Order Without the State: Theory, Evidence, and the Possible Future Of


This essay consists of four parts. Part I is an attempt to define government. Part II is a brief sketch of the evidence on whether and under what circumstances an anarchy, a society without a government, is workable. Part III provides a theoretical discussion intended to help make sense of the evidence, including a discussion of possible reasons for the nonexistence of an anarchy. Finally, part IV discusses under what circumstances and in what form stateless societies might come to exist over the next century or so.


I: What is Government, What is Anarchy?


An anarchy is a society without a government, so a discussion of anarchy requires a definition of government. Government cannot be defined by what it does, because all functions of government, including making and enforcing laws, have been, and most are, performed at some times and places by organizations that almost nobody would call governments.[1] This is a point I will discuss in more detail in parts II and III of this essay. So it makes sense to define government not by what functions it provides but by how it provides them.


Weber famously described the state as “a human community that (successfully) claims the monopoly of the legitimate use of physical force within a given territory.” If we omit the word “legitimate,” no states exist, since no state succeeds in eliminating all use of physical force by others, whether muggers in Chicago or the Mafia in Sicily. But including “legitimate” raises difficult problems. If we interpret “legitimate” as “legal” the definition is circular, since until we know what organization counts as the state we do not know what rules count as laws. Further, there are societies in which law is not viewed as the creation of the state at all–including all traditional Islamic societies–hence where some use of physical force by the state is seen as illegal and some use by non-state actors as legal. Defining “legitimate” in normative terms raises another set of problems.


My preferred solution is to define rights in terms of the set of mutually recognized commitment strategies by which individuals constrain how other individuals act towards them, and to then define a government as an institution with regard to which those strategies do not apply, an institution which can violate what individuals view as their rights with regard to other individuals without setting off the responses by which such rights are normally defended. For a detailed explanation of this view of rights, readers may wish to look at my “A Positive Account of Property Rights.”[2] What I offer here is only a sketch.


I start by considering the way in which property rights are enforced in (some) animal species, especially  many birds and fishes–territorial behavior. An individual in some way claims and marks a territory. If another individual of the same species and gender trespasses on the territory he is attacked by the owner, who will fight more and more desperately the further into the territory the trespasser comes. A fight to the death is usually a loss for both winner and loser, which gives a potential trespasser a good reason to retreat. Thus ownership is enforced by a mutually recognized commitment strategy.


Corresponding behavior by humans, generalized to apply to much more than territory, is observed in many different contexts. A famous experiment in behavioral economics demonstrated that individuals value items they possess much more than items they do not possess, even when the initial allocation of items was random.[3] In international relations, we routinely observe states willing to bear costs in defending their territory out of proportion to the value of the territory defended–the British war with Argentina over the Falklands being a striking modern example. And both casual observation and introspection tell us that while individuals defending what they see as their rights against other individuals are not willing to bear unlimited costs–not many of us are willing to fight to the death against an armed mugger–they are willing to bear costs out of proportion to the value directly threatened. Such behavior makes sense as a commitment strategy, since the knowledge that an individual will bear large costs in defending even small rights provides other individuals with a reason–often although not always an adequate reason–not to violate those rights, and as long as the right is not violated it is not necessary to bear large costs to defend it.


Starting with this definition of rights as based on mutually recognized commitment strategies, we may define a state or government–I will use the terms interchangeably–as an organization against which those commitment strategies do not hold. Most of us are willing to bear substantial costs defending ourselves from a private robber or an attempt by another individual to enslave us, even briefly. Few of us are willing to bear similar costs to defend ourselves from a tax collector or mandatory jury service. Even those who consider taxation morally illegitimate and so have no reservations about concealing taxable income when they believe that doing so is in their interest feel no commitment to do so when it is not.


An anarchy, then, is a society in which there is no such organization, in which the commitment strategies by which we defend our rights, whatever we and other members of that society consider those rights to be, apply to all other individuals.


II. The Evidence


Anarchies, stateless societies, exist in the historical record in two forms. The best known are primitive societies, of which the Comanche indians provide a good example.[4] They functioned for well over a century, they succeeded surprisingly well in defending themselves (and attacking others) despite their eventual defeat by a vastly superior enemy, they had reasonably well defined rules of behavior enforced by private action, a system of private norms entirely substituting for a legal system.


