Against Anarcho-Capitalism

The Problems with the Private Protection Agencies in "The Machinery of Freedom"'

TL;DR: While David Friedman’s proposal of private law created through a protection agency market seems initially appealing, it makes unfounded assumptions that would otherwise lead to undesirable outcomes. These include assuming non-violence, equality, and a power balance between protection agencies. However, even when these assumptions are granted, the ideal market law becomes no different than the ideal state law.

“An economist, a chemist and a physicist are marooned on a desert island. Their only food is a can of beans, but they have no can opener. What are they to do?

The physicist says, ‘Let’s try and focus the tropical sun onto the lid—it might melt a hole.’

‘No,’ says the chemist. ‘We should first pour saltwater on the lid—maybe that will rust it.’

The economist interrupts: ‘You’re wasting your time with all these complicated ideas. Let’s just assume a can opener.”

Once you assume a problem’s difficulties away, an easy answer suddenly appears. We can see this especially in the social sciences, where people love sneaking unfounded assumptions into difficult issues to defend their preferred solution.

In one way, the social sciences are extremely difficult. Precision is impossible, control trials are difficult, and methodology is haphazard. James Q. Wilson summarized it cleverly by stating that “social scientists should never try to predict the future; they have trouble enough predicting the past.”

Yet in another way, the social sciences are extremely easy. The difficulties with prediction and empirical validation allow any crackpot theory to be plausibly defended. With methodology in dispute, it isn’t too difficult to cherry-pick research supporting some preferred conclusion.

Carl von Clausewitz’s description of war applies as well to the social sciences, which he defines not as “an act of the will aimed at inanimate matter, as it is in the mechanical arts” but rather “an act of the will aimed at a living entity that reacts.” No one can ever be certain, so one’s guess isn’t usually much better than someone else’s.

This is evident from answers to one of the most difficult questions in the social sciences: how should human societies be organized? While responses vary considerably, anarchists boldly answer that “it shouldn’t be.” So what is this provocative reply getting at?

Anarchists are highly diverse and are motivated by a wide range of considerations. Some may view government authority as an incurable wrong against people’s natural rights. Others on the anti-capitalist left might consider the state an obstacle to an egalitarian utopia. And yet others might recognize the errors with the prevailing theories of political legitimacy and default to anarchism for want of a satisfactory justification for authority.

And while anarchists can be found across the political spectrum overall, they tend to overwhelmingly be “ideal theorists.” Ideal theories develop political systems based on ideal assumptions about human nature. Ideal theories might assume that people will never violate anyone’s rights, will always obey the law, are perfectly impartial to everyone else, are never motivated by profit, love complete strangers as much as their close relatives, and are of course, omnipotent.

Anarchy usually isn’t compared with the world as it is, but as it should be (or more accurately, what theorists want it to be).

While these assumptions on human nature would ordinarily discredit any political theory, some moral philosophers still grasp onto their cultivated ideology and instead pin the failure on human nature. It’s us who are in the wrong, not their theory.

Other anarchists may still believe that while the underlying principle of their ideal political order is correct, but admit that anarchism bears no real-world application given human nature.

However, any serious political philosophy must grapple with human nature to avoid becoming foolish utopianism. Any political model can work given certain assumptions, but the model we actually care about works in the real world.

By assuming that humans are angels, we could get anarchism. And if assumed a can-opener when all we have is a sealed can, we could get beans. But humans aren’t angels, and the beans are still trapped in the can.

And each ideological camp selectively chooses where they believe human nature is important and where assumptions can be made.

Socialists examine private agent behavior through behavioral economics research but apply ideal assumptions to public agents. Anarcho-capitalists do the reverse by examining public agent behavior through public choice research, but apply ideal assumptions to private agents. Both views are biased, defective, and useless.

Socialists assume a perfect state; anarchists assume perfect people. And as soon as these theories come in contact with real-world people and conditions, they fail spectacularly and become relegated only to academic thought experiments among the tenured and incompetent and edgy youths craving a logo of non-conformity.

