Cross-sectional dependence and regional labor market dynamics (doctoral thesis)

Research output: Book/ReportBookAcademic


One of the most challenging and fascinating issues faced when analyzing spatial data is the often complex dependencies present between units. This thesis applies and contributes to developments in methods that allow for and explain these dependencies. In particular, a dynamic spatial panel data approach is taken to examine regional labor market dynamics in the European Union. A major strength of this approach is that it provides a useful tool to quantify the magnitude and significance of spillover effects. It is shown that a region-specific demand shock not only beneficially impacts the own region’s labor market, lowering unemployment and raising participation rates, but that the shock is also transmitted to neighboring regions. Furthermore, it is investigated in more detail the causes of variation in participation rates across the EU regions, which has become ever more pertinent due to the prevalence of discouragement effects in the wake of the recent crisis. The results reveal strong evidence for local spillovers, but only weak evidence for global spillovers in explaining participation rates. In addition, a simultaneous modeling approach accounting for key stylized facts observed in regional unemployment rates is proposed which integrates serial dynamics, weak and strong cross-sectional dependence, which generalizes previous sequential approaches that can lead to biased results. Finally, in a contribution to recent critiques on spatial econometric methods, fundamental questions are raised on estimating spillovers and recommendations are offered to introduce more flexibility.
Original languageEnglish
Place of PublicationGroningen
PublisherUniversity of Groningen
Number of pages181
ISBN (Electronic)9789036784191
ISBN (Print)9789036784207
Publication statusPublished - 2016

Fingerprint Dive into the research topics of 'Cross-sectional dependence and regional labor market dynamics (doctoral thesis)'. Together they form a unique fingerprint.

Cite this