Are Neighborhoods with More CCTV Really Safer?

CCTV can help reduce crime, but the effect depends heavily on where the cameras are and how they are run. A meta-analysis combining 41 studies found crime fell 16% on average. In car parks it fell 51%, while in city centers and public housing it fell only 7%, and those smaller drops were not statistically significant. So a neighborhood with more cameras is not automatically a safer one.

Camera counts are the first number people reach for when they talk about neighborhood safety. This post looks at what the two best-known reviews of CCTV research compared, what they found, and how public CCTV in Korea's open data differs from the cameras in those studies.

What the research compared

CCTV evaluations usually compare an area that got new cameras (the experimental area) with a similar area that did not (the control area), before and after installation. If crime fell more in the experimental area than in the control area, the difference is credited to CCTV.

These studies did not compare neighborhoods with many cameras against neighborhoods with few. They looked at what happened to crime in a place after a CCTV program was introduced there. Reading the results as "more cameras means more safety" stretches them beyond what they measured.

The 2008 meta-analysis: 16% on average, very different by setting

The often-cited "16% reduction in crime" comes from a 2008 systematic review by Brandon Welsh and David Farrington for the Campbell Collaboration. It gathered 44 evaluations that had a comparable control area and ran a meta-analysis on the 41 with enough crime data. Across all 41, crime fell 16% more in CCTV areas than in control areas.

That average combines very different results:

Setting	Studies in meta-analysis	Crime reduction	Statistically significant
Car parks	6	51%	Yes
Public transport (subways)	4	23%	No
City and town centers	20	7%	No
Public housing	8	7%	No

A few details matter:

• The car park results include other measures. None of the six car park schemes relied on cameras alone. They also added improved lighting, fencing, or security staff, so the 51% cannot be credited to CCTV by itself.
• Vehicle crime responded; violent crime did not. Across 22 studies vehicle crime fell 26%, while across 23 studies there was no overall effect on violent crime.
• Results differed by country. 34 of the 41 studies were from the UK, where crime fell 19% on average. The other 7 (US, Canada, Norway, Sweden) showed no effect. The UK result itself was driven largely by the car park schemes.

The 2019 update: a similar conclusion

A 2019 meta-analysis, with the same researchers among its authors, widened the scope to 40 years of evaluations. Its conclusions were close to the 2008 review:

• CCTV was associated with a modest but significant drop in crime.
• The largest and most consistent effects were again in car parks, with significant reductions in residential areas too.
• Systems with active monitoring produced larger effects than passive, record-only systems.
• Schemes that combined CCTV with several other interventions did better than CCTV alone.
• The authors recommended targeting CCTV narrowly at vehicle and property crime.

Does crime just move next door?

A common worry is that cameras only push crime into nearby areas without cameras. Criminologists call this displacement. The opposite, where nearby areas also see less crime, is called diffusion of benefits.

In the 2008 review, few studies used the minimum design needed to detect either (a camera area, an adjacent area and a separate control area), and the authors called any conclusion premature. Among those that did measure it, the usual finding was no displacement, but the two subway studies that checked split: one found diffusion and the other displacement. On the 2008 evidence, it was too early to say which way it goes.

How public CCTV in Korea differs from the cameras in the research

In Korea's national CCTV standard dataset (전국CCTV표준데이터), data downloaded 2026-09-20, 67.7% (171,081) of the 252,797 public CCTV sites assigned to a province are for crime prevention. The rest: other purposes 13.3%, child safety 7.0%, traffic (enforcement and information collection) 6.6%, disaster safety 2.8%, and facility protection 2.7%. The provincial breakdown is in Korea's public CCTV by region.

Compared with the research, a few differences stand out:

• The settings differ. About two-thirds of Korean public CCTV sites in the open data are registered as 생활방범 (neighborhood crime prevention), and no record is registered for car parks, the setting where the effect was largest. The 2019 analysis did find significant reductions in residential areas, but they were not as large as in car parks.
• The 2008 meta-analysis is mostly UK evidence. 34 of its 41 studies were British, and results elsewhere were weak. That makes it hard to carry the numbers over to Korean neighborhood cameras.
• A count says nothing about operation. What strengthened the effect in the research was live monitoring and measures like lighting. The open data lists each site's purpose and number of cameras, not who watches the screens or how closely.
• The counts themselves are imperfect. Some records share coordinates or are outdated, so the number on a map can differ from what is actually installed. See How Accurate Is Korea's Public CCTV Dataset? for details.

What to look at instead of the count

For everyday decisions, these tell you more than a total:

1. Is it on your route? Cameras on the street you walk every day, at your door, or near your bus stop matter more than the neighborhood total.
2. What is it for? A crime prevention camera and a traffic information camera point in different directions and cover different areas.
3. Lighting and sightlines. The car parks where CCTV worked best also got better lighting. At night, a bright street with open views matters too.
4. How long footage is kept. Public CCTV footage is usually deleted after about 30 days. If something happens, request the footage right away.

To see which public cameras are along a route you use, search the neighborhood or street on Viewcone. Tapping a site shows its purpose and managing agency. The map shows only cameras in the open data; it does not measure actual safety or show every camera.

FAQ

By how much does CCTV reduce crime?

The 2008 meta-analysis found an average 16% reduction across 41 studies. It varied widely by setting: 51% in car parks, 23% on subways, and 7% in city centers and public housing. Outside car parks, the effects were not statistically significant.

Does CCTV reduce violent crime?

No overall effect was found. Across the 23 studies in the 2008 review that measured violent crime, CCTV did not reduce it, so cameras alone are unlikely to make a dent in violent crime.

How can I see how many cameras are in my neighborhood?

Search a neighborhood, place or street name on Viewcone to see nearby public CCTV sites; the region index lists site and camera counts for each province and district. The full how-to is in How to check CCTV locations near you in Korea.

Sources

• Welsh, B. C., & Farrington, D. P. (2008). Effects of Closed Circuit Television Surveillance on Crime. Campbell Systematic Reviews. The 16%, 26%, 51%, 23%, 7%, and 19% figures, the violent crime result, and study counts in this post come from this review's meta-analysis.
• Piza, E. L., Welsh, B. C., Farrington, D. P., & Thomas, A. L. (2019). CCTV Surveillance for Crime Prevention: A 40-Year Systematic Review with Meta-Analysis. Criminology & Public Policy, 18(1). Summary: U.S. Office of Justice Programs
• National CCTV standard dataset - data.go.kr (Ministry of the Interior and Safety, Korean). Purpose shares use the 252,797 sites assigned to a province in data downloaded 2026-09-20.