An online debate among participants in the University of Arizona Libertarians discussion group prompted this attempt to develop and express my understanding of minarchy, the notion of minimal government: what justifies it, whether it’s more than just a pipe dream, and how it differs from anarchy, the absence of government.

Though the ideas might best benefit from being presented in an academic style, this was written informally for a mixed audience of philosophers and laymen. I have changed the original only to make it a bit more accessible to a wider readership. Perhaps, if it provokes sufficient response, I will rework it to be more suitable for an academic journal.

I apologize in advance to readers who have been studying the issue, reading all the best works by all the anointed philosophers on the subject. I haven’t been there or done that. It may be that I’m reinventing the wheel, or that what I have to say here has been anticipated and demolished by wiser heads. I don’t know. This is just what I’ve worked out for myself.


I’m a functionalist — in a broader sense than the variety discussed in philosophy of mind. I think that even matters of ontology (the taxonomy of what exists) are best explained in functional terms. An object, for example, is best understood in terms of how it interacts, as a unit (as opposed to a collection of independently functioning parts), with its environment — perhaps best exemplefied by the fact that human beings are psychologically disposed (for evolutionary and developmental reasons) to regard it as an object.

Now (if I haven’t lost you already with such a controversial abstraction), how do I apply my functionalism to the concept of government? The concept of rights? Basically, the idea is that something is identified (and is best defined) by what it DOES.


This question is too large to address here in detail. For our purposes, I need to distinguish between moral rights and legal rights. As important background, I’ll tell you that I also distinguish between natural rights and civil rights: Natural rights I take to be general kinds of opportunity — first, those that one would have in the complete absence of other people. When nobody else is around, you have complete freedom to live your life as you see fit, within the bounds of physical possibility. You enjoy life, freedom of movement, property (unlimited use of the things within your control), etc. Extend this concept to include peaceful, voluntary interaction with other people (according them the same rights you would claim for yourself) without the interference of any other external control (e.g., government), and you have freedom of association, contract, and the like. I know this differs from the standard Aristotelian view, but I’ve decided not to be shy about advancing functional definitions to replace traditional ones (as you’ll continue to see). Civil rights are the general kinds of things one reasonably and justifiably expects from a government to which one has consented. For illustration, the most important of these is equal treatment under the law.

Back to the essential distinction for this argument, though. Moral rights are those general kinds of opportunity one has to defend one’s natural rights against violation by others. One has the right to retaliate personally when one’s natural rights are violated, or to contract with an agent to do so indirectly. Legal rights are those general kinds of opportunity one has to defend one’s civil rights. When one has contracted with a government, one can defend either directly or through an agent any violation by the government of that contract.


I think there’s a tendency among anarchists to define government as an illegitimate institution that controls people without their consent — and extorts money from them in order to do so. This conveniently defines out of existence the sort of government mentioned by Jefferson in the Declaration: one whose legitimacy derives from the fact that the governed HAVE consented to it (and, I would add, contribute to it voluntarily — something the Founders probably didn’t think possible). I want to put aside any traditional presuppositions about the nature of government and look at what essential functional attributes it has. What does government DO that, if removed, would eliminate its status as government?

‘Government’, then, as a general term, will be defined by the functional attributes of entities that are included by it. Governments are not individual people (even monarchies have lesser officials to carry out the monarch’s wishes); they are institutions.

Not all institutions are governments. Families, churches, schools, businesses, professional associations, etc. — all are institutions, but are not governments because they lack an essential property that distinguishes governments from other institutions of society: they do not require the use of force or the threat of force to accomplish their essential functions.

Governments are institutions that DO require force or the threat of force to accomplish their essential functions.

Note that the goals of specific governments are irrelevant at this point. Governments might exist to maintain the power of a line of kings, or to empower the proletariat, or to secure individual liberties — whatever. They might fit the anarchist’s conception of the inherently oppressive institution, or they might genuinely serve just to protect individual rights, as the Founders intimated.

One more essential property applies to governments: a claim to legitimacy in the exercise of their power. Without this, gangs of highwaymen (relying as they do on force, yet knowing full well they have no right to what they take) would be governments. The claim to legitimacy might rest on divine right, on popular acclaim, on universal consent, or on voices in some dictator’s head — but no government acknowledges that it has NO legitimacy.

