Bryan Caplan  

Fraud and Punishment

My Current Outline... Glaeser on Housing Policy...
While I'm not a libertarian absolutist, there are plenty of criticisms of strict libertarian doctrine that don't hold water.  Two that have been on my mind lately:

1. Will Wilkinson doesn't see why libertarianism prohibits fraud:
Even when I was a believer in Rand/Rothbard-style libertarianism, I found the 'or' in the "no force or fraud" formulation of the non-coercion principle a bit vexing and suspect. It seems too frank an admission that fraud isn't force or aggression at all. It's another morally questionable way to get someone to do something they might not otherwise choose to do. But there are yet still other morally questionable ways to get people to do things. Why not add more 'or's?


Let me just say that because I think emotional coercion is coercion doesn't mean I think the state should try to stop it. And just because I think fraud isn't coercion doesn't mean I think the state shouldn't try to stop it.
Frankly, I don't see the problem.  If you accept the initial libertarian equation of "coercion" with non-consensual use of others' property, then the impermissibility of fraud follows.  If you offer me a Mitsubishi 5500 projector in exchange for $2000, and hand me a box of straw instead, you are using my $2000 without my consent (which was contingent, of course, on you giving me the projector).

From a slightly different angle, if you buy Will's fraud objection, what about skillful pickpocketing?  Neither involves graphic violence.  Indeed, in both cases, the victim may never be the wiser.  Yet both involve non-consensual property use.  Why would fraud be any more acceptable on purely libertarian grounds?

2. Tyler Cowen insists that not only punishment, but even requiring restitution, is contrary to libertarianism:
So let's say I steal your painting.  Yes, you do deserve your painting back.  It is yours.  But say I steal your painting and lose it or wreck it.  That should be the end of the story.  You never owned the "value of that painting."  You simply owned the physical painting.  You are not due compensation.  If you take my money as compensation for your loss, that is simply another theft.  All this talk about the "doctrine of rights forfeiture" is handwaving.  The forfeiture doctrine is a convenient utilitarian fiction (which I will partially endorse), not libertarian theory.  Rights aggressors do not, in fact, lose their own rights in turn.  Why should they?  Everyone in prison is there unjustly and yes that includes murderers.
Tyler's right that strict libertarianism has a counter-intuitive implication here.  But it's the opposite of what he says.  If a thief steals your painting and destroys it, the correct libertarian deduction is that he still owes you your painting and is morally obliged to return it.  Since it is by assumption impossible for him to fulfill his obligation, his only permissible course of action is to come to the victim, hat in hand, and ask the victim what he will accept as a substitute.  The victim isn't entitled to "the value of the painting"; he's entitled to the painting - though of course he can settle for less.  In other words, you can think of the thief as similar to someone who made a short sale, and has to buy back some stock - no matter what the price - to hold up his end of the contract.

What's counter-intuitive about this?  Well, it does imply that a victim could insist on infinite compensation because you stole his candy bar and ate it.  Pragmatically, of course, it might be in the interest of the victim to accept a "plea bargain."  But strict libertarianism still implies that he has a right to hold out for infinity. 

At this point, I would freely admit that libertarianism is not a perfect moral theory.  At least in many cases, there is a maximum amount of compensation a victim can justly demand, libertarian rights notwithstanding.  Nevertheless, to say that strict libertarianism implies that people should be allowed to get away with murder is plain wrong.

Comments and Sharing

COMMENTS (20 to date)
Michael writes:

Compensation or restitution is not a "right" in the libertarian sense. However, a libertarian society would very likely produce various social structures that would have the same effect, without any further rights violations.

For example, if a man damages another man's property and refuses to submit to arbitration and the possible payment of restitution, he is very likely to be banned from many businesses and residences. If you don't think this is likely or feasible, try cheating at a casino and then see how welcome you are at other casinos.

