THE WORLD BANK GROUP A World Free of Poverty

Helping Countries Combat Corruption: The Role of the World Bank

contents / previous / next

7. Mainstreaming a Concern for Corruption in Bank Activities

If the Bank is to help countries tackle corruption, it needs to take corruption more explicitly into account in its own economic work and in its internal processes. This chapter sets out how the Bank is moving forward in mainstreaming a concern for corruption across the range of its operational work. From the Bank's perspective, anticorruption is best thought of not as a set of new initiatives but rather as a more explicit integration of the problem (of which staff and management have long been aware) into country strategy formulation, Bank lending, economic and sector work, research, and country dialogue.

Economic work

Increasing transparency.   The first requirement in mainstreaming a concern for corruption is for the Bank to bring the issue out of the dark. To a significant extent this is already happening. Discussions about the effects of corruption on development have been intensifying within the Bank in recent years, especially since President Wolfensohn's 1996 Annual Meetings speech. But corruption needs to be addressed more systematically in the Bank's analytical work. Until that happens corruption will not be sufficiently understood or integrated with management decisions. And to some extent the Bank is playing "catch-up," acknowledging that corruption threatens development but doing so in the wake of media attention and international recognition of corruption as a major problem.47

Economic and sector work.   The Bank will not be able to provide governments with sound policy advice if it does not make the required intellectual investment. Economic and sector work has not dealt with corruption much, referring to it elliptically with phrases like "rent seeking," "leakage," and "misappropriation." In the future more effort is needed to address corruption directly and openly in economic and sector work when it is relevant. In particular, poverty assessments will need to discuss petty corruption, since in many countries the poor experience extortion by public officials daily. Public expenditure reviews, with their focus on the allocation of government spending, will also have to grapple with the issue, since corruption may mean that inputs are used to produce outputs and have outcomes quite different from those represented in budgets approved by the legislature. And country economic memorandums, sector studies, and environmental assessments will need to address corruption where it represents an impediment to sectoral and economywide performance. Private sector assessments, which already regularly survey corruption issues, should continue to focus on this dimension. In some countries EDI's service delivery surveys, which ask beneficiaries about the quality of government services, have also yielded information on corruption.

Research.   Outside the Bank, research on corruption has grown considerably. A substantial body of academic literature is starting to shed light on the complexities of corruption and its effects on economic performance. There is also a growing body of work on how corruption might be controlled. Little of this work is empirical, but this is beginning to change as survey data accumulate. Inside the Bank little explicit research on corruption is under way. Still, research in several areas, such as property rights and aid effectiveness, touches on the issue.

The Bank will devote more resources to research on corruption. The Development Research Group and the Poverty Reduction and Economic Management Network are developing a research program on corruption, which can be carried out by Bank staff or contracted out. As this research yields insights, they will be reflected in the Bank's policy advice and in its contributions to international discussions on corruption.

Dissemination.   As discussion and analysis of corruption becomes more explicit in economic and sector work, in staff dialogue with country counterparts, and in project design, a body of best-practice knowledge will be built. It will be important to capture this knowledge and disseminate it through the Knowledge Networks and in staff training. An advisory service on anticorruption is being created in the Poverty Reduction and Economic Management Network, drawing on the knowledge and insights of other networks. The Network's staff, with the assistance of the Legal Department, will provide guidance to staff on the treatment of corruption issues in economic and sector work, on the design of projects, and on how to help countries design anticorruption strategies.

Country assistance strategies and the country dialogue

A review of current country assistance strategies reveals that they seldom address corruption as a factor affecting Bank projects or as a development effectiveness issue. The reasons are similar to those for not addressing corruption in economic and sector work—the complexity of the subject, the lack of direct evidence, borrower sensitivity, and uncertainty by managers on how to tackle the issue. In addition, staff are unclear on how to deal with systemic corruption when it is found. For all these reasons, Bank staff feel more at ease assessing a government's economic policy record than whether its programs are affected by corruption.

But a failure to address corruption in cases in which it undermines a country's development prospects and the effectiveness of Bank activities diminishes the value of the country assistance strategy as a mechanism for programming Bank lending and economic and sector work and for raising important matters with the government. To rectify this, future country assistance strategies will discuss any serious effect that corruption has on a country's development prospects and, in turn, define ways in which the Bank could help mitigate the problem, consistent with the Bank's mandate. This does not mean that every strategy should include a paragraph on corruption, regardless of country circumstances. Rather, there should be a substantive discussion of corruption and its implications for Bank operations in any country in which corruption affects Bank projects and development prospects more generally.

