Empirical Research on Criminal Justice, Policing and Law & Economics

Mentor Areas

Professor Abrams is a leading economist working in empirical law and economics, and an expert in the areas of criminal justice and intellectual property. His work covers a range of topics, united by the goal of understanding and measuring how individuals respond to incentives in various legal contexts.  In the area of criminal justice, Professor Abrams has investigated a variety of questions, including whether longer sentences deter crime, how defendant race impacts judicial decisions, to what extent attorney skill affects case outcomes, and how much individuals value freedom. In intellectual property, he has explored the effects of NPEs on innovation, investigated the determinants of patent value, researched the expected impact of the America Invents Act, and examined the effect of patent duration on innovation. Professor Abrams has additional interests in law and health economics, labor economics, and corporate finance. His research has appeared in a number of top peer-reviewed journals and law reviews, including the Stanford Law Review, The University of Chicago Law Review, the University of Pennsylvania Law Review, American Economic Journal: Applied Economics, and The Journal of Legal Studies. He is past president and current board member of the Society for Empirical Legal Studies as well as past chair of the Law and Economics section of the American Association of Law Schools.

Description:

What Police Reforms Work?

There is unprecedented interest in reforming police departments to decrease police shootings and racial disparities, while maintaining efficacy against crime. Broad-based evidence on effective police reform is scant, and this project attempts to address this gap. 

The biggest single driver of police reform over the past two decades has been the Department of Justice which entered into 40 reform agreements with local police departments between 1997 and 2017. While a handful of these agreements have been studied, there has been no comprehensive approach to quantitatively estimating the impact of specific changes made by police departments. 

This project will code each reform agreement by type of changes required by the police department. Outcomes include police violence, complaints, crime rates, property values, bond ratings and other measures when available. A natural experiment, and synthetic control design will be used.

Learning about Crime Fundamentals from the COVID Pandemic

The COVID pandemic has had a massive impact on criminal activity. The changes the pandemic has led to in normal activity, policing, the judicial system and jails may give us insight into how crime is generated and deterred. This project has multiple parts.

First, I plan to examine how crime has changed within cities. I will examine characteristics like racial composition, income, employment, pre-existing crime rates, and others to help obtain a more nuanced understanding of the crime impact of the pandemic.

The second piece of this project stems from the fact that crime declined prior to stay-at-home orders, and happened closer in time to changes in mobility. It raises the question of whether some types of mobility data may be used to predict crime at a local level.

The third piece of the project is the most ambitious. Using data on crime, policing, jails, job loss, and mobility, is it possible to build a structural model that does a good job of predicting crime in a broad range of contexts, well beyond the pandemic? The sudden changes due to COVID may allow the estimation of such a model in a way not previously possible.

Race-Neutral Policing and Stop and Frisk in a Pandemic

A major focus of policing reform for the past decade has been stop-and-frisk practices, where police stop individuals or vehicles suspected of involvement in criminal activity.  A major concern is that the practice may be applied in racially disparate ways, but detecting this can be difficult. If police seek to maximize the likelihood of finding contraband (such as weapons) then the yield (or “hit rate”) of contraband should be equal by race, even if stop rates are not. One difficulty of testing this theory is that the marginal detainee is not observed, only the average. The onset of the pandemic led to a major drop in stops in a number of cities, and provides an opportunity to get closer to the marginal stop and test a major assumption of this theory.  This has critical policy relevance as it is one of the most widely used tools in evaluating police departments for race-neutral policing.

citycrimestats.com

This website was begun in July, 2020 to go along with research into the effect of the pandemic on crime. It has already received widespread media coverage (New York Times, NPR, The Economist) and generates additional interest daily. The site covers over 25 large cities and includes data on crime incidents, arrests, police stops, traffic stops, jail population, COVID incidence and mobility.  The data is updated frequently, via APIs, scraping, manual download, and email and phone requests. The coverage and data sets are continually being expanded. 

I need assistance to maintain and improve upon this resource, which has shown to be useful for journalists, researchers, and the public.

Preferred Qualifications

Most of these research assistant (RA) positions will benefit from prior quantitative RA work (in the quantitative social sciences, statistics, engineering or science). Programming experience is a huge plus, as is experience with Stata or R. There may be a position for someone with less technical interest who is very interested in the subject matter and willing to work hard. Hard work and diligence is essential for all positions.

If interested, please email Professor Abrams with the following:

• Explanation of interest in these projects (specify which).
• One paragraph description of prior RA work,
• Coursework in the following fields: Economics, Statistics, Math, Hard Sciences
• Programming experience (including R or Stata) if any
• Please attach a resume and include an unofficial Penn transcript (for sophomores or above)