A View from Jason Pontin
TED Day 1: Esther Duflo
An MIT economist wants to use the techniques of randomized trials to study international aid.
I am in Long Beach, California, at TED, the invitation-only event where technologists and scientists, venture capitalists and investment bankers, artists and designers (as well as a handful of bemused celebrities) meet every year to be educated and moved. TED (which stands for “Technology, Entertainment, and Design”) is intellectual theater of the highest order: for three days, attendees watch short, 18-minute presentations by passionate speakers on a bewildering but exciting range of themes. It is, by far, my favorite event of its sort.
On the first day I watched Daniel Kahneman, the founder of behavioral economics, ask what is the difference between the remembering and experiencing selves; William Li, an oncologist, argue that anti-angiogenesis is the most promising mechanism for preventing cancer; Dan Barber, the chef of Blue Hill Farm in Manhattan, insist that we can feed nine billion mouths with sustainable farming; and I heard Jake Shimabukuro play Queen’s “Bohemian Rhapsody” on the ukulele.
Amongst such an interesting range, it is hard to select just one speech for particular attention, but I most enjoyed the presentation by Esther Duflo, a 2009 MacArthur Fellow and a developmental economist at MIT, who has discovered an intellectual breakthrough in measuring the effectiveness of international aid. At MIT’s Poverty Action Lab, Duflo has applied the techniques of randomized drug trials to study what actually works and what doesn’t in funding programs in the poor and developing world.
Rather than trying to answer the big question, “Does aid work?” (a highly contentious, much-studied subject, which Duflo argues is essentially unanswerable), Duflo asks small questions. Will poor villagers in Rajasthan be more likely to be immunized if they are paid to get their shots? Will East Africans be more likely to use antimalarial bed nets if they are given the nets for free, or if they are asked to pay for them (on the grounds that poor people will value them more). Did microfinance in Hyderabad, India, actually work? She studies the questions by running parallel experimental programs in the regions. The answers were surprising. In the case of micofinance, for instance, it turns out that if a family already had a small business, microfinance made the family richer. But other families became poorer. And, contrary to years of hyperbole about microfinance, Duflo found “no impact on measures of health, education, or women’s decision-making.”
You can watch Duflo speak about poverty here.
AI is here. Will you lead or follow?
Join us at EmTech Digital 2019.