The World Bank has a public face and a private face. The public face I call the Rhetorical Bank. This is the part of the Bank we are all familiar with – the Bank that produces World Development Reports, gives advice, and applies pressure for structural adjustment. The Rhetorical Bank is the philosophical arm that preaches the wonders of the free market and encourages efficiency, liberalization, deregulation, and debureaucratization. The private face is what I call the Real Bank. This is the part of the Bank that designs projects, negotiates loans, and disburses the money. The power and substance of the World Bank is in the Real Bank, while the Rhetorical Bank takes care of public relations. Indeed, these two components are even separated physically at the Bank's headquarters in Washington, D.C. With few exceptions, the staff that work in the country missions, such as the Bank office in Jakarta, are part of the Rhetorical Bank.

The distinction between the Rhetorical Bank and the Real Bank is important for at least two reasons: first, because few people are aware of the Real Bank and even fewer understand how it operates, and second because the Real and Rhetorical Banks contradict each other in their functions and objectives. The Rhetorical Bank stresses a diminished role for the state in developing countries. The Real Bank, especially in places like Indonesia, has played the leading role in feeding state bureaucracies while displaying very little capacity and even less desire to channel project and development funds past or around the state apparatus. The Rhetorical Bank stresses efficiency, transparency, and accountability. The Real Bank focuses on moving as much money as possible (a "culture of approvals," as one internal Bank report put it), with virtually no effort to audit where funds go or how they are used. The Rhetorical Bank criticizes corruption. The Real Bank colludes with its clients to ignore rampant corruption of the Bank's own funds (which come primarily from taxes in industrialized countries). The Rhetorical Bank preaches about quality control. The Real Bank makes loan after loan for projects that are dysfunctional or severely mismanaged. The Rhetorical Bank portrays itself as an institution for development (especially when asking for more money from the U.S. Congress). The Real Bank says it is a bank and not a development agency.

Structural Problems The problems with the World Bank are deeper than mere schizophrenia. The reformer-President of the Bank, James Wolfensohn, has quickly discovered that the pathologies of the Bank are ingrained in its institutional apparatus, and worse, in the mentality of the staff in charge of loans and projects. Wolfensohn left Wall Street thinking he could take over the Bank and run it like a corporation. Instead, he realized that the Bank is more like an entrenched U.N. agency with a resistant hierarchy of bureaucrats who advance their careers within the Bank not by how much they accomplish with their loans (quality), but by how large their loan portfolio is (quantity). The internal Wapenhans Report, published in 1992, reviewed about 1,800 Bank projects in 113 countries and found that 37.5% had failed, adding that staff and managers "focus on lending targets rather than results on the ground." A 1996 audit by the United States General Accounting Office (GAO), the independent research arm of Congress, found no subsequent improvement. In fact, according to GAO data, because task managers for Bank projects report on their own progress and accomplishments (or use unaudited self-reports from client governments), the number of failed projects is actually 9% higher than the 37.5% figure in the Wapenhans Report. The word in Washington, D.C. is that Wolfensohn is frustrated with the Bank's resistance to internal reform, and wants to return to the private sector.

Collusion and corruption - The case of Indonesia

Indonesia is considered to be a "model" debtor country by the World Bank, and the Bank is very proud of the relationship it has built with Indonesia since the earliest days of the New Order. Indonesia needs the Bank to help continue the large annual injections of credit from the Consultative Group on Indonesia (CGI), while the Bank needs Indonesia as a showcase of lending and developmental success. As countries around the world fail or refuse to repay their debts, Indonesia can always be counted on to pay. Despite the fact that the Bank's figures on poverty reduction in Indonesia were "negotiated" with the Indonesian government (i.e., they were reduced significantly), the Bank puts its full credibility behind the official poverty rates in Indonesia without ever explaining how the number of poor in Indonesia was reached. And despite severe difficulties with some very high profile Indonesian projects for rural development and the environment, the Bank continues to refer to these projects as symbols of how the Bank is able to contribute to development at the grassroots level.

