significantly negative' estimates (Fitzenberger and Speckesser, 2000, Table 3). The German evidence suggests therefore that public programs have done little for their participants, relative to having remained unelnployed - notwithstanding their exceptional intensity and cost. The point is underlined by a similar absence of generally favorable effects in studies of outcomes at more aggregate levels, such as regional employment rates (Schmid et al., 1999; Fitzenberger and Speckesser, 2000, Table 5). More needs to be known about the effects of German programs, including their effects on pay, the comparative effects of the different training categories in Table 13.1, and the role of vocational qualifications in mediating labor market outcomes. The evidence provides little support, however, for the hypothesis that poor prograln results, as commonly found for the US, reflect low levels of investment in each trainee (Grubb and Ryan, 1999). Not even programs that aim as high as in Germany can apparently generate benefits for participants sufficiently strong to be visible in slack labor markets through the fog created by small samples. If the problem with apprenticeship training in Germany has been that it is hard to make it work in other countries, including Korea (Jeong, 1995), public programs apparently face a more fundamental difficulty: making them work in Germany itself. Unfavorable evaluation findings have increased doubts about public training programs for the unemployed in the East, and encouraged their contraction despite enduringly high unemploy- ment (section 2, above). Findings: UK There have been fewer evaluation studies for the UK than for East Germany, but their scope has been wider and their findings more encouraging (Table 13.6). Each of the three major programs has been evaluated by Joan Payne and colleagues. In all three studies, labor market outcomes for participants (here, pay as well as employment) were compared to those for matched members of a comparison group, and statistical methods were used to counter differences between the attributes of the two groups, including unobserved ones in the case of pay.I3 Although the quality of both the matching and the statistical methods used rose across the three studies, continuity of personnel and methods make the findings unusually co~nparableacross programs. The availability of sample sizes several times larger than those typically available for East Germany also increases the l 3Controlling for selection processes is potentially particularly important in comparing these three programs. The quality of the participants is likely to have been lower for ET and TFW than for JTS, insofar as the former but not the latter were targeted upon the long-term unemployed; and for JTS and ET than for TFW, insofar as only the latter has involved output- related funding. which encourages training providers to 'cream' for prospective success when selecting participants. 282 Labor Market Refornzs in Korea: Policy Optionsfor- the Fzftzlre informational content of the evidence. All three programs show substantial and statistically significant improve- ments in employment-related outcomes, including both the incidence of employment at a particular point and cumulative time spent in employment. The increase in employment rates ranged from around 14% after one year for JTS to around 22 and 12% after three years for ET and TFW respectively (Table 13. 6). Short-term (one year) effects were substantial only for JTS, but they built up considerably over three years in the other two programs, for which longer post-program data were available. The pattern of pay effects differs sharply by program. JTS participants gained an estimated 22 % in hourly pay overall. The two-thirds of participants who found a job in the occupational area for which they had been trained gained 31 %. These effects are large, even implausibly large: the private benefits of formal schooling are typically less than 10% per year of full-time study. By contrast, ET and TFW participants as a whole gained little or nothing. Significant benefits were present in both cases for particular participant categories: notably, those who received off-the-job training in ET, and work-based training and an associated vocational qualification in TFW. The estimated pay gains in these two categories- six and nine % respectively -arecomparabletotheprivatereturn onotherinvestments inhumancapital. Overall, however, it appears likely that JTS did more for participants' earning power than did ET or TFW. Interpretation One interpretation of the UK evidence might be that all of the types of training covered by these programs can improve participants' employment prospects, but only occupational training, as in JTS, can systematically improve their pay. This interpretation has some a priori plausibility. It is also consistent with the limitation of pay gains under the other two programs to participants who took off-the-job training and acquired a vocational qualification, under ET and TFW respectively. In the German evidence the interpretation is consistent the apparent rise in the benefits of training when the crisis-generated short courses were withdrawn in 1992. The interpretation is however open to doubt on two scores. Firstly, the quality of both the data and the matching of participants and non-participants appear to have been lower in the JTS evaluation than in the later ET and TFW ones. Secondly, the long and expensive public programs of occupational training that predominate in Germany appear not to have worked there either.I4 " That finding may however reflect the limitation of the German evaluation literature to employment outcomes. (Two of the evaluations in Table 13. 5 act~~allyevaluated effects on pay as well as employment, but neither published the results.) In Iieavily overstocked labor markets, the benefits of occupational training may well be greater for pay than for employment: trained ex-participants may find it hard to find a job but, when they do, their enhanced skills should mean higher pay. 284 Labor hlarliet Refon~lsin Korea: Policy Options fol- the F Z L ~ I L I - ~ An alternative interpretation might be that the greater e~nploynlenteffects estimated for British than for East German programs reflect the prominent role of the workplace in two of the UK programs (ET and TFW) and its marginal role in Gennany. An employer who sponsors a trainee, whether directly or simply by providing work experience within a training program, enjoys Inore scope to tailor training to its own needs and to screen the trainee for subsequent employment than is possible in 'classroorn' training. Work- based training might therefore increase employment prospects for participants by more than does other training. The interpretation is also consistent with the difference between the employment effects of British and East German programs in the 1990s, given the greater part played by work- based training in the former than in the latter.'' Again, other evidence casts doubt on the interpretation. The UK results suggest that JTS, which had no workplace component, appears to have done more for its trainees, particularly in terms of pay, than did its more work- based successors, ET and TFW. A further consideration is that, even were work-based training shown definitively to improve employment prospects for participants, that would have to be set against a greater probability of extensive displacement of non- participants by participants, both during and after training. To the extent that work-based training involves productive labor, employment options during the program are reduced for non-participants. To the extent that employers use the program to train and screen future employees, employment options for non-participants decline after the program as well. Employment effects are then less in aggregate than the sum of those for participants - and may ' even, as in some Swedish and German evidence, be zero in the aggregate (Skedinger, 1995; Fitzenberger and Speckesser, 2000). All in all, it is difficult to derive a coherent story from the German and British evaluation literatures combined. Publicly funded occupational training appears to have worked in the UK, but not in Gennany. The greater severity of labor market problems, and the greater inherited supplies of skilled labor, in East Germany than in the UK might well explain that but - that cannot account for the failure of publicly (as opposed to privately) finnded occupational training in West Germany as well. Work-based, job- oriented training appears to have helped participants to find employ~nentin the UK, but not to a greater extent than has occupational training conducted off-the-job, and it is more prone to displacement-related losses in the aggregate. Equity The assistance provided by British training programs to participants may be " Work-based training. confined largely to Einarbeittmg, accounted for only 8 % of East German trainees in 1993 (Table 13.1, above). as compared to 44 and 39 % in ET and JTS respectively (Payne et al., 1996. Table 2.1: 1999. Table 2.2.). PUBLICLY FUNDED TRAINING FOR UNE MP LO YE D ADULTS 285 valued on grounds of equity as well as efficiency, to the extent that participants are drawn from the socioeconomically disadvantaged. Similarly, even if East German programs have not improved labor market prospects for participants, they may at least have made them better off during the program than they would otherwise have been.I6 In such cases, equity-related benefits are particularly important, although by themselves they do not justify providing training, as opposed to simply income support, which costs less. The presence of equity benefits depends on two conditions. Were participants disadvantaged, econo~nicallyand socially? If so, did the program make them better off, taking account of the participation period, not just subsequent labor market experience? Evaluation research often provides evidence on the first question, by way of the participation equations estimated as part of controlling for selection bias. A mixed picture emerges. In the UK, participants have been character- ized disproportionately by an absence of vocational qualifications, a history of frequent and long-term une~nploymentand membership of a racial minority - criteria all associated with disadvantaged status. At the same time, selection into the two programs with a work-based component (ET and TFW) was slanted towards the more educated, reflecting the interest of sponsoring employers in picking the more promising trainees. Similarly, JTS participants included many female re-entrants to the labor force, whose economic circumstances were unlikely to have been as dire as those of the long-term unemployed. In East Germany, participation was slanted towards females and the previously unemployed - groups particularly prone to unemployment. Again, however, participants tended to be more highly educated than non- participants, consistent with the high educational requirements of occupational training in Germany and the widespread need of even skilled workers for training to help them deal with the changed requirements of the post-unification job structure. Neither the British nor the East German programs have therefore been as strongly oriented towards disadvantaged workers as have most of their US counterparts, although taken as a whole they too have probably redistributed resources towards the less advantaged. The second question concerned the effect of participation on income during training. Evidence on this has typically not been provided by European evaluations, unlike their American counterparts. In the US, the widespread absence of entitlements to unemployment or welfare benefits means that programs that provide training allowances tend to increase incomes during participation. In other countries, including Sweden and Korea, participation in training can improve current income by renewing or '"he Supported Work Demonstration program for disadvantaged youth in the US showed similar attributes: no improvement in subsequent labor market prospects, but a gain in income during the program (Ryan and Bhchtemann, 1996, Tables 2,3). 286 Labor Market Reforn~sin Korea: Policy Optionsfor the Fn[zue extending entitlement to unemployment benefits. In East Germany and the UK, however, the rules for benefit eligibility suggest that participation probably had little effect on incomes. In Germany, participants remained eligible for their normal unemployment-related benefits. Only the few who gave up a job that might have been expected to last - an unattractive move, particularly under the circumstances of the time and place - in order to take training therefore faced lower incomes. In the UK, under JTS, although fully 83% of participants received a public training allowance, 39% saw their living standards as having been reduced by participation." Under ET and TFW, things were simpler: participants received any unemployment benefit to which they were entitled plus a training premium of Ł10 (around 20,000 won) per week - which should have made them better off than had they not participated, but not so strongly as to induce many to participate purely to receive the income. In sum, German and British programs show equity benefits, in that they tended to serve the disadvantaged disproportionately and make them better off than they would otherwise have been. Their main contribution may well have been to stir the unemployment stew, breaking down the 'dumplings' of long-term unemployment into the 'lumps' of short-term unemployment. To the extent that the pain of unemployment is distributed less unequally as a result, equity improves; to that extent that the total pain caused by unemployment is also reduced, efficiency does too. Korean Training Programs Like East Germany in 1990, Korea experienced in 1998 a rapid growth in unemployment in the aftermath of a major external shock. Employers made use of their enhanced discretion to lay off workers, particularly those employed on temporary and daily contracts. Unemployment rose to 8.4 % early in 1999, since when it has declined markedly (KLI, 1999). Although the peak unemployment rate was well below those of the UK in the 1980s and East Germany in the 1990s, the low coverage, limited benefits and tight eligibility requirements of the emergent Employ~nentInsurance System meant large increases in inequality, poverty and hardship (Y.-b.Park 1999). As in East Germany, public intervention curbed the increase in unemploy- ment. Active labor market policies alone covered during 1998 more Korean workers than were formally unemployed (1.60 and 1.46 million, respective- ly; KLI, 1999; OECD, 1999, Table 39). In Korea too, training played an important but not a dominant role in labor market policy. Participation in training as a whole expanded eight-fold between 1997 and 1998 (Table 13. 7). Training accounted in 1998 for 21% of participants in, and 36% of public ''Data taken from an unpublished interim report, kindly provided by Joan Payne. expenditure upon, active labor market policy (;Did.). Program Attributes These training programs built upon and extended the existing training system. Established programs, including Employment Promotion (here EPTU) for the uninsured unemployed and Manpower Development, expanded strongly (Table 13.7). The fledgling program for the insured unemployed, Re-employment Training (here RETU), expanded particularly dramatically, moving during 1997-98 from the smallest to the largest training program for the unemployed. New training programs were created specifically to cater to displaced urban workers prepared to retum to farming. those who had just lost their jobs, and those interested in forming their own businesses. In practice, these programs overlap considerably. The two largest, RETU and EPTU, differ more in the constituencies they serve and the income they provide to participants than in the services on offer. Taking the three main programs as a whole, breakdowns by participant attributes indicate that they have served a broad clientele, in terms of sex, age, schooling and previous occupation, and that a wide range of destination skills has been involved (Table 13.8). Has the distribution of training been equitable amongst the unemployed? A comparison of the attributes of trainees in mid-1998 with those of the unemployed in 1998 as a whole suggests that participation was broadly representative of the unemployed in terms of sex; that it was slanted Table 13.7: Training activity by program in Korea, 1997 and 1998 - -- 1997 1998 Program (persons) (persons) Re-Employment Training For The 1,900 162,400 Unemployed (RETU) Employment Promotion Training For The 25,900 96,600 Unemployed (EPTU) Other : 0 4,100 Agricultural Returners Newly Unemployed 0 36,600 Business Startups 0 11.600 Manpower Development Total Source. OECD (1999),Table 40. 288 Labor Market Rejbrnzs in Korea: Policy Optionsfor rl~eFLI~ZII-e Table 13.8: Attributes of training participants and of unemployed workers in Korea, 1998 -- Trainees') Unemployed (%I (%) June 30, 1998 1998 (stock) (annual mean) Sex Male Female Age <30 years 30-49 50+ Schooling 0-12 >I2 Program RETU 43.9 EPTU 41.7 Manpower dev. 14.4 Occupation Office/clerical 57.0 (previous) Production 43.0 Occupation Machinery 20.1 (training) IT 16.7 Service 16.6 Management 17.2 Other 5.6 - Note : na.:not applicable or not available 1) participantsin the three principalpublic training programsonly (Table13.7, above) Sources:Ministry of Labor, Korea(January1998). Training for the Unemployed. OECD (1999),Table 35. moderately towards the middle-aged rather than older workers; and that it was slanted quite strongly towards the more educated (Table 13.8). The two latter differences marked East German programs as well. Differences in participation rates by age and education may reflect differences in interest in training, but they may also reflect differences in the availability of training, amongst eligible workers. Further information on the attributes of Korean training programs is available from a recent evaluation study (Kang and Lee, 2000), covering the training received during 1998 by samples of around 1,000 participants in the two main programs (RETU and EPTU)." Nearly two-thirds of sample ''The sample's representativeness of trainees as a whole remains unclear. Althougli the authors sought a randomized selection of participants. they depended on local Employment Service offices to provide names of EPTU participants. At that stage, randomization may not have prevailed over any inclination on the part of public officials to select successful or otherwise more attractive participants.The likelihood that sampling was non-random is increased by the disproportionately high share of females, and low share of young people, in the resulting participants were married, around one-third were heads of households. Roughly one half of trainees had been out of work for at least six months before participating, though many, particularly feinale EPTU participants, had been economically inactive. As many as nine-tenths of the previously employed had worked under temporary or daily contracts. In terms of training itself, the average duration of participation in the two main programs amounted to 5 months, a period similar to that of the UK but less than half that in East Germany. Completion rates averaged around 90 %. Approximately one-fifth of RETU participants, but almost no EPTU ones. took two training courses. Around two-thirds of participants were trained by private institutions, with RETU and EPTU diverging markedly, at 36 and 90 % respectively. As has traditionally been the case for Korean vocational education and training (OECD, 1998), almost nobody was trained on the job or at the workplace. Only one in ten trainees reported having received guidance or counseling from the public employ~nentservice, and only one in four froin a training provider. Around one in four expressed dissatisfaction with their access to particular counseling services, while more than one half criticized the job search and introduction services, if any, that had accompanied their training (ibid., Tables 7, 9, 12, 13, 17, 18, 19,22). Income during training appears to have differed greatly according to program. RETU participants received their benefit entitlement under the EIS system. Participation allowed them to extend their period of benefit eligibility, providing a potentially important motive for participation amongst unemployed workers without interest in or need for training (Yoo, 1999). The same applies, albeit to a lesser extent, to EPTU participants, who do not enjoy EIS coverage, and who have been eligible only for a low training allowance. Those who accept training for traditional manual work in '3-D' occupations have received a special training allowance intended to counter aversion to that kind of work. Although training allowances have not been high relative to average earnings, the limited alternatives available to many of the jobless mean that their training-based incomes probably made many participants better off during training. Some potentially important attributes of training remain unclear - notably the extent to which trainees acquire recognized vocational qualifications, which may help them find work and raise their pay; the balance between upgrade training, retraining and remedia1training; the quantity and quality of training received; and trainee satisfaction with training. On the latter score, a mixed picture may be inferred from the finding that, while 93 % of public trainees interviewed by a 1998 survey rated the quality of their training as 'good' or better, 83 %criticized its scope or effectiveness (Kim et al., 1998). In sum, Korean training programs have resembled their East German counterparts of the early 1990s in terms of the rapidity of their expansion, sample compared to the overall trainee population (Kang and Lee. 2000, Tables 3,6,7; Korean M~nistryof Labor 1998). 290 Labor Market Refoi,ms in Korea: Policy Optionsfor the Fzltui-e their initial emphasis on breadth of coverage at the expense of depth of learning, their reliance upon training divorced from the workplace, and the range of participants covered and of training courses provided. The parallel is ominous, given that German training programs appear not to have improved participants' labor market prospects. Program Evaluation Have Korean training programs worked? Employment rates immediately after training have been low, indeed no higher than employment rates for the non-participant unemployed. Policy failure has been widely inferred in Korea as a result (section 1, above). The evidence is indeed less than encouraging, but gross outcomes, such as post-training employment rates, cannot indicate the contribution of training (section 3, above). More direct evidence has recently been provided by a quasi-experimental evaluation of the main program (RETU; Kang and Lee, 2000).'9 The evaluation focuses on an employment-related outcome, the transition rate to employment nine to fifteen months after training, for 1,000 unemployed workers who had separated from employment between mid-1997 and the end of 1998 and who had participated in a RETU course during 1998. The comparison group, comprising around the same number of unemployed workers, comprised those who had also separated from a job in the same period but who had not taken training during 1998." As the participant and comparison groups differed significantly in terms of age, education, marital and household status ([bid., Table 6), regression analysis was used to control statistically for those differences, within a proportional hazards model of the time path of the probability of leaving unemployment for employment after training. In terms of gross outcomes, the two groups' experiences were quite similar: roughly 40 % of each was in employment at the end of the observation period, and around 50 % had held a job by then (ibid., Tables 18,19). Regression analysis suggests however that, controlling for the observed attributes of the two groups, the probability of moving into a job after training is significantly higher for trainees than for non-participants I" The absence of any administrative data from which information on eligible EPTU non- participants might be derived made it impossible to evaluate EPTU and RETU together, as would clearly have been preferable. ''I It is not known how many of those workers received other services, including non-RETU training and job creation; to the extent that some did. the benefits of RETU training are underest~~nated(contamination bias). Similarly, the decision to exclude non-participants who found a job during the time that participants were in training. as well as to ignore job finding during training by participants (the 'no job search during training assumption' which ignores the 4 % of sample participants who dropped out to take employment; Table 17), distorts the comparison of outcomes between the two groups - biasing the estimated benefits of training upwards and downwards respectively. (ibid.,Table 24). The training effect is found however only for ferslales, for whom training is estimated to have increased the probability of moving into work by more than one half. For men the estimate is positive but insignificant. The evidence suggests therefore that Korean training programs have indeed helped their participants, at least the females amongst them. But it cannot be treated as conclusive. The load upon regression analysis in correcting for large differences in measured attributes of the participant and comparison groups is increased by the absence of any close matching on those attributes. Moreover, while the set of measured attributes is no smaller than in most evaluation research, it also excludes such potentially important ones as ability and motivation, around which selection into training may well have occurred. As the econometric model does not address that endemic problem, uncontrolled selection bias may well account for part or all of the benefits attributed to the training it~elf.~' Korea's first proper evaluation study has underlined the inadequacy of short-term gross outcomes and suggested that training may have worked, in the specific sense of improving access to employment for participants. The robustness of the conclusion is open to question, and, even if it is firmly grounded, consideration would also have to be given to both any losses incurred by displaced non-participants and to the costs of training before giving public programs the 'thumbs up'. But the case for public training programs at least remains alive in the Korean context - as does that for further evaluation research, building on a promising start. Conclusions The merits of public training programs as a response to mass unemployment and long-term unemployment remain uncertain, despite a voluminous evaluation literature. This paper has surveyed the evidence for two advanced economies with extensive experience in the field: Germany and the UK. The emphasis has been placed on the policies addressed to mass employment in the 1990s in East Germany and long-term unemployment in the 1980s in the UK. The results prove mixed. The East German training effort appears to have failed its participants, notwithstanding its high cost and its orientation to occupational skills. British programs have, by contrast, helped their participants to find work and - when accompanied by occupational training and vocational qualifications - improve their earning power. In both countries, training programs have tended to make a clientele that exhibited a The findings may also be distorted in favor of training by the lack of any correction for the time that trainees spend in training, but non-participants do not, during which re-employment rates were presumably depressed for trainees relative to non-participants. 292 Labor Market Reforins in Korea: Policy Optionsfor the Fzltzlre variable degree of socio-economic disadvantage better off while partici- pating in training than it would otherwise have been in unelnployment - though simple income support offers a cheaper way of doing so, were that the only benefit. The pattern of results is however too varied by country and type of training to generate a coherent account spanning the two countries' experiences. The ilnplications for Korean policy are correspondingly limited. Given that, the advent of evaluation research devoted to Korea itself is particularly welcome. The first formal evaluation of a Korean program suggests that training has increased ernploy~nent prospects for its participants, particularly for females. The conclusion is potentially sensitive to choice of data and modeling technique, and proble~nsof aggregation and cost-benefit comparisons remain to be addressed. The door has however been opened to further research, which may provide a Inore coherent picture than has emerged for Germany. The pessimistic alternative, however, has to be acknowledged: unless a well-designed social experiment is used, evalua- tion research, however sophisticated, cannot lay claim to the robustness required for effective policy making. i 'I LABOR MARKET AND GENDER 44Were Korean Women More X Affected by the Asian Crisis?* Kye Woo Lee" Kisuk Cho" Introduction Between June and December 1997, a group of newly industrialized Asian countries were hit by a severe financial crisis. Compared with the recent financial crises in other regions, the depth and the costs of this crisis were much more serious. The stock market tumbled, and the national currencies depreciated drastically. During the crisis period, equity prices in Korea dropped by 51 %, and those in Indonesia and Malaysia declined by 45 %. Nominal exchange rates in Indonesian rupiah increased by 122 %, and in Korea, the won depreciated by 80 %. Non-performing loans accounted for between 35 to 75 % of the banking system's total loans, with resolution costs estimated at 45 to 60 % of GDP, and Korea's GDP itself declined by 6.7 % in 1998. The contagious Asian financial crisis aggravated the already distressed Russian economy, negatively affected Latin American economies, and subsequently threatened the stability and growth prospects of the world economy (K. Lee, 1999). The crisis, which started in the financial sector, quickly spread to the real sectors of the economy, which in turn adversely affected the labor market. The stable unemployment rate of 2.0 % in 1995 and 1996 rose to 2.6 % in 1997 and 6.8 % in 1998, and finally reached 8.6 % in February 1999. Accordingly, the number of unemployed people increased from the 1996 ' The authors are indebted to Drs. Pilhwa Chang, Hyung Cho. and Jun,vl Yun for their constructive comments on an earlier draft. The authors also gratefully ack~iowledgethe prelim~naryanalysis made by Dr. Sun Ju Lee and Hye Won Yu. and etficient research assistance provided by Hye Yun Park, Haedong Park, and Young Reeln Clieong, who are all staff members of the GSIS. Ewlia LVomans University. However, the a~~thorsalone are responsible for any remaining errors and shortcomings. -* Ewlia Womais University, Korea. 