The problem with this argument is it assumes the differences are due to discrimination, but when we try to measure that we almost can't detect the effect of discrimination because it's so small. The differences are not skin deep like that, but usually have deeper roots. To be clear I'm not saying there are no biases like racism, but that it doesn't seem to be the dominant factor causing differences in outcome. Often the roots are deeper, like higher divorce rates, worse schools, differences in parents' socioeconomic status. Between men and women the salary difference is almost entirely a difference between women with children and men and women without.