Libertarian Liability Law

by Roy Halliday

The Problem

Libertarians do not agree on rules for determining the suitable punishment or compensation due for crimes and torts. Murray Rothbard advocated that criminals be punished in proportion to their crimes and be forced to compensate their victims for damages.1 Libertarian legal scholar Randy Barnett opposes punishment but insists that criminals be forced to compensate their victims, even if doing so requires putting them into work prisons.2 Also, according to Rothbard and Barnett, noncriminal debtors and tortfeasors should pay what they own, but they should not be imprisoned or forced to work. Christian libertarians, in so far as they follow the teachings of Jesus, are opposed to coercive punishment and compulsory compensation.3 In their view, we should forgive our debtors and those who trespass against us.4

Legal scholars in tune with current American law believe criminal cases should be treated differently from torts, so we have criminal law and tort law (civil law) with different procedures and different remedies. In criminal cases, the state is the prosecutor and defendants may be imprisoned before, during, and after their trials and, in some jurisdictions, those found guilty may be subject to fines, prison sentences, or corporeal punishment (even death). In civil cases, private citizens initiate the law suits and the defendants are usually not arrested or detained, and if they are found guilty they have to compensate the victims, but they do not have to suffer imprisonment or corporeal punishment. Because defendants in criminal cases stand to be punished more severely, the rules of evidence and the standards for establishing guilt are more strict in criminal cases than in civil cases.5

Libertarians who advocate retribution have a different view of rights from libertarians who oppose it. Retributionists generally believe that criminals lose some of their rights by committing crimes and that the worse the crime the more rights a criminal gives up. Libertarians who oppose retribution generally deny that criminals lose their rights and, therefore, to punish a criminal against his will is itself a crime. Then there are the utilitarian libertarians who dismiss natural rights completely and evaluate punishment and compensation schemes based on the social costs and social benefits. I see no way to reconcile these conflicting positions, which start from different fundamental beliefs.

Social utilitarians, who do not believe in individual rights, can justify punishments that deter crime and protect the public, regardless of whether the person being punished is guilty. Utilitarian calculations can lead us to punish some minor offenders more severely than some major ones. In contrast, libertarians who believe in individual rights, including those who believe in retribution, all agree that it is wrong to punish innocent people, and they believe that even if punishment serves other purposes such as deterrence of crime, protection of the public, or rehabilitation of criminals, the right to punish criminals and the extent of the punishment due in a particular case must be derived from a theory of justice that distinguishes between the rights of criminals and noncriminals.

Not only are libertarians divided into different categories such as retributionists, Christians, and social utilitarians based on their fundamental beliefs about individual rights, libertarians within these broad categories do not all agree. Some retributionists advocate corporeal punishment, some advocate fines, some advocate imprisonment. Within these subgroups people disagree on the appropriate kinds of corporeal punishment, the amount of the fines, and the length of the prison sentences. Those who follow the teachings of Jesus must forgive all criminals and tortfeasors and not punish or demand compensation from them. Only the social utilitarians offer an “objective” way to determine the appropriate kind of punishment or the amount of punishment. Namely, try different schemes and see which ones work best. Unfortunately, social utilitarians do not know how to determine what works best other than by relying on democratic consensus, which is no more than an aggregate of subjective opinions.

How could a new libertarian nation reconcile these differences about the morality of imposing punishment on criminals and using force against tortfeasors to make them compensate their victims?

The Simple Solution

The easiest way to solve the problem is to establish the libertarian nation as a single proprietary community and to require everyone who wants to live there to sign a lease that includes the rules for assessing and enforcing criminal and civil liability. People who disagree with the rules would not sign the lease and would not be permitted to stay in the country. Then it wouldn’t matter whether the rules were based on Islamic law, Jewish law, Roman Catholic law, common law, Napoleanic law, utilitarian law, or any other system of law, because whatever rules the lease defined would have been agreed to voluntarily by all parties in advance. Children born in the proprietary nation would be the responsibility of their parents until they claim their independence, at which time they would either have to agree to abide by the law specified in the lease or leave the country.

A single-owner proprietary nation would be difficult to establish, but once it is in operation it could easily solve the problem of conflicting theories of law by requiring everyone to abide by one code, and, unlike a state that imposes laws on its subjects without their consent, the code specified by the proprietary nation would be accepted voluntarily by all residents.

