Young Driver Research Program - A review of information on young driver crashes

Federal Office of Road Safety - Contract Report 128

Abstract:

A theoretical model of the determinants of young driver crash risk is outlined and some methodological issues concerning types of available data, measures of crash risk and measures of exposure to risk are discussed. Within this framework, published literature on young driver crashes in Australia and other countries is reviewed. The general nature and magnitude of the problem is described in terms of data on fatal and serious injury crashes derived from mass databases; information from more detailed studies of sub-sets of crashes is used to provide a more comprehensive description of the patterns of occurrence and characteristics of young driver crashes. Information is discussed in relation to driver age, experience and gender, time of crash occurrence, passenger characteristics, crash type, pre-crash vehicle manoeuvres, driver errors, and alcohol and seatbelt offences.

Executive Summary

The report outlines a theoretical model of the determinants of young driver crash risk. According to the model, driver characteristics are sub-divided into skill factors and motivation factors. Exposure to crash risk is determined by quantitative factors (distance driven), and qualitative factors (nature of the physical and social environment). Differences associated with driving at night versus driving during the day are seen to interact with many other variables within the model to affect crash risk. The literature underlying much of the theoretical model is reviewed in a separate report (Macdonald, 1994).

In the present report, published mass database information is used to document the magnitude of the young driver problem in terms of fatal and serious injury crashes. The extent of young driver involvement in less serious crashes is not discussed, since the available information is inadequate. Published information from more detailed studies of sub-sets of crashes is used to provide a more comprehensive description of the patterns of occurrence and characteristics of young driver crashes.

It is evident throughout the discussion of information about crash rates and their patterns that interpretation of crash data in terms of causal factors is difficult or impossible without more extensive and detailed information on the quantity and quality of driver exposure to crash risk than is currently available.

Results of the review are summarised below in relation to each of the following variables: driver age, driver experience, driver gender, time of crash occurrence, passenger characteristics, crash type, pre-crash vehicle manoeuvres and driver errors, alcohol and seatbelt offences.

Driver age

The relationship between driver age and crash occurrence is the central factor underlying the current research project. In general terms, young drivers are known to be over-involved in crashes. The extent and age-related pattern of young drivers' over-involvement varies somewhat in accord with the particular measure employed: crash involvement rates per head of population, per licensed driver, or per distance driven.

In terms of the first type of measure, the most common pattern is for drivers aged 1824 to be at higher risk than older drivers. However, the pattern is by no means uniform, and it is evident that knowledge of numbers of casualties per head of population in different age groups is an unreliable index of the crash risk of young drivers actually on the road. When rates are calculated in terms of numbers of young drivers licensed, the general pattern of an increased risk for young drivers appears to be common to all jurisdictions where, as in Australia, young drivers have a significant level of exposure to crash risk.

However, calculated crash risk for young drivers relative to the numbers of young licensed drivers does not provide a satisfactory picture of relative risk between different age groups or places, because there is evidence that the quantity and quality of drivers' exposure to risk (distance driven, types of traffic environments) differs considerably between different age groups in different ways in different places. Unfortunately, there is relatively little information on young drivers' crash risk calculated to take account of the quantity and nature of their exposure to risk, relative to other age groups.

The conclusion from those studies which take account of distance driven is that crash risk is highest for young drivers, at a minimum for those in middle age, and rises again for older drivers. It remains to be seen, however, how relative risk levels of drivers of different age groups might be changed if allowances were made for the quality of their exposure as well as its quantity.

Driver experience

In countries such as Australia, driver experience overlaps to a large extent with driver age, so the effects of these two factors are generally confounded. There is relatively little information from which the effects of experience independent of age can be determined.

It was tentatively concluded that while both age and experience may be significant determinants of crash risk, age tends to predominate among young males. For older people, and probably for females, experience has a larger effect than for young males.

Driver gender

Mass crash data are commonly reported separately for males and females, with males typically being involved in more crashes than females. Such differences are at least partially explicable in terms of differences in distance driven; that is, the quantity of exposure, and psychosocial differences between the genders, probably affect the quality of their exposure to crash risk. There is some evidence that such gender differences have decreased in Australia, North America and the U.K. over recent years.

In terms of driving behaviour, there is evidence that young male drivers are more likely than young females to be involved in crashes due to excessive speed, while young female drivers appear to be involved more often than males due to inadequate driving skills.

Time of crash occurrence

It is clear that exposure to risk is again a dominant factor determining variations in crash rates at different times of the day and the week for drivers of different ages. In studies which make no allowance for exposure, young drivers (particularly males) have very much higher crash rates at nights and weekends. However, when an attempt is made to allow for the proportion of time spent by different age groups driving at night and weekends, the pattern is much less clear cut.

Methods used to estimate exposure vary between different studies, and this may partly explain the differences in results obtained. Undifferentiated 'distance driven' data for different age groups is not sufficient; information on distance driven at different times is required, together with information on other aspects of exposure 'quality'.

Passenger characteristics

There is little which can be concluded from available evidence on the possible role of passengers as causal factors in young driver crashes. Information is needed on the numbers and characteristics of passengers normally carried by drivers of different ages at different times and places. Without such information no conclusions can be drawn concerning the possible causal role of passengers in young driver crashes.

Crash type, pre-crash vehicle manoeuvres and driver errors

The youngest and most inexperienced drivers (usually late teenagers) are involved in a greater proportion of single vehicle crashes than older drivers, particularly young male drivers at night. Such single-vehicle crashes typically involve running off the road, often due to excessive speed or 'reckless driving', and are more common in rural areas.

Other crash types in which some studies have found young drivers to be over-represented are head on, overtaking, and rear-end (young driver in rear vehicle). However, there is conflicting evidence on the latter type: some studies have specifically noted that young drivers were not over-represented in crashes due to 'following too closely'.

Very young drivers appear to be involved in crashes due to poor vehicle control skills more often than in the case of older drivers, and are more often responsible for causing the accident. (The oldest drivers are also 'at fault' in a high proportion of their crashes).

Inattention or failure to anticipate was identified as a problem in the older group of young drivers (aged in their early 20s), especially males; this problem appears to be associated with perceptual and cognitive errors. Drivers in this age group also tend to have a high rate of alcohol involvement.

Alcohol and seatbelt offences

The highest percentage of alcohol-related fatal crashes in Australia appear from available data to be in the 21-29 years age group; the youngest drivers (16-20 years) have a relatively low rate of alcohol involvement. This pattern is generally consistent with that in North America and the U.K.

There is very little information on the relationship between driver age and seatbelt wearing in the context of crash occurrence. There is some evidence that wearing rates are relatively low among young drivers (although not among the youngest drivers), and that non-wearing is associated with other non-driving behaviours which increase risk in some way (eg. poor vehicle condition, excessive BAC, greater number of passengers).