Outreach D E V E L O P M E N T 33746 P U T T I N G K N O W L E D G E T O W O R K F O R D E V E L O P M E N T s M A R C H 2 0 0 4 CLIENT POWER Making Services Work FOR THE POOR W O R L D B A N K I N S T I T U T E Promoting knowledge and learning for a better world ABOUT THIS ISSUE Editorial Board SWAMINATHAN S. AIYAR ECONOMIC TIMES OF INDIA, NEW DELHI, INDIA W MICHAEL COHEN hen the World Bank published its World NEW SCHOOL UNIVERSITY, NEW YORK, USA Development Report on Poverty in 2000/2001, PAUL COLLIER it refocused the development debate on the THE WORLD BANK, WASHINGTON, DC, USA plight of the world's poor. The report pre- JOHN GAGE sented a policy framework for addressing global pover- SUN MICROSYSTEMS, PALO ALTO, CALIFORNIA, USA ty by calling for action in three areas: opportunity, JOSEPH K. INGRAM THE WORLD BANK, GENEVA, SWITZERLAND empowerment, and security, and demanded immedi- ate and concerted action to address poverty by both KWAME KARIKARI SCHOOL OF JOURNALISM AND COMMUNICATIONS, developed and developing countries. THE UNIVERSITY OF GHANA, LEGON, GHANA The most recent World Development Report (2004), VIRA NANIVSKA "Making Services Work for Poor People," brings prac- INTERNATIONAL CENTER FOR POLICY STUDIES, KIEV, UKRAINE ticality to the policy debate introduced four years ago. PEPI PATRON This recent work notes that the poor not only suffer CATHOLIC UNIVERSITY, LIMA, PERU from lack of income but from a failure in basic service J. ROBERT S. PRICHARD TORSTAR, TORONTO, CANADA delivery far greater than for those of the non-poor. Education, health, water, sanitation, and electricity RAFAEL RANGEL SOSTMANN MONTERREY TECH UNIVERSITY SYSTEM, MONTERREY, MEXICO services simply do not work or are unavailable to the ADELE SIMMONS world's 2 billion poor people. If we cannot improve the CHICAGO METROPOLIS, CHICAGO, IL, USA delivery of these basic services by 2015as called for in VIVIENNE WEE the Millennium Development Goals (MDGs)--what CENTRE FOR ENVIRONMENT, GENDER AND DEVELOPMENT, SINGAPORE have our efforts at poverty alleviation amounted to? Development OUTREACH is published three times a year by the World Have they really been effective? Bank Institute and reflects issues arising from the World Bank's many What is needed, then, to improve these services? As learning programs. Articles are solicited that offer a range of viewpoints from a variety of authors worldwide and do not represent the title of our issue points out, "client power" plays a official positions of the World Bank or the views of its management. key role in empowering people to demand, and partici- SUNETRA PURI MARY MCNEIL pate in, adequate service delivery. The WDR 2004 out- EDITOR-IN-CHIEF EXECUTIVE EDITOR lines a triangle of accountability between the state, citi- ANNA LAWTON zens/clients, and providers that must be strengthened MANAGING EDITOR to improve services. But it also acknowledges that the MOIRA RATCHFORD weakestlinkinthistriangleisthecitizen/client.Itbold- PUBLICATION DESIGN ly calls for "putting poor people at the center of service PHOTO CREDITS Cover: Moira Ratchford; Pages 4­5: Faucet: World provision: by enabling them to monitor and discipline Bank/Eric Miller; School: World Bank/Curt Carnemark; Medicine: Richard B. Levine/Photographer Showcase; Telecom Worker: Frederic service providers, by amplifying their voice in policy- J. Brown AFP; Page 8: Reuters/Daniel Leclair; Page 11: AFP; making, and by strengthening the incentives for Page 15: World Bank/Trevor Samson; Page 18: World Bank/Eric Miller; Page 22: Reuters/Zohra Bensemra; Page 25: AFP/Prakash providers to serve the poor." This should come as no Singh; Page 27: World Bank/Ami Vitale; Page 30: Reuters/Ali Jarekji; surprise to those of us who believe improved efficiency Page 36: World Bank/Curt Carnemark. can result from strengthened accountability to the poor. On a personal note, this is the last issue of Development OUTREACH for which I will be Executive Editor. As Founding Editor, I will continue to con- This magazine is printed on recycled paper, with soy-based inks. tribute to issues of particular interest, especially those focused on social development and poverty reduction. ISSN 1020-797X © 2004 The World Bank Institute I'd like to thank our readers for their support over the W O R L D B A N K I N S T I T U T E past four years, and look forward to watching the mag- Promoting knowledge and learning for a better world azine continue to improve and grow. World Bank Institute www.worldbank.org/wbi Frannie Léautier, Vice President www.worldbank.org/devoutreach The World Bank devoutreach@worldbank.org 1818 H Street NW Washington, DC 20433, USA Mary McNeil E X E C U T I V E E D I T O R Outreach D E V E L O P M E N T V O L U M E S I X , N U M B E R O N E s M A R C H 2 0 0 4 P A G E 1 1 P A G E 2 2 P A G E 2 6 2 Development News 18 Pro-Poor Health Services: The Catholic Health Network in Uganda SPECIAL REPORT DANIELE GIUSTI, PETER LOCHORO, JOHN ODAGA, CLIENT POWER: MAKING SERVICES WORK AND EVERD MANIPLE FOR THE POOR This article demonstrates how pro-poor ethos, supported by soft regulations and technical assistance, can induce a process of change in a network of faith-based providers. 3 The Path to Education: A Multi-Dimensional Approach 22 Scaling up Drinking Water Services MAMPHELA RAMPHELE JUNAID K. AHMAD, DAVID SAVAGE, AND VIVEK SRIVASTAVA Three case studies in Australia, India, and South Africa 4 Making Services Work for Poor People provide a common lesson: the success of water delivery depends on whether service providers are accountable to Guest Editorial citizens. RITVA REINIKKA One must put poor people at the center of service provision 26 Randomized Evaluations of Interventions by increasing client power. in Social Service Delivery 8 Voice and Accountability in ESTHER DUFLO, RACHEL GLENNERSTER, AND MICHAEL KREMER Service Delivery This article focuses on randomized evaluation of educational programs, which offer both substantive and methodological ANNE MARIE GOETZ AND ROB JENKINS lessons. This article outlines trends in accountability relationships, while highlighting limitations of voice-based approaches. 30 Aid Agencies and Aid Effectiveness 11 Citizen Report Cards: BERTIN MARTENS An Accountability Tool The author aims to highlight some of the structural and organizational problems that occur in the aid delivery process SAMUEL PAUL and affect aid effectiveness. In Bangalore citizens assess the city's public services through CRC, and provide useful feedback on their quality, efficiency, 35 VOICES FROM THE FIELD and adequacy. A Call for More Community-Driven 15 Applying the Service Delivery Triangle and Integrated Approaches to Care and Treatment for HIV/AIDS PATTI PETESCH Examples of proposals on education worked out within SIGRUN MØGEDAL communities in Colombia. The Service Delivery Triangle is a tool for exploring key interaction among state, providers, and citizens. It is applied here to the challenge of scaling up care and 38 KNOWLEDGE RESOURCES treatment of AIDS. 39 BOOKSHELF 40 CALENDAR OF EVENTS DEVELOPMENT NEWS News highlights on development issues from around the world Conclusions of the 2004 World Asia. While AIDS fatalities in Western Europe have fallen Economic Forum significantly since the mid-1990s, thanks to antiretroviral drugs that retard the progression of HIV infection into At the closing session of the World Economic Forum deadly AIDS, the picture in Eastern Europe and Central Annual Meeting, held in January in Davos, Switzerland, the Asia makes for grim reading. A World Bank report pub- co-chairs declared that corporate, government, and civil lished last fall said current efforts to "curb HIV-AIDS in society leaders need to establish a more effective frame- the region are too small to have an effect on the course of work in order to interpret and manage the risks--and the the epidemic." It called for spending on HIV-AIDS pre- perceptions of risk--as part of their partnership for pros- vention and care, estimated in the region at $300 million perity and security for the future. Over 2,100 participants in 2001, to rise to $1.5 billion by 2007. from 94 countries were asked to consider questions such For more information, visit: www.eu20004.ie as: Should corporations focus more on the short term or the long term? How do we develop more corporate respon- Sweden Integrated sibility at a time when businesses are becoming more glob- Development Policy al? And most critically, how do we reconcile the varying speed of clocks between the public and the private sectors? Sweden has become the first nation in the world to pass into It was suggested that the answer could only be found in a law an integrated global development policy. The country's systemic approach based on partnership. trade, defense, agriculture, environment, migration and For more information, visit: www.weforum.org other policies must now, by law, align to fight poverty and promote sustainable development. The new policy takes as New LICUS Trust Fund its starting point the Millennium Development Goals and the Monterrey Consensus. It draws on eight fundamental In January the World Bank created a $25 million Trust Fund components: democracy and good governance; respect for to strengthen institutions, support early efforts at policy human rights; equality between men and women; environ- reform, and build capacity for social service delivery in the mental protection; economic growth; social development; world's poorest countries. These countries, collectively conflict management; and global public goods such known as Low Income Countries Under Stress (LICUS), are as financial stability, disease, and terrorism. With the new characterized by very weak institutions and governance, bill, Sweden is also committed to increasing its develop- and constitute the most difficult environments in which to ment assistance and encouraging the rest of the EU to do so. use aid effectively. The Bank's LICUS initiative aims to support selective basic governance reforms and innovative Afghan Poppy Farmers Need Viable mechanisms for social service delivery. The Trust Fund will Alternative target those LICUS countries that are not eligible to receive International Development Association (IDA) funding due The Food and Agriculture Organization (FAO) said that to their arrears with the Bank. Afghanistan's problematic and large-scale opium pro- For more information, visit: www1.worldbank.org/ duction could only be curtailed if poverty-stricken farm- operations/licus ers have a reasonable alternative. FAO has requested $25.5 million to be disbursed over a five-year period to EU Conference agricultural development projects Focuses Attention geared to offer farmers in four on HIV-AIDS Threat provinces an alternative to growing in Eastern Europe, Asia Send your views and comments opium poppies. Production boomed on Development OUTREACH to: to 3,600 metric tons last year, mak- A two-day, 55-nation conference devoutreach@worldbank.org ing Afghanistan the source of three- took place in Dublin in February, Visit us on the web at: quarters of the world's opium. seeking closer cooperation in fight- www.worldbank.org/devoutreach For more information, visit: ing HIV-AIDS in Europe and Central www.fao.org/newsroom 2 Development Outreach WORLD BANK INSTITUTE SPECIAL REPORT The Path ed approach is the recognition that ownership must reside as close to the ultimate beneficiaries as possible. In the best settings, the balance of empowerment in education deci- to Education sion-making should tend to favor those whose futures are most immediately affected by the opportunity to acquire the skills and behaviors that quality schooling imparts. The sec- A Multidimensional Approach ond and third elements in this new approach fall more to donors and the external community. The shift toward deep- er ownership on the part of national and local stakeholders must be matched by a dedication among donors to harmo- nization and capacity building. That is, donors must lessen BY MAMPHELA RAMPHELE the cost to recipients of doing business while at the same time viewing every dollar, yen, or euro given in the name of EDUCATION AT THE PRESENT TIME sits squarely in the improved education as an opportunity to build the local center of the development agenda. Education's influence on capacity that will help move recipient countries from poverty reduction, inequality, and economic growth are now dependency through ownership towards successful and sus- widely accepted in the development community. We know tainable development. that greater education for girls has strong positive impacts Essential to this approach are the aspects of develop- on the health of infants and children, on immunization ment which bear on education, but which are not tradition- rates, family nutrition and the next generation's school al "education concerns" within the discipline. I refer to, attainment. We understand that an educated and skilled among others, the multi-dimensionality of development workforce is one of the pillars for a dynamic knowledge- needs and specifically the effects of areas such as infra- based economy, and we see correlations between general structure, public health and sanitation on education out- levels of educational attainment, gender parity in education, comes. In Morocco, for example, girls attendance at school and democratic governance. more than doubles with the existence of a paved road. In Truly sustainable development rests on a base of rising Bangladesh, school attendance increases by 15 percent with educational attainment and high levels of human capital in access to piped water and lower water collection times. all countries. With this notion comes an ineluctable chal- Likewise in health: on the positive side de-worming stu- lenge--to make vastly improved education available to all the dents led to a 15 percent increase in school attendance in world's children. We are called on to show progress in the Kenya. Mothers who have completed primary education are short and medium term, to produce demonstrable results 50 percent more likely to immunize their children, and evi- that education for all is going from a rallying cry to a reality. dence is emerging that education can be an effective means To achieve results we must keep two simple facts in mind. to lower HIV/AIDS infection rates. On the negative side, First, progress through education comes from improving each year Zambia loses half as many teachers as it trains to access and learning outcomes for all students at all levels HIV/AIDS, and overall about 20 percent of the EFA financ- under well-chosen curricula. It is not enough to increase ing gap (some $975 million) is attributable to costs associ- enrollment; it is not enough even to be successful at one ated with the HIV/AIDS pandemic. educational level only. We must aim high, seeking broad We see then, that part of the path to education for all is gains in access and quality that rest on deepening and with a coordinated approach to provision of infrastructure maturing educational systems and structures. and health services. For donors and stakeholders alike, it Second, we must be brutally frank in assessing the effec- means creating multi-sectoral teams of donors and partners tiveness of our past interventions. This frank assessment who can appreciate the interrelationships and spillovers that must guide our actions. If it is lack of roads that is keeping work in one area of development brings to another. s children from attending school, we need to put the roads in. If it is lack of access to drinking water, pipes with clean Mamphela Ramphele is Managing Director for Human Development, water are needed. If it turns out to be lack of local educa- The World Bank. tional policy expertise and implementation experience, coupled with weak data, then we need to fix this situation. We must have a comparable standard for adequacy of edu- Excerpted from the 2003 Kneller Lecture Delivered at the Annual cation quality, one that can be brought about only if stake- Conference of the Comparative and International Education Society, holders take a comprehensive and coordinated approach to New Orleans, Louisiana, March 2003. the current constraints. The first element of this coordinat- SPECIAL REPORT MAKINGSERVICES WORK THE EIGHT MILLENNIUM DEVELOPMENT GOALS for poor WITH STARTING POINTS IN 1990, EACH GOAL IS TO BE REACHED BY 2015: people 1 ERADICATE EXTREME POVERTY AND HUNGER Halve the proportion of people with less than one dollar a day. Halve the proportion of people who suffer from hunger. 2 ACHIEVE UNIVERSAL PRIMARY EDUCATION Ensure that boys and girls alike complete primary schooling. 3 PROMOTE GENDER EQUALITY AND EMPOWER WOMEN Guest Editorial Eliminate gender disparity at all levels of education. BY RITVA REINIKKA 4 REDUCE CHILD MORTALITY Reduce by two-thirds the under-five TOO OFTEN, SERVICES FAIL POOR PEOPLE--in access, in mortality rate. quantity, in quality. But the fact that there are strong examples 5 IMPROVE MATERNAL HEALTH where services do work means governments, citizens, and Reduce by three-quarters the maternal donors can do better. How? By putting poor people at the cen- mortality ratio. ter of service provision: by increasing "client power" to enable them to monitor and discipline service providers; by amplify- 6 COMBAT HIV/AIDS, MALARIA AND OTHER ing their voice in politics and policymaking; and by strength- DISEASES Reverse the spread of HIV/AIDS. ening the incentives for service providers to serve the poor. Poverty has many dimensions. In addition to low income 7 ENSURE ENVIRONMENTAL SUSTAINABILITY (living on less than US$1 a day), illiteracy, ill-health, gender Integrate sustainable development into inequality, and environmental degradation are all aspects of country policies and reverse loss of being poor. This is reflected in the Millennium Development environmental resources. Goals, the international community's unprecedented agree- Halve the proportion of people without ment on the goals for reducing poverty (Box 1). That five of access to potable water. these eight goals concern health and education signals how Significantly improve the lives of at least central human development is to human welfare. 