The Complex Solution

If the libertarian nation contains more than one land owner and the land owners do not believe in the same rules for assessing and enforcing criminal and civil liability, the nation will have competing systems of law. Can such a nation survive?

In “The Anticrime Industry in a Free Nation” (Formulations Vol IV, No. 1) I argued that a stateless nation that allowed private citizens and businesses to punish criminals as they see fit would be unstable because (1) private defense firms could not know whether their clients have ever committed crimes for which they deserve to be punished, and a privately administered punishment could look exactly like a criminal attack, so a defense firm would risk losing its own rights if it used force to defend a client from an apparent attacker who was actually administering allowable retribution; (2) some classes of people such as pretty girls would be singled out for public punishment because it would be more profitable; (3) punishments for equal crimes would be unequal and unpredictable; (4) all of this would be unacceptably offensive to many people and would cause some of them to demand that a state be created to impose a more uniform, orderly, and predictable penal code across the nation.

Hans-Hermann Hoppe has described an anarcho-capitalist system of justice that allows competing systems of punishment while satisfying my objections.6 Although Hoppe follows Rothbard in believing in proportional punishment, Hoppe does not require everyone in his ideal nation to submit to his brand of justice. In Hoppe’s system, instead of having a uniform code of law imposed on everyone by the state, competing insurance firms would offer defense services tailored to customers who agree to specific rules of conduct, such as not to commit suicide and not to provoke attacks, and specific standards for such things as rules of evidence and assignment of awards and punishments. Each citizen would be free to make a defense contract with a specialized insurance firm that enforces the legal code of his choice. Muslims might choose to be judged by Islamic law, Catholics might choose Canon law, Jews might choose Mosaic law, and others could choose from a variety of secular legal codes. Cases involving parties who subscribe to the same legal code would be settled according to the laws of that code.

Hoppe gives reasons for believing that in a libertarian nation people would naturally choose to segregate themselves by living among and interacting mainly with like-minded people, so that in many cases both parties to a dispute would subscribe to the same legal code. Nonetheless, conflicts will sometimes arise between people who subscribe to different legal codes, and in some of these cases the competing legal codes will arrive at different conclusions regarding the case in question.

We should not expect either side to simply give in and allow the case to be decided by the “foreign” system of law. Instead, insurance contracts would include provisions for settling disputes between groups, and:

because the intergroup law provisions that an insurer offered to its clients could appear creditable to them, and hence a good, only if and insofar as the same provisions were also accepted by other insurers (and the more of them, the better), competition would promote the development and refinement of a body of law that incorporated the widest—intergroup, cross-cultural, etc.—legal-moral consensus and agreement and thus represented the greatest common denominator among various competing law codes.”7
Hoppe goes on to argue, persuasively, that the incentives inherent in an unhampered insurance market would result in more efficient protection against damages and swifter and more equitable settlement of disputes.

Maybe Hoppe is right. Maybe a nation consisting of people with conflicting beliefs about criminal and civil liability could work as he describes. Maybe the number of libertarians it takes to establish a regime of justice is the same as the number of libertarians it takes to change a light bulb—None. The market will take care of it.

On the other hand, the free market involves voluntary transactions only. Insofar as punishment and compensation are voluntary, they violate no one’s rights and the market can provide them efficiently. But if we impose punishment on someone or force him to pay compensation against his will, we are not engaging in a voluntary transaction, which means we are acting outside the free market. There can be no such thing as a free market for involuntary punishment or involuntary compensation. A slave market is not a free market.

Hoppe’s solution is a free-market solution only insofar as the fines and penalties involved are agreed to at the time by all the parties involved, especially the parties who have to pay. Suppose I sign a contract with an insurance company that administers Islamic law because I think it will give me good protection. Now suppose I run afoul of Islamic law and am convicted and sentenced to corporeal punishment. If I suddenly announce that I am no longer a Muslim and I no longer consent to the Islamic system of justice, would enforcing my sentence still be a free-market transaction? Would I be an aggressor if I resisted having my hands chopped off after I had signed a contract in which I agreed to it?

