Measuring changing rates of violence in government surveys by including the experience of high frequency victims

• In Event: Patterns and Trends in Victimization Risk

Abstract

In both the National Crime Victimization Survey (NCVS) and the Crime Survey for England and Wales (CSEW), the number of crimes reported by a victim which are included in official estimates of crime is capped or trimmed to a maximum value. In the NCVS, any count greater than 10 for a serial crime in six months is capped at 10; in the CSEW serial crimes are capped at five per 12 months. While such capping does not change the prevalence crime rate, it introduces considerable bias into the incident crime rate, especially for violent crime. We discuss the concept of capping and suggest that this methodology is flawed. We investigate the effect of removing the cap on violent crime rates in England and Wales, instead including all reported crimes as the basis for estimating national crime rates, and using other methods to reduce year to year volatility. We show that when the full extent of the victimization of high frequency victims is made visible, then an increase in some violent crime rates is discovered, namely for violent crime against women and domestic violent crime. The increase is estimated to start from 2008/9, the start of the financial crisis.

Authors

• Jude Towers, Lancaster University, UK

• Sylvia Walby, Lancaster University

• Brian Francis, Lancaster University, UK