Three Essays in Applied Microeconomics

This dissertation examines three topics in applied microeconomics. In the first two chapters, novel data sets are used to explore two labor economics topics – how worker output varies under an alternative payment scheme, and the effect of discrimination on the prospect of being hired. The final chapter resides at the intersection of environmental and urban economics, and uses restricted data to estimate the impact of resource booms on the incidence of local crime. The first chapter, written with Austin C. Smith, examines how alternative payment structures can lead to both disincentive and selection effects that reduce worker output. Liquidity constraints can distort efficient investment across a variety of domains, for both firms and individuals. While debt financing is often used to address liquidity constraints, especially at the individual level, there has been a recent push towards Income Share Agreements (ISAs) – equity contracts in which individuals can raise money today by selling shares of their future income. While ISAs eliminate the need for traditional collateral and allow liquidity constraints to be addressed in new markets, this tool brings with it the classical problems associated with asymmetric information – adverse selection and moral hazard. Individuals may select into an ISA if they have private information that their expected income is lower than it appears to investors (adverse selection), and participants may rationally choose to exert less effort or to deviate from otherwise optimal behavior, given that they only reap part of the reward (moral hazard). Using a novel panel data set that tracks individuals as they participate in very short-term ISAs – we find evidence of both issues, with a relatively larger decline in earnings due to moral hazard. In the second chapter I investigate how the relationship between employer- and employee-type affects the probability of being hired by a particular firm; type is determined by the reputation of the college that the employee or employer attended. Statistical theories of discrimination show that asymmetries in information regarding a worker’s productivity can lead to adverse outcomes for both workers and firms. Screening discrimination, a type of statistical discrimination, predicts that employers are more likely to hire an applicant of their own type rather than an applicant of another type. That is, unlike standard statistical discrimination models, screening discrimination allows minority-type employees to be preferred to majority-type employees when the employer is also the minority-type. This paper tests this central prediction of screening discrimination and uses college reputation as the basis for employer- and employer-type. I find that employers are 24% more likely to hire a candidate when the college-type of employer and candidate matches. Further, this probability increases as the opportunity cost of making a hiring decision increases. Finally, I find mixed evidence as to whether or not the initial employee-employer type match was predictive of future productivity. The final chapter of this dissertation, written with Patrick Gourley, focuses on how resource booms can lead to changes in incidences of local crime. The 21st century oil and gas boom is drastically changing life in the American West. While previous literature has examined how resource booms affect household income and infant health, the effects of resource booms on crime remain largely unstudied. We develop a simple model that demonstrates an oil and gas boom could increase or decrease crime as employment opportunities, inequality, and other aspects of the local economy change. Combining well data provided by drillinginfo.com and FBI incident-level crime data, we examine intra-county changes in both property and violent crime in Colorado as wells open and shut down. Over a large range of well activity, we find a positive relationship between the number of active wells i