I must admit that for a moment I enjoyed basking in my own victim hood. My failings are not my own but are due to discrimination! Ahhh, that feels good.
Much as I would like to lay my failings at the feet of the system, however, I cannot do so... I... reject these studies out of intellectual consistency.
Alex doubts that markets discriminate against women. Why? Basic economics predicts that in a competitive market, firms that don't hire on the basis of merit don't last. Advanced econometrics tells us that controlling for differences in workers' quality makes apparent discrimination shrink substantially or vanish. If he is theoretically and empirically suspicious about the reality of discrimination against women, asks Alex, how can he give discrimination against conservatives any more credence?
The main problem with Alex's analysis: The economics of discrimination assumes that firms maximize profits, and can't produce at more than minimum average cost forever. Neither assumption holds for universities. Universities are non-profits. If chairmen or administrators figure out a way to get higher quality professors for lower wages, their pay does not go up. In fact, their lives probably get harder, because their "cost-cutting" efforts will provoke a bitter backlash on campus.
Ask yourself: Why don't department chairs fight to replace senile tenured faculty with bright, eager graduate student instructors? First, they pocket essentially none of the savings; second, their lives would become a living hell.
But how can universities produce at higher than minimum average cost forever? In the case of government universities, it's obvious. They've got tax subsidies to keep them afloat. In the case of private universities (and public universities too!), it's almost as obvious: They've got alumni donations to keep them afloat. Either way, in the university industry, it is NOT "sink or swim."
In theory, then, we should not be surprised if universities discriminate against women, conservatives, or anyone else. The only question is: Who does the hiring, and who do they have a "taste for discrimination" against?
Like Alex, I don't think universities discriminate against women. But my reason is not that discrimination is a one-way ticket to university bankruptcy. No, my reason is that the people in charge of hiring professors like - indeed, prefer - to hire women. Maybe not in math and the hard sciences, but most academics want to assuage their liberal guilt. (And even math and hard sciences face pressure from administrators who want to assuage their liberal guilt). I've repeatedly heard academics insist "We really need to hire a woman!" Sounds like a taste for discrimination against men to me.
In contrast, people in charge of hiring professors do not like - indeed prefer not to - hire conservatives, Republicans, or libertarians. Some hate those bastards; others simply feel uncomfortable around them. Dan Klein's findings of overwhelming ideological imbalance in academia speak volumes. I doubt if the median professor has a single close right-wing friend. Again, this is probably not too important in math and hard sciences, but if research and politics connect, your politics affects whether people want to hire you.
1. There is huge selection bias. Alex and I are libertarian professors. You've got to sample over all the right-wingers who wanted to be professors, not just the ones who made it.
2. One reason I chose economics was because I correctly perceived it to have the weakest left-wing bias of any social science or humanity. Otherwise, I might have done philosophy. If economics were as leftist as most social sciences, I might have become a lawyer instead. (Egad!) I can't believe I'm the only potential professor who weighed these factors in his occupational choice.
3. Even if we control for quality of publications, the gatekeepers - journal editors and referees - also feel virtually no financial cost of rejecting articles they find ideologically distasteful. So there is probably more discrimination against right-wingers than the data suggest, not less.