Report No. 29620 Mainstreaming Anti-Corruption Activities in World Bank Assistance A Review of Progress Since 1997 July 14, 2004 Operations Evaluation Department Document of the World Bank 1 Contents Acknowledgements ...................................................................................................................... v Preface ........................................................................................................................................ vii Summary ..................................................................................................................................... ix 1. Introduction ............................................................................................................................. 1 2. The World Bank’s Support for Governance Reform ........................................................ 5 Historical Perspective on the Bank’s Approaches to Governance Reforms ..................... 5 Main Tools in the Bank’s Package o f Assistance ............................................................. 7 3 . Implementation of the 1997 Anti-Corruption Strategy ................................................... 11 Preventing Fraud and Corruption in Bank Projects ........................................................ 13 Procurement ............................................................................................................. 16 Helping Countries that Request Bank Assistance in Combating Corruption ................. 17 Mainstreaming a Concern for Corruption in Bank’s Work ............................................ 20 Mainstreaming a Concernfor Corruption in the CAS ............................................. 21 Mainstreaming a Concernfor Corruption in Lending ............................................. 21 Providing Active Support to Intemational Efforts to Address Corruption ..................... 24 Support for Transparency International (TI) ........................................................... 24 Anti-Bribery Conventions ......................................................................................... 25 Anti-Money Laundering and Combating the Financing o f Terrorism ..................... 25 4 . Assessing Relevance of Bank Support for Anti-Corruption ....................... 27 H o w Relevance i s Defined and Assessed? ..................................................................... 27 Relevance of CAS ...................................................................................................... 28 Relevance of Lending ............................................................................................... 30 TrainingLeaming Activities o f the Governance Unit o f the World Bank Institute . 33 5 . Overview of the Case Studies ............................................................................................. 35 Evaluation Questions ...................................................................................................... 35 Common Factors ............................................................................................................. 36 Country Ratings-Relevance and Efficacy .................................................................... 43 A Gap Between “Relevance” and “Efficacy” ................................................................. 43 6 . Conclusions .......................................................................................................................... Conclusions o f the Desk Review .................................................................................... 47 47 Conclusions o f Case Studies ........................................................................................... 47 7 . Recommendations ..................................................................................... 51 References ................................................................................................................................... 53 Annexes Annex 1: Methodology and Sample Selection ............................................................ 59 Annex 2: Criteria for Rating Compliance o f Instrument for Anti-corruption Strategy ....................................................................................................... 63 Annex 3: Criteria for Rating Relevance o f Instrument for Anti-corruption Strategy . 64 ii Contents (cont ’d.) Annex 4: Degree of Bank Involvement .In Country Workshops and Surveys (FY98-01) ................................................................................................... 63 Annex 5: Degree of Bank Involvement - ESW and Mission Reports (FY98-01) ...... 64 Annex 6: Degree of Bank Involvement - Grant-Based Technical Assistance (FY98-0 1) ................................................................................................... 65 Annex 7: Degree of Bank Involvement - Govemance Related Lending (FY98-01) .. 66 Annex 8: Framework for Evaluation of Relevance: Principal Agent Theory and New InstitutionalEconomics ...................................................................... 67 Annex 9: List of Projects Analyzed ............................................................................ 71 Annex 10: Country Case Studies .................................................................................. 75 Attachments Attachment 1: Management Response to the OED Review ............................................... 97 Attachment 2: Chairperson’s Summary ............................................................................ 109 Text Boxes Box 1.1: Corruption and Development .............................................................................. 1 Box 1.2: Legal Basis of the Bank‘s Mandate to Curtail Corruption and Improve Governance .......................................................................................................... 2 Box 2.1: Impact o f the Bank’s Trade and Financial Liberalization Reforms on Corruption......................................................................................... 6 Box 3.1 Implications of a Policy of Zero Tolerance for Corruption But High Threshold of Proof: Lessons from Project Experiences in Uganda and Bangladesh ..................................................................................... 14 Box 3.2: Mainstreaming a Definition of “Mainstreaming” .............................................. 20 Box 4.1: Example of a Logical Chain for an Anti-Corruption Intervention..................... 27 Box 4.2: Results from Self-evaluationo f the WBI ........................................................... 34 Text Tables Table 2.1: World Bank Assistance for Governance Reforms . FY98 to FYOl .................... 7 Table 2.2: Level of Corruption by Region ............................................................................ 8 Table 2.3: World Bank Lending for Governance Reforms . FY98 to FYOl ....................... 9 Table 2.4: Package of Reforms by Incidence of Corruption ................................................. 9 Table 3.1: Assessing the Risk of Corruption Among the World Bank’s 10 Largest Borrowers ........................................................................................................... 13 Table 3.2: Countries Requesting World Bank Assistance for Anti-Corruption Activities . 17 Table 3.3: The Effect of Anti-Corruption Initiatives on CPARs ........................................ 19 Table 5.1 Basic Indicators ................................................................................................. 35 Table 5.2: Summary of Ratings ........................................................................................... 43 Text Figures Figure 2.1: Degree o f Bank Involvement .Governance Based Lending ............................... 8 Figure 3.1: Implementationof the Anti-corruption Strategy by Corruption Incidence Class. FY98-01 ................................................................................................. 11 Figure 3.2: Extent of Implementationin All Lending Weighted by Disbursements ............ 12 Figure 3.3: Extent of Implementationin All Lending by Level of Corruption .................... 12 Figure 3.4: Extent o f Imdementation in CASs. FY98-01 ................................................... ~ 21 iii Contents (cont 'd.) Figure 3.5: Extent of Implementationin CASs by Level of Corruption .............................. 22 Figure 4.1: Countries with Relevant CAS by Level of Corruption ...................................... 29 Figure 4.2: Relevance of All Lending Weighted by Disbursements .................................... 32 Figure 4.3: Relevance of Lending by Level of Corruption .................................................. 32 Figure 5.1: Factors Affecting CorruptionAddressed by Bank Interventions ...................... 45 Annex Figures Figure A 1.1: Degree of Bank Involvement-Grant Based Technical Assistance ............ 59 Figure A 1.2: Degree of Bank Involvement-ESW .......................................................... 59 Figure A 1.3: Degree of Bank Involvement-In-Country Workshops and Surveys ......... 60 Annex Tables Table A 1.1: ESW Coverage by Level of Corruption (chapter 3) .................................... 57 Table A1O.l: Bank-Supported Activities Covered in the Guatemala Country Study ....... 76 Table A10.2: Factors Affecting Corruption Addressed by the Bank in Guatemala .......... 77 Table A10.3: Bank-Supported Activities Covered in the Kenya Country Study .............. 79 Table A10.4: Factors Affecting CorruptionAddressed by the Bank in Kenya ................. 79 Table A10.5: Bank-Supported Activities Covered in the Latvia Country Study .............. 82 Table A10.6: Factors Affecting Corruption Addressed by the Bank in Latvia ................. 83 Table A10.7: Bank-Supported Activities Covered in the Pakistan Country Study ........... 84 Table A10.8a: Factors Affecting Corruption Addressed by Bank Lending Activities in Pakistan ................................................................................... 85 Table A10.8b: Factors Affecting Corruption Addressed by Bank Non-Lending Activities in Pakistan ................................................................................... 86 Table A10.9: Bank-Supported Activities Covered in the Philippines Country Study ....... 88 Table A10.loa: Factors Affecting Corruption Addressed by Bank Lending Activities in the Philippines ........................................................................ 89 Table A10 . lob: Factors Affecting Corruption Addressed by Bank Non-Lending Activities in the Philippines ........................................................................ 90 Table A10.11: Bank-Supported Activities Covered in the Tanzania Country Study .......... 92 Table A10.12: Factors Affecting Corruption Addressed by the Bank in Tanzania ............ 93 V Acknowledgements The study team i s grateful to Robert Picciotto, Gregory Ingram and Ruben Lamdany for their comments and editorial contributions. This report has also benefited from comments on earlier drafts b y PREM Public Sector Board, Paul Bermingham, Christopher Clague, Peter Eigen, Osvaldo Feinstein, N i l s Fostvedt, Roy Gilbert, Patrick Grasso, Cheryl Gray, Akhtar Hamid, John Heilbrun, Walter Hoffer Daniel Kaufmann, Robert Klitgaard, Susan Rose-Ackerman, Mark Schacter, Miguel Schloss, Frederick Stapenhurst, Michael Stevens, Vito Tanzi, Werner Thut, Gerolf Weigel, Alexandre Widmer, and Hanspeter Wyss. The financial and intellectual support o f the Danish Governance Trust Fund, Norwegian Partnership Fund, and the Swiss Development Cooperation Agency i s also gratefully acknowledged. The team i s also grateful to Agnes Santos for her outstanding editorial and support assistance. vii The desk review (presented in chapters 2 4 ) has been prepared by Anwar Shah (Team Leader) and Theresa Thompson with contributions from Azam Chaudhry. I t draws upon earlier versions of this report prepared by a team led by Anwar Shah (World Bank) and comprising Professor Christopher Clague (University of California-San Diego), Jeff Huther (US.Treasury), and Mark Schacter (Institute on Governance, Ottawa). The first part of the report presents a synthesis of the findings based upon desk reviews conducted by a panel o f evaluators that included: Mark Schacter, Jeff Huther, P.A. Sharafudheen, Mita Marra, Ewa Tomaszewska, Azam Chaudhry, Tugrul Gurgur, Melissa Thomas, Theresa Thompson, Thornton Matheson, Shamit Chakravarti, and Sohail Malik. The framework for this part of the evaluation was developed by the core team for this study comprising Susan Rose-Ackerman (Yale University), Mark Schacter, and Anwar Shah. The case study synthesis (presented in chapter 5 and annex 10) has been prepared by a team comprising Mark Schacter, Anwar Shah (Team Leader), and Theresa Thompson. The evaluation framework for the country case studies was developed by Anwar Shah. The six country case studies are based upon field missions directed by Anwar Shah and conducted by: e Guatemala Jose Garibaldi e Kenya Daniel Kobb e Latvia Juris Dreifelds (Brock University) e Pakistan Rashid Ahmad, Azam Chaudhry, Ghaffar Chaudhry (Joint Director, Pakistan Institute of Development Economics), Abdul Razzak Kemal (Director, Pakistan Institute of Development Economics), Mushtaq Khan (University of London), Zubair Khan, Anwar Shah, Mansoor Shah, and Theresa Thompson e Philippines Rashid Ahmad, P.A. Sharafudheen, Emmanuel de Dios (University of the Philippines), Salvador Enriquez (Center for Public Resource Management, Inc.), Anwar Shah, Karen Sondergaard (PLS Ramboll Management), and Proserpina Tapales (University o f the Philippines) Tanzania Pius Fischer (NADEL-Swiss Federal Institute of Technology), Daniel Kobb, and Robert Makaramba The report i s based upon the following background papers, of which selected papers will be published in the OED Working Paper Series: Chaudhry, Azam. Forthcoming. “Corruption and Trade Liberalization.’’ ... Vlll e Gurgur, Tugrul and Anwar Shah. 2002. “Localization and Corruption: Panacea or a Pandora’s Box,” In Ehtisham Ahmad and Vito Tanzi (eds.), Managing Fiscal Decentralization, pp. 46-67. London and New York: Routledge Press. e Huther, Jeff and Anwar Shah. 2000. Anti-corruption Policies and Programs: A Framework for Evaluation. Policy Research Working Paper Series No. 2501. The World Bank. Rose-Ackerman, Susan. Forthcoming. “Guidelines for Reviewing World Bank Efforts to Control Corruption.” Schacter, Mark. Forthcoming. “Institutions o f Accountability and Corruption: An Analytical Framework.” Schacter, Mark and Anwar Shah. 2000. “Look Before You Leap: Notes for Corruption Fighters.” Presented at the International Conference on Combating Corruption, Seoul, Korea, December 2000. e Shah, Anwar. 1999. “Governing for Results in a Globalized and Localized World.” The Mahbubul Haq Memorial Lecture, The Pakistan Development Review 38:4 Part 1 (Winter 1999): 385-431. Thompson Theresa. Forthcoming. “Corruption and the Economic Rate of Return on World Bank Funded Projects.” Thompson, Theresa and Anwar Shah. Forthcoming. “Transparency International’s Corruption Perceptions Index: Whose Perceptions Are They Anyway?” Tomaszewska, Ewa and Anwar Shah. Forthcoming. “Phantom Hospitals, Ghost Schools and Roads to Nowhere: The Impact of Corruption on Public Service Delivery Performance in Developing Countries.” ix Summary This evaluation examines the early progress made in implementing the World Bank’s anti-corruption strategy that was launched in September 1997. I t assesses the Bank’s anti-corruption activities between FY98 and FYO1. The report provides a comprehensive desk review of all World Bank initiatives in mainstreaming anti-corruption concerns in i t s work. In addition, i t examines relevance and early outcomes of Bank support for anti-corruption activities in Guatemala, Kenya, Latvia, Pakistan, Philippines, and Tanzania based upon field studies. The desk review concludes that the Bank has come a long way in a short time, with corruption concerns increasingly integrated into operational activities and internal processes. Understanding of the causes and dynamics of corruption in different country settings has improved. B y encouraging open debate, carrying out policy research, and disseminating good practice to member countries and international organizations, the Bank has helped to raise the profile of governance issues on the development agenda. As a result, support for anti-corruption work has grown within the development community. The country case studies confirm the results of the desk review. They demonstrate that the Bank has supported highly relevant anti-corruption activities on the ground. However, the Bank has demonstrated only modest success so far in achieving durable outcomes. The short time elapsed since the anti-corruption program was launched, the unusual complexity of the task at hand, and the magnitude o f the challenge account for the gap between relevance and efficacy. The Bank’s anti-corruption activities have addressed key factors affecting corruption, including state intervention, rule of law, greater transparency in public sector operations, and capacity building aimed at good government and a competent civil service. But corruption i s grounded in political contexts and social fragmentation over which the Bank has limited influence. Furthermore, long term capacity development in such areas i s not always the Bank’s comparative advantage. Therefore, greater reliance on development partners may be warranted provided, o f course, borrowers are receptive. A better understanding o f social and political factors at the country level would enhance the quality and impact of Bank advice and improve the design o f the Bank’s anti-corruption interventions. As the Bank moves on to a more mature phase o f i t s anti-corruption work, i t will need to be even more creative in i t s choice o f instruments, the nurturing of partnerships, and the sequencing of operations. Highly differentiated strategies are required to find adequate entry points and address the most important causes of corruption. Corruption w i l l be most effectively addressed through long-term institutional reforms, including the governance environment for public and private sector institutions, as well as government-civil society engagement. To this end, the following challenges will require sustained attention: X Deal with the demand dilemma by fostering demand among a wide spectrum o f stakeholders. The Bank’s current approach targets anti-corruption support to countries that demand such assistance since ownership i s critical in fighting corruption. Unfortunately, the countries (regions and sectors) most in need of “anti-corruption” assistance often tend to be the least interested in receiving it. Where acceptable to public authorities, the Bank should deal with this problem by fostering demand among a wide spectrum o f stakeholders in those countries. Continue assessments o f the governance environment. The Bank should continue to assist clients in carrying out assessments of their institutional environments, including a diagnostic of the main causes of corruption. I t should help develop a long term, holistic and sequenced reform program. The Bank and other donors should work with the government in identifying the areas where they can best help in implementing these reforms. The CAS should articulate the Bank’s role. Define the governance pre-requisitesfor lending. Corruption poses significant risks for the effectiveness of Bank lending, and therefore the Bank should set minimum governance pre-requisites for lending, including governance assessments. This should lead to greater selectivity in lending and increased reliance on advisory activities in high corruption environments where there i s limited anti-corruption demand. Some governments may be reluctant to engage the Bank in a dialogue on corruption issues, because they believe that this i s not needed or helpful. In such cases, the Bank may need to make an independent assessment, to be reflected in the CAS, of whether corruption would impede the effectiveness of lending. Focus lending instruments on accountability for results. A sharper focus on accountability for results would help create appropriate incentives and promote public sector reform. Efforts to improve service delivery can help address corruption and i t s causes. The Bank should increase i t s support for reforms in service delivery to emphasize probity and results rather than the level o f expenditures. Linking development assistance to demonstrable improvements in output performance of the public sector, especially in areas most relevant to the poor, may be the most effective characteristic o f the Bank’s evolving anti-corruption strategy. Provide more assistance for bottom-up reforms. In many countries where corruption i s entrenched, governments lack either the w i l l or the capability to mount effective anti-corruption programs. In such countries, the Bank and/or i t s partners may choose to amplify citizens’ voice and strengthen exit mechanisms so as to enhance transparency, accountability and the rule o f law. Gregory K. Ingram Director-General Operations Evaluation 1 1. Introduction 1.1 Recognizing that corruption often compromises the effectiveness of external assistance (box 1.1), World Bank President James Wolfensohn in 1996 urged the development assistance community to fight against the “cancer o f corruption.”’ He said: ‘‘...the Bank Group will not tolerate corruption in the programs that we support, and we are taking steps to ensure that our own activities continue to meet the highest standards of probity.” In September 1997, the World Bank Executive Board formally recognized corruption as a major development issue by approving the Bank’s anti-corruption strategy. The strategy called for the Bank to address corruption-defined as the “abuse o f public office for private gain”-along four pillars or dimensions: 1. Preventing fraud and corruption in Bank projects; 2. Helping countries that request Bank assistance for corruption; 3. Mainstreaming a concern for corruption into the Bank’s work; and 4. Giving active support to international efforts to address corruption. I Box 1.1: Corruption and Development Negative consequences of corruption on the investment climate are confirmed by empirical studies (see Brunetti, Kisunko, and Weder 1997, Wei 1997a and b, Hines 1995, Keefer and Knack 1995, and Mauro 1995, among others). Empirical studies report a negative relationship between per capita GDP and corruption (see Hall and Jones 1999, Husted 1999, Kaufmann, Kraay, and Zoido-Lobaton 1999, and Paldam 1999). Corruption lowers the quality o f public infrastructure (see Tanzi and Davoodi 1998, and Isham et al. 1995, 1997), education and health services, and adversely affects capital accumulation (Lambsdorff 1999b). Bribes encourage regulatory burden and red tape (Kaufman and Wei 1999). Corruption distorts public expenditures by channeling resources to large-scale capital investments and away from education and health (Mauro 1997). Econometric analysis conducted for this study indicates that public-sector corruption i s associated with lower public satisfaction with health care, roads and transportation (see Tomaszewska and Shah, forthcoming). Corruption undermines the official economy (Johnson et al. 1997), reduces the effectiveness o f development aid and increases income inequality and poverty (Gupta et al. 1998). A review o f Bank projects completed during 1982-1997 found that a one point increase in corruption (according to the ICRG’s rating based on a six-point scale) lowers a project’s rate of return by 2.5 percent (Thompson, forthcoming). This opened the door for the Bank to address corruption head-on as a major development issue. Staff and management had long been aware of the importance of fighting corruption but felt unsure about the Bank’s mandate (box 1.2). 1.2 The Bank views i t s anti-corruption effort as part o f i t s work to improve governance, “not as a set o f new initiatives but rather as a more explicit integration of the problem into country strategy formulation, Bank lending, economic and sector work, ’ Mr. Wolfensohn’s opening address to the Board o f Governors at the 1996 Annual Meetings. 2 research and country dialogue” (World Bank 1997a, p. 47). The 1997 Strategy Paper laid out the general framework and basic details on how the four pillars would be addressed in practice. Subsequent documents provided further guidance on the implementation of the strategy. Most noteworthy i s the Public Sector Group’s strategy paper (Reforming Public Institutions and Strengthening Governance: A World Bank Strategy, published in 2000), where the Bank stressed that i t views corruption as a symptom of dysfunctional institutions, and that its anti-corruption strategy i s to address these dysfunctions in client countries. Other relevant documents that reflect upon the anti-corruption strategy include Europe and Central Asia’s (ECA) Anti-corruption in Transition (presented at the 2000 Prague Annual Meetings), the Africa Region’s report, Can Africa Claim the 21“ Century, the 2001 World Development Report on Attacking Poverty, and the Public Sector Board self-evaluations including Helping Countries Combat Corruption: Progress at the World Bank Since I997 (2000) and Reforming Public Institutions and Strengthening Governance: A World Bank Strategy Implementation Update (2003). Box 1.2: Legal Basis of the Bank’s Mandate to Curtail Corruption and Improve Governance As early as 1991, The Bank’s General Counsel had developed an explicit legal basis for its work on governance. Mr. Ibrahim Shihata, then the General Counsel o f the Bank, argued that: “the concept o f governance in the sense o f overall management o f country’s resources cannot be irrelevant to an international financial institution which at present not only finances projects but also i s deeply involved in the process o f economic reform carried out b y i t s borrowing members. Clearly the concern here i s not with the exercise o f powers in the broad sense but specifically with the appropriate management o f the public sector and creation o f an enabling environment for the private sector. I t i s a concern for the rules which are actually applied and institutions which ensure the appropriate application o f these rules, to the extent that such rules and institutions are required for the economic management o f the country and in particular the sound management o f i t s resources” (Shihata 1991, p. 54). H e further observed that the Bank’s Articles o f Agreement authorize the Bank to address corruption with care: “. ... The World Bank can hardly insulate itself from major issues of international development policy. Corruption has become such an issue. I t s prevalence in a given country increasingly influences the flow of public and private funds for investment in that country. The Bank’s lending programs and in particular its adjustment lending take into account factors which determine the size and pace o f such flows. From a legal viewpoint, what matters i s that the Bank’s involvement must always be consistent with i t s Articles of Agreement. The Bank can in my view take many actions to help the fight against corruption. I t can conduct research on the causes and effects of this worldwide phenomenon. It can provide assistance, by mutual agreement, to enable its borrowing countries to curb corruption. I t may take up the level of corruption as a subject of discussion in the dialogue with i t s borrowing members. And, ifthe level of corruption i s high so as to have an adverse impact on the effectiveness of Bank assistance, according to factual and objective analysis, and the govemment i s not taking serious measures to combat it, the Bank can take this a s a factor in its lending strategy towards the country. The only legal barrier in this respect i s that in doing so the Bank and its staff must be concerned only with the causes and effects and should refrain from intervening in the country’s political affairs. (emphasis added) While the task may be difficult in borderline cases, i t s limits have been prescribedin detail in legal opinions endorsed by the Board” (World Bank 1990). The Bank has also observed that i t s anti-corruption strategy i s motivated b y its fiduciary duty to member countries, which states that “the Bank shall make arrangements to ensure that the proceeds o f any loan are used only for the purposes for which the loan was granted” (Article 1 11, Section 5(b)). 3 1.3 The first part o f this report (contained in chapters 2 4 ) i s based on a desk review o f the implementation o f the anti-corruption strategy during FY98-01 covering all Bank programs, policies and analytical advisory services dealing with governance reform issues. I t assesses the design o f the strategy relative to i t s objectives, clients’ development priorities, and the Bank’s corporate goals. The focus i s on the second and third pillars (helping countries that request Bank assistance for corruption and mainstreaming a concern for corruption into the Bank’s work); with less emphasis on the Bank’s fiduciary obligations (preventing corruption in Bank projects) or i t s support to other international efforts. The case studies (introduced in chapter 5, with details in annex 10) provide field assessments o f the early experience in helping six borrowers combat corruption: Guatemala, Kenya, Latvia, Pakistan, Philippines, and Tanzania. 1.4 The remainder o f this study i s organized in the following manner. Chapter 2 briefly describes the evolution o f Bank approaches to governance reform and then presents the current menu o f activities in support o f governance reform. Chapter 3 examines the implementation of the four dimensions or pillars o f the anti-corruption strategy. Chapter 4 assesses the relevance o f selected Bank activities in addressing corruption. Chapter 5 briefly summarizes the case studies. Chapter 6 brings together the conclusions o f the desk review and case studies. Chapter 7 concludes with recommendations. 5 2. The World Bank’s Support for Governance Reform 2.1 This section briefly looks at the evolution of Bank approaches to governance reform and then describes the Bank’s main activities in support of governance reform since the inception of the anti-corruption strategy.2 I t also presents the distribution of these activities among countries with different degrees of corruption. Historical Perspective on the Bank’s Approaches to Governance Reforms 2.2 The Bank had always been concerned with the negative impact of poor g~vernance,~ and in particular of high corruption, on the effectiveness o f external assistance. There i s wide agreement that public-sector corruption cannot be properly understood or addressed as an independent phenomenon. Rather i t i s a symptom of a failed governance at the country level.4 The presence and prevalence o f corruption therefore depend on factors relating to the governance environment, such as the quality of public sector management, the nature of the accountability relationship between the government and its citizens, the legal framework for development, and the degree to which public sector processes are accompanied by transparency and the dissemination of ~ report looks at explicit anti-corruption assistance as well as those i n f ~ r m a t i o n .This governance reforms that are likely to bear on anti-corruption. Therefore, in addition to taking precautions in i t s projects through close supervision, procurement, and disbursement rules and audits, the Bank tried to attack corruption through i t s longstanding support to governance reform. Also, the reforms that the Bank supported in the 1980s and 199Os, including trade and financial liberalization, fiscal and regulatory reforms, and institutional strengthening (box 2.1) were partly aimed at reducing the incentives for corruption. But the launching of the 1997 strategy was an important change that put the fight against corruption at the forefront of the Bank development agenda. ’ This study takes the view argued by the Public Sector Board that “evaluating anti-corruption efforts as distinct from the Bank’s core operations of governance i s inappropriate” as the causal relationship between the two run in both directions (PSB Comments 11/06/2000). Thus, all Bank assistance for governance reforms was reviewed through the anti-corruption lens regardless o f whether they are explicitly or implicitly directed at curtailing corruption. This report focuses on the period from FY98-01. We recognize, however, that the Bank’s implementation o f the anti-corruption strategy has continued to evolve since then with the addition of new tools and scaling up of others that this report has not been able to evaluate. Some o f these new activities (introduced after FYO1) include public expenditure tracking and benchmarking, enhanced emphasis on Community Driven Development (CDD) projects in the Africa Region, WBI’s work with civil society in Guatemala, and expanded provincial adjustment lending in India and Pakistan. The World Bank (1989) defined governance as “ the exercise of political power to manage a nation’s affairs” (p. 60). U.N. Secretary General Boutros Boutros Ghali (1992) defined good governance to include “sensible economic and social policies, democratic decision making, adequate governmental transparency, financial accountability, creation of a market friendly environment for development, measures to combat corruption, as well as respect for the rule of law and human rights” (paragraph 59). 4 For example, see Gray and Kaufmann (1998); Huther and Shah (1998,2000); Gurgur and Shah (2002); Schacter and Shah (2000); and World Bank (2000d). This follows from the Bank’s own definition of “governance.” See World Bank (1994). 6 Box 2.1: Impact of the Bank’s Trade and Financial Liberalization Reforms on Corruption World Bank anti-corruption efforts have been in existence long before they were formalized in the 1997 anti-corruption strategy. Trade and financial liberalization have long been championed by the Bank as a way to promote economic development, but they are effective anti-corruption tools as well. Trade policies create the opportunity for discretion by public officials and are more open to rent seeking. High tariffs, non-uniform tariff regimes, and non-tariff barriers can create strong incentives for bribery, under invoicing, mis-declaration and smuggling. Programs that reform exchange rate mechanisms, eliminate export barriers, reduce tariffs, and shift trade barriers from non- tariff barriers to tariffs can increase transparency. A background paper for this study (see Chaudhry, forthcoming) analyzed the impact of trade liberalization on corruption using data from 85 countries for the period 1980-1997. This paper concluded that countries that successfully pursued trade liberalization achieved lower levels of corruption than countries that did not pursue such policies. Also, policies focusing on exchange rate liberalization significantly reduced the black market premia on foreign currency, thereby reducing opportunities for corruption. Financial liberalization can reduce corruption by reducing the monopoly rents that accrue from financial market rigidities. Financial sector reforms fall into four categories: (1) Removing distortions by introducing flexible interest rates and reducing credit distortions, (2) Increasing competition by eliminating public sector monopoly of banks and allowing foreign entry into the financial sector, (3) Reforming institutions by introducing laws, regulations, administrative practices to strengthen capital markets and corporate governance, and (4) Strengthening the financial sector by improving central bank supervisory practices and increasing the technical capabilities o f financial institutions. A review o f 88 financial reform projects in 23 countries reveals that only 3 1 percent of the projects had satisfactory outcome ratings. In view o f this, the full potential of financial liberalization programs in reducing corruption was probably not achieved. ~ Source: Background research carried out for this report. 2.3 The impact on corruption of governance reforms was stressed in many Bank studies since the 1980s, including the World Development Report 1983, the 1989 Africa Long Term Perspective Study, and the 1992 and 1994 discussion papers on ‘Governance and Development’ and ‘Governance: The World Bank’s Experience.’ In 1994, OED’s Annual Review o f Evaluation Results showed the negative association o f corruption with Bank project outcomes. 2.4 In the late 1980s the Africa Region (AFR) was the first operational unit in the Bank to identify governance issues as necessary for the success of i t s development agenda. AFR recognized that “underlying the litany of Africa’s development problems i s a crisis o f governance” (World Bank 1989, p. 60). AFR argued for a “systematic effort to build a pluralistic institutional structure, a determination to respect the rule o f law, and vigorous protection o f the freedom of the press and human rights” (World Bank 1989, p. 61). T o further this agenda, AFR management directed more efforts at curtailing corruption, as well as promoting greater accountability, transparency, openness, respect for rule of law, and human rights in borrowing countries. Public sector management and civil service reforms were given greater emphasis, and analytical work was launched on military expenditures-a controversial matter in Bank-country dialogue-(Serageldin 1990). The AFR lead was followed by the Latin America and the Caribbean (LAC), which supported civil service and judicial reform and provided assistance for decentralization. Since then, all regions have been addressing governance issues. 7 2.5 Bank effectiveness in this area was limited by the lack of appropriate staff s k i l l s and of a policy framework, and an unclear institutional mandate. The situation has improved in recent years. The Bank has now over 200 staff dedicated to public sector reform issues. I t has integrated institutional analysis into i t s diagnosis work and into country assistance strategies. Finally, the Bank’s mandate in this area has been strengthened since 1997. The Public Sector Group’s strategy paper entitled Reforming Public Institutions and Strengthening Governance outlined a policy framework and a broad operational approach. The Regions have been adapting this framework to suit regional and country circumstances (see World Bank 2000d).6 The Comprehensive Development Framework (CDF) and the Poverty Reduction Strategy Paper (PRSP) initiatives also treat governance reform as critical for development assistance; nine out of the first 11 PRSPs identified governance as a major pillar. Table 2.1: World Bank Assistance for Governance Reforms FY98 to FYOl - (50%) (63%) (38%) (63%) Total 230 54 180 63 117 52 175 70 Source: World Bank staff calculations - see annex 1 for details. Main Tools in the Bank’s Package of Assistance 2.6 Table 2.1 summarizes the regional distribution o f some of the governance related activities, namely: (a) raising awareness about corruption with in-country workshops and surveys, (b) identifying weaknesses in governance through economic and sector work, and (c) providing technical and lending assistance to introduce reforms in a large number of countries. Countries are classified into four categories (low, modest, substantial, and high) based on the perceived levels of corruption (table 2.2 and annex 1). This classification i s based on the Bank’s Country Policy and Institutional Assessment (CPIA) ratings which are highly correlated with other publicly available corruption indices (e.g., For example, the EAP Region developed a handbook on anti-corruption to assist i t s task managers in combating corruption in Bank projects. 8 Corruption Perceptions Index (CPI) and the International Country Risk Guide (ICRG)). Most of the Bank’s efforts in terms of these activities have gone to countries with modest or substantial corruption (figures 2.1, and A l . 1 through A1.3). This reflected the corresponding client demands, as well as the Bank’s assessment of the expected effectiveness of these operations. Table 2.2: Level of Corruption by Region High Corruption Substantial Corruption Modest Corruption Low Corruption AFR 9 27 9 2 EAP 2 12 5 0 ECA 5 14 8 2 LAC 5 14 7 2 MNA 2 6 3 0 SAR 0 5 1 1 Total 23 78 33 7 Source: World Bank staff calculations based on CPIA. - Figure 2.1: Degree of Bank Involvement Governance Based Lending 1 0 Negligible Involvement ( ~ $ per 5 capita) El Modest ($5-15 per capita) Substantial ($15-20 per capita) High Substantial Modest Low m0 High (>$20 per capita) Corruption Corruption Corruption Corruption Level of Corruption (CPIA) Source: World Bank staff calculations - see annex 1 for details. 2.7 In-country Workshops and Surveys are often the first activities to address corruption concerns, as they help develop a consensus for reform and serve to break the ice in introducing the subject of corruption in Bank-country dialogue and aiding in corruption diagnostics. Over the period of FY98-01, the Bank has supported at least 230 governance related workshops and surveys in 54 countries, mainly in AFR and ECA. Over the period FY98-01, the Bank was able to engage one-third o f the 23 highly corrupt and half of the 78 countries with substantial corruption in a dialogue on corruption through these workshops and surveys. An example of the diagnostics include the Public Expenditure Tracking Surveys (PETS) to track public spending from the budget to service delivery. The World Bank Institute (WBI) also carries out various surveys o f households, firms, and public officials. Responding to country requests, workshops and 9 surveys were also organized in countries with modest and low incidence o f corruption, which served to transfer knowledge about success stories to other countries undertaking governance reform efforts (annex 4). 2.8 Economic and Sector Work: Governance related economic and sector work (ESW) builds country specific knowledge so that anti-corruption activities can be tailored to local needs. There were 180 examples of governance related ESW prepared for 63 different countries. The SAR, ECA, and EAP produced about four reports for each of the countries studied. In each of the corruption categories about half of all countries were covered. But as can be expected in a mostly demand driven initiative, only in two of the high corruption countries did the Bank prepare more than two studies (annex 5). Table 2.3: World Bank Lending for Governance Reforms - FY98 to FYOl Total Governance Lending per Governance Total Lending . Lending capita Lending per capita (in US$ billion) (in US$ billion) (US$) (US$) AFR 10.5 2.1 16.29 4.21 EAP 24.5 3.1 13.34 1.69 ECA 16.2 4.3 34.25 9.00 LAC 23.1 2.9 45.55 5.11 MNA 4.0 0.52 13.10 1.80 SAR 11.8 0.83 8.81 0.62 All Regions 90.1 14.35 Source: World Bank, various data bases. 2.9 Grant Based Technical Assistance: Grant based technical assistance provides support to anti-conuption agencies, ethics commissions, and parliaments; as well as for reforming legal systems, civil service, financial management functions, and tax and customs administration. This technical assistance may help countries upgrade capacity for fighting corruption. About 117 instances of grant based technical assistance (TA) related to governance were received by 52 countries. About 40 percent o f countries in the high, substantial and modest corruption categories received grant based technical assistance (annex 6). Table 2.4: Package of Reforms bv Incidence of Corruption I High Substantial Modest Low Reform Program Corruption Corruption Corruption Corruption I I LegaVJudicial Reform 4 Source: World Bank staff calculations - see annex 1 for details. 10 2.10 Lending Programs: During FY98-01, Bank loans have supported 175 governance related projects worth US$14 billion in 70 countries, comprising more than 15 percent o f Bank lending (table 2.3). According to the Public Sector Board, in FYOl there were 22 projects that were considered “core” public sector r e f ~ r m . In~ the five years since the approval of the anti-corruption strategy, 40 percent of the high corruption countries have received governance-related lending. This type o f lending has been greater in countries in the substantial and modest categories (figure 2.1, annex 7), where the demand for i t was higher. 2.1 1 Targeting the assistance package to specific country conditions. Tailoring of programs to the country specific causes of corruption i s in i t s initial phase and i s not fully reflected in the sample of projects reviewed (covering through FYOl). Table 2.4 shows that three of the most common project objectives (budget administration, privatization, and civil service reform) are shared by the high, substantial, and modest corruption countries. There are some small differences, e.g. judicial reform efforts have targeted substantial corruption countries and trade liberalization efforts have targeted both substantial and modest corruption countries. ’ Core public sector reforms address administrative and civil service reform, decentralization and intergovernmental fiscal relations, public expenditure and financial management, revenue policy and administration, legal and judicial reform, public enterprise management, and the development of accountability institutions. 11 3. Implementationof the 1997 Anti-Corruption Strategy 3.1 This chapter examines the implementation of the four dimensions or pillars of the anti-corruption strategy. The discussion of the first pillar i s drawn from previous evaluations such as those conducted by the Audit Committee and the United States General Accounting Office that have addressed fiduciary concerns extensively. The current study aims to “fill the gap” by focusing on the second and third pillars, i.e., the implementation of the strategy at the country level. The anti-corruption strategy document did not provide a monitoring and evaluation framework or describe indicators for evaluating i t s implementation. This i s to be expected given the evolving nature of the Bank’s approach to anti-corruption. Therefore, a system of ratings was adapted from the OED’s evaluation methodology for purposes of evaluating how well corruption concerns have been integrated into the Bank’s work. The extent of implementation at the country ZeveZ was determined by looking at ESW and other advisory services and by analyzing the extent of integration of governance/corruption considerations in the design of Country Assistance Strategies (CASs) and the lending program.’ For example, the existence of one or more pieces of ESW was sufficient for the anti-corruption strategy to be implemented in a modest corruption country. In the case of Bank support to countries where corruption was rated “substantial” or “high,” an adequate implementation o f the anti-corruption strategy at the country level would require that at least the CAS reflect corruption concerns (see annex 1 for details of the methodology). The component ratings are described more carefully in the subsequent sections on CASs and lending (also see annexes 2 and 3). 3.2 There were 82 countries for which sufficient data and ratings were available. The desk review found that there has been good progress in incorporating corruption concerns in Bank assistance over the past few years. In more than two-thirds of the countries, the Bank has adequately implemented the anti-corruption strategy. Figure 3.1 presents the Figure 3.1: Implementation of the Anti-corruption Strategy by CorruptionIncidence Class, FY98-01 100% 80% 60% 40% 20% 0% 1/ High Substantial Modest LOW Level of Corruption (CPIA) Source: World Bank staff calculations - see annex 1 for details. The Anti-corruption Strategy Paper (World Bank 1997a, p.61) argued for explicit treatment o f corruption in all Bank activities. However, explicit treatment in each and every product i s not practical, nor i t was considered necessary as long as the intended approach to curtail corruption i s made explicit within the framework of the most appropriate product or activity. 12 distribution of countries with adequate implementation according to the corruption categories (as determined by the 2000 CPIA).9 As can be expected, the figure shows that the level of implementation i s higher in countries with lower levels of corruption which i s a reflection o f the ownership of these efforts by the corresponding clients. Half of all the lending (weighted by disbursement) has ratings of high or substantial in implementing the anti-corruption strategy, with another third rated as modest (see figure 3.2). Eighty-five Figure 3.2: Extent of Implementationin All Lending Weighted by Disbursements 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% U U 0% High Substantial Modest Negligible Extent of Implementation Source: World Bank staff calculations - see annex 1 for details. Figure 3.3: Extent of Implementationin All Lending by Level of Corruption 100% Implementation % 90% - a Q) 80% IA Substantial E 70% .o, 60% g ! j 50% m0 High gjg 5 % 40% -5 5 30% 5 20% 8 10% no/, High Substantial Modest Level of Corruption (CPIA) Source: World Bank staff calculations - see annex 1 for details. We also looked at a three-year average of the CPIA, but i t did not lead to any significant change in the results. 13 percent of lending received by substantial corruption countries was rated as high or substantial in the implementation of the anti-corruption strategy. Even in countries with high corruption the Bank has made some progress in integrating governance considerations in i t s lending (see figure 3.3). Fiduciary risk assessments that help in the proper selection of lending instruments and conditions have been completed (or will be shortly) for the Bank’s ten largest borrowers. Still, further attention i s needed to strengthen country focus, as also called for by Reforming Public Institutions and Strengthening Govemance: A World Bank Strategy, for which more progress on Institutional and Governance Reviews (IGRs) would be a welcome addition (see table 3.1). Table 3.1: Assessing the Risk of Corruption Among the World Bank’s 10 Largest Borrowers I IBRD and IDA Cumulative Existence of Recent Anti-Corruption Instruments I Lending, (Produced between 1996 and 2001) as o f June 30,2001 (in billions of current US$) CAS CPAR CFAAKPFA PER IGR India 56.3 200 1 CPFA J .._ 1997 (progress report China 35.5 1998) I/ CPFA I/ 1999 (progress report MCxico 33.2 200 1) J 2000 (progress report Brazil 30.4 200 1) J Indonesia 28.7 1997 and 2001 I/ CFAA J Argentina 18.2 2000 I/ CFAA J Turkey 16.7 2000 I/ CFAA I/ Rep. of Korea 15.8 None CFAA 1995 (progress reports Pakistan 125 1997,1998,2001) J CPFA J - 1999 (progress report Russia 12.2 200 1) I/ I/ Source: World Bank, various databases. Preventing Fraud and Corruption in Bank Projects 3.3 Bank policy calls for “zero tolerance” for corruption in Bank-financed projects (see box 3.1 for implications of this policy). Actions to prevent fraud and corruption in projects have included the establishment of the Department of Institutional Integrity. The Department has a staff of 40 professionals and i s charged with the responsibility of investigating allegations of fraud and corruption and of staff misconduct. The Department has established a protected hotline and a post office box to receive reports o f fraud and corruption and i s working to raise staff awareness of the signs of fraud and corruption through training and outreach. Another measure undertaken in the effort to prevent fraud and corruption in projects has been the establishment of the Office of Ethics and Business 14 Box 3.1: Implications of a Policy o f Zero Tolerance for Corruption But High Threshold of Proof: Lessons from Project Experiences in Uganda and Bangladesh The Bank has a policy of zero tolerance for corruption in i t s projects. I n practice this policy raises difficult dilemmas for Bank staff when in their supervision missions they observe cost overruns but poor quality o f construction and sometimes even hear undocumented stories about corrupt practices. The problem i s that the threshold o f proof requirements in a law based investigative trigger are very high and the burden of proof to initiate this process are on the Bank rather than the project managers. This makes i t very difficult and costly to initiate any remedial action as demonstrated by experiences from two projects, one in Uganda and the other in Bangladesh. Fortunately a results based approach to project assistance has the potential to break this impasse. The Northern Uganda Reconstruction Project (Cr. 2362-UG) (FY92-99) had some success in helping to rebuild remote northern areas in the country that had been ravaged by civil war through support to agriculture, education and essential infrastructure. The physical separation of project financial management from the location of the project provided opportunities for mismanagemento f funds. This was because the financial management staff in the capital knew how the funds were being spent but not much about what was being executed on the ground. I n the North where project implementation was taking place, those involved in the project knew what was being built, but were unaware o f the costs being incurred. Regular project supervision found cost overruns of 50 percent, although i t i s uncertain whether these were due to corruption, inefficiency, or a combination of the two. The Bank supervision missions discovered irregularities that led the Bank to call for further audits which ultimately provided sufficient evidence to enable the Bank to request an investigation by the Government of Uganda. Ultimately a project manager was indicted by the Ugandan judicial system. The goals o f the General Education Project in Bangladesh (Cr. 2118-BD) (FY90-96) were to increase access and quality o f education. The project succeeded in constructing many schools and teacher training institutes. Problems did arise, though. Project supervision reports indicated early on that some o f the construction being performed was sub-standard and that there were cost overruns both for construction of schools and teacher training institutes. While these shortcomings were observed by Bank staff, further remedial actions could not be taken due to a lack o f sufficient evidence. The dilemmas raised by these projects could have been overcome if the Bank had followed a results-oriented approach to i t s project assistance. A results-oriented approach links financing to the delivery of outputs with specified standards and quality. The incentives implicit in this approach encourage competition and innovation for quality enhancement and cost efficiency. Cost overruns are simply excluded by design and independent monitoring by stakeholders i s encouraged due to transparency and contestability. The supervision focus would also shift from input controls to output (service delivery) performance. Sources: OED Staff. Conduct, which focuses on values based integrity training and provides counsel and resources to uphold the World Bank Group's ethical standards. Other important World Bank initiatives to prevent fraud and corruption in Bank-financed projects include: 0 An amendment of staff rule 8.01 to provide for mandatory termination for staff who defraud the Bank; 0 Intensive audits of selected projects; 0 Strengthening Bank oversight of client procurement processes through commissioning of a systematic post review o f project procurement; and 0 Creation of a Corporate Committee on Fraud and Corruption Policy. 3.4 A 1998 Operational Memorandum established a Sanctions Committee to decide on appropriate action on the sanctioning o f f i r m s having engaged in fraud or corruption in relation to Bank-financed contracts. In August 2001 the Bank adopted revised procedures for the Bank's Sanctions Committee. The Sanctions Committee i s charged 15 with recommending to the President appropriate actions upon a finding of reasonably sufficient evidence that f i r m s and contractors subject to the committee’s jurisdiction have committed corrupt or fraudulent acts. The Department of Institutional Integrity has been given the responsibility to prepare Notices of Debarment outlining the evidence obtained and represent the Bank before the Sanctions Committee. 3.5 As o f September 2002’77 f i r m s or individuals had been sanctioned, and two f i r m s had been given reprimands. Currently more than 300 complaints on fraud and corruption are under investigation by the Department of Institutional Integrity. 3.6 Beyond punitive actions, the Bank has also made efforts to improve internal controls related to financial management, audit, and procurement. Improvements in financial control capacity are based upon a set of financial management standards for financial institutions referred to as COSO (an acronym which refers to the Committee of Sponsoring Organizations of the Treadway Commission). COSO establishes common management controls for all Bank business units and a standard to assess improvements in these controls. Adoption of COSO by the Bank has been instrumental in increasing the level of awareness and commitment to controls throughout the Bank. 3.7 The Bank has also sought to increase financial transparency in loan administration and strengthen borrowers’ financial management capacity under a Loan Administration Change Initiative (LACI). While the objective of COSO i s to reform the Bank’s management of internal business, L A C I targets the improvement of loan administration through reform of related institutions in the Bank and Borrower countries. Initially, the Bank introduced the Project Management Report (PMR) which included a project financial statement, a project progress report and a procurement management report. The PMR was intended to assist borrowers in managing projects and also serve as a project monitoring tool for the Bank against which disbursements would normally be approved. I t aimed to ensure transparency in transactions and constant flow o f information. 3.8 COSO and L A C I exemplify the Bank’s emphasis on reforming systems and institutions within the Bank and those in the borrowing countries for greater transparency and accountability even though they are not initiatives under an explicit anti-corruption program. Many of the reforms under COSO and L A C I target existing institutions and will take a long period to bring about results. A synthesis o f background work conducted for this study and Bank self evaluation yields the following conclusions: COSO and L A C I initiatives are aligned with the anti-corruption strategy. The effectiveness of L A C I depends on strengthening implementing institutions in borrowing countries (where the Bank has less control over implementation). The control environment within the Bank i s expected to improve further. The Project Financial Management System and the new team o f financial management specialists have provided the Bank with the opportunity to exercise tighter control over projects and to achieve higher fiduciary standards.” loThe GAO (April 2000 report, p. 5) pointed out that: ‘The Bank has made significant progress in meeting each of the five standards of an effective management control system. However, challenges remain and further action will be required before the Bank can provide reasonable assurance that project funds are spent according to the Bank’s guidelines.” 16 The PMR, however, has had limited adoption by borrowing countries reducing i t s effectiveness in streamlining loan disbursement, assisting the borrowers to manage projects, and helping the Bank to monitor projects. The expected changes, especially in accounting systems and project auditing at the borrower level have not yet materialized, and i t may take a long time to reform accounting systems in Borrower countries. 3.9 This report agrees with the conclusions of the Audit Committee in their review of the progress of LACI. The Audit Committee found that financial management in the Bank had been strengthened by the hiring of specialized staff and that the timeliness and completeness of project financial statements had improved. On the other hand, i t found that some aspects of L A C I were less successful, e.g., few borrowers were opting for PMR based disbursement, and i t questioned the sustainability o f improved financial management. In response to these concerns, the PMR was redesigned and renamed Financial Management Report (FMR). While FMRs maintain the same basic content as PMRs (including procurement, financial information, and physical progress o f implementation), they can be customized to particular projects. Procurement 3.10 Fraud and corruption were recognized in the 1995 revisions to the Bank’s procurement guidelines. In 1997, the Bank embarked on a major procurement reform aimed at: (a) improving the effectiveness of fiduciary oversight; (b) implementing a systematic approach to help clients improve their national procurement systems; and (c) strengthening internal capacity to discharge both tasks. This revision o f the procurement guidelines sought to strengthen fiduciary oversight b y incorporating a pledge by bidders against fraud and corruption in Bank-financed projects, and created a mechanism for bidders to lodge complaints if they felt that the bidding process had been tainted by corruption. The role o f the procurement specialists and the procurement accredited staff in project planning, implementation, and evaluation was strengthened. New procurement staff have been hired, two-third of which are located in country offices. The Bank i s also conducting special procurement audits in high-risk countries. Work on the harmonization o f procurement documents with other multilateral development banks i s also progressing. T o upgrade borrower procurement capacity, the Bank has conducted seminars and workshops on Bank procurement policy and provided technical assistance. 3.11 The Bank has made significant progress in strengthening fiduciary oversight, but upgrading borrower capacity and streamlining of procurement instruments remain unfinished tasks. Upgrading borrower capacity in the area of procurement could take some time to accomplish. Procurement capacity assessments o f project implementing agencies have led to increasing assistance to countries, and more generally the outcome of Country Procurement Assessment Reports (CPARs) has often been to trigger a reform of procurement institutions. Also, the Legal Department i s advising on procurement legislation and has been involved in procurement training. 17 Helping Countries that Request Bank Assistance in Combating Corruption 3.12 The anti-corruption strategy stipulated that the Bank should respond positively to requests from borrowing countries for help in design and implementation of anti-corruption programs. As of June 2000, 108 countries had requested assistance for broader public sector/institutional reform, out of which 60 had specifically requested anti-corruption assistance. Table 3.2 indicates that two-thirds of the requests were from countries classified as having substantial corruption. This amounted to half of the substantial corruption countries. This i s encouraging since, as mentioned above, i t i s in these countries where Bank assistance has the highest potential for impact on corruption (see box 1.1). Table 3.2: Countries Requesting World Bank Assistance for Anti-corruption Activities Total number of Number o f countries that Percent that have countries have requested assistance requested assistance High corruption 23 9 39% Substantial Corruption 78 40 51% Modest Corruption 33 11 33% 'Low Corruption 7 0 0% Source: World Bank, various sources. 3.13 The Bank offers a large menu of services to analyze the causes and consequences of corruption, to share experiences across countries, and to support the design and implementation of anti-corruption strategies. These include Institutional and Governance Reviews, Public Expenditure Reviews, Country Financial Accountability Assessments, Country Procurement Assessment Reviews, and Country Economic Memoranda. W B I '' i s involved in conducting diagnostic surveys (including surveys o f households, enterprises, and public officials), the development o f governance indicators, and training programs and i s involved in discussions on the design of lending assistance. Other Bank work has delved into understanding state capture, measurement o f the investment climate, and the political economy of governance reform.'* This review found that there are significant regional differences in the degree o f mainstreaming corruption concerns in ESW. In 88 percent of countries in SAR, corruption concerns were addressed in at least one ESW report. While coverage in EAP was only 35 percent, Bank-wide the average was 53 percent. As in many other areas, the focus was on countries with substantial corruption (with a coverage of 74 percent)-where the potential for impact i s high. On the other hand, ESW paid significant attention to corruption only in 27 percent of high corruption countries-where demand i s usually weakest. The integration o f corruption concerns in these products and services i s discussed in the following paragraphs. 3.14 An Institutional and Governance Review (IGR) i s the most comprehensive tool for analysis of corruption offered by the Bank. IGRs assess the status of institutional " Other (more recent) activities include the joint exercise of the Africa Region with the IMF to benchmark the quality o f expenditure accountability systems i n the 22 African HIPC countries in FY02. This exercise has been viewed by the Region as a means to engage countries with high corruption. The EAP Region notes that it i s working to integrate PERs, CPARs, and CFAAs in an effort to make them more effective. The ECA Region notes that the Bank's contributions to the analysis of corruption such as Anti- Corruption in Transition: A Contribution to the Policy Debate (2000) and other ESW has had an important impact on the dialogue on corruption in the Region. 18 reforms and related efforts to strengthen public management and governance. They help sharpen the institution building focus of the CAS and contribute to the Bank’s dialogue with the government and civil society. As of December 1,2001, the Bank had completed IGRs for Armenia, Bangladesh, Bolivia, Burkina Faso, Eastern Caribbean Region, Ethiopia, and Peru.13 3.15 Country Financial Accountability Assessment (CFAA) reports provide useful analysis o f public financial accountability, especially in relation to systems o f internal control and audit. The Bank completed 57 CFAA reports between September 1997 and November 2001. Coverage by region ranged from 17 percent of AFR countries to 11 percent in ECA and MNA. The Bank has already prepared CFAAs covering half of all countries with substantial corruption ratings, but the countries with high corruption ratings have not yet been reviewed. 3.16 Earlier CFAAs focused on monitoring acceptance of international auditing and accounting standards. More recent CFAAs acknowledge that formal acceptance of standards may have limited practical significance if countries are unable to meet those standards due to a weak enabling environment, In view of this, recent CFAAs have begun to focus on the role o f broader institutions of accountability, such as auditors general. The Bank i s committed to the ambitious goal of having current CFAAs that are less than five years old for all major borrowers by FY04. There i s also an effort to integrate the CFAA with the CAS and other ESW.I4 3.17 The Bank uses Public Expenditure Reviews (PERs) and Public Expenditure and Institutional Reviews (PEIR) to identify inefficiencies and inequities in public spending. A PER can be a helpful tool in discovering corrupt practices by tracking expenditures and analyzing service delivery performance. The Bank completed 68 PERsPEIRs during FY98-01. O f these, 25 PERs were reviewed for this study. Among the sampled PERsPEIRs, seven explicitly address corruption. Another ten refer to “rent seeking,” “abuse of office,” and “waste.” All 25 PERs draw some attention to the problem of poor accountability in public institutions. The governance and corruption focus o f PERs varies by region with a high of 38 percent of PERs devoting significant attention to these issues in SAR and a low of 7-9 percent of the PERs doing the same in L A C and AFR. In all regions, more recent PERs, especially PEIRs, have begun to analyze public sector service delivery performance and accountability and responsiveness issues. Most PERs cited lack of transparency as a major issue. Twelve PERs recommended civil service reform as a way of making the public sector lean, efficient and accountable. 3.18 Country Procurement Assessment Reports analyze a country’s public sector procurement structure and the institutional and organizational risk associated with the procurement process. The Bank completed 45 CPARs between September 1997 and November 2001. The Bank’s anti-corruption strategy affected the structure and substance of CPARs. Post- 1997 CPARs devote significant attention to anti-corruption measures adopted by the country in public procurement. They also discuss at length l3The Public Sector Board update Reforming Public Institutions and Strengthening Governance: A World Bank Strategy (2003) also mentions other IGRs which are at different stages o f completion. l4The Africa Region notes that 17 CFAAs have been produced in FYOl and FY02, and that work programs on expenditure accountability (spurred on by PRSP and HIPC) have been put into place in many countries. 19 misprocurement findings and possible remedies (see table 3.3). Three common themes contributing to corruption are: (a) inadequate legal and administrative frameworks for public procurement processes; (b) non-standardization of bidding documents and processes; and (c) lack o f procurement skills. Table 3.3: The Effect of Anti-Corruption Initiatives on CPARs Discussion of Discussion of Discussion of Anti- Discussion of Discussion of Bidding Procurement Suggestions Corruption Mis- Procurement Documents, Staff in to Reduce Period Measures procurement Laws Procedures Country Corruption Pre-1997 Negligible Modest Substantial Substantial Modest Modest Post-1997 Substantial Substantial Substantial Substantial Substantial Substantial Source: Background work carried out for this report. 3.19 There i s substantial variation in regional coverage with E C A having completed CPARs for half of i t s clients, SAR for 38 percent of clients and all other regions having about a quarter of their clients’ procurement systems reviewed. Grouping countries by level of corruption, CPARs have been carried out for 20 percent of high corruption countries and 70 percent of countries with substantial corruption. 3.20 An analysis of 40 Country Economic Memoranda (CEMs) completed during FY98-01 shows that CEMs are increasingly being used to advance the Bank’s governance reform agenda. While only 13 of the CEMs made explicit reference to the existence of corruption, and seven of the CEMs explicitly mentioned the Bank’s anti-corruption strategy, all of the CEMs focused on policies with potential anti-corruption effects. Major areas of attention for the CEMs during the past four years have been: the regulatory environment, civil service reforms, economic liberalization, financial sector reforms, decentralization, judicial reforms, and tax administration. 3.21 WBI’s anti-corruption approach has undergone a transition during the last four years.15 Until FY98 i t focused on raising awareness of corruption through seminars and workshops. Starting in 1999, a new program of action-oriented learning was piloted in seven African countries-Benin, Ethiopia, Ghana, Kenya, Malawi, Tanzania, and Uganda-and recently extended to Latin America. This program embodies an integrated approach to controlling corruption and improving governance. Under this approach, the WBI works with countries that explicitly request i t s help in combating corruption. To ensure that there i s a strong commitment from the national leadership to a transparent, open and participatory reform process, the W B I requires that a written request for its support come from the Head of the State or Government and that i t should indicate a commitment to address corruption as part of an integrated governance improvement program. Following the request for assistance, a steering committee i s formed of government officials, civil society activists, and private sector representatives who participate in integrity workshops (anti-corruption core courses). Client acceptance o f a participatory steering committee triggers the formulation of a program which typically includes data collection, political leadership training, capacity building of accountability institutions, and l5 The following discussion i s based upon a content analysis of W B I documents, as well as semi-structured interviews with W B I task managers. 20 promotion o f collective action. Such activities spur debate on corruption, raise awareness, and actively involve many political actors. The intended product of this process i s an anti-corruption action plan that has broad civil society support. WBI’s governance group has done a significant amount of empirical research and developed various governance indicators (called GAC or Governance and Anti-Comption tools) based on surveys of households, businesses, and public officials and (in collaboration with DEC and the Regions) have made these tools accessible through a “governance databank” available on the Bank’s website. The WBI considers these governance indicators to be empowering tools for civil society and aid in the formation of country-specific anti-corruption strategies. Service delivery surveys may also serve as an entry point for the Bank to address corruption. The WBI has also emphasized that a major thrust of its new strategy i s the emphasis on voice, external accountability mechanisms, transparency tools, and participatory consensus-building strategies involving key stakeholders, rather than the former approach of focusing on internal rules and laws. Mainstreaminga Concern for Corruption in Bank’s Work 3.22 This section reviews the extent of mainstreaming anti-corruption concerns into country assistance strategies and lending programs (see box 3.2). l6 This evaluation considers that corruption has been addressed when the Bank has undertaken activities Box 3.2: Mainstreaming a Definition of “Mainstreaming” Since one o f the four pillars of the Bank’s anti-corruption strategy i s “mainstreaming a concern for corruption in the Bank’s work,” a clear understanding o f what mainstreaming entails i s critical to the present report’s evaluation o f the Bank’s efforts. The mainstream i s “the principal course of activity” or “the major current of opinion.” Mainstreaming, on the other hand, indicates “a perturbation in the natural order of things,” and therefore implies the subversion o f the old order in order to initiate change. In the mainstreaming process, vested interests will be challenged, and support for a new order must be solidified. At the same time, traditions may be broken and stakeholder loyalty weakened. In this light, mainstreaming a concern for anti-corruption should be seen as a difficult process that will take a lot of effort and a long time to fully implement. Source: Picciotto (2001). such as strengthening the rule of law, enhancing public participation in public-sector decision making, and moving public-sector decision making closer to citizens, in addition to those cases when there i s an explicit discussion o f corruption considerations. Integration of corruption concerns in CAS documents i s rated b y taking into account the incidence of corruption in the country and proposed Bank responses in dealing with it. For example, for a highly corrupt country, integration o f the anti-corruption strategy into the CAS would be rated as high if corruption i s a major theme and the proposed lending program has strong anti-corruption elements (on one end o f the scale), and negligible if the CAS has little or no mention o f corruption in either the discussion o f the country’s main development issues nor i s the proposed lending progradadvice focused on corruption. Implementation of the anti-corruption strategy in lending i s rated by taking l6 See Picciotto (2001) for a description o f obstacles to mainstreaming. 21 into account the incidence of corruption in the country and whether or not that calls for design features to curtail incentives and opportunities for malfeasance. For example, for a project in a high corruption environment, the level of implementation would be considered high when the project has a major component geared toward reducing corruption and improving governance (on one end of the scale), and negligible when the project has almost no component geared toward reducing corruption. For both CASs and projects, ratings of “substantial” or “modest” are given to the intermediate cases (also see annex 2). Mainstreaming a Concernfor Corruption in the CAS 3.23 The integration of corruption concerns in the CAS i s critical for mainstreaming corruption concerns in the Bank’s work. The Bank’s anti-corruption strategy calls for the CASs to examine governance failures relating to corruption, and since January 1999, this has been a requirement. About 75 percent of the 113 CASs (CASs, Interim Support Strategies, Transitional Support Strategies, CAS Progress Reports) prepared from FY98 to FYOl recognized corruption concerns explicitly or implicitly. Figure 3.4 shows the distribution of CASs according to the extent of implementation of the anti-corruption strategy by FY2001. The CASs for Bangladesh, Benin, Cambodia, and Ukraine provided excellent discussions of the causes of corruption. Figure 3.5 shows that integration of corruption concerns has been greater in CASs of borrowers with modest or substantial levels of corruption. In the future, CASs of high corruption countries may need to pay greater attention to causes of corruption and possible remedies. Mainstreaming a Concernfor Corruption in Lending 3.24 To assess the integration of conuption concerns in lending, a stratified sample of 93 projects was analyzed among 1,093 projects listed in the Imagebank between September 1997 and June 30,2001. This sample includes 49 o f the 100 core public sector reform projects identified by the Public Sector Board and includes all public sector Figure 3.4: Extent of Implementationin CASs, FY98-01 1O O % / A 9O%Y 80% I 70% El Modest Implementation 60% E4 Substantial 50% UUU High 40% 30% 20% 10% O%Y FY 1998 FY 1999 FY 2000 FY 2001 Source: World Bank staff calculations - see annex 1 for details. 22 Figure 3.5: Extent of Implementation in CASs by Level of Corruption 100% 90% 80% 70% 60% 50% 40% W Substantial 30% 20% 10% 0% High Substantial Modest Low Level of Corruption (CPIA) Source: World Bank staff calculations - see annex 1 for details. projects identified as “best practices” by the Public Sector Board. In addition all judicial reform projects approved between 1999 and 2001 were inc1~ded.I~ Projects were rated b y the extent by which they implementedthe anti-corruption strategy. An explanation of the rating criteria used to analyze this sample i s detailed in annex 2. 3.25 Public Sector Management projects are intended to address the budget process, fiscal policy formulation, public expenditure management, and problems in the provision of public services. Generally, these projects support institution building, administrative strengthening, and civil service reforms. Other projects promote privatization, regulatory reform, and ways to support private sector development which are strong anti-corruption interventions because they help limit opportunities for inappropriate use o f discretion by public officials. Two projects (Nicaragua Economic Management Technical Assistance and Bolivia Institutional Reform) try to promote honesty in the public service through support to National Integrity Committees. Other projects fund tools for better management. These include computers for better tracking of expenditures, taxes, and customs; surveys of taxpayers to build capacity for tax administration, technical assistance and training for civil service reform, and developing models for better budget forecasting. The Uttar Pradesh Fiscal Reform and Public Sector Restructuring project attacks corruption ” The projects are divided into four categories: Public Sector Management (PSM), Judicial Reform, Structural Adjustment Lending, and Sectoral Investment Lending. The P S M category includes projects that were classified as P S M in project documents and those classified as “core P S M ’ b y the Public Sector Board. Adjustment loans that have been classified as P S M are included in both the P S M and the S A L sections. Judicial reform projects are also PSM, but have been placed in their own category because o f their special nature. Except for judicial reform, the initial sample o f projects were selected using stratified sampling so that each region would be represented. Responding to concerns o f the Public Sector Board, the sample was modified to include the projects they have identified as “best practice.” W e did not focus on C D D projects for several reasons. First, OED i s doing a formal evaluation o f C D D projects due to be delivered in FY05. Also, these projects did not make the l i s t o f core projects concerned with governance reform by the Public Sector Board. 23 on multiple fronts by addressing tax and expenditure policy, civil service reform, decentralization, state owned enterprise (SOE) reform, and an anti-corruption strategy. Three-quarters of public sector management projects (weighted by disbursement) have ratings o f modest or higher in implementing the anti-corruption strategy. 3.26 Judicial Reform: In the past few years, the number of projects focused on legal and judicial reform has increased. A weak judicial system may be both a cause and a result o f corruption. Therefore, judicial reform can be a tool to reduce corruption both in the judiciary and in society as a whole. The Bank’s judicial reform projects intend to reform judicial systems through improvements in the administration of civil justice to improve the quality, efficiency, integrity, and access to the judiciary. Reforms have addressed criminal, labor, and commercial justice as well. Common project components include improvements in legal educationhraining, case-management systems (including computerization), and access to legal information especially for judges. Several projects promoted alternative dispute resolution (ADR) as a way to expedite the handling of cases and also attempt to address issues of independence in the judiciary. The Legal Department i s attempting to enact a holistic approach to fighting corruption that includes providing judicial training, access to materials (such as legal databases), and new computerized tools for case management. All of the judicial reforms that have been put in place so far received implementation ratings of modest or higher. 3.27 Structural Adjustment Lending cover projects from pension reform (Moldova), to public sector management reform (Thailand), to surveys and diagnoses (Cambodia), to administrative reforms for the proposed elimination of “ghost workers” (Central African Republic). Structural adjustment loans (SALs) promote numerous public sector and economic reforms, including tax reform, decentralization, civil service reform, regulatory and legal reform, privatization, tariff reduction, adoption o f anti-corruption strategies, and budget reform. In the last few years, Programmatic Structural Adjustment Loans (which consist of multiple adjustment loans) have been utilized to implement broader institutional reform programs. Bank guidelines suggest the risks associated with weak governance and corruption should be taken into account in country risk analysis, decisions on lending and disbursements, portfolio supervision, and assessing the ability of borrower to carry out the programs. These considerations guide the decisions on adjustment lending and i t s conditionality on a case-by-case basis. This study found strong results for adjustment lending, 85 percent of SALs (weighted by disbursement) received ratings of modest or higher in implementing the anti-corruption strategy. 3.28 Sectoral Investment Lending: Corruption can affect all sectors o f an economy, impairing service delivery and provision of infrastructure. Therefore i t i s important that these types of projects are examined, even though they may not be considered formally to be public sector reform projects. Eighteen projects were reviewed from various sectors including transportation, finance, social protection, education, water supply, health, and agriculture. Six percent of these projects (weighted by disbursement) paid careful attention to anti-corruption concerns. The Indonesia Decentralized Agriculture and Forestry Extension Project was notable for the decentralization of implementation, ensuring the participation of stakeholders. 24 Providing Active Support to International Efforts to Address Corruption 3.29 Bank assistance for international efforts to curtail corruption has come mainly in the form of sponsorship of major conferences, support for Transparency International (TI), and for adoption of the 1999 OECD Convention on Combating Bribery of Foreign Public Officials in International Business Transactions. The World Bank i s also collaborating with the IMF in developing a joint approach to enhance the global effort against money laundering. Other partnerships that the Bank has developed include those with OECD D A C (Development Assistance Committee) and PUMA (the Public Management Program of OECD), GOVNET (a network of donors supporting analytic work), the Anti-Corruption Network, SIGMA (now called T I Online Directory for anti-corruption programs in Central and Eastern Europe), UNDP, and many bilateral donors including DfID (Department for International Development) and the governments of the Netherlands, France, and Denmark. The Bank has also participated in the Harmful Tax Practices Initiative. Supportfor Transparency International (TI) 3.30 Since 1996, Transparency International (TI) has been a partner with the World Bank in the fight against corruption.” T I was represented on the Anti-corruption Strategy Committee after participating in a seminar with Mr. Wolfensohn in June 1996. T I also had a role in promoting the OECD Convention on Combating Bribery of Foreign Public Officials in International Business Transactions. 3.3 1 TI i s conducting a significant part of i t s work in partnership with the development assistance community. For example, it has received assistance from United States Agency for International Development (USAID) for i t s anti-corruption work in nine of the countries in which T I operates. The World Bank was a partner to the IOth International Anti-Corruption Conference held in Prague in October 2001, which was organized by TI. Since 1998, the EAP Region has worked directly with T I and other international advisors from developing and developed countries to assess corruption risks Transparency International (TI) was founded by a former Bank staff member in 1993 as a non-profit organization with the goal o f bringing worldwide attention to the prevalence of corruption and the harm that it causes. The founding board members came from different backgrounds including government, the World Bank, and the private sector. Today, T I i s functioning in at least 77 countries, and many o f these have national chapters. National chapters conduct activities such as: publication o f newsletters, journalist training, and opinion pollings. The World Bank i s cited as a contributor on TI’Swebsite. One o f TI’S most prominent contributions to the fight against corruption i s the Corruption Perceptions Index. While the CPI has had a major impact in raising awareness about corruption since annual publication of the index began in 1995, i t i s also controversial. The CPI i s a complex standardized average of corruption indicators drawn from several sources, oftentimes reflecting foreign and expatriate opinions. There i s concern that these indices are used inappropriately for empirical research, given that even T I recognizes that comparisons between different years CPIs should not be made. (For details, see Thompson and Shah, forthcoming. “Corruption Perceptions Index: Whose Perceptions Are They Anyway?”) T I has expanded i t s efforts in recent years to the “supply side” of bribery through the Bribe Payers Index and the Integrity Pact. The Bribe Payers Index (BPI) was calculated from the results o f a survey of businessmen in 14 emerging markets about the propensity of leading exporting countries to bribe. So far, two editions of the BPI have been published: in 1999 and 2002. The Integrity Pact i s an attempt to reduce corruption in public procurement projects by having the government promise not to accept bribes and bidding companies not to offer them. According to TI, some Bank staff are skeptical about the impact o f such pledges on curtailment o f corruption because o f a lack o f enforcement mechanisms i n case o f a breach. 25 and evaluate anti-corruption programs across borrowing countries. The Region found that the private, informal meetings of the EAP’s Anti-Corruption Advisory Group (ACAG) was a useful forum for looking at cross-country experiences and obtaining practical advice on governance reforms. Anti-Bribery Conventions 3.32 In the recent past, bribing foreign officials was considered a cost of doing business in the developing world and condoned through tax deductibility of bribes in some OECD countries. The U.S. Corrupt Foreign Practices Act, 1977, represented a major turning point as i t criminalized such activities. The OECD Convention on Combating Bribery of Foreign Officials has been under discussion for several years including active participation by Bank staff in the Working Group entrusted with the formulation of the Convention. The Convention was finally agreed upon by OECD members in December 1997 (according to TI, in part due to the personal support of World Bank President Wolfensohn). Subsequently the Bank has participated in the dissemination of this convention. The Bank also consults with the Organization of American States on implementation issues arising from the Inter-American Convention Against Corruption (adopted in Caracas, Venezuela in 1996) and i s involved with anti-corruption initiatives undertaken by the Global Coalition for Africa. Anti-Money Laundering and Combating the Financing of Terrorism 3.33 Money laundering i s defined by the Financial Action Task Force (FATF-a task force established by the G7 in 1989 to develop a coordinated response to money laundering) as “processing of , , , criminal proceeds to disguise their illegal origin” (see FATF website at www 1.oecd.org/fatf/MLa~ndering-en.htm).~~ The IMF estimates that money laundering may equal 5 percent of world gross domestic product (GDP), or about $1.5 trillion using 2001 figures. On April 13,2001, the Executive Directors of the IMF and the World Bank endorsed the view that money laundering was a matter of global concern and directed their institutions to develop a joint approach to support international action. They noted that dealing with money laundering “will require a cooperative approach involving many different institutions given the cross-cutting agenda- encompassing financial sector supervision and regulation, good governance, judicial and legal reform, and effective law enforcement.” 3.34 The World Bank Executive Board noted that although the Bank can support efforts against money laundering, i t must remain within i t s mandate and thus avoid connections with law enforcement activities. The FATF 40+8 Recommendations have One of the first tasks of this body, which comprises twenty-six member countries and jurisdictions, two international organizations and three observers, was to spell out measures that national governments should take to combat money laundering. These became known as the “FATF-40” recommendations. They cover the criminal justice system and law enforcement, as well as the financial system and its regulation. In addition, international organizations such as the European Union and the Organization o f American States have established anti-money laundering standards for their member countries, while the Caribbean, Asian and Eastern European countries have created regional “FATF-type groupings.” I n October 2001, eight special recommendations related to the fight against the financing o f terrorism were incorporated, making them the “FATF-40+8” recommendations. 26 been conditionally approved by the Boards of the Fund and the Bank for use in their operational work. The Bank’s Board has directed i t s staff to collaborate with international and regional groups concerned with money laundering, recognize money laundering issues in diagnostic work for clients, provide technical assistance, contribute to a better understanding of the development costs and impact of money laundering, and provide support for international action. 3.35 Bank staff have responded to these directives b y developing methodologies for assessing the implementation of anti-money laundering requirements for banks, insurance companies, and other capital markets. This agenda has recently been expanded to include technical assistance related to anti-terrorism financing. These methodologies w i l l be field tested, as for example in the production of financial sector assessment program (FSAP) reports. A progress report for presentation to the Fund’s Board was presented in September 2002 to the Development Committee, and the A M U C F T (Anti-Money Laundering and Combating Financing of Terrorism) assessment methodology was approved by the Executive Board in November 2002. A twelve-month pilot project testing the methodology was also approved at that time. 27 4. Assessing Relevance of Bank Support for Anti- Corruption 4.1 This section assesses the relevance o f CASs, lending, and training in pursuing the Bank’s anti-corruption goals. This analysis utilizes the same sample o f projects and CAS documents as used to measure the level o f implementation o f the anti-comption strategy in chapter 3 (see para 3.23-3.24). A relevance criterion i s first defined and then applied to WBI programs, CASs, and lending projects. How i s Relevance Defined and Assessed? 4.2 OED defines relevance as “the extent to which the project’s objectives are consistent with the country’s current development priorities and with current Bank country and sectoral assistance strategies and corporate goals.” In particular, anti-corruption programs are judged to be relevant if they have the potential to achieve their objectives given a country’s existing institutional and policy environment, and the Bank’s corporate goals. This evaluation i s guided b y the New Institutional Economics (NE) approach (see annex S), which considers corruption as a manifestation of institutional failures and i t s remedy i s sought b y influencing values, culture, and organizations through changes in incentives and rules. Country specific knowledge and proper sequencing o f reforms are therefore critical to reduce corruption.” The relevance o f these anti-corruption interventions i s assessed in terms o f the quality o f the diagnosis o f the major causes o f corruption in each specific context, and of the matching o f interventions to those causes (see box 4.1). Relevant anti-corruption measures are likely Box 4.1: Example of a Logical Chain for an Anti-Corruption Intervention Consider a Bank-supportedjournalist training program (one of the first types of anti- corruption activity to have been supported by the Bank). The logic behind i t might go as follows: 1) An important cause of corruption in Country X i s inadequate flow o f information to the public on corruption in public administration. 2) News media in Country X are potentially an important conduit o f such information, but lack the capacity to fulfill this role. 3) Journalist training supported by the Bank would raise the capacity of the news media. 4) After training, journalists would do a better job of reporting on public-sector corruption. 5) Better reporting would generate public pressure on the government for honest public administration. 6) The government would respond to pressure by undertaking to reduce corrupt practices. 7 ) Government action would lead to reduced corruption. If any one o f assumptions ‘1’ through ‘6’ appears to be significantly flawed, then i t becomes unlikely that the ultimate assumption, ‘7’, will hold. For example, assumption ‘3’ will break down if there are problems with the quality or the management o f the Bank-supported training program. Assumption ‘4’ will break down if the owners and managers of news media organizations i n Country X refuse to allow their journalists to report on corruption, or if journalists were unable to gain access to reliable information on public-sector corruption. Assumption ‘5’ will break down if there i s a weak culture of civic action in the country, etc. A logical chain may help Bank staff “unpack” their assumptions and assess, at the early stage of project development, potential weaknesses with technical relevance. Source: Schacter and Shah, 2000. ”See Huther and Shah (2000). 28 to reduce the transactions costs in the economy, thus reducing the cost of doing business and the cost of public service delivery. 4.3 CAS relevance i s rated as high if i t has a strong diagnosis of the causes of corruption and the advice and planned lending activities are appropriate for the country’s governance level. The document i s rated negligible if there i s neither a diagnosis of corruption nor i s the proposed lending progradadvice appropriate for dealing with that country’s level of corruption. Ratings of “substantial” or “modest” are given to the intermediate cases (see annex 3). 4.4 Relevance of lending i s rated as high when the anti-corruption objectives play a key role in the country’s development priorities and the Bank’s current CAS, and are consistent with the Bank’s sectoral strategy and corporate goals and policies. Such a project would have great potential to achieve i t s anti-corruption objectives due to the project correctly targeting the country’s specific causes of corruption. At the other extreme, a project would be rated as having negligible relevance when the project’s anti-corruption objectives are absent, or are mostly inconsistent with the country’s development priorities, or the Bank’s sectoral strategies, or the Bank’s current corporate goals. Similarly, a project with reasonable potential to address causes o f corruption i s rated as having substantial relevance and a project with some potential to address the same i s rated as having modest relevance (also see annex 3). An important step in rating a project’s relevance i s determining i t s potential to achieve i t s anti-corruption objectives due to the project correctly targeting the specific causes of corruption in the country. Relevance of CAS 4.5 There i s a sequence of events from analysis of corruption to prescribing the correct measures that must hold for a CAS to be effective in fighting corruption. First the types and causes of corruption in a particular country must be discerned. Secondly measures that are appropriate for fighting corruption in the country should be proposed depending on whether a country has a higher or lower level o f corruption. The assessment of relevance i s based on the type of measures promoted, given a country’s quality of governance and perceptions about local levels of corruption. Overall, about two-thirds of CAS were rated as having modest or higher relevance. This ratio i s high for CASs of countries in the modest and substantial corruption categories, but lower for the high corruption countries (see figure 4. 1).21 4.6 There are CASs, e.g., the Romania CAS, which contained a detailed discussion of the types of corruption present in the country (administrative corruption and corruption in the provision of state services caused by weak enforcement and excessive taxes and regulations) and also provided solid advice and an appropriate loan program. Ukraine CAS identified fundamental causes of institutional weaknesses and poor governance (poor contract enforcement, capture of public institutions by private interests, poor management of public funds etc.) and integrated these concerns into Bank-country dialogue. India’s CAS also had a thoughtful diagnosis and proposed assistance *’ Relevance was not rated for low corruption countries, since corruption i s not considered an important impediment to development in these countries. 29 appropriate to reduce corruption. On the other hand, some CASs provided a detailed analysis of corruption (such as Bangladesh’s CAS) but the necessary actions for reducing corruption were not fully specified. Conversely, other CASs did not provide a detailed analysis o f corruption, but suggested broadly the appropriate measures for the country’s level of conuption. This was the case with the CAS for Turkey, which only mentioned corruption, but proposed a suitable lending program given the country’s circumstances. Finally, a few CASs s t i l l do not appropriately incorporate corruption considerations in their design. Figure 4.1: Countries with Relevant CAS by Level of Corruption 100% 90% Source: World Bank staff calculations - see annex 1 for details. 4.7 Interviews were conducted with 30 Bank staff to ascertain their views regarding the Bank’s anti-corruption strategy and i t s linkage with the CAS work. Most task managers described corruption as being widespread, and identified patronage, political favoritism, and cronyism as being pervasive. But less than one-third of CASs mention patronage or report systemic corruption. This disconnect may be due to the difficulty of overtly discussing public-sector corruption in the dialogue between the Bank and governments. Still, i t seems that staff are more willing than in the past to raise this issue in their discussions with clients and they welcome the public nature of the anti-corruption campaign. Of some concern i s the view reported by some staff that the anti-corruption discourse within the Bank i s having little impact on the ground. The Bank’s strategy, from this view, i s a series of top-down procedures that do not differentiate between countries’ individual conditions. This study concluded that s t i l l many task managers have untapped knowledge of countries’ governance, political, and cultural conditions that would contribute to better program design and choice of the policy instruments. An important challenge i s to bring this knowledge more openly into the CAS’S analyses. 4.8 There i s strong evidence that levels of corruption are correlated with the quality of public services. Econometric analysis conducted for this study (see for example Tomaszewska and Shah, forthcoming) show that corruption in the public sector i s 30 associated with lower quality public service outcomes as measured b y infant mortality rates, child mortality rates, satisfaction from public health care provision, and satisfaction from roads and transportation. While CASs commonly discuss service delivery issues, the linkage with corruption has not received great attention. I t i s possible that a more explicit treatment of this linkage may induce governments to address the corresponding aspects of corruption, and increase the effectiveness of Bank assistance in both, improving service delivery and reducing corruption. Relevance of Lending 4.9 In the following, the relevance of a sample of Bank lending projects with anti-corruption objectives i s analyzed using OED’s standard methodology as described in the first section of this chapter. 4.10 Public Sector Management Reforms have the potential to be highly relevant to anti-corruption efforts through institution building, administrative strengthening, and civil service reforms. More than half of the PSM projects reviewed were given a relevance rating of modest or above of which about 25 percent were rated as being substantially relevant to the anti-corruption strategy. The projects often provide inputs that may be used to build capacity and improve public sector management, but they may not (by themselves) be enough to eliminate the incentives that lead to corruption. This may be the case with some projects which have components to increase automation, such as the Budget Systems Modernization Project in Algeria and Yemen’s Public Sector Management Adjustment Credit. In the case of the Algerian project, automation and training i s intended to enhance the ability of the Ministry o f Finance to plan and present the budget. Yemen’s adjustment credit stipulated as a pre-condition to disbursement that civil service personnel files be computerized so that ghost workers could be eliminated. Broad stakeholder involvement, as an integral part of public-sector reform projects may be needed to ensure sustainable reductions in corruption, e.g., if members o f parliament, the media, and the public are given access to the data. The Guinea Capacity Building for Service Delivery Project likely had an impact on corruption through i t s focus on service delivery. 4.11 Judicial Reform: All the Judicial Reform projects rated thus far have been rated at least modestly relevant for reducing corruption. Most common components were improvements in legal educationhraining, improvements in case management systems (including computerization), and improvements in access to legal information (especially so that judges have access). Several projects also promoted ADR as a way to expedite the handling of cases. Providing judicial training, access to materials and new computerized tools for case management gives courts the ability to better carry out their duties. These improvements in judicial infrastructure w i l l have a significant impact if infrastructure i s a key constraint to judicial reform.22 Promoting ADR would have some impact as i t provides people with an alternative to the formal legal system. Some, but not all projects attempt to strengthen the enforcement o f judicial decisions, e.g., in Albania, 22 I t i s the Legal Department’s view that improving case management can have a major impact on corruption by eliminating grease money and “behind closed door” deals. This i s because computerization and improvements in the handling of case documents means that papers are not lost, cases are handled on schedule, and judges are denied the opportunity to meddle in case administration. 31 Armenia, and Georgia. People are more likely to use the formal judicial system if they feel that a judgment i s more likely to be enforced. 4.12 In countries where there are weak legal systems, people rely more on trust, repeated interactions, and informal means of enforcement. Judicial reform needs to address the incentives o f judges and court employees to engage in corruption. Only once these issues are dealt with and meaningful reform takes place w i l l people begin to trust the judicial system and utilize it. Some of the judicial reform projects that have been put in place, while they are rated as substantial in implementing the Bank’s anti-corruption strategy, are only modest in their relevance because for the most part they focus on inputs (computers, training) for the judicial system and do not address the fundamental causes of corruption in the judicial system. A small cohort of newer projects have begun to address more fundamental issues to establish the rule o f law. 4.13 Structural Adjustment Lending: SALs show a fairly high concern for corruption and three-fourths of SALs (weighted by disbursement) were relevant for fighting corruption, but about half only modestly so. Some of the projects seem promising; Jordan’s Third Economic Reform and Development Loan and Cambodia’s Structural Adjustment Credit are examples of relevant structural adjustment loans. The project in Jordan includes privatization and trade liberalization as well as strengthening the banking sector in preparation for greater foreign investment. Structural adjustment in Cambodia focused on reducing opportunities for corruption in customs by eliminating the inspection of import shipments through the pre-payment of duties. Sustainability of these measures, however, require local ownership and sustained political commitment which are uncertain. The relevance of SALs i s significantly lower for lending to high corruption countries (47 percent relevant) than loans to substantial and modest corruption countries (about 75 percent relevant). In the absence of a proper accountability framework, the increased availability of resources may increase the likelihood that funds might be used for unintended purposes. This suggests that the overall public expenditures and fiduciary framework should be subjected to a far higher level of scrutiny in high corruption countries than in low corruption countries, especially if adjustment lending operations are envisaged. 4.14 Sectoral Investment Lending: Projects in this category, as stated previously, have the potential to be relevant for fighting corruption, since all sectors o f the economy are susceptible to corruption. However, they were rated as having less relevance for fighting corruption than projects in the other three categories. Only about 5 percent o f these projects were rated as having modest or substantial relevance for fighting corruption. The Bangladesh Export Diversification Project seemed promising; there were a few elements of economic liberalization in the project (such as the removal of procedural and infrastructural constraints to export), which may have an impact on corruption. On the other hand, some components of the same project tended to increase the role of the government in the private sector (such as providing “selective assistance” for product and market development), which may actually increase the incidence o f corruption. Another project, the Sri Lanka Central Bank Strengthening, has components o f increasing automation. The automation i s in the form of a real time gross settlement system for payments, and this can have an impact on corruption in that insiders have less opportunity to use delays to profit. 32 4.15 Overall Relevance o f Lending: This evaluation focused on a set of projects that explicitly or implicitly deal with corruption. The result confirms that most o f these projects have some relevance for curtailing corruption (see figure 4.2). The relevance of lending i s significantly lower in high corruption countries (where about half o f lending i s relevant) than in countries classified as having substantial or modest corruption (see figure 4.3). I t i s notable that four-fifths of lending to substantial corruption countries i s at least modestly relevant to the anti-corruption strategy. Figure 4.2: Relevance of All Lending Weighted by Disbursements 100% 90% 80% 70% .- u C 60% C 3 50% Lc O 40% 8 30% 20% 10% 0% High Substantial Modest Negligible Level of Relevance Source: World Bank staff calculations - see annex 1 for details. Figure 4.3: Relevance of Lending by Level of Corruption 90% E l Modest Relevance 80% H Substantial 70% 60% 50% 40% 30% 20% 10% 0% v High Substantial Modest Level of Corruption (CPIA) Source: World Bank staff calculations - see annex 1 for details. 