100 million slum dwellers. To reach these goals, economic growth is essential. But it will not be enough. The projected growth in per capita GDP 8 DEVELOP A GLOBAL PARTNERSHIP will by itself enable five of the world's six developing regions FOR DEVELOPMENT to reach the goal for reducing income poverty. But growth will Raise official development assistance. enable only two of the regions to achieve the primary educa- Expand market access. tion goal and none of them to reach the child mortality goal. If the economic growth projected for Africa doubles, the region will reach the income poverty goal--but still fall short of the health and education goals. More resources, domestic and foreign, have therefore been called for. But increased public spending does not necessarily mean In this issue of Development Outreach we have collected arti- more services and better outcomes (Figures 2 and 3). Why? cles which reiterate the main findings of the World First, the non-poor typically enjoy a disproportionate share of Development Report 2004 in making services work for poor peo- public spending. In Nepal 46 percent of education spending ple. The articles focus on those services that have the most accrues to the richest fifth, only 11 percent to the poorest. In direct link with human development--education, health, India the richest fifth receives three times the curative health water, and sanitation--and highlight key relationships of care subsidy of the poorest fifth. Even though clean water is accountability between clients, policymakers, and providers. critical to health outcomes, in Morocco only 11 percent of the An introductory note by Mamphela Ramphele stresses the poorest fifth of the population has access to safe water, while importance of the notion of multidimensionality when everybody in the richest fifth does. expanding education services and improving their quality. Second, even when public spending can be reallocated In the first article of the Special Report, Anne-Marie Goetz toward poor people--say, by shifting to primary schools and and Rob Jenkins write about the politics of service delivery, with clinics--the money does not always reach the frontline service a strong focus on the accountability relationship between poli- provider. In the early 1990s in Uganda the share of nonsalary cymakers and citizens and the effectiveness of citizen voices. spending on primary education that actually reached primary To make services work for the poor requires increasing schools was 13 percent. This was the average: poorer schools their "client power" over providers to monitor and discipline received even less than the average. them, mimicking the short route. Samuel Paul illustrates the Third, even if this share is increased--as the Ugandans use of an innovative accountability tool--the citizen report have done--teachers must be present and effective at their card--in Bangalore, India, and its results over the past decade. jobs, just as doctors and nurses must provide the care that Sigrun Møgedal extends the World Development Report's patients need. But they are often mired in a system where the framework of accountability relationships to multisectoral incentives for effective service delivery are weak, wages may HIV/AIDS services, highlighting the complexities involved in not be paid, corruption is rife, and political patronage is a way addressing this enormous challenge facing the world. of life. Highly trained doctors seldom wish to serve in remote Highlighting the relationship between policymakers and rural areas. Since those who do serve there are rarely moni- providers,DanieleGiusti,PeterLochoro,JohnOdaga,andEverd tored, the penalties for not being at work are low. A survey of Maniple discuss the role of not-for-profit providers in health primary health care facilities in Bangladesh found the absen- care in Uganda and how they have been able to adapt to changing tee rate among doctors to be 74 percent. circumstances. The article by Junaid Ahmad, David Savage and By no means do all frontline service providers behave this Vivek Srivastava offers insights into scaling up drinking water way. Many, often the majority, are driven by an intrinsic moti- supply in Australia, India, and South Africa. One lesson is that vation to serve. Be it through professional pride or a genuine policy making, regulatory, and service provision functions need commitment to help poor people (or both), many teachers to be kept in separate organizations and processes. and health workers deliver timely, efficient, and courteous Innovating with service delivery arrangements will not be services, often in difficult circumstances. The challenge is to enough. Societies should learn from their innovations by sys- reinforce this experience--to replicate the professional ethics, tematically evaluating and disseminating information about intrinsic motivation, and other incentives of these providers what works and what doesn't. Only then can the innovations in the rest of the service work force. be scaled up to improve the lives of poor people around the The fourth way services fail poor people, is the lack of world. In their article, Esther Duflo, Rachel Glennerster, and demand. Poor people often don't send their chil- dren to school or take them to a clinic. In Bolivia 60 percent of the children who died before age FIGURE 1--SHORT AND LONG ROUTES OF ACCOUNTABILITY five had not seen a formal provider during the ill- ness culminating in their death. Sometimes the reason is the poor quality of the service--missing materials, absent workers, abusive treatment. At other times it is because they are poor. Publicservicesareprovided,notthroughmarket transactions or the "short route" of accountability, but through the government taking responsibility (Figure 1). That is, these services make use of the "long route" of accountability--by clients as citizens paying taxes and influencing policymakers, and policymakers influencing providers and setting incentives for them. When the relationships along thislongroutebreakdown,servicedeliveryfailsand human development outcomes are poor. When the Source: WDR 2004 relationships are strong, good outcomes result. 6 Development Outreach WORLD BANK INSTITUTE Michael Kremer show how we can learn from randomized tri- trials revolutionized medicine in the last century. als about the most effective ways to increase school participa- Foreign-aid donors should reinforce the relationships of tion and improve the quality of education. Creating a culture accountability, not undermine them. It involves changing the where rigorous randomized evaluations are promoted, way much foreign aid is transferred. Harmonizing aid around encouraged, and financed, they argue, has the potential revo- the recipient's service delivery system would be one impor- lutionize social policy in the 21st century, just as randomized tant way. Bertin Martens looks at incentives in donor coun- tries. He argues that aid effectiveness does not depend on the recipient's performance alone, but has a lot to do with donor FIGURE 2--CHANGES CAN DIFFER GREATLY IN PUBLIC SPENDING agencies and their institutional behavior. ON EDUCATION AND PRIMARY COMPLETION RATES Despite the urgent needs of the world's poor people, and the many ways services have failed them, quick results will be hard to come by. Many of the changes involve fundamental shifts in power--something that cannot happen overnight. Making services work for poor people requires patience. But that does not mean we should be complacent. Hubert Lyautey, the French Marshall, once asked his gardener how long a tree would take to reach maturity. When the gardener answered that it would take 100 years, Marshall Lyautey replied, "In that case, plant it this afternoon." s Ritva Reinikka is Research Manager, Development Research Group (DECRG), The World Bank, Co-Director of the 2004 World Development Report "Making Services Work for Poor People," and Development OUTREACH Guest Editor. rreinikka@worldbank.org Public spending on education (1995 US$ per child) and primary completion Source: WDR 2004 FIGURE 3--CHANGES IN PUBLIC SPENDING ON HEALTH AND CHILD MORTALITY ARE ONLY WEAKLY RELATED Per capita public spending (1995 US$) on health and child mortality Source: WDR 2004 M A R C H 2 0 0 4 7 LENCAS CHILDREN WANT MORE ATTENTION TO HEALTH CARE AND EDUCATION. NO MORE COMPROMISES. SPECIAL REPORT Voice and Accountability in Service Delivery BY ANNE MARIE GOETZ AND ROB JENKINS ciations around the world have long enabled clients of public schooling to contribute to performance improvements. What ACCOUNTABILITY, BY DEFINITION, implies voice--the is new is a considerable amount of conceptual innovation and accountable agency is answering a question articulated practical experiment in citizen-led accountability struggles. through a voice. Agencies must also face the other aspect of Beginning with electoral institutions, there are numerous accountability, being vulnerable to sanction--another place at examples of citizens engaging directly in the vertical accounta- which a different type of voice is exercised. bility relationship between voters and representatives. The idea of a relationship between voice and accountabili- Improving electoral accountability was the objective of the ty, however distant, is central to everyday understandings of Election Watch project in the north Indian state of Rajasthan, democratic systems. World Development Report 2004 mentions a where during state elections in December 2003 a coalition of "long route" to accountability: citizens provide mandates to civic groups sought to improve public awareness of candidates' policy-makers to design services to respond to citizens' backgrounds. The Supreme Court had, earlier in 2003, ruled needs. If these needs are not met, this could result in electoral that candidates could be required to disclose information about or other forms of political backlash, including demonstra- their personal assets and histories. Volunteers crosschecked tions and legal proceedings. information supplied by candidates about their assets, their A huge range of citizen-led efforts around the world has outstanding debts, their criminal records, and their education- been experimenting with various means of engaging directly al attainments. This is similar to a project run by the Poder with accountability institutions that were once closed to non- Cuidano (PC), or Citizens' Power, a civic organization seeking state entities. Many civic groups are participating in new, cleaner political competition in Argentina. The PC monitors hybrid forms of accountability, either in partnership with campaign finance norms, broadcasts information about the state organizations or, more frequently, through protest assets of politicians, and accumulates evidence that can be used action, which increasingly takes the form of civil society-ini- to expose political corruption. Both cases reveal ordinary peo- tiated public hearings, people's courts, and commissioned ple, almost by default, performing a function they might justly reports that mimic official state proceedings. Through these have expected the state's electoral commission to discharge, to and other means, ordinary people and their associations are say nothing of another great institution of intermediation, the getting intimately involved, demanding the impartial media. Both organizations represent the vertical (citizen-state) enforcement of punishments for corruption, criminal negli- dimension of accountability intruding upon the horizontal (or gence, and ineffective performance. state-to-state) dimension of accountability­in the process, As with market-based reforms to public services, voice- creating a set of locally adapted hybrid forms. based initiatives stress disintermediation­people are pursu- Public audit functions are notoriously closed to citizen ing individual cases without the aid of auditors and inspec- engagement, but citizens have in some instances assumed the tors-general. Those involved in pursuing what we have called responsibility of verifying whether local government spend- a "new accountability agenda" seem disproportionately to find ing claims are accurate. The citizen-managed public audits of themselves working within shorter accountability time-hori- local government spending conducted by the Mazdoor Kisan zons, the cycle from mandate to sanction radically com- Shakti Sangathan (Workers and Peasants Power Association) pressed. This article outlines some of the salient features of in Rajasthan have led to positive knock on effects: legislative these trends in accountability relationships while also high- change obliging local government officials to supply, on lighting limitations of voice-based approaches. demand by any citizen, photocopies of all local public-spend- ing records, including supporting documentation. These offi- Conceptual innovations and practical cial accounts are assessed by villagers who can verify whether experiments or not, for instance, minimum wages were paid on a road- building employment-generation scheme, or whether the ORDINARY PEOPLE COMPLAINING about abuses of power, or stipulated quality of materials was used to construct a village participating in service delivery in order to better monitor school, or whether a community centre is being used for its providers, are hardly new occurrences. Parent-teachers' asso- official purpose. Citizen engagement in public-expenditure M A R C H 2 0 0 4 9 management is to be found in participatory municipal budg- state-led voice experiments, which often seem driven by pub- eting in some cities in Brazil. This offers an opportunity for lic-relations concerns. At times they manage to inform offi- citizens to express their needs and begin the accountability- cials about public perceptions of government behavior. But seeking process ex ante­demanding explanations and justifi- they rarely include formal obligations on officials to supply cations from policy-makers regarding their plans and pro- answers, accounts, or other information­nor, for that matter, posals, even before decisions are taken. provisions for investigations to be automatically triggered by In countries like India and the USA, long-established prima facie indications of corruption or poor performance. In democracies where rights are constitutionally guaranteed, short, voice is being over-sold, particularly as a means of citizens and their associations have used litigation as a way of securing accountability to the poor. Formal institutions are bringing themselves directly into judicial accountability profoundly biased against socially excluded groups, denying processes. Using the tools available through litigation, indi- them access and meaningful participation. This diminishes vidual citizens and activist groups become part of an official not only their prospects for collective action, but even their fact-finding process. Discovery motions can lead to the avail- ability to formulate common policy positions. ability of government-held information that can incriminate But expecting citizen-initiated voice initiatives to fill the officials, who may never have expected such detailed scrutiny gap left by distorted or underdeveloped expressions of voice is of their decisions. Courts in many countries are being similarly problematic. For a start, the suggestion that those reformed to increase the direct participation of poor people. worst afflicted by abuses of power ought to be directly involved The most important innovation from this point of view has in checking those abuses is hugely unfair. Why should those been an expanded jurisdiction for Public Interest Litigation most likely to suffer adverse consequences from a challenge to (PIL) in some countries, enabling people's organizations to elite privilege be forced to lead the charge? It is, moreover, prosecute abuses of the rights of socially excluded groups and unrealistic to expect relentlessly valiant behavior of ordinary to hold power-holders more directly accountable to the poor. people, who often lack the voice resources of other social groups­elite connections, education, media savvy. In many Citizen-led initiatives cases, voice initiatives­for instance, certain democratic decentralization programs and right to information laws­have CITIZENS' GROUPS HAVE BEEN INSTRUMENTAL in creating resulted neither from pressure by the poor, nor from pure new jurisdictions and institutions for accountability-seeking. altruism. Yet in the long run they can end up making public This process is assisted by the growing legitimacy of a global institutions more accountable to the poor. human rights regime. The non-governmental Coalition for an Citizen's accountability initiatives, particularly those that International Criminal Court (CICC) played an important role end up establishing scrutiny processes parallel to, rather than in establishing the International Criminal Court (ICC). The in partnership with, official horizontal accountability institu- CICC mobilized global public opinion, influenced the diplo- tions, have their own problems. The substitution of citizens' matic procedure that led to the 1999-2000 Rome treaty, and informal institutions for state accountability institutions contributed important amendments to address rights abuses inevitably runs into problems of legitimate authority, controls against vulnerable groups, particularly women. The on power, and at the same time, limited impact. International Campaign to Ban Land Mines (ICBL), a diverse State-society accountability partnerships also must avoid NGO coalition that was critical in creating the Ottawa Mine Ban pitfalls. These should be designed explicitly to increase the Treaty of December 1997, is another high-profile example. capacity of public oversight institutions to answer to the poor. This experimentation is altering popular understandings of Such partnerships must provide rock-solid assurances to citi- accountability and how it should be achieved. It is expanding the zen participants with recognized roles in public oversight range of accredited accountability-seekers. It is exposing dissat- institutions, that they will possess certain rights­most notably, isfaction with weak or consultative versions of accountability, to access the same documentary information as their state those that de-emphasize the enforcement element of accounta- counterparts, to have procedural complaints independently bility relationships. It is drawing attention to the need for de facto assessed, to be able to cross-examine witnesses, and to issue accountability in the actions of non-state power-holders who dissenting reports to a higher body (such as the legislature). substitute voluntary self-regulation for enforceable adherence to Voice without these basic tools of answerability and enforce- normsofsocialjustice.Anditispromotingbroad-baseddeliber- ability is a recipe for disenchantment and disengagement. s ation on changing standards of probity and justice in the actions of power-holders. The results are new standards of what officials Anne Marie Goetz, Institute of Development Studies, oughttobeaccountablefor.Increasinglythestandardisnolonger University of Sussex. adherence to procedure, but the achievement of outcomes, Rob Jenkins, Birkbeck College, University of London assessed in terms of their value for poor and vulnerable groups. The ideas expressed in this article are drawn from Goetz and Jenkins, Pitfalls Reinventing Accountability: Making Democracy Work for the Poor. Palgrave Macmillan, forthcoming. THE PREOCCUPATION WITH "reinventing accountability" has come, in no small measure, from dissatisfaction with 10 Development Outreach WORLD BANK INSTITUTE SPECIAL REPORT Citizen Report Cards An Accountability Tool BY SAMUEL PAUL USER FEEDBACK is a cost-effec- tive way for a government to find out whether its services are reach- ing the people, especially the poor. Users of a public service can tell the government a lot about the quality and value of a service. Strangely enough, this is not a method that is known to or used by most develop- ing country governments. The con- tinuing neglect of the quality of services is in part a consequence of this gap. This is in sharp contrast to the practice of seeking "customer feedback" in the business world, or at least among those who produce and sell goods in the competitive market place. The "take it or leave it" attitude one comes across-- especially at the lower levels of the bureaucracy--is no doubt due to the fact that government is the sole supplier of most essential services. But the disinterest among the high- er levels of political and bureau- cratic leadership in seeking public feedback on the quality and responsiveness of service providers reinforces this tendency. What is Citizen Report Card? WHEN A GOVERNMENT is indif- ferent, the initiative for change must come from civil society. Citizens who elect and pay for gov- ernments cannot and should not remain quiet when essential serv- ices are in disarray and public M A R C H 2 0 0 4 11 FIGURE 1--DECLINE IN PROBLEM INCIDENCE LEGEND 1999 2003 PROBLEM INCIDENCE ACROSS REPORT CARDS 100 BMP: The City Municipal Corporation 80 BESCOM: The Electricity Authority 66 BWSSB: The Water & Sanitation 60 Board incidence 40 39 BDA: Land Development % 29 31 27 31 Authority 20 16 16 19 20 20 13 15 BSNL: Telecom Department 5 8 7 12 3 0 BMTC: City Transport Company BMP RTO POLICE: City Police BSNL BDA BESCOM BWSSB POLICE BMTC RTO: Motor Vehicle Office HOSPITALS Source: PAC, Bangalore, India GOV. FIGURE 2--DECLINE IN FIGURE 3--RISE IN SATISFACTION LEVELS CORRUPTION LEVELS (routine transactions) 1994 1999 2003 OVERALL SATISFACTION ACROSS THREE REPORT CARDS SPEED MONEY INCIDENCE 100 94 96 ACROSS REPORT CARDS 92 85 80 73 78 73 73 77 67 25 60 23 20 satisfied 47 paid 20 40 41 42 % 34 34 32 32 who 15 25 n/a n/a 20 11 16 14 10 5 6 4 9 0 1 5 percentage BMP BSN BDA RTO BMTC 0 BESCOM BWSSB POLICE 1994 1999 2003 HOSPITALS Source: PAC, Bangalore, India Source: PAC, Bangalore, India GOV. accountability is lacking. It was against ices is not just one more opinion poll. from similar backgrounds in terms of this background that I, as a private citi- Report cards reflect the actual experi- education, culture, and so forth, are like- zen, launched a "citizen report card" ence of people with a wide range of pub- ly to use comparable standards in their (CRC) on public services in Bangalore, lic services. The survey on which a assessments. But these standards may be a large city in Southern India, in 1994. report card is based covers only those higher for higher income groups than for The CRC represents an assessment of who have had experiences in the use of the poor whose expectations about public the city's public services from the per- specific services, and interactions with services tend to be much lower. Dividing spective of its citizens. The latter are the relevant public agencies. Users pos- households into relatively homogenous the users of these services and can pro- sess fairly accurate information, for categories is one way to minimize the vide useful feedback on the quality, example, on whether a public agency biases that differing standards can cause. efficiency, and adequacy of the services actually solved their problems or and the problems they face in their whether they had to pay bribes to offi- The Bangalore interactions with service providers. cials. Of course, errors of recall cannot experiment When there are different service be ruled out. But the large numbers of providers, it is possible to compare responses that sample surveys generate THE PUBLIC AFFAIRS CENTRE (PAC) in their ratings across services. The lend credibility to the findings. Bangalore has done pioneering work on resultant pattern of ratings (based on Stratified random sample surveys CRCs over the past decade. The first user satisfaction) is then converted into using well-structured questionnaires are report card on Bangalore's public agen- a "report card" on the city's services. the basis on which report cards are pre- cies in 1994 covered municipal services, A citizen report card on public serv- pared. It is generally assumed that people water supply, electricity, telecom, and 12 Development Outreach WORLD BANK INSTITUTE transport. Since then, PAC has brought out report cards on sev- The credibility of those who craft CRCs is equally important. eral other cities, rural services and also on social services such as The initiators of the exercise should be seen as non-partisan health care. But since it has tracked services for a longer period and independent. They need to maintain high professional in Bangalore, we shall refer to this experiment in detail below. standards. The conduct of the survey and the interpretation of The findings of the first CRC on Bangalore were most the findings should be done with utmost professional integri- striking. Almost all the public service providers received low ty. A report card does not end with the survey and its publica- ratings from the people. Agencies were rated and compared in tion. Much of the advocacy work that follows will draw upon terms of public satisfaction, corruption and responsiveness. the report card findings. The CRC thus is a starting point, to The media publicity that the findings received, and the public be followed by further action through organized advocacy discussions that followed, brought the issue of public services efforts, including civic engagements and dialogues with the out in the open. Civil society groups began to organize them- relevant public agencies. selves to voice their demands for better performance. Some of the public agencies responded to these demands and took Conclusion steps to improve their services. The inter-agency compar- isons and the associated public glare seem to have contributed WHEN A GOVERNMENT ON ITS OWN improves its services to this outcome. When the second report card on Bangalore and accountability, initiatives such as CRCs may not be neces- came out in 1999, these improvements were reflected in the sary. Even under these conditions, a report card can be an somewhat better ratings that the agencies received. Still sev- effective means for civil society groups to monitor the per- eral agencies remained indifferent and corruption levels con- formance of government and its service providers. Public tinued to be high. agencies can on their own initiate report cards on their per- The third CRC on Bangalore, in 2003, has shown a surprising formance as indeed some in Bangalore have done. But when a turnaround in the city's services. It noted a remarkable rise in government is indifferent to these concerns, the report card the citizen ratings of almost all the agencies. Not only did public approach can be an aid to civil society groups that wish to goad satisfaction improve across the board, but problem incidence the government to perform better. s and corruption seem to have declined perceptibly in the routine transactions between the public and the agencies (Figures 1, 2, Samuel Paul is Chairperson, Public Affairs Center, Bangalore, India 3). It is clear that more decisive steps have been taken by the agencies to improve services between 1999 and 2003. The Handbook on Public Lessons Sector Performance Reviews edited by Anwar Shah, World Bank WHAT ACCOUNTS FOR thisdistinct turnaround in Bangalore's public services? And what lessons can we learn from this exper- MAKING THE PUBLIC SECTOR RESPON- iment? Needless to say, without deliberate interventions by the SIBLE, RESPONSIVE AND ACCOUNTABLE government and the service providers, no improvement would have taken place in the services. But the key question is, what Including tools for: made them act? A whole complex of factors seem to have been at Implementing citizen-centered governance work. The new Chief Minister who took over in 1999 was very Measuring and evaluating government concerned about the public dissatisfaction with the city's serv- performance in the delivery of local public services Governing for results and designing governments ices. He set in motion new mechanisms such as the "Bangalore for performance Agenda Task Force," a forum for public-private partnership that Implementing decentralized public management helped energize the agencies and assist in the upgrading of the Understanding governance approaches to services. The civil society groups and the media supported and managing conflict Rooting out corruption and malfeasance monitored these efforts. What is significant is that the initial Reforming bureaucracy trigger for these actions came largely from the civil society ini- Ensuring accountability when there is no tiative, "citizen report cards." bottomline What are the pre-conditions for such civil society initiatives Learning newer approaches to public sector evaluations to work? It is obvious that these initiatives are more likely to Tracking the success of health, education, succeed in a democratic and open society. Without adequate and infrastructure investments space for participation, CRCs are unlikely to make an impact. A Learning about better practices in tax, expenditure, and transfer policies tradition of activism within the civil society would also help. And much more! People should be willing to organize themselves to engage in advocacy and seek reforms supported by credible information. Volumes 1-3 available on the internet library at Political and bureaucratic leaders must have the will and w w w.decentralization.org and on CD-ROM. resources to respond to such information and the call for Contact Theresa Thompson at improved governance by the people. Tthompson@worldbank.org Volumes 4-6 will be coming soon! M A R C H 2 0 0 4 13 OUTLINE OF SOCIAL ACCOUNTABILITY MECHANISMS WBI's Community Empowerment and Social Inclusion Learning Program and the Bank's Social Development Unit are undertaking a global stocktaking exercise in the area of social accountability. This table reflects topics to be covered by the study. For more information, see www.worldbank/wbi/communityempowerment COMPONENTS OF SA MECHANISMS DEFINITION DESCRIPTION AND EXAMPLES INDEPENDENT Refers to the research, advocacy and BUDGETARY PROCESS ANALYSIS: evaluating the knowledge of and general BUDGET ANALYSIS dissemination of information on transparency surrounding the budgetary process (e.g. ADVA in Israel) (IBA) issues related to official budgets by GENERAL ECONOMIC ANALYSIS: Several groups look at whether the civil society and other actors macroeconomic assumptions and forecasts made to guide medium term independent of the government. policies like the PRSP are realistic or not. The goals are to: BUDGET EXPENDITURE ANALYSIS: includes tracking public spending to nanalyze the implications of compare the "planned" budget to the actual budget government budgets for different n Sector-Specific Analysis: examine the implications of the budget on stakeholder groups, particularly specific sector or area of public expenditure (health, education..) the poor and underprivileged n Analysis of Effects on Different Population Groups: such as expenditure nraise the overall level of budget oriented to the poor, the elderly or women(e.g. FUNDAR in Mexico). literacy amongst the general public n Revenue Analysis: a) assessing the macroeconomic effects of tax increases ninform legislatures and policy or cuts on the distributional effects of the tax system, b) connecting makers themselves so that they revenue and expenditure analysis, and c) evaluating the tax collection can engage in a more informed capacity of the government. and efficient budget policy debate. n Evaluating Public Services or Programs: assessing the quality and effectiveness of public services. n Advancing Budget Literacy and Providing Budget Training: providing accessible materials and training on budget-related issues. PARTICIPATORY BPET involves the use of civil society to BPET allows a continuous public involvement in the exercise. It is the actual BUDGET AND track how the public sector spends the users or beneficiaries of services (like parents of school-going children) that PUBLIC money that was allocated to it. This is collect data on inputs and expenditures, rather than some technical agency, usually done by `input-tracking' rather bureaucrat or external consultant. The results of the exercise are immediately EXPENDITURE than tracking of actual expenditures, disseminated to the public either via the media, or through publications in TRACKING since for most developing country the local language. This continuous transfer of information into the public contexts the most readily available domain either through an accompanying media campaign, public awareness data is on inventory records. drive, or via information dissemination and mobilization by CSOs is an integral part of BPET, which differentiates it from the more orthodox methods like public expenditure reviews (PERs). PARTICIPATORY The performance of the selected n Community Score Cards. This is a community based qualitative PERFORMANCE public funded agencies or projects are monitoring tool that draws on techniques or is sometimes called a hybrid MONITORING assessed and monitored by the of social audit, community monitoring and citizen report cards. The process community organizations/groups or is also an instrument for empowerment and accountability as it includes NGOs using predetermined indicators an interface meeting between service providers and the community that and systems that they themselves allows for immediate feedback. decide upon. It also includes n Citizen Report Cards. One of the more analytically robust and powerful communities or community groups instruments to monitor and evaluate public performance draws on a tracking of progress of projects and private sector practice of soliciting feedback from citizens and compiling sub-projects which they have `report cards'. (e.g. Bangalore Report Cards and the Bank-funded Filipino designed and implemented using Report Card (2000)) their own funds or government funded n `Participatory' or community based monitoring and evaluation. It is a (e.g. social investment funds). process of monitoring and evaluating projects that have been designed by the community or community groups themselves or by government agencies. PARTICIPATORY This involves direct citizen/CSO n Participatory policy formulation has become an increasingly common BUDGETING participation in formulating public trend, particularly with the introduction of the PRSPs at the national level budgets and community driven development initiatives at the local level. n Participatory budget formation, is less common and usually occurs at the local level (as in over 100 municipalities in Brazil or in Sirajganj district in Bangladesh). But is also theoretically applicable at higher levels. Another approach to participatory budget formation is when civil society actors prepare alternative budgets (such as South Africa's Women's Budget or Canada's Alternative Federal Budget) with a view to influencing budget formulation by expressing citizen preferences. SPECIAL REPORT Applying the Service Delivery Triangle to Care and Treatment for HIV/AIDS BY SIGRUN MØGEDAL HIV/AIDS in a health system context THE SERVICE DELIVERY TRIANGLE as presented in the 2004 World Development Report (WDR) offers a tool for exploring key AIDS IS TODAY A CHRONIC DISEASE that can be treated, with interactions in service delivery that can be usefully adapted to a growing caseload that weighs heavy on weak health systems. various contexts, services and service policies (Figure 1). This Access to care and treatment is both a right for the individual article is an attempt to apply the triangle concept to the chal- and a public good, with widespread benefits for society. This is lenge of scaling up care and treatment for AIDS. to a varying degree true also for other components of the M A R C H 2 0 0 4 15 health care package. But health systems are often poorly national HIV/AIDS authority with exceptional policy and co- equipped to cope with these basic needs and rights, and face ordination responsibilities, in addition to the regular func- competing demands. tions of the state, generates new challenges in terms of Much attention has in the past been given to reforming accountability (UNAIDS 2003). health system structures. In striving to reach the Millennium For service providers, for the AIDS response it is critical to Development Goals (MDGs), there is now a growing concern have clear contracts with policymakers, as it brings added plu- about performance and equity. This brings questions of access rality in service provision. In many cases the financial and barriers to access to the forefront. Most of the access bar- resources for "outside-the-box" decision-making and provi- riers are systemic in nature and therefore affecting all servic- sion are far greater than what is available for a regular health es, including AIDS. Through its association with stigma, rejec- sector service package. There are new challenges in terms of tion, and issues of power, poverty and gender discrimination, the supply chain, both in quality and continuity. AIDS kills AIDS adds