Estimating Labour Market Transitions and Continuations using Repeated Cross Sectional Data

This paper proposes a new approach for estimating transition (or continuation) probabilities using repeated cross sectional data. A cross sectional method is necessary when good panel data are not available. The proposed approach is uniquely designed for repeated cross-sectional analysis, and as a result, the identifying assumptions are relatively mild and easy to interpret. Canadian Labour Force data and U.S. Current Population Survey data are used to compare proposed and existing approaches. I find that the choice of method can make a difference at the inference stage.