Because my intended audience is familiar with libertarian philosophy, I won’t offer much detail for the next several foundational points for my views here.

First, the claim to legitimacy of government must also rest on its use of force being limited to responding to violations of individual rights (abiding by the libertarian prohibition against the initiation of force — the defining principle of libertarianism).

Next, I’ll assert without argument that protecting individual rights is the proper role/goal of government (if it has any at all), and that consent of the governed can confer legitimacy on it. There’s general agreement among libertarians about this (and even among many anarchists) — the question being whether minarchy, so construed, can actually fulfill that role without simultaneously violating it, and whether consent of the governed can be achieved or maintained, and if there’s any principled way to handle exceptions — people who don’t consent, yet appear to be governed nonetheless.

To answer this question, I first want to reject what I think is an arbitrarily narrow definition of the “ultraminimal state” by Robert Nozick, which was cited in the precipitating discussion:

An ultraminimal state maintains a monopoly over all use of force except that necessary in immediate self-defense, and so excludes private (or agency) retaliation for wrong and exaction of compensation; but it provides protection and enforcement services only to those who purchase its protection and enforcement policies. People who don’t buy a protection contract with the monopoly don’t get protected.

The element of consent here is handled well, but it’s not necessary for the state to have a monopoly on force. I see no reason that several governments couldn’t exist within the same “market” — perhaps even with some people subscribing to more than one of them. I think governments will naturally tend to seek monopoly status, but we needn’t acquiesce to that tendency. More interestingly, though, a de facto government monopoly may turn out to be the most beneficial arrangement to be had.


This should be short, if not sweet. Anarchy is the absence of government. Period. Nobody in charge, no established conventions one can’t break with impunity, no restraints on behavior beyond the worry of how someone might retaliate. Anarchy requires no presuppositions about widespread moral conventions in society, agreements about right or wrong, concern for liberty, justice, etc. It most definitely includes no presuppositions about the uniform rationality of mankind, or any other such pathway to reaching a just, stable, peaceful modus vivendi through the action of market forces.

Humans might not be savants, but at least they’re generally clever — maybe even rational — about dealing with economic concerns. Unfortunately, when it comes to politics, the vast majority of us are merely hair-deficient chimpanzees — seeking out our optimum status in a tribal hierarchy, constantly jockeying for position, and ever willing to sacrifice principle (if we can even identify one) for narrow personal gain. The claim, rather widespread among anarchists (it seems to me), that people without government will quickly hash out a highly effective and efficient set of conventions for peaceable co-existence (one that does NOT constitute government) strikes me as little more than anthropocentric wishful thinking.


Anarchists hold that the individual has the right to retaliate against force initiated against him. Typically, they accept that the individual can contract with some independent agency to retaliate on his behalf. The presumption is that no right of the aggressor is violated when retaliation takes place: retaliation never requires the consent of the aggressor to be legitimate.

This brings up another point about the nature of government. People might think that a defining characteristic of government is that one can’t escape its authority (at least as far as the government is concerned). Maybe so, but I think the real reason for this is that one can’t, morally, escape the authority of a victim (or his agent) to retaliate. Therefore, this inescapable authority is really not an attribute of government (as the retaliatory agent of its citizen-clients), but of the individual rights that create, empower, and justify government.

While an individual has the right to retaliate against aggression, and the aggressor has no right to avoid a just response to his aggression, we have a problem of determining what constitutes a just response. Not only that, but we also have the problem of verifying that aggression has acutally been committed in the first place, since that’s not always clear. Under anarchy, the ultimate judge in both cases is the individual who perceives a violation of his rights. This judgment might be contracted out to a third party — at the risk of reaching a judgment with which that individual disagrees. If we believe that a complainant is the best judge of facts concerning whether a violation of rights has taken place, and who committed it, then anarchy is eminently reasonable. If we think that the passions of a complainant are likely to result in retaliation against innocent persons, or in disproportionately severe retaliation against guilty persons (violating their rights), then we might want to consider alternatives. I know I do.

