Report No. 24408-BR Brazil Jobs Report (In Two Volumes) Volume I: Policy Briefing December 20, 2002 A Joint Report by The World Bank Instituto de Pesquisa With the Support of Brazil Country Management Unit Econ6mica Aplicada the Government of Brazil Latin America and the Caribbean Region Ministry of Labor and Employment Document of the World Bank CURRENCY AND EXCHANGE RATES (R$S$) Currency Unft - Real (R$) December 2000: R$1.95 December 2001: R$2.31 November 2002: R$3.53 WEIGHTS AND DiEASURES The Metric System is used throughout the report. FISCAL YEAR January 1 to December 31 Vice President LCR David de Ferranti Director LCC5C Vinod Thomas Director LCSHD Ana-Maria Arriagada Lead Economist Joachim von Amsberg Task Manager Indermit S. Gill ACKNOWLEDGEMENTS This report is the result of collaboration between Brazil's Institute for Applied Economic Research (IPEA) and the World Bank between 2000 and 2002. The task was jointly managed by Indermit Gill (Lead Economist, World Bank) and Ricardo Paes de Barros (Director of Social Studies, IPEA-Rio de Janeiro), with the support of Andreas Blom (Economist, LCSHD) and Jose Marcio Camargo of the Pontifical Catholic University of Rio de Janeiro. The report was prepared under the overall guidance of Roberto Martins (President, IPEA), Gobind Nankani and Vinod Thomas (Country Directors during 2000-2002), and Xavier Coll and Ana-Maria Arriagada (Sector Directors during 2000-2002), and Suman Bery and Joachim von Amsberg (Lead Economists during 2000-2002). Volume I consists of the Policy Report, prepared by Indermit S. Gill, Andreas Blom, Ricardo Paes de Barros, Carlos Henrique Corseuil (Economist, IPEA) and Mirela Silva (Economist, IPEA). Volume II, prepared by Andreas Blom and Indermit Gill, contains the background papers commissioned for the report, that diagnose labor market developments in Brazil, present relevant intemational experience, and discuss the policy implications. The authors of the papers are Jose Marcio Camargo (PUC-Rio de Janeiro), Francisco Galrao Carneiro (Universidade Catolica de Brasilia), Jorge Saba Arbache (Universidade de Brasilia), Ricardo Paes de Barros, William Maloney (Lead Economist, LCRCE), Wendy Cunningham (Economist, LCSPR), Norbert Fiess (Economist, LCRCE), Marco Fugazza (Consultant, CERAS, Ecole Nationale des Ponts et Chausses, Paris), Dorte Domeland-Narvaez (Consultant, Universidad Pompeu-Fabra, Barcelona), Amit Dar (Senior Economist, HDNSP) and Andreas Blom. Edward Amadeo (former Minister of Labor, Brazil), Marcelo Neri (Fundac,o Getdlio Vargas-Rio de Janeiro), Rosane Mendon,a, and Lauro Ramos (IPEA-Rio de Janeiro) provided valuable feedback and guidance in Brazil, and Marito Garcia (Education Sector Manager, LCSHD), Gillette Hall (Social Protection Economist, LCSHD), Robert Holzmann (Director, HDNSP), Truman Packard (Social Protection Economist, LCSHD), Mark Thomas (Senior Economist, LCC5C), Dorte Vemer (Senior Economist, LCC5C) and Joachim von Amsberg (Lead Economist, LCC5C) at the World Bank. The participants of a two-day workshop in Washington on January 9-10, 2002 to discuss the main findings and policy messages influenced the form and contents of both volumes. The team expresses its appreciation to Gabriela Falconi (LCSHD) and Andreas Blom for organizing the workshop and facilitating IPEA's participation, and to Truman Packard for moderating the discussions effectively. Participants in two workshops organized by HDNSP also provided useful feedback; the critical commentary of Paul Lundall (Deputy Director, Development Policy Research Unit, University of Cape Town) is especially appreciated. The peer reviewer for this task is Gordon Betcherman (Senior Economist, HDNSP), who provided detailed comments on both the concept paper and an earlier draft, and to whom the team owes a debt of gratitude. Finally, we wish to acknowledge the continued collaboration of the Government of Brazil, especially the support of the Minister of Labor, Mr. Paulo Jobim Filho, Mr. Francisco Domelles, the former Minister of Labor, and Mr. Roberto Martins, President, IPEA, for this task. TABLE OF CONTENTS FOREWORD AND SUMMARY .....................I The Debate in Brazil, and How This Report is Organized ...............................................1 ..................... ..I Labor Markets as an Important Link ........................ 2 Principal Features of Brazil's Labor Market ...................................3............................................. 3 A Road Map to the Report .................... . . 4 This Report's Main Messages ..................................................5....... ....................... 5 1. Tirr; ROI,E OF GOVFRNMENT IN TIIE LABOR MARKET- .....................7 Institutions Must Evolve ........................ 7 Regulations Should Promote Employment and Productivity. ....................... 8 Interventions Are Needed, Even in Efficient Labor Markets ....................... 9 2. THE NEED FOR LABOR MARKET REFORM IN BRAZII ....................................... II The Outcomes of Interest ............................................. 11 Insufficient Job Creation ................ .......................................1........ .. .............................. ... ........ ...... ... 12 Declining Participation in Work and Rising Unemployment .................................................. 13 Low Productivity Levels .................................................. 14 Inadequate Income Security ................................................... 15 3. MACROIECONOMICS ANI) LABOR DEMAND ............................................ 16 Macroeconomic Fluctuations Affects Labor Demand ................................................... 16 Stabilization and the Labor Market ................. ............ ................ ... : 17 Trade Opening and the Labor Market .................................................. 18 Macroeconomic Adjustment and Labor Policy Go In Tandem .................................................. 19 4. QUANTITIY AND QUALITY OF LABOR SKUPPI,Y ............................................ 21 A Diversified Labor Supply Requires a Spacious Labor Market .................................................. 21 Strong Labor Market Signals: Quality Matters in Labor Supply .............................. .................... 22 A Closing Labor Quality "Window of Opportunity .................................................. 23 5. How BRAZIL'S LABOR MARKET FUNCTIONS ............................................ 25 Salient Features of the Brazilian Labor Market. .......................... ........................ 25 Is Labor Turnover in Brazil Too High? .................................................. 26 Is Labor Litigation in Brazil Extraordinarily Frequent? .................................................. 27 Is Informality in Brazil Unusually Elevated? .................................................. 28 6. PRICING; LABOR CORRECrLY ............................................ 29 Mandated Minimum Wage and Non-Wage Benefits ...2............... ............................................9...............29 Mandates Price Low Wage Workers out of Formality .................................................. 30 Allowing Unions to Negotiate the Mandated Minimum May Help .................................................. 32 Reducing the Mandated Minimum Package May be Necessary ............. ..................................... 34 7. ALIGNING INCENTIVES BETTER ............................................ 36 Mandated Severance Benefits .................................................. 36 Severance Laws and Labor Court Functioning Distort Behavior .................. ................................ 37 Encouraging In-contract Negotiations May Reduce Litigation .................................................. 39 Redesigning FGTS Will Reduce Perverse Incentives .................................................. 40 8. IMPROVING INCOME SECURIIY ............................................ 43 Unemployment or Poverty? What We Should Care More About .................................................. 43 Improving Effectiveness of Active Labor Programs May Help .................................................. 44 Better Targeting of Income Support Will Help Poorer Workers .................................................. 47 9. SUMMARY ANID POIlACY PRIORITII,S ............................................ 49 What's New About this Report .................................................. 49 Summary of Main Findings .....0............................................ 50 Eliminate Subsidies for Labor Turnover .................................................. 51 Move Negotiations from Labor Courts to the Workplace ........... ................................. ...... 52 Lower the "Minimum" Mandate Down From the Middle .................................................. 52 Widen the Safety Net for Workers .................................................. 53 Policy Recommendations, Effects on Functioning, and Impact on Outcomes ..... .............................................. 54 SELECTED REFERENCES ................................................... 55 LIST OF FIGURES Figure 1 Framework of Analysis and Structure of Volume I Figure 2 Litigation, Informality and Turnover Have Increased Since the 1980s Figure 3 Brazilian Firms View Labor Regulations and Taxes as Especially Onerous Figure 4 The Brazilian State Intervenes Frequently Figure 5 Contribution to Growth of Brazil's Per Capita Income, 1993-1999 Figure 6 The Working Age Population Has Risen Faster Than Employment Figure 7 Had Participation Not Fallen, Unemployment Rates Would be Twice Today's Levels Figure 8 Brazil Has a 50 Percent Shortfall in Labor Productivity Figure 9 Brazilian Workers Cannot Count on Future Earnings Figure 10 At the Aggregate Level, Brazil's Labor Market Has Adjusted Well to Fluctuations Figure 11 Falling Inflation Implies More Adjustment Through Employment Figure 12 Trade Openness and Labor Market Flexibility are Complements Figure 13 Real Exchange Rates and Labor Market Segmentation in Brazil Figure 14a Women Increasingly Enter the Labor Force. Figure 14b Women Hold Different Job Positions Than Men Figure 15 Higher Education, Higher Wages Figure 16 The Rise in Wage Premiums to Educated Workers Has Been Muted in Brazil Figure 17 More Than Half of All Jobs are Now in the Service Sector Figure 18 Job Tenure is Lower in Brazil Than in the US Figure 19 A Busy Labor Justice System Figure 20 Informal Wage and Salaried Workers Outnumber the Selfemployed in Brazil Figure 21 Almost Half of Infomal Workers Earn Less Than the Minimum Wage Figure 22 Minimum Wages are More Binding in the Poorer Northeast Figure 23 Union Members are More Educated Than Non-Unionized Labor in Brazil Figure 24 Why Workers Litigate: Reasons for Taking Employers to Court Figure 25 Poorer Workers Take Their Employers to Court More Often Figure 26 Poorer Workers Stay Unemployed Longer, and Increasingly More Figure 27 Informal Workers are Most Likely to Become Unemployed, and Increasingly So Figure 28 Access to Income Support Programs LIST OF TABLES AND TEXT ]BOXES Table 1 Contents of Volume H-Brazil Diagnostics and International Experience Table 2 Labor and Other Constraints Faced by Firms in Brazil Table 3 Labor Policy Priorities in Different Macroeconomic Scenarios Table 4 Mandates Raise the Cost of Hiring Workers by 65 Percent Table 5 Severance Benefits Can Add Up to a Year's Wages Table 6 Income Support Programs for the Unemployed-Summary and Policy Implications Table 7 International Experience with Active Labor Market Programs Table 8 Policy Recommendations, Effects on Functioning, and Impact on Outcomes Box 1 Economic Growth and Labor Reform in Argentina Box 2 Productivity-an Important but Difficult Concept Box 3 The Themes Underlying this Report Box 4 Institutional Reform-Beware of "Best Practice" Box 5 The CLT-Patemalistic Principles Remain Intact from the 1930s Box 6 Earnings Inequality, Education Inequality, and Cohort Size Box 7 There are Always Negotiations: How Labor Courts Function in Brazil Box 8 Supreme Court Decision Forces Increase in FGTS Contribution and Fine Box 9 Objectives and Effects of Active Labor Programs-The Theory Box 10 Effects of Active Labor Programs in Brazil-the Evidence The Brazil Jobs Report - 1 - Foreword and Summary _03= The Debate in Brazil, and How This Report is Organized 1. Brazil's labor market has intrigued analysts and policymakers for decades. It has been characterized both as flexible and efficient and-at the other extreme- as segmented and inequitable, and almost everything in between. Those who say Brazil's labor market works well point to its relatively low unemployment rate, its generally pragmatic labor unions, its low minimum wage and a high rate of job turnover. For those of this view, labor reform is not a priority. Those who _ i say it does not work well point to the high rate of informality, the frequency of Country Millions litigation in labor courts, high inequality of earnings, and the limited reach of Brazil 170.3 labor laws and government interventions to assist workers who need help. For Araentina 37.0 such people, labor reforms are important. Chile 15.2 2. This report, conducted jointly by researchers in Brazil and at the World Colombia 42.3 Bank, aims to address this debate on how the Brazilian labor mnarket functions. It Mexico 99.0 does so not by focusing on labor market functioning but on its outcomes. After Portugal- 10.0 all, the really important questions are not whether Brazil's labor market is overly Italy 57.5 informal, or excessively litigious, or hyperactive in that people change jobs much Germany 82.0 too frequently. What is central are labor market outcomes, such as adequate Netherlands 15.9 employment growth so that job-seekers can find gainful employment, acceptable UK . 59.2 worker productivity levels that are fairly compensated, and reasonable income USA 275.5 security for workers and their households. Korea 47.2 3. This shift in focus from functioning to outcomes does, however, complicate Thailand 62.1 our work. While one can study labor market functioning without paying too *or latest. much attention to non-labor factors, such "partial equilibrium analysis" is impossible when the subject of study is outcomes. Employment growth depends on macroeconomic factors at least as much as labor laws; worker productivity levels depend as much or more on investments in physical and organizational capital by firms and on investments in human capital through education and training as they do on labor market institutions and interventions; and income security depends as much on factors such as trade patterns and social safety nets as it does on labor market regulations. But we believe that this added complexity is necessary for a fuller understanding of Brazil's labor market. To simplify matters somewhat, we focus here on the most relevant relationships, between labor market institutions, its functioning, and outcomes. The report is structured accordingly (see Figure 1). Figure 1: Framework of Analysis and Structure of Volume I Labor Market Outcomes I. .~ -- Chapter 2 I ...A Macroeconomics Labor Market Labor Force (Demand) Functioning Quality and Quantity Chapter 3 Chapter 5 (Supply) Chapter 4 Labor Market Institutions Regulations Interventions Chapters 6, 7 and 8 The Brazil Jobs Report - 2 - Foreword and Summary Labor Markets as an Important Link 4. Because labor market issues are so controversial, it is easy to forget that the demand for labor is derived entirely from the demand for commodities and services. The latter depends a lot more on macroeconomic factors that spur economic growth than on labor markets. Again, while it is important that the labor market provides accurate signals for investment in skills, the functioning of education and training systems may be far more important for the supply of skilled labor and the quality of jobs. Labor markets, serve only as a link between the factors that affect demnand for labor and those that affect its supply. Country Gini 5. But they are an important link. Without well-functioning labor markets, Brazil 59.1 economic growth may not translate into jobs, investments in education and new Argentina 53.0 technologies may not translate into higher productivity and pay of labor, and Chile 57.5 incomes may remain uncertain even as economies become more stable. In an Colombia 57.1 important sense, therefore, labor markets should be as passive a link as possible Mexico [ 51.9 to qualify as well-functioning, serving to facilitate rather than influence the Spain 32.5 redistribuation of resources. Portugal 35.6 Italy 27.3 6. For atl these reasons, it is important to keep expectations from labor market Germany 30.0 reform realistic. It is difficult to identify the extent to which labor market Netherlands 32.6 functioning is responsible for less than desirable outcomes. While this report UK 36.1 will identify the rationale for labor reform and help in prioritizing the efforts in USA 1 40.8 this difficult area of public policy, it could be argued that these outcomes can be Korea 1 31.6 attained through other policy instruments. But if there is any doubt that the Thailand 4 41.1 outcomes themselves are not satisfactory, consider two facts relating to Source: World Bank efficiency and equity, respectively. Productivity levels in Brazil are low, compared with what should be expected of countries at its per capita income levels (such Chile, Malaysia and Mexico). If Brazil were a particularly equitable country, one could argue that the price in terms of lower efficiency and productivity may be worth paying. In fact, inequality in Brazil is noteworthy for being among the highest in the world. Policy reforms that increase productivity or reduce inequality should be helpful in Brazil. Labor reforms-which can do both-should therefore be seriously considered. BaoKI: Economic growth antidlabor ref§rnl in Argentina ,Eiiloyment growth over.the last quarterc6ntury in Argentina has been uneven. Employment grew 1.3 percent annually between 1974. and, 1994, but at 3.4 percent between 1995 and 1999. The increase in employment growth ;sdc, 1995 was becauie of two'reasons-resumption of economic growth and reduction in labor costs. The latter incluied a general deqcreaspe in payroll'taxes, and,ihe introduction of a new type of contract under the modalidades Sprom?gvWdas, which allowedihiring pf w6rkers without paying labor taxes and with fewef penalties for firing, Half of allw,rkers hired in 19961998 were under these special contracts. By 1998, more than 12 percent of forral ,empl,ojees had been 4entr,acted under the hnodalidades promovidas. Rough calculations indicate that total employm, ent would hav& beeq 13 percent' less if these,refo,rms had not been introduced. Unemployment rates fell from alriost 20 perce'nt to less than, 12 percent over this period. 'Thie.initiative amply demonstrated.tthe positive' effects of-mneasures to make hiring and firing of workers less ,expensive; But the les'son learned by Mgriitini's government appeared to have been the opposite. In,1998, in the run-up,.to the elections, the modalidades prompvidas initiative was revoked, the probationary period 'for standard 'dontrics was halved frorn tweiveim'eonth's to six,' and the;sys#em of collective bargaining was further' centralized to plcacteA,rgentina's vocal labor tinions. By 2091, 'this retreat from labor reform combined with falling growth rates '-led, to., unemploymen't rates climbing back to 1995 levvels. With- the economic crisis caused by, the debt default, ' unemployment rates went up eren higher in 2002. 6 ' §.o4rceArgentina: Lbor Markets in the New;Millennium, World Bank, 2000. The Brazil Jobs Report - 3 - Foreword and Summary Principal Features of Brazil's Labor Market 7. In examining Brazil's labor market, it is worthwhile to keep some simple statistics and dcfinitions in mind. In 2000, Brazil had a labor force of about 80 million that-given that two-thirds of its population of 170 million is of working age-implies a labor force participation rate of about 70 percent. With an unemployment rate of about 8 percent, this means that employment is about 73 million. Of this, about 3 mnillion are employers, and self-employed workers number about 17 rnillion. The remaining 53 million is split evenly between formal and informal wage and salaried workers. 