Three Essays About Economic and Behavioral Responses to Government Interventions

Abstract

This dissertation studies three different policy interventions and the subsequent labor marketresponses across different times and contexts. In Chapter 1, I study changes in the relative wages for women in manufacturing between 1940 and 1950. World War II saw an unprecedented influx of women onto factory floors. While most previous literature focuses on the effects on female labor supply via geographical variation in military mobilization, Rose (2018) highlights the importance of production demand in driving female wartime employment. Using data on the wartime employment of women from Rose (2018), I revisit the framework in Acemoglu et al. (2004), and estimate the impact on relative wages for women, due to both state-wide and industry-wide changes in production demand during WWII. I find that wages increased for women in 1950 compared to 1940 in Durable Manufacturing by 35.4-35.9% in the industry with the largest change in the relative demand for women during WWII, whereas impacts of state-level changes in demand are not significant. Impacts on wages in Non-Durable Manufacturing are statistically insignificant. The relative wage gains are highest for women with 12 or more years of education, suggesting that the increased demand during WWII allowed some women a “foot in the door” into prized manufacturing jobs. My work helps to reconcile the prior literature connecting WWII to gains in employment for women, and recent work highlighting the importance of the War Production Effort in increasing female employment, by showing how changes in the relative industrial demand for women during World War II significantly increased relative wages for women. In addition, by focusing my analysis on Manufacturing industries, (which saw the largest changes in wartime demand), I consider finer industry variation nationally than any previous work in this context. In Chapter 2, (co-authored with Dheeraj Chaudhary), we test if the intra-state deregulation of banks between 1970–2000 had any impacts on fertility rates. U.S. states deregulated their banking sectors in a staggered fashion between 1960-1999, increasing efficiency through competition between banks and boosting economic growth within a state. We find that deregulated states saw a decrease of approximately 4–6% in their average fertility rates (in both state-level as well as individual-level data) using a classic difference-in-differences strategy leveraging the staggered timing of deregulation across states. In updating our results with recent econometric literature to account for differences in treatment timing, we find that our results are robust for the sample of observations before 1989 but not for later years. Women aged 20-44 saw a decrease of approximately 2-3% on average fertility rates post-deregulation (in both state-level and individual-level data) between 1970 and 1988. We test different possible mechanisms and find that a likely mechanism could be the increased opportunity costs of having children in a growing job market for women, especially in non-white and poorer households. In my third and final chapter, (co-authored with Nolan Pope), we look at the academic impacts of a recent large-scale AC installation program in Chicago Public Schools. Since growing evidence demonstrates that heat impairs student learning, a potential policy solution is clearly investing in air-conditioning. Making use of the timing of roll-out of AC across schools, in a $135 million AC installation program undertaken by Chicago Public Schools between 2013–2017, we analyze the effects on student outcomes. We find no evidence AC installation improved students’ end-of-year test scores or grade retention, and find marginal improvements in attendance. These results indicate that improvements in test scores (or other student outcomes) with AC installation could be region-dependent with the impacts of heat on learning, and considering the returns can help school districts better optimize their often limited budgets when striving to improve student performance.