53857 Governance and Anti-Corruption Ways to Enhance the World Bank’s Impact 2 Governance and Anti-Corruption Ways to Enhance the World Bank’s Impact Evaluation Brief 2 July 2006 The World Bank http://www.worldbank.org/ieg Washington, D.C. ©2006 Independent Evaluation Group Knowledge & Evaluation Capacity Development The World Bank 1818 H Street, NW Washington, DC 20433 USA E-mail: eline@worldbank.org Telephone: 202-458-4487 Fax: 202-522-3125 http://www.worldbank.org/ieg All rights reserved This Evaluation Brief is a product of the staff of the Independent Evaluation Group (IEG) of the World Bank. The findings, in- terpretations, and conclusions expressed here do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denomina- tions, and other information shown on any map in this work do not imply any judgment on the part of the World Bank or IEG concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. IEG encourages dissemination of its work and will normally grant permission to repro- duce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to eline@worldbank.org. The lead author for this Evaluation Brief is Ajay Chhibber, Director, IEG-World Bank. Support was provided by Rachid Laajaj and Betty Bain. Contents 1 Governance and Anti-Corruption: Ways to Enhance the World Bank’s Impact 5 Background Analysis 7 Governance and Anti-Corruption: The Efforts So Far 9 Progress Has Been Slow 13 Better Understanding of the Relationship between Governance and Development Is Needed 19 Endnotes Governance and Anti-Corruption: Ways to Enhance the World Bank’s Impact T he Bank has made significant efforts since corporate stance toward governance and anti- the mid-1990s to highlight the harmful corruption. effect of corruption on development and has developed a number of mechanisms to help I. Improving Support to Countries countries improve governance and fight corrup- • Go Beyond Process Reforms and Follow tion. However, progress on outcomes has been the Money. Despite great efforts over the past slow. It should be noted at the outset that 10 years, there are indications that corruption improving governance is a complex, long-term is showing no signs of improvement, and could process. All evidence shows that the quality of even be worsening. Past efforts were heavily fo- governance is strongly correlated with broader cused on processes such as safeguards, pro- measures of development (in levels), suggesting curement, financial management, and public an interdependence, so that anti-corruption sector reform. Progress on some of these meas- initiatives in the developing world will have to be ures when the country is receptive has been a part of, and dependent on, each country’s positive, but public sector reform in general has broad, complex, and long-term state-building shown relatively weak outcomes. Visible process, supported by a strong commitment by progress has been achieved, for example, in sev- developed countries to tighten their policies as eral European Union (EU) accession countries. well. But in many instances, countries have met these requirements on paper, and even agreed This note summarizes suggestions for enhanc- to the “cosmetic” changes that the Bank re- ing the Bank’s work on helping countries deal quires, while the underlying incentives and op- with poor governance and corruption and in portunities for corruption remain unaffected. ensuring that funds disbursed by the World Bank The Bank should expect to help countries to be are protected from abuse. Its purpose is to more transparent about large financial flows inform the ongoing discussions at the Bank on to complement the existing process approach. its governance and anti-corruption strategy. The • Work More Actively to Increase Demand findings are generally anchored in past evalua- for Better Governance and Improved tions described in the attached note. New Country Practices. The Bank’s governance evaluations are under way in the Independent work focuses heavily on top-down rules and Evaluation Group (IEG) on governance and regulations and on systemic processes such as public sector reform, legal and judicial reform, public expenditure management systems, civil decentralization, and doing business indicators service reform, anti-corruption commissions, that will shed deeper light on these complex and the like. These reforms will be ineffective issues. unless demand for reform comes from more aware citizens within the country. More active The findings (so far) and suggestions are approaches are needed to encourage demand summarized under three broad categories: how for change through greater transparency, en- to improve engagement and support to couragement of civil society, freedom of the countries, how to improve Bank-financed press, and public information disclosure in operations, and how to enhance the Bank’s close collaboration with local institutions. 