Symposium on Huemer's Problem of Political Authority, Libertarianism

Michael Huemer Responds to Critics, Part 2

On Monday, I responded to Kevin Vallier’s and Massimo Renzo’s critiques of my book, The Problem of Political AuthorityBelow, I respond to Christopher Morris and Bas van der Vossen.

Comments on Chris Morris’ posts:

1. General metaphilosophical point: in my view, the aim of political philosophy is to figure out how society ought to be arranged, what makes for a good or bad society, or something like that. The drawing of distinctions is not an end in itself, nor is the giving of definitions or the taxonomizing of alternatives. Furthermore, we have limited time, so we should prioritize discussions that are most relevant to those evaluative questions. Therefore, I generally avoid talking about the meanings of words, conceptual distinctions, and taxonomy, unless I can see how discussing these things would help significantly in figuring out how society ought to be arranged.


2. Thus, I don’t want to debate the meanings of “authority”, “coercion”, “state”, “government”, “anarchy”, or “constitution”, because I don’t see how those debates are going to help me figure out the best social arrangements. What matters is this: Are the things that I use these words to pick out interesting and important for the question of how society should be arranged? Are the claims I have made using these words true, when the words are understood as I intended? I think the answer to both questions is “yes,” whether or not my usage of words matches that of other people.


3. Could there be a state that did not coerce? Maybe. And then perhaps I would not object to that state. But again, why should we care? Would Chris actually advocate that we try to create such a non-coercive state? If not, then I don’t see why we need to discuss this idea.

If so, then, well, here is why I think it won’t work. In any realistic human society, laws will be violated fairly often, and a fair number of people will make a determined effort to refuse punishment. If violence is not deployed against these people, they will get away with their law-violations. When this becomes generally known, the number of lawbreakers will start to rise. Moreover, even a small number of sociopaths who are convinced that no one is prepared to use force against them could create extremely serious problems. All of this rests upon sufficiently pervasive features of human nature and human society that it would be overly utopian to talk about a society in which these things wouldn’t happen.

Of course, the problem could be solved if someone other than the state – for instance, private protection agencies – were prepared to deploy violence against the lawbreakers.


4. Chris says there are other alternatives besides “state” and “anarchy”; for instance, “empire” is one such alternative. I don’t know what the other alternatives are nor why “empire” doesn’t count as a kind of state – but I just don’t see the import of discussing this, unless someone actually thinks that one of these alleged alternatives is what we should pursue.


5. In his section “The case for anarchy,” Chris raises objections to anarcho-capitalism. On my reading, there are three objections. The first is that, because “the mob employs murderers,” it is likely that protection agencies under anarcho-capitalism will do likewise. I think Chris may be suggesting that the “protection agencies” will be just like criminal organizations today, and that they will be in the business of predation rather than protection.

In section 10.8 in the book, I argued that organized crime would be much smaller under anarchy than it is at present. Briefly, this is because criminal organizations derive the vast majority of their revenues from selling goods and services that are banned by the state but that would be legal under anarcho-capitalism. In section 10.9, I explained why it would be more profitable to sell (genuine) protection than to run an extortion ring. And in section 10.4, I explained why it is more profitable to protect crime victims than to protect criminals. (This is assuming an anarcho-capitalist society. In a government-dominated society, these things may not be true.)


6. The second objection that I think Chris is hinting at is that in primitive societies, there has been a great deal of violence, including warfare (more violence per capita than in modern states). These primitive societies have been anarchic. So it looks like we have empirical evidence that anarchy leads to violence.

I see two problems with this argument. The first problem is that there are a number of confounds – factors other than the presence of government that affect rates of violence. In my section 9.2.2, I mentioned three factors (social values, economic prosperity, and weapons technology). To evaluate the effects of government on violence, one would have to control for these other factors, which means that we cannot compare primitive anarchies with advanced, prosperous, liberal states. As I suggested in section 9.2.2, once we have liberal values, economic prosperity, and advanced technology, the motives and costs of violence are radically altered, and they may well be altered in such a way that a state is no longer needed to prevent violence (if it ever was needed). We haven’t tested this, because no society with liberal values, economic prosperity, and advanced technology has yet tried anarchy.

Second, as I observe in section 8.1.3, just as there are different kinds of government, there are different kinds of anarchy. And just as it would be wrong to evaluate democratic government by looking at the history of communist dictatorships, it would be wrong to evaluate anarcho-capitalism by looking at primitive Indian tribes. To evaluate anarcho-capitalism empirically, one would need to examine an actual society with the sort of institutions that anarcho-capitalists advocate (e.g., a system of competing arbitration companies and competing protection agencies). Unfortunately, there are no such societies.


7. The third objection – which seems to be Chris’ main objection to anarchy – is that anarchy cannot survive, because an anarchic society will be taken over by states. I devoted chapter 12 to showing why this need not be the case.

But Chris might say: “Look, it’s very nice that you have these theoretical arguments about how an anarcho-capitalist society might survive. But empirical evidence trumps theoretical arguments. And the empirical evidence is that every habitable land mass on Earth has in fact been taken over by states.”

Two responses to this. First, the arguments in chapter 12 were not purely theoretical. They included a good deal of empirical evidence. Of course, this evidence didn’t consist of observations of actual anarcho-capitalist societies; it was evidence concerning such things as the factors that cause war, the success of non-governmental organizations in resisting governments, and actual countries that have no military. I can’t reproduce the arguments of chapter 12 here; I would simply ask the reader to read that chapter before drawing conclusions.

Second: true, all habitable regions on Earth have been taken over by states. But that happened a long time ago. And in fact, quite a lot has changed in ways that are very relevant to the prospects for violence, conquest, and the need for military defense. Today’s society is vastly different from the societies of even 100 years ago – first, in terms of economics and technology, and this is very relevant to the prospects for violence (again, see section 9.2.2). But the change I find most interesting is the change in people’s beliefs and values, which I discuss in section 13.4 in the book. See also my recent TEDx video on “The Progress of Liberalism” for a more entertaining presentation. This change can hardly be overstated. If an average person of today were transported back in time 400 years, he would immediately be the most wild-eyed, raving liberal on the planet.

So at least part of my response to the empirical argument is this: maybe anarchy was no good for the people of yesterday. But the people of yesterday were terrible. People today are much, much better.



