Program Profile: New York City Police Department Body-Worn Cameras’ Effects on Civility and Lawfulness of Police–Citizen Encounters and Policing Activities

This program equips police with body cameras to record encounters with civilians. The program is rated No Effects. Camera use had no statistically significant effects on number of arrests, arrests with force, and summonses issued after 1 year. Officers wearing cameras had statistically significant reductions in complaints filed against them and made more stop reports. Camera use resulted in a statistically significantly reduced likelihood of lawful stops and frisks but fewer subjects searched.

A No Effects rating implies that implementing the program is unlikely to result in the intended outcome(s) and may result in a negative outcome(s).

This program's rating is based on evidence that includes at least one high-quality randomized controlled trial.

Program Goals/Target Population

The federal court settlement of Floyd v. City of New York, 959 F. Supp. 2d 668 mandated that the New York City Police Department (NYPD) implement a series of reforms to address unlawful stop, question, and frisk patterns and practices, under the supervision of an independent monitor (Floyd v. City of New York, 2013). Among other changes, the remedial order required the department to implement a pilot body-worn camera program to visually and audibly record certain interactions between uniformed members of the NYPD and the public, for official law enforcement purposes. The goals of the NYPD’s body-worn camera program are to create objective records of stop-and-frisk encounters, encourage lawful and respectful police–citizen interactions, alleviate mistrust between the department and the public, and help determine the validity of accusations of police misconduct.

Program Components

In September 2014, before the implementation of the court-ordered body-worn camera pilot, the then–NYPD commissioner announced the launch of a volunteer pilot program to test the use of body-worn cameras in all five New York City boroughs. From December 2014 through March 2016 the department conducted this small pilot with 54 volunteer police officers across six different commands. In April 2017, after reviewing the results of the volunteer pilot, the NYPD began the larger-scale, court-ordered pilot. Soon thereafter, the department started a citywide effort to outfit all uniformed patrol officers and officers working selected specialized units with body-worn cameras. The department contracted with a police body-worn camera technology provider to equip officers with the devices, and the cameras were rolled out on a precinct-by-precinct basis (Civilian Complaint Review Board, 2020). In March 2019 the NYPD announced that all uniformed patrol officers in New York City—including police officers, sergeants, and lieutenants assigned to every precinct, transit district, and Police Service Area—were equipped with body-worn cameras (NYPD, 2019).

NYPD officers are required to activate a body camera when conducting investigative and enforcement actions such as making an arrest, issuing a summons, conducting a search, or stopping a person on the street in the course of an investigation. All officers equipped with body cameras have received training on how the cameras function, how to use the video management software, and the department’s body camera policies. As part of their training, officers also participate in role-play scenarios to acclimate themselves to the proper use of the cameras (NYPD, 2019).

Program Theory

Public awareness theory provides a theoretical explanation of the potential benefits of placing body-worn cameras on police officers (Farrar, 2013; Ariel, Farrar, and Sutherland, 2015). The theory posits that when people know they are being watched, they are better able to compare established social norms with their current behaviors, conform to set rules, and behave in socially desirable ways (Duval and Wicklund, 1972). Therefore, body-worn cameras may encourage police officers and citizens to consider societal norms and standards during their interactions and adjust their immediate behaviors in socially desirable ways.

Another theoretical foundation of body-worn camera use is deterrence theory, which posits that individuals decide not to commit criminal acts when the perceived costs of these acts outweigh the perceived benefits (Zimring and Hawkins, 1973). Research suggests that deterrence is most effective when people perceive that there is a sufficient risk of apprehension and certainty of punishment (e.g., Nagin, 2013). Body-worn cameras are an intervention that can potentially stimulate deterrence through the threat of having inappropriate and illegal behaviors captured on video and, in turn, modify officer and citizen perceptions of sanction risks (Ariel et al., 2017). As a result, police officers outfitted with body-worn cameras should be deterred from taking actions that violate the law and departmental policy. Deterrent impacts on antisocial behavior during police–citizen encounters may be generated by the threat of formal punishment by the criminal justice system, and by concerns over informal punishment by family, friends, and the general public through shaming on social media if a captured video “goes viral” (Braga et al., 2020).

Study 1

Braga, MacDonald, and McCabe (2021a) used a cluster randomized controlled trial to measure the effect of New York City Police Department’s (NYPD’s) body-worn cameras on the civility of police–citizen encounters (as measured by civilian complaints and arrest reports listing officer use of force) and policing activity (as measured by domestic incident reports, arrests, summonses issued, and stop reports submitted by officers) from 1 year preintervention to 1 year postintervention. The NYPD used a staggered rollout of the body-worn cameras over 7 months, beginning in April 2017 and ending in November 2017. Despite varying start dates, all treatment precincts used the cameras for 1 full year, with the experimental intervention period ending in November 2018.