to systemic barriers and creates additional ones. health personnel and weakens the manpower base. At the In spite of major efforts to simplify treatment routines, same time the multi-stakeholder response causes increased care and treatment for AIDS is more demanding than other competition for skilled manpower, moving health personnel elements of the basic health care package. It is demanding in from the public to the private, from the periphery to the cen- terms of continuity in the health care chain from periphery to ter, and from low-capacity countries to countries with better referral, both because of its demand for diagnostics and clin- terms of employment. This situation is a major challenge to ical judgment and because of overall cost--although the drug policymakers and provider organizations and calls for codes cost has now been dramatically reduced. of conduct to accompany market and purchase. The service delivery triangle is easily adaptable to this context, but Client power and the may need to be interpreted through a FIGURE 1--ACCOUNTABILITY RELATIONSHIPS AIDS response special "AIDS lense" in order to offer clues on options and choices. CLIENT POWER THROUGH PARTIC- IPATION, watch, and purchase, is a HIV/AIDS response key entry point for application of the and stakeholders service delivery triangle offered by the 2004 World Development Report to THE SERVICE DELIVERY TRIANGLE different types of services. The sets out key relationships of "eight sizes fits all" decision tree that accountability between the client, goes with the triangle is central for the provider, and the policymaker. Source: WDR 2004 determining which service delivery These stakeholder groups are basic arrangement would most improve to all service delivery and obviously provision, along variables such as also to the AIDS response. But the way the stakeholders inter- centralized/decentralized decision-making and public-pri- act and the forces and drivers in each of the three corners of vate provision (figure 2). When applying this decision tree to the triangle are not the same for all the elements of health AIDS, each step needs a modifier reflecting whether the poli- services. cy environment is AIDS-sensitive or not. On the demand side, this is reflected in heterogeneity in With AIDS-sensitive government policies I mean policies the client group. HIV/AIDS adds heterogeneity. Generally, that give visible political leadership to the HIV/AIDS one will expect to find important differences between demand response, fight stigma and discrimination, and are expressed for curative and preventive services and between "high status" in AIDS action frameworks and coordinating structures inclu- and "low status" diseases. Demand will also reflect gender and sive of non-governmental partners, particularly communities age differences, social and economic groups, ethnicity, dis- living with the disease. tance from the provider and so forth. AIDS adds new dimen- The first step of the WDR decision tree is to determine sions of marginalization, aggravates the gaps between the whether government is pro-poor or clientelist. While in terms non-poor and the poor and creates new and competing inter- of the AIDS response this is a useful distinction, in the case of est groups and patterns of demand. pro-poor policies it makes a big difference whether these At the level of policymakers, the challenges of an emer- policies are "AIDS-sensitive" or not. gency HIV/AIDS response in most heavily affected countries The basic assumption must be that providers are able to have led to the establishment of semi-independent, multi- carry our the AIDS response as an integrated component, which stakeholder national authorities, linked up to the government would be exceptionally designed, visible, and have high priori- structures at the prime minister's or presidential level. These ty. This means that in countries with pro-poor policies that are authorities serve as custodians of an exceptional national not AIDS-sensitive the health system cannot be the main serv- HIV/AIDS action framework, including policy co-ordination, ice provider. In such cases central government contracting or umbrella functions for various partnerships mobilized for central government provision without direct client input or AIDS action, and for monitoring and evaluation. Bringing in a client monitoring will have to be ruled out. These options may 16 Development Outreach WORLD BANK INSTITUTE not even be viable in a context of AIDS-sensitive policies, shows us that client participation and client power are critical because of the need for broad involvement of multiple stake- to make health services work, both for poor people and for an holders to deliver and manage an effective AIDS response. In effective AIDS response. countries with heavy AIDS load client groups become more het- erogeneous and thereby call for decentralized options. AIDS Demand and priority setting in the will also tend to make services harder to monitor. context of HIV/AIDS The mix of clientelist and AIDS-sensitive policies raises challenges, which do not fit easily into any of the boxes of the THERE ARE MAJOR QUESTIONS to address in the match decision tree. While seemingly a contradiction, it is important between supply and demand in the context of HIV/AIDS. to remember that government policies are often not coherent Individual demand for treatment and care may overpower and may be contradicting--across sectors and levels as well as public health imperatives, such as prevention. Stigma and within them. Country-level analysis along the lines of poverty discrimination require special measures to ensure that treat- orientation and AIDS-sensitivity is therefore important. ment is available for marginalized groups, even if not It can be assumed that both poverty-orientation and AIDS- demanded. In addition, the purchasing power, interest, and sensitivity are required in order to achieve the Millennium needs of the individual have to be weighed against well estab- Development Goals (MDGs). MDG monitoring may therefore lished criteria for effective public health in order to serve pur- be a useful tool to address issues of policy coherence. poses of common good. Strategies for basic health services as well as care and treatment for AIDS will need to be assessed as to whether they will useful- There is a definite risk that separate provision of treatment ly drive coherence, compatibility, efficiency, and sustainability for disease will risk reducing the demand for prevention and in order to achieve the MDG. Clientelism, combined with may lead to reduced coverage of the basic health package. And AIDS-insensitive policies in countries heavily affected, poses a while accountability between the provider and the client may serious risk scenario for any public health effort. Applying AIDS-sensitivity to the decision tree (Figure 2) S e r v i c e D e l i v e r y T r i a n g l e c o n t i n u e d o n p a g e 2 9 FIGURE 2--EIGHT SIZES FIT ALL Source: WDR 2004 M A R C H 2 0 0 4 17 SPECIAL REPORT Pro-Poor Health Services The Catholic Health Network in Uganda BY DANIELE GIUSTI, PETER LOCHORO, ed through umbrella organizations, such as the Catholic, JOHN ODAGA, AND EVERD MANIPLE Protestant, and Muslim Medical Bureaus and the Uganda Community Based Health Care Association. MOST PRIVATE NOT-FOR-PROFIT (PNFP) health providers During the era of socio-political upheaval and economic in Uganda are faith-based. They account for a sizeable pro- decline in the 1970s and 1980s, the PNFP sector continued to portion of the health services delivered in the country operate using several coping mechanisms aiming at cost con- (Hutchinson 2001) and have as their prime concern the pro- tainment--such as underpayment of personnel, reliance on vision of services to the poor. During the last century this sec- unqualified staff, increasing working hours, and disregard for tor expanded substantially, especially in rural areas, providing depreciation of capital assets and their maintenance--on the services at subsidized prices, thanks to the solidarity of sister one hand, and at increasing reliance on support from external churches and denominations. These providers are coordinat- charities, on the other hand. 18 Development Outreach WORLD BANK INSTITUTE By mid-1990s both approaches showed their limitations. management take a critical look at their performance and Attempts to restore physical and human capital occurred at the compare it to that of the whole sub-sector, its peer group, and time when new service standards were introduced and when outliers. In the absence of realistic "gold standards" of per- the civil service reform was increasing public sector salaries, formance, this approach stimulates comparisons and triggers creating considerable pressure on the PNFP sector. processes of steered change in an environment long dominat- Furthermore, the AIDS crisis became apparent, and inputs ed by crisis management. from private charities started decreasing. The only possible At the end of 2001/02, the following picture could be option was a heavier reliance on user fees. But this caused a detected from the performance indicators: rapid decrease of utilization, accompanied by efficiency losses. n The volume of service outputs, measured by standard units of This article documents the experiences of the Catholic outputrelativetooutpatients(SUOop),wasincreasingsteadi- health network in Uganda and its umbrella organization, the ly, but so was the cost of the services produced. SUO op is a Uganda Catholic Medical Bureau (UCMB) in making health composite activity index weighted by the cost of each activity. services work for poor people. It demonstrates how the pro- n The productivity of staff (SUO op per staff), remained stable, poor ethos--derived from a longstanding tradition and the mis- and in some cases showed a decline. sion of "healing by treating and preventing diseases, with a n Fees (per SUO op), which had been decreasing for three preferential option for the less privileged"--supported by "soft" years, had started increasing again. regulation and technical assistance from the umbrella organi- The interpretation that the UCMB gave to these observa- zation can induce a process of change in a network of providers. tions was that the expansion of service volume had not been accompanied by the necessary efficiency gains, and that Preventing a crisis access by the poor was still a problem. BY MID-1990s many providers in the PNFP health sector were Accelerated reduction of user fees no longer able to cope with the increasing cost of service pro- duction. The UCMB estimated that, on average, the cost of AN EARLIER STUDY commissioned by the UCMB had revealed producing one unit of output was increasing by 20 percent per that a small group of hospitals in the network was operating at year. Given the importance of the PNFP sector, a major crisis higher efficiency levels (Amone and others, 2000). A common in health care countrywide was to be expected. A particular factor for them was that they had introduced and maintained concern was that higher fees reduced access for the most vul- lower and flat fees for selected target groups (children, moth- nerable groups: women, children, and the poor. A clear sign of ers, and female patients). In 2000 a severely underutilized the crisis was the decreasing volume of services delivered, hospital that had reached the verge of closure, was encour- expressed in standard units of output relative to outpatients aged--on the basis of the evidence gathered through the study-- (SUO op). Data prior to 1997/98 showed a steep decline of var- to flatten and lower fees for children and pregnant women. The ious activity indicators; in several hospitals utilization was in a only support the UCMB was able to provide was a moderate free fall. Also recognizing the signs of a crisis, the government degree of managerial assistance and the guarantee that finan- agreed to support the PNFP sector by grants disbursed direct- cial losses, if occurring, would have been met through external ly to individual providers (hospitals and health centers). Since aid. The experience of this hospital was a shocking revelation. then, subsidies have increased from about 5 percent of the In a few months patients started flocking back, without causing operational cost of the sector to about 35 percent. Today the any financial shortfall (Santini, 2002). sector relies on user fees for about half of its operational costs. Given this experience, the Bureau launched in 2002 a This intervention--a private-public partnership--was able to "strategy for accelerated reduction of user fees," proposing it to prevent a fatal crisis, reduce reliance on user fees, and, as a all PNFP hospitals. In a nutshell, hospitals were asked to: result, increase utilization. n Pay a closer look at fees actually paid by patients, with a sys- tematic approach (that is, a baseline survey and follow-up); Developing monitoring tools n Reduce and/or flatten fees for children, mothers, and women; and THE UCMB WAS AWARE of the need to monitor closely devel- n Monitor utilization and provide this information to the opments in the delicate phase of transition. It was also aware UCMB for aggregation across hospitals. that it needed to account for the public subsidies. For these reasons the Bureau undertook, since 1997/98, a systematic Results data collection effort from the affiliated hospitals, covering both inputs and outputs. The analysis of these data has per- IN LESS THAN ONE YEAR from the consensus reached among mitted the observation of trends in critical parameters (listed all affiliated health units in the Catholic health network, the below). Since 2000/01 each hospital has received an annual following results could be documented: feedback report, containing information on its own perform- n The annual increment of outputs has been the highest reg- ance compared to other hospitals for each of the selected indi- istered in the last 5 years cators. The aim of the feedback report is to help the hospital n For the first time in 5 years unit cost has decreased (Figure 1) M A R C H 2 0 0 419 n Productivity of staff has increased in the last year FIGURE 1--MEDIAN VALUES OF COST PER 1 SUO op IN 27 UCMB HOSPITALS (Figure 2) Ugandan schilling n Fees per unit of output, which had started increasing again in 2001/02, have decreased and are at the lowest value in 5 years ( Figure 3). Results are all the more remarkable if one considers that, contrary to the above pilot hospi- tal, the UCMB was not able to offer any guarantee that financial shortfall could be met by external aid. Hospitals had to carry the risk themselves and financial subsi- dies from government, which had been growing rapidly in the previ- ous years, were leveling off. These results point to the fact that intrinsic motivation, or the FIGURE 2--MEDIAN VALUES PRODUCTIVITY OF PERSONNEL IN 27 UCMB HOSPITALS "ethos" of these hospitals, matters. A similar inference was made for the PNFP lower level units by a comparative study of PNFP, private for-profit, and public health units in Uganda (Reinikka and Svensson, 2003). When an appeal to this ethos is accompanied by a rational argument, evidence, and some know-how, major changes are possible. The results docu- mented thus far also provide some evidence that the hospitals have become more pro-poor. Although fee reductions were not targeted at the poor, it is known that utiliza- tion by the poor is more elastic to fee adjustments (McPake, 1993). This implies that fee reductions disproportionately favor the poor. FIGURE 3--MEDIAN VALUES OF FEES CHARGE PER 1 SUO op IN 27 UCMB HOSPITALS Since the largest majority of Ugandan schilling Catholic health units operate in rural environments and a sizeable number of them in war-torn areas, itcouldeasilybeinferredthattheir increased utilization included more patients from poor socio- economic categories. Improving information TO PROVIDE further docu- mentation of the processes under way, strengthen the rational argument, and improve the know 20 Development Outreach WORLD BANK INSTITUTE how in view of limiting the degree of risk that hospitals have to accept, the UCMB with the other denominations' bureaus, went a step further and carried out a study in a sample of affiliated References hospitals (Odaga and Maniple, 2003). Although the study did Amone, Joseph, Salome Asio, Adriano Cattaneo, Annette Kakinda not yield conclusive results, it provided additional insight in the Kweyatulira, Anna Macaluso, Gavino Maciocco, Maurice Mukokoma. 2000. dynamics at play and led to some interesting findings, summa- "User fees in private not-for-profit hospitals in Uganda." Health and Development, Supplement to no. 2. CUAMM Organization, Padua, Italy. rized here, and communicated to all hospitals: Republic of Uganda. 2001. "Facility-based private not-for-profit health n Themajorityofthehospitalshaverespondedbyreducingfees providers: A quantitative survey." Kampala: Ministry of Health. and adopting a flat fee structure. Only 8 percent of hospitals havenotyettakenadecisiontodoso.Mosthospitalshavetar- Hutchinson, Paul. 2001. "Combating illness," in Uganda's Recovery: The geted services towards pregnant mothers and children. Role of Farms, Firms, and Government, Ritva Reinikka and Paul Collier, eds. Regional and Sectoral Studies, Washington, D.C.: World Bank. n Fee reductions have generally resulted in increased utiliza- tion of all targeted services, and have especially favored McPake, Barbara. 1993. "User charges for health services in developing countries: A review of economic literature," Social Science and Medicine, children. vol. 36, no.11. n Responsiveness in maternal services was consistently low, Odaga, John and Everd Maniple. 2003. "Faithfulness to the mission: Effect implying presence of other important barriers. of reducing user fees on access to PNFP health services." Study Report, n Flattening of the fee structure reinforced the effects of fee Uganda Martyrs' University. October. reduction. However, the flat fee concept was not appreciat- Reinikka, Ritva and Jakob Svensson. 2003. "Working for God?" Evaluating ed the same way in all the hospitals. service delivery of religious not-for-profit health care providers in Uganda." n Response was stronger where the community could under- Policy Research Working Paper 3058, World Bank, Washington, D.C. May. stand (and perhaps predict) the new fee structure. Although Santini, Stefano. 