Comments on Bas van der Vossen’s posts:

1. Bas seeks to challenge my general methodology. But the Giant story does not really challenge my methodology, because my methodology is not based on the assumption that the same moral obligations apply to states as apply to individuals. Why not? There’s a small point here and a more important point.


2. Small point: Most of my hypotheticals in the book involve individuals. But the parallel I’m drawing is not between state and individual action per se. It is between state and private (or non-state) action. In other words, I’m not saying, “If x would be wrong for an individual, then it’s also wrong for the state.” I’m saying, “If x would be wrong for any private agent, then it’s also wrong for the state.” For example, in normal circumstances, it’s wrong for me to extort money from other individuals in order to give the money to a charity. But not only that; it’s also wrong for the Exxon corporation to do this. Or the Catholic Church. Or the University of Colorado. Or anyone who isn’t a government. So, I think it’s also wrong for the government.


3. Now, the important point: what I am actually assuming is the following.

a. The same obligations and constraints apply to states as apply to non-state agents, unless there is a morally relevant difference between them that would explain the different obligations/constraints.

b. The person who believes there is such a difference at least has the burden of naming that difference.

Thus, when I point to the wrongness of some private action and draw an analogy to some governmental policy, I am not saying that this suffices to prove that the governmental policy is wrong. I am saying that this establishes a presumption against the policy, so that at a minimum, the policy’s advocates must explain why they think the government is special, such that it is entitled to do this thing. If we can’t find any way in which the government is special, we should reject the policy.

That is why I confronted several philosophical accounts of authority in the first half of the book. To dispute my methodology, Bas would have to be saying that we don’t even need an account of authority; it’s fine to hold that the state is entitled to do things that no other agent may do, while having no account at all of why the state is different from any other agent. But I don’t think Bas is suggesting that. (I don’t know of anyone who believes that.)


4. Bas’ story of the giants can be viewed, rather than as a challenge to my methodology, as an attempt to satisfy it. That is, it might be taken as pointing to a possible account of why the state is special: someone might say that the state has authority because the state is better at adjudicating disputes and exercising justified coercion (with less risk of error, etc.) than non-state agents.


5. Okay, let’s say the giants, as Bas says, “tend to make judgments that are more temperate, more balanced, and less biased than [ours].” Furthermore, “their temperance carries over to their uses of force.” In that case, Bas says, “we should ask the giants to do our justified coercing for us.” I see three problems:

a. Maybe the giants tend to be better at resolving disputes than humans. But that’s compatible with there being some humans who are especially good at resolving disputes. Rather than relying on giants, why couldn’t we rely on humans who were unusually good at this?

Here’s the political relevance: sure, most private agents are no good at resolving disputes. Maybe (?) they’re worse at it than a government court. But professional arbitrators who work for (private) arbitration companies are especially good at resolving disputes. So why not rely on them instead of government courts?

To address this issue, I think one has to add the stipulation that actually, the giants are better at resolving disputes and exercising justified coercion than any non-giant-involving mechanism.

b. Still, I don’t see that it would be obligatory to use giants. If some people wanted to stick with human mechanisms, then as long as the human mechanisms were good enough, even if they were not the best, I think this would be permissible. (Analogously: anarchists should be able to reject the state’s services and rely on private providers.)

To address this, I think one has to again modify the hypothetical. We have to stipulate that actually, no human mechanism is even acceptable. All human ways of resolving disputes and/or exercising coercion are just terrible. The giants, by contrast, are good or at least acceptable.

c. And yet I still don’t see how this gets the giants anything like the moral status that is commonly ascribed to the state. In particular, I don’t see that the giants would ever acquire the power of creating obligations merely by telling people to do things (things that we were not independently required to do). And I don’t see that the giants could justly force people to obey their commands (in cases where a human could not justly force people to obey similar commands).


6. And of course, as a factual matter, it is extremely doubtful that the state is in a position analogous to the giants in this story. The state is not clearly better than all non-state mechanisms for resolving disputes. The state and its agents are commonly biased and intemperate, just like normal humans, because – unlike the giants in the story – the state’s agents actually just are normal humans.

In sum: Bas is right that if giants have different properties from humans, they might have different obligations and constraints, and we might have different obligations towards them. But (a) I don’t see how the giants would ever have anything like political authority, and (b) the state is not relevantly like a giant.


7. What is political legitimacy? As Bas says, for a state to be legitimate does not require that it is permissible for that state to make and enforce any and all laws that it chooses. I stated that legitimacy is an entitlement on the part of the state to make and enforce “certain sorts of laws.” But this isn’t very clear. So here are some clarifications of how I understand this:

  1. a. First, I am not seeking to capture the ordinary usage, or even the usage in political philosophy or political science, of the word “legitimacy.” I don’t care about that. Here is what I’m interested in: there is a class of government activities that (a) would be considered wrong by almost anyone if they were performed by any private agent, but (b) most people consider acceptable. I think that behavior is in fact wrong. I introduce the term “political legitimacy” to express the difference between my view and that of the majority of my society: I say that those others believe in “political legitimacy” (they believe that “the state is legitimate”), whereas I disbelieve this. The term should be interpreted in the light of that dialectical purpose.

Now, I don’t want to saddle my opponents with obviously unreasonable views, so I allow qualifications that render their view plausible, without removing the controversy. In particular:

  1. b. “Political legitimacy” may be conditional. That is, one may think that the state has to satisfy some conditions (other than being a state) in order to retain legitimacy. Perhaps it has to be democratic, do a reasonably good job of protecting rights, etc.
  1. c. “Political legitimacy” may have limited scope. That is, one may think that there is only a certain limited sphere within which the state is entitled to make and enforce its commands.
  1. d. “Political legitimacy” may be defeasible. That is, one may think that the state is entitled to make and enforce commands (only) provided that there are not some specially strong reasons against its doing so that would override its normal entitlement. For instance, perhaps the command must not be grossly unjust, it must not result in the destruction of the human species, etc.

Having said that, however, I would not allow the doctrine of legitimacy to be so watered down as to make it uninteresting. Thus, I would insist on the following:

  1. e. “Political legitimacy” is not limited to the correct laws. That is, if you think the state is only entitled to make and enforce the laws that are objectively correct by some independent standard, then, in my terminology, you do not accept political legitimacy.

Note that one can maintain that an agent should do A, or has most reason to do A, but it is still permissible for the agent not to do A. (“You should be polite to other people. Of course, you don’t have to do this, but you really should.”) Just so, believers in political legitimacy can maintain that the state should not have made some law, but it was still permissible for it to do so, and permissible to enforce the law once made.

  1. f. “Political legitimacy” is a special property of (some) states. I would not allow “political legitimacy” to be interpreted in such a way that it is a power that ordinary people have. It is an entitlement to exert coercion in at least some circumstances in which it would be wrong for a private agent to do so.

I hope that sufficiently clarifies what I had in mind in speaking of political legitimacy.

Published on:
Author: Michael Huemer
  • Jason Brennan

    This post was even tastier than the Georgetown Cupcakes I just shared with my freshman seminar.

    • Sean II

      Ha! I haven’t finished all of the required reading for this one, but I can say one thing. Huemer’s style of thinking and speaking is so clean and crisp, he’s like some strange personification of reason. The guy’s a Vulcan.

      • I know what you mean :-). Part 1 of his book was so clear and well-written that it’s hard to believe that a layman statist could not be greatly influenced by it.

    • Sean II

      So the presence of a down-vote on this comment means one of two things:

      a) Someone hates Michael Huemer.
      b) Someone hates Georgetown Cupcakes.

      Either way, we have failed as a species . I’m afraid it’s time for humanity to be annihilated in a cleansing fire so the universe can get a fresh start.

  • wophugus

    The mob doesn’t just commit crime to prey on people, it also offers legitimate protection services to its voluntary and paying members. You don’t attack, rob, or cheat a made man.

    The reason why you don’t is that mafiosos will murder you. But I don’t know, maybe you think the kind of negotiations that happen when a senior member of one mob (protection agency) wants to revenge kill a less important member of another mob (protection agency) are a model of justice.

    Also “you should reject your system that has been proven to work better than all others tried for mine, which I think will work, sure. I’m smart, trust me.” remains a weak argument. The point isn’t that it proves anarcho capitalism wrong, the point is that you would need an enormously high risk tolerance to actually try anarcho capatalism based on the evidence you provide. “It will work just like the common law! Only owling and hearsay will be legal this time! Also, no king!”

    If your claim is “this might work” then fine, it’s unfair to dismiss it simply because it is unproven. If your claim is “I think anarcho capitalism will work, make policy moves in that direction” then yes, it is totally fair to dismiss it on the grounds that it is insanely risky and unproven. If your claim is “anarcho capatalism is the only just form of government” then spend less time arguing that others are unjust and more time arguing that yours is just (it is possible, after all, that no system is just. For example, if our basic moral impulses are contradictory and irrational (which they are!) or if human nature makes no just system feasible (which it does!) if it can be true that no system is just, proving all others unjust doesn’t prove yours just). For example, i know it’s bern mentioned but you really need a good grounds-up defense of property rights. And if your argument is “I don’t have to prove property rights because everyone accepts them” then no. As a philosophical matter plenty of consequentialists accept property as a good idea rather than a right; as as a legal matter the right to property is the most controversial first generation right (I am pretty sure it isn’t in the two big covenants on civil and political or economic social and cultural rights, for example).

    In other words, without proving that property rights are good there is no reason to prefer anarcho capatalism to anarcho communism (for example) on deontological grounds, and without more evidence there is no reason to prefer anarcho capitalism to liberal democracy on utilitarian grounds.

    I don’t have my copy of your book yet, so to the extent these criticisms are insanely unfair I’m sorry. I do have a question which, if you answer in the book, you can ignore: why wouldn’t people just join protection services with rules insanely favorable to the kind of legal position they are likely to be in in a dispute. It seems like a protection agency that targeted the preferences of rich people could command more money and more power without being more just. If the top quintile forms a cartel to enslave the bottom quintile, who stops them? Are you just counting on it never being profitable to enslave anyone? Less dramatically, businesses have deep pockets and worry more about committing torts than being torted. Why wouldn’t they just form a protection agency with no respondeat superior, no negligence, weak damages for intentional torts (the same way they bankroll legal reform lobbying to that effect in our society)? And why wouldn’t their deeper pockets shift the law in that (unjust) direction? It kind of seems like any bad legal innovation deep pockets lobby for in our society they would pay protection agencies to push in your society, only without democratic or rights-based judicial checks. This is less of a problem in contracts — choice of law is negotiated like everything else in a contract and free market forces help make the chosen law not-to-crazy — but a huge problem for torts (negligence, battery, conversion, etc) where negotiation is impossible (anyone can tort you and you can tort anyone, you can’t negotiate with everyone). And indeed, arbitration is successful and widespread in the world of contracts but I don’t know of a system that has pulled it off in the world of torts.

    • Michael Huemer

      ‘Also “you should reject your system that has been proven to work better
      than all others tried for mine, which I think will work, sure. I’m
      smart, trust me.” remains a weak argument.’
      Yeah, in fact, it’s not even an argument at all. It’s a good thing I didn’t say that, and that instead I provided actual arguments.
      Now, here is an argument: “In this society, right now, things are going very well by historical standards. Any radically new social organization entails risks, however good the theoretical arguments may be. Because things are going well now, it’s not worth taking a risk.” Of course, it might still be worth it for some society where things are not going well.
      a. The status quo is extremely risky. There is a pretty good chance that the human species will go extinct within 1,000 years. The most likely cause of this extinction is weapons of mass destruction, created by some government (most likely the U.S. government, since it accounts for 40% of the world’s military spending).
      b. Anarchy does not have to be tried all at once. Incremental moves can be made, and we can observe the results.

      c. Similarly, anarchy does not have to (and would not) happen everywhere at once. It could be tried in a small area.

      • 1. What incremental steps can be taken in privatizing tort law/criminal law? I mean intentional, knowing, reckless, and negligent damages to people you have no contractual relationship with? Because that is where I think the argument for anarchy is weakest, that is very untried, and that is an all at once thing. 2. Since you stipulate that anarchy needs good conditions to work, trying it in a small place would, in fact, entail replacing a well functioning with an untried social model. If you do it in a new community or in the developing world it’s just setting up for another failure. 3. Adopting an anarchy doesn’t eliminate the risk of state-sanctioned nuclear war. One possible (likely?) outcome of anarchy is that it fails and you get a new, possibly worse, possibly more nuke happy state. If the first state was very good, it increases the chances that the replacement state will be worse.

