As of July 2007, many restaurants in New York had to post caloric information for their food. From a study by Julie Downs et al in the American Economic Review 2009:
To test whether this information would affect food choices, we collected data at three locations (a coffee shop in Manhattan and two hamburger restaurant outlets of the same chain, one in Manhattan and the other in Brooklyn), both before and after implementation of the legislation. Researchers stood outside each restaurant during lunch hours. As customers approached, they were informed that they could get paid for turning in their receipt and completing a short survey when they exited. …
At the coffee shop, there was no impact of the legislation. … For the Manhattan hamburger restaurant, there was again no significant effect of the legislation …
At the hamburger restaurant in Brooklyn, however, fewer calories were consumed after the legislation went into effect …. Furthermore, at the Brooklyn location, in contrast to the sandwich study, there was a marginal interaction between dieting status and legislation…such that dieters tended to be helped more than nondieters by the information.
The sandwich study they refer to is an experimental trial (described in the same paper) in which caloric information was provided to some people, and the researchers found “some evidence of a perverse, calorie-increasing effect of providing this information to dieters.” (Apparently the dieters were overestimating the calories in the sandwiches, so giving the information made them eat more!)
Hat tip to Andrew Leigh’s summary piece on the economics of obesity.
David Evans' personal blog 