Pro-poor growth in the 1990s: Lessons and insights from 14 countries

Executive summary: Pro-poor growth in the 1990s: Lessons and insights from 14 countries 1 Part 1: Poverty, growth and inequality 14 the 14 countries 19 Basic trends in poverty, growth and inequality 21 the relationships between poverty, growth and inequality 25 Growth was good for the poor 26 Growth does not explain all the variation in poverty 28 notes 32 Part 2: Increasing the participation of poor people in growth 34 Making agricultural activities more productive for poor households 36 Factors affecting growth and agricultural earnings 38 agricultural growth and poverty reduction 41 how did poor households participate in agricultural growth? 43 helping poor households take advantage of nonagricultural and urban employment opportunities 52 Factors affecting the growth of nonagricultural earnings 53 nonagricultural growth and poverty reduction 54 how did poor households participate in nonagricultural growth? 57 notes 69 Part 3: Pro-poor growth strategies that reflect country conditions 72 Policy lessons for pro-poor growth 74 Country conditions affect pro-poor growth priorities 75 Country analysis of growth-poverty linkages 76 areas for further research 77 iv Pro-Poor Growth in the 1990s: Lessons and insiGhts from 14 Countries annex 1: ten lines of enquiry for country analysis 80 annex 2: Interpreting growth elasticities of poverty—a difficult task 81 Statistical appendixes 83 Bibliography 96 Boxes 1.1 two definitions of pro-poor growth 19 1.2 Macroeconomic stabilization and growth in the 14 countries 24 2.1 Public transfers and pro-poor growth 37 2.2 hIV/aIDS hits the labor force in Zambia and Uganda 40 2.3 agricultural growth reduces poverty 42 2.4 agricultural subsidies in India—who benefits and at what cost? 51 2.5 Informal employment and poverty reduction—it can be quite effective, but not always 56 2.6 Investment climate indicators 59 2.7 Remittances reduced poverty but in some countries also increased inequality 67 2.8 targeting public investment in Vietnam 68 Figures 1 Growth reduces poverty 2 1.1 the majority of the poor lived in rural areas in the early 1990s, except in Brazil 22 1.2 Urban poverty fell more rapidly than rural poverty except in Burkina Faso and countries that experienced no growth… 22 1.3 … but poverty reduction occurred mostly in rural areas, except in those countries with high urbanization (Brazil, el Salvador and Bolivia) and with growth concentrated mainly in urban areas (Senegal) 23 1.4 Growth recovered in the 1990s 23 1.5 Inequality down for 6—up for 8 26 1.6 Growth reduces poverty 27 1.7 Consumption by the poor generally grew slower than average consumption 27 1.8 Growth and inequality components of poverty reduction— complementing or offsetting 29 1.9 national averages in Ghana mask significant regional variation in the contributions of growth and inequality to poverty reduction 29 1.10 Changes in growth and inequality are related 30 1.11 Significant poverty reduction but rising inequality for countries growing faster than 3 percent a year 31 2.1 trends in agriculture productivity reveal 3 country groups 38 2.2 Production of cash crops rising in three african countries—food crops declining 40 2.3 Cocoa, coffee and cotton prices up, then down 41 2.4 nonagricultural growth was five times that of agriculture 53 2.5 Urbanization was a driver of nonagricultural growth 54 2.6 Faster income growth for those with secondary education in rural Bangladesh 61 2.7 Faster income growth for those with postsecondary education in urban Bangladesh 62 2.8 Most african countries still have not achieved universal primary education 63 2.9 Secondary enrollments remain low in both the asian and especially the african countries 63 2.10 Spending on secondary education for the poorest quintile rose in Ghana, Vietnam and Indonesia but fell in Uganda 64 2.11 non-farm income became a disequalizing force in the late 1990s in rural Bangladesh 67 2.12 Public spending on infrastructure declined in africa 69 Tables 1.1 trends for the 14 countries 21 2.1 Decomposing poverty reduction in Ghana and Uganda by sector of activity 43 2.2 Many agricultural producers in Sub-Saharan africa lack market access, even in areas of good potential 46 2.3 In Zambia expansion into cash crops was mainly by medium-scale farmers 47 Map 1.1 the 14 case countries 20 Contents v this report was prepared under the auspices of the operationalizing Pro-Poor Growth research program co-sponsored by agence Française de Développement (aFD), Bundesministerium für wirtschaftliche Zusammenarbeit und entwicklung (BMZ), kreditanstalt für wiederaufbau entwicklungsbank (kfw), Deutsche Gesellschaft für technische Zusammenarbeit (GtZ), Department for International Development (DFID) and the world Bank. the report was drafted by a team led by louise Cord and comprising Binayak Sen (Bangladesh), Stephan klasen, Melanie Grosse, Rainer thiele, Jann lay, Julius Spatz and Manfred wiebelt (Bolivia), naércio Menezes-Filho and ligia Vasconcellos (Brazil), Michel Grimm and Isabel Gunther (Burkina Faso), José Marques (el Salvador), andrew Mckay and ernest aryeetey (Ghana), timothy Besley, Robin Burgess and Berta esteve-Volart (India), Peter timmer (Indonesia), Radu Gheorghiu, wojciech Paczynski, artur Radziwill, agnieszka Sowa, Manuela Stanculescu, Irena topinska, Geomina turlea and Mateusz walewski (Romania), Jean-Paul azam, Magueye Dia, Clarence tsimpo and Quentin wodon (Senegal), John a. okidi, Sarah Ssewanyana, lawrence Bategeka and Fred Muhumuza (Uganda), thomas Bonschab and Rainer klump (Vietnam), James thurlow and Peter wobst (Zambia), Derek Byerlee, Chris Jackson, Peter timmer and Radhika nayak (agriculture and rural development), Stephan klasen (Gender), omar azfar (Institutions), Sabine Bernabè and Gorana krstic´ (labor Markets), humberto lopez (Macroeconomics), Vera wilhelm and Ignacio Fiestas (Public expenditures). Communications Development Inc., led by Bruce Ross-larson, assisted the team in conceptualizing the report and was responsible for the book’s design, editing and production. the members of the oPPG research program include Christian Rogg, Manu Manthri, Mandy Chatha, tom Crowards, will Gargent (DFID), Christian Flamant, François Pacquement, Jean Marc Chataigner and Jacky amprou (aFD), Birgit Pickel and Daniel alker (BMZ), Ulrike Acknowledgments Part 1: Poverty, Growth and inequaLity vii Maenner, hartmut Janus and Julius Spatz (GtZ), annette langhammer and henning andresen (kfw) and louise Cord, humberto lopez, Ignacio Fiestas and Sabine Bernabè (world Bank). the work was carried out under the direction of adrian wood (DFID), luca Barbone, Sudhir Shetty and Danny leipziger (world Bank). the study reflects comments received from a workshop with the country authors in Frankfurt (June 2004), and from workshops in london (December 2004) and washington (February 2005) with the authors of the sectoral papers, the core donor team and world Bank and DFID staff. these workshops also included academics, nGo representatives and representatives of other donor agencies. the study also reflects feedback from participants of the world Bank’s PReM Conference sessions, Equity and Pro-Poor Growth and Making Growth Pro-Poor: Cases and Policies, organized jointly with the 2006 World Development Report team. Many others provided helpful comments, including Gary Fields (Cornell University), Peter timmer (Center for Global Development), alan Gelb, Daniela Gressani, tamar Manuelyan atinc, John Page, edgardo Campos and Martin Ravallion (world Bank), nancy Birdsall (Center for Global Development), lionel Demery and John toye (Consultants), Max everest-Phillips and arjan de haan (DFID) and Marc Raffinot (aFD-DIal). nelly obias and esteban hernandez were responsible for processing the manuscript and provided invaluable assistance to the team during its preparation. the operationalizing Pro-Poor Growth (oPPG) program was initiated in 2003 by the aFD, BMZ (kFw/GtZ), DFID and the world Bank to better understand the options for policymakers to increase the impact of growth on poverty reduction and how they vary depending on policies and country conditions. the goal was not to provide a specific policy framework for pro-poor growth. It was to explore the channels for the poor to participate in growth and the country context and initial conditions affecting the efficiency of growth in reducing poverty. the oPPG work adds to the literature by drawing on 14 country case studies: Bangladesh, Bolivia, Brazil, Burkina Faso, el Salvador, Ghana, India, Indonesia, Romania, Senegal, tunisia, Uganda, Vietnam and Zambia. the countries had at least two household surveys in the 1990s and early 2000s that offered comparable methodologies, consumption aggregates and poverty lines. the country studies systematically analyzed the distributional pattern of growth and how it was affected by country policies and conditions, thus overcoming some of the wellknown shortcomings of cross-country econometrics. the case studies shared a common empirical methodology to analyze the distributional impact of growth that built on Ravallion (2004a). the studies looked at four broad policy areas and how each affected the ability of poor people to participate in growth: the macro framework and the composition of growth; agriculture and nonfarm income; labor markets and employment; and public expenditure policies. the role of gender and institutions in affecting policies and their outcomes were addressed as cross-cutting themes. these four areas were not meant to be comprehensive. access to financial markets and health services, as well as voice and empowerment, were not explicitly addressed in each case study, though they were mentioned in some. to draw out the key lessons, seven thematic papers were prepared covering: macro stability and pro-poor growth, growth and inequality, labor markets and employment, agriculture, public expenditures, institutions and gender. the thematic papers draw on the case studies, other Preface Part 1: Poverty, Growth and inequaLity ix