Is Government a Mistake? Exploring the Anarchist Option

By Jan Narveson

Introduction: Bastiat on Value

I was fascinated by Mr. de Guenin’s paper about Bastiat at last year’s ISIL conference in Canada, and then and there realized that this is a writer about whom I should know more. Reading through Bastiat’s Economic Harmonies[hereinafter, EH] [1] has been something of a revelation, and I heartily recommend it to all.

Bastiat’s great contribution to economics, in his own view, was his identification of service as the source of economic value. What is anything worth to anybody? In the cases where we are not dealing with what our fellow men do for us, the answer is to be found in its utility – how much the thing contributes to our satisfaction. In the case where we deal with our fellows, we are interested specifically in what they can do for us, that is, how much service they can render us – how much they can or could, by their actions, contribute to the purchaser’s satisfactions. If the providers are free beings, then the recipient will obtain those satisfactions only by agreement with that provider. It is to be assumed that the purchaser will get the best outcome he can: he will pay the lowest price feasible, for the option that is better than any known alternative.

Everyone is aware that price, which is the measure of economic value, is a function of supply and demand. But to say so is virtually to imply Bastiat’s insight: for it is people who do the “supplying” that economics is concerned with, and people who “demand” what is supplied. To say that someone has satisfied my demand is to say that he has done me a service, a good turn. That is what I pay him for. And this applies not only to the purchase of apples or automobiles, but also to the purchase of stocks and bonds, symphony performances, or incantations at the church of one’s choice. If we add to this that presumably the purchaser looks around and supposes that the deal he is about to make is as good as any he can expect at the time, even taking into account the possibility of waiting til later and concluding that it is not worth waiting, then we have a decent approximation to modern writers’ invocation of the idea that the economic agent “maximizes his utility”.

Bastiat refutes Marx before Marx even writes. No version of the “labor theory of value” can work, because I, the customer whose demands create this entire field, do not frankly care how hard you have worked, how many calories you have expended. What I do care about is that you can give me something I want, and will do so at a lower cost than I can get it for from anyone else (including myself). This tosses the Labor Theory of Value into the dustbin, along with Marx’s refinement on it, the view that value is “socially necessary labor” (a formulation which, by the way, already gives away the game, cleverly tucking into the formulation an element of the very thing the labor theory is supposed to see through and supply a more profound understanding).

Bastiat is a thoroughgoing liberal. He has no truck with proposals to take the deciding power over any given individual’s life away from that individual himself. The question is whether the state can possibly do anything else than just that.

The State and Protection

Bastiat’s basic insight into the nature of economic value has, I think, a lot to do with our subject of whether the State is necessary or desirable.

Like liberal-minded thinkers after him, Bastiat thinks of the proper domain of the state as being limited to the provision of protective services. Indeed, in his short book, The Law [2], he defines the law as “the collective organization of the individual right to lawful defense”. If we think of government as promulgating and working under the standard of law, then his definition effectively entails that government is the provider of collectivized protection service for all.

Such services may, alas, be necessary in human affairs from time to time. So long as we confine ourselves to people voluntarily exchanging things, and respecting the voluntary nature of the exchange, it would appear that we actually have no need for government. (Some think otherwise; I’ll return to that later.) But if there are people among us ready to use violence to promote their ends, things will be different. Bastiat notes (EH, 117-118) that a group of people working at some productive activity might find themselves among unpleasant neighbors, such as wild animals – or wild people – who have no respect for their activities, and that group will find it useful to defend itself. In so doing, they might well achieve economies through the division of labor, hiring some few people to learn to use guns and to carry them generally, and so on. And this, Bastiat points out, leads to the possibility that instead of confining themselves to the protection of their retainers from the violence of outsiders, they instead get into a Protection Racket – the business of protecting their customers from the agency itself:

… since they receive from the community services that are proportionate to the community’s need for security … they foment a sense of insecurity and … involve their fellow citizen in continual warfare. [EH, 117-118]

Bastiat does not say so at this point, but here we can discern both the origin and the fundamental nature of the state. People, for some reason, tend to wax mystical about the state, often identifying it with the Law, for example, and goodness knows what other virtuous entities. In Canada where I live, it is all but impossible to find a newspaper or television story that is not, in one way or another, about the doings of this or that official of government, an agency implicitly taken to be the repository of everything that is good in the community. Very few if any Canadians actually think about this – they just do it. But perhaps they should think about it.

Is the State a Gang of Thieves?

A basic hypothesis about the state was advanced much earlier on by Plato’s figure Thrasymachus in The Republic, who tells us that Justice is the Interest of the Stronger Party. This “stronger party” is the entity that we thought we were hiring to protect us. But once they have the guns, as Thrasymachus points out, the smart will discover that the threat to use them provides a wonderful way to make a living: namely, at other people’s expense. The “service” the state does for us was supposed to be that of protecting us from our real enemies – robbers, murderers, rapists, and conmen. But for the most part what we get instead is its promise to refrain from shooting us with these very guns that we have purchased for them, thinking they would be used to protect us, so long as we do what they tell us to. In Thrasymachus’ view, not only is the state a gang of thieves, but rightly so.

The ancient problem of political theory was how to distinguish between the two: the Gang of Thieves, our enemies, and the great Protector, our friend, the state. The challenge of Thrasymachus is that we are deluded in making this distinction, for the state, in truth, is merely the gang to end all gangs, the top-level mafia, able to lord it over all those below it.

Should we accept the Thrasymachean analysis? It can be advanced on different levels (and confusion among them led Thrasymachus rapidly into traps set by the wily Socrates in his ensuing discussions with him in Plato’s Republic.) Let’s sort some of these out.

1. It could be held as some kind of necessary truth about the concept of justice.

At this level, I need hardly add, it fails miserably. We obviously do not mean by ‘just’ ‘conducive to the interests of the strong’, whatever else we might mean by it. If Thrasymachus is to make out his thesis, it will have to be at some other level than the surface semantics of the term ‘justice’.

2. It might be held as a conceptual truth about states.

At this level, there is a fair amount to be said for it. We might surmise that people necessarily act in their own interests; and so, when they assume positions of political power, how could they not use it thus?

While there is something to be said for this, we must appreciate that whatever is a priori about this cannot be used to make any very interesting points about the state. This is because we ordinarily make a distinction between self-interested action, narrowly so called, benevolent actions at various levels, and pure disinterested actions, if there are any such. People as we know them tend not only to minister to their own bodily appetites, for example, but also to those of their children, lovers, and various others for whom they feel sympathy for one reason or another. And finally, there are people who appear to act on principle, too.

Now, there’s nothing to keep people in the state from acting in any of those ways, all of which are compatible with the hypothesis that people act only on their “own interests”. Everything depends on which sorts of interest the people in government actually do have and act on. And we can’t know that a priori.

3. Finally, it might be advanced as an empirical hypothesis about the state.

This, I believe, is by far the most interesting, most plausible, and most fruitful – but most difficult. Our question is: will those administering the state generally speaking be expected to do so

(a) in a way that promotes the legitimate objects of the state? – Or will they instead tend to

(b) line their own pockets at our expense? – Or

(c) promote alien ideologies that have no business being promoted by the state?

Our thesis can now be that we can expect either or both of (b) and (c) to predominate.

In all this, to be sure, I am assuming that we have adequate reason to reject several familiar ideologies. For example, some people think that the state should try to promote Equality, through progressive taxes and so forth. Others think it should promote multi-culturalism; or may that it should try to stop multi-culturalism; still others have other ideas. I assume we are united in thinking that the state should not be in the ideology business, but should instead be, simply, the servant of each and every citizen, insofar as possible. The anarchist hypothesis is that it can do that best by ceasing to exist.


Before arguing further for this, however, let’s consider the bearing of political democracy on the matter. It is, so far as I can see, almost universally supposed that Democracy solves this problem. After all, with democracy the state is us, right? How can we be supposed to be bent on robbing ourselves?

But this idea is also illusory, and was put down effectively by John Stuart Mill – who fails to appreciate how deeply he has driven a stake through the heart of democratic theory. Notes Mill:

The “people” who exercise the power are not always the same people with those over whom it is exercised; and the “self-government” spoken of is not the government of each by himself, but of each by all the rest.

And after a brief discussion, he observes,

The likings and dislikings of society, or of some powerful portion of it, are thus the main thing which has practically determined the rules laid down for general observance under the penalties of law or opinion. [John Stuart Mill, On Liberty, Ch. 1.]

The trouble with democracy is easy to trace. Democracy inevitably is majority rule. And what constrains the majority? Either (1) the Constitution, or (2) Nothing. Mr. de Jasay has beautifully explained why the two answers are so very nearly equivalent [3]. And in the absence of constraint, the majority will do its best to make everyone conform to its wishes concerning their behavior, regardless of any considerations of mere morality. On top of that, however, the majority proceeds by electing representatives and those representatives, who control the distribution of weapons among the police, are mostly interested in reelection, the perks of office, and perhaps of getting their own personal view of what the nation should be like translated into law. The possibility that government should be devoted to the common good seems, under the circumstances, rather remote.

The Law

Governments proceed by making laws, and then seeing to it, more or less, that people are compelled to obey those laws. So far, there is nothing to distinguish government from a gang of thieves. Any such distinction must depend on a further separation, between law as the decrees of the possessors of political power, and law as something else – law as the expression of a rational understanding of what people may reasonably expect and require of each other.

Many philosophers have had a go at trying to identify this law, and fortunately, some have succeeded. To begin with, the general form of law has been beautifully defined by St. Thomas Aquinas, who concludes, after a careful analysis, that law is  … an ordination of reason for the common good, promulgated by the one who is in charge of the community. [Summa Theologica, I. II, Q. 90.] Law is to be for the common good, not the good of some at the expense of others. But Aquinas himself fails to take into account the diversity of views about the good among humans. The consequence of doing so leads us in the direction of the great English liberals [4], Hobbes and Locke, and later of Kant and Mill. The idea of this abstract general law is that we are all to refrain from pursuing our lives at the uncompensated expense of others. Instead, we are to confine ourselves to activities that either leave other people as they are, or else improve their situations, as viewed by those people themselves.

M. van Notten has recorded the law among the Somali clans that is, so far as I can see, an excellent spelling-out of this libertarian law [5]. Michael van Notten, From Nation-State to Stateless Nation: The Somali Experience. See the ISIL 2001 website. The Somalis, it seems, recognize five rights: to have, use, and exchange property within the bounds of others’ rights; to own hitherto unowned items one has found by one’s own efforts; to make agreements which are then binding; to defend oneself, and help defend others if their rights are under attack; and to exact restitution from offenders. I will take this as a good working model for these purposes. The question is: what sort of social or political institutions can effectively administer these rights?

This presents a problem whose practical solubility is open to question. Rationally, I think, we would band together to disallow violence among ourselves and ward it off from any others. That was the classical idea: the state would be “minimal”, tending only to the protective needs of its people. And it still is the idea, a few centuries later, among many political thinkers. But…

It Didn’t Work

When a state is formed, this is not what happens. Instead of banding together for defense, they band together to try to see to it that everyone worships Allah, or refrains from taking marijuana or from driving in excess of 100 km/h, or any of thousands of other things of types we are all too familiar with. Bastiat points this out, often, and in a splendid way, classifying the many deviations under the heading of “legal plunder”:

Now, legal plunder can be committed in an infinite number of ways. Thus we have an infinite number of plans for organizing it: tariffs, protection, benefits, subsidies, encouragements, progressive taxation, public schools, guaranteed jobs, guaranteed profits, minimum wages, a right to relief, a right to the tools of labor, free credit, and so on, and so on. All these plans as a whole – with their common aim of legal plunder – constitute socialism. [The Law, 22]

If I give you a gun, and refrain from retaining one myself, what can I reasonably expect? Evidently, that henceforth I will do only what you allow me to do. If I don’t give you a gun, of course, there is the chance that you will avail yourself of one anyway. The reasonable solution, if available, is to negotiate mutual disarmament. That has very often not been the way of mankind, and the question is whether we can approximate to it in a satisfactory way. The answer has been mixed. Most contemporary societies are ones of substantial but imperfect mutual disarmament. As soon as this is so, our problem is that the few armed would seem to have an enormous advantage over the many unarmed.

But while this may seem to be so, everything depends on why the armed men are armed, and this may turn on many things. One is the possibility that they have somehow conquered the power they have, by aiming their guns at various people who were in consequence disposed to do what they wished. Another is that somebody hired them to have those guns for particular purposes. It is this latter that interests us.

On Hired Guns

The classic justification of the state has it that we all hired the state for our protection (to guarantee our right to life and activity). As we have seen, that has not worked too well (some think it has, of course…). But there is an alternative possibility: that each of us would hire whoever he thought would do the best job, given our budgets, to protect him from whatever he thought he needed protecting from. This is the idea that appeals to the anarchist. Instead of one agency for all, there would be an agency for each person who feels the need of an agency for the purpose, strongly enough to do what it takes to create or support it. Such agencies would be formed either asbusinesses offering their services for a price, or of neighborhood self-help associations, or combinations of those.

