Anarchy Rebound

At Cato Unbound, Peter Leeson has reignited the intra-libertarian debate on minarchism versus anarchism. Bruce Benson, Dani Rodrik, and Randall Holcombe all have responses up.

When I argue with libertarian anarchists, it’s often like an out-of-body experience: I suddenly feel like I know how others see me. I regard libertarian anarchists in much the same way that non-libertarians must regard libertarians in general: kinda wacky, often very smart and well-read, idealistic, lacking in political realism, and a little too confident in the power of their own ideas. It gives me a dose of humility about my own perspective.

That’s not intended as a serious criticism of anarchism, or of any particular anarchist (since some are more reasonable than others). Just as I think non-libertarians ought to take libertarianism seriously, libertarians ought to take anarchism seriously, at least as a test of the limits of our own ideas. The fact that a position is “extreme” is not an argument against it. But at the end of the day, I’ve just never bought it.

My primary objection to anarchism is that I don’t believe it’s sustainable – or as Holcombe puts it, it’s not a stable equilibrium. I’m unpersuaded by the examples of medieval Iceland and Ireland – first, because those were island societies relatively isolated from the outside world; and second, because those societies had much less anonymity than does modern society, allowing for social sanctions that would be much less effective now.

Moreover, these societies have to be put in the context of the rest of the history of civilization, in which states have emerged out of an initially anarchic context. I will sometimes ask anarchists, “What is it that makes the modern world not an example of a highly evolved anarchy?” States have not always existed, after all. As I see it, a stateless world is, by definition, anarchy – whether or not it has the desirable features that anarchists would like it to have. And out of that world, states arose, grew, and largely displaced other forms of organization. Even in Iceland and Ireland.

In response to this challenge, libertarian anarchists with whom I have argued usually say that the initial conditions for anarchy include more than the mere absence of states. There need to be a certain number of already-existing (and perhaps equal-sized) protection agencies, or power sources, or something like that. What this says to me is that anarchism is not an equilibrium that is likely to evolve or come into being spontaneously; those initial conditions have to arise through luck (Somalia) or deliberate choice (states voluntarily dissolving themselves?). This seems an odd position for people who place so much emphasis on the notion of spontaneous order to take.

So, in my estimation, libertarian anarchy is (a) unlikely to emerge in the first place, and (b) likely to evolve into states even if it somehow gets established to begin with. The best response I’ve heard from anarchists is that the minimal state has the same flaws. After all, there have been no truly minimal states in history, and those that have come close have all expanded into something much larger. Sadly, this is true. It is not, however, an argument for anarchism over minarchism; it’s an argument against both. Both varieties of libertarianism have an “ought implies can” problem. The anarchists seem to be saying, “Between two fantasy worlds A and B, I prefer A.” Where does that get us?

I think it gets us to Holcombe’s main argument, which is that anarchism is a dead letter in the policy debate. Maybe the minimal state is, too. But we can certainly try to move policy in a more libertarian direction. Part of that process is explaining to others how our ideas could possibly work in the real world. And here I must differ from Holcombe, who thinks anarchism is useful because it frames minimal-statism as a kind of middle ground: “[B]ecause anarchists extend the bounds of the political debate on the role of government, arguments to eliminate this tax or that regulation become more of a middle-of-the-road proposition.” Sometimes this kind of framing strategy works, but I don’t think this is one of those times. Instead, I think libertarianism of the kind that normal people might accept gets found guilty by association – “See, those libertarians really don’t think we should have government at all!”

If explaining how privatization of the police and military could work would help to advance libertarian ideas more generally, I could see its usefulness. But as sophisticated anarchists like David Friedman have recognized, they are not same ideas. The argument for the efficiency of a competitive market in (say) computers, set against the background of relatively well-enforced rights of property and contract, does not translate into an argument for the efficiency of a market whose function is the establishment and enforcement of those background rights. There’s a chicken-or-the-egg problem, and to overcome it the anarchists need different arguments. As a result, arguments for the viability of markets in rights protection – however convincing or unconvincing they may be – do little to move the debate forward in other areas.