If you plot women's share of doctorates in different fields, it's pretty telling that CS is the only field in which the fraction of women dropped significantly

If you plot women's share of doctorates in different fields, it's pretty telling that CS is the only field in which the fraction of women dropped significantly between 1920 - 1960. In 1920, 20% of mathematics and CS PhD degrees went to women, by 1960 it was down to 5%. It took until around the turn of the century for it to get back up to 20%. (See http://www.nsf.gov/statistics/nsf06319/ http://www.nsf.gov/statistics/nsf06319/)