        Anarchy does arguably reduce our ability to provide public goods, though, such as programs to avoid asteroids or non-excludable non-rivalrous public health programs to resist catastrophic plague. That’s probably why most small government economists are libertarians!

        • jtkennedy

          “I mean intentional, knowing, reckless, and negligent damages to people you have no contractual relationship with? ”

          We already have a model for this. Suppose an American slips across the border, murders a Canadian, and then hurries home. The criminal and the victim are not subject to any common state, yet we see that their “defense agencies” typically come to peaceful agreements about how to deal with such crimes. Private defense agencies would have the same incentives to come to peaceful solutions as states do -war strongly tends to be unprofitable.

        • No, it’s not true that that situation has been worked out. Lots of states don’t routinely extradite criminals to lots of other states. As a result, states have dealt with the problem you describe with emigration and immigration restrictions. We’ve “solved” the problem by giving protection agencies sovereignty over areas and letting them keep their criminals in and other criminals out.

          More generally, choice of law in an international context is a giant headache that frequently frustrates justice.

          Lastly, the situation you describe is made enormously easier because Canadians and Americans don’t chose their defense agency, and because both agencies are democracies. You have two “defense agencies” responsible for policing every segment of their society with every segment of society having a voice. The result is that the systems have similar goals, so our choice of laws problem is between compatible systems.

          The problem I fear in an anarcho-cspatalist world is the people with property joining the protection group hyper-protective of property, the likely tortfeasors joining the group hyper-protective of tortfeasors, the poor people joining the (poorly funded) protection group all about economic and social rights, etc. The problem of getting these groups to agree is way hard (what is the middle ground between “negligence is a tort” and “no, it isn’t?”).

          It is also clearly not true that dominating people with the threat of violence is more costly than negotiating with them. If it were you would not need protection agencies in the first place, no one would use violence ever.

          • “The problem I fear in an anarcho-cspatalist world is the people with property joining the protection group hyper-protective of property, the likely tortfeasors joining the group hyper-protective of tortfeasors, the poor people joining the (poorly funded) protection group all about economic and social rights, etc. The problem of getting these groups to agree is way hard.”

            Yes, the problem of getting them to agree is difficult, and perhaps it is so difficult that they will violently fight out their disagreements instead of peacefully resolving them.

            Then again, the costs of violence are great and the incentives that these groups with conflicting interests will thus have to avoid violence may thus be strong enough to make it very likely that they will resolve their disputes peacefully.

            “what is the middle ground between “negligence is a tort” and “no, it isn’t?””

            I think you’re asking the wrong question. The question you should be asking, if a dispute arises in an anarcho-capitalist society between two people or groups of people who hold these opposing views, are they likely to resort to violence to end their dispute or are they likely to find a peaceful way to resolve their dispute to avoid the high costs of violence?

            For example, suppose you and I were living in an anarcho-capitalist society. You choose to be a member of a dispute-resolution organization that favors arbitrators who are strongly biased towards “negligence is a tort”. On the other hand, I am a very reckless negligent person and consequently choose to become a member of an agency that does its best to insistent that “negligence is not a tort”. One day, I accidentally fly my remote-control airplane into your kitchen window (my phone rang distracting me and I failed to land my plane before answering it as a responsible person would). You claim I owe you some money in restitution since I damaged your property, but I claim otherwise since it was accidental. Our agencies also take these conflicting positions on the question of whether negligence is a tort.

            You asked, “what is the middle ground between “negligence is a tort” and “no, it isn’t?”” There is no middle ground. But, there is middle ground in our dispute. For example, I could pay you *some* money–less than I should if negligence really is a tort, but more than I should if negligence isn’t a tort.

            Perhaps rather than our two agencies fight each other out for possession of the money in my (or my agency’s) possession equal to the cost of repairing your broken window, our agencies might instead decide to avoid such a costly violent struggle by coming to an agreement. For example, they could (1) mediate and decide that I must give you *some* money or (2) they could agree to flip a coin and abide its outcome (if heads, I must pay for the broken window damages, if tails I don’t have to pay) or (3) they could agree to abide by the decision of a third arbitration firm that they both believe will resolve the dispute in a way that is more pleasing to them than violence.

          • But then I can force you into mediation on any ground. My protection agency could have a rule that blasphemy is wrong, or that protectionism is right, or that people we don’t like owe us money. And now your protection agency has to flip a coin, or make you pay have damages, or submit you to arbitration. Bad for you!

            Worse, even if negation is preferable to force, force will form a background to negotiation. A very well armed protection agency (maybe its laws benefit rich people so it has lots of money) will have a trump card in negotiations the same way a powerful nation does. They can always say, “at a certain point, crushing you like a bug will be less costly than giving into your demands” and put a floor on those demands.

          • “But then I can force you into mediation on any ground.”

            Yes. (Although I need not agree to have a third party resolve the dispute. I could just hope you revoke your absurd claim in the dispute.)

            “My protection agency could have a rule that blasphemy is wrong, or that protectionism is right, or that people we don’t like owe us money.”


            “And now your protection agency has to flip a coin, or make you pay have damages, or submit you to arbitration.”

            No. My protection agency *could* do these things, but you left out other possibilities that it would most likely choose from.

            For example, if you claimed “people we don’t like owe us money” then (supposing I’m one of those people) my protection agency could just say, “No,” and do nothing. Then *you* would be forced to choose between using force to extract money from me, or persuading me to give you money voluntarily, or persuading me to let a third party arbitrator resolve the dispute, or doing nothing about your belief that I owe you money.

            “Bad for you!”

            No, you claiming you own my money or whatever doesn’t hurt me unless you are willing to bare the costs of enforcing that claim with violence in the case that I refuse to give you my money. If you are willing to engage in violence then I will either (1) say, “Fine, here’s my money,” or (2) defend myself / my money with violence of my own, or (3) choose a path somewhere in the middle by finding a possible future that is more desirable for me than 1 or 2 and is more desirable for you than 2.

            Note that (3) above would presumably usually be the case in an anarcho-capitalist society whenever two dispute resolution organizations have a dispute about how a dispute between their respective customers should be resolved. Of course, (2) would occur more often if the dispute were something like “give me your money–just because”. Obviously nobody would give in to this merely because someone stated those words.

            “Worse, even if negation [sic] is preferable to force, force will form a background to negotiation.”

