What's in a Glass? Study Shows Wine Drinkers Fail to Gauge Serving Size

With an ever-growing arsenal of flutes, snifters, goblets, copitas and glasses tailored to every region and variety under the sun, you could be forgiven for losing track of just how much wine a glass of wine contains. A new study confirms you wouldn't be alone: Wine drinkers routinely pour different amounts of wine based on the glass. The study also identifies which situations put wine drinkers at risk of unintentionally "overpouring" a single serving.

For the study, published in the journal Substance Use & Misuse, Doug Walker and Laura Smarandescu of Iowa State University and Brian Wansink of Cornell University recorded 73 wine-drinking subjects pouring themselves a glass of wine. Or rather, 504 glasses of wine: The researchers wanted to test whether subjects poured more into a larger glass or a wider glass compared to a more "standard" glass; poured more red wine versus white; poured more while holding the glass or with the glass on a table; and poured differently with a small place setting, large place setting or no place setting at the table. They then quizzed the participants on whether these factors might have caused them to overpour.

The Alcohol, Tobacco, Tax and Trade Bureau (TTB) defines a single serving of wine as 5 fluid ounces. But can drinkers, even experienced ones, really gauge how much wine that represents? "I was looking at two glasses just yesterday, my glass of water and my glass of wine, and they all have different shapes, and it's very difficult," Smarandescu told Wine Spectator. "I was honestly thinking, how much more do I have in this glass versus this glass?"

Her team's research indicates she's not the only one. The average pour into a wide glass was 11.9 percent higher than into a standard one. Subjects holding the glass poured 12.2 percent more than when the glass was placed on a table. White wine pours were 9.2 percent more generous than red. And subjects correctly guessed which factors might skew their judgment. The presence of place settings and the "large" rather than "wide" glass did not significantly affect pour sizes. (Smarandescu hypothesized that if there were food on the table, place setting results might have been different.)

Citing earlier research, the study explains the psychology behind these cues. "When estimating volume, people tend to place a larger emphasis on container height than on its width," so wide glasses get inordinately larger pours. "When standing, it is easier to pour into a glass that is being held than in a glass that is placed on the table," and so participants poured more.

And then there's a phenomenon called Delboeuf’s Illusion that causes people to consume more food or drink if it has a lower color contrast with the container it is served in; because white wine is closer to transparent glass than red, subjects poured more.

In May, after a decade of pressure, the TTB issued a ruling that would allow producers of alcoholic beverages to state Serving Facts on their labels: size per serving, number of servings per container and alcohol content per serving, as well as some nutritional information. Small wineries fear that this voluntary measure, costly for small producers, will lead to a mandatory one. But proponents argue that drinkers have a poorer grasp of serving sizes than they think. The new study supports that argument.

"I think it would be helpful to inform people. It seems like there isn't enough knowledge of what is a standard serving, so I think people attempt to count their drinks to make sure they don't get into trouble," said Smarandescu. "But with things like wine pours it becomes more difficult. So I think it would help to know how many servings are in a bottle."

The study cites previous research that shows bartenders who are made aware of their overpouring still overpour, albeit less. The subjects of this study, though they recognized after the fact when they might have overpoured, still did so. "If participants were trying to be accurate in each pour, they were not, and also post hoc they proved to be aware of the fact that they were inaccurate," wrote the authors.

Were the participants still unable to gauge the correct pour size—or did they simply not care? Smarandescu felt it was the former: "I think people cannot tell."