In the US, the percentage of obese in the slimmest state is higher now than the percentage obese in the fattest state in 1991. Incredibly, obesity has been increasing even faster in many other countries. The literature suggests a number of reasons for this, principally increased wealth and cheaper calories. A number of recent papers, including this one, consider how to lower that rate by policy, in particular with “libertarian paternalism”. The authors of this recent AEJ:Applied run an experiment in a fast food restaurant where menus are altered such that a small cost (such as having to write down a sandwich name rather than check a box) is added for unhealthy menu items, or caloric information is added directly to the menu. They find that both inconvenience and better information significantly lower calorie consumption. The worry with such a paper, however, is the small sample size (300 per trial, in a nonrepresentative sample of the US with significant nonresponse) and the difficulty of measuring compensation due to Hawthorne (I may order the healthy sandwich when I know the lab guys are in the restaurant, then get an ice cream on the way home to “make up” for it). Nonetheless, any information on reducing obesity is good information – it is by far the biggest public health problem in the US today and does not appear in the data to have peaked.
http://pubs.aeaweb.org/doi/pdfplus/10.1257/app.2.2.164 (Gated; no publicly available copy appears to exist)
A Fine Theorem 