Law as a normative discipline

Law is a set of commands; the principles of tort or criminal law, which we shall be dealing with, are negative commands or prohibitions, on the order of “thou shalt not” do actions X, Y, or Z. In short, certain actions are considered wrong to such a degree that it is considered appropriate to use the sanctions of violence (since law is the social embodiment of violence) to combat, defend against, and punish the transgressors.

There are many actions against which it is not considered appropriate to use violence, individual or organised. Mere lying (that is, where contracts to transfer property titles are not broken), treachery, base ingratitude, being nasty to one’s friends or associates, or not showing up for appointments, are generally considered wrong, but few think of using violence to enjoin or combat them. Other sanctions, such as refusing to see the person or have dealings with him, putting him in Coventry, and so on, may be used by individuals or groups, but using the violence of the law to prohibit such actions is considered excessive and inappropriate.

If ethics is a normative discipline that identifies and classifies certain sets of actions as good or evil, right or wrong, then tort or criminal law is a subset of ethics identifying certain actions as appropriate for using violence against them. The law says that action X should be illegal, and therefore should be combated by the violence of the law. The law is a set of “ought” or normative propositions.

Many writers and jurists have claimed the law is a value-free, “positive” discipline. Of course it is possible simply to list, classify and analyse existing law without going further into saying what the law should or should not be. But that sort of jurist is not fulfilling his essential task. Since the law is ultimately a set of normative commands, the true jurist or legal philosopher has not completed his task until he sets forth what the law should be, difficult though that might be. If he does not, then he necessarily abdicates his task in favor of individuals or groups untrained in legal principles, who may lay down their commands by sheer fiat and arbitrary caprice.

Thus, the Austinian jurists proclaim that the king, or sovereign, is supposed to lay down the law, and the law is purely a set of commands emanating from his will. But then the question arises: On what principles does or should the king operate? Is it ever possible to say that the king is issuing a “bad” or “improper” decree? Once the jurist admits that, he is going beyond arbitrary will to begin to frame a set of normative principles that should be guiding the sovereign. And then he is back to normative law.

Modern variants of positive legal theory state that the law should be what the legislators say it is. But what principles are to guide the legislators? And if we say that the legislators should be the spokesmen for their constituents, then we simply push the problem one step back, and ask: What principles are supposed to guide the voters? Or is the law, and therefore everyone’s freedom of action, to be ruled by arbitrary caprice of millions rather than of one man or a few?

Even the older concept that the law should be determined by tribal or common-law judges, who are merely interpreting the custom of the tribe or society, cannot escape normative judgments basic to the theory. Why must the rules of custom be obeyed? If tribal custom requires the murder of all people over six feet tall, must this custom be obeyed regardless? Why cannot reason lay down a set of principles to challenge and overthrow mere custom and tradition? Similarly, why may it not be used to overthrow mere arbitrary caprice by king or public?

As we shall see, tort or criminal law is a set of prohibitions against the invasion of, or aggression against, private property rights; that is, spheres of freedom of action by each individual. But if that is the case, then the implication of the command, “Thou shall not interfere with A’s property right,” is that A’s property right is just and therefore should not be invaded. Legal prohibitions, therefore, far from being in some sense value-free, actually imply a set of theories about justice, in particular the just allocation of property rights and property titles. “Justice” is nothing if not a normative concept.

In recent years, however, jurists and “Chicago school” economists have attempted to develop theories of value-free property rights, rights defined and protected not on the basis of ethical norms such as justice but of some form of “social efficiency.” In one such variant, Ronald Coase and Harold Demsetz have asserted that “it doesn’t make any difference” how property rights are allocated in cases of conflicting interests, provided that some property rights are assigned to someone and then defended. In his famous example, Coase discusses a railroad locomotive’s blighting of nearby farms and orchards. To Coase and Demsetz, this damage of a farmer’s crops by the railroad is an “externality” which should, according to the tenets of social efficiency, be internalised. But to these economists, it doesn’t make any difference which of two possible courses of action one adopts. Either one says that the farmer has a property right in his orchard; therefore the railroad should have to pay damages for his loss, and the farmer should be able to enjoin the railroad’s invasive actions. Or the railroad has the right to spew forth smoke wherever it wishes, and if the farmer wishes to stop the smoke, he must pay the railroad to install a smoke abatement device. It does not matter, from the point of view of expenditure of productive resources, which route is taken.

For example, suppose the railroad commits one hundred thousand dollars worth of damage, and in Case One, this action is held to invade the farmer’s property. In that case, the railroad must pay one hundred thousand dollars to the farmer or else invest in a smoke abatement device, whichever is cheaper. But in Case Two, where the railroad has the property right to emit the smoke, the farmer would have to pay the railroad up to one hundred thousand dollars to stop damaging his farm. If the smoke device costs less than one hundred thousand dollars, say eighty thousand dollars, then the device will be installed regardless of who was assigned the property right. In Case One, the railroad will spend eighty thousand dollars on the device rather than have to pay one hundred thousand dollars to the farmer; in Case Two the farmer will be willing to pay the railroad eighty thousand dollars and up to one hundred thousand dollars to install the device. If, on the other hand, the smoke device costs more than one hundred thousand dollars, say one hundred twenty thousand dollars, then the device will not be installed anyway, regardless of which route is taken. In Case One, the railroad will keep pouring out smoke and keep paying the farmer damages of one hundred thousand dollars rather than spend one hundred twenty thousand dollars on the device; in Case Two, it will not pay the farmer to bribe the railroad one hundred twenty thousand dollars for the device, since this is more of a loss to him than the one hundred thousand dollars of damage. Therefore, regardless of how property rights are assigned — according to Coase and Demsetz — the allocation of resources will be the same. The difference between the two is only a matter of “distribution,” that is, of income or wealth.

There are many problems with this theory. First, income and wealth are important to the parties involved, although they might not be to uninvolved economists. It makes a great deal of difference to both of them who has to pay whom. Second, this thesis works only if we deliberately ignore psychological factors. Costs are not only monetary. The farmer might well have an attachment to the orchard far beyond the monetary damage. Therefore, the orchard might be worth far more to him than the one hundred thousand dollars in damages, so that it might take one million dollars to compensate him for the full loss. But then the supposed indifference totally breaks down. In Case One, the farmer will not be content to accept a mere one hundred thousand dollars in damages. He will take out an injunction against any further aggression against his property, and even if the law allows bargaining between the parties themselves to remove the injunction, he will insist on over one million dollars from the railroad, which the railroad will not be willing to pay. Conversely, in Case Two, there is not likely to be a way for the farmer to raise the one million dollars needed to stop the smoke invasion of the orchard.

The love of the farmer for his orchard is part of a larger difficulty for the Coase-Demsetz doctrine: Costs are purely subjective and not measurable in monetary terms. Coase and Demsetz have a proviso in their indifference thesis that all “transaction costs” be zero. If they are not, then they advocate allocating the property rights to whichever route entails minimum social transaction costs. But once we understand that costs are subjective to each individual and therefore unmeasurable, we see that costs cannot be added up. But if all costs, including transaction costs, cannot be added, then there is no such thing as “social transaction costs,” and they cannot be compared in Cases One or Two, or indeed, in any other situation.

Another serious problem with the Coase-Demsetz approach is that pretending to be value-free, they in reality import the ethical norm of “efficiency,” and assert that property rights should be assigned on the basis of such efficiency. But even if the concept of social efficiency were meaningful, they don’t answer the questions of why efficiency should be the overriding consideration in establishing legal principles or why externalities should be internalised above all other considerations. We are now out of Wertfreiheit and back to unexamined ethical questions.

Another attempt by Chicago school economists to make legal public policy recommendations under the guise of Wertfreiheit is the contention that over the years common-law judges will always arrive at the socially efficient allocation of property rights and tort liabilities. Demsetz stresses rights that will minimise social transaction costs; Richard Posner stresses maximisation of “social wealth.” All this adds an unwarranted historical determinism, functioning as a kind of invisible hand guiding judges to the current Chicago school path, to the other fallacies examined above.

If the law is a set of normative principles, it follows that whatever positive or customary law has emerged cannot simply be recorded and blindly followed. All such law must be subject to a thorough critique grounded on such principles. Then, if there are discrepancies between actual law and just principles, as there almost always are, steps must be taken to make the law conform with correct legal principles.

Physical invasion

The normative principle I am suggesting for the law is simply this: No action should be considered illicit or illegal unless it invades, or aggresses against, the person or just property of another. Only invasive actions should be declared illegal, and combated with the full power of the law. The invasion must be concrete and physical. There are degrees of seriousness of such invasion, and hence, different proper degrees of restitution or punishment. “Burglary,” simple invasion of property for purposes of theft, is less serious than “robbery,” where armed force is likely to be used against the victim. Here, however, we are not concerned with the questions of degrees of invasion or punishment, but simply with invasion per se.