So why be an ideal theorist? Strangely, so many thinkers are wedded to their preferred utopian theory when they must be aware of utopia’s impossibility. They might be motivated by social signaling, personal aesthetics, or laziness. But regardless of personal motivation, ideal political theories—and all of their assumptions—are worthless if they bear no practical application.

However, David Friedman’s “The Machinery of Freedom” (hereafter TMOF) takes the practical application of anarchy seriously. Its proposal of a stateless society is based on an initially convincing economic analysis grounded in individual self-interest rather than transcendental rights or perfectly moral humans. You can’t get less idealistic than assuming only pure self-interest.

However, despite the practical focus on economic incentives, Friedman still slips in ideal assumptions about human nature. Either these assumptions must be relaxed, and the resulting consequences be openly acknowledged, or this anarchist theory must be diminished to the uselessness of ideal theory. This article will review the fatal assumptions made in what I consider to be the best defense of anarcho-capitalism.

I should note that I feel uncomfortable writing this review for two reasons.

First, I view a book's merit not in its correctness, which often takes the form of obvious facts and unoriginal arguments, but in its ability to make the reader consider unique ideas. This is where TMOF succeeds.

Second, this book has received unfair criticism resulting from both a misunderstanding of economics and a status-quo bias in favor of the state. I’m sympathetic to many of the arguments in TMOF and was fascinated enough with the argument on behalf of private legal systems that I thought the idea deserved closer scrutiny.

This review won’t address the provision of public goods. I think the private sector’s ability to provide and price public goods is very underrated. Rather, I will argue that despite assuming rational self-interest, TMOF’s proposal for a private government makes assumptions about human nature that, when relaxed, lead to immoral outcomes that anarcho-capitalists will either have to address or accept.

But, given that I’m writing about moral and political philosophy, I have to examine the merits of what I consider to be the best theory of anarchism available—since it tries to develop a system of lawless organization by taking humans as they really are.

Friedman’s Protection Agencies

To oversimply, the main premises of Friedman’s private protection agencies (hereafter P.A.s) are:

1. Private parties will contract with P.A.s for protection and dispute resolution services.

2. P.A.s will develop their own system of law, police, arbitration, and insurance that governments currently monopolize.

3. Disputes between individuals of different P.A.s would be peacefully mediated because violent conflict is expensive.

4. P.A.s that fail to provide justice for their customers will be uncompetitive in the marketplace, whereas successful ones will grow.

5. The competitive nature of the P.A.s will incentivize P.A.s to create a system of laws and enforcement that is more efficient than the government.

Friedman’s description of his P.A. system is brief and represents an extremely small portion of TMOF, which remains one of the definitive works of libertarian thought.

And his argument is extremely seductive. If you (1) accept that the private sector is generally more efficient than the public sector, given the feedback mechanisms of the market, and (2) that law is a good that can be provided at differing levels of efficiency, then (3) you should accept that private sector will generally produce law more efficiently than the public sector.

I agree with this argument, but “more efficiently” bears some nasty consequences that optimistic anarchists often overlook.

Assuming Non-Violence

The fatal flaw that brings Friedman’s P.A. proposal back to ideal theory is the assumption of non-violent resolution. Whenever there is a potential conflict, Friedman assumes that P.A.s would always prefer to cooperate rather than fight:

wars are very expensive, and Tannahelp and Dawn Defense [protection agencies] are both profit-making corporations, more interested in saving money than face.

Saying that warfare would be rare due to its cost is like saying production or innovation would be rare due to its costs. Costs alone aren’t determinative of rational decision-making, but expected profit—revenue minus expenses—is.

To assume that violence would never be an option under the self-interest model that Friedman uses, you would need a model of warfare to show why it would be a suboptimal decision that a free market would push out of existence.

Phillip T. Hoffman provides one such model in “Why Did Europe Conquer the World?” in his Tournament Model of state warfare. Hoffman models the expected net benefits of warfare as a function of the prize value (wealth, security, glory, etc.), the costs of mobilizing resources, and innovation benefits from participation (the more tournaments you’re in, the better your future chances are).

