Libertarians, Logic, and "Free Liberals"

Andrew Sullivan shares a great post by Tim Lee contrasting Hayek with Rand and Rothbard. Lee writes:

"One of the more pernicious influences of Rand and Rothbard on the libertarian movement was their tendency to treat every policy problem as almost reducible to a logical syllogism. Too many libertarians act as though they don’t need to know very much about the details of any given policy issue because they can deduce the right answer directly from libertarian principles. The practical result is often to shut down internal debate and discourage libertarians from thinking carefully about cases where libertarian principles may have more than one plausible application. Hayek seems to have written “‘Free’ Enterprise and Competitive Order” with the explicit purpose to combat that kind of dogmatism. He thought it “highly desirable that liberals shall strongly disagree on these topics, the more the better.”"

I've critiqued the Austro-libertarian fetish for deduction on here before, and Lee makes largely the same point here. I think you have to start a step back, though, and simply point out that human society is immensely complex. You can take a few first principles that you make up, like mathematical axioms, and derive a logical system from it that you can apply to situations that seem to be relevant to those first principles (like physical matter). That's a deductive project that can have real payoffs. But you simply can't do that with human society. Too much is going on. Even if you have a true axiomatic claim about human action, you're very likely to miss lots of other first principles and contingencies that could completely change the picture. Imagine if we just dropped half the field axioms in mathematics. The set of deductions you would be able to come up with would be incomplete! You wouldn't be able to demonstrate nearly as much, and you'd certainly be missing substantial portions of the way classical mechanical systems behave. Or, as another way of putting it, think of what happens when you add additional field axioms in higher mathematics. We have a hard time getting to the bottom of physical reality this way, and yet many Austro-libertarians go through life as if they can rule out empirically derived social science because their deductions from their axioms don't come to those conclusions.

I've recently been in contact with Jan Helfeld - a fairly well known libertarian interviewer - that covers these same frustrations that Lee is sharing. Jan uses a Socratic Interviewing method based on a few underlying claims that he uses to try to demonstrate inconsistencies in the arguments of various politicians and media personalities. Kevin Rollins (editor at Econ Journal Watch and a GMU student) suggested I debate with him and exchange some thoughts. We're still trying to set that up, but I sent him some written thoughts I had on his whole project. This is a selection of what I sent him that is relevant to this point about libertarians and their use of logic:

"Your interviewees note this [the problems with his axioms], but they don’t always articulate it in the best way (probably because they are caught off guard or just haven’t thought about these issues). But they do realize there are fundamental problems with the deductions that you make. My concern with some of your interviews is that your interviewees tell you this – they tell you your questions miss the point – but you never change or try to think about what may be wrong with your questions, or work with them to figure out a better way of approaching the problem. You keep asking the same question over and over again instead of considering the possibility that you’re asking the wrong questions. You may catch them in logical inconsistencies, but if the logical structure you impose on them is meaningless that doesn’t really prove much except that you were able to trap them in a word-game. Is the point to prove Congressmen are not the greatest logicians or that they don’t understand basic principles of a liberal society? You may succeed at the former – it’s not clear to me you succeed at the latter.

It reminds me of something Keynes said about Hayek. He wrote that one of Hayek’s books was an “example of how, starting with a mistake, a remorseless logician can end up in bedlam”. You strike me as this sort of “remorseless logician”. You take a first principle that you don’t appear to closely interrogate (or at least that you don’t appear to think has the problems that I see in it) and then you persist at it until you elicit a reaction from your interviewees."

Jan seems like a great guy but I feel like the method that strikes him as being so effective can be very misleading. People will nod their heads to claims like "individuals don't have the right to initiate violence" and "individuals can't delegate rights they don't have". These sound like pretty reasonable assertions. There's nothing particularly absurd about them. So his interviewees nod their heads in agreement. Jan then takes those axiomatic building blocks, shorn of their qualifications and problems, and makes a quick case for libertarianism out of them. But that seems to just gloss over the really hard questions! What are rights really? Is government just an institution that individuals delegate their rights to or is it something else? What are some extenuating circumstances where an initiation of violence might be justified? Are rights themselves an initiation of violence (a la Proudhon)? Is government action that seems on the surface to be an initiation of violence really an initiation of violence? Most of the time these secondary questions (which come up because of flimsy "axioms") don't even come up - or if they do come up they get plowed through as if the interviewee is changing the topic.

The confidence that libertarians get from this deduction-fetish, which Lee points out in Rand and Rothbard, is very dangerous when they start talking about organizing society. It raises the specter of libertarian social engineering, which I've critiqued at length here. I have no quarrel with social engineering, but this sort of large-scale social engineering and the risks that it involves simply isn't conceivable for a more empirical framework that does not insist that its deductions are universal and unimpeachable (except by another logician's ability to see a flaw in the logic).

One solution is simply to be more open minded and pluralistic when it comes to libertarian thought. That sounds like a "no duh" solution, but the simple willingness to be speculative, entertain alternative possibilities, and impute significance to competing priorities can go a long way. A good example of this is Kevin Rollins' own "Free Liberal" project*, which is outlined at his website and carried out socially in a variety of venues. Kevin's approach is interesting - I've had lots of opportunities to correspond with him and his group at this point, and their willingness to embrace a pluralist inquiry and perhaps most importantly a speculative inquiry is impressive. They are quite libertarian, but their discussion seems to be a lot more productive than the discussion you see in other venues. This is the sort of thing that is going to move liberal thought forward. Not every libertarian is a Misesian or a Rothbardian or a Randian either (and these three gangs don't always get along with each other). There are lots of libertarians that don't use these logic games, and there are lots that like the logic-games, but don't rely exclusively on them.

*I am hereby issuing an open invitation to Kevin to offer a guest post here exploring Free Liberalism further.