If no man may invade another person’s “just” property, what is our criterion of justice to be? There is no space here to elaborate on a theory of justice in property titles. Suffice it to say that the basic axiom of libertarian political theory holds that every man is a selfowner, having absolute jurisdiction over his own body. In effect, this means that no one else may justly invade, or aggress against, another’s person. It follows then that each person justly owns whatever previously unowned resources he appropriates or “mixes his labour with.” From these twin axioms — self-ownership and “homesteading” — stem the justification for the entire system of property rights titles in a free-market society. This system establishes the right of every man to his own person, the right of donation, of bequest (and, concomitantly, the right to receive the bequest or inheritance), and the right of contractual exchange of property titles.

Legal and political theory have committed much mischief by failing to pinpoint physical invasion as the only human action that should be illegal and that justifies the use of physical violence to combat it. The vague concept of “harm” is substituted for the precise one of physical violence. Consider the following two examples. Jim is courting Susan and is just about to win her hand in marriage, when suddenly Bob appears on the scene and wins her away. Surely Bob has done great “harm” to Jim. Once a nonphysical-invasion sense of harm is adopted, almost any outlaw act might be justified. Should Jim be able to “enjoin” Bob’s very existence?

Similarly, A is a successful seller of razor blades. But then B comes along and sells a better blade, teflon-coated to prevent shaving cuts. The value of A’s property is greatly affected. Should he be able to collect damages from B, or, better yet, to enjoin B’s sale of a better blade? The correct answer is not that consumers would be hurt if they were forced to buy the inferior blade, although that is surely the case. Rather, no one has the right to legally prevent or retaliate against “harms” to his property unless it is an act of physical invasion. Everyone has the right to have the physical integrity of his property inviolate; no one has the right to protect the value of his property, for that value is purely the reflection of what people are willing to pay for it. That willingness solely depends on how they decide to use their money. No one can have a right to someone else’s money, unless that other person had previously contracted to transfer it to him.

In the law of torts, “harm” is generally treated as physical invasion of person or property. The outlawing of defamation (libel and slander) has always been a glaring anomaly in tort law. Words and opinions are not physical invasions. Analogous to the loss of property value from a better product or a shift in consumer demand, no one has a property right in his “reputation.” Reputation is strictly a function of the subjective opinions of other minds, and they have the absolute right to their own opinions whatever they may be. Hence, outlawing defamation is itself a gross invasion of the defamer’s right of freedom of speech, which is a subset of his property right in his own person.

An even broader assault on freedom of speech is the modern Warren-Brandeis-inspired tort of invasion of the alleged right of “privacy,” which outlaws free speech and acts using one’s own property that are not even false or “malicious.”

In the law of torts, “harm” is generally treated as physical invasion of person or property and usually requires payment of damages for “emotional” harm if and only if that harm is a consequence of physical invasion. Thus, within the standard law of trespass — an invasion of person or property — “battery” is the actual invasion of someone else’s body, while “assault” is the creation by one person in another of a fear, or apprehension, of battery.

To be a tortious assault and therefore subject to legal action, tort law wisely requires the threat to be near and imminent. Mere insults and violent words, vague future threats, or simple possession of a weapon cannot constitute an assault; there must be accompanying overt action to give rise to the apprehension of an imminent physical battery. Or, to put it another way, there must be a concrete threat of an imminent battery before the prospective victim may legitimately use force and violence to defend himself.

Physical invasion or molestation need not be actually “harmful” or inflict severe damage in order to constitute a tort. The courts properly have held that such acts as spitting in someone’s face or ripping off someone’s hat are batteries. Chief Justice Holt’s words in 1704 still seem to apply: “The least touching of another in anger is a battery.” While the actual damage may not be substantial, in a profound sense we may conclude that the victim’s person was molested, was interfered with, by the physical aggression against him, and that hence these seemingly minor actions have become legal wrongs.

Initiation of an overt act: Strict liability

If only a physical invasion of person or property constitutes an illicit act or tort, then it becomes important to demarcate when a person may act as if such a physical invasion is about to take place. Libertarian legal theory holds that A may not use force against B except in self-defense, that is, unless B is initiating force against A. But when is A’s force against B legitimate self-defense, and when is it itself illegitimate and tortious aggression against B? To answer this question, we must consider what kind of tort liability theory we are prepared to adopt.

Suppose, for example, that Smith sees Jones frowning in his direction across the street, and that Smith has an abnormal fear of being frowned at. Convinced that Jones is about to shoot him, he therefore pulls a gun and shoots Jones in what he is sure is self-defense. Jones presses a charge of assault and battery against Smith. Was Smith an aggressor and therefore should he be liable? One theory of liability — the orthodox “reasonable man” or “reasonable conduct” or “negligence” theory — says he should, because frowning would not rouse the apprehension of imminent attack in a “reasonable man.” A competing theory, once held and now being revived — that of “strict liability” or “strict causal liability” — agrees because it should be clear to a judge or jury that Jones was not an imminent aggressor. And this would hold regardless of how sincere Smith was in his fear of attack.

Two serious flaws in the “reasonable man” theory are that the definition of “reasonable” is vague and subjective, and that guilty aggressors go unpunished, while their victims remain uncompensated. In this particular case, the two theories happen to coincide, but in many other cases they do not. Take, for example, the case of Courvoisier v. Raymond (1896). In this case, the defendant, a storekeeper, was threatened by a rioting mob. When a man who happened to be a plainclothes policeman walked up to the defendant, trying to help him, the defendant, mistaking him for a rioter, shot the policeman. Should the storekeeper have been liable?

The trial court decided the case properly — on the basis of strict liability — and the jury decided for the policeman. For it is clear that the defendant committed a battery by shooting the plaintiff. In strict liability theory, the question is causation: Who initiated the tort or crime? An overriding defense for the defendant’s action was if the plaintiff in fact had committed an assault, threatening an imminent initiation of a battery against him. The question traditionally then becomes a factual one for juries to decide: Did the plainclothesman in fact threaten battery against the storekeeper? The jury decided for the policeman. The appeals court, however, reversed the trial court’s decision. To the court, the storekeeper acted as a “reasonable man” when he concluded, though incorrectly, that the plainclothesman was out to attack him.

When is an act to be held an assault? Frowning would scarcely qualify. But if Jones had whipped out a gun and pointed it in Smith’s direction, though not yet fired, this is clearly a threat of imminent aggression, and would properly be countered by Smith plugging Jones in self-defense. (In this case, our view and the “reasonable man” theory would again coincide.) The proper yardstick for determining whether the point of assault had been reached is this: Did Jones initiate an “overt act” threatening battery? As Randy Barnett has pointed out:

In a case less than a certainty, the only justifiable use of force is that used to repel an overt act that is something more than mere preparation, remote from time and place of the intended crime. It must be more than “risky”; it must be done with the specific intent to commit a crime and directly tend in some substantial degree to accomplish it.

Similar principles hold in innocent-bystander cases. Jones assaults and attacks Smith; Smith, in self-defense, shoots. The shot goes wild and accidentally hits Brown, an innocent bystander. Should Smith be liable? Unfortunately, the courts, sticking to the traditional “reasonable man” or “negligence” doctrine, have held that Smith is not liable if indeed he was reasonably intending self-defense against Jones. But, in libertarian and in strict liability theory, Smith has indeed aggressed against Brown, albeit unintentionally, and must pay for this tort. Thus, Brown has a proper legal action against Smith: Since Jones coerced or attacked Smith, Smith also has an independent and proper action for assault or battery against Jones. Presumably, the liability or punishment against Jones would be considerably more severe than against Smith.

One of the great flaws in the orthodox negligence approach has been to focus on one victim’s (Smith’s) right of self-defense in repelling an attack, or on his good-faith mistake. But orthodox doctrine unfortunately neglects the other victim — the man frowning across the street, the plainclothesman trying to save someone, the innocent bystander. The plaintiff’s right of self-defense is being grievously neglected. The proper point to focus on in all these cases is: Would the plaintiff have had the right to plug the defendant in his self-defense? Would the frowning man, the plainclothesman, the innocent bystander, if he could have done so in time, have had the right to shoot the sincere but erring defendants in self-defense? Surely, whatever our theory of liability, the answer must be “yes”; hence, the palm must go to the strict liability theory, which focuses on everyone’s right of self-defense and not just that of a particular defendant. For it is clear that since these plaintiffs had the right to plug the defendant in self-defense, then the defendant must have been the tortious aggressor, regardless of how sincere or “reasonable” his actions may have been.

From various illuminating discussions of Professor Epstein, it seems evident that there are three contrasting theories of tort liability interwoven in our legal structure. The oldest, strict causal liability, apportioned blame and burden on the basis of identifiable cause: Who shot whom? Who assaulted whom? Only defense of person and property was a proper defense against a charge of using force. This doctrine was replaced during the nineteenth century by negligence or “reasonable man” theory, which let many guilty defendants off the hook if their actions were judged reasonable or did not exhibit undue negligence. In effect, negligence theory swung the balance excessively in favor of the defendant and against the plaintiff. In contrast, modern theory emerging increasingly in the twentieth century, anxious to help plaintiffs (especially if they are poor), seeks ways to find against defendants even if strict cause of physical invasion cannot be proven. If the oldest theory is termed “strict causal liability,” the modern one might be termed “presumptive liability,” since the presumption seems to be against the defendant, in flagrant violation of the Anglo-Saxon criminal law presumption of innocence on the part of the defendant.

