54

To lower unemployment, the French government decided in 1998 to reduce the working week to 35 hours for full time workers. This measure required firms to modify their choices with respect to employment, working hours, organization, and production. This paper evaluates ex post the effects of these various choices on firm performance. A sample of over 100,000 firms was collated from various datasets. Firms’ decisions to shift to a 35-hour week is not the result of a random process. Effects related to the measure and to firms’ growth characteristics must therefore be distinguished when considering changes in firms’ performance, which a non-parametric method allowing for the correction of a selection bias. Econometric results stress the fact that the shift to a 35-hour week led to a significant increase in employment increase of at least 5 percentage points without worsening firms’ economic position, as was expected by many ex ante theoretical viewpoints.