8. When referring to informal employment in Brazil, people usually add up the Per Capita Income, 17 million self-employed and the 27 million informal wage workers. In fact, 200O' these two groups have more differences than similarities. What they do have in Country US$ common, though, is that neither group pays social security dues. In the Brazil ___ 3,570 vernacular, they are sem carteira de trabalho assinada, or without a signed Argentina 7,440 workung card. In contrast, formal wage and salaried workers have working cards Chile 4,600 that have been signed by their employers, who thereby agree to abide by all labor Colombia 2,080 laws. These are the more fortunate or plucky com carteira assinada. The Mexco8 question of what to do about informality-often viewed as a growing problem Spain 14,960 (see Figure 2)-figures prominently in discussions about labor policy in Brazil. Portugal 11,060 9. Of the institutions that affect labor market functioning in Brazil, none is more Germany 25,050 influential than its system of labor courts, known as the justiVa do trabalho, a Netherlands 2,140 special section of the judicial branch of govermment that employs judges, lawyers UK 24,500 and others charged with resolving labor disputes between employers and USA 34,260 workers. By all measures, they are extraordinarily busy-in 2000, Brazil's labor Korea __ 8,910 courts entertained about 2 rnillion complaints, an order of magnitude larger than Thailand 2,010 in the US (17,000) and Japan (1,500), which have larger workforces. Labor Sorlatest. courts are believed to profoundly alter work relationships, and play a central part in the labor policy debate. 10. Among labor economists, few statistics are as widely cited in Brazil as those on job turnover. Fully one-third of the labor force changes jobs every year, a number that is presented as evidence of the dynamnism of Brazil's labor market, and the ease of firing and hiring under its labor laws. But there is evidence that labor legislation may actually be increasing turnover beyond what would occur if these laws were more neutral. One of the reasons for this hyperactivity is believed to be the design of individual severance accounts system mandated since 1968, the Fundo de Garantia por Tempo de ServiVo (henceforth FGTS). Figure 2: Litigation Informality and Turnover have Increased Since the 1980s 170 Lltigation 1 60 -Htotmaity ,; < C> 1500 Jot, Tuv E- C) 1 90 TI - _. i8 o - {- ----- -;-------- May-82 Macy-84 May-86 May-88 May-g0 May-92 May-94 May 96 May 98 Maby 00 Note Latgation E measured as nurber of court cases per 1000 workers Source. Supremo Tribunal Federal and PME fromn IBCE The Brazil Jobs Report - 4 - Foreword and Summary A Road Map to the Report 11. This report addresses these labor market issues using existing research and work commissioned for this task. Volume 11 contains a detailed examination of the issues of interest in Brazil and the relevant international experience (see Table 1). This volume summarizes the main findings and policy implications. 12. Just as technology and skills need to be updated as economic conditions change, so do labor market institutions. The regulations that govern employment relations and the institutional framework that facilitates and supports them have seen only a few changes over the sixty years since the Consolidacao das Leis do _. * *. - * * Trabalho (henceforth CLT, or the consolidated labor code) was adopted. Chapter 1 argues that labor laws-that were probably more suited to conditions Country_ l % in the l950s and 1960s-have begun to show signs of obsolescence. Chapter 2 Brazil 1 25.4 shows how this is reflected in deteriorating outcomes. Key indicators- Argentina l - employment growth, labor force participation, unemployment rates, and income Chile 18.4 security-all point to worsening labor market functioning since the mid-1990s. Colombia 28.7 Mexico 34.8 13. The report then examines how changed macroeconomic circumstances call .Spain- for changes in labor market institutions, regulations and interventions. In Portugal - particular, using a characterization of the economy in which informality has a Ital n - central role, Chapter 3 illustrates the correspondence between the three main Gerrnany l - macroeconomic phenomena of the 1990s-greater openness, stabilization, and Netherlands - fiscal adjustment-and Brazil's labor reform priorities. USKA _ 14. In examining the supply side, Chapter 4 concludes that the labor market has Korea 2.0 signaled the shortage of educated workers since the 1990s, and the onus is now Thaliand 28.2 on the education and training systems to respond. Because of rapidly declining Source: World Bank population growth, the current cohort aged 5-20 years will be the largest ever. *or latest. Brazil faces a window of opportunity in that it can ensure a supply of skilled workers for many years to come if this generation acquires a strong education base. There are signs that the education system is responding. 15. Analysis of how Brazil's labor market functions m Chapter 5 points to evidence that indicates that Brazil's poorer workers and smaller firms are especially disadvantaged by how the labor market functions. The report identifies three sets of priorities for reform: changes in mandated non-wage benefits and minimum wage setting to price labor correctly and encourage employment growth (Chapter 6), changes in severance legislation and functioning of labor courts to better align incentives and increase productivity (Chapter 7), and improvements in interventions to increase income security for all workers (Chapter 8). Chapter 9 summarizes and highlights the main policy implications. Bob i: ]Productivityan mipiootnat but difficult concept The two most common concept~s of productivity are labor and total factor productivity. Labor productivity is.most cn,mo, nly calculated by dividing value added,by. employment. This is the measure used in this report. Because value'addedfig,res are often not available outside of manufacturing, labor productivity statistics are, usallyavailable for a snalLpart of the workforce. Labor productivity can go up for many reasons-improvements inn ffinction.ing of the labor. m,arket,that improves the efficiency of labor use is just one of them. Others are higher human capital or capital per worker, icreased'use of natu`raes6urces, and&technical progress. T,otalfacttr' productivity is the productivity of all factors.(labor, capital and other inputs) combined, calculated by ..dividing,tput by an aggregate index of inputsclncieiies in total factor productivity-under some assumptions such as perfect competition-can be viewed as technological change. ,Productivity may go up in the event of "jobless growth" , i.e., if output increases but employment does not. We therefore use measures of employment growth alongside labor.productivity growth in assessing the effectiveness of labor reform. - l The Brazil Jobs Report - 5 - Foreword and Summary This Report's Main Messages 16. The next few paragraphs offer an even more abbreviated summary. Readers interested in just a summary of the findings and the principal policy recommendations can go straight to Chapter 9, which can be seen as an executive summary. For those interested in more detail, Volume II contains in-depth examination of the issues of interest in Brazil and the relevant international experience (see Table 1), on which Chapters 1 through 8 of this volume are based. 17. While Brazil's labor market displays macroeconomic flexibility-reflected e m - in moderate unemployment (although increasing), high turnover rates, and Index, 2001- adjustment through wages-it also reveals symptoms of microeconomic Country Index inefficiencies-low productivity and job hyperactivity. The interaction of Brazil 49.7 Brazil's labor institutions-especially the labor justice system and the network of Argentina 37.5 unions-with its regulations-especially for wage and non-wage benefits and for Chile 59.8 severance payments-combined with low education levels of its workforce Colombia 32.8 results in low labor productivity when compared with benchmark countries, and Mexico 43.7 sluggish formal sector employment growth. What makes this even worse is that Spain 60.1 labor market functioning is "regressive"-its influence on outcomes is Ital 48.4 particularly pernicious for the poor. Germany 74.0 18. As remedies, the report proposes more appropriate pricing of labor, reduced Netherlands 81.5 rnisalignment of incentives faced by workers and firms, and steps to improve UK 64.8 income security for workers without unduly distorting prices and incentives. The USA 100.0 priorities for reform are: Korea 51.1 Thailand 42.7 * Reforms to reduce the inadvertent subsidization of turnover, especially Source: IMD the reform of the Fundo do Garantfa por Tempo de Servi o. (FGTS). * Improved coverage of income security programs, especially the redirection of seguro desemprego (unemployment insurance) to reach informal wage and salaried workers. * Moving negotiations from labor courts to the workplace, first by reforming collective bargaining laws (especially unicidade sindical) to improve representation and then by reforming the justiga do trabalho (labor courts) * Lowering mandated benefits down from levels at which they distort the behavior of more than half of the workforce to levels more consistent with the quality of Brazil's workforce and the nature of its institutions. Box 3: The themes underlying'this report .T,her,eare three recurrent hemes in this repor. spanning ecooinic efficiency, socio,economic equity ,and.political 'feasibility.concems: - First, the effects of current design of labor legislation institutions have begun showing up cleariyJin worsening ,economy-wide indicators of employment growth,;labor force participation, unemployment and coverage of social insurance programs. Worsening,outcomes-during the1990s point to the increasing urgency of tabor reforms to improve econoomic efficiency I; Second; the design and enlorcemenh of labor legislation hurns workers and firms with low productivity the most, and the reforms that have-been proposed as compromises again help them the least Labor reforms are needed.to imnprove socioeconomic equality. - Third, Brazil need not reform,its labor market institutions, regulatieis 'and interventionis ll at once. The propo,sed changes fall into three differentaras- pricinglabor correctly, better.aligmng incentives and redirectinig social insurance programs. Even within these areas, the reform enda c,an be partitioned and undertakeni with rnore or less vigor, so that labor reforms meet the condition of poltucalfeasibility - The Brazil Jobs Report - 6 - Foreword and Summary Ch | Atuthor-' T > -lop Nature 1 Andreas Blom Road Map to Volume II Introduction Summary of findings Norbert Fiess, Marco Labor Markets and Macro-Labor 2 Fugazza, & William Economic Adjustment: The aBrazil dagnostics Maloney Brazilian Case irges 3 William Maloney & Re-examining the Informal . International Wendy Cunningham Sector experience 4 Jorge Saba Arbache Unions and the Labor Collective Brazil diagnostics Market in Brazil Bargaining level data Amit Dar & Francisco Economic Impact of Unions Collective International 5 Galrao Cameiro and Collective Bargaining: Bargaining experience International Experience . ~~~~~~~Brazil diagnostics 6 Jose Marcio Camargo Fake Contracts: Justice and Labor Courts based on Labor Contracts in Brazil adbonisurative data administrative data Francisco Galrao An Overview of the Effects Brazil diagnostics 7 Carneiro of the Minimum Wage on Minimum Wages based on household- the Brazilian Labor Market level data The Poverty Implications of Interational 8 Wendy Cunningham Minimum Wages in Minimum Wages experience Developing Countries Dorte Domeland & Unemployment and Income Support Brazil diagnostics Norbert Fiess Unemployment Insurance for Unemployed bed o us l level data | Ricardo Paes de Barros, The Efficiency of Active Active Labor Brazil diagnostics 10 g Mauricio Cossio, & Labor Market Policies for Market Policies based on household- _ Jorge Teles Poverty Reduction in Brazil M level data Active Labor Market Active Labor interational 11 Amit Dar Policies: Evaluative Market Policies experience Evidence Volune lI-is avaiable as a sepainblicatio - The Brazil Jobs Report - 7 - Labor Market Governance 19. The market for labor, like any other market, has institutions that govern its workings. The precise definition of "institutions" varies. World Bank (2001) defimes them as rules, enforcement mechanisms and organizations. In contrast, for Burki, Perry and others (1998), rules and enforcement mechanisms constitute institutions, in contrast with organizations-the latter pursue their interests within an institutionai structure. In this report, we distinguish between institutions (e.g., the labor courts), regulations (e.g., rules governing dismissal of workers) and interventions (e.g., unemployment insurance programs). For our purposes, this Unemployment Rates, seems to be the most sensible classification. 2000* Country % 20. The institutions that are of the most importance in Brazil are the federal Brazil 9.6 constitution, the system of labor courts, and the network of labor unions. The Argentina 12.8 regulations that concern us here are the mandated wage and non-wage benefits Chile 9.9 for individuals while employed, severance mandates including the FGTS, and the Colombia 20.1 rules governing collective bargaining. The interventions we examine in this Mexico 2.0 report are the unemployment insurance system, public job search assistance and Spain 14.1 training programs, and public work schemes. Portugal 3.8 Italy 10.8 Institutions Must Evolve Germany 8.1 21. Like the technologies used by firms and the skills possessed by individuals, Netherlands 3.6 the institutions of countries have a "vintage". And just as changed circumstances UK 5.3 make technologies and skills obsolete, changes in economic conditions can make USA 4.1 institutions outdated. Brazil has changed from a high growth, public-sector Korea __ 4.1 dominated, closed, high inflation economy in the 1960s and 1970s to a low Thailandol growth, private sector led, relatively open, low inflation country since the mid- Sor latest. 1990s. It is hard to imagine a more changed set of circumstances. The rules, regulations and organizations that possibly served Brazil's labor market well in the 1960s and 1970s may not have helped much in the 1980s, and may have even hindered progress since the 1990s. Brazil needs to keep what is working, and change what is not. While the constitution has been changed several times, it would be no exaggeration to say that institutions governing litigation, arbitration and collective bargaining have not changed since their inception in the 1940s (see Boxes 4 and 5). Box 4: Institutional Reform: Beware of "Best Practice" Burki, Perry and others (1998) provide some -guidelines for promoting institutional reform: paying systematic attention to the nature of prospective, winners and losers, crafting compensation schemes that are politically .viable, *and empowering and providing choice to. beneficiaries. The main lessons in World Bank (2001) for building -,effective institutions are to design them to c6mpldiment what exists to-innovate to.identify institutions that work, and to drop those that do not. This is all sensible'enough. But the relevafit lessons for Brazil may be even more obvious. First, while it is tempting to view foreign institutions that are perforniiig well as models for adaptation, attempts to transplant institutions have rarely led to success. And where they have, the gains, are often short-lived. Second, institutions must be updated, conservatively but constantly. Even appropriate institutions becorne obsolete. Brazil's own experience with labor market,institutions is not inconsistent.with this... President Getulio Vargas imported Italy's corporatist labor structures during the 1930s, driven,perhaps as much by.political.expediency as by an objective assessment that they would mnatch Brazilian habits-and hopes, These institutions have not been changed even as economic and political changes have radically altered the relationship between employers and workers in Brazil and the rest of the world..On the contrary, many of these features, have been ."hardwired" into the Constitution. Sources: Building Institutions for Markets, World Bank, 2001 and Beyond the Washington Consensus: Institutions Matter, by Burki, Perry and others (1998).. - - The Brazil Jobs Report - 8 - Labor Market Governance Regulations Should Promote Employment and Productivity 22. The principal aim of labor regulations is to ensure the smooth functioning of the market so that labor is allocated to its most productive use, and wage is as high as the quality of labor warrants. This also ensures that employment is maximized. When imbalances arise-say, the presence of more job-seekers than jobs-regulations should not impede adjustments in wages to correct imbalances. But such aggregate efficiency should not be the only objective of labor regulations. They should also ensure (i) a degree of fairness in that employment contracts that are entered upon voluntarily are honored by both employers and workers, (ii) that the market is competitive in that employers and workers do not Country Monthly collude or discriminate in negotiating contracts; and (iii) that conditions of work wage meet minimum safety standards. Societies generally go further and try to Brazil 12.9 legislate minimum pay and job security. Thus we have minimum wage laws, Argentina 3.0 restrictions on ending contracts and-where contracts must be terminated- Chile j 3.4 stipulations for payment of severance benefits. Colombia __ 10.7 23. It can be argued that the aim of regulations should not be to guarantee job MexSico 3.2 security but to facilitate employment and productivity growth, and that a well- pain 3.2 functioning labor market is the best instrument for income security, and not job- Italy 3.2 security regulations that lock the worker into his/her current position. But we do Gertany 1.1 not wish to enter that debate here. Nor is it necessary. Regulations in Brazil do Netherlands - 0.9 not entirely promote employment and productivity, nor guarantee income UK 1.4 security. Unsustainable regulations show up in the form of high informality USA 0.0 among the wage and salaried, a preference for self-employment among others, Korea 3.6 and low productivity and employment growth in regulated labor markets. Figure Thailand - 4 shows that finms in Brazil view labor regulations as burdensome; and that labor Source: Heclanan and taxes also are a significant portion of the tax burden. Pages(2001) 24. The reform of regulations in Brazil is especially difficult because many of the rules have been "institutionalized"-Brazil is one of the few countries where labor laws are written into the constitution. Changes in labor laws often require constitutional reform, and attempts to change the CLT without changing the constitution lead to inconsistencies that exacerbate litigation, further straining labor courts. This has led reformners to contemplate using the labor unions to circumvent this problem, e.g., by allowing workers to collectively bargain away some of their constitutionally guaranteed rights as individuals. As we will see in Chapter 6, however, this is neither easy nor without the risk of collateral damage. Figure 3: Brazilian Firms View Labor Regulations and Taxes as EspeciaUly Onerous .. . - _ ._ S B0 LSource: Batra (1999) The Brazil Jobs Report - 9 - Labor Market Governance Interventions Are Needed, Even in Efficient Labor Markets 25. Appropriate institutions and regulations improve labor market functioning and result in better outcomes, but this does not imply that government interventions, like training and unemployment insurance are unnecessary. Interventions may be a second best option to remedy market failures, such as the absence of credit to micro and small enterprises. Further, as we will see in chapters 3 and-4, labor market outcomes depend also on macroeconomic policies and the state of the education and training systems. Interventions may be necessary to improve employment and productivity if it is difficult to improve macroeconomic policies, or if education and training system improvements are slow to yield fruit. An example of such interventions is public training schemes to assist the unemployed. 26. In fact, interventions may be needed to promote earnings security even if education and macro policies are fine. The failure of financial markets to provide insurance against the loss of income due to unemployment-while understandable because of information asymmetries-is perhaps the most important reason for government intervention in well-functioning labor markets. Government initiatives to provide instruments for workers to augment their insurance, precautionary saving, and self-protection efforts can raise the well-being of workers. 27. While a tradeoff between employment creation and income security objectives is unavoidable, these interventions should not unduly raise costs of employment nor inadvertently restrict productivity growth. Striking the right balance is difficult. In the case of remedial interventions such as micro-credit programs and adult literacy and training schemes for the less educated, there is probably no alternative but to constantly evaluate the impact and adjust the design while addressing the root causes of the problem. For interventions to correct chronic market failures, the principle for efficient design should be to provide an instrument that the market would provide if it had not failed. Thus, providing a saving vehicle that is protected from risks of inflation and fraud such as FGTS accounts may be appropriate if the incidence of unemployment is high. And providing an instrument such as Seguro Desemprego to pool the risk of earnings loss may be justified if unemployment is a relatively rare event. 