2002. "The Naggalama Initiative: How to revitalize a hospi- most hospitals reported to have displayed their fees publicly tal," UCMB Bulletin, vol 4, no. 2, Kampala, Uganda. in their premises, this alone was not effective in informing the community. Methods that reach out to the community were found to be more effective. n Many hospitals still lack the technical competence to mon- itor the process of user fee reductions, even though most of them reported to have them in place. Conclusion MAKING HEALTH SERVICES WORK for poor people is not easy. The 2004 World Development Report provides an extensive analysis of the actors and factors at play. It has also suggested that with altruistically motivated providers the inherent pro- poor ethos can be banked upon if combined with regulation and support. This article has documented how this can indeed happen. We would only like to add that managerial know-how is necessary to limit the risk that service providers take when they decide to be more pro-poor, in an environment where poverty is widespread and resources are extremely limited. Umbrella organizations can collect and make use of informa- tion to strengthen the pro-poor ethos and enhance self-regu- lation. They can also provide the much-needed additional managerial capacity so often wanting at the implementation level in a resource constrained environment. s Daniele Giusti, MD MPH, and Peter Lochoro, MBChB MSc HSM, Uganda Catholic Medical Bureau (UCMB); Everd Maniple, MBChB MPH and John Odaga, MBChB MPH, Uganda Martyrs' University, Department of Health Sciences. dgiusti@ucmb.co.ug M A R C H 2 0 0 4 21 SPECIAL REPORT Scaling Up Drinking Water Services third richest nation in the world to the fifteenth. This decline BY JUNAID K. AHMAD, DAVID SAVAGE, AND drove successive governments in the 1980s and 1990s to ini- VIVEK SRIVASTAVA tiate wide-ranging economic reforms, including reform of public utilities in the infrastructure sector. WORLDWIDE THERE HAVE BEEN MANY SUCCESSES in the In this context, a compact was signed in the Council of delivery of drinking water services. The reform of Phnom Australian Governments between different tiers of govern- Penh's public utility in Cambodia and Cartaghena's water pri- ment--the federal tier, the Commonwealth, and the states--to vatization in Colombia are outstanding examples. But, there create an economically viable and ecologically sustainable are also cases of dramatic failures--Cochabamba, Bolivia, or billions water industry. The states decided to restructure their public Bangladesh's arsenic crisis. These successes and failures offer agencies in the sector on the basis of agreed upon principles. a host of lessons for reformers, but one stands out in particu- US$ These included introducing commercially viable and justifi- lar: the success of service delivery depends on whether insti- able water pricing; costing and transparency of (cross) subsi- tutions of service provision are accountable to citizens. The dies; institutional reforms of public monopolies to achieve challenge is thus "not to fix the pipes, but to fix the institu- separation of key institutional roles--policy, service delivery tions that fix the pipes." and regulation; performance monitoring; intensive public Institutional changes, which ensure that service providers consultations and education; and measures related natural are accountable to all citizens, are a highly political endeavor. resource sustainability, including allocation and trading in Undertaking such changes on a pilot basis is difficult enough; water entitlements. scaling it up across jurisdictions and sustaining it over time, is a changes In view of Australia's federal structure, the actual reform daunting challenge. Yet, it is precisely the scaling up of institu- path--the implementation of the principles--was left to the tional change that is needed to ensure that the goal of universal discretion of the states. Different states and territories are access to basic services can be realized. How can governments regulatory introducing the reforms at different rates and in different ensure that all citizens have access--and quickly--to a basic level of ways. For examples, some states have viable state-wide utili- of water and sanitation services, that these services are provid- ties, others use local government utilities, some have state ed on a sustainable basis, that service providers respond to a Numberwide regulators other rely significantly on performance mon- variety of consumer preferences across income levels, and can itoring, and so on. These differences in implementation adapt endogenously to changing circumstances? approaches reflect the differences in the initial legislative, The historical experience of three countries--Australia, economic, and political conditions of the separate jurisdic- India, and South Africa--offer important insights into the tions. But in all cases, the institutional choice adopted by the process of institutional change in service delivery and the states is defined by the agreed upon principles. challenges of scaling up of service provision. The country set- As an incentive, the Commonwealth makes fiscal payments tings--demographics, political systems, geographical size, to the states and territories for achieving the milestones of economics, and income levels--are different; but it is precise- reform. Although not large--less than one percent of the over- ly this diversity that enables us to draw common lessons and all budget of the Commonwealth and states--the fiscal trans- principles from their experiences. fers provide sufficient incentive at the margin for the states to stay committed to the reform program. The initial reform The case of Australia timetable was optimistic and underestimated the complexity of the reform program requiring extensive research and RESTRUCTURING IN THE AUSTRALIAN WATER SECTOR was analysis for effective implementation, the need for extensive initiated in the context of broader economy-wide reforms. consultative and educative processes and the demand that Between 1960 and 1992, Australia slipped from being the these reforms placed on governments, institutions, and M A R C H 2 0 0 4 23 stakeholders. But, after more than a decade of sustained insti- and now to the proposed local government and community tutional reform, Australia's water sector has become an management of water and sanitation, reflects changes in the important model for infrastructure reform. politics of India. While federal in structure--with a central government and several states--India adopted a centralized The case of rural India economic and a political model after independence. Over time, broader economic liberalization and strengthening of RURAL INDIA , WITH 700 MILLION PEOPLE, is currently the federal system, including the introduction of local govern- undergoing a major reform in its drinking water and sanita- ments into the Constitution about a decade ago have changed tion sector. A state subject under India's Constitution, rural the overall view of how services should be managed. Water and water services have traditionally been provided through state sanitation will not be an exception to these fundamental water boards and departments. Water has been treated as a changes and success in scaling up service delivery will depend social good by the state agencies, supplied without any user on the political, fiscal, and administrative relationship charges or local stakeholder involvement in the delivery between the different tiers of governments. process. Top down in their approach, the boards were profi- cient at delivering hardware, but less inclined to undertake The case of South Africa the operations and maintenance and manage service delivery. This was not surprising, as the boards were underwritten by SOUTH AFRICA'S WATER SERVICES SECTOR has undergone Central and State budgetary outlays regardless of perform- far-reaching transformation since the advent of democracy in ance. The bias was for greater spending and often spending 1994. This transformation process has been underpinned by a captured by political interests. strong commitment to eliminate the inequities of South Drawing on the lessons learned from smaller projects--often Africa's apartheid past, but driven by a broader program to financed in partnership with donors--central government in decentralize functions and finance to local government. 1999 piloted the Sector Reform Program across all the states Although water services are defined as a local government covering 70 million people. In the SRP, communities were function in the Constitution, the national government initially mobilized through user group committees to determine their took the lead in expanding access to services. The Community choice of service standard and process for managing service Water Supply and Sanitation program was developed as a provision. Capital costs were shared between central govern- national investment program, implemented directly by the ment and communities, but communities were expected to pay Department of Water Affairs and Forestry (DWAF) or through fully for operations and maintenance. Water was treated as an nationally owned water boards or NGO's. Initially, in the con- economic good--valued for its use--thus increasing accounta- text of weak local government authorities, community man- bility to communities and enhancing their sense of ownership. agement structures were established--although considerable The SRP experience showed greater sustainability of services emphasis was placed on large regional scheme investments. and more responsiveness to consumer preferences. But, scaling The restructuring of the local government system through up of community delivery systems has proven to be a challenge-- re-demarcation of municipal boundaries and a redefinition of theproblemsofmanagingcollectiveaction,capacityconstraints, local government functions created a structural conflict and addressing the political context of existing state agencies are between community management structures and elected local important policy issues that need to be addressed. Increasingly, government authorities. This occurred at a time of growing localgovernments--panchayats--arebeingseenasthekeylinkto concern over the cost and sustainability of the national invest- managing a process of scaling up the SRP. ment program. A program of fiscal decentralization, which State governments are now being asked to develop a plan involved the consolidation and decentralization of operating for shifting drinking water services in rural areas to a com- and capital transfers to local government, sought to address pletely demand-driven, participatory approach where com- these problems through placing accountability for investment munities and local governments are partners in the provision decisions with the operating authority and ensuring greater of drinking water. In the process, states are being asked to re- coordination between investment programs of all spheres of define the role of the water agencies. The principles are fixed, government. but the state governments have the full flexibility to develop Although DWAF has begun to transfer schemes it has built their individual approaches to implementation. Once and operated to local government, concerns about their endorsed by the political leadership of the states, the plans capacity to manage water distribution remain. Recognizing will form the basis of a Memorandum of Understanding the positive relationship between functional assignments and between the state and central government, where the center actual capacity DWAF has responded to the new environment will support the funding of the MoU against pre-agreed mile- through re-positioning itself as a "developmental regulator" stones of change proposed by the states. In addition, the cen- of water services, rather than an investment agent or service ter will provide funding for capacity support and clear moni- provider. This focus on capacity building efforts at local gov- toring and evaluation of the process with the possibility of ernment level includes a policy framework that distinguishes benchmarking performance of the states. between "authority" and "provider" functions and thus allows local authorities to engage other agencies, whether public or The transition from state agencies, to community systems, private sectors, to undertake actual provision of services 24 Development Outreach WORLD BANK INSTITUTE under contract to the local authority. DWAF has also placed communities becomes essential. considerable emphasis on building local planning and infra- Separation of roles is also an essential element of the structure investment capacity and on developing effective reform process. Within the sector, the policy making, regula- monitoring and evaluation systems. tory, and service provision functions need to be kept in sepa- rate organizations and processes. Equally important, separat- Some Lessons ing the powers and roles of different tiers of government is essential to enable the center to play a role of providing incen- ACCOUNTABILITY TO CLIENTS is at the heart of successful tives for change. But this requires that the fiscal rules of the service provision. How to expand this relationship of account- game between different tiers of government are clear and ability across the sector is central to the challenge of scaling up binding. In Australia, and increasingly South Africa, this is drinking water services. How can this be brought about? the case. In India, the reforms of rural drinking water may Australia, India and South Africa offer many lessons. well be delayed because the rules of India's fiscal federalism To begin with, scaling up of services is a political process. are still in a state of flux. Importantly, the evolution of multi- In Australia the broader forces of liberalization started it; in tiered governments may well facilitate the separation of poli- India it is the gradual opening up of the economy and the cymaking from regulation and service provision, which is democratic forces of federalism; and in South Africa it was the essential to the sector--the two processes are interlinked. end of apartheid and the emergence of a completely new dem- With the separation of roles emerges the potential for ocratic setting. These changes are complex and take time. independent benchmarking of providers and tiers of govern- None were big bang approaches to service reform and none ment responsible for service delivery. In effect, information saw the politics of reform in the water and sanitation sector becomes an important tool for catalyzing and sustaining come from within the sector. reforms, and one, which is being widely used in Australia and Where the politics are not conducive to scaling up reforms, increasingly in the other countries. innovative pilots may have a demonstration role to play. But as Finally, reforms require mobilization of citizens and the case of SRP in India suggests, scaling up pilots is not sim- capacity support to governments and communities to under- ply a replication of the pilots themselves. While pilots suggest stand and undertake the changes. Central governments can important principles, the wider implementation of the prin- play a proactive role in designing such systems of capacity ciples may require a very different approach, as exemplified support. But the lesson is clear: such support is best delivered by the shift from SRP to a MoU between different tiers of gov- in the context of on-going reforms. s ernment--similar to Australia's federal-state compact In all countries, valuing water as an economic good and Junaid K. Ahmad is with the Water and Sanitation Program, introducing some level of user charges were essential to get- South Asia (WSP-SA). ting providers to become more accountable to consumers-- David Savage formerly with National Treasury, South Africa, rich and poor alike. This may well be a sine qua non of achiev- now works in WSP-SA. ing universal services. Once the value of water is reflected in Vivek Srivastava is with the Africa Region, The World Bank. the transaction, allowing choices about standards and service wspsa@worldbank.org delivery organization to be dictated by local preferences of M A R C H 2 0 0 4 25 SPECIAL REPORT Randomized Evaluations of Interventions in Social Service Delivery BY ESTHER DUFLO, RACHEL GLENNERSTER, many reasons. Non-governmental organizations (NGOs) and AND MICHAEL KREMER governments can use this evidence to focus their limited budgets on those programs that are most effective. With wide- WHAT IS THE MOST EFFECTIVE WAY to increase girls' partic- spread cynicism about the effectiveness of aid, providing clear ipation in school? How can we reduce the high absence rates evidence on the impact of different programs can also help of teachers in many developing countries? We know surpris- galvanize support for more development assistance. ingly little about these questions, but finding answers is cru- cial to improving the quality of education in developing coun- Improving access to social services: the tries and the effectiveness of aid. Every year, millions of dol- example of school participation lars are spent on evaluating development programs but these evaluations tend to focus on process: did the money go where THE MILLENNIUM DEVELOPMENT GOALS call for universal it was meant to? How many teachers attended training cours- primary school enrollment by 2015. However, until recently es? How many textbooks were delivered to schools? While there were no good assessments of how best to achieve tracking performance at this level is important, we should also increased participation in education or how much it would cost. be evaluating programs at a more fundamental level to find out Recent research suggests that a simple means of increasing whether, for example, training teachers or buying textbooks school participation is by reducing the costs of schooling, or does more to raise test scores. even paying for attendance. The PROGRESA program in These fundamental questions, however, are hard to Mexico provided cash grants to families if their children answer. Imagine, for example, that a new headmaster arrives attended school regularly and received preventative health at a school full of enthusiasm and new ideas. He wants to get care like vaccinations. Schultz (forthcoming) takes advantage the parents involved and sets