Neighbourhood income sorting and the effects of neighbourhood income mix on income : a holistic empirical exploration

An econometric model is specified in which an individual's income and the income mix of the neighbourhood in which the individual resides are endogenous, thus providing a holistic model of phenomena that previously have been fragmented into neighbourhood effects and neighbourhood selection literatures. To overcome the biases from selection and endogeneity, the parameters of this model are estimated using instrumental variables in a fixed-effect panel analysis employing annual data on 90 438 working-age males in Stockholm over the 1995-2006 period. Evidence is found of both neighbourhood effects and neighbourhood selection, but more importantly, it is found that the magnitudes of these effects are substantially altered when taking selection and endogeneity biases into account, compared with when only controlling for selection. When taking endogeneity into account, the apparent impact of neighbourhood income mix on individual income is magnified and the effect of individual income on the percentage of high income in the neighbourhood is magnified.