Policy discontinuity and duration outcomes
Causal eﬀects of a policy change on hazard rates of a duration outcome variable are not identiﬁed from a comparison of spells before and after the policy change, if there is unobserved heterogeneity in the eﬀects and no model structure is imposed. We develop a discontinuity approach that overcomes this by considering spells that include the moment of the policy change and by exploiting variation in the moment at which diﬀerent cohorts are exposed to the policy change. We prove identiﬁcation of average treatment eﬀects on hazard rates without model structure. We estimate these eﬀects by kernel hazard regression. We use the introduction of the NDYP program for young unemployed individuals in the UK to estimate average program participation eﬀects on the exit rate to work as well as anticipation eﬀects.