Even for crimes like murder, involuntary incarceration is never warranted. The harshest "punishment" a libertarian society can give is something like exile, which can be just as effective at protecting a society from unrepentant criminals as prisons-- without the expense (borne in part by the victims!) nor the injustice prisons entail.

scott clark writes:

I've had a similar issue with lumping fraud with force. Bryan describes a definite breach of contract, and would be justified in asserting his right to have the contract fulfilled or enforced, especially if the contract had specified mechanisms for enforcement. Other classes of fraud, I am no so sure about. Like Madoff just lying about the investments, or Enron and Worldcom cooking the books, these are things that don't warrant violent retribution in my opinion. If I lie to you to get you to turn over your money and I tell you you're making a fortune and everything is cool, you still chose to trust me, to deal with me. when the lie is discovered you will feel burnt, but i didn't force you to give me your money, i just told you what you wanted to hear. you should exile me for that, like michael says, don't deal with me again and tell everybody else never to deal with me but you shouldn't lock me up for that unless we both agreed that if i am lying about this, you can have be locked up for a certain amount of time and under certain conditions.

Adam writes:

In regards to fraud, what you're really talking about in your example is a contract. Where it can get sticky is where there isn't an explicit contract, such as in an advertisement. At some point and ad takes on aspects of a contract, and I would think that working out such issues would be a major job of arbitration (if not the biggest job).

I've been thinking about fraud lately, and the conclusion that I have come up with is that there's no need to mention fraud in the same breath as force. All fraud issues can be broken down to be either contract violations or not, so it's kind of irrelevant to the whole issue.

Jacob Oost writes:

It's my opinion that without God (i.e. a higher authority above all men), all rationales for government fall apart, and anarcho-capitalist strict libertarian doctrine would be the only ideology left standing.

No, I'm not an anarcho-capitalist strict libertarian. I'm an econotarian, a Republitarian, a Friedmanite, etc.

Renato Drumond writes:

In fact, when I commented Will's post, I used the same argument, but actually agreeing with Wiil:

"I would even say that not all forms of theft are agression. I also have a problem with confusing agression against someone with 'agression' against someone's property."

The problem I see here is that you can't equate 'non-consensual use of others' property' with 'initiation of force'(which I called agression). You don't need to initiate force to use other's property without their consent.

Gil writes:

I personally believe fraud is only defined when the victim/'victim' cannot reasonably ascertain the truth from fiction. In other words, if an ounce of due diligence on the 'victim's' part would have shown the deal was 'too good to be true' then no fraud has taken place. It's akin to the dimwit who buys birdseed with the intention of growing birds (even the shop owner did say so but the dimwit couldn't apparently pick up the sarcastic tone).

Jacob Oost writes:

Well, if anything, I think this discussion shows the importance of NOT letting large central authorities define fraud, but to allow more localized authorities to do so.

Bill R writes:

Maybe I disagree with the interpretation.

Once a property is homesteaded I believe Rothbard argues it is entitled to the extension of the individual rights and protection. To (ironically) paraphrase Braveheart "An attack on the King's property is an attack on the King himself".

For example: if you are playing football and another person pulls out a crowbar and whacks you over the head with it (or less dramatically punches you in the face like the Defensive MVP did in the Superbowl). Did you consent to the punch because you "loaned yourself out" for a physical violence? Whether the attack is considered "contract fraud" or an act of violence isn't distinguishable. I argue they are both violence.

Same goes for Rothbard's "Yelling fire in a theater" example. You breach the agreement in terms of physical property and person then you could be considered an aggressor. A contract where no homesteaded property (ie a trade secret) would be imo more of a reputation issue and probably have no claim for restitution b/c property wasn't exchanged.

Also if coercion doesn't apply to property can I take your car out for a spin while your sleeping and return it after a night on the town? Or how about surgeon rape while you're under the knife(does a falling tree make a sound if no one hears it?)? I'd say it's still aggression even if you don't realize it. I may not be able to recover much damages but I'd still have the right to kick your tail (i.e. prevent the theft) if I caught you in the act.

Skimming the relevant chapter in Ethics of Liberty (far superior to FANL on these matters, although it's been a while for both) I think I've represented the Rothbardian position accurately (

As for crime and proportionality Rothbard argues that the theft would actually be double plus the costs of enforcement. The first portion being your own property and the second being punishment. In other words I lose $1000 dollars to you and now you must suffer losing the maximum of $1000 to me. That's a total of $2000 for the thief but it's "even" in the sense that you are both were "equally" degraded by losing $1000. And yes Rothbard would "enslave" the thief by garnishing the wages until they're paid up.

I also don't see how this is inconsistent with individual rights...otherwise you couldn't defend yourself against a rapist/aggressor b/c you'd be violating his person in delivering your return blows. I don't buy Cowen's premise. If the argument is that libertarianism is the only consistent political ethic with human nature then only fair to say: If you act inhumanly you lose your rights to the degree of your beastliness. Since only human's have rights it is just and not "hand waiving" to assert them to the degree you are human. Rothbard, for example, is willing to apply this standard to children I believe; and I'd imagine the mentally handicapped have leniences and restrictions as well.

So yes I can punch back and degrade the aggressor to point he's degraded me (i.e. two punches to his one punch...see above). Cowen, however, is welcome to voluntarily be everyone's "Tolstoyan" (as Rothbard puts it) punching bag.

As for his "not owning the market value" argument... Caplan's response withstanding, I really can't currently think of an effective counter reply. Fair enough... I'd only have the utilitarian recourse of protecting my property better and screwing the thief over in a similar manner (i.e. tossing infinite amounts of his stuff in the ocean or in the belly of Mordor). Why I could even have a bunch of socialists gleefully join in and it'd be much more fun and effective than simple ostracism! The issues with regard to contract fraud in homesteaded property above still would hold because you've explicitly agreed to a "monetary value" in exchange for that specific property.

All in all I think we could all (myself included) benefit from giving the most straight forward natural law libertarian theorist, Murray Rothbard, and particularly his Ethics of Liberty a closer reading.

Jon writes:

I'm not sure that you've quite grasped Wilkinson's point. I think the argument is that the use of force or fraud to obtain another's property is illegitimate because the exchange is non-consensual. Consent in law is only valid where it is informed and freely given, hence the apparent consent of the victim in the typical fraud is invalid. So what about other varieties of invalid consent, e.g. consent given under duress? Let's say your employer offers to buy your car from you at half its market value, with the implied threat that he will fire you if you don't sell -- is this OK under libertarian principles? If it is OK, it seems that libertarianism is blind to real-world power structures in failing to recognise that such a threat is just as much an abuse of power as the school bully's use of his fists. If it's not OK, then haven't we opened the door to full-blown liberalism? (This is not my opinion, btw, but my take on Wilkinson's opinion).

As to Cowen: I agree with other commenters in that his premises aren't those of any version of libertarianism I've ever seen.

RobbL writes:


". It's akin to the dimwit who buys birdseed with the intention of growing birds (even the shop owner did say so but the dimwit couldn't apparently pick up the sarcastic tone)."

So your version of libertarianism says that selling magic beans to the gullible is A-OK? How about selling stuff to the mentally retarded? How about aluminum siding to the elderly?

Oh...alright, I can't stand it; How about ALT-As to a banker?

Lech Wilkiewicz writes:


The difference between the boss and the bully in your scenario is that your body is yours and the bully has no right to do anything to you without permission. You do not have a right to a job, however; the employer own's the wages he pays you (until you have earned them), the factory, etc. So, even if it makes him a bad guy to push his weight around and try to get you to sell your car, it is not coercion in the sense described by Bryan.