I have already mentioned that my view of minarchy requires that individuals agree to be represented by the state when they believe their rights have been violated by others (excepting, of course, acts of self-defense and situations where the state has not performed its duty). I have also mentioned, both in earlier posts and here, that no aggressor can claim a right to consent to retaliation against his aggression. What I haven’t mentioned yet is that those who subscribe to the protection of the state also must agree to be governed by the state (that is, accept its judgments against them). Naturally, we will carefully pick our state — or states (remember, no monopoly required) — according to whether we think its/their judgments concerning us will be fair and just.

But this is a significant departure from the anarchist positon, for the anarchist’s concern need not extend beyond protecting his own interests. Even if anarchists express a broader concern for everyone’s rights, the only mechanism they can offer is that individuals arrange for the defense of their own rights.

What makes this view different? The minarchist voluntarily subjects himself to judgments against him by the state to which he subscribes. This comes from a willingness to protect the rights of others, even if you turn out to be the violator of those rights.

Wait! One might argue that anarchists will do the same, by voluntarily accepting the authority of private security agencies to which they subscribe. But such such security agencies are institutions requiring force or the threat of it to accomplish their essential functions. They are, by my definition, governments! To dispute this, then, one needs to explain why my functionalist definition should be abandoned.

Now, simple functional definitions such as mine often come at a cost; in this case, the cost is accepting that private security agents are governments even though that doesn’t fit well with the popular preconception of government. Still, I think that cost is minimal. Private security agencies are similar in their relation to their clients to feudal lords (whose rule one would typically call ‘governing’); both provide protection from aggressors at a price, and citizens need have no voice in the operation of their government (save voting with their feet). Furthermore, I think that cost is offset by the benefit of having our term ‘government’ apply to a useful concept (with real explanatory power) rather than to an outdated, mistaken one.

One answer, then, to the question of how minarchy is different from anarchy, is that it isn’t — much: anarchy depending on institutions (rather than strictly individual action) for the defense of individual rights is actually minarchy — a strange breed with lots of “agency-states” wielding force in retaliation against people, with the consent of the complainant, but not necessarily with the consent of the accused.

Who is governed by such a state? If we take the normal meaning, where governing involves imposing control, it’s not the subscriber who is governed (at least, not in the role of subscriber), but the accused, whether or not he subscribes. A better variety of minarchy would include the consent of the governed as well — and I think my version does this.

Now, even private security agencies that don’t get their clients’ agreement to be bound by their judgments (and this is the model I hear anarchists promote) would still rely on force, and therefore be governments. But the client is at some remove from that government, and can’t really be said to belong to that state; he has merely contracted with it for services, and has no obligation to it other than to pay for services rendered. Perhaps even more problematic is the fact that such agencies don’t necessarily have the consent of anyone against whom they act in protection of clients’ rights. In cases where no error has been made, and the person retaliated against was actually guilty of a rights violation, this might offer few problems; retaliation is morally justified, and it has been authorized by the individual who has the right to retaliate.

Even when a client makes no specific agreement to be subject to a contracting agency’s enforcements of other clients’ rights, there is a sense in which he is logically bound nonetheless. If (1) he expects his agency to render justice when he has been wronged, and (2) he acknowledges that his agreement with the agency differs in no significant way with similar agreements other clients have with the agency, then he accepts, by implication, being bound to judgments by that agency against him, on behalf of other clients. He consents, in other words, not only to being protected by the agency, but to being subject to retaliation by it on behalf of its other clients.

In fact, so long as one acknowledges the justice of retaliation against aggression, one accepts by implication the authority of individuals (and, by extension, their agents, whether styled as private agencies or as governments in the traditional sense) to claim that justice when their rights have been violated. One must then accept the legitimacy of retaliatory actions by the individuals (or their agents, including security/government agencies to which one doesn’t personally subscribe) against whom one has trespassed. Maybe this is the common anarchist model. If so, it presumes that individuals do accept the principle of justice, and are willing to abide by it — an assumption that non-anarchists aren’t willing to make. I’d rather not call this view ‘anarchism’, since I think the term implies an unpredictable variety of attitudes concerning how interpersonal actions should be constrained.