Extending our discussion from crimes against the person to crimes against property, we may apply the same conclusion: Anyone has the right to defend his property against an overt act initiated against it. He may not move with force against an alleged aggressor — a trespasser against his land or chattels — until the latter initiates force by an overt act.

How much force may a victim use to defend either his person or his property against invasion? Here we must reject as hopelessly inadequate the current legal doctrine that he may use only “reasonable” force, which in most cases has reduced the victim’s right to defend himself virtually to a nullity. In current law, a victim is only allowed to use maximal, or “deadly” force, (a) in his own home, and then only if he is under direct personal attack; or (b) if there is no way that he can retreat when he is personally under attack. All this is dangerous nonsense. Any personal attack might turn out to be a murderous one; the victim has no way of knowing whether or not the aggressor is going to stop short of inflicting a grave injury upon him. The victim should be entitled to proceed on the assumption that any attack is implicitly a deadly one, and therefore to use deadly force in return.

In current law, the victim is in even worse straits when it comes to defending the integrity of his own land or movable property. For there, he is not even allowed to use deadly force in defending his own home, much less other land or properties. The reasoning seems to be that since a victim would not be allowed to kill a thief who steals his watch, he should therefore not be able to shoot the thief in the process of stealing the watch or in pursuing him. But punishment and defense of person or property are not the same, and must be treated differently. Punishment is an act of retribution after the crime has been committed and the criminal apprehended, tried, and convicted. Defense while the crime is being committed, or until property is recovered and the criminal apprehended, is a very different story. The victim should be entitled to use any force, including deadly force, to defend or to recover his property so long as the crime is in the process of commission — that is, until the criminal is apprehended and duly tried by legal process. In other words, he should be able to shoot looters.

The proper burden of risk

We conclude, then, that no one may use force to defend himself or his property until the initiation of an overt act of aggression against him. But doesn’t this doctrine impose an undue risk upon everyone?

The basic reply is that life is always risky and uncertain and that there is no way of getting round this primordial fact. Any shifting of the burden of risk away from one person simply places it upon someone else. Thus, if our doctrine makes it more risky to wait until someone begins to aggress against you, it also makes life less risky, because as a non-aggressor, one is more assured that no excited alleged victim will pounce upon you in supposed “self-defense.” There is no way for the law to reduce risk overall; it then becomes important to use some other principle to set the limits of permissible action, and thereby to allocate the burdens of risk. The libertarian axiom that all actions are permissible except overt acts of aggression provides such a principled basis for risk allocation.

There are deeper reasons why overall risks cannot be reduced or minimised by overt legal action. Risk is a subjective concept unique to each individual; therefore, it cannot be placed in measurable quantitative form. Hence, no one person’s quantitative degree of risk can be compared to another’s, and no overall measure of social risk can be obtained. As a quantitative concept, overall or social risk is fully as meaningless as the economist’s concept of “social costs” or social benefits.

In a libertarian world, then, everyone would assume the “proper burden of risk” placed upon him as a free human being responsible for himself. That would be the risk involved in each man’s person and property. Of course, individuals could voluntarily pool their risks, as in various forms of insurance, in which risks are shared and benefits paid to losers from the pool. Or, speculators could voluntarily assume risks of future price changes that are sloughed off by others in hedging operations on the market. Or, one man could assume another’s risks for payment, as in the case of performance and other forms of bonding. What would not be permissible is one group getting together and deciding that another group should be forced into assuming their risks. If one group, for example, forces a second group to guarantee the former’s incomes, risks are greatly increased for the latter, to the detriment of their individual rights. In the long run, of course, the whole system might collapse, since the second group can only provide guarantees out of their own production and incomes, which are bound to fall as the burden of social parasitism expands and cripples society.

The proper burden of proof

If every man’s proper burden of risk is to refrain from coercion unless an overt act against his person or property has been initiated against him, then what is the proper burden of proof against a defendant?

First, there must be some rational standards of proof for libertarian principles to operate. Suppose that the basic axiom of libertarianism — no initiation of force against person or property — is enshrined in all judicial proceedings. But suppose that the only criterion of proof is that all persons under six feet tall are considered guilty while all persons over six feet tall are held to be innocent. It is clear that these procedural standards of proof would be in direct and flagrant violation of libertarian principles. So would tests of proof in which irrelevant or random occurrences would decide the case, such as the medieval trial by ordeal or trial by tea leaves or astrological charts.

From a libertarian point of view, then, proper procedure calls for rational proof about the guilt or innocence of persons charged with tort or crime. Evidence must be probative in demonstrating a strict causal chain of acts of invasion of person or property. Evidence must be constructed to demonstrate that aggressor A in fact initiated an overt physical act invading the person or property of victim B.

Who, then, should bear the burden of proof in any particular case? And what criterion or standard of proof should be satisfied?

The basic libertarian principle is that everyone should be allowed to do whatever he or she is doing unless committing an overt act of aggression against someone else. But what about situations where it is unclear whether or not a person is committing aggression? In those cases, the only procedure consonant with libertarian principles is to do nothing; to lean over backwards to ensure that the judicial agency is not coercing an innocent man. If we are unsure, it is far better to let an aggressive act slip through than to impose coercion and therefore to commit aggression ourselves. A fundamental tenet of the Hippocratic oath, “at least, do not harm,” should apply to legal or judicial agencies as well.

The presumption of every case, then, must be that every defendant is innocent until proven guilty, and the burden of proof must be squarely upon the plaintiff.

If we must always insist on laissez-faire, then it follows that such a weak standard of proof as “preponderance of evidence” must not be allowed to serve as a demonstration of guilt. If the plaintiff produces evidence adjudged in some sense to weigh a mere 51 percent on behalf of the guilt of the defendant, this is scarcely better than random chance as justification for the court’s using force against the defendant. Presumption of innocence, then, must set a far higher standard of proof.

At present, “preponderance of evidence” is used to decide civil cases, whereas a far tougher standard is used for criminal cases, since penalties are so much stiffer. But, for libertarians, the test of guilt must not be tied to the degree of punishment; regardless of punishment, guilt involves coercion of some sort levied against the convicted defendant. Defendants deserve as much protection in civil torts as in criminal cases.

A few judges, properly shocked by the dominant view that a mere fifty-one percent of the evidence may serve to convict, have changed the criterion to make sure whoever is trying the case — judge or jury — is convinced of guilt by the preponderance of evidence. A more satisfactory criterion, however, is that the trier must be convinced of the defendant’s guilt by “clear, strong, and convincing proof.” Fortunately, this test has been used increasingly in civil cases in recent years. Better yet were stronger but generally rejected formulations of certain judges such as “clear, positive, and unequivocal” proof, and one judge’s contention that the phrase means that the plaintiffs “must … satisfy you to a moral certainty.”

But the best standard for any proof of guilt is the one commonly used in criminal cases: Proof “beyond a reasonable doubt.” Obviously, some doubt will almost always persist in gauging people’s actions, so that such a standard as “beyond a scintilla of doubt” would be hopelessly unrealistic. But the doubt must remain small enough that any “reasonable man” will be convinced of the fact of the defendant’s guilt. Conviction of guilt “beyond a reasonable doubt” appears to be the standard most consonant with libertarian principle.

The outstanding nineteenth-century libertarian constitutional lawyer, Lysander Spooner, was an ardent advocate of the “beyond a reasonable doubt” standard for all guilt:

the lives, liberties, and properties of men are too valuable to them, and the natural presumptions are too strong in their favor to justify the destruction of them by their fellow men on a mere balancing of probabilities, or on any ground whatever short of certainty beyond a reasonable doubt. (Italics Spooner’s)

While the reasonable doubt criterion generally has not been used in civil cases, a few precedents do exist for this seemingly bold and shocking proposal. Thus, in the claim of an orally offered gift in a probate case, the court ruled that the alleged gift “must be proven by forceful, clear and conclusive testimony which convinces the court beyond a reasonable doubt of its truthfulness.” And in a suit to revise a written contract, the court ruled that the mistake must be “established by evidence so strong and conclusive as to place it beyond reasonable doubt.”

Strict causality

What the plaintiff must prove, then, beyond a reasonable doubt is a strict causal connection between the defendant and his aggression against the plaintiff. He must prove, in short, that A actually “caused” an invasion of the person or property of B.

