Temporal dimensions and measurement of neighbourhood effects

We conduct a panel analysis quantifying the degree to which the mixture of low-income, middle-income, and high-income males in the neighbourhood affects the subsequent labour income of individuals, and test the degree to which these effects vary by timing (lagging up to three years), duration (one to four years), and cumulative amount of exposure and to what extent these effects are persistent. We employ a fixed-effects model to reduce the potential bias arising from unmeasured individual characteristics leading to neighbourhood selection. The empirical study applies individual-level data for the working-age population of the three largest cities in Sweden covering the period 1991 - 2006. The analyses suggest that there are important temporal dimensions in the statistical effect of neighbourhood income mix: recent, continued, or cumulative exposure yields stronger associations than lagged, temporary ones, and there is a distinct time decay (though some persistence) in the potential effects after exposure ceases, though with some gender differences.