The other sort of stateless societies are embedded societies–subcultures that succeed in enforcing their own rules on their own members while to varying degrees obeying or evading the rules of the state within which they are located. Modern examples include the Rominchal gypsies in England[5] and the inhabitants of modern-day Shasta County California.[6] In these cases again, the equivalent of legal rules are enforced by decentralized private action, with no specially privileged enforcer.


In addition to stateless societies, primitive or embedded, we also observe many examples, ancient and modern, of private arrangement substituting for what are usually viewed as core functions of the state. Both the making and the enforcement of law, for instance, exist in private as well as public forms. Commercial arbitration has been and is used to resolve a considerable proportion of legal disputes, both in contexts where it does and in others where it does not depend on enforcement by the state.[7] Contracts create legal rules between the contracting parties. Protection against crime is provided legally by Brinks guards and the manufacturers of burglar alarms, illegally–but in a form much closer to the form provided by the state–by the Sicilian Mafia.[8] As recently as the 19th century, much of the circulating money of both Scotland and the U.S. consisted of privately issued banknotes.


What we do not observe are modern anarchies in the sense of modern, developed, fully independent stateless societies. That fact is, in my view, the strongest argument against the practicality of a stateless society replacing a modern state. To see why the argument, while strong, does not entirely rule out the possibility of future non-primitive anarchies, it is worth considering a somewhat different set of hypothetical institutions from a perspective two centuries in our past


Suppose that in 1800 some political theorist, inspired perhaps by the French and American revolutions, proposes a new and different form of political and social organization. Its features include not merely a democratic government with all adult male citizens having the right to vote, already an unusual arrangement, but voting rights for women as well, along with complete legal equality between men and women. Its government taxes and spends between a third and a half of all income. The polity has no state supported religion, considerable diversity of religious beliefs, and a sizable part of the population believing in no religion at all. To add additional, if implausible, interest to this imaginary creation, it has no serious restrictions on the sexual behavior of adult citizens, no legal and few social penalties for bastardy, and little more than token enforcement of either the stability or fidelity of marriage.


A well informed 18th century scholar might be able to find precedents for one or another of the features of this bizarre hypothetical–whether utopia or distopia could be a matter of dispute. He would, I think, have difficulty citing any past society that combined very many of them.[9] He would thus have about as good reason to reject as impossible what is now the dominant social and political system of the developed world as a modern scholar would have to reject the possibility of a future stateless society.


III: Theory


In constructing a theoretical picture of a modern stateless society and using that theory to explore the possibility of future stateless societies, one faces two sets of problems that such a society must deal with. The first is how a legal framework for trade, contracts, human interaction could exist without a state to maintain it, and the associated stability problem–how such a framework could be prevented from collapsing into either chaos in one direction or a state in the other. The second is how, without a state, various useful functions of current states could be provided–a monetary system, protection of rights, the production of public goods, including defense against foreign states.


Framework and Stability


A legal framework has to solve the problem of creating and enforcing legal rules that will bind both parties to a dispute. I sketched one possible solution over thirty years ago.[10] In the system I then described, each individual is the customer of a private rights enforcement agency which, in exchange for an annual fee, agrees to arrange for his disputes with other individuals to be settled and the verdicts enforced. It does so by contracting with every other such agency whose customers have any significant chance of interacting with its customers to specify, for each pair of agencies, a private court whose verdicts both agencies agree to accept and enforce.


In the absence of a sovereign, what enforces the contracts between the agencies? The answer is that they are enforced by reputation and the discipline of repeat dealings. An agency that adopts a policy of only abiding by court verdicts when they favor its customers will soon find no other agency willing to go to court with it. Since violence is both more expensive and less predictable in its outcome than litigation, such an agency finds itself at a serious competitive disadvantage.[11]


In such a system, law is produced on the market by private courts, each of which seeks to create legal rules that the customers of the enforcement agencies will wish to live under.[12] I have argued elsewhere[13] that the incentives created by that system would lead to a more attractive legal system than what is or can be expected to be created by the incentives of those–judges and legislators–who create current law. My purpose here is more modest–merely to establish the possibility of an enforced legal framework without a state.


Would such a framework be stable? That depends in large part on whether the equilibrium of the market I have described contains few or many agencies. That, in turn, depends on the size of the market and the extent of economies of scale in the business of settling disputes and enforcing rights. If, in equilibrium, there are only two or three rights enforcement agencies, there is an obvious temptation either for them to merge into a government in order to improve their ability to exploit their customers or for the situation to devolve into civil war as one or more concludes that it can establish a monopoly by force.[14] If, in equilibrium, there are hundred such agencies in a territory similar to that of a modern state, with several sufficiently local to be available to the typical customer, then neither approach is likely to be workable.


What determines economies of scale in rights enforcement? The obvious answer is that, as in other industries, larger firms have both advantages, such as greater internal division of labor, and disadvantages, such as the increased difficulty of knowing and controlling what one’s subordinates are doing as the number of layers between the corporate president and the individual worker increases. Casual observation suggests that large police forces have no obvious advantage in cost of operation or quality of service than small ones. Perhaps more relevantly, the Sicilian Mafia, whose chief line of business is private rights enforcement, is made up of many small firms, not a few large ones.[15]


There is one special factor in the rights enforcement industry, however. Consider two firms with somewhat different customer bases, bargaining over what court’s legal system to agree on. Firm A prefers one legal system, say one that permits capital punishment for murder. Firm B prefers a different system, one that does not permit it.


Each firm estimates the value to its customers, and from that the increase it can expect in its revenues, if it can provide them with its preferred legal system. We expect, along conventional Coaseian lines, that they will agree on the system that maximizes their combined benefit. But while this provides a solution to the allocational problem–what legal rule will exist–it leaves open the distributional problem. If it turns out that a system with capital punishment is optimal for the pair of firms, is it achieved by firm A paying firm B to agree to it or by firm A rejecting firm B’s offer to pay it to accept the alternative? Put differently, what is the starting point from which the two firms bargain?[16]


The theoretical answer, I think, is that while allocation comes out of a bargaining game leading to an efficient legal system, distribution comes out of a prior mutual threat game among the rights enforcement agencies. In such a game, one relevant factor is an agency’s ability, if bargaining fails, to employ force. Under many, although not all, military technologies, there are considerable economies of scale in warfare–large armies can not only defeat small armies, they can do so while inflicting larger costs than they receive. Hence if the ability to bully other agencies, or resist bullying by other agencies, is a major part of what determines how well a rights enforcement agency functions, that might lead to substantial economies of scale which might make the system unstable.


My guess is that the ability to use force against other agencies will not be an important element of what rights enforcement agencies produce in a funcitoning society of this sort, mostly because observation suggests a great deal of inertia in such mutual threat games. Consider the obvious example of international relations, which one could argue have very much the structure I have described. We do not observe that national boundaries shift a mile or two in one direction or the other every time one country launches a new battleship or raises or lowers its military budget. Nor do we observe that smaller nations routinely pay tribute to larger nations.


If I am correct then, once the sort of system I have described is well established, the distributional outcome will be determined not by a current mutual threat game based on current abilities to employ force but by the result of such a game played out in the distant past. Given a starting point, the obvious equilibrium is one in which any bargaining among agencies takes the continuance of the existing terms as the default. An agency that tries to use the threat of military force to unilaterally modify the rules in its favor is then seen as a threat to the other agencies which, in total, greatly outnumber it.[17] The system should be stable, provided that the economies of scale in the business of rights enforcement itself do not go far enough to produce a market equilibrium with only a few firms.


Even if economies of scale in the application of force are not an important issue in an existing anarchy, however, they may provide a barrier preventing such an anarchy from coming into existence, since at that point threats of force are likely to play a much larger role.


Private Substitutes for Public Production


At a theoretical level, the problem of replacing useful government services with privately produced services comes down to the problem of market failure–of situations where individual rationality fails to lead to group rationality.[18] Familiar examples are the problem of producing a public good, such as national defense, basic scientific research or a radio broadcast, and the problem of dealing with externalities. It is often argued that since market failure means that some desirable goods and services will not be produced on the private market, they must be produced instead by the state. This argument purports to both explain the existence of the state and justify it.


