Combating Corruption in the Multilateral Development Banks

Testimony before the Senate Foreign Relations Committee
United States Senate
Washington, DC

July 21, 2004

Chairman Lugar, members of the Committee, I would like to thank you for inviting me to address the Committee on this important topic. In contrast to just a decade ago, it is no longer necessary to make the argument that corruption impedes economic development in many poor countries and that addressing it should be a priority in international efforts to combat poverty. Despite the change in attitudes, however, corruption remains a serious problem, and the Committee is to be commended for focusing attention on the important role that the multilateral development banks (MDBs) should play in combating corruption. I would like to begin by discussing the broader context in which assessments of that role should be placed. I will then discuss how different approaches could be tailored to different categories of lending and to different types of countries and conclude with some comments on the special problems related to extractive industries.


Assessing Multilateral Development Bank Performance on Corruption Issues

Since President Wolfensohn's 1996 speech emphasizing the need for the international development community to address the “cancer of corruption,” substantial progress has been made in raising the profile of the issue and in developing strategies to combat it. The MDBs, led by the World Bank, have beefed up internal controls to prevent corruption in projects they finance, and, more significantly, they now explicitly address corruption and governance issues as impediments to development at the country level. Thus, estimates of how many billions of dollars have been lost to corruption since the World Bank was created must be adjusted to reflect more recent experience.

In addition, while the MDBs should not be excused for taking so long to confront corruption, these institutions operate in an environment that is influenced by the broader political and foreign policy priorities of major donor countries. During the Cold War, security concerns often trumped development objectives. For example, it was no secret that Zaire was pervasively corrupt in the 1980s under the leadership of Mobutu Sese Seko. But US policymakers at the time were more concerned about maintaining Mobutu's cooperation against the Soviet Union and access to strategic resources in Zaire than they were about corruption or economic development.1

Since September 11, 2001, there is more recognition that economic development often is a security objective and that poorly governed and failed states can provide a haven for terrorists. Still, there is some fear that corruption and broader governance concerns are once again being put aside in some cases as part of the war on terror and that pressures are being brought to bear on the international financial institutions to lend to certain countries for foreign policy purposes, regardless of the likely development effects.

Finally, in assessing the MDBs' performance in combating corruption, there is a question as to the appropriate standard against which to measure progress. World Bank officials, as well as outside observers, frequently speak of a zero tolerance approach to corruption. This is appropriate for framing institutional attitudes towards corruption, particularly given the MDBs' past failures. But cost-benefit analysis should be applied to anti-corruption policies, just as it is to policies in other areas. This is not intended to revive the old arguments that some forms of corruption—for example, “greasing the wheels” of an unwieldy bureaucracy—are beneficial. (No one taking that position has ever explained how a political and social environment that permits “efficient” corruption can be structured to prevent other more pernicious forms also developing.) But it is also important to keep in mind that controlling and reducing corruption requires time, money, and other scarce resources that cannot be used for other priorities.


Addressing Vulnerabilities in Different Types of Lending and Countries

Despite the gains made over the past decade, corruption remains a serious problem in many countries where the MDBs operate. Much of the testimony presented to the committee in May noted this as well and emphasized that the opportunities for corruption, and for controlling it, differ by type of lending and by sector. Corruption related to nonproject lending is harder to detect and to control than corruption in projects that the MDBs directly fund and oversee. More broadly, corruption will inevitably be harder to control in countries with weak institutions and governments that are able to avoid transparency and accountability to the public.

Though not perfect, improvements in internal controls appear to have been relatively effective in preventing large-scale corruption in projects that the MDBs directly fund. The World Bank generally requires international competitive bidding and has changed bidding documents to beef up anticorruption provisions. It now posts on its website a list of firms that have been blacklisted because of corruption, and it has created hotlines for reporting corruption and whistleblower protections for staff who want to report corruption. The World Bank is often ahead of the regional development banks, however, and the US Treasury should push for further harmonization of anti-corruption procurement policies—for example, mutual recognition of blacklists. For large projects, exceptions to international competitive bidding should be rare and fully explained with as much disclosure as possible.

