![Sample our Geography journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5937/TOC-Subject-Sample-2021-Geography.jpg"})
ABSTRACT
This study investigates the effect of local manufacturing on start-up activity across knowledge-intensive business services (KIBS) industries by relying on the natural experiment of German reunification. KIBS did not exist in the former socialist East Germany. The findings show that local manufacturing affected the co-location of new KIBS firms in East Germany while there was no such effect for West Germany, where the KIBS sector was already well established. The East German effect was fading over time when the KIBS sector became more established. The result suggests that strengthening the industrial base in peripheral regions could induce knowledge-intensive start-up activity.
ACKNOWLEDGEMENTS
The author thanks three anonymous referees and the editors for helpful suggestions.
DISCLOSURE STATEMENT
No potential conflict of interest was reported by the author.
Notes
1. A growing literature in economics identifies causal effects by relying on the natural experiment of German separation and reunification in a similar way as the present draft (e.g., Alesina & Fuchs-Schuendeln, Citation2007).
2. Mean comparison tests indicate that there are significant differences between East and West Germany for all independent variables (see Table A2 in Appendix A in the supplemental data online for summary statistics). This can certainly be traced back to the East German transition and the fact that KIBS industries were newly emerging in the former GDR.
3. The much larger planning regions, which are large functional economic regions, are not used for analysis; they might be too large to measure location attributes reasonably, as proximity to clients is important for KIBS. NUTS = Nomenclature of Territorial Units for Statistics.
4. During privatization, which peaked in 1991/92, new KIBS establishments stemming from outsourcing business services from manufacturing firms due to legal arrangements and political decisions cannot be disentangled from start-ups in the data.
5. Start-up rates are defined as the number of start-ups in KIBS industries over the local employment in the KIBS sector. Wyrwich (Citation2013) found similar results by using start-up data from a different data source.
6. The analysis resembles the result of an earlier analysis that only considered the 1990s (Wyrwich, Citation2014).
7. Interacting the share of highly skilled employees in R&D-intensive manufacturing with year indicators shows that significant effects for East Germany are only visible in 1994 and 1995. The results can be obtained from the author upon request.
8. The regional stock of knowledge is only weakly significantly related to new P-KIBS formation. This suggests that the influence of regional knowledge might be partly absorbed by the measures for high-quality manufacturing. Excluding the controls for high-quality manufacturing reveals a strikingly positive relationship between the regional stock of knowledge and new P-KIBS formation in the Poisson models as well. The results can be obtained from the author upon request.
9. There is also a positive relationship between the employment share and new P-KIBS formation when not controlling for the highly correlated number of KIBS employees. The results can be obtained from the author upon request.
10. Additional analysis of the density effect for every year reveals that the coefficient is relatively constant over time, as is the statistical power of the estimate. The results can be obtained from the author upon request.
11. This holds when separately introducing the highly correlated population density and the share of high-skilled employees, which proxies the regional stock of knowledge. The results can be obtained from the author upon request.
12. There is no firm explanation for this pattern. It might be that the prospects of starting a T-KIBS firm depend on the positive development of the economy to higher degree.
13. Such as for P-KIBS, there are specific-year effects for East Germany for the period 1994–96. In addition, there is a general positive effect for East and West Germany around the year 2000. The results can be obtained from the author upon request.
14. The results in and can be confirmed when restricting the analysis to East German NUTS-III regions in the period 1995–2000 (see Table A2 in Appendix A in the supplemental data online).