77976 ResearchDigest World Bank VOLUME 7 L NUMBER 1 L FALL 2012 Benchmarking Global Poverty Reduction Goals for future poverty reduction when the corresponding poverty rate IN THIS ISSUE should be challenging, but not was more than 50 percent. In the 20 Benchmarking Global Poverty impossible years from 1990 to 2010 the developing Reduction … page 1 world cut its overall poverty rate by Setting goals for poverty reduction helps I half, from 43 percent to 21 percent. in both motivating efforts and assessing nternational development institu- The optimistic trajectory would entail performance. What is achievable in the tions and almost all developing halving it again in another 10 years. next decade? countries track poverty measures, A very ambitious but still imagin- and the results are watched keenly. But able benchmark for even better perfor- Rethinking the State’s Role how do we judge performance? When mance would be a 3 percent poverty in Finance … page 2 do we say that it is good or bad? Little rate by 2022. This could be achieved What is the state’s proper role in financial attention has been given to these by a 1 percentage point higher growth development? The inaugural edition of a questions. rate than we have seen recently, but new World Bank series takes up the A closely related issue is how to it would also require that there be no question set goals for future performance. further deterioration either in overall Coordinating Housing and Transport Setting goals can motivate achieving inequality or in the household sector’s Policies in Urban South Africa … page 3 them. But to motivate extra effort, the (survey-based) share of national Getting housing and transport policies goals that are set must represent real income. right can give people greater choice in progress but not be impossible. Nor Alternatively, moderately lower in- where to live can they be too easy. equality (achieved by bringing overall A new paper by Ravallion draws on inequality back to its level in 1999) Making Public Sector Reforms recent research on the measurement could attain the ambitious target Work … page 4 of poverty and on global economic without higher growth. While the 3 More effective support for public sector prospects to try to identify measures percent target for 2022 is clearly ambi- reforms requires a better understanding of and seemingly defensible benchmarks tious, it appears to be within the range the role of political economy factors for judging progress going forward. of experience for what is possible with Helping Women Get to Work … page 5 The measures proposed span both a concerted effort toward equitable absolute and relative poverty. The growth. How effective are interventions aimed at improving women’s employability and benchmarks for each rest on explicit Allowing for social inclusion, the quality of work? A new paper reviews the assumptions about future growth and optimistic target for 2022 implied by evidence distributional change, though they are the past trajectory would be to ensure assumptions informed by knowledge that no more than 40 percent of people Export Superstars … page 6 of recent past performance against in the developing world are either How much do “superstars” drive trade poverty and current expectations about absolutely poor or poor by standards patterns in developing countries? And growth prospects across the develop- typical of the country they live in. The how do they become superstars? ing world. ambitious target corresponding to the For absolute poverty the proposed 3 percent target for absolute poverty Why Did Inequality in Latin America optimistic benchmark for the $1.25 would be to ensure that at least two- Decline in the 2000s? … page 7 poverty rate in 10 years’ time is 9 per- thirds of the developing world’s A new study looks at factors contributing cent. Achieving this would clearly be a people are neither absolutely poor nor to changes in inequality in Argentina, very good performance—maintaining relatively poor in 10 years. Brazil, and Mexico an impressive trajectory of poverty re- duction going back to the early 1980s, (continued on page 8) 2 World Bank ResearchDigest Rethinking the State’s Role in Finance What should be the role of the state involvement. This does not mean a The third area relates to direct in financial development? What withdrawal of the state from oversee- interventions, such as lending by lessons can we learn from the global ing finance. On the contrary, the state state-owned banks. The report finds has a very important role, especially in that such lending can help stabilize financial crisis? providing strong supervision, ensuring aggregate credit in a downturn, but it T healthy competition, and enhancing can also lead to resource misalloca- he global financial crisis has financial infrastructure. tion and deterioration in the quality prompted many to reassess The report’s underlying message of intermediation. The report presents state interventions in financial is that in the financial sector incen- evidence that lending by state-owned systems—from regulation and super- tives are crucial. The main challenge banks tends to be less procyclical and vision of financial institutions and of financial sector policies is to better that some state-owned