Abstract: This paper studies the health consequences of rural-to-urban migration, leveraging twenty years of longitudinal microdata on Kenyan individuals and a "mover" design. Controlling for individual fixed effects, urban migrants self-report better health and substantially lower incidence of fever, both for themselves and their children. These gains are not offset by increases in cough, diarrhea, or depressive symptoms, despite concerns about pollution, poor sanitation, and social isolation in growing African cities. Urban migrants also report increased healthcare expenditures, consistent with greater accessibility of care. Finally, they gain substantial weight, with corresponding increases in spending on protein and calorie-rich items such as meat, dairy, and processed foods. Overall, these results suggest that Africa's "urban health advantage" may at least partially reflect place effects, rather than selection of healthy individuals into migration.

Abstract: Taller people earn more, especially in low- and middle-income countries. We present among the first evidence of this phenomenon in Africa, using longitudinal microdata on a cohort of middle-aged Kenyan adults. We document a substantial height/earnings premium: controlling for gender, age, and other socio-demographics, monthly earnings increase by 1.07% per centimeter (or 2.72% per inch). Nearly half this effect can be explained by differences in cognition, measured from an unusually rich battery containing 27 modules. Additional shares of the premium can be attributed to measures of physical strength and non-cognitive ability. In contrast to prior work, we find little role for occupational sorting: conditional on cognitive and non-cognitive ability, taller people do not appear more likely to work in higher paid sectors. Leveraging repeated measures of height and an instrumental variables specification, we find suggestive evidence that measurement error may be attenuating the estimated relationship.

Abstract: In 1984, USAID launched the Demographic and Health Surveys program, an international effort to improve data on health and fertility in the developing world. I leverage the program's staggered introduction across countries, evaluating its effects on the geography of economic research. The DHS expanded the geographic coverage of economics, increasing the likelihood of a publication on a country each year by 7 percentage points, 26% of the control mean. However, it decreased the likelihood of a publication with a domestic coauthor. This suggests that access to data opens research opportunities in neglected regions, but reduces reliance on local collaboration.

Abstract: This paper studies how access to local research infrastructure shapes the production and dissemination of scientific research. In particular, I study how the Abdul Latif Jameel Poverty Action Lab (J-PAL) and Innovations for Poverty Action (IPA) — two organizations supporting randomized experiments in economics — have shaped the direction of research and policy, leveraging geographic and temporal variation in their office openings and a generalized difference-in-differences design. The opening of a J-PAL (IPA) office increases the number of RCTs published in a given year about a country by 3.7 (0.95) on average. This effect is robust to a variety of controls, estimators, and alternative modeling assumptions, and is not driven by RCTs published by the organizations’ founders. This marginal evidence diffuses to policymakers. After office openings, country governments publish substantially more policy documents referencing experimental evidence. These results suggest that investments in local research infrastructure can shape the direction of science and the ensuing policy response.