Does Reporting Deter Burglars?-An Empirical Analysis of Risk and Return in Crime

T HE demand for private protection against crime1 and the interaction between private security and public safety have been analyzed in recent studies. While the deterrent effect of public law enforcement has been treated extensively in the literature, systematic empirical analyses of the effect of private enforcement on crime are few.2 Our emphasis is on a form of private behavior that may substantially affect the success of public law enforcement: victims' reporting of crimes to the police. Because apprehension of criminals crucially depends on the reporting of crimes, and because reporting increases the chances of apprehension and conviction, reporting should deter potential offenders. To study this deterrent effect, we chose to concentrate on residential burglary. Our data, collected in the National Crime Panel (NCP) victimization surveys, reveal that only about half of all burglaries were reported to the police. Our basic premise is that among potential victims who offer the same expected gross returns, those perceived by burglars as more likely to report are less attractive targets. This perceived reporting probability is then a victim-specific deterrent variable. The NCP victimization survey data provide opportunities to improve upon previous studies of deterrence that use only aggregate crime statistics. First, they allow us to construct a victim-specific deterrence variable, rather than to assume that all victims in a state or city impose on the offender the same amount of risk. Second, information on loot from burglary enables us to measure directly the illegal returns. Third, in contrast to police-recorded crime statistics, the NCP surveys include information on all crimes, reported or not. Finally, our data on individuals make it easier to avoid the simultaneity problem encountered in studies that employ data on aggregates.3 Consequently, this study tests the deterrence hypothesis more directly.