I just discovered that my pal, Duke Poli Sci chair, and Social Change Workshop faculty member Mike Munger has a blog. He appears to be doing a fantastic job as the voice of reason in Duke's l'affaire de Kurian, which David Bernstein is heated up about over at the VC.
Check it out(I'm pretty sure that's a photo of Buchanan & Tullock in his header). Mike is wickedly smart, thinks he's real funny, and is a mensch of a good-ol-boy.
Munger Blogs
Will Wilkinson 
This goes back to Nietzche’s slave v. master morality doesn’t it? The issue, which you touch on, is that we have a tendency to go along with our peers because that works well for the 99% of decisions we make every moment of our lives. The problem is that many of the points of contention are in that 1% where going along with your peers is unwise (that is not to say they are correct, but that there is little to no reason to suspect they are correct)
I’d have to argue that one’s tendency to want a herd about one is more than pragmatic, however. More fundamentally, isn’t it that we want to be valued, to not be rejected (to feel alone)?
After all, without difference between yourself and another, you can’t feel threatened by him/her. Hence Locke’s view of the body politic. (If we all pretend we’re one “body,” we’ll all be less likely to feel threatened by each other, and thus less likely to get into altercations).
Consider the strut. It is a signal of dominance, but if you to use it without the power to back it up, you may be severely punished. Same with trying to offer distinctive opinions when you aren’t respected enough for others to take them seriously.
Intellectual mirroring can serve useful purposes, such as increasing trust and enabling cooperation. The decision to mirror or not to mirror should be a pragmatic one, driven by considerations like the likelihood of future cooperation, the anticipated benefits of dominant behavior, the risk of being wrong, etc.
In most cases this is hard to figure out so I think a reasonable heuristic is simply to make a more or less random decision, and then compensate for earlier errors based on how you feel. If you regret being overly cooperative, be annoying about the next thing that comes along.
In other words, minimize risk by choosing a randomized blend of different strategies.