The paper addresses two problems in the analysis of regional convergence in the European. First, typical convergence studies use data that is only available for some of the regions in the EU. This introduces a selection bias, as the lack of data is likely to be systematically linked to the...