Unemployment Benefits and the Timing of Redundancies: Evidence from Bunching
Résumé
Most of the empirical literature related to unemployment insurance (UI) has focused on its impact on outflows from unemployment rather than on inflows. In this paper, I show that workers respond to the design of UI while being employed. I exploit a discontinuity in the level of UI benefits at a particular value of tenure at current job. Using French administrative unemployment data, I analyse the concentration of workers in the tenure distribution at the relevant notch, a phenomenon known as bunching. The bunching mass is used to compute an elasticity of employment spell duration with respect to unemployment benefits.
I find an estimate equal to 0.014 in my preferred specification, translating into a 0.5 day of extension for a 10% increase in the replacement rate. This estimate measures strategic behaviours attenuated by optimisation frictions. I identify the underlying mechanism as bargaining between employers and employees who maximise their joint surplus thanks to a state transfer. I find that the elasticity is the highest in the population facing the strongest incentives and in the highest occupations. This heterogeneity can be related to differences either in ability to bargain or in preferences.
Origine : Fichiers produits par l'(les) auteur(s)
Loading...