Is the spatial persistence of deprivation dependent on neighbouring areas?

Abstract In the context of poverty and deprivation alleviation by national and local government, we use census data to explore 50 year trajectories of area deprivation to provide insight into how the spatial context of a community influences these trajectories. Using the temporal and spatial framework of spatial Markov models, we provide evidence for England and Wales on whether changing levels of deprivation over time are conditioned by areas having different types of neighbour. We find that, independent of the nature of their neighbours, there is high persistence of Most Deprived communities remaining deprived. Moreover, after conditioning by the type of neighbour, there is little likelihood of a Most Deprived neighbourhood improving when its neighbours are also the Most Deprived. However, Most Deprived neighbourhoods with Less Deprived neighbours have a greater likelihood of improvement. Communities that are the Least Deprived and have neighbours that are mostly Least Deprived, most likely remain Least Deprived. In terms of policy implications, targeting Most Deprived areas that have mostly Least Deprived neighbours can be considered ‘quick wins’. It will also be resource efficient to target spatial clusters of Most Deprived communities rather than a similar number of isolated Most Deprived communities. This raises ethical questions around investing in some Most Deprived areas, but not in other, potentially more deprived, communities.