ABSTRACT
This article empirically analyses the link between market potential and regional wages in the enlarged EU. We contribute to the existing literature in several ways: (1) we analyse the link between market potential and wages for the EU27 and (2) deconstruct total market potential into several geographical components and analyse their respective contributions to explaining the geographical wage structure. We correct for existing spatial autocorrelation and endogeneity by using an instrumental variable generalized spatial two-stage least squares (IV GS2SLS).
Acknowledgements
We would like to thank M. Pfaffermayr, M. Pflüger, our colleagues at the BCL and ECB, our discussant T. Berglund and the participants at the SNEE conference in Mölle, seminar participants at Clark University and the University of Luxembourg. The views expressed in this article are personal and may not necessarily reflect those of the respective institutions the authors are affiliated with.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1 ‘First nature’ geography examines the physical characteristics, climate and natural resources of an area.
2 For a review of NEG and the empirics with regard to agglomeration and trade, see Overman, Redding, and Venables (Citation2003), Head and Mayer (Citation2004) or Redding (Citation2009).
3 Please see Appendix for variable definitions.
4 Moran’s I results are available upon request.
5 In our data, the overdispersion parameter α ranges between 0.41 and 0.54 and is always statistically significant.
6 Island regions are not included in the following regressions as binary contiguity spatial weight matrices cannot be calculated for regions without contiguous neighbours.