During my two-year time in Cambridge I hadn't had too much opportunity to interact with her. But I learned much about her works and MIT-based Poverty Action Lab, the organization he leads with Abhijit Banerjee and Racherl Glennerster. Her most cited work was, of course, her dissertation-turn-seminal paper on measuring the return on education using Indonesian data. In the paper, she exploited the fact the SD INPRES policy in the 1970s could serve as a exogenous quasi-experiment. By comparing the average years of education and wages between the pre- and post-INPRES cohorts, she corrected the endogeneity and omitted variable biases.
In 2008 I had a privilege to get involved as one of the te aching assistants in the Poverty Action Lab's short course in Bali. There I had some chance to discuss some technicalities in doing randomized experiment studies with her and her co-workers. I also learned that she likes to go hiking and mountain climbing. In fact, after the Bali course finished, she and Pascaline Dupas went to Lombok to climb Rinjani.
This week the American Economist Association awarded her the John Bates Clark Medal (see the story here and here). It is a highly prestigious award, given every 2 years to "American economist under 40 who have had a significant contribution to economic thought and knowledge." (Note: Prof. Duflo is now 37). Former medalists include Milton Friedman, Joseph Stiglitz, Paul Krugman, and recently Steven Levitt, Daron Acemoglu and Susan Athey. Duflo is the only second medalist after Susan Athey. And in the past, Clark Medal is a good predictor for winning the Nobel Prize.
Prof. Duflo had her Ph.D from MIT, and has been working there since then. In the US academic job market (in economics), it is very unusual for a Ph.D to work in the same institution immediately.
If I can summarize her contribution, it will be 'helping to understand which development policy or initiative work, and measure the impact, using randomized experiments.' True, there has been so many debates over the use of randomized experiments. The fact that is has been debated only shows its importance. Although Prof. Duflo was not the inventor of the technique, she and her co-workers are the ones to make it popular not only among academia, but also for policy makers and pracitioners.
For those who have always been skeptical, even hostile, to economics and economists, please spare some time to take a look at her and the Lab's website, and see how diverse the discipline has been.