Research

Working Papers

The Impact of Conflict Exposure in Utero on Cognitive Test Scores (Job Market Paper)

This paper studies the persistent effects of in-utero exposure to violent conflict on human capital formation. I exploit temporal and spatial variation of violent conflict in Kenya over the period from 1989 to 2011 to estimate the impact of conflict exposure in utero and early life on cognitive test scores of school-age children. A unique data set of standardized test scores for more than 330,000 children allows me to control for household fixed effects, thereby eliminating various selection biases. I find that children exposed to conflict have significantly lower test scores in math, reading English and Swahili. Exposure to conflict in utero decreases test scores by 3% of a standard deviation, while effect sizes double for the most intense periods of violence in Kenya. Negative impacts of conflict exposure are largest for in-utero exposure, diminished for exposure before age two, and close to zero after age two. Additionally, I provide evidence for adverse effects on child health. Experiencing conflict in utero decreases height-for-age and increases the risk of infant mortality. In particular, the third trimester is identified as the most sensitive period for effects on height-for-age.

Ethnic Favoritism and Cabinet Positions in sub-Saharan Africa

This paper studies how the distribution of political power across ethnic groups interacts with political institutions to promote ethnic favoritism in the provision of public goods. Using a data set with 142 ethnic groups from 12 countries in sub-Saharan Africa, I estimate the effect of changes in cabinet shares of ethnic groups on education and infant mortality outcomes over the period from 1960 to 2004. I find a positive effect of cabinet positions on primary schooling outcomes, but no effect on infant mortality. A 10\% increase in cabinet shares is estimated to increase the probability of attending and completing primary school by 3\% on average. There is substantial variation in effect size across countries. Political institutions shape incentives and constraints of ministers of cabinet to engage in ethnic favoritism. Paradoxically, the ethnic favoritism effect of cabinet positions gets stronger as countries become more democratic or less autocratic.