The Three Leviathans
Two years ago, at our Spring 1994 Forum on Systems of Law, I suggested that those seeking to build and maintain a Free Nation would face three problems, which I called “the three Leviathans”:
“Leviathan Past (that is, the dangers posed by the state presently occupying the territory within which the Free Nation is to arise), Leviathan Present (that is, the dangers posed, once the Free Nation has arisen, by the threat of other states existing outside the Free Nation’s territory), and Leviathan Yet to Come (that is, the dangers posed by the unwanted but all-too-possible emergence of a state within the Free Nation’s territory — that is, the possible evolution of the … Free Nation into a [statist régime].” 
With regard to Leviathan Past, I noted that there are two ways of getting an existing régime to give up its power and turn libertarian: force and persuasion. Arguing that force was impractical, I suggested three possible modes of persuasion: a) convert the rulers of a country to libertarianism (a daunting prospect); b) convert the ordinary citizens to libertarianism and get them to vote in a libertarian system; and c) pay the rulers to relinquish sovereignty over some portion of their territory. More recently, I have returned to the topic of Leviathan Past, and argued that the vestiges of the old régime in a fledgling libertarian society might be successfully bought off and so discouraged from engaging in obstructionism. 
With regard to Leviathan Present, I have since argued that this threat can be met through a combination of voluntary contributions and for-profit defense agencies. ,
What about Leviathan Yet to Come? In my initial Forum presentation I argued that this threat was minimal. My argument relied on traditional libertarian class analysis, which maintains that a ruling class cannot achieve or maintain power except through the mechanism of the state. Thus a ruling class is the product of governmental institutions and not vice versa.
More recently, however, I have rethought this point. In fact, ruling classes have managed throughout human history to survive and dominate even in stateless or near-stateless societies, through a combination of economic patronage and religious status.  A state may dramatically increase the power of a ruling class, but it is not an absolute prerequisite for the existence of such a class. Still, I’ve argued, there are reasons for optimism: modern society cannot sustain the kind of religious climate needed for the moral support of a ruling class, , and a genuine free market would eliminate the kind of economic dependence that makes people vulnerable to intimidation by the wealthy. , I’ve also argued that attempts by some groups to form cooperative schemes to oppress other groups are likely to backfire in a completely free market. 
So there are reasons for optimism. But there are also reasons for caution. We know from history that the Leviathan virus is a robust one. If it finds a niche, it will cling, and grow. Eternal vigilance is the price of liberty. It is appropriate, then, for those who seek to build a Free Nation to return again and again to the question of how to stay free.
Why Do People Want Leviathan?
Statist régimes exist because people want them. This is not to say that such régimes arise, and maintain themselves in existence, solely through the deliberate choices of individuals. On the contrary, the growth of government is often a spontaneous and quite unintended side effect of human actions pursuing quite different goals. Nevertheless, if the result were entirely unwelcome I do not think it would long survive. Why, then, do people so often desire a powerful coercive state?
There are any number of reasons, of course. Let me mention just four: two reasons that depend on intellectual mistakes, and another two that depend on more intractable facts about human psychology.
One intellectual mistake is the idea that certain desirable goals can be achieved only through coercive authority. Those on the left do not see how people could be protected from poverty, pollution, or discrimination except through the benevolent arm of the state; likewise, those on the right see governmental intervention as a prerequisite for the preservation of moral and cultural values. People who actually lived in a Free Nation, however, would see these benefits being provided without any government involvement; so this intellectual mistake would be less likely to occur, once freedom was achieved.
A second intellectual mistake underlying statism is bit more slippery, however. Statists generally think that one’s willingness to enforce a moral claim is a mark of the importance one attaches to that claim. Even if a free-market system does an excellent job of supplying food to the poor, the fact that such a system does not recognize any right to be fed shows its insensitivity to the importance of hunger relief. When libertarians claim that we have a right to drive a motorcycle without a helmet, but no right to be protected from starvation or discrimination, the statists conclude (not without justification, in the case of some libertarians!) that libertarians regard freedom from helmets as more important than freedom from starvation or discrimination. And so the statists, reasonably enough given their premises, dismiss the libertarian position as absurd.
