As many readers of this blog are likely aware, the World Bank (and the other multinational development banks) have their own procedures for identifying and sanctioning firms that engage in unethical behavior (corruption, but also fraud, collusion, etc.) in Bank projects.  At the World Bank, responsibility for addressing corruption and other unethical practices by Bank contractors and partners is handled by the Integrity Vice Presidency (INT), which has investigative and quasi-prosecutorial functions, and the Office of Suspension and Debarment (OSD), an independent adjudicative body, as well as the Sanctions Board, an appellate body.

A few weeks ago, the OSD released a comprehensive report on its office’s activities and performance over its first seven years in operation (fiscal years 2007-2013).  It’s a very useful report, and well worth reading.  It includes a clear, succinct summary of the World Bank’s sanctions and procedures (including both their history and current structure), and also–most notably–a great deal of descriptive quantitative data about the OSD’s activities.  In many ways, the report is a model of transparency, allowing observers both inside and outside the Bank to understand the activities of OSD (and, to a lesser extent, INT and the Sanctions Board), and perhaps to identify weaknesses and areas for improvement.

But because no good deed goes unpunished, my main initial reaction to the report is to wish there were even more data provided!  Here are a few open questions that the data in the OSD report does not address, but that OSD might consider providing in future reports:

First, although the report helpfully breaks down each case by type of offense (and, within the “fraud” category, by type of fraud), it does not include the magnitude of the offenses–that is, the total dollar amounts allegedly misappropriated (in cases of theft), or the total size of the contracts (in cases where there aren’t allegations of misappropriation, but rather some form of misrepresentation to improperly secure a contract).  It would be very helpful to get a sense of the severity of the alleged offenses in each case, for a number of reasons–including that it might help internal and external assessors evaluate whether the office is allocating its resources effectively.

Second, it would be very useful — both for research and practical reform — to have more data on the characteristics of the firms and transactions involved in these cases, including: (1) size of the firm; (2) ownership structure (private, closely held, publicly traded); (3) industry; (4) firm’s country of origin (and whether the firm is primarily domestic or multinational); (5) country/countries where the project giving rise to the alleged wrongdoing took place (or if that would be too sensitive, perhaps the region and/or WGI “control of corruption” score for the country). This sort of information would be helpful for several purposes.  Identifying correlates of wrongdoing might (in conjunction with other data and some clearly-specified assumptions) allow researchers to develop better predictive models (along the lines of a working paper I discussed in an earlier post on FCPA violations, as well as another paper I’ve posted about on predicting procurement fraud).  It might also help the Bank assess whether it’s targeting its investigative resources effectively, particularly if one could compare those cases where the Bank imposes sanctions to informed perceptions of where corruption/fraud in Bank projects is the biggest problem.  Surely there would be other uses as well.

Third, though the OSD report provides very helpful data on the cases that make it to OSD, to understand the whole process it’s important to know more about how INT processes the cases, particularly because INT does not move ahead with all the cases it looks into.  OSD can’t be faulted for failing to provide this information in its report; after all, OSD and INT are deliberately separate and independent — and moreover the confidentiality concerns regarding cases that INT looked into and decided not to refer are even more acute.  Yet it would be very helpful to researchers and evaluators if INT and OSD could cooperate in a future report to provide as much information as possible about INT’s processing of cases.

Fourth, as the OSD report notes, and as Rick discussed in an earlier post, when the Bank (in particular, INT) finds evidence of illegal wrongdoing, in addition to whatever sanctions the Bank may impose, it will refer the matter to law enforcement authorities with jurisdiction.  Given that debarment sanctions become public, and law enforcement proceedings are also (usually) a matter of public record, it should be straightforward (though perhaps time consuming) for an outside party to figure out which referrals have resulted in enforcement action by one or more national governments (at least in those cases that result in Bank sanctions).  But INT/OSD might do the world a small service if it included, as part of the same data set on Bank enforcement, information on (1) which cases were referred to national governments, and (2) the outcome of the referral (no action, ongoing investigation, sanctions, etc.).