GALLUP NEWS SERVICE
PRINCETON, NJ -- Alcohol is a double-edged sword for society. Although the moderate enjoyment of alcohol may have health and lifestyle benefits for drinkers, the effects of alcoholism and drunken driving can be devastating. Gallup asks two questions that provide rough gauges of the negative effects of alcohol on Americans. Both were recently updated in Gallup's annual Consumption Habits survey, conducted July 7-10, 2005.
Gallup finds that 21% of drinkers -- equivalent to 13% of all adult Americans -- admit that they sometimes drink too much. This is down from the early 1990s, particularly among young adults.
From 1992 to 1994, nearly half of young adults (aged 18 to 29) said they occasionally drank too much, making them the most binge-prone age group. However, in the last two years, that figure has averaged just 30%, comparable to the percentage of middle-aged Americans who say they sometimes drink too much.
Three in 10 Americans report that alcohol has been the source of trouble within their own families. This figure is higher than it was a decade ago, and significantly higher than in the 1970s.
One in Five Occasionally Overdoes It
The percentage saying they sometimes drink too much averaged 30% between 1985 and 1994, but has averaged 23% since.
Although this long-term decrease in reported occasional overindulgence is encouraging, the fact that it has not decreased further since 1996 could explain why the rate of drunken-driving fatalities has also been at a stalemate for the past decade. Government statistics show a steep decline in alcohol-related highway deaths between 1982 and 1994. This corresponds with the emergence of anti-drunken-driving campaigns such as Mothers Against Drunk Driving (MADD), which spawned stiff new drunken-driving penalties nationwide. But since 1994, the number of alcohol-related highway deaths has changed little.
Three in 10 Perceive Drinking as a Cause of Trouble in Family
A different picture emerges from Gallup's trend on drinking-related family problems. The percentage having this experience inched up from the 12% to 15% range recorded in 1947 and for much of the 1970s. By 1981, this figure was typically over 20%, and by 1997, it was typically over 30%.
Does this represent a true increase in the effect of problem drinking on families, or merely an increased willingness to admit that such problems exist? With relatively small demographic differences in reported experience with problem drinking, the answer is not clear.
Fewer Young Adults Drinking to Excess (or So They Say)
Drinking too much on occasion is moderately correlated with age. Gallup's data from 2004-2005* show that those under age 50 are about twice as likely as those aged 50 and older to say they sometimes drink too much. The current rate is 30% among those aged 18 to 29 and 29% of those 30 to 49, but only 14% among those 50 and older.
This pattern was quite different a decade ago. From 1992 to 1994, nearly half of young adults said they occasionally drank too much. By contrast, a comparison of 1992-1994 and 2004-2005 data finds little or no change in the percentage of middle-aged and older Americans who say they sometimes drink too much.
Shrinking Gender Gap
Gallup's aggregated trend data show that, over the past two years, men have been slightly more likely than women to say they sometimes drink too much, 27% vs. 20%. But this gender gap is narrower than it was in the early 1990s when 35% of men, compared with 22% of women, sometimes drank more than they should.
Apart from age, another major predictor of overindulgence is the type of alcohol one tends to drink. Those who say they most often drink beer are somewhat more likely than those who prefer liquor to occasionally drink too much. Wine drinkers are the least likely to report occasionally exceeding their limits.
The percentage of beer drinkers who sometimes drink too much has declined from 40% in 1992-1994 to 29% in 2004-2005. This mostly results from the reported decrease in drinking too much among young adults, currently the only age group that names beer as its most preferred alcoholic drink.
Few Americans Consider Alcohol Beneficial to Health
Some medical studies in recent years have suggested potential health benefits associated with moderate drinking. However, the Gallup measures on this indicate that most Americans don't embrace this finding. Americans are nearly as likely to believe that moderate drinking is bad for one's health (22%) as believe it is good for one's health (25%), while the largest segment believes it makes no difference (51%).
Since this question was first asked in 2001, there has only been a slight change in public perceptions, with a minor shift toward the view that alcohol is beneficial.
Perceived Health Benefits of Moderate Drinking
A common clinical definition for moderate drinking is one drink per day for non-pregnant women and two drinks per day for men. Because Gallup's question defines drinking in moderation as "one or two drinks a day," some women could perceive this as being over the suggested daily allowance. The net perception among women is that drinking is bad for one's health (27% say bad vs. 19% say good), while the net perception among men is that drinking is good for one's health (31% vs. 17%). However, even among men, the prevailing view is that drinking in moderation makes no difference to one's health.
These results are based on telephone interviews with a randomly selected national sample of 1,006 adults, aged 18 and older, conducted July 7-10, 2005. For results based on this sample, one can say with 95% confidence that the maximum error attributable to sampling and other random effects is ±3 percentage points. For results based on the sample of 658 adults who drink alcoholic beverages, the maximum margin of sampling error is ±4 percentage points.
In addition to sampling error, question wording and practical difficulties in conducting surveys can introduce error or bias into the findings of public opinion polls.
*Note: Demographic trends are reported on the basis of two-year averages in order to increase sample size, and therefore, statistical reliability.