This motivation for statism would not necessarily vanish simply as a result of the statists’ seeing that the needs they regard as important are indeed met in a Free Nation. They would still insist that the importance of these needs be underscored by having their provision enforced by an agency speaking for the community as a whole. (Even libertarians have been known to succumb to this sort of reasoning; for example, it was on the basis of considerations rather like these that Robert Nozick was led, in The Examined Life, to repudiate the libertarian position he had defended in Anarchy, State, and Utopia.) 
The only cure for this mistake is education. The statist must be brought to see that the libertarian position on the use of force is based on reciprocity, not on assessments of importance. Evils involving force may legitimately be fought by means of force; evils not involving force must be fought by other means. (It would certainly help matters if libertarians themselves would refrain from speaking as though coercion were a more serious evil than any other. Stealing a grape is an act of force, persistent emotional and psychological abuse is not, but the latter is a far greater evil than the former.) To some extent, though, this second mistake is supported by the first, in that statists may find the libertarian position more plausible when they come to recognize that there are effective non-coercive ways of fighting evils.
But the hankering after Leviathan also rests on two psychological factors that cannot be so easily eradicated: the desire to control, and the desire to be controlled.
The desire to control can take the form of a love of power for power’s own sake. But it needn’t. People also seek power as a means to other ends. Whenever we seek some goal that requires the cooperation of others, and those others refuse to cooperate, there is an opening for the temptation to force them to cooperate. And when many people succumb to the temptation to compel the cooperation of many other people, we are well on our way to statism. This is probably a permanent aspect of the human condition. The best we can do is:
a) provide a moral climate in which this temptation will be easy to resist, by teaching people to regard it as shameful and ignoble to live by violence rather than persuasion;
b) point out, as well, that any attempt to establish a state is likely to backfire, as people with aims other than one’s own may end up holding the reins of power; and
c) for those who prove impervious to moral suasion and politico-economic analysis alike, make it clear to such people that their attempts to control others will not be tolerated.
The desire to govern is easy enough to understand, to the extent that it is the byproduct of a more general desire to see one’s ends fulfilled. The desire to be governed is more puzzling. How could such a desire arise?
I suspect that the desire to be governed is the result of an evolutionary trade-off. Animals at the bottom of the evolutionary ladder (insects, fish, and the like) operate almost entirely on instinct. Very little of their behavior is learned; for the most part it is encoded in their genes, and passed from one generation to another through biological reproduction. As we pass to more advanced species, however, we find the ratio of learned to instinctual behavior steadily increasing, until we reach human beings, whose ability to learn is tremendous — and whose repertoire of instinctive behavior is minimal.
Reliance on learning rather than instinct makes for a more flexible and versatile organism; when environmental conditions change, animals whose behavior is not pre-programmed can adapt more quickly. Moreover, animals with the capacity to learn can acquire new, successful behavioral strategies by imitating one another. They do not have to wait for a random mutation in order to acquire the new behavior; and in addition, knowledge can now be transmitted to all the other members of one’s species, not solely one’s immediate descendants; cultural reproduction is thus more efficient than biological reproduction. Species that rely heavily on instinct, by contrast, are less flexible, and so rely on quantity rather than quality in their reproductive strategies; rather than raising just one or two offspring and investing time and effort teaching them the skills needed to survive, as occurs in the higher animals, the lower animals produce hundreds of offspring at once, and turn them loose with little or no guidance.
