Economists and economic historians attribute long-term underdevelopment in former colonies to either persistent, bad institutions or to initial factor endowments that predisposed the colony to inequality. This paper uses a puzzle in Haiti’s economic history to show that these frameworks are incomplete. In the early 20th century, large numbers of Haitian workers migrated abroad to work on plantations, even though lots of fertile land sat idle at home. This puzzling fact reflects two land institutions developed after Haiti’s independence in 1804. First, lineages had joint claims over the alienation of land; one household could exploit land, but to sell it they needed agreement from a large number of extended kin. Second, the early-nineteenth century Haitian government had distributed land and banned land ownership by foreigners, preventing the creation of large holdings and establishing a checkerboard of landholdings with multiple claimants. To assess how this checkerboard led to idle land, I use data on 5,700 plots adopted over 22 years under a government rental program. A simple model of the optimal allocation of labor (between Haiti and elsewhere) and land (between subsistence holdings and plantations) implies that the checkerboard reduces land adoption and attenuates the extent to which new plantations develop after a migration cost shock. Data from settlement patterns in Haiti and a massacre in the Dominican Republic confirm these predictions. Using the institutional histories of the Dominican Republic and Jamaica as counterfactuals drives this point home: both countries lacked Haiti’s checkerboard pattern and developed plantation agriculture. Haiti’s experience shows that the conditions created by colonial governments did not necessarily constrain future institutional development. Haiti’s post-independence institutional innovations undermined the conditions for development.
Measuring the Effect of Strengthening Property Rights on Political Support in the Presence of Voter Fraud
Many developing countries have not established secure property rights even though research has overwhelmingly concluded that they significantly affect economic development. But leaders might only implement reforms when they can capture some of the benefits since they hold different asset portfolios and face different property rights. To test this hypothesis, we can look at the effect of reforms on political support. Measuring political support in this context is difficult because countries with weak property rights protections also likely have weak democratic institutions. I show that in the presence of fraudulent voting we can only identify the elasticity of support with respect to the property rights reform. I then examine a cattle registration program in Haiti where the government in 2014 provided titles to cow owners. Cattle and goats are the two most important livestock in a typical Haitian’s asset portfolio, and households with more cows benefited more from the reform. I find that in the 2015 election, areas with more cows relative to goats provide more supported to the incumbent party. Because the ratio of cows to goats is exogenously affected by terrain (since goats have a comparative advantage in mountainous areas), I use the region’s slope as an instrument for the ratio of cows to goats and confirm the causal effect. The government improved property rights because it had an incentive to maintain power.
We use a differences-in-differences design to measure the effect of UberX on public transit ridership. We find that UberX increases transit ridership for smaller transit agencies, specifically by increasing bus ridership, and has no measurable effect on train ridership or larger transit agencies.
Works in Progress
Refugees and the Provision of Public Goods: Haiti and the 1937 Refugee Influx
With millions of refugees seeking resettlement, we must understand how refugees affect the receiving countries. I contribute to this literature by looking at how refugees to Haiti affected the provision of public goods. Escaping the 1937 Trujillo massacre in the Dominican Republic, thousands of refugees arrived in Haiti and needed government services. The unexpected influx strained the government’s resources and forced refugees to wait years to receive services. After the government increased income taxes, it cleared its backlog and improved program administration. Thus, the refugees provided a short-term strain on public goods, but in the long-run improved government services.
Council Composition and Public Good Provision: Evidence from Kerala, India (with Sabyasachi Das)
College Attendance on the Intensive Margin: Credit Constraints, Student Employment, and the Return to Education (with Joseph Altonji and Seth Zimmerman)
Ratings and Revenues: Evidence from Movie Ratings, (with Joe Price and Jared Shores) Contemporary Economic Policy, January 2013, Vol 31, Issue 1, pp 13-21
Taxing the Opposition: Cactus League Attendance and the Efficiency of the ‘Cubs Tax,’ (with Michael Davis and Joe Price) International Journal of Sport Finance, May 2013, Vol. 8, No. 2, pp 157-17
What Matters in Movie Ratings? Cross-country Differences in which Content Influence Mature Movie Ratings. (with Joe Price and Doug Gentile) Journal of Children and Media, February 2014, Vol. 8 , Iss. 3, pp 240-252
Technological change, relative worker productivity, and firm-level substitution: Evidence from the NBA (with Grant Gannaway, Joe Price, and David Sims) Journal of Sports Economics October 2014; vol. 15, 5: pp. 478-496