1 E V A L U AT I O N B R I E F S • Concentrate Greater Attention on Coun- and country performance—for example, eco- tries with Poor Governance and Weak nomic growth—is more varied and diverse. Performance. While there are many countries The positive link with Bank projects is not sur- where results are not being achieved and in- prising, since Bank loan proceeds (like the dicators of governance and corruption are funds of other multilateral development banks poor, there are just as many where results are [MDBs] and aid agencies) largely go through being achieved, despite poor indicators of gov- the public sector. Therefore, it makes sense to ernance. Given limited internal resources, one direct anti-corruption efforts toward Bank op- approach is to devote greater attention to the erations of all types. Governance and anti-cor- former to achieve win-win outcomes on gov- ruption are seen largely as a Poverty Reduction ernance and results, while trying to better un- and Economic Management (PREM) Network derstand the dynamics between governance activity, but must now become central to all our and results in countries with good results along operations. Some of the networks are moving with poor governance. In countries with weak in this direction, and this shift should be ac- governance, establish clearer benchmarks of celerated and strengthened. In fact, the Bank progress and get a much better understanding has often achieved better results on gover- of the root causes of poor governance—es- nance and anti-corruption through sector re- pecially its historical, social and political as- forms in banking, energy and other utilities, and pects. In countries with a history of weak on procurement, when government owner- governance, and especially where there is an ship was present than it has through attempts abrupt turn-around in the political regime, at wider public sector reforms and visible anti- lend prudently while staying engaged. corruption measures, such as the establish- • Encourage Fiscal and Financial Trans- ment of anti-corruption commissions.1 parency on Revenues from Extractive In- • Re-examine Guidelines for Project-Re- dustries (EI). In countries where a substantial lated Corruption. The Bank’s policy is zero share of economic activity comes from ex- tolerance for corruption in lending operations, tractive resources, fiscal and financial trans- but there remains considerable lack of clarity parency is typically weak. The World Bank in how this policy should be applied when al- Group (WBG) should vigorously pursue coun- legations of corruption surface in Bank proj- try-wide and industry-wide disclosure of rev- ects. Moreover, the burden of proof to initiate enues from EI and related contractual actions when there are allegations of corrup- arrangements such as production sharing tion is more often on the Bank than on bor- agreements, concession, and privatization rower authorities. This makes it difficult and terms. The Bank should support disclosure of costly to initiate remedial action. The Bank EI revenues and their use in resource-rich and the borrower authorities should be allies, countries, following the U.K. government’s not adversaries, in the campaign against cor- Extractive Industries Transparency Initiative ruption. New staff guidelines are needed to pro- (EITI). The International Finance Corporation vide greater clarity on the proper response to (IFC) and the Multilateral Investment Guar- allegations of corruption and to remove ad antee Agency (MIGA) should also consider re- hoc treatment across projects. quiring their private sector clients to publish • Reduce Opportunities for Corruption in their payments to governments and encourage Infrastructure. There appear to be greater op- the PWYP (publish what you pay) initiative. portunities for corruption in infrastructure projects, related to procurement, contracts, II. Improve Bank Operations and regulatory capture, given the large sums of • Corruption Affects Projects in All Sec- money involved. With public-private partner- tors. Better governance and lower corruption ships, which involve large contracts handled by are positively correlated with the success of low-wage public servants, such opportunities Bank projects. The link between corruption intensify. These opportunities can be reduced 2 G O V E R N A N C E A N D A N T I - C O R R U P T I O N : W AY S T O E N H A N C E T H E W O R L D B A N K ’ S I M PA C T by convincing countries to avoid complex fi- international banks do not accept stolen funds, nancial structures, practice greater transparency, and return them if discovered. The Organi- and employ enhanced competition. One use- saiton for Economic Co-operation and Devel- ful avenue would be to compare and publicize opment (OECD) and UN conventions (UNAC) costs of delivering services and service delivery could be used more aggressively to work with times to foster better benchmarking. countries and implement their agreed-upon • Conduct Physical