So we’re lucky to be human. Hooray for us. There is a downside to all this, though. Because we depend so heavily on learning and are so ill-equipped with instincts, it takes us a long time to acquire the ability to survive on our own. Many insects begin life entirely alone, with the parents long since dead or flown off; the insect knows instinctually how to go about surviving. As we climb the evolutionary ladder, dependence on parenting increases; but even here we find, e.g., that colts can stand and walk, albeit shakily, from the day they are born. Human beings, because of our heavy learning-to-instinct ratio — that ratio that is our glory, that makes us what we are — also have the longest childhoods, the longest period of dependence. Thus we (master learners that we are) learn early that we need someone to take care of us, to make our decisions for us. And what we learn earliest is the most deeply ingrained, the most difficult to unlearn. Hence the desire for the State, as a replacement for the Parent.
So human beings have, all too often, a desire to be controlled. Fortunately, we also have a desire not to be controlled, a desire to make our own decisions; and this desire manifests itself very early as well. These two desires conflict, and circumstances may influence the outcome of the conflict. Here too, then, a crucial role of education is to reinforce our libertarian impulses and to discourage our statist impulses. (And to the extent that the desire to be controlled cannot be suppressed, perhaps it can be channeled into less destructive manifestations; e.g., this desire might be satisfied by joining a religious cult rather than setting up a powerful government.)
As for the intellectual mistakes that support the yearning for Leviathan, perhaps the most important task for educators in a Free Nation (whether parents, teachers, or what have you) is to help people learn how to think straight. I recently had an opportunity to reread the Constitution of North Carolina, and I was struck by the difference, not so much in content as in language, between the original 18th-century provisions and the more recent amendments. (As in most state constitutions, the original wording and the later amendments are all jumbled together, rather than clearly differentiated as they are in the U.S. Constitution. Still, it’s easy enough to distinguish the earlier parts from the later ones.) The founders of the United Sates, men of the 18th-century Enlightenment, used words with a grace and precision rarely met with today (and least of all in today’s politicians!). They took language seriously. They wrote clearly, carefully, incisively. When they set down a sentence, they meant something definite by it, not some vague mush. By contrast, the more recent provisions are full of inanities like “everyone shall have a right to the privilege of education” — a sure sign that the writers are navigating the hallways of thought by touch, not by sight.
Television is often blamed for having shortened people’s attention spans and their ability to deal with complex issues; but there is nothing inherent in the nature of television as a medium that requires such a result. Rather, I think public education is the primary culprit here, for making such a boring and painful process out of reading, writing, and thinking that all too often the student’s intellectual muscles are crippled, his curiosity deadened, his pilot light effectively snuffed out.
The case for libertarianism is complicated. It rests on very broad generalizations, drawn from history and theory alike. It requires a capacity to think in principled terms, be the principles moral or economic. If the citizens of a Free Nation do not understand the case for liberty, they will not support it. Thus, a libertarian society cannot hope to survive unless the educational system is radically transformed. (I hope to return to this subject in a future article.)
So far I’ve been talking about how to maintain freedom in “a” libertarian society. But of course there are different possible models for a libertarian society, with different political and legal structures, and the threat from Leviathan Yet to Come may take rather different forms, and so require correspondingly different measures, in libertarian societies differing in structure.
There seem to me to be three main varieties of libertarian structure: the constitutionalist model, the proprietary model, and the pure market model. (There are also subvarieties of each of these.) Do these structures differ at all in their susceptibility to the Leviathan virus?
The Constitutionalist Model
Under the constitutionalist model, a single agency is charged with protecting libertarian rights within a given territory; this agency is then designed in such a way as to make it as difficult as possible for it to transform itself into Leviathan.
The constitutionalist model comes in both minarchist and quasi-anarchist varieties. In the minarchist version, the single agency holds a coercive monopoly over legal services within the territory; no competitors are permitted. In the quasi-anarchist version, competitors are not prohibited, and a few may function around the edges, but the dominant agency nevertheless holds a monopoly or near-monopoly, not through any coercion on its part, but because other nations regard it as the legitimate government and will deal only with it, so that it becomes more important for residents of the Free Nation to influence that agency’s policies than to influence the policies of rival agencies. (The dominant agency may, for example, be the holder of a 99-year lease of sovereignty on the Free Nation’s territory.)
