Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
We consider the problem of measuring transition probabilities when panel data is not available and retrospective data is measured with error. We establish point-identification conditions of transition probabilities and propose a method to estimate them using time series of cross sections. We illustrate the application of our results using the European Union Labor Force Survey.