Even the first fact is a bit biased. In controlled experiments that rarely go beyond changing names on resumes, you observe bias against women.
When you observe actual practice (with official and unofficial AA, hiring committees, and of course various statistical problems), you tend to observe bias in favor of women.
When you observe actual practice (with official and unofficial AA, hiring committees, and of course various statistical problems), you tend to observe bias in favor of women.
http://www.nap.edu/openbook.php?record_id=12062&page=R1
The picture of what real life looks like is far from clear.