Research

I study the causes and consequences of disparities in human capital within developing economies and the policies that can remedy them. I focus on the historical and social contexts of these inequalities so that development policymakers can properly tailor programs to the communities they serve.

Working Papers

Colonial Institutions, Marriage Markets, and HIV: Evidence from Mozambique

During the colonial period, several types of extractive institutions organized much of economic activity in Sub-Saharan Africa. Historians argued that two of them—one pushing men into circular migration and one restricting their mobility—had different effects on marriage markets with bride price but not on development. Specifically, young men in a migrant-sending institution could make bridewealth payments, narrowing the spousal age gaps that would later increase HIV risk. But much of their wages was captured by elders through inflated bride prices instead of being saved, limiting circular migration’s development potential. To compare these extractive regimes, I exploit the arbitrary border within Mozambique that separated the two institutions for a half-century (1893-1942). In the colonial era, spousal age gaps were smaller in the migrant-sending region, even after the border was erased and circular migration rates converged. Today, HIV prevalence is substantially lower in this area, likely due to narrower age gaps between partners, but development outcomes are similar. These results show how different forms of colonial extraction affect health and wealth in Africa, and that marriage markets are a channel through which historical events shape the present.

Work in Progress

Childhood Health, Marriage Markets, and Young Women's HIV: Evidence from Deworming in Zimbabwe

Young women comprise a disproportionate share of new HIV infections in Africa, largely due to features of the region's marriage markets like age-disparate relationships. Because marital prospects are shaped by human capital, the stock of which is influenced by childhood health, can improving the health of older girls lower their chances of contracting HIV as young women? I examine a nationwide school-based deworming program in Zimbabwe (2012-17) that substantially reduced rates of urogenital schistosomiasis, a parasitic worm disease that can limit learning and school attendance. Using a difference-in-differences design, I find that 3 years after deworming began, HIV prevalence among women 18 to 20 had fallen 4.3 percentage points (p.p., 47 percent) more in formerly high-schistosomiasis districts than in low-morbidity ones. Human capital's effects on marriage markets appear to explain the results: these young women were 7.1 p.p. (51 percent) more likely to still be in school, age gaps with their partners fell by an additional 2 years (30 percent), and HIV risk factors associated with age-disparate relationships decreased more as well. These results show that a cheap treatment for a common childhood disease can also be a highly cost-effective method of combating one of the modern world's deadliest pandemics.

Disease, Disparities, and Development: Evidence from Chagas Disease in Brazil (with Eduardo Montero)

Childhood human capital accumulation is a major channel through which disease affects economic development, but its benefits are not fully realized for decades and they may not be enough to justify interventions on their own. Therefore, it is important for policymakers to understand if disease control has development-relevant impacts in the short run and in other domains in the long run. We provide evidence of such benefits by studying Brazil's initial campaign to eliminate vectorial transmission of Chagas Disease (1984-89), an illness causing 1 to 3 months of acute symptoms after contracting the parasite, and chronic cardiovascular problems 10 or more years later in a large share of those infected. Using a difference-in-differences strategy, we make pre- and post-campaign comparisons across states and municipalities with varying levels of pre-treatment vector prevalence. We find that adults' employment rates rose shortly after spraying began, cohorts treated as children had higher incomes as adults, and both of these effects were much larger for non-white Brazilians. Using a triple-differences strategy, we also show that Brazil's government-run health care system--which consumes 4 percent of GDP--spent substantially less on hospitalizations due to circulatory diseases than other causes following the decline in chronic Chagas Disease. These results show that disease control has important short- and long-run impacts on developing economies not presently accounted for in cost-benefit analyses, and that combating neglected tropical diseases can help to speed convergence in societies with large racial disparities.

The Promises and Pitfalls of Using DHS Data for Comparisons across Countries and National Borders

Blood Taxes: The Development Effects of Military Conscription in Colonial French West Africa (with James Allen IV)

Data collection in progress at the Archives Nationales du Sénégal, Dakar, Senegal