Can Work-Sharing Work?

This paper assesses the effectiveness of work-sharing as a tool against unemployment, within a simple model where unemployment is induced by a binding minimum wage. To be effective, policies such as working time reduction or early retirement require that the workforce be homogeneous enough: The unemployed can then easily replace the freed up hours of work. We calculate the effects of work-sharing and of classical payroll tax reduction policies, as a function of some homogeneity measure and of the relevant elasticities. On the basis of existing empirical results, we estimate that in France, 2.5 hours of work must be removed to free up one hour of work: labor force heterogeneity makes work-sharing much less effective than a naive analysis would predict, and payroll tax reductions are more cost-effective than work-sharing with income compensation