Libertarian Law and Legal Systems Part Four – Liability for Wrongs
By Duncan Whitmore
The fourth part of our survey of libertarian law and legal systems will explore causative events of legal liability arising from wrongs – that is a breach of some obligation owed by one legal person to another without the necessity of a pre-existing relationship such as a contract.
There are two issues that demarcate the approach of a libertarian legal system towards wrongs as opposed to that of a contemporary legal system. First is the definition of a wrong and second is the standard of liability – that is, the point at which the defendant becomes legally liable for a wrong.
Libertarian Definition of a “Wrong”
In contemporary legal systems, a wrong is some sort of act on the part of an individual that is viewed as being subject to legal sanction. Unfortunately, we have to start off with such a vague tautology as, looking at the variety of acts that are subject to legal regulation, this is about as precise as we can get. In many cases, of course, the wrong will be some form of harm caused by one individual to another which serves as the causative event to generate a legal response.
“Harm” is very broadly defined and can include violent and physical inflictions, such as murder, serious bodily injury, or damage and destruction to property, all the way to more ethereal harms that may include nothing more than speaking one’s mind such as “defamation” and causing “offence”. However, events currently classified as legal wrongs needn’t have a victim at all as the act may either be wholly unilateral or take place between consenting individuals. As an example of the former we can cite nearly all offences related to drug possession and dealing, and of the latter the criminalisation of certain sexual practices owing either to their nature (such as in sadomasochism) or to the gender of the participants (i.e. homosexual intercourse). Basically, it is no exaggeration to admit that a wrong, legally defined, in our contemporary, statist legal systems means nothing more than some act that the ruling government or legislature doesn’t like and wishes to outlaw, to the extent that even quite innocuous behaviour may find itself being subjected not only to legal regulation but to criminal sanction.
As we outlined in part one, no legal liability is generated in a libertarian legal order unless the wrong, or the “harm”, consists of a physical invasion of the person or property of another – in other words, only those actions that violate the non-aggression principle are subject to legal regulation.
In most cases, a “physical invasion” (i.e. aggression) will consist of the defendant moving an object or a part of his body so that it strikes the person or property of the plaintiff. Stabbing, shooting, and shoplifting, for instance, are all examples of where there is an actual “clash” of matter between the two parties. However, physical collision is not a requirement in order for the defendant’s action to be classified as aggressive. Directing light beams, sound waves, microwaves, heat, or other forms of energy which lack a corporeal form can be just as invasive. The key is whether an action of the defendant interferes with the physical integrity of the plaintiff’s property, not whether anything controlled by the defendant makes actual contact. Thus, if the defendant possessed a magic wand which, when waved, could change the plaintiff into a frog then we would have no trouble in regarding this as an aggressive action. More realistically, if the defendant employed a device to suck all the air out of the plaintiff’s bedroom so that the latter suffocated in his sleep then this too would be aggressive, in spite of the fact that it consists of a withdrawal of matter rather than its imposition.
Actions falling short of the physical violation of the person or property of another are left untouched by the law and however unpleasant, unkind or distressing these may be one cannot use the force of law (i.e. legitimate physical force) to repel them1. A fortiori, there can be no legal wrong where the act involved has no victim, no other individual whose person or property has been invaded. Anything you do with your own person and property to yourself or to those others who furnish you with consent is no business of the law’s, however immoral or repugnant it might otherwise be. Moreover, actions which change merely the value or appreciation that one has for one’s property are, similarly, not aggressive. Let’s say that you possess a gold watch which, by virtue of its scarcity, commands a high price in the marketplace and, as a status symbol, brings you respect and admiration. If I steal or break this gold watch then I am guilty of an aggressive action. If, however, I flood the market with hundreds of other gold watches causing the price of your watch to plummet (with its value as a status symbol becoming equally negligible), then I would be guilty of no aggression at all. Similarly, if you possess what you think is a Rembrandt then this belief may serve to enhance the pleasure the painting brings you. If, however, I point out that the painting is a fake then the fact that you now regard it as worthless results in no legal impact upon me. Thus, the work of any professional critic, which can damage the reputation of an individual or firm by pointing out all of its flaws and failings, is, likewise, not aggressive. In fact, any attempt to ban the competing vendor of gold watches or silence the critic would constitute aggression against them.
By an “action” it is important to realise that we mean the deliberate choice to devote physical means in pursuit of an end – in other words, the action causing aggression must result from human volition. The mere movement in space of an individual’s body where he is acting as an involuntary automaton, or the movement of physical matter that is caused wholly by nature, are physical phenomena that are outside of the scope of social rules, and no individual can be held responsible for them. This realisation will become important when we discuss the issue of causation below.
There are certain actions that would appear to cause great harm (or have the potential to do so) but do not appear to be invading anyone else’s property. We can cite littering, driving under the influence of alcohol or drugs, and shouting “fire” in a crowded space as examples. The victimless aspect of these acts, however, is caused by the fact that they take place on public property that is owned by no identifiable individual. In a libertarian world however, where all property is privately owned, the property owner would demand standards of conduct (i.e. through contract if you are, for example, paying to drive on a road) and breaches of these standards would then be invasions of the property that could be subject to legal sanction. Indeed, as we saw in part three, contract is a method of preventing and apportioning responsibility for aggression where there is a pre-existing relationship between the parties.
An action which causes an invasion to another person’s body should be contrasted with an omission, i.e. a “failure” to act in order to prevent such an invasion. The latter should not normally be regarded as aggressive unless there is a pre-existing arrangement between the parties that compels one of them to intervene (such as a contract with a doctor to provide you with treatment). The difference between the two is that an action creates the very circumstances that are sufficient to cause the invasion; with an omission, however, sufficient circumstances are pre-existing – the defendant simply refuses to intervene so as to neutralise the danger to the plaintiff when it is within his power to do so. For example, if I deliberately poison your food then I am guilty of murder if the poison kills you; if, however, I refuse to give you food which I own and you subsequently die of starvation then I am legally responsible for nothing. If I hold your head under the water in a lake and drown you then, similarly, I am a murderer. If, however, you are already struggling to swim in the lake and I walk by without helping you then, once again, legal responsibility for your subsequent death could not be pinned on me. In certain circumstances the decision to withhold help may, rightfully, attract severe moral criticism, but the reason why omissions result in no legal liability (which, with a few exceptions, is the case in English law) is that the opposite would itself be aggressive. If there was a positive duty upon you to act then the law would be compelling you to divert your person and property from serving your ends to serving the needs of other people, a situation that amounts to forced charity. This differs only in degree from the kind of redistributive tax slavery under which we all languish now. The law should stop everyone else from forcing new problems onto you, but it should not allow you to simply offload the costs of your existing problems onto them.
Some libertarian scholars2 argue that the ambit of unlawful actions in libertarian legal theory extends beyond those that are physically aggressive. Most of what could be classified as unlawful actions under this approach concern some element of untruthfulness – slander, libel, false accusation of criminality, and spreading lies to devalue property in the hope of buying it cheaply, for instance. As will become clear when we discuss causation, it is the case that ostensibly non-aggressive actions can be regarded as ultimate legal causes of a physical aggression to the plaintiff’s property. However, it is not the case that there should be any basis for a legal response when physical interference to a plaintiff’s property is completely lacking and the plaintiff has suffered only some loss of reputation or value. Space precludes us from examining in detail the ultimate justification for equating unlawfulness with physical aggression, but a few words are warranted to illustrate why expanding the category of unlawfulness away from physical aggression is not only undesirable but also, most likely, not even possible in a free society.
There are many behaviours of individuals that contribute towards creating a free, civilised society – for instance, non-violent conduct, truthful statements, manners, displays of respect, use of a common language, a fidelity to law, culture, customs and traditions, and so on. In other words, an absence of aggression is not the only element required to maintain a peaceful society and, indeed, the motivation to avoid aggression is likely to be impaired if actions which are contrary to any of these elements become widespread. Hence, we can properly view such actions as de-civilising and disrupting behaviour.
As we argued in Part One, the cultural, customary and social elements peculiar to each individual society will be needed not only to determine the precise boundaries between actions which will be regarded as “aggressive” as “non-aggressive”, but also courts need to refer to them in order to interpret the meaning behind people’s actions. The adherence to, or breaking of, a social custom may, for instance, help to indicate whether a defendant’s intentions are, respectively, friendly or hostile in a given situation. If you are found at the dead of night on another person’s property dressed all in black, wearing a mask, and carrying a bag labelled “SWAG” then the court might not be prepared to believe that you accidentally “lost your way”. They might believe you, however, if you were dressed in clothes that are typically appropriate for the time and place. If courts need to draw certain conclusions from your language and from your behaviours then such a fact helps to incentivise good and “normal” behaviours indicative of conviviality while discouraging those that are bad and disruptive. It is not the case, therefore, that confining legal protection to aggressive acts fails to afford some indirect legal protection against other decivilising behaviour. Indeed, adhering to non-aggression on the one hand and to other civilised behaviour on the other are not hermetically sealed practices, and aggression often goes hand in hand with selfishness, deception and underhandedness, if not outright hatred and misanthropy.
However, it is also not the case that using the law to protect civilising behaviours directly is itself a civilising process – quite the opposite, in fact. Indeed, we don’t seek to protect the integrity of language through the law nor do we throw a man into jail for refusing to hold open the door for a lady, however upset she may be. One of the great achievements of libertarian scholarship has been to expunge from political philosophy vague, ethereal definitions of “harm” by separating the realms of objectively definable physical action from subjectively definable thoughts, feelings and values. The result is that force may be met with force while non-force must be met only with non-force. In other words, guns can be fought with guns, but bad words should be met with good words, lies with truth, bad manners with rebuke, errors with correction, etc. The symmetry and equilibrium provided by this stance has been a major civilising factor in society. For however well intentioned such an alternative approach may be – and however upsetting, distressing, and ill-motivated the non-violent actions of others may be in individual cases – you cannot, in the long run, get around the fact that seeking protection of the subjective appreciations of our minds with the objective presence of a barrel of a gun will create more problems than it solves.
Some of these problems are fairly obvious. If, for instance, we protect people’s property values and reputations from lies and falsehoods then should we not protect other intangibles also? Why not people’s self-esteem, their emotional well-being, their happiness, or their feelings in general? Surely injuring these things can be equally harmful, if not more so? And if we do decide to go down this route, then where are the boundaries within all of these things? How much value has to be lost? How much of a reputation? How much can we “upset” someone? How can we be sure to set boundaries at points where people are seeking redress for genuine grievances and are not simply co-opting the law into forcing upon everyone else what happens to be their own thoughts, feelings and values? Our current experience with the seemingly one-sided enforcement of so-called “hate speech” laws should show that they are a recipe for creating more de-civilising discord, disharmony and disunity than they could ever hope to dispel.
But even if we try to constrain legal regulation to the realm of protecting, say, reputations and property values from deliberate lies, this invites more subtle problems. For surely such an approach must presuppose that an individual is entitled to his pre-existing reputation or property value, i.e. the reputation he had or the value of his property before the lies were told? But how do we know this? Reputations and values can be built by lies as much as they can be torn down by them, and surely each of these effects can be as “harmful” as the other? Should it be the case, then, that fledgling competitors of a successful person can sue that individual on the grounds that his reputation is not deserved because it is built on commonly accepted falsehoods? Can failed authors sue successful authors on the grounds that everyone is wrong to believe that the latter’s books are more enjoyable? If not, and we maintain that the law should respond only to lies that harm reputations and values but not to those that enhance them then other people would still have to use their critical faculties to determine whether a given reputation is deserved in the first place. So why can they not do the same when confronted by a lie that is harmful to values and reputations?
