Going against Duflo

By on 10/9/2013 in Uncategorized

In the past, I’ve linked to pieces that challenge the ethically untroubled, and systemically timid work of  Abhijit Banerjee and Esther Duflo, of Poor Economics fame. If you’re unfamiliar with their work, the quick summary is this: they start with the idea that poor people aren’t as dumb as development theories have tended to make them out. This is laudable, though not particularly original. Poor people have, for quite some time, been telling people that they are professors of their own poverty. Banerjee and Duflo suggest that perhaps the ideas that poor people come up with to survive poverty are better than the policies crafted for them in Washington. So far, so good.

The trouble is that when it comes to fixing poverty, the theories of poor people don’t seem to matter much. What is most important is What Works. Banerjee and Duflo’s rigorously clinical approach – they use randomized controlled trials to test What Works – transcends ideology to reveal that bed nets and deworming are good ideas.

Again, nothing terribly objectionable here. Except that this approach limits the extent to which poor people are allowed to be clever. The extent of their intelligence is reduced to rats in a maze, figuring out what works under the benign and compassionate supervision of ‘practical visionaries‘at MIT. If there are ways that poor people want to change their relation to power, we’ll never know. These are not subject to randomized controlled trials.

Let’s call this fixation with What Works “Millennial Economics”. To be obsessed with What Works is always to exclude consideration of how the context of what works might be different. What works when you’re being beaten is to move away from the dude with the baton. It’s a rational solution that many poor people have arrived at using their own faculties of reason.

But surely the police shouldn’t have been beating you in the first place. Protesting against the police, demanding that the police be held accountable, this too can work. But it requires politics, not technocracy. It means ideology, not expertise. For a generation – for a society – that looks to Washington and sees politics hopelessly divided, who wouldn’t want to stick with What Works? Except, of course, that Washington isn’t terribly interested in much beyond a tiny spectrum of difference between plutocrats whose bickering has brought the very idea of politics into disrepute. It’s a context in which Banerjee and Duflo’s reasonableness is refreshing. Even if their own politics crushes those of poor people. And even if what you’re left with is a few splats of policy ideas, because the theory that might have connected and made sense of them has been erased, because methodologically suspect and politically polluted.

If you want to read a general critique, have a look at Sanjay Reddy’s piece, one of the best in the genre. And if you want to see how Banerjee and Duflo get it wrong on food, here’s something hot off the presses. An old friend and teacher – Porus Olpadwala – has just published a fine analysis in Social Scientist. He takes Banerjee and Duflo to task for their 2011 Foreign Policy piece. The PDF’s below the fold. Social Scientist Sept.-Oct. 2012