Does aid work at the macro level? Highly recommended paper
- The "relationship between aid and growth" question is one that I've been finding a lot of literature on. It appears to be heavily studied relative to more micro questions like "what specific programs have worked?"The best single paper I have seen so far on the subject is "Macro Aid Effectiveness Research: A Guide for the Perplexed" available at http://www.cgdev.org/content/publications/detail/15003 . It's good because it is written for the non-expert; it has clear non-technical explanations of the conceptual issues behind the technical (statistical/econometric) debates; it's recent (12/07); and it appears to be fairly "comprehensive" in terms of mentioning the major arguments on different sides. (It includes every paper from the Easterly and Radelet reviews I discussed before, I believe.) It's only 21 pages and I recommend it if you're very interested in this macro-level question.The bottom line is that there are legitimate arguments and major experts both in the "aid is associated with more growth" and in the "aid is not associated with more growth" camp. All studies have potentially major flaws and there is plenty of room for judgment calls. The author of the paper (Roodman) mostly thinks that the potential flaws are big enough that we don't really know anything about the relationship between aid and growth.Points of consensus: author says there's consensus that aid has often worked; that it has often not worked; and that the macro question of "does aid work?" is very hard to answer (much harder than micro questions about specific projects).I also read in implied points of consensus that (a) aid is not significantly destructive at the macro level (at least not in terms of growth); (b) aid does not have tremendous returns at the macro level. To quantify this better, I'd have to run through all the studies and just write down the magnitudes of the estimated impacts of aid and growth, which would allow me to give a range (and frequency distribution) for estimates. We may do this later.As a semi-side point but one that I found very interesting and important: Roodman believes that academics are personally biased to "find" and publish studies showing effects rather than no effects. The discussion of this issue, which he calls "publication bias," is extremely sobering and a good thing to keep in mind for all research on effectiveness of social programs. See pgs 13-15.