All 77 NYPD precincts were ranked according to the 2012–2015 mean yearly counts of citizen complaints handled by the New York City Civilian Complaint Review Board (an independent, all-civilian agency with 13 board members who review and make findings on complaints of officer misconduct that allege the use of excessive or unnecessary force, abuse of authority, discourtesy, or the use of offensive language). The top-ranked 40 precincts were then matched into 20 pairs based on simple matching exercises to ensure baseline similarities. Other relevant variables were then considered, including 2012–2015 mean yearly counts of arrests, uses of force, major reported crimes, sworn officers, 2014–2015 mean 9–1–1 calls for service counts, 2015 New York City Housing Authority resident population data, and an overall concentrated disadvantage index for the neighborhoods that made up the precincts, based on census block data. A randomization algorithm was then used to assign which precinct within each of the 20 pairs would receive the body-worn cameras.

The NYPD provided data on 3,889 officers who worked uniformed third platoon (the 3:00 p.m. to midnight shift) or applicable plainclothes anticrime assignments and appeared on the monthly precinct rosters during the 1-year intervention period in the 20 matched precincts. In the treatment group, body-worn cameras were provided to all uniformed patrol officers working the third platoon and plainclothes Anti-Crime Unit officers working a majority of their shift on the third platoon in the treatment precincts (n = 1991). The control group was also composed of uniformed patrol officers working the third platoon and plainclothes officers in Anti-Crime Units working the majority of their shift on the third platoon (n = 1,898). The concern of contamination (that officers with body-worn cameras influence the behavior of officers without cameras if they work simultaneously in the same area) was limited by randomly allocating groups of officers, rather than individuals, to have cameras.

The matching and randomization procedure generated balanced treatment and control precincts. Comparisons of group characteristics did not reveal any statistically significant differences in officer rank, sex, race/ethnicity, age, and years on the job, for officers included in the treatment and control groups. There were also no statistically significant differences between officers in the treatment and control groups in their average work activities, arrests that involved use of force, and Civilian Complaint Review Board complaints during the 12 months before their inclusion in the cluster randomized controlled trial (the preintervention period). Officers in the treatment group were on average 29.8 years old, predominately male (84.2 percent), and were 47.2 percent white, 30.7 percent Hispanic, 13.2 percent Black, and 8.9 percent Asian or other race. The majority of the treatment group officers (89.7 percent) were assigned to the third platoon patrol and had on average worked 4.3 years on the job. Control group officers were on average 29.9 years old, also predominately male (84 percent), and were 47.0 percent white, 28.9 percent Hispanic, 14.8 percent Black, and 9.3 percent Asian or other race. The majority of control group officers (90.1 percent) were assigned to the third platoon patrol and had on average worked 4.4 years on the job. During the 1-year preintervention period, officers in the treatment group submitted an average of 0.7 stop reports (which involve the use of stop, question, and frisk practices), made an average of 10.4 arrests, made 0.1 arrests with force, served 8.9 summonses, filed 0.3 domestic incident reports, and had 34.4 crime complaint reports and 0.3 Civilian Complaint Review Board complaints against them. Officers in the control group submitted an average of 0.7 stop reports, made 10.7 arrests, made 0.2 arrests with force, served 8.2 summonses, filed 0.2 domestic incident reports, and had 33.4 crime complaints and 0.3 Civilian Complaint Review Board complaints against them.

Intention-to-treat analyses based on the initial random assignment were used so that all treatment and control group officers were included in the analysis. Encounter civility and policing activity outcomes were measured for officers in both treatment and control groups during 1-year preintervention (before the adoption of cameras by treatment officers) and 1-year intervention (treatment officers wearing cameras) periods.

The impact of body-worn cameras on treatment officer outcomes relative to control officer outcomes was estimated by difference-in-differences methods. The difference in differences estimates the difference in treatment officers’ postintervention outcomes at time t compared with their preintervention outcomes, relative to the same difference for the control officers, where the unit of analysis was officer-observation periods. Precinct pair fixed effects were included, and standard errors were clustered by individual officers to address heteroskedasticity and overdispersion. No subgroup analysis was conducted.

Study 2

Braga, MacDonald, and McCabe (2021b) used a stratified random cluster sampling methodology to select representative samples of NYPD stop reports made by third-platoon officers and Anti-Crime Unit officers during the 12-month intervention period in the treatment and control precincts as described in Study 1 (Braga, MacDonald, and McCabe 2021a) to assess the effect of body-worn cameras on police lawfulness. The random selection of stop reports from pairs of experimental precincts started at the beginning of the cluster randomized controlled trial in April 2017.

