Using administrative, longitudinal data on felony arrests in Florida, we exploit the discontinuous increase in the punitiveness of criminal sanctions at 18 to estimate the deterrence effect of incarceration. Our analysis suggests a 2 percent decline in the log-odds of offending at 18, with standard errors ruling out declines of 11 percent or more. We interpret these magnitudes using a stochastic dynamic extension of Becker's (1968) model of criminal behavior. Calibrating the model to match key empirical moments, we conclude that deterrence elasticities with respect to sentence lengths are no more negative than -0.13 for young offenders.

Machine learning can be used to help guide and regulate adjudicator decisions, increasing the reliability and overall quality of decision making. The first chapter provides an analytic and normative overview of what I refer to as ``statistical precedent." It explains how statistical models of previous decisions can help assess and improve the reliability of an adjudication system. The subsequent chapters elaborate on and empirically illustrate two of the techniques introduced in the first chapter. Chapter two, using an original dataset of Ninth Circuit Court of Appeals decisions, presents a method for estimating the amount of inter-judge disagreement. Chapter three, using an original dataset of California parole hearings, demonstrates the potential of synthetically crowdsourced decision making.

Fraud against the federal government can be extremely costly, as the Department of Health and Human Services estimated it made $70 billion in improper payments in 2010 under Medicare and Medicaid. Combating fraud is a similarly costly affair, and the federal government has revitalized a Civil War statute to help in the fight. Under the False Claims Act, the government pays private enforcers a bounty to file cases against defendants committing fraud. These private enforcers, typically whistleblowers, have helped the government recover over $18 billion as of 2010.

Private enforcement has a long history under the common law tradition, but its role in a the modern, public enforcement state is less certain. The False Claims Act's private enforcement system, known as its qui tam provisions, specifies government review of private actions in an attempt to curtail past abuses.

In the first part of this dissertation, I evaluate the role of private enforcers and their attorneys in this qui tam system. The government review process allows a standardized reference point for comparing private performance. I argue that it also provides private actors the opportunity to slack by pursuing a “filing mill” strategy. From the data available, however, I do not find law firms aggressively submitting cases under such a high volume, low effort strategy. Rather, law firms and the government appear to be cooperating as intended under the statutory design.

In the second piece of this dissertation, I address the question of the optimal bounty percentage for a finder's fee in the False Claims Act. The statute current specifies a minimum 15% bounty for information leading to a successful prosecution of fraud. I consider the responsiveness of private enforcers to variation in the bounty percentage. Using variation from a 2004 change in the tax code, I find evidence that private enforcers are more willing to bring new cases valued under $440,000 given an approximate 23% increase in bounty.

For the final piece of this dissertation, I consider the bounty percentage for private litigation in the qui tam process. Unlike the finder's fee bounty, I suggest that a 100% litigation bounty may be useful for both compensatory and deterrence purposes. Although a government agency concerned about compensation for fraud losses might initially be concerned about granting a 100% litigation bounty, I argue that the threat of such a litigation bounty may result in additional compensation. The Department of Justice should have the discretionary power to grant high litigation bounties.

This dissertation presents three empirical studies of the American intellectual property system.

In the first chapter, my coauthors and I explore the extent to which copyright’s incentive function is aligned with the economic realities of the commercial music industry. Using a novel panel dataset of music sales and streams, we find that that most commercial music—even blockbuster releases—has an extremely short commercial “half-life.”

In the second, I investigate the predictability of patent litigation, the outcome of which is typically thought to be difficult to forecast. Using a computational model of the text of asserted patent claims and other characteristics of the patent and litigation, I find that decisions about patent validity are more predictable than previously recognized.

Finally, I replicate several leading descriptive studies of patent litigation, and find that their substantive conclusions are sensitive to coding choices and the period over which the sample is drawn.

Law and politics are inseparably connected, with numerous studies showing that politics plays an influential role in legal decision-making. And law schools primarily exist to transmit and produce knowledge. So the political demographics of the legal academia has the potential to significantly influence the production of legal knowledge in the United States. Yet while previous studies have shown that the political demographics of the legal academy tilt strongly in one direction, no one has yet sought to understand what some of the mechanisms might be that cause this disparity. Further, no one has attempted to document the production of knowledge in an increasingly partisan area of the law: religion.

This dissertation aims to begin filling those gaps. It does so through three empirical studies. The first tests several hypotheses as to why there are so few conservative or libertarian law professors: the Brain Hypothesis (lack of intellectual ability), the Greed Hypothesis (a desire for better paying jobs than academia), the Interest Hypotheses (a lack of interest in academic pursuits), and the Discrimination Hypothesis (discrimination based on political orientation). It does so by examining the publication and citation rates of law professors at the 16 highest-ranked law schools in the country. Using statistical matching techniques, the study examines a unique dataset to find that conservative/libertarian professors publish and are cited at nearly double the rate as their peers, a statistically significant difference. These findings are more consistent with a story of discrimination than the other potential hypotheses.

The second study leverages models of discrimination developed by Gary Becker and Kenneth Arrow to measure the “rank gap” in law school hiring based on political orientation discrimination. The study draws on a unique of all newly-hired law professors from 2001-2010, and finds, using statistical matching techniques, that conservative/libertarian law professors are hired law schools averaging 12-13 ranks lower (i.e., less prestigious) than their liberal peers, after controlling for other predictors of the ranking of the law school one is initially hired at.

Finally, the third study examines in a type of quantitative intellectual history the portrayal, or treatment, of religion in legal scholarship over a watershed 15-year period as relates to religion in American law, politics and society: 1998-2012. The study finds that religion is increasingly treated as something that is problematic, as compared to something positive, that not all religions are treated the same with some getting more favorable treatment in legal scholarship than others.