Publisher's Synopsis
This work investigates a methodology for small area social change which distinguishes between localised and general processes and specifically models the temporal dependence underlying change. Spatial persistence over time is modelled using a conditional regression model relating later to earlier values of social indices to obtain a genuine measure of change independent of initial values. Thus the inconsistencies between methods which are not base free can be resolved. This methodology is applied to intercensal changes (1971 to 1981) in socio-economic, age and ethnic structure for London electoral wards. The ability to determine the relative importance of structural and localised change allows an assessment of the extent to which processes such as gentrification are London wide or primarily local. The investigation is extended to recent trends in unemployment and mortality with the intention of assessing whether changes in spatial differentiation point to a widespread increase in inequality and how far changes in the deprivation rank order of wards reduce spatial persistence over time.
