Retention and disclosure of old criminal records - data protection or protecting the public?

Objectives // This presentation discusses the statistical and criminological issues involved in the retention of old criminal records for individuals who are now in their 30s and 40s but who had an early conviction when they were a teenager. The statstical issue is when do ex-offenders become like non-offenders in terms of their risk of reconviction. // Method/Models // Data from the England and Wales Offenders Index is used to investigate this issue, combined with England and Wales estimates of population. A dicrete-time hazard analysis of the hazard of reoffending is compared with the hazard for non-offenders. // Results and Conclusions // The analysis shows that the hazard curves become similar after a period of around 15 years. The use of this evidence by the Information commissioners in a recent Information Tribunal and the subsequent Court of Appeal case is discussed.