In a brilliant analysis of causation in the law, Professor Epstein has demonstrated that his own theory of strict tort liability is intimately connected to a direct, strict, commonsense view of “cause.” Causal proposition in a strict liability view of the law takes such form as, “A hit B,” “A threatened B,” or “A compelled B to hit C.” Orthodox tort theory, in contrast, by stressing liability for “negligence” rather than for direct aggression action, is tangled up with vague and complex theories of “cause,” far removed from the commonsense “A hit B” variety. Negligence theory postulates a vague, “philosophical” notion of “cause in fact” that virtually blames everyone and no one, past, present and future for every act, and then narrows cause in a vague and unsatisfactory manner to “proximate cause” in the specific case. The result, as Epstein trenchantly points out, is to vitiate the concept of cause altogether and to set the courts free to decide cases arbitrarily and in accordance with their own views of social policy.

To establish guilt and liability, strict causality of aggression leading to harm must meet the rigid test of proof beyond a reasonable doubt. Hunch, conjecture, plausibility, even mere probability are not enough. In recent years, statistical correlation has been commonly used, but it cannot establish causation, certainly not for a rigorous legal proof of guilt or harm. Thus, if lung cancer rates are higher among cigarette smokers than noncigarette smokers, this does not in itself establish proof of causation. The very fact that many smokers never get lung cancer and that many lung cancer sufferers have never smoked indicates that there are other complex variables at work. So that while the correlation is suggestive, it hardly suffices to establish medical or scientific proof; a fortiori it can still less establish any sort of legal guilt (if, for example, a wife who developed lung cancer should sue a husband for smoking and therefore injuring her lungs).

Milton Katz points out, in a case where the plaintiff sued for air pollution damage:

Suppose the plaintiff should claim serious damage: For emphysema, perhaps, or for lung cancer, bronchitis or some other comparably serious injury to his lungs. He would face a problem of proof of causation…. Medical diagnoses appear to have established that sulphur dioxide and other air pollutants often play a significant role in the etiology of emphysema and other forms of lung damage. But they are by no means the only possible causative factors. Emphysema and lung cancer are complex illnesses which may originate in a variety of causes, for example, cigarette smoking, to name one familiar example. If and when the plaintiff should succeed in establishing that the defendants’ conduct polluted the air of his home, it would not follow that the pollution caused his illness. The plaintiff would still have to meet the separate burden of proving the etiology of his lung damage.

Thus, a strict causal connection must exist between an aggressor and a victim, and this connection must be provable beyond a reasonable doubt. It must be causality in the commonsense concept of strict proof of the “A hit B” variety, not mere probability or statistical correlation.

Liability of the aggressor only

Under strict liability theory, it might be assumed that if “A hit B,” then A is the aggressor and that therefore A and only A is liable to B. And yet the legal doctrine has arisen and triumphed, approved even by Professor Epstein, in which sometimes C, innocent and not the aggressor, is also held liable. This is the notorious theory of “vicarious liability.”

Vicarious liability grew up in medieval law, in which a master was responsible for the torts committed by his servants, serfs, slaves, and wife. As individualism and capitalism developed, the common law changed, and vicarious liability disappeared in the sixteenth and seventeenth centuries, when it was sensibly concluded that “the master should not be liable for his servant’s torts unless he had commanded the particular act.”

Since the eighteenth and nineteenth centuries, however, the vicarious liability of masters or employers is back with a vengeance. As long as the tort is committed by the employee in the course of furthering, even if only in part, his employer’s business, then the employer is also liable. The only exception is when the servant goes “on a frolic of his own” unconnected with the employer’s business. Prosser writes:

The fact that the servant’s act is expressly forbidden by the master, or is done in a manner which he has prohibited, is … usually not conclusive, and does not in itself prevent an act from being within the scope of employment [and therefore making the master liable]. A master cannot escape liability merely by ordering his servant to act carefully…. Thus instructions to a sales clerk never to load a gun while exhibiting it will not prevent liability when the clerk does so, in an effort to sell the gun…. [T]he master cannot escape responsibility no matter how specific, detailed, and emphatic his orders may have been to the contrary. This has been clear since the leading English cases (Limpus v. London General Omnibus Co., [1862] 1H. & C. 526, 158 Eng. Rep. 993) in which an omnibus company was held liable notwithstanding definite orders to its driver not to obstruct other vehicles.

Even more remarkably, the master is now held responsible even for intentional torts committed by the servant without the master’s consent:

In general, the master is held liable for any intentional tort committed by the servant where its purpose, however misguided, is wholly or in part to further the master’s business.
Thus he will be held liable where his bus driver crowds a competitor’s bus into a ditch, or assaults a trespasser to eject him from the bus, or a salesman makes fraudulent statements about the products he is selling.

Prosser is properly scornful of the tortured reasoning by which the courts have tried to justify a legal concept so at war with libertarianism, individualism, and capitalism, and suited only to a pre-capitalist society.

A multitude of very ingenious reasons have been offered for the vicarious liability of a master: He has a more or less fictitious “control” over the behaviour of a servant; he has “set the whole thing in motion,” and is therefore responsible for what has happened; he has selected the servant and trusted him, and so should suffer for his wrongs, rather than an innocent stranger who has had no opportunity to protect himself; it is a great concession that any man should be permitted to employ another at all, and there should be a corresponding responsibility as the price to be paid for it…. Most courts have made little or no effort to explain the result, and have taken refuge in rather empty phrases, such as … the endlessly repeated formula of “respondeat superior,” which in itself means nothing more than “look to the man higher up.”

In fact, as Prosser indicates, the only real justification for vicarious liability is that employers generally have more money than employees, so that it becomes more convenient (if one is not the employer), to stick the wealthier class with the liability. In the cynical words of Thomas Baty: “In hard fact, the reason for the employers’ liability is the damages are taken from a deep pocket.”

In opposition, too, we have Justice Holmes’s lucid critique: “I assume that common sense is opposed to making one man pay for another man’s wrong, unless he has actually brought the wrong to pass…. I therefore assume that common sense is opposed to the fundamental theory of agency.”

One would expect that in a strict causal liability theory, vicarious liability would be tossed out with little ceremony. It is therefore surprising to see Professor Epstein violate the spirit of his own theory. He seems to have two defenses for the doctrine of respondeat superior and vicarious liability. One is the curious argument that “just as the employer gets and benefits from the gains for his worker’s activities, so too should he be required to bear the losses from these activities.” This statement fails to appreciate the nature of voluntary exchange: Both employer and employee benefit from the wage contract. Moreover, the employer does bear the “losses” in the event his production (and, therefore, his resources) turn out to be misdirected. Or, suppose the employer makes a mistake and hires an incompetent person, who is paid ten thousand dollars. The employer may fire this worker, but he and he alone bears the ten thousand dollars loss. Thus, there appears to be no legitimate reason for forcing the employer to bear the additional cost of his employee’s tortious behaviour.

Epstein’s second argument is contained in the sentence: “X corporation hurt me because its servant did so in the course of his employment.” Here Epstein commits the error of conceptual realism, since he supposes that a “corporation” actually exists, and that it committed an act of aggression. In reality, a “corporation” does not act; only individuals act, and each must be responsible for his own actions and those alone. Epstein may deride Holmes’s position as being based on the “nineteenth-century premise that individual conduct alone was the basis of individual responsibility,” but Holmes was right nevertheless.

A theory of just property: Homesteading

There are two fundamental principles upon which the libertarian theory of just property rests:

1. Everyone has absolute property right over his or her own body; and
2. everyone has an absolute property right over previously unowned natural resources (land) which he first occupies and brings into use (in the Lockean phrase, “Mixing his labour with the land”).

The “first ownership to first use” principle for natural resources is also popularly called the “homesteading principle.” If each man owns the land that he “mixes his labour with,” then he owns the product of that mixture, and he has the right to exchange property titles with other, similar producers. This establishes the right of free contract in the sense of transfer of property titles. It also establishes the right to give away such titles, either as a gift or bequest.

Most of us think of homesteading unused resources in the old-fashioned sense of clearing a piece of unowned land and farming the soil. There are, however, more sophisticated and modern forms of homesteading, which should establish a property right. Suppose, for example, that an airport is established with a great deal of empty land around it. The airport exudes a noise level of, say, X decibels, with the sound waves traveling over the empty land. A housing development then buys land near the airport. Some time later, the homeowners sue the airport for excessive noise interfering with the use and quiet enjoyment of the houses.

Excessive noise can be considered a form of aggression but in this case the airport has already homesteaded X decibels worth of noise. By its prior claim, the airport now “owns the right” to emit X decibels of noise in the surrounding area. In legal terms, we can then say that the airport, through homesteading, has earned an easement right to creating X decibels of noise. This homesteaded easement is an example of the ancient legal concept of “prescription,” in which a certain activity earns a prescriptive property right to the person engaging in the action.

On the other hand, if the airport starts to increase noise levels, then the homeowners could sue or enjoin the airport from its noise aggression for the extra decibels, which had not been homesteaded. Of course if a new airport is built and begins to send out noise of X decibels onto the existing surrounding homes, the airport becomes fully liable for the noise invasion.

