Presentation of entrepreneurship data and aspects of spatial modeling

Abstract

Positive effects of new firms on the job market, technology transfer, and contributions to structural change has turned political attention to start-ups. Every year the number of new firm creations increases, where on the other hand the number of major enterprises decreases.4 More and more firms have no employees and a trend to small-scale self employment can be recognized.5 Political and economic support programs try to revoke regional discrepancies of business activity, firm development and foundation activity. These programs have to be evaluated and improved continuously. An meaningful curatorial foundation statistic for entrepreneurship research and a statistic to assess the entrepreneurial activity in Austria for political decision making does not exist until now.6 Solely the Federal Economic Chamber of Austria (WKO) reports a statistic of foundation activity of commercial firms every year.7 This statistic permits to observe a trend of firm foundations, 4 Cp. Wirtschaftskammer Osterreich [16]: 23. 5 Cp. Schwarz and Grieshuber [11]: 103ff. 6 For the statistical situation in Germany see e.g. Fritsch et al. [3]: 2f. For the effort to the development of the curatorial statistical system in Germany cp. Struck [14]: 41ff. 7 The Federal Economic Chamber of Austria is the legal representation of interests of Austrian entrepreneurs. In its founding statistic the number of new start-ups are calculated from new entrants into the membership database of the WKO. To exclude pseudo foundations and multiple data set entries the database has been revised. A detailed description of data revision can be found in Wirtschaftskammer Osterreich [16]. but does not map the overall Austrian foundation activity. Firms which are not in the scope of the WKO are not included in this statistic.8 Few other Austrian public services cumulate data from newly founded firms, but the access to these data sources is limited and the data is not appropriate for research.9 To ensure a continuous and complete evaluation of supporting programs it would be wise to design a monitoring system for all Austrian enterprises which includes all commercial and noncommercial firm foundations and closings. Such a system would be a valuable source for entrepreneurship research, but suitable statistical measures and methods for presenting and modeling non-normal spatial entrepreneurship data are needed to build such an overall monitoring system. This paper briefly reports measures and methods commonly used in descriptive statistics for presenting entrepreneurship data in reference to its spatial distribution. We illustrate these by an example of numbers of new start-ups in Austria in the year 2001.10 By applying different measures we show how sensible presenting regional differences in foundation activity can be. Further we will give a brief introduction of spatial general linear models [5] and the hierarchical Bayesian models for count data [15] for modeling non-normal spatial data. © 2009 Springer Berlin Heidelberg.