In the Dutch national statistics for traffic accidents (AVV; BRON) a sharp decline in the number of fatalities in passenger cars has been ascertained during the years 2004 and 2005. It is considered useful to know the cause of this sharp decline and for this reason the ministry of Traffic, Public works and Water management has formulated two questions: 1) Is the decline of the number of traffic victims attributable to increased vehicle safety; 2) If increased vehicle safety has an effect, which part of the decline can be attributed to this fact? In order to answer these questions TNO Automotive has computed the crashworthiness (protection from injury due to an accident) of passenger cars. To this end the first year of registration of the car at the RDW (the Dutch road traffic department) is used. This is not the same as the year of introduction of a car model, which would have been better. Unfortunately the model year is not included in the national statistics. Based on the crashworthiness of a car the expected number of fatalities in a given year can be calculated. To this end the vehicle fleet that was involved in registered accidents in that specific year was used. The in this research observed change in the number of traffic accident victims may be attributed to a renewed and more crashworthy vehicle fleet. The year 2001 was used as a reference year for the calculations of the expected change, because a match between vehicle registration and vehicle type has been available in the national statistics since 2001. Due to the lack of a match between vehicle registration and vehicle type in national statistics before 2001, nothing can be said on the influence of vehicle charactaristics on injury severity in the pre 2001 period. Compared to 2001 a reduction of 16% of the number of fatal car-car accidents has been calculated for 2004. This follows the trend that was ascertained for car - ear accidents. For 2005 a reduction of 20% was predicted based on the involved vehicle fleet, which is also in the declining trend. The observed number of fatal car - car accidents in 2004 and 2005 is lower than the predicted number, however there is no longer an extreme difference with the prediction and could therefore be explained in part by random permutations in the course of the years. It follows from these calculations that the decline in the number of fatal car - car accidents can be explained by an increase in crashwcirthiness. No influence of crashworthiness on the number of fatal car - object accidents has been found. The ascertained decline in this category must therefore be attributed to other factors. It's possible that among others primary safety aspects such as electronic stability programs have contributed to the decline. This possibility has not been subject to further study. For 2004 a decline of approximately 37 fatalities and 287 hospitalized victims due to car accidents can be attributed to improved crashworthiness of cars. For 2005 this number is approximately 48 fatalities. This entails that an observed reduction in 2004 and 2005 of respectively 81 (46%) and 140 (34%) fatalities can be explained by improved crashworthiness. The reduction in the number of severely injured victims in 2005 by improved crashworthiness can not be calculated, because accidents with slight injuries and property damage only were not yet available. For 2004, 42% of the real observed reduction of the 678 registered hospitalized victims can be explained by improved crashworthiness. As indicated in the report, the calculation of the crashworthiness is conservative, so that a small extra reduction (smaller than 2%) may be attributed to improved crashworthiness. The analysis of PENDANT data indicates that, based on the available data, it can not be stated that cars with four or five EuroNCAP stars have a significantly better safety performance than cars with two or three EuroNCAP stars. The number of accidents is potentially large enough to indicate an effect, if this effect were present. The fact that no significant effect on safety can be found means that: either a more advanced analysis method that requires more time has to be used, or that the effect is smaller than anticipated. Even though (Lie et al., 2000) indicates that three and four star cars are 30% safer than one and two star cars, it is not immediately possible to judge the reliability of this result. Keeping in mind the small number of accidents investigated the result may be optimistic. However, it does appear to be the case that there is a relation between EuroNCAP stars and safety. Concluding: newer cars have a high crashworthiness than older cars, but the difference in crashworthiness between new cars - the difference in EuroNCAP stars - is not yet reliable. (Author/publisher)
Abstract