33 4.16 The projects reviewed in this part of the evaluation all tended to focus more on inputs into better public sector management and economic management than on outputs in terms o f improved service delivery and better governance. For instance, as was described above in the section on public sector management reform, the projects in Algeria and Yemen focused on automation (an input) to help improve public sector management. The judicial reform projects are similar in that judicial training, legal materials, and computers for case management are all inputs needed for a quality judicial system. Similarly, several adjustment loans for provinces in Argentina have as requirements the maintenance of certain levels of health and education expenditures. But provision o f inputs does not guarantee better governance. While the Bank’s work has traditionally focused on service delivery, this has not been widely used as a key entry point for the Bank’s governance and anti-corruption work for which there i s some evidence o f effectiveness (Paul 1998). Training/Leaming Activities of the Govemance Unit of the World Bank Institute 4.17 WBI’s anti-corruption program focuses on diagnostic and awareness-raising activities, and capacity building among civil society. The relative emphasis o f each activity varies by country, and the interventions are developed as part of comprehensive national anti-corruption strategies. Diagnostic work i s a critical step to design an anti-corruption strategy and i t has often been the first WBI activity in a country. I t serves to identify the key causes of corruption, to understand the public’s perceptions of corruption and to raise awareness over i t s consequences. The awareness-raising efforts may help build a popular base at the country-level for anti-corruption work and to reduce resistance by the national leadership. Since 1999, WBI gradually shifted i t s orientation from awareness raising toward activities aimed at understanding and addressing the local factors that affect c ~ r r u p t i o n . ~ ~ 4.18 WBI’s focus on building the capacity o f parliament and the news media to hold the government accountable i s targeted to key causes of corruption, namely poor availability of information about public sector operations and low capacity o f citizens and their representatives to demand more effective and honest public administration. The effectiveness of these activities, however, may be limited if these institutions are “captured” by groups that benefit from corrupt practices. T o mitigate this risk, WBI i s careful in selecting partners, and i t notes that sometimes i t has declined official requests to provide media or parliamentary support when i t had concerns about capture. 4.19 The effectiveness of anti-corruption activities depends critically on the extent of country ownership.24 But ownership i s usually weakest in high corruption environments, where these activities are needed the most. This “demand dilemma’’ constrains the relevance of anti-corruption activities supported by the Bank. The diagnostic, 23 A shift also recommended i n Leeuw, F. and C. Klein Haarhuis (2000). “Measuring and Explaining Variance between Anti-Corruption Policies,” paper presented at the European Evaluation Society Conference, Lausanne, October 12-14,2000. 24 Johnson and Wasty (1993) formulated a set of four criteria for discerning borrower ownership for adjustment lending including: locus o f initiative, level o f conviction among key policymakers, expression of political will by top leadership, and efforts toward consensus-building among constituencies. These criteria provides a useful starting point to assess anti-corruption interventions. 34 awareness-raising and capacity building activities of WBI may play an important role in confronting this dilemma, as they foster broad based demand for better governance. (See box 4.2 for WBI’s self-evaluation.) The WBI i s working now to further increase the relevance o f i t s work by raising the level of local participation in i t s work.25 Box 4.2: Results from Self-evaluation of the WBI The W B I from time to time conducts self-evaluation of i t s activities. In this context, an important evaluation of i t s anti-corruption activities was carried out by Professors Frans Leeuw, Ger van Gils, and Cora Kreft on the WBI’s behalf in 1998. A summary of the report’s findings i s given below: O n WBI’s goals and awareness raising: The WBI’s goals to build social capital, civic society, transparency and accountability had merit, but relying on raising awareness as a means to fight corruption was optimistic since there i s not necessarily a link between awareness o f a problem and action to correct it. On media workshops: The media workshops helped improve the quality of newspaper articles dealing with corruption, but the legal environment in Africa was s t i l l a major constraint for a free press. On integrity workshops: The attempts of the integrity workshops to solicit the support of the public and civil society and improving enforcement o f anti-corruption laws were innovations in the two countries’ (Uganda and Tanzania) anti-corruption efforts. The integrity programs coming out of the workshops were strong in that the action plans including “quick wins,” were clear about the need to fight corruption, had concrete action points, and good underlying logic. The report found little indication that the integrity pledges and workshops for members of Parliament that were designed to win the commitment of the political leadership were having an effect. Since the public did not perceive many visible results of the anti-corruption drive thus far, the report expressed concern that the campaigns would lose credibility and support. fl Source: “EDI’s Anti-corruptionInitiative in Uganda and Tanzania” ED1 Evaluation Studies (1998). 25The W B I notes that i t s Governance and Anti-Corruption (GAC) diagnostics follow a locally-driven, participatory approach through which the local stakeholders, with the W B I as facilitators, build action programs. 35 5. Overview of the Case Studies 5.1 This chapter gives a brief overview of the case studies which are presented in full in annex 10, The case studies follow from the desk-based review of the World Bank’s 1997 anti-corruption strategy presented in chapters 2-4 of this report. The desk review assessed Bank implementation of the anti-corruption strategy and evaluated the relevance of the anti-corruption interventions supported by the Bank and covered programs, policies, and analytical advisory services launched during FY98-01. The case studies complement this work by analyzing the Bank’s recent field experiences in carrying out the anti-corruption strategy. We examine the relevance and efficacy of the Bank’s support for anti-corruption activities in Guatemala, Kenya, Latvia, Pakistan, Philippines, and Tanzania. All six countries have high levels o f public-sector corruption. Four have very low human development indicators. Three are classified by the Bank as “low income,” and three are considered “low middle income” (table 5.1). Table 5.1: Basic Indicators GDP per capita Human Development Index Ranking Population (current US $) (out of 174) (000,000) Latvia 2,920 50 2.4 Guatemala 1,680 108 11.4 Philippines 1,040 70 75.6 Pakistan 440 127 138.1 Kenya 350 123 30.1 Tanzania 270 140 33.7 Source: World Bank, various sources. 5.2 Three factors guided our choice of countries. First, we looked for countries, across a range of regions, where the Bank had an intensive level of anti-corruption engagement. Second, we considered advice from Bank managers about data availability and willingness of individuals in various countries to talk about corruption. Third, we sought countries where there was prima facie evidence that Bank-supported anti-corruption efforts had been relatively successful. The case studies covered lending and non-lending services, and focused on activities launched after the formal approval of the anti-corruption strategy. We also looked at selected activities launched as long ago as FY90,26recognizing that Bank involvement in anti-corruption work pre-dated formal adoption of the anti-corruption strategy. Evaluation Questions 5.3 The Bank’s anti-corruption strategy has four elements: e Preventing fraud and corruption in projects and programs financed b y the Bank e Providing assistance to countries that seek the Bank’s help in curbing corruption 26Because corruption i s a symptom of failed public-sector governance and failed accountability, we attributed anti-corruption objectives to actions that targeted public sector governance reforms. Core public sector reforms address the role of government, rule o f law, administrative and civil service reform, decentralization and intergovernmental fiscal relations, public expenditure and financial management, revenue policy and administration, legal and judicial reform, public enterprise management, and the development of accountability institutions. 36 0 Mainstreaming anti-corruption efforts into Bank’s work Contributing to international efforts to fight corruption. 5.4 The desk study in chapters 2-4 covered all four elements o f the anti-corruption strategy. The case studies, with their emphasis on field experience, deal with the second element: Bank performance in responding to requests for assistance. Under that general heading, the case studies address four questions: 0 What was the nature of the corruption problem that the Bank was attempting to help solve? What does corruption look like in the case-study countries? 0 Has the Bank’s anti-corruption support been relevant? 0 Did the Bank’s assistance further i t s goals of curtailing corruption? Has the Bank helped “make a difference?” 0 What can we Zeam from the country cases that would strengthen the Bank’s anti -corruption work? 5.5 This becomes the basis for an evaluation of the Bank’s performance in annex 10, which covers the second and third questions above. I t describes and analyzes the Bank’s anti-corruption assistance and evaluates i t s relevance and efficacy.27 The summary results are presented later in this chapter. Common Factors 5.6 What does corruption look like “on the ground?” What was the problem that anti-corruption interventions supported by the Bank either were attempting or should have been attempting to solve? This section addresses those questions, Despite divergent local circumstances there i s significant commonality across the case study countries in terms o f types and causes of corruption. Instead o f providing individual analyses of each country, we therefore focus on common corruption-related factors. These twelve corruption factors are described below along with examples from the case studies. In addition, annex 10 contains a breakdown for each country displaying which of these factors were addressed in individual ESW products and projects. (i) Public-sector corruption is endemic in all of the six countries. 5.7 Corruption i s pervasive and routine, Both “petty” and “grand” corruption are widely practiceda2*Although statistics on corruption are often questionable, the available data suggest that corruption accounts for a significant proportion of economic activity. In *’ This report focuses on the period up to FYO1. We recognize, however, that the Banks’ implementation of the Anti-Corruption Strategy has continued to evolve since then with the addition o f new tools and scaling up of others that this report has not been able to evaluate. Some o f these new activities (introduced after FYO1) include work with civil society in Guatemala and provincial adjustment lending in Pakistan. 28 “Petty corruption” refers to routine, uncoordinated corrupt transactions. Examples include police demanding bribes in exchange for non-enforcement of traffic violations, clerks demanding “speed money” for processing files, tax collectors “negotiating” for lower tax liability. Individual transactions are small, but petty corruption i s significant i n the aggregate. “Grand corruption” involves theft or misuse of vast amounts of public resources by a relatively limited group of state officials typically associated with the politicaYadministrative Clite. Kickbacks in major public procurement deals, and siphoning o f foreign exchange into private offshore bank accounts are examples. 37 Kenya, “questionable” public expenditures noted by the Controller and Auditor General in 1997 amounted to 7.6 percent of GDP. In Latvia, a World Bank survey found that more than 40 percent of Latvian households and enterprises agreed that “corruption i s a natural part o f our lives and helps solve many problems.’’ In Tanzania, Service Delivery Survey data suggest that bribes paid to officials in the police, courts, tax services, and land offices amounted to 62 percent of official public expenditures in these areas. In the Philippines, the Commission on Audit estimates that $4 billion i s lost annually due to public-sector corruption. Two of four governments that have fallen in Pakistan since 1990 were ousted on charges of corruption. (ii) The notion of a “modem” state that stands above private interests is weak or absent in the case-study countries. The legitimacy and authority of the government as a guardian o f ‘)public interest” is contested. 5.8 There i s at best only modest acceptance in the case study countries o f the idea that the state i s the legitimate guardian of a broadly shared “public interest.” Instead, the state i s seen as furthering private interests of favored stakeholder groups. “Clientelism”-the notion that public office holders serve client groups linked to them by ties o f ethnicity, family, geography, politics or shared economic interests-shapes the public landscape and creates conditions ripe for corruption. In the Philippines, in a not untypical situation, one northern city has members of a single family holding the key positions of mayor, provincial governor and congressman. In Guatemala, the authority of the government i s undermined by the social and economic rift between the minority of European descent that controls the country’s wealth, and the indigenous Amerindian population that has relatively limited access to political representation. In Latvia, there i s a rift between the Latvian-speaking majority that dominates major state institutions and the Russian-speaking minority that dominates the private sector. Corrupt behavior, under these circumstances, becomes a major avenue for addressing problems and meeting demands. 5.9 The legitimacy and authority of the government i s thus contested in the case-study countries. The line between what i s “public” and what i s “private” i s blurred, and so “abuse of public office for private gain”-the Bank’s definition of corruption-occurs routinely. Throughout the six countries, many civil servants operate as private businesspersons within their public offices, using inputs furnished by the state. In Latvia, the Soviet legacy includes the widely accepted notion that “public” services are frequently provided as private favors in return for payment. Anecdotal evidence in Pakistan paints a picture of routine collusion between public servants and politicians in support of particular favored interest groups. In Kenya, officials of state-owned enterprises collude with private banks in schemes o f mutual private enrichment. In many of the countries, private-sector actors conspire to “capture” organs o f the state to further their own private interests. (iii) The “rule of law” is weakly embedded. 5.10 Where rule of law obtains, laws apply equally to all citizens, weak or powerful. Legal, judicial, and law enforcement systems function predictably, are independent from political influence, and protect citizens against abuses of power. Where public-sector 38 corruption i s endemic, i t i s safe to assume that rule of law has broken down. Public-sector corruption thrives where laws apply to some but not to others, and where the law i s used to further private interests rather than protect the public interest. In such environments, abuse of police powers i s often used as a shield for corrupt activities. A common scenario in the case-study countries has law enforcers-police, for example- operating as lawbreakers. A survey in Kenya found that 20 percent of respondents regarded police officers as the most corrupt public servants. Police in Tanzania routinely stop motorists for invented traffic violations and then coerce them into paying “fines.” Systemic corruption pervades the public attorney’s office in Guatemala. 5.1 1 The capability and independence of the judiciary-a pillar of the rule o f law-is at issue in the case-study countries. In Pakistan, decades of military rule undermined judicial independence, leaving judges subservient to the government. In Latvia, the judiciary bears the scars of the Soviet era when party stalwarts ran the system. The Ministry of Justice suffers from low budgets and low prestige. In the Riga district, there i s a two-year backlog of criminal cases. Similar delays plague the system in the Philippines. (iv) Public-sector corruption is deeply rooted in political factors and social fragmentation that are partially or entirely beyond the reach o f technically oriented interventions. 5.12 The endemic corruption observed in the case study countries i s symptomatic of systemic failure by the state to uphold public over private interests. Systemic state failure i s deeply rooted in political factors and social fragmentation. Factors observed in the case study countries included a high degree of social and economic fragmentation (Latvia, Guatemala), the legacy of artificial systems of statehood and governance imposed by departing colonial powers (Kenya, Tanzania, Pakistan, Philippines), institutional turmoil created by transition from war to peace, or from a command to a liberal economy (Guatemala, Latvia), a culture of repression and disdain for independent public institutions resulting from successive military regimes (Pakistan, Guatemala), as well as the absence of effective public interest advocates in civil society (Pakistan, Latvia). (v) Institutions of accountability are weak and ineffective. 5.13 Where levels of public-sector corruption are low, one normally finds strong institutions of accountability that control abuses o f power by public officials. These institutions are either created by the state itself (e.g. auditors-general, judicial systems, and legislatures) or are non-state-based (e.g. news media and organized civic groups). Public-sector accountability works best when the two types o f institutions complement each other. Pressure from non-state-based institutions of accountability compels the state-based institutions to do their job; conversely, non-state accountability actors (such as civic groups) are motivated to act when they have a reasonable expectation that the formal state institutions will respond meaningfully to their demands. 5.14 There are glaring weaknesses in state-based institutions o f accountability in all the case study countries. In Kenya, the Public Accounts Committee o f Parliament (PAC) 39 should play a critical role in overseeing public expenditure. But the P A C itself observes that many o f i t s recommendations to the government “are not implemented at all.” PACs exist in principle in Pakistan, but rarely operate. In the Philippines, traditions of “pork-barreling” and extravagant election-campaign spending induce legislators to indulge in corruption-behavior that undermines their credibility as “watchdogs” over the executive. The judiciary, as noted, i s compromised in many of the case study countries. Auditors General also fall far short of exercising their potential for holding governments accountable. Even where Auditors General are relatively competent, their effectiveness i s limited by the unwillingness of political authorities to respond to criticism and suggestions. Pakistan’s Auditor General reports to the President rather than Parliament, which undermines i t s independence. In Kenya, the government routinely ignores the Auditor General. 5.15 In some countries there are signs o f strength in non-state institutions of accountability. News media actively report on corruption in the Philippines, Kenya, Latvia, and Guatemala. Civil society activism i s a counterweight to official power in the Philippines, where the ouster of two Presidents over the past 20 years was triggered by popular disenchantment with corruption. The leverage that can be exerted by non- state-based institutions of accountability i s however limited by failures o f the state-based institutions. (vi) Information-accurate, timely and complete information about public expenditures and the performance of public servants and public programs-is scarce. 5.16 Lack of information breeds corruption. When improper transactions evade scrutiny, institutions of accountability cannot hold the government accountable. Systems for generating and disseminating information on public sector operations are problematic in the case study countries. In Kenya, the Controller and Auditor General i s four years behind in reporting on the government’s accounts, while Departments routinely fail to produce accounts, maintain ledgers and document expenditures. “Pork barrel” allocations to politicians in the Philippines for spending on local development projects are loosely controlled, leading to widespread abuse. Controller and Auditor General reports in Tanzania indicate accounting anomalies in transactions equivalent to nearly one-quarter of all public expenditure during the period 1991 to 1999. Apart from financial information, information about the performance o f public servants and public programs i s scarce or absent in the case study countries. (vii) Heavy state intervention in the economy through intrusive regulatory regimes and state-ownership of industrial enterprises provides ample opportunities for corruption. 5.17 Regulation i s meant to serve the public interest, but in many of the case study countries, regulation provides an opportunity for predatory behavior b y the state. Overlapping and/or archaic rules governing business licensing, import/export activity, and taxation generate millions of interactions between officials and citizens-each one an opportunity for an unscrupulous official to extract a private benefit. Entrepreneurs in Pakistan reported spending 12 percent of their time dealing with tax and regulatory 40 requirements that frequently involved requests from officials for bribes. Five distinct taxation authorities in Tanzania impose approximately 150 different taxation, permit, fee and licensing requirements on private entrepreneurs. I t i s common practice for tax and customs officials in Pakistan, Philippines and Tanzania to impose unreasonably high assessments on businesses and then bargain down to a lower assessment in return for a payment. A private entrepreneur in Tanzania related the following anecdote: Ifthere is a [shipping] container, the ‘extra duty’ is assessed at 80 million shillings. I n the negotiations, 20 million is paid as a bribe, and an extra 20 million goes to the government. This is the standard breakdown. 5.18 In many of the case study countries, state-owned enterprises become major sources of corruption when they are used as vehicles for funneling state resources into the hands of politically favored clients. (viii) Conditions in the civil service are conducive to corruption. 5.19 A “merit culture”-characterized by recruiting, promoting, and rewarding people based on qualifications, experience, and performance-is absent from civil services in the case study countries. High levels of politicization of the civil service, low responsiveness by civil servants to citizens, and a virtual absence of systems for tracking and evaluating the performance of public servants and public programs, are typical problems. The over-riding failure to pay attention to merit, performance and accountability, combined with the high level of political influence, are a powerful recipe for corruption. 5.20 L o w civil service salaries are cited in some analyses as an important cause of corruption in the case study countries. A Bank study o f the Philippines concluded that “higher level staff in the civil service are significantly underpaid.” A study conducted by Price-Waterhouse in Kenya indicated that private sector salaries exceed public sector salaries by 450 percent for comparable professional positions and 300 percent for non-professional positions. However, these comparisons are not useful as these studies do not take into account the perks and privileges that public sector employees receive. Nonetheless, research on public-sector corruption suggests that there i s no simple relationship between low public sector salaries and high levels of corruption. Raising salaries in an environment where there i s no merit culture, and/or where there i s little probability of corrupt acts being detected or punished, i s unlikely to have any important impact on corruption (see Gurgur and Shah 2002). However, a merit based system can only be successful if civil servants receive a wage that covers basic living costs. (ix) The commitment of high-level political leaders to combating corruption is questionable. Corruption is as much an “ownership ’’ issue as it is a technical problem. 5.21 The Philippines provides an interesting case of the challenge posed to anti-corruption efforts by the behavior of political leaders. In 2000 the Bank, at the behest of President Estrada, completed the study Combating Corruption in the Philippines. Estrada received the report favorably and requested that a national anti-corruption strategy be developed. But even as Estrada was considering the Bank’s report, the Philippine Center for Investigative Journalism was launching an investigation 41 into Estrada’s own involvement in high-level corruption. In early 2001 a popular uprising related to the corruption charges led to Estrada’s removal from power. 5.22 The Estrada case i s a dramatic example of a situation relevant to all the case study countries. The Bank operates at the request of governments who are supposed to take the lead in promoting and implementing anti-corruption interventions. But where corruption i s endemic-as i t i s in all of the case study countries except Latvia-it i s reasonable to suspect that i t touches the highest levels of government. When corruption i s endemic, i t i s a symptom of a deeper institutional problem. Endemic corruption exists either because national decision-makers are content with the status quo, or are unwilling to change it. This casts doubt on the sincerity of formal high-level statements of commitment to the fight against corruption, which may constitute “window dressing” rather than a real desire to reform. This raises a difficult problem as often the countries (regions and sectors) most in need of “anti-corruption” assistance may be the least interested in receiving it. There are no easy answers to this “demand dilemma.” One way that the Bank has been dealing with this problem i s fostering demand by working with a wide spectrum of stakeholders. A clear diagnostic of the problem, identification of good entry points for intervention, and selectivity are key to enhancing Bank effectiveness in i t s anticorruption activities. (x) Fraud in public-sector procurement stands out as a major source of corruption. 5.23 A constant in the country studies i s the perception that public-sector procurement i s one of the most-if not the most-significant sources of corruption. The government’s position as sole or predominant buyer for certain goods and services, combined with the large sums of money at stake, create huge opportunities for corruption. An audit commissioned by the Bank in Pakistan in the late 1990s found that over 90 percent o f procurement contracts were problematic in some significant way. A report of the Public Accounts Committee in Kenya lists 23 problematic public contracts, the aggregate value of which i s equivalent to 2.5 percent o f GDP. Formal “pork barrel” schemes in the Philippines, where funds intended for procurement at the district level are managed by local politicians, are notoriously vulnerable to corruption. A report b y the Philippines Commission on Audit found that up to 60 percent of allocations for district-level procurement o f books and medicines were skimmed off to pay kickbacks. A separate report estimates that half of the country’s publicly funded capital expenditure was lost through corrupt procurement practices. (xi) Laws covering corrupt activity are on the books, but enforcement is weak. 5.24 Lack of enforcement i s a symptom of weak government commitment to root out corruption. A recent study on corruption in Guatemala refers to a pervasive “culture o f impunity” in the public sector, founded on the perception that corrupt acts, even if detected, consistently go unpunished. Kenya provides a case in point. A lively and critical news media, as well as an active Public Accounts Committee and Controller and Auditor General generate information (albeit tardily)29on high-level corruption, but the government’s response i s far from commensurate with the level o f corruption reported. 29 The Controller and Auditor General’s 2000 report covered the 1996197 financial year. 42 In Tanzania, the government’s Prevention of Corruption Bureau produces only about six convictions a year (mostly against low-level functionaries) in a public-sector environment rife with corruption. Pakistan provides a contrast. The National Accountability Bureau (NAB) established in 1999 has obtained convictions of 48 officials of corruption, summarily dismissed 100 others and i s pursuing hundreds of investigations. I t remains to be seen, however, whether the NAB represents a systematic effort to clean up government, or whether i t (like similar institutions established in other developing countries following coups d’dtat) i s simply part of an strategy to discredit the previous regime. NAB does not have the mandate to investigate corruption in the powerful and influential military. (xii) Decentralization can help to increase accountability and therefore reduce corruption. However, one must pay attention to the institutional environment and risks of local capture. 5.25 I t i s often argued that decentralization i s an antidote to corruption because i t brings government “closer” to citizens. Decentralization as a means to make government officials more accountable can help reduce corruption and improve local public service delivery. Efforts to improve service delivery usually force the authorities to address corruption and i t s causes. Bank support for bottom-up reforms in service delivery helps advance this agenda by involving stakeholders in monitoring performance and demanding corrective actions. Improvements in service delivery are also more readily measurable than changes in levels of corruption. 5.26 However, in environments typical of the case study countries, where political interference in public administration i s the norm, merit culture and management systems in the civil service are weak, and institutions of accountability are ineffective, decentralization may in some ways increase opportunities for corruption. The issue i s significant in Pakistan and Philippines, and relevant to Guatemala and Tanzania (where more limited decentralization i s underway). Pakistan has launched a decentralization program involving the creation of 7,000 local and sub-national governments. Given systemic politicization of public services in Pakistan, decentralization may intensify rather than reduce pressures for political/bureaucratic collusion, although this may be mitigated by further administrative decentralization, giving local elected officials the power to hire, fire, and set terms of employment of civil servants in their jurisdiction. Identical concerns are pertinent in the Philippines, where legislation in 1991 devolved to regions and localities powers to provide services and raise revenue. A study of local government procurement in the Philippines revealed that (see Tapales 2001, p. 21): “contractors admit to paying mayors o f the towns where they have projects because, they say, the ofJicials can delay the work by withholding necessary permits or harassing the workers. Municipal mayors get seven percent while the barangay [village] captain is given three percent. The heads o f implementing agencies - usually the district, municipal or city engineer - get about 10 percent. ” 43 Country Ratings-Relevance and Ef‘ficacy 5.27 Each country’s anti-corruption programs are rated for their “relevance” and “efficacy” for fighting corruption, based on the analysis of i t s lending and ESW. “Relevance” i s judged in the same way as in chapter 4 (also see annexes 4 and 8 for the definition o f relevance). 5.28 Efficacy: OED defines efficacy “the extent to which the project’s objectives were achieved, or expected to be achieved, taking into account their relative importance.” In our present context, therefore, by “efficacy” we mean the extent to which the anti-corruption activities supported by the Bank achieved or are expected to achieve the objective o f either reducing the incidence of corruption or affecting contributory factors that are directly associated with corruption. The efficacy question i s o f the form: “Did the anti-corruption activities supported by the Bank lead either to reduced levels of corruption, or a reduced presence o f factors contributing to corruption, or an increased presence o f factors inhibiting corruption?” 5.29 “Relevance” and “efficacy” are rated as “high,” “substantial,” “modest” or “negligible.” Some lending interventions were not complete at the time that this evaluation was prepared. I n those cases our judgments about eficacy are based on assumptions regarding the most probable outcome. Exogenous political and other factors may affect negatively the outcome of even the most highly relevant anti-corruption interventions, thereby resulting in potentially differing relevance and efficacy ratings. A Gap Between “Relevance” and 5.30 The Bank has done a good job of supporting anti-corruption activities that are relevant to the corruption problem on the ground. The Bank’s actions in all six countries had “substantial” relevance. This i s to be expected given that in our sample selection we looked for countries where the Bank had an intensive level o f relatively successful anti-corruption engagement. Table 5.2: Summary of Ratings Country Relevance Rating Long term Efficacy Rating Guatemala Substantial Modest Kenya substantial Unevaluable Latvia substantial substantial Pakistan substantial Unevaluable Philippines substantial Substantial Tanzania substantial Unevaluable Source: Annex 10. 5.3 1 However, the country studies reveal a gap between the relevance and the efficacy of anti-corruption activities supported by the Bank (table 5.2). In view of the short time elapsed, the Bank had only modest success in translating i t s good performance with 30 The detailed ratings for each country are presented in annex 10. 44 respect to relevance into efficacy in achieving corruption reduction objectives. Latvia and the Philippines receive a higher rating o f substantial for the potential long term efficacy due to increasing civil society involvement that will, over time, lead to an increase in the accountability o f the country’s leaders. The potential efficacy rating for Pakistan i s unevaluablehncertain and depends on how well the government’s program for governance i s implemented and if the judiciary i s able to restore i t s credibility as an independent arbiter. Similarly such a long term efficacy for Kenya and Tanzania i s unevaluable as government commitment to initiate and sustain difficult governance reforms appear weak. 5.32 The complexity of the task at hand and the enormity o f the challenge and the short time period accounts for the gap between relevance and efficacy. The relevance o f the Bank’s anti-corruption actions i s clustered around a set o f factors related to information about public sector operations, state intervention, rule o f law, institutions o f accountability and civil service reform (see figure 5.1).31These are all major drivers o f corruption, and so i t i s appropriate that the Bank focuses on them. The skewing o f the Bank’s relevance toward a limited set o f factors affecting corruption i s an important clue to understanding the gap between relevance and efficacy. The factors that the Bank focused on do not operate in isolation. The resistance to change is in large part attributable to the weak enforcement o f sanctions against corruption, the state’s l o w legitimacy in the eyes o f citizens, and weak commitment by political leaders to fighting corruption. Exogenous factors may affect negatively the end result o f even the most highly relevant anti-corruption interventions. So, for example: 0 progress in building effective institutions o f accountability i s likely to be thwarted b y weak political commitment to fighting corruption, and b y a l o w level o f state legitimacy in the eyes o f citizens; 0 progress in establishing rule o f law i s likely to be impeded b y weak enforcement mechanisms, and b y deeply embedded political factors and social fragmentation, such as the dominance o f the military in political life that one sees in Pakistan and (until relatively recently) Guatemala, or the enduring legacy o f a command economy and one-party rule as in Latvia. 