Addressing corruption in the country dialogue and the country assistance strategy requires sensitive handling. It will be important to view corruption not in moral terms but as a constraint to development. The more corruption can be presented in terms of its effect on the outcomes of government policies, programs, and projects, the more constructive the dialogue is likely to be. The objective is not to rate countries on the extent of corruption but to shed light on how corruption affects government objectives, particularly in areas that we deem important for sustainable development. Cross-country indices of corruption open up avenues for empirical research but are not themselves a measure of economic impact.

Lending program

Lending allocations.   The Bank will take corruption into account more explicitly in its decisions on country operations. In the past, donor agencies, including the Bank, have been criticized for continuing to support countries in which corruption was pervasive and a clear obstacle to development. However, it would be wrong to move from a position in which little notice was taken of corruption to one in which the level of corruption alone determined the size of the lending program. The Bank's interest in addressing corruption arises because it can be a constraint to development. From this perspective, corruption, like inadequate skills, is something standing in the way of a well-performing public sector and the achievement of a country's development goals.

Assessing corruption's effect on development will seldom be a precise exercise, and it will require the exercise of judgment based on the available evidence. Corruption comes in many forms, and its impact varies greatly among countries, according to circumstances. One type of bribery may be relatively benign in one setting but economically crippling in another. Moreover, the harm may be mitigated or magnified by the government's macroeconomic or sector policies.48 Thus corruption should always be assessed in the context of other factors affecting development. Furthermore, if corruption is directly addressed in economic and sector work, insights will be gained, evidence on its effects will accumulate, judgments will become more informed, and its costs will be better articulated to governments.49 These tools will help governments prepare and implement strategies to combat corruption.

Corruption should be explicitly taken into account in country risk analysis, lending decisions, and portfolio supervision if it affects project or country performance and the government's commitment to deal with it is in question. The following considerations will guide lending and, with due respect to legal obligations, disbursement decisions:

  • Whether Bank projects are likely to be affected by corruption during design or implementation, or thereafter.
  • The extent to which the achievement of development objectives is compromised by corruption.
  • The willingness of the government to act to control corruption if it threatens the effectiveness of Bank projects and/or economic and social development.

The response in individual cases should be tailored to the nature of the problem. If corruption is found to taint a particular Bank project, current Bank procedures will be applied and should in general be adequate to address the issue. If corruption is thought to taint the achievement of development objectives in an entire sector, the design and size of Bank assistance in that sector should be reconsidered. In extreme cases of far-reaching systemic corruption that affects most sectors, Bank assistance may need to be focused on a few unaffected sectors, limited to institution-building or nonlending activities, or curtailed altogether. This cannot be a mechanistic assessment but will always call for careful judgment based on accurate information and the specifics of the situation. Every effort will be made to ensure even-handed treatment of countries. And in all cases, the Bank's actions will be governed by the Articles of Agreement and will focus on the economic issues within its mandate.

Within this broad framework and consistent with existing loan agreements country operations may need to be adjusted according to the extent to which corruption affects both Bank projects and development objectives and to reflect the efforts of government to control it. This can be done without changing IDA and IBRD allocation criteria, which are based on country and portfolio performance.50 However, this revised approach does imply a greater capacity on the part of staff to acknowledge the costs of corruption and factor them into performance assessments when they are significant. In some cases corruption may be a country risk factor that will require special Bank attention.

Project design and portfolio management.  Projects need to be prepared in a way that integrates institutional considerations and assesses corruption risks more explicitly at the design stage. Many Bank staff already do this, but experience is not widely shared, and thus institutional learning has been low.

The control of corruption can be supported through portfolio management. The increased emphasis by staff on looking at the Bank's portfolio on a countrywide basis and regularly discussing portfolio implementation issues with borrower governments provides an opportunity for raising corruption issues and agreeing on actions to address them.

Lending instruments.  The Bank should be prepared to use a wide range of lending instruments to support countries that are seeking to control corruption. The Institutional Development Fund has been used extensively to finance innovative public management activities, and we expect this to extend into anticorruption work under the umbrella of the proposed new Development Grant Facility. Technical assistance loans, often accompanying structural adjustment loans and sectoral adjustment loans, have been a mainstay of Bank efforts to help countries implement public management reforms.

Structural adjustment loans and sectoral adjustment loans can continue to be used to support economic policy reform. New lending approaches—such as slow-disbursing fiscal support lending—might also be considered to help the Bank more effectively support the broad institutional changes needed to improve public sector performance. Within a clearly articulated medium-term framework, such loans could support the restructuring of government policies and programs consistent with a changed role of the state and foster transparency and accountability in financial management and in the conduct of public business.

