Modelling spatial variation in the determinants of neighbourhood family migration in England with Geographically Weighted Regression

There have been many studies which have modelled internal migration in the UK. However, most of these have used data at geographical scales that conceal the majority of migration flows between neighbourhoods. They have also tended to use Ordinary Least Squares (OLS) regression or spatial interaction models. The latter are computationally unfeasible for migration flows between a large number of neighbourhoods. This paper uses a spatial modelling technique called Geographically Weighted Regression (GWR) to model family out-migration from neighbourhoods in England. GWR can take account of the spatial variation in the relationship between migration and its associated factors which are not accounted for using OLS. The variables included in the model are derived from theory and empirical research and include housing, socioeconomic, and environmental factors. The results show that the proportion of private renting, terraced housing, worklessness and non-domestic building land space in a neighbourhood each affect out-migration at varying levels across the country. For example, the effect of worklessness on out-migration is much stronger in neighbourhoods in the South East than the North of England. Therefore, all other things held constant, a successful intervention to reduce worklessness, initiated to discourage out-migration, would have a greater effect on out-migration in neighbourhoods in the South East compared with neighbourhoods in the North. © 2013 Springer Science+Business Media Dordrecht.