|Before computers, workers just stared at their desks.|
demonstrates that it is not possible now, and doubtless will never be possible, to determine reliably whether any portion of the observed gender wage gap is not attributable to factors that compensate women and men differently on socially acceptable bases, and hence can confidently be attributed to overt discrimination against women.
In plain English: we can't prove that any of the wage gap is the result of sexism.
But that may be because CONSAD didn't look in the right places, as Barry Deutsch, aka Ampersand, explains in a post on Alas, a blog that critiques the methology and the findings of the CONSAD study. Notably, he points out, CONSAD's analysis ignores the issue of occupational segregation, which is at the root of much of the wage gap, as properly measured. And it ignores many other, more subtle kinds of discrimination:
When discussing direct employer discrimination, it’s more realistic to discuss elements like selective hiring, training, promotion ladders, and other things that are a good deal more complex than CONSAD’s vision of the labor market allows for. Given two equally able applicants for a $40,000 job, one male, one female – which one will employers tend to prefer? Once hired, who is more likely to get mentored? Who is more likely to be given the assignments that lead to promotion? Who is more likely to be perceived as doing good work, all else held equal? And if these factors mean that women are rewarded less than men for identical labor market participation, to what degree does that reduce women’s incentive to participate equally in the labor market? All of these are ways that sex discrimination actually happens in the marketplace — and none of them are detectable by by CONSAD’s methods.
There's much more to his argument than this; I'd suggest reading the whole thing.
For more on the issue -- including excerpts from and links to other useful posts on the wage gap by Barry -- see my own Further Reading post on the subject.