Some scenarios in anarchy

On a thread at UR, Nick Szabo gave a series of specific questions regarding the functioning of anarchy, so I thought I would answer him. Before jumping into it, let me clarify two things.

First, I am no prophet of AC. Perhaps you ask David Friedman, you get different answers. Perhaps neither of us is farsighted enough to exactly predict what would happen. Again, as I have said before, anarchy is defined not by the function but the form: market provision of protection services, including competition among agencies with overlapping territory.

I usually discuss AC for simplicity using what I have found in Friedman and elsewhere: protection is provided by protection agencies (PAs) , which are a sort of cross between police agency and insurance company. Adjudication is supplied by separate corporations specializing in it. Everyone has an ongoing relationship with one PA (or maybe more if needed). But is it possible for the system to work without having existing "insurance" client relationships? I think it would be... and that might even be general. It is just hard to say what sort of forms would happen in a free market, because it is so different than what we have now. (If you want some taste of weird arrangements of governance, then hie thee over to Nick's blog and read about jurisdiction as property.)

Second, there is a general property of in anarchy that is worth reiterating, since I have not seen any anarchists making this point other than me. PAs are weaker than states. Therefore, any institution which serves to rein in the state, must also serve to limit PAs, unless somehow it relies on the monopoly of coercion that define the state. Thus, any of the numerous techniques people have invented to moderate the badness of states will also work on PAs. (I discuss this more here.) Because of this, I expect the normal PA (and/or its adjudication agency) to have a constitution (that it actually follows!), an extensive bill of rights, perhaps some sort of limited democracy, etc. Honestly: would you submit yourself to a PA that promised you nothing of that sort, if you could at little or no additional cost, submit yourself to a PA that did? I wouldn't.

Anyway, here are answers to the specific questions Nick asked, which do serve to bring out some of the functioning of anarchy, as I see it.
A factory under the protection of one PA emits pollution, which probably causes extensive damage to the value of properties of hundreds of people, customers of several dozen different protection agencies, downwind.
Probably? Well. Let's just assume it is damaging. So, the aggrieved parties get together and take the factory-owner to court. Presumably some settlement is reached. Then, further assuming that the decision goes against the factory-owners, the PA enforces it on them if they do not comply.

Why should a PA enforce laws against its own paying customers? Profit. I picture every PA as existing in a state of bilateral treaty or contract with every other PA which it might interact via conflicts of customers. This treaty might be explicit (written), or implicit. Either way, for the PA to unilaterally pull out of its treaty with any other PA is to throw its honorable reputation in doubt. This alone seems strong enough that a PA would rarely be tempted. But beyond mere reputation is the potential for collective action by other PAs. They could embargo the PA in many ways that would damage its profits. For example, most obvious would be to declare all judgments going against customers of the PA in question to be voided or in abeyance until it submits.

Trade sanctions are not effective in the state system, because there is no way to pressure the ruling elite enough without drastic effects across the society. And also, because the elite are not motivated by money, but power. Most ACs picture PAs as corporate (as I do), so they are motivated by profit. And their "interstate" trade is orders of magnitude higher.

A PA security officer suspects that a person he sees driving away from the scene of a burglary probably committed that crime, but is not at all sure. The officer arrests the suspect and locks him in prison. A week later the real criminal is found and the suspect is released. But meanwhile the suspect is fired from his job because he failed to show up to work that week.
I doubt this would end up in any way other than what currently obtains. But it is true that states are far more arrogant in power than I picture PAs being; so there is some chance that a victim of such high-handedness might be able to sue and collect damages.

The economic tradeoff that it comes down to is this: customers want more rights, but more rights cost profits. So, ceteris paribus a PA with a policy of "you can sue us" will get more customers. If the average man is willing to pay for the extra dollop of protection he gets, then the whole thing is viable and you see it. (Probably widely, since people's desires are not that variable.)

Note that such economic tradeoffs are of no interest to a state, which has every incentive to declare itself sovereignly immune. What, are you gonna move?

