Search

This dissertation examines several aspects of labor markets and public policy. Chapter 1 examines the prevalence of unobserved sub-state policy in the US and provides a novel method for estimating consistent effects of these unobserved policies. Chapter 2 estimates the spillover effects of minimum wage policies on self-employed workers. Chapter 3 explores a new method for estimating differences between genders in incentive responsiveness. Chapter 1 provides evidence on the effects of treatment contamination and non-compliance in estimates of the impact of state-level public policy. When treatment may varying within jurisdiction, estimating the effects of policy and failing to control for within-jurisdiction variation will bias estimates away from their true values. Even when the non-compliant population is small estimates may not obtain the true effect. This chapter shows that combining standard intention-to-treat corrections for non-compliance with two-sample instrumental variable methods can obtain consistent estimates of the effect of interest, even when the non-compliant population is unobserved in the data. Chapter 2 studies the effects of minimum wages on unincorporated self-employed workers in the US using the 1988-2020 Current Population Survey. Standard state-level difference-in-differences estimates of employment and earnings elasticities find that increasing the minimum wage tends to decrease self-employment but has little effect on hours worked or earnings in the year following the change. Instrumental variable estimates are consistent with these findings but show large earnings gains for these workers. Using a simple model of labor market search, I show that minimum wage increases are potentially welfare improving and the welfare effects can be identified by changes in self-employment. Over this period empirical estimates of minimum wage effects indicate that these policies were on average welfare improving. Chapter 3 examines gender differences in responsiveness to incentives by modeling players' continuation value in head-to-head competition directly using real-world betting odds and data from the 2011-2019 tennis Grand Slam tournaments. In early round match-ups men tend to exert more effort than women given the same continuation value, but when examining competitors of similar ability across the whole tournament differences in responsiveness disappear. For a 100,000 unit increase in the expected value of winning a match, men tend to increase their probability of winning by 0.9 percent, versus 0.7 percent for women. These results highlight the importance of the value of remaining in the competition verses the contemporaneous round prize in these settings.