Defending Liberty: Essays in Honor of David Gordon

CHAPTER 7

Anarchy, Nozick, and Gordon

Aeon J. Skoble

Only a system that renounces force altogether … qualifies as fully libertarian.

—David Gordon

I have never met David Gordon, but he nevertheless has played a role in my intellectual development, so I am happy to contribute to this volume of essays in his honor. That role is indeed the subject of this essay. Although I did not know it at the time, Gordon was one of the anonymous reviewers of the manuscript of my first monograph, Deleting the State: An Argument About Government, and Gordon’s review was largely positive, yet he disagreed with my interpretation of Nozick’s defense of the minimal state.1 For my contribution to this festschrift, I will discuss how Gordon’s interpretation of Nozick differs from mine, but why in either case Nozick’s argument doesn’t quite succeed in defending the minimal state against individualist anarchism. I also discuss how Nozick’s argument can be repurposed to Gordon’s advantage in the debate between minimal state libertarians and anarchist libertarians.

WHAT I THINK NOZICK WAS DOING

Robert Nozick’s 1974 book Anarchy, State, and Utopia advances a robust theory of individual rights as a moral principle which must limit state power. If people have rights that cannot be overridden, then the coercive power of the state will be difficult, if not impossible, to justify. This might be taken to imply the logical necessity of anarchism, the idea that no rulers can be justified. The anarchist position is that since governments necessarily violate rights, the legitimacy of state authority is inconsistent with the idea of rights as a moral principle. The State produces commands which much be followed on pain of imprisonment or death, and this will frequently violate rights. The State finances itself through nonconsensual taxation, also violating rights. For many people, this inference is just obviously false, as anarchism is inconceivable to them and appears as a reductio ad absurdum for rights theory. But Nozick thought the argument was worth taking seriously. He recognized that a robust theory of individual rights is difficult to reconcile with most forms of government that we’re used to—whether via majority rule or kings and parliaments. So before arguing against certain models of political authority, chiefly socialism and redistributive welfare-state liberalism, Nozick argues for the legitimacy of the “minimal state”; that that can be justified even if more expansive conceptions of government cannot. “Since I begin with a strong formulation of individual rights,” Nozick says, “I treat seriously the anarchist claim that … the state must violate individuals’ rights and hence is intrinsically immoral.”2 Nozick sets himself the task of showing how the minimal state could arise without violating anyone’s rights.

In Deleting the State, I interpreted Nozick in the following way: “the services of protection and conflict resolution would be provided more efficiently with the minimal state than without it, and thus the state is inevitable. [Nozick] defends the extension of the right to individual self-defense to one of collective self-defense in the sense that many might collectively engage the same means of exercising their right to self-protection, which means it would be morally legitimate to pay other people or companies to protect us. Then Nozick explains how, as a matter of economic efficiency, one of these protective agencies would come to dominate the others, at least in a particular geographic area, through a non-coercive process of mergers and acquisitions. This ‘dominant protective agency’ would then fit the standard definitions of a state, as it provides protection for all and monopolizes this service.”3

Nozick appreciates the Lockean argument that if we have rights at all, we also have the right to contract for assistance in protecting or enforcing our rights. That I have a moral claim that others not steal from me does not entail that no one will steal from me. I have a right to try to prevent others from stealing, and if they do steal, I have the right to seek redress. Since I may have difficulty doing this on my own, it may be advantageous, and certainly morally legitimate, to contract with one or more agents to assist me in this endeavor. In the classical liberal tradition, this is one way to justify the entry into political association. Anarchists have responded to this by noting that such a service could be provided by any number of private firms, which might in turn compete for customers. So if the state is classically conceived of as being the agent of rights-protection, there’s no reason why we shouldn’t have a market among several protective agencies. Nozick’s move is to suggest that one such protective agency would inevitably become predominant, not via conquest and violence, but via non-rights-violating market practices.

Of course, it is possible that all these competing firms might coalesce into one dominant protective agency, but the difficulty is, once this has happened, why shouldn’t another competitor emerge, either to compete on rates or to offer some innovation (as Netflix challenged Blockbuster)? And what if someone in that region wanted to opt out of subscribing to that agency (you’re not required to watch television in the first place)? It is here where Nozick needs to engage the anarchist argument.