28. Brazil has not yet arrived at a mix of regulations and interventions that balances the objectives of employment, productivity and security well. A survey of Latin American firms found that labor institutions, regulations, and interventions are viewed as especially onerous in Brazil (see Figure 3). Figure 4: The Brazilian State Intervenes Frequently (Question: How Often does the Government Intervene in Firm Decisions?) 4.5 - 4- (3.5 . 3 - 2. 5 -7 2 -~ 0.5 0 B Suc:Batra (1999) t*'.. 4 The Brazil Jobs Report - 10 - Labor Market Governance -Box 5: The CLT---Intervenloniist Principles Remain Ilintact From the 1930s Brazil's labor code, the CLT (Consolida,co das Leis do Trabalho), became effective on May 1, 1943. It consolidated existing labor laws implemented during President Getulio Vargas' first term in power in the 1930s. The core principle of the code is a benevolent state that regulates almost all aspects of the relation between employer and employee. The legislation was influenced by corporatism, the dominant political thinking of the depression era. When the CLT was passed, workers obtained rights previously unheard of, such as a maximum of 8 hours of work per day and the freedom of association. However, at the same time, the law curtailed the room of maneuver of the labor movement, e.g., requiring state approval of leaders of labor unions and their spending. During its 60 years of existence, the CLT has been modified occasionally. After two decades of rising union militancy, the military junta that took power in 1964 reinforced state intervention in labor relations. The government started decreeing national-wide wage increases and heavily suppressed strikes, which eradicated the role of civil society in job relations. Labor protests in the late 1970s succeeded in gaining more influence in the collective bargaining process and the constitutional revision in 1988 further opened the opportunities for labor's self-determination of regulations. The amendments abolished both the prohibition of cooperation among trade unions and the government's right to dissolve a union in case it was found to be inhibiting the government's economic policy. Despite these changes, the CLT still determines many specific terms of employment with little space for negotiations between the worker and the employer. The Constitution has several articles, which accord workers rights that cannot be negotiated-not even by mutual consent-and that are enforced by a system of labor courts. The CLT allows for only one union for each occupation in a given municipality. A union levy is collected from all (forrnal) workers regardless of union membership, of which 60 percent is transferred to the designated union. Unions have neither the incentive nor the opportunity to improve working conditions of the workers. Sources: Arbache (2002), Barros et. al (2001), Camargo (2002), Carneiro (2002) and Salewicz (1993). Constraints By Size By S ctor Overall Small Medium Large Manuf. Services Taxes and Regulations 3.6 3.7 3.4 3.6 3.6 3.61 o High Taxes 3.8 3.9 3.7 4.0 3.8 3.88 o Labor Regulations 3.3 3.5 3.3 3.5 3.5 3.46 o Tax Regulation/Admin. 3.4 3.5 3.2 3.6 3.4 3.46 Policy Instability 3.4 3.5 3.5 3.5 3.5 3.48 Exchange Rate 2.3 3.0 3.2 3.1 2.9 2.94 Crime and Disorder 3.3 2.7 2.5 2.8 2.7 2.74 Inflation 2.6 2.8 2.6 2.6 2.7 2.70 Financing 2.5 2.8 2.3 2.8 2.6 2.69 Corruption 2.6 2.8 2.4 2.8 2.3 2.62 Judiciary 2.3 2.5 2.8 2.5 2.6 2.54 Anti-competitive Practices 2.3 2.5 2.5 2.2 2.5 2.43 Organized Crime 2.8 2.4 2.1 2.5 2.4 2.42 Infrastructure 1.6 2.2 2.5 2.0 2.2 2.20 Ranking: 1=no obstacle, 2-minor obstacles, 3=modernte obstacle, 4=major obstacle Source and Sample: This table is from a survey of private enterprises conducted in Brazil in August 1999, summarized for this report by Geeta Batra. The survey was administered under the supervision of the World Bank Group and the Inter-American Development Bank (IDB) by Gallup Inc. to a sample of 201 firms. The Brazil Jobs Report - 11 - Labor Market Outcomes 29. The debate on labor reforms in Brazil revolves a lot around issues such as high informality, the high frequency of labor disputes, and the high rate of job turnover, and the interconnections between these phenomena. To be sure, these are pertinent issues. But it can be argued that these are not necessarily bad things; as long as the welfare of workers and the profits of firms are reasonable, how the labor market functions is a secondary issue. However, it is difficult to decide what "reasonable" welfare and profit levels are, so we rely on other -E measures that reflect labor market outcomes. In this report we propose that employment, productivity, and income security provide a parsimonious set of i b indicators of labor market outcomes. And their evolution over the last two Country _ decades provides cause for concem. Brazil 2.9 Argentina 4.3 The Outcomes of Interest Chile 6.8 30. A word on why we use these measures is necessary. The reasoning goes as Colombia 3.0 follows. The per capita income of countries is a good reflection of the well-being Mexico 23 of their citizens, and its growth perhaps an even more reliable measure of how Portuaal 2.6 this is changing. The labor market principally affects per capita growth through Italy 1.5 population growth, labor utilization and productivity. Labor utilization is jointly Germany 1.5 determined by the level of employment and the participation of the population in Netherlands 2.9 the labor market. Productivity depends upon on a series of factors, including UK 2.5 investments in capital and skills, and thus in factors such as turnover. Although USA 3.4 these labor outcomes depend upon many factors unrelated to the regulation and Korea 5.7 functioning of the labor market (such as macro-economic changes and Thailand __4.2 educational attainment of the labor force, which will be discussed briefly in Source World Bank Chapter 3 and Chapter 4, respectively), this parsimonious set of indicators allows us to reliably assess labor market performance. Figure 5 accounts for how these factors affected income for the average Brazilian from 1993 to 1999. 3 1. We believe there is another aspect of the economy that workers care about- income or earnings security. To put it simply, most individuals prefer a steady stream of earnings of R$500 per month to another in which income fluctuates randomly between 0 and R$1,000 but averages R$500. And if we care especially about people falling below some poverty line, the case for paying attention to income security around this level is even stronger. Figure 5: Contribution to Growth of Brazil's Per Capita Income, 1993-1999 2.0% v I15% - 1 0% 0 5%- ___ 0 0% - -0.5% - -10% E IGDP growth per c apita oDemographics (W orking-age population/ total population) oLabor Utilization Total (Employment/working-age population) [3Productivity (GDP per person employed) Source. fPEA l)ata on-line The Brazil Jobs Report - 12 - Labor Market Outcomes Insufficient Job Creation 32. For the labor market to promote prosperity, it should assure that demand for labor is met by supply at a price that is acceptable for both parties. This coincidence of wants is what creates employment. The benefits of steady increases in the labor force and employment can be substantial. During the 1990s, for example, higher labor force participation in Ireland and Netherlands raised national income by more than 1 percent of GDP annually, while Sweden and Finland saw declining employment that reduced annual GDP growth by more than 1 percent (OECD, 2001). Job creation is not only important from a oJ~Jj)~a * economic point of view, but also from a social and ethical point of view. Being Country I _____ without a job can undermine the self-esteem of workers and the well-being of Brazil 775 their families, even if they are provided financial support. Argentina 69.9 33. Figure 6 presents the evolution of employment and the economically active Chile 64.3 population for the main metropolitan areas during the last decade. The labor Colombia - market created 2.3 million new employment opportunities-an increase of 15 Mexico 82.4 ak Spain 2 .0 percent-which fell short of the 7.1 million increase in the economically active Portugal n . population. The picture is not very different when national data are used instead Italy T 612 of metropolitan. There has been no change in the trend rate of growth of the Germany j- economically active population. The changes have been in employment growth. Netherlands - Until 1996, the labor market appeared to have created sufficient jobs for the UK i 8.9 expanding population. But since then, job creation has been insufficient to USA | 31.5 provide employment for the growing labor force. Korea 1 45.0 34. Clearly, some of the drop in employment growth does not have much to do Thailand 51.9 with labor market functioning. In 1998, for example, unemployment rate rose Source: ILO from 6 to 9 percent in the course of 4 months-labor market functioning could hardly change so quickly. What is more worrying is that it has not come down to its pre-1998 level. Despite the resumption of job creation in the last two years, the unemployment rate remains 2 percentage points above the unemployment rate prevailing prior to 1998. Put in more alarmist terms, this corresponds to an increase of 25 percent in "structural" unemployment over the decade. Figure 6: The Working Age Population Has Risen Faster than Employment During the Last Decade 135 130 -Working Age Population 125 - Employment o 120 - .115 co '110 a 105 x e) 100 - '95- 90- Jan-91 Jan-92 Jan-93 Jan-94 Jar-95 Jar-96 Jan-97 Jar-98 Jan-99 Jan-00 Jan-01 Source: PME from IIBGE The Brazil Jobs Report - 13 - Labor Market Outcomes Declining Participation in Work and Rising Unemployment 35. Labor utilization is equally determined by the size of the potential labor force and its participation in production. Urban Brazil-comprising more than three fourths of the population-experienced a marked decline in labor force participation during the 1990s, implying that fewer workers must sustain the living standards of the population. Labor force participation in Brazil's metropolitan areas-the percentage of 15 to 65 year olds who worked or searched for employment. In 1991, 61 out of 100 participated in the labor market. By 2001, this had fallen to 56 out of 100, Ramos (2002). At the national -: level, a large inflow of women seems to have counterbalanced the falling X F , * . - participation in metropolitan areas. Overall, participation remained constant at 70 percent. C ntry % 36. But declining participation in work can be because of both good and bad Brazil 70.9 reasons. People may postpone entry to invest more in education, so that when Argentina __4.5 they do enter the labor force they are more productive. Alternatively, people in Chile 59.9 the relevant age group-15-65 years-may now view work to be less rewarding Colombia 71.8 relative to other pursuits. Analysis of the data indicates that increased enrolment exico 63.9 in secondary and tertiary education does in fact partially explain the decline, and Poriugal 70.6 this is encouraging. But this is not the whole story. There has also been a Italy 59.8 decline in labor force participation because of "bad" reasons. This component of Germany 72.2 reduced participation in work could be interpreted as an indictment of the Netherlands 73.6 performance of the labor market, broadly defined to include the design of the UK 76.3 social security system. In the most alarming terms, if we assume that labor force USA 77.2 participation had remained constant at its 1991 level, metropolitan Korea 63.9 unemployment rates would be 15 percent in 2001, as depicted in Figure 7. Thailand 76.7 Source: ILO 37. If even half of the decline in participation rates is because of an exodus from * or latest. the labor market due to a judgment by workers that the quality of jobs has fallen, the unemployment rate would now be in double digits. Figure 7: Had Labor Force Participation Not Fallen During the Last Decade, Urban Unemployment Rates Would be Twice Their Current Levels 17% 16% - Unemployment rate at constant labor force 15% Iartdipation Noe1h 40had are illusraesthesharyeof tewrfocthtwudavbennmpydifheatcptonrateha 13%ho 12912% 11% .9% 9 E ca 8%- 6% 5%- 4% - Jan-91 Jan-92 Jan-93 Jan-94 Jan-95 Jan-96 Jan-97 Jan-98 Jan-99 Jan-00 Jan-01 Source: iPEA Data on-line based on PME from IBGE Note: The shade area illustrates the share of the work force that would have been unemployed if the participation rate had remained at the 1991-level (61 percent). The Brazil Jobs Report - 14 - Labor Market Outcomes Low Productivity Levels 38. Productivity is a difficult concept to interpret. In one sense, it is the flip side of the employment coin: if output increases by 10 percent but employment increases by 5 percent-say because labor is expensive compared to capital due to distorted labor markets-productivity goes up. But this can hardly be a good thing. Again, because of increases in investment in education and training, labor productivity levels generally increase. While this is good, these productivity increases cannot all be attributed to well-functioning labor markets. The -. - - - productivity measure that is appropriate for our purpose, viz., as a desirable _UM labor-related outcome-must capture the concept of microeconomic efficiency of labor use. That is, it should provide a quantitative indicator of how well- matched workers are to jobs. Job-specific investments by firms and workers that Argentina | 4.9 improve how well the work is done will add to productivity. A well-functioning Chile i_44.7 labor market therefore not only ensures that job-seekers find gainful employment Colombia 17.9 and get reasonable income security, it also allows workers to be allocated to their Mexico j -0.9 best use by firms. This implies higher productivity for the firms, and higher Spain ] 38.6 wages for workers with growing employment levels. Portugal 35.6 39. By most measures of productivity, Brazil is not doing well. Value added per Italy ' 38.2 worker was actually lower in 2000 than it was in 1980; in contrast Chile's value Germany - added rose by almost 50 percent, Thailand's more than doubled, and Korea UK 40.8 almost tripled over the same period. Per capita income growth over this period in USA 31.0 Brazil thus seems to have come from increased use of resources (capital and raw Korea 196.2 materials) and not from improved productivity of workers. Increases in income Thailand 115.8 levels are almost always an unmixed blessing, but these numbers indicate that it Source: no will be difficult for Brazil to sustain these increases over time unless more is done to increase productive efficiency in the workplace. 40. Another revealing exercise is to compare labor productivity levels in Brazil today with other countries, after factoring in the obvious advantages of workers in richer countries (e.g., better infrastructure, higher education levels, more capital per worker). We can do this somewhat crudely by estimating what value added per worker in Brazil should be given its level of per capita income, which reliably proxies the advantages that come with greater wealth (see Figure 8). By this measure, Brazil has a "labor productivity shortfall" of between 33 and 50 percent compared with countries such as Malaysia and Mexico which have similar income levels. Figure 8: Brazil Has a 50 Percent Shortfall in Labor Productivity (Industrial value added per worker, 1999) m 60,000 - - 50,000 - - France , 40,000 - d Kingdom . , Germany 0 30,000 - -j - --i *- ! @ / MK~~~~~ Arserdnana-zI- a. X 20,000 - Colomb Xyia "510,000 - > O- 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 GDP per capita (tn US$) Source: ILO KILM(2001) The Brazil Jobs Report - 15 - Labor Market Outcomes Inadequate Income Security 41. Adequate employment opportunities for workers at growing productivity and wage levels are two of the most desirable policy goals. The third is income security. 42. Like productivity, this is a complex concept. But the complexity arises more from misinterpretation rather than mis-measurement. Income security is often equated with job security, but their labor policy implications could not be more different. Policies to increase job security include more incentives and opportunities for workers to upgrade their skills, or disincentives to employers to dismiss workers even if their skills or work habits no longer warrant continuing in the job. In general, the emphasis of labor policy has been the latter, in the form - l * 3 of restrictions or penalties on firing workers that increase with tenure in the firm. Country % Income security as a goal has different policy implications, with the emphasis Brazil 3.0 being on income replacement or consumption smoothing. Policies appropriate for Argentina 5.1 increasing income security are "passive" programs such as saving accounts, Chile 3.8 unemployment insurance and assistance, and "active" programs such as job Colombia 3.0 search assistance, training programs, public works, and support for self- Mexico 3.6 employment. So while job security policies mostly take the form of regulations, Spain 1.7 income security policies are more in the nature of interventions. Portugal 1.8 Italy 1.0 43. With good programs, variability in incomes and consumption will be more Germany 1.9 dampened than GDP growth volatility. During the last two decades, Brazil's Netherlands 1.1 GDP growth volatility has been no higher than its Latin American neighbors, but UK 1.8 considerably higher than countries of the OECD. What is more troubling, USA 1.5 though, is that while consumption growth volatility is lower than GDP growth Korea 4.9 volatility in OECD countries, this is reversed in countries such as Brazil (De Thailand 6.6 Ferranti, et. al., 2000). This hints at the relative ineffectiveness or absence of Source: IMF income support programs in developing countries. 44. Also revealing for the nature of income insecurity in Brazil as compared with other countries is the importance of wage insecurity as relative to employment insecurity, as measured by unemployment (see Figure 9). Brazil has experienced relatively high wage volatility compared with the likelihood of unemployment. To the extent that this is due to the high level of informality in Brazil (where incomes are more variable than employment), this points to the need to extend income security programs to the informal sector. Figure 9: Brazilian Workers Cannot Count on Future Earnings (Indicators of Income Uncertainty for Latin America) ~jś. .- - *.-'.~~.-'~ki)gh income -crtanl * E s;rg; jg -~~r t~ lt S-8 54 V i-t1 -wit*-T ; : _ _ _- __ _.. _ X _5~ _ O, ~enezueia ~~~~~~t ~ ~ Ra Wag Growtha Volti ty, t90 Source: World AL E 0-- 2."i V4.A.T'"i .68 1 .Lo' InAicomie uncertaintly X I Mexico Real Wage Growth Volatility, 1990s Source: World Bank and CEPAL The Brazil Jobs Report - 16 - Labor Demand 45. While labor market institutions have not changed much in Brazil over the last decade, the economic environment in which they operate has been radically altered since the late 1980s. Three main changes have taken place: inflation rates are now down to single digits, barriers to international trade and investment have been reduced, and an ongoing fiscal adjustment is reducing the economic weight of government. All this implies that labor demand has changed considerably. In L - - the next chapter we will see that the nature of labor supply in Brazil has also seen Increase in important changes over the last two decades. If labor market outcomes had not Unemploy0ment, 199- deteriorated, there would still be little cause for concern. But Chapter 2 outlined Country 0 how employment, productivity and income security indicators now display Country , stagnation or deterioration. This chapter examines the possible implications of Aranin 7.7 macroeconomic changes for Brazil's labor reform agenda. Chile 1.4 Macroeconomic Fluctuations Affects Labor Demand Colombia 9.2 Mexico -0.5 46. It is well-known that the temporary ups and downs of the economy strongly Spain -2.1 influence labor demand independently of labor market policy. Figure 10 shows Portugal [ -0.7 how employment and earnings changed with the economic cycle in Brazil. Using Italy | 1.5 GDP growth, periods of economic expansion (1984-1986, 1993-1996 and 2000) Germany 3.1 and contraction (1983, 1990-1992 and 1998) are distinguished. Employment Netherlands -3.1 growth, decline in unemployment, and wage growth take place in good times; in UK -1.5 bad times, the brunt of the adjustment is borne by wage decreases, with USA - employment growth staying positive, while unemployment increases only Korea marginally. This is the finding that leads many observers to give labor market Thailand _ reform low priority. This report provides evidence that this "macroeconomic Source: CEPAL and ILO flexibility" masks a microeconomic rigidity that results in low productivity levels Note: For Latin Amencan and poor income security. But more on that in Chapters 6 7 and 8. countries: urban unemployment a p 47. It is important to understand that the labor market outcomes-in the short run-are dictated by business cycle fluctuations, but in the long run are determined by changes in macroeconomic regimes, or what are commonly called structural reforms. The rest of this chapter focuses on the how macroeconomic reforms that the government of Brazil have undertaken have permanently altered labor demand and the functioning of the labor market. Figure 10: At the Aggregate Level, Brazil's Labor Market Has Adjusted Well to Economic Fluctuations 25% 20% =j Good .YXa.