Nathanael Snow writes:

"At this point, I would freely admit that libertarianism is not a perfect moral theory."
Why not? Because it requires infinite compensation? Because there is no implicit compromise?
These are challenges to the practicality of Libertarianism, but not to the theory. The theory works fine: do not encroach on others or their property, and do all that you promise to do. If you don't keep these, you are history, outlawed, and subject to whatever consequence the wronged individual chooses to impose. Such harsh rules would discourage all encroachment and breach of contract. They would also be very expensive, and inefficient.
The theory works fine. It just doesn't jive with our empathy, or our desire for efficiency.

Jon writes:


Bryan's sense of coercion is the "non-consensual use of other's property". The common law has long recognised that consent not freely given is no consent at all -- that's why, for example, contracts entered into under duress are voidable. So either the employer's behaviour is coercive, or someone needs to devise a suitable redefinition of "consent".

Will Wilkinson writes:



First, libertarianism is not Rothbardism, and it's confusing to encourage this confusion.

You start...

"If you accept the initial libertarian equation of "coercion" with non-consensual use of others' property..."

But argument is precisely that this identification is an arbitrary, ideological reinvention of the meaning of a natural language term. The claim is that "non-consensual use of property" is not coextensive with "coercion" or "aggression". And it's obvious that it isn't to anyone who hasn't undergone Rand/Rothbard reconditioning. One can accept a ban on the non-consensual use of legitimate property, and still reject as both too broad and too narrow a ban on the initiation of coercion/aggression, ordinarily understood.

Then we are left with two questions. One is about the kinds of actions that count as genuine consent. The other is about the bounds of legitimate property. I claim that good answers to those questions leaves the "no non-consensual use of (legitimate)property" rule miles away from the traditional understanding of "the principle of nonagression." One has to define terms tendentiously at every turn to produce anything approximating Rothbardism.

"Reconditioning" is a very strong word, Mr. Wilkinson, let me just say that. "Coercion" is a typically Hayekian term, and as characteristically Hayekian, a poor/imprecise one. Coercion is nothing that us libertarians oppose. After all, justice requires some use of force (criminals don't turn themselves in in most cases, you will agree). But the use of force to transform/invade others and their property, when they have not violated any rights themselves, is considered aggression.

There is a space of course for Mengerian institutions arising, since both culture and technology are the context where universal principles are to be applied. In other words, the "no non-consensual use of (legitimate)property", when property is understood as in a *causal* and not only accidental relation to individuals, leads directly to the "principle of nonaggression." It is not enough to reject legislation, but also it is necessary to reject purely custom-based law systems too. Natural law, based on human nature that is, provides the analytical core that judges will apply according to cultural and technological context.

One has to define terms very precisely and rigorously, to produce anything approximating Rothbardianism, Aristotelianism or Randianism. The "problem" Hayek's followers encounter when facing Rand or Rothbard followers, is that the latter acknowledge humans as beings capable of justice, that work in a cause-effect framework, etc. That leaves less space for fuzzy relativism, of course.

Inquisitor writes:

"First, libertarianism is not Rothbardism, and it's confusing to encourage this confusion. "

Ipse dixit. Rothbard is Mr Libertarian. Seems the Hayekians just can't get over it.

"One can accept a ban on the non-consensual use of legitimate property, and still reject as both too broad and too narrow a ban on the initiation of coercion/aggression, ordinarily understood."

One would be in error...

"Let's say your employer offers to buy your car from you at half its market value, with the implied threat that he will fire you if you don't sell -- is this OK under libertarian principles? If it is OK, it seems that libertarianism is blind to real-world power structures in failing to recognise that such a threat is just as much an abuse of power as the school bully's use of his fists. If it's not OK, then haven't we opened the door to full-blown liberalism? (This is not my opinion, btw, but my take on Wilkinson's opinion)."

You have no "right" to a job, so the answer is to the extent that it's "unfair", you may boycott him, inform others he's a prick etc. But he has every right to fire you, and sorry, but that isn't "coercion". Deal with it.

Inquisitor writes:

BTW, I didn't actually see an answer to Caplan's point, namely that why on Rothbardian grounds fraud is illicit...

Inquisitor writes:

Fraud - exchange on grounds the defrauded party would not consent to = theft - retroactive slavery. Yes, painstaking deduction flowing from proper definitions. If it's counter-intuitive, I again suggest people get over it.

Regarding the contention that libertarianism doesn't prohibit fraud.

This was argued years ago by non-libertarian James Child. In this post, The Problem with "Fraud": Fraud, Threat, and Contract Breach as Types of Aggression, the "Fraud" section, I explain what is wrong with standard libertarian views on fraud (and with the criticisms of libertarianism in this regard, based in part on loose talk about "fraud). The problem is that people use fraud to mean basically "dishonesty," and in this usage it's hard to see why it's a type of aggression. If one has a coherent understanding of the nature of contract (a title-transfer theory along the Evers-Rothbard line) and property rights, then one has to understand fraud as some kind of misrepresentation that vitiates the consent needed for a title transfer to be effective. But I go into this in greater detail in that post, and in the articles linked therein.

(Regarding the term "coercion," I think this too is a term misused by libertarians--coercion is not a synonym for or even a subset of aggression; it's a type of force, and force can be justified or not. SEe my post "Coercion" is annoying, but coercion is neutral.)

Regarding the contention of Cowen "that not only punishment, but even requiring restitution, is contrary to libertarianism", because it's impossible to get restitution in some cases. Thus, punishment is unlibertarian, since it does not make the victim whole--so the argument goes.

The problem here is a common libertarian mistake of making restitution the goal of justice. Then someone like Cowen quite rightly points out that restitution really means making someone whole, but that this is a utopian, unattainable goal; and therefore, there's nothing to be done in such cases.

The mistake is in thinking restitution is primary and punishment is only secondary, or even impermissible--that the only force that is permissible is that used to enforce some kind of restitution, or perhaps as some kind of extended self-defense (putting down a standing threat, as Randy Barnett argues (see pp. 80- and n. 11, in the section "Standing Threats," in my Inalienability and Punishment); or as some kind of incapacitation)--but never pure "retribution" or punishment.

As I have laid out in detail elsewhere (Punishment and Proportionality: The Estoppel Approach; also Inalienability and Punishment; Defending Argumentation Ethics; and New Rationalist Directions in Libertarian Rights Theory), I believe the proper approach is to realize that justice is about giving someone their due, and what a victim is due is being allowed to respond in kind to the aggressor, within the limits of proportionality.

That is, libertarianism opposes aggression, the initiation of force. But it does not hold that the opposite of aggression is unjustified. The opposite of aggression--the initiation of force--is not "defensive" force, or "force used to enforce restitution"--but rather, "responsive" force--force in response to aggression. Force is thus either initiated, or it is in response to intiated force. "Responsive force" is justified; aggression is not.

Responsive force may also be referred to as retaliation, even punishment or retribution, but the latter concepts are probably best viewed as a type of responsive force, or one possible purpose of responsive force. As I note in Inalienability and Punishment (see the section "The Right of Proportional Punishment," at p. 84; see also Punishment and Proportionality),

an individual has a right to use force against an aggressor in response to aggression. This right to use force can be utilized for a variety of purposes: for self-defense during or before the act of aggression, for revenge, to obtain restitution, to prevent the aggressor from committing further crimes, or to deter others from committing crimes. What the victim wants to use the right for is his business. But the reason why a victim has a right to retaliate or defend against an aggressor is that the aggressor cannot sensibly withhold his consent to retaliatory, defensive, or restitutive force (these may be considered different types of responsive force, that is, non-initiated force, force which is in response to initiated force). To use related legal terminology, the aggressor is "estopped," or precluded, from denying the victim’s right to use (proportional) responsive force, since such a denial would contradict the aggressor’s view that the use of force is permissible (the view demonstrated by the act of aggression).

In other words, it is retaliation--the right to respond with proportionate force against the aggressor--that is the primary right the victim has under libertarian justice. Restitution is then seen not as some utopian, unattainable goal of making the victim whole (which is impossibe), but simply the ransom paid by the aggressor pursuant to negotiation backed by the victim's threat of imposing the rightful amount of responsive force he is entitled to impose.

Thus, in the example given about the tolen and destroyed painting, the victim has the right to do something similar to the aggressor--take the aggressor's property and destroy it (or not--up to the victim). This does not rest on any fallacious notion that there are property rights in value (there are not, as Hoppe shows--see here and here). But there is no reason to take into account the consequences to the victim, that are a result of an act of trespass (aggression), when determining the proportionality of the response.

Finally, let me note that just because the right to forcefully respond, including use of force for not only defensive or restitutive purposes, but also for retaliation or retribution, does not mean that we would expect a libertarian society to actually employ punishment often in practice. As I argued in Knowledge, Calculation, Conflict, and Law, a review essay of Randy Barnett's The Structure of Liberty, "It is ... more costly to seek punishment than to seek restitution. For this and other reasons, restitution would probably become the predominant mode of justice in a free society."

However, as I explain there and elaborate in Inalienability and Punishment and Punishment and Proportionality,

Nevertheless, acknowledging (and justifying) the theoretical legitimacy of punishment can be useful. For example, punishment (or a theory of punishment) may be utilized to reach a more objective determination of the proper amount of restitution, because a serious aggression leads to the right to inflict more severe punishment on the aggressor, which would thus tend to be traded for a higher average amount of ransom or restitution than for comparatively minor crimes. Especially offended victims will tend to bargain for a higher ransom; and richer aggressors will tend to be willing to pay more ransom to avoid the punishment the victim has a right to inflict, thereby solving the so-called “millionaire” problem faced under a pure restitution system (where a rich man may commit crimes with impunity, since he can simply pay easily-affordable restitution after committing the crime).

Moreover, even if punishment is banned (de facto or de jure) and is not an actual option—because of the possibility of mistakenly punishing innocents, say--an award of restitution can be based on the model of punishment. To-wit: a jury could be instructed to award the victim an amount of money it believes he could bargain for, given all the circumstances, if he could threaten to proportionately punish the aggressor. This can lead to more just and objective restitution awards than would result if the jury is simply told to award the amount of damages it "feels" is "fair."

This latter point is significant because, as noted above, restitution based on the idea of restoring the victim is, as Cowen notes, often impossible, so meaningless; for this reason, those advocating restitution usually are vague about the proper standard (since there is no proper standard), or just "punt" it to the juries or courts, much like Congress does when it uses vague terms in statutes or Constitutions such as "accommodate" in the Americans with Disabilities Act or "privileges or immunities" in the Fourteenth Amendment.

One final note. To claim that a murderer who is imprisoned is there unjustly is a confusion. It does not violate his rights, as I argued in Inalienability and Punishment. It is unjust for other reasons--for one, it's done by the state, which is inherently criminal and unjust; for another, it's paid for by tax dollars stolen from citizens; for another, it violates the rights of the victim by depriving them of either personal vengeance or a type of restitution.

Robert Capozzi writes:

Interesting banter on this thread.

Part of the challenge is that some believe there is such a thing as "libertarian theory." Obviously, there is no such thing. Instead, there are a set of ideas about a virtuous society and social order that can be labeled in the "libertarian" camp. Attempting to build up a construct is an interesting exercise, but this thread illustrates the ultimate folly of such an exercise.

Whether "fraud" is "coercive" or the "initiation of force" or something else seems beside the point. I suggest: Test the opposite, and see if it's true. If a social order allows for wholesale fraud in all human economic interaction, does that point to a healthy, functional civil society?

I'd like the hear the case for "yes," because I can't imagine one. So let's stipulate "no."

Law is like price discovery. It evolves. Law should (and does!) evolve, ever imperfect, but -- it seems -- it's the best we can do, given the circumstances.

Dissuading fraud seems like an excellent objective for the law, all things considered.

Comments for this entry have been closed
Return to top