Who Guards the Guardians?

Of course there continues to be the same question about the smaller agencies as there is about the state: who is to control the “guardians”? In the case of anarchism, the special form of the question is: What’s to keep them from becoming the Mafia? The democrat will claim that the necessity of being re-elected puts a limit on the mafia-aspirations of the state. But it’s a very generous limit. Within the extremely wide limits democracy imposes, the state can and does plunder shamelessly and ubiquitously.

However, for anarchism, there is, unlike in the political case, a very good answer, in principle: competition. My agency will protect me not only from you, but also from your agency, should it be inclined to get grabby. And vice versa, of course. And in turn, we can expect that these agencies will be much less interested in shooting at each other than in shooting at the occasional thief.

Variable Law?

The “purposes” for which one needs protection are spelled out, in effect, as codes of law. My laws are what I want enforced; yours are what you want enforced; and so on. Terrific… But of course, laws are useless if they are issued by Hiram Smith only and not paid any attention to by Shirley Jones. This being so, the idea would be to get people with similar and compatible sets of laws interacting, while keeping those groups which would not be able to agree separate from each other. If the boundaries (spatial or otherwise) can be clearly set, then the administrative problem is easier, for it is then boundary-crossing that becomes the object of prosecution.

To an extent, we do this now, to be sure. The main useful social device for this purpose is, not surprisingly, private property. If we all respect each others’ possession of certain bits of the world, and know what these are and who has them, then the people who agreed on Law System X can interact mostly with others of similar mind. Families, for example, live in their own houses or apartments, and others enter only on invitation from the owners. This confines the family activities and arguments to members of the family, who are in considerable agreement about various things (at least).

Now the question arises whether a society in which there were multiple protection-agencies would “work”. By ‘work’ is meant that overwhelmingly, people would feel and be safe in this society, on an on-going basis, one that lasts: in short, a stable peaceable society in which people interfered minimally with the pursuits of others. I take it that a consequence of this is general prosperity; by and large, if people aren’t prospering, then there is almost certainly some aspect of their governments that is to blame.

Violence Control

There is an obvious, general recipe for achieving this: those who are inclined to violence must find it probable that their resorting to violence will not work. The price of initiated violence must be high enough so that would-be initiators do not resort to it.

Again, van Notten illustrates a workable system from the Somalis, where the behavior of all persons, and therefore all criminals, is something that their extended families are responsible for. But it should be possible, though difficult, to have an analogue in private non-family associations. When victims’ families or protective associations are owed compensation by the family or association of the miscreant, then they have a huge incentive to identify and deal with that miscreant, and meanwhile the members of the victims’ family or association has no further incentive to hound and punish the miscreant.

Alternatively, people would have no such inclinations in the first place. We make a large step toward the ideal of general disinclination to violence by bringing children up nonviolently, and in such a way as to grow up nonviolent, in the first place. But while this is true, it is not yet clear that it is reliably possible in all cases, and more important, in the world we know, many people will emerge from childhood with those impulses intact. So the question is how to deal with those others. And that in turn leads to the question, who is to deal with those others?

Which, to come to cases, leads to our immediate question here: suppose that it is always privately acting agents who deal with them. Could we expect this to work out better than if the job is left to an elected monopoly?

There are two things that could make private enforcement work, and one that would get us back to the miseries of state-supplied justice. The two we need are (1) the principle that the expenses incurred in apprehending aggressors will be paid by the aggressors; and (2) the presence of extensive competition for provision of these services.

The thing that could wreak it, perhaps, is collusion with resulting de facto monopoly.

I conclude with a few words about each.

Affording Justice

1) How will a poor man be able to get justice? The answer has to be that the man who commits violence toward the poor must pay not only to compensate for what he has taken, but also for the costs of tracking him down and bringing him to justice, including his lawyer if needed. If we can’t be fairly sure of retrieving costs, so that it costs us more to get justice than the benefits of it if we do not get it, then only those able and willing to expend further money (or time, etc.) of their own will be able to afford the protection we all seek.

