Fernando Teson has an excellent post up at Bleeding Heart Libertarians (itself an outstanding blog, or at least a blog with an outstanding name):
He begins with something that might sound familiar:
[I]n the political arena, a person often supports a policy, not because of the effects he thinks that policy will have, but because his supporting it has symbolic value for himself or others. Supporting the minimum wage is an act that stands for a value such as concern for the poor. The person who is concerned for the poor wants to express that concern, and there are acts that socially symbolize that concern: praising the New Deal, announcing that you voted for a Democrat, supporting public schools, criticizing Bush.
Symbolic behavior, I hasten to say, is not exclusive
ofto [sorry, couldn’t help it — B.] progressives. In libertarian circles someone may oppose environmental regulation for symbolic reasons. That position evinces a hostile attitude toward government regulation in general which he wants to express. In his haste to send the right signals he overlooks (say) the problems of externalities and market failure.
It shouldn’t shock us that people take irrational (Teson, charitably, writes symbolic instead) positions — supporting the minimum wage, opposing environmental regulation* — in order to signal their positions on various social issues. (“I’m concerned about poverty!” “I’m concerned about regulatory capture!”) In that sense, it’s perfectly rational for someone who opposes poverty and wants everyone to know it to support the minimum wage, especially if the latter dominates the former. It usually does: no matter how much I dislike poverty, I’m vastly unlikely to see it eliminated in my lifetime; on the other hand, if I signal socially that I dislike poverty, I can reap immediate benefits.
The problem is, signaling cuts both ways. Just as I’m unlikely to take seriously anyone who crows about the social justice of Zimbabwe’s monetary policy or expresses
monetarist mercantilist [ed: Holy shit, how’d I fuck up that bad?] sentiment without apparent shame, someone who’s concerned about poverty and supports the minimum wage to express that concern is unlikely to take arguments against the minimum wage at face value. (It is perhaps especially damning that those arguments happen to be well-founded.) It’s a pretty straightforward deduction: if I support position P because of axiom A, and you express position not-P, you must hold axiom not-A. (It’s also wrong; causation is not bijective.) If your support for the minimum wage is essentially an expression of your objection to poverty, you’re likely to interpret my argument against minimum wage as an expression of my support of — or at least acquiescence to — poverty. At the very least, you’re going to be extremely suspicious when I say that I really hate poverty as much as you do, because in your experience the proposition minimum-wage-bad is inconsistent with the axiom poverty-bad. Which, given the data coming from the real world, is unfortunate.
What’s particularly perverse is that this turns “I’m anti-poverty and I don’t like the minimum wage” into yet another fucking social signal rather than a simple statement of fact. (In this case it means “Ah, ha ha, I’ve read Tyler Cowen, and if you’re smart enough to read the blogs I like you’ll recognize my erudition.” Well, at least it’ll piss off Paul Krugman.) So because we thrice-damned social bipeds like communicating on an infuriatingly-encrypted side band, even the discussion of actual statistical data becomes a method of in-group identification.
This is why I like dogs better than people.
* Yeah, yeah, I know: in anarchotopia we won’t have environmental regulations. Anarchotopia won’t happen until transaction costs are trivially small.