PrécisW O R L D B A N K O P E R A T I O N S E V A L U A T I O N D E P A R T M E N T A U T U M N 2 0 0 0 N U M B E R 2 0 2

Poverty Reductionin the 1990sThe World Bank Strategy

IN 1990, WHEN THE BANK FIRST FORMALLY ARTICULATEDits strategy for poverty reduction in the World DevelopmentReport, the prospects for global poverty reduction looked

These factors have increased the challengefacing the Bank and the development assis-tance community over the past decade. Theoverall volume of official aid has declined(until the past fiscal year), while the agendaof development assistance has increased sig-nificantly in scope and complexity. Withinthe Bank, projects have evolved to take intoaccount the complexities of the implementa-tion environment. An increased focus on theinstitutional framework for development

promising. But the international economic environment has changedradically. Although there has been a modest reduction in the pro-portion of people living on less that $1 a day, the absolute numberof poor has increased because of continuing high population in-creases and economic growth that failed to meet expectations, ashortfall that was aggravated by the uneven distribution of growthin many regions. Further fueling the escalation in numbers were theeconomic transition in Eastern Europe and Central Asia; financialvolatility and its impact, particularly in the South Asian economiesin the latter part of the 1990s; the AIDS pandemic; civil war; and achain of natural disasters, particularly in Sub-Saharan Africa andLatin America.

assistance has been combined with a movetoward greater policy-based lending andsector-wide investments. The Bank hasincreased its focus on the need to improvedevelopment effectiveness. These are movesin the right direction. Yet a great dealremains to be done.

Building A StrategyIn 1990 a review of global experience hadindicated that a three-part strategy would

2 World Bank Operations Evaluation Department

be most effective in the fight against poverty: (1) encour-age a pattern of growth that promotes the productive useof labor, the most abundant asset of the poor, throughpolicies that harness market incentives, social and politi-cal institutions, infrastructure, and technology; (2) provideimproved access to education, nutrition, health care, andother social services to improve welfare directly and toenhance the ability of the poor to take advantage of theopportunities of broad-based growth; and (3) programwell-targeted transfers and safety nets to support those liv-ing in remote and resource-poor regions or groupsaffected directly by economic transition.

How Well Has It Worked?OED evaluated the 1990 strategy as part of the overalleffort to revise the Bank’s poverty reduction approach. Itfound that there had been substantial progress, includ-ing significant improvement in the social indicators inmost Regions. The strategy performed the essential rolein highlighting the importance of broad-based growthand bringing into focus the significance of the equitableprovision of social services. It also helped to reshapeboth the composition of Bank assistance and the priori-ties of other development agencies. And it spurred a sig-nificant improvement in the poverty knowledge basethat now spans the development community. Improve-ments in the quality of Poverty Assessments and the useof combined methods of gathering nationally and inter-nationally comparable poverty data have broadened theBank’s conceptualization of poverty and its awareness ofthe priorities of the poor.

But this is not the full story. Monitoring of resultshas been insufficient, leaving unanswered questionsabout the benefits and costs of much of what the Bankdoes. Implementation of the central elements of the strat-egy has lagged, and although the 1990 strategy was a

positive step, it did not assemble all thepieces of the puzzle.

The Country as the FocusThe Country Assistance Strategy (CAS),the Bank’s basic planning document,sets out the goals and objectives ofBank assistance. A sample of CASs cov-ering the period 1994 to 1999 evalu-ated by this study show an upwardtrend in relevance and poverty focus,particularly after 1996; the latest CASsgive additional attention to the relatedproblems of governance, institutionbuilding, and participation by the poorand other stakeholders.

A Focus on GrowthGrowth is at the center of all CASs,and the focus on social service provi-

sion is present across the board. But the Bank has foundit difficult to move from the generalities of the 1990strategy to specific, concrete recommendations tailoredto individual country contexts and designed to addressspecific social and structural constraints. Over 90 per-cent of evaluated CASs recommend the same growthstrategy: macro-stability, liberalization, and trade andtariff reform. The linkages between broad-based growthand poverty are not always clear, and CASs rarelyassign priorities to policy choices or address possibletradeoffs and the likely impact on the poor. Many CASsstill make inadequate arrangements to monitor andevaluate progress toward poverty reduction objectives,so that it is not possible to determine results. Until

Box 1. Will the International DevelopmentTargets be Met?