It should be clear that the same theory should apply to air pollution. If A is causing pollution of B’s air, and this can be proven beyond a reasonable doubt, then this is aggression and it should be enjoined and damages paid in accordance with strict liability, unless A had been there first and had already been polluting the air before B’s property was developed. For example, if a factory owned by A polluted originally unused property, up to a certain amount of pollutant X, then A can be said to have homesteaded a pollution easement of a certain degree and type.

Given a prescriptive easement, the courts have generally done well in deciding its limits. In Kerlin v. Southern Telephone and Telegraph Co. (1941), a public utility had maintained an easement by prescription of telephone poles and wires over someone else’s land (called the “servient estate” in law). The utility wished to string up two additional wires, and the servient estate challenged its right to do so. The court decided correctly that the utility had the right because there was no proposed change in the “outer limits of space utilised by the owner of the easement.” On the other hand, an early English case decided that an easement for moving carts could not later be used for the purpose of driving cattle.

Unfortunately, the courts have not honored the concept of homestead in a noise or pollution easement. The classic case is Sturgis v. Bridgman (1879) in England. The plaintiff, a physician, had purchased land in 1865; on the property next to him the defendant, a pharmacist, used a mortar and pestle, which caused vibrations on the physician’s property. There was no problem, however, until the physician built a consultation room ten years later. He then sued to enjoin the pharmacist, claiming that his work constituted a nuisance. The defendant properly argued that the vibrations were going on before the construction of the consultation room, that they then did not constitute a nuisance, and that therefore he had a prescriptive right to keep operating his business. Nevertheless, defendant’s claim was denied.

Consequently, we have such injustice as compulsory changes of character in a business and a failure to provide prescription through first use. Thus, Prosser notes that “the character of a district may change with the passage of time, and the industry set up in the open country may become a nuisance, or be required to modify its activities, when residences spring up around it. It will acquire no prescriptive right.” A just law would tell the later arriving residents that they knew what they were getting into, and that they must adapt to the industrial ambience rather than vice-versa.

In some cases, however, the courts have held or at least considered that by the plaintiff’s “coming to the nuisance,” he has voluntarily entered a pre-existing situation, and that therefore the defendant is not guilty. Prosser states that “in the absence of a prescriptive right the defendant cannot condemn the surrounding premises to endure the nuisance,” but our whole point here is that the homesteader of a noise or a pollution easement has indeed earned that right in cases of “coming to the nuisance.”

Dominant court opinion, as in the case of Ensign v. Walls (1948), discards or minimises “coming to the nuisance” and dismisses the idea of a homesteaded easement. But minority opinion has strongly supported it, as in the New York case of Bove v. Donner-Hanna Coke Co. (1932). Plaintiff had moved into an industrial region, where defendant was operating a coke oven on the opposite side of the street. When plaintiff tried to enjoin the coke oven out of existence, the court rejected the plea with these exemplary words:

With all the dirt, smoke and gas which necessarily come from factory chimneys, trains and boats, and with full knowledge that this region was especially adapted for industrial rather than residential purposes, and that factories would increase in the future, plaintiff selected this locality as the site of her future home. She voluntarily moved into this district, fully aware of the fact that the atmosphere would constantly be contaminated by dirt, gas and foul odors; and that she could not hope to find in this locality the pure air of a strictly residential zone. She evidently saw certain advantages in living in this congested center. This is not the case of an industry, with its attendant noise and dirt, invading a quiet, residential district. This is just the opposite. Here a residence is built in an area naturally adapted for industrial purposes and already dedicated to that use. Plaintiff can hardly be heard to complain at this late date that her peace and comfort have been disturbed by a situation which existed, to some extent at least, at the very time she bought her property.

Nuisances, visible and invisible

An invasion of someone else’s land can be considered a trespass or a nuisance, and there is considerable confusion about the boundaries of each. For our purposes, the classic distinction between the two is important. Trespass occurs when “there is a physical entry that is a direct interference with the possession of land, which usually must be accomplished by a tangible mass.” On the other hand, “contact by minute particles or intangibles, such as industrial dust, noxious fumes, or light rays, has heretofore generally been held insufficient to constitute a trespassory entry, on the ground that there is no interference with possession, or that the entry is not direct, or that the invasion failed to qualify as an entry because of its imponderable or intangible nature.”

These more intangible invasions qualify as private nuisances and can be prosecuted as such. A nuisance may be, as Prosser points out:

an interference with the physical condition of the land itself, as by vibration or blasting which damages a house, the destruction of crops, flooding, raising the water table, or the pollution of a stream or of an underground water supply. It may consist of a disturbance of the comfort or convenience of the occupant, as by unpleasant odors, smoke or dust or gas, loud noises, excessive light or high temperature, or even repeated telephone calls.

Prosser sums up the difference between trespass and nuisance:

Trespass is an invasion of the plaintiff’s interest in the exclusive possession of his land, while nuisance is an interference with his use and enjoyment of it. The difference is that between … felling a tree across his boundary line and keeping him awake at night with the noise of a rolling mill.

But what precisely does the difference between “exclusive possession” and “interference with use” mean? Furthermore, the practical difference between a tort action for trespass and for nuisance is that a trespass is illegal per se, whereas a nuisance, to be actionable, has to damage the victim beyond the mere fact of invasion itself. What, if any, is the justification for treating a trespass and nuisance so differently? And is the old distinction between tangible and invisible invasion really now obsolete as Prosser maintains, “in the light of modern scientific tests?” Or, as a Columbia Law Review note put it:

The federal court … suggested that historically the reluctance of courts to find that invasion by gases and minute particles were trespassory resulted from the requirement that to find a trespass a court must be able to see some physical intrusion by tangible matter; it then found that this difficulty no longer exists because courts may today rely on scientific detecting methods, which can make accurate quantitative measurements of gases and minute solids, to determine the existence of a physical entry of tangible matter.

The distinction between visible and invisible, however, is not completely swept away by modern scientific detection methods. Let us take two opposite situations. First, a direct trespass: A rolls his car onto B’s lawn or places a heavy object on B’s grounds. Why is this an invasion and illegal per se? Partly because, in the words of an old English case, “the law infers some damage; if nothing more, the treading down of grass or herbage.” But it is not just treading down; a tangible invasion of B’s property interferes with his exclusive use of the property, if only by taking up tangible square feet (or cubic feet). If A walks on or puts an object on B’s land, then B cannot use the space A or his object has taken up. An invasion by a tangible mass is a per se interference with someone else’s property and therefore illegal.

In contrast, consider the case of radio waves, which is a crossing of other people’s boundaries that is invisible and insensible in every way to the property owner. We are all bombarded by radio waves that cross our properties without our knowledge or consent. Are they invasive and should they therefore be illegal, now that we have scientific devices to detect such waves? Are we then to outlaw all radio transmission? And if not, why not?

The reason why not is that these boundary crossings do not interfere with anyone’s exclusive possession, use or enjoyment of their property. They are invisible, cannot be detected by man’s senses, and do no harm. They are therefore not really invasions of property, for we must refine our concept of invasion to mean not just boundary crossing, but boundary crossings that in some way interfere with the owner’s use or enjoyment of this property. What counts is whether the senses of the property owner are interfered with.

But suppose it is later discovered that radio waves are harmful, that they cause cancer or some other illness? Then they would be interfering with the use of the property in one’s person and should be illegal and enjoined, provided of course that this proof of harm and the causal connection between the specific invaders and specific victims are established beyond a reasonable doubt.

So we see that the proper distinction between trespass and nuisance, between strict liability per se and strict liability only on proof of harm, is not really based on “exclusive possession” as opposed to “use and enjoyment.” The proper distinction is between visible and tangible or “sensible” invasion, which interferes with possession and use of the property, and invisible, “insensible” boundary crossings that do not and therefore should be outlawed only on proof of harm.

The same doctrine applies to low-level radiation, which virtually everyone and every object in the world emanates, and therefore everyone receives. Outlawing, or enjoining, low-level radiation, as some of our environmental fanatics seem to be advocating, would be tantamount to enjoining the entire human race and all the world about us. Low-level radiation, precisely because it is undetectable by man’s senses, interferes with no one’s use or possession of his property, and therefore may only be acted against upon strict causal proof of harm beyond a reasonable doubt.

The theory of homestead easements discussed earlier would require no restriction upon radio transmissions or on people’s low-level radiation. In the case of radio transmissions, Smith’s ownership of land and all of its appurtenances does not entitle him to own all radio waves passing over and across his land, for Smith has not homesteaded or transmitted on radio frequencies here. Hence, Jones, who transmits a wave on, say, twelve hundred kilohertz, homesteads the ownership of that wave as far as it travels, even if it travels across Smith’s property. If Smith tries to interfere with or otherwise disrupt Jones’s transmissions, he is guilty of interfering with Jones’s just property.