In the Somali system, this is mostly a nonproblem because payments are handled by the respective families. But it seems that the same could be done by protective associations, or (what comes to the same, really) insurance companies. Perhaps not everyone would want to join such a company. Yet, as with the Somalis, the consequences of not doing so could be very severe. Anyone who got into major trouble would have nobody to fall back on and would be on his own. He could end up in a very bad way indeed.

Competitive Provision of Protection

2) It is hardly necessary to tell this audience about the benefits of competition. If rival agencies compete for the victim-consumer’s dollar, we can expect that only pretty effective ones will survive. Such is the idea anyway.

An analogue to the Somali case would be where membership in families was voluntary rather than biological. Of course, it is generally both in the sense that people usually like their families and don’t want to leave. Nonfamilial associations would have to proceed by carefully rigging their fee and payout structures in such a way as to maximize customer loyalty.

In contemporary governed states, competition is forbidden, and indeed the first and often only concern of a government or official police force is to establish and defend its turf, rather than concern itself with the protection of persons and property. This can be traced directly to the fatal problem of the state: letting somebody think he actually has authority over everybody.

The Danger of Collusion

3) But against this, we have the thesis, pressed by Nozick [6] and later elaborated by Tyler Cowen [7], that erstwhile rival agencies will end up colluding, or being bought out by other agencies, leaving the consumer with a monopoly on his hands, no choice, and no control. But that is what we have now! Even in the democratic system it is so. It would seem to be of critical importance who is right about this [8].

Cowen argues that various enforcement agencies would not be able to resist the temptation to collude with each other. Recalling that these agencies are armed, would they perhaps get into battles? Not too likely: they’re in it to make money, not to get themselves killed [9]. But the fact that company A could trounce company B or otherwise give them a hard time might lead to collusion.

I think the Somali example well illustrates the problem with Cowen’s thesis. In Somalia there are hundreds of families and clans, and their incentive for colluding with each other appears to be trivial, except in regard to danger from outside, in which case members of different clans act together to drive the invaders out. But their attempts to re-impose government on the whole country have for the past ten years or so been failures.

In an Appendix to this talk, I have sketched a most interesting idea about this, from John Hasnas, who suggests a sort of ultra-minimal-state device for keeping watchdog on the Cowen problem, as we might call it, and acting accordingly [10].

As I see it, the problem with Cowen’s argument is that it understates the power of the consumer, who has a clear interest in maintaining competitiveness among his suppliers. One can envisage any number of scenarios – e.g., a Consumers Research type organization that would keep tabs on protection agencies and blackball ones that weren’t doing good service.

A Criterion

Do we need the state? What would decide? There is a natural criterion to propose, one with a very strong claim to being the fundamental and necessary one for the purpose. The state lays down various rules and forces people, if need be, to go along with them. From the point of view of any individual who comes within the scope of those rules, the question is this: is this individual better off, given the rule and its status as enforceable, than he would be in its absence?

This requires a cost-benefit analysis. The rule is imposed, and so there are costs. Those who would have preferred to do what the rule forbids are paying by having to forego that preferred option. Still, if it is generally and effectively enforced, there is the possibility that he would be worse off without it. The rule against killing innocent people just because it pays the killer, for example, is one whose merits are easily appreciated by potential victims. If having the rule preserves one’s life as compared with not having it, then most of us want it. And, again for most of us at least, what we lose by being forbidden to kill the people we might want to kill, if any, is far less than what we gain – namely, life. For almost all of us, the benefit of this rule far outweighs its cost. Arguably there really aren’t any exceptions, for that matter. Murderers do not think that their own lives would be better if murder were generally allowed, like wearing laced shoes or parting one’s hair on the left. It’s simply too easy for others to do us in, if they should happen to take a notion to do so.

But now we have to ask: yes, but the very appeal of such a rule brings up the question whether centralized enforcement is the best way to enforce it. With centralized, tax-financed enforcement, after all, there are people whose official duty is to apprehend murderers and such, but their salaries aren’t paid by the people whose loved ones or associates are the victims and who are concerned about it. They are instead paid by everyone who has any money, whatever those people’s relations to possible or actual murderers may be. And the police draw their salaries whether they find the murderer or not. For that matter, as we know, the state has a way of prosecuting people without sufficient, or sometimes any, evidence, and when they do, what or who is to stop them? Alas, there is no alternative.