5.33 The relevance o f the Bank’s anti-corruption efforts i s commendable. I t has gravitated toward issues-public sector management, public financial management, regulatory reform, privatization, judicial sector reform-that invite technocratically oriented interventions such as provision o f training, technical assistance, and financial and information technology systems. This i s perhaps fitting, given the nature o f the Bank’s mandate, i t s project orientation, and the skills o f i t s staff. But as is indicated b y the gap between the relevance and the efficacy o f the Bank’s actions, these reforms may have a greater impact in the long run than in the short run. Corruption i s a complex issue, 3’ In the previous section o f this chapter, 12 factors affecting corruption were identified that were common across some or (in most cases) all o f the case study countries. We incorporated these factors into the relevance analysis in annex 10, noting for each country which o f the 12 factors were addressed b y the Bank’s lending and analytical work (tables A10.2, A10.4, A10.6, A10.8a and A10.8b, A10.10a and AlO.lOb, and A10.12). Figure 5.1 shows the result o f aggregating across &hesix countries all the tick- marks in the country tables. 45 and incremental policy changes in the present may yield greater returns in the future as a cultural change in the public sector starts to take hold. Addressing corruption rooted in political factors and social fragmentation will further enhance the impact of technocratic reforms. Figure 5.1: Factors Affecting Corruption Addressed by Bank Interventions 40 35 30 25 20 15 10 5 0 Source: Annex 10. 47 6. Conclusions Conclusions of the Desk Review 6.1 The desk review assessed the progress of the World Bank in carrying out i t s anti-corruption strategy that was launched in September 1997. Because the implementation of this strategy i s at an early stage, the desk review only assessed i t s relevance and implementation. 6.2 The World Bank has come a long way in a short time in addressing corruption as a development issue. Signs of progress include increased attention to the incorporation of corruption concerns into: (i) internal administrative operations; (ii)country assistance strategies and economic and sector work; (iii) research; (iv) training and dissemination of knowledge on effective anti-corruption programs; (v) open dialogue with member countries; (vi) support for international conventions and conferences and (vii) projects. These developments are promising and the Bank has already accumulated a strong record in engaging countries and international organizations in discussing these issues openly. Most of these initial efforts were focused on building awareness about and support for anti-corruption work both inside the Bank and developing countries. The Bank has also made important strides in building i t s understanding about the causes and dynamics of corruption in various kinds of country settings. 6.3 As the Bank moves on to a more mature phase o f anti-corruption support, the continuing relevance and efficacy of i t s work will hinge upon greater selectivity (identifying countries and areas for priority attention) and developing differentiated strategies to address the most important causes of corruption in each country. This can be done by finding adequate entry points for anti-corruption efforts, such as service delivery performance and the involvement of a variety of non-state stakeholders and development partners. These stakeholders and partners are a means to foster demand and curtail corruption in highly corrupted environments where government commitments for a serious dialogue on anti-corruption efforts may not be forthcoming (see paras 2.6 through 3.2). In some cases (such as Kenya and Indonesia), Bank lending volumes have been curtailed in response to corruption concerns. In addition, partners can be involved (in any environment, whether highly corrupt or not) in areas where the Bank does not have a comparative advantage. A good example of this practice i s how the Bank directed the Latvian government to the Soros Foundation for requested assistance in campaign finance reform. This i s an approach that can continue to be developed. Conclusions of Case Studies 6.4 The main conclusions of the case studies can be summarized as follows: (i) The Bank’s continuing focus on deepening i t s understanding o f country specific causes of corruption with added emphasis on social, political, and institutional origins of corruption at the country level w i l l improve i t s policy advice and assistance dialogue with developing countries. A comprehensive analysis of the public and private governance environment would include 48 analysis o f the formal and informal institutions o f participation and accountability, law enforcement, civil service organization and culture, public financial management, and public sector service delivery performance. (ii)The Bank itself cannot directly affect the roots of corruption derived from political and social fragmentation. However, there are at least two ways corruption can be attacked indirectly, by looking for ways to facilitate actions by others, often external partners and local non-governmental stakeholders, and by strengthening the mechanisms that empower citizens to demand greater accountability for service delivery. 6.5 The Bank i s in a bind. I t s normal interlocutor i s the government, with whom i t must do all i t can to build a solid anti-corruption partnership. And yet in many o f the countries where the Bank would like to pursue an anti-corruption agenda, the government’s commitment to that same agenda i s weak or absent, and i s subject to shifting political currents. This creates a “demand dilemma’’ such that countries most in need of anti-corruption measures are the least interested in engaging in serious reform. 6.6 In countries where corruption i s endemic, many political and administrative leaders may have a vested interest in the status quo. The partnership against corruption must therefore include actors from outside the government. The Bank has recognized the critically important role to be played by other development partners and non-state stakeho1ders-e.g. civic groups-in applying pressure on the government to act firmly against corruption. Entities outside the government, whether they be organized civic groups, private sector organizations or the news media, comprise non-state-based institutions of accountability that have the potential to act as a check on abuses o f power by the government. The question for the Bank i s how, while acting within the terms of i t s mandate, to involve these outside actors in the fight against corruption. In addition, participatory processes in which the Bank i s already involved at the country level such as the CAS and the PRSP, which give priority to cross-cutting governance issues such as corruption, provide an important entry point for non-governmental stakeholders.. The Bank can use these processes to maximum advantage in building a broad-based coalition against corruption. 6.7 Service delivery performance i s useful as an entry point to addressing corruption. There i s strong evidence that levels of corruption are correlated with the quality o f public services. Therefore, any serious effort by the government to improve service delivery outcomes will have an impact on reducing corruption. Econometric analysis (Tomaszewska and Shah, forthcoming) shows that corruption in the public sector i s associated with lower quality public service outcomes as measured by infant mortality rates, child mortality rates, satisfaction from public health care provision, and satisfaction from roads and transportation. Improvements in service delivery are also more readily measurable than changes in levels of corruption. The Bank’s support of tools including citizen surveys of service delivery performance, budget transparency, and e-government allow people to hold government accountable. (iii) Country dialogue can be as important as lending operations in the fight against corruption. 49 6.8 In Latvia and the Philippines, some of the success that the Bank has had in advancing the anti-corruption agenda has been due in large part to the quality of the Bank’s dialogue with the government-as well as the impact of that dialogue on the broader discussion between government and civil society about governance. In Latvia, the quality and usefulness of the Bank’s country analysis became the basis for a strong relationship with the government, which saw the Bank as a trusted partner in anti-corruption work. The variety of analytical and awareness-raising outputs related to corruption that were supported by the Bank in Latvia helped initiate a change in attitudes among Latvian decision makers and non-governmental stakeholders. In the Philippines, the Bank’s study on corruption-produced at the government’s request, and followed by a government-led anti-corruption initiative-had a similar impact. In Kenya, possibly the Bank’s most successful effort to advance the anti-corruption agenda may well have been i t s decision to suspend lending for two years-a decision i t took partly because o f concerns about corruption. The Bank’s “disengagement” from Kenya sparked a high-level, public debate on corruption and likely played a role in the government’s decision to initiate a range of anti-corruption measures. (iv) Decentralization i s an important instrument for improving accountability and service delivery to citizens when the institutional factors that may lead to local capture are dealt with suitably. 6.9 Decentralization as a means to make government officials more accountable can help reduce corruption and improve local public service delivery. As mentioned above, efforts to improve service delivery usually force the authorities to address corruption and i t s causes. Bank support for bottom-up reforms in service delivery helps advance this agenda by involving stakeholders in monitoring performance and demanding corrective actions. Improvements in service delivery are also more readily measurable than changes in levels of corruption. 6.10 The central government i s normally the center of the Bank’s attention. There i s indeed no shortage of corruption problems at the central level in countries where the Bank operates, and so i t i s appropriate that the Bank devote anti-corruption resources there. Decentralization creates thousands of new public authorities, each one with powers to tax, spend and regulate that may be abused in environments where accountability i s weak and governance i s unhealthy. As the cases o f the Philippines and Pakistan indicate, the Bank can help countries address and harness the power o f decentralization to reduce public-sector corruption. 51 7. Recommendations 7.1 The desk review concludes that the Bank has come a long way in a short time with corruption concerns increasingly integrated into operational activities and internal processes. Understanding of the causes and dynamics of corruption in different country settings has improved. B y encouraging open debate, carrying out policy research and disseminating good practice to member countries and international organizations, the Bank has helped to raise the profile of governance issues onto the development agenda. As a result, support for anti-corruption work has grown within the development community. 7.2 The country case studies confirm the results of the desk review. They demonstrate that the Bank has supported highly relevant anti-corruption activities on the ground. However, the Bank has achieved only modest success so far in achieving durable outcomes. The short time elapsed since the anti-corruption program was launched, the unusual complexity of the task at hand and the enormity of the challenge account for the gap between relevance and efficacy. 7.3 The Bank’s anti-corruption activities have addressed key factors affecting corruption, including greater transparency in public sector operations, state intervention, rule of law, and capacity building aimed at good government and a competent civil service. But corruption i s grounded in political factors and social fragmentation over which the Bank has limited influence. Furthermore, long term capacity development in such areas i s not always the Bank’s comparative advantage. 7.4 Therefore, greater reliance on development partners may be warranted provided, of course, borrowers are receptive. A better understanding of political factors and social fragmentation at the country level would enhance the quality and impact of Bank advice and improve the design of the Bank’s anti-corruption interventions. As the Bank moves on to a more mature phase of i t s anti-corruption work, i t will need to be even more creative in i t s choice of instruments, the nurturing of partnerships and the sequencing of operations. Highly differentiated strategies are required to find adequate entry points and address the most important causes of corruption. Service delivery performance, an area of Bank’s comparative advantage, serves as a useful tool for detecting corruption and may serve as a good entry point to curtail it. Efforts to improve service delivery can lead authorities to address corruption and i t s causes. 32 7.5 To this end, the following challenges require continued attention:33 32 Service delivery embodies other concerns such as enabling environment for the private sector-the so called investment climate and “state capture” which implies that service delivery will not be consistent with local preferences but would serve the narrow interests o f the elite. 33 Many o f the recommendations here are consistent with those of the Reforming Public Institutions and Strengthening Governance: Implementation Update by the Public Sector Board (2003), which advises to deepen efforts to: a) understand and measure governance through (participatory) upstream diagnostic work, b) monitor the impact of Bank projects and programs in improving governance, c) mainstream governance concerns across sectors, d) balance a stronger focus on anti-corruption and governance with the need for ownership and the imperative of poverty reduction, particularly in weak governance environments, and e) practice selectivity by focusing efforts where likelihood o f success i s strong. 52 - Deal with the demand dilemma by fostering demand among a wide spectrum o f stakeholders. The Bank’s current approach targets anti-corruption support to countries that demand such assistance since ownership i s critical in fighting corruption. Unfortunately, the countries (regions and sectors) most in need of “anti-corruption” assistance tend to be the least interested in receiving it. Where acceptable to public authorities, the Bank should deal with this problem by fostering demand among a wide spectrum o f stakeholders in those countries (see para 3.12,3.21,3.23, and 4.19). - Continue assessments of governance environment. The Bank should continue to assist clients in carrying out assessments of their institutional environments, including a diagnostic of the main causes of corruption. I t should help develop a long term, holistic and sequenced reform program. The Bank and other donors should work with the government in identifying the areas where they can best help in implementing these reforms. The CAS should articulate the Bank’s role (see para 3.2,3.14,3.23 and 4.3-4.5). - Define the “governance”pre-requisites for lending. Corruption poses significant risks for the effectiveness of Bank lending and therefore the Bank should set minimum governance pre-requisites for lending, including governance assessments. This should lead to greater selectivity in lending and increased reliance on advisory activities in high corruption environments where there i s limited anti-corruption demand. Some governments may be reluctant to engage the Bank in a dialogue on “corruption” issues, because they believe that this i s not needed or helpful. In such cases, the Bank may need to make an independent assessment, to be reflected in the CAS, of whether “corruption” would impede the effectiveness of lending (see paras 4.10 through 4.16). - Focus lending instruments on accountabilityfor results. A sharper focus on accountability for results would help create appropriate incentives and promote public sector reform. Efforts to improve service delivery can help address corruption and i t s causes. The Bank should increase i t s support for reforms in service delivery to emphasize probity and results rather than only the level o f expenditures. Linking development assistance to demonstrable improvements in output performance o f the public sector, especially in areas most relevant to the poor, may be the most effective characteristic of the Bank’s evolving anti-corruption strategy (see para 4.10 and box 3.1, para 5.7 and 6.8). - Provide more assistance for bottom-up reforms. In many countries where corruption i s entrenched, governments lack either the will or the capability to mount effective anti-corruption programs. In such countries, the Bank and/or i t s partners may choose to amplify citizens’ voice and strengthen exit mechanisms so as to enhance transparency, accountability and the rule o f law (see paras 4.16 through 4.19).34 34Moving the decision making closer to people poses significant risks for state capture b y local elites in societies that lack democratic participation. Even in democratic societies, local government officials might engage in corruption. However, in the latter societies, the electoral process and taxpayer accountability work to restrain corruption in the long run (see Gurgur and Shah 2002 for a synthesis o f empirical evidence on this issue). 53 References Ablo, Emmanuel, and Ritva Reinikka.1998. Do Budgets Really Matter? - Evidence From Public Spending on Education and Health in Uganda. Policy Research Working Paper No. WPS 1926, Washington, D.C.: The World Bank. Brunetti, Aymo, Gregory Kisunko, and Beatrice Weder. 1997. Credibility of Rules and Economic Growth - Evidencefrom a Worldwide Survey o f the Private Sector. Background paper for the World Development Report 1997. Policy Research Working Paper No. WPS 1760, Washington, D.C.: The World Bank. Chaudhry, Azam. Forthcoming. “Corruption and Trade Liberalization.” OED Working Paper. Washington, D.C. Clague, Christopher. 1993. “Rule Obedience, Organizational Loyalty, and Economic Development.” Journal o f Institutional and Theoretical Economics 149(2) (June): 393-4 14. Das-Gupta, Arindam, and Dilip M. Mookherjee. 1998. Incentives and Institutional Reform in Tax Enforcement. New Delhi: Oxford University Press. Gray, Cheryl, and Daniel. Kaufmann. 1998. “Corruption and Development.” Finance and Development 35( 1): 7-10. Gupta, Sanjeev, Hamid Davoodi, and Rosa Alonso-Terme. 1998. Does Corruption AfSect Income Inequality and Poverty? Working Paper No. 98/76. Washington, D.C.: International Monetary Fund. Gurgur, Tugrul, and Anwar Shah. 2002. “Localization and Corruption: Panacea or Pandora’ s Box?” In Ehtisham Ahmad and Vito Tanzi (eds.), Managing Fiscal Decentralization, pp. 46-67. London and New York: Routledge Press. Hall, Robert E., and Charles I.Jones. 1999. “Why Do Some Countries Do Much More Output Per Worker Than Others?’ Quarterly Joumal o f Economics 114( 1): 83-1 16. Hines, James R., Jr. 1995. Forbidden Payment: Foreign Bribery and American Business afer 1977. NBER Working Paper, No. w5266. Cambridge, M A : National Bureau of Economic Research. Huntington, Samuel. 1968. Political Order in Changing Societies. New Haven: Yale University Press. Husted, Bryan W. 1999. “Wealth, Culture, and Corruption.’’ Journal of lntemational Business Studies 30(2) :339-60. Huther, Jeff, and Anwar Shah. 1998. Applying a Simple Measure of Good Govemance to the Debate on Fiscal Decentralization. Policy Research Working Paper No. 1894. Washington, D.C.: The World Bank. Huther, Jeff, and Anwar Shah. 2000. Anti-corruption Policies and Programs: A Framework for Evaluation. Policy Research Working Paper No. 2501.Washington, D.C.: The World Bank. Isham, Jonathan, Daniel Kaufmann, and Lant Pritchett. 1995. Governance and Retums on Investment: An Empirical Investigation. Policy Research Working Paper No. 1550. Washington, D.C.: The World Bank. Isham, J., D. Kaufmann, and L. Pritchett. 1997. “Civil Liberties, Democracy, and the Performance of Government Projects.” The World Bank Economic Review 1l(2): 219-42. 54 Johnson, John H., and Sulaiman S. Wasty. 1993. Borrower Ownership of Adjustment Programs and the Political Economy of Reform. World Bank Discussion Paper No. 199. Washington, D.C.: The World Bank. Johnson, Simon, Daniel Kaufmann, Andrei Shleifer, Marshall IGoldman, and Martin L. . Weitzman. 1997. ‘The Unofficial Economy in Transition.” Brookings Papers on Economic Activity (2): 159-239. Kaufmann, Daniel 1998. “Challenges in the Next Stage o f Anti-corruption.” In New Perspectives on Combating Corruption. Berlin: Transparency International and Washington, D.C.: The World Bank. Kaufmann, Daniel, Aart Kraay, and Pablo Zoido-Lobaton. 1999. Governance Matters. Policy Research Working Paper No. 2196. Washington, D.C.: The World Bank. Kaufmann, D., and S.-J. Wei. 1999. Does “Grease Money” Speed up the Wheels of Commerce? NBER Working Paper No. w 7093. Cambridge, MA: National Bureau o f Economic Research. Khan, Mushtaq, and Kwame Sundaram Jomo. 2000. (eds.) Rents, Rent-Seeking and Economic Development. Cambridge, MA: Cambridge University Press. Klitgaard, Robert 1988. Controlling Corruption. Berkeley and Los Angeles: University o f California Press. Knack, Stephen, and Philip Keefer. 1995. “Institutions and Economic Performance: Cross-country Tests Using Alternative Institutional Measures.’’ Economics and Politics 7(3): 207-28. Krueger, Anne. 1974. “The Political Economy o f the Rent-Seeking Society.” American Economic Review 64(3): 291-303. Lambsdorff, Johann Graf. 1999a. “Corruption in Empirical Research - A Review.” Paper presented at the 9” International Anti-Corruption Conference, Durban, South Africa, October 9-15, 1999. Lambsdorff, Johann Graf. 1999b. “The Impact o f Corruption on Capital Accumulation.” Unpublished Paper, Gottingen University. Leeuw, Frans L., and Caroline Klein Haarhuis. 2000. “Measuring and Explaining Variance between Anti-Corruption Policies.” Paper presented at the European Evaluation Society Conference, Lausanne, Switzerland, October 12-14,2000. Marmorstein, Victoria. 1978. “World Bank Power to Consider Human Rights Factors in Loan Decisions.” The Joumal of International Law and Economics 13: 113-136. Mauro, Paolo. 1995. “Corruption and Growth.” The Quarterly Journal of Economics llO(3) 681-712. Mauro, Paolo. 1997. “The Effects o f Corruption on Growth, Investment, and Government Expenditure: A Cross-Country Analysis.” In Kimberly Ann Elliot (ed.), Corruption and the Global Economy 83-107. Washington, D.C.: Institute for International Economics. OECD (Organization for Economic Co-operation and Development). 1999. Public Sector Corruption: An International Survey of Prevention Measures. Paris: OECD. OED (Operations Evaluation Department). 1994. Annual Review of Evaluation Results. Washington, D.C.: The World Bank. OED. 1998. Financial Sector Reform: A Review of World Bank Assistance. Washington, D.C.: The World Bank. 55 OED. 200 la. Annual Review of Development Effectiveness 2000. Washington, D.C.: The World Bank. OED. 2001b. OED I D A Review: Governance - The Critical Factor, IDA10-12.1 Washington, D.C.: The World Bank. Paldam, Martin. 1999. The Big Pattern of Corruption: Economics, Culture and the Seesaw Dynamics. Department of Economics Working Paper No. 1999-11. Denmark: University of Aarhus. Paul, Samuel. 1998. Making Voice Work: The Report Card on Bangalore’s Public Service. Policy Research Working Paper No. 1921. Washington, D.C.: The World Bank. Picciotto, Robert. 1995. Putting Institutional Economics to Work: From Participation to Governance. World Bank Discussion Paper No. 304. Washington, D.C.: The World Bank. Picciotto, Robert. 2000. “Poverty Reduction and Institutional Change.” Seton Hall Journal of Diplomacy and International Relations. Inaugural Issue. Picciotto, Robert. 2001. “The Logic of Mainstreaming.A Development Evaluation Perspective.” Keynote address for the American Evaluation Conference, St. Louis, Missouri, November 7-10,2001. Pope, Jeremy. 2000. “Confronting Corruption: The Elements of a National Integrity System.” T I Source Book 2000. Berlin and London: Transparency International. Rose-Ackerman, Susan. Forthcoming. “Guidelines for Reviewing World Bank Efforts to Control Corruption.’’ OED Working Paper. Washington, D.C. Schacter, Mark. Forthcoming. “Institutions of Accountability and Corruption: An Analytical Framework.” OED Working Paper. Washington, D.C. Schacter, Mark, and Anwar Shah. 2000. “Look Before You Leap: Notes for Corruption Fighters.” Presented at the International Conference on Combating Corruption, Seoul, Korea, December 2000. Segal, Philip. 1999. “Dealing with Devil: The H e l l of Corruption.” Impact. Spring. Washington, D.C.: InternationalFinance Corporation. Serageldin, Ismail. 1990. “Governance, Democracy and the World Bank in Africa.” Address to the World Bank Legal Department Seminar on Governance, November 30, 1990. Shah, Anwar. 1999. “Governing for Results in a Globalized and Localized World.” The Pakistan Development Review 38:4 Part I (winter 1999):385-431. Shah, Anwar. 2000. “Comments on Governance and Anti-corruption: New Insights and Challenges,” by D. Kaufmann. In Evaluation and Poverty Reduction. Proceedings from a World Bank Conference, Operations Evaluation Department, Washington, D.C.: The World Bank. Shihata, Ibrahim. 1991. The World Bank in a Changing World: Selected Essays. Dordrecht: Nijhoff (Kluwer Academic Publishers). Tanzi, Vito, and Hamid Davoodi. 1998. “Road to Nowhere: How Corruption in Public Investment Hurts Growth.” Economic Issues No. 12. Washington, D.C.: IMF. Thompson, Theresa. Forthcoming. “Corruption and the Economic Rate of Return on World Bank Funded Projects.” OED Working Paper. Washington, D.C. 56 Thompson, Theresa, and Anwar Shah. Forthcoming. “Transparency International’s Corruption Perceptions Index: Whose Perceptions Are They Anyway?” OED Working Paper. Washington, D.C. Tomaszewska, Ewa, and Anwar Shah. Forthcoming. “Phantom Hospitals, Ghost Schools and Roads to Nowhere: The Impact of Corruption on Public Service Delivery Performance in Developing Countries.” OED Working Paper. Washington, D.C. U.N. Secretary General. 1992. An Agenda for Peace. New York: United Nations. U.S. General Accounting Office. 2000. World Bank: Management Controls Stronger but Challenges in Fighting Corruption Remain. April 2000. Wei, Shang-Jin. 1997a. How Taxing is Corruption on International Investors. NBER Working Paper No. 6030. Cambridge, MA: National Bureau of Economic Research. Wei, Shang-Jin. 1997b. Why is Corruption So Much More Taxing than Tax? Arbitrariness Kills. NBER Working Paper No. 6255. Cambridge, MA: National Bureau of Economic Research. Williams, Robert, and Robin Theobald. 2000. (eds.) Corruption in the Developing World (The Politics of Corruption, 2). U.K.: Edward Elgar Publishers. World Bank. 1983. World Development Report 1983. New York, N.Y.: Oxford University Press. World Bank. 1989. Sub-SaharanAfrica: From Crisis to Sustainable Growth: A Long Term Perspectives Study. Washington, D.C: The World Bank. World Bank. 1990. The Long-Term Perspectives Study of Sub-Saharan Africa: Background Papers. Washington, D.C.: The World Bank. World Bank. 1992. Governance and Development. Washington, D.C.: The World Bank. World Bank. 1994. Governance: The World Bank’s Experience. Washington, D.C.: The World Bank. World Bank. 1997a. World Development Report 1997. New York, N.Y.: Oxford University Press. World Bank. 2000a. Anticorruption in Transition: A Contribution to the Policy Debate. Washington, D.C.: The World Bank. World Bank. 2000b. Helping Countries Combat Corruption: Progress at the World Bank Since 1997. Washington, D.C.: The World Bank. World Bank. 2000c. Can Africa Claim the 21” Century? Washington, D.C.: The World Bank. World Bank. 2000d. Reforming Public Institutions and Strengthening Governance. Washington, D.C.: The World Bank. World Bank. 2000e. World Development Report 2000/2001. Attacking Poverty. New York, N.Y. Oxford University Press. World Bank. 2001. Initiatives in Legal and Judicial Reform. Legal and Judicial Reform Unit. Washington, D.C.: The World Bank. World Bank. 2003. Reforming Public Institutions and Strengthening Governance: A World Bank Strategy Implementation Update. Public Sector Group. Washington, D.C.: The World Bank. 57 Annex 1 Annex 1: Methodology and Sample Selection The report provides a comprehensive review o f Bank products since September 1997. The following presents a summary o f the data sources and time periods for the Bank products analyzed in the report. CAS: All CAS documents (CASs, Interim Support Strategies, Transitional Support Strategies, CAS Progress Reports) from September 1997 until June 30,2001 were studied, totaling 113 documents. There were 92 countries in this sample (some countries had more than one CAS document). Lending: In all, there were 93 projects analyzed among 1,093 project documents in Imagebank between September 1997 and June 30,2001. Of the 93 projects that were rated for the anti-corruption report, 49 o f the approximately 100 public sector reform projects (as designated by the Public Sector Board) were included. Projects were selected so that each region was represented in the sample. In the case o f judicial reform, all projects approved between 1999 and 2001 were included. Economic and Sector Work (ESW) (chapter 3): CPAR: 18 reviewed (out of 45) CEM: 40 reviewed (out o f 86) CPFNCFAA: 22 reviewed (out of 57) PER: 25 reviewed (out o f 68) IGR: 2 reviewed (two o f the two available were reviewed although 7 were completed in draft form) Table A1.1: E S W Coverage by Level of Corruption (chapter 3) Sample as a Sample by Incidence of Corruption percent of Total High Substantial Modest LOW CFAA 39% 73% 23% 4% PER 37% 15% 73% 12% CPAR 40% 78% 17% 5% CEM 45 % 16% 51% 27 % 6% IGR 29 % 100% ESW 43 % 9% 67% 20% 4% Implementation of the Anti-corruption Strategy at the Country Level (chapter 3): The necessary data was available for 82 countries to be included. Implementation in lending, CAS, and ESW were taken together to determine the level o f integration o f the anti-corruption strategy at the country level. For a high or substantial corruption country, either implementation in lending (having a project with modest or higher compliance) or implementation in a recent CAS (compliance of modest or higher) was needed for implementation at the country level. Implementation would be considered adequate at the country level for a modest corruption country if there was implementation in ESW. A country’s highest project rating and their highest CAS rating for compliance were used, and ESW compliance at the country level was determined separately (see below). Annex 1 (continued) 58 Implementation in ESW, at the Country Level (chapter 3): The Bank would be considered compliant in ESW for a country if i t had: i) Any CFAA, CPAR, or IGR, or ii) Implementation in a PER or C E M (determined by examining the document). Country Level Relevance (chapter 4): The necessary data was available for 76 countries to be included. Relevance ratings for CASs and lending were averaged to calculate relevance at the country level. A country’s highest project rating and their highest CAS rating for relevance was used in this exercise. Assessment o f World Bank Financial Liberalization Initiatives (box 2.1): This was a review o f 87 projects in 24 countries. Assessment o f World Bank Trade-Related Initiatives (box 2.1): This analysis used OED outcome ratings from 115 trade related projects. Governance Related Bank Activities (Workshops and Surveys, ESW, Grant based TA, Lending) from section “Main Tools in the Bank’s Package of Assistance” (chapter 2): 140 countries were included. The information on the number of governance related Bank activities (workshops and surveys, ESW, grant based TA, and lending) was taken from the report “Helping Countries Combat Corruption: Progress at the World Bank Since 1997” and was updated to FYOl by the regions. The ratings o f the “Degree o f Bank Involvement” for Grant Based TA, ESW and Mission Reports, and In-Country Workshops and Surveys were based on the following scale: Number of Activities Degree of Bank Involvement 5 or more activities High 3-4 activities Substantial 1-2 activities Modest 0 activities Negligible The Degree o f Bank Involvement in Governance Lending was determined separately. The following scale was used: I Governance Lending Per Cauita I Degree of Bank Involvement Greater than $20 per capita High $15 - 20 per capita Substantial $5 - 10 per capita Modest Less than $5 per capita Negligible For table 2.4-Package of Reforms by Incidence of Corruption: Using the summaries o f governance related projects for countries where there was greater than $5 per capita in governance lending, the number o f times a theme appeared in the governance related loans was tallied. Those themes that were most prevalent are listed in table 2.4 o f the report. 59 Annex 1(continued) Figure Al.1: Degree of Bank Involvement-Grant Based Technical Assistance 0 Negligible El Substantial Bm High High Substantial Modest Low Corruption Corruption Corruption Corruption Level of Corruption (CPIA) Figure A1.2: Degree of Bank Involvement-ESW 100% 90% 80% 70% Involvement 60% 50% 40% 30% El Substantial 20% 10% l lHigh 0% High Substantial Modest Low Corruption Corruption Corruption Corruption Level of Corruption (CPIA) Annex 1 (continued) 60 Figure A1.3: Degree of Bank Involvement-In-country Workshops and Surveys 0 Negligible Involvement (NONE) El Modest Elsubstantial 0% Y High Substantial Modest Low Om High Corruption Corruption Corruption Corruption Level of Corruption (CPIA) CPIA: The 2000 CPIA rating for question 20, “Transparency, Accountability, and Corruption in the Public Sector” was used as a country’s corruption rating throughout the report. The number scale given in the CPIA was transformed into the categories of high, substantial, modest, and low corruption according to the table below. Using a three-year average o f the CPIA did not lead to any significant change in the results. CPIA Score Level of Corruption 1-2 High Corruption 2.5 - 3 Substantial Corruption 3.5 - 4 Modest Corruption 4.5 - 6 L o w Corruption Workshop: There was a workshop conducted in July 2000, which included participants from OED, PREM, WBI, University o f Virginia, and the Center for Democracy Governance. The workshop was designed to discuss the framework for evaluation o f anti-corruption programs and case study proposals. The OED evaluation framework for anti-corruption programs was presented at this time. Surveys and Interviews: There were two sets o f interviews and surveys conducted for the background research o f this report. In the first, 22 E C A staff were interviewed, and a survey was conducted. The interviewees included staff from country offices, PREM, EDI, and from the sectors. In the second, 30 task managers were interviewed regarding the Bank’s anti-corruption strategy and i t s linkage with CAS work. Those interviewed represented task managers from all regions and different sectors. Seminars: Three Bank-wide seminars were held on the research that had been conducted for the report, regarding trade liberalization, financial liberalization, and the CAS and lending projects review. 