56809 FAST TRACK BRIEF June 29, 2009 The IEG report "Earnings Growth and Employment Creation: An Assessment of World Bank Support in Three Middle-Income Countries," was discussed by CODE on June 29, 2009 Earnings Growth and Employment Creation: An Assessment of World Bank Support in Three Middle-Income Countries This assessment reviews earnings and employment outcomes in Colombia, Tunisia, and Tur- key during 1998-2007, as well as five policy areas (the MILES framework) likely to affect those outcomes: macroeconomic conditions, investment climate, labor regulations, education, and so- cial protection. The three countries had very different experiences during the period: a deep re- cession during 1998-99 that increased cyclical unemployment in Colombia to unprecedented le- vels; a high unemployment rate that resisted efforts at reducing it for decades in Tunisia; and a sharp rise in structural unemployment in Turkey as it implemented structural reforms. Employment-related outcomes in the three countries were mixed, with notable progress in eco- nomic growth, earnings and poverty reduction, but not in the employment-to-population ratio or unemployment rate. This underscores the desirability of focusing on the full set of employment- related variables--GDP, poverty, employment, unemployment, and earnings--in an integrative fashion rather than just on employment when setting the objectives of Bank support. This focus will need better employment-related statistics, an area where the Bank can help further. Bank objectives in the three countries focused more on MILES components than on employ- ment itself. This suggests that employment issues, which straddle three Bank Networks (PREM, HD, and PSD), may fall by the wayside unless the Country Partnership Strategy serves as a vehicle for diagnosis, priority setting, and programming cross-sector work. Employment- related issues in particular--such as structural unemployment--are largely cross-sectoral in na- ture. From this perspective, the Bank could usefully strengthen its role as knowledge provider, broker, and facilitator of knowledge creation. Bank support in the three countries achieved differential progress in the individual MILES components, with the most progress on macro stabilization, followed in approximate order by progress on the investment climate, education, social protection, and labor taxation and regula- tions. The experience of the three countries illustrates how analytic and advisory activities can be the main instrument of support in those areas (such as labor regulations and taxes) where progress in reform is difficult and the need for building engagement and consensus is critical. D uring 1998-2007, Colombia, Turkey, and Tunisia faced difficult employment challenges, as shown in the figure below. Over the first half of the review period, Colombia and Turkey experienced sharp increases in unemployment rate as deep recessions reduced employment rent global crisis. The review looks at Bank support for the five areas embodied in the MILES framework developed by the Human Development (HD) and Poverty Reduction and Eco- nomic Management (PREM) networks: macroeconomic, in- vestment climate, labor market, education, and social protec- demand, with Colombia's unemployment declining and Tur- tion policies. These areas cover the key factors that affect the key's remaining higher as these economies recovered. De- demand for labor, the workings of the labor market, and the spite Tunisia's economic stability, it had the highest unem- supply of skills. While the demand for labor is expected to re- ployment rate of the three countries, although unemployment spond positively to economic growth, the two need not move declined moderately during the review period (Panel A). Fur- in tandem, as the sources of economic growth include growth thermore, the employment-to-population ratio ceased grow- in skills, in physical capital and in total factor productivity, in ing in Colombia during the review period after many years of addition to growth in employment. increases in female participation, and continued declining in Turkey. Tunisia's employment-to-population ratio increased Although the MILES framework was not explicitly used in by a few percentage points (Panel B) because female partici- formulating the objectives of the country programs reviewed in pation increased from a very low level. Earnings declined this report, the framework can be used to identify binding con- and poverty increased during the recessions in Colombia and straints to employment creation and earnings growth. It thus Turkey, but these improved toward the end of the review offers an organizing device for the evaluation, into which the period. Earnings in Tunisia increased and poverty declined. three country programs can be retrofitted and assessed. This approach thus differs in some (though not all) respects from the standard approach in many IEG evaluations of "evaluating Employment Performance in Colombia, Tunisia, against objectives," where the outcome of the Bank program is and Turkey assessed relative to objectives stated at the outset. In this case, A. Unemployment Rate: 1997-2007 although the programs in the three countries sought to help increase employment, the employment-related objectives were 20 not always formulated and unbundled into more proximate objectives. 