literature covering related themes in the 14 countries, and in some cases additional empirical work generally on a subset of the 14 countries. this report provides an overarching synthesis of all this work. In looking at the distributional impact of growth on the poor, the study adopts an income-based metric of poverty reduction (based on national poverty lines). (the role of nonincome dimensions in increasing the impact of growth on the poor and helping households to escape poverty will also be covered by an ongoing world Bank study that will be executed in FY06–07 titled, “Moving out of Poverty.” the oeCD PoVnet group also commissioned a study on the nonincome dimensions of pro-poor growth.) But institutions and nonincome dimensions of poverty are considered highly relevant determinants of the distributional impact of growth on income poverty and are discussed where relevant. It is thus primarily focused on the first Millennium Development Goal for the reduction of income poverty. the country cases and the synthesis paper focus on the 1990s, but overall poverty, growth and inequality trends of the decade are viewed, where possible, within the countries’ broader historical experience. Partly data-driven, the limitation on the 1990s reflects the relatively limited time horizon available to many policymakers. It also allows us to investigate how the economic developments of the 1990s may have affected the relationships between poverty, growth and inequality. 2 Pro-Poor Growth in the 1990s: Lessons and insiGhts from 14 Countries Policymakers who seek to accelerate growth in the incomes of poor people, and thus reduce overall poverty, would be well advised to implement policies that enable their countries to achieve a higher rate of overall growth. evidence from the 14 countries in this study confirms that the pace of overall economic growth is the main factor that determines how quickly poverty declines. a successful pro-poor growth strategy should have, at its core, measures to achieve sustained and rapid economic growth. these include macroeconomic stability, welldefined property rights, a good investment climate, an attractive incentive framework, well-functioning factor markets and broad access to infrastructure and education. the country studies demonstrate the strong link between overall economic growth and the speed of poverty reduction. the incidence of poverty fell in the 11 countries that experienced significant growth during the period, and rose in the three countries that saw little or no growth (Zambia, Indonesia and Romania). on average, a 1 percent increase in GDP per capita for these countries reduced poverty by 1.7 percent during this period (figure 1). the reduction in poverty was particularly spectacular in Vietnam, where poverty fell by 7.8 percent a year between 1993 and 2002, halving the poverty rate from 58 percent to 29 percent. other countries with impressive poverty reductions include el Salvador, Uganda, Ghana, India and tunisia, each with declines of 3 to 6 percent a year. The country studies demonstrate the strong link between overall economic growth and the speed of poverty reduction In growing countries, most of the absolute reduction in poverty was in rural areas, where the majority of poor households lived. the proportional decline in poverty rates was more marked, however, in these countries’ urban areas, characterized by higher growth. For instance, trade liberalization, market-oriented reforms, export incentives and massive increases in infrastructure and education helped to reduce urban poverty by 11 percent a year in Vietnam between 1993 and 2002. In the Sub-Saharan countries, where most agricultural growth came from export crops, the deepest cuts in poverty were for those growing them (cotton in Burkina Faso, coffee and cotton in Uganda, cocoa in Ghana). But given the predominance of poor households in producing foodcrops, they accounted for the greatest share of poverty reduction even in these countries. Driving these overall reductions in poverty was the rebound in growth that began for most of the countries in the mid-1990s. the median GDP growth rate for the 14 countries was 2.4 percent a year between 1996 and 2003. the success of macro stabilization measures was integral to this recovery in many of them, particularly those initially underperforming (Uganda, Senegal, Bolivia, and Ghana). efforts to cut inflation and reduce external and internal imbalances (including competitive exchange rates) were particularly effective in stimulating nonagricultural growth, which grew on average by 3.1 percent in the 1990s and early 2000s, compared with 0.6 percent for agriculture. Price liberalization and trade reforms also stimulated exports of manufactures and agricultural products for the asian and several of the african countries (Burkina Faso, Ghana and Uganda). Improved structural and sectoral policies also contributed to higher growth. the devaluation of the CFa franc in Senegal stimulated higher levels of investment and strong urban