Only if the radio transmissions are proven to be harmful to Smith’s person beyond a reasonable doubt should Jones’s activities be subject to injunction. The same type of argument, of course, applies to radiation transmissions.

Between tangible trespass and radio waves or low-level radiation, there is a range of intermediate nuisances. How should they be treated?

Air pollution, consisting of noxious odors, smoke, or other visible matter, definitely constitutes an invasive interference. These particles can be seen, smelled, or touched, and should therefore constitute invasion per se, except in the case of homesteaded air pollution easements. (Damages beyond the simple invasion would, of course, call for further liability.) Air pollution, however, of gases or particles that are invisible or undetectable by the senses should not constitute aggression per se, because being insensible they do not interfere with the owner’s possession or use. They take on the status of invisible radio waves or radiation, unless they are proven to be harmful, and until this proof and the causal connection from aggressor to victim can be established beyond a reasonable doubt.

Excessive noise is certainly a tort of nuisance; it interferes with a person’s enjoyment of his property, including his health. However, no one would maintain that every man has the right to live as if in a soundproofed room; only excessive noise, however vague the concept, can be actionable.

In a sense, life itself homesteads noise easement. Every area has certain noises, and people moving into an area must anticipate a reasonable amount of noise. As Terry Yamada ruefully concedes:

An urban resident must accept the consequences of a noisy environment situation. Courts generally hold that persons who live or work in densely populated communities must necessarily endure the usual annoyances and discomforts of those trades and businesses located in the neighbourhood where they live or work; such annoyances and discomforts, however, must not be more than those reasonably expected in the community and lawful to the conduct of the trade or business.

In short, he who wants a soundproof room must pay for its installation.

The current general rule of the civil courts on nuisance suits for noise is cogent:

A noise source is not a nuisance per se but only becomes a nuisance under certain conditions. These conditions depend on a consideration of the surrounding area, the time of day or night when the noise-producing activities take place and the manner in which the activity is conducted. A private nuisance is compensable only when it is unreasonable or excessive and when it produces actual physical discomfort or injury to a person of ordinary sensibilities so as to interfere with the use and enjoyment of the property.

Owning the technological unit: Land and air

In our discussion of homesteading, we did not stress the problem of the size of the area to be homesteaded. If A uses a certain amount of a resource, how much of that resource is to accrue to his ownership? Our answer is that he owns the technological unit of the resource. The size of that unit depends on the type of good or resource in question, and must be determined by judges, juries, or arbitrators who are expert in the particular resource or industry in question. If resource X is owned by A, then A must own enough of it so as to include necessary appurtenances. For example, in the courts’ determination of radio frequency ownership in the 1920s, the extent of ownership depended on the technological unit of the radio wave — its width on the electromagnetic spectrum so that another wave would not interfere with the signal, and its length over space. The ownership of the frequency then was determined by width, length, and location.

American land settlement is a history of grappling, often unsuccessfully, with the size of the homestead unit. Thus, the homesteading provision in the federal land law of 1861 provided a unit of one hundred sixty acres, the clearing and use of which over a certain term would convey ownership to the homesteader. Unfortunately, in a few years, when the dry prairie began to be settled, one hundred sixty acres was much too low for any viable land use (generally ranching and grazing). As a result, very little Western land came into private ownership for several decades. The resulting overuse of the land caused the destruction of Western grass cover and much of the timberland.

With the importance of analysing the technological unit in mind, let us examine the ownership of airspace. Can there be private ownership of the air, and if so, to what extent?

The common-law principle is that every landowner owns all the airspace above him upward indefinitely unto the heavens and downward into the center of the earth. In Lord Coke’s famous dictum: Cujus est solum ejus est usque ad coelum; that is, he who owns the soil owns upward unto heaven, and, by analogy, downward to Hades. While this is a time-honored rule, it was, of course, designed before planes were invented. A literal application of the rule would in effect outlaw all aviation, as well as rockets and satellites.

But is the practical problem of aviation the only thing wrong with the ad coelum rule? Using the homesteading principle, the ad coelum rule never made any sense, and is therefore overdue in the dustbin of legal history. If one homesteads and uses the soil, in what sense is he also using all the sky above him up into heaven? Clearly, he isn’t.

The ad coelum rule unfortunately lingered on in the Restatement of Torts (1939), adopted by the Uniform State Law for Aeronautics and enacted in twenty-two states during the 1930s and 1940s. This variant continued to recognise unlimited ownership of upward space, but added a superior public privilege to invade the right. Aviators and satellite owners would still bear the burden of proof that they possessed this rather vague privilege to invade private property in airspace. Fortunately, the Uniform Act was withdrawn by the Commissioners on Uniform State Laws in 1943, and is now on the way out.

A second solution, adopted by the Ninth Circuit Federal Court in 1936, scrapped private property in airspace altogether and even allowed planes to buzz land close to the surface. Only actual interference with present enjoyment of land would constitute a tort. The most popular nuisance theory simply outlaws interference with land use, but is unsatisfactory because it scraps any discussion whatever of ownership of airspace.

The best judicial theory is the “zone,” which asserts that only the lower part of the airspace above one’s land is owned; this zone is the limit of the owner’s “effective possession.” As Prosser defines it, “effective possession” is “so much of the space above him as is essential to the complete use and enjoyment of the land.” The height of the owned airspace will vary according to the facts of the case and therefore according to the “technological unit.” Thus, Prosser writes:

This was the rule applied in the early case of Smith v. New England Aircraft Co., where flights at the level of one hundred feet were held to be trespass, since the land was used for cultivation of trees which reached that height. A few other cases have adopted the same view.

The height of the zone of ownership must vary according to the facts of each case.

On the other hand, the nuisance theory should be added to the strict zone of ownership for cases such as where excess aircraft noise injures people or activities in an adjoining area, not directly underneath the plane. At first, the federal courts ruled that only low flights overhead could constitute a tort against private landowners, but the excessive noise case of Thornburg v. Port of Portland (1962) corrected that view. The court properly reasoned in Thornburg:

If we accept … the validity of the propositions that a noise can be a nuisance; that a nuisance can give rise to an easement; and that a noise coming straight down from above one’s land can ripen into a taking if it is persistent enough and aggravated enough, then logically the same kind and degree of interference with the use and enjoyment of one’s land can also be a taking even though the noise vector may come from some direction other than the perpendicular.

While there is no reason why the concept of ownership of airspace cannot be used to combat air pollution torts, this has rarely been done. Even when ad coelum was riding high, it was used against airplane overflights but not to combat pollution of one’s air, which was inconsistently considered as a communal resource. The law of nuisance could traditionally be used against air pollution, but until recently it was crippled by “balancing of the equities,” negligence rules against strict liability, and by declaration that “reasonable” air pollution was not actionable. In the classic case of Holman v. Athens Empire Laundry Co. (1919), the Supreme Court of Georgia declared: “The pollution of the air, so far as reasonably necessary to the enjoyment of life and indispensable to the progress of society, is not actionable.” Fortunately, that attitude is now becoming obsolete.

Although air pollution should be a tort subject to strict liability, it should be emphasised that statements like “everyone has the right to clean air” are senseless. There are air pollutants constantly emerging from natural processes, and one’s air is whatever one may happen to possess. The eruption of Mt. St. Helens should have alerted everyone to the ever-present processes of natural pollution. It has been the traditional and proper rule of the common-law courts that no landowner is responsible for the harm caused by natural forces originating on his property. As Prosser writes, a landowner

is under no affirmative duty to remedy conditions of purely natural origin upon his land, although they may be highly dangerous or inconvenient to his neighbours…. Thus it has been held that the landowner is not liable for the existence of a foul swamp, for falling rocks, for the spread of weeds or thistles growing on his land, for harm done by indigenous animals, or for the normal, natural flow of surface water.

In sum, no one has a right to clean air, but one does have a right to not have his air invaded by pollutants generated by an aggressor.

Air pollution: Law and regulation

We have established that everyone may do as he wishes provided he does not initiate an overt act of aggression against the person or property of anyone else. Anyone who initiates such aggression must be strictly liable for damages against the victim, even if the action is “reasonable” or accidental. Finally, such aggression may take the form of pollution of someone else’s air, including his owned effective airspace, injury against his person, or a nuisance interfering with his possession or use of his land.

This is the case, provided that:

1. the polluter has not previously established a homestead easement;
2. while visible pollutants or noxious odors are per se aggression, in the case of invisible and insensible pollutants the plaintiff must prove actual harm;
3. the burden of proof of such aggression rests upon the plaintiff;
4. the plaintiff must prove strict causality from the actions of the defendant to the victimisation of the plaintiff;
5. the plaintiff must prove such causality and aggression beyond a reasonable doubt; and
6. there is no vicarious liability, but only liability for those who actually commit the deed.

