The state is often ascribed a special sort of authority, one that obliges citizens to obey its commands and entitles the state to enforce those commands through threats of violence. This book argues that this notion is a moral illusion: no one has ever possessed that sort of authority.


Anarchists face a catch-22: most people will not give anarchism a serious hearing because they are convinced that the position is crazy; they are convinced that the position is crazy because they do not understand it; and they do not understand it because they will not give it a serious hearing.

Bryan Caplan, who went to Berkeley with Huemer, writes:

I’ve read almost every major work of libertarian political philosophy ever published. In my view, Michael Huemer’s new The Problem of Political Authority: An Examination of the Right to Coerce and the Duty to Obey is the best book in the genre.

Michael Huemer’s genius is that he so often presents the kind of arguments that aren’t on your radar at all but are obvious as soon as they’re presented to you. The Problem of Political Authority is like that. Unlike Nozick or other deontological libertarians, he doesn’t begin with any strong premises about property rights. He just begins with common-sense beliefs:

  1. A presumption against coercion—People shouldn’t attack, kill, kidnap, steal from, or defraud other people outside of a few special circumstances (like with consent, or to prevent those other people from harming third parties, or to prevent much worse consequences).
  2. A recognition that the government is coercive—While direct physical violence isn’t a routine punishment, the threat of violence underpins the incremental series of penalties that governments do routinely impose. For example, drivers who run red lights on empty streets might be fined $200, but if they repeatedly fail to pay, their driver’s licenses may be revoked, and if they continue to drive anyway, they may be ultimately imprisoned (kidnapped?).
  3. The imposition of the burden of proof on the government—If the state claims to have an authority to do what ordinary individuals can’t, then this claim has to be justified by some special property of the state per se.

The first two are clear, and the second is introduced with a story (Nozick tells a similar story in Anarchy, State, and Utopia called “The Tale of the Slave”) about a random vigilante who resolves to deal with local vandals by capturing and imprisoning them in his basement. When he goes door to door demanding payment from confused neighbors, on threat of similar imprisonment, the neighbors obviously have no obligation to pay for this scheme planned without their consent. This obligation doesn’t materialize even if the vigilante is particularly effective, if he operates with the support of 67% of your neighbors, or if he has been operating for a long time.

So the statist is put on the defensive, forced to justify the morally relevant difference between the vigilante and the state that would (1) uniquely entitle the state to enforce broad, content-independent rules on its citizens and (2) create an obligation on citizens to obey the government even when they wouldn’t be obligated to obey the same commands from an individual.

Huemer distinguishes philosophical anarchism, the view that there are no political obligations, from political anarchism, the view that the government should actually be abolished. Political obligation is not on strong footing and the dominant view towards it in the academy today is skepticism, but Huemer wants more. He aims to prove that no one has any obligation to obey any modern state, and that they should all be abolished.

In the first half of the book, he refutes all of the common arguments for political authority, leaving the only justification for the state’s existence the consequentialist argument that an anarchist society would be much worse. The second half of the book tackles this problem by outlining how essential functions that are currently government-operated—the police, courts, and national defense—could be effectively privatized.

The first half is the smoother one. Huemer comprehensively destroys theories that justify political obligations through a social contract or democratic ideals of fairness. First, the social contract theories all purport that citizens somehow meaningfully consent to their political obligations:

A Real Social Contract: Locke argues that we are currently constrained by an unwritten covenant on the land in our borders that was made explicitly at the founding of our society in the distant past. American high school civics courses describe the social contract along these lines as well. But this is reserved for historical interest (or maybe in the future for the colonization of outer space?) since there are no modern states that weren’t founded on “usurpation of conquest.”

An Implicit Social Contract: You consent through one of a few non-explicit means.

  1. Passive Consent: You consent by not objecting, the same way you’d consent to a choice of restaurant by not objecting when I ask the group “Any objections?”
  2. Consent by Accepting Benefits: You consent by receiving government services, the same way you consent to pay a restaurant simply by eating their food, since it’s well known that you have to pay for the food afterward.
  3. Consent by Participation: You consent by voting or running in elections, the same way you consent to pay into a lottery by entering your name, if that scheme was well known beforehand.
  4. Consent through Presence: You consent by simply staying within the country’s borders, the same way you consent to follow my rules about removing your shoes when you’re at my house party.

These all fail to satisfy the normal understanding of consent, which requires reasonable ways of opting out, including explicit dissent. If anarchists explicitly decline the services of the state, the state will insist that they have political obligations anyway. Receiving government services or participating in the political system can’t be taken as consent to the social contract since taxes and obedience to the state would still be enforced otherwise.

The strongest of the passive consent arguments is consent by presence, even though it requires the dubious claim that the state literally owns all of the territory over which its claims jurisdiction, in the same way that I own the house in which I conduct my house party. This still has two issues:

First, that most people don’t think of the state as literally owning all of the land in their country in the same way that citizens do, but merely that states have a kind of political authority over it. In fact, most (all?) modern states would struggle to justify ownership of their territory since they gained it through conquest (this is different from criticizing individuals for indirectly owning land that was once conquered unjustifiably, since the state claims to be the same entity that did the conquering).

Second, and less importantly, I might be unjustified in enforcing arbitrary rules on my party guests if it was well known that they would likely die or lose their whole livelihood in leaving the party, which could be a better analogy for emigration from some states.

Hypothetical Consent: While you can’t dissent to the social contract, you would certainly have consented under certain hypothetical conditions. This is akin to how a doctor would be justified in operating on an unconscious patient in an emergency, since any reasonable patient would have consented to this procedure if awake.

Notice that this is already suspicious. For example, if it was well known that the patient had a stated objection to undergoing surgery under these circumstances, then the doctor would at least be unable to use any kind of consent-based argument to justify operating.

A Hypothetical Social Contract: While you can’t dissent to the social contract, this is fine because you would consent under “ideal conditions of deliberation,” if you were well-informed, skilled at reasoning, disposed to make fair agreements, and not motivated by self-interest. Rawls says that if you weren’t influenced by your position in society, then you would consent to some social contract.

This doesn’t make sense either. Rawls, the most important political philosopher of the last century, argues that people with different religious, moral, and philosophical views would somehow agree on a political system, but doesn’t really explain how this would happen, unless you suppose that differences in political philosophy are driven entirely by self-interest. Huemer takes offense:

Anarchist thinkers do not, as a rule, appear particularly less rational, informed, or reasonable than partisans of other political views. They do not, for example, refuse to offer reasons for their views, refuse to consider objections, or refuse to take into account the interests of others. It is therefore difficult to identify any non-question-begging rationale for excluding them from the class of people whose agreement is sought. Unless anarchists are to be simply excluded from the agreement, hypothetical social contract theorists owe us an account of how political anarchists could be convinced to accept government.

. . .

Rawls’s conclusion does not follow from his stated premises. Rawls assumes that, once all particular inclinations and all individual characteristics (or knowledge thereof) are excised, all reasonable and rational people will be convinced by the same arguments. This assumption rests on a particular diagnosis of the phenomenon of widespread intellectual disagreement: that such disagreement is due entirely to such factors as ignorance, irrationality, and biases created by knowledge of one’s individual characteristics.

Maybe even more pressingly, the fact that a reasonable person would agree to your contract doesn’t by itself justify you imposing that contract by force.

Imagine that an employer approaches a prospective employee with an entirely fair, reasonable, and attractive job offer, including generous pay, reasonable hours, pleasant working conditions, and so on. If the worker were fully informed, rational, and reasonable, he would accept the employment offer. Nevertheless, the employer is not ethically entitled to coerce the employee into working for him in the event that the employee, however unreasonably, declines the offer. The reasonableness of the offer, together with hypothetical consent, would bear very little ethical weight, at most slightly mitigating the wrongness of imposing forced labor.

Similar judgments apply to other exercises of coercion that would normally require consent: it is not permissible for a physician to coercively impose a medical procedure on a patient, even if the patient was unreasonable to refuse the treatment; nor for a vendor to extort money from a customer, even if the customer was unreasonable to refuse to buy the vendor’s product; nor for a boxer to compel another boxer to fight, even if the latter was unreasonable to reject the offer of a match.

“Arjun Panickssery is the greatest book review writer in the history of book review writers, maybe ever.” -Aristotle of Stagira

Next are the democratic arguments for political obligation:

Thomas Christiano has developed the Argument from Equality as an argument for political obligation, roughly as follows:

1.  Individuals are obligated to treat other members of their society as equals and not to treat them as inferiors.
2.  To treat others as equals and not as inferiors, one must obey democratic laws.
3.  Therefore, individuals are obligated to obey democratic laws.


Why should we accept the premises of the Argument from Equality? Begin with premise (1). Christiano advances the following subargument, in paraphrase:
1a.  Justice requires giving each person his due and treating like cases alike.
1b.  All members of one’s society have equal moral status.
1c.  Therefore, justice requires treating other members of one’s society as equals.

Next, why should one accept premise (2)? There appear to be two subarguments for this. The first appeals to the idea of placing one’s judgment above that of others:
2a.  To disobey a democratic law is to place one’s judgment above that of other members of one’s society.
2b.  To place one’s judgment above that of others is to treat those others as inferiors.
2c.  Therefore, to disobey a democratic law is to treat other members of one’s society as inferiors.18 (from 2a, 2b)

The second subargument appeals to the obligation to support democracy:
2d.  Treating others as equals requires supporting the equal advancement of their interests.
2e.  Democracy is crucial to the equal advancement of persons’ interests.
2f.  To support democracy, one must obey democratic laws.
2g.  Therefore, treating others as equals requires obeying democratic laws.19 (from 2d–2f)

This argument seemed so silly to me that I was concerned I misunderstood it. Huemer for his part attacks the idea that “respecting others’ judgments” can generally override individual rights, that “treating others as equals” requires deference to their opinions on coercive laws, and that there is any strong duty to “support the equal advancement of others’ interests.” This last claim would imply that just as Christiano would support an obligation to obey a $50 tax instead of spending it on personal consumption, he would have to support an obligation to donate $50 to an effective antipoverty charity instead of personal consumption for the same reason, to promote equal advancement of everyone’s interests.

Even more important is that deliberation by itself doesn’t change the unacceptably coercive nature of the state. Huemer presents a thought experiment:

Imagine the following scenario, which I shall call the Bar Tab example. You have gone out for drinks with a few of your colleagues and graduate students. You are all busy talking about philosophy, when someone raises the question of who is going to pay the bill. A number of options are discussed. A colleague suggests dividing the bill evenly among everyone at the table. You suggest that everyone pay for his own drinks. A graduate student then suggests that you pay for everybody’s drinks. Reluctant to spend so much money, you decline. But the student persists: ‘Let’s take a vote.’ To your consternation, they proceed to take the vote, which reveals that everyone at the table except you wants you to pay for everybody’s drinks. ‘Well, that settles it’, declares the student. ‘Pay up.’

The only option left is a consequentialist account of political authority. Governments deter crime, promote social welfare, provide rules of social conduct, and defend against external threats, and if obedience to the state was necessary to prevent a much worse outcome on these dimensions, then this could create political legitimacy for the state and political obligations on its citizens.

Huemer brings up the Drowning Child case: you have an obligation to save other people from great harms if you can do so at minimal cost to yourself. This might be analogous to an obligation to obey the state in order to prevent the great harms if the state were to collapse, except that your disobedience of the law won’t actually cause the state to collapse. A closer analogy would be a drowning child already being approached by several muddy strangers who will rescue him regardless of what you do.