growth. high levels of infrastructure spending in Indonesia, Vietnam and Bangladesh fueled agricultural and nonfarm growth, particularly in rural areas. and better human development outcomes (health and education) in most countries outside africa (where aIDS resulted in reversals on several dimensions), improved their growth prospects. Rising capital inflows from foreign direct investment (albeit from a very low base), aid (particularly in the african countries) and remittances also fostered higher growth in the 1990s. while the power of growth in reducing poverty is undeniable, the experience of the 14 countries in this study also shows that growth was more powerful in reducing poverty in some countries than others. Greater poverty reduction was observed where policies were in place to enhance the capacity of poor people to participate in growth. Several of these policies are the same as those required to foster higher growth. exeCutive summary—Pro-Poor Growth in the 1990s: Lessons and insiGhts from 14 Countries 3 Greater poverty reduction was observed where policies were in place to enhance the capacity of poor people to participate in growth 4 Pro-Poor Growth in the 1990s: Lessons and insiGhts from 14 Countries For example, trade liberalization and incentives for manufacturing enterprises helped expand employment opportunities for semiskilled and unskilled labor (and particularly women) in Bangladesh, el Salvador, tunisia and Vietnam. the liberalization of imports and marketing of agricultural inputs allowed poorer Bangladeshi farmers to expand their use of low-cost irrigation pumps, in turn facilitating their use of higher yielding Green Revolution technology. Coffee sector reforms at a time of rising world prices helped lift a significant number of rural Uganda households out of poverty. the country studies also illustrate the value of viewing growth through a pro-poor lens for analyzing and addressing the constraints that poor households face in participating in growth.1 Depending on the country circumstances, this may mean that access to electricity and secondary education should increase not only in the capital city but also in small towns, peri-urban areas and villages. In other contexts, it may call for an emphasis on strengthening institutions that help to deliver titles that build on customary tenure systems. these interventions increase the quantity and quality of poor people’s productive assets and their ability to participate on an equal footing in product and factor markets. In so doing, they help poor households to increase their agricultural and nonagricultural employment (both wages and self-employment) and benefit from rising earnings associated with the growth process. Increasing the participation of poor households in growth Making agricultural activities more productive Because the vast majority of poor people live in rural areas and draw their livelihoods from agriculture, the discussion begins with factors that can raise the incomes of poor people who continue to rely on agriculture. Bangladesh and Uganda typify some of the policies and constraints that affected the agricultural earnings of poor farmers. In Bangladesh liberalizing imports of agricultural inputs and machinery improved access to low-cost irrigation, which along with greater investments in flood infrastructure and safety net programs, led many poorer farmers to adopt Green Revolution technology, raising their productivity and incomes. In Uganda poorer farmers benefited from rising coffee prices in the mid-1990s. But since the late 1990s agricultural earnings have stagnated, particularly for poor farmers, and rural poverty reduction has slowed significantly. what constrains earnings growth for poor farmers? thin input markets, despite liberalization. extension and microfinance services that are still accessible mainly to larger farmers. and The country studies illustrate the value of viewing growth through a pro-poor lens land market and use rights that remain unclear, reducing incentives for smallholder farmers to invest. Five policy interventions were important in helping to raise the agricultural earnings of poor households in the 1990s. • Improving market access and lowering transaction costs. • Strengthening property rights for land. • Creating an incentive framework that benefits all farmers. • expanding the technology available to smallholder producers. • helping poorer and smaller producers deal with risk. among the countries where agricultural earnings increased for the poor, these policies were not in place to the same degree. and not all of them had the same effect on increasing the ability of the poor to participate in growth, reflecting the different initial conditions and other influences in individual countries. Improving market access and lowering transaction costs were essential in Indonesia, Bangladesh and parts of Vietnam for increasing the agricultural earnings of smaller and poor farmers. Market access for them was facilitated by significant investments in rural