up a parent committee. Word of the fact that the program was phased-in in different areas spreads that the new headmaster is good, and some children randomly to assess its effectiveness. He finds an average transfer into the school from other local schools. If an attempt increase in enrollment of 3.4 percent for all students in grades were made to evaluate the impact of the parents' committee, it 1 through 8, and 14.8 percent among girls who had completed would be extremely difficult to disentangle the effects of the grade 6. In part because the randomized phase-in of the pro- parents' committee, the impact of the headmaster's enthusi- gram made the benefits so clear, the Mexican government asm on the other teachers in school, and the influx of new stu- expanded the program, and similar programs are now being dents who might be more motivated than average. introduced elsewhere in Latin America. The cleanest and clearest way to establish how a program Randomized evaluations performed in the same setting would work in an average school is to run a randomized trial. provide an opportunity to compare the cost effectiveness of Choose 100 representative schools, initially establish a par- various interventions. A series of evaluations in Kenya is a ents' committee in half, phase in the program later in the good example which provides comparative costs of different other half, and in the meantime compare the outcomes in the ways to increase school attendance. The most cost-effective two groups of schools. This approach, however, requires that approach was to implement a twice-yearly school-based mass evaluation be built into the design of the original program and treatment with an inexpensive deworming drug. As the chil- that data be collected on all 100 schools, which can be expen- dren's health improved, so did their attendance, and it cost sive. However, this is what we do if we want to know whether a only $3.50 per additional year of schooling induced (Miguel drug or vaccine is effective, and new research is showing this and Kremer 2004). Providing school meals to pre-schoolers technique can teach us a lot about development. cost an average of $36 per additional year of schooling Having strong evidence about what works is important for induced, and also improved test scores in schools where 26 Development Outreach WORLD BANK INSTITUTE teachers were well-trained prior to the program (Vermeersch randomized evaluations suggest that providing additional 2002). In contrast, even under optimistic assumptions, a pro- inputs (such as textbooks or flipcharts) to under-resourced gram which provided free school uniforms (among other schools can improve learning and test scores (Glewwe and inputs) cost $99 per additional year of schooling induced others 2002; Glewwe and others, forthcoming). However, (Kremer and others 2003). results from a randomized evaluation point to a subtler pic- Overall, these results suggest that school participation is ture: provision of textbooks increased test scores only among quite elastic to cost and that school health programs may be students who had scored in the top 40 percent on pre-tests one of the most cost-effective ways of increasing school par- prior to the program, and did not affect scores for the bottom ticipation. Is access to health care similarly elastic to cost? We 60 percent of students. Flipcharts had no impact on test will only know for sure when we have conducted a randomized scores. This result shows how misleading the results of non- evaluation of the policy. randomized evaluations can be when, for example, inputs are more available in richer schools (as was the case with Improving the quality of social services flipcharts in this study). It would be wrong to conclude from these studies that pro- WHILE THESE RESULTS SHOW that improving access to social viding inputs is necessarily ineffective. Banerjee and others services is relatively easy, other evaluations have shown that (2002) conducted a randomized evaluation of a remedial edu- improving the quality of education is more difficult. Non- cation program run by an Indian NGO. The program hired M A R C H 2 0 0 427 young local women from the communities to provide remedial reserved for these minorities. To avoid any possible manipu- education to students who were identified as lagging behind in lation, the reserved positions were randomly allocated, allow- traditional classes, and was found to have substantial positive ing for a clear evaluation of the program. Chattopadhyay and impacts on learning, particularly for the weakest students. On Duflo (forthcoming) find that in villages reserved for women, average, student test scores increased by 0.39 standard devia- the public goods chosen better reflect women's needs, and in tions, and gains for the bottom third of students were 0.6 stan- villages reserved for the scheduled castes, a larger fraction of dard deviations, a very large impact. The program was shown to the public goods is allocated in the scheduled caste hamlet. be at least 6 times more cost-effective than computer-assisted learning implemented (and evaluated) in the same schools. Conclusion These studies show that intuition about what works and what does not can be misleading. Substantial amount of money can THE RESULTS DESCRIBED ABOVE offer both substantive and be saved by finding out which inputs work and which do not. methodological lessons. Inexpensive health programs, reduc- ing the cost of schooling to households, or providing meals Improving governance can substantially increase school participation. Given the existing governance problems in many developing countries, ONE OF THE REASONS WHY PROVIDING MORE INPUTS is simply providing more resources may have a limited impact not always effective is that social service delivery in developing on social service quality, unless the resources are carefully countries is plagued by high absence rates and low effective- allocated. Systemic reforms (such as school choice, or decen- ness. If teachers don't show up to school, the benefit of more tralization and involvement in decision making by disadvan- textbooks may be limited to the few who can read well at the taged groups) have important impacts. start. How to reform the governance of social services, however, These results also show that randomized evaluations of is a complicated and hotly debated question. Several recent policy programs can be implemented successfully. They can studies involving randomized evaluations shed light on policies take place in the context of small NGO or pilot programs, lot- such as school reform, decentralization, and incentives. teries, or where a policy is phased-in over time. The results A Colombian program provided vouchers for private can be in sharp contrast to conventional wisdom and the schools through a random lottery (due to budgetary con- results of more traditional evaluations. straints), which thus allowed for credible estimates of program Finally, good evaluation promotes good policy. The positive impact (Angrist and others 2002). Lottery winners were 15-20 results found through the rigorous evaluation of the PROGRE- percent more likely to attend private school, 10 percent more SA program led to the implementation of similar programs in likely to complete 8th grade, and scored the equivalent of a full other Latin American countries. The NGOs that have partici- grade level higher on standardized tests (note that the vouchers pated in the evaluations discussed in this article have used the were renewable, conditional on satisfactory academic per- results to focus their resources where they are most effective. formance). The effects of the program Creating a culture in which rigor- were larger for girls than for boys. ous randomized evaluations are pro- Winners were substantially more moted, encouraged, and financed has likely to graduate from high school the potential to revolutionize social and scored higher on high school policy during the 21st century, just as completion or college entrance More detail on randomized randomized trials revolutionized exams. Overall, the benefits of this evaluations can be found at medicine during the 20th. program to participants clearly www.povertyactionlab.com, the exceeded the additional cost relative website of the new center at the Esther Duflo is Professor of Economics, to the alternative of providing places Massachusetts Institute of Department of Economics, MIT, and one in public schools. Technology devoted to encouraging of the founders of the Poverty Action Lab In India, an important effort to the use of randomized evaluation as at MIT, eduflo@mit.edu. decentralize social service delivery a way to improve the effectiveness Rachel Glennerster is Director of the started in the mid-1990s. Local vil- of poverty programs in advanced Poverty Action Lab at MIT, lage councils, elected every five years, rglenner@mit.edu. have the power to decide how to allo- and developing countries. Michael Kremer is the Gates Professor of cate expenditure on local public Randomized evaluations have been Developing Societies, Department of goods. The constitutional amend- applied in many areas besides social Economics, Harvard University, ment that set up this program also service delivery, such as savings and mkremer@fas.harvard.edu. required that one-third of all posi- credit, discrimination, and tions be reserved for women, and that agricultural extension. a share equal to the representation of disadvantaged minorities (scheduled castes and scheduled tribes) be 28 Development Outreach WORLD BANK INSTITUTE References Shultz, T. Paul. "School Subsidies for the Poor: Evaluating the Mexican PRO- Angrist, Joshua, Eric Bettinger, Erik Bloom, Elizabeth King, and Michael GRESA Poverty Program," Mimeo, Yale University, 2001, forthcoming in Kremer. "Vouchers for Private Schooling in Colombia: Evidence from a Journal of Development Economics. Randomized Experiment." American Economic Review, 92(5): 1535-1558, Vermeersch, Christel. "School Meals, Educational Achievement, and School 2002. Competition: Evidence from a Randomized Experiment," Mimeo, Harvard Banerjee, Abhijit and Michael Kremer; with Jean Lanjouw and Peter Lanjouw. University, 2002. "Teacher-Student Ratios and School Performance in Udaipur, India: A Prospective Evaluation," Mimeo, Harvard University, 2002. Chattopadhyay, Raghabendra and Esther Duflo. "Women as Policy Makers: Evidence from an India-Wide Randomized Policy Experiment," NBER Working Paper #8615, 2001, forthcoming in Econometrica, 2004. Glewwe, Paul, Michael Kremer, and Sylvie Moulin. "Textbooks and Test Scores: Evidence from a Prospective Evaluation in Kenya." Mimeo, Harvard University, November, 2002. Service Delivery Triangle Glewwe, Paul, Michael Kremer, Sylvie Moulin, and Eric Zitzewitz. c o n t i n u e d f r o m p a g e 1 7 "Retrospective vs. Prospective Analyses of School Inputs: The Case of Flip Charts in Kenya," NBER Working Paper #8018, 2000, forthcoming in Journal of Development Economics. Kremer, Michael, Sylvie Moulin, and Robert Namunyu. "Decentralization: A be strengthened by increasing the client's purchasing power, Cautionary Tale," Mimeo, Harvard University, March, 2003. this measure in itself may lead to a service that offers only the Miguel, Edward and Michael Kremer. "Worms: Identifying Impacts on Health components clients want to buy. and Education in the Presence of Treatment Externalities," Econometrica, Private provision for one disease, such as AIDS, in a sys- 72(1): 159-217, 2004. tem with largely public provision for other diseases will also require careful monitoring to avoid major inefficiencies and imbalances in investments (infrastructure and technology) and skilled personnel. Conclusion THE SERVICE DELIVERY TRIANGLE presented in the 2004 WDR does not offer solutions to the many specific challenges of the AIDS response, but helps clarify them and link them to options for general service provision. The role of the policy- makers, politicians and the state in setting out an enabling policy framework is essential, also for the AIDS response. The need for corrective measures and watch through client and community mechanisms that can guide choices on the basis of quality and the common good is strengthened in the context of AIDS. AIDS adds challenges to transparent priority setting and makes it harder for clients to hold providers' and policy- makers accountable. But it also has the potential for directing policymakers' and providers' attention to the need for becoming more responsive to demand and overcoming barri- ers to access. In this way the AIDS response offers opportuni- ties for new solutions and new alignment of interests to pro- mote better accountability and service delivery. s Sigrun Møgedal is Senior Adviser, NORAD, and Senior Policy Adviser, UNAIDS sigrun.mogedal@eunet.no References UNAIDS 2003. Coordination of National Responses to HIV/AIDS. Guiding principles for national authorities and their partners. M A R C H 2 0 0 4 29 SPECIAL REPORT Aid Agencies and Aid Effectiveness BY BERTIN MARTENS AID EFFECTIVENESS IS OFTEN LOOKED AT from the point of view of performance in the recipient country: are programme objectives achieved in a sustainable manner, have poli- cy reforms been implemented, and do they produce results in terms of economic growth and poverty reduction? This view is biased in the sense that it clears the aid suppliers of all responsibility and puts the burden of effec- tiveness squarely on the recipient. It is also simplistic because it assumes that the sole objective of all individuals and organizations involved in the aid delivery process is the realisation of the program objectives. While some of these may identify closely with the stated program objectives, others may pur- sue different objectives. Commercial suppli- ers and consultants pursue profit motives. Academic experts pursue their academic interests. Political decision-makers aim for political success, and civil servants seek to advance their careers. This article presents an approach to aid effectiveness that takes into account an array of different motives. It aims to highlight some of the structural and organizational problems that occur in the aid delivery process and affect aid effectiveness. A broken feedback loop WHEN A SUPERMARKET ADVERTISES "we aim to serve you," the shopper knows that it really aims to maximize profits and returns for its shareholders. But she also knows that, to achieve its aims, the supermarket will make efforts to keep her satisfied and deliver the products and services that she wants. The shopper pays for these products, and that gives her leverage over the supermarket. In 30 Development Outreach WORLD BANK INSTITUTE the case of aid, the situation is different. The beneficiaries do tangible outputs delivered by their agents, such as roads built, not pay for the products and services they get. Donors in other than for less tangible outputs, such as institutional reforms. countries foot the bill and decide on the terms and conditions Also, it is generally easier to verify project inputs than outputs under which the aid is given. Decisions are taken in donors' and impact. The longer is the aid delivery chain, and the less political constituencies where beneficiaries of aid have no vot- tangible the outputs, the higher is the risk of diversion from ing rights. In other words, the feedback loop between benefici- the original aid objectives. aries and decision-makers is broken (Figure 1). Unlike super- market customers, beneficiaries of aid have little leverage over Why not shorten the aid delivery chain? the quality and quantity of aid delivered. Naturally, donor coun- try politicians and civil servants in donor agencies feel insulat- IF THE LENGTH OF THE AID DELIVERY CHAIN is part of the ed from the wishes of their "customers". No wonder then that effectiveness problem, why not shorten it? One extreme solu- recipients do not always appreciate the products and services tion would be to cut the entire chain and transfer aid directly donors deliver, or that recipients' response is not what donors from donors to recipients. But this would run into transaction expect. cost problems. Donors would have to spend considerable resources on identifying potential recipients, assessing their Many principals and agents needs and delivering the funds. Aid agencies benefit from economies of scale in information gathering about potential MODERN ORGANIZATION THEORY can help us understand recipients and handling of transfers on behalf of donors--they the aid delivery process and aid effectiveness. It looks at large- reduce transaction costs. However, if that would be the only scale organizations that are composed of individuals who are role of aid agencies, we could still do with a very short chain of linked through hierarchies and specialise in specific sets of delivery. The finance ministries of donor countries could sim- tasks. Typically, a "principal" delegates tasks to an "agent." The ply send a cheque with the agreed amount of aid to their coun- agent carries out the task in return for a reward (a salary or the terparts in recipient countries. In reality, aid is not transferred prospectofapromotion).Becauseofdelegationoftasksandspe- as a "free" amount of money. It comes with an elaborate set of cialization between individuals, no single individual has full constraints attached to it. Donor agency and recipient organi- information on all tasks. Information is asymmetrically distrib- zation need to agree on programmes, including objectives, uted in organizations. This gives rise to two potential problems. implementation arrangements, procurement procedures, and First, the agent charged with a task may use his privileged infor- accounting. Often, aid agencies even carry out the projects mation for his own benefit, and not necessarily to the full bene- themselves. In short, what aid agencies do is to reduce the fit of the principal. This is called moral hazard. Second, the prin- degrees of freedom in the use of aid by the recipient. cipal may not receive all information that is relevant to his deci- Why would donors want to do that? If donors and recipi- sion to delegate a task. As a result, he may take a sub-optimal ents fully shared the same objectives and preferences in decision. This is called adverse selection. Moral hazard and spending aid, there would be no need to do so. Donors would adverse selection are inherent in large-scale