Chris Blattman argues that we can't expect too much from countries in need of aid:
Donors’ three mistakes in fragile states, by Chris Blattman: You’re the Finance Minister in a country just coming out of conflict. Or maybe you’re disaster-struck like Haiti. Donors line up and make big pledges. UN agencies arrive and occupy whole blocks of office buildings. Each come in with a template. It looks reasonable. It’s certainly well-intentioned.
None of you know it yet, but you’re setting yourself up to fail ... with three mistakes donors will probably make.
1. Let’s set high standards for governing and disbursing public money. Bad idea. Bureaucracies need procedures, norms and experienced personnel. If a Mozambique or Liberia improves its bureaucracy at the fastest rate in human history, it will have the sophistication of an India or Pakistan in 20 years.
2. Invest quickly in education, health and infrastructure. Actually, these aren’t the country’s first priority. Law, order and security come first. Unfortunately, freedom from violence, or access to justice, are not MDGs [Millennium Development Goals]. Your donors are focused (and evaluated) on human development and poverty alleviation. That’s also what they know how to do best. Security sector reform and justice? Less so.
3. Get NGOs to deliver aid directly. Since the state bureaucracy can’t meet high standards, you can forget direct budget support. But how to build schools and clinics and roads? Enter the NGOs and contractors. Unfortunately, this direct delivery is not going to help you build bureaucratic capability. It might even undermine it.
So what’s the solution?
Set goals for the rate of bureaucratic improvement, not the level of standards. In the meantime, this or that Deputy Minister is going to need to send pork to his constituents. And money is going to get mismanaged or diverted.
Keep education and poverty on the table, but make certain that law and order are first not fourth on the agenda. ...
Finally, in place of direct aid, there’s a nice new trick: community-driven development. Rich countries give the state a big pot of money, then the state defines simple local procedures for disbursement. The donors love it: it sounds all participatory and pro-poor (and often it is). But most of all, it lets a weak state actually disburse cash without a ridiculous amount of accounting, with lots of room for pork and (diminishing over time) diversion of funds to ruling party coffers.
The short story: shoot for the possible, not the impossible.