15 Percent 10 The table below shows that the largest amounts of lending Year support in the three countries were for macroeconomic poli- 5 cies (including financial sector policies) followed by investment climate. 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 Colombia Tunisia Turkey Bank Lending (US$ million) and AAA (number of tasks) B. Employment-to-Population Ratio: 1997-2007 by MILES Area: 1998-2007 Country Colombia Tunisia Turkey 55 MILES Area Loans AAA Loans AAA Loans AAA Macro conditions 1,235 9 562 11 4,988 6 Investment climate 1,100 4 436 17 1,961 7 Percent 50 Labor regulations 45 857 6 0 11 500 7 Education policy 662 3 385 10 704 2 Social 40 protection 341 4 0 3 1,715 1 Year 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 Total 4,195 26 1,383 52 9,868 23 Source: Appendices D, I, and N. Colombia Tunisia Turkey The following paragraphs briefly present the Bank's support and its outcome. Sources: The World Bank (World Development Indicators); Colombia Statistical Office (DANE); Turkey Statistical Office (Turkstat). Colombia This report evaluates World Bank support for employment and Throughout the period, the Bank was engaged in the five earnings growth in Colombia, Tunisia, and Turkey during areas of the MILES framework in varying degrees. Following 1998-2007. It aims at drawing findings that may help guide the the 1998-99 crisis, the Bank supported the government's ef- Bank's work in this area, including its response during the cur- forts to strengthen fiscal performance (revenue collection, 2 efficiency in the provision of health services, contingent lia- · Intermediate outcomes: the restoration of fiscal stability, sub- bilities, public administration, and public expenditure and stantial progress in the investment climate, gains in the debt management) and the financial system. These improve- coverage and financing of education, and broader cover- ments were critical in redressing growth and earnings and in age of the safety net. dampening a surge in cyclical unemployment. They did not, · Bank interventions: significant Bank contributions in tax however, achieve a reduction in structural unemployment, administration, bank restructuring, business climate im- which remained high due, inter alia, to an excessive minimum provements, labor market diagnostics, coverage and fi- wage and high payroll taxes. nancing of education, and design of the conditional cash transfer (CCT) system. There were shortfalls in resolving Beginning in 2002, reflecting the government's priorities and Colombia's revenue and expenditure rigidities, as well as concerns that the economic recovery was slow to yield better in lowering the minimum wage and the burden of non- employment and social outcomes, especially for the poor, wage costs, relative to objectives. There was also lower there was increased emphasis by the Bank on improving growth and higher structural unemployment. competitiveness (through investment climate strengthening and labor market reforms) and on improving the social risk Tunisia management system. In Tunisia, the Bank was also engaged in the five pillars of the MILES framework, though there were important differences Bank-supported objectives in the MILES areas were largely of emphasis. In macroeconomic policy, where performance achieved, although the Bank over-promised on the elimina- was the best among the three countries, the Bank's role was tion of fiscal structural rigidities given the political economy positive, albeit marginal on the fiscal front. It was most im- constraints, and did not adequately factor in the fiscal impli- portant on trade liberalization and facilitation and financial cations of the conditional cash transfer program. Nonethe- sector reform, both of which benefited from AAA and lend- less, Colombian counterparts acknowledged the contribution ing, including high-quality diagnostic financial sector work of the Bank in several areas, such as tax administration. In undertaken jointly with the IMF. Regarding the investment some areas the Bank missed certain opportunities to use its climate, Bank support addressed important reforms, including convening power. A number of consultations suggested, for a reduction in the regulatory gap that favored the offshore example, that the Bank (as well as the IMF) had been too over the onshore economy, improving competition law, timid in pressing the policy implications of its labor market opening the information and communications technology diagnostics (particularly on the minimum wage and labor tax- sector, and developing the regulatory framework in infra- es), which were of good quality. In general, the Bank's inter- structure. ventions were relevant and AAA, technical assistance, and lending operations adapted to the government's changing The Bank had little impact on policies governing the labor priorities. market. The Bank conducted AAA of good quality, but the government did not adopt any of the Bank's key recommen- Bank-supported policies improved employment performance. dations. In education, a great deal was achieved at the prima- The improvements in macroeconomic policy and in the in- ry and secondary levels, and the Bank played a major role in vestment climate helped increase average yearly GDP growth this. The results of Bank support were more muted at the from 0.5 percent during 1997-02 to 5.6 percent during 2002- tertiary level, where there was less-than-full agreement on 07, primarily