An anarchist system for handling this would be different. Here is the family or friend or associate of the victim, and these people are anxious to track down the murderer. If found, that murderer owes them a lot – perhaps even his life. And they have a right to it. Sending the fellow to prison for a long time may not strike these people as much compensation. And the time and trouble that their hired detective had to go to in order to apprehend the culprit? Who pays for that? Ideally, the answer is that the criminal pays for it, too. After all, it’s his fault. Of course, not all murderers will be able to make the necessary payment, even if we put him to involuntary labor. Some victims’ friends will prefer execution anyway – that is their “payment”, such as it is. Others will have taken out insurance against this kind of eventuality. (Should the murderer at least be required to pay his victim’s premiums? It’s a thought….)

The State as Inevitably Inefficient

With most of the laws that all governments come up with, there’s a more trenchant objection. Laws forbid actions X, Y, and Z, and they do so on the ground that X or Y or Z will bring about certain evils, E1, 2, 3… But some or perhaps many or most or almost all people will be able to do X and yet avoid the evil. For example, they might exceed the speed limit and still be very safe. Yet they will get fined anyway.

In fact, what the state will do is punish people for disobedience, rather than for inflicting damage on other people. Indeed, the state will set itself up as a sort of mini-god, jealously concerned that people do its will and not ask questions. But we are asking questions, and are in fact demanding certain answers. The answer we always demand is this: would the community (as distinct from the government) be better off if this law weren’t on the books, but instead one in which this case, and cases like it, would not be prosecuted?

For example, should jaywalking be prohibited? Not unless it would cause inconvenience or danger to drivers. That, of course, is often, but not quite always. The correct version of the law would result in only a very few jaywalkers ever being charged. In point of fact, that is close to what we have. Police, fortunately, do not usually charge jaywalkers even though they are obviously jaywalking within the meaning of the current law. But the same is not true of speeding, running red lights, or having two bathroom sinks which feed into the same drain (illegal in my community – would you believe? — Well, Yes, you would, I’m sure!).

In a proper anarchy, people would not be allowed to transgress on the properties of others, of course. But that is all they would not be allowed to do. No one would be able to prosecute anyone for things such as we have just described.

— Or wouldn’t they? If the streets were privately owned, presumably the owner could issue rules just as silly as the ones we actually have when they are publicly “owned”. However, if he did, he might find himself with very few users of his road or street. This question, at any rate, of the ownership and management of what are usually public facilities such as roads is plainly a very important one. But it should be noted that there are other ways to do it besides having a profit-making firm own the roads. There could be road co-ops, for example. But such questions are very tricky and cannot be explored here.

Public Goods and The State

Now to return to our attempt to produce a general formula. What we want is a situation in which all and only those who are actually guilty of significantly harming, damaging, or in general imposing restrictions on the legitimate liberties of others are found guilty, and required to make amends. The criterion of “amends” has to be that the situation of the victim is made to be, as nearly as practicable, as it was prior to the incursion on his liberty, with the miscreant and no one else being the one who bears the cost.

When if ever would a state-like system be justified? The classic proposals along this line are (as you may be surprised to hear) public-goods justifications. The argument is that there are certain kinds of goods such that private production of these goods is sure to be inadequate, in the nature of the case. These will be, it is argued, those cases in which the good’s benefits cannot be confined to those who pay the costs, perhaps because everybody must get the benefits whether they paid or not. In such cases, so the argument goes, we need a compulsory system, where everyone is required to pay.

The classic example of the public good in question is nonviolence. The assassin, the robber, the rapist, collect benefits from others without paying for them; the cost all is borne by the victim. To fix this, we must have a system in which everyone bears the cost of a policing establishment, which will make such people pay, and by threatening to do so, induce them not to commit these crimes in the first place. Nowadays, people argue this way about all sorts of things: pollution, garbage collection, and even the education of the public are held to be goods which must be produced by governments if they are to be had at all. (I mention these last examples because they are so obviously bad ones. Obviously garbage can be collected by private agencies who are prosecuted by any landowners they spill the garbage on without their permission, and this sometimes happens.)

The question, then, is whether the public-goods argument ever works. Nonviolence is a public good in the requisite sense, to be sure; but can it be provided without government?

