Different approaches to measuring specific deterrence : some examples from speeding offender management

Traffic law enforcement sanctions can impact on road user behaviour through both general and specific deterrence mechanisms. General deterrence is typically measured in terms of community-wide changes in offending behaviour, while specific deterrence is measured by re-offence or recidivism rates. However, recidivism can be conceptualised in a number of different ways. This paper compares and contrasts different measures of recidivism within the context of managing speeding offenders with regard to absolute and marginal specific deterrent effects. As part of a larger study evaluating the effects of penalty increases in 2003 on speeding offenders in Queensland, the research team considered four definitions of recidivism across two cohorts of drivers detected for speeding prior to (2001) and after (2003) the speeding penalty changes: 1) the proportion of offenders who re-offended in the follow up period; 2) the overall frequency of re-offending in the follow up period; 3) the length of delay to re-offence among those who re-offended; and 4) the average number of re-offences during the follow up period among those who re-offended. Any reduction in speeding will have road safety benefits, however, the ways in which a ‘reduction’ is determined deserves greater methodological attention and has implications for countermeasure evaluation more generally. Overall, results suggested evidence of an absolute deterrent effect of penalty changes in that there were significant reductions in the proportion of drivers who re-offended and the overall frequency of re-offending, consistent with prediction. However, there was no evidence of a marginal specific deterrent effect among those who re-offended. Contrary to prediction, there was a significant reduction in the length of time to re-offence and no significant change in average number of offences committed. Speed camera and radar data were analysed to explore whether the changes in re-offence patterns may have been influenced by changes in speed enforcement activity. This analysis indicated that the speed enforcement operational hours increased by 43% from the pre to post-penalty change period, suggesting that the reduction in offences observed in the follow-up period was not due to an enforcement effect. Limitations regarding the use of routinely-collected data, as well as issues of exposure, enforcement levels and public education accompanying penalty changes are discussed and areas for future research are outlined