9

This paper is about estimating gender wage discrimination using firm-level data, covering the 2002-2010 period, for the Belgian private economy. Compared to worker-level wage data, firm-level data present the advantage of containing an independent measure of productivity. Using the framework assembled by Hellerstein-Neumark, they permit separate estimations of gender-wage and gender-productivity gaps; and also — something crucial for the evaluation of gender wage discrimination — of the degree of (non) alignment of these two gaps. Results are essentially twofold. First, gender wage discrimination estimated using firm-level evidence is small compare to worker-level earnings-regression estimates. Second, in the case of Belgium’s private economy during the 2000s, it is only statistically significant for female blue collars. JEL Classification: J24, C52, D24