-s m Bad Years Bnployrnent Decline in tnemployment (%-point) Wage grow th Source: PNAD 1999 from IBGE The Brazil Jobs Report - 17 - Labor Demand Stabilization and the Labor Market 48. Three major shifts in Brazilian macroeconomic policy took place during the late 1980s and the 1990s-greater openness, stabilization, and fiscal adjustment. The Collor government initiated the trade reforms in the late 1980s, while the successful inflation stabilization begun in 1994 under the Cardoso government, which equally embarked on fiscal adjustment in 1998. 49. Figure 11 below shows how the inflation stabilization affects labor demand. It displays how the labor market adjusts to changes in economic activity in several countries in the hemisphere. Adjustment in employment, "Employment U -i Okun coefficients", show how much a given change in output changes 3 - - employment, while adjustment in* wages, measured by "Wage Okun ' E - coefficients", indicate the degree to which the change in output results in changes Country In %-o oint in real wages. As we move from countries that have high inflation to those with Brazil 340.9 more stable price levels, the reliance on adjustment through quantities increases. Argentina 344.7 The reason is not complicated-by obviating the need to reduce nominal salaries, Chile 13.6 which is difficult (Brazilian labor laws forbid this for individual contracts), Colombia 8.3 inflation facilitated reduction in real wage levels in response to economic Mexico 31.7 contractions. When inflation falls, this channel of adjustment is narrowed and Spain 4.1 employers have to respond by cutting jobs. Portugal 9.6 Italy 3.3 50. The analysis indicates that the labor reform priorities during a period of Germany 1.0 successful stabilization such as the Real Plan are the institutions and regulations Netherlands -0.7 that impede downward wage flexibility. During the late-1990s, however, there UK 2.4 was no reform of collective bargaining institutions, no change in regulations USA 1.3 forbidding nominal wage adjustments, and a substantial increase in real Korea 1.6 minimum wage levels in Brazil. As growth slowed, both labor market Thailand -0.7 functioning and outcomes have changed in response-employment growth has Source: IMF slowed, unemployment has risen, and the share of informal (self-employed and wage) workers has increased. Figure 11: Falling Inflation Implies More Labor Market Adjustment Through Employment and Less Through Real Wages 2.50 * Bolivia *Mexico 2.00- n X 1.50 Peri cJ * Paraguay * Costa Rica a 1.00 * Argentina * Uruguay 10_ 9 co ChCile n ˘ 0.50 - Brazil A C " Venezuela *United States * Panama 0.00- uull 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 Unemployment Okun Coefficients Source: De Ferranti, Perry, Gill and Serven (2000) The Brazil Jobs Report - 18 - Labor Demand Trade Opening and the Labor Market 51. The opening of the Brazilian economy began in the late 1980s. In less than five years since 1990, all non-tariff barriers were removed and the average import tariff fell from 32 percent to less than 13 percent. The Real Plan also used the exchange rate as a stabilization tool a deliberate choice by the government aimed at increasing the competitive pressures in the market for tradable goods. Imports were to play the role of adjusting aggregate demand to aggregate supply. And intermediate and consumption good imports duly increased after 1995. 52. These changes had three main effects. The first was a rise in the price of I____ Country i non-traded goods-principally services but also construction-relative to traded Brazil 10.6 goods originating in the manufacturing, mining and agriculture sectors which Argentina 9.8 were exposed to the discipline of world markets. The second, related, effect was Chile 29.0 the nse in informality-both for wage- and self-employed workers. The third Colombia 18.0 effect was a rise in productivity in the tradable goods sectors, driven mainly by Mexico 31.0 lower employment rather than improved production techniques. Spain 1 27.6 53. This last findmg is important. In countries that have improved their Portugal 31.0 international linkages through trade reform, the demand for educated workers has Germany 25 generally increased, as new technologies are adopted which are complementary Netherlands 60. 6 with skilled workers (Sanchez-Paramo and Schady, 2002, and Arbache, 2001). IehK ad 25.8 But this has increased more in countries that carry out trade reform with labor _ _USA I 11.1 markets that facilitate this reallocation of workers (see Figure 12 from Cohen, [Korea 42.1 2001). As we will see in the next chapter, compared with Chile and Mexico, for Thailand 58.5 example, the increase in demand for educated workers in Brazil has been quite Source: World Bank muted. The finding highlights the complementarity between trade and labor * or latest. reform, two politically charged public policy issues in Brazil. 54. The rise in informal employment has not outweighed the fall in formal sector employment in the traded goods sectors and-despite a fall in labor force participation rates-unemployment has risen from an average of 6 percent in the early 1990s to about 8 percent since 1996. Not surprisingly, the rise in joblessness is greater in industrial states such as Sao Paulo and Minas Gerais. The increased exposure to external influences may also have increased insecunty of tenure due to an increase in informality, a rise in competitive pressures in formal manufacturing sector jobs and, possibly, an increase in GDP growth volatility. But it is not clear whether these are permanent changes or temporary adjustments during the transition from a closed to an open economy. Figure 12: Trade Openness and Labor Market Flexibility are Complements 1 2% LL 10% < s - - = E 6% 0% X -2% rigid -4% ebl Labor Markets open closed Trade Regime The Brazil-Jobs Report - 19 - Labor Demand Macroeconomic Adjustment and Labor Policy Go In Tandem 55. Greater openness appears to leave the functioning of the labor market relatively unchanged, but demands greater employment flexibility. Opening the economy to external competition exposes differences between the traded and non-traded goods sectors. Generally, this implies first a reduction in (unskilled) employment in- tradable goods sectors, and then an increase in demand for (skilled) workers as these industries begin to compete. Since the de7gree of informality varies between the traded and nontraded sectors, there will also be a shift in the formal-informal "fault-line". The evidence indicates that only the first l change has taken place so far-Figure 13 hints at the decrease in the share of formal (tradable sector) employment relative to the non-tradable sector. But if the labor market is integrated or were made more so, flows between the sectors Brazil - would bridge this rift. Accordingly, labor policy priorities during trade reform are Argentina -.9 to reduce the burden of severance payments on firms, strengthen income support Chile 1_ 0 programs in the formal sector, and extend coverage of social protection programs Colombia -4.7 to the informal sector. Mexico -0.9 56. Stabilization, on the other hand, changes the functioning of the labor market Spain -3.2 because-unless accompanied by labor reform-it reduces real wage flexibility. Portugal -2.8 As discussed above, this exposes the difference between sectors where Italy -3.8 downward wage adjustments are not permitted (e.g., the formal sector) and where Germany -2.0 these adjustments are commonplace (e.g., among the self-employed and in family Netherlands -0.9 firms) Accordingly, labor policies should aim to reduce these differences US -0.8 through measures that can broadly be classified as "deregulation": liberalize USA -0.4 unionization rules, eliminate legislation that restricts downward wage Thailand - adjustments, and reassess minimum wage setting mechanisms. Source: IMF (2002) and World 57. Fiscal adjustment highlights the divide between the taxed and untaxed sector. Bank database Labor taxes form a significant part of the tax revenues (about 8 percent of GDP, or almost 30 percent of tax revenues-see OECD 2001), and social security programs form a large part of govemment expenditures (about a third of current non-interest spending, and responsible for about a quarter of accumulated debt- see World Bank, 2001). Labor reform priorities during a period of fiscal adjustment can be classified as "formalization": reform of payroll tax funded programs combined with increased enforcement of labor laws and, perhaps, temporary micro-credit programs for small firms, given the high interest rates at this stage of adjustment. Figure 13: Real Exchange Rates and Labor Market Segmentation in Brazil Brazil 23 1.5 V2.1- 1 4 a) 1.9 1.3 x1~1 ''''* .-.>/1.7 A - *J xt W tD r- X O O _ N Ch * tD (D Z1.2 a).5 co .1.3 0.9 0.x.9~ - -- - - - - ..9 0.7 . . .. . . . . .. . . .0.8 C\ , CO 1t LO co F- OD CD 0 - C'.J C') LO1 co I~- co OD co co co co OD OD co mD CD CD C C CY) CD CD CD CD CD CD CD CD - CD C CD- C) a) CD CD CD 0D 0D CD CD ..I. Re exchanige rate - w age F/SE - ~ FoDrnaVSE The Brazil Jobs Report - 20 - Labor Demand 58. The analysis done for this report indicates that labor reform has become more pressing since the 1990s. The analysis also provides guidelines to determine labor policy priorities at each stage of macroeconomic adjustment, since it is possible to determine the specific labor policy reforms needed to complement each of the stages of macroeconomic adjustment by tracing the effects of macroeconomic changes in the labor market (see Table 3). Labor policy has, however, has not been high on the reform agenda of the Cardoso government. Because it has been through three structural macroeconomic reforms but has not made any progress in labor policy, one could say that Brazil today faces a large "backlog" of labor policy reform. 59. But while policy reform in trade or monetary policy in a vigorous democracy such as Brazil is difficult, it is even harder to undertake durable reform of policy in contentious areas such as labor markets. Despite this reform backlog, therefore, it would be unwise to expect Brazil to change its labor policies in a hurry or all at once. Fortunately, it appears that this rushed, all-or-nothing approach is not necessary. This report proposes that even a piecemeal approach to labor reform-if reasonably well thought-out-can be effective in improving outcomes. Macroeconomic Trade Liberalization Price Stabilization Fiscal Adjustment Policy Macroeconomic Exchange rate closer to Lower money supply Lower government changes market and reduced tariffs growth and expected spending or higher tax and quotas inflation revenues Interest rate effects Ambiguous effects lower real interest rates Faling interest rates as government borrowing requirement falls and repayment credibility rises Wage effects Indirect: Wages will Direct: Real wage Indirect: Cost of labor reflect labor intensity of flexibility declines as rises relative to capital new product mix inflation falls Employment Direct: Reduced Indirect: Employment Direct: Reduced public effects employment in falls in sectors with rigid employment as noncompetitive sectors real wages government expenditure falls Relevant informal Production of non- Downward wage Evasion of taxes on sector feature tradeables flexibility income and labor Emphasis of Labor Compensatory (subsidies) Cost (wage costs) Compliance (non-wage Policy costs) Labor Policy Lower severance costs, Collective bargaining Payroll tax reform; Priorities strengthen income reform; eliniinate wage social security reform, support programs adjustment rules; assess improve enforcement of minimum wage setting labor regulations The Brazil Jobs Report - 21 - Labor Supply ___0 60. Just as demand for labor changes over time, so do offered labor services. Two facets of the Brazilian labor supply are especially noteworthy. The first is a rise in female labor force participation, which increases specialization of labor and thereby improves labor market outcomes. The second is low attainment of education, which leads to low employability and impedes productivity improvements. This chapter explores the interactions between labor quantity and quality in Brazil. It also tries to answer the question: Do changes in labor supply require a change in the way the labor market is regulated? A Diversified Labor Supply Requires a Spacious Labor Market %Countr *_____% 61. It would not be an exaggeration to characterize the existing labor code as Brazil** 35.5 being designed for a labor force consisting of full-time employed (male) Argentina 33.2 household heads who work under a formal contract eight hours a day in a Colombia 3386 manufacturing firm. However, the labor force continuously grows more diverse Mexico 33.2 in terms of -working hours, job-status, gender and sector of employment. Spain 37.2 62. In Brazil, women have increasingly entered the job market. In 1976, less Portugal 44.0 than four out of ten women in the working age earned a salary. Currently, the Italy 38.5 labor force participation exceeds 59 percent (see Figure 14a). This compares to Germany 42.3 86 percent for men. The large increase in female LFP has primarily occurred Netherlands 40.6 outside of the metropolitan areas, where LFP decreased substantially in the last UK 44.1 decade, see chapter 2. As in other countries, Brazilian-women often search for USA 46.0 more flexible, less career-oriented, part-time jobs in the service sector that allow Thailand 46.3 them to take care of household responsibilities (Volume U, Chapter 3 and Source: ILO Figure 14b). These changes are important for economic development, since they * or latest increase family earnings and national production. ** In the national household survey, PNAD, from 1999, the 63. As job-relations becomes increasingly diverse, the need for labor market female share of the workforce regulations to follow suit and become more flexible increases. As will be argued reaches 40.7 percent. in chapter 6, the existing regulations are designed for a narrow spectrum of workers and often push other workers into informality, where it is difficult to design interventions to help increase income security. The need for a more spacious labor market grows ever stronger. Figure 14a L Women Have Increasingly Entered the Figure 14b: Women Hold Different Jobs Than Men Labor Force 60%- C 65% 40% - 0 600/ 591° _ 24.4% - 7 45/ ~ t E 20% - 1 5 V - od 35°/ __- m ______,___ _ T ^' _ . -60% D i n cK S3.6% 1E 55/0 1 Source: PNAD fom IBEvarousyarsSorce:NAD199frmIBG0% CD500/o 7 % ,j ' 0 E?~~~~~~~~~~ -20% --54 - LL 45% W~~~~~-4% -15.49 . 40 6 0 8~~~~~~~~~~~~~-0 L-a 30% nfe 60 1976 1981 1987 1993 1999 aE . CJ2 (. ~ E E Source: PNAD from IBGE various years Source: PNAD 1999 from IBGE The Brazil Jobs Report - 22 - Labor Supply Strong Labor Market Signals: Quality of Labor Supply Matters 64. The human capital aspects of Brazil's labor supply remain underdeveloped and fragile. According to Barro and Lee (2001), the average Brazilian has completed less than five years of schooling. The low education level of workers places an effective upper limit on labor market outcomes. In Brazil, graduates from higher education are 20 percent more likely to participate in the labor market than workers with elementary schooling only. They are also 38 percent less prone to experience unemployment and in such an event, six times faster in __ finding a new job (Volume 11, Chapter 9). Moreover, it is a well-known fact (population above 25 that productivity and innovation crucially depend on availability of educated years) people (for the case of Brazil see Arbache and Sarquis, 2001). Does the Brazilian labor market function in such a way that these and a long list of other important Country Years of social benefits from education are not transferred to the individual workers and Brazil 4.6 therefore to blame for the inadequate human capital level? Argentina 8.5 65. The answer to this question is unambiguously 'no'. Brazil's labor market Chile 7.9 strongly signals the high and rising value of education: the private return to an Colombia 5.0 additional year of education has risen to about 12 percent and the demand for Mexico 6.7 advanced skills has risen by about 30 percent since 1982 (see Figure 15). What is Portugal . interesting, however, is that the ratio of wages of workers with some college has Italy 7 0 risen relative to those with secondary school education much less in Brazil since Germany 798 1990 than in countries such as Mexico and Colombia (see Figure 16, from Netherlands 9.2 Sanchez-Paramo and Schady, 2002). This is despite a fall in the relative supply UK 9.2 of workers with tertiary education. These trends emerging from the job market USA 12.3 are hardly subtle, and suggest that the reason for under-investment in education Korea 10.5 lies outside of the labor market-Brazil's inability to engender rapid Thailand 6.1 technological progress, and its inability to increase the supply of more educated Source: Barro and Lee (2001) workers. Upgrading the quality of labor supply through educational reform is an integral strategy for improving the main labor market outcomes-employment, productivity and income security. Figure 15: Higher Education, Higher Wages (Private returns to schooling in Urban Brazil by education level, indexed 1982=100) 130- 120 - : 110 - -*-- Higher 100 - 3-Secondary 9--- Primary 80 -)(-- Elementary 70 60 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 Source: Blom, Holm-Nielsen and Verner (2001) based on PME-data The Brazil Jobs Report - 23 - Labor Supply A Closing "Window of Opportunity" for Labor Quality 66. Income inequality is notoriously high in Brazil. As the following chapters will show, the labor market institutions, regulations and interventions reinforce rather than diminish the existing disparities. So improvements in labor policies will help to reduce income inequality. Nevertheless, as Paes de Barros et. al. (2001) argue, the inequalities generated by labor market functioning are far exceeded by the unequal distribution of education and other assets in the population. 67. The benefits of education are so far-reaching that-as this report later C ' argues-it can even negate the adverse effects of some labor policies that are not suited for Brazil. Minimum mandated non-wage benefits are more binding for _ , unskilled workers. As workers become more skilled and productive, these laws Country 28% become less distortionary. The problem of course is that it takes a long time to Argentina a27. raise education levels in the work-force (see Box 6). Chile 28.3 68. Besides, education is not the only instrument for increasing the skills of Colombia 32.9 workers-on-the-job training is another. There are good reasons to believe that Mexico 33.2 labor laws in Brazil do not encourage investments in training, and the functioning Spain 14.6 of the system of tax-financed but ostensibly employer-led training in Brazil is a Portugal 17.0 subject of some debate. Brazil pioneered what is now known as 'S' system of in- Italy 14.3 service training-for example SENAI, SENAR, and SENAC, the network of Germany 15.5 training institutions administered by employer confederations for industry, Netherlands 18.3 agriculture, and commerce respectively. An assessment of this system was UK 18.8 outside the scope of this report. But it is now well-established that while USA 21.5 employers often make claims that appear to indicate substitutability between Koreaia 25.2 education and in-service training, they in fact treat education and training as Source: ILO complements-study after study has shown that more educated workers receive more firm-sponsored in-service training. 