How to Spend It (Paul Collier on Budget Support)

I managed to miss this a couple of months ago*; Paul Collier proposed at an ODI meeting that Budget Support should be given according to a simple decision rule, based upon an independent assessment of country public financial management (PFM).

Mick Foster's response is very interesting, arguing that "It is about managing an aid relationship, not certifying PFM and turning on auto-pilot ... Aid donors are the right people to do this – not agents who don’t have to live with the consequences of their verdicts."

I do like the principal that decisions be made objectively according to a transparent rule, but it is easy to imagine problems. A donor could easily be embarrassed by a democratically elected recipient government with good PFM systems deciding to make spending decisions which look somewhat inappropriate to a British taxpayer. But then that is the whole point of budget support - ceding control over spending decisions to a recipient government which is accountable to local people. Which could get tricky if local people hold values which might be different to the donors' (e.g., "The British government says it will cut aid to African countries that persecute LGBT people").

Foster argues that the economic case for budget support is clear, providing "the most cost effective way to deliver services, including allowing for leakages," which contrasts with the perspective of politicians who are "absolutists, not a penny can be stolen, clearly unrealistic." The trouble is, our politicians are also democratically accountable. To an electorate which is also pretty absolutist about corruption. And kind of understandably so. Being stolen from generates an emotional response.**  

Imagine you want to give £100 to a Kenyan family. Would you rather spend £25 of that on legitimate administration expenses and have £75 go to the family, or have £80 get through to the family, only £10 spent on administration, and £10 stolen by a politician and spent on his new Mercedes Benz? We know from a large literature on psychological experimental games that individuals often value procedural "fairness" above  monetary gain. People can choose inferior outcomes when they think that they are getting stiffed. Does this extrapolate to aid? I can certainly understand both a rational economic acknowledgement that corruption often isn't the main issue, alongside a kind of emotional moral revulsion which makes me want to absolutist.

If the economic case really is so strong compared to the political constraints, how can supporters of budget support make it more effectively? 

Read the ODI meeting summary here, and Mick Foster's response slides here.

*Thanks for the reference to Stephen Peterson at a recent Oxford Seminar.
**And as an aside, it's interesting that proponents of budget support argue that the economic case needs to trump the political one in this case, when they are often the kind of folks who criticise economists for focusing only on the economic case with scant regard to political constraints.