Mechanisms to reduce charges on low wages were originally introduced in 1993. They were significantly extended in 1995 and 1996 to the point of becoming a central employment policy instrument today. Their impact on the economy has hitherto been forecast using traditional macroeconomic models. However, for the first time, a post evaluation of the effect of reduced charges on employment is put forward based on business data. This evaluation focuses in particular on the initial mechanism's significant extensions introduced in 1995 and 1996. Information on the distribution of wages in businesses in 1994 is used to calculate, business by business, the ex-ante reduction in the cost of labour due to this extension. The evaluation then compares the growth in employment and other performances of businesses benefiting ex ante differently from the mechanisms from 1994 to 1997. This evaluation uses and extends methods recently put forward to evaluate economic policies. In particular, the effect of the measures is separated from the effect of business characteristics (size, line of business, exposure to international competition, etc). Such characteristics are likely to affect the extent of the ex-ante cost reduction and the growth in employment («other things being equal» approach). The mechanisms to reduce charges triggered numerous job creations from 1994 to 1997. The employee growth rate that can be put down to these mechanisms is 2.6% in industry and 3.4% in the service sector. These measures are therefore thought to have created or saved 460,000 jobs in the economy from 1994 to 1997. Half of the jobs are reported to be unskilled jobs. These job creations can be explained by substantial replacements of skilled employees with unskilled employees and, to a lesser extent, capital with labour. This hence reinforces the idea often put forward that the job-richer growth observed over this period is associated with reduced charges on low wages. Volume effects due to price ...