Wouldn't it be very cost-effective to fund a charity that effectively campaigns for more or better aid? Let's imagine there exists a charity, Advocaid, whose sole purpose is to campaign for developed country governments to increase their funding of a highly cost-effective aid intervention- Against Malaria Foundation's distribution of insecticide treated bed-nets (ITNs). And let's say Advocaid are pretty good at what they do: for every marginal dollar they receive in donations, they generate $10 of developed-country government aid for AMF's programs. Brilliant! If Advocaid exists already, it should be our number one recommended charity. It it doesn't exist, we should create it! Right?

Unfortunately, there remain several reasons to doubt that Advocaid would be more cost-effective than AMF, and even reasons to think it could do harm. In this and my next blog post, I will explain one of the most important and counter-intuitive reasons: the counter-factual of health aid spending on a particular project.

Remember Advocaid works by raising government aid for ITN distribution. We have to ask, "Where do this government aid come from?" and "How would the money have been spent had Advocaid not been successful in moving it to ITNs?"

First, where does the money come from?

The money can either come from somewhere within the government's budget for health aid, other areas of the aid budget (beyond health), the government's non-aid budget, or tax-payers. We tried to work out which is the case. It turns out, it's very difficult to tell.

On the one hand, it seems likely that governments' aid budgets are sensitive to perceptions of aid effectiveness, and that a campaign that successfully shows the merits of ITN-delivery would affect such perceptions- ITNs becoming something of a 'poster child' for the potential of international development in general, for example (we should like to ask from RESULTS and the Global Network for NTDs, both of whom work in aid advocacy, whether they have any evidence for such dynamics). In this case, the money Advocaid raised would come from non-aid aspects of the national budget, or from tax-payers.

On the other hand, in the short run at least, a country's aid budget is usually set 'in advance' by the government, and a second body or process then decides how that set amount of money will be distributed amongst competing projects. At least in the short run, then, increasing funding for ITNs should result in less funding for other, competing, aid interventions.

Perhaps you can see where this is going. If money raised for ITNs is effectively shifted from other aid projects, then the result of Advocaid's work is not just that ITNs gets more, but also that other aid projects gets less (or, their funding expands by less than it would have done). To estimate the good done by Advocaid, we need to subtract the harm done from this latter process from the good done by the former.

In email conversations with us, USAID representatives have claimed that, in the US at least, the projects that would receive less funding as a result of successful advocacy for a health aid intervention (including ITNs, but also TB, Neglected Tropical Diseases, micronutrients and so on) would tend to be other health interventions. This is because the budgets of broad sectors like 'health' are fixed in advance of the advocacy; the advocacy only moves money within them.

Not only do we not know how well this model generalises across countries, we also still have a 'long-run / short-run' problem: we don't know whether, despite short-term costs to competing aid projects, in the long run, successful advocacy for ITNs would lead to improved perceptions of the cost-effectiveness of health aid (or aid), and hence increased budgetary support for health aid (or aid) as a whole.

We are going to ignore this long-run uncertainty for now and explore short-term dynamics, mainly because we can with a little more ease.

Our initial research suggests that health aid programs tend to be highly cost-effective, with a mean and median cost-effectiveness similar to that of ITN-distribution. If this is indeed the case, a charity like Advocaid simply moves money from one highly cost-effective intervention to another, and may thus do little good, or even do harm. Thus, our initial enthusiasm would sadly be unwarrented, and in general, we should be wary of recommending charities that work in the field of aid advocacy.

In my next blog, I will explain how we estimated the cost-effectiveness of health aid. Since our estimation is extremely tentative, the above blog post certainly should not be considered the final word on the cost-effectiveness of aid advocacy.