One response is that this is unfair, the same way it would be unfair to rely on others to bail water in a lifeboat simply because they could clearly succeed without your specific contribution. Huemer lists relevant features that make this unfair:

i) There is a large good being produced by the actions of others – in this case, that the boat remains afloat. In contrast, if the others were doing something harmful (say, scooping water into the boat), useless (say, praying to Poseidon), or merely of trivial value (say, entertaining each other by telling stories), then you would not be obligated to help.

ii) The others assume a cost that is causally necessary to the production of the good. In this case, the cost is the effort involved in bailing water.

iii) You receive a fair share of the benefit being produced. In this case, you avoid drowning

iv) Your participation in the cooperative scheme would causally contribute to the production of the good.

v) The costs to you of participation would be reasonable and not significantly greater than the costs undertaken by others.

vi) Your participation would not interfere with your doing something more important. For example, suppose that instead of bailing water, you decide to tie down the supplies on the boat to prevent their being thrown overboard. Assume that this is more important than helping to bail water. In this case, it is not unfair to refrain from helping to bail water.

Political obligations aren’t very much like this. Most laws are unrelated to the essential functions that would we would be much worse without, like protection from crime or invaders. Item (vi) above would also be refuted by cases where you could safely evade $1,000 worth of taxes and spend it in a more socially valuable way than the government, such as to an effective private charity.

As for political legitimacy, consequentialist accounts are even weaker at justifying any modern state. While coercion is intuitively justified to prevent much worse outcomes in a variety of situations—stealing a car to urgently drive someone to a hospital or brandishing a gun to force lifeboat passengers to bail water with you—Huemer paints a more unhinged scenario to illustrate how real states operate:

Let us extend the story of the lifeboat a little further. You have forced the other passengers to bail water out of the boat, thus saving it from sinking. While you have your gun out, you decide you might as well accomplish a few other desirable goals. You see a passenger eating potato chips, which will elevate his risk of heart disease. Pointing the gun at him, you order him to hand over the chips. Then you notice a pair of passengers at the other end of the boat playing a card game. When you see that they have bet money on the game, you threaten to hurt them if they don’t stop gambling. Another passenger has some expensive jewelry, so you take it from her and distribute it to some of the poorer passengers. You also collect $50 from everyone and give it to your friend Sally. You threaten to shoot any other passenger who tries to do the same things you are doing. Then you decide that it would be nice to have some art, so you force the other passengers to hand over some of their belongings so you can make a sculpture out of them. Finally, you have an uneasy feeling about one of the passengers – you don’t like the way he looks – so you order the other passengers to throw him overboard.

Consequentialist arguments fail to justify the political obligations and legitimacy of real states. True political obligations are broadly content-independent whereas fairness-related obligations are particular to each situation. Political legitimacy is generally held to be comprehensive but the lifeboat analogy only justifies very narrow coercion. Political authority is also supreme, but there’s nothing to prevent any lifeboat passenger from being equally justified in forcing the others to bail water.

The second part of the book explains how an anarcho-capitalist society would function. Part II of David Friedman’s The Machinery of Freedom, appropriately titled “How to Sell the State in Small Pieces,” is a better treatment. Huemer meanders around S.L.A Marshall’s work on fire rate (which I think has been refuted) and Milgram’s obedience study (which I think has replicated) but focuses on police-style security, which would be privatized under security agencies, and the court system, which would be privatized under arbitration firms. He addresses the most common immediate objections but is necessarily a little handwavy since there isn’t much empirical data to go off (most of these are verbatim):

  • Would security agencies be in constant wars? Violence is very costly and security agencies wouldn’t have an incentive to fight wars.
  • Would criminals form their own security agencies? Protecting ordinary people is more profitable than criminals, who by definition enter conflicts on purpose.
  • Would the poor have access to security agencies? Most industries are dominated by production for low- and middle-income customers.
  • Would arbitration firms just unfair judgments? They would depend on a good reputation to attract customers.
  • Could criminals reject arbitration? Security agencies would refuse to protect clients who reject arbitration.

He does well at drawing attention to the fact that the standard of comparison is the modern state, in which most crimes are not solved, the poor are already disadvantaged, and most people can’t cheaply navigate the court system. Private agencies can be expected to outperform the government in the same ways that it does in other industries.

He claims that other countries would have less incentive to invade an anarchic society since the anarchic society wouldn’t pose as much of the threat to them, and that they would be more costly to invade since there would be no existing government structure to co-opt. I’m not sure what to make of this.

He imagines a timeline where, after most people in a country embrace anarchism, the government gradually removes paternalistic laws, rent-seeking regulations, immigrations restrictions, and then gradually integrates more arbitration into the legal system, abolishes the standing military, privatizes the police, and finally “someone would probably figure out how to make the politicians go home.”

New to LessWrong?

New Comment
22 comments, sorted by Click to highlight new comments since: Today at 2:10 AM

Anarchists always miss the argument from logical necessity, which I won't actually make because it is too much effort, but in summary, politics abhors a vacuum. If there is not a formal power you must consent to, there will be an informal one. If there isn't an informal one, you will shortly be conquered.

In these proposals, what is to stop these security forces from simply conquering anyone and everyone that isn't under the protection of one? Nothing. Security forces have no reason to fight each other to protect your right not to belong to one. And they will conquer, since the ones that don't, won't grow to keep pace. It is thus the same as the example given of a job offer you can't refuse, except that here the deal offered likely is terrible (since they have no reason to give you a good one.).

Why give up a modern, functional government, where you have an actual say, for an ad-hoc, notably violent one where you have no say? I have a lot of problems with the way governments operate, but this isn't better. You can always just support getting rid of or reforming nonfunctional and bad governments, and not be an anarchist.

In these proposals, what is to stop these security forces from simply conquering anyone and everyone that isn't under the protection of one? Nothing. Security forces have no reason to fight each other to protect your right not to belong to one. And they will conquer, since the ones that don't, won't grow to keep pace. It is thus the same as the example given of a job offer you can't refuse, except that here the deal offered likely is terrible (since they have no reason to give you a good one.).

Channeling Huemer, I'd say that the world's states are in a kind of anarchy and they don't simply gobble each other up all the time. 

Except that is clearly not real anarchy. It is a balance of power between the states. The states themselves ARE the security forces in this proposal. I'm saying that they would conquer everyone who doesn't belong to one.

Yes, anarcho-capitalists accept that ~everyone will hire a security agency. This isn't a refutation of anarchism.

The point is that security agencies have incentive to compete on quality, whereas current governments don't (as much), so the quality of security agencies would be higher than the quality of governments today.

The point is that security agencies have incentive to compete on quality,

Yes, but they also have an incentive to use coercion directly, ie. to offer protection from themselves.

So the example given to decry a hypothetical, obviously bad situation applies even better to what they're proposing. It's every bit the same coercion as they're decrying, but with less personal benefit and choice (you get nothing out of this deal.). And they admit this?  This is self-refuting.

Security agencies don't have any more reason to compete on quality than countries do, it's actually less, because they have every bit as much force, and you don't really have any say. What, you're in the middle of a million people with company A security, and you think you can pick B and they'll be able to do anything?

Both are half true. States do gobble each other up some of the time, and there is some sort of world order, not pure anarchy.

He claims that other countries would have less incentive to invade an anarchic society since the anarchic society wouldn’t pose as much of the threat to them

An anarchistic society that doesn't enforce patent rights and other intellectual property rights, that might very well create threats to other nations. The anarchistic society also lacks a good way to coordinate against harboring terrorists which might seem threatening as well. 

In general, political movements are better at explaining the problems of solutions proposed by their opponents, than at defending their own solution. That's because the world is complicated and everything is imperfect, which works for you when you are attacking a solution, but works against you when you are defending one.

Yes, the theories trying to derive the legitimacy of the state from the first principles are bullshit. The "social contract" only works if we stretch the meaning of "contract" into something that would not be accepted in any other context. Even if we assume the necessity of the state to do some things, the states are eager to expand way beyond that. Yes, this all sucks. It is trivial to prove, because in real life, everything sucks.

Now if we look closely at the anarchist theories, they are usually based on some form of "Copenhagen interpretation of ethics". If something very bad happens for "natural" reasons, it might be sad, but it is morally irrelevant. If someone prevents the very bad thing from happening... and then charges you 10 cents for the service, without obtaining your consent first, hundreds of pages will be written explaining in detail why this is the actual evil that we all need to fight against.

