Evaluation of the Crash Effects of the Queensland Speed Camera Program

A speed camera program was introduced in Queensland from 1 May 1997 utilising overt deployment of cameras in white commercial vans marked as a speed camera unit at sites chosen on the basis of crash history. Sites at which cameras are operated at on a particular day are chosen using randomised scheduling and have grown in number from 500 at program commencement to over 2,500 by June 2001. This study has investigated the crash effects of the speed camera program in Queensland over the period from its introduction in May 1997 to the end of June 2001 in areas within 6km of speed camera sites that had been used up to the end of the study period. When operating at maximum coverage, the Queensland speed camera program was estimated to have produced a reduction in fatal crashes of around 45% in areas within 2km of speed camera sites. Corresponding reductions of 31%, 39% 19% and 21% were estimated for hospitalisation, medically treated, other injury and non-injury crashes respectively. This translates to an annual crash saving in the order of 110 fatal, 1100 hospitalisation, 2200 medically treated, 500 other injury and 1600 non-injury crashes. In terms of total annual road trauma in Queensland, these savings represent a 32% reduction in fatal crashes, a 26% reduction in fatal to medically treated crashes combined and a 21% reduction in all reported casualty crashes. The benefit cost ratio estimated for the program over the period from its introduction to June 2001 was 47. Comparison of t he estimated crash reduction associated with the Queensland speed camera program and program operational measures gave insight into the mechanisms of program effectiveness. Variations in estimated crash reduction over time were strongly related to the size of the overall program and the density of enforcement. Periods of program growth were also associated with larger crash reductions beyond that expected from the increasing size of the program alone. Higher levels of true randomness in selection of speed camera sites for operation was also associated with higher levels of crash reduction when comparing differential performance of the program across police regions in Queensland. This report is disseminated in the interest of information exchange. The views expressed here are those of the authors, and not necessarily those of Monash University