296 Lcrbor Market Refornzs in Korea: Policy Optionsjor the Fzrtzrre base figure of 426,000 to about 1.5 million in 1998. By comparison, the World Bank's unemployment forecast was around 850,000 persons. This enormous gap between the forecast and actual unemployment levels is indicative of the severe impact of the crisis on the labor market. This incremental unemployment resulted from bankruptcies, downward adjust- ment of existing firms, and the economy's inability to absorb new entrants into the labor force. From the onset of economic reconstruction, experts from international organizations advised the Korean government to pay special attention to the strengthening of social safety nets, especially for the most vulnerable groups of society, such as wornen, children, and the poor (E. Lee, 1998). As observed in previous crises elsewhere, these groups might bear the brunt of adjustment costs. The measures to ease the social costs of adjustment would complement the other structural reforms and help mobilize social and political consensus on the adjustment process. Scholars and social activists, as well as feminist leaders, expressed concerns about the socially underprivileged groups, especially women, and called for special action programs to protect them. The purpose of this paper is to analyze the effect of the Asian economic crisis on women in the Korean labor market in co~nparisonwith the pre- crisis situation. Particular emphasis will be placed on two questions: (a) Were women more adversely affected than men? and (b) If so, did men discriminate against women? In the following section, we will compare labor market experiences between male and female workers. For this purpose, nation-wide official survey statistics will be analyzed. This study has found that it is difficult to state categorically that female workers were more adversely affected than male workers. The crisis affected both male and female workers adversely. However, one can safely conclude that female workers had different experiences in the labor market, especially in tenns of involuntary disengagement and rehiring status, and were continuously treated differently with regard to labor earnings compared to before the crisis. From these findings, some observers may quickly conclude that the different gender experiences reflect gender discrimination. Therefore, we will next briefly review literature on the theories of gender discrimination to find out which theory most accurately explains the different gender experiences in the Korean labor market. The ensuing sections will test the theories on the basis of nation-wide official survey data published by the National Statistics Office and the Ministry of Labor, as well as a set of regional data collected through a field survey specially designed for this study. The final section will suggest future labor market policies based on the findings. Women in the Labor Market To compare the labor market experiences of males and females before and after the eruption of the economic crisis, five labor market indicators are selected for analysis: labor force participation, unemployment, employment, involuntary disengagement (in particular, dismissals), and labor earnings. Through the analysis of these indicators, it should be possible to state whether female workers experienced greater hardship than their male counterparts. Labor Force Participation Before (in 1996 and 1997) and after (in 1998) the Korean economic crisis, the size of the population 15-years of age and older increased by about 1.5 % per year. However, the economically active population (labor force) decreased by 1.0 %, in contrast to an increase of 2 % in 1996 and 1997 (National Statistics Office, 1998 and 1999). This drop originates mainly from the drastic decrease of 3.7 % (329,000 persons) among females, which more than offset the increase of 0.9 % (121,000 persons) among males. Accordingly, the labor force participation rate decreased from 62.2 % in 1997 to 60.7 % in 1998, with this decrease being more pronounced among females (from 49.5 % to 47.0 %), than among males (from 75.6 % to 75.2 %). As a result, the increase in the population of econo~nicallyinactive females was 6.5 % (585,000 persons), compared with 3.4 % (141,000 persons) for males (Table 14.1). Unemployment One of the most noticeable changes between the pre and post economic crisis periods was the sharp increase in unemployment rates from 2.6 % to 6.8 %. This increase was more dramatic among males (from 2.8 % to 7.6 %) than among females (from 2.3 % to 5.6 %). Consequently, females' share of total unemployment declined from 36.7 % in 1997 to 32.7 % in 1998 (Table 14.1). On the basis of unemployment statistics alone, therefore, it is difficult to accept the argument that the econo~niccrisis affected female workers more adversely than male workers. However, some labor market analysts argue that these unemployment rates mask a more serious unemployment situation among females than males for several reasons. First, employment decreased faster among females (by 602,000 persons) than males (by 510,000 persons), i.e., by 6.9 % for women versus 4.1 % for men between 1997 and 1998. Consequently, females' share of total employment declined by one %age point. Second, the stark increase in the economically inactive population among females versus males (by 6.5 % vs. 3.4 %) may reflect a !bountiful 298 Labor Market Refornzs in Korua: Policj~Options for file Fzltz~re increase in the hidden female unemployed -- female workers who were so discouraged by employment conditions that they withdrew from the labor force entirely. This may explain the sharper drop in female labor force participation rates than those of males in 1998. Some researchers estimate hidden unemployment at as high as 90 % of those who lost jobs, with only 10 % of them remaining as open unemployment in the labor market (D. Kim, 2000). Other researchers estimate that if hidden unemployment were included, the female unemployment rate in 1998 would increase to 11.4 % and would be higher than the male une~nploymentrate of 9.5 % (T. Kim and Y. Moon. 1999) (Table 14.2). Table 14.1: Economically activepopulation,employmentand unemploy- ment in Korea, 1996-1998 Change over the year 1996 1997 1998 1995-96 1996-97 1997-98 (000's) -- Population>l5 years-old 34,285 34,842 35,362 621(1.8) 557f1.6) 520(1.5) Male 16,612 16.888 17,150 - 276(1.7) 262(1.6) Female 17,674 17,955 18,211 - 281(1.6) 256(1.4) Econ. active pop. 21,243 21,662 21,456 390f1.9) 419(2.0) -206(-1.0) Male 12,636 12,772 12,893 - 136(1.1) 121(0.9) Female 8,607 8,891 8,562 - 284(3.3) -329(-3.7) Employment 20,817 21,106 19,994 - 284(1.4) -1112(-5.3) Male 12,345 12,420 11,910 - 79(0.6) 510(-4.1) Female 8,472 8,686 8,084 - 205(2.4) 602(-6.9) Unemployment 426 556 1,461 - 130(30.5) 905(162.8) Male 291 352 983 - 61(21.0) 631(179.3) Female 135 204 478 - 69(51.I) 274(134.3) Econ. inactive pop. 13,043 13,180 13,906 232(1.8) 137(1.1) 726f5.5) Male 3,976 4,116 4,257 152(4.0) 140(3.5) 141(3.4) Female 9,067 9,064 9,649 80(0.9) -3(-0.0) 585(6.5) Labor force (62.0) (62.2) (60.7) (0.1P) (0.2P) (-1.5P) participation rate ('A) Male (76.1) (75.6) (75.2) (-0.4P) (-0.5P) (-0.4P) Female (48.7) (49.5) (47.0) (0.4P) (0.8P) (-2.5P) Unemployment rate (%) (2.0) (2.6) (6.8) - (0.5P) (4.2P) Male (2.3) (2.8) (7.6) - (05.P) (4.8P) Female (1.6) (2.3) (5.6) - (0.7P) (3.3P) Female's share (31.7) (36.7) (32.7) (5.0P) (-4.OP) Note: Numbers in parentheses are changes in %, and "P" indicates %age points. Source: National Statistics Ofice, Korea (1997-1999). 1996-1998 Annual Report on the Economi- cally Active Population Suwey. Table 14.2: Unemployment(includingHidden Unemployment)in Korea, 1996-1998 Total Male Female Total Male Female Female share ('000) (%) (%) Source: Kim and Moon (1999) However, if hidden unemployment were included, female unemployment rates would be higher than male rates even before the economic crisis, and females' share of total unemployment would decline from 48.4 % in 1997 to 44.3 % in 1998 since male unernployment increased more sharply than female uneniployment. Therefore, on the basis of unemployment rates alone, it would be difficult to come up with an unequivocal judgement on whether the 1997 economic crisis affected female workers more severely than male workers. Using panel data constructed on the basis of the economically active population of 1995-1998, D. Kim (2000) reached more or less similar conclusions. He also shows that if the unemployment spell is measured not only by the employment rate of the unemployed, but by the transition to the economically inactive status, including hidden unemployment (those un- employed who decided to drop out of the labor force entirely), female unemployment spells were shorter than those of males (Tables 14.3 and 14.4). Table 14.3: Estimated unemploymentspell based on employmentrates alone in Korea,1996-1998 1996 1997 1998 (months) Total Male Female Source: D. Kim (2000) Table 14.4: Estimated unemploymentspell (including transition to economi- cally inactive population)in Korea,1996-1998 1996 1997 1998 (months) Total Male Female Source: D. Kim (2000) 300 Labor Market Refirms in Kor-ea: Policy Optionsfir the Ftlttll-e On balance, one can conclude that although unemployment rates do not support the hypothesis that female workers were more adversely affected than male workers during the Asian crisis, female workers also suffered acutely. Employnient Although unemployment rates were higher among male workers than females after the economic crisis, the level of female worker employment declined more rapidly than that of male workers, suggesting that the Asian crisis inflicted greater damage on the level of female worker employment. Between 1997 and 1998, male employment declined by 510,000 workers (- 4.1 %), but female employment dropped by 602,000 workers (-6.9 %) (Table 14.2). Consequently, female's share of total employment declined by one %age point. These seemingly contradictory phenomena of a sharper decline in female employment, yet a lower feinale unelnployinent rate arose partly because new male entrants into the labor force experienced greater difficulty in getting jobs than their female counterparts (Table 14.7). But a more important aspect of the explanation is that once female workers lost their jobs, they were more inclined to withdraw from the labor market than their male counterparts, as shown in the sharper increase in the economically inactive population among females than males. D. Kim's (2000) panel data support this argument. His panel data study confirms that since the outbreak of the economic crisis, employment rates among the unemployed declined among both male and female workers and reached more or less the same level (around 22 %) in 1998. One difference is that the reemployment rate was higher among female workers before the crisis. Therefore, after the crisis, female reemployment rates declined more rapidly than those of male workers. One more important difference is that job-seeking rates were lower among unemployed females than among unemployed males since un- employed female workers were more likely to withdraw from the labor force as their employment rate shrank. Interestingly, female employment fell relatively deeper in those industries and occupations where female workers were more represented, and male workers' share in those industries and occupations increased after the economic crisis. This is indicative of some degree of substitution of male workers for female workers during the economic crisis. For example, while employment of male workers increased in the electricity, transportation, storage, and finance industries, as well as in those businesses and services where female workers were relatively more represented, the number of female workers in these industries declined (Table 14.5). By occupation, while inale workers increased in administration, service, and sales occupations, where female workers were predominant before the crisis, the number of female workers employed in these occupations declined sharply Table 14.5: Employed persons by industry in Korea, 1997-1998 Change over the year, 1997-1998 Total Male Female Change Change Change Change Change Change ('000) rate (%) ('000) rate (%) ('000) rate (%) Total -1,112 -5.3 -510 -4.1 -602 -6.9 Agr., forestry&fish. 95 4.0 63 5.1 31 2.7 Mining & manuf. -589 -13.1 -335 -11.5 -253 -15.8 Manufacturing -584 -13.0 -332 -1 1.5 -252 -15.8 Social overhead -618 -4.3 -238 -2.9 -381 -6.4 capital & other services Constmction -426 -21.3 -352 -19.7 -74 -34.1 Wholesale/retail -234 -4.0 -34 -1.2 -201 -6.6 restaumnts & hotels Electricity, tmnspolt -8 -0.4 32 2.3 -41 -7.1 Storage,finance Business,pn'vate & 49 1 .I 113 4.9 -66 -3.2 publicservices andothers Source: National Statistics Office, Korea (1999), 1998 AnnualReport on the Economically Active PopulationSurvey. Table 14.6: Employed persons by occupation in Korea, 1997-1998 Change over the year, 1997-1998 Total Male Female Change Change Change Change Change Change ('000) rate (%) ('000) rate (%) ('000) rate (%) Total -1,112 -5.3 -510 -4.1 -602 -6.9 Professionals, technicians 48 1.3 32 1.2 15 1.4 Administrative workers & clerks -1 54 -6.0 78 6.5 -232 -16.9 Se~ice/salesworkers -132 -2.7 24 1.3 -156 -5.2 Skilled agr. and fish. workers 91 4.0 62 5.2 30 2.8 Craft, plant & machine -965 -12.5 -706 -12.9 -258 -1 1.6 operators, elementary occuoations Source: National Statistics Office, Korea (1999), 1998 AnnualReport on the Economically Active PopulationSurvey. since the crisis (Table 14.6). By employment status, male workers maintained their status before and after the crisis. However, among regular (permanent) workers, the propor- tion of female workers declined (from 37 % to 32 %), while among daily workers, female workers' share rose substantially (from 16 % to 21 %). This suggests that some of the former"regular" female workers might have shifted their employment status from "regular"to "daily" workers (Table 14.7). 302 Labor Market Refor-msin Kor-err: Policy Optionsfor the Fut~it-e Using the panel data, Y. Ahn (2000) confirmed similar findings in a dynamic context. In the case of males who lost regular jobs, 38 % once again found regular jobs. By contrast, only 3 1 % of female workers landed regular jobs. While 22 % of the males who lost regular jobs found non- regular jobs, 26 % of females landed non-regular jobs. This non- regularization of female workers has been reinforced by the trend among those who lost non-regular jobs. Only 6 % of the unemployed workers who lost non-regular jobs subsequently found regular jobs, and as many as 59 % of them obtained non-regular jobs. For maIes, these rates were 8 %and 53 %, while for females, they were 5 % and 63 %, respectively, demonstrating that female non-regular job holders found it more difficult to acquire regular jobs. These findings indicate that, after the economic crisis, female workers must have increasingly occupied a greater portion of non-regular jobs, unless this shift was offset by an opposing trend among new entrants into the labor force. Unfortunately, the study does not report on the total number ofemployed Table 14.7: Employmentstatus in Korea, December 1997 and December1998 December1997 December 1998 Total Male Female Total Male Female (Yo) (%) Economically active 100 100 100 100 100 100 population Employment 97 97 97 92 91 93 Unemployment 3 3 3 8 9 7 Employment 100 100 100 100 100 100 Wage earners 64 65 61 62 63 60 Non-wage earners 36 35 39 38 37 40 Wage earners 100 100 100 100 100 100 Regular 52 63 37 51 63 32 Temporary 34 24 47 33 24 47 Daily 14 13 16 16 13 21 Non-wageearners 100 100 100 100 100 100 Self-employed 77 96 52 77 96 51 Family workers 23 4 48 23 4 49 Unemployment 100 100 100 100 100 100 Joblosers 58 63 50 75 77 72 New entrants 42 37 50 25 23 28 Source: National Statistics Office. Korea (1998-1999), Monthly Reporl on the Economic Active Population Survey. workers by employment status. However, the nation-wide statistics on total wage earners report that the proportion of daily female workers has steadily increased from 16 % in December 1997 to 21 % in December 1998 and to 25 % in November 1999. Therefore, based on both the nation-wide and the panel data, one can safely conclude that, since the economic crisis, female workers' employment status steadily deteriorated toward non-regular jobs, which are characterized by lower wages and poorer working conditions. Involuntary Disengagen~ent Another study tries to analyze gender differences in unemployment, especially in the causes of unemployment, using unemployment insurance statistics (S. Cho, 1999). When a worker loses insurance membership for voluntary reasons such as old-age retirement, transfer to self-employment, disengagement due to marriage, childbirth and child rearing, continuing education, and enlistment, helshe would not be eligible for unemployment benefits. In contrast, when a worker loses membership for involuntary reasons such as redundancy, resignation at the employer's suggestion, and honorary early retirement, the worker would be eligible to receive insurance benefits. Cho, therefore, tries to analyze whether female workers became unemployed involuntarily for fair and justifiable reasons vis-6-vis male workers. According to the official statistics, male workers became involuntarily unemployed inore often than female workers and therefore appear to be more adversely affected by the econo~niccrisis (Table 14.8). Cho rightly argues, however, that absolute numbers of involuntary unemployment distort the more severe effects of the economic crisis on female versus male workers. The important numbers are each gender's relative share of the 1998 involuntary unemployment among the respective total membership before the economic crisis in 1997. Based on an analysis of the relative share of both male and female involuntary unemployinent by industry, occupation, age, educational attainment, and scale of enterprises, Cho argues that female workers disproportionately suffered from involuntary unemployment, and therefore suspects that there must have been gender discrimination in the course of enterprise reorganization and structural adjustment. The share of female involuntary unemployment in 1998 among the total female members of the unemployment insurance system in 1997 (24.7 %) was greater than that of male involuntary unemploy~nent(19.4 %). This is particularly so in those cases where workers fell into involuntary unemploy- ment because of reorganization redundancy, suggested resignation, and other types of dismissals. In this narrow sense of involuntary unemployment, females' share was about 16 %, compared with 12 % for male workers. This suggests that female workers were selected first and more often as redundant workers in structural reorganization. 304 Labor- Market Refornzs in Korea: Policy Optionsfor the F1rtlli.e Table 14.8; Involuntarydisengagementin Korea, 1997-1998 Total Male Female Total insurancemembership(1997) 4.280.0 Total involuntarydisengagement(1998) (%) change from 1997 (%) of total membershipin 1997 Total involuntarydisengagement Includingmarriage, childbirth,nursing (1998) 935.6 (%) of total membershipin 1997 21.9 Dismissals (1998) (%) of total membershipin 1997 Redundancy ('A) of total membershipin 1997 Suggested Resignation (%) of total membershipin 4997 Other Dismissals (%) of total membershipin 1997 Source: Cho, Soon-kyung (1999) Labor Earnings According to the 1997 and 1998 Survey Reports on Wage Structures (The Ministry of Labor, 1998 and 1999), female workers' average compensation was only 61.7 % of that of male workers in 1998. Although this is a slight improvement over 1997, when female workers' average compensation was 60.1 % of male workers' remuneration (Table 14.9). the gender gap in the average wage remained substantial between the pre and post crisis periods. Moreover, this gender gap in compensation was observed in all occupational groups. Gender differences in compensation are also observed in the distribution of labor earnings. In 1998, while male workers with low levels of monthly compensation (below 500,000 won per month) accounted for 1.3 % of the total male workers, as many as 6.6 %of female workers belonged to this low level coinpensation group. Moreover, while male workers with high levels of total monthly compensation (above 2 million won per month) constituted 26.9 % of the total, only 5.5 % of female workers fell into to this group. The key issue is whether these compensation gaps between male and female workers come from gender differences in reasonable determinants of compensation (such as educational attainment, age and on-the-job experience, hours worked, etc.) or from gender discrimination. Table 14.9: Average labor earnings by gender in Korea, 1997-1998 Special annual Total Index Monthly earnings earnings compensation (Male=100) (won in thousands) 1997 Total Male Female 1998 Total Male Female Source: Min~stryof Labor, Korea (1998-1999). In conclusion, the impact of the Asian economic crisis on women in the Korean labor market can be summarized as follows: First, female labor force participation rates declined more drastically than male participation rates. Second, although female unemployment rates increased, male unemploy- ment rates rose more sharply and reached a higher level than female unemployment rates. Third, when hidden unemployment is taken into account, female unemployment rates still stood higher than those of their male counterparts. However, the same trend was observed even before the crisis. Moreover, male unemployment rates rose faster than those of their female counterparts, and consequently, females' share of total unemployment declined after the outbreak of the Asian crisis. Fourth, unemployment spells were similar between male and female workers, and were even a bit shorter among female workers when their transitions to the status of economically inactive population were included. Fifth, female workers with lower levels of education and with jobs in female-crowded industries/occupations experienced unemployment in a greater proportion than their male counterparts, and some male workers substituted female workers in those industries and occupations. Sixth, the female employment level declined at a faster rate than the male employment level and unemployed female workers tended to withdraw from the labor force more easily. Seventh, when female unemployed workers were reemployed, they 306 Labor Market Refornzs in Korea: Policy Optionsfor. the Fzltu/-e tended to obtain non-regular (daily) jobs more often than their male counter- parts. Eighth, incidences of involuntary unemployment were higher among males than females; however, the relative share of involuntary unemploy- ment in total unemployment insurance memberships of each gender group was greater among females, and they suffered more from enterprise reorganization and structural adjustment. Ninth, the average earnings of female workers were lower than those of their male counterparts, just as before the crisis. From these findings, it is difficult to draw a definitive conclusion as to whether or not female workers suffered more from the Asian economic crisis. Both male and female workers were adversely affected by the crisis. However, one can safely conclude that male and female workers were treated differently in the labor market, especially in terms of involuntary disengagement and rehiring status during the crisis, and in labor earnings both before and after the crisis. Some observers therefore argue that the difference in the effects of the economic crisis on men and women in the labor market originates from gender discrimination in the labor market. Hence, as a basis for formulating future labor market policies, it is natural that we review the various theories of gender discrimination and find out which theory best explains the gender differences in the Korean labor market. Theories of Gender Discrimination in Labor Markets The existing literature on the theories of gender discrimination fall into four categories. First, we are concerned with the neo-classical economics approach, which stresses the importance of human capital. We then move on to the institutional approach, which emphasizes the structure of labor markets, in general, in terms of political and institutional factors. Third, we consider the feminist approach, which also explains the structure of the labor market in terms of political and institutional factors, but with particular emphasis on gender discrimination. Finally, we discuss the risks and costs approach in explaining gender discrimination. Neo-Classical Economics Approach The neo-classical economics approach regards the labor market as just a commodity market where workers sell their labor and are paid according to their productivity, which is affected by human capital such as education, qualifications, training, skills, and experience. It is assumed that the behavior of rational individual workers is to maximize their wages and invest in the development of their own human capital because they expect a WERE KOREAN WOMEN MORE AFFECTED BY T H E ASIAN CRISIS? 307 fair return on this investment in terms of future earnings (Bayes, 1988). According to this precept, neo-classical economics theorists try to explain gender discrimination in the employment sphere in terms of women's rational choices and preferences in the labor market. Becker (1985) argues that women's taste and preference for household work would make them spend less energy and time on paid work than their male counterparts, and they would therefore invest less in human capital to raise their productivity. Hence, their lower earnings can be explained. For Becker, domestic division of labor operates to achieve a full equilibrium within the household; that is, it reflects household rationality by stipulating that women are responsible for household work and other non-economic activities, and men for economic activities. Thus, crowding occurs in a relatively small number of occupations where women find it easier to combine paid work and household responsibilities, and women compete with one another in those occupations. Abundant supply of women's labor in certain occupations is a predominant factor in gender segregation of occupations. This in turn pushes their wages down in the labor market (Mincer and Polacheck, 1974; Becker, 1985; Dunn, 1996; Figart and Mutari, 1999). However feminist critics argue that a fundamental problem with neo- classical economics theories lies in the assulnption of a competitive labor market in which workers are paid according to their worth (Walby, 1990; Crompton and Sanderson, 1990). According to their arguments, neo- classical economics theorists ignore the ways in which power relations structure the labor market unequally for men and women. Occupational segregation and gender-based wage differentials appear not as a result of the rational choice of women, but as the reflection of the labor marlcet's structural problems (Hanson and Pratt, 1995). Institutional Approach The institutional approach challenges the neo-classical economics approach by contending that the structure of the labor market is explained not by equilibrium between supply and demand but by a number of political and institutional factors. This approach argues that through a set of administra- tive rules and procedures employers create segmented labor markets, between which competition is limited, as part of their strategy to enhance profit levels (Doeringer and Piore, 1971; Gordon et al.. 1982; Burchell et al., 1994). However, feminists have criticized that the labor market segmentation theory has developed without any significant concern given to gender since it regards that structured inequality between workers and capitalists is much more crucial than inequality between men and women (England, 1992). Moreover, it treats all female workers as secondary without considering differences among women (Hartmann, 1987; Hakim, 1995). 308 Labor Market Reforms in Korea: Policy Options,for the Fz~tzltr Feminist Approach Feminists argue that the structure of labor markets is determined by political and institutional factors (Crompton and Sanderson, 1990; Bradley, 1989). However. feminists are divided into various groups in explaining those political and institutional factors. Some place emphasis on the capitalist system of production, and others on patriarchy. Still others emphasize cultural sex-typing. Marxist feminists see gender inequality as rooted in social class inequality. It is the capitalistic market system of production that generates social class inequality as well as women's economic dependence on men. They argue that women's inferior position and lower pay in the labor market are formed by the capital-labor relation, and that employers use women as a reserve army of labor to exploit women's labor by regarding it as cheap and disposable (Beechey, 1979). Such a strategy leads e~nployersto replace the male labor force with women in times of economic recession. However, scholars who explain the segmentation of Labor markets and occupational segregation have criticized this reserve army thesis. They argue that women's labor has not competed with or replaced men's labor because a largc number of women have been employed in the new or expanding areas of the economy, rather than having been employed in the core areas that were traditionally recognized as being occupied by male workers (Hartmann, 1979; Collinson et al., 1990). Another group of feminists suggests that the structure of the labor market must be analyzed and explained in relation to patriarchy (Hartmann, 1979; Walby, 1990). Employers and male-dominant unions largely exclude women from certain occupations by generating gender-based internal labor markets and controlling access to jobs and training. Female labor does not compete with male labor due to patriarchal strategy of employers and unions. Other feminists consider such structures of labor markets in relation to sex-typing. Cultural distinctions between male and female work bring about discrimination in the workplace because the work done by women is devalued (Bradley, 1989; Crompton and Sanderson, 1990; England, 1992; Ridgeway, 1997). Thus, women's career ladders tend to be shorter than those of men, and a majority of women are concentrated in the secondary labor market. Labor Market Risks and Costs Approach A theory of statistical discrimination is set up under the premise that rational employers do not have any specific taste or preference, but rather are generally oriented towards maximizing profits. Gender discrimination occurs as they estimate a degree of risk in a world of imperfect information. For employers, it is difficult to measure individzral worker's productivity, but differences in average productivity between groups of workers are WERE KOREAN WOMEN MORE AFFECTED BY T H E ASIANCRISIS? 