The first thing to point out is that if our question is whether it can work, the answer is, obviously, that it can. For notice that it is provided without government almost all of the time, every day. Very few of us commit crimes, and we do not refrain from committing them because we think we will be caught and punished, but because we have a sense of decency and mutual interdependence. We do not want to have our goods stolen or be raped or murdered, and realizing this, we simply see the point of refraining from doing these things to others even if, as would rarely be the case, we could see any real advantage to doing so. Does the state’s presence do anything to improve this situation? Many have thought so, but I think the jury is still out on that one. The state’s presence creates a population that is quite sensitive to the demands of the state, but does it create a more moral population? One can doubt that. One has only to look at the recent history of Africa, with its litany of horrible governments which have murdered millions of their own people and brought nation after nation to ruin. They were able to do that because those dictators faced extremely compliant populations which would, by and large, do anything they were told to do by people in uniforms.

Meanwhile, if it is generally agreed that these things are wrong and that people do have a right to exact compensation from those who perpetrate such violence, so that in fact many people set up mechanisms for dealing with the presumably few who violate people’s rights, can we expect as good or better performance, relative to costs, as with the state?

The State as an Inefficient Supplier

There is ample evidence of the state’s inferior performance at prevention. Most people have locks on their doors, possibly closed-circuit TV’s, electronic sensors, and in the case of businesses, hired guards to prevent robbers. Why do they do all these things, given the presence of the state’s police?

Answer: they do not have the “presence” of the state’s police. Instead, there is a police station somewhere down town, and police officers having doughnuts, and so on. But their protective efforts are often grossly inadequate, even while their efforts are being all too adequate on the highways where they harass people for infractions of trivial or pointless rules. It is not difficult to believe that private agencies could do better. Indeed, it seems that they already are.

Pollution is more difficult, perhaps. But it is not clear how much more difficult it is. Once we know the damage being done to individual people by specific kinds of pollution, we are in a position to track down the polluters. And when we do that, we are in a position to sue them for the damage they do. These are not easy things to find out, of course – but neither are they easy when the state tries to do them for us.

What we do know, however, is that the state will do its best to impose an agenda on somebody. The Democratic State is a system for finding out which way the wind is blowing, and then exaggerating it – Zero Tolerance!, for instance. This gives the people who wanted that sort of thing a lot more than they wanted, at the expense of everyone, including the people who don’t want it.

The Last Word on Arguments for the State

Suppose that if the state provides service, it will be able to save a great deal of cost in administration. Example: socialized medicine as in Canada, where you pay for nothing above the counter. Administrative overhead, in billing people and getting the bills paid, is much lower than in private insurance schemes in the U.S. So we are told: I’ll bet anything that the high costs of administration of such schemes in the U.S. is owing to intense regulation by the government. But all right, suppose this is so. And suppose that everyone wants this service. Finally, suppose that the cost to the consumer who pays for it in taxes is actually lower than the cost would be under private medicine. Would we then have a good argument for public intervention?

Let us suppose we would. But even if it is, are its premises ever fulfilled? I don’t think so.

1. In the first place, the differences among people are so great that it is certain that not everyone would find any particular insurance package worth paying for. Private insurances companies would exploit these differences. So would charities, who will treat people for nothing, without imposing any costs on anyone.

2. The state has ample scope and motivation for spending its money on other things besides administration in the narrow sense – the savings are illusory. In Ontario, for example, the government manages to spend almost $3,000 per person – infants, children, young adults, the lot – when hardly any of them spend more than a day in hospital in a given year. (I don’t think I’m extraordinarily healthy; I last was in a hospital in 1974, for 4 days. The operation I had would now be performed in the morning and I’d be home in time for dinner.)

3. We have to remember that if it didn’t profit the providers, they wouldn’t be in this business. With an assured involuntary income, you don’t have to worry overly much about making the next dollar. Why wouldn’t you spend the money on luxuries – unneeded employees of all kinds, for example?

There is only one respectable argument for the state: that everyone (literally) would do better, on his own terms, than in its absence. This requirement is so strong that I suspect no argument for the state could ever meet it. As soon as we see that protective services are, as Bastiat recognized, services, then the plausibility of having these services provided by a monopoly, including an elected one, becomes very low indeed.

The state is, then, almost certainly a mistake. Or alternatively, the state is a fraud: it is in fact a gang of thieves, but it masquerades as a benevolent association.

The theft is of two types: (a) just plain, as in depriving us of our money, and (b) more subtle and less tangible, in imposing ways of life that people don’t want.

