Young Driver Research Program - Mass crash data analyses: Overview of results from Australian and USA mass crash database analyses

Federal Office of Road Safety - Contract Report 131(11)

Full report in .pdf format [8.5MB]

Abstract:

This report presents an overview of results tabled in other reports in this series which separately analysed Australian and USA mass crash databases. The focus was on identifying differences between the crash patterns of young and mature car drivers which occur consistently across databases, as well as consistent differences between daytime and night-time crash patterns. Results showed that day/night differences were often large, and usually of similar magnitude for both young and mature drivers. Higher young driver compared to mature driver involvement was shown for single vehicle crashes, crashes occurring on curves or slopes, travelling above the speed limit or at excessive speed, and crashes where 'loss of control' was identified as a causal factor.

Executive Summary

INTRODUCTION

The Monash University Accident Research Centre was commissioned by the Federal Office of Road Safety to undertake a Young Driver Research Program as part of a Federal Government Road Safety Initiative. One of the research projects in this Young Driver Research Program focussed on Young Driver Crashes. As part of this project, the following Australian and USA databases were analysed:

• New South Wales: casualty data 1986 to 1990
• Victoria: casualty data 1984 to 1989
• South Australia: casualty data 1986 to 1990
• Federal Office of Road Safety (FORS): 1988 Fatality File containing data for New South Wales, Victoria and South Australia
• USA General Estimates System (GES): 1989 data files. These databases provide a probability sample of USA road crashes, compiled by the NHTSA from the original police accident reports.

PURPOSE AND NATURE OF OVERVIEW ANALYSIS

This report presents an overview analysis, incorporating results from all of the above databases. Results from each database are presented separately (in appendices) in the form of crash frequencies and group percentages for each crash descriptor variable, calculated within age groups (young/mature) and time of day (day/night) subcategories. Cross-tabulations combining results from the appendices are presented in Tables 1 and 2. Table 1 shows differences between age groups, for day and night separately; Table 2 shows differences between day and night for each age group separately. Magnitudes of differences are shown in terms of absolute number of percentage points difference between the young/mature and day/night percentage figures presented in the appendices.

Focus is on identifying differences between the crash patterns of young and mature drivers which occur consistently across databases, and subsidiary to this, consistent differences between daytime and night-time crash patterns. The conceptual framework of the research identified day/night variation as a factor likely to interact with driver age and other factors in influencing crash risk (see Figure 1, Macdonald, 1993b). The overview analysis explores this issue within the bounds of information available from mass databases.

RESULTS

Major results are presented below, grouped within the main categories of crash variables.

Type and location of crash

Proportions of young drivers were higher than the corresponding proportions of mature drivers for the following types of crashes:

• on curves and on slopes, as opposed to straight and level road sections, particularly at night

Proportions of drivers involved in crashes on curves rather than on straight road sections were higher at night than during the day for both young and mature drivers; the difference was a little greater for young drivers.

For the Australian Fatals database, proportion of drivers in crashes on slopes was considerably higher at night than during the day for young drivers, but for mature drivers the proportion on slopes was considerably lower at night. The only other databases having this variable were the USA ones, where no trend in either direction was apparent.

• off the road, or on the road shoulder

Proportions of drivers in 'off road/shoulder' crashes were much higher at night, for young and mature drivers equally.

• off footpath, on straight - particularly at night

• off footpath, on curve or turning - particularly at night

Proportions of drivers in 'off path on straight' and 'off path, curve/turning' crashes were higher at night for both young and mature drivers; the relative increase was a little greater for young drivers.

• crashes in which the impact was with an object other than another vehicle particularly at night

Proportions of drivers involved in these crashes were always considerably higher at night than during the day for both young and mature drivers; the night-time increase was considerably greater for young Australian drivers, but for USA drivers the night-time increase was a little greater for the mature drivers.