Tournaments are expensive, risky, and usually negative-sum, given that many participants risk getting nothing despite their expended costs. From a social perspective, they’re inefficient. But high-stake contests can be worth it if the prize is big enough and the costs can be held down. A tournament’s expected return for a participant can be positive, as history has shown and successful P.A.s would find out.

Anarchists have to do better than assume that no one will want to participate in a tournament. Violence may have a cost, but it’s not costs alone that matter, but expected return.

The early trading corporations (see the V.O.C. and East India Company) have demonstrated that profit-seeking can go hand-in-hand with warfare. World trade, writes historians Ronald Findlay and Kevin O'Rourke, often came "from the barrel of a Maxim gun, the edge of a scimitar, or the ferocity of nomadic horsemen."¹ War and trade have historically been complements, not substitutes.

The reality of warfare should be examined and modeled rather than ignored for the sake of theory. One might argue that modernity has reduced incentives for violence since cooperation is linked to wealth—mutual gain reduces violence. However, this assumption, too, conveniently disregards warfare.

There is this widespread misconception that violence is motivated by poverty or want. Yet violence isn’t only (or even primarily) wealth-motivated. Some of the world’s wealthiest cities also have the highest murder rates. Positive-sum transactions aren’t enough to eliminate human warfare, as the existence of violent conflict since the industrial revolution has shown.

Instead, as Thomas Hobbes cited, status and security are actual and often overlooked explanations for violence. Crime is often motivated by self-help and one's personal sense of justice. To the perpetrator, crime isn't a violation of another's natural rights but an enforcement of their rights.² And it’s been the rise of the state and police enforcement that has reduced the need for self-help and, therefore, violence.³ Despite humanity unlocking positive-sum growth since the dawn of capitalism and the industrial revolution, there has been no shortage of motivations for human violence.

However, we can see an anarcho-capitalist system on display in American prisons (ironically), where some of TMOF’s predictions prove correct, but only to an extent.

Despite the state controlling every facet of prisoners’ lives, the state has no authority over prisoner disputes. Even speaking to guards is taboo, and snitching on another inmate is practically a death sentence. Yet living in close quarters requires a dispute resolution system. Where the state has been deemed illegitimate, prison gangs have taken their place.

These gangs help ensure mutual protection for one another, like a P.A. and tend to be split along racial lines. While seemingly primitive, race is a signal of a group membership. If someone of one race starts trouble with a member of another, the perpetrator’s color would inform his victim of who he should go to for the matter to be addressed.

And when there is a dispute between members of different prison gangs, these gangs wouldn’t likely engage in costly warfare. They generate income from selling drugs, which becomes much harder during murder-related prison lockdowns.

Importantly for Friedman’s argument, there is a fair bit of evidence that prison gangs reduce violence by resolving disputes and allocating public goods. P.A.s can, in fact, take the roles of a state.

Yet it shouldn’t be surprising that prison gangs are also extremely violent, and power plays an outsized role in organization and mediating disputes. Despite prison gangs’ incentive for peace, there are still clear incentives for brutality that must be addressed. Sometimes, people would rather save face than money.

So violence isn’t a fair assumption to make. At the very least, an anarchic society may engage in violence. Now the below will argue why this is very probable.

Assuming Equality

Legal equality, as we understand it currently, would not exist under a P.A. market regime. Our legal status would depend on our contract with the P.A. rather than something recognized inherently in all humans.

P.A.s would generate income through customer premiums and pay costs when there are disputes. So naturally, it will set its prices low enough to attract customers and minimize costs associated with conflicts and protection.

And just as health insurance companies prefer to insure the young and healthy over the old and ill, a P.A. would prefer to insure the wealthy and powerful over the weak and crime-prone. If not, anarchists must provide a rationale for why P.A. would choose to undertake such costs for the sake of morality—while also arguing that behavior will be driven by self-interest.

