Abstract
This paper compares the results from two procedures for estimating wage differentials. Using Brazilian data, the standard procedure of Krueger and Summers (Econometrica, 56, 259–193, 1988) is shown to provide an overestimation of standard errors and a biased estimation of the measure of wage dispersion, in comparison to the methodology proposed by Haisken-DeNew and Schmidt (Review of Economics and Statistics, forthcoming, 1997).