The broader category of 'single vehicle crashes' (not confined to collisions) also showed higher proportions of night-time relative to daytime drivers involved, in this case for young and mature drivers fairly equally.

• in vehicle/vehicle crashes when the young driver was in the striking vehicle, rather than the struck - but only in daytime (USA data only)

Proportions of drivers in vehicle/vehicle collisions (two or more vehicles) were lower at night than during the day for both young and mature drivers; the night-time decrease was greater for young drivers when they drove the striking vehicle, and greater for mature drivers when they drove the vehicle struck.

• in urban rather than rural areas, and in 60 kph speed zones, for both day and night-time fatal crashes (Australian Fatals database only - not in other databases)

• where there were no controls, as opposed to where there were traffic lights, stop/give way signs or other forms of control

Proportions of drivers who were crash-involved at stop or give ways signs tended to be lower at night than during the day, equally for young and mature drivers

• when the vehicle was 'moving along roadway' as opposed to turning, reversing or stationary

• particularly at night the night-time increase was greater for mature than young drivers

Proportions of drivers who were crash-involved while turning/reversing or stationary, were lower at night than during the day for both young and mature drivers.

• left turns, USA - daytime only

• pedestrian - daytime Australian fatals only

The converse of young drivers' relatively higher proportions of the 'single vehicle' types of crash were their relatively lower proportions of vehicle/vehicle collisions, including right-angle, adjacent directions, rear-end, and head-on.

Proportions of drivers in vehicle/vehicle crashes were generally lower at night than during the day - again, the converse of the pattern for single vehicle crashes.

As would be expected in view of this pattern of day/night variation in vehicle/vehicle crashes, proportions of drivers at most types of intersection crashes were lower at night than during the day for both young and mature drivers, and proportions of non-intersection crashes were higher at night.

Vehicle speed (data only from NSW and Australian Fatals databases)

Proportions of young drivers were higher than the corresponding proportions of mature drivers for the following types of crash:

• speeds greater than 60 kph, particularly at night

• for both young and mature drivers, proportions were higher at night than during the day

• 'definitely over the speed limit', particularly at night; the increase in night-time proportions was greater for young than for mature drivers

• combination of all 'speeding' categories: 'possibly over limit', 'definitely over limit' and 'within limit but excessive for conditions'

Proportions were much higher at night than during the day, equally for young and mature drivers

• 'excessive speed' identified as a causal factor, mainly day (Australian Fatals)

Control of vehicle

Proportions of young drivers were higher than the corresponding proportions of mature drivers for:

• 'loss of control' identified as a causal factor, particularly at night

• the night-time increase was greater for young drivers (data only for NSW)

Inattention, failure to observe

Proportions of young drivers were higher than the corresponding proportions of mature drivers for:

• crashes when driver 'in attention' was judged to be a factor, both day and night (SA database only)

Proportions of 'inattentive' drivers were higher at night than during the day, for young and mature drivers equally.

Proportions of drivers coded with 'failure to observe person/vehicle' recorded as a causal factor were lower at night than during the day for both young and mature drivers (Australian Fatals database only).

Responsibility for crash

Proportions of young drivers were higher than the corresponding proportions of mature drivers for crashes where:

• driver designated as responsible for the crash (Australian Fatals database only), both day and night

Proportions of 'responsible' drivers were higher at night than during the day for both young and mature drivers, but more so for young drivers.

Weather, dry/other road surface

Proportions of young drivers were lower than the corresponding proportions of mature drivers:

• when weather conditions were wet or non-clear - daytime only

• for young drivers only, proportions were higher at night than during the day

Proportions of both young and mature Australian drivers were higher at night than during the day when the road surface was wet; this pattern did not occur in the USA databases

Alcohol involvement

Proportions of young drivers were lower than the corresponding proportions of mature drivers for:

• drivers in 'alcohol-involved' crashes at night - USA data only; proportions of drivers in alcohol-involved crashes were higher at night than during the day, equally for both young and mature drivers

