By Jeffrey Fagan and Lila Nojima

Theories of rational behavior assume that actors make decisions where the benefits of their acts exceed their costs or losses. If those expected costs and benefits change over time, the behavior will change accordingly as actors learn and internalize the parameters of success and failure. In the context of proactive policing, police stops that achieve any of several goals—constitutional compliance, stops that lead to “good” arrests or summonses, stops that lead to seizures of weapons, drugs, or other contraband, or stops that produce good will and citizen cooperation—should signal to officers the features of a stop that increase its rewards or benefits. Having formed a subjective estimate of success (i.e., prior beliefs), officers should observe their outcomes in subsequent encounters and form updated probability estimates, with specific features of the event, with a positive weight on those features. Officers should also learn the features of unproductive stops and adjust accordingly. A rational actor would pursue “good” or “productive” stops and avoid “unproductive” stops by updating their knowledge of these features through experience. We analyze data on 4.9 million Terry stops in New York City from 2004–2016 to estimate the extent of updating by officers in the New York Police Department. We compare models using a frequentist analysis of officer behavior with a Bayesian analysis where subsequent events are weighted by the signals from prior events. By comparing productive and unproductive stops, the analysis estimates the weights or values—an experience effect—that officers assign to the signals of each type of stop outcome. We find evidence of updating using both analytic methods, although the “hit rates”—our measure of stop productivity including recovery of firearms or arrests for criminal behavior—remain low. Updating is independent of total officer stop activity each month, suggesting that learning may be selective and specific to certain stop features. However, hit rates decline as officer stop activity increases. Both updating and hit rates improved as stop rates declined following a series of internal memoranda and trial orders beginning in May 2012. There is also evidence of differential updating by officers conditional on a variety of features of prior and current stops, including suspect race and stop legality. Though our analysis is limited to NYPD stops, given the ubiquity of policing regimes of intensive stop and frisk encounters across the United States, the relevance of these findings reaches beyond New York City. These regimes reveal tensions between the Terry jurisprudence of reasonable suspicion and evidence on contemporary police practices across the country.

New York, NY: Columbia Public Law School, 2023, 61p.

By Michael German and Kaylana Mueller-Hsia

The failure of the Federal Bureau of Investigation (FBI) and other law enforcement agencies to anticipate and prepare for the January 6, 2021, attack on the U.S. Capitol by far-right insurrectionists has elicited proposals to expand the bureau’s authority to investigate domestic terrorism.

The FBI already received expansive new powers after the 9/11 terrorist attacks, and its current guidelines place few limits on agents’ ability to search broadly for potential threats. Confusion about the current scope of the bureau’s powers is understandable, however, as FBI leaders have regularly misstated their authorities in public testimony.

These misstatements deflect FBI accountability by focusing overseers on filling perceived gaps in its authority rather than examining how the bureau uses, misuses, or fails to use the tools it already has.

The real problem is not that the FBI’s authorities are too narrow, but rather that they are overbroad and untethered to evidence of wrongdoing. After 9/11, the Department of Justice (DOJ) reduced or eliminated reasonable evidentiary predicates to justify broader collection and sharing of Americans’ personal information. This new domestic intelligence process replaced evidence-driven investigations of suspected criminal activities with mass data collection and untriaged reporting of speculative harms unsupported by facts. The sheer volume of threat reporting resulting from this system suffocates effective intelligence analysis, flooding law enforcement leaders with thousands of specious threat warnings a day. In addition to unjustified invasions of privacy, the high rate of false alarms that this process produces naturally dulls the response, and the disconnect from evidence of criminality opens the door to bias-driven law enforcement responses. As they have in the past, the FBI’s unbridled authorities have resulted in abuses of civil rights and civil liberties without improving its ability to identify and mitigate real threats.

Misinformation from FBI officials has confused the policy debate. When senators investigating the January 6 attack asked Jill Sanborn, then the assistant director of the FBI’s Counterterrorism Division, whether FBI agents monitored the multitude of threats made in public forums prior to the attack, Sanborn replied, “It’s not within our authorities.”

Sanborn claimed that the FBI cannot collect information involving First Amendment–protected activities without a predicated investigation or a tip from a community member or law enforcement officer. These statements are inaccurate, yet they featured prominently in the Senate’s report on the security, planning, and response failures regarding the attack on the Capitol.

New York: Brennan Center for Justice at New York University School of Law, 2022. 21p.

By Oeindrila Dube, Sandy Jo MacArthur and Anuj Shah

Prevailing views suggest that adverse policing outcomes are driven by problematic police officers and deficient policies. This study highlights an overlooked factor – cognitive demands inherent in police work. A training designed to improve officer decision-making under stress and time pressure, two key cognitive demands, lead to 23% fewer uses of force and discretionary arrests, and 11% fewer arrests of Black civilians. The results demonstrate the power of leveraging behavioral science insights to make policing more effective and equitable.