A constitutionalist model is extremely risky. Nothing is better positioned to transform itself into Leviathan than a government (or quasi-government), however minimal. Being the dominant provider of protective services has been a stepping-stone to governmental power many times in history. In ancient times, that’s how Rome and Athens acquired their empires; in mediæval times, that’s how Aelfred King of Wessex became Aelfred King of England.
But the constitutionalist model has its advantages as well. While it may be highly susceptible to the virus of Leviathan Yet to Come, it is perhaps the most impervious to the viruses of Leviathans Past and Present. A libertarian society that can turn a governmental face toward the outside world, that looks to other nations like a state and can negotiate as a state, is more likely to be taken seriously and treated with respect in the community of nations. By contrast, if it looks to outsiders as though “nobody is in charge,” hostile powers may take this as an invitation to invade in order to “restore order,” and world opinion will put up little protest.
Suppose we do opt for the constitutionalist model; how should we design our minimal state (if we go the minarchist route) or our dominant protection agency (if we go the quasi-anarchist route) so as to minimize the likelihood that it will grow and seize power?
This is a topic I’ve addressed in a number of earlier articles, , , , , so let me just quickly recapitulate the main points of those earlier discussions:
- The libertarian state must consist of a central government, highly restricted in its powers so as to keep it from mischief, and a large number of competing local cantons, less restricted in their powers, so as to force political pressure down to the competitive canton level lest it otherwise simply shatter the central government, or bypass it, or shape it to its will.
- The cantons should be “virtual” rather than physical; that is, membership in cantons should not be tied to geographical location; thus changing from one canton to another will be costless, thereby limiting the ability of cantons to oppress their members; anyone can change canton membership at will, and any sufficiently large number of citizens can start a new canton.
- Cantons should be almost entirely self-governing, appealing to the central government to solve inter-canton disputes only as a last resort.
- The central government should have a bicameral legislature — one house of canton representatives, empowered to pass laws but only by a supermajority, and a second house of popular representatives, empowered to repeal laws, with only a superminority being needed for such repeal.
- Laws should also be subject to repeal, and public officials subject to dismissal, by popular referendum.
- A plural executive should be instituted, to provide a check on presidential power. The plurality should be three rather than two, to resolve deadlocks in an emergency.
I’ve discussed other provisions, but these strike me as the most essential ones.
The Proprietary Model
A different libertarian structure is the proprietary community. In this case, the agency providing protective services is also the owner of the territory in which it will operate. This model too comes in two subvarieties, depending on whether a single person or firm owns all the territory and everyone else leases from the owner, or whether instead the territory is jointly owned by all the inhabitants. (The latter arrangement is called a commune when the inhabitants are hippies, and a condominium when the inhabitants are yuppies.)
This model has some disadvantages from a libertarian standpoint. One of the frustrations of statist society is that individuals have no place to stand on and call their own, no private property on which to do as they please without asking leave from others. Yet in a proprietary community, one’s home is not really private; it belongs either to the landlord or to the collective. Since the arrangement is contractual, it satisfies libertarian standards of rights and justice — but perhaps not libertarian yearnings for independence.
Of course, it is possible to write the contract, and libertarians forming a proprietary community probably would so write it, in such a way as to make the leaseholds simulate private property as far as possible, to leave ample sphere for individuals to go their own way. But what guarantees that the contract will be respected? After all, the firm offering protective services, presumably including contract enforcement, is itself a party to the contract, and might decide to alter the terms unilaterally to its own advantage — thus turning into an oligarchic Leviathan. (Or, on the condominium model, the majority might decide to impose their will on the minority in defiance of the terms of the contract, thus turning into a democratic Leviathan.) Proponents of the proprietary model like to appeal to the example of hotels or apartment complexes; but hotel customers do not fear being oppressed by the manager, since they know there is a background of law enforcement to which the hotel manager must answer. But in a proprietary community, the hotel manager is also the chief of police.
