Bryan Caplan  

EconLog Book Club: For a New Liberty, Chapter 12

Merton on the Financial Crisis... Hayek Interviewed...
A common reductio ad absurdum of libertarianism is that it implies anarchism.  Rothbard now eagerly bites this bullet.  Indeed, this chapter is the most detailed defense of anarcho-capitalism that he ever wrote.

Rothbard begins by sketching how a free market would handle "police, law, and the courts."  In a nutshell: Police would sell subscriptions; courts would be hired by police firms to resolve their disputes; the law would emerge from judicial firms' quest to attract customers with their wise, streamlined legal codes.

Rothbard answers many questions along the way, but he devotes the most time to...
...a final nightmare which most people who have contemplated private protection agencies consider to be decisive in rejecting such a concept. Wouldn't the agencies always be clashing? Wouldn't "anarchy" break out, with perpetual conflicts between police forces as one person calls in "his" police while a rival calls in "his"?
Rothbard has one answer for "honest disagrements," and another for "outlaw protectors." 

Honest disagreements, he says, would rarely lead to violence, because it is in businesses' interest to hammer out peaceful ways to resolve disputes:
Every consumer, every buyer of police protection, would wish above all for protection that is efficient and quiet, with no conflicts or disturbances. Every police agency would be fully aware of this vital fact. To assume that police would continually clash and battle with each other is absurd... To put it bluntly, such wars and conflicts would be bad--very bad--for business... [A]ll conflicts of opinion would be ironed out in private courts, decided by private judges or arbitrators.
Doesn't this just push the problem back a step?  No.  Rothbard offers the reader a complete flow chart:
Mr. Jones is robbed, his hired detective agency decides that one Brown committed the crime, and Brown refuses to concede his guilt. What then?... Jones would go to the Prudential Court Company to charge Brown with theft.

It is possible, of course, that Brown is also a client of the Prudential Court, in which case there is no problem. The Prudential's decision covers both parties, and becomes binding...

What, however, if Brown does not recognize the Prudential Court? What if he is a client of the Metropolitan Court Company? ...The case will then be heard by Metropolitan. If Metropolitan also finds Brown guilty, this too ends the controversy and Prudential will proceed against Brown with dispatch. Suppose, however, that Metropolitan finds Brown innocent of the charge. Then what? Will the two courts and their arms-wielding marshals shoot it out in the streets?

Once again, this would clearly be irrational and self-destructive behavior on the part of the courts... Thus, an essential part of any court's service to its clients would be an appeals procedure. ...The appeals judge would make his decision, and the result of this third trial would be treated as binding on the guilty...

An appeals court! But isn't this setting up a compulsory monopoly government once again? No, because there is nothing in the system that requires any one person or court to be the court of appeal...

But suppose Brown insists on another appeals judge, and yet another? Couldn't he escape judgment by appealing ad infinitum? ... In the libertarian society, there would also have to be an agreed-upon cutoff point, and since there are only two parties to any crime or dispute--the plaintiff and the defendant--it seems most sensible for the legal code to declare that a decision arrived at by any two courts shall be binding.
OK, what about "outlaw protectors"?  Rothbard's answer: the free market provides built-in "checks and balances" that make dishonesty and open violence hazardous to the perpetrators' health:
What would keep the free-market judges and courts honest is the lively possibility of heading down the block or down the road to another judge or court if suspicion should descend on any particular one... These are the real, active checks and balances of the free-market economy and the free society.

The same analysis applies to the possibility of a private police force becoming outlaw, of using their coercive powers to exact tribute, set up a "protection racket" to shake down their victims, etc... [I]n contrast to present-day society, there would be immediate checks and balances available; there would be other police forces who could use their weapons to band together to put down the aggressors against their clientele. If the Metropolitan Police Force should become gangsters and exact tribute, then the rest of society could flock to the Prudential, Equitable, etc., police forces who could band together to put them down.
Rothbard ends the chapter with a brief discussion of national defense.  Even if the Soviet Union were "hell-bent on on attacking a libertarian population," he opines that decentralized deterrence might work better than what we've got.  And if that falls through, we've always got guerilla warfare:
[T]he occupying Russians' lives would be made even more difficult by the inevitable eruption of guerrilla warfare by the American population. It is surely a lesson of the twentieth century--a lesson first driven home by the successful American revolutionaries against the mighty British Empire--that no occupying force can long keep down a native population determined to resist.
Critical Comments
This chapter that has literally converted thousands of highly intelligent, economically informed people to anarcho-capitalism.  The first time I read it, I thought it was crazy.  By the tenth time I'd read it, Rothbard had made a believer out of me.