Policing practices have increasingly come under public scrutiny, spurring widespread calls for police reform. To date, prevailing views have attributed adverse policing outcomes (such as excessive force and unnecessary arrests) to factors such as prejudiced officers and deficient departmental policies. In this paper, the authors introduce a new focus that explores the role of cognitive demands in contributing to poor policing outcomes. Police work often involves making complex decisions in situations that produce stress, trigger many emotions, and require officers to act quickly. These cognitive demands make it more likely that officers will act without sufficient deliberation and that their actions will be driven by cognitive biases. In this paper, the authors explore this overlooked perspective, which suggests an additional avenue for improving policing outcomes.

WORKING PAPER · NO. 2023-118 . Chicago: University of Chicago, 2023. 99p.

By Deepak Premkumar

After a spate of protests touched off by high-profile incidents of police use of force, there has been a renewed focus on whether public scrutiny shapes policing behavior, otherwise known as the Ferguson Effect. This question has gained additional urgency as the country grapples with increases in murders. This paper provides the first national analysis showing that after police killings that generate significant public attention and scrutiny, officers reduce effort and crime increases. The effects differ by offense type: Reduced police effort yields persistently fewer arrests for low-level offenses (e.g., marijuana possession) but limited changes in arrests for violent or more serious property crimes. I show that decreased interaction with civilians through police stops may be driving the results. However, the increase in offending is driven by murders and robberies, imposing significant crime costs on affected municipalities. The effects only occur after there is broad community awareness of the incident. These findings are robust to numerous changes in empirical specification, transformations of the dependent variable, and varying levels of fixed effects that control for changes in state law and treatment spillovers. I also present evidence that suggests these effects are not driven by a pattern-or-practice investigation or a court-mandated monitoring agreement. To distinguish between the potential effects that may simultaneously impact arrest levels following a high-profile police killing, I develop a theoretical model that provides empirically testable predictions for each mechanism. I find that the reduction in low-level arrests corroborates public scrutiny as the causal channel. Finally, I provide evidence that the increase in offending is driven by both a response to the reduction in policing effort and a reaction to the police killing itself, suggesting that measures to reduce use of force should be prioritized.\

Unpublished paper, 2019, revised 2022. 132p.

By Daniel AC Barbosa Thiemo Fetzer Caterina Soto & Pedro CL Souza

We provide experimental evidence that monitoring of the police activity through body-worn cameras reduces use-of-force, handcuffs and arrests, and enhances criminal reporting. Stronger treatment effects occur on events classified ex-ante of low seriousness. Monitoring effects are moderated by officer rank, which is consistent with a career concern motive by junior officers. Overall, results show that the use of body-worn cameras de-escalates conflicts

Warwick Economics Research Papers . Coventry, UK: University of Warwick, 2021. 56p.

By Jessica A. Reichert, Kaitlin Martins, Bruce Taylor, and Brandon Del Pozo

Police encounter individuals with opioid use disorder (OUD) during their routine work and are often called to the scene of overdoses. Despite this frequency, officer knowledge and attitudes about addiction, treatment, and harm reduction vary. Views held by officers, and the extent of their knowledge, can impact the decisions they make regarding people with OUD, yet our understanding of these factors is limited. Using stratified random sampling, we surveyed 248 officers from 27 Illinois police departments on their knowledge of addiction and the means to address it. We performed descriptive and regression analyses to examine differences based on officer characteristics. We found a high proportion of officers lacked knowledge of addiction, treatment, and harm reduction. Our findings suggest the need for police training to improve understanding of addiction. Community collaboration and coordination of resources may give officers the tools to better address OUD, reduce harm, and decrease overdose.

Chicago: Illinois Criminal Justice Information Authority 2023. 22p.

By Julia Anne Yesberg and Ben Bradford

Collective efficacy is a neighbourhood social process that has important benefits for crime prevention. Policing is thought to be one antecedent to collective efficacy, but the mechanisms by which police activity and officer behaviour are thought to foster collective efficacy are not well understood. This article presents findings from a rapid evidence assessment conducted to take stock of the empirical research on policing and collective efficacy. Thirty-nine studies were identified and examined. Overall, trust in police was the aspect of policing most consistently associated with collective efficacy. There was also some evidence that community policing activities, such as visibility and community engagement, predicted collective efficacy. Police legitimacy, on the other hand, was relatively unrelated to collective efficacy: a finding which suggests perceptions of police linked to the ‘action’ of individual officers may be more enabling of collective efficacy than perceptions of the policing institution as a whole. Implications and directions for future research are discussed.