69. Good education policies should be seen as a complement to good labor market and training policies, not a substitute. Box 6: Earnings Inequality, Education Inequality and Cohort Size A recent World Bank reJort6on inequality in Brazil shows that reductions in. inequality through improved .distribution of education occur via the entry of cohorts of school-graduates into the labor force because younger generations have less unequal distributions of education. But because of a decrease in population growth over time, younger cohorts are smaller. The improvement on overall inequality thus depends upon two factors, the size of the entering generation (quantity), and this generation's equality of schooling (quality). This-is both good news and bad news. The good news is that (basic) schooling has become increasingly egalitarian during the Cardoso years-The'bad news is that current and future generations of school-leavers count :less's as a share of the workforce, because of decelerating population growth. The reduction in disparities will therefore take longer time to.occur. At best, reduced inequality of schooling will take two decades to noticeably ,affect the existing income-gap between the rich and poor. However, the window of opportunity is fast closing. Demographic patterns indicate that the diminution of successive cohorts of labor market entrants will accelerate over time. The longer time it takes the education system to provide every Brazilian child with equal sch6oling opportunity, the smaller will the benefiting generations be, and the more sand will run into the hourglass before Brazil's income inequality is reduced to 'socially sustainable levels. - ' Source: Brazil Inequality Report, World Bank (2002) The Brazil Jobs Report - 24 - Labor Supply Figure 16: The Rise in Wage Premiums to More Educated Workers Has Been Muted in Brazil Compared with Other Latin American Countries Such as Mexico and Colombia (Dark line is supply of college educated workers relative to secondary educated workers, light line is relative wages) RolatIvo Wago eand Supply by Educotlon Laval -Tartiary and Socondary Education Urban Brazil - Malo and Femaloo 2100 I no 1 40 120 so I s- 40 20 1982 1983 1984 1985 18ane IIC? 198 1D99 1990 1991 1sD2 1993 1994 199s 109C 1907 198 1999 Year af Burley Relativo Wogee and Supply by Education Lavel - Tortiary and Secondary Education Mexico - Males and FomaloD 200 lee- 140- 90 ao 60 20 40 1997 l9os 1999 l990 1991 1992 1993 1994 1995 199e 1997 1998 1999 Year oa Surney Relativn Wogos and Supply by Educatlon Level-Tortlary and Secondary Educatlon Colambia Males and Females 200 160 140 120 'no - - ---- --- F~~~~~~~~~~~~~~~~~~~~~~- WP _-1PsnelSe 90 60 40 20 0 s992 193S 1994 1909 1999 1097 1999 1989 1990 1991 1992 1991 1994 1995 199D 1997 1998 199B 2000 Year at a uerce The Brazil Jobs Report - 25 - Labor Market Functioning 70. This chapter discusses the main features of Brazil's labor market (informality, turnover, and litigation), relates them to outcomes, and flags the policies that may require reform in order to improve functioning and outcomes. Salient Features of the Brazilian Labor Market 71. Several developments in labor market indicators deserve notice. The first is that the sectoral composition of labor has changed radically over time. Most countries have seen a shift in employment first towards industry from agriculture and then towards services, and Brazil is no exception. As Figure 17 shows, X employment in services has been considerably greater than in industry since the C - 1960s, and has exceeded agricultural employment since the 1980s. Today, it Brazil 56.5 exceeds employment in industry and agriculture combined. And it appears that Argentina 74.0 this trend will continue (see sidebar). The second, related, change is that the share Chile 60.0 of the urban labor force has risen. The third change is that unionization rates Colombia 74.5 have fallen, especially during the last two decades. Mexico 53.4 72. These developments are not unique to Brazil. But labor market functioning in Spain 61.9 Brazil is somewhat unique. The likely reason is that countries differ in their Portugal 52.2 institutions, and the regulations and interventions that result because of these Italy 62.2 institutional differences. Regulatory reform in Brazil has not kept pace with Germany 62.6 changes in the nature of work. The following example is revealing of the Netherlands 72.8 slowness with which regulations have adapted to these changes. In 1998, USA 74.5 pressured by farmers, the government introduced a special contract in agriculture Korea 61.1 where the harvests can be labor-intensive but planting and growing are not-the Thailand 33.1 contrato de safra-which allows farmers to hire workers at termns different from Source. ILO those in the CLT. This is the only sector where the law government recognizes * or latest by law that the standard contract do not suit work conditions in every part of the economy. 73. By the logic of the contrato de safra, a new "standard contract" should be formulated to regulate service sector work as well. Of course, given the heterogeneity of service sector employment, this is close to impossible. In the meantime, the "standard contract" in services is the one arrived at informally. Even in industry, it appears that more and more workers and employers are finding the standard legal contract unreasonably restrictive. Figure 17: More than Half of All Jobs Are Now in Services (Employment by Sector 19o0-l 999) 35,000,000 . .~ I +Agiicullure _ 30,000,000 -U- +hIdusiry - _- 25,000,000 _ Serv,ces - __ , _ 3 20,000,000 - o-- -'i'---A -- 2 15,000,000 - a - .S i - 10,000,000 - -- -A 5,000,000 - - - - - 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 1999 Source: Staff calculations, using data from Castelar et al. (2002) The Brazil Jobs Report - 26 - Labor Market Functioning Is Labor Turnover in Brazil Too High? 74. Labor turnover is an integral part of the labor market. Lifetime employment is now rare, since workers' skills and finns' labor demand change during the 40- odd years that a worker typically participates in the labor market. Job turnover arises due to job-separation and job-creation, which is part of the creative destruction of a well-functioning market economy that constantly seeks to match available skills with demand and thereby increase productivity. 75. Nevertheless, job-separation entails a cost for employers and employees. Job-specific knowledge and skills that an employee acquires during employment are lost by job-separation, which can lead to lower productivity and wages. Furthermore, if employers expect high turnover, this reduces the incentive for the employer to invest in the skills of the individual worker (Ramos and Carneiro, *Vi 2002). Second, frequent job-loss may also increase income insecurity. This Years aspect of high job turnover is perhaps best accommodated through well-designed Brazl 6.3 social security programs. The starting point, though, should be to determine Argentina 8.9 whether government policies and programs are inadvertently subsidizing or Chioble - encouraging turnover. Mexico - 76. While the ideal rate of labor turnover for welfare is unknown due to both Spain __ 10.9 positive and negative welfare implications, it is desirable that labor market Portugal 10.4 regulations are neutral in the sense that they avoid distorting the behavior of Italy 11.2 workers and employers. Labor turnover in Brazil does seem to be unusually Germany 10.8 high. One out of three workers changes jobs every year. Moreover, labor Netherlands 10.3 turnover has increased by 60 percent from the recession in 1992 to 2001. A UK 9.0 comparison with intemational statistics confirms this "hyperactivity" in the USA I 9.2 functioning of the Brazilian labor mnarket. The question this report attempts to Korea i - answer is whether turnover m Brazil is made artificially high by labor market Thailand _ - functioning. Paradoxically, in most other countries the question is usually the Source: Maloney (1998) based opposite, since labor regulations that make dismissals costly for employers are on OECD and CEPAL * or latest. generally believed to "freeze the market" or lead to artificially low turnover. 77. Many reasons may be behind this pattern, such as a younger population. But as Figure 18 shows, the proportion of the Brazilian labor force with low tenure for the 25-65 year-old workforce is markedly below even that found in the US, the most active labor market among developed countries. The tendency of Brazilian workers to frequently change employers-confirmed by many studies, e.g., Paes de Barros et al.(1999) and Maloney (1998)-indicates that a Brazil- specific institution or regulation increases turnover. As will be further developed in Chapter 7, the design of the severance fund, the FGTS, seems indeed to subsidize labor market turnover and thus lead to a hyperactive labor market partially explaining Brazil's poor performance in labor productivity. Figure 18: Job Tenure in Brazil and the US 60 so 36.4 4. g 400 at 40- =_.__=_ 27.4 30s----- 20 Low tenure High tenure 25 to 64 old |1Br1azil US7| Source: OECD and PNAD 1999 The Brazil Jobs Report - 27 - Labor Market Functioning Is Labor Litigation in Brazil Extraordinarily Frequent? 78. In Brazil many job separations result in disputes between the employer and the employee, and end up in court with the workers claiming that their rights as listed in the CLT were not honored. The system of labor courts, a special branch of government called the Justica do Trabalho, is charged with conciliation, arbitration and judgment. Scholars debate whether labor courts also have the power to make labor policy (poder normativo) instead of simply interpreting the law, and whether the court strictly upholds precedence. The fee paid to the court is low, but is always borne by the employer. Employers bear the burden of providing proof that the standard contract was honored. The system works as - follows: after a worker files a lawsuit, the case goes to conciliation. If not settled, the case goes to the judge. The average length of time to reach a decision is 31 months (Castelar et al., 2000, and Volume 1, Chapter 6). Parties - - * to the lawsuit can appeal the verdict twice, and many do (see Figure 19). Country r_%___ 79. Labor courts exist in most countries, but nowhere does the labor justice play Brazil 4.3 as prominent a role as in Brazil. Every year, about 2 million lawsuits are filed Argentina 4.3 against current or former employers by workers. That is, more than six percent of Chile 4.6 all salaried workers go to court every year (Figure 19). Four out of five court Colombia 4.4 cases that companies are involved in are labor-related. Workers have resorted to Mexico 4.5 legal action with increasing frequency: lawsuits have increased by 60 percent Spain 4.6 over the last decade. These figures are high even when compared with countries Italy 4.4 thought to be overly litigious, such as the US. Germany 5.3 80. Given the high incidence of labor litigation, it is of no surprise that the labor Netherlands 6.1 justice significantly influences the outcomes of the labor market. The courts UK 5.4 impose both direct and indirect costs to the users. The direct costs of the USA 5.0 extraordinarily frequent litigation are fees payable to the state and the lawyers, Korea 3.5 which are shouldered by the firms. These costs are non-negligible. However, the Thailand 4.8 indirect costs through changed behavior are far greater and fall upon both Sourcep Global Compettiveness employees and employers. The frequent litigation implies that the costs of labor Harvard University and WEF in terms of future pay, fines, and costs of lawyers is uncertain to the employer at (2001) the time of employment. The real price of labor is only known five years after job-separation-after which the justice system no longer accepts claims. Faced with this uncertainty, employers react rationally and are cautious in hiring new workers, increasing screening efforts and raising costs of hiring as well. While there is no systematic evidence, it is likely that this has reduced employment (Volume II, Chapter 10). Figure 19: A Busy Labor Justice System 3,000,000 O Federal (3rd instance)| 2,500,000 * Regional (2nd Instance) , , E]~~E3Locai (1 st instance) | i i i i 2,000,000 - -5 1,500,000 co 1,000,000__ 500,000 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Source: Supremo Tnbunal Federal The Brazil Jobs Report - 28 - Labor Market Functioning Is Informality in Brazil Unusually Elevated? 81. A large fraction of the Brazilian labor force works outside of the formal labor market, where mandated benefits in the CLT such as the 13h salary, paid vacations and overtime bonuses are not necessarily honored, and nor are social security dues paid on behalf of the worker. This is where most jobs have been created during the last decade, adding to the large arny of informal workers. 82. Traditionally, informality has been viewed as an inferior, low productive option, where workers only accepted employment if there were no jobs in the formal sector. Recent research contests this view, suggesting that some informal - workers may deliberately choose to be in the unregulated sector because it offers greater independence, for example, in term of working hours. Surveys of 000 | m J l " informal workers in Brazil and other Latin American countries show that both the Country % traditional, dualistic view and this opposing view are right. What is wrong is to Brazil 27.5 think of informal workers as a homogeneous lot. Supporting this view, roughly Argentina - 15.0 four out of every five self-employed Brazilians prefer an informal job to one in Chile 13.4 the formal sector. But the traditional view is not without merit: informal wage Colombia 1 workers appear to be rationed out of regulated employment, and would-if they Mexico 13.8 had the opportunity-rather work in the formal sector. (Volume [1[, Chapter 3 of Portual - this report). Italy - 83. Being an informal worker always has one major drawback: little or no access Germany - to income support programs such as abono salarial and unemployment Korea - insurance. Informal workers hence face higher income insecurity (Volume 1[, Thailand - Chapter 8 of this report). This excessive insecurity is a prominent feature of the Source: ILO Brazilian labor market, and implies lower welfare levels for workers. *or latest. 84. More than half of the labor force works outside the regulated sector, with the informal salaried at 28 percent of the labor force and the self-employed at 23 percent (Figure 20). Both types of informality increased noticeably through the 1990s from a low of 17 percent each in 1990 to the current levels. Today, while the share of self-employment in Brazil is normal when compared with countries at similar income levels (see Volume 1[11, Chapter 3), the share of informal salaried is double that found elsewhere in Latin America. The extraordinarily large share of informal wage and salaried workers suggests that the regulations are either undesirable for a large and growing fraction of the workforce have significantly reduced the demand for formal contracts (Carneiro, 1997). Figure 20: Informal Wage and Salaried Workers Outnumber the Selfemployed in Brazil 29% 27% - informai Salaied Workers v A1 25% - SeHf-employed (D 223% - .21% 19/o- 17% 15% D CD o co OD D OD OD c c c $ v o co co co co g S rc : . cn O) -. W co 0 CPMEa ta 4 0 -I Source: PME-data from IBGE The Brazil Jobs Report - 29 - Pricing Labor Correctly 85. The CLT and the Constitution are clear on what constitutes an acceptable employment contract. When employers sign the working card (carteira de trabalho), they are seen by the Ministry of Labor as accepting a "standard contract". Under this contract, employers must pay at least the legal minimum wage; they must give a mandatory Christmas bonus, a month of paid vacations, an allowance for the education of children, and so on; and they must contribute to a worker's severance account and the national social security system. Workers may negotiate "nonstandard" contracts through individual or collective - . - bargaining, but the law stipulates that these contracts must be at least as generous - in every single dimension as the standard contract. This chapter provides some evidence that these regulations may well be hurting the least productive workers Country in terms of job opportunities and income security, and provides a graded set of Brazil y6 recommendations to improve outcomes. Argentina _ Mandated Minimum Wage and Non-Wage Benefits Chile Colombia 86. The main wage-related mandates on employers for individual contracts are Mexico 14 minimum wage laws (saldrio minimo), compulsory payment of a thirteenth Spain 52 salary (decimo terceiro saldrio) to all workers at the end of the year, payment of Portugal 38 a family/education allowance (saldrio famflia), contribution to workers' training Italy 68 (saldrio educa,do); other laws forbid reductions in nominal wages, and mandate Germany 66 higher wages for overtime and night-time work. For collectively bargained Netherlands 60 contracts, employers must pay at least what is mandated in the individual labor UK 26 code; another law forbids new contracts from reducing wage and benefit levels USA ____ 33 agreed to in the previous collective contract. Because these benefits are Korea 18 constitutionally guaranteed, they cannot be waived even by mutual consent. Nor Thailand can they be bargained away by labor unions. For a typical contract, these Source: OECD (2001) "Taxing benefits amount to about 35 percent of the basic salary (see Table 4). Wages" and Volume 11 of this report. 87. The main non-wage mandates are paid vacations, contributions to the FGTS Note. Calculation for OECD severance fund accounts (for most employees, other than, e.g., household countnes based on a married couple with two children and two workers), and employer contributions to the national social security system. For earners. the typical worker, these cost the employer another 35 percent of the basic salary. In case of firing, workers always get a month's notice and often are paid a fine of 50 percent of the FGTS account balance (discussed more in Chapter 7). All of these mandates add up to between 70 and 90 percent of the basic salary. Table 4: Mandates Raise the Cost of Hiring Workers by 65 Percent Component Percent Total Basic Wage 100.0 Annual Bonus (decimo terceiro saldrio) 8.3 108.3 Paid vacations 11.3 119.6 Severance Fund (FGTS) 8.5 128.1 Other mandatory benefits 10.0 138.1 Social Security Contributions (INSS) 20.0 158.1 Other Contributions (Work Injury etc.) 4.7 162.8 Employer associations (In-service Training etc.) 3.1 165.9 Source: Volume II, Chapter 6. The Brazil Jobs Report - 30 - Pricing Labor Correctly Mandates Price Low Wage Workers out of Formality 88. This calculation assumes that the worker's productivity warrants a wage greater than or equal to the minimum wage. If the worker's productivity is valued at less than the minimum salary, these mandated benefits add more than 80 percent to the basic salary. For example, for workers with productivity levels 20 percent below the legal minimum wage, these mandates would add 125 percent of the workers productivity. Similarly, for an employee with a productivity level that warrants a salary 33 percent less than the minimum wage, these mandates would amount to 170 percent of the basic salary. The lower the productivity j level of the worker, the more onerous is the burden of the legislation on the 1 . j employer. Under these circumstances, it would not be surprising to see low Country % productivity workers being pushed into informality. And the data seem to provide Brazil 25 support: Figure 21 shows that, in fact, more than a third of the informal workers Argentina 35 in Brazil earn less than the legal minimum wage. While this is not the only Chile 21 reason for informality (e.g., self-employment may be a desirable option for many Colombia 27 skilled workers), it is a plausible explanation for the high informality among Mexico 22 wage and salaried workers. Spain 28 89. For relatively high productivity workers, minimum wage legislation is not Ital 28 binding. So employers can adjust the wage level so that the overall amount Gerany 34 (wage plus non-wage benefits) is appropriate. But there are some mandates that Netherlands 37 imply a cost for the employer, but are valued at less than this cost by the UK 17 employee. Social security contributions are the most important in this regard, USA 14 and there may be others (such as the contributions to SENAIISENAC, FAT, Korea 13 PIS/PASEP contributions, and so on). This is another plausible reason for Thailand - informality, and it is likely that the recent social security reform-which tightens Source: OECD and Packard the link between contributions and benefits-will improve matters. But it will (2001). remain an important factor contributing to informality since the total (employer or latest. plus employee) social security contributions add up to a sizeable 30 percent of wages. International experience suggests that contribution rates above 15 percent are associated with noticeably higher rates of evasion. Figure 21: More than a third of all Informal Workers Earn Less Than the Minimum Wage Y Cumulative Wage Distribution for informal and formal workers 0 Minimum Wage Informal workers 5- _______ _ ___________. ; - Formal workers 0 - 0 5 10 twage Brazil, 1999 Source: PNAD from IBGE The Brazil Jobs Report - 31 - Pricing Labor Correctly Regionalizing Minimum Wages May Reduce the Problem 90. The first legal minimum wage was established in 1940. Originally, there were 14 different regional minimum wages, and by 1963, this rose to 39 levels. In the 1970s and early 1980s, Brazil had an activist wage policy, and the minimum wage was used to guide adjustments in wages. (Volume II, Chapter 7). In 1984, the minimum wage level was standardized nationwide, at the (higher) levels prevailing in the south and southeast. In 1987, the role of the minimum wage as a benchmark for the determination of other wages was officially revoked in order to give minimum wage policy a poverty-reducing focus. But the practice has continued informally, and there is some evidence- nicknamed the "lighthouse effect"'-that minimum wage adjustments still CountY influence other wages. And in 1988, a (largely non-contributory) pensions Brazil program that uses the minimum wage as a benchmark was also instituted. Argentina 91. The debate on minimum wages in Brazil has revolved around the question of Chile 32 its effectiveness as a poverty-reducing policy. Three chapters in Volume H of Colombia 40 this report aim to shed light on this issue. In general, the international evidence Mexico 32 is not conclusive: there are few if any examples of effective use of minimum Spain 354 wage policy as a poverty-reducing policy (Volume II, Chapter 8). Principally, Portugal 7 this is because of the employment-reducing (especially in developed countries) 73Ya and informality-causing (especially in developing countries) effects of mmnimum Netherlands 58 wage increases. Brazil's own experience is consistent with that of other UK 40 countries (Volume II, Chapters 7 and 10). USA 38 92. Since 2001, minimum wage policy has reverted to the principle of having Korea different values for each region, with a federally set minimum. This is a welcome Thailand ___ change, because minimum wages reasonable in the South and Southeast are too Source: Maloney and high for the poorer Northeast (see Figure 22). But the effectiveness of regional Nuflez(2001) minimum wages in bringing poorer workers into formality depends on keeping the federal minimum low. The real minimum wage increased from R$120 to R$180 between 1995 and 2001 (in November 2001 prices). The impact of this on the deficit in Old Age Pensions has been a help in face of populist pressures to raise the level even more-each R$1 increase in the minimum wage increases the social security expenditure by about R$160 million (Volume II, Chapter 7), so the R$60 increase has cost the government almost R$10 billion. Given the "natural restraint" built into this linkage, and the effectiveness of the pension program in lowering old age poverty, this report does not recommend de-linking the minimum pension benefit from the minimum wage. Figure 22: Minimum Wages are More Binding in the Poorer Northeast 40 -. 