Aid is in the news at the moment. The Accra Aid forum took place last week in advance of the UN’s Financing for Development meeting in Doha later this year (on Accra see Simon Maxwell on ‘High Drama at the High Level Forum’ over on the ODI blog).
Meanwhile, in Martin Wolf’s Economists’ Forum at the FT Adrian Wood argues for capping aid. He writes:
“… one can have too much of a good thing. Some developing countries, most of them in Africa, have had high levels of aid dependence – in excess of 10 per cent of gross domestic product, or half of government spending – for decades. It is questionable whether this has been helpful”.
“I therefore propose that donors collectively set an upper limit on the amount of aid they give to any developing country. This limit should be 50 per cent of the amount of tax revenue that the aid-receiving government raises from its own citizens, by non-coercive means and excluding revenue from oil and minerals”.
This proposal has mileage. Countries certainly don’t benefit from excessive dependence on aid. But Adrian’s idea needs refinement to make it work, as I point out in a response posted on the FT blog. In particular, I worry that it increases the pro-cyclical nature of aid. That is, donors reward governments in good times (when they need aid least) and reduce aid in bad times (when they need it most). (On the evidence that aid is pro-cyclical see: John Thornton in the Journal of African Economies and the IMF on the macroeconomics of managing aid).
Capping aid at 50% of the country’s tax revenue, as Adrian suggests, could exacerbate the budgetary impact of negative shocks, either external (such as energy costs) or internal (such as drought). How? Shocks reduce GDP and therefore tax revenue. This is especially so for indirect taxes as market sales fall, and tariff revenues as import volumes decline (on which many low-income countries are still very dependent). As revenue falls, aid will be automatically reduced under Adrian’s proposal. The proposal would punish governments hit by shocks that are not of their making. In sum, this exacerbates an existing and undesirable bias in the timing of aid.
Here is my solution: governments and donors will need to agree a time frame (say 5 years) over which to monitor the aid/tax ratio (i.e. a period matching the country’s business cycle: different economies might need different time frames). Then agree a level of aid over the next five years, based on the outcome for the first five. And make sure that most of this extra tax revenue is spent wisely, for example on the “seekers” that Bill Easterly has written about (NGOs working to find better solutions to chronic poverty).
So, Adrian has an interesting idea, but it needs refinement.
But a bigger question is: how do you get donors to co-ordinate, so all their aid adds up to less than 50% of tax revenue? Getting donors to act together is like herding the proverbial cats: this is despite the 2005 Paris Declaration (as this FT report discusses). And that’s the OECD-DAC donors, who are supposed to coordinate. What about the aid donors who don’t participate in the DAC, most importantly China and India – the so-scalled ‘new donors’ (although China was giving substantial aid to Africa back in the 1970s). Can you design incentives to encourage their co-operation? (on the non-DAC donors see Peter Kragelund’s paper in the DPR here).
Adrian’s proposal has received a lot of comment (go to the CGD blog for a collection of these).