309 known to employers. Thus, a profit-maximizing employer seeks the most productive workers to fill jobs which require future training, and helshe would prefer to hire workers who have a higher probability to stay on the job longer and contribute themselves to a rise in productivity (Perlman and Pike, 1994). Since the majority of women will have children and, hence, take maternity leaves or give up their jobs, employers may believe that the average turnover rate of women will be higher than that of men (Banton, 1994). The main point of this argument is that women are discriminated against in comparative labor markets as their work discontinuity, caused by childbirth and child rearing, may bring about, statistically. a greater risk of low productivity, and impose high turnover costs on the employer. A corollary to this theory of statistical discrimination is the theory of motherhood protection costs. During the Asian crisis, some argued that the ever increasing costs of protecting female workers might have prevented employers from treating female workers equally and resulted in de facto discrimination against women (The World Bank, 1999). As a basis for their hypothesis, they argue that the legal provisions for protecting women workers in Korea are above the international standards. Moreover, the laws place the burden of financial responsibilities for protecting motherhood exclusively on employers. This indicates a stark contrast with international practices, by which financial responsibilities for most protective measures are shared among the state, employers and employees under a social insurance scheme. Thus, they have argued that employers shy away from hiring women or pay women less to avoid the ever-increasing costs of complying with the legal provisions for protecting motherhood. Indeed, with the advent of the civilian government, which took power in 1987, the outspoken trade unions' and proliferating women's inovements put huge pressure on the government. The late 1980s thus witnessed a series of legislative actions to enhance motherhood protection and gender equality as a sharp reaction to employers' negligence or exploitative practices in the past (E. Lee, 1996). In 1987, the Equal Employment Law was promulgated to improve gender equality in recruitment, training, promotion, assignment, retirement, and dismissal. The amendment to the Equal Employment Law in 1989 ensures equal pay for work of equal value, which was not declared in the 1987 law. The 1989 amendment also broadened its coverage from those enterprises with more than ten workers to those with more than five j~orkers. The 1989 and 1995 amendments not only guarantee gender equality in employment, but also make gender discrimination by employers legally punishable. The Law prohibits employers not only from demanding certain standards of appearance, weight and height (article 6), but also froin discriminating against marital status or gender (article 8). For motherhood protection, the Equal Employment Law guarantees women (or their spouses) with a child under one year of age "childcare leave" of up to one year, which is counted as a working period when the worker's eligibility for other benefits is considered (article 11). This has significant repercussions on 3 10 Labor 'IfarketRefornzs in Koi*en:Poiicy Options for the Fzrttl~-e other types of benefits such as severance and pension payments. The law also requires employers to provide childcare facilities or allowances to help women continue working after childbirth (article 12). According to International Labor Organization (ILO) conventions (No. 156), this responsi- bility is imposed not on employers but on the government, which is to develop and promote public and private facilities. The 1989 amendment to the Labor Standards Law also made it more difficult for employers to assign night work to women by requiring not only the approval of the Labor Minister, as in the past, but also the consent of employees. T h o ~ eprovisions of the Equal Employment Law are over and beyond those of the Labor Standards Law, which have already required more of employers than do the ILO conventions. For example, under the Labor Standards Law, employers should provide a 60-day maternity leave with I00 % regular earnings (article 72). This compares with the ILO convention (No. 103) which stipuIates that two-thirds of regular earnings should be provided not by the employer, but rather either by the means of compulsory social insurance or public funds (E. Kim, 1994). The law also requires employers to provide female workers a one-day menstruation leave per month with full regular earnings (article 71), or extra compensatory payments in lieu of the leave. This provision is a result of the 1989 amendment, which forces employers to grant such leave irrespective of a female worker's request or not. Neither ILO conventions nor provisions of many advanced countries mention the menstruation leave, let alone payments. Only a few countries require such leave by law (e.g., Japan and Indonesia) and define it as an unpaid leave. Some feminist scholars and advocates suggested that menstruation leave should be eliminated, and instead a fetal check-up leave should be introduced (The Women's News, 1999). In addition, the Labor Standards Law banned women's night work, work on holidays, and underground work in any mine, without taking into account any specific circumstances. While ILO conventions (Nos. 45 and 89) allow exceptional cases, the Labor Standards Law permits exceptions only with the consent of the workers concerned and approval of the Ministry of Labor. Although the aims of such provisions are to protect family life and the social reproductive system, the financial responsibilities for implementing such provisions in Korea are imposed exclusively on employers. While provisions for social welfare have increasingly caught up with the trend of advanced countries and have even exceeded the international standards in some cases, the mode of financing the welfare system has not kept up with international practices and has placed exorbitant costs on employers. Therefore, it may be inevitable for a rationaI entrepreneur to avoid hiring more women and to treat female workers differently from male workers, especially when the going gets tough as during the Korean financial and economic crisis. WERE KOREAN WOMEN MORE AFFECTED BY THE ASIAN CRISIS? 311 Empirical Tests of Discrimination Theories Test of the Neo-Classical Approach To test the validity of the neo-classical economics approach, representative variables for human capital were obtained from the government's 1998 and 1997 St~rveyReports on Wage Stvuclt~ves. A function of the human capital approach to labor incomes is expressed as follows: That is, earnings of individual i in period j (Y,,) are expressed as the sum of the returns (r,i)of all hislher previous net investments (C,,)and the earnings from hislher original endowment (Xj,). In the current application, the schooling and on-the-job experience model is employed, following Mincer's approach (Mincer, 1979). That is: where Y is the average annual earnings, E is the number of years of formal schooling, T is the number of years of on-the-job training or experience, and G is a dummy variable for gender. This equation has been estimated for each and all occupations in 1997 and 1998. Nine occupational categories have been used in the analysis: managers, professionals, technicians, clerks, service workers, agricultural and fishery workers, crafts, plants and machine operators, and elementary occupations. The results for all occupations in 1997 and 1998 are as follows (numbers in parenthesis are t-statistics). As illustrated, approximately 56 % and 57 % of the observed variations in earnings have been explained by differences in educational attainment, on- the-job experience, and gender. Coefficients of all variables are significant at 3 12 Labor- Market Refornzs in Korea: Policy Options,forthe Fzlttrre a high level of probability; at less than either the .01(**)or .001("') level and only a few are at the .05 (') level. Those unexplained parts (44 % and 43 %) may be due to the innate ability of workers, scale effects of enterprises, difference in hours worked, etc. The Ministry of Labor has not yet cross- classified the scale effects with the level of education, on-the-job experience, and gender. As to equations (14.3) and (14.4), educational attainment has a highly significant effect on earnings. The level of earnings generally rises with the level of education. An extra year of schooling increases labor earnings (measured in natural log) by about 7 % across all occupations in both 1997 and 1998. This suggests that men and women who acquire a higher level of education would be better off econo~nically.Similarly, an additional year of on-the-job experience or training improves one's labor earnings (measured in natural log) by about 8 %, contributing more than that of an additional year of education. T2 is significant across all occupations. The results may reflect that the productivity level of workers gradually increases up to a certain point in the life cycle and then generally starts to decline as they get older. Therefore, the findings may underscore the rationality of employers' behavior to dismiss inefficient older-aged workers first and more often during the economic crisis. The gender variable, G, has a statistically significant and negative coefficient in both sets of equations. The regression analysis indicates that, first, although the observed average labor earnings gaps between Inen and women were about 40 % and 38 % in 1997 and 1998, respectively, as indicated by the nation-wide wage statistics (Table 14.9), these gaps cannot entirely be attributable to discrimination. And a large part of them (5 % points and 14 % points in 1997 and 1998, respectively) is explained by differences in human capital, as the neo-classical economics approach argues. Second, even after the differences in human capital were taken into account, there still remains a substantial degree of gender gap in average workers' earnings (about 35 % and 24 % in 1997 and 1998, respectively). Women earn less than their male counterparts. Third, the labor earnings gap between Inen and women, after accounting for the differences in human capital accumulation, was closing by about ten %age points before and after the economic crisis. By comparison, this gap was either five % or 25 %, depending on whether women were with or without children, respectively, in the U.S. labor market in 1991 (Furchgott-Roth and Stolba, 1999). The neo-classical economics approach (or human capital theory) attributes this remaining gender gap in workers' compensation to the rational behavior (taste or preference) of female workers, who try to combine paid work and household responsibility in an optimum way by concentrating on a few occupations and thus lowering their earnings vis-a-vrs those of male workers. It is here where various theories of gender discrimination differ in offering reasons for this earnings gap and different treatments between male and female workers for other employment conditions. Contrary to the neo- classical approach, all other theories attribute this and other differences to gender discrimination by employers in the workplace. Unfortunately, nation- wide official statistics do not shed any light on this issue. To test the validity of these other theories, we will have to construct a new database showing employers' behavioral motives. A New Survey Database To date. most empirical studies on gender discrimination have focused on workers. However, if one suspects discrimination in the workplace, focus should be placed on the agent who exercises discrimination. Therefore, in this study, a new database has been constructed to explain the behavior of employers despite the difficulty of data collection. Data were collected through personal interviews, which were conducted by Hyundai Research Institute from January 15 to February 20, 2000. The unit of survey was individual firms and respondents were personnel managers of the firms. A sample of 800 firms located in the metropolitan Seoul area was selected from the Ministry of Labor's annual establishment survey database, using a systematic sampling method. The initial sample list contained 800 firms, but the final sample size was reduced to 114 firms. from which all data were collected. This sharp reduction in the sample size was caused not only by a high refusal rate, but also by the high rate of bankruptcy, closure and relocation of many firms. The sample covered five industries with a relatively high proportion of female workers according to the nation-wide statistics. They are: the manu- facturing industry (25 firms, 42 % female); wholesale and retail trading industry (I7 firms, I0 % female); real estate industry (23 firms, 9 % female); finance industry (21 finns, 8 % female); and transportation and communi- cation industry (21 firms, 4 % female). Although a greater %age of female workers are employed in the health and social services industry, this has been excluded from the sample since it contains mainly gender-wise segregated occupations. The sample also contains more or less the same number of firms distributed by following scale: 30-99 employees (28 firms), 100-499 employees (36), 500-999 employees (21), and above 1000 employees (29). To test the validity of various discrimination theories, the following hypotheses should be tested. Neo-Classical Approach Revisited Hypothesis 1: Female workers are concentrated in a few sectors to combine work and household responsibilities in an optimum way. Female workers were found to be most concentrated in the manufacturing, 3 14 Labor Market Refornzs in Korea: Policy Optionsfor the Futut-e wholesale and retail trading, real estate, finance, and transportation1 communication industries, in that order (Table 14.10). By size, females were more crowded in smaller firms (25 % in firms with 100-500 employees) than in larger size firms (7 % in firms with 500-1000 employees and 10% in firms with over 1000 employees) (Table 14.11). By occupation, females were more concentrated in such occupations as administrative and clerical jobs, sales and service jobs (Table 14.12). However, there is no firm ground to believe that women perceive the female-crowded sectors as more convenient for combining work and family responsibilities. Investigation of female worker's opinions was outside the purview of this study and must be followed up in a future study. Hypothesis 2: The earnings gap between genders is larger in those female-crowded sectors, after controlling for the level of education and on-the-job experience. When the starting earnings of men and women who have no previous work experience and have only a high school or university level of education were compared, there was no statistically significant gender difference in earnings, especially among high school graduates. Statistically, the only significant variation found was that the eamings of female university graduates differed by industry: the highest pay being in finance and manufacturing, and the lowest in real estate (Table 14.10). For university graduates, both female and male earnings were correlated with the size of the firms (p<.025 and p<.002, respectively) (Table 14.1 1.). The smaller the size of the firm, the lower the earnings were for both male and female workers. However, this variation did not show any earnings gap between genders. Nor did the order of the industries have any systematic relation with female concentration. Therefore, the neo-classical hypothesis that the earn- ings of female workers in those female-concentrated sectors are lower than the earnings of male workers is rejected. Again, the neo-classical approach Table 14.10: Starting earnings level by industry in Korea High school graduates College graduates Industry - Male Female Male Female (won) (won) Manufacturing 844,099(48) 806,341(24) 1,104,452(45) 1,107.151(21) Wholesale 876,201(30) 860,814(16) 1,101,737(29) 1,086,680 (15) Realestate 742,433 (32) 700,866 (19) 976,689(25) 960,935 (14) Transp. & comrnun. 975,524(49) 951,615 (26) 1,056,182(42) 1,065,059 (20) Finance 856,795(35) 847,426(18) 1,140,450(38) 1,125,582(19) Total 867,779 (194) 839,197 (103) 1,082,484 (179) 1,075,176 (89) Significance Pc.110 Pc.754 Pc.191 Pc.002 Note: Figuresin parenthesesare the number of persons WERE KOREAN WOMEN MORE AFFECTED BY THE ASIAN CRISIS? 315 Table 14.11: Starting earnings level by size of firm in Korea High schoolgraduates College graduates Employees Male Female Male Female (won) (won) 30-99 813,106 (44) 762,341 (24) 952,471 (40) 916,414 (20) Total 867,779(194) 839,197(103) 1,082,484(179) 1,075,176 (89) Sianificance P<.191 P<.222 P<.002 P<.025 Note: Figures ~nparentheses are the number of persons. Table 14.12: Occupational concentration of female workers and gender earnings gap in Korea Female concentration Female earnings (%) (male=100) All occupations 100 100 61.7 Managers 0 0 82.0 Professionals 12 13 Technicians Adm.1clerks 16 14 64.4 Crafts 10 10 53.8 Plantlrnachineoper. 12 11 61.4 Elementary 9 15 70.5 Sources: Ministry of Labor, Korea (1999); National Statistics Office, Korea (1999 b) is convincing in explaining the gender difference in earnings on the basis of the human capital theory only up to a point, but cannot fully explain it. Test of the Marxist Feminist Approach Hypothesis 3: The cheap and disposable reserve army of female workers replaced male workers during the Asian economic crisis. This theory argues that employers' strategy to exploit female workers has segregated cheap and disposable female workers from male workers, and during the economic recession, employers replace male workers with female workers to maximize their profits. However, during the Asian crisis, exactly the opposite situation developed. As seen above, during the Asian crisis, employment of male workers increased in the electricity, transportation, 316 Labor Market Reforms in Koi-ea: Policy Optionsfor the Fzltzlre storage, and finance industries, while the number of female workers in those industries actually declined. Likewise, employment of male workers increased in administration, service, and sales occupations, while the number of female workers in these occupations declined sharply (Tables 14.5 and 14.6). Therefore, contrary to the Marxist feminist theory, male workers replaced female workers to some extent during the econornic crisis, especially in those industries and occupations where female workers have held a relatively greater share vis-a-vis male workers. Test of the Patriarchal Approach Hypothesis 4: Trade unions are predominated by male employees; and also, trade unions strengthen gender discrimination because of male dominance. Patriarchal feminists argue that male-dominant e~nployersand labor unions create patriarchal structure by generating gender-based internal labor markets and controlling job entry and access to training. Trade unions appeared to be predominated by male employees, who occupied 79 % of union membership. However, males on average accounted for 81 % of the total employees of the firms with trade unions, hence, the hypothesis is not sustained. Hypothesis 5: Trade unions strengthen discrimination against women. An analysis was made on the correlation between the existence of trade unions and discrimination practices in wage determination, disengagement of workers, recruitment, and motherhood protection. Our findings are as follows: (a) We failed to find any statistically significant difference in the starting earnings of the males and females who graduated from high schools or universities regardless of whether or not a trade union exists in the firm. (b) We also failed to find any statistical difference (p<.44) in the rate of disengagement between men (20.1 %) and women (23.3 %) in the presence of a trade union. Neither was there statistical difference in the rate between men (20.6 %) and women (20.6 %) when trade unions were absent. (c) We again failed to find any statistical difference in the rate of recruitment between males and females irrespective of the presence of a trade union (Table 14.13). (d) The only statistically significant relationship found was between the presence of trade unions and motherhood protection practices. Trade unions tended to promote the practice of menstruation and maternity leave. Of the 60 firms that had a trade union, 50 firms (83.3 %) offered menstruation leaves to their female workers. Of the 54 firms that did not have a trade union, 31 firms (57.4 %) practiced menstruation leave. Likewise, of the 60 firms with a trade union, 34 firms (56.7 %) offered maternity and childcare leaves to female workers, whereas of the 54 firms without a trade union, only 19 firms (35.2 %) practiced these kinds of leave. This last finding on the positive role of trade unions on motherhood WERE KO REAN WOMEN M O R E AFFECTED BY T H E ASIAN CRISIS? 317 Table 14.13: Recruitment by firms withlwithout trade unions Firms Firms with trade unions without trade unions Mean Mean Firms Sig.(P) Firms Sig.(P) (persons) (persons) 1997 HS grad Male New recruit Female 1998 HS grad. Male Experience Female 1997 college grd. Male New recruit Female 1998 college grd. Male Experience Female protection practices is consistent with the results of previous research (H. Chung, 1996). Cultural Gender-Typing Approach Hypothesis 6: Women are less loyal and less committed to the company in comparison with men. Cultural feminists maintain that gender-stereotyping prejudice devaluing women's work brings about discri~ninationin the workplace. For example, it is a generally accepted notion in Korean society that women are less loyal, less committed to the job, and less able than men. If this notion is not proved in the workplace, it is really a cultural gender-stereotype. For measuring commitment, we investigated whether there existed any gender difference in the degree of absence, tardiness, taking time off, and working overtime. There was no significant difference in mean attitudes toward work between men (53.2 times per year) and women (39.9 times per year) in terms of tardiness (p<.69) and in taking time off (2.78 times per year for men and 2.20 times per year for women, at p<.46). Regarding absence, however, there was a statistical difference between men (18.5 times per year) and women (2.8 times per year) at p<.015. Men were more likely than women to be absent from work. For measuring loyalty to the firm, we examined whether women were as participatory as men in meetings and meals after work and in overtime work. Although it is debatable as to whether these measures are good indicators of loyalty, woinen were as participatory as men, and there was no statistical difference between men and women in attending after work meetings and meals. Women (2.47 on a scale of 1 to 4), however, were slightly less willing to participate in overtime work than men (2.79). Hence, there is no ground to assume that women are less loyal or committed to the company, and the hypothesis is merely a 318 Labor ;Market Refirnu in Korea: Polic,v Optionsfoi- the Ft~fui*e culturally gender-biased pre.judice. Hypothesis 7: Cultural prejudice works adversely against women's recruitment and disengagement, and earnings level. To test whether the cultural stereotype adversely affects women in the workplace, we investigated differences between Inen and women in recruitment. The results suggested that there were no statistically significant differences in recruitment criteria between men and women, except for the category of qualifications. For both men (34.9 % of the firms) and women (39.7 % of the firms), interview scores were the most highly valued criterion in recruitment, and recruitment exam scores were the second highest, 14.1 % for men and 13.2 % for women, respectively. Statistical differences were observed, however, in the criteria of certificatesilicenses and appearance. Men's (1 1.2 %) qualifications were more important than that of women (5.9 %) at p<.OOI, whereas women's appearance (6.8 %) was more considered than that of men (4.8 %) at p<.024. The results imply that gender-type bias might differently affect the recruitment criteria for male and female workers. In response to one direct question to employers whether they believe men and women have different talents that are suited to different types of work, 46 firms responded positively and 43 firms answered negatively. However, the survey data did not show that the cultural gender-type adversely affected women in recruitment, disengagement, and earnings level. Rather, firms that responded that there were no gender differences in work ability actually hired more women (15.2 and 10.7 persons per company on average) than men (7.8 and 7.1 persons per company on average) at significance level of p<.006 and p<.043, before (1997) and after (1998) the economic crisis, respectively. Labor Market Risks and Costs Approach Hypothesis 8: Low productivity and high turnover rates of female workers due to childbirth and child rearing are the major causes of female disengagement in the workplace. Data showing discontinuity of female workers are prevalent. The nation- wide aggregate data show that female workers have, on average, a shorter on-the-job experience with their current employer than do their male counterparts (4.3 years vs. 6.8 years). The current survey confirms this as well. While male workers have worked 7.2 years on average with their current employer, female workers have worked only 4.7 years. Cho's study (1999) shows that a greater percentage of female workers (5.7 %) than male workers voluntarily disengaged during the Asian crisis (5.7 % vs. 5.1 %). Contrary to the nation-wide statistics; the current sample survey data show that 28.4 % of male workers voluntarily left their current employer while only 20.6 % of female workers left their job voluntarily. However, the current survey data show that the major reason for female workers to cut their career short during the economic crisis was marriage, childbirth and child rearing (68.4 % of total firms), while for male workers, a better offer from other firms was cited. The survey also reports that in those female disengagement cases due to marriage and childbirth or rearing, it was not the firm, but female workers (93.9 %) themselves who initiated the disengagement. However, when 66 firms dismissed female workers due to childbirth or rearing, the major reasons were low productivity (29.8 %), difficulty of replacement (19.4 %) or costly motherhood protection (3.5 %). The rest of the firms do not specify the reason. It can be construed that the firms may foresee risks in hiring women due to discontinuity of work and turnover costs. It is also possible that since the majority of female workers know of the employer's negative perception of their productivity after marriage and childbirth, they may have often voluntarily quit their jobs. Hypothesis 9: Legal provisions demanding motherhood protection above the international standards burden employers with financial responsibilities and encourage employers to discriminate against female workers on the job. The survey data show that motherhood protection costs are not a major concern of employers. The costs do not bring about discrimination in earnings, recruitment, and disengagement. For example, the menstruation leave practice does not encourage gender discrimination in recruitment. Rather, more women (17.2 O h )were hired than men (12.01 %) in companies with the menstruation leave practice in 1998, whereas companies without it do not exhibit any gender differences. The results seem to directly refute the theory of motherhood protection costs. The effect of maternity leave is consistent with that of menstruation leave. Companies with the maternity leave practice hired more women (14.5 and 14.3 persons per company on average) than men (7.8 and 8.0 persons per company on average) in 1997 and 1998, respectively. Companies with this protection practice are more favorable toward hiring women. At least, it does not appear that motherhood protection costs are the direct source of gender discrimination in recruitment. These findings contradict our observation based on the nation-wide statistics that there was some substitution of female workers by male workers during the economic crisis, especially in the female-crowded industries and occupations. One can suspect that the size of the co~npanyis correlated with affordability, the presence of a trade union, and the motherhood protection practice. The motherhood protection practice is more often observed in larger firms, which offer a higher level of earnings in general and also often have trade unions lobbying for the motherhood protection practice. Others may suspect that those firms that practice the motherhood protection measures might have predominantly female workers. Still others may suspect that since costs are one part of the definition of efficiency, employers' negative perception about motherhood protection costs is already subsumed in their perception of the risk in hiring women. 320 Labor Market Rejornls in Korea: Policy Optionsfor the Future Such risk was clearly borne out in this survey in terms of inefficiency and discontinuity of female worker employment, as shown in the test of hypothesis 8. In sum, our survey data do not conclusively support any of the discrimi- nation theories. Nonetheless, they render more support for the cultural gender-type theory and the statistical discrimination or labor market risk theory. From these results, we can infer that these two discrimination theories are the potential candidates to be confirmed with further examination and studies once we improve some defects of the design of our survey. We call for further studies in this area. Lessons Learned and Policy Recommendations Although this study has not brought out any conclusive findings about discrimination theories, it has shed informative light on the issues related to women in the labor market. We may be able to draw lessons useful not only for Korea during the econo~niccrisis period, but also for many developing countries in general. Special Labor Market Action Progranqs for Women During an economic recession or crisis, the government may want to take special measures focused on tackling the problems and issues of female workers separately from the general approach taken for all workers. Although this study could not support the general statement that female workers were more adversely affected by the Asian crisis, there is no doubt that female workers were acutely affected nonetheless. Moreover, this study shows that female workers have gone through a different labor market experience from that of male workers, especially in the incidence of hidden unemployment and involuntary disengagement, reemployment status, concentration of employment, and labor earnings. Special measures that should be taken by the government may include: (i) labor market information systems with special emphasis on monitoring the situation of female workers, (ii) special job counseling for women so as to help them stay in the market and actively search for a job, (iii) special employment services and training programs and facilities for female workers, and (iv) a reinforced labor inspection system with a focus on motherhood protection. Investnzent in Female Capital For an efficient and equal utilization of the human potential of both genders, both feminists and the government should focus on the gender difference in human capital investments in terms of educational attainment and on-the-job training. The human capital theory accounted for more than a third of the observed labor earnings differentials (14 % out of about 38 %) between male and female workers in 1998. This implies that employers have not been completely arbitrary or discriminatory, but have been behaving rationally to some extent. The basis for the Korean employers' rational behavior has been gender differentials in human capital accumulation. 