Essays in labor and health economics

This dissertation consists of three essays in empirical labor and health economics.

The first chapter examines how the amount of time devoted to a leisure activity varies in response to temporary changes in the price of that activity. Specifically, I estimate the effect of changes in expected winnings in an online poker game on the probability that players quit playing. I find that expected winnings have a large negative effect on the probability that a player quits playing poker. A one dollar increase in expected winnings decreases the probability that a player quits playing altogether by 0.5 percentage points, compared to the mean of 1.1 percentage points. This corresponds to a price elasticity of demand for poker of -0.14.

The second chapter develops and tests a model of how college students choose their field of study. The model combines features from learning and human capital models and captures several stylized facts from the empirical literature on choice of college major. I test the model's predictions using High School and Beyond data. I find three results that generally agree with the model's predictions. First, students with higher levels of ability choose majors with higher average earnings. Second, students who receive low grades in college are more likely to change their field of study. Third, students who switch majors in college subsequently earn less than students who do not change majors, but this difference is primarily due to major-switchers obtaining degrees in low-paying fields.

The third chapter, coauthored with Abhijit Banerjee, Esther Duflo and Gilles Postel-Vinay, provides estimates of the long-term effects on height and health of a large income shock experienced in early childhood. Phylloxera, an insect that attacks the roots of grape vines, destroyed 40% of French vineyards between 1863 and 1890, causing major income losses among wine growing families. We exploit the regional variation in the timing of this shock to identify its effects. We find that, at age 20, those born in affected regions were about 1.8 millimeters shorter than others. This estimate implies that children of wine-growing families born when the vines were affected in their regions were 0.6 to 0.9 centimeters shorter than others by age 20. This is a significant effect since average heights grew by only 2 centimeters in the entire 19th century. (Copies available exclusively from MIT Libraries, Rm. 14-0551, Cambridge, MA 02139-4307. Ph. 617-253-5668; Fax 617-253-1690.)