and Financial Audits standards. on a Sample of Social Fund and Com- • Encourage Partners and MDBs to Follow munity-Based and -Driven Projects. Com- Similar Procedures. The Bank and its share- munity-based and -driven approaches present holders should try to adopt similar procedures a very powerful tool to circumvent corrupt and policies on governance and anti-corruption government and get funds directly into the across all aid, and not just at the World Bank. hands of communities. However, evaluation of An MDB Task Force is looking into this issue both community-based and -driven projects for better coordination across the MDBs. Many and social funds shows that these approaches bilateral aid programs focus on governance, but do not control overall corruption—only the na- the approaches and objectives vary widely. The ture of corruption changes. They are not magic Bank could help harmonize approaches to bullets. Evaluation also shows that audits (phys- governance and anti-corruption. ical and financial), and even the threat of an • Establish a Governance and Anti-Corrup- audit, reduce corruption. Conduct such spe- tion Advisory Council. In view of the sensi- cial audits on a sample of such projects and tivity of governance and anti-corruption issues, keep them available as part of the toolkit to dis- and the potential concern that the Bank can be courage corruption. charged (even if incorrectly) with political in- terference, the Bank might consider establish- III. Enhance Corporate Policies ing a Governance and Anti-Corruption Council • Encourage Developed Country Institu- made up of highly respected individuals from tions and Businesses to Tighten Their across the world. The Council would advise Practices. Corruption is a two-way street, and the president and senior management and the briber is as responsible as the bribee. In this communicate with political leadership in coun- regard, the role of international business in tries on governance and anti-corruption issues. bribing, and of international banks in shelter- ing bribe-based capital flight, needs much IEG is looking at this issue from several different greater scrutiny and attention. The local private angles, including an ongoing evaluation of the sector should also be enrolled in the fight Bank’s work on low-income countries under against corruption through IFC and MIGA work stress (LICUS), fiduciary work, governance and with partners. The developed countries should public sector reform, legal and judicial reform, also play a role by ensuring that their busi- and decentralization. IEG will report further as nesses are punished for paying bribes and that we get more detailed results. 3 Background Analysis 5 1—Governance and Anti-Corruption: The Efforts So Far G overnance2 and anti-corruption are • Governance forms a key component of central elements of the World Bank IDA allocations through the Country Policy Group’s work in supporting economic and Institutional Assessment (CPIA), in which development. Much progress has been made assessment of governance forms a key com- over the past decade in bringing governance and ponent. In addition, a special weight is at- anti-corruption to center stage in development. tached to governance in International The Bank’s work on governance became explicit Development Association (IDA) allocations.4 and more systematic with the 1996 Annual Governance, therefore, gets a double weight Meeting Speech of the President and the 1997 in IDA allocations. World Development Report on the role and the • Financial management plans are required effectiveness of the state. of every project, along with the existing procurement oversight. Core diagnostics Bank Has Made New and Intensified related to public expenditure, procurement, Efforts poverty, and financial accountability have been introduced and are required of all IDA coun- Since then the Bank’s governance work has tries. Most country teams have, over the past focused on seven broad areas: five years, completed these assessments. • Developing indicators of governance and • Community-driven and -based approaches doing more analytical work on aspects of gov- have become more popular as mecha- ernance and the business climate.3 Most of nisms to help money reach people di- these are survey-based perception indicators. rectly, circumventing corrupt govern- They typically ask business groups and citi- ments. Such approaches are commonly used zens about aspects of governance—such as in conflict-ridden countries as well as in coun- delivery of services, bribery, rule of law, and gen- tries where corruption is egregious and wide- eral perceptions of government effectiveness. spread. The main idea is to provide Some new work is also developing direct quan- communities a role in decision making (com- titative indicators, but this remains very limited. munity-based) and sometimes direct control • Governance and anti-corruption are now (community-driven) in