The noisiest debates on poverty focus on the amount of money that needs to be put to work to help the poor—be it rock star Bono’s campaign to get rich countries to give more aid to Africa or the very public declarations of the Indian government to spend more on education, health and infrastructure to ensure inclusive growth. Unfortunately, these campaigns and policy statements tell us little of value about how money is to be actually spent. Too much money has gone down the drain.
That’s where a relatively new research tool can be of great use—randomized trials. The Abdul Latif Jameel Poverty Action Lab at the Massachusetts Institute of Technology (MIT) has been one of the pioneers in the use of a technique borrowed from the world of medicine. Doctors who want to understand how well a certain drug works as an antidote to a certain illness will divide a group of patients selected at random into two clusters. One will be given the new drug and the other will not. The progress of the two groups of patients is monitored to see whether there is a significant difference between the health parameters of those who were given the drug and those who weren’t. That helps establish the efficacy of a new drug.
MIT economists such as Abhijit Banerjee and Esther Duflo, as well as Sendhil Mullainathan of Harvard, have adapted this technique to study various poverty-related issues. Take the issue of education. What does it mean when we say that something needs to be done about the horrid state of municipal education in India? Should their budgets be increased? Are more teachers needed? Would computers do the trick? The standard “let’s throw some more money at the problem” does little to clarify the policy choices.
Two experiments conducted by the Poverty Action Lab gives us clear clues. In one done with Pratham, an NGO, we get a clear sense of what can be done. Pratham hires young women living in local communities to provide remedial education to children who have moved to Class II and III without acquiring the basic literacy and numerical skills that they should have in Class I. The Poverty Lab research showed that the remedial education provided by these young women (called balsakhis) helped improve exam scores. A cost-benefit analysis showed that it was between 4.5 and 6.7 times more cost-effective than hiring a new teacher for those kids.
A similar piece of research in Vadodara showed that computer-assisted learning could provide similar improvements in marks, but that it was far less cost-effective than the balsakhi programme. Such randomized trials give us clear signs that balsakhis are more effective that either employing more teachers or giving computers to government schools. God is in such details.
Many of the Poverty Action Lab’s most interesting projects are being conducted in India. So it is fitting that the lab is now setting up base in Chennai. Infosys chief mentor N.R. Narayana Murthy will inaugurate the lab’s new South Asia centre on 13 July
Randomized trials are not without their critics. Some have questioned, the ethics of such trials, since one group gets “help” and another doesn’t. How fair is it to, for example, give iron supplements to one group of poor village women and not give them to another? The other set of criticisms are of a different nature. Nobel economist James Heckman, for instance, says that such trials do little to explain why a particular policy or intervention succeeded or failed. (His 1995 essay on this predates the setting up of the Poverty Lab.) The point is: balsakhis do more for remedial education than new teachers. But why? Heckman also says that these trials do little to estimate structural parameters, such as the labour market response to a tax cut.
Despite these criticisms, there is little doubt that the techniques used by the Poverty Lab are a great leap forward from the rather empty homilies that often pass off as development policy in India. I am reminded here of something said by William Easterly, a thorn in the side of the development economics establishment, in a speech he gave in January 2006. “Historically, poverty has never been ended by central planners. It is only ended by searchers, both economic and political, who explore solutions by trial and error… A Planner thinks he already knows the answers: he thinks of poverty as a technical engineering problem that his answers will solve. A Searcher admits he doesn’t know the answers in advance: he believes that poverty is a complicated tangle of political, social, historical, institutional and technological factors.”
The Indian development discourse is all about budgets—how much will be spent? There is little clarity on what needs to be done to improve education, health, access to finance, the use of subsidies, etc. Social policy is still dominated by what Easterly would call the planning mentality. We need a bit more searching.
Your comments are welcome at cafeeconomics@livemint.com