Wednesday, July 28, 2010

Anarchy and Efficient Law

Earlier this year I gave a talk in Brazil based on an old book chapter. The central argument is that the producers of law in a system such as that I described in The Machinery of Freedom, unlike the producers of law in a conventional political system, have an incentive to generate economically efficient law—for reasons similar but not identical to the reasons why other producers in private markets have an incentive to produce the products that consumers most value. In the talk I also discuss the reasons why the law generated will fall short of being perfectly efficient, due to imperfections of that particular market.

It has now been webbed.


StrangeLoop said...

Thanks for continuing to web new media!

1. I'm glad to see you speaking at a Mises/Austrian event! It seems there isn't a strong enough coalition between Austrian School and Chicago School/neoclassical anarcho-capitalists. I like the Austrians... even though they don't believe in market failure, literally!

2. For other readers, here are the two links for the talks by your son, Patri:


There's an excellent demonstration of incentives in the Q&A!

3. What's your beef with Pareto-efficiency? Kaldor-Hicks better?

Once again, thank you. You were the one intellectual that was capable of converting me to anarcho-capitalism.

David Friedman said...

I see both Pareto and Hicks-Kaldor as ways of doing interpersonal utility comparisons while pretending not to. I prefer to follow Marshall, who doesn't pretend.

I believe I have a discussion of the issue in my Price Theory, which is webbed on my site.

StrangeLoop said...

To quote from chapter 15, "I prefer to use the Marshallian approach, which makes the interpersonal comparison explicit, instead of hiding it in the 'could be made but isn't' compensating payment. To go back to the example given earlier, a change that benefits a millionaire by $10 and costs a pauper $9 is a potential Pareto improvement, since if combined with a payment of $9.50 from the millionaire to the pauper it would benefit both. If the payment is not made, however, the change is not an actual Pareto improvement. The 'potential Paretian' approach reaches the same conclusion as the Marshallian approach and has the same faults; it simply hides them better. That is why I prefer Marshall. From here on, whenever I describe something as an improvement or an economic improvement, I am using the term in Marshall's sense unless I specifically say that I am not.

It is worth noting that although a Marshall improvement is usually not a Pareto improvement, the adoption of a general policy of 'Wherever possible, make Marshall improvements' may come very close to being a Pareto improvement. In one case, the Marshall improvement benefits me by $3 and hurts you by $2; in another it helps you by $6 and hurts me by $4; in another . . . Add up all the effects and, unless one individual or group is consistently on the losing side, everyone, or almost everyone, is likely to benefit. That is one of the arguments for such a policy and one of the reasons to believe that economic arrangements that are Marshall efficient are desirable."

StrangeLoop said...

For clarification, if you have time, how does your proposal differ from the following, "Under the Kaldor-Hicks definition of efficiency . . . a reallocation of resources is efficient if it enables the gainers to compensate the losers, whether or not they actually do so. This is equivalent to wealth maximization" (Richard Posner).

proclear said...

David, my wife thinks you look like a movie director! Any chance of a career change? :-)

StrangeLoop said...

proclear, I know I'd like to see him on the set of "Atlas Shrugged" (!

David, sorry to possibly overwhelm you with questions, but the debate over "economic efficiency" has always left this non-economist a bit befuddled, particulary since so many textbooks present the topic differently.

In an article, Edward Stringham, an Austrian anarcho-capitalist, states, "Kaldor-Hicks efficiency is an unusable standard. To implement Kaldor–Hicks-efficient policies, government would require wisdom beyond its grasp. Prices only provide evidence of past willingness to pay and cannot be used for formulating future policies. To truly figure out willingness to pay, the government would need to read minds to determine how much every single person would value every possible state of the world, a feat that can never be attained. For these reasons, Kaldor-Hicks should be rejected as a means of judging policy."