I’ll call these four types of customers that the P.A. would be dealing with the “more powerful,” “less powerful,” “vulnerable,” and “crime-prone.”

Friedman doesn’t consider unequal laws as seriously as he should. While he recognizes that private government would create different laws for different members, he believes these discrepancies would be largely similar to varying types of legal regimes currently.

Before labelling a society in which different people are under different laws chaotic and unjust, remember that in our society the law under which you are judged depends on the country, state, and even city in which you happen to be.

However, this is a false comparison. All levels of American government are bound to provide their citizens with due process and equal protection under the laws. There are certain things that U.S. states and localities can’t do.

But P.A.s would have no incentive to provide universal due process and equality under the law unless they would increase profits. Those principles represent costs to the P.A.s when dealing with consumers at different power levels. Airlines wouldn’t provide everyone first-class service, so why would we expect private companies to provide equal law?

Any anarchist theory must admit outright that a free market in protection will provide different services at different prices for different types of consumers. Some people would be better off under this free-market system of security, but other people will also be made worse off.

People who could protect themselves (i.e. the more powerful) will be ideal customers. They may have inherent strength or may have enough wealth to purchase the services of those who do. People generally wouldn’t threaten them, and they could probably handle smaller disputes on their own, so the P.A. wouldn’t have to expend much in resources to protect them. P.A.s will also want to hire the services of their more powerful clients to enforce their rules.

People who are entirely vulnerable or crime-prone will only be accepted at a prohibitively high cost, which, given their position, likely means they will be priced out of the P.A. system (vulnerable and crime-prone people aren’t exactly wealthy).

The less powerful would still be customers but would get much less in services. They may be able to only afford an “emergencies only” plan from the P.A., whereas wealthier patrons may get 24/7 security and personal guards.

The main point is that a P.A.s interests will align more closely with some people than others.

A profitable P.A. will want to know a client’s likelihood of violent behavior, level of vulnerability, and inherent strength. They may require information about one's genetics, gender, I.Q., family income, and general social environment to come up with an acceptable price. And a P.A.’s prices will adjust according to the choices, lifestyle, and circumstances of their clients.

So now we have crime-prone and vulnerable individuals who have been priced out of P.A. market, and the more and less powerful clients obtaining different levels of status within P.A.s. Can we really assume there will be no warfare?

And if disputes arise between P.A. members, who will a P.A. side with? P.A.s who care about profit will favor their most powerful and valuable clientele. And if more powerful members can get away with violating the rights of the less-powerful lower class, why would they refrain from doing so?

The less powerful will have to accept unfavorably rulings when committed by the more powerful. They could try going elsewhere, but other P.A.s will have the same arrangement of privileged, more powerful customers at the expense of lower-power ones. All the moral P.A.s have gone out of business.

Less-powerful customers will have to give up if they want to contract with a P.A., like maybe signing a long-term contract binding them permanently to a specific P.A.?

To be more direct, anarcho-capitalism would likely lead to a form of a market-based feudal system of protection. Powerful individuals will pay less, the powerless will pay more, the vulnerable will be excluded, and roving criminal bandits who no one wants to contract with will be a constant threat.

But don’t take this as a moral judgment. Maybe reasonable people would accept this type of society. Yet that’s a separate question. What should be noted is that inequality in power would be the outcome. So long as violence is valuable and the potential for violence is distributed unequally, then value would be distributed unequally in a P.A. market. Some people would be ideal clients, others less ideal clients, and others, not even clients at all. And P.A.s would put the interests of their ideal clientele first.

In no scenario would a P.A. treat all clients equally. Everyone has relevant differences that would fetch different prices in a P.A. market. And any insurance agency would know better than to charge high-expense individuals the same as low-expense ones. Price discrimination would not be a method of if, but how.

So how did we come to this assumption of human equality? Why do anarchists assume that P.A.s will negotiate with parties on equal terms? To a large part, violence has been responsible for the current state of legal equality that we all take for granted.

State violence has historically created liberties for the lower classes, from Ancient Greece to the era of mass military mobilization (which led to the progressive income tax and the welfare state). Violence has been a leveler throughout history, and it is hard to take any ideology seriously which fails to take violence seriously.⁴

The state grew because of violence, and legal equality between different groups of people has been enforced through violence. To develop a system of human organization, we can’t take violence out of the equation.

Assuming a Power Balance

There will undoubtedly be hierarchies within P.A.s between the most and least powerful since P.A.s have stronger incentives to look after the interests of the former, even at the expense of the latter. And if P.A.s themselves are also unequal in power among one another, then we can expect a similar power hierarchy between P.A.s where "justice" is the will of the stronger P.A.⁵

Friedman assumes that P.A.s would always negotiate with one another rather than fight, given the costs of war and the benefits of trade. Yet warfare is cheap when one party is much stronger than the other. And warfare generates more benefits than trade. Why exchange when you can take? Why contract when you can enslave?

The stronger P.A. would be leaving money on the table if it didn’t oppress the weaker P.A. It might be immoral, but Friedman isn’t assuming morality, only self-interest. If it sticks to its morals about peace and autonomy, its clients will exit and gravitate towards a P.A. that prioritizes profits.

Especially given that P.A.s primarily receive income from their clients, their loyalties to other P.A.s would be severely limited. They wouldn’t favor a competitor over their clients.

A power balance between P.A.s is possible, but it would require giving more value to power. P.A.s will compete even harder for more powerful clients to serve as their protectors or aggressors. Again, violence is valuable.

A system of justice (as we understand it) between self-interested agents would require equality in strength. As the Athenians said to the Melians, "the question of justice arises only between parties equal in strength, and that the strong do what they can, and the weak suffer what they must.”

We shouldn’t take justice, fairness, and autonomy for granted. Once we assume organization only by self-interest, moral principles represent costs to be minimized. P.A.s will only be as fair and peaceful as a cost-benefit analysis will lead them.

Friedman also assumes that any conflicts between P.A.s can be settled peacefully through a mediating, free-market court, which would always be honest and just.

Several objections may be raised to such free-market courts. The first is that they would sell justice by deciding in favor of the highest bidder. That would be suicidal; unless they maintained a reputation for honesty, they would have no customers—unlike our present judges.

Again, the obvious flaw isn’t taken as seriously as it should. These courts would make decisions based on profitability, not on honesty or abstract notions of justice. And interest and morality aren’t always congruous. The court will want to make decisions that serve the interests of the wealthier, or more powerful, P.A., for fear of losing their business to another court or the stronger P.A.’s threats of violence. Although there are different ways these courts can be arranged, any method would have problems.

If these courts were independently powerful by being financed through their P.A. clientele, then its subject to the same problems as above. Its justice will be whatever the better-positioned P.A. would demand. If it sided with who was right rather than what the client wanted, then it would lose out on business. Its P.A. clients won’t care about honesty but their interests. The “justice of the strong” problem is still present.

But if these courts didn’t rely on clients, then they wouldn’t have power and, therefore, wouldn’t be in a position to enforce their decisions, like the current International Court of Justice. Successful P.A.s will have no incentive to obey an opinion from mediators when that opinion is against their client’s interests. The mediators are free to make decisions, but who can enforce them?

We see this all the time in the anarchic setting of international relations.⁶

For instance, say Elizabeth is the C.E.O. of the England P.A., and Phillip is the C.E.O. of the Spain P.A., a threatening competitor of the England P.A.

One day, a customer of the England P.A., who we’ll call Drake, steals some precious cargo carried by customers of the Spain P.A. Drake then shares a portion of the stolen loot with Elizabeth, which she redistributes to the rest of her clientele to keep them happy. Then, Phillip demands that Elizabeth return the stolen loot and extradite Drake to the Spain P.A. for trial.

Elizabeth could either: (1) Return the loot, give up Drake to her threatening Spanish competitor, and risk backlash from her clientele, or (2) go tell Phillip to pound sand.