In other words, the anarchist solution seems morally superior because it uses a clever accounting trick where the bad things the anarchism doesn't care about are excluded from the equation. A scenario where the bears rip and devour your entire family is perfectly neutral from this perspective, because no human was acting against your consent. A scenario where a group of human thugs does the same is still okay-ish in some sense, because no state was involved; at least you have avoided the horrors of non-consensual taxation! You may be a victim of an actual theft or worse, but at least you will be free from the metaphorical theft.

Would security agencies be in constant wars? Violence is very costly and security agencies wouldn’t have an incentive to fight wars.

This is an argument against a constant war of balanced armies. But if a stronger army could eliminate a weaker opponent in one decisive strike, it might be a profitable investment.

Also, you could harass your weaker competitor's customers. If the competitor sends their soldiers to fight against your soldiers, both sides will pay the costs of the war, but only your side will profit from the customers who changed their mind and decided to start paying you for the protection instead.

Would criminals form their own security agencies? Protecting ordinary people is more profitable than criminals, who by definition enter conflicts on purpose.

Translation: Only the rich criminals could afford that. The anarchist utopia will protect you against the crime, under the condition that the crime never pays off significantly.

He claims that other countries would have less incentive to invade an anarchic society since the anarchic society wouldn’t pose as much of the threat to them, and that they would be more costly to invade since there would be no existing government structure to co-opt.

Of course, only a fool would invade an anarchic society. Instead, you would take its territory piece by piece. Each piece small enough that in short term it is more profitable for the security agency to give it up, rather than risk fighting your army.

What happens, from the economical perspective, if Putin puts his army next to your farm, without attacking yet? If the security agency you have a contract with is serious about potentially defending your farm no matter what, they will expect more expenses compared to their competitors, so they will have to increase your payments. At some moment, you will lose your farm even if Putin actually never attacks, because you will not be able to afford the expensive insurance anymore. The market price of the farm will drop, until it just isn't worth insuring it. Then Putin will take it without resistance. -- But at least we have avoided the war, isn't that a great thing? Sure it is, but 20 or 50 years later, your society will exist no more, taken apart one farm at a time.

"A scenario where a group of human thugs [rips and devours your entire family] is still okay-ish in some sense, because no state was involved; at least you have avoided the horrors of non-consensual taxation!"

Sorry, this doesn't pass the ITT.

Suppose that in the anarchist utopia, someone overconfident about their self-defense skills refuses to hire a protection agency, and the next day a group of thugs murders the entire family. I assume the answer would be: "well, they clearly made a mistake, that's exactly what the protection agencies are for." The dead family would be a tragedy, but it would not be considered a tragedy caused by the regime.

Compare to the statist utopia, where the police protects everyone, and citizens pay taxes. If someone doesn't pay, they get thrown in jail. "They clearly made a mistake, you are supposed to pay taxes..." "How dare you!" Every person in the jail (or killed resisting the arrest) is a tragedy caused by the regime.

So, even if hypothetically the former regime had orders of magnitude more tragedies, in the anarchist calculus it would be morally preferable, because those tragedies don't count as an argument against it.

Michael Huemer and David D Friedman primarily employ consequentialist arguments in favor of philosophical anarchism (especially Friedman). My understanding is that you’re assuming their arguments are rooted in applying a blanket action/omission asymmetry on the part of state actors, implying that the fewer actions states take, the better. I think this view substantially misinterprets their actual arguments though, as I don’t think they lean heavily on this asymmetry in any part in their books.

this doesn’t pass the ITT

Can you say why?

What if we don't accept the starting assumption that non-consensual, unequal relationships are bad? I read the arguments against the social contract and democratic theories of the state, and I agree with the author that they are inadequate defenses for state authority, but that's because they're trying to construct the state starting from consensualist and egalitarian premises.

But consensualism and egalitarianism are false starting points! Relations which are both non-consensual and unequal are sometimes right and good, and my argument for this is that such relations are an ineradicable feature of physical reality. The prime example of this is the relation between parents and children, but many more examples may be found.

(Constructing an argument for the state beginning from the position of non-consensualism and inegalitarianism is left as an exercise for the reader.)