If Hoppe’s system doesn’t allow people to withdraw their consent to their sentences at the last minute, it is not truly a consensual, free-market system of justice. If it is a free-market system that allows people to change their minds at the last minute, would it still work? The rest of this paper presents arguments for a voluntary system of liability law based on the nonaggression principle—libertarian liability law.8

Questions for Liability Theory

The major questions for liability theory are: Who decides that a tort or crime has been committed by a particular person? How do they make this determination? Who determines the appropriate sentence? How do they determine the appropriate sentence? Who enforces the sentence? How do they get the right to enforce the sentence? What do the enforcers have the right to do to those who resist?

If a specific amount of punishment or compensation is proper, wouldn’t it be wrong to inflict more than that amount? If we cannot determine the proper amount of punishment or compensation for an offense, wouldn’t we would risk becoming criminals if we impose any punishment or compensation at all?


Retributionists regard punishment, like justice, as an end in itself. In fact, they regard punishment as a requirement of justice. Retribution is the philosophy that the offender deserves to suffer. The physical harm inflicted on the offender or the damage done to his property is not a means to some other end such as compensation, deterrence, or reform. Punishment itself is the end.

The great moral philosopher Immanual Kant believed that murderers ought to be executed and that it would be wrong not to execute them, regardless of the circumstances:

Even if a civil society were to dissolve itself by common agreement of all its members (for example, if the people inhabiting an island decided to separate and disperse themselves around the world), the last murderer remaining in prison must first be executed, so that everyone will duly receive what his actions are worth and so that the bloodguilt thereof will not be fixed on the people because they failed to insist on carrying out the punishment; for if they fail to do so, they may be regarded as accomplices in this public violation of legal justice."9
If justice means giving everybody what they deserve, and if offenders deserve retribution, then it is unjust to fail to punish them. If an offender isn't caught until 50 years after his offense, the mere passage of time has no bearing on what he deserves, so he should be punished mercilessly even if he has led a good and productive life in the mean time and has a family who depend on him for support.

Retribution theories generally require a correspondence between a particular offense and a particular punishment. If an offender deserves penalty X for offense Y, he deserves penalty X for offense Y each time he commits the offense. Punishment for retribution must be carefully measured out in exact amounts that correspond to the offense. To punish too little or too much is to be unjust.

Corporeal punishment, imprisonment, or even confiscation of property without consent would normally be regarded by libertarians as a crime. Retribution theories make exceptions to the libertarian definition of crime by postulating that criminals lose their right to not be punished, so it is not an act of aggression (in the libertarian sense of the word) to impose punishment on them without their consent. However, retribution theories open up an array of moral dilemmas that have never been answered logically and objectively. If I say a thief deserves twenty lashes and you say he deserves thirty days in jail., how can either of us prove the other wrong? We don’t even agree on a unit of measurement much less a method for relating units of punishment to specific crimes. Who has the right to decide what the punishment should be? When does punishment become excessive? Do onlookers have the right to intervene if they think the punishment is excessive?


Tort law can be somewhat more objective than penal law because almost everyone agrees that the goal of tort law is to compensate victims for their losses, and the relationship between a loss and a suitable compensation is generally more clear than the relationship between a particular offense and a particular penalty. Compensation presupposes a victim. It cannot be used to justify punishment of an offender who violates a statute when there is no person who is victimized by the infraction. Consequently, compensation has a strong appeal to the libertarian sense of fairness. If we agree that the offender insulted or trespassed on the rights of the offended party, then we naturally sympathize with the victim, and we would like to see the violation undone somehow. Compensation is like a debt that the offender owes to the offended party. For minor offenses, the offender may merely owe an apology. For serious offenses that are more likely to be litigated, the offender may owe a significant amount of money or services.

Compensation can be a substitute for repossession when repossession is impossible, or it can be a supplement to repossession when repossession fails to restore the repossessor to the condition he was in before the offense. For example, if A borrows B's car, has an accident that damages it, and then returns the damaged car, B is likely to feel that A owes him compensation. The purpose of compensation is to restore the condition of the offended party, not to hurt the offender. Usually a sentence will consist of monetary debts owed by the tortfeasor to his victims.

The idea of compensation is to end or relieve the hardship and suffering that the victim is experiencing as a result of the offense. This can be accomplished legitimately through an insurance policy, through voluntary donations from relatives, friends, and neighbors, or through voluntary atonement by the offender. The focus of compensation is on improving the well-being of the victim of an offense. However, when compensation is extracted by force, it is self-defeating, because it creates another victim. When compensation is extracted by force from the public, the taxpayers are robbed, and when it is extracted by force from the offender, the offender and his family become victims.

Compensation Instead of Retribution

Randy Barnett does not trust the natural urges that make us want to see criminals punished, but he has complete trust in the sense of fairness that makes us want to force criminals to make restitution. So in his legal system compensation supersedes punishment. Although he does not question our right to coerce criminals to make them pay restitution, he does worry about convicting innocent people by mistake. In the end he decides that requiring criminals to compensate their victims is better than punishing criminals and better than not forcing criminals to make restitution, even though some innocent people will be convicted in error and wrongly forced to pay:

Surely a rule—'no restitution for injustice'—resulting in the certainty of injustice to every innocent victim is inferior to a rule creating only a chance of an injustice [to] an innocent accused.10
... the only alternative to imposing this risk [of extracting compensation from innocent people because of error], is to guarantee that every innocent victim of crime will suffer an injustice.11
... to eliminate all restitution would be to guarantee that injustice will be done in every criminal case.12

Barnett’s argument assumes a false dichotomy between no restitution and imposed restitution. He leaves out voluntary options. In some cases stolen goods can be recovered and returned to their rightful owners without imposing any risk on innocent victims. In some cases the criminal will voluntarily atone for his crime (social pressure, ostracism, boycott, shame, and conscience can all play a part in this). And in some cases the victim will have an insurance policy that will require his insurance company to compensate him for his loss.13

Liability for Accidents

Suppose Mrs. O'Leary’s cow knocks over a lantern that starts a fire that burns down Miss Ember's house. In fairness, Mrs. O'Leary should replace Miss Ember's house or compensate Miss Ember in some way. Who decides what the compensation should be? If Mrs. O'Leary and Miss Ember agree on what the compensation should be, does anyone else have the right to demand that Mrs. O'Leary pay more compensation? or less? If Mrs. O'Leary and Miss Ember do not reach agreement on compensation, does someone else then acquire the right to decide what the compensation should be? How could someone else acquire this right unless Mrs. O'Leary and Miss Ember delegate it to them? What if Mrs. O'Leary and Miss Ember do not delegate the right to decide? Does Miss Ember have the right to determine by herself what the compensation should be and to take it by brute force? Does Mrs. O'Leary have the right to defend her property from Miss Ember?

It is hard to calculate the market value of Miss Ember's house. It is immeasurably more difficult to calculate the sentimental value of Miss Ember's house. Subjective values cannot be measured or compared interpersonally. Intangible losses are real, but they can't be measured and compensated for with the exactitude that justice requires. Miss Ember's evaluation of her home may be much greater than (or less than) its market value or its insured value. How can Miss Ember possibly be compensated, except by a miraculous restoration of her house from the ashes?

If we concede that the accidental rights violator (Mrs. O'Leary) does not lose all her property rights, then it is essential that we be able to determine exactly what rights she retains. In cases that are not clear-cut, and, unfortunately, this is a high percentage of the cases, we are in danger of violating the culprit's rights if we coerce her into compensating the victim. On the other hand, if we defend the culprit's right to resist paying what she believes is excessive compensation, then we run the risk of violating the victim's right to just compensation.

The only safe course of action (in the sense of being sure that you violate no one's rights) is to not interfere in the dispute. But, if outsiders cannot safely interfere, Miss Ember cannot subrogate her right to compensation, and institutions cannot garnish wages or set up employment programs, or in any other way coerce Mrs. O'Leary to make restitution. Also, if Mrs. O'Leary owes compensation to Miss Ember or anyone else, then Mrs. O'Leary cannot subrogate his right to self-defense, because it is not clear what property Mrs. O'Leary or her agent has the right to defend.

Suppose we have a system for objectively calculating compensations and we apply it to the case of Mrs. O'Leary versus Miss Ember, but Mrs. O'Leary does not have enough money or resources to pay the full compensation. Does the process of calculating compensation somehow transfer some of Mrs. O'Leary's future earnings to Miss Ember? How can title to things that do not even exist be transferred now?

