Time vs. distance as measures of exposure in driving surveys.

A survey of drivers carried out in Ontario in 1988 has provided data on time spent driving as well as the distances driven for licensed drivers of both sexes in six age groups and three regions. Substantial differences were found in times, distances, and distance/time ratios among these groups. Men drove 50% greater distances, but spent only 30% more time driving than women; speed, averaged over each day's driving, was lower for older drivers than for younger drivers. Differences in speed reflect differences in the driving done in urban or rural areas, and differences in the opportunity for road crashes; such differences, whether based on units of time or distance, will also affect both the comparisons of accident rates and the perceptions of risk among different groups of drivers. A definition of exposure to risk of road crash is required that considers both time and distance appropriately.

The role of exposure in comparisons of crash risk among different drivers and driving environments.

Crash rates based on drivers, driver-kilometers, and driver-days in the denominator were compared, using survey estimates of time and distance driven and the annual frequency of traffic crashes in Ontario. Rates by age, sex, and region were computed for all crashes and for crashes resulting in injury or fatality. Young male drivers remained at high risk for all types of denominator; older women had high rates when distance was included in the denominator. When time spent driving was substituted, men and women drivers over 60 had very similar rates. For comparisons of rural residents with urban and northern residents, time and distance give equivalent results. These findings suggest that apparent differences in crash risk per kilometre, whether for older women or for urban drivers, is explained by differences in typical driving speed and environment. Exposure time is better than distance to explain crash risk among drivers and regions with very different driving patterns and environments.