• driving under the influence at night (Australian Fatals only)

Drivers apportioned to crashes where 'driving under the influence' was a factor were much higher at night than during the day: a little more so for mature than young drivers

Proportions of drivers with a BAC above 0.05 were much higher at night, particularly in fatal crashes: a little more so for mature than young drivers

Proportions of drivers with zero BAC were much lower at night, equally for young and mature drivers (NSW data only)

Proportions of drivers involved in crashes where the highest BAC was above 0.05 were much higher at night, equally for young and mature drivers (NSW data only)

Legal aspects

Proportions of young drivers were higher than the corresponding proportions of mature drivers for crashes where:

• legal action taken against driver, or driver charged with violation, particularly during the day (NSW, USA data only).

Proportions of drivers involved in crashes where legal action taken against a driver were higher at night than during the day, particularly for mature drivers.

• driver charged with 'negligent driving' (NSW data only)

Proportions of non-licensed drivers (no USA data) were slightly but consistently higher at night, equally for young and mature drivers.

Proportions of provisional/probationary licensed drivers (no USA data - young drivers only) were a little higher at night.

Proportions of drivers charged with 'proscribed alcohol content' were higher at night, equally for young and mature drivers (NSW data only).

Seatbelt wearing

Proportions of young drivers were higher than the corresponding proportions of mature drivers for:

• driver not wearing a seatbelt - night-time (found only in Australian Fatals data (no USA data, trend not present in other two databases)

Proportions of drivers not wearing seatbelts were higher at night than during the day, particularly for young drivers and particularly in fatal crashes.

Vehicle driven

Proportions of young drivers were higher than the corresponding proportions of mature drivers for:

• cars aged 9 to 13 years (8 to 12 years, for Fatals and the USA databases)

• cars aged 14+ years (13+ years for Fatals and USA databases) - trend much stronger in the Australian data

Proportions of drivers with cars in the oldest category (14+ or 13+ years) were higher at night than during the day for both young and mature drivers.

Proportions of young drivers were lower than the corresponding proportions of mature drivers for:

• cars in the newest category (<] to 3 years - Australia, or <1 to 2 years Fatals and USA)

Proportions of drivers with cars in the newest category were lower at night than during the day, for both age groups but marginally more so for young drivers

Numbers of people/injuries/deaths in vehicle/crash

Data on these variables are recorded in a wide variety of forms, in different combinations for different databases. They include numbers of..

• vehicle occupants
• people involved in crash
• people injured/killed in vehicle
• people injured/killed in crash.

Not surprisingly, many differences were found between different age groups and between daytime and night-time. For example:

• proportions of drivers involved in daytime crashes with four or more people involved were higher for mature drivers than for young ones

• proportions of drivers involved in daytime crashes with two or more people injured did not vary with age group

• proportions of drivers involved in night-time crashes with two or more people injured were higher for young than mature drivers

• proportions of crashes with driver plus two or more passengers in the vehicle were higher for young drivers than mature ones, particularly at night.

Day/night differences in numbers killed and injured generally reflect the differences in types of crash, as reported above. The full set of such results was presented in the previous section and will not be discussed further, since a major determinant of the observed differences is likely to be the varying patterns of exposure of different age groups at different times of the day and night. In very general terms, young driver crashes involve more people, and result in higher levels of injury, than those of mature drivers.

Times of crash

As for numbers of vehicle occupants, people injured, etc, the temporal pattern of road crashes probably reflects to a large degree the different temporal patterns of exposure.

Proportions of young drivers were higher than the corresponding proportions of mature drivers for:

• weekend crashes
• during night-time darkness (night = 6 pm-6 am)
• for weekend crashes: noon-6 pm, and most strongly, midnight-6 am
• for weekday crashes: in the midnight-6 am time period

Proportion of drivers in weekend crashes was higher at night than during the day for both young and mature drivers, slightly more so for the young.

