Compounding these concerns is the lack of consistent evidence on the subject. As of 2020, more than 30 studies have attempted to measure the impact of BWC programs on officer behavior, primarily focusing on the effect of BWCs on the use of excessive force by officers, police passivity, arrests, and complaints by citizens. Results are mixed. Some studies find that citizens make fewer complaints against officers equipped with BWCs, whereas others find no meaningful relationship between BWCs and citizen complaints. Results are even more ambiguous with respect to the use of force. Nearly half of studies find officers wearing BWCs use force less often, a roughly equivalent number show no significant change, and at least one has found that BWCs lead to an increase in police use of force. As to whether BWCs make police officers more passive, the evidence is also muddy. Findings from the approximately 15 studies that examine the impact of BWCs on officer arrest behavior show no clear pattern of outcomes related to arrests and citations. Some researchers observe fewer arrests, others report that arrests increase, and many observe no significant change.
Perhaps of greater concern is the lack of direct applicability that prior work has to the decisions confronting many law enforcement agencies today, particularly those servicing major metropolitan areas. Although some research has focused on larger departments (e.g., Phoenix, Las Vegas, Orlando, Denver), participants often represent a small subset of the officers on those forces over a predefined period. For example, the study conducted by the New York City Police Department (NYPD), before the full rollout of BWCs that we consider here, included only those patrol officers operating on one shift across the 40 precincts that exhibited the greatest number of civilian complaints, a sample representing fewer than 3 percent of all NYPD officers. In other studies, officers self-selected into participation, potentially biasing results. Still other studies were conducted in small locations like Rialto, California; Mesa, Arizona; and Spokane, Washington, whose police forces and overall populations bear little resemblance to those in America’s larger urban areas. The types of crime, volume of calls, training of police officers, and other factors are likely to differ between urban and nonurban police departments.
This work builds on past studies by revisiting the question of how BWCs affect the behavior of officers in large metropolitan police departments. We examine the phased rollout of BWCs by the NYPD. The NYPD’s BWC program breaks new ground and offers benefits over prior work. First, to date, no study has analyzed an American law enforcement agency near the size of the NYPD—the largest police department in the most populous city in the nation—even though it has been a popular context for examining the effect of information technology on policing.
Second, the NYPD rollout is one of few in which every patrol officer was assigned a camera, alleviating concerns about biased results from officers self-selecting use of BWCs. In this regard, our study also benefits from explicit efforts that have been undertaken to ensure officer compliance— the NYPD Monitor (an independent oversight team) has issued extensive guidance on BWC compliance protocols and auditing processes enacted by the Quality Assurance Division of the NYPD.
Third, our analysis is contemporary by covering a three-year period ending in December 2019. The passage of time diminishes the applicability of prior work, much of which was done several years before our research. Moreover, past work largely takes the form of exploratory pilots, with BWC deployment conducted for a preadvertised, relatively brief time span, such as six months to a year. In our setting, we observed a three-year period of officer behavior, under a scenario in which officers have no expectation that the BWC program will cease. This aspect is important, because the fact that BWCs will be used only temporarily may induce different behavior from officers and citizens. As the NYPD’s BWC program had been advertised from the outset as a permanent effort, observed responses are likely to reflect what other cities should expect from permanent deployment, as neither citizens nor officers can “wait out” the BWC program.
To identify the effects of BWCs, we exploited the phased rollout of the BWC program into different precincts at different times. We constructed a data set consisting of counts of officer-initiated stops, arrests, and citizen complaints against officers between January 2017 and December 2019. We also conducted several secondary analyses to better understand the nuances of BWC effects on police behavior and interactions with citizens. Specifically, we explored the effects of BWCs on different categories of stops and types of complaints. Finally, we examined the effect of BWCs on officers’ use of force, leveraging quarterly reports of such use-of-force incidents by precinct.
Our findings yielded three notable results. First, BWCs have increased the number of investigative stops conducted by the NYPD. As the officers involved knew they were being filmed, one would expect most of these stops to constitute proper investigative activity. This finding runs counter to the fears expressed by some researchers and commentators that BWCs will chill legitimate police activity. To the contrary, it appears that officers view BWCs as a safeguard that provides them with protection to engage with the public. Interestingly, this increase in stops appears to be primarily driven by less dangerous stops; new stops tend to be nonviolent and involve no weapons. Second, BWCs decrease the number of citizen complaints against NYPD officers. Interestingly, the decline in complaints appears to stem mostly from a reduction in abuse-of-authority allegations. An increase in investigative activity (which, as mentioned, is likely to be legitimate) coupled with a decrease in allegations that the police have abused their authority suggests that BWCs are a positive development. Third, BWCs decrease the number of arrests officers make.
There are at least two potential explanations for this third finding. First, BWCs may change civilian behavior during interactions with police. Civilians may modify their behavior to be less likely to provide police with grounds to make an arrest when they know they are on camera. One could argue such change is beneficial as it increases officer and citizen rapport and decreases the likelihood of violence during stops. Alternatively, BWCs may cause police officers to regulate their own behavior. Knowing that the encounter is being filmed, police may be less likely to make questionable arrests. Although we were unable to observe which explanation is dominant, either would suggest a positive impact, as confrontations are less violent and less likely to yield an abuse-of-authority complaint against the officer (whether spurious or legitimate).
NOTE
This research brief is based on Mitchell E. Zamoff, Brad N. Greenwood, and Gordon Burtch, “Who Watches the Watchmen: Evidence of the Effect of Body-Worn Cameras on New York City Policing,” Journal of Law, Economics, and Organization 38, no. 1 (2022): 161–95.