Does the accident by which Mrs. O'Leary destroyed Miss Ember's house, in itself, transfer any property titles from Mrs. O'Leary to Miss Ember? If Mrs. O'Leary had an insurance policy that covered this situation, then title to $X might be transferred from Mrs. O'Leary's insurance company to Miss Ember as the result of the house burning. Similarly, if Miss Ember had an appropriate insurance policy, title to $Y might be transferred from Miss Ember's insurance company to Miss Ember as a result of the house burning. The terms of the specific insurance policies determine what property titles are transferred when Miss Ember's house burns. The market value of the house, or its sentimental or historic value, or its value in God's eyes, may have little relation to the insured value ($X or $Y). The burning of the house is the occasion for the transfer of $X or $Y because of the terms of Mrs. O'Leary's or Miss Ember's insurance policy. In the absence of such insurance policies and any private bets that may have been made about the burning of Miss Ember's house, no property title transfers are occasioned by the burning of Miss Ember's house.

Liability for Fraud

Fraud can be used to achieve results similar to theft. The legitimate property of one person can come under the unjust control of another by fraud or by theft. Neither fraud nor theft can entitle its user to property obtained by these means. The ownership remains the same as it was before the fraud or theft.

If a thief steals your car, the car remains your property. He has no right to use or sell the car. If the thief sells the car to a car dealer who knows the car is stolen, the car dealer is a thief too. If the car dealer sells the car to a customer who does not realize that the car is stolen, the car dealer commits fraud. Because no rights can be lost or obtained by fraud, the car dealer has no right to the money that he obtained fraudulently from the customer, and the victim of the fraud has no right to the car. If the customer then sells the car without realizing that it is stolen property, this is a mistake due to misinformation, rather than a dishonest act. Nevertheless, the sale is not legitimate.

The thief, his accomplices, and the defrauder are all aware of what they are doing. They are not likely to return the stolen property to its rightful owner voluntarily. It would, therefore, be legitimate to forcibly seize the stolen property from them and return it to the owner. No compensation is due to them.

Someone who deals in stolen goods unwittingly should be informed of the fact and given the opportunity to return the property voluntarily. If he refuses to do so, then it is legitimate to use brute force against him to return the property to its rightful owner. The victim of fraud retains title to the money he paid for the stolen property. He has the right to use brute force to take the money back. Consequently, there is a moral basis for insurance against fraud similar to insurance against theft.

Liability for Crimes

Even in simple criminal cases it is not clear what the correct compensation is. Suppose Mr. Hijack steals Mr. Worthy's car. The car still is Mr. Worthy's property, so he can take it back from Mr. Hijack or he can hire someone to do this for him. If Mr. Hijack refuses to give up the car peacefully, it is legitimate to use brute force against him to repossess it. This much restitution is allowed by the right to self-defense against invasion, which is consistent with the nonaggression principle. Can any more compensation legitimately be demanded by brute force?

Suppose Mr. Worthy hires someone to repossess his car. Does Mr. Worthy have the right to coerce Mr. Hijack to pay the repossession costs? If the case comes to trial, and Mr. Worthy proves that Mr. Hijack stole his car, can the court demand any more than that Mr. Hijack return the car to Mr. Worthy? The answer depends on the policy of the court that Mr. Hijack and Mr. Worthy agree to use.

Suppose we deny that Mr. Hijack can be compelled to do any more than return the car. Then, if Mr. Worthy hires someone to repossess the car, Mr. Worthy would have to bear the costs. If Mr. Worthy has auto-theft insurance, his insurance company might take action to repossess the car and provide Mr. Worthy with a substitute car in the interim at no additional cost to Mr. Worthy. If Mr. Worthy's insurance company operated in this way, it would include these kinds of costs as part of its operating expenses and take them into account when establishing the prices for its policies. Mr. Hijack's theft would have no significant impact on Mr. Worthy's insurance company or on Mr. Worthy's insurance premiums. So, it is possible for the market to provide some restitution beyond mere repossession, without having to coerce criminals more than is clearly warranted.

Similarly, insurance policies could provide the same kind of voluntary compensation for other aspects of crime. Insurance policies could pay the court costs and pay the victim for his inconvenience.

Suppose we have a foolproof system for objectively calculating the exact amount of compensation commensurate with a particular crime, but the culprit is unwilling to pay it. How does the victim or his agent get the right to take the compensation by brute force? Does the process of calculating compensation somehow transfer property titles from the culprit to the victim without requiring the culprit's consent?

