Targeted reductions in employers' social security contributions are conceived as a key policy instrument used to facilitate job creation when labour cost is so high that it may deter companies from hiring new employees. Among the different measures implemented in France, the set of instruments implemented in the West Indian "Départements" is the most accomplished form as the rates of exemption as well as the base and scope of these meausures have reached their maximum there. This paper seeks to determine to what extent these instuments contribute to job creation looking at the growth rate in the number of employees through the use of a balanced panel of business entities with at least one employee, drawn from a matching between several administrative data sources from 2004 to 2011. We studied the differentiated effects of the payroll tax using a quantile regression for panel data estimation technique. We show that the impact of the exemption rate and of the intensity of use of the various measures on changes in the number of employees differ according to the establishment growth rate. They tend to be negative on the left side of the distribution and positive on the right side. However, these effects may significantly differ according to the size class and the industry in which the business operates. Large ones tend to be advataged compared to the ones whose total number of employees is fewer than eleven, whereas the estimated correlation between growth and exemption rate is higher for most of the entities in the manufacturing industry but only for a small part of those in business services.