'There has been a substantial difference in the level of educational attainment and on-the-job experience between male and fernale workers. For all occupationaR groups, the average educational attainment of male workers was 12.8 years, but fernale workers, on average, had only 11.8 years. Likewise, the gender gap in on-the-job training and experience shows 6.8 years vs. 4.3 years for males and females, respectively. On a national scale, this gap was further widened in each individual occupational group and in different sizes of firms. Of course, significant progress has been made in females' educational attainment. For example, there is practically no gender difference in enrollments through the high school level, and female's share of bachelor degree earners markedly increased from 37.0 % of all bachelor degree earners in 1985 to 46 % in 1999. However, more progress is needed at higher degree levels. In 1999, female's share of master's and doctoral degree earners was only 31 %and 20 %, respectively. More studies should be carried out on the gender difference in the motives for making different types of human capital investment, and on the ways to help guide females to attain equity not only in the level but also in the type of human capital investment. Besides the gender difference in the level of education and training attainment, differences also exist in the content and field of studies at every level of human capital investment, which would lead to the different gender status in the labor market. Males may focus more on those investment opportunities (e.g., specialties that would bring higher earnings) in which they may be able to enhance their productivity efficiently, while females may do it less efficiently. For example, at the tertiary level, female and male enrollments specializing in social studies have attained equity over time (from 14.6 % vs. 32.5 % between females and males in 1985 to 25.5 % vs. 26.8 % in 1999). However, female students are still concentrated in humanities, arts, and teacher training (21.8 %, 13.01 %, and 7.5 % of all female enrollees, respectively), while much fewer male students are enrolled in these fields (9.4 %, 5.3 %, and 2.3 %, respectively). Male students are specializing more in natural science and engineering (53.0 % in 1999), with which they can earn a higher level of earnings. Although the proportion of female students specializing in this field increased, it has reached only half the proportion of male students (27.5 % vs. 53.0 % in 1999). When the gender gap in a field of studies is substantially closed, such as in social studies, medicine, and pharmacology, labor earnings differentials would be substantially reduced. This is because the gender gap in labor earnings at managerial and professional levels is much smaller than in other 322 Labor Market Reforms in Korea: Policy Optionsfor the Fziture occupational groups, and the field of studies predominated by male students commands a higher level of earnings. Changes in Cultural Gender-Typing The government should prepare a comprehensive plan to change the cultural gender-typing among citizens, especially employers. managers and employees. This study reveals that more than half of the observed labor earnings difference (24 % to 35 %) in the labor market could not be accounted for by the human capital theory. Therefore, the remaining gender difference should have been explained by other theories of discrimination. The study indicates that the cultural gender-type theory and the labor inarket risks and costs theory are more likely to explain the remaining gender difference in labor earnings and other employment conditions. This study confirms that strong cultural gender-typing is prevalent among employers, which may have contributed to some extent to the concentration of female workers in certain occupations and unequal treatment of female workers in the labor market. The government plan should include not only the narrow legal approach, but also formal and non-formal education programs. It should also include a partnership with non-governmental organizations for public education and behavioral changes. Successful cases of active and equal participation of female workers in productivity increase should be identified and disseminated. Changes in Motherhood ProtectionSystems The government should amend the social security system so as to share the motherhood protection and child rearing costs among employers, employees and government. Motherhood protection, a sort of public good, is justified for the betterment of family life and social well being, and therefore the costs should be shared among all members of society in an equitable manner. This study construed that although motherhood protection costs have not been independently shown as a major cause of discriminatory practice, they may have contributed to the employers' negative perception of marriage,. childbirth and child rearing, and about the risks of high female turnover rates. The government should, first, exonerate employers from the heavy burden of financing the motherhood protection costs alone. Continuation of the current system has had a boomerang effect on the equal employment of women. Second, the government should amend the legal provisions for motherhood protection to a more realistic level, and then enforce them rigorously with strong sanctions. The legal provisions should lay down only the minimum baseline, over and beyond which employers and employees should strive to reach an agreement on higher standards through negotiations. Reinforcing the motherhood protection measures ever more generously by legislative actions is not a panacea. Currently, motherhood protection practices are observed in so few firms, especially among small-sized firms. Unrealistic and impractical legal provisions have created a tendency among employers to eschew hiring of female workers and to treat them unequally in personnel management. They have also created a non-compliance culture, and yet no sense of guilt among employers. Third, the Ministry of Labor should resume the periodic general inspection and monitoring of compliance by employers, which it has ceased to do since 1992, thus limiting itself to reacting passively to those cases filed by workers as violations of the labor law provisions. ,rip y r i THE KOREAN EXPERIENCE IN A N INTERNATIONAL CONTEXT i5Labor Adjustment, Non-Standard Work, and Employment Programs: Korea in an OECD Context Gordon Betcherman* Amit Dar* Amy Lirinstra* Makoto Ogawa' Introduction The 15 years prior to the crisis were years of unprecedented growth for Korea. Per capita GDP grew at 7.5% per annum, life expectancy rose by 20%, and enrollment in primary and secondary education reached close to 100%. Korea went from being a lower-middle income country to a nation with a living standard appropriate to the (lower) ranks of OECD countries (with a per capita GDP of US$10,550 in 1997). On the labor market front, employment grew strongly; real wages increased by over 50% between 1990-97; and unemployment rates remained around 2%. This transformation touched all aspects of society and the economy, including the labor market. As development proceeded through the 1970s, 1980s, and 1990s, the need for formal labor market policies and institutions increased accordingly. Indeed, by the mid-1990s, the foundations of an "industrial-era" labor market regime had been put in place including a vocational training system, employinent services, a collective bargaining framework, and uneinploy- ment insurance. The onset of the crisis in 1997 marked an important event for labor policy-makers. Certainly, the crisis raised some issues particular to the extraordinary situation, most notably the need to provide income support to the rapidly expanding pool of impoverished workers. However, the spotlight of the crisis also exposed more generic development concerns about the labor market - for example, inadequate protection offered by the new unemploy~nent insurance system; limitations of existing employment services and programs; and difficulties in sustaining constructive social dialogue between labor, employers, and the government. Fortunately, Korea now appears to have emerged from the crisis with ' Social Protection Unit. Human Development Network, the World Bank. 328 Labor Mnrket Rejornzs in Korea: Policy Options jor the Futzlre impressive momentum. After a decline of 5.8% in 1998, real GDP growth in 1999 was 9.5% and is forecast to expand by 6% in 2000. Unemployment has now fallen below 5%. Real wages have turned around as well with the 1998 decrease of 9.3% followed by a rapid increases in 1999, with estimates as high as 11.1%.I While the immediate hardships of the crisis for Korean workers may have been moderated (if not eliminated) by the recovery, the more durable labor market concerns revealed during the crisis remain as significant issues for the future. But the policy agenda does not stop there. As Korea directs its focus forward, it must now address a new set of issues associated with the "knowledge-based" economy. Future improvements in living standards will increasingly be tied to knowledge-intensive industries and services and success in these sectors will depend in part on a high-quality labor force and on appropriate labor market policies and institutions. Korea thus faces a dual task: completing the industrial-era reforms while preparing for the knowledge era. In the process, policy-makers will need to learn from more advanced OECD countries while assessing the opportunities for "leap- frogging" their experience. Within this context, three labor market challenges are critical for the ongoing development of the Korean economy: First, unemployment will need to be addressed as a regular feature of the labor market. Unemployment rates may not approach the levels that characterize some European countries but the days of 2 % rates appear to be over. Consequently, a long-run strategy for mitigating unemployment shocks through a mix of active and passive policies will need to be established and implemented. Second, developing a skilled and innovative labor force will have to be a priority. In addition to what this means for education, it also will call for an emphasis on lifelong learning and on organizational innovation in the deployment of human resources. Third, the benefits of growth will need to be broadly shared through the workforce. This is closely linked to the preceding two challenges but also to the importance of minimizing the dualism between regular and non-regular employees in terms of labor market regulation and social protection. This chapter considers aspects of these challenges from an international perspective. In addressing these issues, policy-makers will find the experience of OECD countries more relevant than that of World Bank client countries even in East Asia. This corresponds to the reality that, after the decades of growth, the Korean labor market more closely resembles the former group of countries (even with the impressive growth in many of the Iztter countries). However, Korea's particular development path, culture, and institutions are obviously important factors as well. We begin in the next section with a statistical overview of Korean labor market trends. First, we review some of the main Korean structural ' Thrs nas reported In a press release from the Ministry of Labor. February 22. 2000 LABORADJUSTMENT, NON-STANDARD WORK, AND ... 329 characteristics and indicators in an international context. We then summarize how the Korean labor market adjusted to the financial crisis and compare this to the adjustment paths of other affected countries in the region and then selected OECD countries that have also experienced aggregate shocks. The main point of this section is to establish that, even as the Korean labor market has been transformed to a point where it increasingly resembles (some) OECD countries, the crisis demonstrated that it remains a very flexible labor market. In the following two sections, we turn to selected labor market policy issues that are central to the broader challenges identified above. In each case, the discussion of Korean policy is grounded in the OECD experience. First, we look at the growth of non-standard employment, an issue of growing concern in Korea. The second topic discussed is active labor market programs and how they can best contribute to the employability and future earnings of une~nployedor low-income workers. Conclusions are offered in the final section. Korean Labor Markets in a Comparative Context This section starts off by offering a brief comparison of Korea's econo~nic and labor force characteristics with that of its East Asian neighbors and other OECD countries. This is followed by an overview of recent tumultuous labor market developments in Korea and a comparative analysis of the labor market impacts ofthe crisis with those of countries in the region which were also impacted by the crisis (Malaysia and Thailand) as well as other OECD countries at the time they were affected by declines in GDP (Canada and U.K. in the early 1990s and Japan in the late 1990s). While the magnitude of the shock was significantly more dramatic in Korea (and the other East Asian countries) than in the other OECD countries chosen (Table 15. l), it is Table 15.1: International comparisonof GDP growthrates ') Year Year Year Year -1 0 1 2 (%I Korea 7.1 5.5 -5.8 9.5 Malaysia 8.6 7.8 -7.5 8.0 Thailand 5.5 -0.4 -10.0 4.0 Canada -0.2 -2.0 -0.5 2.1 Japan 3.9 0.9 -2.8 United Kingdom 0.4 -2.0 -0.5 2.1 Note: 1) Year -1 refers to the year prior to economic slowdown; Year 0 refers to the year in wh~ch economic growth slowed significantlyldeclined,Year 1 refers to the year immediately following the declineand Year 2 refers to two years following decline. Source: Various sources 330 Labor Market Reforins in Korea: Policy Options for the Futzrr-t. nonetheless useful to compare how labor markets have adjusted sincc in many ways Korea may be more similar to other OECD countries than to other countries in the region as demonstrated below. Econonlic Trendsand Labor Force Characteristics; A Conparalive Picture Table 15.2 provides a comparison of economic indicators for Korea relative to other OECD and East Asian countries included in our analysis. From this table it is clear that Korea is rapidly catching up other OECD countries. The structure of the Korean workforce is rapidly becoming more similar to that of Canada, Japan and the U.K. - with low and shrinking agricultural employment and a growing majority of the employment in the services sector. It should be noted that a similar trend has also being observed in Malaysia and Thailand although at an earlier stage. Unlike these two countries though, where a significant proportion of the population still resides in rural areas, almost 80% of Korea's population is urban - a proportion similar to that in Canada and Japan and just slightly lower than in the U.K.. Korea's per capita GDP is also significantly higher than that of Malaysia and Thailand (though it is still significantly lower than that of thc three other OECD countries listed here). Overall labor force participation rates are highest in Thailand due to high participation rates among women ('Table 15.3). While they are lower in Korea, participation rates in Korea are roughly the same as in Canada and Japan (even though the participation rates among women are still quite low as compared to other OECD countries).' The skill composition of thc Korean labor force is also relatively high. In 1995, just over a quarter of the labor force had attained primary education or less as coinpared to 50% in Malaysia and 80% in Thailand. As the table above shows, the educational attainment of the Korean labor force compares favorably to that of high-income OECD countries, making it well placed to tackle the challenges ofthe knowledge-based economy. We now discuss how Korea's labor market has respondcd to the crisis in comparison to the selected countries. Labor Aajtlstnzent to Crisis Table 15.4 highlights trends in unemployment, underelnploylnent and employment in the years ilnlnediately prior to and following the crisis. * In the first year of the East Asian crisis participation rates fcll in all three crisis countries discussed here - while participation rates fell for both men and women, there was a disproportionate decrease in women's labor force participation rates. These trends wcrc somewhat reversed in Canada and the U.K. when their GDP's declined in the early 1990s. Women's participation rates rose slightly while those for men fell to a small extent. LABOR ADJUSTMENT, NON-STANDARDWORK, AND ... 331 15.2: Summary of economic trends in selected countries,1980-1998 GDP Labor force structure Urban GDP per growth % distribution. 1996 population capita (annual (% change since 1980) (% of total) (US$) ave.)(%) 1990-98 Agriculture lndustw Services 1998 1998 Korea 6.1 11.6 32.5 55.9 80 8,600 (-5.9) (29.0) (37.0) Malaysia 7.4 19.4 32.2 48.4 56 3,670 (-7.8) (33.6) (25.1) Thailand 5.7 50.0 20.8 29.1 21 2,160 (-9.4) (101.9) (54.0) Canada 2.2 4.1 22.8 73.1 77 19,170 Japan 1.5 5.5 33.3 60.8 79 32,350 United 2.2 2.0 26.8 71.O 89 21,410 Kingdom Sources.World Bank, Wor/dDevelopmentIndicators;ILO,KeyIndicatorsof theLaborMarket Table 15.3: Selectedlabor force characteristics, in crisis countries, 1995-1997 Labor forceparticipation Laborforceparticipation Educationalattainmentofthe labor ratesby gender(1997) ratesby gender (1998) force(% distribution,1995) Female Male All Female Male All < Iyear Primary Sec. Tertiary Korea 49.5 75.6 62.2 47.0 75.2 60.7 8.7 18.2 51.9 21.1 Malaysia" 46.8 80.8 63.8 41.6 79.1 60.7 16.7 33.6 42.8 6.8 Canada 57.4 72.5 64.8 58.1 72.4 65.1 20.8 31.2 48.1 Japan 50.4 77.7 63.7 50.1 77.3 63.3 20.5 48.8 30.6 United 43.0 56.0 49.2 43.0 55.7 49.2 13.5 45.7 22.8 Kingdom3' Notes:1) Malaysia labor forceparticipation rates are for 1995. 2) Thailand educational attainment figures for 1990. 3) United Kingdom's labor Participation rates are based on the entire population, not just individuals aged 15+.Education Attainment does not include "Education not defined by level" 18.1%. Sources: International Labor Organization (2000). Key Indicators of the Labor Market; Worton and Mazurndar(2000). 332 Labor Market Refornrs in Kor-en: Policy Optionsfor the Futz~re Quantity-related adjustments were greatest for Korea and Thailand. In both countries, unemployinent rates - which had been in the 2% range at the time of the crisis - more than doubled in the year immediately following the crisis. The adjustment was more muted in the other OECD countries, where unemployment rates rose by 20-30% from their initial - albeit higher - levels. Similar changes are observed when we look at the statistics on underemployment. In terms of employment, Korea saw the highest decline in the year following the crisis (5.3%). All the other countries examined also saw declines in employment ranging from 3.3% in the U.K. to 0.7% in Japan. What seems to distinguish Korea from these other countries is the rapidness of recovery from the crisis. While the 1999 data (second year post- crisis) from Malaysia and Thailand are unavailable for comparative purposes, Korean data show a reversal of the trends for the past two years -- by the third quarter of 1999, unemployment declined by 1.8 % points and employment increased by 1.4%' (though a significant proportion of this increase in employment is in non-permanent jobs). This ii in strong contrast to Canada, Japan and the U.K. which were all experiencing growth in unemployment and underemployment as well as declines in employment - albeit at generally slower rates as compared to the first year - in the second year following the crisis. This is undoubtedly related to the rapid economic recovery - as Table 15. 1 shows GDP grew by about 10% in Korea in the second year after the crisis, while the corresponding growth rates in other OECD countries were about 2%. Similarly, price-related adjustments were also significant in Korea (Table 15. 5). After growing at 6.7% in the year preceding the crisis, real wage growth slowed dramatically and real wages fell by 9.3% in the first year following the cri~is.~Other countries had varying responses - real wages feel quite steeply in Thailand in the year following the crisis and also fell more modestly in Malaysia, Canada and Japan. In the U.K. on the other hand, adjustment to their GDP decline seems to have been entirely quantity-related, as real wages grew by over 4% in the first year after declines in GDP. Wages recovered faster than quantity-related indicators in all countries - wages in the U.K. continued to rise, Canada experienced positive wage growth within two years of its downturn and while wages declined in Japan they did so by less than 1%. However, again, Korea's recovery on the wage front was very swift. In 1999, real wages grew by 11.1% thus reversing any decline that took place as a result ofthe crisis. Korean government forecasts for 2000 are also optimistic - employment is expected to increase by 3.6% and unemploy~nentis expected to fall to 4.5% (Republic of Korea, 2000). ' The wage decline following the crisis has been partly attributable to significant cuts in bonuses and overtime payments - wh~chaccount for a large part of wages - following the crisis. These payments increased in 1999 leading to robust wage growth. However. it should be mentioned that bonuses and overtime payments are usually not part of the remuneration for non-regular workers, who, as we shall see in Section Ill, are a growlng proportion of Korea's workforce. Table 15.4: Summary of unemployment, underemployment, and employment trends in selected countries') Unemployment Underemployment 41 Employment r 2 U1 % Yo Yo ' O/o Yo Yo 0 change change Change Change Year 0 Year 1% Year 2 change year change Year Year o Year 1 Year 2 year Year Year Year .d 0-1 1-2 0-1 1-2 0-1 1-2 L C V) Korea (1996-) 2.0 2.6 6.8 5.0 161.5 -7.4 7.3 9.3 29.2 -5.3 1.4 -I 3 m Malaysia (1996-) 2.6 2.6 4.0 53.8 7.3 7.9 8.2 -2.7 z -I ThailandS)(1996-) 2.0 2.2 5.2 136.4 0.9 1.2 33.3 -2.8 2 = Canada (1989-) 7.5 8.1 10.4 11.3 28.4 8.7 17.0 18.1 18.5 6.5 2.2 -1.9 -0.8 0 cn Japan (1996-) 3.4 3.4 4.1 4.7 20.6 14.6 20.8 20.9 21.5 0.5 2.9 -0.7 -0.8 1 United Kingdom 7.3 7.1 8.8 10.1 23.9 14.8 20.1 20.7 21.5 0.5 3.9 -3.3 -2.2 i (1989-) >0z Notes: 1) Thailand figures for February; Korea and the Philippines are averages of quarterly estimates. 0 2) Year -1 refers to the year prior to economic slowdown; Year 0 refers to the year in which economic growth slowed significantlyldeclined, 2 3) Year 1 refers to the year immediately following the decline and Year 2 refers to two years following decline 0 4) Underemployment (as % of employed population) defined as: Canada, Malaysia, and UK less than 30 hours per week; Japan. Non-Agriculture, less than 35 hours, Korea, 35 hours or less per week; Thailand, less than 40 hours per week and available for more hours. > 5) Working age population in Thailand defined as 13 years of age and older. z Sources: Mahmood and Arya (2000) ; Kang etal. (2000) ; Mansor et a/. (2000) ; Kakwani (1996) ; World Bank (January 2000) ; Thailand Economic Monitor ; 0 For Japan, Labour force survey ; For Canada and UK, /LO (1999). Key Indicators of Labour Market 334 Labor Market Reforms in Korea: Policy Optionsfor the Fzltzri-e Table 15.5: Real wage trends in selected countries % change in real wages Year -1 Year 0 Year I Year 2 Korea (1996;) 6.7 2.4 -9.3 11. l5 Malaysia (1996-) 2.1 3.3 -1.1 Thailand (1996-) 2.3 1.4 -7.4 Canada (1989-) 0.5 -0.6 -0.4 1.9 Japan (1996-) 1.6 0.4 -2.0 -0.9 United Kinqdom (19894 2.3 0.1 4.2 3.6 Sources: For UK, ILO Year Book of Labour Statistics ; For Japan, Ministry of Labour, Monthly Labor Survey ; For Canada, OECD, Manufacturing wage "Historical Statistics 1990-1995': These data indicate that the Korean labor market has demonstrated significant flexibility - both in terms of prices and quantity during the crisis. This is true in comparison to other east Asian countries and especially true in relation to OECD countries. Significant downward and upward adjustments, especially on the wage front, indicate a high level of responsiveness to shocks (it should be noted though that growth rates fell and rose much more sharply than in other OECD countries). The economic strategies that Korea adopts in the aftermath of the crisis as well as the ability of the labor market to respond to shocks will go a long way in determining if and when Korea will join the ranks of the high-income OECD countries. Non-Standard Employment The quality of employ~nenthas become a growing concern in recent years. Between the fourth quarters of 1997 and 1998, total wage employment declined by 8.3%, with the only growth in the wage sector in daily work. Even during the initial phases of recovery in the labor market, the number of workers in regular employment has continued to decline while all employment gains have been in the temporary and daily categories. This development has raised a number of questions. Is the growth of "non- regular" employment merely linked to the crisis and its immediate aftermath or does it reflect an underlying structural change in the economy? If it is a more fundamental trend, what is driving it? Should policy-makers respond and, if so, how? In this section, we consider these questions by reviewing the This growth rate is based on numbers in a press release from the Korean Ministry of Labor (February. 2000). While fixed monthly wages grew in 1999 by 6.1%, overtime wages and bonuses grew by 30.1% and 28.3% respectively, which contributed to an overall wage growth of II.I%. LABOR ADJUSTMENT, NON-STANDARDWORK, AND ... 335 Korean experience with non-regular employment and then considering this experience within the international context. Ultimately, more time and more data will be required to conclusively assess the significance of the recent trends. What Do We Know About Non-Standard Eniploynlent? The issue of non-regular work has had a high profile in many OECD countries over the past decade or so. Various labels have been applied to the phenomenon (such as "non-standard" work and "atypical"employment) but the underlying notion is work forms that depart from the traditional concept of a full-time position of indeterminate length with a single employer. Definitions of "non-standard" employment have also varied but in the OECD region, it has generally included part-time workers, short-term or temporary workers, temporary-agency or "dispatch" employment, and the "own-account" self-empl~yed.~As we will see below, these categories represent roughly two-thirds of all employment in Korea, a proportion that is relatively high by OECD standards. The composition of non-standard employment is also somewhat different in Korea than in many other OECD countries, with a relatively large share in self-employment andl unpaid family work. There is a general perception that non-standard work has been increasing in many industrialized countries. This reflected in the attention that non- standard work has received from academics and policy-makers.7In fact, the empirical trends are less definitive than might be expected. The incidence of part-time work - generally the largest category -- has been increasing in most countries but only marginally.' The share of self-employment has been rising moderately in about one-half of industrialized countries and declining moderately in the other half.9Trends in short-term, temporary contract, and temporary agency (dispatch) work are the inost difficult to establish empirically in many countries. At least through the mid-1990s, the incidence of temporary employment had not changed significantly with some OECD countries reporting increases (Australia, the Netherlands, France, and Spain) and others reporting decreases (Belgium, Greece, Luxembourg, and Portugal). Job stability measures such as tenure and retention rates also do "his latter group includes self-employed workers who themselves have no employees ' For example. the annual OECD Eriiployment O~irlookhas devoted a number of chapters to aspects of non-standard work over the past decade. At the national level. there has also been considerable attention. "he proportion of OECD employment in part-time jobs was 14.9 % in 1997, compared to 14.4 % in 1990. In 8 countries (Canada, France. Ireland, Japan, Luxembourg, New Zealand, Spain, and the UK), part-time shares did increase by 2O hpoints or more (OECD 1998). " Of 21 industrialized countries where data are available for 1990-97. 10 had increas~ngshares of self-employment, I0 had decreasing shares. and one had no change (ILO 1999). 