35,__ _= - _ I_ 30 -c_ -- --- 3,.,-. 0 25 ,'---.=_|J| ' ' CD 51 4 1 L10 -: _ . - South Sao Paulo Southeast Northeast Other OShare in Formnal Sector Employment U Share in Workforce Eaming Minimum Wage| The Brazil Jobs Report - 32 - Pricing Labor Correctly Allowing Unions to Nlegotiate the Mlandate'd lMJinimum Mlay Help 93. The most serious of the proposals to make labor markets more flexible in Brazil is the recent initiative to allow collective bargaining agreements to supersede the labor code. Very simply put, the proposal would allow unions to negotiate wage and employment conditions "from scratch", i.e., unconstrained by the individual labor code or by their own previous agreements. The aim of this proposal is to make contracts reflect enterprise-specific conditions rather than straitjacketing every firm and worker to conform to the mandated minimum. C - * 94. This discussion is a useful one, but there are two caveats. The first is that Country % this would only help if collective bargaining laws are changed. Under the CLT Brazil 17 as originally formulated, unions were to be responsible for harmonizing relations Argentina 25 between labor and capital, and to help in implementing the government's Chile 16 economic policies (for a detailed account of the evolution of collective Colombia 7 bargaining in Brazil, see Volume IIl, Chapter 4). These principles implied close Mexico 31 links with government, but weakened union links with workers and firms. Labor Spain 19 unions were organized by occupation, but employer federations by economic Portugal 39 sector. In 1988, some restrictions on union formation were lifted, but the law Italy 39 continues to give unions monopoly power: only one union in a municipality is Germany 29 recognized as legal for an occupation, and every worker pays a mandatory union UK i 34 fee more than half of which is transferred to these unions whether or not the UKA 34 16 worker is a member of the union. These regulations imply that unions have less USA 16 Korea i - incentive to represent labor's economic interests through mutually beneficial Thailand - bargaining with employers. So the proposal to allow collective agreements to Source: OECD supersede the individual labor code would make the labor market more efficient * or latest. only if the monopoly power of unions (unicidade sindical) is revoked and if payment of unions dues is made voluntary. 95. The second caveat may be even more troublesome. While the experience with minimum wage regulations in Brazil is similar to that in other countries, Brazil's collective bargaining institutions are unique in some key features (Volume 1[1, Chapters 4 and 5). About 17 percent of the workforce is unionized, down from 22 percent in 1986-both the level and trend is not unusual for developing countries. Unlike other countries, however, union members in Brazil are both more educated and otherwise skilled: while 47 percent of union workers have at least a secondary degree, this ratio is just 27 percent for non-union workers, and unionization rates among managers (30 percent) and professionals (42 percent) are higher than for other occupations- see Figure 23. In terms of its economic effects, unionization raises earnings by about 10-15 percent in Brazil, again not unusual given international experience, The Brazil Jobs Report - 33 - Pricing Labor Correctly but inconsistent with Brazil's own law that automatically extends collectively bargained contracts to non-union workers. Unlike other countries, wage dispersion in Brazil's unionized workforce is greater than among the non- unionized. 96. These results suggest that at least in the foreseeable future, unions in Brazil cannot be expected to represent the interests of low productivity workers, precisely those for whom the individual labor code is most binding. Therefore while the proposal to allow collective bargaining to supersede the individual labor code may help in reducing the cost of labor to employers, it is unlikely to help increase the productivity and formality, and income security of poorer workers. Country Brazil "100" Argentina 73 Chile 13 Colombia Mexico Spain 78 Portugal Italy 82 Germany 92 Netherlands 81 UK 47 USA 18 Korea 11.2 Thailand 2.9 Source: ILO and OECD (1999) Figure 23: In Brazil, Union Members are More Educated Than Non-unionized Labor, and are More Likely to be Managers and Professionals than Production Workers 100 90 -_ _ _ _ 80 70 _ _ _ 60__ __ O~~~~~~~~~~~~~~ 50 40 30- 20 0 ~~ '~~~~~~- ~ ~ ~ ~ 0 CO Ml E r 2 j, 3_ C L X- U. At a a. Vom I C 4 Source: Volume 11, Chapter 4. The Brazil Jobs Report - 34 - Pricing Labor Correctly Reducing the Mandated Minimum Package May be Necessary 97. In the sixty years since the CLT was adopted, three developments deserve mention. First, the composition of labor supply has radically changed. The main changes have been an increase in the share of women in the labor force from less than 15 percent to about 40 percent, and a rapid rise in the share of the services sector from about the same as industry and agriculture each to more than agriculture and industry combined. Second, the structure of labor demand has changed considerably. The demand for more skilled workers has increased. Third, in contrast, Brazil's labor institutions and regulations have not changed by Law Is .. much. The two main institutions-collective bargaining rules and organizations (1 =never enforced, and the labor justice system-continue to be based on principles appropriate for 7=strongly enforced) the post-Depression era. And the regulations mandating minimum wage and Country l Rating other benefits remain essentially unchanged. Brazil 5.1 98. The result is a growing dissonance between labor market forces and the Argentina 3.1 institutions that govern their interaction. While it has not helped richer, educated Chile 5.8 workers much, it has definitely hurt those workers that all governments vow to Colombia 5.5 help-low-skilled, poorer workers. Since broad-based reforms that frontally Mexico T 5.4 address such contradictions are political non-starters, stealth appears to be the Spain 5.7 principle underlying many of the proposed changes in labor institutions and italygal 5.0 regulations. One proposal is to give unions more responsibility while making the Gerrany 5.1 market for collective representation more contestable by reforming unionization Netherlands 5.7 rules. There is some merit in such proposals, but they do not improve the UK 5.5 prospects of poorer workers to find jobs which give them the chance to increase USA 6.3 their productivity and pay through hard work and secure tenure. Korea 4.8 99. The best way for Brazil to help the weakest segments of the population is to Thailand 5.2 gradually reduce these mandates.. Minimum wage increases should be restrained, Source: Global Competiuveness and non-wage benefits required by the law should be scaled back. This may be Report (Employer survey) by Harvard University and WEF seem contradictory, but consider this: today, fewer wage and salaried workers get (2001) these benefits than are denied these rights-in 2000, the number of informal wage workers in Brazil surpassed employment in the formal sector. It is also illustrative that the most frequent cause for labor disputes is non-wage benefits (about two-thirds of all cases-see red bars in Figure 24), and that the average court judgment results in about 40 percent of the contested amount being paid the worker (Volume H, Chapter 6). It may be time for the government to recognize de jure what labor courts appear to judge de facto: that the minimum mandate is too high. Reducing these mandates would be the surest way to lower informality among the poor, reduce labor litigation, and thus to increase productivity, employment, and earnings security. Figure 24: Why Workers Litigate: Reasons for Taking Employers to Court 70- -6 60 60 so E 40 u) 30 E 20- l 2 1_ _ i 0S1 0 'X a),- ~~~ - C .0 'EU C_l)0 ) 05 0 D C 0 A ~~ 0 LLa~L C C Z5 0) ,, - The Brazil Jobs Report - 35 - Pricing Labor Correctly Box 7: There Are Always Negotiations: How Labor Courts Function in Brazil J' , ' 4 ,44 '' , S'S' ' 8 , ' jThe Brazilian Labor Court' systemhas ,three levels: Conciliation and Judginent Boards;, Regional Lbor Cpurts, ,and the Superior Labor Court., Until recently, ,the Conciliation and Judginent Board was' composed by a labor- ,"Ja'yer, and a representative 'ech'of employees and of employers (know'n 'as class re'presen,tatives).. These last two members were norninated by the presidenttof the Superior L'ab6r, Couri where the Board is to,flinction. The, Regi6nal Labor Court was 6omposedlar,gely by.labor lawyers and by a tinority of representatives of employees and employers. The class' representatives were The federal government already,in the midsi of a painful fiscal adjustment-clearly did not have'the resources 4,ojpay Nhis amount all at once-,,A,peiod of protracted negotiations between the unions and the gdvernment resulted in a compromisesnluiion. The account holders would be compensated gradually, and the total sum ofg ,compensation was less than' R$40'billion. The compensation is to'be financed by an increase in the employer's? FGTS contribution from 8:0 percent to 85 percent, and by an increase in the fine for dismissal without just cause. fromn 40 to 50 percent. That is,.the a'dditional 10 percentage points are to'go to.the government instead of the R fired wurkr. _ ,,-,5=$ms. >-j,y.kwWiZrKprce ge .ppin ts are, tog to, th gove r n i The Brazil Jobs Report -42 - Aligning Incentives Better Program Nature Advantages and Disadvantages Policy Implication Individual Self-insurance-no Low labor market efficiency Should be considered by countries that saving pooling of risk costs, but welfare reduction have high unemployment, especially accounts especially for poorer workers where labor reforms are only a distant ____________ possibility Severance Pooling over small Almost no advantage. Little Possibly the worst form of unemployment pay group- pooling of risk, entails labor support in a globalized economy globalization market inefficiency, makes labor makes group even relations contentious, and is smaller administratively challenging Public work Market-type Can reach informal sector Should be considered for a part of work programs insurance workers and poor; but can entail force, but not a universal scheme. elements-implicit high leakages in the form of Permanent schemes allow for better pooling of risk non-labor costs when investment balance between consumption smoothing element is made a priority and investment over the economic cycle Unemploy- Market-type Most pooling of risk, can be Should be considered by governments ment insurance-explicit used both to address that have camed out comprehensive insurance pooling of risks idiosyncratic and aggregate risk econornic reforms; labor market and hence serve as a "automatic disincentive effects can be reduced by fiscal stabilizer", generally keeping benefits frugal and "mimicking politically popular the market" as much as possible in design Source: Securing Our Future in a Global Economy, De Ferranti, Perry, Gill and Serven (2000). The Brazil Jobs Report - 43 - Improving Income Security __ 120. FGTS reforms can result in better or less distortionary income security for the unemployed manner, and even increase the size of the formal sector. But informality will be a fact of life in Brazil for many years to come, and interventions should be designed keeping this in mind. Unlike the uniqueness of labor market functioning in Brazil due to the interaction of its institutions and regulations, in the mater of interventions Brazil's experience has been the same as what evaluations in other countries appear to show. Accordingly, Brazil's priorities are to improve the effectiveness of its training, job search and micro- credit programs, and to extend the coverage of passive income support programs 9 .. to cover more of the unemployed. Unemployment or Poverty? What We Should Care More About Brazil 25.4 121. An important issue in the design of interventions is whether one should Argentina care less about unemployment and more about poverty in Brazil, given the C ie 18.4 general belief that in developing countries unemployment is a luxury that only Colombia 28.7 the relatively wealthy can afford. Brazilian data indicate the following: Mxc34 * PME and nationwide surveys reveal that between 70 and 80 percent of the Portugal _ unemployed are from poor households. So concerns about unemployment Italy _ in Brazil are not inconsistent with being concerned about the poor. Germanyd Netherlands * But the magnitude of poverty in Brazil dwarfs the problem of UK unemployment. In 2000, there were between 40 and 60 million poor in the USA country (depending on the poverty line used), and 6 to 10 million are Korea 2.0 unemployed (depending on the definition of unemployment). Thailand 28.2 Source: ILO * In fact, about 75 percent of Brazil's poor have jobs, and about 20 percent * or latest. are inactive. In the large metropolitan regions, however, these ratios are about 55 percent and 40 percent respectively. Labor-related policies to reduce poverty would, therefore, need to emphasize measures to increase labor force participation in large metro regions, but earnings of the working poor in smaller cities and rural areas. Figure 26: Poorer Workers Stay Unemployed Longer, and Increasingly More 19 17o-<1 SM 17 -..o..a 1-2SM 15 3-5SM -I--z-- >5SM 13 9- 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 Source: Volume 11, Chapter 9. The Brazil Jobs Report - 44 - Improving Income Security 122. An important question for the design of interventions is the issue of target group, i.e., some groups such as women, youth and the (less) educated have higher unemployment rates than others. Brazilian data show that: 0 Unemployment rates of women are almost a third higher than for men, and those for young job-seekers (aged 15-17 years) are more than twice as high as for prime aged (25-29 years) workers. -7<3E o, ° Between 1991 and 1997, unemployment rates were highest for job-seekers with 5-8 years of education, but since mid-1998, unemployment rates of people with Country | % 9-11 years have been the highest. Unemployment rates have remained the Brazil 21.2 lowest for people with more than 12 years of schooling. Argentina 18.3 o Between 1991 and 2000, unemployment rates rose for all groups, suggesting a Chile | - structural element, which has to be addressed not by temporary, targeted Colombia |_ 7.7 interventions but by sustained economic recovery and (permanent) changes in Mexico 7.4 labor legislation. At the same time, the poor need help the most. Figure 26 Spain - shows that the duration of unemployment is highest for low wage earners. Portugal 28.5 Italy 14.2 0 Unemployment rates among informal sector workers are high and increasing, Germany 32.9 despite the common belief that informality is itself a safety net (Figure 27). Netherlands 8.7 This has been observed for other countries as well (Volume II Chapter 3). UK 9.9 USA - Improving Effectiveness of Active Labor Programs May Help Korea 3.4 123. The main interventions used to increase income security are micro-credit Thailand 12.3 programs, professional training, and labor market intermediation. Some Source: ILO government intervention in these areas may be necessary, as they are important for income generation among the poorest families, for whom markets tend to be imperfect. But the fact that credit, professional skills and labor intermediation are important for reducing unemployment and increasing productivity-and consequently for tackling poverty--does not mean such services have to be provided publicly or even subsidized; and even less so that they should actually be produced by the state. This section briefly considers why and how the state should intervene. It then surveys the experience in Brazil and suggests remedies. Figure 27: Informal Workers are Most likely to Become Unemployed, and Increasingly So 0.14 - 0 -.~ --fsemf -errploy ->unerrployed 0.1 0.08 - 0.06 - 0..~~~~~~~~~~~~~~~~). o-~~~~~~~~~~~~~~~~~~~~4 0.04 0.02 0 1982 1984 1986 1988 1990 1992 1994 1996 1998 Years Source: Fiess and Domeland (Volume 11, Chapter 9). The Brazil Jobs Report - 45 - Improving Income Security 124. If it were possible to provide access to these services without subsidies for their consumption and direct state involvement in their production, would there still be a need for subsidies and intervention? While direct state intervention in production is hard to justify, there are at least three arguments justifying the existence of subsidies. First, given the importance of these services to the incomes of poor families, they may be "merit goods" that justify a compensatory policy of subsidized provision to low-income population segments. Second, as these services can help in helping workers displaced by structural changes to rejoin the labor market, their subsidized provision can be seen as a social contract to compensate those made worse off by such changes. Last, there may be positive externalities in both professional training and labor intermediation. A -. ; 5 worker who is trained may pass on at least part of the training to his colleagues, at no cost. The more intensively workers search, the lower are firms' recruitment Country _ s costs. In the presence of positive externalities, the private demand for such Brazil 63 services will be below what is socially desirable. The only justification for direct Argentina 48 state intervention in the production of these services arises from technical or Chile 28 logistic reasons involved in ensuring access to the poor, though it is not clear Colombia 48 what these might be. Mexico 67 Spain 82 125. On the whole, the evaluations of the real impact of such policies on Portugal 76 unemployment and productivity in Brazil are far below the theoretical Italy 62 expectations outlined in Box 9 (see Box 10). But they are line with international Germany 42 experience (see Table 8). These limited impacts may partly reflect intrinsic Netherlands __ 31 shortcomings in the policies as instruments for combating'unemployment, low UK 4 labor productivity and poverty, but they may also be a result of shortcomings and USA 4 imperfections in their implementation. Korea 27 Thailand 35 126. Although there is a wide range of factors that could explain why the impact Source: Djankov (2001) of the government training program, PLANFOR, on unemployment and * or latest. productivity is below its potential, the most important one seems to concern the selection of courses and clientele. Course selection is made on the basis of assessments of demand in local labor markets made by municipal and State employment commissions. These are consolidated by the State Employment Commission and the State Labor Secretariat into a State Skill Training Plan (PEQ). The selected courses are then contracted out to a set of executing agencies. Once supply has been defined, trainees are then sought to fill the available vacancies. As widespread publicity for the courses on offer and their requirements is seldom "Box 9: Objectives and Effects of Active Labor Programs: The Theory,- MwIrocredit programs aim to reduce unemployment and raise'labor productivity.' Microcredit acts as a weapon against' *structural unemployment, since it leads to an expansion of output and hence greater demand for labor. As-if has no impact on,i "workers' skills or oh available inf6rmation, microcredit cannot address-fricti6oal or nismatch unemployment. Microcredit is al: useful tool for improving the 'quality ofjobs, since it4leads'to'.investnient,inphysical capitaliand encourages incorporation of new% mtechnol1gies. But it is of'little relevanice in raising productviity if this-is limited either by workers' skills'or by the q'ualivy of Job training also aims to combat both iinemployment'and labor force productivity. It reuces' mismatch unemployrnent, to the extent that expanding-the -capacity of some workers gives them access to' aset of j6bs that would not'be available to them 6without additional training. Training cannot comnbat frictional unemployment and, as: it does not generate jobs directly, its,˘ Pimpact ,on structural employment is small and indirect. Training ipacts productivity mostly through increases in workers' - skills, whereas its impacts on the quality of jobs or worker ifirm matches 'are minor and inidirect. 