posted on 2009-03-19, 13:28authored byFabrizia Mealli, Stephen Pudney
We illustrate the difficulties raised by four features of realistic transition models in labour economics: dimensionality, institutional constraints, persistence and sample attrition. We estimate a multi-spell multi-state transition model using longitudinal data on the 1988 cohort of male school-leavers in North-West England. The model predicts transitions between college, the government Youth Training Scheme (YTS), employment and unemployment, allowing for endogenous sample attrition and persistent cross-correlated heterogeneity. We simulate the impact of YTS, allowing for endogenous YTS selection induced by heterogeneity. The main endings are a strong positive effect of YTS participation on employment prospects and a large negative impact of early drop-out from YTS.