Video surveillance is widespread across U.S. law enforcement, but adoption varies by technology, agency size and the denominator used. The available evidence spans 2013–2023, from department surveys and body-worn-camera policies to selected CCTV evaluations and federal facial-recognition reviews.
Contents
- Police adoption
- Community size and coverage
- Body-worn video policies
- Storage, privacy and cost barriers
- CCTV crime and clearance outcomes
- Integrated analytics and facial recognition
How widely U.S. police departments use video surveillance
The Bureau of Justice Statistics (BJS) reported that nearly 90% of U.S. local police departments used some type of video-camera technology in 2013. In the same survey, about 49% used video cameras to surveil public areas, an estimated 17% used automated license-plate readers, less than 1% used unmanned aircraft systems, and about 4% used gunshot-detection systems. These are department-use estimates, not equipment inventories or measures of continuous coverage. Local Police Departments, 2013: Equipment and Technology
The later BJS 2020 survey shows how different video categories can coexist. Among U.S. local police departments, 61.8% used body-worn cameras, 42.5% used fixed-site public surveillance, and 61.7% used patrol-car video cameras. The percentages describe department use; they do not show how many cameras each department owned, how often cameras recorded, or what share of public space was covered. Local Police Departments, Procedures, Policies, and Technology, 2020 – Statistical Tables
The 2016 BJS survey used a broader agency denominator: U.S. general-purpose law-enforcement agencies. It found that 83.3% had at least one recording-device type, 69.0% had dashboard cameras, 47.4% had body-worn cameras, and 38.1% had personal audio recorders. The categories overlap, and the recording-device total includes audio, so it should not be read as a video-only adoption rate. Body-Worn Cameras in Law Enforcement Agencies, 2016
Key fact: In 2020, departments covering 79.1% of local police officers used body-worn cameras, but that officer-weighted estimate does not mean every officer had a camera. Local Police Departments, Procedures, Policies, and Technology, 2020 – Statistical Tables
How adoption varies by community size and officer coverage
Large departments reported especially high adoption in the 2020 BJS tables. Departments serving 1 million or more residents reported 76.9% use of fixed-site public surveillance and 100% use of body-worn cameras. Among departments serving 500,000–999,999 residents, 67.7% used mobile surveillance. For departments serving 250,000–499,999 residents, 94.3% used body-worn cameras.
Smaller-community figures vary by technology rather than moving in a single direction. Departments serving 100,000–249,999 residents reported 59.2% use of cameras on aerial drones. Departments serving 25,000–49,999 residents reported 68.6% use of patrol-car video cameras, while departments serving 2,500–9,999 residents reported 8.1% use of cameras on aerial drones. These are department-level adoption measures, not deployment completeness.
The same 2020 source provides an officer-weighted perspective. Departments using body-worn cameras covered 79.1% of local police officers. It also reported one body-worn camera for every 1.7 officers across departments, compared with 2.2 officers per body-worn camera among departments serving 2,500–9,999 residents. The ratios use full-time-equivalent sworn officers and reported cameras; they are not measures of simultaneous use or recording frequency. Local Police Departments, Procedures, Policies, and Technology, 2020 – Statistical Tables
| Measure | Population and period | Result | What it means |
|---|---|---|---|
| Body-worn-camera use | U.S. local police departments, 2020 | 61.8% | Department adoption |
| Officer coverage by adopting departments | U.S. local police departments, 2020 | 79.1% | Officer-weighted reach, not universal assignment |
| Cameras per officer | U.S. local police departments, 2020 | 1 per 1.7 officers | Reported equipment ratio |
| Cameras per officer in smaller departments | Communities of 2,500–9,999, 2020 | 1 per 2.2 officers | Stratum ratio, not recording frequency |
What body-worn video policies record and who can access footage
In the 2016 BJS survey, agencies with body-worn cameras and event policies most often required recording during traffic stops: 92.9%. The same policy population reported requirements for officer-initiated citizen contacts at 84.5%, arrests or search-warrant execution at 84.5%, firearms deployments at 84.0%, routine service calls at 78.9%, criminal investigations at 71.9%, public-order policing at 53.7%, and offender transport at 50.1%.
These are written-policy requirements, not observed compliance rates. The categories are non-mutually-exclusive, so one agency can require recording for several event types. They also apply to agencies with body-worn cameras and event policies, rather than to every law-enforcement agency in the country. Body-Worn Cameras in Law Enforcement Agencies, 2016
Access also varied by role. Among agencies with body-worn cameras in service, 88.0% gave the chief executive direct access to BWC video, 59.8% gave the recording officer direct access, and 19.6% gave a district attorney’s office direct access. In this survey, direct access meant access without a formal request, and the categories could overlap.
