Use of age–period–cohort models to estimate effects of vehicle age, year of crash and year of vehicle manufacture on driver injury and fatality rates in single vehicle crashes in New South Wales, 2003–2010

• Age–period–cohort methods are used to explain changes in vehicle based crash rates.
• The risk of crashing increases with vehicle age.
• There is small decline in crash rates with crash period.
• Vehicle cohort year appears to be associated with a strong decline in crash rates.

A novel application of age–period–cohort methods are used to explain changes in vehicle based crash rates in New South Wales, Australia over the period 2003–2010. Models are developed using vehicle age, crash period and vehicle cohort to explain changes in the rate of single vehicle driver fatalities and injuries in vehicles less than 13 years of age. Large declines in risk are associated with vehicle cohorts built after about 1996. The decline in risk appears to have accelerated to 12 percent per vehicle cohort year for cohorts since 2004. Within each cohort, the risk of crashing appears to be a minimum at two years of age and increases as the vehicle ages beyond this. Period effects (i.e., other road safety measures) between 2003 and 2010 appear to have contributed to declines of up to about two percent per annum to the driver-fatality single vehicle crash rate, and possibly only negligible improvements to the driver-injury single vehicle crash rate. Vehicle improvements appear to have been responsible for a decline in per-vehicle crash risk of at least three percent per calendar year for both severity levels over the same period. Given the decline in risk associated with more recent vehicle cohorts and the dynamics of fleet turnover, continued declines in per-vehicle crash risk over coming years are almost certain.