Each quarter the NYPD Quality Assurance Division sent the NYPD monitor team a list of all audits conducted on all commands’ stop reports during the previous quarter. The monitor team then randomly selected commands until reaching at least 300 stop reports to be independently audited by team members. Beginning in the first quarter of 2017, the monitor team stratified the ongoing random cluster sample to ensure that the first 10 commands selected were from the 40 precincts included in the randomized controlled trial (Study 1). The first strata included five experimental precincts randomly selected from these 40 commands, and the accompanying matched precinct within their pair were included to ensure balanced representation of stop reports in treatment and control precincts in subsequent analyses (10 precincts total, comprising 5 matched pairs of treatment and control precincts). The second strata included the random selection of nonexperimental commands until the desired 300 stop reports required for citywide compliance monitoring was reached.

The stratified cluster sampling process yielded an audit of 474 total stop reports made by body-worn camera treatment group officers (n = 289 stop reports) and control group officers (n = 185 stop reports) during the 1-year intervention period. In the included stop reports, stopped citizens were mostly male (89.1 percent of treatment group stop reports, and 94 percent of control group stop reports) and an average of 26.5 years old in treatment group stop reports, compared with 28.4 years old in control group stop reports. Individuals stopped by treatment group officers were statistically significantly more likely to be non-Hispanic Black subjects (61.6 percent of treatment group stop reports, compared with 50.8 percent of control group stop reports) and less likely to be non-Hispanic white subjects (4.8 percent of treatment group stop reports, compared with 9.7 percent of control group stop reports). There were no other statistically significant demographic differences in treatment and control group stop reports.

Once the quarterly stop report data from randomly selected precincts included in the randomized controlled trial (Study 1) were provided to the NYPD monitor team, a structured process was followed to assess the lawfulness of the stops, frisks, and searches in the included reports. The Quality Assurance Division and monitor team reviewers of the sampled stop reports assessed the narratives that described specific officer actions during these encounters relative to the legal standards in the NYPD mandatory officer training on the federal and New York state requirements for lawful stops, and proper documentation of those stops. The multistage process began with two monitor team reviewers (reviewer 1 and reviewer 2) independently assessing the lawfulness of the stop, frisk, and search described in the narrative of each stop report. Reviewers 1 and 2 then shared their independent assessments with each other and discussed their findings. This discussion generated a list of stops at which two types of disagreements existed: a) both monitor team reviewers disagreed with the assessment of lawfulness made by NYPD’s Quality Assurance Division on either the stop, frisk, or search and b) the monitor team reviewers disagreed with each other on the lawfulness of the police actions described in each report. The two monitor team reviewers then considered each other’s assessments and subsequently made any rating changes they deemed appropriate. The list of both types of disagreements was then sent to reviewer 3, who assessed the lawfulness of the stop, frisk, and search of the stop reports under disagreement and then shared those views with reviewers 1 and 2. All three reviewers discussed their views, and a final list of stop reports with disagreement (either with the Quality Assurance Division or among the team) was created. The final list of stop reports with disagreements was sent to the Monitor and Deputy Monitor for their review and assessment of lawfulness. All five monitor team members (Monitor, Deputy Monitor, and reviewers 1, 2, and 3) then discussed all stop reports with disagreement and made a final decision on the lawfulness of the encounter described in the narrative of each stop report. The final list of disagreements was subsequently sent to the NYPD for review. The monitor team and representative from the NYPD Risk Management Bureau and Quality Assurance Division then met to discuss the stop reports identified by the monitor team that did not articulate lawfulness in either the stop or frisk or search. After the monitor team meeting with the Risk Management Bureau and Quality Assurance Division, a final assessment was made of the lawfulness of police actions in the included stop reports for that quarter.

Analyses of the police lawfulness outcomes focused on the differences in stop reports in the treatment precincts relative to control precincts during the intervention period for each pair. The monitor team assessments of lawfulness of stops, frisks, and searches served as the key outcome variables. Multivariate logistic regressions were used to estimate the impact of the body-worn cameras on officer actions and lawfulness assessments of those actions, controlling for characteristics of the stops. No subgroup analysis was conducted.

These sources were used in the development of the program profile:

These sources were used in the development of the program profile:

Following are CrimeSolutions-rated programs that are related to this practice:

Body-Worn Cameras’ Effects on Police Officer Behavior

This practice involves the use of body-worn cameras by law enforcement. The aim of this practice is to record interactions from an officer’s point of view to improve accountability and positively affect police officer behavior. The practice is rated No Effects for its effects on officer use of force, officer injuries, officer-initiated calls for service, traffic stops, field interviews, and arrest incidents.

Evidence Ratings for Outcomes

	Justice Systems or Processes - Use of force
	Crime & Delinquency - Assault on officer/officer injuries/resistance
	Crime & Delinquency - Multiple crime/offense types
	Justice Systems or Processes - Officer-initiated calls for service
	Crime & Delinquency - Traffic stops/traffic tickets
	Justice Systems or Processes - Field interviews/stop and frisk

Why might a practice's outcome ratings differ from the ratings of specific programs encompassed by that practice?