A few month's ago, Ruth Levine recommended that I read Steve Berkman’s book “The World Bank and the Gods of Lending” and yesterday I had the chance to listen over the phone while he gave an informal talk about it at the Center for Global Development. The talk was just as disturbing as the book - in fact, it is one of the most worrisome books about foreign aid that I have ever read. It provides a convincing look at specific project accounts and patterns of disbursement to show that embezzlement and theft of World Bank funds may be the rule rather than the exception in a wide range of programs and countries. Berkman challenges anyone to prove him wrong when he estimates that 30% to 40% of World Bank lending is stolen (not lost to inefficiency but actually stolen).
His concern is not that World Bank staff are corrupt but rather that the pressure to disburse funds makes it easy for corrupt people in borrowing countries to divert and steal from aid programs with impunity. Essentially, Berkman describes what happens when the culture of disbursement meets the culture of corruption.
I found the book to be disturbing in large part because Berkman holds up a troubling mirror to the community of specialists working on governance and corruption (like me). While Berkman commends the World Bank for putting the issue of corruption on the international agenda, he also points out how workshops, conferences and projects aimed at “reducing vulnerabilities” or “reforming governance” are so much more palatable to the organization’s managers than digging up evidence of wrongdoing and pressuring governments to prosecute and recover funds. Unless investigation and prosecution are part of the strategy, though, all our workshops and technical assistance may simply provide diversions that allow criminals to get away with their loot.
Some people find fault with the book, saying that it overstates its case and relies on a relatively small and biased number of cases. However, I find his argument convincing – in part because he includes examples from relatively well-governed countries like Ghana but also because his experiences echo those that I have seen in many places.
The key issue, though, about the scale of corruption is that no one really knows how large it is. Without credible estimates of how much is being diverted from its projects through corruption, the World Bank – and indeed most development banks and aid agencies – cannot defend themselves against a book like Berkman’s. They may be following accepted practices by creating channels for complaints and doing investigations, but this only reveals a small share of problems. (Berkman can even cite credible complaints that the World Bank's investigative division never had the time to look into.)
A reliable estimate of the extent of corruption requires picking a random representative sample of projects; doing thorough performance audits; and extrapolating to the full population of projects. Until the World Bank starts doing such systematic estimates, it will be extremely difficult to counter the disbursement pressures that keep the money flowing, even when it is gushing down the drain.