Outcome Evaluation of the Effectiveness of the Safe Routes to Schools Program

The Safe Routes to Schools (SRTS) community based road safety program has operated in primary schools throughout Victoria since 1990. The program includes education, engineering, encouragement and enforcement initiatives, aimed principally at reducing casualty crash frequency and severity for children as pedestrians, bicyclists and passengers. This paper describes the methodology and results of an evaluation of the crash effects of the Safe Routes to Schools program in Victoria. The evaluation examines the crash effects associated with the program across a number of road user groups at all times of day and at the times at which children are likely to be travelling to and from school. The analysis was conducted using a quasi-experimental, before and after, treatment and control design. The treatment areas were defined by the postcode areas surrounding the treatment schools and control areas were defined as all other postcode areas within the same Local Government Areas (LGAs) as the treated schools. The results of the analyses indicate that the estimated average yearly net effect of the program over the post-implementation period was beneficial in safety terms. The largest percentage reduction in casualty crashes was identified for primary school-aged pedestrians and bicyclists travelling during school travel times only (17.9%). Crashes involving primary school-aged pedestrians and bicyclists at all times, and crashes involving primary school-aged children at all times, were estimated to have fallen by 12.6 and 12.7 percent respectively. In addition, a 4.8 percent reduction in crashes involving all road users at school travel times was identified. Attempts to estimate the effect of the program in each year following the implementation of the program were statistically inconclusive, most likely because of insufficient data. In respect of the severity of crashes involving the relevant road users, no statistically reliable reductions in fatal and serious injury crash frequency could be identified at the five-percent level.