The first edition of this book got one good review, defined not as a favorable review but as a review that makes the author think. The reviewer was James Buchanan, later my colleague at the public choice center of VPI. The review, while generally positive, pointed out a crucial hole in my analysis of a stateless legal system. This chapter is my attempt to fill it.

Back in Chapter 29, I considered the case of two rights enforcement agencies whose customers disagreed about a legal rule, with customers of one supporting and of one opposing capital punishment. Each agency calculates how much it is worth to its customers to have their preferred rule apply to disputes with the customers of the other. The agencies then bargain; whichever is willing to offer more to get its preferred rule gets it. My conclusion—which I qualified a little in Chapter 54—was that the result would be an efficient set of legal rules in the usual economic sense, a set of rules that maximized the summed benefit to all affected.

There was one question that I forgot to ask: What is the default rule, the starting position that the parties are bargaining from? In order for the agency that wants capital punishment to get its preferred rule does it have to pay the other agency to agree to have capital punishment or only turn down the other agency’s offer to pay it to agree not to?

One way of answering the question is with another. If no agreement is reached, what happens? The obvious response is that, if no agreement is reached, the two companies will be, in the terminology of the Icelandic sagas, out of law with each other. Each conflict between their customers will have to be settled by ad hoc bargaining or violent conflict.

That suggests that one important determinant of the terms of agreement, if agreement is reached, is how well each side thinks it would do under those circumstances. To put the point differently, underlying the market for law is an implicit threat game. The worse I expect to do if it does come down to violence, the more willing I will be to either agree to your terms or pay you to agree to mine, the less likely I will be to insist on doing it the other way around.

There is a sense in which the rules of any society are built on an underlying threat game of the same sort. There are always potential conflicts, with a government or without it, over the legal rules and the rules that determine the legal rules. If one faction finds current rules sufficiently unsatisfactory, there is always the potential recourse to force, whether in the form of civil war, domestic terrorism, or merely large scale private violation of the law. The point was demonstrated most recently by the antics of the Occupy movement, which engaged in various illegal activities as a way of pressing for what its members saw as desirable change. It was demonstrated on a larger and much more violent scale by recent conflicts in Libya and Syria.

Seen from this point of view, one can view the social contract not as a voluntary agreement, which it obviously is not, but as a peace treaty. Each individual and group in the society has things it believes it is entitled to. Some of those claims are inconsistent with each other. Slave holders believe they are entitled to own and control their slaves, slaves (and some others) believe they are not. Libertarians believe that freedom of association gives everyone the right to refuse to hire, serve, rent to anyone he wishes on any grounds he wishes, including race. Liberals believe that everyone has the right not to be discriminated against on illegitimate grounds, such as race. A multitude of other examples, different for different societies at different times and places, could be offered.

Each person knows that he does not have sufficient resources to compel everyone else to grant him everything he believes he is entitled to. An actual society, a civil order, embodies a set of compromises, giving each participant enough of what he wants and believes he is entitled to get so that he does not find it worth trying, with his allies, to forcibly overthrow the system and substitute one closer to his desires. One of the things determining just what that set of compromises is will be the ability of the various factions to win a civil war if it happens, to successfully violate legal rules its members disagree with but others want and get, to win out if the order breaks down. In that sense, a mutual threat game underlies not only the stateless order I proposed in part III but every society.

Libertarians, going back at least to Lysander Spooner, find a social contract insufficient as a justification for state action. So do I. A peace treaty is a contract formed under duress; my moral intuition, like current law, fails to see such a contract as binding. If a mugger points a gun at me and offers me the choice of my money or my life it may be prudent to give him my money. But I have no morally binding obligation to do so, no obligation to tell him about the money that is not in my wallet but my money belt. Viewing the social contract as a peace treaty formed in response to the implicit or explicit threat of violent conflict explains what it is but does not give it moral force.

The implication of this analysis for my proposed stateless society is that rights enforcement agencies are producing two different sorts of products, only one of which was included in the analysis of part III of this book. That one is the service of enforcing rights, negotiating legal agreements, settling disputes. The other is the service of threatening other enforcement agencies in order to get the legal rules their customers prefer on the best possible terms—or, alternatively, to make sure that if they do not get them they get compensated for going along with the terms preferred by the customers of other agencies.

It follows that my reasons for believing that economies of scale in the industry will not be sufficient to produce very large firms are weaker than I thought. Even if economies of scale in rights enforcement and related activities are not sufficient to produce large firms, economies of scale in the ability to threaten other agencies might be. Judging by the historical record, economies of scale are common in that industry. God is often, as Napoleon is supposed to have said, on the side of the big battalions. If so, the equilibrium size of agencies might be large enough, the number few enough, to threaten the stability of the system, to create a risk that government will be reestablished by a cartel of rights enforcement agencies. Possibly a worse government than we started with.

This should be an issue mainly in the early stages of a stateless society, when the legal system is forming. Judging again by historical evidence, there is a lot of inertia in the equilibrium of mutual threat games; national borders do not shift out by a few miles every time a country launches a new battleship or increases its military budget. Once the system has established itself and legal rules have been agreed on between each pair of agencies, they should be reasonably stable against non-consensual change, due to the same forces that established stability in the two person society of Chapter 51 and the larger society of Chapter 52. An agency that insists on changing the rules without offering adequate compensation will be seen as an aggressor to be resisted even at substantial cost, enough of a danger to other agencies to generate a defensive coalition against it. The Schelling point is the status quo.

If so, enforcement agencies in an established stateless order should be primarily producing rights enforcement for their customers, which means that their efficient size will be mainly determined by the technology of that industry. With luck agencies will be sufficiently small and sufficiently numerous to make a cartel unlikely, bringing us back to the optimistic conclusion of my earlier discussion.

But they may still keep a few tanks in the basement, just in case.