This study looks at the effects of participating in active labour market policy programs for persons with work disabilities. More in detail, it draws attention to the importance of taking into account the timing of when a work disability has been registered for persons attending ALMP programs. We argue that not controlling for time of events, i.e. the exact time a participant starts a program, as well as the time that person is being registered with a work disability, implies a risk of ending up with biased estimations with respect to the effectiveness of certain ALMP programs.
work disability ALMP programs endogenous selection post registration bias