61 Annex 2 Annex 2: Criteria for Rating Compliance of Instrument for Anti-corruption Strategy Znstrument Criteria for Rating Compliance of Instrument for Anti-Corruption Strategy Lending A project’s compliance with the anti-corruption strategy i s rated taking into account the incidence of corruption in the country and whether or not that calls for design (93 projects features to curtail incentives and opportunities for malfeasance. For example, for a reviewed between project i n a high corruption environment compliance rating i s done as follows: FY98 and FYO1) High when the project has a major component geared toward reducing corruption and improving governance. There may be specific reference to corruption and governance concerns i n the project. Substantial when the project has substantive components geared toward reducing corruption. There may be specific reference to corruption and governance concerns in the project. Modest when the project has minor components geared toward reducing corruption. Negligible when the project has almost no component geared toward reducing corruption. CAS CAS compliance i s rated taking into account the incidence of corruption in the country and proposed Bank responses in dealing with it. If the country has low incidence of (113 documents corruption and CAS does not address corruption issues, i t will be considered compliant reviewed-all with the Anti-Corruption Strategy. On the other hand, for a highly corrupt country, CASs between CAS i s rated as follows: July 1998 and June 2001) High when the CAS fully complies with the Bank’s anti-corruption strategy. When corruption i s a major theme and the proposed lending program has strong anti-corruption elements, then, the CAS receives a high compliance rating. Substantial when the CAS includes most elements o f the anti-corruption strategy. When there i s a substantial discussion o f corruption and the lending progradadvice has a significant amount of anti-corruption measures, the CAS receives a substantial compliance rating. Modest when the CAS includes some elements of the anti-corruption strategy. This means that the CAS has some discussion of corruption, and that the lending progradadvice i s somewhat related to governance reform. Negligible when the CAS includes few or none o f the elements o f the anti-corruption strategy. This means that the CAS has little or no mention o f corruption in neither the discussion o f the country’s issues nor i s the proposed lending progradadvice focused on corruption. Economic and Compliance of ESW i s rated taking into account the incidence of corruption i n the Sector Work country concerned and the identified causes and policy responses in the ESW work. For example for a high corruption country, ESW i s rated as follows: CPAR (18 reviewed) High if corruption and governance are major themes o f the document. CPFNCFA A Substantial i f corruption and governance are rather important themes o f the document. (22 reviewed) Modest if corruption and governance are somewhat important themes o f the document. PER Negligible if corruption and governance issues are not mentioned in any meaningful (25 reviewed) way, or not at all. IGR (2 reviewed) CEM (40 reviewed) Annex 3 62 Annex 3: Criteria for Rating Relevance of Instrument for Anti-corruption Strategy 1 Instrument Criteria for Rating Relevance of Znstrument for Anti-Corruption Strategy Lending A project’s relevance i s rated as: High when the project’s anti-corruption objectives play a key role in the country’s (93 projects current development priorities and the Bank’s current CAS, and are fully consistent reviewed between with the Bank’s current sectoral assistance strategy and corporate goals and policies. FY98 and FYO1) In this case, the project has great potential to achieve i t s anti-corruption objectives due to the project correctly targeting the country’s specific causes o f corruption. Substantial when the project’s anti-corruption objectives are mostly consistent (minor shortcomings only) with the country’s current development priorities, the Bank’s current country and sectoral assistance strategies, and the Bank’s current corporate goals and policies. In this case, the project has reasonable potential in achieving i t s anti-corruption objectives due to the project correctly targeting the country specific causes of corruption. Modest when the project’s anti-corruption objectives have one or more significant inconsistencies with the country’s current development priorities, the Bank’s current sectoral assistance strategies, or the Bank’s current corporate goals and policies. In this case, the project has some potential in achieving its anti-corruption objectives. Negligible when the project’s anti-corruption objective are absent, or are mostly inconsistent with, and possibly counterproductive to, the country’s current development priorities, or the Bank’s current sectoral assistance strategies, or the Bank’s current corporate goals and policies. In this case, the project does not have potential to achieve anti-corruption objectives because the project i s not targeting the causes of corruption in the country. CAS CAS relevance i s rated as: High when the lending programsladvice play a key role in the country’s current (113 documents development priorities, and the Bank’s current sectoral assistance strategy and reviewed-all corporate goals and policies. When a CAS has a strong diagnosis o f the causes of CASs between corruption and the advice and planned lending activities are appropriate for the July 1998 and country’s governance level, the document i s rated as highly relevant. June 2001) Substantial when the lending programstadvice are mostly consistent (minor shortcomings only) with the country’s current development priorities and the Bank’s current corporate goals and policies. When a CAS has a good diagnosis of the causes of corruption and the proposed lending progradadvice i s for the most part appropriate given the country’s level o f corruption, the document i s rated as having substantial relevance. Modest when the lending programstadvice have one or more significant inconsistencies with the country’s current development priorities, the Bank’s current sectoral assistance strategies, or the Bank’s current corporate goals and policies. A CAS with modest relevance suffers a disconnect between an understanding o f the causes of corruption and the prescription. There may be a fairly good diagnosis (but a weak program) or vice versa. Negligible when an anti-corruption strategy i s absent, or i s mostly inconsistent with, and possibly counterproductive to, the country’s current development priorities, or the Bank’s corporate goals and policies. There i s neither a diagnosis o f corruption nor i s the proposed lending progradadvice appropriate for dealing with that country’s level of corruption. Note: Relevance is not applicable for low corruption countries. 63 Annex 4 0 Annex 5 64 65 Annex 6 0 0 5: W v f 3 W Annex 7 66 0 0 m vr 3 0 m I , , , I e m 0 vr 3 cr, m h id I .e a id u ki a 67 Annex 8 Annex 8: Framework for Evaluation of Relevance: Principal Agent Theory and New Institutional Economics Moral Hazard Issues a. One intellectual framework often used for analyzing corruption i s principaVagent theory, as developed in the economics literature in recent decades. In this framework, there i s a principal (say a supervisor) dealing with an agent (say an employee) in a situation o f asymmetric information. Specifically, the agent knows more about the details o f his work than does the principal. The objectives of the principal and the agent are normally not identical. The principal can collect various types o f information (with specified costs of information collection) and she can specify the criteria according to which the agent w i l l be rewarded or punished. A rich literature has developed spelling out the kinds o f reward systems and information collection mechanisms that will enable the principal to maximize the attainment o f her objectives, subject to the resource and information constraints. I t i s generally assumed in this literature that both the principal and the agent are self-interested, rational, and calculative. b. One way to apply this framework to corruption in public bureaucracies i s to suppose that top-level politicians and administrators want public services to be delivered without corruption on the part o f the employees. In that case, the top-level decision-makers would set up a meritocratic civil service along the lines described long ago by M a x Weber. Recruitment would be b y examination, salaries would be generous, employment would be secure as long as the rules were not broken by employees, retirement pensions would be generous but would be contingent on following the rules, and so forth. Of course there would be clear lines o f authority, and accounting systems and auditing procedures would be put in place to monitor employee performance and uncover corruption. c. . I f we look at the incentives facing the employee, we see that there are benefits and costs o f engaging in corruption. The rational, self-interested bureaucrat calculates the expected net benefits to himself o f breaking the rules and engaging in illicit exchanges with private interests, extortion o f private actors, and theft o f public assets. As discussed above, the benefits can be monetary or non-monetary, and easy or hard for outsiders to detect. The expected costs to the bureaucrat are the probability of paying a penalty times the magnitude o f the penalty. The penalties can take the form o f fines, imprisonment, loss o f j o b and pension, public humiliation, and so forth. d. The top-level decision-makers can manipulate the benefits and expected costs in various ways. The bribes that public officials can extract are strongly influenced by the value to private interests o f the services the officials have to offer, In an economic regime o f price controls, exchange restrictions, directed credit, import quotas, industrial licensing, public officials have a very strong bargaining position vis-&vis private interests. Thus economic liberalization (“neoclassical” economic policy reform) can be a powerful tool for curbing corruption. The expected costs o f corruption depend on the probabilities o f detection, prosecution, and exaction of a penalty. The probability o f detection can be increased by inspections and audits. The most corrupt countries are most in need o f these agencies. Inspections and audits can only be effective when the agencies conducting them are not corrupt themselves. This i s less likely to be the case the more corrupt the country is. The penalty o f dismissal and loss o f pension can be exacted by the public agency itself, while the probability o f prosecution and judicial punishment depend on the competence and motivations within the judicial system. e. A widely used formula for analyzing corruption control i s Corruption = Monopoly + Discretion - Accountability (Klitgaard 1988). Corruption can sometimes be reduced by Annex 8 (continued) 68 permitting competition among public agencies (in granting licenses or permits, for example). Corruption can also be controlled by limiting the discretion o f public employees, for example by allowing an employee only certain types o f actions. Accountability can be enhanced by inspections, audits, and follow-up procedures that result in penalties for discovered violations. I t should be noted, however, that discretion i s not necessarily the enemy o f organizational efficiency. Organization theorists and practitioners have found that the most efficient organizations are not those that follow the rules rigidly in a Weberian manner, but those that permit and encourage their members to exert effort, exercise initiative, and take responsibility (as discussed in Clague 1993 and many other places). Understanding this type o f behavior requires us to move away from the strictly rational self-interest assumption o f principaYagent theory, and to recognize the possibility o f group identity, internalization o f group goals, and esprit de corps. Organizations that succeed in eliciting effort, initiative, and responsible behavior from their employees are found much more frequently in the private sector o f developed countries than in their public sectors, and they are probably extremely rare in the public sectors o f less-developed countries. But the possibility o f creating such organizations in the public sectors o f poor countries should not be ruled out, and this has implications for designing rules to curb corruption in these organizations (see the discussion o f reforming the Mexican intemal revenue service in Das-Gupta and Mookherjee 1998). f. PrincipaYagent theory can also be used when we relax the assumption that top-level decision-makers are interested in the cormption-free delivery o f public services. In this analysis, the principal may be the public and the political leaders may be their agents. This i s known as the theory of common agency. The degree o f control that the public exerts over political leaders i s obviously much less strict than the control o f a supervisor over her employee. However, some of the same principles apply: the principal should collect information and reward and punish agents according to their performance. A rich body o f democratic theory has been devoted to this problem, and the outlines of a properly functioning democratic system are well known. All countries fall short o f the ideal o f democratic accountability, but the shortfalls are especially severe in most less-developed countries, most o f which cannot be considered to be democracies at all. Even most o f those that meet the minimal conditions o f electoral democracy do not have good systems o f accountability o f politicians and public officials. Strengthening institutions of accountability i s one important strategy o f corruption control, but principauagent theory i s not particularly helpful in analyzing this problem. For this, we turn to institutional analysis. Institutional Analysis of Corruption g. The New Institutional Economics (NE), a development o f the last few decades, helps to explain how institutions affect economic development and how institutions change over time. In addressing these issues o f the determinants and the consequences o f different institutional arrangements, the NE has broadened i t s assumptions from those o f traditional neoclassical theory to include a wider range o f causal factors. At the same time, some branches o f political science have developed a style o f analysis similar to economics and they have also devoted much attention to the role o f institutions in political outcomes. Thus there i s a new style o f institutional analysis that has been developed b y both economists and political scientists. While it i s often called the New Institutional Economics, that name i s perhaps no longer appropriate, and in any case the analysis recognizes that economic development can only be understood b y taking politics seriously. In other words, a country’s economic progress depends critically on i t s political as well as i t s economic institutions. h. In the NE, institutions are the humanly devised rules o f the game. Institutions form a complex set o f rules, some fundamental for the society and others applicable only to particular 69 Annex 8 (continued) organizations. Organizations are institutions themselves, and they are players in economic and political games that are conducted under other institutions. An important idea in the NE i s that the games can have multiple equilibria. For example, within an organization or within a political framework, if other people follow the rules, then i t i s in each individual’s interest to do so too. On the other hand, if rule breaking i s widespread, individuals find it to their advantage to ignore the formal rules. 1. A situation in which no rules prevail can be called one that i s not institutionalized. A prominent analysis o f politics in less-developed countries characterizes these countries as lacking in institutionalization (Huntington 1969). While not all institutions are beneficial to society, the near absence o f institutions i s considered to be a very serious problem for development. j. Institutional analysis can take place on different levels. One can take as given the more fundamental institutions o f society and examine the choices at a lower level. For example, given a functioning market economy, one can analyze the organizational forms best suited to a small biotech firm or to a national retail chain. Alternatively, given a society’s culture and democratic constitutional framework, one can analyze how different campaign finance rules affect political outcomes. At a deeper level, one can visualize the constitutional political order or the cultural characteristics o f societies as institutions that have emerged out o f the largely self-interested actions o f individuals. As one progresses from smaller-scale to larger-scale institutions, the role o f conscious planning o f the institution diminishes. The important economic and political institutions o f most societies are not the result o f conscious planning b y small groups o f people but are the unplanned outcomes, over long periods o f time, o f the actions o f very many individuals and groups, all seeking their own goals. k. An important concept in the NE i s that of “coordinating on the rules o f the game.” Societies tend to function better when there are predictable rules o f behavior. There are few situations worse for societies than anarchy. Thus whatever rules prevail at a given time, there are normally many individuals and groups with a strong interest in seeing that the rules are respected. One implication i s that the basic institutions o f society normally change very slowly over time. However, the idea o f coordinating on the rules contains the possibility o f rapid changes from one political order to another. Clearly this i s more likely to occur in lower-level institutions than in higher-level ones. For example, individual companies in a market environment can often be “tumed around” quickly, and individual agencies in the public sector can experience a fundamental change in the rules o f behavior and in the accompanying attitudes of employees (Das-Gupta and Mookherjee 1998). Moreover, societies sometimes experience fairly rapid change from one political order to another. One example i s the European colonization o f Africa, which changed the top-level political order in a couple o f decades in many colonies. The collapse of communism i s o f course another example. 1. In the NE framework, corruption i s a manifestation o f institutional failure, and i t s remedy i s sought mainly in strengthening certain institutions (and perhaps in weakening certain others). One part o f institutional analysis i s to describe a set of institutions that would make an economy or a society function well. The other (and more difficult) part i s to determine how institutions have changed through history and how interventions might help to move them in a desirable direction. In this latter endeavor, we do not have precise theorems, but rather generalizations about pattems o f development observable in history. This task i s made more difficult by the fact that institutions often change slowly and there i s an element o f path dependence influencing the manner in which institutions change. In other words, where a society has been influenced not only by the current situation, but also where i t w i l l be and how i t w i l l change in the future. Annex 8 (continued) 70 m. An example o f the effect o f institutions on an economic system arises in the discussion of transaction costs. Transaction costs are defined as the costs, other than price, which are incurred when trading goods and services and encompass the costs of using the mechanisms o f production and exchange. Thus, transaction costs can be viewed as frictions in the economic system, and are affected by the institutions in the environment. In particular, a system characterized b y corruption can suffer from inadequate mechanisms o f contract enforcement, weak judicial systems, and inadequate provisions for public safety. This in turn w i l l raise the transaction costs in the economy, raising the cost o f investment and the cost o f public service delivery. Alternately, an economy that reduces the level o f corruption can reduce the level o f transaction costs. 71 Annex 9 Annex 9: List of Projects Analyzed I Country I Project I FY I project^^ I Albania Structural Adjustment Credit Project 1999 55160 Albania Public ExpenditureSupport Credit 1999 65825 Albania Legal and Judicial Reform Project 2000 57182 Albania Public Administration Reform Project 2000 69939 Albania Emergency Road Repair 2000 68853 Albania Water Supply Rehabilitation 2000 6649 1 Albania Trade and Transport in SE Europe Project 200 1 70078 Algeria Budget Systems ModernizationProject 2001 6492 1 Argentina Provincial Reform Adjustment Loan-Tucuman 1998 6006 Argentina Provincial Reform Adjustment Loan-Rio Negro 1998 5 1569 Argentina Provincial Reform Adjustment Loan-Salta 1998 51693 Argentina Provincial Reform Adjustment Loan-San Juan 1998 5 1694 -~ Argentina Model Court Development Project 1999 507 13 Argentina Social and Fiscal National ID System Program 1999 55461 Argentina Special SAL 1999 62992 Argentina Catamarca Provincial Reform 2001 044447 Argentina Cordoba Provincial Reform Adjustment Loan 200 1 68344 Argentina Santa Fe Provincial Reform Adjustment Loan 2001 699 13 Armenia Judicial Reform Project 200 1 57838 Bangladesh Legal and Judicial Capacity Building Project 2001 448 10 ~ Bangladesh Export Diversification 200 1 44810 Benin Public Expenditure Reform Adjustment Credit 200 1 61577 Bo1ivia Public Sector Management 1999 62790 Bolivia Enhanced HIPC Initiative 2000 56368 Programmatic Structural Adjustment Credit for Bolivia 200 1 068 134 Decentralization Program Brazil Rio de Janeiro State Reform Privatization Project 1998 39197 Annex 9 (continued) 72 Country Project FY Project ID Guatemala Judicial Reform 1999 47039 Guatemala IntegratedFinancial Management Project 1999 48657 Guinea Capacity Building for Service Delivery Proiect (Vol. 1) 2000 49716 Uttar Pradesh Fiscal Reform and Public Sector Restructuring India Credit and Loan 2ooo 6547 1 India Karnataka Economic Restructuring Loadcredit Project 2001 055490 Indonesia SAL - 2nd Policy Reform Support Loan 1999 60258 IDecentralizedAgric. & ForestryExtension I 2000 I 59930 1 Jordan Third Economic Reform and Development Loan 1999 45676 Jordan Public Sector Reform Loan 2001 66121 IKazakhstan ILeeal Reform Proiect I 1999 I 46046 I Kenya Economic and Public Sector Reform Credit Project 2001 069501 Korea Second Structural Adjustment Loan 1999 56521 Latvia Public Sector Management - State Revenue 1999 55585 Latvia Programmatic Structural Adjustment Loan 2000 66153 Lebanon Community Development Project 2001 71113 Madagascar Second Health Support 2000 51741 Maldives Third Education and Training Project 2000 55944 ~ Mexico DecentralizationAdjustment Loan 2000 66867 Mexico Structural Adjustment Loan to the Estado de Mexico Project: 2001 070479 Moldova Third Structural Adjustment Credit Project 1999 61496 Morocco Legal and Judicial Development Project 2000 I 63918 Mozambiaue Enterprise Development Proiect (Vol. I) 2000 I 49874 Nepal Road Maintenance 2000 45052 Nicaragua Financial Sector Adj, Credit 1998 7777 Public SectorKivil Service - Economic Management Technical Nicaragua 2ooo 49296 Assistance Project (V01.l) Nicaragua Second Basic Education 2000 50613 Niger Public Finance Recover Credit 2001 69568 Pakistan Structural Adjustment Loan 1999 59323 Pakistan Structural Adjustment Credit Project 200 1 71463 Peru Urban Property Rights Project 1999 39086 Public Sector Management - Social Expenditure Management Philippines 2ooo 65113 Project (Vol. 1) Philippines National Road Improvement 2000 39019 ~ Russia Treasury Development Project 2000 64508 Sao Tome and Principe Public Resource Management Credit and TA Project 2001 69909 lsierra Leone IEconomic Rehabilitationand Recovery I 2000 I 35637 I S r i Lanka Legal and Judicial Reforms Project 2000 I 44809 S r i Lanka ~~ Central Bank Strengthening 2001 I 71131 Tajikistan Primary Health Care 2000 49894 Tanzania Privatization and Private Sector DevelopmentProject (Vol. 1) 2000 49839 Tanzania Public Sector Management Reform Program Project (Vol. 1) 2000 60833 73 Annex 9 (continued) Country Project FY Project ID Thailand Economic Management Assistance Project 1998 54799 Thailand Public Sector Reform Loan Project 1999 56522 Thailand Economic and Financial Adiustment Loan I 1999 54801 Thailand Economic and Financial Adjustment Loan I1 1999 58536 Tunisia Water Sector Investment Project 2000 70250 Uganda Poverty Reduction Support Credit 2001 50438 Venezuela Supreme Court Modemization Project 1998 44325 Venezuela Public Sector Management - Public Expenditure 1999 57601 lYemen ]Legal and Judicial Development Project I 1999 I 57915 I Yemen Public Sector Management Adjustment Credit Project 1999 62837 Yemen Civil Service Modemization 2000 50706 ~~ Zambia Public Sector Reform and Export Promotion Credit 1999 35641 Zambia Public Service Capacity Building Program 2000 50400 Zimbabwe Land Reform Support 2000 63096 ~ Zimbabwe Third SAC 2000 3334 75 Annex 10 Annex 10: Country Case Studies 1. This annex provides a more detailed account of Bank-supported interventions-lending and non-lending-in the six countries. The program o f each country i s rated on i t s “relevance” and “efficacy” for fighting corruption. Definitions: Relevance and Efficacy 2. Relevance: OED defines the relevance o f projects as “the exte,nt to which the project’s objectives are consistent with the country’s current development priorities and with current Bank country and sectoral assistance strategies and corporate goals.” Extending this definition to anti-corruption programs, they are judged to be relevant if they have the potential to achieve their objectives given a country’s existing institutional and policy environment and the Bank’s corporate goals. B y “relevance” we mean evidence o f a logical relationship between the interventions supported by the Bank and factors that are known or reasonably believed to be contributing to corruption in the country. Highly relevant interventions would, in other words, be well targeted to known or imputed causes o f corruption. The more relevant an intervention, the higher the likelihood that the intervention w i l l have an effect (see below) on corruption. 3. “Relevance” also includes the idea that, all things being equal, anti-corruption interventions should be targeted to forms o f corruption that are known or believed to be doing the most harm to social and economic development. A full understanding o f “relevance” therefore ’ incorporates two criteria: relevant interventions are well targeted to known or imputed causes o f corruption and to forms o f corruption believed to be most harmful. For practical reasons, the second criterion i s often not fully applicable. Our knowledge o f the relative impacts o f different kinds o f corruption i s s t i l l so limited that it i s rarely possible to say with assurance that certain kinds o f corruption are significantly more harmful than others. Furthermore, even in cases where we feel reasonably sure about the relative harmfulness o f certain types o f corruption, i t may be unfeasible (due to local political factors) to attack the most harmful kinds o f corruption. This evaluation therefore bears in mind the desirability o f attacking the most harmful types o f corruption, but puts more weight on the first criterion regarding the targeting of interventions to any type o f public-sector corruption. 4. Efficacy: OED defines efficacy “the extent to which the project’s objectives were achieved, or expected to be achieved, taking into account their relative importance”. In our present context, therefore, by “efficacy” we mean the extent to which the anti-corruption activities supported by the Bank achieved or are expected to achieve the objective o f either reducing the incidence o f corruption or affecting contributory factors that are directly associated with corruption. The efficacy question i s o f the form: “Did the anti-corruption activities supported by the Bank lead either to reduced levels of corruption, or a reduced presence o f factors contributing to corruption, or an increased presence of factors inhibiting corruption? ” 5. “Relevance” and “efficacy” are rated as “high,” “substantial,” “modest” or “negligible.” Some lending interventions were not complete at the time that this evaluation was prepared. In those cases our judgments about efJicacy are based on assumptions regarding the most probable outcome. Exogenous political and other factors may affect negatively the outcome o f even the most highly relevant anti-corruption interventions, thereby resulting in potentially differing relevance and efficacy ratings. Annex 10 (continued) 76 Guatemala 6. Relevance: The Bank’s program in Guatemala was targeted to the following factors that affect corruption (as described in chapter 5): 0 weak rule o f law 0 weak institutions o f accountability 0 inadequate information about public-sector finances and operations, andor the performance o f public servants 0 opportunities for corruption created by state intervention in economic activity 0 weak law enforcement, leading to a “culture o f impunity” Lending Date (FY) Judicial Reform 99 Reconstruction and Local Development 99 Social Investment I1 99 Integrated Financial Management I1 98 Rural and Main Roads 98 Tax Administration 98 Private Participation in Infrastructure 97 Reform of Basic Education 97 Non- Lending CAS 1998 7. “Combating corruption” i s identified as a specific objective o f the Judicial Reform project. I t emphasizes judicial independence and promotion o f ethical behavior in the judiciary. I t includes measures aimed at improving resource allocations to the judicial sector and addressing capacity gaps. The project aims to build a stronger culture o f professionalism in the judiciary and create systems increasing the likelihood that judicial misbehavior w i l l be detected and punished. 8. The Integrated Financial Management project aims at reinforcing the Comptroller General’s Office, strengthening public investment functions, and modernizing public debt management. The project has a clear anti-corruption orientation, as it aims to strengthen an important institution o f accountability (Controller General) and to make i t more difficult for corrupt activities to go undetected by strengthening financial management systems that produce and disseminate information about public financial management. The project also seeks to create a basis - through provision o f better information on government operations - o f strengthening the accountability o f public servants leading to improved delivery of public services and public programs. 9. Anti-corruption objectives are not central to the Tax Administration project. The project i s however relevant to an anti-corruption agenda because i t aims to increase the transparency and enforcement in relation to tax administration - thereby reducing opportunities for collusion between citizens and tax collectors. The Private Participation in Infrastructure project aims, among other things, to curtail opportunities for corruption by reducing the state’s direct involvement in provision o f basic public infrastructure. 10. In the (on-going) Social Investment I1and Reconstruction and Local Development Projects, the Bank promoted the development of education, basic health, water and sanitation, and access roads by enlisting the active participation o f the beneficiary communities themselves. 77 Annex 10 (continued) 11. Overall, the relevance to corruption o f the Bank‘s program in Guatemala was substantial. 12. Efficacy: The efficacy o f the Bank’s work in Guatemala i s likely to be compromised by powerful factors in the broader political and institutional environment that are beyond the reach o f interventions supported b y the Bank. 13. For example, reforms to the judicial sector are likely to have only a limited effect on corruption in the absence of reforms to the criminal code. Existing sanctions for those found guilty o f corrupt activity are weak, and, in any case, only a limited range o f behaviors that would normally be considered “corrupt” are actually covered by the criminal code. Finally, judicial reform i s incapable o f reaching much corruption within the Guatemalan administration, as Guatemalan legal doctrine denies jurisdiction to the judiciary in these areas and the law does not include provisions for how such cases should be pursued. However, since criminal law i s an area that the Bank generally does not operate, the Bank can direct Guatemala toward external partners, should the government be willing to address this issue. 14. In any case, the judicial system w i l l not have access to corrupt public servants until they are apprehended and brought to trial. But mismanagement in the public sector i s such that the vast majority of misdeeds w i l l continue to go undetected, even with anticipated improvements in information flows expected to result from the Integrated Financial Management project. Table A10.2: Factors Affecting Corruption Addressed by the Bank in Guatemala 15. The Integrated Financial Management project shows strong potential to have a significant impact on corruption, but continuing institutional weaknesses cast doubt on whether the potential w i l l be realized. The IFM project generates information that could be used to identify and prosecute corrupt public officials, but whether in fact that information would be used effectively as a weapon against corruption remains an open question. The Comptroller General, the key actor in the system who would have to be aggressive in using the information generated b y the IFM project, i s under-funded, under-staffed, and regarded as having relatively little power and Annex 10 (continued) 78 influence within the public administration. In addition, as noted above, the current legal and punitive regime governing corrupt behavior by public officials i s badly in need o f strengthening. This undermines the usefulness o f lFh4 information as a corruption-fighting tool. 16. One sees a similar pattern in relation to the Bank’s support for reform o f tax administration. Improved monitoring and transparency o f tax administration are positive developments, but their ultimate impact will be affected by limited capacity for judicial enforcement. Regarding the Private Participation in Infrastructure project, i t s efficacy has been limited given the significant opposition in the country from various quarters, and concems about bid rigging and corruption in the privatization process. 17. The Social Investment I1and Rehabilitation and Local Development projects have some weaknesses in design that may limit their effectiveness in having any long-term effect on corruption in service delivery. For example, even though they require community involvement, the projects by-pass local governments by using independent (but central) agencies (Fondo de Inversion Social - FIS and Project Coordination Unit - PCU) to appraise and monitor the community sub-projects. This limits the ability of the projects to have any impact on the country’s institutions. 18. Other factors that are likely to dampen the ultimate efficacy o f Bank-supported anti-corruption initiatives in Guatemala include: 0 Questionable ownership. The government has so far failed to show consistent signs o f a high level o f ownership of the anti-corruption agenda. 0 No progress in civil service reform The continuing presence o f a poorly managed and politicized civil service i s a major obstacle to effective implementation o f anti-corruption activities. 19. On the other hand, there i s a significant bright spot in the broader institutional environment. The independent press enjoys credibility among the population and may perhaps be the country’s most effective institution o f accountability. I t w i l l undoubtedly have a positive influence on the ultimate outcome o f anti-corruption activities supported by the Bank.35 20. Overall, the efficacy o f the Bank’s program in curtailing corruption in Guatemala was modest. Kenya 21. Relevance: The Bank’s program in Kenya was targeted to the following factors that affect corruption (as described in chapter 5): 0 the widespread and deeply entrenched nature o f corruption 0 weak rule o f law 0 corruption derived from political factors and social fragmentation 0 weak institutions o f accountability 0 inadequate information about public-sector finances and operations, and/or the performance o f public servants 0 opportunities for corruption created by state intervention in economic activity 35 Recently, the WBI has been engaging civil society (with support from the Govemment) through conferences and meetings such a s the Primer Taller Nacional Sobre Transparencia Global (FY03), the meeting on the Anti-Corruption Strategy (FY02), and a module on citizen participation in the recent distance learning event on Controlling Corruption and Judicial Reform. Since these initiatives are so recent, they have not yet been evaluated. 