With these principles in mind, let us consider the current state of air pollution law. Even the current shift from negligence and “reasonable” actions to strict liability has by no means satisfied the chronic special pleaders for environmental plaintiffs. As Paul Downing says, “Currently, a party who has been damaged by air pollution must prove in court that emitter A damaged him. He must establish that he was damaged and emitter A did it, and not emitter B. This is almost always an impossible task.” If true, then we must assent uncomplainingly. After all, proof of causality is a basic principle of civilised law, let alone of libertarian legal theory.

Similarly, James Krier concedes that even if requirement to prove intent or unreasonable conduct or negligence is replaced by strict liability, there is still the problem of proving the causal link between the wrongful conduct and the injury. Krier complains that “cause and effect must still be established.” He wants to “make systematic reallocation of the burden of proof,” that is, take the burden off the plaintiff, where it clearly belongs. Are defendants now to be guilty until they can prove themselves innocent?

The prevalence of multiple sources of pollution emissions is a problem. How are we to blame emitter A if there are other emitters or if there are natural sources of emission? Whatever the answer, it must not come at the expense of throwing out proper standards of proof, and conferring unjust special privileges on plaintiffs and special burdens on defendants.

Similar problems of proof are faced by plaintiffs in nuclear radiation cases. As Jeffrey Bodie writes, “In general the courts seem to require a high degree of causation in radiation cases which frequently is impossible to satisfy given the limited extent of medical knowledge in this field.” But as we have seen above, it is precisely this “limited extent of knowledge” that makes it imperative to safeguard defendants from lax canons of proof.

There are, of course, innumerable statutes and regulations that create illegality besides the torts dealt with in common-law courts. We have not dealt with laws such as the Clean Air Act of 1970 or regulations for a simple reason: None of them can be permissible under libertarian legal theory. In libertarian theory, it is only permissible to proceed coercively against someone if he is a proven aggressor, and that aggression must be proven in court (or in arbitration) beyond a reasonable doubt. Any statute or administrative regulation necessarily makes actions illegal that are not overt initiations of crimes or torts according to libertarian theory. Every statute or administrative rule is therefore illegitimate and itself invasive and a criminal interference with the property rights of noncriminals.

Suppose, for example, that A builds a building, sells it to B, and it promptly collapses. A should be liable for injuring B’s person and property and the liability should be proven in court, which can then enforce the proper measures of restitution and punishment. But if the legislature has imposed building codes and inspections in the name of “safety,” innocent builders (that is, those whose buildings have not collapsed) are subjected to unnecessary and often costly rules, with no necessity by government to prove crime or damage. They have committed no tort or crime, but are subject to rules, often only distantly related to safety, in advance by tyrannical governmental bodies. Yet, a builder who meets administrative inspection and safety codes and then has a building of his collapse, is often let off the hook by the courts. After all, has he not obeyed all the safety rules of the government, and hasn’t he thereby received the advance imprimatur of the authorities?

The only civil or criminal system consonant with libertarian legal principles is to have judges (and/or juries and arbitrators) pursuing charges of torts by plaintiffs made against defendants.

It should be underlined that in libertarian legal theory, only the victim (or his heirs and assigns) can legitimately press suit against alleged transgressors against his person or property. District attorneys or other government officials should not be allowed to press charges against the wishes of the victim, in the name of “crimes” against such dubious or nonexistent entities as “society” or the “state.” If, for example, the victim of an assault or theft is a pacifist and refuses to press charges against the criminal, no one else should have the right to do so against his wishes. For just as a creditor has the right to “forgive” an unpaid debt voluntarily, so a victim, whether on pacifist grounds or because the criminal has bought his way out of a suit or any other reason, has the right to “forgive” the crime so that the crime is thereby annulled.

Critics of automobile emissions will be disturbed by the absence of government regulation, in view of the difficulties of proving harm to victims from individual automobiles. But, as we have stressed, utilitarian considerations must always be subordinate to the requirements of justice. Those worried about auto emissions are in even worse shape in the tort law courts, because libertarian principle also requires a return to the now much scorned nineteenth-century rule of privity.

The privity rule, which applies largely to the field of products liability, states that the buyer of a defective product can only sue the person with whom he had a contract. If the consumer buys a watch from a retailer, and the watch does not work, it should only be the retailer whom he can sue, since it was the retailer who transferred ownership of the watch in exchange for the consumer’s money. The consumer, in contrast to modern rulings, should not be able to sue the manufacturer, with whom he had no dealings. It was the retailer who, by selling the product, gave an implied warranty that the product would not be defective. And similarly, the retailer should only be able to sue the wholesaler for the defective product, the wholesaler the jobber, and finally the manufacturer.

In the same way, the privity role should be applied to auto emissions. The guilty polluter should be each individual car owner and not the automobile manufacturer, who is not responsible for the actual tort and the actual emission. (For all the manufacturer knows, for example, the car might only be used in some unpopulated area or used mainly for aesthetic contemplation by the car owner.) As in the product liability cases, the only real justification for suing the manufacturer rather than the retailer is simply convenience and deep pockets, with the manufacturer presumably being wealthier than the retailer.

While the situation for plaintiffs against auto emissions might seem hopeless under libertarian law, there is a partial way out. In a libertarian society, the roads would be privately owned. This means that the auto emissions would be emanating from the road of the road owner into the lungs or airspace of other citizens, so that the road owner would be liable for pollution damage to the surrounding inhabitants. Suing the road owner is much more feasible than suing each individual car owner for the minute amount of pollutants he might be responsible for. In order to protect himself from these suits, or even from possible injunctions, the road owner would then have the economic incentive to issue anti-pollution regulations for all cars that wish to ride on his road. Once again, as in other cases of the “tragedy of the commons,” private ownership of the resource can solve many “externality” problems.

Collapsing crime into tort

But if there is no such entity as society or the state, or no one except the victim that should have any standing as a prosecutor or plaintiff, this means that the entire structure of criminal law must be dispensed with, and that we are left with tort law, where the victim indeed presses charges against the aggressor. However, there is no reason why parts of the law that are now the province of criminal law cannot be grafted onto an enlarged law of torts. For example, restitution to the victim is now considered the province of tort law, whereas punishment is the realm of criminal law. Yet, punitive damages for intentional torts (as opposed to accidents) now generally are awarded in tort law. It is therefore conceivable that more severe punishments, such as imprisonment, forced labour to repay the victim, or transportation, could be grafted onto tort law as well.

One cogent argument against any proposal to collapse criminal into tort law is that, in the reasoning against allowing punitive damages in tort cases, they are “fixed only by the caprice of the jury and imposed without the usual safeguards thrown about criminal procedure, such as proof of guilt beyond a reasonable doubt [and] the privilege against self-incrimination.” But, as argued above, standards such as proof beyond a reasonable doubt should be applied to tort law cases as well.

Professor Epstein, in attempting to preserve a separate realm for criminal law as against a proposed collapse into tort law, rests much of his case on the law of attempts. In criminal law, an attempted crime that for some reason fails and results in no damage or invasion of the rights of the victim, is still a crime and can be prosecuted. And yet, Epstein charges, such an attempted crime would not be an invasion of rights and therefore could not be a tort and could not be prosecuted under tort law.

Randy Barnett’s rebuttal, however, is conclusive. Barnett points out, first, that most unsuccessful attempts at invasion result nevertheless in “successful” though lesser invasion of person or property, and would therefore be prosecutable under tort law. “For example, attempted murder is usually an aggravated assault and battery, attempted armed robbery is usually an assault, attempted car theft or burglary is usually a trespass.” Secondly, even if the attempted crime created no invasion of property per se, if the attempted battery or murder became known to the victim, the resulting creation of fear in the victim would be prosecutable as an assault. So the attempted criminal (or tortfeasor) could not get away unscathed.

Therefore, the only attempted invasion that could not be prosecuted under the law of torts would be one that no one ever knew anything about. But if no one knows about it, it cannot be prosecuted, under any law.

Furthermore, as Barnett concludes, potential victims would not be prevented under libertarian law from defending themselves from attempts at crime. As Barnett says, it is justifiable for a victim or his agents to repel an overt act that has been initiated against him, and that in fact is what an attempt at crime is all about.

Joint torts and joint victims

So far in discussing invasions of person or property, we have confined ourselves to single aggressors and single victims, of the “A hit B” or “damaged B” variety. But actual air pollution cases often have multiple alleged aggressors and multiple victims. On what principles may they be prosecuted or convicted?

When more than one aggressor has contributed to a tort, it is generally more convenient for the plaintiffs to join the defendants together in one suit (“joinder”). Convenience, however, should not be allowed to override principle or rights, and in our view the original common-law rule of joinder was correct: Defendants can be compulsorily joined only when all the parties acted in concert in a joint tortious enterprise.

In the case of truly joint torts, it also makes sense to have each of the joint aggressors equally liable for the entire amount of the damages. If it were otherwise, each criminal could dilute his own liability in advance by simply adding more criminals to their joint enterprise. Hence, since the action of all the aggressors was in concert, the tort was truly joint, so that

“all coming to do an unlawful act and of one part, the act of one is the act of the same part being present.” Each was therefore liable for the entire damage done, although one might have battered the plaintiff, while another imprisoned him, and a third stole his silver buttons. All might be joined as defendants in the same action at law.