roads and marketplaces (often implemented under food for work programs), by high population densities and by the fact that smallholder export and foodcrops were often the same (rice). But high transaction costs did constrain agricultural earnings in the more remote regions of the asian and latin american countries, where rural poverty is disproportionately high (such as the upland regions of Vietnam, northeast Brazil and Bolivia). In some of these countries, policy options could include providing skills that would enable the poor to profit from economic opportunities elsewhere. among the low income african countries in the sample, high transaction costs and low market access were among the most important constraints on expanding agricultural earnings, especially for small farmers and those in remote areas. with food markets in africa expected to be the fastest growing of all agricultural markets in the continent over the next 20 years (Commission for africa 2005), it will be important to link rural farmers to local and regional markets with better infrastructure and marketing associations. Contract farming with nGos and the private sector has facilitated market access in several african countries, particularly when complemented by organized involvement at the grassroots. Strengthening property rights for land improved the incentives to increase production and diversify into higher value crops in Vietnam. In 1988 land was decollectivized, and under the 1993 land law certificates of use were issued to all rural households, stimulating the intensification and diversification of agricultural production into higher exeCutive summary—Pro-Poor Growth in the 1990s: Lessons and insiGhts from 14 Countries 5 Five policy interventions were important in helping to raise the agricultural earnings of poor households in the 1990s 6 Pro-Poor Growth in the 1990s: Lessons and insiGhts from 14 Countries value-added crops. For the poorer farmers in the african case studies, clear tenure and transparent land markets were important. weak land market institutions—often reflecting the partial implementation of land laws (Uganda, Burkina Faso), the lack of full transparency in local land management decisions (Zambia) and the difficulties in gaining access to land for non-community members (Ghana)—were key constraints on the ability of all farmers, but particularly poorer farmers, to invest in their land. the lack of secure tenure and of legally recognized ownership rights, particularly for inheritance, affected poor rural women in the african countries, who often are the primary producers of foodcrops. In many african countries, improving the security of land tenure for poorer farmers will require developing formal systems that strengthen and complement customary land practices. In Brazil and Bolivia, access to land is a major issue because of very unequal land distributions. large-scale land reform is not politically viable, but expanding the access of smallholders and poorer farmers to long-term financing, and in some cases grants for land purchases, have been successful. In Bangladesh and India continuing restrictions on land rental markets to protect ownership rights make it difficult and costly for smaller farmers (particularly women and the landless) to rent land. In Indonesia land rights remain fairly undefined at the local level, particularly for forest (land records cover only 20 percent of all land in Indonesia). opaque and costly systems of land administration and allocation in rural Indonesia are serious obstacles to expanding agricultural earnings, particularly for poorer farmers (Deininger and Zakout 2004). Creating an incentive framework that benefits all farmers was an important part of the structural reforms by the african countries, Bangladesh, Vietnam and Romania. the impact has varied depending on the size of production units, access to capital, technical assistance, and markets (or transaction costs) and the crops they grow. trade liberalization, along with land reform, promoted Vietnam’s rapid emergence as a major world exporter of rice and coffee in the 1990s, greatly benefiting smallholders. trade liberalization in Bangladesh facilitated imports of low-cost inputs, increasing their use by poor farmers. Foodcrop farmers in africa generally benefited less than export crop farmers, whose poverty rates fell sharply. But except for coffee producers in Uganda, export farmers tended to make up a small share of the total and were mainly the better-off. the private sector often did not fill the void left by reforms in foodcrop marketing, leaving many poor producers in remote areas of africa without market access. Subsidies and protection in India, Indonesia and tunisia characterized agricultural production, redirecting public resources and incentives Creating an incentive framework that benefits all farmers was an important part of the structural reforms by the African countries away from higher value production toward less labor-intensive basic food grains. In India the reform of agricultural subsidies has been diffi- cult, in large part because of their political appeal (keefer and khemani 2003). In Indonesia the tariff on rice imports raised prices for rice producers (many of them are smallholders and poorer farmers) but hurt rice consumers and slowed poverty reduction. But in reforming agriculture it will be important to consider the transition costs to small farmers: they may receive only a small share of total subsidies, but these subsidies are a significant share of their total income. and in implementing trade and price reforms more generally, it will be important to understand how they will affect different types of households and to provide poorer households with roads, financial services and marketing associations so that they can take advantage of the new opportunities. Expanding the technology available to smallholder producers was a critical driver for the Green Revolution to raise agricultural earnings in asia. In Indonesia Green Revolution technology and massive investments in agriculture catalyzed high rates of pro-poor growth from the 1960s to the 1980s. In Sub-Saharan africa, the lack of adequate technologies for arid climates was a severe constraint on producers, particularly those in foodcrops, where the poor are concentrated. Increasing financial support to african research institutions and improving the delivery of extension services to food crop farmers; and in particular women with private firms and nGos, could lift agricultural earnings for poorer farmers. Helping poorer and smaller producers deal with risk has stimulated the adoption of higher yielding agricultural techniques. Investments in flood infrastructure and flood season safety nets for poorer farmers in Bangladesh (along with greater access to private irrigation) reduced risk and created incentives for diversification. Information and communication technologies can provide smallholders with market information (mobile phones in Uganda). and improving access to market storage facilities can help to smooth seasonal fluctuations (Burkina Faso). In general, expanding the use of targeted safety net programs (where administrative capacity exists or can be reinforced) would help avoid severe deprivation from output and price variations and encourage farmers to adopt riskier technologies that offer higher returns. Taking advantage of nonagricultural and urban employment opportunities Urbanization is continuing at a fast pace. Understanding the factors that can allow poor households to take advantage of nonagricultural jobs in rural areas and job opportunities in urban areas is crucial for a exeCutive summary—Pro-Poor Growth in the 1990s: Lessons and insiGhts from 14 Countries 7 Helping poorer and smaller producers deal with risk has stimulated the adoption of higher yielding agricultural techniques 8 Pro-Poor Growth in the 1990s: Lessons and insiGhts from 14 Countries pro-poor growth strategy. In Vietnam trade liberalization and export promotion in labor-intensive manufacturing—combined with rising domestic demand stimulated in part by fairly high rates of agricultural growth—increased nonagricultural employment and earnings for poor households in urban and more connected rural areas. In Burkina Faso informal employment expanded rapidly in services, but earnings fell, due to insufficient demand, reflecting slow agricultural growth, a weak investment climate and difficulties in accessing international markets. More generally, the country cases underscored four broad policy options to enhance access to nonagricultural earnings for poor households. • Improving the investment climate. • expanding access to secondary and girls’ education. • Designing labor market regulations to create attractive employment opportunities. • Increasing access to infrastructure. as with policies to expand agricultural earnings for the poor, the relative priorities and the appropriate design and scope of these policy options vary across countries. Improving the investment climate stimulated growth, influencing the size of the formal sector and the composition of formal employment. In Vietnam, Bangladesh and tunisia, investment climate improvements, trade liberalization and special incentives for manufacturing industries significantly increased unskilled manufacturing employment, particularly for women. In Senegal, rising urban employment reflected high levels of investment, a competitive exchange rate and a fairly good investment climate. By contrast, policy uncertainty and macro instability constrained nonagricultural investment and employment in Zambia, increasing urban poverty. In Ghana, private investment remained low (undermined in part by persistently high inflation and a poor investment climate), causing manufacturing employment to contract in the late 1990s. Expanding access to secondary and girls’ education has become more important with the rising skill bias of nonagricultural employment. In India and Brazil poor educational outcomes reduced growth among different states and the impact of that growth on poverty reduction. Female literacy, also important in reducing poverty, was the most important determinant of interstate differences in the efficiency of nonfarm growth in reducing poverty in India (Ravallion and Datt 1996). educational differences were associated with rising inequality in Bolivia and Uganda: those with more education were better placed to take the more attractive nonagricultural jobs. although most of the african The country cases underscored four policy options to enhance access to nonagricultural earnings for poor households countries in the sample now have near-universal primary school enrollment, secondary enrollments hardly increased, greatly constraining the development of nonagricultural activities. In tracking educational outcomes, it is also important to examine quality and transparency. In Uganda the massive increases in primary school enrollments in the 1990s undermined quality. In addition, there was extensive leakage of educational funds—leakages since reduced by increasing the transparency of budget management, particularly at local levels. Designing labor market regulations to create attractive formal employment for poor workers helps expand their nonagricultural earnings, particularly in countries with fast growth. labor market regulations, often designed to protect the interests of poor workers, can restrict formal labor markets and the access of poor workers. In India states with “proworker” legislation recorded lower growth rates and less efficiency in reducing poverty. In Bolivia and Romania “pro-worker” regulations, encouraged by unions and the economic elite, kept employment in the formal labor market below levels otherwise possible. During Romania’s negative growth period, workers were forced to return to agriculture in large part because labor markets were inflexible, due to high payroll taxes, a cumbersome bureaucracy and tight labor regulations (especially for unemployment benefits and the minimum wage). By contrast, Indonesia’s high degree of labor market flexibility during the Suharto years promoted formal employment and labor-intensive growth. But since the 1997 asian crisis, minimum wage increases prompted by union activity have left almost all employment growth to the informal sector, at wages below those in the formal sector. three caveats: First, labor market regulations are only one of a set of factors that affect the investment climate and the willingness of a firm to formalize. other critical constraints include policy uncertainty, fiscal burdens, cost of finance, corruption and the quality of courts (world Bank 2005c) (see earlier discussion on investment climate, p. 8). Second, loosening labor market regulations in some regions, particularly africa, may have little impact on labor markets, especially if employment is mainly in agriculture (Burkina Faso). third, labor market regulations, though imperfect, constitute a form of social protection. the extent of labor market regulation needs to reflect a balance between workers’ needs and employers’ needs, a balance that hangs on a country’s labor market conditions and level of development. Increasing access to infrastructure (especially roads combined with electricity) linking rural areas to small towns and urban centers, along with strong nonagricultural growth, contributed to rising informal sector employment in rural Bangladesh, India, Vietnam and el Salvador. exeCutive summary—Pro-Poor Growth in the 1990s: Lessons and insiGhts from 14 Countries 9 Labor market regulations can restrict formal labor markets and the access of poor workers to these markets 10 Pro-Poor Growth in the 1990s: Lessons and insiGhts from 14 Countries In contrast, the lack of infrastructure in africa constrained access to attractive informal employment in rural areas, and kept the rural poor engaged in more traditional and lower return nonfarm activities linked to agriculture. as such, lifting infrastructure constraints to improve market access, as well as increasing access to electricity and education in high density rural areas and small towns, may raise nonagricultural earnings for the poor. But improving access to infrastructure requires more than expanding public investments—it also requires higher institutional quality. Poor institutions in Uganda may have prevented improvements to the power infrastructure (keefer 2000). Pro-poor growth strategies that reflect country conditions this summary identifies several policies that can help poor households take advantage of growth opportunities. the analysis also underscores that the priority-setting and phasing of these policies will differ across countries—according to their conditions. Successful pro-poor growth strategies need to be built on a thorough analysis of what limits the participation of poor households in the growth process in specific countries. ten lines of enquiry for such analysis are attached to the report (annex 1). as with growth strategies, the binding constraints that need to be addressed to enhance the