the use of resources al- central features in every country assistance located for their betterment. program of the World Bank, with public sector • For countries with severe governance and judicial reform as key elements. This work problems, the special category of LICUS is seen as the responsibility of the PREM Net- was created. These are countries with a CPIA work but has not been addressed with equal rating below 3 and a governance rating below priority in the work of the Bank’s other net- 3. The objective of creating a special category works, although well-designed sector reform of LICUS countries5 was to try and find special programs and projects in such areas as trans- ways to deal with countries with very poor port, finance, energy, and the like lead to bet- governance. These countries are often con- ter governance and reduced corruption, with flict ridden, with state authority that has col- strong government ownership. lapsed to a level where even the basic functions 7 E V A L U AT I O N B R I E F S of government, such as law and order, are dif- investigations are triggered by allegations of ficult to maintain. corruption. While this unit does not have any • An Institutional Integrity Unit was es- judicial powers, its investigations can lead to tablished to help detect fraud in the actions against corruption among Bank staff, Bank and in Bank-financed projects. This sanctions on companies involved in fraud in unit was established to investigate corrup- Bank-financed projects, and provision of in- tion among Bank staff as well as issues of cor- formation to countries investigating fraud in ruption in Bank-financed projects. Its Bank-financed projects. 8 2—Progress Has Been Slow Between 1996 and 2005, important initiatives effort rather than outcome, because the CPIA were taken on governance and anti-corruption, is an in-house indicator, not based on surveys. but there remains a perception that real Bank staff who produce the CPIA sometimes progress has been slow. There are several rate an improvement when certain policy ac- reasons for this perception: tions are taken, such as passage of new legis- • Indicators for governance show deterio- lation, without clear evidence that the actions ration since the mid-1990s, or at least do not have actually been implemented on the show significant improvement6 despite all the ground. The Business Environment and En- efforts made so far to help countries improve terprise Performance Survey (BEEPS) shows governance (see figure 1a, b). The most com- some improvement in governance indicators, prehensive governance indicators are being but it is restricted to one Region, Europe and assembled at the World Bank Institute (WBI) Central Asia. European Union (EU) accession by combining a number of external and inter- has been a major driver of institutional im- nal indicators, including the Bank’s own in- provements in several countries in the Region. ternal indicator, the CPIA. Overall they show a • Public Sector and judicial reform has led deterioration, or at least show no improve- to institutional reforms, but in many in- ment. Only the CPIA shows a small improve- stances countries appear to be willing to ment in governance, but this could indicate go through the process of reform, while Figure 1a. Corruption Remains Deeply Rooted in Developing Countries 6 5 4 Rating 3 2 1 0 1990 1992 1994 1996 1998 2000 2002 2004 Source: 2004 Annual Review of Development Effectiveness (IEG, 2004, Washington, DC, World Bank). Note: ICRG: Index of corruption. 9 E V A L U AT I O N B R I E F S Figure 1b. Governance Indicators for LICUS Countries and Non-LICUS Low-Income Countries (LICs) Ϫ0.25 Ϫ0.50 Indicators Ϫ0.75 Ϫ1.00 Ϫ1.25 1996 1998 2000 2002 2004 LICUS Non-LICUS LICs Source: Based on Kauffman, Kraay, and Mastruzzi (KKM). underlying governance deteriorates, or at best sciousness that delivery of services is affected remains as poor as before. Many countries have by institutional and governance issues. But embarked on institutional reforms, such as civil the change is very slow. Projects are being de- service reforms, public expenditure manage- signed using checklists for safeguards, pro- ment reforms, public enterprise reforms, and curement, and fiduciary, but these are often judicial reforms. Project performance in the seen as requirements rather than as exercises area of public sector reform appears to be rel- that add value. There is limited interaction be- atively weak, and sector-wide assessments show tween the project work and the PREM sector- a relatively lower success rates than any other led anti-corruption work, although some of sector. Moreover, such reforms will take a long this is now changing. time to show results and are sometimes more • The procurement and financial control cosmetic than real. In the meantime gover- framework is well regarded but it is still nance deteriorates because the control of the insufficient. The