I interpreted him as saying something like the following: “the protection of ‘independents’ who do not want to subscribe to the dominant protective agency might be subsidized by members of the dominant protective agency who pay a higher fee for a different kind of service….Nozick argues that it would not be rights-violating for this ‘de facto monopoly’ to coerce payment for the services. Presumably, he arrives at this conclusion because each of the smaller protection agencies was voluntarily funded, so the new ‘parent company’ is not violating rights to exact payment after the ‘merger.’ The question, then, is whether or not coercion is involved after the formation of the dominant protective agency. If someone who, despite having voluntarily subscribed to one of the smaller companies, disliked the operation of the new dominant protective agency wanted to ‘opt out,’ would this be permitted by the agency? Would he be entitled to secure this service from someone else? If the dominant protective agency must use coercion to bar market entry of competing services, and may force dissenters to continue paying them, then Nozick will be in error in claiming that no rights have been violated.”4

Besides the fact that it protects “independents,” the dominant protective agency’s prohibition of competing protective agencies is what makes it a state. To Nozick, this means he has shown how the state could arise legitimately, without violating rights, and this is the aspect of Anarchy, State, and Utopia I had expressed disagreement with. I was attempting to argue that Nozick can only make this move if he’s tacitly appealing to a kind of Hobbesian worry that if the dominant protective agency has competition, it would produce a decline in social order as they found themselves in conflict. If everyone’s rights are made less secure by the entry of a new firm, then maybe it’s not wrong to prohibit, and if they are to prohibit, they must compensate. As Nozick himself explains: “A protective agency dominant in a territory does satisfy the two crucial necessary conditions for being a state. It is the only generally effective enforcer of a prohibition on others’ using unreliable enforcement procedures [i.e., independent enforcement]….And the agency protects those nonclients in its territory whom it prohibits from using self-help enforcement…even if such protection must be financed (in apparent redistributive fashion) by its clients. It is morally required to do this by the principle of compensation, which requires those who act in self-protection in order to increase their own security to compensate those they prohibit from doing risky acts which might have turned out to be harmless [i.e., seeking independent enforcement] for the disadvantages imposed on them.”5 This seems like a fear that a polycentric system of dispute resolution would be incapable of stable cooperation. I go on to explain why I think Nozick is mistaken to find this an insurmountable objection, but that’s not particularly what led Gordon to disagree with my analysis.

WHAT GORDON DISAGREES WITH ME ABOUT

Gordon argues that if my interpretation of chapters 4 and 5 of Anarchy, State, and Utopia were correct, the rest of my argument against Nozick would follow, but that in fact I have misconstrued what Nozick is arguing there. “But,” Gordon says, “Nozick does not say that the dominant agency can compel those who prefer other agencies to join it. Neither can it compel people to pay for its protective services. Further, although Nozick does discuss preemptive attacks, he does not at all assume that agencies will fear such attacks by other agencies.”6 A lot rides on the following passage from Nozick: “It is enough to have rebutted the charge we imagined earlier that our argument fails because it ‘proves’ too much in that it provides a rationale not only for the permissible rise of a dominant protective association, but also for the association’s forcing someone not to take his patronage elsewhere or for some person’s forcing someone not to join any association. Our argument provides no rationale for the latter actions and cannot be used to defend them.”7 Gordon cites this passage as evidence that I’ve gotten Nozick wrong, whereas I had deployed it as evidence that Nozick’s argument doesn’t work. It’s clear that this is a difference in reading with great import. I had gone on to note that if Nozick has no argument for the state beyond that one could conceivably arise without violating anyone’s rights, merely a logical possibility as opposed to an inevitability, then that means he thinks that no competitive set of such agencies could be fair and feasible; that is, he rejects polycentrism because of a Hobbesian fear of social chaos. This is where Gordon argues I have misread.

Gordon continues by clarifying his disagreement. “Skoble has allowed his conception of what a minimal statist must believe to govern his interpretation of Nozick. In Nozick’s system, the dominant agency can successfully forbid other agencies from imposing risky decision procedures on its clients. Because in doing so it disadvantages these independents, it must provide them with low cost or free protective services. These features, in Nozick’s view, suffice to make the dominant agency a minimal state. If Skoble does not agree that this is enough for a state, he may well be right; but that is Nozick’s contention. He does not say, as Skoble thinks, that preservation of society necessitates rights-violating coercion.”8 To clarify, I wasn’t disputing that this makes the dominant protective agency a state; I was disputing that such a state would not be rights-violative. But Gordon’s other point is more solid: I did indeed infer that Nozick thinks the rights violations are justified because otherwise the minimal levels of social cooperation would not obtain—while I didn’t use the expression “preservation of society,” that is not an unreasonable summary of what I had argued. That is why I had characterized the concern as a Hobbesian fear.