Perhaps it will be said that the owners of a proprietary community will be restrained from abusing power by the fear of losing customers. But there is more than one way to prevent losing customers; one thinks of the Berlin Wall, for example.
The proprietary community model is also at a disadvantage, relative to the constitutionalist model, in dealing with other nations, which will treat a for-profit business enterprise with less respect than they would something that looks more like a sovereign state.
But the proprietary model has its advantages as well. Having a single firm control the territory can simplify the decision-making process, and income from leaseholds provides needed revenues to the “state” without the need for taxation. Perhaps most importantly, for those considering building a libertarian community at sea (or in space, for that matter), some kind of proprietary model may be unavoidable, at least for the territorial nucleus.
How can a proprietary community guard against the rise of Leviathan? There are several possibilities. One is to build into the contract something like the political structure of a virtual-canton constitution, as described above. Another is to separate the provision of security from the ownership of the territory; leaseholders might contract individually with a security company (or, preferably, with several competing security companies) rather than purchase their security through the landlord. Above all, it would be extremely foolish for residents in a proprietary community to contract away either the right to communicate freely (both with one another and with the outside world) or the right to own and carry weapons for self-defense. These two freedoms are the essential bulwark of liberty. To be sure, landlords have the right to place restrictions on the activities of their tenants; they can legitimately demand that everyone on their premises (other than themselves) be disarmed and refrain from unauthorized communication. But any tenant who agrees to live in the proprietary community on those terms is taking an unacceptable risk, in my view.
How is a proprietary community to gain respect in the eyes of other nations? One possibility is to build a community around a research station or a university; in conflicts with aggressor states, such a community would win more sympathy in world opinion than a community built around, say, a casino. Nor would it be the first time that an educational institution has acted as an autonomous political entity; the mediæval University of Bologna, for example, had its own student-run legal system, separate from that of the town of Bologna, and exercised civil and criminal jurisdiction over its own members. 
The Pure Market Model
The third possible structure for a libertarian society is the pure market model. On this model, there is no central agency in charge — be it a minimal state, a dominant protection agency, or a landlord. Individuals own their own homes, and provision of legal services is not monopolized. I have defended this model in previous articles. ,,,
The pure market model seems more vulnerable to Leviathans Past and Present than the constitutionalist and proprietary models, since it has nothing at all resembling a governmental face to turn to the outside world. Hence a libertarian society following such a model might have to be quite populous and powerful in order to succeed. This seems a serious disadvantage to market anarchism, at least in the short run.
On the other hand, the pure market model seems less vulnerable to Leviathan Yet to Come than does either the constitutionalist or the proprietary model, since those models all involve some monopolistic or near-monopolistic agency that is perfectly situated to turn itself into an oppressive state, whereas the pure market model involves no such agency. Nevertheless, many critics of the pure market model have argued that Leviathan would inevitably re-emerge.
Most versions of the pure market model envision a number of different agencies, specializing in protective services, and competing for clients. (As we shall see below, this is not the only form a pure market model might take.) Robert Nozick has argued, however, that any system of competing protection agencies would soon collapse into a monopolistic state. 
Nozick argues as follows: Competing protection agencies operating in the same territory will sometimes have conflicts. They will resolve these conflicts either by arbitration or by force. But in either case, the state will re-emerge.
Suppose two protection agencies resolve their disputes by resorting to force. Then either they are evenly matched, or one is stronger than the other. If one is stronger, then it will defeat the weaker one, either eliminating it or else making it subordinate to the victor. Where there were two agencies sharing one territory, there is now one agency for that territory; we have returned to territorial monopoly.
If instead the two agencies are evenly matched, and yet they continue to fight, clients of different agencies living in the same area will be motivated to relocate for security’s sake. Two separate “turfs” will emerge, with one containing mostly clients of agency A, and the other mostly clients of agency B. Where there were two agencies sharing one territory, each agency now has its own territory; once again, we have returned to territorial monopoly.
In the first case, monopoly is achieved by uniting the agencies (or by eliminating one of them); in the second, it is achieved by dividing the territory. As many different protection agencies fight it out, each individual conflict between any two will resolve itself in one of the two ways outlined above, Nozick thinks; and the cumulative result of all the different conflicts will be some number of agencies, each holding a territorial monopoly. In other words, states.
Proponents of the market model generally regard it as unlikely that protection agencies would resolve their disagreement by force. In a competitive market, such agencies need to attract customers, and an agency that settles its disputes by expensive means, such as war, will have to charge higher premiums, and so will attract fewer customers, than a agency that settles its disputes by less expensive means, such as arbitration. Accordingly, such theorists argue, arbitration rather than violence will predominate.
Nozick denies none of this. But he argues that if protection agencies do opt for arbitration, the state will still re-emerge. For whatever system the agencies set up for resolving their disputes — courts of appeal, or what have you — will amount to the formation of a single legal system for the entire territory, and the individual protection agencies will then become no more than branches of this new structure. Where there were many agencies sharing one territory, there is now a single overarching agency for that territory; here too, then, we have returned to territorial monopoly.
To meet Nozick’s argument, market anarchists must maintain that protection agencies could set up a dispute-resolution system that would be effective enough to prevent conflict, but would nevertheless fall short of constituting a new monopolistic agency. But a recent article by Tyler Cowen maintains that this is a vain hope.  Cowen argues that once competing protection agencies have set up a dispute-resolution network, the members of the network once it is established will be able to collude successfully to put competitors out of business. Ordinarily, such collusion would fail in a libertarian society, because new firms would have free entry into the market. But, points out Cowen, a new protection agency that is not part of the network cannot compete successfully with network members, since being a member of the network ensures that an agency will not have to go to war to secure its clients’ claims. Customers will want the assurance of network membership before they sign up with a protection agency. But existing firms can simply choose to exclude from participation in their network any newcomer who doesn’t toe the line. Hence, monopoly.
The most convincing reply I’ve seen to Cowan’s argument is from Bryan Caplan.  Caplan makes two points against Cowan. First, Cowan neglects to consider the possibility of competing networks. In the credit card market, Visa providers have to cooperate with one another through the Visa network, and no new firm can succeed in acting as a Visa provider if it is excluded from the network; but the Visa network still has to compete with the Mastercard network and so forth. Likewise, one network of protection agencies might compete with two or three others. Cowan might reply that these networks too will need to cooperate with each other, and that this will lead to a new, meta-network; but Caplan disagrees. Protection agencies need to join a network to cut down on the transaction costs of contracting bilaterally with all the other agencies individually. But the number of networks is much smaller than the number of agencies, so transactions costs would not be high enough to warrant a new network; and with only bilateral contracts rather than a network, the capacity for collusion is quite small.
Caplan’s second objection to Cowan’s collusion scenario is that even if there were just one network, Cowan is too quick to assume that its attempts at collusion would be successful. Suppose two protection agencies, Titanic Defense and Hindenburg Security, come into conflict; Titanic is a member of the network, Hindenburg is not. But Hindenburg suggests submitting the dispute to arbitration. What is Titanic to do? The rules of the collusion agreement suggest that Titanic should refuse, that Titanic should resort to force instead. But this is expensive. And Hindenburg is even offering to pay the costs of arbitration. In such a case, Titanic has a strong incentive to defy the agreement and cooperate with Hindenburg. Of course, the other members of the network could boycott Titanic for doing this; but such a boycott is not in their interest either. Thus, Caplan suggests, attempts at collusion among protection agencies are likely to fail. Or, as I would put it, collusion among protection agencies is a form of selective cooperation, and so is likely to be undermined for the reasons I set out in my collective-action article. 
