Essays in Economic History and Applied Microeconomics


This thesis consists of three papers studying applied microeconomics and economic history. The chapters are broadly organized around two different topics: the political effects of media and the consequences of disruption to human capital. In the first chapter, joint with Susan Ou, we study the role of media in the transmission of ideology during the Cultural Revolution. We develop a novel identification strategy by interacting the strength of radio signals and linguistic compatibility of local dialects to the broadcast language, Mandarin. A stronger signal is found to increase revolutionary intensity in counties where Mandarin was better understood. Through investigation of participation in the Send Down Movement, we provide evidence that one mechanism underlying our findings is the direct effect of exposure on individuals, even absent differences in local policies induced by media. The effects of propaganda are persistent, as evidenced by Communist Party membership in later life. The second chapter examines the career of Ronald Reagan to study the political consequences of television stardom. I exploit quasi-experimental variation in television reception to provide causal estimates of the effect of celebrity exposure on political outcomes. I find that Ronald Reagan's tenure as the host of a 1950s entertainment television program translated into electoral support during his initial bids for the presidency nearly two decades after the show's last airing. The effect is especially pronounced in Republican primaries relative to the general election and dissipates entirely in locations where Reagan was a known political entity. I find evidence consistent with two distinct, but not necessarily exclusive, mechanisms. For politically uninformed voters, non-political media acted through the channel of name recognition. For the uninformed and informed alike, celebrity exposure idealized perception of Reagan's traits and personalized political considerations in elections featuring him. In the last chapter of my dissertation, I study broadly speaking how criminals cope with the loss of their criminal human capital. It is widely hypothesized that legalization disrupts illicit markets and displaces illegal suppliers, but the consequences for those who are displaced remain poorly understood. In this paper, I use comprehensive administrative data on the universe of offenders in three states that legalized marijuana to study the effect of the policy change on the subsequent criminal and labor activity of convicted dealers. I find that marijuana legalization increased the 9-month recidivism rate of marijuana offenders by 5 percentage points relative to a baseline rate of 11 percent. The results are not explained by changes in enforcement. Rather, the increased recidivism is driven by substitution to the trafficking of other drugs, which is consistent with a Becker-style model where individuals develop human capital specific to the drug industry. Using the NLSY97, I show evidence of legalization-induced displacement even amongst non-convicted dealers. In contrast, the transition to formal employment appears much more modest. To learn about potential mechanisms behind these results, I use transaction-level data to estimate the effect of legalization on average prices and price dispersion. I provide suggestive evidence that both the price level and residual variance declined following legalization, consistent with legalization eroding rents earned in the illicit marijuana market. Overall, the results in this paper suggest that an unintended consequence of selective legalization is a re-allocation of drug criminals to other illicit activity.

Alternate Identifier
Date created
Resource type
Rights statement