But the Pew Survey one: wow, that was surprising. The survey summary shows that the percentage of scientists who identify as Democrat goes down if you look at those who are working not in universities but in private industry (where it's just under the rate found in the general public; see quote below). But for the overall rate to be only 6% Republican: that's so low and so weird! http://people-press.org/2009/07/09/section-4-scientists-poli...
From the section 'Scientists and Politics':
Overall, 52% of the public identifies as Democratic or leans Democratic, while 35% identifies as Republican or leans Republican. Majorities of scientists working in academia (60%), for non-profits (55%) and in government (52%) call themselves Democrats, as do nearly half of those working in private industry (47%).