336 Labor Market Reforrns in Korea: Policy Optionsjol- the Fz~tui-e not show any major or systemic shifts in the industrialized countries (OECD 1997). Where non-standard employment has increased, a number of explanatory hypotheses have been offered: O One focuses on changes in labor supply and, specifically, the increasing female and (and to a lesser extent) youth participation rates in some countries. This argument is based on the fact that these groups have a relatively high propensity for certain non-standard work arrangements (part-time and short-term/temporary) because of the flexibility they offer to combine work with other activities (e.g., family responsibilities, school, etc.). O A second hypothesis emphasizes changes in the nature of economic activity that may be shifting optimal labor contracting towards non- standard employment forms. This is the "new economy" or "know- ledge-based economy" hypothesis. For example, while the scale, capital investment, and production process in many manufacturing industries call for standard work arrangements, many service activities require or are amenable to part-time, short-term, or independent contractor arrangements. At the same time, the revolution in information and communication technologies has opened up new possibilities for organizing work in non-traditional ways. Finally, globalization and technology have raised the premium on flexible production which can be served by a variety of non-standard employ~nentpatterns. O Third, labor regulations and other policy variables can affect non- standard employment. In some countries, there have been restrictions on part-time work or temporary jobs. Where these restrictions have been lifted, it has been typical to observe large increases in the relevant non-standard category (e.g., Spain and France with temporary work). There are also many examples where the differential application of labor standards or payroll tax contributions creates incentives towards different forms of employment (e.g., where short-term or part-time workers are outside the system). O Finally, there can be cyclical elements in non-standard employment patterns. During recessions, employers will want to limit labor costs and, as a consequence, it can be expected that their preference for part- time and short-termltemporary work forms will increase. Also, when labor demand is slack, some unemployed workers are likely to respond to a lack of job opportunities by becoming self-einployed.1° '" In fact, there are competing forces that can neutralize these cyclical te~lde~lcies.For example. seniority rules may protect standard employees with the result that layoffs are concentrated among short-term or temporary workers. Examining the patterns of self-employment in the 1980s, the OECD (1992) found that a "pull" factor (i.e.. interest in setting LIPbusinesses when times were good) could counteract the "p~1s11"factor (of scarce job opportunities) LABOR ADJUSTMENT, NON-STANDARDWORK, AND ... 337 Korean Evidence on Non-standard Employment Pends Our analysis of non-standard employment trends in Korea is primarily based on employment status data from the Economically Active Population Survey (EAPS). This covers self-employment, unpaid family work, and three categories of wage employment (regular, temporary, and daily)." Our analysis. based on the EAPS data, does not include part-time and temporary agency (dispatch) jobs as separate categories though they are largely embedded in the employrnent status numbers." We begin in this sub-section by comparing wage vs. non-wage employment since 1980. Following this, we turn to the trends in non-standard employment. Chart 15. 1 compares trends in wage and non-wage employment between 1980 and 1999. The former category includes regular, temporary, and daily workers while the latter includes the self-employed and unpaid family workers. This chart shows an enormous shift over this period from non- wage to wage employment. In 1980, less than half of Korean workers were in wage employment but by the end of the 1990s, about two-thirds were wage employees. This change was particularly dramatic during the 1980s when wage employment grew from about 6.5 million to almost 11 million. The transformation of the Korean labor market su~ntnarizedin Chart 15. 1 reflects the rapid urbanization and industrialization (especially during the 1980s) that remains a remarkable story in international development. In fact, when we look at employment status trends separately for the farm and non- farm sectors, the balance between wage and non-wage employment has changed very little. In the farm sector, self-employment and unpaid family work accounted for 84% of total employment in 1980 and 82% in 1999. In the non-farm sector, corresponding tigures are 29% and 26%. Table 15.6 presents the 1990 and 1999 shares of non-standard work on an aggregate basis and then for the farm and non-farm sectors. Standard work includes regular wage employment and nonstandard work includes self- employment, unpaid family work, and temporary and daily jobs." Overall, the data indicate a high proportion of non-standard employment (around I ' It also includes employers (i.e., the self-employed with employees) but this group has been excluded from the analysis. We use the term "self-employment," then, to mean the seltl employed who do not themselves have employees (i.e.,"own-account') " The only exception would be part-time work where the employee has regular (i.e.,permanent) status. This group is likely to be very small. Part-time work. in Fdct, has had a relat~vely10% incidence in Korea. In 1997, it accounted for 5.1% of total employment. Temporary agency or dispatch work has been a relatively small category histor~callyin part because of legal restrictions. The official estimate at the end of 1998 was 42.000 workers employed in temporary agencies (Kim era/. 2000). This may well change as restrictions are loosened. 17 Regular" wage employment refers to jobs where en~ployeeswork Inore than one year, .c "temporary-' workers are employed for a fixed period of time (usually less than one year); and "daily" workers are employed on a daily basis (though they may be with a particular employer for more than a day). 338 Labor Market Refbrnls in Korea: Policy Optionsfor the Flltii~~ Figure 15.1: Wage and non-wage employment growth in Korea, 1980, 1990, 1997 and 1999 two-thirds) with some modest increase over the decade. However, this overall picture of relative stability masks two counteracting trends. On the one hand, the shift froin non-wage to wage ernployrnent already discussed has supported the growth of standard employment. However, this has been offset by the growing non-standardization of wage e~nploylnentin Korea during the 1990s.'' Table 15.6: Proportionof non-standardemployment in Korea, 1990 and 1999 1990 1999 (%) Total 64.9 68.0 Farm 93.3 94.4 Non-Farm 57.7 64.0 Male 55.2 58.1 Female 78.0 81.3 --- Source: Bureau of Labor Statistics, Korea (VariousYears), EconomicallyAcbve PopulationSurvey. This non-standardization of wage e~nployinentin evident in Chart 15. 2 which presents the 1990 trends in inore detail. During a decade when aggregate ernployinent increased by 11.7%, regular ernployrnent grew by " Non-standard figures cannotbe calculated fiom the EAPS prior to 1989because 110dislinction was made betwcen rcgular and temporary wage employment. LABOR ADJUSTMENT, NON-STANDARDWORK, A N D ... 339 just 1.9% while temporary and daily jobs increased by 31.9% and 24.4%, respectiveiy. It is these trends in wage elnployment that are at the heart of concern regarding non-standard employment. Figure 15.2: Percentage change in employment by status in Korea, 1990-1999 Self-employed Family workers Regularemployee Temporary Daily workers employee In order to assess the significance of the growth in non-standard wage employ~nentand to draw appropriate policy implications, it is important to understand how much of the trend is due to the immediate impacts of the crisis and how much to Inore fundamental underlying factors. As we have noted, in many other OECD countries, one non-cyclical explanation for increasing non-standard work has focused on labor supply and specifically fernale and youth participation. As Chart 15. 3 indicates, the aggregate pattern of relative declines in regular wage e~nploy~nentand increases in temporary and daily work has been slightly Inore inarked in the case of women. However, a crucial element in the labor supply hypothesis is rising participation rates for women and for youth - yet these rates remained relatively unchanged in Korea during the 1990s." The other key structural (non-cyclical) hypothesis for growth in non-standard work relates to technological innovation, sectoral change, and the growth of the knowledge- based economy. This is Inore difficult to assess e~npiricallyand we do not address this in this paper. Female participation rates were 47.0% in 1990 and 47 4% rn 1999. Corresponding figures for the 15-24 year age group were 35 0% in I990 and 31.3% in 1999. 340 Labor- Market Refornzs in Korzea. Pol~cyOptions for the Fzitzlr-e Figure 15.3: Employmentchange by gender and status in Korea, 1990-1999 Self-employed Family workers Regular employee Temporary Daily workers employee Turning to cyclical factors, it has been widely recognized that non- standard wage e~nployment has grown substantially in the immediate aftermath to the crisis. Between 1997 and 1998, all forms of wage employment declined -temporary and daily as well as regular (Table 15. 7). Table 15.7: Percentage change in employment by status in Korea, 1990-1999 1990 to 1997 1997 to 1998 1998 to 1999 Status (%) (Yo) (%I Total 15.1 -4.6 1.8 Self-employed 11.5 0.1 2.5 Family workers -8.1 6.8 -5.4 Regular employee 20.4 -9.7 -6.3 Temporary employee 31.9 -4.4 4.6 Dailv workers 2.8 -8.3 31.9 Source: Bureau of Labor Statistics, Korea (Various Years), Economicaliy Active Population Survey. However, as the recovery was established in 1999 and aggregate job growth resumed, all of the gains in employment were accounted for by non- standard work forms. This was particularly dramatic in the case of daily work which increased in 1999 by 31.9% while regular wage ernployrnent continued on a downward trend declining by 6.3%. LABOR ADJUSTMENT, NON-STANDARDWORK, AND ... 341 Is increasing non-standard employment likely to be an ongoing feature of the Korean labor market? There are two views on this, leading to different answers. One posits that as the recovery takes hold, employers will adopt longer-term planning horizons and will start to invest more in their workforces, including creating regular jobs. There is limited support for this view in the experience of the other OECD countries that have had real GDP declines. In both Canada and the UK, for example, part-time work increased during the recessions and, in fact, continued to increase in the first few years of recovery before starting to recede again.16 However, in each case, the post-recession declines were only modest. The alternate view is that, while the crisis may have exacerbated the situation, Korea has a more fundamental issue with high rates of non- standard employment. There are two pieces of evidence that support this argument. First, the shift from regular to non-regular wage jobs was already occurring prior to the crisis (Chart 15. 4). By the mid-1990s, when the economy was still growing at substantial rates, employers were already substituting non-regular (largely temporary) jobs for regular ones. In part, this may reflect responses to mounting competitive pressures. Kim et al. (2000) makes the argument that the cost competitiveness of Korean labor had been declining since the late 1980s. These conclusions are supported by World Bank estimates which found that unit labor costs in Korea increased in the late 1980s,continuing at least through the early 1990s. Figure15.4: Standardand nonstandard employment trends in Korea, 1989-1999 ' V n Canada, the share of part-time work was 17.0% in 1990 (immediately prior to the recession), 18.1% in 1991 (the recession year) and 18.5%, 19.176, 18.8%. and 18.6% in the four subsequent years. In the UK, the pattern was similar: 20.1% (pre-recession), 20.7% (recession), then 21.5%, 22.1%. 22.4%, and 22.3%. 342 Labor Market Reforms in Korea: Policy Options for the Ftftt~re The second piece of evidence is that, relative to other OECD countries, Korea has a high level of non-standard employment (and did even before the crisis). This is true even when comparisons are made with countries at the lower end of the OECD developinent continuum. For example, while regular wage jobs account for only about 30% of Korea's total employment, the corresponding 1998 figures for Greece and Portugal are 48% and 59%; the incidence of non-standard wage employment is 7% in Greece and 12% in Portugal compared to over 30% in Korea." On balance, the Korean labor market seems to have a propensity towards the creation of non-standard employment that cannot be~fullyexplained by the crisis nor by the country's stage of development. Labor market institutions, including compensation practices, labor standards, and social insurance contributions, seem to create incentives for non-standard job creation at the expense of standard work forms. Four aspects seem especially relevant. First, regular workers receive a large portion of their total compensation in the form of bonuses which are not generally paid to temporary or daily workers. Second, working conditions and contract provisions mandated by the Labor Standards Act primarily apply to regular workers. Severance pay arrangements (applicable to workers with at least one year of tenure) are relatively generous in comparison to many other OECD countries. Third, dismissal procedures for regular workers are somewhat cumbersome and c o ~ t l y .Finally, the Employ~nentand Insurance ' ~ System (EIS) implemented in 1995 initially required payroll contributions for each regular employee in enterprises over 30 workers. Coverage of the EIS has expanded to enterprises of all sizes and some temporary workers, but still does not apply to part-time or daily workers. Recent policy changes may alter the incentives towards non-standard employment. On the one hand, the loosening of restrictions on temporary agencies obviously will lead to increases in this form of work. On the other hand, the current thrust of labor policy is to extend social insurance and regulations to all firms and more categories of workers. This is evident in the recent changes to both Employment Insurance and the Labor Standards Act. However, the continued importance of bonuses in the overall compensation of regular employees and the generous severance system remain as significant incentives for employers to meet staffing needs through non- " 'Two indicators of stage of development are the sectoral composition of e~nploymentand GDP per capita. In terms of the former, these two countries are roughly similar to Korea. In terms of the latter, Greece and Portugal have slightly higher GDP per capita (In 1998, US$11,740 and US$10,670 compared to US$8,600 in Korea). The non-standard figures for Greece and Portugal come from unpublished OECD data. 'Termination for non-performance, allowable under the law but rare, must meet high standards of ':just cause." A Tripaltite Cornmission agreement In 1998 allowed einployers to lay off workers for economic reasons but continued to require a 60-day advance notice. coasultations with worker representatives and, in the case of large firn~s.notification and justification to the hlinistry of Labor. LABOR ADJUSTMENT, NON-STANDARDW O R K , A N D ... 343 standard arrangements. To the extent that the playing field can continue to be leveled for labor institutions, social insurance, and labor regulation, Korea will not only improve the social protection offered non-standard workers but will also reduce the incentives towards these work arrangements. The question is what the price will be in terms of the job creation and labor market flexibility that have been so characteristic of the Korean economic success. Active Labor Market Programs" Unlike most OECD countries, a strong tradition of Active Labor Market Programs (ALMPs) does not exist in Korea. One of the main reasons has been the low rates of unemployment prior to the crisis. However, the crisis has led to a significant expansion of these programs as the increases in unemployment, the difficulties facing many laid-off workers in reestabli- shing employlnent and earnings, and the obstacles facing young people trying to enter the labor force have all underscored the need for policy- makers to consider new options. In many OECD countries, ALMPs have been an important employment policy tool for over four decades. This policy envelope includes a wide range of activities, intended to increase the quality of labor supply (e.g., retraining); to increase labor demand (e.g., direct job creation); or to improve the matching of workers and jobs je.g., job search assistance). They are often targeted at workers with special difficulties in the labor market. The objective of these measures is primarily economic -- to increase the probability that the uneinployed will find jobs or that the underemployed will increase their productivity and earnings. More recently the case for active labor market policies has also emphasized the potential social benefits in the form of the inclusion and participation in the labor force. While ALMPs have become an attractive option for labor policy-makers, evaluative evidence on the effectiveness of the programs in the OECD demonstrates that designing and implementing effective ALMPs poses many challenges. This section looks at these issues in more detail. We compare active labor market programs provided in OECD countries and Korea, summarize the evidence on their impacts, and discuss some of the key issues that should be considered as policy makers aim design and implement more effective programs in Korea. As Korea moves towards a knowledge-based economy, the importance of skills upgrading is going to become increasingly important The likelihood of worker transitions will also increase as activities will become more project-based. The transition to knowledge-based economy will have some '"This section draws heavily on Betcherman el. ul. (2000). 344 Labor A4arket Refbrnzs in Korea: Policy Options,forthe Ftlttl,u implications for active labor programs in terrns of new delivery techniques (greater participation of the private sector and civil society), increasing reliance on technology, and innovations in financing of these programs. As discussed later in this section, these transformations are already evident at the cutting edge of some OECD ALMPs. What are ALMPs?" The more common ALMPs are: Udirect job creation (public works schemes, micro-enterprise development, wage subsidies); O public employment servicesljobsearch assistance agencies; O traininghetraining for unemployed adults and those at risk of unemploy- ment; The OECD has collected statistics on public expenditures by member countries on active labor market programs since the mid-1980s. Average national expenditures (as a % of GDP) increased in the 1990s as cornpared to the 1980s - from about 0.6% to over 1%. This increase reflects both an increasing preference on the part of governments for ALMPs and the higher unemployment in most countries in the 1990s compared to the 1980s. OECD countries generally see active and passive programming as complements rather than substitutes. Where spending is relatively high in the former area, it is also likely to be relatively high in the latter. In 1990, the correlation coefficient between national spending on active and passive programs was .60 (OECD, 1993). After diverging in the early nineties when income support jumped to acco~nmodateworkers laid off in the recession, the strong correlation resumed in 1993, with spending on active programs slightly increasing relative to passive. However, there are major differences across OECD countries in terms of the level and composition of spending on ALMPs. Annex Table 1 highlights these differences for a subset of countries using the latest expenditure data available. While increasing focus has been paid to ALMPs in Korea following the crisis, expenditures on ALMPs remain low relative to other OECD countries. In 1993 (the first year for which data was available), Korea spent 0.08% of GDP on ALMPs. By 1998, this had gone up to 0.18% (along with expenditures of 0.02% of GDP on passive programs). It is anticipated these expenditures will.decline 30-40% in 2000 as the crisis subsides (Republic of Korea, 2000). '"For a more thorough discussion also see Martin (2000) LABORADJUSTMENT, NON-STANDARDWORK, AND ... 345 ALMP 'sin Korea and Other OECD Countries: A Description Korea adopted an Employment Insurance System (EIS) in 1995. The purposes of the EIS are two-fold: (a) to help une~nployedworkers by providing unemployment benefits to them and; (b) to enhance employment stability and job competency of workers through active labor market policy measures. The EIS has three major components: the employment stabili- zation scheme, the vocational competency development scheme and unemployment benefits. Each of the three schemes has its own premium. The premium for unemployment benefits is 1.0 % (shared by employers and employees) of total payroll. The premium for the employment stabilization scheme is 0.3 % of total payroll financed by employers. The premium for the vocational competency development scheme, also entirely financed by employers, varies from 0.1 % to 0.7 % oftotal payroll, depending on the size of the firm (Kang et. al., 2000)." This section will focus on the active labor market programs. Employment Services Employment services serve brokerage functions, matching jobs with job seekers. This assistance comprises many different types of activities: for example, initial interviews at employment offices, in-depth counseling during the unemploy~nentspell, job clubs, etc. In most OECD countries, workers have access to job referrals, job counseling, skills assessment, job search training, resume preparation and job clubs. In most of these countries, public and private services coexist, usually serving different clientele - public employment services target the disadvantaged and the long-term unemployed while private agencies usually target the employed, skilled, and white-collar workers (Fretwell and Goldberg, 1994). Often employment services are integrated with the other ALMPs, as well as passive programs - e.g. the employment office provides job search assistance as well as providing unemployment benefits. Korea has a long, though modest, history of public employment services. Following the crisis, the Korean government expanded and reinforced public employment service agencies for job placement and created a nationwide network for job information. Private employment services have always played a larger role than public services in job placement since the 1960s when they were legalized and the crisis has also seen a significant expansion in the number of such centers (Table 15.8.) " When the EIS was created, its coverage extended to firms with over 30 workers (with some exceptions). However, after the onset of the crisis coverage was extended to smaller firms. and since October I998 coverage has been prov~dedfor all workers -with certain exceptions (e.g. part-time workers, older workers, government officials). However, owing to these exceptions, only 65% of all workers are eligible to avail of the EIS and only 70% of these are actually insured. 346 Labor illorket Refor-nzs in Koiaen: Policy Options for the Ft~ttlre Table 15.8: Employmentservice agencies in Korea,1997 and 1999 Number of... Public Private 1997 1999 1997 1999 - - - Agencies 337 427 1,432 1.756 Staff 1.210 ... ... ... Source: Ministry of Labor. Korea Since 1998, in order to improve the quality of public employment services. the government has merged various active and passive labor market functions to create employment centers that are based on the concept of "One-Stop Service"and are designed to provide job seekers with all kinds of information and services, from job vacancy information to vocational training. Lrnemployed workers can also receive unemployment benefits at these centers. The government has also recently moved towards using information technology for employment services. Public employment offices in different areas are now able to share information about job seekers and vacancies via networks and the internet through the recent launching of an electronic labor exchange system, "Work-Net,'' which provides information on job vacancies, vocational training, career guidance, employment policies, employment insurance and labor market statistics. However, this system is yet to be evaluated (Kang et al.,2000). Training and Re-g Most OECD countries focus on three types of training programs: (a) Retraining aiined at the long term-unemployed (e.g., unemployed for more than 12 months); (b) Retraining displaced workers, especially those dis- placed en masse as a result of enterpriselindustrial restructuring; and (c) Training programs aimed at young people, often with special attention to school drop-outs. Support can come in the form of direct provision of training, financial support for trainees, or provision of infor~iiationon training services. In many OECD countries, governments are moving away from the role of direct provision of training and focusing Inore on addressing market failures in information and financing, while leaving more of the delivery to private providers. Korea has long-standing experience with vocational education and pre and in-service vocational training. In this section, we will not focus on these issue^^^ limiting our attention to training and retraining measures that have been implemented by the government for the unemployed. Since the onset of severe economic difficulties, the government has set up training programs 2' For a more ~ntensivediscussion of these issties see Gill and Ihm (2000): Kiln (2000) LABOR ADJUSTMENT, NON-STANDARD WORK, AND ... 347 available to the unemployed as a social relief strategy. Individuals are allowed to receive traininglretraining for a duration of up to one year in new skills and the government pays for the cost of training. Almost a quarter of all the unemployed have received government sponsored training. A pilot program of training vouchers was also launched in 1998. The objective of the training voucher system was to provide the unemployed with more choices and promote competition among institutions providing training. Measures that have been implemented in 1999 include efforts to make training more demand-oriented; expansion of training in knowledge-based, high technology fields; improving une~nployedworkers' transition from training programs to jobs; enhancement of information systems; and extension of private sector participation in training (Kang et. al.,2000). Direct Job Creation In Korea, job creation measures have been the primary labor market policy in response to the economic crisis. Public works programs, support for small and medium sized businesses, internship programs for new labor market entrants, and wage subsidy programs are among the job creation efforts planned by the government since the crisis. Interventions that have been designed in Korea have been somewhat similar to those in other OECD countries. Public Works Programs The idea behind these programs is generally to help disadvantaged and long- term unemployed groups to regain contact with the labor market, thereby minimizing the probability of stigmatization, skills obsolescence, and marginalization. Public works programs have been instituted in Korea at both the local and central government levels primarily since the crisis to provide temporary employment opportunities for the unemployed and provide them with a safety net. These works were generally of three month duration and any individual could not participate in more than three such projects consecutively (Kang et. al., 2000; Republic of Korea, 2000). Micro-Enterprise Development Technical assistance, credit, and other support can contribute to the creation and promotion of small-scale new businesses and self-employment. Micro- enterprise development assistance (MEDA) programs have been offered both on a universal basis or to particular groups - the newly unemployed or the long-term unemployed. Program conditions also vary. Participants may receive assistance to set up their businesses as a lump-sum payment or periodic allowances. Often there is "screening" whereby potential benefi- 348 Labor- Market Rejor-n.1~in Korva: Policy Optionsfor the F7zltzir*e ciaries undergo a rigorous assessment which evaluates their likelihood of success (Wilson and Adams, 1994). In most cases participants may also receive business advisory services and counseling. While such assistance has increased in Korea since the crisis, it is still quite insignificant. Individuals have been provided with financial support (3- 500 million won per business) to start up small- and mediurn-sized businesses and venture enterprises. Business Incubators have also been established to provide advisory services to assist entrepreneurs with management training. WageIEmployment Subsidy (WES) In OECD countries, these programs have been targeted at the long-term unemployed and youth. They aim to reduce social exclusion, that is, to help these individuals regain contact with the world of work by providing employers with an incentive to hire these individuals. They are instituted under varying economic conditions, though usually during slack periods (NERA, 1995). The level and duration of these subsidies varies significantly between programs and countries. Following the crisis, Korea's subsidy programs expanded significantly. In 1998, wage subsidies were provided for 6 months for the following firms: O Those that were closed for at least two days a month: 112 to 213 of the temporary shutdown allowance that must be paid to employees was subsidized. O Those that reduced regular working hours by more than 10%: 6-10% of the wage bill prior to the reduction was subsidized. 0 Those that provided training to workers: 50-66% of the wages along with training expenses subsidized. Firms that granted paid leave to their employees and those that switched to a new line of business while retaining 60 % of their employees were also paid a subsidy. Evaluations ofALMP 's:Evidence@on?OECD and Ko~ea'~ In spite of the large public expenditures on active labor market programs, rigorous evaluations of their impacts have been limited. However, policy- makers are increasingly realizing the importance of good evaluation in improving program design. They want to know what programs accomplish, what they cost, and how they should be designed to be cost-effective. Over the past few years increasing emphasis is being placed on scientific 2' For a more rigorous discussion of evaluation techniques and results from evaluations of OECD ALMPs, also see Martin. 2000; and Dar and Tzannatos (1999). LABOR ADJUSTMENT, NON-STANDARD WORK, AND ... 349 evaluation techniques --which compare the outcomes of the "treatment" group of program participants relative to a "control" group of non- participants - to determine program outcomes. As a result there is a growing body of evidence on the effectiveness of these programs - especially in OECD countries. In this section, we provide a brief overview of the evidence from evaluations of active labor market programs in Korea and other OECD c~untries.'~While these programs have not been evaluated with the same rigor in Korea, some preliminary evaluative evidence is presented. Job Search Aassistance !JSA)/Employment Services Evaluations of about 20 of these programs suggests that JSA is one of the most successful active labor market programs: it costs little to provide and the program is not any less effective than alternative and more expensive ALMPs. However, much depends on whether the economy is growing and on the availability of public funds (which can be scarce during a recession). Most evaluations indicate positive results. Less successful programs are generally associated with periods of recessions and rising unemployment (Fay, 1996). Conversely, the effectiveness of job search assistance seems to increase when economic conditions improve and when new jobs are being generated (OECD, 1993; 0' Leary, 1998a, b; Fretwell et. al., 1999). Job search assistance is one of the most-cost effective of the ALMPs. For example, Leigh (1995) finds that JSA measures cost 2-4 times less than training, but appear equally effective in terms of impacts. While rigorous evaluations of public employment services has not been done in Korea (and little evidence exists on the effectiveness of private employment services), evidence is more discouraging than in other OECD countries. Employment services may not be highly successful in placing workers - the rate of employment has never exceed 20% and declined to 7.4% in 1998 (Table 15. 