Drinking behavior by youth also has changed substantially since 1982. The most thorough and consistent data come from the Monitoring the Future surveys of high school seniors, conducted annually since 1975 (Johnston, O'Malley, and Bachman (1999)). As part of an extensive investigation of drug use, these surveys ask several questions about alcohol ranging from lifetime use (81.4 percent in 1998) to "been drunk daily" (1.5 percent in 1998).
The measures that appear most relevant to drinking and driving are "use within the past 30 days" and "5 or more drinks in a row within the past two weeks." These will be called "30-day drinking" and "binge drinking," respectively. In 1998, about half of all high school seniors reported 30-day drinking (52.0 percent) and about one-third reported binge drinking (31.5 percent).
Figure19 shows the trends since 1982 in these two drinking measures. The question used to measure 30-day drinking changed in 1994. The 1994 survey included both old and new questions, and the response to the new question was 2.4 percentage points lower than to the old. To account for this change, the 30-day drinking data for each year after 1994 have been adjusted by adding 2.4 percentage points.
30-Day and Binge Drinking by High School Seniors
The 30-day drinking and binge drinking trends are similar: a steady decrease from 1982 until about 1993, then a slight increase. The trends appear generally similar to the trend in drinking driver fatal crash involvements shown in Figures 1 and 7. To compare these trends more accurately, Figures 20 and 21 plot the changes in each from a base of 1982 = 100 percent.
High School Seniors' 30-Day Drinking and Youth Drinking
Drivers in Fatal Crashes; Change from 1982
High School Seniors' Binge Drinking and Youth Drinking
Drivers in Fatal Crashes, Change from 1982
From these Figures, the similarities and differences are apparent. Both 30-day and binge drinking decreased 22 percent from 1982 to 1998, while drinking driver fatal crash involvements decreased 61 percent. Both 30-day and binge drinking increased gradually since 1993, while drinking driver fatal crash involvements have not.
While Monitoring the Future data are not available for each state, they are published for four regional groupings of the 48 contiguous states: North East (7 states: New York, New Jersey, Pennsylvania, and New England), North Central (12 states from Ohio west to the Dakotas, Nebraska, and Kansas), West (11 states: Montana, Wyoming, Colorado, New Mexico west to the Pacific) and South (the remaining 16 states and the District of Columbia, from the Atlantic west to Oklahoma and Texas).
Figures 22 and 23 show regional trends in 30-day and binge drinking, respectively. Trends for individual regions are difficult to distinguish in these figures. In general, drinking decreased similarly in all regions; slightly more in the North East and North Central, but these two regions had higher drinking levels in 1982 than did the South and West. By 1998, drinking levels in all four regions were very similar. Thus the Monitoring the Future data suggest that regional differences in drinking by high school seniors were relatively small in 1982 and have become smaller - drinking behavior by high school seniors is now very similar all across the country.
Youth Drinking by Region
Youth Binge Drinking by Region
Three studies provide more fragmentary evidence on drinking changes for youth aged 18 and above since 1982.
Hanson and Engs (1992) surveyed approximately 4,000 students at 65 four-year colleges in 1982, 1985, 1988, and 1991. Between 1982 and 1991 they found that annual drinking had decreased slightly but significantly, from 82 percent to 78 percent, while binge drinking (more than five drinks at one sitting at least one a week) had increased slightly but non-significantly, from 24 percent to 27 percent.
Caetano and Clark (1997) surveyed about 4,500 persons aged 18 and older in 1984 and 1995. They report drinking behavior in five classes: abstain, infrequent (less than once a month), less frequent (one to three times a month), frequent (weekly), and frequent heavy (five or more drinks weekly). Caetano and Clark group all persons aged 18-29, so youth under 21 cannot be examined separately. However, they note that:
"We examined trends in frequent heavy drinking for 18-20 year olds and 21-29 year olds, separately. Our data showed that the patterns of frequent heavy drinking within ethnic and gender categories did not differ for the two age groups and were accurately reflected in the trends for the 18-29 year old age category."
Caetano and Clark found that drinking at least weekly had decreased by about one-quarter for white and black men and by almost one-half for Hispanic men; by almost one-third for white and black women but had risen very slightly for Hispanic women (from a low level).
Midanik and Clark (1994) reported drinking changes from 1984 to 1990, again for persons aged 18-29. The number of persons reporting drinking at least weekly dropped from 40.1 percent to 32.2 percent, a decrease of about 20 percent; the number reporting weekly binge drinking (five or more drinks weekly) decreased from 10.3 percent to 7.0 percent, a decrease of about one-third.
The evidence from these three studies is not completely consistent. However, taken together, they suggest drinking decreases among persons aged 18-29 that are generally consistent with the 22 percent drop in both 30-day and binge drinking recorded for high school seniors by the Monitoring the Future surveys and are considerably smaller than the 58 percent decrease in fatal crash involvements of young drinking drivers.
Beer is generally regarded as the "beverage of choice" among young adults. Figure 24 shows the trend in U.S. per capita beer consumption over the years 1982-1997. The figure indicates there has been a generally downward trend over this period with per capita consumption in 1997 approximately 10 percent lower than in 1982. Total per capita data, of course, do not isolate the trend for persons under the age of 21. Nevertheless, they support the perspective that alcohol consumption has decreased in the entire population including youth.
U.S. Per Capita Beer Consumption
While youth drinking has decreased, it certainly hasn't disappeared. Table 13 presents results from five sources on recent drinking levels. Drinking level definitions are approximately similar across the five studies.
|Monitoring the Future||1998||High school seniors||74%||52%||32%|
|Wechsler et al||1993||College||84%||44%|
|Hanson and Engs||1991||College||78%||27%|
|Caetano and Clark||1995||18-29||55-68%*||16-18%*|
|Midanik and Clark||1990||18-29||73%||7%|
These studies suggest that the majority of youth drink at least monthly and 20-40% regularly drink at least five drinks in one sitting.
The data presented in this section suggest several conclusions.