Beginning in 1984, the Monitoring the Future surveys included questions on drinking and driving behavior. The drinking and driving questions were "during the last two weeks, how many times (if any) have you driven a car, truck, or motorcycle after drinking alcohol?" and " ... after having five or more drinks in a row?" The survey also asked about riding with a driver who had been drinking and about friends' attitudes regarding driving after drinking (O'Malley and Johnston, 1999).
Figure 25 presents these self-reported drinking and driving trends and compares them with the drinking trends from Figure 19. Both driving after drinking measures have decreased substantially since 1984: driving after drinking dropped 48 percent, from 31.2 percent in 1984 to 16.3 percent in 1998; driving after binge drinking dropped 45 percent, from 18.3 percent to 10.0 percent. Each measure decreased steadily until about 1995 and rose slightly since then. The 30-day drinking and binge drinking trends reached their lowest level slightly earlier, in about 1993.
Youth Drinking and Driving after Drinking
O'Malley and Johnston (1999) note that, for high school seniors, friends' disapproval of driving after drinking has risen at the same time. In 1984, 30 percent of high school seniors reported that their friends would strongly disapprove of their driving after having 1 or 2 drinks. By 1998, 48% reported this level of disapproval. As O'Malley and Johnston suggest, the substantial decrease in drinking and driving may have occurred because of a substantial decrease in the social acceptability of drinking and driving among young people.
Figure 26 compares the relative changes in 30-day drinking, driving after drinking in the past two weeks, and drinking driver fatal crash involvements, from a base of 1984 = 100 percent. Figure 27 does the same for binge drinking and driving after binge drinking. The two drinking and driving trends follow the fatal crash involvement trend very closely through about 1995. Since 1995, self-reported drinking and driving has increased while fatal crash involvements have not. The two drinking trends, on the other hand, decreased considerably less.
Youth 30-Day Drinking, Driving after Drinking,
and Drinking Drivers in Fatal Crashes
Youth Binge Drinking, Driving after Binge Drinking,
and Drinking Drivers in Fatal Crashes
Table 14 summarizes these Figures by comparing overall percentage decreases from 1984 to 1998. In the Table, drinking drivers in fatal crashes include all drivers under 21 years of age as recorded in FARS. The driving after drinking and the drinking data apply to high school seniors as reported in Monitoring the Future. All four drinking and driving measures have decreased more than twice as much as either of the drinking measures.
|Youth Drinking, Driving Involvement||Change, 1984 to 1998|
|Drinking drivers in fatal crashes||-61%|
|Drinking drivers in fatal crashes per population||-58%|
|Driving after drinking||-48%|
|Driving after binge drinking||-45%|
Hanson and Engs (1992) report evidence that drinking and driving has decreased more than drinking among college students. As noted previously, they found a slight decrease in annual drinking but no significant change in binge drinking between 1982 and 1991. They also report on 17 problems during the past year related to drinking, such as having a hangover, fighting after drinking, and missing a class. Four of the measures involved driving and 13 did not. Among students who drank at least once a year, there were statistically significant increases in 10 of the 13 non-driving measures and a significant decrease in only one. In contrast, there were significant and substantial decreases in three of the four driving measures. Table 15 summarizes these changes.
|Driven after several drinks||59%||43%||-27%|
|Driven after too much to drink||41%||32%||-21%|
|Driven while drinking||48%||32%||-32%|
|Arrested for driving while intoxicated||1%||1%|
Three other sources of evidence suggest that youth are separating their drinking from their driving. The first comes from a study comparing driver BAC levels and drinking histories. Roeper and Voas (1999) measured BAC levels at the roadside on Friday and Saturday nights in three communities from July 1992 to June 1996. They also recorded the drivers' drinking history over the previous 28 days. They found that drivers under 21 years of age were more frequent binge drinkers, but less frequent drinking drivers, than older drivers. More precisely, drivers under 21 were more likely to have consumed six or more drinks at a sitting during the previous month than older drivers. But the proportion of drivers under 21 with a positive BAC was less than half that of older drivers: 6.8 percent for drivers under 21 compared to 18.7 percent for age 21-34 and 14.7 percent for drivers 35 and older. When drivers were disaggregated by drinking level, measured as the maximum number of drinks on any one day in the previous month, the result was the same: at each drinking level, fewer drivers under 21 had a positive BAC than older drivers. Drivers under 21 drank more, but drove after drinking less, than older drivers.
Balmforth (1998) reports on a national telephone survey of drinking and driving issues conducted every two years since 1991 by NHTSA. Each year, about 3,300 adults are contacted in a nationally-representative telephone survey. Results are reported for the age groups 16-20, 21-29, 30-45, 46-64, and 65 and above.
From 1991 to 1997, Balmforth found that the number of youth who reported driving within two hours of drinking in the past year had remained about the same (10 percent in 1991, 13 percent in 1993, 9 percent in 1995, and 12 percent in 1997). However, the average number of times these drivers drove after drinking within the past month decreased from 1.6 in 1993 to 1.2 in 1997. The study's sample size for drivers age 16-20 was relatively small, so these changes may not be statistically significant. Consistent with Roeper and Voas, Balmforth found that youth aged 16-20 drink more per sitting but drink and drive less frequently than older persons. For example, in 1997 youth averaged 4.6 drinks on their most recent drinking-driving occasion, compared to 3.1 for drivers aged 21-29, but only 7 percent of youth drove after drinking in the past month, compared to 18 percent of drivers aged 21-29.
One method that can be used to separate drinking from driving is the use of a designated driver: a person who will drive those who have been drinking and who agrees (in principle) not to drink. DeJong and Winsten (1999) report on designated driver use by college students in 1993 from the same national survey used by Wechsler at al (1993). They found that, in the past 30 days, 33 percent of all college students reported having served as a designated driver and 37 percent of students who drink had ridden with a designated driver. Of drinkers, over half did not drink when they were the designated driver and only 2 percent had five or more drinks. Of students who usually binge drink, only 5 percent binged when they were the designated driver. Barr and MacKinnon (1998) report even higher designated driver use at a single college: 86 percent of students had ridden with a designated driver, 84 percent had been a designated driver, 37 percent "always" use a designated driver when they are drinking and another 28 percent do "most of the time." It's clear that the designated driver is a socially accepted practice on college campuses. While the designated driver may not abstain from drinking, it's also clear that he or she only rarely drinks to excess.
The evidence in this section on drinking and driving practices supports and provides a link between the fatal crash data and the information on drinking behavior.