Bank’s procurement and fi- political elite over resource flows is largely un- nancial management systems have high-qual- affected by such reforms.7 In some cases, one ity processes in place, but in countries or suspects that the political elite agrees to go systems where institutions are weak, increas- along with such reforms because they are well ingly sophisticated mechanisms are allegedly aware that the process of reform will take a used to circumvent the Bank’s safeguards. long time, and their control of overall resources While hard evidence is difficult to come by,8 will be unaffected in the meantime. there is a perception that even such elaborate • The core sector work of the Bank in control frameworks are being circumvented. health, education, agriculture, infra- Very sophisticated collusion among bidders is structure, and the like is not directly difficult to detect, and the Bank needs to find linked to anti-corruption. Systemic anti- ways to reduce the avenues for corruption by corruption issues are largely dealt with by bringing greater transparency to the process PREM, while sector staff in other networks are and employing tougher sanctions once prob- not as focused on governance and corruption lems have been detected. issues in their sectors. This is changing to some • The community-driven approaches are extent, because there is now a growing con- useful for getting funds to people but ap- 10 G O V E R N A N C E A N D A N T I - C O R R U P T I O N : W AY S T O E N H A N C E T H E W O R L D B A N K ’ S I M PA C T pear unable to stem corruption. For quite system based on performance, such countries some time the Bank has held the view that would have received very limited assistance, one way to reduce corruption is to deliver but by classifying them into a separate cate- money directly to the local communities and gory, such countries have continued to receive to increase their involvement in how funds in- IDA allocations. The bulk of the increase in IDA tended for their benefit will be spent (com- assistance to these countries has gone to post- munity-based development) or to given them conflict countries. So far the Bank does not ap- direct control over resources (community- pear to have found a viable way to help these driven development). The logic is that if com- countries improve their poor governance munities have more say or involvement, it is a record. The approach has merely allowed the self-policing mechanism against corruption, Bank to remain engaged with the low-per- because they are unlikely to steal resources in- forming and weakly governed LICUS countries, tended for their own benefit. But even such ap- while maintaining a performance-based resource proaches have to function within the local allocation system in the non-LICUS countries. political reality. This is the same political real- Lately the Bank has shifted focus in the ity that leads to underdevelopment: the con- LICUS/fragile states to the concepts of state- trol of economic resources by a narrow, local building and peace-building.11 But these ap- elite is often difficult to circumvent by a donor- proaches, which are central to the LICUS/fragile driven, community-based approach. Where states approach, need greater elaboration if genuine grassroots community initiatives exist, they are to be useful for operational work. building on them has worked well. But where • The threshold of proof for initiating ac- such a grassroots mobilization does not exist, tion on corruption in Bank-financed proj- donor projects—including those from the ects is a law-based investigative trigger.12 World Bank—have been usurped by local elites. The Bank follows a policy of zero tolerance in Impact evaluation of even the flagship com- the projects it finances, but the guidance to staff munity-driven development project—the Ke- on how to respond to allegations remains un- camatan Development Project in Indonesia— clear and is therefore applied inconsistently. indicates that the community-driven develop- Moreover, the onus of responsibility appears ment approach has changed the nature, but not to be on Bank staff rather than on the bor- the level, of corruption.9 Such approaches rower. The responses of Bank operational staff must be carefully implemented, because they appear to vary from Region to Region. A more may even undermine governance, in the long- consistent approach is needed. More broadly, term, by weakening further the institutions of there is a need to tackle project-related cor- local government.10 ruption as a development issue, which requires • The LICUS approach has not yet helped im- remedial action in partnership with country au- prove governance. The governance indicators thorities. Once allegations of corruption sur- for LICUS countries have also not shown any im- face, the onus of responsibility must shift to the provement—if anything, they seem to be de- borrower to show evidence of actions taken to teriorating. The LICUS countries have now been address the issue. If no remedial action is taken, termed fragile states. Under an aid-allocation penalties can then be applied. 