Esther Duflo, Clark Medalist 2010


Duflo has distinguished herself through definitive contributions to the field of Development Economics. Through her research, mentoring of young scholars, and role in helping to direct the Abdul Latif Jameel Poverty Action Lab at MIT, she has played a major role in setting a new agenda for the field of Development Economics, one that focuses on microeconomic issues and relies heavily on large-scale field experiments. Much of her work addresses questions of politics, gender, and education. She has written extensively on India, but has also studied Indonesia, Cote d'Ivoire, South Africa, and Kenya. This bio summarizes some of the highlights of her research.

Policy Evaluation through Randomized Field Experiments

Duflo has been a leader in using randomized field experiments to address important questions concerning public policy in developing countries. In one series of papers, she and various coauthors study the impact of female political leadership on local government spending and attitudes toward women by examining a policy that required one-third of India’s villages, selected at random in each election cycle, to choose a woman as council head. She finds, for example, that villages forced to choose female council heads shift local government spending away from education and towards drinking water and (in some areas) roads. Another series of papers measures the effects of various randomized educational interventions, such as the introduction of teacher aids (local women with some secondary education but no formal teacher training), a computerassisted math program, reductions in pupil-teacher ratios, ability grouping, various forms of monitoring and incentives to reduce teacher absenteeism, and programs designed to encourage community participation. For example, in work with Abhijit Banerjee, Rukmini Banerji, Rachel Glennerster, and Stuti Khemani, she investigates the effects of three randomly assigned programs targeting community participation. The first informed villagers about opportunities to participate in school governance and monitoring committees, the second trained villagers to use a testing tool, and the third organized literate villagers to hold remedial reading classes for illiterate children. The study concludes that the interventions had no impact on community involvement in the schools, no impact on teacher effort, and no impact on students' achievement in school.

In work with Emmanuel Saez, Duflo has also used field experiments to study issues concerning U.S. pension policy. One article reports the results of a study in which a random sample of employees in some departments of a major university were provided with monetary incentives to attend a benefits fair. The incentives increased attendance of both the targeted employees and employees in the same departments who were not targeted, and significantly raised enrollment in tax deferred accounts for both groups. Thus, the effects of the intervention appear to have been transmitted through social channels.

Policy Evaluation Using Non-Experimental Methods

Duflo’s contributions to Development Economics also include a collection of important studies that use nonexperimental methods. Her earliest work evaluates the impact of a massive school construction program in Indonesia, which the government initiated during the 1970s. She asks whether construction projects improved local educational and employment outcomes relative to areas in which no construction occurred. She finds only weak effects on educational attainment and early wages; moreover, the effect on wages disappears rather quickly as the cohort ages. Duflo attributes this finding to a reduction in the returns to education caused by the rapid increase in the supply of better-educated workers. In another paper, Duflo asks whether an increase in household income benefits children more if the income recipient is female. She exploits the fact that South Africa dramatically increased pensions for elderly blacks in 1991. She finds that if the family contained an elderly female, the increase in pension benefits coincided with a significant gain in the body mass index of girls. No such increase occurs if the family contains an elderly male. Finally, in a 2 paper that generated considerable popular attention, Duflo and Rohini Pande examined the effects of dam construction on economic development in India. The study finds beneficial effects on poverty rates and agricultural productivity for downstream districts, which presumably benefited from improved irrigation, but detrimental effects for the districts in which the dams were built, possibly due to displacement and degradation of water and soil. Their calculations suggest that the construction of large dams in India has marginally positive returns to the entire economy but negative distributional consequences, including an overall increase in India's poverty rate.