43rd Congress of the European Regional Science Association: "Peripheries, Centres, and Spatial Development in the New Europe", 27th - 30th August 2003, Jyväskylä, Finland
Differences in the spatial patters of the intensity of new firm formation have attracted the intrest of researchers for a long time. Usally birth-rates or sometimes count-data is used to explain the spatial pattern with a variety of independend variables. Starting from a Shift-Share-analysis we examine the regional share of the number of newly founded firms in 74 West-German planning regions between 1987 and 1997. The regional share has the advantage that effects of different regional industry-structures as well as different size-structures are excluded. Therefore, by analysing the regional share, the factors determing the number of newly founded firms apart from industry-structure and regional size can be examined very clearly. The structure of the data allows us to apply a panel-regression appraoch with fixed effects. There are four main results of our estimations. First agglomeration-effects are of great importance for the number of newly founded firms. Regions with a high density of population provide a better environment for entrepreneurs. Furthermore, agglomeration-effects shape the correlation between the regional share and other regional characteristics, too. Usually a high share of employees working in SMEs and especially a high proportion of R&D in SME are rated as signs for the existence of a ”seed-bed” for entrepreneurs. But in our analysis these variables show a negative influence on the number of newly founded firms. Our conclusion is that this is because these variables indicate missing agglomeration-effects. SMEs are typical for rural and peripheral locations. Secondly, there is a tendency that the properties that are favourable for starting new businesses are less favourable for the survival of newly founded firms. We are the opinion that this is at least partially the result of a high degree of competition between newly founded firms. Thirdly, high rates of unemployment and growing numbers of unemployed result in growing numbers of newly founded firms. Because of missing alternatives on the labour market, people tend to start their own businesses more easily than in regions with a lesser degree of unemployment. Fourthly, we found a high degree of spatial autocorrelation. ”Neighbouring regions have much in common” that result in similar values of the regional share. But it could be shown that this is not due to common factors that are missing in our analysis.