Repo man working for a bank under PA2 repossesses a car owned by customer of PA1 who is late on his payments to the bank. Security officer from PA1 believes he sees a theft in action and arrests and imprisons the repo man.
Again, this sort of thing can happen now within a state. I.e., a repo man from one jurisdiction gets caught by some other. This is an example of where there is almost certain to be some sort of adjudication already in place. So, after some delay perhaps to check up on the repo man's authenticity, he is released. Again, whether or not he can sue the PA for mistaken arrest is not clear, and may vary by PA.

A murderer, belonging to PA1, which forbids the death penalty, is caught and convicted by PA2 that liberally applies the death penalty. PA2 executes the murderer over the strong objections of PA1.
Easy: he's dead. But you said that. Really, this outcome assumes there is already agreement between the two (which again, there almost certainly would be), and that the terms worked out by PA1 and PA2 allow PA2 to kill convicted murderers.

Continuing the above example, PA1 in retribution disavows its contracts with PA2.
Then they are edging towards war. Perhaps a bit much for a murderer. But surely possible. In any case, highly unlikely given that they are profit-maximizing, as in the first example. Their trade with PA2 may be minimal and worth sacrificing to score points with their own customer base. But their trade in general, with all other PAs, is probably not minimal. Therefore, PA2, if they can, would be advised to seek out collective action to get PA1 to back down.

But what if nothing happens? Then, unless PA1 are willing to go to war, which is unlikely for any profit-maximizing body, then that is that. PA1 may well advise its customers who want to murder people that they should check to see if the victims are PA2-people first. I kid. The point here is that unless the customers of PA1 are truly fanatical in their anti-death-penalty stance, then it is unlikely that they are willing to pay the costs necessary to prosecute a war or even lose the trade possibilities with PA2-people. (I don't know of anyone that fanatical, not even progressives at their worst.)

A customer takes the opportunity to leave his protection agency. In retribution, the protection agency kills him, and threatens any other customers who leave with a like fate, although one can still leave the PA voluntarily after paying a 50% wealth tax as a final payment for services rendered. To justify these acts, only recently implemented, the PA cites a long-ignored passage, obscurely worded and in fine print in the 500-page PA contract but absent from all other PA materials describing their services, by which the customer agreed to give the PA a number of powers of legal procedure "and all other sovereign powers" over the customer.
Great profit opportunity for other PAs here. Collectively, or individually, they sue in whatever means seem presentable, perhaps "class action". The point is that the PA is a fraud, and just murdered a man. But they need not wait for the law's wheels to grind. Rather, they announce their near-certainly of the outcome, and that they are willing to go all the way to war against the PA if it attempts any such thing on any other customers that wish to defect. Perhaps this is backed with something like an injunction, which they can get in short order.

The next day, the PA has no customers left. Perhaps the CEO hangs. He did, after all, kill an innocent man and attempt to semi-enslave his customers. But more likely, the threat of mass abandonment by jealously free customers would make all PAs with any significant competition think more than twice about any such slaving scheme.

PA2, armed with nuclear weapons, seizes an important oil field owned by the customer of nuclear-free PA1, claiming, PA1 officials and the owner believe falsely, that the owner reneged on an obscure pledge in the fine print of his mortgage regarding religious practices.
Back to court, and the first scenario. In this case, I think Nick is hinting at the idea that if a crime is profitable enough, perhaps it would be worthwhile to sacrifice relations with every other PA, and lose most of your customers over it. (Or perhaps you just form a mercenary army and forget about being a PA as such.) Well. Then you might possibly organize the rest of the world against PA2, and refuse to buy their oil or do any other trade. Could the blok hold? I don't know. What this boils down is a state forming. It would not be a nice state, at all. Who would want to live in a hermit nation? But it is at least possible that AC might end up with slavers sitting on particularly valuable resources, at least so long as the resources hold.

For Neocameralism, Against Morality

Morality, a consistent morality, was one of the reasons that I long ago became a libertarian, and then an anarchist. I took the precepts of morality as I received them, and pushed them hard. Thus, I accepted (and I still accept) the libertarian argument that taxation is theft. If taking property by force is wrong, then voting about it does not make it right. You can see such arguments widely in the libertarian ideosphere; I am certainly not alone. (I would guess that most libertarians are highly intelligent moralists, usually self-educated.)

There is a big problem, though, with morals as a foundation for politics. It is this: it may be that a moral society cannot last. That is, that there is a "tragedy of the commons" with morality itself. It may be the case that a consistently moral society cannot compete against immoral societies. One example of this is the common idea that a anarchic protection agency would not be able to effectively defend territory against state incursion. Another example is the conservative idea that civilization itself is created by patriarchy. If that is so, then a progressive civilization is not possible in the long run. Yet another is the idea that the welfare state is not compatible with free immigration.

To take a case I know I have written about some years ago: taxation. To many people, it seems commonsensical that that a state that will not tax will be beaten militarily by other states not so squeamish. But here is me, 15 years ago:

[Q:] How do libertarians feel about taxes?

The state must initiate coercion to tax. As such, taxes should be abolished.

[Q:] I'm for cutting taxes, but as a practical matter, how do we do it?

This is not a practical FAQ. Morally speaking, we should end all taxes first and figure out how to solve the resulting economic mess later.

[Q:] Aren't you going too far?

For those in the statist mindset? Yes. But you see, they DON'T BELIEVE in one of our basic moral axioms.

For us? No. Once you identify what is moral -- what is acceptable and not -- then logic compels you to accept the social and political effects of your morality.


I became an anarchist, in part, because anarchy offers a partial solution to the problem of suicidally moral government. Anarchists conceive of protection agencies as government-like, but still corporations subject to market forces, including most particularly competition. So, they can have policies, which would be immoral if forceable imposed. The moral objectionableness is reduced, at least, because the affiliation of the customer with the agency is (to some degree) voluntary.

Of course, there are still moral problems, even with anarchy. For example, what if a customer only has one protection agency to choose, or two, even. Two does not seem like much choice -- might be a choice between "pay 30% income tax" and "pay 31% income tax". And of course, there is the lingering concern that anarchy itself is unstable; that the agencies would not be able to defeat a protostate because they are hamstrung morally.

With relatively modest moral assumptions, neocameralism offers us a way out of this conundrum. Note that the defining feature of democracy, including the chimerical "limited government", as well as anarchy, is that government is based on the will of the governed. Thus, it is transparent to the morals of its citizens or customers. That is, there can be no unpopular law. (This is in the abstract. In the real world, "friction" of various kinds means there can be unpopular law down to a certain minority level of support. However, abstractly at least, the principle applies.)

By contrast, in neocameralism the source of state power is not the subjects; it is an earlier conquest. The end of power is profit to the stockholders. Note two things. First, both of these are invariant. Because they are stable (unlike popular opinion), the law will also be stable. Second, note that neither the justification for power, nor the ends to which it is put, are derived from moral reasoning. They simply are. They bear no relationship to morality at all.

Thus, it is possible for a neocameral society to have laws that are not possible in democracy or in anarchy. This would be any law which is profit-enhancing, but which is immoral. Obviously, if people are unanimously against the law, it will be abolished; but even a minority who support would be sufficient to uphold it so long as the sovcorp's owners wanted it. Or, alternatively, if it was not sufficiently immoral that men would refuse to enforce it.

For example, consider laws controlling immigration. It is certainly possible to be against them in the abstract while nonetheless enforcing them, or supporting their enforcement, as upholding the rule of law. That is, while they may be objectionable, at the same time, it is even more objectionable to have a body of unenforced law.

Thus, it is possible for a neocameral society, even composed of entirely moral men, to implement laws that no person holds as moral!

Now, most of us intuitively sense this, at least in the case of bad laws, immoral laws. This is our progressive education speaking: cherchez la genocide, it whispers. What, the state can fire up the ovens, and nobody can stop it?? Well, yes, it can. (That it has incentive not to -- well, progressive education does not teach anything about incentives and their effects.)

But I think few us consider the case of good laws, or at least, necessary laws, which are nonetheless hard to square morally. In progressive terms, the category is empty; it's oxymoronic, because the law and righteousness (social justice) are one. As anti-progressives, we should reject such reasoning, and look at the case with fresh eyes.