that's not noise, that's the sound of freedom - oh, by the way, I voted Libertarian

Libertarianism, by which I mean yer bog standard comments thread North American subtype, is irrational. This struck me in the context of this post of Charlie Whitaker's, which was of course criticised by the local libertarian in the usual terms. First of all, let's define terms; the standard set-up of modern right-libertarian arguments is based on two layered arguments. The first is a weak-form argument, which derives from the Austrian tradition in economics and, therefore, eventually from classical liberalism. It goes a little something like this: As Hayek and von Mises pointed out, we can't possibly know enough to manage the whole economy, and therefore, we should rely on market mechanisms as far as possible.

The notion that GOSPLAN wasn't a triumph is hardly controversial, and for most of the political spectrum, the debate is really about how far we can rely on market mechanisms and what the alternatives to them should be. But the libertarian argument does a hop, a skip, and a little leap of faith here; it extends this argument to claim that any nonmarket organisation will inevitably be less efficient than any market alternative, and that it will be so much less efficient that the costs, whatever they are, will always be worthwhile. We can characterise this as the argument from efficiency.

What's worth pointing out here is the degree to which the volume has been turned up. All the points of doubt in the economic argument have been replaced by conviction. The argument against central planning is built on scepticism, drawing its strength from flexibility like a great bridge or the backbone of a ship; but suddenly, there's a lot of massive but brittle certainty about. What may very well be very small efficiency gains, which may well come at great cost, are being used to legitimate a very strong normative claim indeed.

This is something libertarianism shares with conservatism; the belief that 1) the government will always be incompetent at allocating investment, but 2) it is perfectly competent to de-allocate it. Ministers and their officials cannot under any circumstances be trusted to choose projects to fund, but they have perfect knowledge of what to cut. Now, this isn't actually as stupid as it may sound - as Daniel Davies says, it's much easier to identify stupid proposals than it is to come up with intelligent ones. But economic decisions have a fundamental duality - the decision to buy X involves not buying Y, the decision not to buy Y involves going without Y. Cutting, privatising, deregulating - these are the same activities as spending, nationalising, or legislating.

Clearly, there is a need for a stronger philosophical foundation.

This we find in the doctrine of the illegitimacy of taxation. The ability of the state to collect tax depends on its monopoly of force, and therefore taxation is actually a form of theft. Again, a hop, a skip, and a little leap of faith, and we arrive at the position that as all the activities of the state depend eventually on tax, therefore an NHS blood transfusion is as illegitimate as, say, being shot seven times in the head by a policeman. The immediate counter-argument is to knock the ball forward from Hobbes up to Locke, and to appeal to the social contract. It's entirely possible to enter into a voluntary association to which you pay dues and in which you agree to accept its internal discipline; and here we come up against the rock. You didn't choose to be born into a democratic society - and therefore the contract is invalid.

Of course, at this point most people will lose their temper with the obvious stupidity on display. The horrors - to be born into a democratic society rather than, say, Somalia or the eastern Congo! What an appalling imposition! But then, we parted company with empiricism way back when we decided that Railtrack was necessarily a better idea than any other way of organising the railways.

But there's another, more fundamental point here. The doctrine of the total illegitimacy of the state is a huge, extraordinary claim. It is at least as extraordinary as the claim that a central planning commission with 1920s information technology could manage an advanced economy with better results, both in terms of equality and in terms of economic growth, than any alternative. (And, going back to the DRC for a brief holiday, it has arguably amassed a comparable pile of corpses.) Extraordinary claims demand extraordinary evidence, but then, we're dealing with libertarianism here.

Here is the point; if we're being rational about this - and we better be, as it's absolutely indispensable for any libertarian argument to work whatsoever - things should change as soon as we accept this claim. In fact, even deciding that such an extraordinary statement is worth considering, rather than noting that it is at least as crazy as Stalinism and passing it straight to /dev/null, should lead us to change our minds about other extraordinary statements. If we decide it's worth bothering with, we must have changed the criteria we use to assess these things. We need to update our Bayesian priors.

If minarchy makes the cut, why doesn't anarchy, an Islamic caliphate, libertarian Marxism, deep ecology, or the joke from The Onion about the political scientists who discover a new form of government which has features of both anarchy and fascism? Further, libertarianism is very excluding - it denounces everything, all the way across the centre ground and well over into conservatism. If you can find a place for it in your worldview, your priors have to be very odd indeed to exclude social democracy.

And therefore, it is, indeed, irrational; all the strength in it comes from the prior assumptions, and these assumptions have to take a very strange form indeed. Further, accepting a zone of political possibility large enough to hold libertarianism forces a rational person to accept a wide range of views as possible that libertarianism demands that you denounce and attack at every opportunity.

"Prior assumptions", after all, is a polite way of saying "prejudices", and irrational forms of politics inevitably become ways of expressing prejudice.

As a reward for putting up with this drivel: