Victoria collects a Motorcycle Safety Levy from all registered motorcycles from which funds are directed into a Motorcycle Blackspot Program (MBP) that provides treatments to improve motorcycle safety at locations throughout the state with a history of motorcycle crashes. VicRoads commissioned ARRB to conduct an evaluation of the MBP in terms of its crash reduction effects and the associated economic returns. Data were received for 176 treatments, made up as follows: 9 barrier protection treatments, 4 intersection treatments, 61 long route treatments, 92 loss-of-control treatments, 4 roundabout trial treatments and 6 variable message sign trial treatments. Crash records were matched to sites using the ArcView GIS software. A quasi-experimental design was followed, with road sections adjacent to the treatment sites adopted as the control sites, except in the case of the long route treatments where other routes which were broadly similar to the treatment routes were used. The crash reductions and their significance were estimated by fitting a mixed generalised linear negative binomial model with sites nested within sub-programs. This procedure takes into account changes in the number of crashes at the control sites, an essential step since there was an upward but fluctuating trend in motorcycle travel over the life of the program. Statistically significant crash reductions were found for the program overall with an estimated 27% reduction in casualty crashes and an estimated 31% reduction in fatality and serious injury crashes after adjustment for changes at the control sites. When the different treatment types were considered separately, there were substantial crash reductions although only one of these was statistically significant. This was the barrier protection treatment, which produced a highly significant reduction of 74% in FSI crashes. Results for the other treatments were highly variable from site to site; results were not statistically significant, but the FSI crash reductions were substantial in the case of the long route and loss-of-control sites, 29% and 42% respectively, while the intersection sites showed a 69% reduction although the numbers were much smaller. More detailed examination of the crash data showed that the types of crash which had reduced corresponded with what would be expected given the nature of the countermeasures. There was no evidence of a crash migration effect. Best performing sites were identified and discussed. The cost of the program was $32 million. Considered from the point of view of all motorcycle casualty crashes, the program has reduced all casualty crashes by 27%, which is statistically highly significant. Considered in these terms, the BCR ranges from 7.6 to 6.3 and the NPV ranges from $211 million to $170 million, depending on the discount rate adopted. Considered from the point of view of motorcycle FSI crashes, the program has reduced FSI crashes by 31%, which again is statistically highly significant. Considered in these terms, the BCR ranges from 8.5 to 7.1 and the NPV ranges from $240 million to $195 million, depending on the discount rate adopted. The average cost of preventing an FSI motorcycle crash was estimated at almost $80 000. The main conclusions of the study were: 1. The program has been successful in reducing motorcycle casualty crashes (by 27%) and FSI crashes (by 31%), both these reductions being statistically highly significant. 2. The program also showed good economic returns. 3. The barrier protection program has been particularly effective in reducing FSI crashes (by 74%), and shows the best economic returns. 4. The long route treatments and the loss-of-control treatments have both been successful in reducing crashes and show good economic returns. In both cases, sufficient numbers of sites have received the treatments to allow confidence in the results. 5. The intersection treatments also showed good reductions in motorcycle crashes, but the number of sites is small; trials at more sites are needed before full confidence can be placed in it. (Author/publisher)
Abstract