Procurement and financial management

Procurement.   The Bank is redesigning its Country Procurement Assessment Reviews (CPARs), so that these can become more effective tools for helping countries strengthen public procurement systems. Strengthening capacity to procure goods, works, and services is good not just for the control of corruption but for government performance in general.

Financial management.  In keeping with the changes under way in the Bank's disbursement systems, greater emphasis will be placed on borrower accountability and government financial management systems. This means increasing the intensity of Bank work on financial management and carrying out Country Financial Accountability Assessments (CFAAs) for a wider range of countries. These assessments look at financial management capacity, skills, and standards in both the private and public sectors and offer governments recommendations on how capacity in this critical area could be strengthened.

As the Bank places more responsibility on managers for achieving results on the ground, simplifies internal processes, and shifts functions to resident missions, the Bank must ensure that the risks of fraud and corruption in Bank-financed projects remain well controlled. As noted above, review of large-scale procurement decisions will remain a headquarters function, while more staff based in the field should strengthen project supervision and portfolio management.

External agency practices and public sector management

Some of the Bank's (and bilateral donors') modes of operation might impede national anticorruption efforts, and certain Bank practices associated with "enclaving" may need to be changed. In a broader sense, it is also possible that high levels of aid to poor countries may, in ways not fully understood, slow the emergence of institutions of state accountability that otherwise might develop (as they did in industrial countries) if states were more dependent on local revenue sources. This is not an argument against aid but rather a caution about the risks of externalizing accountability where aid flows are a significant part of government revenues.

Enclaving.   A frequently debated issue in the Bank is the tendency to "enclave"—safeguarding Bank operations from the weaknesses of a borrower's public sector by insisting on an implementation environment separate from prevailing government management systems. A long-acknowledged (and now less common) example is the project implementation unit, appropriate in some circumstances but now viewed as overused. An egregious example is salary supplements (or "leave of absence" contracts). Social funds, if they are used for what properly are central government functions, may weaken public management.51 The reasons for enclaving are understandable, given the Bank's fiduciary responsibilities and the desire for smoothly running projects in situations in which government budgeting, accounting, auditing, procurement, and personnel management systems are weak. To some extent these reasons are inherent in any project-centered approach to lending. But they reduce the pressure on government to reform, and they may weaken domestic systems by replacing them with donor-mandated procedures.

This tension may never be fully resolved, but it needs to be better recognized. Bank staff need to pay more attention to the government's budgeting, accounting, auditing, procurement, and personnel management systems and to how they affect the sustainability of Bank projects. Doing so would focus management and staff attention more explicitly on the condition of those systems and on how the Bank might help governments improve them. In turn, this would help governments control fraud and corruption.

Soft budget effects of aid.   Concern has been voiced that the way aid traditionally has been provided by the Bank and bilateral donors may undermine governance in countries receiving large amounts of aid, by expanding government's reach, softening budget constraints, and discouraging a rationalization of functions and redeployment of resources. Donor preferences for funding projects in public investment programs, channeled through a dual budget structure, may have helped bloat government spending when coupled with increasing aid flows. Although donor projects now support the rehabilitation of existing services more than their expansion, the de facto fragmentation of budgets in aid-dependent countries resulting from dual budgeting still discourages a full consideration of the tradeoffs involved in the allocation of resources. Furthermore, when poorly managed, aid softens the budget constraints that should discipline policy choice.52 When accountability is externalized, national accountability mechanisms are weakened, which may allow corruption to flourish.

One way to promote discipline in policymaking and better financial management in countries that depend on aid could be to shift a portion of aid flows from project- to program-based financing and to conditional general budgetary support, within a medium-term expenditure framework. Such budget support lending could be conditioned on the installation and operation of transparent budgeting and financial management systems (with the control of corruption as one objective), just as earlier policy-based lending supported economic policy reform. As part of its efforts to consider new lending instruments, the Bank will consider the feasibility of such an approach.

Staffing implications

Staff training.   The experience and expertise of staff are vital for the success of the Bank's efforts against corruption. Experience will build over time as staff begin helping countries, and it can be augmented through training. Addressing corruption is not an additional set of activities but an underlying factor in the Bank's assessment of development prospects, and it should be integrated into Bank work. Current training proposals for mainstreaming the control of corruption involve adapting existing training courses as well as launching new ones.