Public access demand was measurable but uneven. In 2016, 17.7% of agencies with body-worn cameras said the public had never requested BWC footage. Separately, 44.8% reported zero public footage requests in an average month, while 20.8% received an average of 1–5 requests per month. The figures are agency-reported and concern requests, not the amount of footage released.
Video storage, privacy and cost barriers
Agencies with body-worn cameras identified several operational obstacles in 2016. Privacy was an obstacle for 34.7%, video-storage procedures for 33.8%, greater-than-expected costs for 32.5%, technical difficulties for 21.0%, and information security for 20.0%. Because agencies could report more than one obstacle, these percentages should not be added together.
The reasons reported by agencies without body-worn cameras point to acquisition and retention pressures. Video-storage or disposal costs were cited by 76.6% as a reason not to acquire BWCs, and hardware costs by 74.4%. Among agencies without BWCs, 62.1% had dashboard cameras and 32.3% had personal audio recorders. The BJS tables report no dollar cost for these barriers, and the reason categories can overlap.
Taken together, the survey evidence separates ownership from the work required after recording. A department may have a dashboard camera, audio recorder or other device while still lacking body-worn cameras. A department with BWCs must also manage privacy, storage, security and access policies. These findings describe U.S. general-purpose law-enforcement agencies in 2016 and are not household or commercial-security cost estimates. Body-Worn Cameras in Law Enforcement Agencies, 2016
What CCTV evaluations show about crime and clearances
The NIJ CrimeSolutions profile summarizes a random-effects meta-analysis of international CCTV evaluation sites drawing on literature from 1978–2017 and published in 2019. The pooled overall-crime estimate was an odds ratio of 1.14, corresponding to 14% lower odds in treatment areas. The pooled property-offense estimate was an odds ratio of 1.16, corresponding to 16% lower odds, and the pooled vehicle-crime estimate was also an odds ratio of 1.16, corresponding to 16% lower odds.
Those odds-ratio results are not incident-count percentage changes. The profile reports setting-specific variation. Car-park CCTV treatment areas had a statistically significant 37% crime reduction, while residential CCTV treatment areas had a statistically significant 12% crime reduction. These are subgroup meta-analysis estimates, not universal rates for every camera system. Practice Profile: Closed Circuit Television (CCTV) Surveillance
The Philadelphia public-camera evaluation, published in 2015, found a 13% reduction in crime after controlling for long-term trends and seasonality in Philadelphia public-camera viewsheds. The publication abstract reports mixed site-level results, so the estimate should be understood as a site evaluation rather than a general national effect. The Crime-Reduction Effects of Public CCTV Cameras: A Multi-Method Spatial Approach
Milwaukee provides a separate outcome measure. During the January–December 2018 intervention at high-crime, high-traffic intersections, the enhanced CCTV program had a 14% higher crime-clearance rate than comparison intersections. In that evaluation, clearance means arrests, and the result comes from a quasi-experimental site evaluation. Program Profile: Milwaukee’s (Wis.) Closed-Circuit Television Camera Program
Growth of integrated analytics and facial-recognition surveillance
Milwaukee’s public surveillance system illustrates how a local program can expand beyond a small fixed-camera network. The system grew from 42 cameras at 40 locations to 87 cameras between 2007 and 2018. In 2018, the expansion installed 24 panoramic cameras, 12 pan-tilt-zoom cameras and 9 automatic-license-plate-recognition cameras. These are program-specific counts, and the listed categories do not exhaust the 87-camera system. Program Profile: Milwaukee’s (Wis.) Closed-Circuit Television Camera Program
Federal use of facial recognition is documented differently. A 2023 GAO review examined seven selected DOJ and DHS agencies and reported that all seven used facial-recognition systems owned by others, while three owned facial-recognition technology. The review is not a prevalence estimate for all federal, state or local agencies.
The selected agencies reported four nongovernment facial-recognition services used from October 2019 through March 2022. They also reported about 60,000 facial-recognition searches before training requirements were in place. That cumulative figure uses agencies with available data; it is not a count of matches, an accuracy rate or a national total. Facial Recognition Technology: Federal Law Enforcement Agency Efforts Related to Civil Rights and Training
Across these sources, the measurement units matter. Department adoption, officer-weighted reach, equipment ratios, written-policy requirements, public-records requests, odds ratios, site-level crime estimates and selected-agency searches describe different parts of video surveillance. None alone measures the total number of cameras, continuous coverage, compliance with recording rules or a universal return on investment.