What's the appeal?   In all honesty, its sheer eloquence is hard to resist.  I still get chills from passages like:
And, indeed, what is the State anyway but organized banditry? What is taxation but theft on a gigantic, unchecked, scale? What is war but mass murder on a scale impossible by private police forces? What is conscription but mass enslavement? Can anyone envision a private police force getting away with a tiny fraction of what States get away with, and do habitually, year after year, century after century?
But there's a lot more to this chapter than soaring sentences.  While Rothbard doesn't use this term, his main intellectual achievement here is to show that anarcho-capitalism is a plausible Nash equilibrium.  If a free market in police, courts, and law were established and expected to continue, then Rothbard persuasively argues that no major actor would have a strong incentive to undermine it.  If one police firm stopped peacefully settling disputes, its costs would skyrocket and most of its customers would abandon it.  If a judge became corrupt, he'd ruin his reputation and lose most of his business.  If one firm tried to become the new government, the most likely result is that it's employees would quietly fail to show up for work.

Alas, world history shows that there are many other equilibria, and it's extremely difficult to explain how to move to the good equilibrium that Rothbard sketches.  If the U.S. government dissolved itself overnight, almost everyone would expect violent chaos, and this expectation would be self-fulfilling.

Also of note: Even Rothbard lacks the chutzpah to confidently claim that decentralized "national defense" would better deter a Soviet invasion than the Department of Defense.  He makes up for this chutzpah shortfall, though, with the single most ridiculous argument in the entire book:
But suppose--just suppose--that despite all these handicaps and obstacles, despite the love for their new-found freedom, despite the inherent checks and balances of the free market, suppose anyway that the State manages to reestablish itself. What then? Well, then, all that would have happened is that we would have a State once again. We would be no worse off than we are now, with our current State. And, as one libertarian philosopher has put it, "at least the world will have had a glorious holiday." Karl Marx's ringing promise applies far more to a libertarian society than to communism: In trying freedom, in abolishing the State, we have nothing to lose and everything to gain.
I don't think I've ever seen paragraph with a greater disparity between beauty and truth.  Have you?

Comments and Sharing

Twitter: Bryan Caplan @bryan_caplan

COMMENTS (21 to date)
Barry Cotter writes:

While I agree with you that anarcho-capitalism is probably a Nash equilibrium I think the section on guerilla warfare is well, bollocks.

If you're willing to kill everybody to keep the patch of ground, you won't need to kill everybody. The Jihad only had to kill about 30% of the Afghan population (on average) to win them over, and then as now, Afghanistan was a collection of very, very clannish tribes, living in hilltop forts, engaged in warfare and banditry as a way of life.

Now at any technological stage where most of the value of a country is knowledge workers, (post WW2 for Europe, probably) war makes no economic sense whatsoever, even on relative basis, if you're aiming to make a profit, but it might if you're a hegemon or Great Power that's dominant in your region. It makes fairly obvious sense if you're invading to say "No, you may not have a civil war, here have five years off from democracy while we sort this crap out." if the invadee is nearby and you'd have to deal with refugees and lawlessness otherwise.

david writes:

If a free market in police, courts, and law were established and expected to continue, then Rothbard persuasively argues that no major actor would have a strong incentive to undermine it.

It's quite startling to do so without any actual numbers! There are dramatic economies of scale in, say, operating a company of mercenaries. Marines operate in certain numbers for a reason.

Besides that - how are customers supposed to freely switch between service providers in the face of organised firepower? Suppose that I, Armed Mob Private Limited, seize four city blocks and extract rationally optimized rent. Another armed mob could plausibly take me on, but my new peons can't possibly pay them more than what I can extract from my peons by force (hint: see how real-life gangs sort out territory). And I can stop anyone from leaving! This seems like a much more likely equilibrium.

caveat bettor writes:

The police subscription sounds interesting, but I suspect that conflict-mediation will still be similar to Law Firm vs. Law Firm, not unlike the MAD Magazine Spy v. Spy series or the approximate justice that we see today. I fear that justice will not be blind to whoever has the deeper pockets in Rothbard's system.

spencer writes:

We have a real life example of what kind of system would evolve if we tried your theory-- it is called the Mafia and it does not operate anything like you theorize.

Blackadder writes:

I found this to be the least persuasive chapter so far (and the only application chapter so far where I've been less convinced of the libertarian position after reading it than before). It's not that Rothbard doesn't have some good points, but there are too many instances of hand waving (e.g. protection companies providing free protection to people who aren't clients in exchange for a voluntary donation afterwards). And the end of the chapter is almost tantamount to an admission that libertarianism can't actually provide national defense (how could a anarchist society defend us against the Soviets? Well, assume that there aren't any Soviets; problem solved).

Alex J. writes:

Spencer, the mafia thrives in activities that are specifically banned and suppressed by the government. Competing systems of justice are currently banned and suppressed by the government. The theory has not been tried and found to result in mafias.

Alex J. writes:

If you tried to seize four city blocks, you'd have to watch your back in those blocks and all the neighboring ones, because the residents would rightly fear you.

david writes:

... how does that present an obstacle to someone who has already seized control of such city blocks?

El Presidente writes:

It's only a stable equilibrium until you begin to get hostile takeovers and market consolidation. Then you move closer to oligopoly, duopoly, monopoly, and a state. The incentives still exist, given enough time, to seek a monopoly on the use of force.

A "holiday" is hardly a holiday if it leads to massive instability in the interim AND in the end produces nothing better than what we already have. The presumption that this new freedom would be glorious is nice, but if we cannot say what it would look like with any certainty we are hard-pressed to justify that proposition with anything other than normative principles. That's fine. We all have them. But it is quite a different argument.

At the risk of confusing the subject, a similar pattern can be seen in religious apologetics. Authors start from a sacred principle and then create arguments for why it must be so. If it is a true principle, the argument does not enhance the principle. If it isn't, the argument likewise cannot enhance it. The argument is a means of persuasion. Rothbard has chosen his principle and made his argument. It is eloquent, but I think his principle needs balance from some additional principles that others find valuable.

sean writes:

I hope we move to a market anarchist system so that I can employ my boyish good looks, dashing charm, and charismatic eloquence to finally become a war lord.

darjen writes:

This chapter definitely made a believer out of me as well... despite its shortcomings. I still think voluntary militias could provide adequate national defense. The American colonies beat the British without anything resembling national defense as we think of it today. Besides, the military is hardly ever used to actually defend America.

Zac writes:

This chapter didn't do too much to convince me. Maybe this is because by the time I first read FANL (and certainly now), I had read much better economic justifications for anarcho-capitalism. Rothbard's definitely got some emotional appeal, but his defense of a free market for law is weak at best, self-destructive at worst (like the quoted paragraph).

I do appreciate that he addresses decentralized 'national defense.' There's no really good response to this issue besides: its an empirical question. There are some good reasons to think it is not as big an issue as some suggest, but it is also not crazy to think of it as a total dealbreaker for libertopia. It seems clear to me though that there's at least a good chance that no external force would successfully engage in conquest of libertopia, and in fact the success of libertopia would be likely to foment revolution in those states and encourage people to reject their geographic defense monopolists and demand private law and defense of their own. I think we could have had a similar discussion in the middle ages about the prospects of eliminating serfdom.

A good way to empirically test this? Create a real market for government services and see what happens.

pedro writes:

Bryan's first impression was the correct one. Rothbard's assertions require Marx-scale leaps of faith.

Bob Murphy writes:


Please excuse the self-promotion, but I try to answer a bunch of the objections above:

* For chutzpah, see 2nd essay here (pdf).

* For a shorter take on just the warlords, see this article.

* For the mafia objection, see this.

Bryan, 5 years ago I would have agreed with you about that argument from Rothbard, but now I don't think so. I assume you mean that the State that takes over after the brief an-cap period would be awful. But I don't see how, if it's domestically generated. I.e. such a State would necessarily have to be really teeny weeny in the beginning, because it would grow from scratch.

And since I think an an-cap society could defend itself much better than a statist one, then any outside conqueror would have a harder time imposes tyranny on the ancappers.

Glen writes:

To think about the stability of the Rothbardian equilibrium, I think it's important to consider network externalities. Part of the appeal of any given protection agency to customers will be the number of other customers it has, since it will presumably be lower cost for a company to deal with intra-firm disputes than external disputes. That means larger companies will be able to offer lower prices (even without economies of scale). As a result, while there might be an equilibrium of N equal-sized firms, if any firm gets a little larger than the others it will have a tendency to grow. This suggests the equilibrium is unstable.

Political Observer writes:

This is one part of libertarian economics that I have always struggled with. It just seemed to me that in this area you were just substituting one form of government for another. The only difference was how you arrived at that government.

To a degree we have some forms of this model in operation in our current economic system. It is not uncommon for parties in contract to have an agreement for private arbitration of issues that cannot be resolved by the contracted parties. We also see in some countries - private police forces that operate with tacit police powers.

Unfortunately we also have living examples in the extreme of this economic model (Somolia, the Mexican drug cartels, etc.)

My point is that going the full route as suggested by Rothbard is putting us on the path to anarchy. While the small and numerous "minor" disputes will be solved in a non-violent manner (as one would expect rational beings to do) the larger issues will ultimately be resolved by force (which unfortunately rational beings also do).

Arare Litus writes:

A few points:

(1) In a limited manner we *already have* Rothbard's anarchy - the relationship between nations is anarchic. We can look at the current, and past, relationships as examples of equilibria that evolved. Subscriptions to the better "police" can be canceled, and taken up with another (I believe all OECD countries allow you to give up citizenship), the worse police subscriptions do not allow this. Sure, subscription is spatially determined, but this is one equilibria. The problems and advantages we have now are useful for study of anarchy - we have a rich data set of (a limited type) of anarchy available right now to inform our thinking.

(2) guerrilla warfare & voluntary militias: don't forget your basic economics. Specialization is important. Imagine some guerrillas against some marines - now imagine those marines not constrained by their home country in battling (i.e. imagine something more like Russian special forces). Yes, guerrillas would have specific victories, kills, etc, but it would not be a fun fight for them. I would have more faith in specialized police forces being the deterrent.

(3) Related to #2, @Zac: "a good chance that no external force would successfully engage in conquest of libertopia, and in fact the success of libertopia would be likely to foment revolution in those states and encourage people to reject their geographic defense monopolists and demand private law and defense of their own." Your second point would give incentive to attack libertopia, and as conquest need not be a goal - only partial destruction, one could easily get a "Iraq war II situation", where a country claims preemptive self defense and kicks down the place. One can observe how easy it was for the US to kick the place down, staying and policing (with lots of constraints) is much harder - but presumably a kick down strategy would be more in line with an enemy of libertopia.

Zac writes:

Several commenters have made the appeal to defense as a "network industry," and might be interested in reading Tyler Cowen's "Law as a Public Good: The Economics of Anarchy," Economics and Philosophy, Fall 1992, 8, 249-67, David Friedman's reply (and Cowen's rejoinder, 1994), and also our host's reply, available online.

Kurbla writes:

The main problem with Rothbard is of course, that modern countries satisfy his definition of anarchy. The simplest example, Saudi Arabia is private property of patriarchal Saudi family. Saudi king writes the laws on his property, if you do not like, leave it. Sure, you didn't see Saudi Arabia when Rothbard spoke about cinema example - but there is no difference - except in size. Furthermore, if king decides to transfer its property on all citizens collectively - you have modern democracy. The world Rothbard imagines does not result from his principles. He imagines existing world but with increased granularity - states are just small, and many states are just individual properties, and families live on these properties. Nothing in Rothbard's theory prevent mini states growing. But we can still discuss that imagined world.

One problem is that protection agencies cannot work without breaking property rights. They must break property rights to search for evidences of crime. I do not speak about abuse, but about normal work of the system. Instead of one agency (or one court) that search my house without my consent - the state, I ended with system where everyone can rightfully do that. An alternative is that no agency or court can search my house for evidences, without my consent. It is more in the spirit of the private property, but in that case, of course, no criminal will ever give his consent. Hence, system can work only if it is broken to certain degree, if property rights are protected to the level which is not defined by system but by practice.

Another problem is protection of those unable or unwilling to protect themselves. Ignoring poor is ugly - but lets assume that charity is the solution. However, what about children abuse - preventing such abuse seems to require laws superordinate to the property laws. If parents or adoptive parents neglect health of their children, or if they even abuse them, anarchocapitalist society cannot do anything about that.

OK, I'll stop here.

Tracy W writes:

Every consumer, every buyer of police protection, would wish above all for protection that is efficient and quiet, with no conflicts or disturbances.

Why do we have wars at all then if everyone wishes above all for no conflicts or disturbances? How about people who are risk loving?

and since there are only two parties to any crime or dispute--the plaintiff and the defendant--

How about when a dam fails and floods multiple farms and houses? How about when a drunk driver swerves into a crowd of pedestrians?

--it seems most sensible for the legal code to declare that a decision arrived at by any two courts shall be binding.

Who writes this legal code that establishes the meta-principle that two courts judgments' shall be binding? After all, I don't know anyone who regards the judgment of other courts as binding. You could probably find two courts somewhere in the world who would agree that criticising Islam is a crime, but does anyone really believe that Christian missionaires or committed atheists would accept those rulings?
The only way a legal code can effectively declare that a decision arrived at by any two courts shall be binding is if somone is willing to use force to back up that legal code and can bring a lot of force to bear.

Richard Garner writes:

This chapter didn't make a convert from me, because I was already converted by reading David Friedman, or Bryan Caplan's Economics of Non-State Legal systems. Nevertheless, I thoroughly enjoyed the chapter and agreed with much of it.

In some ways, I think Friedman's model is more plausible: In Friedman's model, police firms go to the same court, agreed upon in advance, to resolve a dispute. Under Rothbard's, the police firm bringing the charges goes to its prefered court, whilst the defendant goes to his. If the courts reach different conclusions, which I think they are bound to, only then will the police firms, or the courts, go to a mutually agreed upon third party. In other words, Rothbard's system ends up working like Friedman's.

Under Friedman's model, courts adopt the legal codes for which their is the most consumer demand, consumers being police firms that pay competitors to adopt their prefered legal system. This has serious ethical problems, most clearly because it could mean that an unlibertarian legal code be adopted: A police firm representing somebody who wants to beat up people simply because they have read hair would charge this anti-redhead guy enough for them to pay police firms that would otherwise protect his victims to get them to agree to a court that uses a legal code that permits anti-red head violence.

Rothbard clearly thinks this should not be the case, and of course, I agree with him. It should not be the case... but that is different from saying that it will not be. I agree with Friedman that the market will supply the laws that there is most demand for. The other approach is to ask what laws a state is likely to supply, what laws there is likely to be a demand for. If there is likely to be more demand for unlibertarian laws from a state, and less demand for libertarian laws, than there would be on a market, then anarchism still wins. We should remember, though, that anarcho-capitalism is not libertarianism. A consistent libertarian should be an anarcho-capitalist, because state's cannot exist without doing what libertarians think people shouldn't do. However, a libertarian anarchist can still be critical of an unlibertarian anarcho-capitalism.

The comments on consolidation, mergers, and the size of the firm are important and correct. However, actual existing security firms tend to be small to medium sized companies, with fewer than fifty employees. Exceptions, like Blackwater, are firms that do most of their business with governments. It is not self-evident that the economies of scale are there that would lead to a conclusion that the market would be dominated by a small number of large firms. Of course, if the tendency in the industry is towards war and conflict, then size will probably pay. But for that tendency to be the case, surely it would already have to be the case that there are a small number of large firms anyway, since otherwise conquest of one firm would gain little in market share for the victor, whilst increasing opposition against the aggressor.

On Saudi Arabia, I suggest you just go back and read the chapter on property. Rothbard clearly distinguishes between things people own and things they rightfully own.

Comments for this entry have been closed
Return to top