Gallup's Latest Alcohol Preferences Survey: Noisy

The last couple years, Gallup has sent me a link to their annual alcohol preferences survey. This year they didn't and it fell to Alan to alert me to the latest numbers. One of the things that really lept off the page last year were the number of women who were apparently flocking to beer. I illustrated the numbers with this handy graph:


Women Citing Beer as Preferred Beverage

_____________________2009____2010__Change
All
__________________21%_____27% ___+6%
Under 49 years old
___25%_____35% ___+10%
Over 50 years old
____15%_____18% ___+3%

Great, right? Well, this year's numbers show a substantial reversal--still positive over '09's numbers, but only modestly so.

Women Citing Beer as Preferred Beverage

_____________________2009____2010___2011
All
__________________21%_____27% ___22%
Under 49 years old
___25%_____35% ___28%
Over 50 years old
____15%_____18% ___17%
The numbers are worse for beer across the map. Beer was favored by 41% of respondents last year and only 36% this year (wine gained 3% to 35% and liquor 2% to 23%)--tied for the lowest ever measured. And it was down particularly badly among young people, who cited it as their favorite alcohol in just 39% of cases (down from 51%).

Of course, it's pretty clear we're just seeing poll "noise." These surveys have a margin of error of ±4%. When you look at the graphs, you see a lot of zig-zagging year-to-year. It's hard to think that such large populations are swinging in preference one year to the next. Rather, there are trends that become exaggerated by single data points.


If you consider the averages, things appear a little different. Over the past ten years, beer is cited as the favorite by an average of 41% of respondents (high of 46, low of 39). That is off the average from the previous decade's average of 45% (high 47%, low 42%). Wine has picked up some the slack, cited by 33% of people over the past decade, up from 30% the decade before. The remainder goes to liquor.

So if we go back to the number for women, I think we could assume the same phenomenon. (Unfortunately, we don't have numbers going back two decades, so we can't see the trendline.) Probably things are moving in the right direction, but far more modestly than it appeared after last year's survey.