There are two responses. The first is that, in many cases, an ingenious entrepreneur can find a profitable partial solution to the problem of market failure. A radio or television broadcast, for example, is a pure public good in the conventional sense–the producer cannot control who gets it and consumption by one individual does not interfere with consumption by another. Nonetheless such broadcasts are routinely produced privately, paid for by the ingenious trick of combining a public good with positive cost of production and positive value to the consumer with a public good with negative cost of production and negative value and giving away the package: program plus advertising. Basic research may be produced either as a source of valuable reputation for scientists, universities and firms or because those producing it will be able to appropriate some of the benefits through first mover advantages–the scientists who develop a new idea will understand it better and be better able to exploit it than others. This is a point that has been explored at greater lengthy by Terence Kealey.[19] It follows that while market failure on private markets will lead to suboptimal production, it may not, and frequently does not, lead to the complete failure to produce. There is some inefficiency but usually less than one would at first expect, since one’s first thoughts are unlikely to generate all of the possible ways of overcoming market failure. The classic example is provided by Cheung’s “The Fable of the Bees,”[20] which demonstrated that a particular problem of market failure which J. E. Meade, a prominent economist who later won a Nobel prize, had offered as an example of the sort that could not be privately dealt with, had been being privately dealt with on a routine basis for many decades before Meade discovered that doing so was impossible.


Other private solutions occur through voluntary non-commercial arrangements. One example would be systems of private norms.[21] Another would be open source software such as the Linux operating system. In each case, even though individuals are not in a position to appropriate the full value of what they do, they are in a position to get sufficient benefit from doing things to make it in their interest to do them.


The first response to the problem of market failure is that although it is a problem on private markets, it need not be a catastrophic problem. The second response is that although market failure is indeed a problem, government is not  a solution–because market failure is not limited to private markets. Market failure arises in situations where one individual’s acts produce costs for others that they need not assent to or benefits for which they cannot be charged. Such situations exist in private markets but are endemic in political markets. A voter who spends time and effort figuring out what candidate will best serve the public interest is producing a public good for a very large public–and because such goods are underproduced, rational voters are rationally ignorant. Similarly for legislator, lobbyist, and judge. Each is making decisions most of whose costs and benefits go to others, hence most have little incentive to make the optimal decision. So shifting production from political to private markets is likely to reduce, not increase, problems due to with market failure.[22]


While this provides an argument against the general superiority of states to anarchies, there remains the possibility of narrower arguments on the other side. The obvious one is the problem of national defense. Defense against nations is a public good, it is successfully–although perhaps inefficiently and/or suboptimally–produced by many existing states, and it is a public good whose inadequate provision may be catastrophic, since it can lead to the disappearance of the inadequately protected society.


Like other public goods, national defense may be privately provided. Under current institutions, it is common for supporters of a threatened state to volunteer for military service on unattractive terms, buy war bonds for patriotic rather than financial reasons, contribute privately owned weapons if there is an inadequate supply of publicly owned ones. In a stateless society, one can imagine a variety of other mechanisms by which some level of defense might be privately produced.[23] How successful such mechanisms would be depends in part on the society, in particular its level of patriotism–if the term can be applied to a patria without a state–and on the size of the external threat it must defend against relative to its own resources. It is easier to imagine a stateless U.S. defending itself against Mexico today than a stateless France defending itself against Germany in 1914.


Before leaving the subject of replacing services provided by government, it is worth noting that not all of the things governments do, not even all of the things that practically everyone expects governments to do, can be plausibly represented as the production of pubic goods or solutions to other forms of market failure. The protection of an individual against criminals, for instance, is not a public good; a police force can decide who is or is not to be protected, and in practice police forces often do. Similarly a private agency can, and private agencies do, protect those who pay them and not those who do not. Like many private activities, private protection against crime may produce externalities, positive or negative–my protection agency may incapacitate the criminal whose next offense would have been against you or it may deter him into attacking you instead of me. But as long as the external costs and benefits are not too large compared to the internal ones, we can expect private arrangements to produce something close to the optimal result–closer than we have any good reason to expect from political production.


A monetary system is not a public good either; currency, like other useful commodities, can be privately produced and exchanged. For a hundred and fifty years in the 18th and 19th centuries Scotland maintained a private monetary system using banknotes issued by competing private banks. Insofar as circumstances have changed since then, the changes probably make such a system more workable, not less, since our technologies for communicating the information needed to keep track of the reliability of money issuers are very much better than those available to the contemporaries of Adam Smith.[24]


This brief discussion of the possibilities for a stateless society suggests two possible reasons why such societies could not exist in the modern world:


1. Under current circumstances, economies of scale in the rights enforcement industry go to a size sufficiently large to make the system unstable.


2. Under current circumstances, a stateless society is unable to adequately produce the public good of defense against nations.


It also suggests several possible reasons why they do not exist but could.


3. Economies of scale are large when such a system is coming into existence, for reasons discussed above, but much smaller once such a system has been for some time established.


4. A stateless society in the process of forming cannot defend itself, but an established one can, at least if it has the good fortune not to be threatened by powerful and aggressive neighbors.


5. More generally, it might be that the system is stable in the sense of Nash equilibrium, but that it does not come into existence for much the same reason that English drivers do not shift to driving on the right side of the road, despite the obvious advantages of doing so: the move to a new equilibrium requires coordinated change, and so is a mistake for anyone who does it if others do not.


In The Moral Basis of a Backward Society, Edward Banfield compared two towns, one in Italy and one in the U.S., and argued that they differed in the sets of mutually consistent strategies followed in each. Each set–betray and don’t trust, trust and don’t betray–was self reinforcing, making both alternatives stable. Similarly, given the set of existing commitment strategies followed by the inhabitants of existing states, there may be no simple way of shifting to the different set in which prevention against rights violation is produced by voluntary interaction among peers rather than by an involuntary relation between sovereign and subjects.


6. Stateless societies do not come into existence for ideological reasons–because too many people believe in the necessity of the state.


If 1 or 2 is the case, then we will not get stateless societies unless and until social and/or technological change makes it no longer the case. If 3, 4, 5 or 6 is the case, we might get such societies if a pathway exists, or comes to exist, for moving to them from where we now are.


The Future


In this section I will explore three different possibilities for achieving a future stateless society. The first assumes that the problem is not in maintaining such a society but in bringing it into existence, and considers how that problem might be solved. The second describes how an embedded anarchy within modern societies might come into existence–arguably is coming into existence–and might eventually become the dominant form of organization. The third considers possible technological changes that might reduce or eliminate factors that currently prevent stateless orders.


Fabian Anarchy


In The Wealth of Nations, published in 1776, Adam Smith wrote that “To expect, indeed, that the freedom of trade should ever be entirely restored in Great Britain is as absurd as to expect that an Oceana or Utopia should ever be established in it.” Within less than seventy years, due in part to Smith’s influence, England had moved to free trade.


Currently, the governments of developed countries are large, actively interventionist, and democratic. The degree of intervention has varied over the past fifty years in both directions–increased intervention in land use due largely to environentalist concerns, reduced regulation of transport industries both in the U.S. and elsewhere. One factor causing such changes is the opinion of academics, through their influence on both public and expert opinion, another is the public view, based in part on experience, of the relative advantages of government and private production.


One can imagine a future in which the prevalent views of some such societies, like those in England between 1776 and 1840, shift steadily in a direction opposed to government involvement. Consider schooling, currently the largest activity of state and local governments in the U.S. A full blown voucher system in which schooling subsidized by the state could be provided privately is well within the bounds of political possibility, although it has not yet come into existence. If such a system were introduced in one state and its success lets to its introduction in others, it might, over a period of a few decades, virtually eliminate public schooling. Once most parents were sending their children to privately run schools, there would be a strong argument for restricting public subsidy to poor parents, thus eliminating most of the government role in schooling.


One can imagine similar developments in other areas, including the replacement of much of the existing system of police and courts by private arrangements–either private arrangements sanctioned by the state, such as current arbitration and the private police forces maintained by many colleges, universities, and gated communities, or unsanctioned arrangements as in modern Sicily.


Following out this line of development, one gets a future, sometime in the next century, in which government still exists but controls relatively little, with the de facto legal rules coming out of agreements among a network of private enforcers and arbitrators as descriped in the previous section of this essay. At some point in that future, government, whether or not it still exists on paper, no longer exists in fact. For an entertaining, if not very serious, picture of such a future, see Neal Stephenson’s novel Snowcrash.