The right to repossess stolen goods follows from the entitlement to one's own property. Coercive punishment, on the other hand, is outlawed by the nonaggression principle. This leads to apparent paradoxes. What is to be done with murderers? Are they to go free? Isn't there an imbalance in this scheme that allows murderers to pay nothing while thieves are required to return what they have stolen?

If someone destroys your property without your permission, it is reasonable to seek compensation from him. If a criminal steals your car and wrecks it, a reasonable compensation might be a car equivalent in market value to the stolen car before it was wrecked, or enough money to buy such a car. Such compensation would help to restore your wealth to what it was before the theft. It would be unreasonable to demand that the thief's car, if he has one, be wrecked in order to establish a balance. Wrecking the thief's car does not restore the situation to what it was before the crime. It only makes the thief's situation worse than it was. Such punishment cannot be justified as compensation or restitution, it can only be justified by a theory of retribution or by social utilitarianism that overrides or dismisses the nonaggression principle.

Suppose one man destroys another man's eye. Suppose also that a doctor could replace the destroyed eye with a transplanted human eye. It could reasonably be decided that the culprit ought to provide a suitable eye for the transplant operation as compensation to the victim. In such a case, the culprit might even donate his own eye if it was of the right type. If so, it would be fair to give an eye for an eye. But this would differ from the Old Testament idea, because, in this case, the culprit's eye is not destroyed to punish him. Instead, the culprit's eye is preserved and transplanted to compensate the victim. It is immaterial whether the culprit donates his own eye or arranges for someone else to voluntarily donate one.

Suppose medical science cannot restore sight to the victim of the eye gouger. Then the culprit could not make full restitution. Nonetheless, procedures could be developed whereby impartial people could recommend some other form of compensation. Not everything has to be stipulated in advance by a signed document. Social customs can play a role in situations where it is reasonable to assume that the custom was accepted by the parties involved. Those who make recommendations for compensations should be guided by the principle that the purpose of compensation is to restore the victim's property rights, as much as possible, to what they were before the crime or accident. In cases like this, only simulated compensation is possible. If the culprit has enough money or credit, he should pay for any medical treatment resulting from his actions and any loss of income that can be directly attributed to him. However, the compensation must be paid voluntarily, because involuntary compensation violates the rights of the involuntary compensator.

In the case of murder or manslaughter, the property destroyed is a person. There is no way to compensate the victim at all. You can't transfer any rights to a dead man. Yet surely, a killer owes more compensation for his deed than an eye gouger owes for his. Death is the ultimate deprivation. It destroys the victim's eyes, arms, legs, brain—everything that was his.

But to whom does the killer owe compensation? If the victim has heirs, the killer may try to compensate them by paying the victim's medical and funeral expenses and any other financial costs that can be directly attributed to the death. Beyond these costs it is impossible to determine any suitable compensation in the general case. Particular cases may entail obvious specific modes of compensation, but we cannot analyze them all ahead of time.

It does no good to kill a killer or to put him in prison (unless he is threatening to kill again). Such punishments violate the killer's rights. We may hate killers and we may show it in many imaginative ways. We can be nasty to them. We can publicize their deeds and humiliate them. We may disassociate ourselves from them. If everyone so chooses, killers may effectively be banished from society. But we have no right to physically punish them without their consent.

Justice as defined by the nonaggression principle supports forcible restitution to the extent that the exact same property that was taken without the owner's permission be returned to him: nothing more, and nothing less. When this cannot be done, justice cannot be done. Principles of fairness or retribution or social utility should not be imposed by brute force as a substitute for justice. The principles of justice are, by definition, the only principles that may legitimately be enforced by violence.

A criminal has the same rights as anybody else. He does not gain or lose any rights by committing a crime. Because a thief does not gain any rights by stealing, it is OK to take back what a thief has stolen, whether the thief agrees or not. But, because a thief does not lose any rights by stealing, if you want to do more than take back what was stolen, if you want to punish the thief or make him pay, you must restrict your actions to those that you have a right to do to anybody. No matter how much you think the criminal deserves to suffer, you have no right to directly harm him or his property without his permission, except to stop his crime.