through an increase in total factor productivity some key policy issues (such as financing and cost recovery (TFP) growth. This allowed the unemployment rate to de- reforms) and it is believed that aggressive expansion in cover- cline and earnings to grow despite a slowdown in employ- age has been responsible for a deterioration of quality. ment growth. The key policies that improved GDP, employment, and earn- IEG rates the outcome of the Bank program in Colombia ings growth in Tunisia were macroeconomic stabilization and with respect to progress in employment and earnings growth trade integration with the EU, the latter largely established as moderately satisfactory. Key achievements can be summarized before the review period. Improvements in the investment as follows: climate may have further contributed to the growth in earn- ings suggested by rising formal sector wages through im- · Higher-order outcomes: increased earnings growth and lower provements in productivity growth. These improvements poverty with stable employment growth and mitigation have not yet had a discernible effect on investment levels. of cyclical unemployment, particularly during the second While education has contributed to rising productivity and part of the review period. earnings, it has also led to increased unemployment among the educated due to both skills mismatches and low invest- 3 ment levels. Unemployment has been persistently high, but ment in the education sector was less successful than that in declined moderately during the review period. Inadequate the other areas relating to employment. skills appear to be behind the high level of unemployment. The latter may also reflect the effect of minimum wages. Bank-supported policies, particularly those in the macroeco- nomic area, helped reignite growth and the demand for em- The outcome of the Bank program in Tunisia relating to em- ployment, with the employment ratio stabilizing somewhat ployment and earnings growth is rated satisfactory. Key after a long-term decline. Higher productivity triggered by achievements can be summarized as follows: some of the supported reforms helped increase earnings of the poor. On the other hand, structural adjustment, particu- · Higher-order outcomes: increased earnings growth and lower larly migration away from agriculture, continued to decrease poverty with stable employment growth, lower structural labor participation, while increasing productivity. On bal- unemployment, and increased GDP growth. ance, average earnings declined until about 2003 and then · Intermediate outcomes: sustained macroeconomic stability, recovered. significant progress in trade facilitation and the business environment, substantial progress in the investment cli- The outcome of the Bank program in Turkey with respect to mate, and gains in coverage of education. employment and earnings is rated moderately satisfactory. Key · Bank interventions: significant Bank contribution to trade, achievements can be summarized as follows: the investment climate, and coverage in education. How- ever, there were shortfalls in making labor markets more · Higher-order outcomes: increased earnings, lower poverty, flexible, in improving education quality, and in streng- and higher GDP growth. thening public sector governance. · Intermediate outcomes: the restoration of fiscal stability and gains in the coverage of education. Turkey · Bank interventions: significant Bank contributions towards In Turkey, the Bank was heavily engaged in the five areas of macroeconomic stability and the CCT program. There the MILES framework to varying degrees. In the earlier were shortfalls in the effectiveness of support on the years, the Bank concentrated on macroeconomic stabilization business climate, in lowering non-wage costs, and in re- and accompanying structural reforms. Bank support was laxing labor regulations (though some progress was delivered primarily through policy advice and development achieved the year after the end of the review period). Ef- policy lending. Following the 2001-02 crises, the Bank sup- fectiveness in contributing to education gains was also ported reforms to strengthen fiscal performance and the fi- limited. Employment growth declined and unemploy- nancial system. This set of reforms was critical in redressing ment increased. economic growth and earnings growth in Turkey. But they did not help reduce unemployment, which in fact increased. Findings In later years, greater attention was devoted to employment Three sets of findings are worth highlighting. The first set and competitiveness issues. Within the scope of the MILES concerns outcomes. In the three countries, progress in eco- framework, the remaining problem areas are labor market nomic growth and earnings was notable but not in the em- regulations, education, and the investment climate. ployment-to-population ratio or the unemployment rate. This finding underscores that higher economic growth does Bank-supported objectives were largely achieved. Stabilization not always lead to lower unemployment. It also underscores and related structural reforms were achieved, although the the need to consider the full set of employment-related out- support was initially unsuccessful because weak macroeco- comes, not just employment, wherever employment objec- nomic stabilization undermined financial sector reforms. On tives feature prominently in Bank strategies and programs. investment climate matters, active Foreign Investment Advi- The full set would cover employment growth, the employ- sory Service (FIAS) participation supported important ment-to-population-ratio, the unemployment rate and earn- changes in foreign direct investment legislation. The more ings growth, as well as the impact of these on poverty and the recent Investment Climate Assessment (ICA) has, however, uncertainties of that impact. This means that policies and been less successful thus far and would benefit from stronger programs need to be understood in terms of their possible client engagement and dissemination. On labor policy issues, joint effects on those variables. As monitoring employment- the Bank provided high-quality advice, although this did not related outcomes requires adequate data, the Bank can pro- translate into policies to ease the rigid regulations and high vide a valuable service by continuing to help client countries labor taxes that had been diagnosed. On education, the first to reduce the long lags with which critical data on employ- operation was unfocused and suffered from procurement ment performance (for example, earnings) are produced and problems, but follow-on lending fared better. Despite these disseminated. operations and several pieces of AAA, the Bank's engage- 4 A second set of findings refer to Bank country strategies and lombia and Turkey and to trade and financial development in programs. In the three countries covered in this evaluation, Tunisia. These efforts were the main factor that helped re- the Bank primarily targeted multiple objectives with a bearing store GDP growth in the first two countries and also contri- on employment (for example, objectives relating to MILES buted to sustaining higher growth in the third. Restored components) rather than explicit employment outcomes. GDP growth increased the demand for labor, resulting pri- This points to the usefulness of situating Bank support for marily in the higher earnings growth documented earlier. employment and earnings growth within an integrative Bank support for investment climate reforms covered the three framework. For client countries where employment-related countries and contributed to improvements in specific policy objectives underlie Bank support, the Country Partnership areas that helped raise investment or total factor productivity Strategy takes on added importance as a vehicle for diagnosis, to varying degrees. priority-setting, and cross-sector work on employment issues, which straddle three networks (PREM, HD, and PSD) and Support for labor market regulations and institutions oc- may not otherwise get the requisite attention. curred primarily through AAA that was relevant and techni- cally competent, but that had unclear impact. Labor rigidity AAA has been an important component of the Bank's assis- and labor taxes remained a constraint to employment, par- tance on employment and earnings. Although the impact that ticularly formal employment, in the three countries. Support it had is unclear, the Bank's AAA was generally of good quali- for education contributed primarily to extensions in access and ty, addressing the right employment issues competently, and to an increased supply of the skills needed to raise output and it was well appreciated by the three clients. This experience earnings growth, but the impact on quality was more limited. suggests that where it is judged that employment-related out- And finally, support for social protection arrangements contri- comes are central to country strategies, it would be advisable buted successfully to conditional cash transfer (CCT) pro- to deploy AAA in the critical areas, including: understanding grams in Colombia and Turkey and less so to the unemploy- and addressing the roots of unemployment, including its cyc- ment insurance programs that are needed to move from job lical, classical, frictional, and structural components; and as- protection to worker protection. The three countries' expe- sessing labor market regulations and institutions and their riences attest to the difficulties of setting up unemployment effects. In particular, the Bank can help improve and extend insurance systems. the coverage of estimates and cross-country comparisons of labor rigidities and labor taxes. From a more general cross-cutting perspective, the Bank could usefully strengthen its role as knowledge provider, bro- In some cases the Bank was not sufficiently engaged on labor ker, and facilitator of knowledge creation. The Bank is in a market issues. The experience of the three countries shows good position to exercise leadership on employment issues as varying degrees of engagement, which in turn are likely to be it has demonstrated capacity to use research for developing driven by differences in the political economy of labor reform operationally relevant policy. In this regard, AAA, not lend- and in Bank proactivity regarding engagement on such issues. ing, is often the best instrument to deal with the more sensi- Where labor reform issues are deemed important to its coun- tive policy issues that affect employment outcomes. As with try strategy, the Bank would be well advised to conduct the the three countries reviewed in this report, the Bank can con- required AAA and find the right windows of opportunity to tribute to understanding employment issues where those