Influence of Calorie Labeling on Food Choice: Evidence from NYC

Research Objective: Obesity is a major public health problem, yet few true population-level solutions have been implemented. The first such policy meant to impact obesity at a large scale is the mandatory labeling of the caloric content of menu items in fast food restaurants. The legislation has been passed or considered in over 25 cities and states, yet we have no evidence of its effectiveness. New York City was the first location to implement mandatory calorie labeling in the summer of 2008. The objective of our study is to examine the influence of calorie labeling on food choice in low-income communities.

Study Design: We utilized this natural experiment via a difference-in-difference design to study the influence of labeling in New York City, utilizing Newark, NJ as a control group. We collected 1,100 receipts from adults outside of fast food restaurants both before and approximately 1-month after labeling was introduced. When entering the restaurants, subjects were approached by study staff and asked to bring us their receipt and answer a short survey upon leaving. The methodology is similar to a street intercept survey. Calories purchased were calculated using the information on the receipt and each restaurants corporate website.

Population Studied: Restaurants from the largest fast food chains in low-income, largely racial and ethnic minority areas in NYC were matched to those similar areas in Newark. Subjects were sampled outside of these restaurants.

Principal Findings: We did not observe a change in the number of calories purchased from the pre- to post-labeling period in NYC or Newark. However, the percentage of people that noticed and reported utilizing the information increased to approximately 15% in the NYC, post-labeling period. We will also report on why the information might not have been valuable to consumers more generally.

Conclusion: While these results do not imply that provision of caloric information ineffective in the long term, it does show the need for greater research before labeling is introduced on a wide scale. The provision of information alone to change behavior could have limits that must be explored. The findings from behavioral economics could be helpful.