Police recorded violent crime (PRC) and the Crime Survey of England and Wales (CSEW) show substantially different trends in the rates of violent crime according to the Office of National Statistics (ONS), with rates rising in police data and falling in survey data. Both the PRC and CSEW have suffered periods in which the UK Statistics Authority has withdrawn their quality approval as ‘national statistics’. This paper investigates possible seven reasons for the disparity in the trend and volume of violent crime between the PRC and CSEW, with a focus on the processes of measurement deployed.
The paper offers a new way to compare the methods and outcomes of the two data sources, by developing an ‘aligned’ data set to support comparison of trends in the PRC CSEW data since 2010. It analyses data from the PRC and from different sections of the CSEW, the main face-to-face module, the self-completion module on domestic abuse, and the children’s module asked of those aged 10-15 years. We offer new estimates of the volume and discussion on the trend in violent crime since 2010. We estimate that there were 5,164,983 violent crimes in 2022/3. This is significantly higher than the estimate provided by the ONS based on CSEW data. The estimate of the trend is uncertain, but challenges over-confidence in the assumption that it is declining.
We conclude that improvements in police accuracy in recording crime explains part of the difference, and the exclusion from sampling of vulnerable groups by the CSEW another part, with the recent reduction in the survey’s response rate to 42% giving further quality concerns. We determine that the CSEW has always underestimated violent crime, and this has become visible now that police data has improved.