Urban Studies

Regional policies may be assessed from at least three separate perspectives: their impact on the regional distribution of economic activity (their primary purpose), their aggregate efficiency effect on the economy (contribution to national output net of resource cost) and their financial impact on the national Treasury (net Exchequer costs). This paper reviews the considerable effort directed since the turn of the 1970s towards evaluation of these effects in Britain. So far as the impact of policy on the regional distribution of economic activity is concerned, the paper classifies studies according to quantitative methods used (simple statistical analysis, shift-share analysis and regression analysis) and emphasises the strengths of the latter over the others, showing that, without as yet being used entirely satisfactorily, models useful for policy assessment are developing. So far as aggregate policy effects (economic and financial) are concerned, there is more ground to be broken, although the outline of a basic model is emerging.

This paper appraises the potential of social area analysis - the identification and description of areas according to their social characteristics - for planning and policy-related research. The development of social area analysis is briefly described, from its roots in pre-war urban sociology into its modern form as a research method based on multivariate statistical techniques, notably principal components analysis and cluster analysis. A number of applications of this method in mainly British planning experience are then described, and the main strengths and weaknesses of social area analysis are considered in terms of the problems of interpreting the results, its accessibility for potential users and the utility of the method for typical planning research problems. A number of further applications are considered and the conclusion is drawn that social area analysis has limited but important uses for a wide variety of planning research exercises.

The manner in which density and crowding were measured in Melbourne, Australia, was found to be of crucial importance in explaining the observed relationships between these variables and human pathology measures. The more theoretically relevant measures of density and crowding displayed higher correlations with human pathology measures than less theoretically relevant measures. It was concluded that contradictory findings in the literature regarding the importance of density and crowding as determinants of human behaviour can be attributed in part to the failure of researchers to specify measures of density and crowding appropriate to the theoretical models they are testing.

The paper argues that criticisms of shift-share techniques are insufficiently strong to affect its application to the analysis of regional growth in the United Kingdom. Evidence for the 11 British regions for 1952-75 is used to demonstrate the robustness of the technique in the context of differing levels of sectoral disaggregation and choice of base years and the technique is then extended to incorporate multiplier and linkage effects. Finally its utility is compared with that of analysis of variance techniques for standardising for industrial structure.