Crypto-Anarchy Online


An increasing part of human interaction, social and commercial, has shifted over the past few decades from realspace to cyberspace. Travel agents have been virtually eliminated by online competitors such as Travelocity.com, bookstores faced with the competition of Amazon.com, movies, amusement parks, television and other forms of entertainment in part replaced by massively multiplayer online role playing games such as World of Warcraft, currently played by something over ten million people.


One attractive feature of the online world is that, within it, physical coercion is impossible. One can fool people or their computers, but one cannot get a bullet through a T-1 line. It is in that sense a world of entirely voluntary interaction. In order to use force against an online competitor or opponent, it is necessary to shift the conflict to realspace–sue him, assassinate him, arrest him.


Current encryption technology provides the tools with which it is possible to almost entirely eliminate that option. As pointed out long ago by computer scientist and sf author Verner Vinge,[25] anonymity is the ultimate defense–you cannot arrest someone if you do not know who he is. Public key encryption makes possible what I have described elsewhere as a world of strong privacy, a world in which it is possible to communicate with strangers in a form that no third party observer can read, establish one’s online identity in a way that no imposter can imitate, link or separate one’s online persona from one’s realspace persona as desired, make online payments with untraceable anonymous digital cash.[26] While the full infracture required for such a world does not yet exist, there is no obvious reason why it cannot come into existence over the next few decades.


Within such a world, government can use force only against those who, by revealing their realspace identity, choose to make themselves vulnerable to it. It does not follow that there will be no ways of enforcing rules. Within a proprietary community such as World of Warcraft or Second Life, the proprietors can make rules and enforce them against those who choose to enter that world.[27] Outside of that framework, the technologies that make possible strong privacy also make possible forms of reputational enforcement of contract that can substitute for existing state enforcement.[28]


It thus seems possible, even likely, that within a decade or two large parts of human life in developed societies will become stateless, a system of voluntary transactions enforced entirely by competing private entities using reputational rather than coercive means of enforcement. At that point, what we will have will be an embedded anarchy available to much, perhaps almost all, of the population of existing developed states. As interaction through virtual reality becomes better, a process already occurring, the amount of life that can be effectively lived in cyberspace expands.


The larger the role of online anarchy in individual life, the weaker the power of realspace states. If most of my social and professional life occurs online, I can move from one state to another at little cost. If I am a valuable citizen–one whose taxes more than pay for the costs he imposes–some states will have an incentive to make an effort to attract me. Competition among states for citizens drives down the ability of states to exploit their citizens. At the same time, a world of strong privacy almost entirely eliminates the state’s ability to control information, since information can be produced, exchanged, bought and sold online where the state cannot observe transactions. And if a large fraction of commercial transactions shift to a world of online anonymity, the result is to substantially reduce the state’s ability to collect taxes.


This suggests a second possible scenario for a continual shift towards a stateless future, driven not by beliefs but by technology.


Other Technological Possibilities


There are other technological developments that might, or might not, make realspace anarchy more likely. Consider, for instance, the division in present society between tort law and criminal law. Both depend on state enforcement of verdicts, but only in the case of criminal law does the state control the process that leads to the verdict; tort cases are privately prosecuted and can be privately settled. A shift from criminal law to tort law, a reversal of the legal trend of the past thousand years or so, would sharply reduce the role of the state in law enforcement and so make easier a transition to a fully private system.


The question of why some offenses are handled by criminal law and some by tort law is a difficult one,[29] but one factor may be the cost of catching and convicting criminals. If it is larger than the amount of damages the loser of a tort case will pay, whether because of limits to his legal liability or limits to his resources, it may not be in the interest of the victim to sue.[30] If that is the problem, then improved surveillance technology, a change currently occurring,[31] may solve it. In a transparent society where everything that happens in public places, perhaps in private places as well, is observed, recorded, and searchable, most legal disputes may be reduced to disputes over the legal implications of known facts, thus eliminating a large part of the cost of current enforcement of law.


