Employment discontinuity and unemployment insurance reform: what effects on jobseekers' income and return to work?[Discontinuité de l'emploi et réforme de l'assurance chômage : quels effets sur le revenu et le retour à l'emploi des demandeurs d'emploi ?]

This research report presents an assessment of the impact on jobseekers of the reform implemented on 1 October 2021, which changes the calculation of the Daily Wage Reference (DWR) used as the basis for calculating the Unemployment Insurance (UI) daily benefit. There are two changes to this calculation method: the reference calculation period (RCP) has been extended from 12 months to 24 months, and the calculation is no longer based solely on days worked, but on calendar days in the reference period. Our research is based on the MiDAS databases (Minima sociaux, Droit d'Assurance chômage et parcours Salariés), from France Travail, DARES and CNAF. We compare two periods: one after the reform was implemented in October and November 2021 and one before the reform was implemented in May and June 2021, the latter in order to avoid the summer period. We defined a group of ‘treated' individuals, i.e. benefit recipients who had a discontinuous career path during their RCP and were therefore likely to be affected by the reform, and another group of ‘untreated' individuals with a continuous career path. Our objective was to estimate the causal effect of the reform on the amount of DWR, the actual unemployment benefit allowance and the overall standard of living of jobseekers, on the one hand, and on the return to work, in particular on the job duration search, the duration of the new contract and the type of contract obtained, on the other. To do this, we used two methods: (i) an DWR simulation method, coupled with a difference-in-differences (DiD) method; (ii) matching using the Coarsened Exact Matching (CEM) method in order to make our two samples (before and after the reform) comparable, coupled with a DiD method using regressions (ordinary least squares, ordered probit and Cox duration models).