I’m unsure what an actual P.A. would choose, but history has proven that the latter strategy worked best.

Friedman does use Medieval Iceland as a historical example of a successful P.A. system. Medieval Iceland was a notable outlier among medieval European governments in its absence of a ruling monarch. However, the vendetta culture as reported in the sagas isn’t what comfortable Americans would want to sign up for, nor what ideal theorists consider an ideal society.

Medieval Iceland's adjudication system was more of a fact-finding case for regulating violent feuds—rather than an ideal system of justice (unless you consider revenge killings an ideal system of justice).⁷

And Friedman is aware the decentralized medieval Icelandic system of kin-based clans and private law relied on a power balance.

A second objection is that the rich (or powerful) could commit crimes with impunity, since nobody would be able to enforce judgment against them. Where power is sufficiently concentrated this might be true; this was one of the problems which led to the eventual breakdown of the Icelandic legal system in the thirteenth century. But so long as power was reasonably dispersed, as it seems to have been for the first two centuries after the system was established, this was a less serious problem.

And a power balance is a strong assumption to make. Given that a power imbalance led to the end of anarchic Iceland— which Friedman uses as an example of a successful P.A. market—anarchists need to address how a balance of power is to be maintained and what should be done when it isn’t.

And historically, power balances haven’t created a stable system of decentralized power—but a stable system of centralized power— such as through a united confederacy or nation. Ironically, the most durable power balance in Europe was in Great Britain—between the central monarch, rural lords, and urban bourgeois.⁸ And, in terms of fiscal capacity, Great Britain then became the most centralized state in human history.⁹

And this centralizing process wasn’t all through consequent and oppression, but also through justice and fairness. In medieval England, localities were biased in favor of the ruling elite for the above reasons. A commoner’s best hope for obtaining justice was by appealing bad decisions to the Court of King’s Bench, where judgments were made without local elite biases. Where a fiefdom’s more powerful clients were given the privilege in a system of “justice,” the less powerful looked to a centralized state for dispensing actual justice.

It doesn’t take an army for power to be centralized as well. In the 12th century, the Pope in Rome consolidated power by settling ecclesiastical disputes that local jurisdictions didn’t have the incentive to adjudicate fairly.

Either an ambitious P.A. topples the persistent power balance, or the power balance persists, in which case a strong central state would likely develop. Regardless of what route is taken, historically, peace has been enforced through top-down control.

A Better Libertarianism

When individuals privately contract with P.A.s, violence will undoubtedly become much more valuable, as it had been throughout pre-state history. The real choice isn’t between a utopian society, where everyone respects everyone else’s natural rights, and a coercive government—but between coercive people and a coercive state.

However, let’s say that all the assumptions I discussed above are actually accurate. Let’s assume that P.A.s will never engage in violence, value all parties equally, and the P.A. market is perfectly competitive.

Then, the most likely outcome would be a market standard among the P.A.s. Everyone would accept a standard contract of adhesion, as exists for many other contracts. Of course, there would be P.A.s with different rules, like how current governments impose different laws. But given the above assumptions, there will be a baseline of legal equality.

We’d all converge on terms everyone would agree to, and no one could find a P.A. offering significantly better or worse terms. Again, we’re assuming a perfectly competitive market that treats all parties equally.

I’m unsure what the real difference would be between people being subject to a fixed market or a fixed government standard. There wouldn’t be actual “consent” to the regime in either case.

The law would be what reasonable people would agree to without regard to their power—basically a “Veil of Ignorance” standard.

The problem with anarcho-capitalism, described above, is that power becomes valuable. Yet in a political system that obeys moral rules, power would be disregarded. There, society would create and obey rules based on principles that parties would agree to behind a veil of ignorance. These would include legal equality, personal autonomy, and even adequate social insurance.

English kings and Catholic popes have used this method of dispensing fair justice as a legitimate tool of centralization. People choose to be ruled by reasonable law, even if it’s by a faraway regime. And especially when local authority tends to be corrupt in favor of the powerful elite.