This leads Gordon to elaborate. “In Skoble’s framework, Hobbesian Fear leads libertarians to accept coercion they would otherwise deem unjustified. Nozick does not think this: he thinks that the dominant agency acts perfectly within its rights in forbidding risky decision procedures to its clients. Skoble, if I have understood him, thinks that the dominant agency can de facto shut down all competing agencies by declaring their decision procedures unacceptably risky. But this is not correct: the dominant agency cannot forbid other agencies from applying such procedures to nonclients. Nozick does not contend that agencies would be unable, without a minimal state, peacefully to resolve their differences about decision procedures: he thinks that they are under no obligation to do so.”9 This would mean, he says, that I have missed an important distinction, and that Nozick is not motivated by Hobbesian-fear-based opposition to polycentrism.

A DECADE LATER,
HOW DO I UNDERSTAND THIS DISAGREEMENT

I do see that much hangs on the modality of whether the dominant protective agency would act a certain way as opposed to might act a certain way. But Nozick certainly seems to be saying that it would have to prohibit. Nozick might be arguing that they have to prohibit, but then might have an obligation to compensate, and that this means, when all is said and done, no rights have been violated. But that doesn’t quite work; when Smith gives Jones fair compensation for having trespassed, it doesn’t imply that Smith didn’t in fact trespass. If the dominant protective agency says to a potential competitor “we will forcibly prevent you from starting a new rival firm, but will also compensate you to make up for it,” the would-be rival’s rights have in fact been violated, the correctness of the compensation notwithstanding. Why would the dominant protective agency feel it necessary to say something like this? Assuming it’s not simply wishing to avoid competition, it would have to be the idea that allowing the rival firm to operate reduces the security of everyone. That would make it a necessity to prohibit, and that is indeed the fear of social chaos.

Consider this passage from Anarchy, State, and Utopia: “when an action is forbidden to someone because it might cause harm to others and is especially dangerous when he does it, then those who forbid in order to gain increased security for themselves must compensate the person forbidden for the disadvantage they place him under.”10 While Nozick appeals here to a principle of fairness to generate the obligation for the dominant protective agency to compensate, it is the specter of open conflict that underlies this special category of prohibition. The concern expressed in this passage continues to seem to me to indicate opposition to polycentrism based on Hobbesian fear; viz., fear that minimal levels of social cooperation would not obtain under polycentrism.

One thing I have come to see differently about Nozick’s argument in the last several years is the extent to which he is confident that very few people would actually be “independents” of the sort that would prompt the advent of rival firms. Given his “framework” conception of the minimal state, in which many sorts of voluntary societies are possible provided that one’s right to exit is preserved, perhaps it might never come up that another “firm” could deliver rights protection better than the one that first arose. But this just means that Nozick’s conception of the minimal state is closer to the concerns of the anarchist than it first appears. It remains the case that Nozick wants the dominant protective agency to be seen as morally in the right when it prohibits competition. Gordon’s remarks did prompt me to reread Nozick carefully on this point.

WHY, EITHER WAY, NOZICK’S ARGUMENT DOESN’T DEFEAT THE ANARCHIST OBJECTION

At the end of the day, I think Gordon remained sympathetic to my main point: “Many people do dismiss libertarian anarchism just because they think that it would lead to chaos. Skoble argues that the Hobbesian Fear lacks adequate grounds: people can resolve disputes without a monopoly state.”11 Whether Gordon is right or wrong about the way I’ve characterized Nozick’s justification for the dominant protective agency prohibiting competition, it remains the case that the dominant protective agency, now the minimal state, is a monopoly that uses coercion to remain that way.12 While in Nozick’s telling, this is entirely benign, it’s of course possible that this minimal state could degenerate into something less attractive, or more specifically, something that didn’t protect people’s rights as robustly. It could be subject to any number of well-documented problems to which states are prone: it could be captured by one faction and adjust its operations to disadvantage another; it could cease recognizing as robust a conception of rights as it once did; it could ramp up what the “user fees” are to the point where they’re not recognizably voluntary at all. Paper constitutions are helpful in averting this sort of “mission creep,” but they are not infallible, as experience has shown. The advantage of polycentrism is its built-in resistance to this. As long as there is more than one firm providing the service of conflict resolution and rights protection, no one firm can overstep its bounds. The idea that these true “checks and balances” would devolve into constant violent conflict is belied by both the history of competition and game-theoretic analysis.13

Gordon addressed this larger argument thus: “A group of people may recognize that if all of them refrained from initiating violence, they would all be better off. No longer would life be nasty, brutish, and short: people could now live peacefully. Unfortunately, this recognition will fail to generate the required agreement. Each person will also recognize that he would be still better off if he himself resorted to violence whenever he deemed it advantageous. If others keep their agreement, so much the better, and if they do not, one clearly is worse off by being the sole person to observe the agreement. Of course, everyone will reason in the same fashion and no one will keep the agreement. In brief, we have here a classic Prisoner’s Dilemma.”14 This is, in large part, why proponents even of minimal-state libertarianism often “settle for” the state despite their recognition of its essentially coercive nature. The minimal level of cooperation necessary for society to function at all would be lacking without centralized authority. But is this correct?

