In a recent blog post Duncan Green of Oxfam briefly introduced COD Aid (for what that is go here) and raised a few good questions (along with a disclaimer that he needs to learn more) about the approach. One concern he raised is whether the approach doesn’t pass the donor’s usual attribution test, i.e. the test of whether the donor’s aid made some positive and measurable difference. We could point out that that’s the case already with about 99 percent of all official aid—including budget support—but that would be surly! (To understand why serious evaluation doesn’t usually happen read the Evaluation Gap and learn about the subsequent founding of the 3IE and the support it provides for recipient countries to do their own evaluations of their own programs, whether financed by their own or donors’ revenues.)
Here’s what we say about the attribution issue, in the form of answering a frequently asked question.
Another way to think about it: With COD Aid the recipient government instead of the donor internalizes the incentives to translate inputs into outcomes, and thus ensures that whatever steps it takes make a difference. (We think the emphasis on results is why COD Aid is appealing to liberals and in the UK to the Conservative Party too.)
Go here for answers to other good questions—including how to handle in a contract the negative weather and other shocks (over which a government has no control).