You might be right, but the female fraction of CS/math PhD's also dropped by a factor of 4 between 1920 and 1960 (see my other post). So the effect extends beyond "unskilled" computer jobs.