Gordon continues: “Skoble brings to bear important work by Robert Axelrod and later writers that undercuts the analysis just presented. The argument that keeping the agreement, despite its recognized advantages to the group, is irrational applies only in a special case. If people do not expect to have further dealings with one another, then defection is the rational course. But people in a society do not find themselves in such a ‘one-shot Prisoner’s Dilemma.’ Quite the contrary, they must deal with one another repeatedly. In such an ‘iterated Prisoner’s Dilemma,’ cooperation, not defection is rational.”15 In other words, competition can co-exist with cooperation. In a way, Nozick’s “framework” presupposes this: the many different voluntary societies that would flourish there are, at their own level, an example of polycentricity. Nozick tacitly recognizes this, but thinks there still needs to be a final authority which is centralized, because it would be too risky, too chaotic to allow polycentricity at that level. But that still seems to me to be a mistake.

There’s no particular reason to assume a polycentric (or non-monopolistic) system of dispute resolution and rights enforcement would degenerate into chaotic strife. If one protective agency’s client had a dispute with a client of another agency, why assume the two agencies would resolve the matter through violence? That’s surely not an efficient or sustainable business model. I suspect that people are thinking of the passions of the individual disputants, who may be so aggrieved as to want blood. But it’s already true in the actual world that some people in disputes have inflamed passions and seek immediate private retribution. Consider this scenario: Smith drives over Jones’ rose bushes, Jones gets enraged, Jones punches Smith. People assume this is what would happen without a state, but of course it happens all the time with the state. Most people prefer to settle their disputes in a more procedurally stable format. In the world with a state, this can involve reporting a crime or filing a lawsuit—or, as it turns out, seeking private arbitration. If there were no state, it wouldn’t follow that none of the things the state does would get done. The tired old chestnut about roads and lighthouses requiring state action is belied by both theory and history. Since private arbitration already exists, and in general the idea of third party adjudication of disputes already exists, there’s no reason to assume they wouldn’t be the preferred approach in a stateless world, just as they are in the real world. No social order, stateless or otherwise, can guarantee that there will be no ill-tempered individuals who prefer violence. The question is, which sort of social order offers the most stable institutions with the least propensity for corruption and the greater incentives for peaceful cooperation. For all the reasons Nozick thinks this points towards the minimal state, it points towards the polycentric legal order envisioned by individualist anarchists.

Gordon’s sympathies appear to fall on the side of polycentricity, though we’ve never actually had a chance to discuss this aspect of it. In any event, Nozick’s argument, for all its many virtues, still seems to fail to refute the anarchist objection. Gordon doesn’t claim otherwise, of course; his main concern being with how I characterized Nozick. Gordon’s comment pressing me to think more carefully about Nozick’s argument ended up helping me be even more confident that the polycentric model withstands Nozick’s objections. Gordon’s analysis and arguments are incisive; his benevolence invaluable.

1The points he made in his referee report were later published as a book review; quotations from Gordon will be to that. The published review is David Gordon, “Review of Deleting the State: An Argument about Government, by Aeon J. Skoble,” The Mises Review 15, No. 1 (Spring 2009).

2Robert Nozick, Anarchy, State, and Utopia (Basic Books, 1974), p. xi.

3Aeon J. Skoble, Deleting the State: An Argument About Government (Open Court, 2008), pp. 43–44.

4Deleting the State, pp. 44–45.

5Anarchy, State, and Utopia, pp. 113–114.

6David Gordon, “Review of Deleting the State: An Argument about Government,” The Mises Review 15, No. 1 (Spring 2009). [Review is unpaginated.]

7Anarchy, State, and Utopia, p. 129.

8“Review of Deleting the State: An Argument about Government,” The Mises Review 15, No. 1 (Spring 2009). [Review is unpaginated.]

9Ibid.

10Anarchy, State, and Utopia, p. 81.

11“Review of Deleting the State: An Argument about Government,” The Mises Review 15, No. 1 (Spring 2009). [Review is unpaginated.]

12That is, the minimal state that is the “framework” is a monopoly. Within the framework, the various voluntary communities/associations may or may not be “minimal,” but in any case they’re all “under” the minimal state which guarantees rights protection, and which can prohibit competition against itself (not between communities).

13Both of which are discussed in Deleting the State, but much more thoroughly explored in the works I drew on, as Gordon rightly notes.

14“Review of Deleting the State: An Argument about Government,” The Mises Review 15, No. 1 (Spring 2009). [Review is unpaginated.]

15Ibid.