banks even markets to state guarantees and state align private incentives with public played a countercyclical role during ownership of banks. But the crisis interest without taxing or subsidiz- the global financial crisis. But the track does not necessarily negate the body ing private risk taking. The design of record of state banks in credit alloca- of evidence on these topics accumu- public policy needs to strike the right tion remains generally unimpressive, lated over the past few decades. It is balance—promoting development, yet undermining the benefits of using important to use both the crisis and in a sustainable way. This approach them as a countercyclical tool. Policy precrisis experience to examine what leads to challenges and trade-offs. The makers can limit the inefficiencies as- went wrong and how to fix it. This is report studies these in depth, focusing sociated with state bank credit by pay- the motivation of the World Bank’s on four main areas. ing special attention to the governance Global Financial Development Report 2013: The first area of focus is regulation of these institutions and schemes Rethinking the Role of the State in Finance, and supervision. The report examines and ensuring that adequate risk the inaugural edition of a new World new evidence from the crisis using a management processes are in place. Bank series. unique survey covering some 730 fea- This oversight is challenging, how- The report reexamines a basic ques- tures of the regulatory framework in ever, particularly in weak institutional tion: What is the proper role of the 143 jurisdictions. One finding is the environments. state in financial development? To ad- importance of first getting the “basics” Finally, the report examines the dress this question, it synthesizes new right—through strong, timely, and an- role of the state in financial infrastruc- and existing evidence on the state’s ticipatory supervisory action comple- ture, particularly credit information performance as financial sector regula- mented by market discipline. In many sharing systems. Evidence points to tor, overseer, promoter, and owner. It developing economies this means giv- a useful role for the state in promot- builds on novel data, surveys, research, ing priority to building up supervisory ing transparency of information and and wide-ranging country experience, capacity. Less can mean more: less reducing counterparty risk. The report with an emphasis on emerging market complex regulations, for example, can shows that the state can facilitate the and developing economies. The report allow more effective enforcement by inclusion of a broader set of lenders in also tracks financial systems in more supervisors and better monitoring by credit reporting systems and promote than 200 economies before and during stakeholders. the provision of high-quality credit the global financial crisis. The second area is competition information, particularly where there The report strikes a cautionary policy. The crisis fueled criticisms are significant monopoly rents that chord. It acknowledges that the global that “too much competition” in the discourage information sharing. financial crisis has given greater cre- financial sector led to instability. But dence to the idea that direct state research presented in the report sug- involvement in the financial sector, gests that for the most part instability such as lending by state-owned banks, is due to other factors, such as a poor can help maintain economic stability, regulatory environment and distorted drive growth, and create jobs. And it risk-taking incentives. With good regu- provides evidence that some of these lation and supervision, bank competi- direct interventions may have had an tion can help improve efficiency and World Bank. 2012. Global Financial Devel- impact, at least in the short run. But enhance access to financial services opment Report 2013: Rethinking the Role it emphasizes that there is also evi- without necessarily undermining sys- of the State in Finance. Washington, DC: dence of potential longer-term nega- temic stability. Rather than restrict- World Bank. tive effects. The report suggests that ing competition, the state needs to For data sets on financial development, financial as the crisis subsides, there may be encourage contestability, improve the regulation and supervision, and state-owned a need to adjust the role of the state flow of information, and strengthen banks, visit http://www.worldbank.org/ from direct interventions to less direct the contractual environment. financialdevelopment. World Bank Research Digest 3 Coordinating Housing and Transport Policies in Urban South Africa People often choose dwellings for investments in housing maintenance access to land and housing closer to their proximity to jobs—but land or upgrading. Why? Households that job opportunities in urban centers. markets and transport infrastructure receive RDP housing are often dis- Many households live in informal satisfied with their dwellings because settlements not because of afford- can be major constraints they are far from employment centers; ability but because of the lack of suit- T the new houses were often built in the able small plots of land in the formal he United Nations projects “old” apartheid locations (deliberately market. that 2 billion people will live in sited far from urban centers and white In light of these findings, the au- slums by 2030. How to manage neighborhoods). In addition, house- thors suggest five policy innovations: slum formation and holds rarely receive title • To remedy the location of the reduce the hazards deeds immediately but RDP-subsidized houses, allow eligible faced by slum dwellers? To undo the are instead allocated individuals and groups to use grants Most policy discus- “geography of houses administratively, to buy land and build or upgrade their sions tend to focus with the condition that own houses or to buy already-built on moving people to apartheid,” South the houses cannot be houses. safe environments or Africa needs to find sold or rented for a • Increase the flexibility of the providing better hous- period of eight years, housing grant structure in such a way ing elsewhere—such ways to enable poor recently reduced to five. that beneficiaries themselves are able as through urban up- workers to gain Even if they do receive to make a trade-off between location grading, resettlement, title, this condition is and value of the housing. Some would housing subsidies, and access to land and attached. And even after opt for a location closer to work while land titling. But many housing closer to jobs five years the govern- others would prefer more housing of these policies do not ment has a right of first value in a location farther away. work. The reason is that in urban centers refusal on the sale of • To create more incentives to im- people do not always the property. So poor prove and invest in the housing asset, willingly trade away a better loca- location and lack of fully tradable substantially reduce or eliminate the tion for a better home with modern property rights limit the benefits of the “no sale and no rental” and “right of utilities. program. first refusal” period. People choose neighborhoods To complement the view from • Increase the supply of subsidized for their affordable services and their the supply side, the authors fielded land and houses by making it easy for amenities—but also for their proximity a small pilot survey in an informal developers and commercial banks to to jobs. But in many developing coun- settlement in Cape Town to investigate provide “sites and services” or fully try cities it can be difficult to live near how much the poor actually pay for the built houses by using them as agents one’s job because land markets have land on which they have their dwell- to manage and disburse the grants. failed: formal housing supply is low, in ings and their willingness and ability • Remove the many land use and part because of restrictive regulations. to pay for different quality of housing. housing development regulations that But it may also be difficult and costly The survey revealed an active informal date to the apartheid era—regulations to commute to work because transport market for land and housing. Among that had the explicit objective of bar- infrastructure fails to connect urban the respondents, 66 percent said that ring blacks from well-located land and neighborhoods. they had paid for, or were renting, the preventing them from accumulating A new paper by Lall, van den Brink, land on which they had built their capital and that continue to have the Dasgupta, and Muir Leresche takes dwelling. Respondents reported pay- originally intended effect. a close look at the interplay between ing prices for their small plots that, urban regulations and public sector on a per-hectare basis, are similar to housing development in South Africa those found in the upscale market for to examine the impact on living stan- undeveloped land. Collectively, then, dards for the beneficiaries of the gov- the very poor could compete with the ernment’s flagship housing program, high end of the property development the Reconstruction and Development market. But land zoning regulations Somik V. Lall, Rogier van den Brink, Basab Das- Program (RDP). and subdivision laws do not allow the gupta, and Kay Muir Leresche. 2012. “Shelter Comparison of RDP beneficiaries supply of such small plots. from the Storm—but Disconnected from Jobs: Les- and similar nonbeneficiaries does To improve the equity and efficiency sons from Urban South Africa on the Importance not show that public housing provi- of cities and undo the “geography of of Coordinating Housing and Transport Policies.” sion generated multiplier effects by apartheid,” South Africa needs to find Policy Research Working Paper 6173, World stimulating complementary private ways to enable poor workers to gain Bank, Washington, DC. 4 World Bank ResearchDigest Making Public Sector Reforms Work Political economy factors can make decade, indicating a significant inter- make reform achievements more ro- providing effective support for public est in such reforms. Only a quarter bust and more likely to be sustained. sector reforms challenging. What of the countries borrowing for public The paper explores efforts made by sector reforms can be considered aid Bank staff to understand levels of gov- are some solutions? dependent (with aid accounting for ernment commitment and incentives S more than 10 percent of