Abstract
The potential for socioeconomic analysis by a comparison of 1981 and 1991 UK Censuses of Population data is very considerable. Spatial analysts have begun to exploit the new opportunities provided by the broadly comparable two sets of Small Area Statistics with the advantages of computerised access. However, the integration of 1981 and 1991 data is far from straightforward because of changes in census geography and definitions, in the administration of the census, and in the format and nature of the data. It is explained how, by a further refinement of population surface modelling concepts, these two data sets can be transformed into a single, consistent geographical database independent of the two original and inconsistent zonal structures. At the same time, the technique allows the effective management of the difficulties associated with ‘restricted’ and ‘special’ census zones. There is no sense in which such techniques can add to the accuracy (real or apparent) of the census data, but they are literally a re-presentation of the same data in a different form. The resulting models permit intercensal analysis with considerable spatial flexibility and a number of computational advantages. Illustration is made of the types of analysis that can be supported, and attention is drawn to the availability of the remodelled data to the academic community.

This publication has 6 references indexed in Scilit: