Abstract
An increasing body of theory suggests that some environments are more favourable than others for new firm foundation (NFF). The characteristics of favourable environments for production firm formation in the UK are isolated and then related to county-scale variations in new business registration rates over the 1980-88 period. Bivariate correlation analysis identified 24 independent variables which have a high statistical association with rates of NFF. In order to identify the combination of factors which best ‘explain’ variations in NFF, data for 16 independent variables found to be associated with rates of NFF at the 0.01 level of significance were standardized and made orthogonal by Principal Components Analysis. Three composite indices or independent variables were abstracted and labelled and the component scores for these independent variables were used in a stepwise multiple correlation and regression analysis. Results from the final model indicate that at a county level a wide range of factors ‘explain’ spatial variations in new business registration. It is inferred from the final model that a high rate of NFF is positively associated with areas having a ‘high level of self-employment, a tradition of smaller scale manufacturing and a political ethos fostering the role of the individual and self–help’ and ‘turbulent and volatile environments with a strong tradition of non-production entrepreneurship’, whilst negatively related to a ‘limited entrepreneurial and educational experience in depressed local environments where individuals have limited access to finance’. It is suggested from this model that no overriding factor can statistically explain the marked spatial variations in NFF during the ‘Thatcher’ years, 1980–88. A combination of factors best explain the spatial unevenness in this phenomenon. The residuals from the final stepwise multiple regression model suggest that targeted intervention by local development agencies has enabled more new businesses to be established in local environments than would have been expected based on the existing social and economic structure alone.

This publication has 7 references indexed in Scilit: