What Should We Expect from Police Data: Can They Tell Us Whether Crime Rates Rise or Fall?

Abstract

Police data on registered crime are incomplete, inconsistent, and susceptible to manipulation. This has long been understood. Less well understood are changes over time in victims’ and others’ patterns of reporting to the police that sometimes make apparent changes in crime figures fundamentally misleading. Police data in recent decades in many countries, for example, overstated increases in violence rates, especially when the real incidence of violence was rising, and understated declines when the real incidence was falling. Changes in victim reporting are understandable; they partly reflect widely recognized changes in social norms and attitudes toward disturbing behavior. Behaviors affected include drunken driving and violence against women. Even more confounding are changes in thresholds of tolerance for violence and other disturbing behavior that shape citizens’ answers to victimization surveys and their decisions to report alleged crimes, and police decisions to register them. Disturbing behaviors, for example, minor violence, sexual misconduct, and impaired driving, that formerly were not viewed as criminal now are. Taken together, changes in victim reporting and in the effects on victims and police of changing thresholds of tolerance have in some European countries generated misleading police crime data that indicate that violent and sexual offending has increased in recent years, or been stable. The best evidence, however, is they are much more likely to have declined.

Keywords: police data, crime trends, changing thresholds of tolerance