79 Annex 10 (continued) 0 conditions in the civil service that contribute to corruption 0 weak government commitment to fighting corruption 0 corruption in public procurement able A10.3: Bank-Supported Activities Covered in the Kenya Country Study Lending Date (FY) Economic and Public Sector Reform 00 Structural Adjustment 97 Institutional Development and Civil Service 95 Reform Parastatal Reform and Privatization 93 Financial Sector Adjustment 93 Non-Lending “Disengagement” 1998 CAS 1999 22. An implicit objective o f the Institutional Development and Civil Service Reform project was to reduce incentives for corruption in the civil service by providing support for increased salaries. As noted before, research on corruption has cast doubt on the notion that there i s a significant direct correlation between civil service pay and levels o f corruption. Current conventional wisdom holds that in the absence o f a merit culture in the civil service (as i s the case in Kenya), there i s no reliable link between civil service pay and levels o f corruption. The relevance to anti-corruption o f the Bank’s support for civil service pay increases in Kenya i s therefore questionable. Bank Lending Activities Non-Lending Economic Institutional Factors and Public Dev ’p and Parastatal Financial Sector Structural Civil Service Reform and Sector Reform Adjustment Reform Privatization Adjustment Disengagement CAS 23. The Financial Sector Adjustment project included among i t s objectives implementation o f a legal and supervisory regime that would make i t more difficult for managers o f state-owned enterprises to exert improper influence over the banking system for the purpose o f extracting Annex 10 (continued) 80 private gain. The Bank’s assumption was reasonable: using legislation to strengthen an institution o f accountability (the Central Bank) and providing technical assistance to strengthen bank supervision, which would generate more and better information about transactions between banks and state-owned enterprises, ought in principle to have an impact on corruption. The Bank’s support o f the Parastatal Reform and Privatization project was also founded in relevant assumptions about corruption. Given the pervasive nature o f public sector corruption in Kenya, it i s reasonable to assume that reducing the scope o f direct state intervention in the economy would also reduce opportunities for public actors to engage in corruption. 24. The Economic and Public Sector Reform project and the Structural Adjustment project support measures aimed at improving transparency and accountability in public financial management and public procurement. The logic here i s sound. Financial mismanagement and corrupt procurement thrive in environments where timely and reliable information i s not made available, and where proper procedures are not widely published and understood. 25. Possibly, the Bank’s most significant recent anti-corruption intervention in Kenya had to do not with i t s provision o f lending or non-lending services, but rather may have been i t s decision to scale buck i t s support to Kenya. In September 1998 the Bank “disengaged” from Kenya-it switched to a “low case” lending s c e n a r i d u e in part to concerns over public sector corruption. This was a rational response by the Bank to a situation where corruption was endemic, being rooted in political factors and social fragmentation beyond the reach o f technical interventions and abetted by weak commitment o f the political leadership to anti-corruption measures. The Bank returned to a normal (“base case”) level o f operational activity in Kenya in October 2000. 26. Overall, the relevance to corruption o f the Bank’s program in Kenya was substantial. 27. Efficucy: In the period immediately following the Bank’s disengagement, the government launched a series o f anti-corruption measures. Although i t i s not possible to attribute these directly to disengagement, i t i s reasonable to assume that disengagement was an important influencing factor. Actions taken by the government included: e appointment by the government to key public sector posts o f a “dream team” o f six respected experts from the private sector e management changes at the Nairobi City Council, the Port o f Mombasa, the Kenya Coffee Board, and the National Hospital Insurance Fund tabling o f an Anti-Corruption and Economic Crimes Bill development o f new procurement regulations drafting o f a Code o f Conduct Bill for public office-holders launching o f 111 cases against public officials (as o f M a y 2000) by the Kenya Anti-Corruption Authority e assigning additional staff to the Controller and Auditor General’s Office e introducing a code of ethics into the judicial system. 28. In Kenya’s dysfunctional governance environment, the ultimate efficacy o f these positive developments i s highly uncertain and therefore ~ n - e v a l u a b l e .Legal ~ ~ and procedural reforms are necessary, but the impact they w i l l have in a setting where rule o f law i s weak, and the 36 Subsequent events demonstrated the fragility of the reforms. The Kenya Anti-Corruption Authority was disbanded after being ruled unconstitutional by the High Court in January 2001. In May 2001 President Moi fired Richard Leakey, head of the civil service and leading member of the “dream team.” In August 2001, the Kenyan Parliament threw out the Anti-Corruption Bill. T h i s was due in part to anger among opposition MPs that the Bill granted immunity to retired politicians and civil servants (assumed to be cronies of the President). 81 Annex 10 (continued) government’s commitment i s uncertain, remains to be seen. Leverage exercised b y the Bank through disengagement disappeared in 2000 when the Bank returned to a normal lending scenario. Triggers prompting the Bank’s return to the base case were founded mainly on statements o f intent, policy and strategy, rather than concrete actions. Nevertheless, the Bank‘s decision to disengage for two years had the important effect o f stimulating a high-level and open dialogue on public-sector corruption. 29. The efficacy o f the Bank’s lending activities i s mixed. I t s involvement in civil service reform through the Institutional Development and Civil Service Reform project appears to have had n o effect on civil service performance. The reforms supported by the Financial Sector . Adjustment project were equally ineffective. Widespread involvement by Kenyan banks in political scandals was evident after the credit closed. The supervisory capacity (for example, a database o f balance sheets) created through the project was not sufficient to stop corruption in the financial sector when the commitment to using these tools was lacking. The Bank’s involvement in the Parastatal Reform and Privatization project yielded some positive results. The number o f “non-strategic” state-owned enterprises was cut by more than 50 percent, presumably leading to a reduction in opportunities for public officials to engage in corrupt transactions. A variety o f measures supported under the Economic and Public Sector Reform project led to changes in rules and regulations affecting, for example, public procurement and the transparency o f public financial management and priority setting (through implementation o f a Medium-Term Expenditure Framework). There has been some short-term evidence o f impact on procurement: the Minister o f Finance has reported that following implementation o f reforms supported b y the Bank, more than 60 percent o f Tender Board recommendations have been overturned. 30. Overall, the efficacy o f the Bank’s program in curtailing corruption in Kenya is unevaluable due to weak political commitment to initiate and sustain dificult reforms to curtail corruption. 31. Relevance: The Bank’s program in Latvia was targeted to the following factors that affect corruption (as described in chapter 5): the widespread and deeply entrenched nature o f corruption citizens’ perceptions about the legitimacy o f the state weak rule o f law corruption derived from political factors and social fragmentation weak institutions o f accountability inadequate information about public-sector finances and operations, andor the performance o f public servants opportunities for corruption created by state intervention in economic activity conditions in the civil service that contribute to corruption weak government commitment to fighting corruption corruption in public procurement weak law enforcement Lending: As one would expect in a transition country such as Latvia, the emphasis in the Bank’s anti-corruption efforts was on minimizing opportunities for corruption created b y direct state intervention in economic activity. Several projects-Rehabilitation, Enterprise and Financial Restructuring, Structural Adjustment-aimed to reduce the level o f state intervention by supporting privatization. Annex 10 (continued) 82 J Lending Date (FY) Programmatic Structural Adjustment 00 State Revenue Service Modernization 99 Public Sector Reform and Capacity Building (IDF Grant) 98 Structural Adjustment 97 Municipal Services Development 95 Enterprise and Financial Sector Restructuring 94 Rehabilitation Loan 92 Non- Lending CAS 1998 Economic and Sector Work (various) from 1996 33. There was also emphasis on improving the collection and use o f information on public financial management and the performance o f public servants. The Structural Adjustment loan, for example, supported a census o f civil servants and the issuance o f taxpayer identification numbers. The Strategic Compact Anticorruption funds supported a study o f the State Revenue Service that assessed i t s vulnerability to corruption. I t reached conclusions about, among others, the need for an internal audit function and for greater transparency in decision making. The ensuing State Revenue Service Modernization Project addressed corruption in tax and customs administration through support to activities including the development o f rules for transparency in recruitment, implementation o f an internal audit strategy (for anti-corruption), taxpayer registration, development o f an automated tax declaration system, and development o f an audit system. The Programmatic Structural Adjustment project, the first loan in a set o f three operations to enact integrated governance reform, supported strengthened reporting on the operations o f government departments, strengthening of the financial audit function, regular reporting on progress in implementing the national anti-corruption program, and improved accountability for organizational performance throughout the public service. The Programmatic Structural Adjustment project also created a secretariat for the Corruption Prevention Council, strengthened the anti-corruption unit to aid in the implementation o f the anti-corruption law, and brought into the budget all expenditures o f budget-financed institutions. 34. C i v i l service reform was also an element o f the Bank’s program in Latvia. The rationale was that opportunities for corruption are minimized in a well managed, performance-oriented civil service. The Programmatic Structural Adjustment project supported capacity-building and streamlining within the civil service, increased efficiency and transparency, and greater clarity around roles and responsibilities within the civil service. There was also support for human resource management, with a view to improving incentives for recruitment and retention o f high-quality staff. 35. Another objective o f the Bank’s program in Latvia involved strengthening the rule o f law. Several projects supported the drafting o f new laws and regulations or the revision o f existing ones in order to strengthen the judiciary, alter the incentive environment in the public sector, improve transparency in the financial sector, and reduce the vulnerability o f the State Revenue Service to corruption. 36. Non-Lending: The Bank also supported a range o f non-lending activities with anti-corruption objectives. These included workshops, conferences, surveys, and analytic studies, many o f which aimed at building a base o f public knowledge and public debate around questions of corruption in the public sector. This type o f public education activity i s helpful in the Latvian environment, where barely more than a decade has passed since the end o f the Soviet system, and 83 Annex 10 (continued) a common understanding of the role and functioning o f a modem apolitical public service i s not yet well established. 37. Overall, the relevance to corruption o f the Bank’s program in Latvia was substantial. 38. Efficacy: The Bank has made an important contribution to the anti-corruption agenda in Latvia through i t s non-lending activities. The various corruption-related conferences, surveys, and analytical studies supported by the Bank have had an effect on the attitudes o f political and administrative leaders, as well as citizens. The raising o f public consciousness about corruption appears to have created pressure on public office holders to be more accountable for appropriate use of public funds and public authority. This i s likely to pay high dividends for the fight against corruption over the long term. 39. Early evidence suggests that the Bank’s efforts in civil service reform in Latvia have contributed to reducing corruption in the civil service and raising levels o f citizen satisfaction. The Bank’s analytical work on the State Revenue Service (SRS) has had a positive influence on public financial management. The government implemented many o f the Bank’s recommendations for reducing the SRS’s vulnerability to corruption, including the establishment of an anti-corruption unit. More broadly, however, some important gaps remain in cleaning up public financial management. A considerable amount o f public spending remains “off budget” and vulnerable to corruption. 40. A substantial proportion o f the Bank’s lending support was related to privatization. Privatization has advanced more slowly than planned in Latvia, and i t s effect on corruption i s questionable, because the privatization process itself has been tarnished by allegations o f corruption. Annex 10 (continued) 84 41. The Bank's support for judicial reform has yet to yield significant results. The dilapidated state o f the judicial sector, which s t i l l bears the scars o f neglect under the Soviet system, far outweighs the relatively modest level o f support from the Bank. Some promising results are starting to emerge; key judicial decisions are now publicly available, and judges are n o longer allowed to meet privately with litigants. However, the judiciary remains a high-risk area for corruption in Latvia. 42. Overall, the long run efficacy o f the Bank's program in controlling corruption in Latvia i s considered substantial. 43. Relevance: The Bank's program in Pakistan was targeted to the following factors that affect corruption (as described in chapter 5): 0 weak institutions o f accountability 0 inadequate information about public-sector finances and operations, andlor the performance o f public servants 0 opportunities for corruption created by state intervention in economic activity 0 conditions in the civil service that contribute to Corruption 0 corruption in public procurement 0 decentralization-increased accountability to citizens vs. local capture I ble A10.7: Bank-Supported Activities Covered in the Pakistan Country Study ' Lending Date (FY) Structural Adjustment Credit Project 01 Structural Adjustment Loan Project 99 Poverty Alleviation Fund 99 Second Social Action Program 98 Banking Sector Adjustment 97 Fourth On-Farm Water Management 94 Social Action Program 94 Public Sector Adjustment 94 Balochistan Primary Education 93 Sindh Special Development 93 Environmental Protection and Resource Conservation 93 Third On-Farm Water Management 91 Energy Sector Adjustment 91 Agriculture Sector Adjustment 90 Financial Sector Adjustment 90 Agricultural Credit Project 90 Non-Lending Reforming Provincial Finances in the Context of Devolution 200 1 Country Financial Accountability Assessment 200 1 Pakistan Economic Report 1999 Public Expenditure Review 1998 Policy Framework Paper 1998 Country Economic Memorandum 1998 CAS 1995 85 Annex 10 (continued) 44. Lending: The Bank was active in supporting anti-corruption activities in Pakistan well before the formal launching o f the anti-corruption strategy in FY98. A series o f sectoral adjustment loans-Public Sector Adjustment, Financial Sector Adjustment (FSAL), Energy Sector Adjustment and Agriculture Sector Adjustment-aimed at minimizing corruption in relation to the mobilization and allocation o f public resources. 45. In the agricultural sector, the Bank made a small effort to empower small farmers. The Third and Fourth On-Farm Water Management projects attempted to give rights o f water allocation and assessment and collection o f water charges to Water User Associations and Area Water Boards so that small farmers would have improved access to irrigation water. The Agricultural Credit Project intended to target credit toward small farmers and women, although a substantial portion o f the loan was cancelled as a result o f violations to the loan covenant. 46. In the urban sector, the Bank attempted to address factors contributing to political interference in land transactions, tax evasion and revenue leakage. The major intervention in the social sector in the mid-1990s-the Social Action Program (SAP 1)-intended to improve the provision o f education, basic health, population welfare, and rural water supply and sanitation. I t also attempted to address institutional factors linked to failures of accountability and corruption. Among other things, it attempted dealing with corruption by including community and N G O involvement. Through the Participatory Development Program (PDP), NGOs and community groups would be funded to provide some o f the social services as well as help supervise the schools. 47. In i t s lending to Pakistan after the formal launching o f the anti-corruption strategy, the Bank has continued to provide support in relation to governance and corruption. I t has focused on attacking corruption through improvement of transparency in financial management and procurement, strengthening the reporting and auditing functions in the public sector, and building the capacity of state-based institutions o f accountability such as the Office o f the Auditor Annex 10 (continued) 86 General. T w o structural adjustment operations ( S A L and SAC) supported the government’s program to reduce bribery in tax administration, political interference in public expenditure and resource allocation, lack o f accountability in public expenditure management, and corruption in social service delivery.37 48. State capture o f the regulatory and supervisory process, weak accountability, and an inefficient judiciary system (which affects loan recovery operations) are main drivers o f corruption in the financial sector in Pakistan. The Bank attempted to address these through the Banking Sector Adjustment Loan (BSAL) project. In the power sector, the Bank attempted to minimize conditions conducive to corruption by promoting (through the first Structural Adjustment project) establishment o f a competitive electric power system. Efforts to use community and NGO involvement as a counterweight to political interference, bureaucratic inertia, corrupt procurement, and non-responsivenessto citizens were continued in the Second Social Action Program project (SAP I I ), although problems with the project (including corruption) led to i t s early closure. 49. Non-Lending: The Bank’s recent - policy- advice to Pakistan covered privatization, resource mobilization, public financial management, procurement, decentralization, and c i v i l service reform. All o f these areas are, in principle, o f significant relevance to corruption in Pakistan. The quality o f the Bank’s treatment o f these issues, in relation to corruption, has been variable. I t s work in the areas of civil service reform and decentralization has been particularly relevant to drivers o f corruption in Pakistan, however their efficacy has been limited. The report on civil service reform was not followed up with Bank assistance, and the 1996 decentralization report was not shared with the government due to a lack o f consensus within the Bank. 50. Overall, the relevance to corruption o f the Bank’s program in Pakistan was substantial. 5 1. Efficacy: Across the board, the Bank has faced huge difficulties in Pakistan in translating the relevance o f i t s lending and non-lending work into actual effect on corruption or on the factors affecting it. The long tradition o f military rule, as well as constant political instability, make for a volatile governance environment where policy reversals are common and a steady and high-level political focus on anti-corruption objectives i s nearly impossible to secure. 37 SAC I1 was approved (FY02) to continue implementation of the government’s PRSP with its focus on engendering growth, improving governance, and human development. 87 Annex 10 (continued) 52. Major governance problems persist across all sectors. Lack o f financial accountability, complicated and unnecessary rules and regulations, excessive state intervention, and routine exploitation o f citizens by public officials continue to be features o f public life. 53. The privatization process supported by the Bank, which could in principle have had a significant effect on corruption, has progressed slowly due to political instability and an inadequate regulatory framework. In revenue mobilization, corruption appears to have become worse in the most recent decade, due to increased levels o f collusion between enterprises and officials through the misuse o f tax holidays, tax credits, accelerated depreciation allowances, and other exemptions. Income taxes are paid by less than 1.5 percent o f the population and evasion o f the sales tax especially at the manufacturing and wholesale level i s rampant. Attempts to reform the Central Board o f Revenue and make i t more autonomous have failed. 54. In public financial management, the state o f the public accounting system remains abysmal. The government i s unable, for example, to present reliable estimates o f i t s financial position. Internal financial controls have broken down to the extent that accounts are often prepared without appropriate reconciliation in key areas. However, the auditing and accounting functions o f the government have recently been separated and ad hoc Public Accounts Committees created, which may in the future improve accountability. The erosion o f rule o f law in Pakistan has not been reversed. Indeed in recent years corruption in the lower judiciary has increased. The level o f public confidence in the judiciary continues to be very low. 55. Heavy-handed government intervention, particularly in the form o f needless and complicated regulations affecting private sector activity, continues to provide fertile ground for corruption. In the urban sector, the Bank’s attempt to address dysfunctional government accomplished little in terms of correcting distortions in land markets, the property tax, and service delivery. In the social sectors, i t appears that the Social Action Program supported by the Bank actually contributed to mismanagement o f resources and became, itself, a vehicle for corruption. Ghost schools, corruption in recruitment and selection o f sites, and absent teachers were some o f the problems that emerged. The government did not implement reforms supported b y the Social Action Program that were aimed at curtailing favoritism in the posting o f teachers and health workers either. The attempted NGO involvement also resulted in increased corruption, as a number o f fly-by-night NGOs sprung up to take advantage o f project funding. Continued problems eventually led to the cancellation o f part o f SAP 1 1. 56. Despite the Bank’s involvement in aiding Pakistan to develop a framework for a private energy sector and fund the Long Term Credit Fund (LTCF) used to finance private investments, i t was widely believed that corruption tainted the government’s decisions in signing contracts with independent power producers. 57. In the banking sector, the Bank’s adjustment operation did not meet i t s objective o f improving legal processes in order to enhance accountability and the nationalized commercial banks (accounting for 90 percent o f bad loans) s t i l l control most o f the banking assets. In the area of public sector procurement, a major source o f corruption in Pakistan, the Bank’s work has had little impact. I t i s too early to determine the impact o f the recently implemented decentralization reforms. 58. Despite the generally negative picture as far as the outcome o f Bank’s efforts on controlling corruption in Pakistan i s concerned, there are a few promising areas: 0 as part o f reforms recommended b y the Bank, the government has strengthened the autonomy and clarified the responsibilities o f the Auditor General and the Accountant Annex 10 (continued) 88 General, and has established the Office o f the Comptroller General o f Accounts; the government has also entrusted the National Reconstruction Bureau with responsibility for developing a framework for fiscal decentralization; the government has devolved most of the local services to local governments and have created a framework for greater local participation in local decision making. I t has also eliminated the executive magistracy; adoption o f reforms advocated by the Bank, including a medium term budgeting framework, are expected to improve monitoring o f the Public Sector Development Program; recommendations by the Bank led to the restoration o f the autonomy of the federal Public Service Commission. 59. Overall, the long term efJicacy o f the Bank’s program in Pakistan remains unevaluable due to political and policy uncertainties. Philippines 60. Relevance: Activities undertaken by the World Bank in relation to the Philippines addressed the following factors that affect corruption (as described in chapter 5): the widespread and deeply entrenched nature of corruption citizens’ perceptions about the legitimacy o f the state weak rule o f law weak institutions o f accountability inadequate information about public-sector finances and operations, and/or the performance o f public servants opportunities for corruption created b y state intervention in economic activity ,le A10.9: Bank-Supported Activities Covered in the Philippines Country Stu Lending Date (FY) Judicial Sector Reform 02 Land Administration and Management 01 Social Expenditure Management 00 Banking Systems Reform 00 Private Enterprise Support 99 LGU Finance and Development 99 Third Rural Finance 99 Second Rural Finance 96 Economic Integration 93 Philippines Tax Computerization 93 Environment and National Resources Management 91 Financial Sector Adjustment 90 Second Municipal Development 90 Non-Lending Growth with Equity 2000 Combating Corruption in the Philippines 2000 CAS 1999 Report Card on Pro-Poor Services 1999 Country Procurement Assessment 1997 Recent Macroeconomic Developments 1994 Fiscal Decentralization 1993 89 Annex 10 (continued) 0 conditions in the civil service that contribute to corruption 0 weak government commitment to fighting corruption 0 corruption in public procurement 0 enforcement o f laws i s weak, leading to a “culture o f impunity” 0 decentralization-increased accountability toward citizens vs. local capture 6 1. Lending: Bank-supported operations targeted four areas that pose particularly important problems with respect to corruption in the Philippines: state-owned enterprises (SOEs), tax administration, local government, and the financial sector. The Economic Integration project was intended to support the government’s objective o f divesting 75 percent of the value o f i t s corporate assets. The Tax Computerization project addressed key drivers o f corruption in tax and customs administration by increasing transparency and reducing room for exercise o f discretion by officials (through the elimination o f tax exemptions and o f certain customs procedures). 62. At the level o f decentralized local government, the LGU Finance and Development project targeted improved financial management in local administration, thereby taking aim at an important source o f corruption. Also at the local level, the Land Administration and Management project aimed at reducing corruption in land transactions by increasing transparency and accountability in the land titling process, where fake and multiple titles are a persistent problem. 63. In the financial sector, the Banking Systems Reform project aimed, among other things, at strengthening the supervisory framework and raising supervisory effectiveness, increasing the likelihood o f detecting corrupt transactions with measures including the adjustment o f accounting standards and requirements o f greater disclosure. The Economic Integration project aimed at restructuring the Central Bank (an important institution o f accountability in the financial sector) and liberalizing the foreign currency market (where regulations were a corruption driver). Table A10.10a: Factors Affecting Corruption Addressed by Bank Lending Activities in the Philippines I I Bank Lending Activities I Annex 10 (continued) 90 64. The recent Judicial Sector Reform project aims at addressing glaring weaknesses in the legal and judicial system, particularly with respect to the slow administration o f justice. Other operations in areas such as natural resource management, education, social expenditure management, rural finance, and private enterprise credit had less relevance to anti-corruption objectives. 65. Non-Lending: The Bank took advantage in the Philippines o f the potential of economic and sector work (ESW) for building a dialogue with the government and other stakeholders on the subject o f corruption in the public sector. The Bank's ESW in the Philippines, particularly since the formalization of the Bank's anti-corruption strategy in 1997, has focused heavily on anti-corruption issues. The report Combating Corrupting in the Philippines 2000-initiated at the request o f the government-was the product o f an effort by the Bank to understand the nature and roots o f corruption in the country and to propose relevant actions. 66. The report Growth With Equity: The Remaining Agenda positioned anti-corruption and governance improvement measures at the heart o f the country's poverty reduction and development agenda. It highlighted the need in the Philippines for a merit-based public service, strengthened management o f public finances and public procurement, and a more honest and effective judiciary. 67. The Philippine Fiscal Decentralization Study, although prepared four years before the Bank formally launched i t s anti-corruption strategy, raised important issues in relation to dealing with opportunities for corruption created by decentralization. I t identified the increased autonomy and discretion granted to local officials as factors having high potential to affect the quality o f governance. 68. B y contrast, the Country Procurement Assessment Report missed an opportunity to make an important contribution to the dialogue around corruption in public procurement, a major issue in the Philippines. Although the report discussed the scope of corruption in public procurement in the Philippines, it focused on the reform o f procurement practices in Bank-financed projects only. 91 Annex 10 (continued) 69. Overall, the relevance to corruption o f the Bank’s program in the Philippines was substantial. 70. Efficucy: Several recent Bank interventions are expected to have an effect on key types and drivers o f corruption in the Philippines. Although privatization has had i t s problems, the Bank’s support has had a positive outcome. One o f the major flaws was that the government considered a corporation “privatized” when the government divested itself o f a majority o f shares and removed government rules on procurement and salaries, even though the government may s t i l l hold a large block o f shares and (politically) appoint directors to the boards. The Tax Computerization Project, while successful at building hardware and software for tax administration, does not appear to have had a major impact on corruption since i t has been reported that “negotiations” s t i l l regularly take place between tax collectors and businesses. The Bank projects intended to strengthen the banking and financial sector had a positive effect on financial stability at the time o f the Asian financial crisis, but did not address some underlying problems. The pensions systems were s t i l l being used to make political loans and a partially privatized bank suffered recurring difficulties. The Bank’s support for the Commission o f Audit appears likely to yield meaningful gains in strengthening the integrity o f public financial management. The Judicial Sector Reform project also has potential to have an important positive impact on the legal system and the judiciary. 71. The Bank’s lending operations targeting corruption drivers in land administration, local government operations, rural finance and environmental management have not been effective. The Bank’s assistance in building LGU capacity did not make a perceptible difference since the framework promoted to increase oversight and accountability did not receive much attention as evidenced in the lack o f change in local government structures. 72. The Bank’s ESW has been influential in stimulating a high-level, well informed and inclusive dialogue among citizens, the government, and donors on the question of corruption. Moreover, the Bank’s ESW has not only stimulated the dialogue but has also played a key role in shaping the terms o f the dialogue on matters such as a merit-based public service, a reliable financial management system, an honest and effective judiciary, and the urgency o f dealing with corruption at the local-government level. 73. The fact that Combating Corruption in the Philippines was prepared at the request o f former President Estrada, who himself left office under a cloud o f corruption charges, might be cause for skepticism about the value o f the report. But, on the other hand, the report was highly regarded in the Philippines and set in motion a process that led to the government developing i t s own anti-corruption strategy. The level o f dialogue about corruption that was stimulated b y the report i s likely to have had a positive impact that w i l l outlive the credibility problems created by the Estrada crisis. 74. On the negative side, the Bank’s anti-corruption interventions appear unlikely to make a significant effect on corruption in public procurement, which remains a serious problem in the Philippines. 75. Overall, the long term efficacy o f the Bank’s program in curtailing corruption in the Philippines i s rated as substantial. Annex 10 (continued) 92 Tanzania 76. Relevance: The Bank’s program in Tanzania was targeted to the following factors that affect corruption (as described in chapter 5): the widespread and deeply entrenched nature o f corruption weak rule o f law corruption derived from political factors and social fragmentation weak institutions o f accountability inadequate information about public-sector finances and operations, and/or the performance o f public servants opportunities for corruption created by state intervention in economic activity conditions in the civil service that contribute to corruption weak government commitment to fighting corruption corruption in public procurement decentralization-increased accountability toward citizens vs. local capture Table A10.11: Bank-Supported Activities Covered in the Tanzania Country Stuc Y Lending Date (FY) Programmatic Structural Adjustment 00 Public Service Reform Project 00 Parastatal and Private Sector Development 99 Tax Administration 99 Parastatal and Public Sector Reform 93 Financial and Legal Management Upgrading 93 Non-Lending CAS 2000 Public Expenditure Reviews 1994-99 Country Economic Memorandum 1996 World Bank Institute - various activities 1995-96 77. Lending: Tax laws and regulations in Tanzania are complex and archaic, providing multiple opportunities for corrupt officials to extract private benefits from business people by bargaining over tax liability. The Bank attempted to address this significant driver o f corruption through a procedural streamlining component o f the Tax Administration project. The streamlining component was, however, only a relatively minor component o f the project. The rest o f the project was devoted to institutional and technical strengthening o f the Tanzania Revenue Authority (TRA), but these other elements largely failed to target root causes o f widespread corruption in the TRA. 78. The dominance, over an extended period, of state-owned enterprises in the Tanzanian economy has also been a significant cause o f corruption. An early project addressing governance in parastatals was the Financial and Legal Management Upgrading Project (FILMUP) initiated in FY93. The FILMUP attempted to develop the legal sector, improve cost recovery, and increase efficiency in the audit o f parastatals by undertaking studies and improving human and material resources. The Bank continued to address this sector through i t s support for the Parastatal and Private Sector Development (FY99) and Parastatal and Public Sector Reform (FYOO) projects. Both projects provided the government with advice and capacity-building support related to the privatization process. The 1999 project, unlike the 1993 project, also paid explicit attention to combating corruption within the privatization process itself, a highly pertinent issue in Tanzania, 93 Annex 10 (continued) where privatizations in the 1990s were tainted by subjectivity and lack o f transparency in the evaluation o f bidders and ultimate decisions about sales o f state-owned enterprises. 