'Job' search assistance or labor, intermediation services aimito reducle-iniformatio'n imperfections in the labor market.'i Consiequently, they can-reduce unemployment if it is frictional,'and'also increase productivity insofar as poor matching reduces- productivity. Labor intermediation does not have a direct impact on workers' skills or on the quantity and quality of jobs, available; nor is it, effective for tackling structural,unemploymnent. Labor intimediation will uiot be effective in raising-, kproductivity when this is limited either by workers' skills or by low quality of ihe job's.,;:. Source Paes de Barros and others (2002); V'me 11 Chaptr lo., The Brazil Jobs Report - 46 - Improving Income Security made, it is generally left to the executing agency to fill class vacancies from its own clientele. 127. This procedure appears to suffer from two problems. First, the failure to use the workers' own preferences as a basic input in the course selection process may result in errors, since labor market conditions are difficult to predict. Second, the current system does not ensure equal opportunity in terms of access to the program. One way to avoid these problems involves setting up a four-step procedure: (a) construct a catalog of courses that could be offered, and evaluate the _ quality of courses and executing agencies; (b) select the program's clientele; (c) ask each client to choose a course and provider, and the time when they would like the course to be held; and (d) direct executing agencies towards the existing demand E . , , X 1 s for skill training to allow them to program their supply. On the basis of this system * it could be possible simultaneously to guarantee equal opportunity to the program l la s - XRi and to incorporate workers' preferences in the selection of courses. Country | Rating 128. In the case of job search assistance, the main operational constraint seems Brazil 2.8 to be direct intervention by the state in production. Intermediation services are also Argentina _ 3.8 produced in four phases: (a) registration of vacancies; (b) registration of Chile 1 3.3 unemployed workers; (c) matching of vacancies and unemployed workers; and (d) Color 3.1 sending workers to the selected jobs. The key problem in public provision arises in Mexico 1 3.0 stage (d). As a vacancy is a "perishable good", once a potential vacancy has been Portugal 3.3 found, it is essential for the worker to be sent there as quickly as possible. If the Italy________ 2.3 vacancy is registered after registration of the worker, he or she has to be notified or Germaly 2.9 required to check regularly to the agency providing the intermediation service. In Netherlands 3.5 either case the costs are high, and there is no guarantee or expectation that the UK I 4.5 process will ever be quick enough. USA 4.9 129. A way of resolving this problem, with less public intervention, would be to Korea 4.1 limit the government's role to registering unemployed workers and developing a Thailand 4.4 computer program to match them to the available jobs. All registered firms could Source. Global Competitiveness use the computer program in searching for workers, with the possibility of Report (Employer Survey), Harvard University and World developing their own selection criteria. In this case, workers would be called Economic Forum (2001) directly by the firm, although part of the costs of calling them may be borne by the state. As the firm itself would be selecting workers, it would also be financing part of the costs. The cost would be less and a larger proportion of the candidates called would actually be hired. Boy, II: Effects of Active Labor JP'rograms in Brazil: The Evidence Impact assessmeriis in Brazil only exist for training and labor intermediation. in the case of trining, there have.been numerous- evaluanons of PLANFOR. including a series that track the subsequent development of course graduates. .Some of these, apart from tracling course graduates, also include informanon on performance of a companson group. The.resuhs of all the analyses with a comparison group,-except one program in Pernambuco, suggest,the impact on unemployment and income is quite small. Ah evaluation in Rio de Janeiro and Fortaleza showed that training programs had a positive and statisuicall) significant impact on unemployment, but not oni incomes of those already employed: those with access to training were 34 percenl more likely to be employed siM to 12 months afterwards, but there-was no impact onwagies of those employed already. As the program costs about RS170 per person, the new jobs for parficipants need to last more than 17 months for the program to have a net positive benefit. - .. Evaluations of labor intermediation services are less common. A recent assessmeni by IPEA and FIPE based on PvfE longirudinal data tracked the,labor market performance of two groups of previously unemployed workers for tuo * months: unemployed workers who sought labor iniermediation in the week of the interie.w and those who did 'not: Those who sought these.performed better tban those who did nbt in just two of the six Metropolitan regiotns investigated (Belo Horizonte and Salvador). Even in these regions, the impact was such that those %kho sought work .-ended with a employment:rate 2-3 percentage points higher, and a rate of formality 3-6 percentage points higher than those that did not use inter,mediation ser%ices. These results suggest that labor intermediation is irnp6rtant in terms of greater access to the formal sector. but not in reducing unemployment. The Brazil Jobs Report - 47 - Improving Income Security Better Targeting of Income Support Will Help Poorer Workers 130. The main difference between consumption smoothing and safety-net oriented unemployment insurance (Ul) systems lies in replacement ratio and depth of coverage. Ul systems which primarily focus on consumption smoothing generally aim at a high replacement ratio of the previous income. Such systems are often restricted to workers who were employed during a substantial period of time before becoming unemployed. UI programs that emphasize a social net objective generally provide low levels of benefits, but cover a large fraction of the population. 131. The Brazilian UI system is the largest in Latin America, serving an - average of 300,000 to 400,000 beneficiaries each month. It is characterized by a low replacement ratio, short benefit duration and the fact that it is restricted to Country formal sector workers, which implies that UT is not accessible to over 50 percent Brazil of the workforce (see also Figure 28). The program's design does not seem to fit Argentina 31.6 its objectives. A further inconsistency with the social safety net objective arrives Chile 56.5 from the fact that receiving UI is conditional on having access to the Fundo de Colombia 35.4 Garantia do Tempo de Serviqo (FGTS). For a long time the FGTS was the only Mexico 31.2 program that provided income to the workers when they were laid off. Even Span a- nowadays, this aspect of the FGTS is important. UI benefits in Brazil are low ortuga and do not exceed two minimum wages. As a consequence, the amount worker Germany receives from their FGTS is likely to exceed unemployment benefits by far. At Netherlands _ the other extreme are those that are not covered by the FGTS. These workers are UK not covered by UI either. If these workers were less likely to become employed USA this would be less of a concern. But Volume H, Chapter 9 provides evidence Korea that the likelihood of becoming unemployed is highest for those earning less than Thailand the minimum wage, who are (almost by definition) informal sector workers Source: Packard (2001) without access to income support programs. Figure 28: Access to Income Support Prograrns 55% - 1 00% 50//0 900/o 450% E 4-o/- 80% CD 35/a (D 30% J\ cn~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~0 gB 25/o% <20%/6- 15% -\ -- - - -Seguro Des. (leti scales FGTS (nght scale) 10%~ ~ 40% 0 10 20 30 40 50 60 70 80 90 100 Household Income Percerntile Source: Paes de Barros and others (Volume II, Chapter 10) The Brazil Jobs Report - 48 - Improving Income Security 132. Under these circumstances, it would be sensible to expand access of informal sector workers to some form of income support during unemployment. It may be argued that it would be difficult to provide insurance against unemployment to the informal because they are mostly self-employed, and it would be impossible to verify that they are indeed without work during the period in which they are claiming eligibility. And unemployment benefits are usually funded by a tax on payroll, which-almost by definition-is impossible to collect from informal sector enterprises. But these arguments do not pass muster in Brazil. First, the informal wage and salaried workforce is larger than the number of self-employed in Brazil. Second, there is considerable evidence from household surveys that a high ratio of the current (formal sector) claimants of UTI benefits is in fact employed while receiving these benefits. Third, unlike most other countries, unemployment insurance in Brazil is financed not by a tax on payroll but by a 0.65 percent tax on revenues of private firms, 1 percent tax on revenues of public enterprises, and a 1 percent tax on costs of nonprofit organizations. 133. Given these facts and the duplication of benefits (or even triplication, if abono salarial benefits are considered), redirecting the current unemployment insurance system to serve the poor may be a low cost way to help Brazil's poorest and most unemployment-prone workers while maintaining income support for the bulk of the workforce through a sensible reform of the FGTS that eliminates its design flaws. Program Appears to Help Comments 1. Public Works Severely disadvantaged Long-term employment prospects not helped: program Programs (17 groups in providing participants are less likely to be employed in a normal job and evaluations) temporary employment earn less than do individuals in the control group. Not cost- and a safety net. effective if objective is to get people into gainful employment. 2. Job-search Adult unemployed Relatively more cost-effective than other labor market assistance/ employment when economy is interventions (e.g. training) - mainly due to the lower cost, youth services (19 improving; women may do not benefit usually. Difficulty lies in deciding who needs help evaluations) benefit more. in order to minimize deadweight loss. 3. Training of long-term Women and other No more effective than job-search assistance in increasing re- unemployed, those laid- disadvantaged groups employment probabilities and post-intervention earnings and are off en masse, and youth generally when 2-4 times more costly. For youth, employment/earnings prospects (47 evaluations) economy is improving. not improved as a result of going through the training. Taking However, little positive costs into account - the real rate of return of these programs is impact on youth. negative for youth. 4. Employment/ Wage Long-term unemployed High deadweight and substitution effects. Impact analysis shows subsidies (22 in providing an entry treatment group does not do well as compared to control. evaluations) into the labor force. Sometimes used by firms as permanent subsidy. 5. Micro-enterprise Relatively older groups, Low take-up rate among unemployed. High failure rate of small Development Programs the more educated. businesses. High deadweight and displacement effects. Cost- (15 evaluations) benefit analysis rarely conducted but sometime show costs to UI I____________I_____ budget higher than for control group. Source: Dar, 2001 (Volume 11, Chapter 11) The Brazil Jobs Report - 49 - Policy Priorities 134. This concluding chapter may also be viewed as the executive summary of the report. It recapitulates the approach used and highlights the innovative features of the report, summarizes the main findings, and discusses the principal policy implications. What's New About this Report 135. Brazil's labor market has received a lot of attention from researchers, and it is fair to ask what this report adds to the large stock of knowledge on the Novel aspects of subject. There are five features that distinguish this report from earlier attempts: * A new approach. First, it is structured to systematically address all the this report include aspects of Brazil's labor market that are central to the public debate in the treatment of the country-labor market governance, labor market functioning, and labor market outcomes. It traces the link from outcomes-employment growth, "third side" of the productivity levels, and income security-to the labor market's functioning-litigation, turnover and informality-to arrive at policy options labor market to improve its governance, viz., its institutions, regulations and interventions. * Macro-labor linkages. Second, in doing so, it also breaks new ground in -institutions- identifying links between macroeconomic policies, its effects on labor along with labor market functioning and outcomes, and labor policy priorities corresponding to different macroeconomic stages. The priority during trade liberalization is demand and to facilitate reallocation of labor-strengthening income support programs while lowering severance costs. During stabilization the priorities are to supply; a special increase wage flexibility-collective bargaining reform and reassessing wage-related mandates such as minimum wages. In periods of fiscal emphasis on adjustment, the policy focus is on non-wage costs-streamlining payroll taxes, social security reform and improving enforcement of labor laws. implications for * Attention to institutions. Third, along with analysis of demand and supply the poor; and a of labor, the analysis of labor market institutions-especially the labor justice system and trade unions-has a prominent place in the report. The report graduated set of thus blends the economic and institutional strands of labor policy analysis. * Concern for feasibility of reform. Fourth, because labor reform is policy options that politically difficult and always undertaken in a piecemeal fashion, the report are sensitive to presents policy options that range from the modest to the radical. The recommendations include institutional reform but, recognizing that this will political feasibility be difficult and time-consuming, other options are also suggested. To guide policymakers, the report also attempts to determine priorities and sequencing. considerations * Strong focus on the poor. Finally, and perhaps most importantly, the report explicitly examines the implications of labor market functioning for poor workers. In particular, it illustrates how Brazil's poor are the most disadvantaged by labor regulations and institutions that are designed to help them, resulting in their exclusion from income security programs and other formal safety nets. The Brazil Jobs Report - 50 - Policy Priorities Summary of Main Findings 136. Functioning. Brazil's labor market functions differently when viewed in the aggregate than from the viewpoint of firms and individuals, and from the viewpoint of the non-poor than of poor workers. It is: ° Increasingly female and service oriented. The nature of employment has changed radically. While in 1970, four out of ten women of working age B 'razil's labor were in the labor force, this ratio is now six out of ten. In 1970, the shares of services, agriculture, and industry were roughly equal; today service sector market is employment is more than that in agriculture and industry combined. characterized by o Flexible in macroeconomic terms. At the aggregate level, changes in real wages have absorbed most of the impact of macro fluctuations during the last macroeconomic decade, with employment growth taking place in both good and bad times. But since stabilization in 1995, this "macro flexibility" has declined. flexibility that o Increasingly hyperactive. Labor turnover has increased by about 50 percent during the last decade, making an already hyperactive labor market even masks more so. Labor turnover is high even when compared with active labor microecoDnomic markets such as that of the US. ° Adversarial, especially for the poor. Labor litigation is an order of rigidiies, magnitude higher than benchmark countries. More than two million lawsuits are filed every year, implying that more than 6 percent of all workers go to reflected in high court every year, and four out of five lawsuits faced by firms are labor- related. Poorer workers-those earning less than two minimum wages-are turnover, about three-quarters of litigants, but less than half of the workforce. frequent o Less unionized. Unionization rates are down to 17 percent from about 22 percent in 1986, a trend that is common around the world. Unlike most other lItigation, and countries, however, union members are more educated than non-unionized labor, and are more likely to be managers and professionals than production rising informality workers in Brazil. Unions are not representative of formal sector workers, and even less so of the pbor among them. ° Increasingly informal. More than half of employment in Brazil is unregulated. While the share of the self-employed has increased since 1990-it is currently 23 percent of the labor force-it remains below benchmark countries. However, the share of informal wage and salaried workers has increased even faster and is now 28 percent of the workforce, much higher than comparable countries. Unlike the self-employed, most of these workers appear to in the informal sector involuntarily. And almost half of these workers earn less than the minimum wage. 137. Outcomes. The report finds signs of deterioration of labor market outcomes, or underperformance relative to benchmark countries, both of which signal the need for labor reforms. The main outcomes are: o Sluggish employment growth. Job creation is increasingly insufficient- while the working age population increased by 25 percent between 1991 and 2001, employment grew by about 12.5 percent. The unemployment rate now hovers around 8 percent, as compared with 6 percent in the early 1990s. The Brazil Jobs Report - 51 - Policy Priorities * Falling participation rates. Labor force participation has fallen in urban areas from 61 to 56 percent. Had it fallen by only half the amount it did, unemployment rates would be in double digits. * Low productivity. Value added per worker in Brazil was lower in 2000 than it was in 1980; in contrast, Chile's labor productivity rose by almost 50 percent and Thailand's more than doubled during this period. Compared The outcome with countries at the same income level such as Mexico and Malaysia, Brazil faces a productivity shortfall of about 50 percent. associated with * Low income security. In 2000, for the first time, the number of workers without access to social security and compensatory programs became larger Brazil's labor than the number of those who enjoy some income security during unemployment, disability, or old age. market 138. Policies. Based on the analysis, this report arrives at the conclusion that institutions, labor reforms are needed. The three policy objectives are (i) to price labor correctly to increase employment growth, (ii) align the incentives that workers regulations and face to raise labor productivity and wages, and (iii) improve income security programs so that workers and their families are better protected from both interventions is systemic and idiosyncratic shocks. The instruments for achieving better outcomes are to eliminate subsidies for labor turnover, move negotiations from that a growing labor courts to the workplace, reduce non-wage benefits while maintaining a moderate minimum wage, and eliminate overlaps in income security programs share of jobs and expand their coverage. offers neither Eliminate Subsidies for Labor Turnover 139. While frequent job changes do not imply poorly functioning labor markets, the prospect of the government should not subsidize turnover, even inadvertently. It appears that ,rdutit and the current regulations and institutions do exactly this. One of the main reasons productivi a for the unusually high turnover is the design of severance regulations, especially wage growth the FGTS scheme. The first reason is that FGTS accounts usually pay less than w market rates of interest; the second is that the 40 percent penalty for unjust nor of adequate dismissal is paid to the worker. 