International Journal of Police Science & Management , 2021. vol.23(4):417-430

By Marie Pryor, Kim Shayo Buchanan, and Phillip Atiba Goff

Swencionis & Goff identified five situations that tend to increase the likelihood that an individual police officer may behave in a racially disparate way: discretion, inexperience, salience of crime, cognitive demand, and identity threat. This article applies their framework to the realities of police work, identifying situations and assignments in which these factors are likely to influence officers’ behavior. These insights may identify opportunities for further empirical research into racial disparities in such contexts and may highlight institutional reforms and policy changes that could reduce officers’ vulnerability to risks that can result in racially unjust actions.

Annual Review of Law and Social Science , 16: 341-360. 2020.

By Yvon Dandurand, Paul Maxim and Darryl Plecas

Strained police relations with visible minorities are reflected in the fact that these minorities are much less likely than other citizens to view the police as legitimate, fair, or trustworthy, or to report crime to the police. Police in and outside of Canada have long understood the importance of improving their relationship with minorities, and in this regard, they have undertaken multiple initiatives intended to improve minority-police relations. Considerable resources and energy were devoted to trying to enhance police relationships with various visible minority groups. These efforts have included extensive outreach initiatives, force-wide sensitivity training for police officers, substantial recruitment and promotion of minorities, and policy changes relating to police practices. Have those efforts made any significant difference in how visible minorities view the police? This study was undertaken as a step toward understanding how the relationship between police services and their host communities has evolved over the years. It examined the extent to which police efforts aimed at improving police-minority relations over the past 20 years have improved perceptions of the police among visible minority groups in Canada (with special attention to British Columbia). More specifically, the study examined the degree to which attitudes of visible minorities over that 20-year period between 1999 and 2019 can be distinguished from those of the overall population in Canada and British Columbia – with special attention to the matter of crime victims’ contacts with police. The core analysis for this study involved a comparison of data from Statistics Canada’s General Social Survey (GSS) panels on Canadians' Safety (Victimization) conducted in 1999, 2004, 2009, 2014 and 2019. The findings of the analysis tell a simple story. The GSS national data collected over a period of twenty years do not show significant improvements in how visible minorities in British Columbia and in Canada perceive the police. Visible minorities hold more negative views of police behaviour than non-minorities. Minorities are less likely than non-minorities to agree that police treat people fairly or do a good job in approaching people. Since the turn of the century, minorities’ views of police behaviour and fairness have generally worsened. Notably, by 2019 those views had become more negative than at any time in the previous twenty years. By then, survey results indicated that less than half of visible minorities agreed that police treat people fairly and do a good job in the way they approach people.

Vancouver, BC: International Centre for Criminal Law Reform and Criminal Justice Policy, 2022.62p.

By Aaron Chalfin and Felipe Goncalves

  How do public sector workers balance their pro-social motivations with private interests? In this study of police officers, we exploit two institutional features that change the implicit cost of making an arrest: arrests made near the end of an officer’s shift are more likely to require overtime work, and arrests made on days when an officer “moonlights” at an off-duty job after their shift have a higher opportunity cost. We document two consequences for officer behavior. First, contrary to popular wisdom, officers reduce arrests near the end of their shift, and the quality of arrests increases. We argue that these patterns are driven by officer preferences rather than departmental policy, fatigue, or incapacitation from earlier arrests. Second, officers further reduce late-shift arrests on days in which they moonlight after work, suggesting that they are, in fact, modestly responsive to financial incentives. Using these results, we estimate a dynamic model that identifies an officer’s implied trade-off between private and pro-social motivations. We find that police officers exhibit high pro-social motivation towards their work. In contrast to prior research showing that law enforcement outcomes are sensitive to financial incentives at an institutional level, the behavior of individual officers — the “street-level bureaucrats” who enforce the law — is not meaningfully distorted by monetary considerations.

Unpublished paper, 2020. 74p.

By Shelley Hyland, Lynn Langton and Elizabeth Davis

This report presents data on the threat or use of nonfatal force by police against white, black, and Hispanic residents during police contact. This report describes the characteristics of the contact, the type of force threatened or used, and the perceptions that force was excessive or the police behaved properly during the contact. It also examines trends in the threat or use of force and the relationship between officer and driver race and Hispanic origin in traffic stops involving the threat or use of force. Data are from the 2002, 2005, 2008, and 2011 Police- Public Contact Surveys, which were administered as supplements to the National Crime Victimization Survey.

Washington, DC: U.S. Department of Justice Office of Justice Programs Bureau of Justice Statistics, 2015. 17