If a thief or tortfeasor loses or destroys your property, he cannot possibly return it. I think it would be fair to take some of his property as compensation, but it would be unjust to do this without his consent. If you have an insurance policy you can get compensation from your insurance company. So, it is possible for the market to provide some compensation beyond mere repossession without having to coerce criminals and tortfeasors more than is allowed by the nonaggression principle.

Nature is not fair. It does not allow a murderer to restore his victim's life. A murderer cannot possibly make any reparations to the victim. He should try to compensate the victim's family. It would be extremely unfair if he refused to offer any reparations, but he has the right to refuse, which means no one may force him to pay for his crime. The same goes for many other serious crimes. Rape, for example, is a crime for which perfect reparation is impossible. It was not fair of nature to make reparation impossible in the most serious crimes. It is also not fair that the nonaggression principle allows criminals to refuse to compensate their victims.

Justice deals with specific rights. It gives us the authority to demand what belongs to us. But justice must be exact. Reparation and punishment cannot be determined by principles of justice—by appeals to rights. Reparation and punishment are concerned with fairness rather than justice. Therefore, they must be agreed to voluntarily according to whatever principles of fairness the parties share.

The nonaggression principle limits the means that we can use to attain fairness. It prohibits us from using violence against unfair people such as racists and bigots who are not aggressing. It prohibits us from using violence to punish criminals. It prohibits us from forcing criminals to make reparation to their victims. None of this is fair.

Justice takes priority.

Justice is not the only kind of morality, but it is the primary kind, and it should not be bypassed. The only way for morality to make sense is to allow people to make moral decisions. For this they need the freedom allowed by the nonaggression principle.

Morality deals with justice, fairness, social welfare, self-fulfillment, and many other things, but these are separate subjects. Social welfare, for example, cannot be a sure guide to justice, because it requires us to determine: (1) what we can do to make others happy, (2) whether the same action that makes some people happy might make others less happy, and (3) which action will maximize happiness overall. We cannot make these determinations with the precision required to justify the use of brute force. Justice must be more certain than this.

It is OK to take the law into your own hands.

If we associate punishment with law, the fear of people taking the law into their own hands is appropriate. There are no objective standards for punishment, so private citizens cannot administer punishment without creating fear and chaos. The state can protect us from lynch mobs. This is the grain of truth in the link between anarchy (the absence of the state) and chaos.

If we separate punishment and law, the idea of allowing people to take the law into their own hands becomes less terrifying. If law is simple justice, as defined by the nonaggression principle, then it is legitimate for people to take the law into their own hands.

In a libertarian nation based on the nonaggression principle, vindictive practices would be restricted, because no one would be forced to submit to any coercive punishments against his will. Anyone who tried to impose physical punishment by brute force would be recognized as a criminal, and everyone would have the right to use force to stop him.

Are compensation and punishment necessary?

One argument against the nonaggression principle goes like this:

  1. If criminals are not punished and not even forced to compensate their victims, they have nothing to lose by committing crimes.
  2. Since they have nothing to lose by committing crimes, more people will become criminals, and each criminal will commit more crimes.
  3. Society will fall apart.
My response to this argument is that the first premise is false. I admit that fear of punishment deters many people from committing crimes. I know this is true because fear of punishment works for me. It not only inhibits me from committing real crimes, it also inhibits me from committing victimless crimes that the state has decided to prohibit. I also admit that forcing criminals to compensate their victims would discourage many people from committing real crimes. But fear of punishment and forced compensation are not the only means to deter crime.

Remember that in the nonaggression system of justice everyone has the right to use violence, if necessary, to defend his rights. We also have the right to help each other to defend our rights and to hire professionals to defend us. So it is not true that under the nonaggression paradigm people have nothing to lose by committing crimes. They could lose their lives.

Under a system of law based on the nonaggression principle people can own weapons. They can hire bodyguards, watchmen, and private investigators. They can install burglar alarms, keep their valuables in vaults, purchase insurance policies, and so on.

Protection agencies have a legitimate place in a libertarian nation whose laws are based on the nonaggression principle. They do not have to know the past history of their clients in order to know that their clients have a right to be protected from all invaders. They only need to stipulate that they will protect each of their clients except while the client is breaking the peace. The protection agencies do not have to know whether their clients have ever committed crimes for which they have not been punished. The protection agencies only have to be concerned with their client's actions now.

Now is when the client wants protection from threats. Now is when all the pertinent facts are evident and can be evaluated. Now is the best time to make moral judgments, not later in a courtroom after the crisis is over and circumstances have changed, when the evidence is old and doesn't matter anymore.

We have the right to deputize others to defend us. This justifies private firms that, for a price, could provide such services as:

  • Registering property titles and wills
  • Doing title searches
  • Retrieving stolen goods
  • Mediating disputes in private courts
  • Defending lives and property by supplying bodyguards, watchmen, burglar alarms, and other devices
  • Drafting, witnessing, and registering performance-bond contracts and wagers
  • Holding performance bonds and gambling stakes
  • Redistributing such performance bonds and gambling stakes according to the terms of the contract or wager Protection agencies might be hired by individuals or by insurance firms acting on behalf of their clients. The insurance companies would have a financial interest in returning stolen property to its rightful owners if they are obligated to pay compensation for stolen goods not returned. Insurance firms might hire detectives to retrieve stolen goods. They might hire guards and watchmen to prevent crime. They might finance the development of new methods to prevent or stop crime. They could hire scientists to invent methods for identifying insured property and finding it when it is lost or stolen. Perhaps an electronic device for detecting objects could be invented that could distinguish any particular registered piece of property from all others and make it easier to track it down.

    If a person wants to be sure his stolen goods will be returned as quickly as possible, he should have an insurance policy that requires his insurance company to pay a high rate of interest on the insured value of the goods until they are returned. If an object is of special sentimental or esthetic value to a person, he should insure it for a great deal of money so that if it is stolen, the insurance company will have a special incentive to search diligently for it.

    If you argue that it is crass to put a monetary price on the things that are most important to you, and that is why you do not take out an insurance policy on them, fine. That is a respectable position. But then don't ask for money in compensation when you lose those precious things. If you say that you value something highly, but have not insured it for a lot of money, why should anyone believe that your demands for a large sum of money for compensation are justified? The time to prove that you treasure something is when you have it in your possession, not after someone has taken it from you.

    If possible, the insurance companies should return the same physical item that was stolen. If this is not possible, or if the property can only be returned in a damaged condition, then the insurance company should pay compensation. The insurance policy should spell out the method for determining compensation. Those insurance firms that gain renown for providing the best protection and the fairest settlements at the best prices will do the most business.

    In this regime, the law is enforced privately by whoever chooses to enforce it. Private individuals and organizations offer to arbitrate disputes and assess compensations. There is no definitive repository of laws or legal rulings. The nonaggression principle is the only law. All other legitimate laws such as the laws against murder, assault, and theft are already implicit in the nonaggression principle. Everyone is free to expostulate on the meaning of the nonaggression principle, but no one's written laws need to be regarded as authoritative.

    In a libertarian nation, no one could be dragged into a courtroom to testify, to judge, or to stand trial against his will. Courts would be set up solely for the convenience of disputing parties who mutually agree to arbitration. No one would be obligated to use such courts.

    Arbitration companies might choose to publish their procedural rules, rules of evidence, and rules for assessing compensations. This would help disputants to decide which arbitrator to go to, if any. Arbitrators will probably find it useful to study prior legal rulings, but they will always retain the option of reasoning directly from the nonaggression principle in each case. They will succeed or fail in their legal careers based on the reputations they earn for the wisdom of their decisions.

    If I were involved in a contract dispute, I would prefer to use a judge who was familiar with the principles of contract law that were developed over the years in common-law courts. Arbitration companies would probably hire legal scholars to sort through case law and legal treatises. Instead of merely looking for legal precedents and loopholes, they would look for sound arguments developed in previous cases that might be useful in the future. They might compile their own databases and select their own rules of procedure from the best procedures used in the past. Competition among arbitration companies would encourage them to find and adopt legal principles and procedures that enhance their reputations for fairness and professionalism.

    Not only can a system of law based on the nonaggression principle work, it can work better than any other system. The free market will produce better methods of enforcing the nonaggression principle and settling disputes than any individual theorist could hope to do on his own.

    Back to Libertarian Essays by Roy Halliday

    This page was last updated on October 1, 2012.
    This site is maintained by Roy Halliday. If you have any comments or suggestions, please send them to