publications

Does Race Matter for Police Use of Force? Evidence from 911 Calls with Mark Hoekstra American Economic Review, 112, 2022 (gated version, ungated) podcast

This paper examines race and police use of force using data on 1.6 million 911 calls in two cities, neither of which allows for discretion in officer dispatch. Results indicate white officers increase force much more than minority officers when dispatched to more minority neighborhoods. Estimates indicate black (Hispanic) civilians are 55 (75) percent more likely to experience any force, and five times as likely to experience a police shooting, compared to if white officers scaled up force similarly to minority officers. Additionally, 14 percent of white officers use excess force in black neighborhoods relative to our statistical benchmark.

Minimum Wage and Search Effort with Camilla Adams and Jonathan Meer (gated version, CATO brief) Economic Letters, 212, 2022

Labor market search-and-matching models posit supply-side responses to minimum wage increases that may lead to improved matches and lessen or even reverse negative employment effects. Using event study analysis of recent minimum wage increases, we find that these changes do not affect the likelihood of searching, but do lead to transitory spikes in search effort by individuals already looking for work. These results are not driven by changes in the composition of searchers, and are concentrated among the groups most likely to be impacted by the minimum wage and in response to larger minimum wage increases.

The Effect of Risk Assessment Scores on Judicial Behavior and Defendant Outcomes with George Naufal and Heather Caspers (updated version) accepted at Journal of Human Resources

Risk assessment has been increasingly adopted in an effort to reduce pretrial detention for poor, low-risk defendants. This paper examines the impact of risk assessment using administrative data from a large Texas County. We identify effects using a regression discontinuity that exploits the overnight implementation of a risk assessment policy. Results indicate this led to a 6.5 percent increase in nonfinancial bond and an 8.5 percent decrease in pretrial detention, though neither effect persisted beyond two months. Additionally, the policy did not increase violent pretrial crime, though there is some suggestive evidence of increases in nonviolent pretrial crime..

Can Text Messages Reduce Incarceration in Rural and Vulnerable Populations? with Emily Owens click here! accepted at Journal of Policy Analysis and Management

Reducing failures to appear (FTA) in court is a top priority for criminal justice practitioners and advocates. However, existing work on reducing FTAs through text message reminders focuses on large urban jurisdictions and defendants who are housed. Using a field study in Shasta County, California, we evaluate whether text message outreach can increase court appearances for housed and unhoused populations. We find housed defendants randomly assigned to the treatment group were 10% less likely to miss their next scheduled court date than defendants in the control group. We find no statistical difference in the FTA rate of unhoused individuals assigned to treatment or control groups, and cannot rule out large positive or negative effects. We find that improving the quality of contact information in court records could lead to large reductions in FTAs. Partnering with local social service providers may not, but more research is needed.

working papers

Do Prosecutor and Defendant Race Pairings Matter? Evidence from Random Assignment click here!

Racial disparities in criminal justice outcomes are well-documented. However, there is little evidence on whether the pairing of prosecutors and defendants is responsible for these disparities. This paper tests for cross-race effects in convictions by prosecutors. To identify effects, I leverage conditional as-good-as random variation in prosecutor race using detailed administrative data on the case assignment process and case outcomes in New York County. I show that the assignment of a cross-race prosecutor leads to a 5 percentage point (9 percent) increase in convictions for property crimes and no strong evidence of effects for other types of crimes.

Does Violence on the Job Change Police Behavior? click here!

There are rising concerns about the quality of policing in high-violence urban neighborhoods in the United States. Residents of these areas are concerned that police officers are failing to reduce crime, but also that when police officers do engage, their tactics are too severe. This paper examines whether risks to officer safety drive both phenomena. To do so, I exploit variation in unprovoked ambushes on police within and across beats using administrative 911 call data from a large American city. Results show that ambushes lead to an 8 percent decline in arrests, an effect that persists for at least three years after an ambush. In contrast, I find no effect on police severity, as measured by use of force and civilian complaints. This suggests police officers respond to increased risk by de-policing, rather than using more aggressive tactics.

The Effect of Field Training Officers on Police Use of Force with Chandon Adger and Matt Ross click here!

Over the past decade, police use of force has become an increasingly charged political issue with growing calls for reform. One of the few reforms where advocates and the policing community have reached a consensus is on the need for improved and expanded training. In this paper, we study an under-researched but nearly universal training approach whereby a recruit is paired with a senior officer during a phase referred to as “field training.” In particular, we consider the link between a field training officer’s prior propensity to use force and a recruit’s subsequent enforcement behavior. We leverage a unique setting where recruits are as-good-as-randomly assigned to field training officers and where we have detailed information on the universe of calls for service. We document meaningful differences across field training officers in terms of their propensity to use force prior to being paired with a recruit. Further, we find that a one standard deviation increase in a field training officer’s propensity to use force (124 percent) leads to a 15 percent increase in their recruit’s subsequent propensity to use force. The effect of having a more aggressive field training officer persists for as much as two years after the recruit completes training.

works in progress

How Much Does Your Prosecutor Matter? An Estimate of Prosecutorial Discretion click here!

Many argue that prosecutors wield more discretion and power than any other agent in the criminal justice system. Despite this widespread belief, little work has documented the extent of prosecutorial discretion. Using data that tracks cases from arrest through disposition, this paper exploits the conditionally random assignment of cases to prosecutors. To quantify prosecutorial discretion, I use a random effects model to estimate prosecutor-specific measures of leniency. Results show that prosecutors are important for case disposition. Replacing a prosecutor with one who is a standard deviation harsher increases the odds of a guilty outcome by 3 percentage points, or 5 percent. This magnitude is similar to the average difference in guilty outcomes between Black and White defendants, or half the effect of a prior arrest. Additional results indicate that prosecutor race and gender are not primary determinants of leniency. However, prosecutors are willing or able to exercise significant discretion for defendants with no prior arrests and less severe crimes.

Pilot Program to Reduce Racial and Ethnic Disparities in Motor Vehicle Enforcement Actions with Matt Ross press release here!

undergraduate publications

Entrepreneurship and crime: The case of new restaurant location decisions with Steven Caudill and Franklin Mixon Journal of Business Venturing Insights, 5, 2016 (gated version)

This paper examines the impact of various violent crimes on restaurant location decisions in a single city, Memphis, TN. Using location information on crimes and newly opened restaurants, we are able to match crimes and restaurants in parcels we construct. We examine the impact of the number of burglaries, assaults, rapes, and murders in each of 400 to up to 1,000,000 parcels occurring from 2009–2013 on the number of restaurants opening in 2014 in models including dummy variables capturing neighborhood effects. Regression results indicate that each crime is positively related to the number of new restaurants in a parcel. This suggests that even with the crime problem, these locations provide sufficient benefits, perhaps population density, which make them attractive to restaurant entrepreneurs.

*email me for gated versions of papers