RESULTS: PHASE 2 QUANTITATIVE ANALYSIS

5 RESULTS: PHASE 2 QUANTITATIVE ANALYSIS

5.1 Primary safety estimates

There are a number of salient features of the crash risk charts that are relevant to the objectives of the study. The most important is that there is no evidence of a trend to increasing crash risk with increasing vehicle age for any of the taxi or hire car types. For the metropolitan taxis, crash risk is lowest at the beginning and end of the vehicle operating lifetime and highest in the mid years. Results of the operator survey showed that most metropolitan taxis are not first registered as a taxi when they are new. Instead they are purchased second hand and registered as a taxi at a median age of between 12 and 18 months old. Because exact compliance data of the vehicle is not recorded in the registration data, the year of compliance was used to estimate vehicle age. This means that a proportion of vehicles will not have been taxis for the full year in their first year of service as a taxi consequently leading to apparently lower risk estimates in the early years as observed in figures. Similarly, maximum operating age for non-WAT metropolitan taxis is 6.5 years meaning again the taxi will not have been registered for the full year in its seventh year of operation leading to lower estimated risk. Discounting the end points of the risk curve shows even less evidence of any age based trend in the crash risk estimates. Trends in crash risk by vehicle age in the peak service, country and urban taxi and hire car fleets are similar to those observed in the metropolitan taxi fleet.

Estimates of risk past eight years for taxis will relate to WATs. Although based on a short time period, there is no evidence of age based trends for these vehicles in their later years of operation. Since WATs could not be identified in the crash and registration data (since they are indistinguishable from other vehicles of the same type also used as regular taxis), it was not possible to study trends in the risk of WATs relative to regular vehicles but it is likely that the lack of age based crash risk trend in their later years of operation would be mirrored in earlier years of operation.

One clear limitation in estimating vehicle crash risk is the lack of data on vehicle travel exposure. Variation in the estimates of crash risk per registered vehicle year is likely to be driven by differential travel by vehicle type and age to some degree. Collecting vehicle odometer readings at times of random and periodic inspections of taxis and hire cars would allow vehicle travel to be estimated and included in the calculation of crash risk. This data enhancement is recommended for the future.

Some vehicle ages in Figure 5.1a-d have no crash risk estimated. These are vehicle types and ages where there were no recorded crashes in the police data. This in combination with the width of confidence limits for crash risk estimates on vehicle over seven years old show that crashes involving taxis and hire cars over seven years old are rare. This is particularly important for WATs, high luxury and stretched hire car vehicles when considering the safety impacts of changing age limits since these vehicle types are rarely involved in serious crashes. This will be further explored in the scenario analysis.

The final notable feature of the crash risk analysis is the difference in crash risk between hire cars and each taxi type and in particular the high average crash risk for metropolitan taxis. The average crash risk for each taxi type is provided in Table 5.1 above and shows the crash risks for metropolitan taxis is nearly four times higher than a hire car. As noted, this is in part due to the different travel exposures between taxi types.

The operator survey response provides estimates of relative mileage between taxi types which is summarised in Table 5.2 along with the annual crash risks per registered vehicle. From these, the exposure adjusted crash risk for each taxi and hire car type has been estimated relative to hire car crash risk. It should be noted that in Table 5.2 there were no specific travel estimates by taxi type so an average has been assigned based on the survey responses. It is likely that ST and PS taxi exposure is lower than that estimated which would mean the estimated relative exposure adjusted crash risk is under estimated. Table 5.2 shows that crash risk per kilometre travelled for each taxi type is around twice that of hire cars, with metropolitan taxis being the highest at 2.28 times the crash risk of a hire car.

Although the vehicle types used as hire cars are different to those used as taxis, although not substantially different with both dominated by large passenger sedans, these results suggest that there are driver based issues affecting crash risk in taxis compared to hire cars. Results suggest the crash risk associated with drivers of taxis is much higher than that for hire car drivers. The potential safety benefits of mitigating these differences are investigated in the scenario modelling.