A Longitudinal Mixed Logit Model for Estimation of Push and Pull Effects in Residential Location Choice

Fiona Steele, London School of Economics and Political Science (LSE)
Elizabeth Washbrook, University of Bristol
William Browne, University of Bristol
Christopher Charlton, University of Bristol

We consider households’ choice of neighbourhood using household panel data linked to neighbourhood data. A type of random effects multinomial logit model is developed to study neighbourhood choice over time, extending previous research that has been restricted to cross-sectional data. We show how longitudinal data on households’ residential location allow separation of the influence of neighbourhood characteristics on the decision to stay in the current area (‘push’ effects) and on the choice of destination among movers (‘pull’ effects). The model is applied in an analysis of residential choice in England, with a focus on how effects of area deprivation and distance from the current area on location choice at year t depend on characteristics at t-1 and life course transitions between t-1 and t. Household random effects are included to allow for unobserved heterogeneity between households in their propensity to move, and in the importance placed on deprivation and distance.