organizations. confidently hand over full and unconstrained ownership of However, the magnitude of these potential problems can be aid to the recipients and the latter would spend the funds in reduced through good organizational design: providing agents the same way, as the donors would have done. The fact that aid with incentives to reveal relevant information. Increasing the is not transferred in this way indicates that donors and recip- sizeoftherewardisusuallynotasolution.Increasingthevisibil- ients have different preferences. Donors may seek commer- ity of agents' performance or putting them in competition with cial gains (tied aid) or have different views on policy reforms each other may be more effective. in the recipient country. Recipient organizations may have The aid delivery chain usually contains many pairs of prin- hidden political objectives, or face policy constraints. The role cipal-agent relationships. Donor-taxpayers delegate deci- of aid agencies as intermediaries is to negotiate a compromise sions to elected politicians. Politicians delegate implementation to an (bilateral or multilateral) FIGURE 1--DONORS HAVE THEIR OWN AID DELIVERY TRIANGLE aid agency. Within an aid agency, several layers of hierarchy and delegation of tasks exist. Agencies often hire commercial suppliers, con- sultants, and experts to implement specific tasks. These contractors deliver their products to recipient country organizations and agencies, also with their own internal chain of hierarchi- cal delegation of tasks. The ultimate beneficiar- ies often find themselves at the end of a long chain of intermediaries. Each of these interme- diaries may benefit from asymmetric informa- tion to advance his position. It is easier for Source: WDR 2004 principals to verify the quantity and quality of M A R C H 2 0 0 4 31 between donors and recipients regarding the constraints and the performance of intermediaries in the aid delivery process conditions to be attached to the aid transfer. That explains and the degree of asymmetry in information between these why they "package" aid transfers in an elaborate set of proce- intermediaries. Good organizational design can go a long way dures rather than simply carry out a financial transfer. It sus- towards avoiding these problems, by providing intermedi- tains domestic political support for aid in the donor country: aries with proper incentives to do what is expected of them. donors-taxpayers feel that their preferences in the spending Second, there are inherent structural limits to effectiveness, of aid are taken into account. The need to package aid also if only because of the broken feedback loop and differences in implies that full ownership by the recipient cannot be preferences between donors and recipients. Any aid package achieved. Aid remains a compromise. necessarily constitutes a compromise between these diverg- The situation is somewhat different for multilateral aid ing preferences. The ideal of a fully unconstrained aid trans- agencies. Bilateral agencies usually spend taxpayers' money, fer is generally out of reach. Third, it is difficult to define a and are therefore subject to close supervision by taxpayers' good measure of aid effectiveness. If effectiveness is meas- political representatives. Domestic political concerns will ured solely in terms of recipient country objectives and pref- influence decisions. To the extent that multilateral agencies erences, and if aid agencies focus overwhelming on these, spend loans rather than tax revenue (grants), they can take they will neglect donor preferences and thereby may reduce somewhat more distance from these concerns. To the extent donor support for aid flows. If supermarkets would really "aim that beneficiary countries are effectively represented in their to serve you", at the expense of their profit objective, they decision-making process, the broken feedback loop that is so would simply go bankrupt and disappear. s typical in bilateral aid may be somewhat restored. Bertin Martens, Economist, European Commission. It takes two to aid effectiveness References: ALL THE ABOVE HAS IMPLICATIONS for the effectiveness of Martens Bertin, Uwe Mummert, Peter Murrell and Paul Seabright. The aid. First, effectiveness is not only determined by the behav- Institutional Economics of Foreign Aid, Cambridge University Press, 2003. iour of recipient countries and organizations. It depends on Ostrom E., et.al. Aid, Incentives and Sustainability: An Institutional Analysis of Development Cooperation, SIDA Studies in Evaluation 2002. 32 Development Outreach WORLD BANK INSTITUTE Index of Articles--Development OUTREACH 2000-2003 Abd al-Hameed, Muhammad. Calvo, Sara. Exchange Rates: The Haws, Chris. Battle for the Planet: Léautier, Frannie. Sustainable Children Worldwide Must Be Internet- Targeting Debate, Vol. 3, No. 1 (Winter Mission Impossible? Vol. 4, No. 3 Development: Lessons Learned and Connected, Vol. 2, No. 2 (Spring 2000) 2001) (Fall 2002) Challenges Ahead, Vol. 4, No. 3 (Fall Adhar Utz, Anuja. Mapping Progress Campbell, Tim E. Unknown Cities: Heikkilä, Pauli. New Information 2002) on the Road to the Knowledge Metropolis, Identity, and Governance Infrastructure: A Knowledge Booster, Léautier, Frannie. Urban Air Pollution Economy: Brazil, China, and India, in a Global World, Vol. 5, No. 3 Vol. 3, No. 3 (Fall 2001) Management, Vol. 5, No. 3 (November Vol. 3, No. 3 (Fall 2001) (November 2003) Herfkens, Eveline. Can We Do the 2003) Agarwal, Anil. The Value of Natural Carothers, Thomas and Marina Right Things? The Future of Technical Leipziger, Danny. Achieving Social Capital, Vol. 3, No. 1 (Winter 2001) Ottaway. Defining Civil Society: The Assistance and Capacity Building, and Political Consensus, Vol. 2, No. 1 Agénor, Pierre-Richard. Why Crises Elusive Term, Vol. 4, No. 1 (Winter Vol. 3, No. 3 (Fall 2001) (Winter 2000) 2002) Are Bad for the Poor, Vol. 4, No. 2 Heyzer, Noeleen. Violence against Litan, Robert E. Accepting Foreign (Spring 2002) Clark, John D. Global Is Good, Global Women: With an End in Sight, Vol. 3, Capital: Lessons from Recent Crises, Ahmad, Nilufar. Opportunity for Is Bad, Vol. 4, No. 1 (Winter 2002) No. 2 (Spring 2001) Vol. 3, No. 1 (Winter 2001) Women in Energy Technology in Clinton, Bill et al. The Global Divide in Hodges, Adrian. The Business of Lubrano, Mike. Why Corporate Bangladesh, Vol. 3, No. 2 Health, Education and Technology, Youth, Vol. 4, No. 2 (Spring 2002) Governance?, Vol. 5, No. 1 (Spring 2001) Vol. 2, No. 2 (Spring 2000) Hoekman, Bernard. More Favorable (March 2003) Aiyar, Swaminathan S. Labor, Dadush, Uri B. Toward a Pro-Poor Treatment of Developing Countries: Lustig, Nora. Broadening the Poverty Governance and the Information Age, Trade Agenda, Vol. 5, No. 2 (July 2003) Toward a New Grand Bargain, Vol. 5, Reduction Agenda, Vol. 2, No.3 Vol. 2, No. 1 (Winter 2000) Dahlman, Carl J. Updating the No. 2 (July 2003) (Summer 2000) Aiyar, Swaminathan S. Using Economic Incentive and Institutional Homans, Hilary Yvonne. Tearing Down MacDonnell, Roderick. Access to Community Empowerment to Reduce Regime for the Knowledge Economy, the Wall, Vol. 4, No. 2 (Spring 2002) Information: The Commercial Side, Poverty, Vol. 2, No. 3 (Summer 2000) Vol. 3, No. 3 (Fall 2001) Hsieh, Nien-Hê, William S. Laufer, and Vol. 5, No. 1 (March 2003) Alonso, Rosa. The Revolt of de Wit, Joop. Dynamics of Mark S. Schwartz. Business Students Madani, Dorsati, Will Martin, and John Argentina's Middle Class, Vol. 4, No. 1 Participation in Bangalore's Slums, Debate Ethical Principles and Social Page. The Two Battles of Seattle, (Winter 2002) Vol. 4, No. 1 (Winter 2002) Impact, Vol. 4, No. 2 (Spring 2002) Vol. 2, No. 1 (Winter 2000) Annan, Kofi. Peace and Development Dollar, David and Aart Kraay. Growth- Hübner, Danuta. Gender and Martin, Will. China and the WTO: --One Struggle, Two Fronts, Vol. 2, Enhancing Policies Are Good for Poor Transition. The Cae of Eastern Europe Policy Reform and Poverty Reduction, No. 1 (Winter 2000) People, Vol. 2, No. 3 (Summer 2000) and the Commonwealth of Vol. 5, No. 2 (July 2003) Aslund, Anders. Radical Reformers Dyson, Esther. On the Internet in Independent States (CIS), Vol. 3, No. 2 Mashelkar, Ragunat A. The Indian Lead the Way, Vol. 2, No.1 (Winter (Spring 2001) Russia, Vol. 2, No. 2 (Spring 2000) Innovation System, Vol. 3, No. 3 2000) Edwards, Michael. More Social Johnson, Ian. Johannesburg and (Fall 2001) Bandyopadhyay, Sudeshna. Growth: Beyond: An Agenda for Collective Capital, Less Global Poverty?, Vol. 2, Matovu, George W. M. Africa and Quantity Versus Quality, Vol. 3, No. 1 Action, Vol. 4, No. 3 (Fall 2002) No. 3 (Summer 2000) Decentralization: Enter the Citizen, (Winter 2001) Ficinski Dunin, Lubomir. Kenny, Charles and Anuja Adhar Utz. Vol. 4, No. 1 (Winter 2002) Bebbington, Anthony. Development Is Decentralizing City Management: A Korean Telecommunications Grow at Mbekeani, George W. M. GATS More than Just Growth, Vol. 2, No. 3 Successful Experiment, Vol. 5, No. 3 Record Speed: A Country Profile, Negotiations Must Focus on Services (Summer 2000) (November 2003) Vol. 2, No. 2 (Spring 2000) Liberalization: The Case of SADC, Bellamy, Carol. We Must Invest in Gage, John. From Digital Divide to King, Elizabeth M. and Andrew D. Vol. 5, No. 2 (July 2003) Children, Vol.4, No.2 (Spring 2002) Digital Opportunity: Business Leaders Mason. Engendering development Report from Davos, Vol. 2, No. 2 through gender Equality, Vol. 3, No. 2 McNeil, Mary. Engaging the Poor, Vol. Berry, Ron. E-Commerce Eludes (Spring 2000) (Spring 2001) 4, No. 1 (Winter 2002) Developing Countries, Vol. 2, No. 2 Meden, Natacha. From Resistance to (Spring 2000) Gama, Hobbs. Africa's Brain Drain Klein, Michael. Where Do We Stand Impacts Health Sector, Vol. 4, No. 3 with Private Infrastructure?, Vol. 5, National Building: The Changing Role Bidani, Benu, Homi Kharas, and (Fall 2002) No. 1 (March 2003) of Civil Society in East Timor, Vol. 4, Tamar Manuelyan-Atinc. Coping with No. 1 (Winter 2002) Crisis, Vol. 2, No. 3 (Summer 2000) Gelb, Alan. Gender and Growth: Koro, Emmanuel. Africa Lacks Africa's Missed Potential, Vol. 3, No. 2 Adequate Expertise to Tackle Bio- Mestrallet, Gérard. Bridging the Birdsall, Nancy and Peter Hakim. The (Spring 2001) technology, Vol. 3, No. 1 (Winter 2001) Water Divide, Vol. 4, No. 3 (Fall 2002) Carrot Is Better than the Stick, Vol. 2, Michon, Louis. The Global Economy: A No. 2 (Spring 2000) Goetz, Anne-Marie and Rob Jenkins. Koro, Emmanuel. Armed Conflicts Gender-Sensitive Local Auditing: Destroy African Environment, Vol. 3, Flawed Ecosystem, Vol. 4, No. 3 Birdsall, Nancy. Human Capital and Initiatives from India to build No. 3 (Fall 2001) (Fall 2002) the Quality of Growth, Vol. 3, No. 1 accountability to women, Vol. 3, No. 2 Molavi, Afshin. Supporting the Private (Winter 2001) (Spring 2001) Koro, Emmanuel. Linking Conservation with Development: Sector, Vol. 3, No. 3 (Fall 2001) Bloome, Anthony. Fighting the Gorostiaga, Xabier. The University Is NEPAD Sets Goals at WSSD, Vol. 4, Naidoo, Kumi. Rethinking Insidious Killer: African Teenagers Responsible for Development: No. 3 (Fall 2002) Governance: The Case of South Africa, Battle HIV/AIDS through ICT, Vol.4, Challenges and Solutions in Latin Vol. 3, No. 1 (Winter 2001) No.2 (Spring 2002) America, Vol. 3, No. 3 (Fall 2001) Koro, Emmanuel. Harare Loses Status of Cleanest City in Africa, Vol. 5, No. 3 Nanivska, Vira. What's Wrong with Brown, Mark Malloch. The Information Harris, Bruce. Street Children: Latin (November 2003) Technical Assistance: The Case of Revolution and Development, Vol. 2, America's Wasted Resource, Vol. 2, Ukraine, Vol. 2, No. 1 (Winter 2000) No. 2 (Spring 2000) No. 3 (Summer 2000) M A R C H 2 0 0 4 33 Narayan, Deepa et al. Voices of the Papic, Zarko. Building Civil Society in Serra, Viktor. Learning from the Poor: Svensson, Bent and Sascha Djumena. Poor, Vol. 2, No. 3 (Summer 2000) Former Yugoslavia, Vol. 4, No. 1 Housing and Urban Land Markets, A Public-Private Partnership to Nassery, Homira, Jennifer Brinkerhoff, (Winter 2002) Vol. 5, No. 3 (November 2003) Reduce Global Gas Flaring, Vol. 4, and Najma Siddiqi. Afghanistan and Penton, Ronald. Roaming the Streets Shalizi, Zmarak. Development at a No. 3 (Fall 2002) Pakistan: At the Crossroads, Vol. 4, of Eastern Europe: The Children of Crossroads: Highlights from the World Terry, Donald F. A New Way to Learn No. 1 (Winter 2002) Gloom, Vol. 2, No. 3 (Summer 2000) Development Report 2003, Vol. 4, and Build Networks, Vol. 5, No. 1 Nellis, John and Nancy Birdsall. The Petkoski, Djordjija B. Developing No. 3 (Fall 2002) (March 2003) Distributional Impact of Privatization, Young Voices, Vol. 4, No. 2 (Spring Siphana, Sok. Mainstreaming Trade Thomas, Vinod et al. The Quality of Vol. 5, No. 1 (March 2003) 2002) for Poverty Alleviation: A Cambodian Growth: Key to Less Poverty and Better Newfarmer, Richard. An International Quaynor, Nii. The Digital Divide in Experience, Vol. 5, No. 2 (July 2003) Lives for All, Vol. 3, No. 1 Investment Agreement: Promise and (Winter 2001) Africa, Vol. 2, No. 2 (Spring 2000) Potential Pitfall, Vol. 5, No. 2 (July Smilevski, Blasko. Shaping the Future: Macedonia Children and Youth Tinker, Robert F. Technology and 2003) Ramphele, Mamphela. Keeping the Education: The Trojan Mouse, Vol. 2, promise: A Better Future for Our Development Project, Vol. 4, No. 2 Nogués, Julio J. Agricultural No. 3 (Summer 2000) Children, Vol. 4, No. 2 (Spring 2002) (Spring 2002) Protectionism: Debt Problems and the Tooker, Gary. Wireless Doha Round, Vol. 5, No. 2 (July 2003) Rosen, Harold. Improved Access to Sostmann, Rafael Rangel. Monterrey Communication: Linking Remote Finance: A Key to SME Growth, Vol. 5, Tech's Virtual University, Vol. 2, No. 2 Ocampo, Luis Moreno. State Capture: Areas, Vol. 2, No. 2 (Spring 2000) No. 1 (March 2003) (Spring 2000) Who Represents the Poor?, Vol. 3, Trifonova, Elika. Hi-tech Booms in No. 1 (Winter 2001) Sachs, Jeffrey D. The New Urban Soule, Jeffrey. Forward Thinking in Russia, Vol. 5, No. 1 (March 2003) Planning, Vol. 5, No. 3 (November Nanjing: An Interview with Jiangsu O'Keefe, Joseph. The Private Sector 2003) Party Secretary Li Yuanchao, Vol. 5, Twose, Nigel and Ziba Cranmer. Drives Economic Growth, Vol. 5, No. 1 No. 3 (November 2003) Responsibility Breeds Success, Vol. 5, (March 2003) Santiago, Irene. Women and Power, No. 1 (March 2003) Vol. 3, No. 2 (Spring 2001) Stapenhurst, Rick. Helping Olorunyomi, Dapo. The African Parliaments Help the Poor, Vol. 4, Van Trotsenburg, Axel. Is HIPC Debt Century?, Vol. 2, No. 1 (Winter 2000) Saul, John Ralston. Globalization: The No. 2 (Spring 2002) Relief Working? Vol. 3, No. 2 Choices before Us, Vol. 2, No. 1 Ouma Jura, Jairo. Africa's Business (Spring 2001) (Winter 2000) Stapenhurst, Rick. Media Gets Insurance Against Political Independence: Is Self-Regulation an Vest, Charles. OpenCourseWare, Risks, Vol. 3, No. 3 (Fall 2001) Savir, Uri. Glocalization: A New Answer?, Vol. 5, No. 1 (March 2003) Vol. 3, No. 3 (Fall 2001) Balance of Power, Vol. 5, No. 3 Oyejide, T. Ademola. Trade Reform for (November 2003) Stern, Nicholas H. Engendering Economic Growth and Poverty Development: A Comment, Vol. 3, Watkins, Kevin. Farm Fallacies That Reduction, Vol. 5, No. 2 (July 2003) Serban, Daniel and Claudia Pamfil. Hurt the Poor, Vol. 5, No. 2 (July 2003) No. 2 (Spring 2001) Getting Involved in Romania, Vol. 4, No. 1 (Winter 2002) Wilson, John S. Trade Facilitation, WTO Rules, and Capacity Building: What's at Stake?, Vol. 5, No. 2 (July 2003) Woicke, Peter. The Private Sector and Sustainability, Vol. 4, No. 3 (Fall 2002) Wolfensohn, James D. Fight Terrorism and Poverty, Vol. 3, No. 3 (Fall 2001) Wolfensohn, James D. Cities and Citizens, Vol. 5, No. 3 (November 2003) Yusuf, James D. Urban Development Needs Creativity: How Creative Industries Can Affect Urban Areas, Vol. 5, No. 3 (November 2003) Zaltman, Ann and Gerald Zaltman. Remembering the Future, Vol. 4, No. 2 (Spring 2002) Zegeye, Abebe. The Identity of South African Youth, Vol. 4, No. 2 (Spring 2002) Zhang, Shengman. Human Capacity Building for the New Economy, Vol. 3, No. 3 (Fall 2001) Zurayk, Rami and Mona Haidar. The Dryland Predicament: Natural Capital, Global Forces, Vol. 4, No. 3 (Fall 2002) 34 Development Outreach WORLD BANK INSTITUTE VOICES FROM THE FIELD A Call for More Community-Driven and Integrated Approaches BY PATTI PETESCH ple struggle to be creative and entrepreneurial, but say there are few rewards. Workers in the urban barrios consider IN THE SUMMER OF 2002, a new Voices of the Poor study Col$10,000 a day (less than US$4) to be a good wage; engaged 942 poor women and men of Colombia in discus- Col$5,000 or less is not unusual. A woman from Medellín sions about urgent problems facing their families and describes the scramble to patch together enough earnings: communities. Despite the wider political violence, a sig- "Peopleseekameanstoearnanycent:sellingtamales,blood nificant finding from this work is poor people's very wide- sausages, performing housekeeping jobs by the day, selling spread view that more cohesive and stronger families and newspapers and handicrafts. I, for example, go out to the local organizations are the bedrock to greater peace and streets on holidays trying to sell anything I can. But things development where they live. Study participants also have become much worse. It's terrible when you work so insisted that much could be gained from tackling their hardandearnlessthanbefore...Youfeeldespair,becauseno problems with more comprehensive approaches. matter how stupid you are, you can see that something is "I believe that if we have a good community organiza- happening and someone should be found guilty for it." tion, things will improve and we will deal with issues and Study participants view the political violence and asso- we will all benefit, without excluding anyone," said a man ciated drug and arms trade as fueling unemployment and from a poor coastal fishing community." He continued, insecure jobs and commerce. The violence is said to trigger "Here, we fish and we're organized, [so] we can have cold unreliable or missing government services; growing juve- storing rooms, marketing channels, and transformation nile delinquency, gang violence and neighborhood crime; activities, and [we can] begin saving and have a permanent vigilante justice; and a cascade of other harmful political, income to improve the condition of our families." social, and economic effects. A