Private Security Confounds Estimates of Public Police and Crime

The effectiveness of public policing has become an important issue given recent discussions of defunding law enforcement. This has also led to a discussion about the role of private security in potentially filling the void of creating public safety in the absence of public security. Thus, it is implicitly assumed that private security and public police are substitutes for one another. Alternatively, these two forms of security could complement each other. Causally determining whether private security and public police are complements or substitutes for one another is complicated by numerous layers of endogeneity. Nevertheless, excluding the presence of private security from a causal analysis of the effect of public police on crime could result in omitted variable bias. In this work we utilize survey data on private security and public police, as well as crime data, by county over the period 2006–2019 to explore the extent that omitting private security confounds an analysis of public security and crime. Our results indicate that it is a non-trivial omission to exclude private security from such analyses.