I’m not sure what the actual difference would be between an ideal market standard and a reasonable state regime. Both versions of the law should be the same: what people would accept under ideal conditions.

This is what the law should be. We can think of this standard as resulting from either an ideal P.A. market or ideal deliberation (i.e. the Veil of Ignorance). Yet, regardless of what thought experiment we use, we would still be nonideal humans trying to produce outcomes that would be created under ideal conditions. And we can’t get there by pretending we are acting under ideal conditions but by holding that ideal outcome as the standard while we reason under a method of reflective equilibrium.

There is a truth to libertarianism: that moral rule is what people would reasonably choose. And if reason doesn’t rule in a lawless society, then power will. Anarchy won’t create self-rule but rule by the strong.

For those still unconvinced, we can use an analogy of an inquisitorial religious regime. While we can agree that religious oppression is evil, say we lived in a world where people killed those who they deemed to be heretics with complete impunity. In that case, we might accept a centralized inquisition regime because it’s better than the alternative of anarchic killings. Religious killing would require a rule-based trial.

This was, in fact, why medieval inquisition institutions developed; oppression had to be taken out of the hands of the people. And the church created reason-based procedures to determine heretical status, despite the depravity of the actual outcome.

However, once you have procedural rules by reason, you’ll soon have substantive rules by reason. And soon, the whole edifice of oppression comes down as “power” gets replaced by “reason.” It is the slippery slope of reason.

First, we had indiscriminate killings of whoever people labeled as heretics. Our power and animal spirits dictated oppression. Then we had an inquisitional legal system, where heretical status and punishments were determined by agreed-upon rules and procedures. Then we had the Enlightenment, where we discovered that reason requires that we do away with the concept of “heretics” altogether.

We had replaced “power” with “reason” first at the procedural stage, then at the substantive stage. The state serves a similar role, by helping us get from anarchic power-based rule to reason-based rule.

And no, everyone respecting everyone else’s religious doctrines isn’t an option. Not if we’re dealing with humans as they are. While that’s the ideal, anarchy would similarly be ideal if men were angels. But if we’re going to develop a political theory, it has to start with how we are. While religious prosecution is an evil, any theory on how to reduce this evil needs to be practical.

Friedman’s work deserves praise for trying to take anarchism out of the useless fringe of ideal theory and into the seriousness of nonideal theory. However, his theory still relies on ideal assumptions. Either anarchists must bite the bullet and accept the undesirable outcomes of their “ideal” state, or they must develop an additional model on human nature and violence that can help alleviate these legitimate concerns.

Anarchism sounds like a very easy answer to a very difficult problem. But that’s only because it’s assuming away exactly what makes the problem difficult. So far, anarcho-capitalism belongs in the same league as other anarchist utopian theories: appealing in theory, but useless in practice.

1. See “Power and Plenty: Trade, War, and the World Economy in the Second Millennium” by Ronald Findlay and Kevin O'Rourke.

2. See a discussion on this point in Steven Pinker’s “The Better Angels of Our Nature: Why Violence has Declined.”

3. See the argument for this point again in Steven Pinker’s “The Better Angels of Our Nature: Why Violence has Declined.”

4. See Walter Scheidel’s “The Great Leveler: Violence and the History of Inequality from the Stone Age to the Twenty-First Century” for a discussion of how warfare has been an equalizing force in human history.

5. Equal power between P.A.s will likely drive the costs of warfare above its expected benefits. Even animals understand that fights on equal terms aren’t worth it.

6. International relations reflect the best model of a Hobbesian state of nature.

7. This brief account of Medieval Iceland comes from Chris Wickham’s “Medieval Europe.”

8. See Chapter 17 of Francis Fukuyama’s “The Origins of Political Order.”

9. Peer Vries makes this argument in “State, Economy and the Great Divergence: Great Britain and China” where he concludes that there can be “no doubt that taxes were much higher in ‘laissez-faire Britain’ than in ‘oriental despotic China.’“