gross national and examines whether more intensive upporting effective public sec- income). Among those with multiple efforts to do so appear to pay off. Staff tor reforms is a big challenge for public sector operations, many are typically seek to understand commit- development agencies. Part of middle-income countries. Global inte- ment, but in most cases such efforts the reason for this is the crucial role gration appears to be the most com- are limited and implicit. There are that political economy factors play in mon driver of public sector reforms. institutional disincentives in the Bank such reforms. A new paper by Bunse World Bank support for public sec- to fully acknowledge risks, especially and Fritz examines this role. The paper tor reforms typically involves areas in country contexts that make success reviews the existing literature to iden- that are narrow and technical (for more uncertain. There is some tenta- tify key propositions about factors that example, revising a chart of accounts tive evidence that operations designed trigger or delay reforms. It also draws or introducing a medium-term expen- with greater explicit awareness of gov- lessons from World Bank public sector diture framework) as well as areas that ernance and political economy drivers operations over the decade 2000–10. are more politically salient or sensitive have been more successful in imple- Political economy incentives to (such as strengthening external audit, mentation, but the number of opera- initiate public sector reforms—such as which is aimed more directly at reduc- tions for which this can be assessed is regional and global trade integration, ing rent-seeking behaviors). Nearly small. fiscal pressures, and, for lower-income three-quarters of all Bank-supported The findings suggest a need to countries, donor demands—are pres- public sector reform projects receive differentiate more clearly between ent in a wide range of countries. But an internal rating of moderately suc- country contexts, so as to develop an disincentives to implement more chal- cessful or better upon completion. explicit understanding of the potential lenging aspects of public sector re- This is higher than the literature would for reform and the risk involved. Based forms are also widespread. Economic expect. Well-performing operations on this understanding, choices and and political conditions for successful are more common in countries with approaches should be calibrated—by large-scale public sector reforms— great­er prior government effective- designing approaches that seek a “fea- including a process of economic ness (as rated by the World Bank’s sible stretch” in environments offering transformation and the presence of Worldwide Governance Indicators) and some opportunity; focusing on incre- programmatic parties or otherwise worse-per­ forming ones at the medium mental but tangible reforms within strongly motivated leadership—are to low end of prior effectiveness. But limited windows of opportunity (such less frequent. moderately successful operations are as the years between elections) while Consequently, while governments found across a wide range of country avoiding wasting time on large-scale in many developing countries face contexts, including postconflict coun­ or complex reform designs; balanc- incentives to initiate public sector tries with the most lim­ ited prior gov- ing reforms that are more technically reforms, at the implementation stage ernment effectiveness. focused with reforms that “bite” and political costs often outweigh poten- Political economy obstacles fig- are politically sensitive (for example, tial gains—and reforms are therefore ure prominently in ex post reviews of by directly addressing rent seeking); abandoned or left to wither. Real Bank-supported public sector reform and working with progressive local breakthroughs have been achieved in projects. For operations whose per- stakeholders to strengthen pro-reform countries experiencing major struc- formance was assessed as moderately incentives. Given the urgent need to tural shifts (such as Brazil) and those unsatisfactory or worse, they are the enable better service delivery, the pa- with a political leadership committed most frequently cited source of failure. per advocates continuing to seek pub- to higher-level goals. In addition, in a In addition, many ex post reviews of lic sector reforms even in unpropitious number of countries incremental gains operations raise concerns about the environments. are observable even if broad, overall sustainability of their achievements reform goals were not achieved (such because of political economy disincen- as in Ghana). tives. Currently, there is no tracking The review of World Bank op- over time to see whether achievements Simone Bunse and Verena Fritz. 2012. “Making erations yields interesting findings. are in fact sustained. Moreover, reviews Public Sector Reforms Work: Political and Eco- Nearly 70 percent of all potential cli- do not specifically address whether nomic Contexts, Incentives, and Strategies.” Policy ent countries borrowed from the Bank and how political economy challenges Research Working Paper 6174, World Bank, for public sector reforms over the could have been better managed to Washington, DC. World Bank Research Digest 5 Helping Women Get to Work Many programs have sought to Romania, the Russian Federation, women’s working behavior by affect- improve employment and earnings and the Slovak Republic there is some ing the availability and pricing of child outcomes for women. Which are evidence suggesting that active labor care. A review of child care programs in market programs have helped work- Argentina, Colombia, and Guatemala effective? ers find jobs more quickly but at the found that the availability of child D expense of lower wages than they care had strong effects on mothers’ eveloping countries have tried would otherwise have been able to rates of working and on the number a range of policy interventions obtain searching on their own. Highly of hours worked. The programs also aimed at improving women’s educated workers on the whole did not had substantial positive effects on employability and quality of work— benefit much from these programs. the nutrition and development of the including active labor market pro- The survey of microfinance pro- young children participating in them, grams, microfinance programs, condi- grams found mixed evidence on their suggesting high benefit-cost ratios. In tional cash transfer programs, educa- effectiveness and on the benefits of Accra, Ghana, however, it was found tion and training programs, entrepre- gender-targeted lending. Many studies that the local supply of child care was neurship and leadership programs, document high rates of return to such not a significant determinant of moth- and programs that facilitate work programs, but the returns can be com- ers’ working behavior. One reason is (child care subsidies, parental leave promised by targeting the loans to nar- that mothers worked mainly in the programs, land titling programs). rowly specified groups. In implementa- informal sector, where it was easier to Common objectives are to reduce un- tion, the relative redistributive benefits combine work with child care. employment, increase wages, provide of targeting loans to women or to poor By contrast, in formerly communist social protection, and increase wom- borrowers need to be weighed against countries that have seen enormous en’s empowerment in the household. the benefits of allocating funds to the changes in the costs of child care— How effective are such programs? A highest-return activities. Using a theo- such as Romania and Russia—there is new survey paper by Todd reviews a retical model, one study analyzes the strong evidence of these costs having a subset of recent programs that have relative effectiveness of (non-gender- substantial effect on working behavior. been subjected to rigorous impact targeted) microcredit and cash trans- In some countries child care went from evaluations and were aimed at directly fers in increasing the consumption and being free and widely available to be- affecting employment and earnings utility of poor people. It concludes that ing expensive and a major determinant outcomes. a microcredit program was more effec- of whether women worked. Two stud- Active labor market programs are tive than a conditional cash transfer ies find women’s labor supply to be often adopted as a way to ease the program in increasing consumption, fairly elastic with respect to the price effects of economic shocks. These because the credit was used for pro- of child care. An important finding is programs are not usually targeted at ductive purposes. that child care subsidies were more women, though women often seek out In recent years conditional cash effective than wage subsidies or family the employment and training services transfer programs have become a com- income subsidies in increasing family they offer. On the whole, the evidence mon strategy for reducing poverty and income levels, in part because of bet- suggests that in Latin America many encouraging investment in children’s ter targeting of women whose labor of these programs have been effec- schooling. Many give the transfers to force behavior was potentially affected tive in reducing labor market frictions mothers, under the assumption that by the policy. and facilitating firm-worker matches, they are more likely to spend the trans- though not very effective in increasing fers on children and that providing worker productivity. But it is perhaps transfers to women improves their rel- unsurprising that productivity cannot ative bargaining position in the house- be greatly enhanced through relatively hold. The effects of such programs on short-term interventions. women’s labor force participation have Only a few active labor market pro- not yet been systematically studied, grams have been subjected to rigorous particularly in the urban context. In evaluation of rates of return, making urban areas of Mexico, however, there it difficult to assess which have gen- is some evidence that some families erated benefits that exceed program did not take up the program in part be- costs. One study finds that program cause working mothers were unable to Petra Todd. 2012. “Effectiveness of Interventions impacts have to be sustained over the meet the time requirements of partici- Aimed at Improving Women’s Employability and longer term (nine years or more) for pation (such as attending meetings). Quality of Work: A Critical Review.” Policy Re- the program to be cost-effective. In the Another class of policies reviewed search Working Paper 6689, World Bank, Wash- transition economy settings of Poland, are those designed to influence ington, DC. 6 World Bank ResearchDigest Export Superstars The top 1 percent of exporters that revealed comparative advantage enter the export sector and become drive comparative advantage and is driven more by having a few giants superstars without a long period of export growth and diversification in than by having more firms. learning. Second, the paper shows the The data do not permit system- developing countries importance of superstars for export atic examination of superstars before L growth and diversification. Over the they began exporting to shed light on arge firms define exports. latest period of three consecu­ tive years their potential previous experience Consider such well-known for which data are avail­ in domestic markets.   examples as Nokia in Finland, able, superstars account Export superstars But for three countries Samsung in the Republic of Korea, for more than half of where the authors could and Intel in Costa Rica. Each of these total export growth as are unique—they identify the superstars accounts for around 20 percent of its well as more than half are born as large by name (Jordan, Peru, country’s total exports. But patterns of the growth driven by and Tanzania), they are not that different elsewhere. On product markets new to exporters rather than researched the origins average across 32 developing coun- the country (the exten- becoming superstars of these superstars to tries, the top firm alone accounted sive margin) during that understand how they for almost 15 percent of total (nonoil) period. This suggests by learning through started as exporters. exports between 2006 and 2008. The that stimulating trade domestic production This made it possible to top 1 percent of exporters accounted growth and diversifica- determine whether the for 53 percent of exports on average tion depends largely on or exporting firms grew slowly in the during that period. The remaining vol- creating an environment domestic market before ume of trade was concentrated mainly where large firms can thrive. becoming exporters and whether they in the next tier of large firms: the top Third, the paper explores the ori- are domestically or foreign owned. 5 percent of exporters accounted for gins of new superstars—defined as It also alleviates potential concerns almost 80 percent of exports on aver- firms that entered the export sector about traders (nonproducers) in the age, and the top 10 percent for almost during the period for which data are sample. The authors find that the ma- 90 percent. available and grew to be superstars jority of superstars are foreign owned Using a novel firm-level panel, by the end of the period. Using the and began operations as exporters and based on highly disaggregated cus- most recent period of three consecu- that only a very small fraction are ex- toms data across all regions, a new tive years for which data are available clusively traders. paper by Freund and Pierola exam- for 18 countries, the authors find This finding further supports the ines the role of these “superstars” in that more than 80 percent of the new argument that superstars are unique— defining trade patterns and uncovers superstars entered the export sector they are born as large exporters rather their origins. The paper makes several very large—in the top 5 percent of than becoming superstars by learning contributions to the literature on firm- exporters. through domestic production or ex- level analysis of trade. For three countries where the time porting. Coupled with the large foreign First, the paper reveals the im- series allow analysis over a decade ownership share in superstars, this portance of export superstars—the (Costa Rica, Morocco, and Peru), the finding also highlights the role of mul- top 1 percent of firms—in defining authors also find that superstars are tinationals in exports. comparative advantage. The authors born relatively large and grow quickly decompose the variation in sectoral into the top 1 percent. Among the new export shares across countries and in- exporters that became superstars dur- dustries into the part due to firm size ing the decade, more than half entered (intensive margin) and the part due the export sector in the top 5 percent to firm count (extensive margin). They of exporters and grew into superstars find that variation in average firm size within three years of entry on average. accounts for two-thirds of the variation In addition, nearly all of the incumbent across countries in export shares, with superstars had been large one decade variation in firm concentration explain- earlier—so cases of exporters that ing the remaining third. In addition, transitioned slowly from the bottom they show that the variation in average to the top of the size distribution are exporter size comes largely from the extremely rare. These results suggest exports of superstars—other export- that superstars start as large exporters Caroline Freund and Martha Denisse Pierola. ers are more similar across countries and grow fast—implying that they are 2012. “Export Superstars.” Policy Research Work- and industries. The results suggest already highly productive when they ing Paper 6222, World Bank, Washington, DC. World Bank Research Digest 7 Why Did Inequality in Latin America Decline in the 2000s? Two main factors drove the drop in of employment of low-skilled workers supply-side constraints. In addition, inequality in Latin America: lower and because of changes in the distri- conditional cash transfer programs skill premiums and more progressive bution of hourly wages, which in turn reduced demand-side constraints by can be driven by changes in the dis- compensating poor households for the social spending tribution of characteristics, especially costs of schooling, including the op- H education and experience, and by portunity cost in the form of forgone igh and persistent inequality changes in returns to those character- child labor earnings. is a distinctive feature of Latin istics, especially educa- The reduction in the America. After rising in the tion. The authors show Income inequality inequality of nonlabor 1990s, however, income inequality un- that in all three coun- income can be attribut- ambiguously declined in the majority tries the changes in unambiguously ed in part to more pro- of countries in the region in the 2000s. the skill premium were declined in the gressive social spend- Of the 17 countries for which there equalizing. The skill pre- ing. Although in Mexico, are comparable data, 13 experienced mium is affected by the majority of countries for example, private a decline between about 2000 and demand for and supply in Latin America in transfers such as remit- 2010, a trend not observed in other of workers with different tances proved to be parts of the world. Existing analysis skills and by institu- the 2000s—a trend equalizing (in particular, suggests that the decline in inequality tional factors such as not observed in other because they narrowed is robust to the selection of the time the minimum wage and the rural-urban income interval, income variable, inequality the influence of unions. parts of the world gap) and became even measure, and data source. The decline Labor demand by skill more so in the 2000s, is nontrivial and statistically signifi- is affected primarily by the characteris- the most important factor in the cant. Moreover, a decomposition of the tics of technological change and inter- decline in nonlabor income inequal- changes in poverty during the 2000s national trade. The composition of the ity was a significant increase in the reveals that the reduction in inequality labor supply is determined to a large equalizing contribution of government accounted on average for 50 percent of extent by the characteristics of edu- transfers. These transfers not only rose the observed decline in poverty. cational upgrading and demographic over time but were large enough to To a large extent, inequality in Latin factors. offset the unequalizing effect of other America is the result of state capture The authors’ results show that the sources of nonlabor income, such as by elites, capital market imperfections, fall in the skill premium in Argentina interest, profits, and rents. These cash inequality in access to good-quality can be explained by a decline in the transfers are a small share of total gov- education, labor market segmenta- relative demand for skilled workers and ernment social spending but go a long tion, and discrimination. Thus the by a rise in the minimum wage and way in redistributing income to the observed decline in inequality is good the power of unions. In Brazil, in addi- lowest-income households. The expan- news both for fairness and for growth. tion to demand-side and institutional sion of their coverage during the 2000s Why has inequality declined? Have factors, supply-side factors played an was key to reducing both inequality the changes in inequality been driven important part. In Mexico institutional and poverty. by market forces such as the demand factors were unimportant and the de- for and supply of labor with different mand for skilled labor may even have skills? Or have governments become continued to rise; the fall in the skill more redistributive than they used to premium was therefore driven primar- be? ily by an expansion in the supply of A recent study by Lustig, Lopez- skilled workers. Calva, and Ortiz-Juarez examines the In Brazil and Mexico then, the de- role of the skill premium and govern- cline in labor income inequality was ment cash transfers in changes in in- linked to the fact that unskilled labor equality in Latin America’s three larg- became relatively less abundant. The est economies (as measured by GDP): significant expansion of basic educa- Argentina (urban areas), Brazil, and tion underlying the change in labor Mexico. The authors’ analysis finds composition by skill in these two that the decline in overall inequality countries seems to be associated Nora Lustig, Luis Lopez-Calva, and E. Ortiz- can be attributed to a decline in both with conscious government efforts. Juarez. 2012. “Declining Inequality in Latin labor and nonlabor income inequality. Higher spending per student in basic America in the 2000s: The Cases of Argentina, A decline in labor income inequal- education and an effort to make edu- Brazil, and Mexico.” Policy Research Working ity can occur because of an expansion cation accessible in rural areas eased Paper 6248, World Bank, Washington, DC. 8 World Bank ResearchDigest (continued from page 1) How exactly we get there will there- countries. Today we use more than 850 Recent Policy Research fore matter, and the best sustainable surveys for 125 countries—more than Working Papers route will vary from country to country. six per country; the latest estimates The growth projections underlying use a “global” sample of 2.1 million 6212 Vyaghranomics in Space and Time: these targets are grounded in the households. Many problems remain, Estimating Habitat Threats for Bengal, economic realities of the countries however. There are persistent lags and Indochinese, Malayan and Sumatran Tigers Susmita Dasgupta, Dan Hammer, Robin Kraft, concerned and of the global economy. uneven coverage. There are continuing and David Wheeler But the policy challenges in ensuring concerns about the comparability 6215 Conditional Cash Transfers, Political Participation, and Voting Behavior that poor households share sufficiently of the surveys over time and across Javier E. Baez, Adriana Camacho, Emily and sustainably in that growth at the countries. And there are continuing Conover, and Roman A. Zarate 6223 Making Up People—The Effect of Identity country level will need to be addressed concerns about underreporting and on Preferences and Performance in a in depth. selective compliance in household Modernizing Society Karla Hoff and Priyanka Pandey Most of the work needed to reach surveys; the rich are hard to interview, 6231 Is Green Growth Good for the Poor? such global targets would be done at and that task is not getting any easier. Stefan Dercon the country level. A similar benchmark- The weak integration of “macro” and 6253 There Goes Gravity: How eBay Reduces Trade Costs ing exercise for individual countries “micro” data is also a long-standing Andreas Lendle, Marcelo Olarreaga, Simon Schropp, and Pierre-Louis Vézina would be desirable. Naturally this concern, warranting far more attention 6256 The Great Recession and the Future of Cities would reflect the specifics of each than it has received. Jean-Jacques Dethier and Curtis Morrill country. One should be cautious in Our collective success in address- 6259 Global Income Inequality by the Numbers: In History and Now—An Overview trying to infer what any one country ing these and other data problems will Branko Milanovic can do through cross-country com- determine how confident we are about 6260 Visualizing Uncertainties, or How Albert Hirschman and the World Bank Disagreed parisons, even when these control for these benchmarks and about our abil- on Project Appraisal and Development observable exogenous differences. ity in the future to assess how close we Approaches Michele Alacevich Ideally the country-level benchmarking are to reaching them. 6261 Agricultural Trade: What Matters in the Doha would come with a reasonably clear Round? David Laborde and Will Martin plan for achieving the targets. Different 6262 The Cost Structure of the Clean Development tools of economic analysis could Mechanism inform such plans, ranging from com- Shaikh M. Rahman, Donald F. Larson, and Ariel Dinar putable general equilibrium models to 6266 Monitoring Export Vulnerability to Changes in micro-simulation tools, all with both Growth Rates of Major Global Markets Claire H. Hollweg, Daniel Lederman, and strengths and weaknesses. José-Daniel Reyes Monitoring performance against 6267 Ahmedabad: More but Different Government for “Slum Free” and Livable Cities these benchmarks would pose serious Patricia Clarke Annez, Alain Bertaud, Marie- data challenges. There has been huge Agnes Bertaud, Bijal Bhatt, Chirayu Bhatt, Bimal Patel, and Vidyadhar Phatak progress in collecting the primary household survey data. When the World Bank’s current global poverty Martin Ravallion. 2012. “Benchmarking Global Working Papers can be downloaded at http://econ.worldbank.org monitoring effort began in 1990, the Poverty Reduction.” Policy Research Working Pa- To download the World Bank Research E-Newsletter, estimates used 22 surveys for 22 per 6205, World Bank, Washington, DC. go to Data & Research at http://www.worldbank.org The World Bank Research Digest is a quarterly publica- The Research Digest is financed by the Bank’s Editorial Committee: Jean-Jacques Dethier (managing tion disseminating findings of World Bank research. Research Committee and managed by DECRS, the editor) and Aslı Demirgüç-Kunt. Research assistance: The views and interpretations in the articles are those research support unit of the Development Economics Alexander Moore; editor: Alison Strong; production: of the authors and do not necessarily represent the Senior Vice Presidency (DEC). The Research Digest is Roula Yazigi. For information or free subscriptions, views of the World Bank, its Executive Directors, or the not copyrighted and may be reproduced with appropri- send email to research@worldbank.org or visit countries they represent. ate source attribution. http://econ.worldbank.org/research_digest. The World Bank 1818 H Street, NW Washington, DC 20433, USA Printed on Recycled Paper