For a different sort of change, consider changes in the technology of violence which might alter the balance between offense and defense or the pattern of economies and diseconomies of scale in the production of military force. Or consider the way in which improved communications and transport may make national boundaries increasingly irrelevant, forcing an increasing fraction of transactions, including legal disputes, into non-state mechanisms. Or consider the effects of increasing levels of wealth and education. It is at least arguable that modern democracies don’t attack each other. If so, and if in the future most power is held by democracies, and if the reasons for their restraint apply to attacks on developed stateless societies as well, the problem of national defense may become less critical in the future.


Finally, consider the possibility of seasteading[32]–floating housing competitive with conventional housing, located either on the high seas or in the territorial waters of some friendly state. Like a shift of life to cyberspace, it increases mobility and thus increases the degree to which states must compete for citizens, weakening the power of the state. And it offers the possibility of floating stateless societies, with secession the ultimate, and easy, sanction.


None of these provide confident grounds to predict stateless societies in the next century–cyber-anarchy is probably the closest–but they suggest a variety of possible ways in which such societies might come into existence.




David Friedman



School of Law

Santa Clara University

Santa Clara, CA 95117

[1] “Sociologically the state cannot be defined in terms of its ends. There is scarcely any task that some political association has not taken in hand, and there is no task that one could say has always been exclusive and peculiar to those associations which are designated as political ones: today the state, or historically, those associations which have been the predecessors of the modern state.” Weber

[2] “A Positive Account of Property Rights,” Social Philosophy and Policy 11 No. 2 (Summer 1994) pp. 1-16. Webbed at: http://www.daviddfriedman.com/Academic/Property/Property.html.



[3] Kahneman, D., J. Knetsch and R. Thaler, 1991, “Experimental Tests of the Endowment Effect and the Coase Theorem,” Journal of Political Economy, 98, 1990, pp:1325-1348

[4] E. Adamson Hoebel, The Law of Primitive Man, Chapter 7.

[5] Walter O. Weyrauch, Gypsy Law: Romani Legal Traditions and Culture, Chapter 3.

[6] Described by Robert Ellickson, Order Without Law.

[7] Lisa Bernstein, “Opting Out of the Legal System: Extralegal Contractual Relations in the Diamond Industry," 21 Journal of Legal Studies 115 (1992). Benson, B. L. 1998. “Law Merchant,” In P. Newman, (ed.). The New Palgrave Dictionary of Economics and the Law, London: Macmillan Press.

[8] Diego Gambetta, The Sicilian Mafia: The Business of Private Protection.

[9] Readers are invited to offer counterexamples.

[10] David Friedman, The Machinery of Freedom: Guide to a Radical Capitalism, part III.

[11] As I have pointed out elsewhere, we can observe a strategically equivalent game being played out in the real world. Just as a rights enforcement agency may deal with a conflict between its customer and the customer of another agency by either inexpensive agreement or expensive violence, so an auto insurance company may deal with an accident between its customer and the customer of a competitor by either inexpensive agreement or expensive litigation. Almost all such cases are settled outside of the courts.

[12] I am describing a polylegal system–the legal rules between A and B need not be the same as between A and C or B and D. The same is true, although less obviously, in our legal system, where the outcome of a legal dispute may depend in part on what state’s courts or what federal district it is settled in, which itself may depend on the state citizenship of the parties, the physical location of the events that led to the dispute, and similar factors. It was more obviously true in medieval legal systems such as those of the Islamic world, where the outcome of a case might depend on the religion or ethnicity of the parties or on which of the four mutually orthodox Sunni legal schools they adhered to.

[13] Friedman, David, The Machinery of Freedom Chapters 29, 31, 32; also Friedman, David, “Anarchy and Efficient Law” in For and Against the State, edited by Jack Sanders and Jan Narveson, 1996.

[14] Arguably this is what eventually happened to the semi-stateless society of saga period Iceland during the Sturlung period. Friedman, David, “Private Creation and Enforcement of Law — A Historical Case,” Journal of Legal Studies, (March 1979), pp. 399-415.

[15] Gambetts, The Sicilian Mafia. Some of the rights enforced by the Mafia are ones many of us disapprove of, such as the right to be protected from competition. But that is true of many of the rights enforced by states as well–arguably of more of them. Both my pieces of evidence involve rights enforcement by organizations with some degree of territorial monopoly, so raise the possibility of natural monopoly within a territory even if optimal territorial size is small. How serious a problem that would be for the sort of system I describe largely depends on how easily individual customers can change their company by changing their residence.

[16] This point was initially raised by James Buchanan in his (1974) "Review of The Machinery of Freedom: Guide to a Radical Capitalism', Journal of Economic Literature, 12: 914-15.

[17] All of this is presented in terms of plausiblity arguments, not proofs, since game theory, in its present state of development, is strikingly short of convincing ways of proving what the result of strategic interactions will be.

[18] For discussions of the general problem see D. Friedman, Hidden Order: The Economics of Everyday Life, Chapter 18, and the corresponding chapter of D. Friedman, Price Theory: An Intermediate Text, webbed at:


[19] Terence Kealey, The Scientific Laws of Scientific Research.

[20] Steven N. S. Cheung,  “The Fable of the Bees: An Economic Investigation,”  Journal of Law and Economics, Vol. 16, No. 1 (Apr., 1973), pp. 11-33.

[21] Eric A. Posner,  Law and Social Norms, Robert Ellickson, Order Without Law.

[22] For a more extensive discussion of this issue, see Friedman, David, Hidden Order, Chapter 19, or the corresponding chapter of Price Theory, webbed at http://www.daviddfriedman.com/Academic/Price_Theory/PThy_Chapter_19/PThy_Chap_19.html.

[23] David Friedman, The Machinery of Freedom, Chapter 34. See also the suggestions in the piece webbed at http://www.daviddfriedman.com/Academic/mps_iceland_talk/Iceland%20MP%20talk.htm

[24] Readers interested in a historical discussion should look at Lawrence White, Free Banking in Britain. For theoretical discussions see Friedman, David, “Gold, Paper, Or...Is There a Better Money?” Cato Policy Analysis No. 17 (1982), webbed at http://www.cato.org/pubs/pas/pa017.html, and Friedman, David and Macintosh, Kerry, “Technology and the Case for Free Banking” in The Half-Life of Policy Rationales: How New Technology Affects Old Policy Issues, edited by Fred Foldvary and Daniel Klein, New York University Press, New York, 2003.


[25] In the short story “True Names,” in True Names … and Other Dangers. Riverdale: Baen.

[26] For details see Friedman, David, "A World of Strong Privacy: Perils and Promises of Encryption," in Scientific Innovation, Philosophy, and Public Policy, edited by Ellen Frankel Paul, Fred D. Miller, Jr., and Jeffrey Paul, Cambridge University Press 1996 (webbed at http://www.daviddfriedman.com/Academic/Strong_Privacy/Strong_Privacy.html). Chapter 3 of my forthcoming Future Imperfect (a draft is webbed at http://www.daviddfriedman.com/Future_Imperfect.html), my  “Privacy and Technology” in The Right to Privacy, edited by Ellen Frankel Paul, Fred D. Miller, Jr., and Jeffrey Paul, Cambridge University Press, 2000 (http://www.daviddfriedman.com/Academic/Privacy%20and%20Technology.html).

[27] Thus a world with many such online proprietory communities looks rather like the Utopia described by Robert Nozick in Anarchy, State, and Utopia–an environment with a  wide variety of different social systems among which each individual is free to choose.

[28] For details, see my "Contracts in Cyberspace," Journal of Internet Law, December 2002. Webbed at: http://www.daviddfriedman.com/Academic/contracts_in_%20cyberspace/contracts_in_cyberspace.htm.

[29] For an extended discussion see my Law’s Order, Chapter 18, “The Crime/Tort Puzzle.”

[30] One possible solution to this problem is to pre-commit as a way of producing private deterrence; see Friedman, David, “Making Sense of English Law Enforcement in the Eighteenth Century,” The University of Chicago Law School Roundtable (Spring/Summer 1995): 475-505.

[31] Brin, David, The Transparent Society. See also the discussion in Friedman, David, Future Imperfect (forthcoming from Cambridge University Press).

[32] For more information, see http://seastead.org/. For a discussion of the high seas as a presently existing stateless order, see The Outlaw Sea: A World of Freedom, Chaos and Crime by William Langewiesche.