IF THE 4 PERCENT ANNUAL GROWTH RATEpredicted for developing countries up to 2015 is accompa-nied by low inequality, the international target of halvingabsolute poverty should be met easily, and poverty halvedby 2006. If high income inequality accompanies growth,however, the target is not achieved. Stimulating even higherrates of growth will help, but only to a limited extent. Shift-ing from a high-inequality to a low-inequality growth pathhas a much larger effect on poverty than adding even 1 per-cent to the growth rate. To affect inequality, countries needto design growth strategies that increase disproportion-ately the incomes of the poorest or redistribute incomethrough taxation.

Source: L. Hanmer and F. Naschold, Can the International De-velopment Targets Be Met? A Preliminary Report (London:Overseas Development Institute, 1999).

National Geographic photo.

Précis 2023

recently safety nets have been something of an after-thought: just over 50 percent of CASs (and only 20 per-cent in the East Asia Region) cover such provisions forthe vulnerable and the ultra-poor. When they were dis-cussed, the focus was generally limited to groupsaffected by public sector reform.

The Pervasive Effects of InequalityNew evidence on income inequality points to its damp-ening effect on growth and poverty reduction. Improve-ments in social indicators have not necessarily been ac-companied by less “income poverty.” While the evidencedoes not undermine the Bank’s strategy, it does suggestthe need to focus more clearly on structuralinequality as central to any long-term strategy for sus-tained growth and poverty reduction. Without such afocus, it is unlikely that the goal of a reduction in abso-lute poverty by 2015 will be achieved (see box 1)

Disparities in Asset Distribution: Treatmentin Two Country CasesThe FY99 Malawi CAS confronted persistently high lev-els of income poverty, rapid population growth, skewedasset holding, and a slowing down of structural reforms.One of the major issues identified in the CAS was thecomparatively low growth elasticity of poverty. Analysisof household-level income distribution data suggestedthat because of the depth of poverty and inequalities inaccess to land, education, and other assets (between poorand non-poor, and between women and men), theresponsiveness of poverty measures to changes in aver-age incomes is comparatively low. In response, the CASadvocated, in addition to measures to support an accel-eration of growth, investments and policies to improveincome distribution, including building numeracy andliteracy skills and access to credit and market informa-tion for microenterprise, as well as policies to increaselabor productivity, promote labor-intensive exports,increase equity in access to social services (especiallyHIV support), adopt a more equal distribution of land,and mobilize community participation.

Ecuador also displays a highly skewed distributionof wealth. Poverty is pervasive, especially in rural areas.The FY96 CAS proposed a strategy focused on socialsector reforms and the provision of basic infrastructure.But in contrast to the Malawi CAS, it limited the assess-ment of the relationships among growth, inequality, andpoverty and their implications for the growth strategy.The Ecuador Poverty Assessment (1996) clearly points toland as a determining factor for rural poverty, arguingthe need to support the existing informal market in landand the titling of the many unregistered farms as a steptoward more formal land transactions. But the CASavoids the issue of land distribution, focusing instead onthe role of broad-based education and health in improv-ing opportunities for the rural poor.

Challenge to the CASFewer than 50 percent of the evaluated CASs recommendgrowth-enhancing actions such as reforming inequitabletax systems, removing constraints to the informal sector,and improving the regulatory environment for small andmedium-size enterprises, all measures that woulddirectly assist the poor. Only 20 percent include explicittreatment of microeconomic constraints and equity issuesthat affect the access of the poor to physical assets andkey markets, while few CASs directly confront the rela-tionship between persistent gender inequalities and thestrategy for broad-based growth. Explicit inclusion ofthese issues, in addition to the removal of macro andsectoral distortions, would clearly accelerate growth andthe reduction of poverty, particularly in rural areas (box2). And growth based on rural development has a no-table impact on overall levels of poverty.

The Lending BalanceBank lending has been broadly compliant with theoperational aims of the 1990 strategy. The share of pov-erty-focused adjustment operations has steadilyincreased, particularly in International Development As-sociation (IDA) lending (designed for the poorest coun-tries). Lending has increasingly favored sectors particu-larly relevant to the poor—rural and urban infrastruc-ture, primary and secondary education, basic health,social assistance, and microfinance. And lending for thethree main social sectors—education; health, nutrition,and population; and social protection—reached 20 per-cent of total Bank lending, compared with just 5 percentin the early 1980s (figure 1).

Box 2. The Interaction of Growth andInequality

WHILE GROWTH THAT TAKES PLACE IN AREASand sectors where the poor live and work may be consis-tent with a labor-intensive growth path, it can also beaccompanied by increasing, declining or static incomeinequality. In Ethiopia, for example, macroeconomicchange looks promising and the first signs are that ruralpoverty has fallen sharply since the change of governmentin 1992. Yet those that have gained have been those withassets, including land, oxen for plowing, education, andaccess to public goods such as roads. Rural householdswith few assets have lost out. Rural inequality has risenbased largely on inequalities in access to assets. Researchon the pro-poor bias of growth in India also finds thatbetween 1973 and 1989 growth was essentially pro-poorin Andhra Pradesh and anti-poor in Uttar Pradesh.Although poverty levels declined in both states, and therural poor benefited, the scale of poverty reduction inUttar Pradesh was significantly reduced by a worseningincome distribution.

4 World Bank Operations Evaluation Department

Selected poverty-focused adjustment loans havemade positive contributions by lessening inflation andprotecting social spending levels. Less has been done toaddress economic distortions that affect the poor, to raisethe real level of resources for spending on essential ser-vices, and to develop targeted safety net programs.

At the same time, the record of investment lendinghas been mixed. While the Bank’s commitment to thesocial sectors has increased, insufficient attention hasbeen given to ensuring that lending for social servicesactually benefits the poor. Investment lending has suf-fered from problems of fragmentation, and opportunitiesfor leveraging pro-poor growth through complementaryinvestment in private sector development, especially inareas of strong public/private overlap such as basic in-frastructure, have not always been taken.

Projects included in the Program for Targeted Inter-ventions (PTI), a tracking mechanism initiated in 1992,have performed better, on average, than other projects inthe same sectors, largely because of their greater empha-sis on beneficiary and community participation at thedesign and implementation stage, and their greater reli-ance on performance monitoring. Lending specifically tocommunity-based projects, including social funds, has per-formed well overall—81 percent of projects approvedsince 1987 have achieved satisfactory outcomes. But in-clusion in the PTI category does notensure that a project delivers benefits directly to the poor.

Aggregate Bank lending in the 1990s was greater incountries showing strong borrower and institutional perfor-mance and clear effort to reach poverty reduction goals.Recent data on country poverty reduction efforts indicatethat during 1997–99, lending commitments were higher incountries that demonstrated better performance on povertyactions (monitoring, targeted poverty programs, and safetynets). But there is still a need to improve both the criteriaand the extent of country selectivity based on assessmentsof country poverty reduction effort.

Nonlending AssistancePoverty AssessmentsAs the link between country information on poverty andthe CAS is strengthened, the availability of national pov-erty data and the quality of Bank Poverty Assessmentsbecome increasingly important. Bank staff and stakehold-ers alike have confirmed that the Poverty Assessmentsbased on a combination of qualitative and quantitativetechniques have done much to strengthen the understand-ing of poverty and related policy issues. The importanceof the quality of analytical work on poverty is confirmedby the finding that the diagnostic treatment of poverty inthe CAS is positively and significantly correlated with therelevance of the Bank assistance strategy overall. Thequality of the Poverty Assessment is also linked to theinfluence it is likely to have on country clients and stake-holders. And recent evidence on Participatory PovertyAssessments suggests that beyond identifying non-incomedimensions of poverty, the methods used can also be effec-tive instruments for shaping policy dialogue with countrypartners and building country ownership.

Despite tangible change, the improvement in thequality of Poverty Assessments is still modest overall, aswell as being variable across the Bank. Although quali-tative and participatory work have entered the Bank’smainstream, the overall framing of Poverty Assessmentsis still based on drawing a poverty line defined in mon-etary terms, which restricts the scope of the resultingstrategy recommendations. Dissemination of povertydata and the analysis contained in Poverty Assessmentshas not always been adequate, and Bank efforts have notbeen focused sufficiently on supporting borrower capac-ity to utilize quantitative and qualitative poverty data.Limited availability of poverty data continues to be aproblem in many countries, particularly in Africa andLatin America, although the Bank has been influential inaddressing this constraint in South and East Asia.

0

2

4

6

8

10

1980–84 1985–89 1990–94 1995–99

Percent of total lending

Education

Health, nutrition,population

Social protection

Watersupply/sanitation

Year

Figure 1. A Shift in Favor of theSocial Sectors

Figure 2. World Bank Lending and CountryPoverty Reduction Effort

FY97FY98

FY99

Low

Medium

High

0

100

200

300

400

500

600

Average lending per country($ millions, by tercile)

CPIA/CPRpovertycomponent

Précis 2025

OED also found that nearly half the Poverty Assess-ments reviewed did not inadequately address the indi-vidual elements of broad-based growth, social serviceprovision, and safety nets, nor did they justify the bal-ance among these elements in strategy recommenda-tions. Many still do not address global or macro link-ages to poverty or discuss sectoral issues in detail; par-ticularly the prominent role of food policy and ruraldevelopment in contributing to poverty reduction. In thesurvey of stakeholders, 30 percent of respondents feltthat the treatment of rural poverty issues in the assess-ment was unsatisfactory. OED’s recent evaluation of theBank’s implementation of the rural development sectorstrategy—From Vision to Action—notes that effectiverural development work is closely correlated with thequality of related analytical work, particularly the focuson rural poverty.

Public Expenditure ReviewsA complementary tool, the Public Expenditure Review(PER), assesses the efficiency, equity, and accountabilityof public expenditures that affect the welfare of the poor.By focusing on intersectoral tradeoffs, the efficiency andeffectiveness of spending, and the cost-effectiveness ofsafety nets and targeted programs, the PER can contrib-ute to the policy dialogue and the framing of a povertyfocused assistance strategy. Thus far, however, there hasbeen inadequate attention to public expenditure issues—only 15 of the 42 CASs reviewed made direct referenceto PER findings. This is partially because PERs were notalways available for the country in question, but alsobecause of the highly uneven treatment of poverty-related issues in the PERs that were done.

The relationship between nonlending work and theCAS points to improved coverage and diagnosis of pov-erty issues, but also demonstrates the uneven quality andavailability of the work, as well as the failure to take fulladvantage of the information that has been gathered.

The Road AheadThe Bank has made significant progress in focusing itsstrategic and operational work on poverty reduction,and the process continues. The key challenge is toimprove implementation:

n Operationalize the 2001 strategy for poverty reduc-tion in full consultation with representatives fromcivil society, the private sector, and the Bank’s devel-

opment assistance partners. This should include clear,monitorable benchmarks for assessing implementationperformance.

n Move beyond a fragmented, project-based frameworkto emphasize the integration between macro- andmicro-level interventions.

n Better integrate quantitative and qualitative dataabout the diversity of the poor—particularly genderissues and indigenous and socially excluded groups—and their informal safely net strategies, and improveex ante analysis of the likely social impact of pro-posed policy changes.

n Step up policy guidance and training to match theincreased focus on poverty and the new directions inthinking and operations. This should be accompaniedby a systematic program of knowledge disseminationand training—which would include borrower staff—on monitoring and implementing poverty reductionstrategies.

n Develop a strategic framework for evaluating povertyoutcomes alongside the new policy framework. Thisshould include Regional strategies for improving thestock of evaluative data on the relationship betweenBank assistance and poverty outcomes and a set ofcriteria for identifying self-evaluation priorities tomaximize learning about poverty reduction.

n Focus on supporting a wide spectrum of capacitydevelopment, both inside and outside government, toprepare, implement, and evaluate national povertyreduction strategies. This should be based on exten-sive consultation and participation. Support forcapacity building should be done sensitively and flex-ibly to ensure that local ownership is not undermined.

Ultimately, whether the Bank’s poverty strategymakes an impact depends not only on how the Bank per-forms, but also on the performance of borrowers andother development partners, as well as the part played byexogenous factors. The challenge of poverty itself is bothmultidimensional and multisectoral. It is the interactionamong policies that sustain long-term growth, improvethe distribution and the stock of human capital, curb cor-ruption, and enhance the social and physical capital ofthe poor that is likely to make the real difference. Oper-ationalization of the WDR 2000 presents a unique oppor-tunity for the Bank to heed the lessons of experience andenhance the poverty reduction impact of its operations.

This Report is based on Poverty Reduction in the 1990s: An Evaluation of Strategy and Performance,by Alison Evans. Available to Bank Executive Directors and staff from the Internal Documents Unit andfrom regional information service centers, and to the public from the World Bank InfoShop.www.worldbank.org/html/oed