ability of poor people to participate in growth will vary depending on country conditions. Six characteristics that were particularly relevant among the case studies are discussed below. • Population density and its degree of urbanization. a country’s population density and degree of urbanization determine the extent to which transaction costs and remoteness preclude rural households from participating in growth and the relative importance of a targeted infrastructure strategy. For example, in Bangladesh transaction costs are less of a constraint than in Uganda, where the population density is much lower. Variations in population density within countries also influence regional priorities. In Vietnam transaction cost may not be a major constraint for rural entrepreneurs in the Southwest, but they are for producers in the remote northern mountain region. • Asset and income inequality. the initial asset and income inequality influences the poverty-reducing impact of future growth. It may also reveal gender or ethnic discrimination or other inequality traps that keep certain groups from having an equal footing in factor or input markets. • Importance of agriculture. the relative importance of agriculture in the economy and the workforce determines the need to raise agri- Successful propoor growth strategies need to be built on a thorough analysis of what limits the participation of poor households in the growth process cultural earnings or encourage mobility. For example, in Uganda where 90 percent of poor households are in rural areas and 80 percent of the workforce is engaged in agriculture, promoting sectoral growth for smallholders and sectoral mobility will be central to the country’s pro-poor growth strategy. • Climate in and across sectors. Climatic instability affects agricultural earnings and the need for risk management initiatives to protect poor farmers and to encourage their investments in higher yielding but riskier activities. • Fertility. the fertility rate indicates how women, particularly poorer women (since they tend to have higher birth rates), can participate in the workforce. where the fertility rate remains high, as for most of Sub-Saharan africa, countries should accelerate girls’ education. • Institutions. the quality and capacity of institutions (accountability, transparency) for service delivery affect how much public investments can be relied upon to link the poor to growth. the experience of the 14 countries in the 1990s underscores three areas of future research to help policymakers understand how to increase the participation of poor households in growth and accelerate poverty reduction. • First, movement from agricultural to nonagricultural employment was important in raising the incomes of poor households in many countries, but there is also evidence that the more educated, better connected workers were more successful in this regard. Understanding how sectoral mobility might be enhanced is an important area for further research. • Second, the impact of growth was uneven across regions within countries. Understanding how to craft public investment strategies that can address subregional growth and poverty is another important area for further analysis. the findings may differ for low and middle income countries and could be particularly important for countries with decentralized governments. • third, political economy considerations often affect the distributional outcomes of structural and investment policies, at times at the expense of poor households. Understanding how to make public policy to enhance the ability of the poor to participate in and influence government processes is also an area for further exploration. the 14 country studies give us useful insights on how to better integrate short-term and long-term policies to increase the impact of growth on poverty reduction. while policymakers cannot systematically “trade less growth for more equity,” they can and should focus on countryexeCutive summary—Pro-Poor Growth in the 1990s: Lessons and insiGhts from 14 Countries 11 The country studies give us useful insights on how to better integrate shortterm and longterm policies to increase the impact of growth on poverty reduction 12 Pro-Poor Growth in the 1990s: Lessons and insiGhts from 14 Countries specific interventions that may make growth more poverty-reducing. ensuring that national planning and strategic processes, such as poverty reduction strategies in low-income countries, take more fully into account the factors discussed here, and how they apply to different countries, will be a key ingredient for reducing poverty more rapidly in the coming decade. Note 1. these messages are broadly similar to the findings of the 2006 World Development Report, “equity and Development,” which calls for growth to increase the opportunities of the poor, using a broader definition of opportunities than this study to include asset endowments (including human capital assets), wealth and power, market access and process fairness (world Bank 2005e).