Young Driver Crash Risk

RISK PER MILE DRIVEN

First and foremost, crash rates for teenagers are much higher than comparable rates for older drivers. This is particularly true for 16 year-olds, most of whom would be in their first year of licensed driving. The rate of serious driver crash involvements for 16 year-olds is estimated at 34.5 per million miles driven; 20.2 for 17 year-olds; and 13.8 for 18 year-olds. This compares with 7.8 for drivers in their 20s and 3.9 for drivers ages 30–69 (Williams, 2000).

ALL OR JUST A FEW?

Some of the earliest work addressing young drivers attempted to determine which teenagers were going to be involved in crashes and how they differed from other teenagers (Harrington, 1971; Kraus et al., 1970). The idea was that we could somehow intervene prior to licensure and make crash involvement less likely. These and other early studies indicated that: young drivers are at greater risk of crash involvement; we can identify at least some of the underlying problems; and we can predict future crash frequencies based on individual characteristics.

Our level of prediction is not, however, sufficiently high to warrant subjecting some young people, and not others, for mandated special treatments. There has not been a prelicense intervention, including selected (or universal) drivers education, which effectively deals with the teenage crash problem.

CRASH RATES FOR NEWLY LICENSED

Crash rates for learning drivers tend to be low (Williams et al., 1997). Crash rates rise dramatically as soon as unsupervised driving (i.e., licensure) is allowed.

In 1992, high school freshmen in four states were recruited to participate in a continuing teenage driving panel. They were surveyed by telephone every 6 months until they graduated high school in 1996. The primary interest was to trace their development, in real time, of driving, driving skills, and related driving exposure including crash involvements (Preusser et al., 1998).

Results indicated that the first few miles of driving during the first few months following licensure were by far the most dangerous. Specifically, the crash involvement rate for the first 250 miles driven following licensure was 3.2 (per 10,000 miles) versus 1.8 for the next 250 miles and 1.3 for the next. These results, shown separately for males and females, are depicted in Figure 1 (McCartt et al., 2003).

Figure 1.

Crash Rate by Cumulative Miles Driven after Licensure — Male Drivers and Female Drivers

Note that the first 250 miles of driving following licensure is not the same as the first 250 miles of driving. Most of these young people accumulated several miles as learners and, in fact, they typically “first drove on a public road” 1 year prior to their state’s minimum age for obtaining a learner’s permit. But, learning driving is relatively safe, whereas the first 250 miles of driving after they are licensed, that is, after they are allowed to drive unsupervised, is particularly dangerous.

Several other studies also have found that crash rates are extremely high immediately following licensure, then decline rapidly during the first few miles and/or first few months of driving. This effect has been found in Norway, Sweden, and Canada (Gregersen et al., 2000; LaBerge-Nadeau, 1997; Mayhew et al., in press; Sagberg, 1998). Declines for each subsequent year following licensure have been reported for young drivers in Michigan (Waller et al., 2001).

NIGHT DRIVING

A second characteristic of youth driving is that crash rates are particularly high at night. Williams and Preusser (1997) reported that 16- and 17-year-old drivers accumulate only 14 percent of their miles driven between 9 p.m. and 5:59 a.m., yet they experience 39 percent of their fatal crash involvements during this time period. Drivers of other ages also show higher fatal crash rates at night, but the crash rate differences are greater by far for young drivers.

Some jurisdictions restricted the night driving of teenagers long before night driving was addressed by GDL. Some of these restrictions were applied statewide (e.g., New York and Pennsylvania), and some were based on city curfew ordinances (e.g., Cleveland, Detroit, and Columbus). These restrictions have been shown to substantially reduce the crash involvement rate of young drivers (Preusser et al., 1984; Preusser et al., 1990). Night restrictions have the added advantage of allowing young drivers to accumulate driving experience during daylight while limiting their exposure to more hazardous night driving conditions.

The night driving problem may be associated with lack of visibility at night. Or, it may simply be a correlated or associated variable combining a host of other factors. Night driving more often is for recreation as opposed to driving to work or school. It is this recreational driving that is most affected by a night driving restriction. Young drivers involved in fatal crashes at night, as compared with young driver daylight fatal crash involvements, more often involve: only a single vehicle (53 versus 33 percent); young driver error (81 versus 76 percent); speeding (43 versus 30 percent); alcohol (34 versus 6 percent): and two or more passengers (36 versus 29 percent), most often teenage passengers.

PASSENGERS

One of the least understood factors associated with teenage crash involvement is passengers, more particularly the presence of multiple teenage passengers. We do know that the presence of teenage passengers leads to greatly elevated teenage driver crash risk. There is some evidence that these passengers serve as a distraction to focused risk aversive driving (Williams et al., 1998).

Preusser et al. (1998) demonstrated that the relative risk of fatal crash involvement for a 16 year-old driver traveling alone, versus a driver age 30–59 traveling alone, was 2.28. That is, the 16 year-old was about twice as likely to become involved in a fatal crash. Add passengers, and the relative risk of fatal crash involvement increases to 4.72. And, if these are multiple teenage passengers, 16-year-old drivers are 7.86 times more likely to become involved in a fatal crash than drivers ages 30–59 (see also Chen et al., 2000).

DISCUSSION

Young drivers have extreme crash rates during their first few months and first few miles of driving. Special restrictions need to be applied to this period. Night driving is particularly hazardous for teenagers as is transporting other teenagers as passengers. Efforts to reduce young driver crash involvements should focus on these three factors.

Notably absent from this list are prelicense education or other interventions and alcohol. There is little or no evidence that currently available prelicense actions reduce postlicense crash rates. Alcohol already has seen substantial progress with minimum age 21 drinking laws and zero tolerance laws for youth drinking and driving. Although more certainly can be done, it now is the case that teenage drinking and driving is less of a problem than drinking and driving of young persons older than 21.

Also absent from this list are seat belts. Belts are not currently a focus in young driver injury reduction efforts, though perhaps they should be. Young driver belt use can be increased through broad population efforts such as passing primary belt use laws and selective enforcement. There also is the possibility that high-school-based programs can increase belt use in much the same way that high-school-based programs have reduced drinking and driving (Leaf and Preusser, 1995).