Poverty: New ideas about an old problem

Poor Economics: A Radical Rethinking of the Way to Fight Global Poverty. By Abhijit Banerjee and Esther Duflo.

At Free Exchange there is a series of commentaries on the new book. E.g., “Health: unpicked low-hanging fruit“. There are some findings here that are sincerely new in development economics. We know of course that “incentives matter”. So what do you do when the rural parents of Rajasthan do not take their children for free immunizations? Well you could offer a small bag of lentils as a ‘reward’:

(…) The authors recount an absolutely fascinating series of experiments designed to get more people to immunise their children in rural Rajasthan. Again: hugely beneficial action, and thanks to the NGO they worked with, readily and freely and reliably available. But also something with a small cost—perhaps a couple of hours off work and the trudge to the clinic, and benefits that are intangible and lie in the future. Result: abysmal immunisation rates.

Their idea: a small bag of lentils given as a sort of “reward”. This was opposed by public health officials, who thought “bribing” people to do what they should do anyway was a bad way to go. Yet it had a dramatic effect—and actually reduced the cost per immunisation to the NGO, because the nurses who had to be paid for the whole day anyway were now busier. Yes, convincing people of the benefits is probably useful in the long run, but this does the trick much better and more quickly—and, possibly, experience with immunisation is a pretty valuable kind of “convincing”. And yes, it’s paternalistic. But a whole host of things are essentially done for us—often by a paternalistic state, which purifies our drinking water and provides sewage systems and so on. There are many, many areas where we simply do not have to take responsibility because stuff is done for us, or made incredibly easy. But the poor must actively decide to “do” them.

The Economist review – excerpt:

The two economists made their names (and remade their discipline) by championing randomised trials. These trials test anti-poverty remedies much as pharmaceutical firms test drugs. One group gets the remedy, another does not. The two groups are chosen at random, so the remedy should be the only systematic difference between them. If the first group does better, the benefit can be attributed to the project and not the many other factors that might otherwise obscure the result.

These trials proved immensely appealing. They promised to sift nuggets of truth from the slurry of received wisdom and wishful thinking that characterises much aid-talk. The hope was that once a trial proved the worth of a project or programme, governments and donors would back it and prescribe it more widely.