11 3—Better Understanding of the Relationship between Governance and Development Is Needed T he first step toward such an approach is countries—especially in Asia—deliver better to try and improve our understanding of services and achieve faster growth than others, the complex nature and types of even in environments of weak institutions (see governance arrangements and to deepen figure 3)? One explanation could be that growth analysis on how governance and corruption is affected by many channels. Remittances, now affect development. several times larger than aid, have also played an important part in helping countries (such as In addition to perception indicators,13 which Albania, Armenia), or have allowed countries to have been criticized for a number of reasons manage despite governance problems (such as (see Arndt and Oman 2006)14 there is need for Philippines or countries in Central America). A physical indicators of performance, costs, and second reason is that nongovernmental organi- quality that would allow countries and agencies zations (NGOs) have also played an important to better benchmark their activities. For role in delivering services, as in Bangladesh. example, in the infrastructure sector we should get more systematic data on costs of construc- It can be argued that the medium-term relation- tion, energy, telecommunications, water, and ship between governance and growth is weak, quality indicators such as time taken to get but that in the long run the relationship goods through ports, customs, and the like. between governance and the level of develop- These data will help sector operations to ment is strong. This is no doubt true. Developed address issues of governance and efficiency and countries have better governance indicators and to better understand why service delivery stronger institutions than developing countries. parameters vary across the world. It will also But this does not help us understand the help borrowers to see where they are not dynamics of growth and governance. On the getting value for their money. Some of this work long-run relationship between governance and is now under way. growth, there are at least two competing hypotheses. One argues that with good institu- While corruption is harmful to development, it tions, checking arbitrary behavior by govern- must be recognized that there are many ment and establishing rules lead to human and developing countries that have experienced physical capital formation and growth and rapid growth (see figures 2a and 2b) and reason- prosperity. The other approach argues that ably effective service delivery15 for periods of growth starts with human and physical capital time, along with poor governance and high accumulation, which leads to growth, and as levels of perceived corruption. Even within the incomes rise, it leads to demands for better same region, these relationships show a very governance and democracy. You end up with weak correlation. We need to understand not good governance being correlated with higher just the level, but also the nature, of corruption levels of development—but causality is more and its links to development.16 Why do some complex. 13 E V A L U AT I O N B R I E F S Figure 2a. Medium-Term Growth and Governance Indicators: Weak Correlation All developing countries 20 15 Average growth rate from 1996 to 2004 10 5 0 Ϫ5 Ϫ2 Ϫ1 0 1 Overall indicator of governance Note: Using Kaufmann, Kraay, Mastruzzi (2005). The former approach has been empirically governance, corruption, and development that verified by a number of studies (Knack and is more country specific and helps identify Keefer 1995;17 Mauro 1995;18 Hall and Jones which aspects of governance need greater 1999;19 Acemoglu, Johnson, and Robinson 2001, attention at different levels of development. In 2002;20 Easterly and Levine 2003;21 Dollar and many cases work on government effectiveness Kray 2003;22 and Rodrik, Subramanian and will provide better results, in others the judicial Trebbi 200223). But the instruments used in and legal system may need attention, and in these studies to establish causality have been some cases bringing greater voice and accounta- questioned, and recent studies have challenged bility will be helpful. the direction of causality and have tried to show that the relationships among growth, human It should be noted that while the relationship capital formation, and institutions goes in the between the medium-term growth and opposite direction (Glaeser and others 2005),24 governance indicators is weak, there is a strong or at least quite complex (Rajan and Zingales correlation between the percentage of projects 2006; Arndt and Oman 2006).25 Given the high with a satisfactory rating and good governance