  • Training programs oriented to a particular sector progressively will include material showing how projects can be designed to minimize the scope for corruption. Case studies will need to be developed over time, building on regional experience and using the network anchors to assemble the material.
  • Public sector management training programs will be modified to include a module on anticorruption, to sensitize staff, and to show how corruption can be addressed within a public sector management and governance framework.
  • Workshops on the analysis of corruption and the development of anticorruption strategies will be organized, aimed initially at Bank staff working on countries that have approached the Bank for assistance in combating corruption.53 Diagnostic frameworks will be developed and provided to staff. Guidance on dealing with this sensitive subject will be included in staff media training. And training courses should help Bank staff understand the political as well as the economic and social dimensions of corruption.
  • Staff guidelines on addressing corruption have been prepared and are being distributed to all staff.

Staffing implications.  The bulk of the responsibility for mainstreaming corruption in Bank work will fall on existing staff. Thus the additional staff required to implement the recommendations are limited, and many needs can be accommodated through training.

In three areas, however, staffing levels are inadequate and will be increased.

  • The number of skilled financial management specialists in operational units has declined in recent years, a trend the 1994 Financial Reporting and Auditing Task Force Report drew attention to. Specialists are overloaded, and new staff are needed to enable the Bank to assess borrowers' financial management systems and to provide assistance to countries seeking to improve them.
  • The Bank also needs more procurement specialists. In theory, some relief may be obtained by shifting the emphasis from prior review to post-audits. But these, too, require budget funds to hire local external auditors. If the Bank is to help more governments improve their procurement systems, there will inevitably be additional staff costs.
  • Additional expertise is needed in public sector reform. Building better institutions helps control corruption. This means that the demands placed on public management and institutional specialists are likely to grow, requiring an increase in their numbers. Lessons of experience and best practice should be duly collected and disseminated through the Knowledge Networks.

Personal security and risks. Finally, this chapter ends with a warning on personal risks to Bank staff, consultants, and counterparts working with them. Previous initiatives of the Bank—on poverty, participation, the environment or private sector development—have not resulted in Bank staff and consultants being directly threatened by those engaged in criminal activities. Anticorruption work is different. Not only is the bribing of public officials illegal in all borrower countries, but in many countries there are close links between systemic corruption and organized crime. The risk to the personal security of Bank staff and consultants and those working with them cannot be lightly dismissed. Close inspection of substandard construction, the persistent checking of project accounts that have been subject to large-scale fraud, pressure on governments to abandon white elephant projects in which politically powerful figures have a financial stake, or work on the banking system in countries experiencing large-scale money laundering can be hazardous activities for Bank staff and local counterparts. Managers need to be sensitive to these risks. They must avoid placing staff in potentially dangerous situations and be prepared to support them, if necessary by withdrawal, in the face of threats. Staff need to be aware of possible danger and to know that managers will support them.


47. An outstanding example of how open governments have become in acknowledging the problem and their preparedness to discuss it is the communiquŽ of the Summit of the Americas, held in Miami in 1995.

48. A 10 percent bribe on the cost of a good public investment project depresses the project's economic rate of return only slightly. A bribe that saddles the country with a white elephant investment may result in economic costs far exceeding the corrupt payment, particularly if the policy environment causes a value-subtracting investment to appear nominally profitable.

49. As noted elsewhere in this report, private sector assessments are yielding information on the extent to which entrepreneurs see bribery as a constraint to business development. Service delivery surveys are starting to provide insights into whether beneficiaries have to pay bribes to receive government services. International price benchmarking can point to whether procurement is affected by corruption. And in one country, Bank staff were able to bring major fraud to the authorities' attention by comparing customs revenue receipts with central bank balance of payments data.

50. IDA resources are allocated to eligible countries mainly on the basis of their performance in meeting economic and social goals. Countries are rated according to three sets of criteria: macroeconomic stability, structural policies, and portfolio management. Pervasive corruption will have an impact on all three criteria.

51. It is appropriate to create a social fund if the activities to be performed are, in public-management terms, best handled by relatively autonomous agencies with separate governance structures at arms length from the government department with portfolio responsibility. Credit to small entrepreneurs or grants to community schemes without recourse to further government funding can be administered effectively by stand-alone agencies. But when the financing of an activity (for example, building a public primary school) results in recurring departmental staffing and maintenance responsibilities or involves policymaking, a social fund may not be appropriate.

52. Faced with an offer of concessional project aid, a government in a poor country has every incentive to accept it even when the government doubts that the domestic budget can sustain project operating costs. Governments have been right so far in their assumption that donors will meet the shortfall, perhaps disguised in a "rehabilitation" project.

53. A pilot workshop took place in the beginning of February 1997, coordinated jointly by the Poverty Reduction and Economic Management Network and the Learning and Leadership Center. This workshop has been evaluated for lessons that can be reflected in future training events.

contents / previous / next