Understanding the Process of Economic Development: Essays on Culture and Institutions

Abstract

This dissertation explores the origins of various institutions and cultural norms and the implications of these institutions and norms for economic development and development policy. In the first paper, I examine how matrilineal kinship systems affect spousal cooperation. In matrilineal kinship systems, lineage is traced through female members. I use experimental and physiological measures and a geographic regression discontinuity design along the "matrilineal belt" in Africa to show that matrilineal individuals are less cooperative with their spouses. However, children of matrilineal women are healthier and better educated, and matrilineal women experience less domestic violence. The results highlight how broader social structures affect bargaining within the household. The second paper, joint work with Eduardo Montero, examines how historical exposure to extractive institutions affects long-run development in the case of the Congo Free State. The CFS granted concessions to private companies that used violent tactics to force people to collect rubber. We use a geographic regression discontinuity design along the concession boundaries to show that greater exposure to extractive institutions causes significantly worse development outcomes. We explore the role of institutions and culture as channels. Chiefs within the former concessions provide fewer public goods and are less likely to be elected. However, individuals within the concessions are more pro-social. The results suggest that events of short duration can affect development through changes in institutions and culture. In the final paper, joint work with Eduardo Montero, we examine the legacy of French colonial medical campaigns for trust in medicine. Between 1921 and 1958, the French military organized medical campaigns focused on treating sleeping sickness. The campaigns forced individuals to participate and used medications with severe side effects. We digitize annual colonial records with the locations of campaign visits for Central Africa. We find large and significant positive effect of historical exposure to campaigns on refusal to consent to a blood test -- a revealed preference measure of trust in medicine. The results provide evidence on how historical events can affect the efficacy of present day health policy.