correlation between income, education, and and less corruption (figure 4a and 4b).26 This institutional quality, it is not surprising that the correlation is even stronger for investment direction of causality is under debate and will loans. The positive association of governance probably remain so. Clearly more work is and corruption indicators with the degree of needed on the complex interaction among success of Bank projects is not surprising, since 14 G O V E R N A N C E A N D A N T I - C O R R U P T I O N : W AY S T O E N H A N C E T H E W O R L D B A N K ’ S I M PA C T Figure 2b. Medium-Term Growth and Perception-Based Corruption Indicators: Weak Correlation All developing countries 20 Average growth rate from 1996 to 2004 15 10 5 0 Ϫ5 Ϫ2 Ϫ1 0 1 2 Corruption indicator Note: Using data from Kaufmann, Kraay, Mastruzzi (2005). Bank loan proceeds (like the funds of other improve governance than its direct anti-corrup- MDBs and aid agencies) largely go through the tion initiatives. public sector. One implication of these results is that it makes sense to direct anti-corruption However, the lack of a clear association between efforts toward Bank operations of all types. medium-term growth and governance suggests Governance and anti-corruption are seen largely that a more nuanced approach maybe needed in as an activity for the economic group, but must helping countries improve governance and fight now become central to all sectors. Some of the corruption. For countries with weak governance sector teams are moving in this direction, and and poor performance, a greater focus on this shift should be accelerated and strength- improving governance is clearly a win-win ened. In many cases the Bank’s sector work on approach. But, in countries where growth is improving financial systems, energy regulation, high despite poor governance indicators, a and greater transparency in tendering processes better understanding of the dynamics between has done more to reduce corruption and governance and development is needed. 15 E V A L U AT I O N B R I E F S Figure 3. Rapid Growth Comes with Variance in Governance Quality Avg. ICRG Change in WBI governance Time ratings avg. ICRG indicators required to Growth in GDP (3 yr. avg. ratings (2002)c establish a per capita 2002–2004)b (1990–2004) percentile business (1990–2003)a percent percentage rank (2004)d Country percent of total change 0–100 days China 8.6 55.7 2.0 43.2 48 Vietnam 5.8 55.6 17.4 37.2 50 Maldives 4.9 n.a. n.a. 65.8 12 South Korea 4.8 69.0 38.0 72.5 22 Lebanon 4.3 54.7 71.5 15.5 46 Chile 4.1 62.1 11.0 87.5 27 Mozambique 4.1 56.5 37.4 73.6 153 Mauritius 3.9 n.a. n.a. 65.2 46 Slovenia 3.9 65.9 n.a. 81.6 60 India 3.8 68.1 86.2 47.9 71 Cambodia 3.8 n.a. n.a. 29.7 86 Malaysia 3.7 67.3 –4.1 39.5 30 Poland 3.7 59.9 11.7 58.3 31 Thailand 3.7 57.6 17.7 61.0 33 Bhutan 3.5 n.a. n.a. 56.2 62 Guyana 3.5 61.8 117.4 44.5 46 Sudan 3.4 58.7 92.4 7.4 38 Malta 3.3 76.2 32.9 85.6 n.a. Lao PDR 3.3 n.a. n.a. 71.4 198 Sri Lanka 3.3 60.9 48.4 50.4 50 Samoa 3.2 n.a. n.a. 66.2 68 Dominican Republic 3.2 60.6 41.1 47.3 75 Belize 3.2 n.a. n.a. 60.4 n.a. Bangladesh 3.0 49.6 122.5 23.1 35 Tunisia 3.0 64.7 14.3 54.9 14 Iran 3.0 61.4 7.7 28.2 47 Cape Verde 3.0 n.a. n.a. 60.1 n.a. Lesotho 3.0 n.a. n.a. 46.4 92 Uganda 3.0 56.7 41.2 29.0 36 a. Source: World Bank Database. The 29 (of 143) developing countries with GDP growth above 3% between 1990 and 2003. b. Note: International Country Risk Guide (ICRG) political risk indicators for government stability, corruption, law and order, democratic accountability, and bureaucracy quality. Percent- age based on a total possible rating of 34 points. c. Source: Kaufmann, Kraay, and Mastruzzi (2003). Governance Indicators for 1996–2002: Composite rating of voice and accountability, political stability, government effectiveness, reg- ulatory quality, rule of law, and control of corruption indicators. d. Source: World Bank Doing Business: Benchmarking Business Regulations Database. The rate shows the average number of days needed to establish a business. 16 G O V E R N A N C E A N D A N T I - C O R R U P T I O N : W AY S T O E N H A N C E T H E W O R L D B A N K ’ S I M PA C T Figure 4a. Percentage of Successful Projects and Overall Governance: Strong Correlation All developing countries 100 Percent 50 0 Ϫ2 Ϫ1 0 1 Overall indicator of KKM 95% CI Fitted values Average success rate Average success rate Figure 4b. Percentage of Successful Projects and Corruption: Strong Correlation All developing countries 100 Percent 50 0 Ϫ2 Ϫ1 0 1 2 Corruption indicator of KKM 95% CI Fitted values Average success rate Average success rate 17 Endnotes 1. See IEG Annual Review of Development Effec- 11. See also Fukuyama, 2005, State-Building: Gov- tiveness, 2006. ernance and World Order in the 21st Century, Ithaca, 2. There is no common definition for the term gov- NY, Cornell University Press. ernance and its definition varies across organizations, 