is- support the government (as was the case in Turkey). sues are a central concern. This would include advising coun- tries how to use available resources judiciously to address All in all, sustained macroeconomic and financial stability in various employment challenges (such as unemployment in- Tunisia and stabilization in Colombia and Turkey as well as surance) while ensuring macroeconomic stability. It would higher education levels explain increased or recovered eco- also include looking at education quality, possibly with greater nomic growth and earnings in the three countries, with some emphasis on addressing skills mismatches, which may be be- additional role attributed to improvement in the investment hind much of the structural unemployment observed in coun- climate. However, progress in these three areas did not re- tries like the three reviewed in this report. duce unemployment rates in Colombia and Turkey and re- duced them only moderately in Tunisia. High remaining un- The Current Global Jobs Crisis employment is likely to reflect skill mismatches, rigid wages, or other structural conditions. The findings presented in this report suggest ways in which the Bank can strengthen its response to the current global A third set of findings concerns the specific areas of support. economic crisis. The crisis has reduced output growth from Bank support for macroeconomic policies contributed effectively 5.2 percent in 2007 to -1.3 percent in 2009 and increased the to country efforts on macro-financial crisis resolution in Co- world unemployment rate from 5.7 percent in 2007 to 6.8 percent in 2009, with the higher increases in developed econ- 5 omies, Central and South Eastern Europe (non-EU) and CIS, and Latin America and the Caribbean. The findings suggest the following three directions: About Fast Track Briefs Fast Track Briefs help inform the World Bank Group (WBG) Country programs can respond by focusing on both em- managers and staff about new evaluation findings and recom- ployment and earnings outcomes, which determine po- mendations. The views expressed here are those of IEG and verty outcomes. Country Directors are well positioned to should not be attributed to the WBG or its affiliated organiza- deploy the Bank's multisector capacity to engage with client tions. Management's Response to IEG is included in the pub- countries on how policies and adjustments to Bank programs lished IEG report. The findings here do not support any general can help improve the full set of employment-related out- inferences beyond the scope of the evaluation, including any infe- comes, not just employment. Such engagement will require rences about the WBG's past, current or prospective overall an understanding of how policies in different areas (including performance. macro, labor market) can redress both employment and earn- ings in specific countries, presumably through increased de- mand for labor and skills and through higher labor productiv- The Fast Track Brief, which summarizes major IEG evalua- ity. Many governments have made adjustments to their tions, will be distributed to selected World Bank Group staff. If policies in response to the financial crisis and/or may need you would like to be added to the subscription list, please email well-tailored advice in these areas. Diagnoses and recommen- us at ieg@worldbank.org, with "FTB subscription" in the dations are likely to differ according to specific country cir- subject line and your mail-stop number. If you would like to cumstances, such as the degree of openness of the economy. stop receiving FTBs, please email us at ieg@worldbank.org, with "FTB unsubscribe" in the subject line. The Bank's effectiveness will be stronger as it integrates across networks to focus on those issues that are likely to be relevant during crises. Examples include improving employment and earnings monitoring; assessing the impact of Contact IEG: global recession on structural unemployment; developing Director-General, Evaluation: Vinod Thomas strategies to promote job flexibility and worker protection; Director: Cheryl Gray (IEGWB) finding the right balance between macroeconomic stability Manager: Ali M. Khadr (IEGCR) and the fiscal cost of other employment-related policies; Task Manager: Mauricio Carrisoza (IEGCR) strengthening the interaction between macroeconomic and microeconomic conditions affecting growth and employment demand; understanding the interface between quality of edu- Copies of the report are available at: cation and absorption of labor services; and assessing the http://www.worldbank.org/ieg/ possible effect of differential changes in employment and IEG Help Desk: (202) 458-4497 earnings on migration. E-mail: ieg@worldbank.org Country programs will be more effective as they focus on areas where progress in the enabling conditions has been slow but where windows of opportunity for reform surface. The area where progress has typically been slow is on making labor regulations more flexible and lowering labor taxes. Conversely, the Bank can also help countries maintain progress where it has been achieved, as was the case with macroeconomic stability in the three countries reviewed. In this regard, the Bank can deploy its multisector capacity to help countries avoid conflicts among employment policies (for example, ensuring that stimulus packages or extended social protection are affordable). 6