140. The "first best" solution is to pay market rates on FGTS accounts, and to income security eliminate the 40 percent penalty for unjust dismissal. But this may be politically difficult-those who currently benefit from access to these low-cost funds would likely resist any reform to pay market rates, and workers may see abolition of the fine as reduction of their legitimate benefits. A compromise-for which there is some precedence since 2001-is for the government to collect the fine, and use it to increase the rate of return on FGTS accounts above the currently guaranteed amounts. This would reduce the perverse incentives under the current scheme, and may even serve as a cyclical stabilizer. The report discusses several other intermediate alternatives. 141. The main effect of these reforms would be to lower turnover to more normal levels, and hence increase productivity as firm-specific investments increase. These changes will also make labor relations less adversarial, and hence promote employment growth. The Brazil Jobs Report - 52 - Policy Priorities Move N'egotiations from Labor Courts to the Workplace 142. The current design of labor regulations, the functioning of labor courts, and the manner in which unions operate all result in most worker-firm negotiations taking place after the contract is dissolved and not-as is the practice in better functioning labor markets-before the contract is signed, or after some time on the job. Encouraging negotiations while on the job and not in courts will reduce litigation and lead to higher employment and productivity. 143. A proposal in Brazil is to increase the role of unions to offset the reach of Brazil's labor labor courts. While a shift in negotiations from courts to unions will help, it is not a universal panacea. Collective bargaining laws in Brazil do not encourage policy priorities unions to be representative of workers' interests. The law that requires all workers to pay dues to unions whether or not they belong to one and the are to eliminate stipulation that there should be only one legally recognized union for each subsidies for occupation and city implies little or no competition in the market for collective representation of workers. So these laws would have to be changed-no easy labor turnover, task because this would require changes in the federal constitution. 144. But even if these laws were changed, labor outcomes will not improve at change the once and for everyone. This report has documented that while the main clientele of labor courts are poorer workers, unions in Brazil traditionally represents better locus of paid, more skilled workers. While there is a law that extends union-negotiated terms of employment to all workers, large union-nonunion earnings differentials ne!gDotiatiOs are evidence that this law is ignored. Therefore, more reliance on unions may from labor still leave poorer workers with no choice but to revert to the current system of bargaining-in labor courts. These changes may reduce litigation among courts to the wealthier workers, but not among the poor; inequality in access to formal sector employment may thus increase. workplace, 145. And inequality concerns in a country as unequal as Brazil are not peripheral but central. If collective bargaining laws were changed, union lower the membership would rise and litigation would decrease. And if the manner in which labor courts function were reformed there would again be fewer lawsuits. minimum Reduced litigation will reduce the uncertainty regarding costs associated with employing workers, and hence increase formal sector employment. To the extent mandate down that this would lower turnover, it would also increase productivity. from the middle, Lower the "Minimum" Mandate Down From the Middle and widen the 146. The standard neoclassical solution proposed by many labor economists also appears to fit Brazil-lower mandated non-wage benefits. This is the one safety net for change that would help on all fronts-employment growth, productivity, income security-and would especially help low wage workers. This may seem workers somewhat paradoxical: after all, is it not the poor who most need protection under the law? The analysis in this report shows that the laws and institutions are hurting those whom they were originally designed to protect. 147. Changes in the demand and supply of labor have resulted in the "standard contract" as stipulated under the CLT and the Constitution becoming more and more unsuitable for workers and firms. Labor institutions and rules designed to regulate employment of heads of households working in large industrial enterprises are now being used to regulate a world of workers employed in small The Brazil Jobs Report - 53 - Policy Priorities and medium sized service-oriented enterprises with a growing share of women. The standard contract is suited for a smaller and smaller share of workers, and the most obvious sign of its rejection is the growing share of workers without a The optimal signed working card (sem carteira assinada). 148. Another important concern in Brazil is the large disparity between the sequencing of poorer northeast and the richer, more industrialized south and southeast. The reforms is report shows that mandating a high federal minimum wage and non-wage benefits unintentionally exacerbates inter-regional differences in the quality of difficult to jobs instead of making labor market outcomes more uniform across the country. 149. While minimum wage setting is now regionalized, the federal minimum determine, but has risen rapidly since 1995. All signs are that this has led to reduced income Brazil should security in the poorer northeast, while not helping workers in the richer parts of the country. Allowing workers and employers to customize contracts will probably start by increase formality and extend the safety net to poorer workers. This can best be accomplished by reducing the minimum mandate regarding non-wage benefits reforming such as vacations, overtime restrictions, and severance payments, and by regionalizing minimum wage setting and keeping the federal minimum wage severance moderate so that poorer, less productive, workers are not pushed into informality. Widen the Safety Net for Workers schemes- 150. Brazil's income support programs have two major shortcomings. The first especially the is that there is duplication of benefits for formal sector workers. Thus, while benefit levels are not always generous, formal sector workers who lose their jobs FGTS-and can access FGTS savings accounts to which their employers have contributed, they usually receive 40 percent of the account balance, and generally also receive expanding the unemployment benefits. Low wage formal sector workers also get a Christmas bonus from the government (abono salarial) and their employer. coverage of 151. The second shortcoming is that-in sharp contrast-informal sector income security workers get nothing. While it can be argued that inforriality is a deliberate choice for some, it is difficult to support the claim that everyone in unregulated programs; employment has chosen this over a job in the regulated sector. In general, many wage and salaried workers are believed to be in the informal sector because these improving are the only jobs on offer for them, as compared with many of the self-employed who may be there by choice. collective 152. The changes proposed above will increase the share of the workforce that bargaining and is formal. But even so, informality will be a fact of life for some years, and the safety net should be extended to increase income security for the informal wage reducing the and salaried. This report offers options for doing so, one of which addresses both the problem of duplication and of insufficient coverage. This is the proposal to level of redirect the current unemployment insurance program-under which benefits are funded by a tax on sales and occasionally paid to those who have found work-to mandated the informal wage and salaried. The table below summarizes the main reconmmendations, effects on functioning, and the principal effects on outcomes. benefits will be more difficult The Brazil Jobs Report - 54 - Policy Priorities Policy Recommendations, Effects on Functioning, and Impact on Outcomes Reform Severance Paymen System Primary: Lower iurnover Primary: eigher produotivity (a) Pay market rates on FGTS accounts Secondary: Lower litigation Secondary: Higher employment (b) Makede fine go to workers' fund to increase interest rate on FGTS accounts (c) Make FGTS fine payable to the Govemment and not the workers (d) Eliminate the fine for unjust dismissal Lower Minimum Package Primary: Lower informality Primary: Better social protection (a) Make use of regional minimum wage setling Secondary: Lower litigation Secondary: Higher employment (b) Reduce maniated non-wage benefits Encourage Negotiations in Workplace Primary: Lower litiuation Primary: Higher employment (a) Greater freedom in creating unions- Secondary: Lower tumover Secondary: Higher productivity eliminate unicidade sindical clause (b) Greater competition for unions-eliminate mandatory union contributions Make Contracts More Flexible Primary: Lower turnover Primary: Higher employment (a) Allow wages to be adjustable downward Secondary: Lower litigation Secondary: Higher productivity (b) Allow collective contracts to be lower than previously negotiated contracts Improve Functioning of Labor Courts Primary: Lower litigation Primary: Higher employment (a) Eliminate inconsistencies between CLT and Constitution (b) Assess the usefulness of poder normativo (c) Assess arbitration rules and faimess of court decisions Widen Safety Net for Workers (FAT) Little impact on labor market Primary: Better social protection (a) Seguro Desemprego redirected to informal functioning Secondary: Higher employment wage and salaried (b) Micro-credit directed to self-employed and micro-enterprises (c) Training and job search assistance directed to formal sector job seekers Improve In-service Training (S system) Little impact on labor market Primary: Higher productivity functioning-works through quality of labor supply The Brazil Jobs Report - 55 - Arbache, Jorge S., and Francisco G. Carneiro (1999), "Unions and Interindustry Wage Differentials," Pergamon, PII: S0305-750X(99)00090-X. Arbache, Jorge S. (2002), "Trade Liberalization and Labor Markets in Developing Countries: Theory and Evidence" in "Economic Growth, Inequality and Migration: National and International Perspectives," edited by Edward Elgar. Arbache, Jorge S. (2001), "Liberalizacao Comercial e Mercado de Trabalho no Brasil." Departamento de Economia. Universidade de Brasilia. Arbache, Jorge S. (2002), "Unions and the Labor Market in Brazil" Volume II of this report, The World Bank. Barro, Robert and Jong-Wha Lee, (2001), "International data on educational attainment: updates and implications", Oxford Economic Papers, 53, No. 3:541-63, July 2001. Batra, Geeta and Andrew Stone (1999), "The Business Environment in Brazil: Results from a Private Sector Survey", The World Bank. Betcherman, Gordon, Amy Luinstra, and Makoto Ogawa (2001), Labor Market Regulation: International Experience in Promoting Employment and Social Protection, The World Bank, SP Discussion Paper No. 0128. Blom, Andreas, Lauritz Holm-Nielsen, and Dorte Verner (2001), "Education, Earnings, and Inequality in Brazil," 1982 - 98, World Bank Policy Research Working Paper 2686. Blom, Andreas, and Carlos Eduardo Velez (2001), "The Dynamics of the Skill-Premium in Brazil: Growing Demand and Insufficient Supply?" The World Bank. Bonelli Regis, and Fonseca Renato (1998), "Notas Sobre o Mercado de Trabalho." Ministerio do Trabalho, Numero 3. Burki, Shavid Javed, and Guillermo Perry (1998), "Beyond the Washington Consensus: Institutions Matter." The World Bank. Burki, Shavid Javed, Guillermo Perry and William Dillinger with Charles Griffin, Jeffrey Gutman, Fernando Rojas, Steven Webb, Donald Winkler (1999), "Beyond the Center: Decentralizing the State." The World Bank. Camargo, Jose Marcio, and Marcelo Neri (1999), "Emprego e Productividade no Brasil na Decada de Noventa." Naciones Unidas, Reformas econ6micas. Camargo, Jose Maircio (2002), "Fake Contracts: Justice and Labor Contracts in Brazil" Volume II of this report, The World Bank. Carneiro, Francisco G, and Andrew Henley (1997), "Wage Determination in Brazil: The Growth of Union Bargaining Power and Informal Employment," The Journal of Development Studies. - 56 - Carneiro, Francisco G. (1997), 'The Changing Informal labour market in Brazil: cyclicality versus excessive intervention", Labour, 11 (1): 3-22. Carneiro, Francisco G., and Joao Ricardo Faria (2001), "Macroeconomic determinants of equilibrium unemployment insurance, Labour, 15 (3). Carneiro, Francisco G., and Joao IRicardo Faria (2002), "Labour", Review of Labour Economics and Industrial Relations, Volume 15, Number 3, Blackwell Publishers, ISSN 1121-7081. Carneiro, Francisco G. (2002), "An Overview of the Effects of the Minimum Wage on the Brazilian Labor Market" Volume II of this report, The World Bank. Castelar, Armando, Bolivar Lamounier, Maria Tereza Sadek, Vladimir Brandrio, Eunice Nunes and Marco Aith (2000), Judiciario e Economia no Brasil. Sao Paulo: Editora Sumare. Cohen, Daniel (2001), "Fear of Globalization: The Human Capital Nexus", working paper, OECD. Cunningham, Wendy (2002), "The Poverty Implications of Minimum Wages in Developing Countries" Volume II of this report, The World Bank. Dar, Amit and Francisco G. Carneiro (2002), "Economic Impact of Unions and Collective Bargaining: Lessons from Intemational Experience" Volume II of this report, The World Bank. Dar, Amit (2002), "Active Labor Market Policies: Evaluative Evidence" Volume II of this report, The World Bank. Dahlman, Carl, Anuja Adhar Utz, Jean-Eric Aubert, and Christine Zhen-Wei Qiang (2001), "Brazil and the Knowledge Economy", Background paper for the Brazil Growth Report. The World Bank. De Ferranti, David, Guillermo Perry, Daniel Ledernnan, and William F. Maloney (2001), "From Natural Resources to the Knowledge Economy: Trade and Job Quality", World Bank Latin American and Caribbean Studies, Washington, DC: The World Bank. De Ferranti, David, Guillermo Perry, lIndermit Gill, and Luis Serven (2000), "Securing Our Future", World Bank Latin American and Caribbean Studies, Washington, DC: The World Bank. Djankov, Simeon, Rafael La Porta, FlForencio Lopez de Silanes, and Andrei Shleifer (2001), "The Regulation of Entry." The World Bank. Fiess, Norbert, and lindermit Gill (2001), "Schooling and Growth in Brazil: the Long and Short of it," working paper. The World Bank. Fiess, Norbert, Marco Fugazza and William Maloney (2002), "Labor Markets and Economic Adjustment: The Brazilian Case" Volume II of this report, The World Bank. The Brazil Jobs Report - 57 - Fiess, Norbert, and Dorte Domeland (2002), "Unemployment and Unemployment Insurance" Volume II of this report, The World Bank. Gill Indermit S., Claudio E. Montenegro and Dorte Domeland ( 2002), "Crafting Labor Policy: Techniques and Lessons from Latin America." The World Bank & Oxford University Press. Gill, Indermit S., and Harry A. Patrinos (2001), "Human Capital Policies for Growth in Brazil", Paper prepared for the Brazil Growth Report. The World Bank. Gonzalez Anaya, Jose Antonio (1999), "Labor Market Flexibility in Thirteen Latin American Countries and United Sates: Revisiting and Expanding Okun Coefficients." The World Bank. Guasch, J. Luis (1999), "Labor Market Reform and Job Creation: The Unfinished Agenda in Latin American and Caribbean Countries." The World Bank. Heckman, James, and Carmen Pag6s (2000), "The Cost of Job Security Regulation: Evidence from Latin American Labor Markets", Inter-American Development Bank, Working Paper Series 430. ILO (2001), "Key Indicators of the Labour Market 2001-2002," International Labour Office. IMF (2002), "World Economic Outlook Database," IMD (2002), "World Competitive Yearbook", IMD International Institute, Switzerland ILPEA (1994), "Conferencia sobre Regulamentacao do Mercado de Trabalho no Brasil", Instituto de Pesquisa Economico Aplicada, Seminarios No. 1/94 Directoria de Pesquisa. Lazear, Edward (1986), "Incentive Contracts", NBER Working Paper No.1917. Maloney, William (1998), "Self-Employment and Labor Turn-over in LDCs: Cross-Country Evidence", World Bank Working Paper, The World Bank. Maloney, William, and Jairo Nufiez (2000), "Measuring the Impact of Minimum Wages: Evidence from Latin America", World Bank Working Paper, The World Bank. IPEA (1998), "A Economia Brasileira em Perspectiva", Volume 2. Menezes-Filho, Naercio Aquino, and Mark Roland Thomas (2001) "Household Income Growth and its Distribution in Brazil". OECD (1994), "Employment Outlook", Organization for Economic Cooperation and Development, HD5701.015 1994. OECD (1995), "Taxation, Employment and Unemployment," The OECD Jobs Study. OECD (1999), "Taxing Wages 1998-1999", Organization for Economic Cooperation and Development, OECD. -58- OECD (1999), "Assessing Performance and Policy", Implementing the OECD Jobs Strategy, OECD. OECD (1999), "Employment Outlook", OECD. OECD (2001), "Education Policy Analysis", OECD. Ogier, Thierry (2001), "Brazil to Boost Spending in Pre-poll budget", FT.com Financial Times. Paes de Barros, Ricardo, Carlos Enrique Corseuil, and Gustavo Gonzaga (1999), "Labor Market Regulations and The Demand for Labor in Brazil". Departamento de Economia Puc- Rio, Texto para discussao No. 398. IPaes de Barros, Ricardo, Carlos Enrique Corseuil, and Saxmir Cury (2001), "Salario Minimos e Pobreza no Brasil: Estimativas que Consideram Efeitos de Equilibrio General", Instituto de Pesquisa Aplicada - IPEA, Texto para Discussao No. 779, ISSN 14154765. Paes de Barros, lRicardo, Carlos Enrique CorseuiD, and Monica lBahia (1999), "Labor Market Regulations and the Duration of Employment in Brazil", Instituto de Pesquisa Aplicada - IPEA, Texto para discussao No. 676, ISSN 14154765. Paes de Barros, IRicardo, Carlos Enrique Corseuil, Daniel Dom,ingues dos Santos, and Sergio Pinheiro Firpo (2001), "Insercao no Mercado de Trabalo: Diferencas por sexo e Consequencias Sobre o Bem-Estar", Instituto de Pesquisa Aplicada - IPEA, Texto para discussao No. 796, SIN 14154765 Paes de Barros, IRicardo, Mauricio Blanco Cossio and Jorge Luiz Teles (2002), "The Efficiency of Active Labor Market Policies for Poverty Reduction in Brazil" Volume II of this report, The World Bank. Ramos, Lauro, and Francisco Carneiro (forthcoming), "Os determinantes da rotatatividade do trabalo no Brasil: instituic6es y ciclos econ6micos", Nova Economia forthcoming. Ramos, Lauro (2002) "O Mercado de Trabalho Metropolitano Brasileiro na Decada de 90", IPEA Discusi6n Paper, IPEA. Salewicz, Stephen (1993), "Corporatism and the Brazilian Labor Movement", Latitudes Volume II, McGill. Sainchez-Paramo, Carolina, and Norbert Schady (2002), "Off and Running? The Rising Demand for Skilled Workers in Latin America." The World Bank. Sarquis, Sarquis J.B., and Jorge Saba Arbache (2002), "Human Capital, External Effects and Technical Change." London School of Economics & University of Brasilia. Sociedade brasileira de econometrfa (2000), 'The Brazilian Review of Econometrics", Revista de Econometria Volume 20 - Numero 1, SIN 0101-7012. Packard, Truman G. (2001), "Is There a Positive Incentive Effect from Privatizing Social Security?", World Bank Working Paper No. 2719, World Bank. The Brazil Jobs Report - 59 - Wagner de Alburquerque Oliveira, Carlos, and Francisco Galrao Carneiro (2001), "Flutuacoes de Longo Prazo do Emprego no Brasil: uma Analise Altemativa de Co- integracao." World Bank (2002), "Argentina: Labor Markets in the New Millennium", The World Bank, Confidential Report No. 19996 AR. World Bank (2002), "Building Institutions for Markets" World Development Report. The World Bank. World Bank (2002), "World Development Indicators" The World Bank. World Bank (2002), "Brazil Inequality Report", The World Bank. World Economic Forum (2001), "Global Competitive Report 2001-2002". Report No.: 24408 BR Type: SR