focus on local organizations and cross-cutting strate- With very high frequency, poor men and women say gies were also central features of proposals to reduce vio- youths are the principal perpetrators and victims of the lence and prepare youth for a better future as well as to rising violence. In all of the study communities, people improve economic opportunities. stress that a large number of poor children and youths are working instead of attending school, leading to very high Problems: Severe and interlocking rates of school drop out and erratic attendance among hardships those who remain in school. Poor people and youths also share concerns about children's extreme vulnerability to IN BOTH URBAN AND RURAL AREAS, the men and women violence by family members or to being abandoned to be who participated in the study report "raised by the streets." that employment opportunities have Schools are seen to be failing poor deteriorated markedly in recent childrenandyouths.Whileeducationis years, and they rely largely on inse- "Voices from the Field" provides deeply valued as a way out of poverty, cure and very low paid informal first-hand insight into issues of poor people indicate that available work. Lack of adequate work is the current concern to the schooling and training opportunities most pressing problem, they say, development community. To are not equipping students with the fueling fear and insecurity in homes participate, send your stories to: skills required by the labor market. and on the streets. devoutreach@worldbank.org. Participants express dissatisfaction With informal work largely Make your voice heard. with the quality of instruction; over- replacing formal employment, peo- crowded classes, high costs associated M A R C H 2 0 0 435 bors, and against those who are community authorities, as a daily occurrence. Proposals: Strengthen capacities and institutions FOR THIS STUDY, poor people put forward 252 proposals of varying length and detail to expand work opportunities, reduce violence, and improve education in their communities. The action recommendations reflect poor peo- ple's sense that these problems are tightly con- nected, and that progress will have to be made simultaneously on all three fronts if any is to be made at all. Households are seen to be the criti- calarenasinwhichdevelopmentproblemsbegin and then spread, and responsive and effective local institutions are terribly important if poor families and communities are to become more harmonious, secure and prosperous. A discussion group of women and men in a barrio of Pasto, which lies in the south of Colombia and contains large displaced popula- tions, developed an initiative to address house- hold-level violence, as shown in Figure 2. Their proposal centers on making psychological assis- tance available to family members. The measure would, they believe, help people recover moral qualities that families used to possess before vio- with keeping a child in school, restricted access to education lence was pervasive. Even if offered, however, study partici- beyond primary years, and selected problems of discrimina- pants recognize that many obstacles will have to be overcome tion and abuse in the classroom. These problems, combined before people will be able to take advantage of the services. with the difficult economy and lure of fast money to be made Forexample,thereisshameinseekingsuchhelp,peoplelack on the streets, especially drive youths to drop out. But teen time, many feel indifferent in the face of so many problems, pregnancyaswellashighcostsandlimitedbenefitsofstaying and there is little communication between parents and chil- in school also contribute. dren. The diverse and numerous challenges to educating chil- In Cali, study participants propose the establishment of dren faced by communities engaged in this study are sum- an integrated program of family support services provided marized in Figure 1, a diagram by a discussion group of men by a single local agency to address pressing problems relat- and women in Medellín on the causes and impacts of school ing to service delivery, political interference in projects, drop out. Taken as a whole, the diagram illustrates that poor insecurity, discord among leaders, discrimination, and people do not see education merely through a sectoral lens. lack of resources. In their view, a focus on providing these Rather, they view education as part of a wider set of prob- services to families would help to make community inter- lems that, without systemic change, hampers their capacity ests and cohesion a priority, and push community leaders to take control of their lives, to develop and prosper. to work together and support decisions based on objective People in the study feel the state has abandoned them, criteria, among other benefits. leaving them defenseless against key threats to their wellbe- Proposals on education span diverse approaches that ing--especially the total lack of public safety and justice, but would improve access to quality education, and especially also increasing hunger, poor work opportunities, and the its relevance to work and family life. Study participants deep vulnerability of poor children and youths. In nine of the make recommendations for an education that is compre- ten communities studied, and particularly in the urban areas, hensive, affordable, and practical--and that will keep chil- people identify a chronic lack of social cohesion in the very dren and youths from the streets. Cali youths developed a heart of their communities as undermining their ability to youth training proposal to build skills in such areas as address urgent local needs. In all six urban neighborhoods, "electricity, refrigeration, and leather work," as illustrated people report physical and verbal aggression among neigh- in Figure 2. According to these youths, and many others in 36 Development Outreach WORLD BANK INSTITUTE the study, work-related training might help to pry apart the propose to create an association of single mothers for interlocking disadvantages of unemployment, community domestic and cleaning work. violence, and low education levels. The overwhelming perception about the importance of Proposals to improve work prospects also emphasize a local organization is a striking finding from a country torn need to build "good organizations" that can help people apart by violence. "Individually we have achieved nothing find or generate work or improve the work they have. In the ... however, if we organize ourselves we can do it," insists a sharecropping community of Girón, study participants community leader in rural Girón. propose to create a farmer association to market their pro- Aligning development efforts with such priorities will duce independently of intermediaries. Participants in require important redirections of development policies and Barrancabermeja put forward recommendations to form a programs to support more community-driven approaches cooperative to produce and market detergent and textile and to meet the needs of poor families in comprehensive items. In the Cauca community, participants plan to mar- ways. Poor people worldwide repeatedly demonstrate that ket organic coffee internationally. In Medellín, women they are willing to invest their scarce time in guiding and overseeing development processes if they feel these actions can bring about positive FIGURE 1. CAUSES AND EFFECTS OF DROPPING OUT OF SCHOOL, A DISCUSSION change. The challenge then becomes one GROUP OF MEN AND WOMEN IN MEDELLÍN of unleashing such energies systematical- ly. While people's aspirations may be Ignorance Rise in most operative at the micro level, they Unemployment Increase Illiteracy in Vices nevertheless require enabling policy frameworks and institutional structures Violence Little at the macro level. It is precisely within Development Migration of the Region these critical national policy arenas, how- ever, where poor people's voices are the Loss of EFFECTS Opportunities least heard. This must change. s SCHOOL DROP-OUT Patti Petesch is a World Bank Consultant with CAUSES Education Loses Importance the Poverty Reduction Group. She is co-author Insecurity of the second volume of Voices of the Poor Irresponsibility and co-editor of the third volume. Economic of Parents Problems Mistreatment Discrimination This article was adapted from Arboleda, J., Family from Teachers Lack of Motivation P. Petesch, and J. Blackburn, Voices of the Disintegration Young Lack of Poor in Colombia: Strengthening Livelihoods, Mothers Work Families, and Communities, Washington, DC: World Bank, 2004. FIGURE 2. ASSESSING OBSTACLES AND OPPORTUNITIES: A YOUTH TRAINING PROPOSAL, DISCUSSION GROUP OF YOUTH IN CALI Many people Violence Quality of life Youth gain Delinquency benefit is reduced improves confidence decreases Agreed-upon Unemployment commitments made Refrigeration is reduced, public families benefit YOUTH VOCATIONAL Electricity TRAINING PROGRAM-- Leatherwork EXCHANGE OF KNOW-HOW Others Lack of communication Delay due to lack of between the No physical Lack of Adults do not have Youth do not get knowledge and different leaders space resources time to teach them along with adults technology M A R C H 2 0 0 4 37 KNOWLEDGE RESOURCES PUBLIC SERVICES RESEARCH CESI, the World Bank Institute's website profiles research-in- Community Empowerment & progress and related outputs on Social Inclusion learning public service delivery in the program, works to help create Development Research Group the conditions that enable the of The World Bank. This work is poor and the excluded to shape organized around three sub- their own development. The topics: a) education; b) health, program looks at the issues of nutrition, and population; and empowerment and governance, c) impact evaluation, finance, and public policy. Each thus making the crucial link to sub-topic has a document library with links to the research strengthen the voices of the poor in influencing public outputs, survey questionnaire, and other resources. policies, as well as in making institutions more accountable Visit: www.publicspending.org and responsive to their needs. Visit: www.worldbank.org/wbi/communityempowerment WORLD DEVELOPMENT REPORT 2004 can be accessed through this POVERTYNET site is maintained web site. The report provides by the Poverty Reduction Group, powerful examples of where part of the Poverty Reduction and services can do work, showing how Economic Management Network governments and citizens can do at the World Bank. The sites better. The main difference highlights several topics and between success and failure is the programs: Social Capital for degree to which poor people themselves are involved in Development--norms and determining the quality and the quantity of the services, networks that enable collective which they receive. The site includes: full text, background action; Understanding Poverty--how poverty can be papers, consultations, core team, archived e-discussions, defined, measured, and studied through a variety of related resources, and order information. indicators; Understanding Impact Evaluation--information Visit: http://econ.worldbank.org/wdr/wdr2004/ and resources for people and organizations working to assess and improve the effectiveness of interventions aimed at reducing poverty; and others. Subscription to PovertyNet Newsletter is available through the site. POVERTY ACTION LAB, Visit: www.worldbank.org/poverty housed at the Massachusetts Institute of Technology serves as a focal point for development and poverty research based on GLOBAL DEVELOPMENT randomized trials. The objective LEARNING NETWORK (GDLN) is to improve the effectiveness uses distance learning of poverty programs by technologies and methods to providing policy makers with scientific results that help facilitate interactive, cost- shape policies to combat poverty. The Lab works with NGOs, effective learning and international organizations, and others to evaluate programs knowledge-sharing for and disseminate the results. It works on issues as diverse as sustainable development and boosting girls' attendance at school, improving the output of poverty reduction. GDLN farmers in sub-Saharan Africa, racial bias in employment in Centers around the world offer a unique set of timely the US, and the role of women political leaders in India. services to development practitioners. Through GDLN, Visit: www.povertyactionlab.com individuals, groups, and organizations design and deliver courses, seminars, and other activities that cover the full range of development issues. Visit: www.gdln.org 38 Development Outreach WORLD BANK INSTITUTE BOOKSHELF HOW TO CHANGE THE WORLD: of property and property rights. Every developed nation in SOCIAL ENTREPRENEURS AND the world went through the transformation from THE POWER OF NEW IDEAS, by predominantly informal, extralegal ownership to a formal, David Bornstein. Oxford University unified legal property system. This system is what allowed Press, 2004. people everywhere to leverage property into wealth. This This book tells the fascinating book revolutionizes our understanding of capital and points stories of a number of social the way to a major transformation of the world economy. entrepreneurs, creative individuals who question the status quo and LIFELONG LEARNING IN THE exploit new opportunities to remake GLOBAL KNOWLEDGE the world. From Brazil to Hungary, South Africa, and the ECONOMY: CHALLENGES FOR USA, the fastest-growing segment of society is the DEVELOPING COUNTRIES, nonprofit sector, as millions of social entrepreneurs are World Bank, 2003. increasingly stepping in to solve problems where This new World Bank reports governments and bureaucracies have failed. Bornstein warns that developing countries shows that with determination and innovation, even a will have little success boosting single person can make a surprising difference. The book economic growth and reducing is both inspirational and a source of practical ideas whose poverty unless they can close a time has come. growing knowledge, or education, divide between themselves and richer countries. PATHOLOGIES OF POWER: Investing in quality lifelong learning, the reports says, can HEALTH, HUMAN RIGHTS, AND help to close this gap by promoting economic vitality, THE NEW WAR ON THE POOR, reducing poverty, and encouraging open and cohesive by Paul Farmer. University of societies. In a global economy which becomes faster and California Press, 2003. more powerful every year, lifelong learning allows people Paul Farmer, a physician and to prosper. This book is a roadmap for policymakers who anthropologist with twenty years of wish to know more about the key issues and challenges of experience working in Haiti, Peru, education in a knowledge economy. and Russia, argues that promoting the social and economic rights of THE INSTITUTIONAL the world's poor is the most important human rights ECONOMICS OF FOREIGN AID, struggle of our times. The author exposes the relationships by Bertin Martens with Uwe between political and economic injustice, on one hand, and Mummert, Peter Murrell, and Paul the suffering and illness of the powerless, on the other. Seabright. Cambridge University Farmer's urgent plea to think about human rights in the Press, 2003. context of global public health and to consider critical This book deals with effectiveness issues of quality and access for the world's poor should be of of foreign aid from a new fundamental concern to everyone interested in perspective and analyzes how development. results can be improved. It claims that the results of aid are not solely a function of recipient country policies but also of the THE MYSTERY OF CAPITAL: policies and actions related to the aid donors and their WHY CAPITALISM TRIUMPHS intermediaries. By pointing to the various agency IN THE WEST AND FAILS problems that may arise throughout the chain of the aid EVERYWHERE ELSE, delivery process, the book also clearly shows that by Hernando de Soto. Basic Books, improving aid performance is far from simple. It explains Reprint edition, 2003. why development banks are different from grant-based The world-famous Peruvian donor agencies, or why bilateral aid is different from economist Hernando de Soto finds multilateral. This book provides a fresh new look at that economic success is not foreign aid and aid agencies that deliver it. It is useful to determined by cultural differences donors and recipient policymakers alike. but depends on the legal structure M A R C H 2 0 0 4 39 CALENDAR APRIL 2004 JUNE 2004 4­6 11th Annual Global Finance Conference 3­5 13th International Symposium on HIV & Las Vegas, Nevada, USA Emerging Infectious Diseases Conference www.globalfinance.csufresno.edu Toulon, France www.avps.org 24­25 IMF/World Bank Spring Meetings Washington, DC, USA 14­15 Symposium for Broadcasters www.imf.org/external/am/ on International Development Washington, DC, USA 24­28 American Planning Association's www.worldbank.org/mediaworkshop/ National Planning Conference agenda.htm Washington, DC, USA www.planning.org JULY 2004 MAY 2004 11­16 XV International AIDS Conference Bangkok, Thailand 10­11 Annual Bank Conference on Development www.ias.se/bangkok/ Economics In Europe Brussels, Belgium 19­21 Traditional Healing & HIV/AIDS http://wbln0018.worldbank.org/EURVP Dakar, Senegal www.africa-first.com/4thictm.asp 26­28 Reducing Poverty, Sustaining Growth Shanghai, China www.worldbank.org/wbi/reducingpoverty/ index.html Outreach SubscriptionOrderForm D E V E L O P M E N T SUBSCRIBERS FROM DEVELOPING NAME COUNTRIES WILL CONTINUE TO RECEIVE THE MAGAZINE FREE OF CHARGE. TITLE SUBSCRIPTION FEE APPLIES TO READERS ORGANIZATION IN THE FOLLOWING DEVELOPED COUNTRIES: ADDRESS Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Greece, Iceland, Ireland, Italy, Japan, Kuwait, Luxembourg, CITY AND STATE OR PROVINCE Monaco, Netherlands, New Zealand, Norway, Oman, Portugal, Spain, Sweden, Switzerland, COUNTRY United Arab Emirates, United Kingdom, United States, and Vatican City. ZIP / POSTAL CODE PHONE FAX Please print or type all information E-MAIL q YES, sign me up for Development Outreach Please do not send cash. Make checks payable to Development OUTREACH/WBI. for 2004 for just $18 q Check no.___________ in the amount of $________ is enclosed. (three issues per calendar year, January- Mail order to: December) Editor, Development OUTREACH The World Bank 1818 H Street, NW, Room J2-200 Washington, DC 20433 USA