In 2006, two researchers in Wisconsin noticed a weird pattern. A number of cities had recently banned smoking in public places, such as at bars and restaurants. The laws were amazingly effective at cutting down on the number of people using cigarettes. Simply making it harder to smoke meant that fewer people puffed away.

But, for some reason, drunk driving fatalities seemed to increase wherever smoking bans were put in place. What, the researchers wondered, was going on?

The researchers’ first hypothesis was that smokers were driving longer distances to find bars in other cities where it was legal to puff. Or, that smokers were driving to obscure parts of town to find bars that ignored the law, or had outdoor seating where smoking was allowed. This reflects a basic axiom: for the drunk driver, the length of your drive matters almost as much as how many drinks you’ve consumed. That’s why rural areas have so many drunk driving deaths. Intoxicated people who have to drive 30 or 40 minutes after leaving a bar are far more likely to fall asleep at the wheel than a drunk with a three minute commute. When the researchers looked at data from states that had recently implemented smoking bans, the data seemed to support their hypothesis. Bans = longer drives = more deaths. Case closed, they figured.

But a few years later, someone else noticed another strange pattern. By then, smoking bans Read More