Neighborhood effects in mental health and health behavior: longitudinal analysis of social causation versus selective residential mobility

Differences in well-being between neighborhoods have long been recognized. These inequalities have been explained by neighborhood effects, in other words characteristics of the neighborhoods that affect well-being. However, the direction of causality in neighborhood effects is still unknown. This dissertation examines whether inequalities in well-being between neighborhoods are caused by neighborhood characteristics social causation, or by selective residential mobility social selection. The participants in these studies were from the ongoing prospective longitudinal population-based Young Finns study. The aims of this study were: 1) to examine whether neighborhood urbanicity and socioeconomic status affected health behaviors, depressive symptoms and source of social support, and 2) to examine how those variables affect residential mobility behavior. The results suggest that people in more urban and affluent neighborhoods are more interested in their health. However, simultaneously those people drink more alcohol, and also people in more urban areas smoke more. People in more rural areas received more social support from their family, whereas people in more urban areas received more social support from their friends. While part of the results were explained by social causation, most of the effects were attributed to social selection. People who received more social support from their friends were more likely to move and to move more frequently. Also, people with better health behaviors moved longer distances. None of the individual level variables were associated with selective residential mobility between municipalities. This dissertation addressed the problem of causality in neighborhood effects. The findings do give some support for social causation in neighborhood effects, but most of the effects are explained by social selection. However, the results also imply that health behaviors or social support do not affect selective residential mobility. Therefore, it is likely that some other individual level variables govern selective residential mobility and the forming of differences in well-being between neighborhoods.