What bracing idealism on an important topic I long thought dead. Unfortunately, taken by itself, it would be naive for the Bank to invest in CBA -- economic analysis -- skills without accompanying institutional measures. In the absence of any institutional incentives for objective and careful CBA -- or to connect such analysis to program design and decision-making, any such investment will merely depreciate rapidly, just as it seems to have done in the past. Careful CBA, integrated in the design and decision process, entails thinking skeptically and broadly about causes and "solutions," considering available evidence and context, and projecting benefit streams realistically rather than according to some project proponent's imagined ideal. Taking it seriously also entails an institutional tolerance for "bad" news -- low estimated ERRs -- which does not currently exist. Neither the Bank as an institution nor its staff have any incentive beyond professional integrity (to counter prevailing incentives) to conduct careful and objective CBA. If used for decision-making (rather than as after-dressing) CBA could provide an opportunity to identify key weaknesses in program logic, consider and discuss alternatives with country stakeholders, and improve the chances of successful interventions. Yet there are no mechanisms to help a TTL to ensure that economic analysis is objective. Consultants who depend on TTLs for their next consultancy surely are not going to risk a less than favorable economic analysis. With all the internal review mechanisms that exist, who will ask the TTL's or their managers about the objectivity of the economic analysis? Managers themselves "know" that TTLs lack the time and financial resources in many cases to produce more careful analysis and solutions. The Board is unlikely to seriously question the basis of a high or acceptable ERR, and if it did, would anyone be at the Board meeting to answer their questions? If an intrepid economist, perhaps unaware of the perils to his or her career, estimated returns to be low, how long before they were labeled a "trouble maker"? "too tough"? or "too critical/negative"?
Although the Bank has issued a new policy saying that economic analysis matters and shall be done -- more often, better, and more systematically perhaps than before, there is no discernible discussion of how to do this. There is no explicit discussion (at least in public) connected to the Bank reform agenda on the institutional factors which would support the new policy. Not only are skills lacking, but so is a clear understanding of what it would take to implement the new policy, the links between CBA to the science of delivery, or to better M&E, learning, and results. When a skill or sphere of knowledge is not valued, it depreciates. When it depreciates, the credibility that its practitioners have also does. The importance of the exercise is devalued, and the cycle continues. It is thus difficult to envisage the Bank building capacity elsewhere in a skill set that it has so badly lost. While I am sure that receiving the message that a favored project has low expected returns is frustrating for some staff, it is preferable to wasting public resources, or even worse -- wasting opportunities to provide better, more cost-effective solutions sooner. Ineffective solutions sincerely embraced only delay a more serious search for effective ones, especially if we never learn that they are ineffective. This holds for HD interventions as well, if not more so.
Other institutions have experimented with different institutional arrangements to incentivize improved CBA and CBA-based decision-making. DfID's Chief Economist's office has a serious review function (which encompasses the whole "business case"). Other entities, such as MCC, are still figuring this out with little to no supporting institutional structures as well. Still, all MCC project teams have an economist with an independent reporting line (and ultimately job security) whose job it is to conduct objective CBA and explain the analysis to anyone who might, one day, care.