At any rate, where there is agreement on the principle of justice, this arrangement is designed both to protect the client against trespasses by others, and to ensure fair treatment when the client is accused by another client of the same, or any other provider — all while maintaining both a strictly retaliatory role for government and the consent of the governed. That’s the essence of the minarchy I support; I’ll get to some nuances in a moment. Before proceeding further, I should note that this really amounts to a social contract — one differing importantly from Hobbesian or Rousseauian versions by obtaining the real (not imaginary or presumed) contractual agreement of citizens.

What, though, of individuals who deny the justice of retaliation, or claim personal exemptions, etc.? Justice against them must be prosecuted without their consent, and any government doing so operates without the “legal” sanction of the “consent of the governed,” and only with the moral sanction of the principle of justice.

These individuals will be what the term “outlaw” originally meant: people who live outside the law — without the protection of government. Having refused the most basic tenet of civilization — justice — they are not owed its benefits, any more than are non-human animals, pirates on the high seas, or attacking foreign soldiers. They can be treated as varmints, terrorists, or warriors when they violate the rights of others. But they will be held accountable for wrongs they commit against anyone who is prepared to defend his rights, directly or with representation. Governments representing their victims do not require, either legally or morally, their consent for retaliation.

That’s my basic position. Now let me try to make my vision of how a minarchy could develop more vivid — and, I hope, plausible in the “real world” sense that a skeptical anarchist might demand.


Without undertaking an exhaustive examination of the options, let me just tell you what sort of arrangement I think will work best — one that might emerge from competition in a free market, but that could also be implemented by design when founding a minarchist state. It will require no special new attributes for people; ordinary psychological dispositions will do the job.

Let’s imagine that we begin with lots of private security agencies, but that they will develop among themselves an agreement of reciprocity. When a client of one agency has a beef with the client of another agency, both of those agencies defer judgment and disposition of the case to a third, uninterested agency (randomly selected or agreed on by both parties); both clients agree, as part of their contracts with their own agencies, to pay the third agency its customary fee. Disputes between a client and his own agency will also be settled by an uninterested agency, and the client and his agency will have agreed in advance that the ordinary costs of the case will be paid according to the instructions of the arbitrating agency.

The interagency agreement of reciprocity will be a streamlined set of rules promoting the just and inexpensive resolution of disputes, for it will be entered by entities who, like the participants in John Rawls’ “Original Position”, have no knowledge of how their particular narrow interests might be better served by a strategy other than one of equality, efficiency, economy, and justice. It will also be effectively an umbrella government, under which the individual agencies act as arms. And, assuming everyone chooses to be represented by a security agency rather than handle the responsibility himself, then everyone will have consented to be governed according to the terms of a contract.

Will such interminarchic reciprocity emerge? Frankly, I think it need only be offered for its benefits to be recognized and come to dominate in a free market. Game-theoretically, I think it will turn out to be what is called an ‘evolutionarily stable strategy’.

What of exceptions — people who take the do-it-yourself approach to protecting their rights and resolving disputes, or agencies that opt out of the reciprocal agreement? In the former case, I think even individuals who are ideally qualified to defend their own rights would still be motivated to subscribe to interminarchic reciprocity because of its stabilizing benefits. And, for those who aren’t fully capable of self-defense, yet decline participation in an interminarchic system, I can only imagine that these would be people who can’t afford the services of a government. (If anyone has an argument that individuals who can affort such representation [with all the safeguards to rights I’ve described above] would reject it, I’d love to hear it.) For the indigent, I believe the ordinary charitable dispositions of people would provide for representation. So, basically, I think this worry would be eliminated by the free market and human generosity.

In the latter case, I believe that non-reciprocating agencies would be quickly eliminated from the market for lack of business. The benefits of reciprocation are so strong that I don’t think they can be rejected without disastrous economic consequences.

There you have it: my defense of minarchy in preference to anarchy. Please feel free to comment, object, and dispute as you see fit. I’m especially interested in feedback on my scheme of interminarchic reciprocity.

&copy 1999 Kent B. Van Cleave
All Rights Reserved.

RR4 Box 287

Spencer, IN 47460