DISCUSSION AND CONCLUSION

The above findings are consistent with the patterns identified in an earlier review of published information on young driver crashes (Macdonald, 1993a). However, the present overview analysis provides much more information than was previously available on day/night differences in crash patterns. It has been found that such day/night differences are often large but, interestingly, they are in most cases the same or similar for young and mature drivers.

Underlying reasons for young drivers' generally higher crash risk remain unclear. A major factor influencing at least some types of over-involvement is likely to be the difference in exposure to risk of young versus mature drivers, including differing proportions of exposure during the day relative to night. Information on actual exposure patterns is needed to determine the extent to which this is reflected in the pattern of crashes. Based simply on crash patterns, it is difficult or impossible to differentiate the relative importance as causes of young driver over-involvement in various types of crash of:

• a greater relative proportion of young drivers on the road in the times and places where, independent of their skills and behaviour, crashes are most likely
• a higher-risk subset of young drivers on the road at some times and places
• most young drivers displaying 'riskier' behaviour at some times and places than at others
• most young drivers, under most conditions, displaying generally 'riskier' behaviour, whether due to lack of skill, lack of appropriate 'attitudes', or both.

One of the clearest features of the present overview analysis is that proportions of young drivers were higher than those of older drivers for all types of single vehicle crashes, most of which involved running off the road and/or hitting an object. They were more likely to crash on curves and slopes. Consistent with this pattern, young drivers were more likely to be travelling above the speed limit or at excessive speed, and more likely to be in a crash where 'loss of control' was identified as a causal factor; such circumstances were more likely at night.

Their higher incidence of single-vehicle crashes, often entailing a loss of control, may suggest inadequate vehicle control skills. Young drivers' less developed vehicle control skills are well documented in the driver performance literature (Macdonald, 1993b). However, another significant factor might be age-related differences in the way in which drivers perceive hazards and associated risks in relation to their own perceived capacity to cope with these hazards. Again, there is clear documentation in the literature that inexperienced drivers have poorer hazard perception abilities than drivers with greater experience. Perhaps young drivers are poorer than older drivers at perceiving and interpreting the available perceptual information concerning the curvature and gradient of the road ahead in terms of appropriate changes to vehicle speed and position on the road. Such questions cannot be answered by reference to information on crashes.

It is frequently suggested that young drivers' higher crash risk may be due in large part to a relatively small sub-set of 'young problem drivers' who, because of their particular combination of personal and socio-economic characteristics, are likely to expose themselves to particularly high risk levels and/or to drive in a particularly risky fashion. The 'Young Problem Driver' issue was briefly reviewed by Macdonald (1993b). There is a little evidence from the present study of relevance to this issue.

In the present analyses, proportions of unlicensed Australian drivers, and proportions of drivers charged with Proscribed Alcohol Content, were both found to be higher at night than during the day, but this was equally so for young and older driver groups. Similarly, proportions of crashes in which the car was in the oldest category were higher at night than during the day for both young and older drivers. These findings are suggestive of a somewhat different driver population at night. There is evidence of a correlation, at least within young driver groups, between level of exposure to risk, the 'riskiness' of driving performance itself, and personal characteristics such as socio-economic status (Macdonald, 1993b, pp.33-35); it could therefore be hypothesised that the apparent differences between daytime and night-time driver populations indicate a probable increase in the riskiness of driving at night. Clearly, more direct evidence on this issue is needed.

The basic questions which were the focus of the present study were:

• do young driver and mature driver crashes have the same set of causes?
• do daytime and night-time crashes have the same set of causes?

Results indicated some differences between young and mature driver crash patterns, and large differences between daytime and night-time patterns, some (but relatively few) of which were related to driver age. To pursue these questions in greater depth, there is a need for more information on the different patterns of exposure to risk of young versus mature drivers, during the day and at night. There is also a need for more information on the driving performance capacities and actual driving behaviour of drivers of different ages and different levels of skill development.