We study the impact of payroll tax subsidies for low-wage workers implemented in France in 1995-1996 on employment at the firm level. This study is based on the computation for each firm in 1994 of the ex ante reduction in labor costs, due to the tax reduction. We adapt the treatment literature to this continuous treatment case and propose an estimation method based on series estimators. We find that, between 1994 and 1997, tax reductions are associated had strong effects on employment and wages. We find that they permit the creation of roughly 400.000 works.