Practice Profile: Police-Initiated Pedestrian Stops to Reduce Crime

Evidence Rating for Outcomes

Crime & Delinquency | Multiple crime/offense types

Crime & Delinquency | Displacement effects

Mental Health & Behavioral Health | Multiple mental health/behavioral health outcomes

Physical Health | Physical health of citizens who come into contact with law enforcement

Date: This profile was posted on December 20, 2023

Practice Goals/Practice Components

Police-initiated pedestrian stops on the street (also referred to as pedestrian stops, stop and frisks, Terry stops, street pops, stop and searches, etc.) are forms of engagement between law enforcement and citizens “whereby an officer stops and questions a pedestrian based on reasonable suspicion that the pedestrian is—or was recently—engaged in unlawful activity” (La Vigne et al. 2012, iv). The tactic involves police initiating a stop of an individual (or individuals) on the street for the purpose of investigation or questioning. Generally, the goal of the stops is to prevent or reduce the occurrence of crime and disorder.

Police-initiated stops can be reactive in nature, usually requiring the police to notice some prior indication of suspicious behavior or criminal activity to make the stop. Depending on the level of suspicion of police, a frisk or search of the stopped individual may be conducted. In contrast, police-initiated stops may also be part of larger proactive police interventions (Petersen et al. 2023). For example, as part of the Philadelphia Foot Patrol strategy (implemented in 2009), police officers patrolled targeted areas in the city for at least 8 hours a day. The officers would engage in various types of activities while patrolling assigned locations, including interviewing pedestrians.

The constitutionality of police-initiated stops was considered in the U.S. Supreme Court Case Terry v. Ohio (1968). The ruling provided police with discretion to conduct stops of individuals, given a reasonable suspicion that the individual has committed a crime or is in the process of committing a crime. The ruling also provided police with discretion to conduct frisks (or pat-downs) if there is reasonable suspicion that the individual is carrying a weapon (Petersen et al. 2023). The court places on the officers the responsibility of articulating a reasonable basis for frisking an individual but implicitly accepts the facts cited by police as reasonable (Sherman, Shaw, and Rogan 1995).

Practice Theory

Police-initiated stops may help prevent crime and disorder through the deterrent effect of police actions. For example, stops by police may increase the perceived certainty of apprehension of those individuals who may commit a crime. Further, if police-initiated stops are part of a larger proactive police response, the presence of additional police in a targeted neighborhood may also serve as a deterrent (Petersen et al. 2023).

Crime & Delinquency | Multiple crime/offense types
Looking at the results from nine place-based studies, Petersen and colleagues (2023) found a statistically significant effect on measures of crime (RIRR = 0.87). The results showed a 13 percent reduction in crime for treatment areas with police- stop interventions, compared with control areas.

Crime & Delinquency | Displacement effects

Looking at the results from four place-based studies, Petersen and colleagues (2023) found a statistically significant effect on measures of displacement (RIRR = 0.93). (Displacement is the relocation of crime from one place, time, target, offense, or tactic to another as a result of some crime-prevention initiative, in this instance, the stops initiated by police.) The results showed a 7 percent decrease in crime for treatment displacement areas, compared with control displacement areas.

Mental Health & Behavioral Health | Multiple mental health/behavioral health outcomes
Looking at the results from eight person-based studies, Petersen and colleagues (2023) found a statistically significant and negative effect on mental health issues (OR = 1.46). The results showed that individuals who were stopped by police in treatment areas reported a 46 percent increase in the odds of mental health issues (such as posttraumatic stress disorder, anxiety, and depressive symptoms), compared with individuals in the control areas. In other words, this finding demonstrates an adverse social effect of the program.

Physical Health | Physical health of citizens who come into contact with law enforcement
Looking at the results from four place-based studies, Petersen and colleagues (2023) found a statistically significant and negative effect on physical health issues (OR = 1.36). The results showed that individuals who were stopped by police in treatment areas reported a 36 percent increase in the odds of physical health issues (such as trouble sleeping and poor health), compared with individuals in control areas. In other words, this finding demonstrates an adverse social effect of the program.

Meta-Analysis Snapshot
	Literature Coverage Dates	Number of Studies	Number of Study Participants
Meta Analysis 1	1975-2021	40	90904

Meta Analysis 1

Petersen and colleagues (2023) conducted a meta-analysis to synthesize the research looking at the impact of police-initiated pedestrian stops on outcomes including crime and disorder (as well as displacement), mental and physical health, individual attitudes toward the police, self-reported crime/delinquency, violence in policeâcitizen encounters, and police misbehavior (the CrimeSolutions review of this meta-analysis focused on outcomes related to mental and physical health issues, crime and disorder, and displacement; there were no eligible studies identified measuring violence in policeâcitizen encounters and police misbehavior). Studies were eligible for inclusion if they included a treatment group that received a pedestrian-stops intervention and a separate comparison group that did not receive a pedestrian-stops intervention (this included randomized controlled trials and quasi-experimental research designs). For the review, the treatment group could include geographic areas or individuals. Eligible interventions could include proactive policing interventions, natural variation in the use of pedestrian stops across area, or natural variation in the use of pedestrian stops across individuals. Studies were excluded if they were focused primarily on traffic stops.

The Global Policing Database (a repository of all experimental and quasi-experimental studies of policing interventions completed since 1950) was searched to identify all published and unpublished studies eligible for inclusion in the review. This was supplemented with additional searches of academic databases (such as Criminal Justice Abstracts, National Criminal Justice Reference Service Abstracts, and Criminal Justice Database), searches of gray-literature sources, and correspondence with subject-matter experts to identify eligible studies available through December 2021.

The search resulted in the inclusion of 40 studies that were published between 1975 and 2021. Most of the studies (33) were conducted in the United States and the remaining (7) were conducted in Europe. Only one study was experimental; the other 39 were quasi-experimental (with matched or unmatched comparison groups). The unit of analysis for 29 studies was individuals, and the unit of analysis was geographic areas for the other 11 studies (the CrimeSolutions review of this meta-analysis focused on findings conducted at the geographic area unit of analysis).

Crime and disorder (and displacement) outcomes were considered eligible if the outcome was measured using official data (such as arrest data or calls for service), unofficial crime data (such as crime reported by civilians), and systematic social observations of crime. All types of crime and disorder (i.e., property, drug, violence crimes) were included. Mental and physical health issues were measured through self-reports taken from individuals with direct police stop experience or from official data (such as injury data from hospitals). Measures at both the individual- and community-level were included.

For the place-based studies that include outcomes of crime and disorder and/or displacement, effect sizes were calculated using logged relative incident rate ratios (or RIRRs). For place-based studies that included outcomes of mental and physical health, the effect sizes were calculated using logged odds ratios (or ORs). Data synthesis for this review involved standard inverse-variance weighted meta-analysis.