Finally, it’s worth noting that both Nozick and Cowan think of the pure market model solely in terms of customers purchasing protective services on a competitive market. But this is not the only form that market anarchism can take. Another possibility is for customers to join together to provide for the common defense, rather than delegating this task to an agency. Such a arrangement is reminiscent of the mutual-protection associations common in English history. The advantage of this self-help model is that it decreases the risk that protection agencies will get together to form a Leviathan. The downside, of course, is that self-help is time-consuming and can involve heavy transaction costs. But having such a self-help system in the background, ready to be mobilized if necessary, might help to keep the protection agencies in line, thus achieving the best of both worlds.
Translations for this article:
- Portuguese, Anarquia: ontem, hoje e amanhã.
1 Roderick T. Long, “Implementing Private Law in a World of States,” Proceedings of a Forum on the Subject: Systems of Law (30 April 1994).
2 Roderick T. Long, “Dismantling Leviathan From Within, Part IV: The Sons of Brutus,” elsewhere in the present issue.
3 Roderick T. Long, “Funding Public Goods: Six Solutions,” in Formulations, Vol. II, No. 1 (Autumn 1994).
4 Roderick T. Long, “Defending a Free Nation,” in Formulations, Vol. II, No. 2 (Winter 1994-95).
5 Roderick T. Long, “Can We Escape the Ruling Class?,” in Formulations, Vol. II, No. 1 (Autumn 1994).
6 Roderick T. Long, “Religious Influence on Political Structure: Lessons from the Past, Prospects for the Future,” in Formulations, Vol. II, No. 3 (Spring 1995).
7 Roderick T. Long, “Who’s the Scrooge? Libertarians and Compassion,” in Formulations, Vol. I, No. 2 (Winter 1993-94).
8 Roderick T. Long, letters exchange with Kevin Whiteacre, “Our Readers Write,” in Formulations, Vol. III, No. 2 (Winter 1995-96).
9 Roderick T. Long, “Good and Bad Collective Action: Can We Nourish One and Squelch the Other?,” in Formulations, Vol. III, No. 1 (Autumn 1995).
10 Robert Nozick, “The Zigzag of Politics”; in The Examined Life: Philosophical Meditations (New York: Simon & Schuster, 1989).
11 Roderick T. Long, “Virtual Cantons: A New Path to Freedom?,” Formulations, Vol. I, No. 1 (Autumn 1993).
12 Roderick T. Long, “Imagineering Freedom: A Constitution of Liberty” (Part I in Formulations, Vol. I, No. 4; Part II in Vol. II, No. 2; Part III in Vol. II, No. 3; Part IV in Vol. II, No. 4).
13 Roderick T. Long, “The Rationale of a Virtual-Canton Constitution,” Proceedings of a Forum on the Subject of Constitutions (2 October 1993).
14 Roderick T. Long, “Assets and Liabilities of the Constitution of Oceania,” Proceedings of a Forum on the Subject of Constitutions (2 October 1993).
15 Roderick T. Long, “A University Built by the Invisible Hand,” in Formulations, Vol. I, No. 3 (Spring 1994).
16 Roderick T. Long, “The Nature of Law” (Part I in Formulations, Vol. I, No. 3; Part II in Vol. I, No. 4; Part III in Vol. II, No. 1).
17 Roderick T. Long, “The Decline and Fall of Private Law in Iceland,” in Formulations, Vol. I, No. 3 (Spring 1994).
18 Roderick T. Long, “Anarchy in the U.K.: The English Experience with Private Protection,” in Formulations, Vol. II, No. 1 (Autumn 1994).
19 Robert Nozick, Anarchy, State, and Utopia (New York: Basic Books, 1974), Chapter 2.
20 Tyler Cowen, “Law as a Public Good: The Economics of Anarchy,” in Economics and Philosophy, Vol. 8 (1992), pp. 249-267. (See also the reply by David Friedman in the subsequent issue.)
21 Bryan Caplan, “Outline of a Critique of Tyler Cowen’s ‘Law as a Public Good’,” unpublished manuscript, July 1993.