12. IEG’s report on Anti-Corruption Activities in but it is broadly defined to refer to the effectiveness World Bank Assistance. of the state in acquiring and exercising the authority 13. Perception indicators are useful—but are sub- to provide and manage public services. Corruption is ject to wide margins of error—and must be carefully an outcome of the failure of the interface between the used. Moreover, changes in perception could come state and society. from greater awareness as well as changes in underly- 3. This is the work of the World Bank Institute ing governance problems. It is also important to unpack (WBI), the work done under the CPIA by OPCS and the indicators for governance—which include a com- the regions and the work done under the rubric of the bination of political, administrative and judicial variables. Business Climate work by the PSD group. Some re- 14. Arndt, Christiane, and Charles Oman, 2006, gional work is also available such as the Europe and “Uses and Abuses of Governance Indicators,” OECD Central Asia Region’s Business Environment and En- Development Centre Studies. terprise Performance Survey (BEEPS). 15. IEG’s recent impact evaluation of health in 4. Governance features twice in the IDA allocation, Bangladesh shows how Bangladesh was able to im- once as part of the CPIA which is the performance prove its health outcomes—such as infant and ma- yardstick for the IDA allocation and separately for the ternal mortality. extra weight attached to governance on top of the per- 16. WDR 1997 showed that it is not just the level, formance allocation itself. but the predictability of corruption that affects growth. 5. The name is now changed to “fragile states.” 17. Knack, Steven, and Philip Keefer, 1995, “Insti- 6. ICRG indicators show deterioration in corrup- tutions and Economic Performance: Cross-Country tion since the mid-1990s. Kaufmann, Kraay, and Mas- Tests Using Alternative Measures,” Economics and truzzi (2005, “Governance Matters IV: Governance Politics 7 (3): 207–27. Indicators for 1996–2004,” World Bank Policy Research 18. Mauro, Paolo, 1995, “Corruption and Growth,” Working Paper Series No. 3630), after surveying the Quarterly Journal of Economics, 110: 681–712. 12 indicators used in their work, conclude that there 19. Hall, Robert E., and Charles I. Jones, 1999, has been no improvement in the governance indica- “Why Do Some Countries Produce So Much More tors since 1996 and, if anything, the weight of the ev- Output per Worker than Others?” Quarterly Journal idence shows a deterioration. of Economics 114 (1): 83–116. 7. IEG’s Kenya CAE update warned of the lack of 20. Acemoglu, Daron, Simon Johnson, and James real institutional change in the country and the need A. Robinson, 2001, “The Colonial Origins of Compar- for caution. ative Development: An Empirical Investigation,” Amer- 8. When problems are identified the Bank has re- ican Economic Review 91 (5): 1369–401; ibid, 2002, course to a variety of options—requesting a review of “Reversal of Fortune: Geography and Development in bids, misprocurement, sanctions, etc. the Making of the Modern World Income Distribution,” 9. See “Monitoring Corruption: Evidence from a Quarterly Journal of Economics 117 (4): 1231–94. Field Experiment in Indonesia,” Benjamin A. Olken, 21. Easterly, William, and Ross Levine, 2003, “Trop- NBER, October 2004. ics, Germs, and Crops: How Endowments Influence 10. Based on IEG’s evaluation of World Bank sup- Economic Development,” Journal of Monetary Eco- port for community-based and -driven development. nomics 50 (1): 3–39. 19 F R O M S C H O O L I N G AC C E S S TO L E A R N I N G O UT C O M E S : A N U N F I N I S H E D AG E N DA 22. Dollar, David, and Aart Kraay, 2003, “Institutions, Human Capital, or Constituencies?” NBER Working Trade and Growth,” Journal of Monetary Economics Paper 12093, National Bureau of Economic Research, 50 (1): 133–62. Inc.; Christiane Arndt and Charles Oman, 2006, “Uses 23. Rodrik, Dani, Arvind Subramanian, and and Abuses of Governance Indicators,” OECD De- Francesco Trebbi, 2002, “Institutions Rule: The Primacy velopment Centre Studies. of Institutions over Geography and Integrationin Eco- 26. The regressions show a strong significant effect nomic Development,” NBER Working Paper 9305, Na- of governance on project performance, controlling for tional Bureau of Economic Research (October). income per capita and education. The same results 24. Glaeser, E.L., R. La Porta, F. Lopez-de-Silanes, were obtained in World Development Report, 1997, and A. Shleifer, 2004, “Do Institutions Cause Growth?” and in Dollar, David, and Levine, Victoria, 2005, “Sow- Journal of Economic Growth 9 (3): 271–303. ing and Reaping: Institutional Quality and Project 25. Rajan, Raghuram G., and Luigi Zingales, 2006, Outcomes in Developing Countries,” World Bank Pol- “The Persistence of Underdevelopment: Institutions, icy Research Working Paper No. 3524. 20