In one of the most recent experiments to test the effects of including calorie information on restaurant menus, Ellison et al found that both menus with calories listed numerically and menus listing calories in the context of the traffic light system, led to fewer (entree) calories purchased.  
    For their study, they manipulated menus in one restaurant for about two weeks. The restaurant was divided into three sections and as people came in to be seated, they were randomly placed into one of these three sections.  Everything was the same except the wording on the menus.  In one section of the restaurant the menu was as usual, in another the menu had the number of calories written next to each item and in the third, the calorie number was accompanied with a color (remember Go Slow Whoa).  Entrees with 400 or less calories were coded green,  with 401 to 800 calories were coded amber and with more than 800, red.  They used smaller calorie ranges to code the side items .
   Since the customers were randomly placed in the sections, each section should have contained people who were otherwise similar (any difference in the people is merely by chance).  This type of design makes it easier for one to say that what they did (put calories on a menu) caused the outcome (purchase less calories).  In fact, that is what they found. The customers who had either calorie menu purchased less calories from entrees than those that had no information.  
   The researchers made a good point in their writing.  I had not thought of this, but which makes perfect sense.  It applies to people like me.  People like me are going to order the low calorie item no matter which section of the restaurant they sit in.  People like me go in with the intention of choosing the lower calorie item and generally know what the low calorie items are.  But this is very good news, the menu information helps the people who need it the most!  The people who were not "health conscious" chose lower calorie items when they were in the restaurant sections with calorie counts and not when they were in the section with regular menus. (a side note - if all the people in the restaurant were like me, then the researcher would have found that the menus made no difference.  That finding would have been misleading, right!?!)

Ellison, B., Lusk, J., & Davis, D. (2013). Looking at the label and beyond: the effects of calorie labels, health consciousness, and demographics on caloric intake in restaurants. International Journal of Behavioral Nutrition and Physical Activity, 10(1), 21.