79. Pay reform in the civil service was also an element o f the Bank’s anti-corruption strategy in Tanzania. The Parastatal and Public Sector Reform project provided support for retrenchment and subsequent salary increases for the remaining public servants. As noted, however, the anti-corruption literature suggests that there i s no obvious direct correlation between public-sector salaries and levels o f corruption. Phase 1 o f the Public Service Reform Program i s attempting to increase accountability and realign incentives within the civil service through changes that include making appointments more merit based and conducting annual surveys o f public service delivery. However, this project i s currently under implementation, and therefore it i s too early to evaluate the project’s impact on corruption. 80. The Programmatic Structural Adjustment project was the first Bank lending operation in Tanzania to draw explicit links between governance and corruption. Among other things, it supported the development o f ministerial anti-corruption plans. Many o f these plans had the appearance o f “shopping lists” proposing a wide range o f expenditures that often seemed to have little direct relationship to curtailing corruption. On the other hand, provision b y the Bank o f support for a process whereby ministries themselves developed their own anti-corruption strategies may have relevance as an attempt b y the Bank to build local ownership o f the anti-corruption agenda. The Public Sector Adjustment project also supported the introduction of greater transparency into the public procurement process, an important corruption issue in Tanzania. 8 1. Nun-Lending: The World Bank Institute undertook a range o f anti-corruption activities in 1995/96. These were well targeted in terms o f the need to build a “critical mass” o f government officials motivated to support the fight against corruption. However, the relevance o f Annex 10 (continued) 94 some o f the activities supported by WBI was fairly modest. That i s because Service Delivery Surveys and participatory workshops were aimed at building awareness about corruption but there i s little evidence o f a link between awareness and willingness to intervene against it. 82. A series o f three Public Expenditure Reviews conducted during the 1990s provided thorough and high-quality analyses o f a range o f corruption related factors (although none o f the studies explicitly addressed corruption). A financial accountability assessment prepared in FYOl paid no explicit attention to corruption, although it addressed issues o f public sector management and Parliamentary oversight o f public expenditure that had some relevance to corruption. A Country Economic Memorandum prepared in 1996 had little relevance to corruption. 83. Overall, the relevance to corruption o f the Bank’s program in Tanzania was substantial. 84. Efficacy: The Bank’s work on streamlining tax administration notwithstanding, there remains widespread evidence in Tanzania that taxpayers continue to suffer from arbitrary assessment and harassment by taxation officials. The Tax Administration project left untouched some major determinants o f corruption within the tax system and could have done more to improve preventative measures. The project invests heavily in computerization which may have limited efficacy given the fact that much goes unrecorded. 85. If we take the overall success o f the privatization program (supported by two Bank projects) as a proxy for the impact o f privatization on corruption, then the judgment i s mixed. A significant amount o f privatization was completed; however, only half o f the targeted enterprises have actually been sold. Moreover, the privatization process itself was tainted b y suspicion o f corruption. On the positive side, a recent study o f privatization in Tanzania concluded that the privatized enterprises showed improved profitability and lower employment levels. This i s consistent with the hypothesis that privatization would lead to less waste attributable to corruption and nepotism. 86. The Bank’s support for retrenchment and salary increases in the public sector appears to have had no impact on corruption since the probability o f corrupt c i v i l servants being caught has not at all changed. Similarly, anti-corruption strategies undertaken with support from the Programmatic Structural Adjustment project appear to have had no significant effect, as the action plans suggested expenditures were either not directly related to corruption or were unlikely to be financed (since the planned expenditures amounted to 25 percent o f tax revenues). On the other hand, the measures supported by the same project to support increased transparency in public procurement are likely to have an effect on reducing corruption. 87. There i s no evidence that the workshops and surveys supported b y the World Bank Institute have had an impact on corruption in Tanzania since there i s likely only a weak link ~ ~ is, however, reason to believe that the between awareness of a problem and i n t e r ~ e n t i o n .There Public Expenditure Reviews have had a significant effect. In the case o f the 1994 PER, it appears that the government has acted on approximately 40 percent o f the recommendations contained in that report. 88. Overall, the long term efJicacy o f the Bank‘s program in Tanzania remains unevaluable due to lack o f local ownership and uncertain policy regime. 38 This conclusion was reached by WBI’s own self-evaluation, “EDI’s Anticorruption Initiative in Uganda and Tanzania” ED1 Evaluation Studies (1998). 95 Annex 10 (continued) Summary Conclusions 89. In conclusion, this part o f the study finds that the Bank has done a good j o b o f supporting anti-corruption activities that are relevant to the corruption problem on the ground. We judged that the Bank’s actions in all six countries had “substantial” relevance. However, due to the short time elapsed, the Bank had only modest success in translating i t s good performance with respect to relevance into efficacy. 90. The complexity o f the task at hand and the enormity o f the challenge and the short time period accounts for the gap between relevance and efficacy. The relevance o f the Bank’s anti-corruption actions i s clustered around a set o f factors that affect corruption related to information about public sector operations, state intervention, rule o f law, institutions o f accountability, and civil service reform. These are all major drivers o f corruption, and so i t i s appropriate that the Bank focuses on them. Corruption also derives i t s roots from political factors and social fragmentation. The Bank has no special comparative advantage or expertise in dealing with these latter factors but if the government was receptive, it could ask i t s development assistance partners to support government efforts in those areas, such as when the Bank directed the Latvian Government to the Soros Foundation for assistance in campaign finance reform. The short time elapsed and the skewing o f the Bank’s relevance toward one set o f factors-ven though these factors are appropriate targets for anti-corruption actions-accounts for the Bank’s modest efficacy. A better understanding o f these social and political factors at the country level w i l l however assist the Bank in improving i t s advice on the design o f appropriate anti-corruption interventions. 91. The generally sound relevance of the Bank’s anti-corruption support in the six countries i s a promising start for the future success o f the Bank’s efforts. In order to increase the likelihood of success, the conclusions o f the case studies are that: The Bank’s continuing focus on deepening i t s understanding o f country specific causes of corruption with added emphasis on social, political, and institutional origins o f corruption at the country level w i l l improve i t s policy advice and assistance dialogue with developing countries. e The Bank itself cannot directly affect the roots o f corruption derived from political factors and social fragmentation. However, there are at least two ways corruption can be attacked indirectly, by looking for ways to facilitate actions b y others, often external partners and local non-governmental stakeholders, and b y strengthening the mechanisms that empower citizens to demand greater accountability for service delivery. The Bank’s support o f tools including citizen surveys o f service delivery performance, budget transparency, and e-government allow people to hold government accountable. Country dialogue can be as important as lending operations in the fight against corruption. e Decentralization i s an important instrument for improving accountability and service delivery to citizens when the institutional factors that may lead to local capture are dealt with suitably. 97 Attachment 1 MANAGEMENT RESPONSE TO THE OED REVIEW MAINSTREAMINGANTI-CORRUPTION ACTIVITIES IN WORLD BANKASSISTANCE: A REVIEWOF PROGRESS SINCE 1997 24,2004 MAY 99 Attachment 1 (continued) MANAGEMENT RESPONSE TO THE OED REVIEW ANTI-CORRUPTION MAINSTREAMING ACTIVITIES IN WORLD BANKASSISTANCE: A REVIEW OF PROGRESS SINCE 1997 I. Introduction.............................................................................................................. 1 11. OED Findings .......................................................................................................... 1 111. Management Comments .......................................................................................... 2 IV. Conclusions.. ........................................................................................................... .4 Annex. Management Response to the Recommendations of the OED Review ................. 5 ABBREVIATIONS AND ACRONYMS CAS Country Assistance Strategy CDD Communi ty-Driven Development CDF Comprehensive Development Framework CFAA Country Financial Accountability Assessment CPAR Country Procurement Assessment Report CPIA Country Policy and Institutional Assessment LICUS Low-Income Countries Under Stress OED Operations Evaluation Department PER Public Expenditure Review PRSP Poverty Reduction Strategy Paper WDR World Develonment ReDort 101 Attachment 1 (continued) 1 RESPONSE MANAGEMENT TO THE OED REPORT ANTI-CORRUPTION ACTIVITIES IN WORLD BANKASSISTANCE: MAINSTREAMING A REVIEW OF PROGRESS SINCE 1997 I.INTRODUCTION 1. World Bank Management welcomes the Operations Evaluation Department (OED) review’ and i t s overall assessment that the World Bank has made rapid and substantial progress on addressing corruption in i t s development assistance and has helped to raise the profile of governance issues in the development agenda. The review analyzes developments between FY98 and FYOl. The Bank’s governance and anticorruption agenda i s moving rapidly. Significant progress has been made since June 2000. Notably, the Comprehensive Development Framework (CDF) has been widely accepted and implemented in many low-income countries through the Poverty Reduction Strategy Paper (PRSP) process. This framework i s a potent instrument to address governance and corruption issues as part of a long-term, comprehensive strategy. The update on implementation of the Public Sector Strategy gives a comprehensive overview of the Bank’s intensified efforts to place at the center of i t s assistance strategy measures to strengthen governance and fight corruption. (Regular progress reports on all sector strategies w i l l be provided through annual Sector Strategy Implementation Updates, starting in FY05.) 11. OED FINDINGS 2. The report i s a desk review of Bank initiatives in mainstreaming anticorruption concerns in i t s work. The review also assesses the relevance of and early outcomes from Bank support for anticorruption activities in Guatemala, Kenya, Latvia, Pakistan, Philippines, and Tanzania. The review finds that the Bank has made rapid progress, increasingly integrating corruption concerns into operational activities and internal processes. B y encouraging open debate, carrying out policy research, and disseminating good practice to member countries and international organizations, the Bank has helped to raise the profile o f governance issues in the development agenda. As a result, support for anticorruption work has grown within the development community. 3. Conclusions. The review concludes that the Bank has supported highly relevant anticorruption activities on the ground, but has so far experienced only modest success in achieving durable outcomes. While the Bank’s anticorruption activities have addressed key corruption factors-transparency in the public sector, state intervention, the rule o f law, civil service competence, and capacity building, the review argues that corruption i s ’MainstreamingAnti-Corruption Activities in the World Bank Assistance: A Review of Progress since 1997, World Bank, OED, forthcoming. Attachment 1 (continued) 102 2 grounded in a country’s cultural, social, and political context-conditions over which the Bank has limited influence. The review also states that the Bank does not always have a comparative advantage in long-term capacity development and should therefore rely more actively on development partners and strengthen i t s understanding of social and political factors at the country level. The review further suggests that the Bank should pursue a variety of strategies in different countries to address the most important causes of corruption and should be more creative in i t s choice of instruments, nurturing of partnerships, and sequencing o f operations. 4. Recommendations. The review recommends that the Bank stimulate demand for anticorruption measures in high-corruption countries, continue to assess countries’ governance environments, tighten governance requirements for Bank lending, and focus lending on results. The review also recommends that the Bank provide more assistance for bottom-up reforms in countries where corruption i s entrenched and where governments lack the will or the capability to mount effective anticorruption programs. COMMENTS 111. MANAGEMENT 5. The review focuses on public sector bureaucratic and administrative corruption. A key factor to which i t gives less emphasis i s the role of the private sector, especially the financial sector and the complex interface between the private and public sectors. Recent analytic approaches to corruption focus on understanding the institutions o f influence, including the forces of state and regulatory capture, and reject the notion that some cultures are more corrupt than others. The recognition that complex forces outside of the public sector bureaucracy may play a key role in shaping misgovernance broadens the framework from corruption to governance. Within this comprehensive framework, the Bank’s approach i s to try to unbundle these influences and directly address specific manifestations. The Bank works to help client countries combat corruption through a broad-based governance agenda, including specific and integral measures to fight corruption. 6. Long-Term Activity. The review states that capacity building i s not always the Bank’s comparative advantage in some key areas of work on anticorruption. Management perceives long-term capacity building, strengthening o f institutions, and governance reform as being areas of i t s comparative advantage. These are core activities in most of the Bank’s development assistance strategies, activities undertaken at the request of borrowers, in close collaboration with partners, and often with the active engagement of country office staff (who provide extensive knowledge of cultural and political influences). 7. Sustainability of Results. The review states that the Bank has supported highly relevant anticorruption activities. I t concludes, however, that the Bank has so far achieved only modest success in terms of durable outcomes. Management believes that judgments about the sustainability o f results of the Bank’s anticorruption efforts are 103 Attachment 1 (continued) 3 difficult after only four years, especially since-as the review rightly states-corruption i s most effectively addressed through long-term institutional reforms. 8. Demand for Anticorruption Programs. OED i s concerned that high-corruption countries are the least likely to request anticorruption assistance from the Bank and that Bank assistance, therefore, i s concentrated in countries where corruption i s perceived to be lower. O E D recommends that in high-corruption countries the Bank, where acceptable to public authorities, should try to foster demand for anticorruption programs among a wide spectrum o f stakeholders. Management notes that the Bank i s actively engaged with a variety o f stakeholders through i t s support for countries’ own participatory, consultative approaches, especially the CDF framework as implemented through the PRSP process, and through participatory approaches of public service delivery projects in particular. As OED notes, this involvement i s exemplified by the fact that nine of eleven PRSPs during the review period identified governance as a major pillar. The program for Low-Income Countries Under Stress (LICUS), adopted since the review period, supports the creation of demand for sound governance. These processes help foster awareness and demand for anticorruption programs. 9. Governance Prerequisitesfor Lending. Management agrees that i t i s important to exercise selectivity in lending to countries with weak governance and high levels of corruption. The Bank has a multifaceted approach to adjust lending levels: CAS triggers often relate to the degree of implementation of reforms in governance standards, transparency, and accountability. In response to corruption risks, lending volumes are frequently reduced or, in extreme cases, suspended. (The review highlights cases where this has occurred.) For IDA countries, a governance discount based on the CPIA indicator i s used as an important factor in determining the level o f IDA allocations. Every CAS i s required to incorporate a careful diagnosis o f governance conditions, including corruption and public financial accountability issues, their impact on the country strategy, and the risks they pose to Bank Group activities. The most recent CAS retrospective shows compliance with this requirement. Public Expenditure Reviews (PERs), Country Financial Accountability Assessments (CFAAs), and Country Procurement Assessment Reports (CPARs) now contain action plans to deal with governance issues and provide the analytical underpinnings for operations, which, where appropriate, pick up fiduciary components coming out of this analysis. 10. Additional Prerequisites. The OED review proposes the establishment of minimum governance prerequisites, including governance assessments. The Bank’s core diagnostic reports already include PERs, CFAAs, and CPARs, which provide important insights into the rules and procedures, transparency, and accountability of countries’ public sector institutions and financial management practices. While recognizing the value of government assessments in appropriate circumstances, Management does not intend to mandate their preparation. 11. Bottom-Up Reforms. Management agrees on the importance of bottom-up reforms grounded in participatory processes with stakeholders and beneficiaries. Complementing extensive assistance to subnational governments, our Community-Driven Attachment 1 (continued) 104 4 Development (CDD) agenda i s one instrument to achieve this objective. (The report did not include a review of the Bank’s CDD portfolio and engagement.) Community-based development processes, often focused on service delivery, are an important instrument to encourage participation, openness, voice, transparency, and accountability. Under the CDF framework, notably as exemplified through the PRSP process, the Bank i s working with client countries to experiment with C D D approaches to address governance issues, with lending operations in a variety o f sectors, medium-term public expenditure processes, and components of governance and anticorruption programs. Again, this bottom-up approach i s a feature of support to LICUS countries. 12. Focus Lending Instruments on Accountability for Results. The review suggests that linking development assistance to output achievements in the public sector would significantly reduce corruption levels in the public sector. The Bank i s working to define processes for measuring results and attributing them to development assistance, and i t intends to introduce better outcome measurement in i t s programs. The 2004 World Development Report2 (WDR) focused on alternative service delivery mechanisms, tailored to country circumstances, that enable the poor to effectively monitor and hold accountable service providers. The Bank i s working to support client countries in using these accountability mechanisms to improve the delivery of services that are the responsibility o f the public sector. But simply focusing on agreed outputs may not be sufficient for effectively reducing corruption, as there can be significant levels of corruption even when agreed outputs have been achieved. As noted in footnote 32 o f the OED review, while corruption i s often an obstacle to effective public service delivery, i t i s a more generic impediment to equitable growth through i t s “state capture” impacts and i t s effects on the private sector investment climate. A singular focus on public service output delivery may not be sufficient to address these systemic corruption issues. However, the Bank i s committed to linking i t s assistance to better public service delivery. Through i t s established legal instruments i t i s well positioned to address corruption in the implementation o f Bank projects. IV. CONCLUSIONS 13. The Bank i s firmly committed to continuing and strengthening i t s focus on governance and anticorruption and Management welcomes this review as a contribution toward this end. Effective development rests on efficient, effective, transparent, and accountable private and public sector institutions. As the OED review emphasizes, the Bank-through i t s analytic and policy assistance agenda, and in close coordination with partners-has made an important start and i s deeply engaged in this rapidly evolving agenda. The Bank w i l l continue to accumulate experience and adjust i t s support for country anti-corruption activities on the basis o f lessons learned. See World Development Report 2004: Making Services Work for the Poor. Washington, D.C.: World Bank. 2003. 105 Attachment 1 (continued) 5 RESPONSE MANAGEMENT TO THE RECOMMENDATIONS OF THE OED REVIEW ANTI-CORRUPTION ACTIVITIES IN WORLD BANKASSISTANCE: MAINSTREAMING A REVIEW OF PROGRESS SINCE 1997 M a j o r monitorable OED recommendations Management response requiring a response 1. Deal with the demand dilemma by fostering Management agrees that, when acceptable to demand among a wide spectrum o f public authorities, the Bank should assist in stakeholders. The Bank’s current approach creating demand for transparent and accountable targets anticorruption support to countries that institutions and processes and specific demand such assistance since ownership i s anti-corruption measures. This i s achieved critical in fighting corruption. Unfortunately, through the C D D programs; through the the countries (regions and sectors) most in need participatory approaches supporting country-led o f “anticorruption” assistance tend to be the CDF framework preparation, notably as least interested in receiving. Where acceptable embodied in PRSPs, through capacity-building to public authorities, the Bank should deal with programs aimed at improved transparency and this problem by fostering demand among a wide access to information on public budgets, spectrum o f stakeholders in those countries. resources, and decisions, involving a variety o f stakeholders. I t i s also an important feature in the LICUS approach in using support for “zero generation” reforms to create a stronger demand for good governance. 2. Continue assessments o f governance Management agrees that the Bank should environment. The Bank should continue to continue to assist clients in assessments o f their assist clients in carrying out assessments o f their institutional environment. Completion o f a PER, institutional environments, including a CFAA, and CPAR are part o f core ESW diagnostic of the main causes o f corruption. I t requirements for each country, and now include should help develop a long-term, holistic and action plans. IGRs are an additional optional sequenced reform program, including an instrument. The Bank i s increasingly studying anticorruption plan. The Bank and other donors the processes o f public sector reforms in client should work with the government in identifying countries, since a better understanding o f the the areas where they can best help in drivers o f change w i l l help the client and the implementing these reforms. The CAS should Bank design reform programs with greater articulate the Bank’s role. chances o f success, particularly during implementation. Holistic approaches are pursued in countries with sufficiently broad based commitment across stakeholder groups. The Bank works with more focused analysis and interventions if opportunities for change exist only in selected sectors. CAS documents regularly entail an analysis o f the governance agenda and the Bank’s role. Attachment 1 (continued) 106 6 Major monitorable OED recommendations Management response requiring a response 3. Define the “governance” prerequisites for The Bank agrees that weak institutions, weak lending. Corruption poses significant risks for governance, and corruption pose significant risk the effectiveness o f Bank lending, especially to development effectiveness. The Bank adjustment lending, and therefore the Bank Exercises selectivity in lending through a should set minimum governance prerequisites multifaceted approach: CAS triggers often relate for lending, including governance assessment to the degree o f implementation o f governance and anticorruption plans. This should lead to standards, transparency, and accountability greater selectivity in lending and increased reforms. Lending volumes are reduced in reliance on advisory activities in high corruption response to corruption risks. For IDA countries, environments where there i s limited a “governance discount” based on the CPIA i s anticorruption demand. Some governments may used in determining the level o f IDA allocations. be reluctant to engage the Bank in dialogue on Every CAS i s required to incorporate a careful “corruption” issues, because they believe that diagnosis o f governance conditions, including this i s not needed or helpful. In such cases, the corruption and public financial accountability Bank may need to make an independent issues, their impact on country strategy, and the assessment, to be reflected in the CAS, o f risks they pose to Bank Group activities. whether “corruption” would impede the Compliance with this requirement has been effectiveness o f lending, especially adjustment good. ‘All CAS documents approved in FYOO and lending. FYOl included diagnosis o f governance and risks o f corruption and more than 90 percent o f CASs integrated anticorruption into TA and lending. The Bank’s core diagnostic work includes a PER, CFAA, and CPAR for each borrowing country. These analytical pieces provide important insights into the effectiveness, transparency, and accountability o f public I institutions. While recognizing the importance o f governance assessments in particular circumstances, Management does not wish to mandate the preparation o f governance assessments as an additional piece o f required core diagnostic work. 4. Focus lending instruments on Management agrees with a stronger focus on accountability for results. A sharper focus on measurement o f results and outcomes and has accountability for results would help create implemented a management for results agenda. appropriate incentives and promote public sector In addition, WDR04 presented options for reform. Efforts to improve service delivery can improving the ability o f the poor to monitor and help address corruption and i t s causes. Bank hold accountable those responsible for service should increase its support for reforms in service delivery, which the Bank i s pursuing in i t s delivery to emphasize probity and results rather support to client countries. The Bank also goes than only the level o f expenditures. Linking beyond a focus on service delivery. Systemic development assistance to demonstrable corruption can prevail even in cases where improvements in output performance o f the agreed outputs have been achieved. Such public sector, especially in areas most relevant to systemic corruption needs to be addressed as i t the poor, may be the most effective characteristic poses significant risks to the development o f the of the Bank’s evolving anticorruption strategy. private sector, financial sector, and rule o f law. 107 Attachment 1 (continued) 7 Major monitorable OED recommendations Management response reauirinp a resvonse 5. Provide more assistancefor bottom-up Management agrees that assistance for reforms. In many countries where corruption i s bottom-up reforms i s an effective instrument to entrenched, governments lack either the w i l l or amplify citizen voice, which enhances pressures the capability to mount effective anticorruption for improved transparency, accountability, and programs. In such countries, the Bank and/or i t s rule o f law. The Bank uses CDD programs and partners may choose to amplify citizens’ voice support for the CDF approach, notably as and strengthen exit mechanisms so as to implemented through the PRSP process, and enhance transparency, accountability, and the intends to further strengthen participatory rule o f law. programs in i t s governance agenda, mainly through these instruments. As noted above, support for “zero generation” reforms i s a key element o f the LICUS initiative. 109 Attachment 2 Chairperson’s Summary Mainstreaming Anti-Corruption Activities in World Bank Assistance A Review of Progress Since 1997 - CODE Subcommittee Meeting M a r c h 17,2004 On March 17, 2004, the Informal Subcommittee to Committee on Development Effectiveness (CODE) met to discuss “Mainstreaming Anti-Corruption Activities in World Bank Assistance: A Review o f Progress since 1997” along with the Draft Management Response. Comments bv OED. OED remarked that the evaluation examined early progress in implementing the Bank’s anti-corruption strategy. The strategy was launched in September 1997 and the evaluation covered activities through 2001. The evaluation concluded that the Bank’s understanding o f the causes and dynamics of corruption in different country settings had improved and the Bank’s efforts had raised the profile o f governance issues on the development agenda. Given the limited time frame, the Bank had made only moderate progress in achieving durable outcomes. The evaluation also found that a better understanding o f social and political factors would enhance the quality and impact o f Bank advice and improve the design o f the Bank’s anti-corruption interventions. The evaluation had a number o f recommendations, including to involve development partners more fully, engage c i v i l society more directly, and foster public-private sector interfaces. More diagnostic work was recommended to assess the governance environment, and the Bank needed to define more clearly the governance pre-requisites for lending in individual countries. And finally, OED suggested the Bank concentrate more o f i t s efforts on supporting bottom-up reforms. In this regard, the evaluation suggested that engaging civil society in the monitoring o f service delivery would help involve them more actively in anti-corruption activities, increase the results focus o f Bank programs, and enhance the demand for the Bank’s involvement. Comments bv Manapement. Management welcomed the evaluation noting that though many challenges remain, much progress has been made since 2001. Achievements include strengthening the CPIA to look more deeply at governance issues, integrating governance as a regular part o f the CAS exercise, core diagnostics that address governance, and using IDFs to work on governance issues. Management further remarked that an internal evaluation done by Management complements OED’s work and looks at the period beyond 2001. The internal evaluation shows that there has been a tremendous increase in the awareness o f governance related issues as a result o f the Bank’s work and major programs have been launched in support o f i t s goals. However, more impact i s needed on the ground and more work i s needed to foster demand, deal with powerful vested interests, entrenched cultural norms, strong political motivations, and mainstreaming anti-corruption activity across sectors and countries. Management stressed that these are deep seated issues which require a long term commitment from the Bank and i t will take time to see real progress. Management intends to continue to strengthen the Bank’s engagement with civil society, maintain the high profile o f i t s work in this area, and build partnerships with others who may have a greater comparative advantage in this area. Main Conclusions and Next S t e m The Subcommittee generally welcomed the OED report and Management’s Response and stressed the importance o f the Bank’s fight against corruption as a critical pillar to sustainable development. However, some members felt that the Bank was overstepping i t s mandate and entering into social and political realms in client countries. The Subcommittee acknowledged that it would take time to see an impact given the deep seated nature o f these issues. They welcomed the work done by the Bank to bring governance issues to Attachment 2 (continued) 110 the forefront o f the global development agenda as well as adjustments made (i.e. core diagnostics) to the Bank’s own practices to address these issues. The Subcommittee emphasized that a dilemma remains fostering demand and creating a constituency for governance reforms and generally supported the evaluation’s recommendation o f broadening the spectrum o f stakeholders and engaging more intensively with civil society as appropriate. There was some disagreement about defining governance pre-requisites for lending. Some members supported this recommendation but stressed that clarity was needed on defining the pre-requisites, while others rejected this recommendation and suggested that the Bank should focus only on indirect mechanisms like finding and working with reform minded sectors o f the Government. Among the specific issues raised by the Subcommittee were: Impact. The Subcommittee generally accepted that it was early to see real impact o f the Bank’s work in this area. However, some members asked for more specific information about the extent o f the Bank’s impact on governance issues. Some members suggested that there was not ample impact for the resources spent on these activities, and rather than following holistic strategies, the Bank needed to find pockets o f reform and work in these areas. Another member asked what an appropriate time frame would be to truly assess impact and whether any further evaluative work was planned in this context. Fostering Demand. The Subcommittee agreed that fostering demand for governance work was the main challenge faced by the Bank. Several members asked for specifics on how Management proposed to address this issue, and stressed that innovative instruments focusing on accountability and results needed to be developed. Bank’s Mandate. Some members noted that corruption i s usually fostered b y socio-economic conditions over which the Bank has no control. They also stressed that the Bank should not be engaging in social and political realms in client countries. In this regard, they questioned how progress would be made in this area given the l i m i t s o f the Bank’s mandate. Lending Pre-Requisites, Some members disagreed with the idea o f setting lending pre-requisites and said this would lead to more conditionality in an era when the Bank was supposed to be simplifying and modernizing i t s approaches. Other members supported this notion, but stressed that such pre-requisites should be clearly defined and applied in a systematic way. One member proposed a system comparable to the governance discount used for IDA countries. Partnerships. Some members agreed with the notion o f building partnerships with c i v i l society, the private sector, and others to make progress in this area. However, one member questioned the relevance o f these entry points given the cultural and institutional elements o f corruption while others questioned the legitimacy o f c i v i l society organizations. Another asked how client countries would be involved as partners in the process. The role o f the private sector was particularly stressed by two members. Staff Incentives. Two members asked what incentives are being created for staff to work in this area. Rosemary Stevenson, Chairperson