9. This compares unfavorably to over 50% employinent rates in most other OECD countries for which such data is available. The declines in placement rates in 1998 are to be expected as there were few vacancies during the crisis. A survey on "Unemployment and Welfare Needs of Workers" found that most workers feel that the employlnent services do not provide timely and locally reliable labor supply and demand information to customers (KLI, 1998). Most job seekers felt that the public employment services failed to provide adequate job opportunities and counseling on career development. Some of the reasons that may underlie the low effectiveness of these agencies include the high workload of job counselors (an average of 4000 workers per counselor in 1997 compared to 364 in Germany and 325 in " Most of the evidence for OECD countries is draw11from the U.S., Canada, Sweden, Germany, France and the U.K.. 350 Labor Market Rejornzs in Korea: Policy Options for the F t ~ t ~ ~ r e Sweden) and the inadequate use of information technology though that is slowly changing with the advent of "Work-Net '. No evidence exists on the 7 relative effectiveness of these programs for participants as compared to non- participants. Table 15.9: Effectivenessof public employment serviceagencies in Korea, 1994-1998 Vacancies Job seekers Employed (C,B) Year (A) (B) (C) Source: Kang et a/.(2000) Trainin? Programs A review of close to 50 programs training programs targeted at long-term unemployed, those laid off en masse and youth yield somewhat mixed results. The success of programs for the long-term unemployed tends to be dependent on the business cycle: programs have performed better when instituted at times of economic expansion. Longitudinal studies indicate mixed results; while in some cases positive effects dissipated within two years of program completion, in some cases the impacts persisted. Programs targeted at those laid-off en masse have resulted in some modest increase in reemployment probabilities, however, this result is often statistically insignificant (Corson, Long and Maynard, 1985). The results for post- program earnings are more discouraging: relative to the control group, wage effects on participants are rarely positive and in most cases are negative. Programs targeted at youth often yield the most discouraging results - training rarely has an effect on earnings or employment probabilities of program beneficiaries (Fay, 1996). While costs are available only in some cases, cost-benefit analysis usually show that costs are so high relative to the benefits of most of the program that, even if the effects persisted for 10 years, the social return of the program may remain negative (Friedlander et. al., 1997). Rigorous evaluations of training programs for the unemployed have not been carried out in Korea in the aftermath of the crisis - evaluations do not compare outcomes of trainees with that of a control group. However, as is the case in other OECD countries, the completion rate of training is extremely low and the re-employment rate for those who completed training also appears to be quite low (Table 15.10). Table 15.10: Training programs for the unemployed in Korea, 1998-1999 1998 1999 (till June 1999) Trainees 362941 181273 Completion Rate (%) 53.4 12.3 Employment Rate (for those who 22.4 completed training) (%) Source:Ministry of Labor, Korea. Public Works Programs/Public Service Employment Public works programs have been evaluated in both OECD and developing countries. Evaluations of close to 20 programs point to some general conclusions. Some scientific evaluations suggest very high displacement effects which can reach 100 %, as they did in Sweden (Skedinger, 1995). Participants tend to have a smaller probability of being etnployed in a non- assisted job after participation in the program, and are likely to earn less than their counterparts in the control group. These programs also do not seem to have a significant impact on reducing long-term unemployment and whatever small short-run impacts may exist tend to diminish over time (Webster, 1998). However, these programs can be effective as short-run anti-poverty interventions. For this reason, some countries -including Korea - have used them extensively in periods of hardship. In 1998, 440,000 people participated in public works projects, with a total government expenditure of almost 1 trillion won. Participation for 1999 has been estimated around one million with a budget of 2.5 trillion won. Preliminary evidence indicates that programs have not been well-targeted. Some public works projects are considered as wasteful (Yoo, 1999). In the early stage of the program, leakage rates were high as some well-off people participated in the program, while many poor unelnployed were excluded from public works because local governments often selected participants on a first-come-first-served basis regardless of the selection criteria. Another criticism is that the program may be distorting the labor market - the daily wage rate for participants is still 30-40% more than the minimum wage - which may have led to crowding out ofjobs in the private sector (Kang et.al., 2000). Micro-Enterprise Development/Self-EmploymentSchemes (MEDA) Evaluations of 15 of these programs lead to the conclusion that MEDA programs have high deadweight loss. Estimates of these losses vary from about 30% in the self-employment experiments in the U.S. to over 50% in Canada and Denmark in the early 1990s (Fay, 1996). Evidence shows that 352 Labor Marker Reforms in Korea: Policy Options,/orthe Fzltz~~.e businesses assisted through mentoring and business counseling are more likely to succeed. However there is only a small ~nultipliereffect - Most surviving businesses create, on average, half an additional job (Wilson and Adams, 1994). Cost-benefit issues have rarely been addressed in the context of these programs and hence it would be premature to draw any conclusions on the cost-effectiveness of these programs. These programs have also not been evaluated in the Korean context to date. WageIEmployment Subsidies (WES) Evidence from 22 such programs in OECD countries suggest that WES programs have rather low take-up rates, high deadweight loss and substitution effects. In the extreme case of Ireland's wage subsidy program, these losses combined totaled over 95% - alternatively, the net incrementality of the program was a meager four % (OECD, 1993). Evaluations of similar programs in Australia, Holland and the U.K. also indicate high deadweight andlor substitution effects exceeding 60%. Equally disappointing are evaluations which compared wages and employment outcomes of participants with those of a control group - with a few exceptions, over time participants are likely to earn less are also less likely to be employed (OECD, 1993). Again, while Korea's programs have not been scientifically evaluated, initial results point to somewhat similarly disappointing performance. In 1998, 1,805 firms received employment maintenance subsidies, covering a total of almost 800,000 workers. Contrary to initial hopes, these employment subsidy programs have been used by less than one % of private employers - though there is a strong positive correlation between the size of the firm and the likelihood of utilizing the subsidy programs. A recent study which investigated the employment effects of these subsidy programs using employer surveys as well as case studies showed low net employment creation of these schemes (Kim et al., 1999). According to this survey, while there were variations by type of subsidy program, on average the deadweight loss was around 70'76,somewhat similar to OECD countries. ALMPs in Korea: The Way Forward The relative scarcity of rigorous evidence on the effectiveness of ALMPs in Korea does not allow a definitive conclusion as to whether such interventions have been justified on economic or social grounds. However, the general conclusion emerging from international evidence as well as preliminary evidence from Korea suggests that large scale application of these programs should be avoided without knowledge of their effects. It is clear that designing and implementing of active labor market programs should take into account the general state of the econorny and set the right objectives and expectations. Different programs are needed for different LABORADJUSTMENT, NON-STANDARDWORK, AND ... 353 scenarios. Also, expectations should be modest: rarely can a program simultaneously meet criteria of economic policy (efficiency and distribution) andalso political considerations. However, policy-makers in Korea should not avoid this area in the future as the country emerges from the crisis. Active labor market programs serve social objectives, as well as economic and little research has been done to examine these social impacts, which may be more positive. Furthermore, the disappointing performance of these programs in the aggregate masks the fact that some program designs do seem to lead to positive outcomes for some types of workers. The challenge is to learn from existing experiences. As policy-makers in Korea aim to improve on the design and implementation of these programs while moving towards a knowledge- based economy, they should keep the following issues in mind: Prioritv Setting. As we have noted, active labor market programs can have various policy objectives including reducing unemployment in cyclical downturns, correcting structural imbalances, improving labor market functioning, and assisting disadvantaged groups of workers. In designing an overall strategy, it is important to identify which of these are the priority objectives since it is the objectives that should determine program choices and program design. The Roles of the Public and Private Sectors. This is a key consideration both in developing an overall strategy and in designing and implementing programs. At one time, governments in the OECD developed and delivered virtually all ALMPs but, increasingly, governments have reconsidered the respective roles of the public and private sectors. In many countries, possibilities have opened up for the other sectors to play important roles, at least in the delivery of services. This seems to be happening in Korea too as in the case of the voucher system pilot for training. This can lead to more diverse and cost-effective services. However, governments must remain responsible for the overall system, ensuring that it remains focused on public priorities. They must also address distributional issues and provide critical public goods. Promoting Partnerships and Dialogue. The identification of priorities for active labor market policy and can benefit from ongoing dialogue between government, business, labor, and other relevant organizations. Where this dialogue is conducted effectively, policy- makers can maintain a close connection with the needs of the labor market and can maximize support for ALMPs. In Korea, the dialogue needs to be carried out both at the national policy making level (e.g., through the 354 Lnbor hlnrket Reforms in KO/-en:Policy Options for the Fufui-e tripartite commission) and locally where programs are delivered. "Infrastructure"for the Labor Market. Infrastructure services are critical if ALMPs are to be a useful policy instrument. By "infrastructure," we mean labor market infonnation, a viable network of employment service offices, and certification and accreditation systems. These services are the cornerstones of an effective system: They inform the program choices that should be made. They provide the bridge between the Iabor market, service deliverers, workers, and employers. As largely public goods, these are the responsibility of governments. Korea is in the process of developing these services and they should continue to be a high priority. Policy and Administrative/Operational Capacity. Designing and implementing ALMPs requires considerable capacity within government. Capacity needs differ significantly by program. For example, employment services require a network of facilities with extensive geographic coverage, the resources (e.g. technology) to generate and disseminate accurate labor market information, skilled counselors, and reliable connections with the employer and educational communities. Training programs require labor market information plus training and occupational standards, monitoring and evaluation capabilities, and capacity to deliver effective programs. Governments must recognize that capacity building is a slow but essential process. Monitoring and Evaluation. This is a key part of capacity and deserves special mention. In spite of the large public expenditures on ALMPs in OECD countries, rigorous evaluations of these programs have been relatively uncommon. Korea is no exception. In an effort to improve the targeting and eff~ciencyof social programs, sound impact evaluation techniques should be used to evaluate these programs. A good evaluation compares labor market outcolnes for individuals who have gone through a particular program with those of a control group of their peers, and also utilizes data on program costs to attempt to answer questions such as: what are the impact estimates of the program on the individual; are the impacts large enough to yield net social gains; and is this the best outcome that could have been achieved for the money spent. LABOR ADJUSTMENT, NON-STANDARDWORK, AND ... 355 Conclusion We have not attempted to provide a comprehensive analysis of Korean labor market issues in this paper. A number of key issues have remained untouched including labor-management relations, unemployment insurance, and women in the labor market. Using international experiences to provide context, we have focused on Korea's adjustment to the economic crisis and two labor market issues of growing importance - active labor market programs and non-standard employment. Our review has highlighted the flexibility apparent in the Korean labor market. It is true that the rapid develop~nentof the economy has transformed the labor market dramatically over recent decades. In many respects, it now resembles labor markets in other OECD countries. However, our analysis of both the adjustment to the crisis and non-standard employment indicates that the Korean labor market is significantly different and, indeed, more flexible than what we see elsewhere in the OECD region. This has been beneficial in terms of job creation (including in the wake of the crisis) but it raises questions for social policy and labor market segmentation. The challenge emerging from our paper is for Korea to improve social protection for all workers, to create incentives for broader access to secure work, and to provide support for workers in transition without constraining job creation. Looking first at active labor market programs, these have the potential to reduce unemployment, enhance human resource development, increase the employability of disadvantaged workers. and improve the overall func- tioning of the labor market. However, rigorous evaluations of these programs in OECD countries show that economic conditions, worker characteristics, and program design contribute significantly to their effectiveness. Otherwise, the potential of ALMPs will not be realized. Korea's transition towards a knowledge-based economy will have implications for active labor programs in terms of institutional capacity building, delivery techniques (e.g. greater participation of the private sector and civil society), increasing reliance on technology, innovations in financing, and monitoring and evaluation of these programs. These transformations are already evident at the cutting edge of ALMPs in some OECD countries and Korea can benefit from learning from those experiences. The analysis on non-standard employment has concluded that Korea has an underlying propensity towards such forms of job creation. While the crisis has exaggerated the situation, there were already tendencies in this direction prior to the crisis, and, from an international perspective, Korea has a high share of non-standard employees. While this has been key to the country's labor market flexibility, we view this as an issue that rightly concerns policy makers. What results is a segmentation between relatively 356 Labor Market Refornzs in Korea: Policy Options for the Ftltur-e well paid regular workers with access to severance, employment protection, and social insurance and non-standard workers in an insecure situation. Continuing in this direction is not beneficial from a distributional perspective. Nor is it likely to be beneficial from an efficiency perspective when, increasingly, the country's growth will depend on a skilled and productive work force. At least in part, the high level of non-standard employment in Korea is due to labor policies and practices that create incentives for these types of jobs. Recent reforms intend to level the playing field by extending the coverage of labor standards and social insurance. Minimizing differential treatment of standard and non-standard workers will alter these incentives. However, further analysis will be required to strike the optimal balance between protection and job creation. Over the past two generations, Korea has experienced dramatic economic progress which obviously has led to huge improvements in the living standards of its people. In a couple of generations, Koreans have gone from a rural agricultural existence through industrialization and now their prosperity increasingly depends on success in service industries and high technology. Many other OECD countries have taken a century to make this transition. While this pace of development has created pressures in the labor market and formidable challenges for policy-makers, workers, and employers, it also raises the prospect of leapfrogging on the experiences of other countries and thus moving up the "learning curve" quickly in preparing for the knowledge-based economy. LABOR ADJUSTMENT, NON-STANDARDWORK, AND ... 357 Annex 15.1: Expenditures on labor market programs in selected OECD countries Australia Denmark France Germany Japan Italy Spain Sweden U. S. Labor market pr~gram (199718) (1998) (1997) (1998) (1997-(1996)(1998)(1998)(1997- 98) 98) Public employment services & 0.21 0.14 0.16 0.23 0.03 0.04 0.07 0.30 0.06 administraCon Labor market training 0.07 1.07 0.35 0.34 0.03 0.01 0.21 0.48 0.04 Trainingunemployedadultsand 0.06 0.73 0.31 0.35 0.03 -0.10 0.47 0.04 those at risk 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 Measuresfor unemployed & - 0.08 0.07 0.06 - 0.04 0.07 0.03 0.03 disadvantagedyouth Apprenticeshipand relatedform: 0.05 - 0.19 0.01 - 0.38 - of generalyouth training Subsidized employment 0.13 0.30 0.52 0.39 0.02 0.61 0.35 0.58 0.01 Subsidiesto employmentin the 0.04 0.03 0.32 0.03 0.02 0.56 0.24 0.15 - privatesector Support of unemployedpersons 0.02 0.04 - 0.03 - - 0.03 0.09 - starting enterprises Directjob creation (publicornon- 0.07 0.23 0.20 0.32 - 0.04 0.07 0.35 0.01 profit) Measuresfor 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 forthe disabled 0.04 - 0.06 0.15 - - 0.02 0.58 - Unemployment compensab'on 1.17 1.86 4.50 2.29 0.43 0.68 1.64 1.91 0.25 Early retirementfor labor market - 1.88 0.35 - 0.20 - reasons Total 1.69 5.63 3.22 3.56 0.52 1.96 2.36 3.93 0.43 Actrve Measures 0.52 1.89 1.37 1.27 0.09 1.08 0.72 2.01 0.18 Passive Measures 1.17 3.74 1.85 2.29 0.43 0.88 1.64 1.91 0.25 f"Korean Labor Market and Social Safety-Net Reforms: Challenges and Policy Requirements* John I? Martin" Raymond Torres" Introduction Korea has long registered one of the most impressive economic records of modern times. During the past four decades, GDP growth averaged more than 9 % on an annual basis, permitting current per capita income to reach 65 % of the OECD average. The financial crisis which started at the end of 1997 brought the growth process to a sudden halt. The crisis forced the government to agree to a rescue package with the IMF. Unemployment rocketed from 2% % to a peak of 8% % in early 1999, and more than one million Koreans were thrown into poverty. The difficult economic and social situation provoked by the crisis forced the authorities to implement quickly a wide range of macroeconomic and structural reforms. The economy has now turned the corner, permitting the unemployment rate to fall rapidly to below 5 %, and short-term prospects point to a continuation of the solid recovery. In 1999, the economy grew by over 10 % and the latest OECD projections show a growth rate of 8.5 % this year, followed by 6 % in 2001. This recovery owes a great deal to market forces -- the significant depreciation of the Won that took place short after the start of the crisis, combined with significant nominal and real wage cuts in 1998, have improved international competitiveness, export performance and firms' This paper draws heavily on the main findings of the OECD Review on Labor Market and Social Safety-Net Policies in Korea (OECD, 2000a). We would like to acknowledge the contribut~onsofthe following colleagues who participated w ~ t hus in the review and provided helpful comments on this paper: Willem Adema. Saehung I.ee. Elena Stancanelli and Peter Tergeist. We would also like to thank Sylvie Jeannot for stat~sticalsupport. The views expressed In this paper are our own and do not necessar~lyreflect those of OECD Member countries. ** Directorate for Education, Employment, Labor and Social Affairs, OECD. KOREAN LABOR MARKET AND SOCIAL SAFETY-NET REFORMS: ... 359 profitability. But government policies have also played a role in ensuring the recovery. The Korean government has implemented over the past three years many structural reforms, not only in the areas of banking, corporate governance and financial markets, but also covering labor markets, industrial relations and social protection. Indeed, the range and scope of these reforms, together with the speed with which they were implemented, is unprecedented by the standards of other OECD countries. Now that the recovery is well underway, it may be tempting to assume that the crisis was just a blip in an otherwise high-growth path and that, as a consequence, a pause in the reform agenda is called for. On the contrary, Section I1 of this paper argues that the Korean economy and society is still facing key challenges on the fronts of industrial relations, labor markets and social protection. Section 111 sets out reform proposals to tackle these challenges. The final section presents some concluding remarks. Outstanding Problems in the Areas of Industrial Relations, Labor Markets and the Social Safety-Net OECD (2000a) highlights several outstanding problems in the areas of industrial relations, labor markets and social protection which need to be tackled if the Korean economy and society is to achieve its stated objectives. This section summarizes the arguments. Reforms to Labor Laws to Bring Them into Line ~ ~ iInternationally t h Accepted Standards When Korea joined the OECD in 1996, the Korean authorities made a commitment to the Organization that they would reform existing laws on industrial relations in line with internationally accepted standards, including those concerning basic rights such as freedom of association and collective bargaining. The OECD's Employment, Labor and Social Affairs Committee has since been monitoring closely progress on labor law reforms in Korea in the light of that commitment. The Korean authorities have taken a number of important steps over the past three years to meet this commitment, steps which have been welcomed by the OECD and the ILO. However, as documented in OECD (2000a) and in several recent reports of the 1LO's Committee on Freedom of Association, there remain some outstanding problems. The most prominent are: O Prohibition of remuneration of full-time union officials: The 1997 Trade Union Act prohibits the remuneration of full-time trade union officers by the employer, as well as any financial payment for regular 360 Labor Market Refornzs in Korea: Policy Options for the Futtue trade union operations. This clause is supposed to take effect by the year 2002.' It is not clear why the law should have to regulate this matter, rather than leaving it to be decided by collective bargaining. A survey of practices in other OECD countries shows that paid working time off for trade union representatives is a cornmon feature of industrial relations systems. Additional time off for works council members (where these exist) is also widespread. In large enterprises, hours credits for both these functions often imply that many union officials devote their full time to union matters while being on the company payroll. In some countries, hours credits and time-off for union officials are regulated in detail through labor law, while in other countries the matter is left to collective agreements. O Union status of dismissed workers: According to the Trade Union and Labor Relations Adjustment Act, an organization will not be regarded as a trade union when persons who are not workers are allowed to join it. In particular, dismissed workers are not allowed to retain their membership in a trade union as soon as the dismissal has become valid. As the long drawn-out saga over the recognition of the second national trade union center, the Korean Confederation of Trade Unions (KCTU) has shown, the fact that some trade union officials are dismissed or unemployed workers can lead to difficulties in securing legal recognition of a trade union by the authorities. If the law remains unchanged, this problem is bound to re- appear as new unions are formed and seek to register with government authorities2. This particular provision of Korean labor law has no counterpart in other OECD countries which consider qualifications for membership as a matter for the union itself to decide and not to be set down in legislation3. 0 Multiple trade unions at the enterprise level: Though the revised 1997 Trade Union and Labor Relations Adjust~nentAct makes it legally ' The law has thus incorporated employer concerns that they will at that time be forced to support large numbers of paid full-time union officials from both FKTU- and KCTU-affiliated enterprise unions. Already under current conditions, employer associations complain that their members are forced to pay an unusually high number of union officers in relation to the workforce inembers they represent compared with the situation in other OECD countries Although the February 1998 Tripartite Commission Agreement did recognize ~~nemployed workers' rights to Join a trade union or remain a union member (altl~oughnot at enterprise level). a corresponding bill subsequently submitted by the government was re.jected by the National Assembly. Despite repeated statements of intent by the government. no further bill has been submitted, as there seems to be disagreement among the Ministries concerned as to which groups among the l~nemployedmight be allowed to join upper-level trade union organizations (in particular, as concerns the length of their unemployment spell). Most trade unions in OECD countries have varying proportions of students, self-employed. unemployed and retired workers among tlie~rranks (OECD 1991, Chapter 4). possible to form multiple trade unions at industrial and national levels, on the enterprise level, multiple unions will only be allowed from January 2002 onwards. O U h t s of civil servants: Civil servants are not allowed to join trade unions. Following agreement in the Tripartite Com~nissionin February 1998, the establishment of workplace associations in the public service is allowed since January 1999, but due to certain restrictions over one third of public service personnel is ineligible to join such associations. The ILO has expressed concern over these restrictions, in pa~ticularthat there is no timetable for allowing union rights to public servants. O Right to strike in the public sector: Strikes by workers in central and local government and those engaged in the production of military goods are prohibited, which is not unusual in OECD countries. However, the law also restricts the right to strike in the more broadly detined public sector, including regular public transportation services; gas, electricity and water; oil refineries and oil supply; hospitals and other medical services; banks and the Mint; broadcasting and communication. The ILO has repeatedly asked the Korean government to ensure that restrictions on the right to strike only apply to essential services in the "strict sense of the term", under conditions of an "imminent threat to the life, personal safety or health of the whole or part of the population" (ILO, 1996 and 1998). Korea is not unique in restricting the right to strike in some public services. However, Korean labor law provisions seem to go farther than in other OECD countries in that they set out an unusually broad definition of "essential services' wherea strike ban covering entire sectors is maintained. 7 Lack of Trust among the Social Partners It is more necessary than ever to create a less confrontational and more consensual system of industrial relations if the Korean economy and society is to modernize and prosper. Trust and co-operation among the social partners can play an important role in fostering a dynamic, knowledge-based economy. Unfortunately, past confrontational attitudes between the social partners have re-surfaced recently in a context of econo~nicrecovery. The result is that consensus on wage moderation and necessary reforms of employment conditions and workforce practices are increasingly difficult to achieve. Moreover, the social partners have not even agreed on a common agenda for discussions by the Tripartite Commission and the prospect for the Com~nissionto find satisfactory solutions to outstanding issues seems rather 362 Labor Market Refornzs in KOI-ea:Policy Options for the Fl~ture remote at present. One manifestation of these confrontational attitudes is the long-standing practice of arresting and imprisoning trade union leaders and members for violations of existing laws. This particular feature of industrial disputes in Korea is a matter of considerable concern to the OECD and the [LO. Successive waves of arrests of trade unionists not only represent a threat to fundamental workers' rights, but they also damage trust among the social partners. During the economic crisis and its aftermath, many union members reacted strongly to the announcement of business restructuring and collective dismissals. This included takeovers of plants and equipment, acts of violence and strike action in some essential services where, as noted above, Korean law prescribes conflict resolution through arbitration. These events gave rise to a sharp upsurge in arrests of workers and union officials. The [LO'S Committee on Freedom of Association has repeatedly urged the Korean government to take appropriate measures so that the persons detained or on trial as a result of their trade union activities are released or that the charges against them are dropped. It has objected, in particular, to the use of the clause in the Penal Code relating to "obstruction of business" as a basis for the arrest of union demonstrators and strike participant^.^ The Korean authorities, for their part, argue that the concerned trade unionists have been arrested on such grounds as violence during strikes, unlawful occupation of premises or serious damage to company facilities. They have also announced a plan to ''minimize'' the number of arrests through focusing mainly on strike leaders and through applying fines, but not detaining other participants. In sum, although it is not possible to assess the extent to which these pending issues have undermined labor-management relations, it is likely that addressing them urgently will contribute to a better climate of trust between the social partners. This, in turn, is an essential element in fostering a less confrontational system of industrial relations. Not only would this benefit Korean workers and firms but it would also increase the attractiveness of Korea as a location for foreign direct investment. A Low Enzployment-PopulationRatio The sharp fall in the open unemployment rate recently (to 4.7 % in March 2000) is most welcome news. However, there remains considerable room for increasing the number of jobs. Indeed, the employment-population ratio in Korea is relatively low by international comparison. In 1999, the overall employment-population ratio was less than 60 % -- 5 % age points below the OECD average (Table 16.1). The employment-population ratios The OECD's Employment, Labor and Social Affairs Committee has repeatedly espressed the view that the definition of "unlawful activities" in Korea is unusually broad and encompasses union activities that would be regarded as lawful in most OECD countries. KOREAN LABOR MARKET AND SOCIAL SAFETY-NET REFORMS: ... 363 of women and youth are particularly low by internationalstandards5. In addition, though the open unemployment rate in currently below the OECD average, there is a substantial degree of labor market slack. This is illustrated by the high numbers of discouraged workers (the total participation rate in 1999 was 1.7 % points lower than in 1997) and participants in labor market programs such as public works. Finally, economic restructuring among the chaebols is still underway and its completion may entail further significant job losses. An Extremely High Degree ofLubor Market Duality More fundamentally, Korea has one of the most "dual" labor markets in the OECD area, as illustrated by the following facts: O In 1999, more than half of all employees had either a temporary or a daily contract, the remaining having a "regular" (i.e open-ended) contract. In international comparisons, Korea has the highest share of non-regular jobs in total employment, followed by Spain, while the number of workers holding a permanent job is the lowest among OECD countries, followed by Turkey (Table 16.2). Younger and higher-educated rnen hold most regular jobs, while the incidence of non-regular work is particularly high among women, older workers and the lower-educated (Chart 16.1). The incidence of self-employment is among the highest for OECD countries together with Mexico and second only to Turkey. Unpaid family workers make up about 10 % of total employment in Korea, while this figure varies between 0 % and 3 % in most other OECD countries. O Reflecting the high incidence of non-regular employment, average job tenure (i.e. the length of stay in the same enterprise) is low by international comparison. In 1998, the average Korean worker stayed slightly over six years with the same employer-- in 1995, the figure was almost ten years in the case of the European Union and over eleven years in Japan (Table 16.3).Average tenure is even lower in Korea than in the United States. Job tenure is especially low for Korean female workers. Average job tenure for men was almost seven years in 1998, while average job tenure for women was just over four years in 1998 -- by contrast, in 1995, the unweighted average for OECD countries was 10.4 years for men and 8.4 years for women. It also appears that workers with a college degree enjoy longer job tenures than lower- educated ones. However, tenure is higher for the lower-educated (those j On the other hand, the employment-population rate of workers aged 55 to 64 is among the highest in the OECD area. Given the modest level of retirement penbions, Koreans tcnd to work until late in their lives. 364 Laboi, Market Refornzs in Korea: Policy Optionsfor the Ftlttl~e Table 16.1: International comparison of employmenffpopulation ratios,'' 1999 Total Men wornen- All aaes2) 15 to 24 25 to 54 55 to 64- All ages2) All agesz) Australia 76.5 59.9 Austria 76.7 59.7 Belgium 67.5 50.2 Canada 75.5 64.7 Czech Republic 74.3 57.4 Denmark 81.2 71.6 Finland 68.4 63.5 France 66.8 52.9 Germany 73.1 56.5 Greece3' 71.6 49.0 Hungary 62.6 49.0 Iceland4' 88.2 80.2 Ireland 73.5 51.3 Italy 67.1 38.1 Japan 81.0 56.7 Korea, 1997 76.0 51.6 Korea,1998 71.7 47.4 Korea,1999 71.5 48.1 Luxembourg 74.4 48.5 Mexico 84.8 39.6 Netherlands 80.3 61.3 New Zealand 77.3 63.0 NorwafJ 82.1 73.8 Poland3) 65.8 52.2 Portugal 75.5 59.4 Spain4) 69.6 38.3 Sweden4' 74.8 70.9 Switzerland 87.2 71.8 Turkey 71.7 32.0 United Kingdom4' 78.4 64.9 United States4) 80.5 67.6 Total OECD5' 76.6 55.4 Notes. 1) EmploymentJpopulat~onratios refer to persons in employment divided by the working-age population for the relevant groups. 2) 'All ages" refers to 15 to 64. 3) Data refer to 1998. 4) Age group 15 to 24 refers to 16 to 24, and "All ages" refers to 16to 64. 5) For above countries only. Source: OECD (June2000). Employment OuNook. Statistical Annex. Tables B and C. KOREAN LABOR MARKET A N D SOCIAL SAFETY-NET REFORMS: ... 365 Table 16.2: Internationalcomparison of employment by status of workers, 1999 " Employees Unpaid Self- Non- family employment permanentlNon workers2) regular regular (% of total employment) Australia 86 63 23 1 14 Austria 86 80 7 3 11 Belgium 82 75 6 3 15 Canada 82 73 10 0 17 Czech Republic 86 80 7 0 13 Denmark 91 81 9 1 8 Finland 86 71 15 1 14 France 89 77 12 0 11 Germany 89 78 11 1 10 Greece 55 48 7 12 33 Hungary 85 81 5 1 14 Iceland 82 73 9 0 18 Ireland 79 72 7 1 20 Italy 72 65 6 4 24 Japan 83 73 10 6 12 Korea, 1997 63 34 29 9 28 Korea, 1998 61 32 29 10 29 Korea, 1999 62 30 32 9 29 Luxembourg 93 90 3 0 7 Mexico 61 51 10 10 29 Netherlands 88 77 11 1 11 New Zealand 80 n.a n.a 1 19 Norway 92 81 10 1 8 Poland3) 70 n.a n.a 5 26 Portugal 71 59 12 2 27 Spain 77 52 25 3 20 Sweden 89 78 11 0 10 Switzerland 87 77 10 2 11 Turkey 44 35 8 25 31 United Kingdom 87 81 6 1 12 United States 92 n.a n.a 0 8 Notes: 1) Data relate to 1997 for Greece, Ireland. Luxembourg and Portugal;and as indicated in the case of Korea. 2) Figures less than 0.5 appearin the column as zeros in italics. n.a)Data not ava~lable. Sources: Korea: National Statistical Office (1997 and 1998), Annual Report on the EconomfcallyActive Population Survey (December 1999), Monthly Report on the EconomfcaNyActive Populatfon Survey Japan: Ministry of Labor (1998). Year Book of Actwe PopulatfonLabour Stat~stics,and for other countries: OECD (1999),Labour Force Statistics(for data on total employees, unpa~d family workers and self-employment), and OECDIDEELSA Stat~sticsand lnd~catorsD~v~sion, temporary-work database (for the breakdown of permanentinon-permanentwork) 366 Labor Market Refon~zsin KOI-en;Policy Optionsfor the Fzrttrre Figure 16.1: Composition of employment by status of work in Korea, 1990-1999 P~n?ulA- Employment by status of work, 1990-1999 l 5-25 30.4, Is-$9 60 1i.g 111.24 Ji.ill 60 Men Women I.*\.lhrn ~lllddlo.il,l,tll Il b ~ l\~I,t.t,l l Cilll~~~ B lll,licrllli C ,l,d,,,",d,Il~ \~l,,><,l11831,,~llt>,>lC~,IIC~C ,,,,,,~..il~ B ~L \< - - Men Women p Sources: Panel A. National Statlstlcal Offlce (1998). Annual Report on the EconornicaNyActive Population Survey, and ( D e c e m b e r 1999), Monthly Report on Me EconomicallyActive PopulationSurvey. Panels B and C. Direct subm~ss~ons the Korean author~t~es by Table 16.3: Job tenure in Korea and selected OECD countries Korea European Union 'I2) Japan1) United Stat& 1996 1997 1998 1995 1995 1995 (average number of years with the same employer) Total employees 5.4 5.7 6.1 9.8 11.3 7.4 Men 6.1 6.3 6.8 10.5 12.9 7.9 Women 3.7 4.0 4.3 8.8 7.9 6.8 Middle school and below 5.7 6.1 6.4 10.1 15.3 5.8 High school 5.1 5.4 5.8 9.4 11.4 7.9 Junior college 4.5 4.6 5.1 College and university and over 10.1 9.5 7.4 6.2 6.5 7.0 Notes: 1) Average for different levels of educational attainment are based on weighted averages of mid-points of tenure classes. The two highest educational levels are not shown separately and correspond to some or completed tertiary education. These data were published in Table 5.6. CECD employmentOutlook. July 1997. 2) Unweightedaverage on the different breakdowns of tenure for the 15 member countries. Sources: For Korea: Direct submissions by the Ministry of Labor on the basis of the report on Wage Structure Survey; For other countries,: OECD(July 1997), Employment Outlook, Table 5.6. Table 16.4: Coverage of the employment insurance system(EIS) in Korea, July 1995 to February 2000 July 1995 January 1998 July 1999 February 2000 ('000s) Paid workers 12824 12500 12603 12819 Eligible workers 4280 5190 8342 8700 Insured workers 4204 4309 5876 6172 As a proportion of eligible (98.2) (83.0) (70.4) (70.9) workers(%) Source: Direct submission by the Korean authorit~es,and Secretariatestimates 368 Labor Mc~~-ket Refbrnzs in Korea: Polic,v Optionsfor [he Fz[ti~l.e with educational attainment of middle school and below) than for those with high-school education or junior college -- signaling problems in the school-to-work transition. O Regular workers tend to work in large firms. This is important because the wage gap between large and small firms has been growing in recent years (Chart 16.2). Working in a large tirm entails other benefits to workers: a) in 1995, the unionization rate in firms with more than 300 workers exceeded 80 %, whereas it was only 5 % in firms with less than 100 workers (and probably almost nil in small businesses); 6) workers in large firms are covered by both the Employment Insurance System and the National Pension System which, despite government efforts to expand coverage, is far from being the case in small firms; and c) in 1999, nearly 40 % of workers in large firms received in-plant training, compared with less than 3 % in small firms. While there are probably many factors which account for such a pronounced degree of duality in the Korean labor market, employment protection is an important part of the story. indeed, regular workers enjoy a relatively high level of employment protection. For instance, severance pay amounts to one monthly wage per year of seniority (with no maximum) and dismissal procedures remain complex and cumbersome despite recent reform^.^ In addition, the preconditions for dismissals based on economic reasons (advance notification to a trade union, verifiable efforts to avoid dismissal, etc.) make conditions in Korea resemble closely those in other high-protection countries such as Japan, Portugal, France or Germany. Finally, the new provisions on notification of collective dismissals to the Ministry of Labor correspond to the strict regulations in force in EU countries. Overall, the OECD has ranked Korea as the second strictest country out of 27 Member countries in terms of protection of regular (or permanent) employment contracts7. By contrast, employers are under no obligation to provide severance pay to temporary and daily workers who stay less than one year in their enterprise. Another possible explanation is the structure of product markets. Competition among large finns is weak (domestically), reflecting cartel-type arrangements. At the same time, large firms have managed to sub-contract work to small ones, to remain competitive. As noted above, regular workers are concentrated in large firms, implying possibilities for rent-sharing between them and their employers. " While the one-month notice per~od.even for workers with high tenure. is low by international standards, the legal severance pay requirements applying to regular workers would seem to be relatively onerous. ' For details, see OECD (1999a. Chapter 2) aql j o 3!is!~aw~1ey:,IeJnpnJls e aq 01 suraas ssausno!~e:,a~d qo[ ' s n q ~ .luau~Xoldura11nj-~eauiCq paz!Jal3enq:, popad P '~6-0661u! q m m aseanu! lou p!p luauriCo1dura ~ e l n 4 aj o~ a:,uap!3u! ayl ley1 passalls aq plnoys ' J ~ A ~ M ~ H'sauo ~ e l n 9 aolu! sqof ~eln8a~-uou1.1a~uo3ol luel3nlaJ ssa1 aq ~I!M ~ Lay1 ' ~ n s ae ~se 'ley] pue sl3adso~d3!urouo~aa w n j lnoqe luapyuo3 alour au[o:,3q [I!M s~a/Co[duraleyl s! no!lnl3adxa ayl 'sanu!~uo:, iC~a~o3a.1ayl j r 'sa!l!loglne uealoy aqj dq palyddns saieu!lsa leuo!s!~oldale 6 6 6 ~pue 8661 J O ~sa~n6ysnleis leuo!iedn330 uo elep 104 aln$!$sulJnoqel eajo>l u o ~ 4uo!ss!luqns paJ!p pue :az!s UJM u o eiep 104 ' A a ~ n s~oqe7A/yjuo~('(sanssl s n o ! ~ e n :luauu!eUe leuo!je3npa pue sdnol6 a6e ) '9apua6 dq e)ep 104 'a~npnqsa b e ~uo godaw A a ~ n s'(sanssl sno!len) Joqel 40 hls!u!w :Sa3JnOS a l o u l o SlaYloM L q 1 1sjuauqs!lqelsa o ~ ol 6 u ! ~ l a ~'E'Z ue43 u! se sluauodwo3 a 6 e ~Ile 6u!pnl3u! s a 6 e ~lenuue a6e~anv(L :ajoN' a ~ (161118661 1661 9661 1661 t661 1661 7661 1661 11661 61161 8861 '1161 'J861 5x61 tS61 !X61 i R h l l X h l 11861 $1'2 ' i t'i - on Temporary Work Businesses in Korea, Seoul: Korea Labor Institute (in Korean). Kraft, K. (1993), "Eurosclerosis Reconsidered: Employ~nentProtection and Employment Adjustment in West Germany", in C. F. Buechteman (eds.), Employment Security and Labor Market Behavior: Inter- disciplinary Approaches and International Evidence, Ithaca: Cornell University Press, pp.297-304. Korea Labor Institute (1999), Qt~arterlyLabor Review, 12 (3). Mills, Daniel Q. (1994), Labor-Management Relations, Singapore: McGraw-Hill. OECD (1999), Employment Outlook, Paris. (1997), Implementing the OECD Jobs Strategy: Member Countries Experience, Paris. ( 1 994), The OECD Jobs Sttldy, Evidence and Explanations, Part 2. Paris. Pak, Po-hi, Moon-Hong Min, and Young-Sook Kang (1999), "Social Safety Net for the Most Vulnerable Groups in the Republic of Korea: A Working Paper prepared for the UNDP", UNDP. Potts, N. (1999), "Flexibility and all that: Understanding the EU Labour Market", European Business Review, 99 (3), pp.1-19. Purcell, J. (1996), "Contingent Workers and Human Resource Strategy: Rediscovering the Core-Periphery Dimension", Jot~rnal of Professional HRM, 5, pp.16-23. Tregaskis, 0 . (1997), "The 'Non-Permanent' Reality!", MCB Employee Relations, 19 (6), pp.I -12. Treu, Tiziano (1992), "Labour Flexibility in Europe", International Labour Review, 13 1 (4-3, pp.497-512. Wilkinson, B. (2000), "Employment Agencies and Labour Market Flexibility in the UK", Draft for World BanWASEM-EU Study. Chapter 8 Ackroyd, P. and P. Procter (1998), "British Manufacturing Organization and Workplace lndustrial Relations: Some Attributes of the New Flexible Firm", British Journal of Indz~strlalRelations, 36 (2) pp.163-183. Araki, T. (1999) "1999 Revisions of Employment security Law and Worker Dispatching Law: Drastic Reforms of Japanese Labour Market Regulations", Japan Labour Bulletin, 38 (9). (1994), "Characteristics of Regulations on Dispatched Work (Temporary work) in Japan", Japan Labour Bulletin, 33 (8). Arulampalam, W. and A. Booth (1998), "Training and Labour Market Flexibility: is there a Trade-off?", British Journal of Industrial Relations, 36 (4), pp.521-536. Bamard, C. and S. Deakin (1999), "A Year of Living Dangerously? EC Social Rights, Employment Policy and EMU", Industrial Relations Journal, 30 (4). Bassett, P. (1986), Strike Free: New Industrial Relations in Britain, London: MacMillan. Beatson, M. (1995), Labour Market Flexibility, Employment Department Research Series No.48. Benson, J. and P. Debroux (1997), "KRM in Japanese Enterprises: Trends and Challenges", Asia Pac~JicBusiness Review, 3 (4), pp.62-8 1. Berggren, C. and M. Nomura (1997), The Resilience of Corporate Japan, London: Paul Chapman. Bosch, G. (1994), Flexibility and Work Organuation, Report to the European Commission of an Expert Working Group. Bronstein, A. (1991), "Temporary Work in Western Europe: Threat or Complement to Permanent Employment?", International Labour Review, 130 (3), pp.291-310. Brown, W. and S. Wadhwani (1990), "The Econolnic Effects of lndustrial Relations Legislation Since 1979", h'ational Institute Economic Review, 131, pp.57-70. Brown, W., Deakin, S. and P. Ryan (1997), "The Effects of British Industrial Relations Legislation 1979-97", National Institute Economic Review, 161, pp.69-83. Burchell, B., Deakin, S. and S. Honey (1999), The Employment Status of Individuals in Non-standard Employment, Report to the Department of Trade and Industry. Casey, B., Metcalf, R. and N. Millward (1997), Employers' Use of Flexible Labour, London: Policy Studies Institute. 428 Labor Market Refornzs in Korea: Policy Opfionfor the F~lfz~re Croucher, R. and C. Brewster (1998), "Flexible Working Practices and the Trade Unions", Employee Relations,20 (5). Cully, M., O'Reilly, A., Millward, N., Woodland, S., Dix, G. and A. Bryson (1999), The 1998 Workplace Employee Relations Survey: First Findings, Report to the DTI, ACAS, ESRC and PSI. Department of Trade and Industry (Employment Agency Standards) (1999), Regulation of the Private Recrzritment fndzistry: A Consultation Doctrment, DTI. (1 998), Employment Rights on the Transfer ofan Undertaking, DTI, PL 699 (Rev 4). (1996), The Employment Agencies Act 1973: Guidance on the Act and Regzilations,DTI. Dore, R. (1987), Taking Japan Seriously, London: Athlone. (1996), "The End of Jobs for Life? Corporate Employment Systems: Japan and Elsewhere", Centre for Economic Perfor- mance Occasional Paper No. 11, London School of Economics, March. Eckert, C. (1990), "The South Korean Bourgeoisie: a Class in Search of Hegemony", Journal of Korean Studies, 7, pp.115-148. Federation of Recruitment and Employment Services (1997/98), Annual Recruitment Industry Survey 1997/8: The FRES Survey. Fenton-Jones, R. (1999), After Outsourcing: Managing Tripartite Business Relationships, University of Bath School of Management, MBA dissertation. Genda, Yuji (1999), "The Changing Employment Structure", Japan Labour Bulletin, 38 (3). Guest, D. (1997),"Towards Jobs and Justice in Europe: a Research Agenda", Industrial Relations Journal,28 (4). Haskel, J., Kersley, B. and C. Martin (1997), "Labour Market Flexibility and Employment Adjustment: Micro Evidence from UK Establish- ments", Oxford Economic Papers, 49, pp.362-379. Hall, B. and J. Willems (1997), "Einployment Law", Law Journal Extra, 13 October. Houseman, S. (1999), Flexible Staffing Arrangen7ents: A Report on Temporary Help, On-Call, Direct Hire Te~nporary,Leased, Contract Company, and Independent Contractor Employment in the United States, Washington, D.C.: US Department of Labour. Institute of Management and Manpower (1996), Survey of Long Term Employment Strategies, London: Institute of Management. IRS (Various Issues), Employee Development Blilletin, Industrial Relations Services, UK. (Various Issues), Employment Review, Industrial Relations Services, UK. (Various Issues), Employment Trends, Industrial Relations Services, UK. (Various Issues), Indzutrial Relations Law Bulletin, Industrial Relations Services, UK. Japan Institute of Labour (Various Issues), "Short Reports on 'Working Conditions and the Labour Market"', Japan Labour Bulletin. Kern, H. and M. Schulnann (1987), "Limits to the Division of Labour", Economic and Industrial Democracy, 8 (2), pp.151-170. Kleinknecht, A. (1998), "Is Labour Market Flexibility Harmful to Innova- tion?', Cambridge Journal of Economics, 22 (3), pp.387-396. Kumazawa, M. and J. Yamada (1989), "Jobs and Skills under the Lifelong Nenko Employment Practice", in Wood, S. (ed.) The Trunsfor- mation of Work?, London: Unwin Hyman. Labour Market Trends (1998), Trade Union Membership and Recognition 1996-97: an Analysis of Data @om the Certljication OfJicer and the LFS, London: HMSO, July. (1997), Temporary Workers in Great Britain, London: HMSO, September. Local Government Management Board (1998), Flexible Working: Local Government and the Wider Economy, London. Ministry of Agriculture, Fisheries and Forestry (1998), Planting, Picking or Packing Agricultural Produce? Your Rights Explained, MAFF Web Site. Ministry of Labour (Japan) (1999), White Paper on Labour: The Rapidly Changing Labour Market and New Job Creation. MINTEL Market Intelligence (1998), Employment Agencies, MINTEL, December. Murray, J. (1999), "Normalising Temporary work: The Proposed Directive on Fixed-Term Work ', Indzrstrial Law Jozlrnal, 28 (3), pp.269- 7 275. Nam II Chong, Kim Joon-Kyung, Kang Yeongjae, Joh Sung Wook and Kim Jun 11,"Corporate Governance in Korea", Paper Presented to the Conference on Corporate Governance in Asia: A comparative Perspective, OECD/Korea Development Institute, Seoul, 3-5 March. National Association of Citizens Advice Bureaux (1997), Flexibility Abzrsed: a CAB Evidence Report on Employment Conditions in the Labozlr Market, NACAB. Oliver, N. and B. Wilkinson (1988), The Japanization of British Indzatry; Oxford: Blackwells. OECD (1986), Flexibility in the Labour Market: the Ctlrrent Debate, Paris. (1977), The Development of Industrial Relutions Systems: Some Implications of the Japanese Experience, Paris. Osawa, M. and J. Kingston (1996), "Flexibility and Inspiration: Restruc- turing and the Japanese Labour Market", Japan Labozlr Bulletin, 35 (1). Park, Duck Jay, Gychang Yu and Jonghee Park (2000), "Assessment of 430 Lnbor Market Reforn~sin Korea: Policy Optioi~jOl-tlie Fzltzrre Labor Market Responses to the Labor Law Changes Introduced in 1998", Draft Report to the World Bank, February (Revised and Included as Chapter 7 of this Volume). (1992), "Industrial Relations in Korea", International Journal of Human Resource Management, 3 ( 1), pp.105-124. Piore, M. and C. Sabel (1984), The Second Industrial Dfvrde, New York: Basic Books. (1986), "Perspectives on Labour Market Flexibility", Indzrstrial Relatrons, 25 (2), pp.146-166. Potts, N. (1999), "Flexibility and All That: Understanding the EU Labour Market", European Btrsiness Review, 99 (3). Purcell, J. (1996), "Contingent Workers and Human Resource Strategy: Rediscovering the Core-Periphery Dimension", Journal of Professional HRM, 5, pp.16-23. Purcell, K., Hogarth, T. and C. Simon (1999), Whose Fle>xibility?The Costs and Benefits of Non-Standard Workrng Arrangements and Contractual Relations, Joseph Rowntree FoundationIYork Publi- shing Services. Rhee Yang Soo (1985), "A Cross-cultural Comparison of Korean and American Managerial Styles: an Inventory of Propositions", in Kim Bun Woong, Bell, D. and Lee Chong Bum, Administrative Dynamics and Development. the Korean Experience, Seoul: Kyobo, pp.78-98. Rothwell, S. (1990), Changing Labour Markets What rs the Role for Employment Agencres? Henley Working Paper Series No.5190, Henley: The Management College. Sako, M. (1997), "Forces for Homogeneity and Diversity in the Japanese Industrial Relations System", in Sako, M. and Sato, H., Japanese Labour and Management in Transition, London: Routledge. Sato, Hiroki (2000), "The Current Situation of 'Family-friendly' Policies in Japan", Japan Labour Bulletin, 39 (2). (1996), "Keeping Employees Employed: Shzrkko and Tenseki Job Transfers - Formation of a Labour Market within Corporate Groups", Japan Labour Bulletin, 35 (1 2). Trades Union Congress (1994), Report of the 126'" Annual Trades Union Congress, Blackpool, 5-9 September. Tregaskis, 0. (1997), "The Non-permanent Reality ', Enzployee Relations, 7 19(6). US Department of Labour (1999), Futurework: Trends and Challenges for the 2I"Century. Whitley, R. (1992), Business Syste~nsrn East Asia, London: Sage. Wood, S. (1999), "Family-friendly Management: Testing the Various Perspectives", National Institute Economic Revrew, 168. Yamakawa, R. (1999), "The Silence of Stockholders: Japanese Labor Law from the Viewpoint of Corporate Governance ', Japan Labotlr 7 Bulletin, 38 (1 1). (1998), "Overhaul After 50 Years: The Amendment of the Labour Standards Law", Japun Labour Bzilletin,37 (11). (1995), "Client Company's Duty to Bargain With Dispatched Workers' Union: Asahi Broadcasting Case", Japan Labour Bzilletin, 34 ( I2). Yasui, T. (1999), "Corporate Governance in Japan", Paper Presented to the Conference on Corporate Governance in Asia: A Comparative Perspective, OECDiKorea Development Institute, Seoul, 3-5 March. Chapter 9 Calmfors, Lars (1994), "Active Labor Market Policy and Unemploy~nent- A Framework for Crucial Design Features", OECD Economic Studies, 22, Paris: OECD. Erhel, Christine, Jerome Gautie, Bernard Gazier and Sylvie Morel (1996), "Job Opportunities for the Hard-to-place," in eci. by Schmid, Gunther et al., International Handbook of Labor Market Policy and Evaluation, Cheltenham and Brookfield: Edward Elgar. Forslund Anders and Alan B. Krueger (1994), "An Evaluation of the Swedish Active Labor Market Policy: New and Received Wisdom", National Bureau of Economic Research Working Paper No. 4802, Washington, D.C.: NBER. Freedman, David H. (1990), "Special Employment Programmes in Developed and Developing Countries", International Labour Review, 129, p.2. Jeong, Insoo, Byoungkoo Kang, Chungwan Shin, Wanshik Yoo, and Hongkeun Chang(2000), An Evaluation of the Unemployment Policies in 1999, Seoul: Korea Labor Institute (in Korean). Lee, Joohee (1999), "Public Works Project", in Korean Unemployment Policy, Seoul: Korea Labor Institute (in Korean). Martin, John P. (1998), "What Works among Active Labor Market Policies: Evidence Form OECD Countries' Experiences", Labor Market and Social Policy Occasional Papers, No. 35. OECD (1998), "Key Employment Policy Challenges Faced by OECD Countries", Labor Market and Social Policy Occasional Papers No. 35. OECD submission to the G8 Growth, Employability and Inclusion Conference London, Feb. 1998. Rose, Nancy E. (1994), Pzit to Work: Relief Programs in the Great Depression, New York: Monthly Review Press. Schmid, Gunther (1996), "Process Evaluation: Policy Formation and Implementation," in ed. by Schmid, Gunther et al., International Handbook of Labor Market Policy and Evaluation, Cheltenham and Brookfield: Edward Elgar. 432 Ltrbor Market Reforms in Korea: Policy Opt~on Jar tile filtzlre Chapter 10 Andenas, K. (1992), Sosialomsorg i gode og onde dager, 0slo.Tano. Baldwin, P. (1990), The Politics of Social Solidarity: Class Basis of the European Welfae States 1875-1975. Cambridge: Cambridge University Press. Choo, H. (1993), Income Distribution and Distributive Equity in Korea, in Krause, L. B. and Park, F. K. (eds), Soc~alissues in Korea: Korean and American Perspectives, Seoul: Korea Development Institute. Ditch, J. and N. Oldfield (1999), "Social Assistance: Recent Trends and Themes ', Journal of European Social Policy, 9 (1 ), pp.65-76. 7 Esping Andersen, G. (1990), The Three Worlds of Welfare Capitulum, Cambridge: Polity Press. European Commission (1999a), Employment in Europe, Brussels: European Commission. (1999b), Joint Employment Report, Brussels: European Com- mission. Gilbert, N. (1995), "Welfare Justice", Restoring Social Equity, New Haven: Yale University Press. and B. Gilbert (1989), The Enabling State. Modern Welfare Capitalism in America, New York: Oxford University Press. Heikkila, M. (1999), A Brief Introduction to the Topic, in Linkrng Welfnre and Work, Luxembourg: European Foundation for the Improve- ment of Living and Working Conditions. Hvinden, B. (1999), Activation: a Nordic Perspective, in Linking Welfare and Work, Luxembourg: European Foundation for the Improve- ment of Living and Working Conditions. Kang, S.H. (1999), Labor Market Outcomes and Policy Responses after the Crisis: The Case of Korea, Seoul: Korea Labor Institute. Katz, M. (1996), In the Shadow of the Poor House, A Social History of Welfare In America, New York: Basics Books. Kuhnle, S. (1983), Velferdsstatens utvikling, Norge i komparativt perspektiv, Bergen, Universitetsforlaget. Kwon, S. (1993), Socialpolicy in Korea. Challenges and Responses, Seoul: Korea Development Institute. Lee, J. (2000), "Assessment of the Cost-effectiveness of Public Works and Loan Schemes", Interim report prepared for the World Bank, Seoul: Korea Labor Institute (Revised and Included as Chapter 9 of this Volume) Leipziger, D.M.(1993), "Korea: issues of distribution", in Krause, L.B. and Park, F.K. (eds.), Social Issues in Korea: Korean and American Perspectives, Seoul: Korea Development Institute. L~demel,I. (1997), The Welfare Paradox: Income Maintenance and Personal Social Services in Norway and Britain 1946-1966, Oslo: Scandinavian University Press. and 13. Trickey (2000), An Offer You Can't Refuse: Workfare in Inrernational Perspective, Policy Press (forthcoming). Ministry of Health and Welfare, Korea (2000), Plans for Implementing Major Welfare Programs. OECD (1996), Employment Outlook, Paris. (1994), The OECD Jobs Study, Paris. Park, Y.-b. (2000), "Study on Feasibility of Introducing a Non-contributory Cash Benefits System for the unemployed in Korea" (Revised and Included as Chapter 5 of this Volume). Peck, J. (1998), "Workfare, a geopolitical etymology", Environment and Planning Society and Space, 16, pp.133-160. Salonen, T. and H. Johansson (2000), "A Bridge or a Side-Track? Workfare and Activation Policies in the Swedish Welfare State", Dzrtch Journal for Work and Particiaption, 21 (2/3), pp.143-156. Shragge, E. (Ed) (1997), Workfare: Ideology for a New Underclass, Garamond Press: Toronto. Spies, H. (1996), "Workfare: Emancipation or Marginalisation?" in M.P.M. de Goede, P.M. de Klaver, J.A.C. van Ophem, C.H.A. Verhaar, A. De Vries (ed.), Yozrth: Unemploymenf, Identity and Policy. Altershot: Avebury. Struthers, J. (1 996), Can Workfare Work? Rejections from history, Caledon Institute of Social Policy. Wilensky, H.L. (1992), "Active Labour Policy: its Content, Effectiveness and Odd Relationship to Evaluation Research", in Heath, A Social Research and Social Reform, Oxford: Clarendon Press. Chapter 11 Bang, H.-n. (1999), Economic Activities of Korean Hotlsehold and Individz~als,Seoul: Korea Labor Institute (in Korean). DEETYA, Australia (1998), Job Seeker Classification Instrument. Eberts, R. W. and C. J. O'Leary (1996), Profiling Unemployment Insurance Beneficiaries, USA: Upjohn lnstitute. Human Resources Development, Canada (1997), "A Strategic Approach to HRDC's Labour Market Information Function", (mimeo). Jeong, 1.