Guidance on using the social cost estimates - The Social Cost of Road Crashes and Injuries June 2008 Update

3. Guidance on using the social cost estimates

3.1 Adjustment for non-reported incidents

Tables 4.1a and 4.1b provide the estimates of average social costs per crash and per injury, without adjustment for under-reporting. These estimates are suitable only for cases where we know the total number of crashes and injuries.

For convenience of application, Tables 4.3a to 4.3c provide the estimates of average social costs per reported crash and per reported injury, after adjusting for the level of non-reporting. If a programme is expected to reduce the number of injuries, but not the number of crashes, use the estimates from Table 4.3c. Otherwise, use the estimates from Tables 4.3a and 4.3b, depending on data availability and the purpose of the analysis.

3.2 Estimates for rural and urban areas

The estimated average social cost per crash for rural areas tends to be higher than that for urban areas, because rural crashes tend to be more severe and often result in a larger number of fatal and serious injuries. Thus, if an intervention affects only one area type (either rural or urban), the corresponding social cost estimates should be used.

Table 4.2 shows the estimated average social cost per crash by area and severity, without any adjustment for non-reported incidents. All other tables with estimates by area include adjustments for under-reporting.

3.3 Estimates by region and vehicle movement

Table 4.4 provides estimates of the average social cost per reported crash by vehicle movement, using crash data from 2003 to 2007. These estimates have been adjusted for the level of non-reporting and are suitable for analysing policies or programmes that focus on specific vehicle movement classifications (eg head-on crashes).

Due to differences in physical locations, sizes of regions, the availability of facilities, and for other reasons, the proportions of injury crashes that are reported to the police differ across regions. Furthermore, the mix of rural and urban crashes also differs across regions. As a result, there are regional variations in the average social costs per reported injury and per crash.

Tables 4.5a and 4.5b provide the regional average social costs per reported crash and per reported injury respectively, using crash data from 2005 to 2007. These estimates have been adjusted for the level of non-reporting and are useful for the evaluation of regional programmes or policies. Table 4.5c provides the regional average social costs per reported injury, excluding the associated property damage costs. These estimates have been adjusted for the level of non-reporting and are useful for estimating the social cost of injuries associated with pedestrians and cyclists at the regional level.

Historical regional estimates using year-specific crash and injury conversion factors are incorporated into the Ministry’s Crash Analysis System and are available upon request.
 

3.4 Estimates for an increase in risk

While the majority of safety programmes or projects intend to reduce crash or injury risk, some programmes or projects could result in an increase in risk but produce other benefits. In this situation, the estimate of social cost for an increase in risk should be based on those derived from the willingness-to-accept (WTA)-based Value of Statistical Life (VOSL). The WTA-based value represents the amount of money the public would need to receive or save in exchange for an increase in risk. In a value of safety study conducted in 1997/98, the WTA-based value was found to be around three to five times the willingness-to-pay (WTP)-based value (Guria et al., 2003).

Tables 4.6a to 4.6c provide the estimates with WTA-based VOSL at three times the WTP value. These estimates have been adjusted for the level of non-reporting and are useful for analysing any programme that may result in an increase in risk of crash or injury to road users.

3.5 Estimates for infrequent events

Some of the social cost estimates in this document include estimates for a combination of crash or injury types: fatal and serious, serious and minor, and all three. These estimates are useful for assessing safety risks that could cause severe injury to road users but have a low probability of occurrence (eg in situations where the crash or injury numbers are small).


<< Previous | Contents | Next >>

• The Social Cost of Road Crashes and Injuries June 2008 update (PDF v7.0 156kb)