“All crimes are not created equal in the harm they cause: homicide is many times more harmful than shoplifting but in crime statistics where offences are counted by number, they appear equivalent. For example, in the UK for the year ending September 2019, there were 3,578,000 incidents of theft and 729 homicides (Office for National Statistics, 2019). An increase of 500 thefts would be a small change in the overall number of thefts and have little impact on police resources. 500 extra homicides would have large consequences both for the harm caused and the impact on police resources. In a number-only count, the additional 500 thefts or homicides would result in the same overall number of crimes, yet clearly the impacts are disparate.

This reality has led to the proposition of a “Harm Index” to measure how harmful different crimes are in proportion to the others. This approach adds a larger weight to more harmful crimes (e.g. homicide, rape and grievous bodily harm with intent), distinguishing them from less harmful types of crime (e.g. minor thefts, criminal damage and common assault). Practically, adoption of a harm index can allow targeting of the highestharm places, the most harmful offenders, the most harmed victims, and can assist in identifying victimoffenders. Experimentally, use of a harm index can add an additional dimension to the usual measures of success or failure, by considering harm prevented as well as reductions in prevalence or frequency. For the police, creation of harm index could allow them to invest scarce resources in proportion to the harm of each offence type.

Sherman, Neyroud and Neyroud (2016) propose that any index needs to meet three requirements in order to be considered a legitimate measure of harm: An index must meet a democratic standard, be reliable and also be adopted at minimal cost to the end user. To meet these requirements, Sherman, Neyroud and Neyroud (2016) opted for using sentence starting points rather than maximum or average actual sentences. The sentencing starting point is used to calculate crime harm as it provides a baseline penalty relative to the crime. We propose that it is a better measure of harm caused by the crime than average actual sentences, which are offender-focused and thus substantially affected by previous offending history.

The Cambridge Crime Harm Consensus proposes creation of seven statistics for counting crime, usefully including separation of historic crime reports, creation of a harm detection fraction and separation of public reported crime and those detected by proactive police activity, with the aim of providing the public with a more reliable and realistic assessment of trends, patterns and differences in public safety.

Counting crime by harm is an idea that has spread beyond the United Kingdom with indices published for Denmark (Andersen and Mueller-Johnson, 2018), Sweden (Karrholm et al. 2020), Western Australia (House and Neyroud, 2018), California (Mitchell, 2017), New Zealand and other countries.”