Unfortunately, for purposes of convenience, the joinder rule has been weakened, and the courts in many cases have permitted plaintiffs to compel joinder of defendants even in cases where torts are committed separately and not in concert. The confusion in joinder for both joint and separate torts has caused many courts to apply the full or “entire” liability rule to each aggressor. In the case of separate torts impinging upon a victim, this makes little sense. Here the rule should always be what it has traditionally been in nuisance cases, that the courts apportion damage in accordance with the separate causal actions contributed by each defendant.

Air pollution cases generally are those of separate torts impinging upon victims; therefore, there should be no compulsory joinder and damages should be apportioned in accordance with the separate causal factors involved. As Prosser writes:

Nuisance cases, in particular, have tended to result in apportionment of the damages, largely because the interference with the plaintiff’s use of his land has tended to be severable in terms of quantity, percentage, or degree. Thus defendants who independently pollute the same stream or who flood the plaintiff’s land from separate sources, are liable only severally for the damages individually caused, and the same is true as to nuisance due to noise, or pollution of the air.

But because the injuries are multiple and separate, it is then up to the plaintiffs to show a rational and provable basis for apportioning the damage among the various defendants and causative factors. If this rule is properly and strictly adhered to, and proof is beyond a reasonable doubt, the plaintiffs in air pollution cases generally will be able to accomplish very little. To counter this, environmental lawyers have proposed a weakening of the very basis of our legal system by shifting the burden of proof for detailed allocation of damages from the plaintiffs to the various defendants.

Thus, compulsory joinder of defendants may proceed on the original common-law rule only when the defendants have allegedly committed a truly joint tort, in concerted action. Otherwise, defendants may insist on separate court actions.

What about joinder of several plaintiffs against one or more defendants? When may that take place? This problem is highly relevant to air pollution cases, where there are usually many plaintiffs proceeding against one or more defendants.

In the early common law, the rules were rigorous on limiting permissible joinder of plaintiffs to cases where all causes in action had to affect all the parties joined. This has now been liberalised to permit joint action by plaintiffs where the joint action arises out of the same transaction or series of transactions, and where there is at least one question of law of fact common to all plaintiffs. This appears to be a legitimate liberalisation of when plaintiffs shall be allowed voluntary joinder.

While permissive joinder of plaintiffs in this sense is perfectly legitimate, this is not the case for “class action” suits, where the outcome of the suit is binding even upon those members of the alleged class of victims who did not participate in the suit. It seems the height of presumption for plaintiffs to join in a common suit and to press a “class action” suit, in which even those other alleged victims who never heard of or in some way did not consent to a suit are bound by the result. The only plaintiffs who should be affected by a suit are those who voluntarily join. Thus, it would not be permissible for fifty residents of Los Angeles to file a pollution suit on behalf of the class of “all citizens of Los Angeles,” without their knowledge or express consent. On the principle that only the victim and his heirs and assigns may press suit or use force on his behalf, class action suits binding on anyone except voluntary plaintiffs are impermissible.

Unfortunately, while the 1938 Federal Rule of Civil Procedure 23 provided for at least one type of nonbinding class action, the “spurious class action,” the revised 1966 rules make all class action suits binding upon the class as a whole, or rather on all those members of the class who do not specifically request exclusion. In an unprecedented step, voluntary action is now being assumed if no action is taken. The residents of Los Angeles, who might not even know about the suit in question, are required to take steps to exclude themselves from the suit, otherwise the decision will be binding upon them. Furthermore, most states have followed the new federal rules for class action suits.

As in the case of voluntary joinder, the post-1966 class action must involve questions of law or fact common to their entire class. Fortunately, the courts have placed further limits on the use of class action. In most cases, all identifiable members of the class must be given individual notice of the suit, giving them at least an opportunity to opt out of the action; also, the class must be definitely identifiable, ascertainable, and manageable. Under this rule, the federal courts generally would not allow “all residents of the city of Los Angeles” to be party to a class action suit. Thus, a suit allegedly on behalf of all residents of Los Angeles County (over seven million persons) to enjoin two hundred ninety-three companies from polluting the atmosphere was dismissed by the court “as unmanageable because of the number of parties (plaintiffs and defendants), the diversity of their interests, and the multiplicity of issues involved.”

Another sensible limitation placed on most class action suits is that common class interests in the suit must predominate over separate individual interests. Thus, a class suit will not be allowed where separate individual issues are “numerous and substantial,” and therefore common issues do not predominate. In the case of City of San Jose v. Superior Court (1974), the court threw out a class action suit of landowners near an airport, suing for damages to their land resulting from airport noise, pollution, traffic, and so on. Even though the airport affected each of the landowners, the court properly ruled that “the right of each landowner to recover for the harm to his land involved too many individual facts (for example, proximity to flight paths, type of property, value, use, and so on)” to permit a class suit.

Thus, class action suits should not be allowed except where every plaintiff actively and voluntarily joins and where common interests predominate over separate and individual ones.

How, then, have the recent class action rules been applied to the question of air pollution? Krier says with dismay that while the 1966 Federal Rule 23 is indeed more liberal than its predecessor in allowing class action, the US Supreme Court has virtually nullified its impact by ruling that class members may aggregate individual claims for federal courts only when they share a common undivided interest. According to Krier, this cogent limitation rules out most class action suits in air pollution cases. He adds that while this restriction does not apply to state suits, these are often even less viable than federal class suits before the new rules. Krier complains, in an unconsciously humorous note, that some class action suits don’t attract any plaintiffs at all.

But the major problem of class action suits for the plaintiffs, Krier concedes, is the manageability and ascertainability rules for suits with a large number of plaintiffs in the class, citing in particular the Diamond v. General Motors case. But whereas Krier attributes the problem solely to the lack of competence and facilities judges possess to balance the various interests, he fails to realise the still larger problem of lack of identifiability and lack of clear proof of guilt and causality between defendant and plaintiff.

Conclusion

We have attempted to set forth a set of libertarian principles by which to gauge and reconstruct the law. We have concluded that everyone should be able to do what he likes, except if he commits an overt act of aggression against the person and property of another. Only this act should be illegal, and it should be prosecutable only in the courts under tort law, with the victim or his heirs and assigns pressing the case against the alleged aggressor. Therefore, no statute or administrative ruling creating illegal actions should be permitted. And since any prosecution on behalf of “society” or the “state” is impermissible, the criminal law would be collapsed into a reconstituted tort law, incorporating punishment and part of the law of attempts.

The tortfeasor or criminal is to be strictly liable for his aggression, with no evasion of liability permissible on the basis of “negligence” or “reasonability” theories. However, the liability must be proven on the basis of strict causality of the defendant’s action against the plaintiff, and it must be proven by the plaintiff beyond a reasonable doubt.

The aggressor and only the aggressor should be liable, and not the employer of an aggressor, provided, of course, that the tort was not committed at the direction of the employer. The current system of vicarious employer liability is a hangover from pre-capitalist master/serf relations and is basically an unjust method of finding deep pockets to plunder.

These principles should apply to all torts, including air pollution. Air pollution is a private nuisance generated from one person’s landed property onto another and is an invasion of the airspace appurtenant to land and, often, of the person of the landowner. Basic to libertarian theory of property rights is the concept of homesteading, in which the first occupier and user of a resource thereby makes it his property. Therefore, where a “polluter” has come first to the pollution and has preceded the landowner in emitting air pollution or excessive noise onto empty land, he has thereby homesteaded a pollution or excessive noise easement. Such an easement becomes his legitimate property right rather than that of the later, adjacent landowner. Air pollution, then, is not a tort but only the ineluctable right of the polluter if he is simply acting on a homestead easement. But where there is no easement and air pollution is evident to the senses, pollution is a tort per se because it interferes with the possession and use of another’s air. Boundary crossing — say by radio waves or low-level radiation — cannot be considered aggression because it does not interfere with the owner’s use or enjoyment of his person or property. Only if such a boundary crossing commits provable harm — according to principles of strict causality and beyond a reasonable doubt — can it be considered a tort and subject to liability and injunction.

A joint tort, in which defendants are compelled to defend themselves jointly, should apply only if all acted in concert. Where their actions are separate, the suits must be separate as well, and the liability apportioned separately. Plaintiffs should be able to join their suits against a defendant only if their cases have a common element predominating over the separate and individual interests. Class action suits are impermissible beyond a voluntary joinder of plaintiffs because they presume to act for and bind class members who have not agreed to join in the suit.

Finally, we must renounce the common practice of writers on environmental law of acting as special pleaders for air pollution plaintiffs, lamenting whenever plaintiffs are not allowed to ride roughshod over defendants. The overriding factor in air pollution law, as in other parts of the law, should be libertarian and property rights principles rather than the convenience or special interests of one set of contestants.