May 22, 2013
This study utilizes a natural experiment—conditionally random dormitory assignments of first-year US college students—to investigate the influence of obesogenic environmental factors in explaining changes in weight and exercise behavior during the 2009–2010 academic year. The design addresses potential selection biases resulting from the likelihood that individuals sort into built environments that match their preferences for exercise and healthy eating. We find some evidence that the food environment, specifically access to campus dining, significantly affected the weight of female students in our study. Females assigned to dormitories where the nearest campus dining hall was closed on the weekends gained about 1 lb less over the course of the year than females assigned to dormitories near dining halls that were open 7 days a week. We also find some evidence that female who lived in close proximity to a grocery store gained less weight over the course of the year. Finally, females who lived closer to campus gym reported more frequent exercise over the course of the year. We do not find significant effects of the built environment on weight changes of males in our sample, but we are cautious to draw strong conclusions from this because the male weight change in our sample was quite small.