Surveillance cameras and crime: a review of randomized and natural experiments

Abstract

Research on the effectiveness of surveillance cameras in reducing crime suffers from potential threats to causal validity. This paper reviews seven studies that address some of these problems using the rigorous research designs of randomized and natural experiments. Included studies that reported changes in total crime found crime reductions ranging from 24 to 28% in public streets and urban subway stations, but no desirable effects in parking facilities or suburban subway stations. Moreover, surveillance cameras may help reduce unruly behaviour in football stadiums and theft in supermarkets/mass merchant stores. These findings indicate that video surveillance can reduce crime in several settings.

Keywords:

• Surveillance cameras
• CCTV
• crime prevention
• randomized experiments
• natural experiments
• review

Acknowledgements

I am grateful to Manne Gerell, Markus Lahtinen, Caroline Olsson, the editor and two anonymous referees for helpful comments on an earlier draft of this paper. Needless to say, any errors are mine.

Notes

1. Consulted experts were: Mikael Priks, Department of Economics at Stockholm University, and Brandon Welsh, School of Criminology and Criminal Justice at Northeastern University. I am grateful for their suggestions.

2. Although one of the consulted experts suggested that ‘it may be possible’ that Webb and Laycock (1992) and Sivarajasingam, Shepherd and Matthews (2003) could be classified as natural experiments, they were excluded since the authors did not provide any reason for believing that camera installations were exogenous.

3. This divergence should not be exaggerated since there are several exceptions. For instance, the study by Priks (2015a) in the present review found no significant crime changes in the subway system as a whole, whereas the study by Webb and Laycock (1992) in Welsh and Farrington’s (2008) meta-analysis found a reduction of robbery.

4. Indeed measurement errors, including some of the ones discussed in the previous paragraph, may also induce endogeneity.