At the risk of being a one-note Charlie, here's the single best thing I've read yet on the bias in academia question, by Tim Burke (hat tip to Ralph Luker at Cliopatria). Money quote number one:

On the other hand, collegiality is a powerful cultural force in many colleges and universities, and its stultifying or comforting effects (take your pick) often have nothing to do with politics in any sense. A conservative or libertarian who is a mensch about his or her views and research may well be admired, even beloved, by liberal or left colleagues, and fondly regarded as valuable because of their views. On the other hand, someone like Daniel Pipes who is running around picking broad-brush fights with everyone whom he perceives as a bad academic, usually based on a paper-thin reading of their syllabi or even just the titles of their research, is going to be loathed, but as much for his behavior as his political views. A liberal or leftist who plays Stalinist Truth Squad in the same way is going to be equally loathed and avoided.

Money quote number two:

12. Kieran Healy rightfully observes that conservatives talking about this issue are making an interesting exception to their general tendency among conservatives to assume that results in the market are probably based on some real distribution of qualifications rather than bias or discrimination. It might be fair to assert in response that academic hiring is a closed or non-market system, and this is precisely what is unfair about it. But if so, it requires that one demonstrate that there is a class of potential, qualified individuals who are being discriminated against at the time of hiring, or that these individuals are being discriminatorily weeded out at the time of initial acceptance for training. If not, then the argument that conservatives are being discriminated against in academic hiring practices is exactly comparable in its logics and evidence to the logic of most affirmative action programs and many other antidiscrimination initiatives, that there is a subtle systemic bias which is producing unequal results that prevents a “normal” sociological distribution of candidates in particular jobs. It behooves conservatives who want to claim this to either concretely explain why this argument only applies to conservatives in academia, or to repudiate the standard conservative argument against affirmative action and other public-policy programs designed to deal with subtle bias effects.

13. On the other hand, most of the people mocking or disagreeing with the claim that conservatives are treated poorly in academia seem to me to be equally at odds with many standard representations of bias effects that are widely accepted by liberals or leftists, namely, that bias is often subtle, discursive, and institutionally pervasive, and that “hostile environments” can exist where no single action or statement, or any concrete form of discrimination can be easily pointed to as a smoking gun. Most of those claiming a bias against conservatives in academia are pointing to exactly these kinds of hostile-environment incidents and moments, and seeing them as causing the same kinds of psychological and inhibitory harms that this type of discrimination is said to cause in other contexts. I accept that people edging away from you in an elevator is a type of bias-effect that is harmful—an often cited instance of the kinds of subtly pervasive discrimination that African-Americans may suffer from in mostly-white institutions. I’ve never experienced myself because I’m white, and had I not read of it in the personal, anecdotal accounts of many African-Americans, I truthfully would never have noticed it. Same here. I don’t understand why it is so hard to accept that self-identified conservative undergraduates, graduate students, and faculty report experiencing many similar forms of pervasive, subtle bias. What I'm seeing from many of those who dismiss these claims is a collective eye-rolling, a sort of "big deal, so your professor sneered at you, get over it". And yet few of those doing that eye-rolling would say the same to a student of color or a woman reporting similar experiences. The grounds on which many critics are doubting that such bias exists would have to, in all honesty, extend to all anecdotal, experiential or narrative claims of bias. The only way to salvage such claims would be if they could be profitably correlated with quantifiable evidence of discrimination—but in this case, we have some evidence to that effect. The only other way to salvage this point is to say, "It's wrong to be biased against people because of their race, gender or sexual orientation, but not because of their politics". A few seem willing to say just that: I can only say I think that's a big, fat mistake on a great many fronts.

All I can say is "I wish I'd written this."