Understanding Neighbourhood Effects: Selection Bias and Residential Mobility

Abstract

This chapter argues that the most severe problem in the identification of causal neighbourhood effects is selection bias as a result of selective sorting into neighbourhoods. People sort themselves into and out of neighbourhoods and selection bias occurs when the selection mechanism into neighbourhoods is not independent from the outcome studied. However, many studies do not include controls in their models of neighbourhood effects for selection bias. As a result it is impossible to say whether correlations between neighbourhood characteristics and individual outcomes are causal effects, or the result of neighbourhood selection. The chapter argues that more knowledge is needed about residential mobility and the selective sorting into and out of neighbourhoods to better understand mechanisms behind neighbourhood effects. An illustration is provided using a case study for Stockholm, Sweden and a conceptual model is presented which links neighbourhood choices made by individuals and households with individual level outcomes. Both real causal effects and selection effects are featured in the model. The conclusion is that in order to further our understanding of neighbourhood effects we should incorporate neighbourhood sorting into our models of neighbourhood effects.