-s. et al. (1999), Une~nploymentMeasures Foct~sedon Kyeong-gi Province, Seoul: Korea Labor Instilute (in Korean). Hunter, John (2000), "The Korean Public Employment Service", Korea Labor Institute1 World Bank (mimeo). Japan Institute of Labor (1998), "Public Employment Services in Japan", A Report to ILO, Kang, S-H. (1999), Unemployment Policies in Korea, Seoul: Korea Labor Institute (in Korean). 434 Lobor Market Refur-nzsin KO/-en.Policy Optronfbr the Flitz/t*e Keum, J. (2000), "Unemployment Structure and Job Search-Recruiting Activities in Korea", Korea-Japan Joint Workshop on Unemploy- ment Structure and Policies, Korea Labor Institute and Japan Institute of Labour (in Korean). (1999a), "The Financial Crisis and Labor Market Services in Korea", Seminar on the Economic Crisis, Employment cmd the Labor Market in East and South-East Asia, WB-ILO-JMOLIJIL, Oct. 1999. ( I 999b), A Survey on Needs ofthe PES: Job Searchers, Firms, and Counselors,Seoul: Korea Labor Institute (in Korean). (1999c), "Mid and Long-term Strategies for the Labor Market Information System in Korea", Seoul: Korea Labor Institute (in Korean) et al. (1998), " A Survey on Unemployment and Welfare Needs", Korea Labor Institute and Korea Institute of Health and Social Affairs (in Korean). (1997), Urban Unemployment in Korea,Seoul: Korea Labor Institute. Korea International Labour Foundation (1999), Unemployment Measures: From Desperation to Reconstruction, Seoul. Korea Labor Institute (1999a), Korea Labor and Inconle Panel Study, Seoul (in Korean). -- (1999b), A Survey on Firrn's Hiring Practices (in Korean). (1998), Korea Labor and Income Panel Sttidy, Seoul (in Korean). Martin, John P. (1998), What Works among Active Labor Market Policies: Evidencefrom OECD Countries 'Experiences, OECD. OECD (2000), Review of Labor Market and Social Safety-Net Policies in Korea, Paris. Roberts, Margaret (1999), "Improving the Basis for Labor Market Informa- tion in Korea: A New Occupational Classification", A Report to World Bank, Wahsington. Yoo, K.-s. (1999), The Employnlent Insurance System in Korea, Seoul: Korea Labor Institute. Chapter 12 Fay, Robert G. (1996), "Enhancing the Effectiveness of Active Labour Market Policies: Evidence from Progra~nineEvaluations in OECD Countries", OECD Labour Market and Social Policy Occasional Papers, No. 18. Paris: OECD. Grubb, W. Norton and Paul Ryan (1999), The Roles of Evaluation for Vocational Edzrcationand Training, Geneva: ILO. Kang, Soon-Hie, Byung-Hee Lee and Mi-Ran Kim(2000), Performance Analysis on Vocational Ability Program in Korea, Seoul: Korea Labor Institute(in Korean). BIBLIOGRAPHY 435 -, Jaeho Keum, Dong-Heon Kim and Donggyun Shin(1999), Labor Market Outcomes and Policy Responses after the Crisis: The Case ofKorea,Seoul: Korea Labor Institute. Martin, John P. (1998), "What Works among Active Labour Market Policies: Evidence from OECD Countries' Experiences", OECD Labour Market and Social Policy Occasional Papers, No. 35. Paris: OECD. Ministry of Labor, Korea (1999), Midterm Unemployment Policy(in Korean). OECD (1991), Evaluating Labor Market and Social Programmes: The State of a Complex Art, Paris. Park, Funkoo and Joohee Lee (1999), "The Social Impact of the Financial Crisis: Labor Market Outcomes and Policy Responses in Korea", Unpublished Paper, Seoul: Korea Labor Institute. Payne, J. C., S. Lissenburgh and M. Range(1998), Work-based Training and Job Prospects for the Unemployed: An Evaltlation of Training for Work, London: Policy Studies Institute, Department for Education and Employment. Yoo, Kil-Sang (1999), The Employment Insurance System in Korea, Seoul: Korea Labor lnstitute. Chapter 13 Bach, H.-U. et al. (1998), "Labor Market Trends and Active Labor Market Policy in the East German Transformation Process, 1990-97", Unpublished Paper, Niirnberg: Bundesantalt fur Arbeit. Bamow, B. (1992), "The Effect of Performance Standards on State and Local Programs", in C.Manski and I.Garfinke1 (eds.), Evaluating Welfare and Training Programs, Cambridge, Mass.: Harvard University Press. Blaschke, D. and E. Nagel (1995), "Beschaftigungssituation von TeiInehmem an AFG-finanzierter beruflicher Weiterbildung", Mitteilungen aus der Arbeitsmarkt- und Berufsforschung, 28 (2), pp.195-213. -, H.-E. Plath and E. Nagel (1992), "Konzepte und Probleme der Evaluation aktiver Arbeitsmarkpolitik am Beispiel Fortbildung und Umschulung", Mitteilungen aus der Arbeitsmarkt-und Berufsforschung, 25 (3), pp.381-404. Cho, E. and Y. Ra (1999), "Training Policies and Programs for the Unemployed in Korea", Working Paper, Seoul: Korean Research Institute for Vocational Education and Training. Dar, A. and I.S. Gill (1995), "Evaluations of Retraining Programs in OECD Countries: Implications for Economies in Transition", Report IDP-156, Washington, DC: World Bank. DfEE (1998), Labor Market Qtrarterly Report, August, p.7, London: 436 Labor Market Rejornzs in Korea: Policy Optionjor tile Ftfttlre Department for Education and Employment. Fitzenberg, B. and S. Speckesser (2000 forthcoming), "Zur wissenscahft- lichen Evaluation der aktiven Arbeitsmarketpolitik in Deutsch- land: ein Uberblick", Mitteilungen aus der Arbeitsmarkt- und Berufsforschung. and H. Prey (1999), "Berufliche Weiterbildung und die Stabilitat der Beschaftigung: Eine Evaluation auf Basis des Sozio- Okonomisches Panels", in D.Beer (ed.), Die wirtschaftlichen Folgen von Aus- und Weiterbildung, Munich: R.Hampp Verlag. Fitzenberger, B. and H. Prey (2000), "Evaluating Public Sector Sponsored Training in East Germany", forthcoming, Oxford Economic Papers. Grubb, W.N. (1996), Learning to Work: The Case for Re-Integrating Job Train~ngand Edtrcation, New York: Russell Sage Foundation. and P. Ryan (1999), The Roles of Evaluation for Education and Training: Plain Talk on the Field of Dreams, London: Kogan Page. Heckman, J.J. and J.A. Smith (1999), "The Pre-program Earnings Dip and the Determinants of Participation in a Social Program: Implica- tions for Simple Program Evaluation Strategies", Economic Journal, 109, August, pp.313-348. , R.J. Lalonde and J.A. Smith (1999), "The Economics and Econometrics of Active Labor Market Programs", in 0. Ashenfelter and D. Card (eds.), Handbook of Labor Economics, Vo1.3, Elsevier: Amsterdam. Hofbauer, H. and W. Dadzio (1984), "Berufliche Weiterbildung fur Arbeitslose", Mitteilungen aus der Arbeitsmarkt- und Berufs- forschung, 17 (2), pp.129-141. -, and W. Dadzio (1987), "Mittelfristige Wirkungen beruflichen Weiterbildung; die berufliche Situation von Teilnehmem zwei Jahre nach Beendigugn der MaRnahme", Mitteilungen aus der Arbeitsmarkt- und Berufsforschung, 20 (2), pp.129-141. Hiibler, 0 . (1997), "Evaluation beschaftigungspolitischer Massnahmen in Ostdeutschland", Jahrbiicher fur Nationalokonornie und Statistik, 216 (2), pp.21-44. Hujer, R. and M. Wellner (2000), "The Effects of Public Sector Sponsored Training on Individual Employment Performance in East Germany", Working Paper, Frankfurt: J.W.Goethe University. Jeong, J. (1995), "The Failure of Recent State Vocational Training Policies in Korea from a Comparative Perspective", British Jotrrnal of Industrial Relations, 33 (2), June, pp.237-252. Johanson, R. (1994), Retraining adults in Germany, Training Policy Study Number 4, Geneva: ILO. Kang, S.-H. and B.H. Lee (2000), "Evaluating Training for the Unemployed in Korea", Unpublished Paper, Seoul: Korean Labor Institute (Revised and Included as Chapter 12 of this Volume). ------, J. H. Keum, D.-H. Kim and D. Shin (1999), "Labor Market Outcomes and Policy Responses after the Crisis: the Case of Korea", unpublished paper, Seoul: Korean Labor Institute. Kim, B., C. Jang, J. Choi and J. Kwan (1998), The Real Picture of Vocational Training and Policy Issz~esunder Mass Unemployment in Korea, Seoul: Korean Research Institute for Vocational Education and Training. Korean Ministry of Labor, Korea (1998), "Training for the Unemployed: 1998.7, Summary Statistics". Kraus, F., Puhani, P. and V. Steiner (1999), "Employment Effects of Training Programs-the East German Experience", Jahrbucher fur Nationalokonomie und Statistik,219 ( I ), pp.2 16-248. Lalonde, R. (1986), "Evaluating the Econometric Evaluations of Training Programs with Experimental Data", American Economic Review, 76, pp.604-620. Lecher, M. (2000 forthcoming), "An Evaluation of Public-sector-sponsored Continuous Vocational Training Programs in East Germany", Journal of Human Resources. (1995), "Effects of Continuing Off-the-job Training in East Germany after Unification", Discussion Paper 95-27, Mannheim: Zentrum fur Europaishce Wirtschaftsforschung. Martin, J. (1998), "What Works among Active Labor Market Policies: Evidence from OECD Countries' Experiences", Labor Market and Social Policy Occasional Paper number 35, Paris. OECD (1999), OECD Economic Surveys 1998-99: Korea, Paris. (1998), Review of National Policies for Education: Korea, Paris. Pannenberg, M. (1996), "Zur Evaluation Staatlicher Qualifizierungsmarjnah- men in Ostdeutschland: Das Instrument Fortbildung und Umschulung, Discussion Paper 38, Halle: Institut fur Wirtschaks- forschung. (1995), Weiterbildungsaktivitaten und Enverbsbiographie. Eine empirische Analyse fur Deutschland, Frankfurt: Campus Verlag (cited in Fitzenberger and Speckesser, 2000). Park, Y.-b. (1999), "Unemployment and the Social Safety Net in Korea", Unpublished Paper, Seoul: Hansung University. Payne, J., Payne C., Lissenburgh S. and M. Range (1999), Work-Based Training and Job Prospects for the Unemployed: an Evaluation of Training for Work, Research Report RR96, London: Department for Education and Employment. -, Lissenburgh S., White M. and C. Payne (1996), Employment Training and Employment Action: an Evaluation by the Matched Comparison Method, Research Series No.74, London: Depart- ment for Education and Employment. (1990), Adult off the Job Skills Training: an Evaluation Study, 438 Labor Market Refornzs in Korea: Policy Option for the Future Training Agency, R&D Series No.57, Sheffield, UK: UK Depart- ment of Employment. Pischke, J.-S. (1996), "Continuous training in Germany", Working Paper 5829, National Bureau of Econoinic Research: Cambridge, Mass. Ryan, P. (1994), "Training Quality and Trainee Exploitation", in R.Layard, K.Mayhew and G.Owen (eds), BritainS Paining Deficit, Aldershot, UK: Avebury. and C. Biichtemann (l996), "The School to Work Transition", in G. Schmid, J. O'Reilly, J. and K. Schoman (eds.), International Handbook of Labor Market Policy and Evaluation, Cheltenham: Edward Elgar. Schinid, G., Mosley H., Hilbert C. and H.Schiitz (1999), "Zur Effektivitat- aktiver Arbeitsinarktpolitik", Mitteilt~ngenLILISder Arbeitsmarkt- tind Berzgfsforschtmg,32 (4), pp.547-563. Skedinger, P. (1995), "Employment Policies and Displacement in the Youth Labor Market", Swedish Economic Policy Review, 2 (I), Spring, pp.135-171. Yoo, K.-S. (1999), The Employment Insurance System in Korea. Seoul: Korean Labor Institute. Chapter I4 Ahn, Chu-Yub (2000), "Recent Econoinic Crisis and Part-time Employment in Labor Market : An Analysis of Unemployment after IMF", in Dea-ill Kim, Chu-Yub Ahn, Jun-Mo Yang, and Kwan-ho Shin (eds.), Economic Crisis and Changes in Unemploymer7tStructure, Seoul: Korea Labor Institute (in Korean). Banton, Michael (1994), Discrimination, Buckingham: Open University. Bayes, Jane (1988), "Occupational Sex Segregation and Comparable Worth," in Rita Mac Kelly and Jane Bayes (eds.), Comparable Worth Pay, Equality and Public Policy, New York: Greenwood. Becker, Gary S. (1985), "Human Capital, Effort and the Sexual Division of Labor", .Journal of Labor Economics, 3 (I), pp.33-58. (1 957), The Economics of Discrimination, Chicago: Chicago -- University. Beechy, Veronica (1978). "Women and Production: a Critical Analysis of some Sociological Theories of Women's Work," in Annette Kuhn and AannMarie Wolpe (eds.), Feminism and Materialism: Women and Modes ofProduction, London: Routledge. Bradley, Harriet (1989), Men's Work and Women's Work, Minneapolis: University of Minnesota Press. Burchell, Brendan, Jane Elliott and Jill Burbery (1994). "Perceptions of the Labour Market: an Investigation of Differences by Gender and by Working-Time," in J. Burbery and F. Wilkinson (eds.), Employer Strategy and the Labor Market, Oxford: Oxford University. BIBLIOGRAPHY 439 Cho, Soon-Kyung (1999), Practice of Dismissing Female Workers and Policy Agenda, Seoul: Presidential Special Commission for Wornen, Korea (in Koran). Chung, Hye-Sun (1996), "In-Pant Motherhood Protection and Trade Unions", The Journal of Korean Woman S~udies,12 (1) (in Korean). Collinson, David L., David Knights and Margaret Collinson (1990), Managing to Discriminate, London: Routledge. Crompton, Rosemary and Kay Sanderson (1990), Gendered Jobs and Social Change, London: Unwin Hyman. Dunn, Danna (1996), "Gender-Segregated Occupations," in Paula J. Dubeck and Kathryn Borman (eds.), Women and Work, New Jersey: Rutgers University Press. Figart, Deborah and Ellen Mutari (1999), "Global Feminization and Flexibie Labor Markets: Gender Discourse in the Opposition to Pay Equity Reform," in Jeanne Gregory and Rosemary Sales, and Arine Hegewisch (eds.), Women, Work, and Ineqzlalily: The Challenge of a Deregulated Labor Market, New York: St. Marin's Press. Furchgott-Roth, Diana and Christine Stolba (1999), Women Figzlres, Washington, D.C.: The AEI Press and Arlington, Virginia: lndependent Women's Forum Hakim, Catherine (1995), "Five Feminist Myths about Women's Employ- ment", The British ,rournal ofSociology, 46 (3). Hanson, Susan and Geraldine Pratt (1995), Gender; Work a ~ dSpace, London: Routledge. Hartmann, Heidi I. (1979), "Capitalism, Patriarchy and Job Segregation by Sex", in Sillah R. Eisenstein (ed.), Capitalist Patriarchy, New York: Monthly Review. Holtzman, Mellisa and Jennifer Glass (1999), "Explaining Changes in Mothers' Job Satisfaction Following Childbirth," Work und Occupations, 26 (3). Joshi, Heather and Pierella Paci (1998), Unequal Pay for Women and Men, Massachusetts: MIT Press. Kim, Dae-I1 (2000), "Recent Economic Crisis and Changes in Unemploy- ment Spells," in Dea-ill Kim, Chu-Yub Ahn, Jun-Mo Yang, and Kwan-ho Shin (eds.), Economic Crisis and Changes in Unemployment Structure, Seoul: Korea Labor Institute (in Korean). Kim, El-Lim, Yun-Hee Choi and Young-A Chung (1994), Comparison between UN, ILO and Labor Law of South Korea Concerning Women,Seoul: Korean Women Development Institute (in Korean). Kim, Tae-Hong and Yu-Kyung Moon (1999), "The Structure of Female Unemployment and Suggestions for Une~nploymentPolicies", Women's Studies, 57, Winter, Seoul: Korean Women Develop- ment Institute (in Korean). Lee, Eddy (1998), The Asian Financial Crisis: Tlze Challenge for Social 440 Labor Market Reforms in Korea: Policy Optionfor the Futzire Policy, Geneva: ILO. Lee, Eun Young (1996), "Women's Legal Status at Work," in Hyung Cho (ed.), Cendec Equality and the Korean Legal System, Seoul: Ewha University Press (in Korean). Lee, Kye Woo (1999), "The Asian Financial Crisis and APEC", The Journal ofAPECStudies, 1 (1). Mincer, Jacob (1979), "The Distribution of Labor Incomes: A Survey", Journal ofEconomic Literature, 7 (1). and Solomon Polachek (1974), "Family Investments in Human Capital: Earnings of Women", .Journal of Political Economy 82, March-April. Ministry of Labor, Korea (1998), 1997 Survey Report on Wage Structure. (1999), 1998 Survey Report on Wage Structure. National Statistics Ofice, Korea (1999a), 1998 Annual Report on the Economically Active Population Survey. -- (1 999b), Monthly Economic Activity Survey. (1998), 1997 Annual Report on the Economically Active Popzllation Survey. Perlman, Richard and Maureen Pike (1994), Sex Discrimination in the Labour Markel: the Case for Comparable Worth, Manchester: Manchester University Press. Ridgeway, Cecilia (1997), "Interaction and the Conservation of Gender Inequality: Considering Employment", American Sociological Review, 62, pp.218-235. The Women's News (1999), December 1, Seoul. The World Bank (1999), "Research Proposal for A Study of the Korean Labor Market and Protection of Workers", (mimeo), Washington D.C. Walby, Sylvia (1990), Theorizing Patriarchy, Oxford: Blackwell. Chapter 15 Arya, G. and M. Mahmood (2000, forthcoming), "An appraisal of the Labor Market and Policy in the Context of the Macroeconomy - Growth, Crisis, and Competitiveness in Thailand", in G. Betcherman and R. Islam (eds.), East Asian Labor Markets and the Economic Crisis: Impacts, Responses, and Lessons, Washington D.C.: A Joint World Bank-ILO Publication. Betcherman, G., Dar A., Luinstra A. and M. Ogawa (2000, forthcoming), "Active Labor Market Programs: Policy Issues for East Asia", in G. Betcherrnan and R. Islam (eds.), East Asian Labor Markets and the Economic Crisis: Impacts, Responses, and Lessons Washington D.C.: A Joint World Bank-1LO Publication. Corson, W., Long S. and R. Maynard (1985), "An Impact Evaluation of the Buffalo Dislocated Worker Demonstration Program", Working BIBLIOGRAPHY 441 Paper, Mathematica Policy Research. Dar, A. and Z. Tzannatos (1999), "Active Labor Market Programs: A Review of the Evidence from Evaluations", Social Protection Discussion Paper 9901, Washington D.C.: The World Bank. Fay, R. (1996), "Enhancing the Effectiveness of Active Labour Market Policies: Evidence From Program Evaluations in OECD Countries", OECD Labor Market and Social Policy Occasional Paper No.18. Fretwell, D., Benus J. and C. J. O'Leary (1998), "Evaluating the Impact of Active Labor Programs: Results of Cross Country Studies in Europe and Central Asia", Social Protection Discussion Paper 9915, Washington D.C.: The World Bank. Fretwell, D. and S. Goldberg (1994), "Developing Effective Employment Services", in Equity, Efficiency and Adjustment in Labor Markets, Discussion Paper No.208, Washington D.C.: World Bank. Friedlander, D., Greenberg D. and P. Robins (I997), "Evaluating Govern- ment Training Programs for the Economically Disadvantaged", Jozrrnal of Economic Literature, 35 (4). Gill, I. and C. Ihrn (2000), "Korea" in I. Gill, F. Fluitman and A. Dar (eds.), Vocational Education and Training Reform: Matching Skills to markets and Budgets, Washington D.C.: Oxford University Press: A Joint World Bank-ILO Publication. Horton and Mazumdar (2000), "Vulnerable Groups and the 1,abor Market: The Aftermath of the Asian Financial Crisis", in G. Betcherman and R. Islam (eds.), East Asian Labor Markets and the Economic Crisis: Impacts, Responses, and Lessons, Washington D.C.: A Joint World Bank-ILO Publication. ILO (Various Years), Yearbook of Labor Statistics, Geneva. (1999), Key Indicators ofLabor Markets, Geneva. Kakwani, N. (1998), Impact of Economic Crisis on Employment, Unemploy- ment and Real Income, Bangkok: Thailand National Economic and Social Development Board. Kang, S., J. Keum, D. Kim and D. Shin (2000, forthcoming), "The Economic Crisis and Labor Market Policies in Korea: Labor Market Outcomes and Policy Issues", in G. Betcherman and R. Islam (eds.), East Asian Labor Markets and the Economic Crisis: Impacts, Responses, and Lessons, Washington D.C.: A Joint World Bank-1LO Publication. Kim, T. (1999), "The Strategy of Vocational Education Reform: The Korean Experience", Paper prepared for the 2"* International Congress UNESCO on Technical and Vocational Education. Kim, D. et al. (1999), "Employment Maintenance Subsidies and Their Employment Effects", Unpublished Paper, Seoul: Korean Labor Institute. Korea Labor Institute (1999), An Evaluation of the EIS, Seoul. 442 Labor Market Reforms in Korea: Policy Option for tile Futcrre Leigh, D. (1995), Assisting Workers Displaced by Structural Change: An fnternational Perspective, W.E. Upjohn Institute for Employment Research. Mansor, N., Chye T., Boehanoeddin A., Said F. and S. Said (2000, forthcoming), "Economic Crisis and Employment Policies in Malaysia", in G. Betcherman and R. Islam (eds.), East Asian Labor Markets and the Economic Crisis: Impacts, Responses, and Lessons, Washington D.C.: A Joint World Bank-ILO Publication. Martin, J. (2000), What Works Among Active Labor Market Policies: Evidence From OECD Countries ' Experiences, OECD Economic Studies, No.30. NERA (1995), "OECD Wage Subsidy Evaluations: Lessons for Workstart, Report for the Department of Education and Employment by NERA", Working Paper. OECD (1998), Employment Outlook, Paris. (1997), Employment Outlook, Paris. (1993), Employment Outlook, Paris ( 1 992), Employment Outlook, Paris. (1991), Employment Outlook, Paris O'Leary, Christopher (1998a), "Evaluating the Effectiveness of Active Labor Programs in Hungary", W.E. Upjohn Institute for Employ- ment Research Draft Working Paper. (1998b), "Evaluating the Effectiveness of Active Labor Programs in Poland", W.E. Upjohn Institute for Employment Research Draft Working Paper. Republic of Korea (2000), Unemployment Countermeustzres, Seoul. Skedinger, P. (1995), "Employment Policies and Displacement in the Youth Labor Market",Swedish Economic Policy Review, Vo1.2(1). Webster, E. (1998), "Microeconomic Evaluations of Australian Labor Market Programs", The Australian Economic Review, Vo1.3 l(2). Wilson, S. and A.V. Adams (1994), "Self-Employment for the Unemployed: Experience in OECD and Transitional Economies", World Bank Discussion Paper No.263, Washington, D.C.: World Bank. Yoo, S. (1999), "Employment Insurance and Social Safety Nets for the Unemployed", in Labor Relations in Korea (eds.), Seoul: Korean International Labor Foundation. Chapter 16 ILO (1998), 30Yh Report of the Committee on Freedom of Association, Geneva. .- (1996), Freedom of Association. Digest of Decisions and Principles of the Freedom of Association Committee o f the Governing Body of the /LO, 41hedition, Geneva. Kang, S-H. (1999), "Vocational Education and Training in Korea", paper prepared for the East Asian Labour Market Project, Seoul: Korea Labour Institute, July (mimeo). Korea Labour Institute (1999), "An Evaluation of the E~nployment Insurance System (EIS)", Seoul, October (in Korean). Martin, J.P. (2000), "What Works among Active Labour Market Policies: Evidence from OECD Countries' Experiences", OECD Economic Stzldzes, 30, pp.79-114. OECD (2000a), Ptwhing Ahead with Reform in Korea: Labour Market and Social Safely-net Policies, Paris. (2000b), Educat~onat a Glance: OECD Indicators, Paris. (1999a), Employment Outlook, Paris. (1999b), The Battle Against Exclusion, Volume 3, Social Assistance in Canada and Switzerland, Paris. (1998a), The Battle Against Exclzrsion, Volume 2, Social Assistance in Belgium, the Czech Republic, the Netherlands and iVorway, Paris. -- (1998b),MucationataGlance:OECDIndicators,Paris. (199 I), Employment Ozrtlook,Paris. PILAT, D. (1993),"The Economics of Catch Up: The Experienc,e of Japan and Korea", Groningen Growth and Development Centre, Monograph Series, 2. Chapter 17 Amsden, A. H. (1 989), Asia's Next Giant: South Korea and Late fndt~striali- zation, Oxford: Oxford University Press. Anker, N.(1998), Eleverne, Virksomhederne og de Nyudlarte, Social- f0rskningsinstitutetPh.D. serie. Bach, Henning Bjerregkd (1998), "Langtidsledige 1 tre kommuner", in Socialforskning - Nyt fra Socialforskningsinstitutet,No.2, July, Socialforskningsinslitutet,Copenhagen, Denmark. Barrientos, A. (1 998), Pension RtIform in Latin America,Aldershot: Ashgate. and P. Lloyd-Sherlock (forthcoming), "Health Insurance Reforms in Latin America: Cream Skimming, Equity and Cost-Containment", in Haagh, L. and C. Helga, Social Pollcy Reform and Market Governance in Latin America, Forthcorning with St. Antony's- Macmillan. Barro, R.J. (1991), "Economic Growth in a Cross-section of Countries", NBER Working Paper, No.3120, Cambridge, Mass: National Bureau of Economic Research, USA. Blank. R.M. and Freeman, R.B. (1993), "Evaluating the Connection between Social protection and Economic Flexibility", Working Paper Series, No.4338, Cambridge, Mass: National Bureau of Economic Research, USA. Boyer, R. (1988), The Search for Labour Market Flexibilily - The European 444 Labor Market Reformsin Korea: PolicyOptron for the Futzlture Economies in Transition, in Boyer, R. (ed.), Oxford: Clarendon Press. Burchill, Scott (1998), "Human Nature, Freedom and Political Community: An Interview with Naom Chomsky", in Citizenship Sttrdies, 2 (1): February. Drkze, J., and Sen (1989), Hunger and Public Action, Oxford: Clarendon Press. Due, J., Madsen. J.S. and M. Mailand (1997), "Arbejdsmarkedets parter. mtzltiniveauregularing og [ransitionelle nrbejdsmarkeder", in ArbejdsmarkedspolitiskArbog, pp.120-144, Copenhagen: Denmark. Eggertson, T., Economic Behcrviour and Institutions,Cambridge: Cambridge Surveys of Economic Literature, 1990. Finegold, D. and D. Soskice (1988), "The Failure of Training in Britain: Analysis and Prescription", Oxford Review of Economic Policy, 4, pp.21-54. Garon, Sheldon (1997), Molding Japanese Minds - The State in Everyday Life, Princeton, New Jersey: Princeton University Press. Gospel, H. (1998), "The Revival of Apprenticeship Training in Britain?", British Journal of industrial Relations,pp.429-457. -- (1 994), "Whatever happened to Apprenticeship Training? A British, American, Australian Comparison", in British Journal of Industrial Relations,32, pp.505-522. (1992), Markets, Firms and the Management of Modern Labour in Britain,Cambridge: Cambridge University Press. Haagh, L (200 I), Citizenship, Labour Markets and Democratization- Chile and the Modern Sequence, Macmillan-St: Antony's. (1999), "Training Policy and the Property Rights of Labour in Chile: Social Citizenship in the Atomised Market Regime", Journal ofLatin American Studies, 3 1, Part II, May. (1998), "Re-Democratisation, Labour Relations and the Develop- ment of Human Resources in Chile", Oxford University D.Phil thesis. (forthcoming), "Human Resources and De-Centralisation in Chile", in Social Policy Reform and Market Governance in Latin America, Edited book with Camilla Helgoe (University of Cambridge). St. Antony's -Macmillan. (forthcoming), "Perspectives on Social Policy and Market Governance in Latin America", in Social Policy Reform and Markel Governance in Latin America, Edited book with Camilla Helgoe (University of Cambridge), St. Antony's -Macmillan. Hayek, F.A. (1980), 1980s Unemployment and the Unions - The Distortion of Relative Prices by Monopoly in the Labotrr Market, Hobart Papers, No.87, London: The Institute of Economic Affairs. Holt, H. (1997), "Transitionelle arbejdsinarkeder mellem arbejdsliv og BIBLIOGRAPHY 445 familieliv" in Arbejdsmarkedspolitisk .drbog, pp.145-159, Copen- hagen: Denmark. ILO (1999), "Denmark: Flexibility, Security and Labour Market Success ", Per Kongshaj-Madsen, ILO Employment and Training Papers, No.53, Switzerland: Geneva. Jeong, J. (1995), "The Failure of Recent State Vocational Training Policies in Korea from a Comparative Perspective", British Journal of Industrial Relations,33 (2) June, pp.237-252. (2000), "Skill formation of Engineers in Large Korean Firms: and Analysis from a Comparative Perspective', in International Journal of Training and Development, 4 (I), pp.66-78. Kongshnrj Madsen, P. (1997), "Transitionalle arbejdsmarkeder - et generelt perspektive pb det rummelige arbejdsmarked" in Arbejdsmarked- spolitisk Arbog, pp.94-101 , Copenhagen, Denmark. Korea Labor lnstitute (Various Issues), Trends in Employment Insurance (in Korean), Seoul. Korean Metal Workers' federation (1999), Koyongkucho Py6nhwawa Nodongchohap-ui KoyongchGng Chaek, Seoul, Korea. Krueger, A. 0. (1 983), Trade And Employment in DevelopingCountries, Vol. 3: Synthesis and Conclusions, Chicago: Chicago University Press. Lagos, R.A.(1994), "QuC se Entiende por Flexibilidad del Mercado de 'Trabajo ?Yn Revista de la CEPAL,No.54. 1,andsarbejdsrbdet (1997), "LandsarbejdsrBdets arbejdsmarkedspolitiske redeggrelse",Denmark: Copenhagen. Landsorganisationen (LO) (1998a), "Plads ti1 alle pa et dynamisk arbejdsmarked ",21 August, 1998. (LO) ( 1998b), 'Korttidsledigheden: Kan der skabes "gratis strukturforbedringer af arbejdrmarkedet?", 25 August, Denmark: Copenhagen. (LO) (1998c), ''LOIS mds~tninger og forslag til en ny arbejdsmarkedsreform", 14,September. Lee, Joohee (2000),"Income Assistance and Employment Creation Through Public Works in Korea", paper presented at the International Conference on Economic Crisis and Labor Market Reform: The Case of Korea, 18-20. May, sponsored by the World Bank, the Korea Labor Institute and KED. Ministry of Labour, Denmark (1997),Arbejdsmarkedspolitisk ,&bog. Ministry of Labor, Korea (Various Years), Yearbook of Labor Statistics. OECD (1992), Technolo~yand the Economy- The Key Relationships, Paris. (1986), Flexibility in the Labour Market - The Current Debate - A Technical Report, Paris. Park, Funkoo and Lee Joohee (1999), "The Social Impact of the Financial Crisis: Labor Market Outcomes and Policy Responses in Korea", Unpublished Paper, 'The Korea Labour Institute, August, Korea: SeouI. 446 Labor Market Refornzs in Korea: PolicyOption for the Futlrre Park, Young-Bum (1994), "State Regulation, the Labour Market and Economic Development: The Republic of Korea", in G. Rodgers (ed), Workers' Instit~rtions and Economic Growth in Asia, Switzerland: Geneva. Piore, M, J. (1986), "Perspectives on Labor Market Flexibility", in Industrial Relations, 25 (2), Spring, pp.146-166. Romer (I 986), P.M. "Increasing Returns and Long-run Growth, Journal of Political Ecnomy, 94 (5), pp.1002-1037. Rosdahl, Anders (I998), "Merbeskzftigelse erller Fortraengning?" in Socia[forskning - Nyt fra Socialforskningsinstitz~tet, No.2 June, Denmark: Copenhagen. Ryan, P. (1995) "Trade Union Policies towards the Youth Training Scheme: Patterns and Causes", British Journal of Industrial Relations, 33 (I)?March, pp.1-35, England: London. (1994), "Training Quality and Trainee Exploitation", in Layard, R., Mayhew, K., and Owen, G., (eds.) Britain's Training De$cit, Aldershot: Avebury. Sen, A. (1999), Development as Freedom,Oxford University Press. Smith, N. (1997), "Det effektive, rummelige og trygge danske arbejdsmarked?" in Arbejdsmarkedspolitisk Arbog, Denmark: Copenhagen. Torfing, J. (1999), "Workfare, Danish-style", Centre for the Study of Democracy Bulletin, 6 (2),pp.11-13, United Kingdom. Williamson, 0.(1985), Tlze Economic Institzrtions of Capitalism, New York: The Free Press. Williamson, O., Wachter, M. and J. Harris (1986), "Understanding the Employment relation: The Analysis of Idiosyncratic Exchange", in L. Putterman (ed.), The Economic Nature of the Firm, Cambridge: Cambridge [Jniversity Press. You, J-I (1997), "Globalization, Labor Market Flexibility and the Korean Labor Reform", in Seoul .Journal of Economics, 10 (4), pp.341- 372. (1994), "Labour Institutions and Economic Development in the Republic of Korea", in G. Rodgers (ed.), Institutions and Economic Growth in Asia, Geneva: International Institute for Labour Studies, and H-J, Chang (1993), "The Myth of Free Labour Market in Korea", in Contributions to Political Economy, 12.