My thoughts go to two places:
1) Conservatives believe in a lot of wrong stuff. Things either empirically disproven or logically faulty to begin with. Creationism is the most obvious example, and we can add in climate denial for good measure. These aren't fringe beliefs for the right, most US conservatives are creationists and/or climate deniers. Academia is about the pursuit of knowledge and truth and you can't expect to progress there by disavowing these. Being wrong about germane stuff is legitimate grounds for employment discrimination. There are a couple successful biologists who don't believe in evolution and while that proves it is occasionally possible to be a competent while rejecting a major portion of a field, such people cannot but be hampered by the baggage of these beliefs. That's what cognitive dissonance does. Scientists who reject very well established theories in their fields should have the onus on them to justify that and not be catered to any more than medicine should seek out Stork-theory-of-reproduction proponents.
I know there is a risk of group-think and I am not arguing for any special exclusion of conservatives, only that I see no reason to privilege the nonsense they insist on believing in over any other unorthodox theories that anyone might have. Academia should value diversity in thinking but seeing as how they are a group of psychologists perhaps they have other ways to ensure diverse thinking that don't involve purposely hiring lots of people known to hold bad ideas as some kind of foil. Don't they have other metrics of thinking styles or personal background that would avoid uniformity of biases?
2) I think Haidt's thinking on this is informed by the artificial moral equality he has assumed between liberal and conservative belief systems. He starts from the premise that they are equally valid viewpoints and if that were true, it would make sense for academia to include both in good numbers. I wrote about this awhile back, when I encountered Haidt's ideas around the differing sources of morality for liberals and conservatives:
[...] Having additional priorities to your moral calculus doesn't preclude your decisions from actually being immoral. Just because it made moral sense to your value system doesn't mean we should all just agree to disagree. Conservatives, according to Haidt's test results place a higher value on authority than fairness or avoiding harm. So they'd rather not weaken a strong President who engages in torture. Perhaps that result is internally consistent with their values, but it is still immoral, if the word is to have any meaning at all.(Quick primer: Haidt's basic idea is to define five basic values: harm reduction, fairness, loyalty, authority and purity. His empirical research shows that conservatives and liberals value these five at different levels, with liberals giving more weight to harm reduction and fairness, while conservatives tend to value all five nearly equally).
Consider what is happening when you attribute moral value to loyalty, by which Haidt mostly means loyalty to your in-group (however defined). That is inherently an exercise in contracting the circle of compassion and leaving some group outside it in order to favour others within. It is the thinking that allows for bigotry, racism, slavery, misogyny and oppression. If it is moral to harm person X because of fairness or harm reduction then it doesn't matter if group loyalty also supports harming X. If neither of these are true, then you must being doing something unfair, that does not reduce harm in order to support group loyalty by harming person X.
Consider two people in a position to disarm a shooter spraying a crowd of innocent people. Person 1 decides to do so because the shooter is killing many people and stopping that is moral. Person 2 decides that the shooter is "like him" in some sense, sharing some type of identity and decides not to disarm the shooter, because that would lead to one of that person's group being killed or put in jail.
Is Haidt really claiming these two value systems are morally equal? This example is admittedly extreme and I am not claiming most conservatives would decline to disarm a mass shooter if they thought the shooter was also a conservative, just that if your value system valued loyalty to the group that much, it can easily lead to absurd results. Conservatives do routinely find that actions are moral when America does them (say, waterboarding) which are not moral when others use them (say, Japan).
It rather appears that loyalty, while a common human attribute and something we value when shown towards us, is not really inherently moral at all. Actions are either moral or not regardless of whether one demonstrates loyalty in doing them. What is a "whistleblower" but a person performing moral yet disloyal acts? What is a "cover up" but group loyalty protecting immoral actors from reprisal?
I might be beating this point to death but this is what happens when you ignore the sources of liberal and conservative morality. Liberal morality preferencing harm reduction and fairness is not some accident, it is the consequence of hundreds of years of philisophical thought and development. It's not that we don't feel any affinity towards the in-group, have any instincts toward purity and authority, but that we consciously try and override such things. A quick trip to the latest Obama picture worship diary on the Daily Kos rec list tells you we don't succeed 100% of the time at this, but at least we try to battle the demons of our worst behaviour.
Ironically, Haidt's implied view that liberal and conservative value systems are equal is itself part of the dogma of entirely subjective morality (Sam Harris has recently taken up against this to much controversy). I hope conservative psychologists enjoy having a moral relativist argue on their behalf. I am a little proud as a liberal of the reaction of the field to Haidt's speech (read the NY Times piece to the end to learn how they have already begun to heed his call trying to positively support conservative psychologists) even though I think it a poor idea, I do wonder how that same call would go over at a gathering of any right wing leaning profession.