Thus, a great deal has been done to guard against corruption in project lending, and most analysts conclude that it is relatively rare in parts of projects that the MDBs directly fund. On many projects, however, the banks provide only a fraction of the overall funding, and they do not monitor procurements for portions of projects that do not use MDB funds. But corruption anywhere in a project can undermine the effectiveness or raise its costs, so it makes sense for the MDBs to require competitive bidding and transparency in all parts of projects for which they provide even partial funding. To be effective, however, such a policy change would have to be coordinated with official export credit and investment insurance agencies in the United States and other OECD countries.

But these procurement safeguards may work less well for smaller, community-based projects, such as building schools, health clinics, or roads, where there may be fewer bidders in the initial construction phase and where ultimate effectiveness depends on ongoing monitoring to ensure that teachers show up and are qualified, that drugs and other supplies are not stolen, and that maintenance is adequate. In these cases, transparency in the procurement process, rigorous auditing, whistleblower protection, and other safeguards are still essential. But it is also crucial to engage local stakeholders to do ongoing monitoring and to hold officials accountable for delivering services as promised.

But in some countries, even the best safeguards may not be enough. Earlier this year, my colleague, Steven Radelet, testified before the House International Relations Committee on the need for further reform of bilateral development assistance to make it more effective in achieving US goals.2 In discussing the need for a more comprehensive and coherent foreign assistance strategy, he recommended that the government develop different approaches for different types of countries. In essence, he argued that how well governed a country is and how effective its institutions are should determine the amount of money for which it is eligible and the discretion that the government is allowed in deciding how to spend it.

Thus, high-performing countries, such as those that have qualified for the new Millennium Challenge Account, should receive more funding, most in the form of nonproject lending and principally allocated to the central government. The client government would be responsible for demonstrating that the money was effectively used but would also have broad discretion in setting priorities and allocating the money accordingly.

Low-performing countries would receive fewer dollars, most of it for specific projects with intensive monitoring and directed to particular agencies that have demonstrated effectiveness in development projects. In the context of US aid, Radelet recommends that in cases where the corruption risks of dealing with the government cannot be mitigated, aid should be channeled through nongovernmental organizations. The only option in cases where corruption risks are unacceptably high is to avoid lending. The World Bank has already moved in the direction of more selectivity in its lending, but the Operations Evaluation Department just released its report on “development effectiveness” recommending that the Bank could do more.3

At the same time, the development banks have also increased the share of loans going to public-sector institutional reform in countries with weak governance. The World Bank website notes that “more than 40 percent of the Bank's lending operations now include public-sector governance components.” In some cases the World Bank has also conditioned nonproject lending on specific reforms, such as the creation of anti-corruption commissions or agencies. But externally imposed anti-corruption conditions are unlikely to be effective in situations where the political will to adopt fundamental reforms is weak or absent. Recent research by the World Bank's chief corruption expert, Daniel Kaufmann, also finds that narrow civil service and technocratic reforms have yielded relatively little in the battle against corruption. Kaufmann concludes that reforms should, instead, focus on “external accountability mechanisms, with new participatory approaches, providing voice and feedback mechanisms to stakeholders outside the executive…” (emphasis in original).4

Such external accountability mechanisms depend on maximum transparency in government operations, an independent media to uncover and disseminate important information, and organized civil society groups. The World Bank has implemented programs in some countries to strengthen the local media, and it is doing more than previously to consult with local groups that will be affected by its projects. Such consultations are typically focused on ensuring that local stakeholders support projects in their area, but they would also be useful in transmitting information about projects, facilitating monitoring by local groups, and providing mechanisms for feedback on potential corruption in projects, as well as other aspects of effectiveness. Kaufmann also recommends that the Bank develop “citizen scorecards” to help in the evaluation of projects and notes that “Transparency-enhancing mechanisms involving a multitude of stakeholders throughout society can be thought of as creating millions of ‘auditors'” (ibid., p. 35).


The Special Problem of Extractive Industries

Finally, extensive research has shown that countries rich in natural resources, if they also have weak political and social institutions, often have poor growth and development outcomes and high levels of poverty. In these cases, corrupt governments often collude with corrupt investors, bankers, and other private sector actors to steal the proceeds of extractive industries, rather than investing them in the country's development. In many countries, weak governance and inadequate oversight by external funders have meant that publicly financed projects in extractive industries have led to environmental degradation, social problems, such as AIDS, and, in all too many cases, human rights violations and violent conflict to gain control over resources.

