[The following is a preliminary draft of comments that will be presented at the APA. See also the author’s paper Why Libertarians Should Be Welfare Liberals, in which the argument sketched below is worked out in greater detail. Feedback welcome on either piece.]
Note: “Libertarian” will be used throughout to refer to those people who are inclined towards minarchism as opposed to anarcho-capitalism.
Against the State: An Introduction to Anarchist Political Theory is a lively, provocative, at times insightful, sometimes rude and often downright hilarious introduction to anarchist thought.
Anarchism/Minarchism: Is a Government Part of a Free Country? is a well-integrated collection of articles that provides a nice introduction to some of the main arguments for and against both libertarianism and anarchism.
So, one might think the topic of this colloquium of little interest to those who are neither anarchists nor libertarians. For, while many mainstream social liberals take libertarianism seriously, few are interested in anarchism. But, as Crispin Sartwell, Tibor Machan, Charles Johnson and others point out in these books, both anarchism and the debates between libertarians and anarchists should be of broader interest to social liberals. Anarchism should be of interest because it plays the role in political philosophy that skepticism plays in epistemology – raising the question of what, if anything, could justify a state in the way that brains in vats etc. raise the question what if anything could justify beliefs. The debate between anarchists and libertarians should be of interest because if the anarchists are right then libertarianism commits one to anarchism. So social liberals who take libertarianism seriously may have to take anarchism seriously too.
Obviously, it is impossible in a few minutes to summarize, never mind critique, two books. So what I propose to do in my commentary is to focus my attention primarily on one of the arguments in the debate between anarchists and libertarians (though I will make a few methodological points along the way).
What I want to do is to suggest that even if the argument I will consider for anarchy over libertarianism goes through, and I think it does, libertarians should not be anarchists, they should be social liberals of a sort. Finally, I’ll suggest that the best anarchists should be welfare liberals too.
The relevant argument (advanced by Roderick Long and Aeon Skoble amongst others in Is a Government Part of a Free Country?) is roughly this: Libertarians (and anarchists) accept a non-aggression principle on which each person should be free as long as they respect the like freedom for others. But a state which exercises a monopoly on coercive force prohibits others from defending their rights and so is illegitimately coercive as well as inefficient and expensive. Rather, competition should be allowed over the provision of protective services.
Machan responds on behalf of the libertarian that it is possible that a state could come into existence with everyone’s implicit consent. People could just buy protective services from the best agency around and that agency could come to have a monopoly on force in a particular area without violating anyone’s rights. To support the point, Machan relies on an analogy with Microsoft which, (well, pre-mac-book), had a virtual monopoly on computer sales.1
Long and others respond that this argument doesn’t explain how it can be legitimate for a state to prohibit competing protective agencies from entering the market in protective force. But Machan suggests that this is no more problematic than allowing a store to set up shop in a certain area as long as people can shop elsewhere; leaving mostly implicit a libertarian argument for open borders in a society of states.
Most businesses have at least a limited monopoly on land. Most businesses do not preclude competition in other locations but they do not allow others to set up shop in the exact same location. As long as people can go elsewhere if they prefer to purchase protective services from someone else, Machan concludes, a state would not violate anyone’s rights. Plus, Machan goes on, anarchists face a serious problem if they want to say geographically located monopolies are illegitimate. This would be like saying only online stores could sell their goods – not geographically located ones.
The anarchist might respond that neither stores nor states can prohibit competitors from operating where they like unless their competitors are violating rights in doing so. Most stores do have property rights in a particular location. So, they can legitimately prohibit anyone (including their competitors) from operating on their property without consent as long as their competitors can operate somewhere. States, on the other hand, do not have property rights to everything within their borders. Normally, they do not have property rights in the property of those who have not consented to give up their property rights. Nor do states have property in those people who do not consent to the state having those rights. So the libertarian state violates rights when it keeps other rights respecting agencies and individuals from protecting rights (just like a store would violate rights if it kept competitors from operating on land its competitors own). Saying people can move elsewhere does not answer the objection.
So, I think the premises in the anarchist’s argument are correct (though I’ll suggest that with a few more relatively uncontroversial premises it is possible to show that libertarians should actually be welfare liberals).
Of course, libertarians could argue that something could qualify as a state without claiming a monopoly on coercive force (if no other entities wanted to enter the market). Machan seems to suggest something like this at points. But then someone could be both an anarchist and a libertarian. For, anarchists would be happy with the mere possibility of entry into a market in coercive force, libertarians happy with the mere possibility of a single entity being the only agent to exercise coercive force in a given area (the agency need not have an in principle monopoly).
But it is hard to believe the main issue between libertarians and anarchists could just be about whether it is acceptable for there to be an in principle monopoly on coercive force. At least, that would make short work of the arguments Sartwell gives in Against the State. For, he is only concerned to show that all actual states are illegitimate and, given a sympathetic reading, he does this primarily by arguing that it is not clear that any actual states fulfill the existing conditions for legitimacy philosophers advance.
I think there is much more room for, in principle, ideal theory arguments in political philosophy than Sartwell allows. I also believe it would be good if Sartwell considered whether states can be more or less legitimate in the real world. But, I think Sartwell is right to suggest that it really matters whether we should have states in the real world. So, I will suppose, in what follows, that anarchists want to deny that we should have minimal states (with an in principle monopoly on coercive force) while the libertarians will disagree.
Before trying to show, however, that libertarians should be welfare liberals, let me make those methodological points I promised. For, if it really matters what we should say about states in the real world, we need to consider the relevance of empirical evidence to the debate between libertarians and anarchists. And this is especially important in the present context since many arguments both for and against anarchism in Is a Government Part of a Free Country? hinge on anarchy’s empirical desirability. Some suggest for instance, that anarchism will lead to an all out war of all against all. These authors argue that to avoid this war, we must have a constitutional state that can act as an impartial arbitrator over disputes. Adam Reed and William Thomas, for instance, claim to provide historical evidence to this effect suggesting that existing anarchies are generally unstable, can only exist in poor, unimportant, under-developed places or lead to great injustices. Although these authors provide some compelling anecdotes (talking, for instance, about the Polish Lithuanian Commonwealth) there is a serious problem with this method of argumentation. To show that something is always, or even generally, the case we need not case studies, but macro-level evidence. Similarly, though the anarchists like Skoble and John Hasnas can defend themselves against the claim that anarchy always has bad consequences with counter-examples, they cannot make a general case for anarchy with case studies. Furthermore, as Jan Narveson and others insightfully point out, even if there were sufficiently high quality macro-level evidence that applied equally to modern developed countries as to developing countries or tribal communities, it would almost certainly be aggregated data. Hence, those who are concerned about each person’s freedom should be unimpressed. Rather, it is likely that some have done and would do better under some anarchies than under some states while others have done and would do worse under some anarchies than under some states. Or to put the point differently, some would probably be better off under almost any state than under almost any anarchy and vice versa.
Which, brings me to a final methodological point. I think both anarchists and libertarians should wonder a bit what the point is in the empirical arguments for and against anarchy. Are the arguments against anarchy, for instance, intended to show, as Narveson suggests, that anarchy is always unreasonable? And, if so, why should the anarchists care? Anarchists might ask why, if they want to be unreasonable, shouldn’t they be? What can justify forcing them to live in a state if they, however unreasonably, prefer anarchy? If what is really motivating the anarchist is a concern for individual freedom, it is hard to see how reasonableness matters. Rather, anarchists might suggest that, and if what I’ve said so far is right, libertarians should agree that, only consent can justify a state.
Which, finally, brings us to the argument I said that I’d provide at the beginning. If what I’ve said so far is right, and libertarians should agree that only consent can justify a state then, as I’ve intimated, I think libertarians should accept some kind of welfare liberalism. To see why, let us suppose that libertarians do not accept anarchy despite the anarchists’ argument, they only accept actual consent theory.
More precisely, let us make two assumptions (defended or implicit above):
1) Libertarians agree that any existing states must be legitimate and some states should exist.
2) Libertarians hold that for any existing states to be legitimate it must only exercise coercive force over (rights respecting) individuals to protect these individuals’ liberty.
From the argument that libertarians should be anarchists we get:
3) Libertarians should agree that for states to be legitimate, they must secure their subjects’ autonomous consent.
I’d just add:
4) For states to secure their subjects’ autonomous consent, they must do what they can to enable their subjects to secure sufficient autonomy to autonomously consent to their rules.
5) To secure this autonomy most people (in all states) must be able to secure some minimal amount of healthcare, food, water, and shelter.
6) So, states must do what they can to enable most of their subjects to secure some minimal amount of healthcare, food, water, and shelter.
Suppose further that the libertarian will accept the following implicit premise (which we can take to imply some commitments with regard to the definition of welfare liberalism):
7) If libertarians must agree that states must do what they can to enable most of their subjects to secure some minimal amount of healthcare, food, water, and shelter, they must be (some kind of) welfare liberals.
We reach the following conclusion: Libertarians should be (some kind of) welfare liberals. 
Although I obviously cannot do full justice to this argument here, and there may be other ways à la Charles Johnson to arrive at social welfare policies from anarchistic principles, let me just say a few words about this argument’s 4th premise (I can share a paper with a longer defense later with anyone who is interested).
The reason libertarian actual consent theorists have to accept the thesis that, insofar as possible, people must be able to secure sufficient autonomy for the states to which they are subject to be legitimate is this. In order for someone to actually autonomously consent to a state that person must be able to do so. But, we must say more to convince libertarians that states must do what they can to enable their subjects to secure sufficient autonomy. Consider an argument for this conclusion. When states subject people who cannot secure sufficient autonomy to coercive rules and do not do whatever is possible to enable these people to secure sufficient autonomy, they act wrongly. This is because such states are not justified in exercising a monopoly on coercive force over those who cannot secure sufficient autonomy. states continue to exercise a monopoly on the use of coercive force over their subjects, legitimacy requires that they do whatever they can to enable these people to secure sufficient autonomy.2 Insofar as they exist, states do continue to exercise such a monopoly. So, states are obligated to do what they can to enable their subjects to secure sufficient autonomy.
One might object that states that subject people to coercive rules, even wrongly, do not thereby acquire an obligation to do what they can to enable these people to secure sufficient autonomy. Consider an analogy. Suppose someone, let us call her Samantha, who is not capable of autonomous consent agrees to give me a large sum of money. I do not thereby have a duty to do what I can to enable Samantha to secure sufficient autonomy. I merely fail to have a contract with her. Samantha has not, by agreeing to give me a large sum of money, incurred an enforceable debt to me. If I were to try to enforce the agreement on Samantha without securing her autonomous consent, I would act wrongly. But, as long as I do not try to extract any money from her, I have no obligation to her. Similarly, one might suggest, libertarians can deny the legitimacy of actual states. Yet, they can maintain that something like a state or protective organization that only enforced the rights of those who actually autonomously consent could be legitimate. Such protective organizations would not need to enable anyone to consent. Libertarians could argue as follows. The fact that legitimate states must secure all of their subjects’ autonomous consent just shows that there should not be states. Rather protective organizations could enforce the rights of those who actually autonomously consent to their rule. They just cannot enforce the rights of those who do not or can not autonomously consent.
But that would commit the libertarian to anarchism and we’re assuming that they are not anarchists; libertarians believe that in principle minimal states can be justified and do not advocate eliminating all states.
One might worry that this response relies on a false premise. According to the response, states can be legitimate only if they do what they can to enable their subjects to secure sufficient autonomy. Perhaps relatives or charities can enable these people to secure sufficient autonomy. Sticking with the analogy, the objection would be this. In order for Samantha to autonomously consent, I need not do what I can to enable her to do so. Perhaps her family or others involved in charitable work can help her instead. I may be able to legitimately enforce the contract without doing what I can to enable Samantha to autonomously consent.
This objection has some truth in it. Others may be able to enable those subject to a state to secure sufficient autonomy. Others may even have primary responsibility for doing so. But the objection misunderstands the nature of enabling. Enabling is like being a lender of last resort. So, in some cases, states may not need to do anything to enable someone to secure sufficient autonomy. If a person secures autonomy on his or her own or with the help of friends and benefactors the state need not do a thing to help this person. States need to step into the breech, however, if help is required. It is only if states do this that all of their subjects who are capable of securing sufficient autonomy will do so; so states must do what they can to enable their subjects to secure autonomy. This is the only way states can be legitimate in our imperfect world.
Perhaps we also require the empirical assumption that voluntary assistance hasn’t and won’t immediately eliminate poverty. But this assumption is minimal. So, if there is a single person only a state could help secure the ability to autonomously consent it must do so.
The anarchist’s argument against libertarianism provides a key premise in reaching the conclusion that libertarians should be some kind of welfare liberals.
I’m out of time but since I promised not only to argue that libertarians should be welfare liberals of a sort but to explain why I think the best kind of anarchists are welfare liberals too let me just say this. An anarchist can accept the following: It is wrong to violate others’ rights, and ideally there would not be a state. But other things matter too (e.g. life, happiness, etc.). So, if the only way to avoid massive suffering and death, for instance, is to violate rights and institute welfare states, doing so may be justifiable. For a world where 18 million people die annually of easily preventable poverty related causes is anything but ideal.
Nicole Hassoun
Philosophy, Carnegie Mellon University
Notes
1 Still, Microsoft never had a complete monopoly and it is hard to see how any protective organization could have such a monopoly as some cannot enter into free contracts at all.
2 I assume here and in what follows that at least some of these subjects respect others’ rights.