Designated drivers aren't always the teetotalers one might hope them to be, a study suggests. Researchers tested 1,071 subjects exiting bars in a Florida city on a series of Friday nights over a three-month period. 165 of the subjects called themselves designated drivers. Of those, 35% had had something to drink: 17% of them had blood alcohol content between 0.02% and 0.049%, and another 18% had BAC of at least 0.05% (the NTSB's newly desired drunk-driving limit).

Subjects' average age was 28 and they weren't a diverse group—most were white male college students—limiting the scope of the study, the New York Times notes. Still, "when you look at evaluations of designated driver campaigns, they're really ineffective," says the head researcher, who observes that too often, the chosen designated driver is simply the person who has had the fewest—but not zero—drinks.