no one really knows how to get a state to start recognizing and addressing the core needs of its population. This outcome seems more likely to happen in countries where a parliamentary system diffuses power, where built-in institutions oversee public policies, and where a strong civil society keeps the state in check. But the recent history of the developing world suggests that these features are not easily obtainable.

…Outside donors have tried to help such fragile states by lavishing them with foreign aid, but as admirable as the intention behind such largess may be, it often ends up entrenching governments that do little to promote development. And when outsiders try to use their leverage to spur political reform, it raises many difficult questions; for one, the legacy of colonialism

…The alternative is to give up on existing state structures. For example, multilateral donors and aid agencies could try to build independent agencies to deliver public services in the developing world, institutions that would operate on a larger scale than NGOs currently do and adopt the mantra of evidence-based policymaking. But if states are to remain legitimate in the eyes of their populations, then such agencies will eventually have to be absorbed by the government. Or perhaps they could remain quasi-independent and democratically governed, the way that school boards and hospital trusts are.

…None of these lessons suggests that there is much low-hanging fruit to be had when it comes to eradicating poverty. Still, economists need to at least try to understand what goes into making states better at fostering development.

Tim Besley in Foreign Affairs reviewing the three micro-revolutionary books in development: Poor Economics, More Than Good Intentions, and Portfolios of the Poor.

I am a fan of all three books (and teach them) but Tim outlines the fundamental limit on all the randomized control trials in the world: the main constraint on good policy is not that policymakers don’t know what to do with public money.

If you don’t actually have a national system with the incentives and accountability to execute good policy, information on better policy is not going to make things right.

And even though I wear the hat of both randomista and political scientist, the last thing I believe is that we can understand how big institutions work through the randomized trial. Too big. Too adaptive. Too heterogeneous.

Tim’s book is here. To be reviewed, when I have the time to finish it. So far it’s very good.

A caveat: One mustn’t only let political economy back in at the macro level. For a nice micro example, everyone in development should read the famous Anti-Politics Machine.