Recognizing these problems, the World Bank Group recently undertook a review of its lending for extractive industry projects, including creating an independent “stakeholder consultation process,” headed by the former Minister for Population and Environment from Indonesia. In part out of concern over global warming and other environmental problems associated with carbon fuels, and in part because of the well-documented problems in controlling corruption in extractive industries, the independent Extractive Industries Review (EIR) recommended that the Bank sharply reduce its lending for oil and gas, as it has already done for coal, and increase lending for investment in renewable energy sources (

In the draft response to the EIR, which has been released for public comment before presentation to the Board of Executive Directors, the Bank management promises to better assess the corruption risks before it provides support for extractive industries, to be more selective in lending to countries “where the risks are deemed too great and cannot be mitigated,” and to require transparency in relation to the revenue from these projects as a condition for lending. But, while Bank management accepted the EIR recommendations on selectivity, it rejected the recommendation to completely phase out support for oil production by 2008 for environmental reasons. But the Bank's strategy of continuing to finance such projects in resource-dependent countries depends crucially on the effectiveness of the mechanisms to guard against corruption. The success of the World Bank-supported Chad-Cameroon pipeline project will be an important test of whether transparency and intrusive resource revenue controls can be effective in controlling corruption in poorly governed countries. Careful evaluation of the results of this project should guide future decision-making in this area.

In addition, MDB policies in relation to extractive industries need to be coordinated with and supported by bilateral official credit and insurance agencies. The Commission on Weak States and US National Security, sponsored by the Center for Global Development (CGD), recently recommended that the US government take steps to ensure that Export-Import Bank and OPIC support for extractive industries be linked to transparency conditions. The Commission recommended specifically that the National Security Council should:

…broker an interagency agreement that outlines basic principles of transparency and accountability in the handling of natural resource revenues that must be met by governments before the US supports public-sector financing of extractive industry projects.5

The Commission recommends that the Treasury Department have responsibility for overseeing the implementation of this agreement.

For countries with extractive industries that do not need public financing, the report recommends that the Treasury Department coordinate an interagency review of the options for regulating multinational corporate payments to such countries. In particular, the Commission proposes giving “serious consideration” to the recommendations of the publish-what-you-pay campaign to have developed-country stock market regulators require full disclosure as a condition of being listed (ibid., p. 57). The US Treasury and State Department could also do more to support British Prime Minister Tony Blair's Extractive Industries Transparency Initiative, which the World Bank Group has already endorsed.



In sum, the World Bank Group has taken important steps to ensure that project funding is not diverted for corrupt purposes, but continued vigilance is essential and some further improvements in transparency and accountability of Bank operations would be useful. The US Treasury needs to continue pressuring the regional development banks to adopt similar reforms where they have not and use the US voice and vote to ensure that increased selectivity in lending is used where anticorruption safeguards are unlikely to be effective. The Treasury should also ensure that the development banks do more thorough and rigorous analysis of the effectiveness of various institutional and governance reforms they support. As recommended by the CGD-sponsored Commission on Weak States, the Treasury should also take steps to ensure that US government support for extractive industries does not undermine efforts to increase transparency and improve governance in resource-dependent countries.



1. Several years ago, the Financial Times (May 12, 1996, p. 1) published a detailed report estimating that the International Monetary Fund, under pressure from the US government and other western donors, lent Zaire more than $1 billion after a senior official warned that much of the money was being wasted through corruption (cited in Kimberly Ann Elliott, “Corruption as an International Policy Problem: Overview and Recommendations,” p. 214, in Corruption and the Global Economy, edited by Kimberly Ann Elliott, Washington: Institute for International Economics, 1997).

2. “US Foreign Assistance after September 11 th ,” Testimony for the House International Relations Committee, February 26, 2004.

3. World Bank, Operations Evaluation Department, 2003 Annual Review of Development Effectiveness, Washington, 2004.

4. Daniel Kaufmann, “Rethinking Governance: Empirical Lessons Challenge Orthodoxy,” Discussion Draft, World Bank, March 11, 2003.

5. Commission on Weak States and US National Security, On the Brink, Washington: Center for Global Development, 2004, p. 56.