Using regression kink design to infer causal effects of unemployment benefits

Nathan Lachapelle

In this paper I exploit a kink in the unemployment benefits schedule to investigate whether higher unemployment benefits leads to lower transition rates to employment and to higher unemployment duration. Using rich administrative data from Belgium I estimate that a 1% increase in the unemployment benefits decreases the transition rate at 12 months by 0.8% and increases the (paid) unemployment duration at 12 months by 0.5%