In this month's lead essay at Cato Unbound, 'Why Aid Doesn't Work,' William Easterly makes a rational case for directing international aid dollars toward programs where the results can be objectively measured by hard scientific methods. He's persuasive, but in the end, the scientific method is still just a patch, a facsimile for what's really missing: legality.
In the developed world, the distinction between government and non-government organizations is meaningful, but not in places where government is corrupt and ineffective. In a lawless state (ie, one where corruption dominates the channels through which people get things done, for good or for ill), both aid organizations and entrepreneurial warlords effectively operate by the same rules. One's moral orientation is not the point. If the basic sphere of operation is illegal (in the deepest sense of the word), then there is little chance that your efforts will result in enlightened, long-term improvement. Will Connors, a Chicago journalist and blogger working in Ethiopia has written compellingly about corruption in the aid community there.
Any aid-before-government aparatus is still going to break down--it will just do so further down the road. As Easterly suggests, you may demonstrably produce healthier, taller, better-educated children by buying them meat to eat twice a week, but if the best they can hope for is to grow up to be a corrupt low-level official, have you really accomplished anything lasting?
Better to see aid dollars spent as investment dollars--sunk into private businesses, rather than programs of any kind. It is possible. It is even possible that the weaker the local (corrupt) government, the greater the opportunity for leveraging capital investment. In Carol Pineau's documentary, Africa: Open For Business, the CEO of Daallo, a Somali airline, marvels that the only reason he's able to thrive is that there's no government at all in his country. No government means no corruption, he observes dryly.
cross posted on Samizdata.