            Why “worse”? This is just reality that we always face–in an anarchist society or a statist one. For example, I have essentially no bargaining power when I try to “negotiate” with the state in the state’s dispute with me that I owe it a bunch of money for no legitimate reason (taxes). This is because the “negotiation” is backed up with force and the state is much, much, much more powerful than I am. But this is only “bad” in the sense that the state’s much more powerful than I am and the state’s claim is illegitimate. It’s not “bad” (“worse”) due to the fact that there is a possible reality in which people cannot choose to attempt to enforce unjust claims against others. No such reality is conceivable. The only reality is the current one, which would remain true in an anarcho-capitalist society: all negotiation has a background of force.

            “A very well armed protection agency (maybe its laws benefit rich people so it has lots of money) will have a trump card in negotiations the same way a powerful nation does.”

            Well no, the very well armed protection agency may have a lot of bargaining power, but it wouldn’t be as powerful as a state.

            Although, for the sake of argument, even if it did, and this were a reason to oppose such protection agencies, the same reasoning would apply to the state and you consequently should oppose the state for having so much power that it can just throw down the “trump card” and demand its citizens to hand over their money (taxation/extortion). So your support of states doesn’t follow from this.

            “They can always say, “at a certain point, crushing you like a bug will be less costly than giving into your demands” and put a floor on those demands.”

            Again, this is more true for the state than for protection agencies in an anarcho-capitalist society, so you’re yet to explain why you think a social order with a state is better than an anarcho-capitalist social order.

            Also, it’s worth pointing out that in this whole discussion you have not been trying to account for political authority, but instead have just been trying to argue that anarcho-capitalism is not perfect. I agree: anarcho-capitalism is not perfect. But (hopefully obviously) that’s not a sufficient reason to oppose anarcho-capitalism.

        • “What incremental steps can be taken in privatizing tort law/criminal law?”

          Huemer touches on this in his last chapter “From Democracy to Anarchy”. In the section “13.2.1 Outsourcing court duties” Huemer points out that the role of government courts can be outsourced to private arbitrators (and indeed, this process of outsourcing to private arbitration has already begun). The outsourcing of the resolution of criminal cases has not yet begun, but this *can* begin in the future (answering your question).

          Huemer writes (p. 326): “The most controversial step would be to outsource the resolution of criminal cases. This step would be more plausible once we began to view criminal cases, not as disputes between the defendant and the state, but as disputes between the defendant and the crime victim. When viewed in this way, there is no reason why these cases, too, could not be handled through private arbitration.”

          For a more in-depth discussion of this, I highly recommend Bruce L. Benson’s book “The Enterprise of Law: Justice Without the State”.

          • Every successful arbitration program I know of has one of two things in common: 1. The relevant law is forced on participants by the state, or 2. The relevant law is pre-agreed to via arms link bargaining.

            The scary thing about torts and crimes is that there is no arms-length bargaining before the dispute, so there is no incentive to not pick a legal regime insanely biased towards your position. If you try to do that in commercial law, the other party won’t do business with you and you will suffer. Just like we don’t have to worry about me selling cheap books for a million dollars and bilking people (no one will buy them), we don’t have to worry about people in commercial disputes picking crazily favorable legal regimes (no one will do business with them).

            But in the law of torts (note that I’m not even really thinking about this in terms of crime) you don’t get that check. You can’t just decide to never tort or never be torted by people who subscribe to law all about advancing their interests. You’ve gotta walk the streets, buy goods, own property. All open you up to torts.

            Just having private people arbitrate criminal disputes isn’t such a big step — laymen did it in republican Rome, laymen partially do it here. It’s saying “the state no longer makes or enforces the law outside of the realm of business. Find your own” that is the big, scary, possibly-too-unethical-to-experiment-with step.

            And to be clear, I am fine with experiments in a libertarian direction. They are more reversible and more plausibly beneficial (to me). It’s going from “we should democratically agree to experiment with moving in a marginally more libertarian direction” to “do that and keep going to anarchy” that scares me.

          • M Lister

            The scary thing about torts and crimes is that there is no arms-length
            bargaining before the dispute, so there is no incentive to not pick a
            legal regime insanely biased towards your position. If you try to do
            that in commercial law, the other party won’t do business with you and
            you will suffer.

            Have you ever seen the sort of arbitration agreements that are in the vast majority of commercial contracts? There is no opportunity to bargain on them- they are take it or leave it, and most people have no choice but to take it. This comment is, I’m afraid, pure fantasy for a huge percentage of arbitration agreements (the ones you “sign” when you buy plane tickets, most consumer goods, software, etc.) If you don’t know this, you shouldn’t be talking about arbitration. Even in agreements between businessmen, the more powerful generally dictate terms. The limits on this are largely set by the state, when there are any at all. I’m not necessarily anti-arbitration, but this comment has very, very little to do with how arbitration actually works. (The same is true, as far as I can tell, of the discussion of arbitration in the book- it’s a discussion of an ideal type that is very, very far from how arbitration does work now or ever has.)

          • “It’s saying “the state no longer makes or enforces the law outside of the realm of business. Find your own” that is the big, scary, possibly-too-unethical-to-experiment-with step.”

            It’s the opposite of scary to me, perhaps because essentially no person or organization has ever used or threatened to use unjust force against me in my entire life, other than governments through laws.

            It’s definitely not unethical to experiment with. What’s unethical is for the state to continue enforcing unjust laws against people, and for people to continue to support said laws (where’s your account of political authority?).

          • This is my account of state authority: “I don’t have an account of state authority. Our basic moral impulses are contradictory and confusing, no one has proven any moral precepts from first principles (to most people’s satisfaction, anyway), my religion is not focused on providing ethical norms, and the world is too confusing for utilitarianism to be a constant guide. This leaves me with a general mishmash of swirling, incomplete, and contradictory moral guideposts, out of which i get the sense that rights constrained democracy seems good? It helps spread out the work of discovering good and moral policy to the rest of the polity and you can tell plausible stories where it maximizes things I care about. It also seems to be spreading and making things better, whatever “better” means. But there are also plausible stories where democracy fails to maximize other things i care about, like freedom from government coercion, so maybe I am wrong. Tie-breaker goes to ‘it gets results.'”

            One of my issues in this debate (likely an issue with my ability to comprehend it, not a fault in other people) is that political philosophers, with the exception of Nozick, seem to have a tradition of talking about ethics with a self-assuredness I find difficult. So Huemer, for example, says “we know it is immoral for people to act in certain ways; when the state acts in those ways it should have some burden of proving why it is different” and I am left thinking, “wait, get back to how we conclusively know certain moral precepts for individual behavior, because I’ve been kind of making it up as I go along. I agree those examples of individual coercion sound bad and that I shouldn’t do them, but I don’t really know why i agree or if any universal principle is at stake or if there are exceptions or if principals should be the same for states and people or what.” There’s some baseline assumptions I see at work (wrongly, perhaps) that I just don’t get, namely 1. We have the ability to extract consistent and meaningful rules from moral impulses triggered by specific hypotheticals (compare that to the law, where we often think it is *bad* to change a rule because of a few, emotionally resonant examples of how it can reach bad results), 2. Ethical rules will be universal, consistent, and fully accord with our sense of justice. If a rule applies here it should apply everywhere. 3. Establishing that there are flaws in the ethical argument for something is sufficient to cast doubt on whether we should do it (I tend to see flaws in the ethical arguments for doing anything and everything), 4. If we come to the conclusion that something is ethical but lots of people disagree, we should have total confidence in our conclusion.

            The result of all this is that I have a much easier time understanding and debating more concrete claims about the efficacy of anarchic government (especially legal claims, since I’m a lawyer) than I do the “give me your moral proof for government legitimacy” type stuff. It’s why I am eager to get and tackle Huemer’s book.

            I know this is rambling and self indulgent, I just thought it might be helpful to see the difficulties someone without a philosophical or political science background has coming to this book. Or not. I am getting increasingly off topic, I know. Sorry for using this as a place to think out loud, in this and other comments.

          • Also, I’ll check out that book (literally). Thanks for the recommendation.

  • Basvandervossen

    Thanks Mike. I agree with everything you say. But I would add, I think, one detail about legitimacy. You write: ““Political legitimacy” is not limited to the correct laws.” I suppose your concern is that at this point the claim that a state is legitimate becomes practically indistinguishable from the claim that it is not. (Correct me if that’s wrong.)

    But here’s an alternative idea. Why not say that a legitimate state has the permission to coerce / right to be obeyed only when it gets things right, but *also* has the right that people don’t try to resist it in ways that undercut its ability to function when it gets things wrong. That gets the state a standing that you and I don’t necessarily enjoy, and without the stupid statism that you so nicely show mistaken. Any thoughts?

    • Michael Huemer

      Yeah, good question. That condition seems more likely to be defensible, and it’s an interesting condition. Does it capture what ordinary people think? Well, I think it captures only part of what ordinary people think, so I think it is still interesting to discuss the stronger notion of legitimacy.

      Now, is the condition satisfied; do states have the right not to be resisted? I don’t think so, but I think a much longer discussion would be required (in which someone first gives the account of why states would have that right).

      Here’s a first try: you shouldn’t resist the state when it gets things wrong, because you’re threatening social order, and it’s not worth it (provided that the state’s error is not very large). But here’s another thought: in any given case of getting things wrong, the least-cost avoider of the problem is the state: they could just not implement the wrongful or unjust policy. Why should I have to give up the right of self-defense that I normally have, to avoid this problem that could be avoided by the state just not acting wrongly?

      (Of course, if the “resistance” is going to involve harming innocent people, then you have constraints against that, even if the state doesn’t have any special rights. I assume that’s not the sort of case you had in mind.)

      • Ted_Levy

        There is, of course, a long history of writing, familiar to the Founding generation, on the right of resistance to unjust laws (distinct from the right of revolution).

      • Basvandervossen

        Agreed. I’m not much interested in this idea of legitimacy capturing what ordinary people think of states. But it does seem both an interesting property states might have and one that I’m open to defending (at least more open than the one you attack). I don’t mean that your exercise is not an interesting one – I just mean that I’m personally interested in thinking about something slightly different.

        About your point that there may be no reason for you to give up the option of self-defense. Two things:

        (1) Sometimes there are different ways of resisting a state when it gets things wrong. When we have the option to resist in ways that do not further undercut the state’s ability to do things right, and the state is legitimate (despite its getting it wrong), then you ought to resist in that way. When a state is illegitimate, it is permissible and perhaps even good to resist in ways that do undercut that state’s ability to function. This I am pretty confident I subscribe to.

        (2) There may be a further difference with respect to other-defense. That is, it might be the case that (a) you are permitted to resist even in ways that are going to harm the state’s marginal ability to function, and (b) you are permitted to help people in illegitimate states in ways that are going to harm their state’s marginal ability to function, but (c) you are not permitted to help people in legitimate states in ways that are going to harm their state’s marginal ability to function. This I am much less confident I subscribe to.

        • Michael Huemer

          (1): That’s plausible. I would add that in reality, your resistance almost never impairs the state’s ability to do things right. There are a few reasons: a) because your own individual action is just not going to make any difference in a large state; b) even if *a lot* of people are resisting some unjust law, the state has the option of just repealing that law — problem solved; c) actually, this means that if anything, you’re making it more likely that the state will come closer to satisfying its legitimate purpose.

      • “Why should I have to give up the right of self-defense that I normally have, to avoid this problem that could be avoided by the state just not acting wrongly?”

        Because if you don’t then social order will collapse.

        It’s unfortunate that the state is not choosing to avoid the problem even though it easily could just by not issuing the command, but this doesn’t change the fact that if your resistance to the state’s bad law causes social collapse then your actions (in addition to the state’s action of enforcing the bad law) are morally reprehensible. You’re thus obligated to obey the bad law.

        The problem with this defense of this kind of “weak political authority” is that I don’t think it’s a defense of political authority at all. Yes, I do think one would be obligated to obey the bad law, but one would also be obligated to obey the unjust command if it were issued by any non-state agent also. This is because independent morality obligates people to obey unjust commands (by any agent) if the realty is that disobeying (resisting) those commands is very likely to lead to a very great harm. Since it wouldn’t make a difference whether a state or a non-state agent were issuing the commands, this doesn’t appear to be a defense of any kind of political authority.

        • Michael Huemer

          Good points. On the first one: consider this analogy. Let’s say I’m in the bank when armed robbers come in to rob it. They order all the customers, including me, to hand over our money too, or else they will shoot one of the bank tellers. Now I’m obligated to hand over my money. Okay. But is this “authority”? Would it be appropriate for me to say, “I gave them my money because I recognized the robbers’ authority”?

          If the account of political obligation is like that, that we have to obey because the government is holding society hostage, then I don’t think we’d want to call that authority.


        Michael (if I may):
        Just wondering if you are familiar with Eric Mack’s essay “Nozickian Arguments for the More-Than-Minimal State” in the Cambridge Companion to Nozick’s ASU (2011), and if so, your reaction?

    • Christopher Morris

      Suppose an individual or a group is bossing others around, requesting that they act as justice requires. The rest of us ought not to undermine their efforts and quite possibly in many circumstances to support them. If enough of us refrain from undermining their efforts or actively support them, they (the individual or the group) looks increasingly like a ruler. Assuming the “ruler” continues to do well, that is, to rule justly and efficiently, then his/their rule becomes more secure.

      We can tell this story about the formation of states too. It is unlikely to support the claim that only the ruler has the right to boss others around. The exclusivity claim of states is hard to sustain. But much else can be supported in this way. (Recall my claim that constitutions can be understood as coordination equilibria. Jean Hampton wrote some very interesting things that are relevant here too.)

  • wophugus

    Re: legitimacy, What if we say “political legitimacy is the ability to make and enforce law when there is some doubt or debate as to what the correct law is.” That somewhat accords with how America operates: for one thing, democracies rarely endorse laws that everyone agrees are definitively wrong and that harm large groups of people. Minorities are not doomed either: if you can prove that there is no rational basis for a law you that harms your group but not others, you can generally mount an equal protection challenge.

    Where we have stronger intuitions about what the law should be, we can demand the state have a higher burden of proof when it makes laws that harm us. This accords with American law, where laws that run afoul of fundamental rights receive strict scrutiny.

    So with all those checks in place, we could assume that the bulk of American law is legitimate, ie America will rarely endorse laws everyone agrees are morally wrong.
    Without those checks (ie, before the equal protection clause got more robust), government action could grow more and more illegitimate (ie, Jim Crow).

    Frustratingly, this means things I think are morally wrong are, if they have sufficient support, arguably legitimate (ie, closed borders). But the alternative is to make me some sort of dictator, declaring that legitimate government is that which accords with my moral impulses even if others disagree. That strikes me as an even more immoral alternative.

    • Michael Huemer

      “So with all those checks in place, we could assume that the bulk of
      American law is legitimate, ie America will rarely endorse laws everyone
      agrees are morally wrong.”

      This is hardly a “check”. “A law is legitimate as long as there is at least one person who thinks it might be okay” seems to me about on a par with “It’s permissible to kill you as long as at least one person supports it.”

      I think that even if everyone else in the world but me thought it was okay to kill me, that still wouldn’t make it okay to kill me. Still less would it be okay merely because at least *one person* thought it was ok. Similarly for lesser acts of coercion. That’s an absurdly low standard for justifying coercion.

      “… the alternative is to make me some sort of dictator …”
      I don’t think that the only possible alternative to “Coercion is justified whenever at least one person thinks it is” is “Coercion is justified if and only if Robert H. thinks it is.”

      • Well again, the goal is for the government to legislate when there is real, not manufactured or irrational, doubt as to what the correct law should be.

        For example: more than one person thinks the earth is flat, but there is not a legitimate debate or controversy as to wether it is. If the goverment made a law that you can’t sail 500 miles east because you will fall off the earth, it would not be passing a law supported by a rational argument. That would not be legitimate coercion.

        Just so, the check isn’t “only pass laws that are arguably correct.” That’s the definition of legimate, not the process to help ensure legitimate law. The check is ” only pass laws in a majoritarian manner and have the judiciary ensure they have at least a rational basis.” “Rational basis” is a term of art, but it means mostly what it says.

        So to the extent you think there is a rational and reasonable argument for killing everyone but one guy, yes, that would be legitimate. And to the extent he could get that bill through congress and the judiciary, well, good for him. Good persuasion, guy. I don’t think there is much of a legitimate argument for that “kill everyone” bill, though.

        There is also the fact that coercive government action could be legitimate but disfavored because it often masks illegitimate action. For example, you could say “there is a legitimate argument for an educated voting public, but historically and logically majorities will try to restrict the franchise to cement their own power. Our judiciary will look at laws forcing voters to take literacy tests with higher scrutiny.”

        In other words, if the polity deems it necessary, government action can be reviewed more stringently than this. Whence rights (in a legal, not moral sense).

        And pretty much by definition you think coercion is only justified when Michael Huemer thinks it is, so you are kind of facing the same question I am: how much weight, when pushing for policy, should I give to the fact that I might be wrong? My answer is that I need to create some kind of mechanism for other people who might be right to put their honest beliefs into practice. Big government liberals should get their say, even if I disagree. Just so, I need some checks to keep them from behaving out of selfish motives or from denying everyone else their ability to have a say. So, rights constrained democracy!

        Your answer is “No, everyone should live in an anarchy and have no effective means to advance their preferred policy if they want more government.”

        • jtkennedy

          “For example: more than one person thinks the earth is flat, but there is not a legitimate debate or controversy as to wether it is. If the goverment made a law that you can’t sail 500 miles east because you will fall off the earth, it would not be passing a law supported by a rational argument. That would not be legitimate coercion.”

          I don’t see how it would be legitimate coercion if the world *was* flat.

          • Paternalism is arguably justified on consequentialist grounds.

      • So I backed into another argument for democracy being legitimate: as long as their is significant and well-meaning disagreement about the scope and coercive power of government, democracy gives us a means to work that out. Anarchies have no government to ratchet the level of government back up, tyrannies have no easy mechanism for the marketplace of ideas to ratchet it back down, both should be avoided or only experimented with in a controlled, reversible manner.

  • Irfan Khawaja

    “true, all habitable regions on Earth have been taken over by states.”

    The Palestinian West Bank has not been taken over by any state. Area A, supposedly under Palestinian civil control, is not in fact under anyone’s civil control, and in any case, even if “it” controlled Area A, Palestine is not a state. Area B, supposedly under joint Israeli security/Palestinian civil control, is in fact under neither. It is subject to sporadic Israeli incursions but there is no de facto Palestinian civil control there.

    What state has “taken over” Waziristan, the Federally Administered Tribal Areas of Pakistan, the northern parts of Khyber-Pukhtunkhwa, or regions just north of there? Despite the legalistic names, Pakistan not only doesn’t control or administer those regions but lost control of them years ago (if it ever had control). That’s why it’s fighting to get them back. But fighting to get them back is incompatible with “having” them or “taking them over.” Tehrik i Taliban is there, but the TTP is not a state and doesn’t even pretend to be one. It’s a militarized “student movement” lacking even a unified command or leadership, much less a monopoly on the use of force in any given area. It doesn’t rule the areas in which it resides; it just resides in them and terrorizes people there. And it’s even misleading to say it resides in “them.” It moves around.

    Both the West Bank and northwest Pakistan are perfectly “habitable.” In fact, the West Bank is idyllic and NW Pakistan used to be a kind of paradise on earth.

    If these places are supposed to be “states,” it’s unclear what “state” means in this discussion. Both places strike me as paradigms of geographic locations lacking states.

    • Les Kyle Nearhood

      Perhaps but they do not lack an indigenous population which are somewhat hostile to outsiders. Nor have they formed their own anarcho-capitalist societies. In so far as they have any law, they are governed by customs and Sharia.

      • I know more about afganistan than pakistan, but i think there is more or less no Sharia in northwest Pakistan (though a lot of the locals think what they are doing is sharia). It’s pretty much all tribal customary law, specifically Pashtunwalli.

        This has been a sticking point with foreign fighters.

  • Les Kyle Nearhood

    Having viewed much of the evidence I pin my hopes for a more libertarian society upon a nation state with regular laws, borders, and a constitution. I realize I am unlikely to get most of what I want, but it is still preferable to getting a whole lot of nothing. I am not saying that forming a lasting anarcho-capitalist society has a probability of zero, but it sure is not a very great probability.

    • “Having viewed much of the evidence I pin my hopes for a more libertarian society upon a nation state with regular laws, borders, and a constitution.”

      By “regular laws” do you mean laws that nearly everyone (including yourself) agrees would be unjust for any non-state agent to enforce? Two follow up questions:

      (1) If so, do you admit you advocate unjust laws or do you deny you advocate unjust laws? If the latter, how do you account for the political authority you believe in?

      (2) If not, how is the “nation state with regular laws” you envision different than a protection agency in an anarcho-capitalist society?

      • Les Kyle Nearhood

        No, I wont play your silly word game. None of that matters a wit to me. I am a consequentialist. What you or some other philosophy major finds to be terribly unjust, I might find to be merely a matter of subjective opinion. I like things that work. Better to have a somewhat libertarian society that works most of the time than to have a pie in the sky nothing, or anarchy that degenerates into gangs and warlords.

  • Two notes to clarify Huemer’s usage of certain terms:

    (1) Huemer uses “criminal organizations” (in this article and his book) in a narrow sense that excludes governments.

    (2) Huemer uses “terrorism” to mean non-government terrorism. On page 219 of his book, he writes, “On 11 September 2001, America suffered the most devastating terrorist attack in history. […] resulting in close to three thousand fatalities. This was many times worse than any other terrorist attack ever suffered by the U.S. or any other country.” This shows that he excludes terrorist attacks committed by governments, such as the U.S. atomic bombing of Hiroshima ( ) in his usage of “terrorism.”

    I agree with Huemer’s point about semantics and making distinctions in his first comment on Chris Morris’ post above.

  • “The Progress of Liberalism” Michael Huemer Tedx Talk:

    (Link in article didn’t work.)

  • Christopher Morris

    A few comments on Michael’s response to my posts.

    – In medieval Iceland if a man told another that his writings tasted better than Georgetown cupcakes there would be a fight.

    – On MH’s 1-2. I agree that conceptual clarity should serve a useful purpose. For instance, suppose we could deter thieves more efficiently by breaking fewer bones, but to do so we would have to increase the number of threats we make. Making a distinction between coercion (i.e., threatening to do something undesirable to someone) and violence (e.g., breaking bones) was meant to facilitate asking about the desirability of this alternative strategy. Similarly, if our polities came to resemble certain medieval communes (cities) or the sort of agricultural communities that developed ten thousand years ago, would that be a move to something less state-like? In empires, much like federal states, governance is importantly “indirect”. That seems preferable to unitary states in many respects; is it less state-like? We need to clarify our concepts to raise these kinds of questions. I thought these sorts of questions would be important to anarchists who don’t necessarily think we can move to anarchy very soon or quickly. But I can understand being impatient with another’s distinctions, especially if they divert attention from the question one is seeking to answer.

    – Regarding 3. The purpose of the thought experiment was merely to show that coercion/force/violence are not part of the concept of the state, as many think. I don’t rest much on it as it is fanciful. But I do think that other features (e.g., territoriality) are more central to states than coercion and that the coerciveness of states is overemphasized in contemporary political thought, certainly by libertarians.

    – on 5-7. The reference to mobsters was not to be taken literally. I just don’t see why should not expect a good number of predatory “protective agencies”. After all, most rulers and governments started as predators, and some would say they remained such. I am not convinced that competition will pacify predatory groups. So I am not making a point about contemporary criminals.

    The second argument is not one that I make (I think). The points made seem quite sensible.

    I am not sure my “third argument” is well understood. Consider that early anarchist communities of thousands of years ago were displaced by communities with more centralized means of control and extortion (“states” in the anthropologist’s sense). Most such communities formed in reaction to the prospect of being conquered by another community. Anthropologists call this process “secondary state formation”. Refraining from centralizing means of control left these communities vulnerable to conquest. Similarly, I noted that the many varied non-state forms of social organization found in late medieval and early modern Europe were displaced by the modern state. In the terms of game theory, these alternatives – and anarchy – are not “evolutionary stable strategies” against states. I should have thought that the victory of the modern state over all alternative forms of political organization would give one pause. I grant that anarchy has many attractive features, but if it one becomes convinced that attempting to achieve it in our world would lead to one’s demise, one might cease being an anarchist. This is not an argument against anarchists who (rightly) think that a lot has to change before we can institute anarchy. But if states offer protection against being devoured by other states, then wouldn’t that be something to be said in their favor?

    I take the many points that can be raised against this line of thought. I just thought that it would make anarchists pause.

    Thank you, Michael, for stimulating a very interesting discussion. I appreciate the effort it must have taken you to reply to the multitude of comments and criticisms. With the notable exception of the man who compared your writings to muffins, the discussion was also very civil.