The Washington-based NGO ACTION has just released a report on the effectiveness of the World Bank’s preeminent instrument for strengthening health sectors in poor countries: the Sector Wide Approach or SWAp. Through a SWAp the World Bank and other donors collectively provide broad financial support to a country’s health sector, in order to foster country ownership and to coordinate all the many parts of the health sector toward improving the population’s health status.

This is one of the best reports on the mysterious workings of World Bank aid modalities that I’ve ever seen. Perhaps I’m not the best judge, since I already know what a SWAp is, but to me the report seems to clearly lay out the reasons the Bank got into the SWAp business and then shows the deficiencies with the way SWAps currently work. Congratulations to the author Richard Skolnik and his colleagues for making a seemingly boring and impenetrable topic come alive. They also demonstrate that an evaluation can be quite persuasive in the absence of a randomized controlled trial or other equally rigorous statistical analysis.

The authors cite CGD’s new report on “cash-on-delivery” (COD) aid in a passage urging that SWAps move increasingly toward results-based financing. They go beyond CGD’s work in stressing that a results-based focus would improve the internal accountability within donor agencies, as well as between donors and their constituencies and within the recipient countries. Here’s a quote I like a lot:

“After many years of paying insufficient attention to results, management within these institutions must provide stronger incentives for staff to focus on achieving results. … If the achievement of key indicators for TB control were a trigger for the disbursement of financing, then it is likely that both countries and their development partners would pay more attention to realizing these results than they do now. (Pp. 22-23)”

Some will argue that it’s not fair to ask that interventions designed to strengthen health systems demonstrate an immediate improvement in health status. I disagree. Assuming that the typical developing country health system in need of a SWAp is less than fully functional, reorganization and system strengthening should improve some of the more narrowly-targeted, front-line programs in the short-term, and positively contribute to health status within the 3 to 5 year framework of a development project. Indeed, knowing that they will be held accountable for improvements in various measures of health status (not just for changes in procedural indicators), donor and recipient country staff will design the SWAp to assure that it focuses first on improving population health in the easiest ways, while simultaneously making changes to the system that will only bear fruit in the outer years of the project. If neither the donor staff nor the recipient government is willing to be held accountable for health improvements during the project’s time-frame, that is not an argument for turning one’s back on results, but rather for lengthening the duration of the project to 8 or even ten years, if necessary.

I have two qualms about this report, one of form and the other of content. On form, my one regret is that the authors were unable to come up with any charts or statistics to show how little attention to results currently occurs in SWAps. I wish there were a few charts one could lift to cite this analysis. Lifting a pithy quote into a PPT is just not very satisfying, at least for me.

My other point of contention is that the authors seem unaware of the tension between their own purpose, which is to assure more attention to tuberculosis in African health sectors, and the challenge of adding results measurement to a SWAp, which has the broader purpose of improving overall health status. They allude to a list of indicators one might develop for the SWAps, against which donors would provide payments. But coming up with such indicators and more importantly the dollar amount of the reward attached to a unit improvement in each, is no trivial task. In the absence of a market mechanism in which the suppliers and consumers arrive at prices which are mutually acceptable, the World Bank and its donor partners would have to negotiate such rewards.

The alternative to long lists of indicators is the purist version of the COD approach, which would use only one indicator for the SWAp, a measure of the population’s overall health status, as measured for example by an age-adjusted mortality rate. Such an approach would depend on all the implementers, on both the donor and recipient sides, to figure out to what degree vertical programs like TB should be “ring-fenced” for protection and to what degree they should be integrated into horizontal systems, all with the ultimate objective of improving overall health status. Such an approach might result in less expenditure on TB interventions than TB enthusiasts like the authors of this report would prefer. I hope that they and other disease-specific interest groups would be willing to agree on a single measure and then let the chips fall where they may.