Private and Public Security: Theory and Evidence from the UK

This paper examines how the private sector responds to reductions in the public provision of a key public good: internal security. It develops an infinitehorizon theoretical model in which police forces and potential victims jointly determine optimal levels of security investment. These theoretical predictions are examined using a novel panel dataset on private security licenses and police staffing across England and Wales from 2009 to 2019, exploiting variation generated by UK austerity policies that produced heterogeneous reductions in police numbers. To address endogeneity in public policing, two-stage least squares estimation is employed, using national police grants and local council tax precepts as instrumental variables. The results reveal substantial crowdingout: each police officer reduction per capita increases private security licenses by 0.52, with each £1 saved on police officers corresponding to approximately £0.41 spent on private security. This response is concentrated in functions closely aligned with policing. The analysis also uncovers important heterogeneity, with public policing becoming complementary to private security in wealthier and more unequal areas. Finally, the findings show that public police reduce all types of crime, whereas private security affects only property crime, underscoring the differential welfare implications of security privatization.