Can anarchy fix this? If we look at functions one by one, the answer is in the affirmative, and this for an obvious reason. Anarchy tends toward the situation in which people get what and only what they pay for. If the state provides you with something you don’t want, you won’t have to get it in an anarchic situation; if it doesn’t provide you with something you do want, all you have to do is buy it, or persuade enough other people to go along with it so as to get the job done. And if you can’t get it that way, then you’ll have to do without, instead of extracting it by force from your neighbors.

Appendix: John Hasnas’ Idea

Recently there has been a brilliant suggestion by John Hasnas on this matter. Hasnas agrees, at least for the sake of argument, that is indeed not clear who’s right as between anarchists and minimal-staters. But perhaps we can proceed in the right direction without presupposing a solution. Cowen, as he says, argues that “the private provision of rule-making, adjudicative, and enforcement services will inevitably result in an abusive monopoly”. This last would be an example of “Market Failure”. And if it indeed occurred, the state would be brought in to remedy that failure. But Hasnas detects a flaw in the argument for the state. What the state would need to do is remedy the market failure; it does not follow, he points out, that it would actually provide the services in question itself. What is needed is a “state with the power to prevent protective agency collusion, not one that required all citizens to purchase protective and adjudicative services exclusively from itself”. And he proceeds to describe such a state. It would be avery minimal state indeed:

The constitution invests the new government with only two powers; the power to prevent dangerous protective agency collusion and the power to subsidize the production of rule-making, adjudicative, and enforcement services. The exercise of these powers is divided among the three branches of the new government. The congress possesses no legislative authority of any kind. It is invested with only two functions: 1) to appoint the members of the Supreme Antitrust Court and the Supreme Public Goods Court, and 2) to impeach any president who exercises domestic power without express authorization from one of these courts. Further, the members of the two courts must be selected from among highly qualified economists who have been certified as such by the American Economic Bar Association.

This government functions only as a watch-dog; it does not itself provide any of the protective, judiciary, or legislative services normally monopolized by governments. If it turned out to be very busy, it might evolve into a more extensive government; if it turned out to have, really, nothing to do, it might eventually dissolve, and anarchism would win.

Hasnas’ idea has the intriguing property of providing a framework for an empirical test of the two contrary arguments about government. It is fairly clear that chances for getting such a framework erected in the near future are very poor, at least in any familiar nation, but the idea is a most interesting one. 


[1] Frederic Bastiat, Economic Harmonies, tr. by W. Hayden Boyers (Irvington-on-Hudson, NY: Foundation for Economic Education, 1996)

[2] Frederic Bastiat, The Law tr. by Dean Russell (Irvington-on-Hudson, NY: Foundation for Economic Education, 1977

[3] For an accessible discussion, see Anthony de Jasay, Choice, Contract and Consent, London: Institute of Economic Affairs, Hobart, Paperback #30 (1991), 114-118. This book, by the way, should be on the shelf of every thoughtful libertarian.

[4] Many readers will be surprised to see Hobbes mentioned in this group, or even for them all to be grouped together. But in fact, the fundamental Law of Nature of Hobbes, that “every man, ought to endeavor Peace, as farre as he has hope of obtaining it; and when he cannot obtain it, that he may seek, and use, all helps, and advantages of Warre” [Leviathan, Ch. XIV] is equivalent to Locke’s fundamental Law of Nature, “reason, which is that law, teaches all mankind, who will but consult it, that being all equal and independent, no one ought to harm another in his life, health, liberty, or possessions” [Second Treatise of Government, #6]; and Kant, in his great late work, the Doctrine of Right (Rechtslehre), the first part of the Metaphysik der Sitten tells us that the Universal Principle of Right is that “Every action is right which in itself, or in the maxim on which it proceeds, is such that it can coexist along with the freedom of the will of each and all in action, according to a universal law” [I., C].

[5] Michael van Notten, From Nation-State to Stateless Nation: The Somali Experience. See the ISIL 2001 website.

[6] Robert Nozick, Anarchy, State, and Utopia (New York: Basic Books 1974), ch. 2.

[7] Tyler Cowen, “Law as a Public Good” in Economics & Philosophy 8 (1992), 249, 250.

[8] David Lamb, The Africans (NY: Vintage, 1984)

[9] We are all indebted to the ceaseless efforts of Bruce Benson to shed light on these matters. See his neoclassic study, The Enterprise of Law (San Francisco: Pacific Research Foundation, 1991), and more recently, To Serve and Protect (New York: New York University Press, 1998).

[10] The paper may for the time being be seen at: