Analytical Vice

Interesting discussion on analytical vices, libertarian and otherwise, started by one of the blogosphere's leading suppliers of public goods, Tyler Cowen-

Tyler starts things off by claiming that a libertarian analytical vice is to assume that the quality of government is fixed and thus reduce all arguments to (fixed) government quality vs. market solution quality.

Henry Farrell at Crooked Timber responds with a mention of Exit, Voice, and Loyalty, saying that allowing exit (in his view, analagous to the market) undermines political voice (by letting dissidents flee instead of being forced to be rabble rousers at home).

Called out by Tyler in his post as an example of the vice in action, Glen Whitman responds by saying that government quality is not a fixed point, its analagous to thinking of government quality as a bell curve distribution that, on the right tail beats some market outcomes, but has a mean quality lower than market outcomes (defined by Glen as "any solution not governmental").

Back at home, Alex Tabarrok counters by saying not only is Tyler wrong, but he's 180 degrees wrong- libertarian analysis has been primarily aimed at improving government quality, and is not about assuming one static level of governmental quality or even one static distribution of governmental quality. And, not only has libertarian analysis been mostly about improving governmental quality, its been a phenomenal success.

Observing all of this and playing the role of inscrutable meta-critic, Don Boudreaux warns of "the contrarian vice"-

Not all contrarians commit the contrarian vice; Tyler doesn't. But the contrarian vice is a hazard of being an accomplished contrarian. Contrarians run great risks of rejecting some piece of wisdom simply because it is widely accepted -- and of confusing the possible for the plausible.

I think this is all fascinating, and it brings to mind a quote I remember from Coase in an interview I read somewhere (bleg for link!), such that for any government there are diminishing marginal returns on governmental action, and if a government does too much it can easily push its marginal return to negative territory; I believe that was Coase's own view at the time, that government ought to pare itself down so that it could be more effective overall. I think he also said that in the 1960s, so, yikes.

Whether or not the 'Coasean effect' above is true, it seems to me that the analytic vice that Tyler points out is or is not a vice depending on the analytic context- I'd think for a given situation or a "short run" analysis you can easily assume that the quality of whatever government process you're looking at is given and do a comparison. As the time horizon changes, the constraints on government quality loosen (Glen's view) until you get into complete abstraction in which case your arguments are about institutions and improving expected government quality (shifting the quality curve rightward, so to speak- Alex's view). It would indeed be a vice to consider government quality always fixed at every level of analytic abstraction, but I don't think that is a vice that libertarian analysts as a whole indulge too often.

And going back to Henry's point, I'm not sure if its his position that voice is better than exit or if he was just making a neutral observation; it would seem rather macabre, given his example, to prefer eliminating exit so that the best and brightest can *suffer* enough to want to 'voice' their displeasure to their authoritarian masters...

Share this