Boosting Offence Solvability and Detections: Solving Residential Burglaries by Predicting Single and Multiple Repeats

Abstract

Predicting repeat burglary incidence promises to make what may be safer burglaries for offenders into very risky ones by facilitating the use of entrapment techniques. These include installing covert CCTV, ‘bugging’ attractive items of goods, and planning rapid patrol responses to triggered silent and delayed audible alarms at dwellings at high risk of repeat victimisation. There is potential to help pick out the incidents at high risk of repeat burglary (single repeats) and then those at risk of yet further re-victimisation (multiple repeats). By isolating the subset of cases that are highly likely to be repeatedly victimised, burglary solvability may be improved and the existing detection rate of 6.3% cost-effectively raised. Twenty-three per cent of sample burglaries are at greatest risk of repeat burglary and just 11% of these face the highest risks of multiple repeat burglary. By using prediction scores derived from the distinctive characteristics of burglary sites selected for single repeat and multiple repeat burglaries, the subsets of dwellings highly liable to be re-burgled can be identified, and some of the potential realised for detecting in excess of an additional 7% of burglaries. With only 7.7% of single repeats and 53% of multiple repeats currently solved, there is considerable scope for improvement.