Measuring Labor Market Transitions in Europe: Identication and Validation Analysis

We consider the problem of measuring transition probabilities across employment, unemployment and nonparticipation when longitudinal data is not available and/or the available retrospective data is measured with error. We establish nonparametric point-identification conditions from time series of cross sections, focusing on the European Union Labor Force Survey (EULFS) microdata released by Eurostat, and assess their validity using auxiliary panel data for Portugal and the United Kingdom. We find that the variables in the EULFS do not satisfy the identification conditions. Consequently, we propose alternative data-releasing solutions allowing users to measure transitions from EULFS data while satisfying the existing legal requirements.