The effect of accompanied driving (2toDrive) on self-reported crashes and violations : a questionnaire survey among all young drivers in the Netherlands. In 2011, a six-year accompanied driving pilot under the name of 2toDrive was launched in the Netherlands: young people can get their driving licence B before they are 18 years old, but until their 18th birthday they can only drive while accompanied by an experienced driver. This way they gain more and more varied experience before driving solo. Experiences in other countries show that accompanied driving can be an effective measure to reduce the crash risk of young novice drivers. To determine whether the measure also has a road safety effect in the Netherlands, SWOV is performing an evaluation. This report describes the third part of the evaluation and compares the crash involvement and traffic violations of young drivers who participated in 2toDrive (the ' 2toDrivers ') and young drivers who followed the "regular" driver training. The research was carried out using a questionnaire among all drivers between 18.5 and 20 years who had been in possession of a driving licence B for at least six months and were registered by the Dutch RDW Vehicle Technology and Information Centre. A total of 97.079 invitations to complete the questionnaire were sent out. Of these questionnaires, 30.763 (31.7%) were completed and returned. Eventually, we were able to analyse the questionnaires of 11.469 2toDrive participants and 18.410 of those who followed the regular driver training. First analyses show that at the time of filling out the questionnaire 2toDrivers had already driven significantly more kilometres as an independent driver than young drivers who had followed the regular driver training. This is partly due to the fact that on average 2toDrive participants had already driven independently for a longer period of time; after all, they could drive independently immediately after their 18th birthday, whereas many non- 2toDrive participants first had to take driving lessons. Furthermore, 2toDrive participants travelled more kilometres in the same time frame. Therefore 2toDrive participants are exposed to traffic hazards and enforcement activities to a greater extent. That is why the differences in crash involvement and traffic violations between the two groups were calculated in the form of a risk, i.e. corrected for the number of months as a solo driver and the distance travelled as a solo driver. The main conclusion is that it was not possible to determine an effect on the risk of a crash. Corrected for the number of months as a solo driver, hardly any differences in crash involvement were found between 2toDrive participants and young novice drivers who had followed the regular driver training. Adjusted for the distances travelled as reported by the 2toDrive respondents, their crash risk seems to be substantially lower than that of the drivers who had followed the regular training. However, statistically this appears to have no meaning. In fact, such great individual differences in distance travelled were observed that we cannot rule out that the difference found in the sample is a chance occurrence and cannot be generalized to the population as a whole. The effect of 2toDrive on committing traffic offences is inconsistent. When corrected for the number of months of solo driving, 2toDrive participants report a significantly higher number of fines than drivers who had followed the regular training. When adjusted for distance travelled another pattern emerges: 2toDrive participants report fewer fines per distance travelled. Unfortunately it was not possible to assess this difference 'per distance travelled' on statistical significance. In addition to the effect on crashes and violations, we also looked at some aspects of the learning phase and at the first experiences as a solo driver. As in an earlier study, we found that a considerable proportion (18%) of the 2toDrive participants under the age of 18 sometimes drive a car without a coach being present. However, we also found that almost half of the young novice drivers who had followed the regular driver training had driven without a driving instructor before they got their driving license. Both driving without a coach and driving without an instructor is usually done on private property or on forest or country lanes; 2toDrive participants do this on public roads slightly more often than learners following the regular training. Furthermore, it seems that part of the novice drivers with a regular training also make their first trips ‘accompanied' and thus apply some kind of selfregulation. Indeed, 40% of them say that they drove accompanied by a more experienced driver just after their driving test. When, after their 18th birthday, 2toDrive participants are allowed to drive independently, 20% initially do this with a more experienced driver in the passenger seat; nearly 60% mainly do this alone. A questionnaire study such as the present one was the only possible way to get an indication of the effects of 2toDrive in a reasonably short period of time. However, in this study we encountered an unforeseen problem: the differences in distance travelled proved to be so large that it was not possible to make the necessary corrections. Further research focuses on developing a model in which such corrections can correctly be made. In addition, this study is limited to only part of the young drivers, namely the 30% who completed the questionnaire. This is a very reasonable response rate, but it cannot be excluded that these are mainly the 'safer ' or, on the other hand, the 'more unsafe ' young drivers. The differences found in risk could therefore best an underestimate or an overestimate. Furthermore, for several reasons self-reporting is not always reliable and could especially bias the number of crashes and traffic fines. Therefore, we would also like to get insight into the real number of (registered) crashes and traffic offences of all 2toDrive participants as well as that of all traditionally trained drivers. This is not possible for traffic offences, because fines are mostly given on vehicle registration and not to individuals. In theory, such insight could be obtained for crashes if we were to add information about the age of passing the driving test (before or after the 18th birthday) to the real number of registered crashes. If possible, we would eventually like to base the evaluation of 2toDrive on this real number of (registered) crashes.
Abstract