The prime example of this is the relation between parents and children

For what it's worth, I would not be surprised if Huemer argued that children have no general obligation to obey their parents.

Relations which are both non-consensual and unequal are sometimes right and good, and my argument for this is that such relations are an ineradicable feature of physical reality.

It might be true that non-consensual relations are sometimes acceptable, but we should probably still have a presumption that they’re generally unacceptable, right? The right question to ask is why it’s permissible in this case, not just whether it’s sometimes permissible.

If I robbed you, and you complained, I think you’d find the defense, “But sometimes it is OK to steal from others” very unsatisfying. That defense completely omits the context, such as my motive or my circumstances. If instead I had told you that I was starving to death and robbing people was the only way I could eat, then you’d probably find my behavior much more reasonable.

I'm not trying to argue for a general inversion of the principle, ie. I'm not suggesting that non-consent is somehow automatically justified. Mostly I was observing the thing where two people on "opposite" sides of an issue nonetheless have major unstated premises in common, and without those premises the contention between them dissolves.

As I alluded to by saying "left as an exercise to the reader", I don't have a full explanation at the ready about the ethics of non-consensuality. Mostly I just wanted to bring the readers' attention to the way in which consensualism is being assumed by the above, and that the argument fails hard in the cases where consensualism is rejected or simply doesn't apply.

(If I were to make a general gesture towards the ethics of non-consent, I would start by talking about the phenomenon of dependency, where one party explicitly requires the cooperation of another party in order to live. Such dependency relations are by definition unequal, and in the natural world they are also often non-consensual, but despite these features they still place binding moral obligations on both parties.)

A presumption against coercion

Most examples of people are citizens,members of a state. We don't have many examples of people in a state of nature to study. That creates some ambiguity between "people aren't allowed to use coercion" and "citizens aren't allowed to use coercion".

Statists would argue that people do have a right to coercion (it's difficult to control crime without it, for one thing), and that the right is yielded to the state (if you're in a state).... and that the idea that citizens are not allowed to use coercion emerges from the state monopoly ..which itself emerges from the natural right to use coercion.

A nice review, the book sounds like an interesting read.

"Protecting ordinary people is more profitable than criminals, who by definition enter conflicts on purpose." > This is a bad argument. X is a bigger market that Y does not imply that X is more profitable. If I am founding a new security agency I might see a gap in the market for one that serves criminals, it saves me competing with all the many agencies that help nice people. Criminals may enter conflicts on purpose, but maybe they will pay a premium.

Would criminals form their own security agencies?


Does the very concept of "criminals" even make sense in this context? How can you distinguish criminals from ordinary people if no one has formal obligations?

Suppose that someone in a gray pinstripe shows up at your door offering "protection" in exchange for a monthly fee. How are the anarchists supposed to handle this situation? If the answer is "paying someone else for actual protection", then this system sounds awfully similar to a world ruled by the mafia, if you just substitute "security agencies" with "families"...

I think the assumption is that exactly this would happen, but your total payment for "protection" and all kinds of services you are currently getting from state, would be smaller than the taxes you are paying now.

And yes, there would be mafia families getting rich from offering "protection", but it would be analogical to other kinds of entrepreneurs getting rich. The assumption is that you would be free to replace the "protection" of a more expensive mafia for a cheaper one with no bad consequences for your health.

This would require a precise degree of coordination between the mafias, strong enough to make sure that no individual mafia would defect against the system by e.g. harassing the customers who want to join a competitor, yet not strong enough for all mafias to form a cartel and increase the costs for all customers. (This is the part that I consider wishful thinking.)

The book claims that authority by government is never justified and you state also that private agencies are expected to outperform the government. I take issue with both of these claims. Free markets fail to solve tragedy of the commons problem for which collective solutions through government are the best solution. Free markets also often fail in solving inequality, externalities, asymmetry of information, monopolies and the free rider problem. A world under anarcho-capitalism would likely lead to a large monopoly by a powerful force controlling security and weapons for which life for the average person would be worse than our current system run by democratic governments. Countries with higher amounts of government spending have higher levels of Quality of Life after adjusting for level of development.