In essence, he makes a case that the "law and economics" movement, as applying notions of efficiency, suffers from the economic calculation problem (

Earlier, in a footnote, he states, "When the public-interested-judge assumption is dropped, Kaldor-Hicks becomes even less reliable."

Would you counter that private enforcement agencies and private arbitration would evolve toward "wealth-maximizing" decisions, thereby suspending worries about the calculation problem?

David Friedman said...

Strangeloop asks about Marshall v Hicks-Kaldor. Under most circumstances they give the same result, merely a different justification. For one exception, see:

David Friedman said...

Strangeloop asks about Stringham's criticism of Hicks-Kaldor. I haven't read it, but from your summary I think it misguided, for two quite different reasons:

1. It makes the best the enemy of the good. The fact that we cannot know something perfectly doesn't mean we know nothing about it. How much I was willing to pay for something yesterday isn't a perfect measure of its value to me today, but it's a pretty good measure.

2. Using efficiency to evaluate legal rules doesn't require solving the calculation problem, any more than proving the efficiency of competitive equilibrium does. You start with revealed preference, and ask, given that individuals will act rationally in their own self-interest, what legal rules will result in their making the efficient choices. Neither legislator nor judge has to know what choices are efficient.

StrangeLoop said...

1. I totally agree that Stringham, in searching for realistic foundations, criticizes the good for an impossible best. This is a problem within Austrian economics in general, in my opinion.

2. One problem he is raising is that "revealed preference" isn't being given; to quote, "This is easier said than done; Rothbard (1956, p. 225) argued the only way an observer can know if someone values a good is by seeing him actually paying for it. Choice demonstrates preference, and without it, it is not possible to tell how much a person values a good. Judges in this situation are faced with making decisions based on costs and benefits that are unknowable. As Cordato (1992, p. 95) stated, 'The arguments that were made in opposition to the Pigouvian tax assessor are equally applicable to the Coasian judge.'"

Overall, I fully agree with you, though. Essentially, I believe Austrian economists raise valid concerns about neoclassical assumptions, but--by pursuing an unrelenting realism--they lose much insight and findings that bolster the discipline and the anti-statist movement. And, of course, I find the praxeological assumptions (or "axioms") of Austrianism to be questionable.

Once again, thanks for the comments and kicking ass for anarcho-capitalism!

David Friedman said...

"Judges in this situation are faced with making decisions based on costs and benefits that are unknowable."

I think you may be misunderstanding the law and econ approach. The proposal is not that the judge, in each case, should decide what the efficient outcome is. It is rather that the judge should decide what legal rule will lead individuals, making choices on the basis of their values, to the most nearly efficient outcome.

Answering that question requires information of various sorts, but it doesn't require the judge to know the utility functions of the plaintiffs.

My Law's Order is webbed on my site, if you want a more detailed account.

Anonymous said...

What do you make of this, David:

Man Finds Ring On Ground, Throws It Away, Gets Sued By Owner

Sounds like it incentivizes the wrong behavior, at first glance anyway.

Doc Merlin said...

You know, open immigration between wealthy societies would have similar effects to competitive governance. The main problem as I see it, is overcoming the initial "hump" and costs associated with switching governments.

1.) We should therefore push for lower transaction costs to switching governments.
2.) Assuming the EU federal government doesn't gain much power, European governments should become more efficient as time goes on now that their borders are open.
3.) This can explain a lot of the success of the US.

StrangeLoop said...

With costs like language acquisition, loss of close proximity to family and friends, etc., I don't think nation-state migration and competition (using "exit" as its called") is a strong competitive push.

I like Randy Barnett's suggestion, "Some libertarians prefer a different legal structure, one that promises to work better than the structure provided by the Constitution. Such a structure would take the principles or strategies embodied in the Constitution farther than did the Framers. These principles are (1) reciprocity of power between the ruler and the ruled that is supposed to be accomplished by voting, (2) checks and balances on power that are supposed to result from federalism and separation of powers, and (3) the power of exit that is provided by free emigration and, formerly, the power of secession. These libertarians merely propose two itsy-bitsy, teeny-weenie changes to the status quo: First, end the government’s power to put its competitors out of business by force (which violates the freedom to contract of those who wish to provide and obtain such services); second, end the government’s power to confiscate its income by force (which violates the freedom from contract of those whose property is taken without their consent). Not much really."

Jonathan said...

StrangeLoop, I've personally lived and worked in six different European countries, and I've known other people with similar histories. But I agree that most people are less mobile.

Your proposal to "end the government’s power to confiscate its income by force" would cut off most of its income and effectively put it out of business, unless it adapted with implausible efficiency to raising money in different ways and operating on a much reduced income. Although I like the idea in principle, it's more of a revolution than a minor tweak, and (unfortunately) no government is likely to adopt it in the foreseeable future.

Anonymous said...

Question for you David: the system you propose has been tried in a limited way here in the United States. No doubt you are familiar with it, it is called the Federal System. The private protection agencies are called States, Municipalities, Counties and so forth. The transaction cost of switching from one agency to another is obviously quite high, but certainly not unprecedented.

Yet the result we see in such a system is that the states all tend to wax fat in the same way any political system does. Laws tend to be pushed out to one central proxy (the federal government), and unified codes tend to get defined by third parties, such as the uniform commercial code.

The states rights system used to be more diverse (consider as one good example abortion), but have ended up as one non competitive mass, with little differentiation.

Can you comment on why you think your competitive private model would produce different results?

Doc Merlin said...

'Question for you David: the system you propose has been tried in a limited way here in the United States. No doubt you are familiar with it, it is called the Federal System. The private protection agencies are called States, Municipalities, Counties and so forth. The transaction cost of switching from one agency to another is obviously quite high, but certainly not unprecedented.'

One difference is that they have a territorial monopoly over a broad range of things. For example, I can't change my fire dep without changing my police dep.

Anonymous said...

"For example, I can't change my fire dep without changing my police dep."

True, but that is also true in many other businesses. You can't use Mac Finder on Windows, you can't use an iPhone with T-Mobile (in the USA.)

StrangeLoop said...

You can't use an iPhone with Sprint, but you aren't forced to use an iPhone, either. Plus, competitors are widely available and impose no serious costs (perhaps you can't claim "I have an iPhone!").

Contrariwise, consider your "choice" for police, judges, law-making, etc. You are much more constrained. Federalism was an attempt at competition, but the monopolistic core of the State overwhelmed the "laboratories of democracy."

Let's put it this way: Apple never "taxed" (forced payment) you for an iPhone and its monthly bills.

Anonymous said...

> Plus, competitors are widely
> available

There are a lot more states than cell phone providers, and certainly a lot more municipalities.

> and impose no serious costs

Not true, there is a considerable cost changing cell phone provider. Less than moving to a different state, but the magnitude of the significance is quite different too.

> Let's put it this way: Apple n
> never "taxed" (forced payment)
> you for an iPhone and its
> monthly bills.

Sure they did. If we assume I want cell phone service, someone demands payment. If I want protection services, someone will demand payment. However, as California can testify, if you raise the protection services charge too high, people move to a different service. (Arizona here I come.)

However, of course the comparison isn't perfect. My point, and my question is simply this: the closest approximation to what David talks about has spiraled into the black hole of statism. Why would David's proposal be any different?

Consider this: we wave a magic wand and have the system David proposes. Everything works well until the protection agencies begin to cartel-ize. Since the nature of protection agencies is the socially legitimate use of violence, they use that violence to prevent any agencies form outside of the cartel. The cartel becomes the equivalent of the feds. What mechanisms would prevent this from happening.

Doc Merlin said...

'Everything works well until the protection agencies begin to cartel-ize.'

I don't see that, cartels are exceedingly rare without state enforcement of them. Also, the larger the number of players the harder it is to keep people from cheating in a cartel.

Kid said...

Of course David's system cannot possibly work.

However, if you would suggest government to me, and I was living under David's system, I would think that government could not possibly work either.

Yes, perhaps the protection agencies could cartelize, but what is stopping the army from taking over the government in our present society? Nothing!

We would have to risk the cartels - I definitely do not think that risk is better than all the risks of government.

Keep in mind that our present government is guilty of murder on a giant scale and does not seem to have any intent of stopping that.

Maybe David's system cannot possibly work, but hopefully it would not work a little bit less than the current system does not work.

StrangeLoop said...

Awesome post, Kid.

Anonymous said...

> I don't see that, cartels are exceedingly rare without state enforcement of them.

Right, I agree in principle with you. Cartels do tend to self destruct because a new player can always try to undermine the cartel to their advantage. However, cartels, when they do exist, are supported by government forces or actually are governments (such as OPEC for example.)

David's proposal is putting these agencies in a pseudo government position, specifically, granting them the socially acceptable right to use force, and so they can certainly exclude competitors using that one advantage.

This is especially so because the network of agreements he talks of are specifically agreements as to when force is legitimate. When a newbie comes along he has to negotiate those same force agreements with his competitors, and it is not obvious why they would participate.

It isn't hard to imagine the PR on this. Newbie protection services is a dangerous new organization threatening violence against the existing social order. People hiring them are just trying to gain an unfair advantage over you. We will shoot them down to protect you all, or loyal clients.

How exactly can Newbie break into the cartel -- something which is necessary to be a viable protection agency? I don't need Saudi Arabia's permission to sell you a barrel of oil. However, Newbie does need permission from existing agencies to enforce and protect its members.

Doc Merlin said...

So, your argument is:
'Defense businesses would effectively need permission from other like businesses to operate, and it isn't obvious why they would give them permission.'

This is only true if the defense orgs are monolithic. If they are not monolithic, the non-violence defense corps (insurance companies etc) have incentive to make deals with new violence corps. And that allows for an 'in' for the newbie orgs.

TGGP said...

Nick Szabo has criticized David's use of the Coase theorem. I'd like to hear David respond.

Anonymous said...

very nice to see "Machinery" posted. Webbed books are going to be much easier to read w/ the advent of the Kindle and e-readers. Be fantastic to see "Future Imperfect" in a more accessible format.

David Friedman said...

"Be fantastic to see "Future Imperfect" in a more accessible format."

What sort of format would be significantly better than the straight html which it's currently webbed in?

Anonymous said...

it looks like I probably spoke too soon. The HTML format may work better than PDF, for Kindles at least. I converted Machinery into a mobi file with Calibre, but it's a bit wonky (mostly the links throwing off formatting). That said, it's readable, just not pretty.

Still pretty new and inept at all this.

But I think Kindles and the like could be very cool for just this sort of thing. Not many people are going to read a couple hundred page on a monitor or print that much, but if they can throw it on a reader, it should work very well.

Anonymous said...


As (continental, West-) European, I don't understand why you think the car metaphor is such a slam dunk in your book and your talk

Granted, anarchocapitalism may well beat pre-1989 communism, just as pre-1989 Fords beat pre-1989 Wolgas and Ladas. But in America, this isn't the status quo your hypothetical legal system would be competing with. Instead, you are competing with a government-run industry within a mixed economy.

So in terms of your car metaphor, your pre-1989 Ford is competing, not with pre-1989 Wolgas, but with pre-1989 Renaults and Volkswagens. Both companies have been mostly privatized since, but were mostly state-controlled then. They made decent cars then, just as they do now.

Once you compare Fords to Renaults and Volkswagens, I no longer see that the Fords are obviously superior. How can you still argue that we should still go with the Ford? And if you can't, what does that tell us about the merits of market-produced law vs mixed-economy, democratic law?