A lot of attention is being paid to a recent study, by David Horowitz, of political affiliations of faculty in law schools and schools of journalism. A major finding is that Democratic faculty are far more numerous than Republican faculty. An underlying concern is that if the overwhelming majority of faculty shares a set of political beliefs, teaching and discussion are likely to be skewed.
The concern is an entirely legitimate one, but here's another: As decades of social science research have shown, like-minded people, engaged in discussion with one another, tend to go to extremes. Suppose that one group of people believes that global warming is a serious problem; another thinks that Harriet Miers is unqualified; another believes that aggressive affirmative action policies are desirable; yet another believes that feminism has "gone too far." After the members of these groups talk to each other, they are likely to shift toward a more extreme version of their original views. More technically: Deliberating groups, after deliberation, usually adopt a more extreme position in the same direction as the median of their predeliberation views.
This is the phenomenon known as group polarization. It is extremely robust, and it has been found in many countries. Juries have been shown to polarize; in experimental settings, juries generate punitive damage awards that are systematically higher than the predeliberation median of jurors. (A book of which I'm one of many coauthors, called Punitive Damages: How Juries Decide, has the relevant experiment.) Even federal appellate judges have been shown to polarize. In many domains of the law, Democratic appointees show especially liberal voting patterns when they are sitting on three-judge panels consisting solely of Democratic appointees. (By "especially liberal voting patterns," I mean far more liberal voting than Democratic appointees show when on a panel with one or two Republican appointees.) So too, Republican appointees often show especially conservative voting patterns when sitting on three-judge panels consisting solely of Republican appointees. (The Chicago Judges Project is studying this and related phenomena; an early set of results can be found in a 2004 paper in the Virginia Law Review, Ideological Judging on Federal Courts of Appeals: A Preliminary Analysis, by Lisa Ellman, David Schkade, and me.) Current work is investigating group polarization on political questions.
Why does group polarization occur? Several factors appear to be at work. 1) If a group's members are predisposed to think that X is true, group members will hear many arguments in favor of X, and few the other way. Exposed to more arguments in favor of X, people tend to become more convinced of X. 2) If a group's members are predisposed in favor of X, those who are skeptical of X might well silence themselves, so as not to seem obtuse or to risk their reputations. Both individuals and groups as a whole are likely to shift as a result. 3) People tend to moderate their opinions when they don't know what others think. If they're surrounded by like-minded others, their initial inclinations are confirmed, and they tend to become more confident -- and probably more extreme as a result.
Here's the upshot: When a college or university is skewed in one direction, group polarization is highly likely. Suppose that most faculty or most students are left-leaning. Their mutual interactions will probably move them further to the left. Of course it's possible that on some issues, the movement leads them in sensible directions. But to say the least, there is no guarantee to this effect. (Some of the bizarre radicalism of student groups in the 1960s was produced by group polarization; there were many causes, but social interactions were a contributing factor.) Group polarization often causes political correctness, and it is also fueled by it.
Of course group polarization occurs within countless groups; left-leaning organizations have no monopoly on it. The conservative furor over the appointment of Harriet Miers, for example, seems to be a case study in group polarization. (This is simply a point about social dynamics, not at all about the merits.) There's no simple "cure" for group polarization, but awareness of the phenomenon can provide at least a degree of inoculation.