Can the Built Environment Reduce Obesity? The Impact of Residential Sprawl and Neighborhood Parks on Obesity and Physical Activity

This paper investigates the impact of the built environment on obesity while looking for causation in the presence of unobserved individual characteristics. Unobserved heterogeneity is addressed with instrumental variables. Two features of the built environment are considered: residential sprawl and proximity to neighborhood parks. Standard single equation results indicate that less sprawl and better park access are both significantly associated with lower levels of obesity. However, these effects disappear in the two-stage specification, implying that the built environment may not causally affect obesity. This conclusion is further supported by the finding that neither parks nor sprawl seems to influence the occurrence of physical activity.