(joint with Pauline Carry, Claire Montialoux, Elio Nimier-David and Alexandra Roulet)
We revisit the longstanding policy debate on working time reductions by studying the effects of France’s 35-hour reform. The policy mandated a reduction in weekly hours from 39 to 35 hours without loss of pay and was accompanied by payroll tax cuts targeted at low-wage workers. Using matched employer–employee records, firm balance-sheet data, and worker surveys, we exploit the staggered rollout of the reform across firm sizes between 2000 and 2002 to identify its causal effects. The reform reduced total hours worked per firm by about 6%, through shorter workweeks for full-time employees and a shift toward part-time jobs. Overall firm employment remained unchanged. However, this average effect masks substantial heterogeneity by skill: among higher-skilled workers, fewer hours per job were fully offset by additional workers, while both hours and employment declined for lower-skilled workers. To isolate the effect of the hours mandate from that of the payroll tax cuts, we develop a policy decomposition design that augments the size-based quasi-experiment with two additional sources of variation—firms’ pre-reform shares of full-time and low-wage workers. We show that payroll tax cuts offset what would otherwise have been sizable job losses from the workweek reduction alone. Finally, the reform led to a modest 1% decline in firm-level value added, partly cushioned by a 3% increase in total factor productivity.