One of the most disturbing aspects of the World Bank's close relationship with the Indonesian government is that for three decades the Bank has allowed a large proportion of its loan funds to leak into the government bureaucracy and disappear. Although it would be disastrous for Bank staff to admit publicly that loan funds are routinely stolen in Indonesia, they quite readily admit privately that such leakage is substantial. In an interview on January 3, 1990 with Atilla Sonmez, the head of the Jakarta mission of the World Bank, I asked the following question: "How much of the Bank's loan funds to Indonesia never see the light of day?" Mr. Sonmez responded "about a third." Since then, half a dozen Bank officials, both in Jakarta and in Washington, D.C., have given the same rough estimate. Indeed, most have pointed out that 30-33% is a conservative estimate, and that such levels of theft and leakage are not uncommon in developing countries like Indonesia.

It is interesting to note how the World Bank has responded to this charge (which was published in a column I wrote in the Far Eastern Economic Review on February 13th, 1997). Rather than deny the estimate of corrupted Bank funds, Michael Walton, chief economist for East Asia and the Pacific for the World Bank, stated: "While we do not know what an estimate of a 30 percent rakeoff in Indonesia is based on, we are well aware of the problem of corruption." This is a particularly revealing statement. First, the Bank is admitting that it is aware that its funds are being stolen. Second, the Bank is saying that it does not know what the 30 percent corruption rate is based on. Both statements are true, in fact. The reason Bank staff conservatively estimate the amount of lost funds in Indonesia (or elsewhere) is because the Bank does not collect and report data on such theft. Oversight of World Bank loans is minimal, and independent auditing is almost nonexistent. In other words, the Bank knows money is being stolen, its staffers have a general sense of how large the problem is, but there is no concrete data on such theft because the Bank does not collect such data. In short, the World Bank has a global policy of "don't ask, don't tell."

When I asked Mr. Sonmez back in 1990 if the level of leakage in the Indonesian system was disturbing to the Bank, he responded nonchalantly that Indonesia repays its debts on time and that it is not the Bank's business or role (as a bank) to follow the money and make sure it is used properly by its clients. Several staffers at the World Bank have made the same argument to me during the years since (most recently in November 1996 at World Bank headquarters in Washington, D.C.).

Conclusions and recommendations

1. The World Bank should be encouraged to overcome its schizophrenia by reforming its lending and its management of project quality. In other words, the primary consumer of the Rhetorical Bank's advice about efficiency, transparency, quality control, and streamlining should be the Real Bank itself. There are two broad ways of achieving this: a strong and detailed internal system for oversight should be created (where promotion and career advancement is based on how many bad projects are detected, reformed, or eliminated); or a strong external body should be created that examines and oversees the Bank's operations and projects. This external body should be given the power to cancel Bank projects found to be mismanaged.

2. The World Bank should insist on much greater involvement by independent organizations and groups in borrowing countries. These local actors can assist the Bank as watchdogs that have a much greater capacity (and motive) to detect fraud and mismanagement. More importantly, these organizations should deal with Real Bank staff, not with Rhetorical Bank staff (as is the current practice).

3. The World Bank should begin immediately to assess how much of the total amount of World Bank loans to Indonesia have been corrupted or stolen. Upon determining the level that has leaked into the bureaucracy, that amount should be forgiven. The rationale for this recommendation is simple: since the World Bank supplied funds in an on-going way that it knew were not being used for their intended purpose, it is unfair that the Indonesian society (especially the poor) should be responsible to pay these funds back (with interest). If the average Indonesian received only two-thirds of the loan funds from the World Bank, why should they be required to pay back 100% with interest? The average Indonesian does not have the power to impose limits on the actions of government officials. But the World Bank does have the leverage, and it should begin to use it immediately.

[Jeffrey A. Winters is an Associate Professor at the Northwestern University, Department of Politics]