Comments on my debate with Jan Helfield


1. Jan repeatedly makes arguments of the form “isn’t such and such a bad outcome possible under your system.” The answer is that since we don’t have an adequate theory of human societies almost no outcome is provably impossible. The issue is whether particular bad outcomes are more likely under my system than his.


In the case of civil war, which he argued could conceivably happen under an anarcho-capitalist system, we have some evidence for the risks of his system. The government set up by the Constitution, a limited government of the sort Jan wants, resulted in a very bloody civil war less than a century after it was established. Similarly, if nuclear weapons exist, whatever the institutions, there is some possibility that they will be used. But we observe that in a world run by states a lot of nuclear and thermonuclear weapons got built and two of them got used.


2. One of Jan’s arguments was that anarchy existed for most of history and civilization only got started when government came along. There are two problems with this. The first is that, as I tried to make clear in my opening comments (and in more detail in things I’ve written), the workability of the institutions I proposed depends on competition, on a society large enough relative to the equilibrium size of rights enforcement agencies so that there are lots of them. You can describe hunter gatherer societies as either anarchys or governments, but they did not have the sorts of institutions I describe.


To put the point differently, Jan would reasonably object if I claimed that the existence of some very bad governments, such as the Khmer Rouge in Cambodia, proved that government is bad—because that isn’t the sort of government he is defending. A hunter gatherer band is not the sort of anarchy I am defending.


The second problem is that he is making a post hoc ergo propter hoc argument. Lots of things changed that led to modern societies including the first agricultural revolution, the second agricultural revolution, the industrial revolution, and a variety of other changes. He thinks they were set off by the invention of the state but has no way of distinguishing that theory from the theory that the surplus produced by the invention of agricultural made it possible to establish states using some of that surplus to maintain the apparatus needed to funnel the rest of it to those ruling the states.


3. Jan keeps making arguments that hinge on his description of how he thinks my system would work without, so far as I can tell, any serious analysis of what institutions would or would not be stable in that environment. I keep pointing out that a firm that wants to defend its customers right or wrong, or even to defend its customers whenever its own court says they are right, faces a problem. Either it gets other rights enforcement agencies to agree or it doesn’t. To get them to agree, it has to pay more than the loss to them of being unable to protect their customers’ rights against its customers, which is unlikely, since rights violation usually benefits the violator less than it harms the victim. If it doesn’t get them to agree, every dispute results in violent conflict and since it is facing that situation in all disputes while the other firms only in disputes with it, its costs are very much higher. Either way, it goes out of business.


4. Jan offers three responses to the argument for why voters will be rationally ignorant:

A. Sometimes, as in the Florida vote in Bush v Gore, a few votes make a difference. Even in that case, a few thousand votes made a difference—there was no voter who would have changed the outcome by changing his vote. And that was one case in one state in all of U.S. history, so the odds for a voter of ending up even in that situation are tiny.

B. You can influence other people’s votes. That is true, but unless you are very influential you cannot change enough votes to have a significant effect on a major election. So the implication is that most voters will be rationally ignorant, a few influential individuals will find out what outcomes are in their interest and try to produce them. There is no reason to expect their interest to be the same as the general interest. You end up with a system that benefits concentrated interest groups, groups that do have enough at stake and enough resources to affect outcomes, at the expense of everyone else.


C. Jan thinks it is inconsistent for me to believe it is in my rational interest to argue for anarchy but that it is not in voters’ rational interest to find out who they should support and do so. What he is missing is that, in both cases, the main incentive to act is not to influence the outcome. People participate in politics because it’s fun, because they like arguing, because it makes them feel important, because it is a way of meeting others with similar views, for a wide variety of reasons almost none of which depends on whether they are on the right side of the issues they get involved in. The problem is not that people have no incentive to vote but that they have little incentive to be well informed voters, and if voters are not well informed their voting does not force government to behave well.

I will probably expand this as more ideas come to me. If Jan writes a similar commentary from his side, I will link to it.

Additional Material:

The Machinery of Freedom. Part III explores how a market anarchy might function.
Chapters for the Third Edition (drafts). Several of these explore relevant issues, in particular the problem of national defense.