The fact is that valuations and reputations are always based on a mixture of truth and falsehood, fact and legend, underappreciation and overappreciation, accuracy and inaccuracy, exaggeration and understatement. There is no way of determining objectively the reputation or valuation to which one should be entitled, and even a cursory glance in this direction similar to the one we have made here should indicate that a just enforcement of laws to protect them would be impossible.
A further problem with such laws is that they would replace the market regulation of truth with legal regulation. The discovery of truth and information, as well as the furnishing of veracity and authenticity, are not free gifts of nature. Investigation, analysis, verification and corroboration, are all costly endeavours that consume real resources, and so people have to decide for themselves how much time and effort is worth devoting to them in a particular instance.3 Legal standards of truth are determined by a specific need in a specific situation – that is, the need of enforcement agencies to be sure that they are justified in taking punitive action against a perpetrator. But there is no reason to believe that a standard of truth appropriate for this situation is what is required or desired in all other situations. Even a law which criminalised only deliberate lies to protect property values would probably result in the reluctance of anyone, at least in a commercial setting, to make any statement that could not be tested in court even if the statement was delivered honestly and there were listeners willing to accept a lower standard of veracity.4 People would, therefore, be de facto banned from learning and acting on information – rumours, gossip, cursory studies, etc. – that they were happy to accept as sufficient to inform their actions. Indeed, the irony is that a law designed to protect property values from mistruths could end up lowering property values if information of a sub-legal standard of veracity is what ignites demand in that particular market. Moreover, any proponent of such a law overlooks the fact that a liar has to risk own reputation when disseminating a lie, for surely the most effective threats to valuations and reputations come when the liar himself has some pre-existing credibility?
Finally, it is not even likely that a legal system would be able to protect property values and reputations from untruths if the latter are able to have such a profound effect, for such harms differ from physical harms in more ways than simple tangibility. In the case of physical aggression, the relationship is bilateral between aggressor and victim – the actions of a single person are sufficient to cause the harm. When discussing, therefore, how a legal system will operate in a free society, we can assume that the majority of people are committed to an array of general values that will help to ensure the legal proscription of such acts. Indeed, actual instances of deliberate aggression are likely to be limited to a relative handful of sociopaths and aberrations. This is not the case, however, with harms caused by lies to reputations and property values, because reputations and property values are produced by the minds of other people rather than being integral to the victim. Thus, such harms are multilateral – they require not only a liar and a victim, but also a plethora of other people to believe the lie before the victim’s reputation or property is devalued. But if it is the case that people in general are willing to base their judgments on shock, sensationalist, headline-grabbing lies while subsequently ignoring truthful rebuttals, then this probably indicates that society places a generally low value on truth in the first place. If this is the case, then we can no longer assume that society in general is committed to values that make a sound legal system a possibility. In our own society today, where our attitudes are very much influenced by Twitter tidbits, soundbites, and superficial analysis – and where the scientific method is being corrupted in favour of political ends (i.e. climate change) – legal standards are being concomitantly diluted. For instance, we have already had the selective removal of habeas corpus, erosion of the right to trial by jury in (at least) complex cases, and dilution of the standard of proof from “reasonable doubt” to one of simply being “sure”. This has been most egregious in rape and sexual assault cases where we can see that feminist political priorities seem to be gradually shifting the burden of proof away from the female accuser and onto the male defendant.
The point, therefore, is that a society that places a high value on truth and veracity is unlikely to require its legal system to protect it; a society that places a low value on truth is likely to find that its legal system is unable to offer such protection. You cannot just assume away certain characteristics from one part of society and then wish them back in again at another part. Indeed, it might be the case that a society in which untruths can reign unchecked is plagued by a lack of cultural cohesion, general antagonism and mistrust to the point where freedom is under threat anyway, if it exists in the first place. Thus, it is likely that we have to pre-suppose certain civilising societal elements in order to discuss how a free society will operate.
Before we proceed onto discussing the elements of liability, we should reiterate that we will not attempt to justify here the basis of self-ownership and private property upon which legal regulation in a libertarian world rests; rather we will merely assume it to be true and examine its consequences for a legal order.
Standard of Liability
The determination of the standard of liability – the extent to which a defendant should be held legally liable for his actions – is a difficult question.
The simplest approach is to view all actions that result in a physical incursion to the person and property of another individual as existing on a scale, with the least violent or aggressive actions sitting on one end while the most violent (such as murder and rape) occupy the other. Defendants would simply be liable according to the level of harm that they caused. If you cause a bruise, you are liable for a bruise; if you kill someone you are liable for a death; and so on.
However, such a simple classification overlooks the fact that the same outcome to the victim – death, for example – can result from a variety of incidents for which the defendant may be responsible. It may be anything from a deliberately and coldly calculated murder perpetrated by an evil and inhuman serial killer all the way down to an unfortunate accident that the defendant, while responsible, regrets bitterly and would never have dreamed of doing. Such accidents can result from not only the most innocuous of actions but also from a moment of absent-mindedness, a weakness which is extant in all humans. It is not likely that libertarian legal systems will categorise these two defendants in the same way. The first person is clearly a murderer whereas to apply this label to the second person would appear absurd. To hold these two individuals to the same standard of liability would not only be unjust but would also cause the legal system to fail to distinguish between those individuals who are (or otherwise have been) a deliberately engineered threat from other individuals from those who are not.
One solution to this problem is to recognise the difference at the remedy stage, so that once liability is established then the remedy can vary appropriate to the individual act of aggression5. Clearly this would be logical when considering the result that the act of aggression caused – a person who caused a scratch wouldn’t be liable for the same degree of compensation (or whatever remedy) as a person who caused a death. Yet when the outcome to the victim is the same – death, for instance – but the methods or motives of the defendant have been different, such an approach causes the confused situation where you have different responses to the same causative event. We do not, in our contemporary legal systems, label those who cause accidents “murderers” and then simply apply a lesser standard of remedy. Rather the murderers and manslaughterers incur liability on different grounds from those who merely cause accidents; different remedies then flow from the differing grounds of liability.
In part one, we stated that the purpose of legal rules and norms is to broadcast, publicly, the rights and obligations of each and every person so that conflicts can be avoided, or otherwise resolved. As “consumers” of the legal system, people will seek the services of jurisdictions where these demarcations are at their most clear so that they can carry on with their lives free from the aggression of others (and can, in turn, prevent their own actions from aggressing against anyone else). In a similar vein people are likely to require the legal system to accurately and specifically categorise those who do (or at one time did) pose a deliberately engineered threat to the person and property of others and those who do not. A prospective employer, for instance, might wish to think twice before hiring you if you deliberately killed someone, whereas if you were the unfortunate cause of a road traffic accident in which a person died then he may have some interest in knowing this but it may not make much difference to your chances of employment. Labelling all of those who cause death “murderers” (or, if we were to go the other way, just “tortfeasors”) would banish the benefits to be gained from this categorisation.
One objection to this might be that such a classification is very broad and there will be more detailed considerations to be accounted for within each category. Within the category of murderers, for example, will be cold blooded and unrepentant serial killers as well as those who acted in moments of passion and bitterly regret what they did. Shouldn’t these possibilities be recognised too?
The answer to this is that the law is qualified only to investigate what you have done and not what you might do; it is the purpose of the law to state that you, say, deliberately murdered someone or that you caused an accident. The law cannot say that you might go on to commit murder or cause an accident in the future. There will be extra-legal standards that might determine if a convicted murderer is still a threat to the public. A prospective employer might, for example, require a psychological test or some sort of guarantee from a sponsor or insurer before they hire the convicted individual.
However, based upon the approach of our contemporary legal systems and the likely requirements of those seeking justice, there is an arguable case that a libertarian legal system would still categorise the past acts of aggressors into something that approximates crimes and torts, with the individuals classified as criminals and tortfeasors respectively, if anything because this is what people, the “consumers” of justice, are used to. We must, however, remind ourselves that these are only suggestions as to how a libertarian legal system might operate. Everything we are stating here is only a speculation or projection of how jurisdictions might respond to the considerations they face in a libertarian way. There is absolutely no reason to suggest that some other way would not be more appropriate or would not be more likely to emerge in a libertarian society.
Anyhow, the upshot of all of this is that there is likely to be something more, some second aspect in addition to the simple results of the act of the defendant in order to determine the standard of that defendant’s liability.
This second element is the mindset (“mens rea”) of the individual defendant – what was in his mind at the time that the act of aggression occurred. Specifically, this means what did he intend for his action to accomplish? All voluntary actions are, of course, “intended” in that they seek to bring about a result from utilising specific means. What matters here, however, is whether the defendant intended the action to cause aggression on the one hand or whether the aggression was an unintended by-product of his action on the other. If the first possibility is true then the defendant, to use some legal jargon, acted with “malice aforethought” or “intention with deliberation”, and is held to a far higher degree of culpability than if the second possibility was true.
The broad classification of wrongs by our contemporary legal systems into criminal liability (crimes) and civil liability (torts) more or less reflects this. Crimes normally require a standard of criminal intent, a higher standard of proof (beyond a reasonable doubt) and what is viewed as a harsher remedy – usually imprisonment, but also, in some jurisdictions, capital or corporal punishment for those at the more serious end of the scale, and fines for those at the less serious. Civil wrongs, or torts, require no such intent, have a lower standard of proof (balance of probabilities), and the remedy is normally restricted to payment of compensation. By the far the most dominant area of civil liability for wrongs in English law today is the tort of negligence, negligence being a mindset far below deliberate intent to cause aggression. (In English law there is, however, a limited number of intentional torts that allow a victim to take advantage of the lower burden of proof in civil proceedings. As we shall see in part five, varying standards of proof will not exist in a libertarian legal order. Thus, it is likely that, in such an order, any intentional aggression will be classified as “criminal” while most cases of negligence and carelessness will be categorised as “tortious”, or with some kind of equivalent labels).
While this goes some way to resolving our problem the risk now we are taking now is that of deferring too much to the conduct of the defendant and not to the actual property invasion that was suffered by the victim – that legal systems will consider only how the defendant behaved and hold him responsible accordingly. Certain, limited circumstances such as duress (to the extent that the defendant could not wholly choose his action) may, of course, serve to partially or completely absolve the defendant from any liability. That aside, however, while the intended objective of a soundly sane adult’s freely chosen action should certainly determine the extent of his liability (i.e. the degree to which he should furnish compensation and/or submit to punishment), it should not absolve such a person from any liability whatsoever. While it might seem “unfair” to hold someone liable for an unintended consequence of his action, it is equally “unfair” to leave the victim, who had no involvement at all – either intentional or accidental – in bringing about the state of affairs, to bear his injury or loss uncompensated. In fact, as we shall see, any attempt to argue the contrary leads to absurdities.
All of this we will now explore in more detail by proposing a likely approach of a libertarian legal system to the question of liability, an approach that we suggest will consist of three stages:
- Can liability be established? In other words, did the defendant initiate an action that was the cause of the physical invasion of the person or property of another?
- If so, what is the extent of the liability? In other words, what was the result of the invasion to the victim?
- What is the nature of liability? I.e., did the defendant intend his action to cause an aggression or was the aggression an unintended by-product?
Establishment of liability in a libertarian legal system is likely to rest squarely on the simple occurrence of aggression – whether an individual, of his own volition, initiated an action that caused the physical interference with the person or property of another individual. Once the fact of aggression is established the question of law of whether the defendant is liable at this stage will be in the affirmative.
The courts will have to determine precisely which actions are and are not aggressive. As we mentioned in part one there are plenty of innocuous physical actions – such as emitting light beams, radio waves, sound waves, even fumes and odours – which, under a certain level of intensity would not ordinarily be categorised as invasive or aggressive. It must be remembered that physical invasions only give rise to social rules so long as they create a conflict and conflicts exist only in the minds of the participating individuals – they are not determined by scientific analyses of colliding matter6. People could invade and physically interfere with each other as much as they liked if they had no problem with it. It is only because such interferences prevent one of the parties from pursuing the fulfilment of his ends that rights come into being and their enforcement is sought.
The test in more difficult, threshold or novel cases, therefore, is likely to be whether the invasive action prevented the plaintiff from pursuing his ends with the property that was subject to the invasion. The court may need to interpret the actions of the victim in regard to the invasion to determine this. Did he, for example, have to interrupt his operations? Did he appeal to the defendant for a cessation of the aggressive act? The courts are likely develop rules in order to accomplish this. They may say, for example, that after a certain period of time a person carrying out a physically invasive act without any complaint gains an easement title and is permitted to continue the act. Nevertheless the courts would have to look at all the facts in each case in order to determine whether the plaintiff’s action is genuine, and it is not likely that any one particular factor will override all others in every single case. They are also likely to look at whether the conflict perceived by the plaintiff was reasonable given the social, customary, and conventional context. For instance, if a neighbour objects to particles of dust blowing off from my laundry hanging on the washing line onto his property then, even if he regards this as a personal annoyance, the court may hold that he, rather than I, should bear the burden of this unusual sensitivity if the rest of the community has no problem with my laundry.
With some important exceptions we will examine below, the initiation of the invading act must be implemented by an intervening act of will of another legal person. A movement of matter that is not initiated by human volition is not an aggressive action but merely a phenomenon of nature from which people must protect themselves with technology rather than social rules (such as umbrellas to protect against the rain).
While in most cases the intervention of will is likely to be contemporaneous with the act of aggression – A punches B; C stabs D, and so on – there is no requirement for it to be so. I may leave the hand brake off on my car and half an hour later the car rolls into someone; I may build a house with insufficient foundations that many years later collapses and falls onto another person. In both instances I would be liable.
It is here that we need to clear up, once and for all, an important confusion concerning the liability of the owners of property that invades that of another, liability that is easily misunderstood. It is true that, in such cases, the owner of the invading good may be prima facie liable for its physical invasion of another person or the latter’s property. However, this fact owes itself to procedural or investigative convenience, and nothing about ownership of an invading object per se attracts legal liability. Given that all original ownership derives from acts of deliberate, physical homesteading which transform an object from one good into another (in other words, any dangerous condition of owned goods is most likely the result of the intervention of will of the owner), and that any current owner as a voluntary successor in title has nearly always assumed full responsibility for this condition via the terms of the contract of exchange, it makes sense to look to the owner of the invading object first in order to find the culprit. In other words, the owner of an object is the most likely person to have used it last and to have used it to have caused the invasion. If your knife is found at the scene of a murder then it is obvious that the police will look to you first – you, as the knife’s owner, are the “prime suspect”, if you like. However, where there is clearly a further intervention of will by another individual then it is this latter individual who becomes liable, and whoever owns the invading object is irrelevant. For example, A uses B’s knife to stab C. Neither A nor anything that belongs to him may ever actually touch C but it was the intervention of his will, beginning with his own body, that initiated the act of aggression, not that of B. B’s knife was merely the intermediate tool that was used to fulfil the act and simply because B owned the knife does not mean that he should be liable for the stabbing, barring some special circumstance7. In cases where someone stabs you with your own knife then the aggression has already begun when the attacker assumes physical control of your knife. In short, the law responds to actions, not ownership – you are liable for what you do, not for which you own, and you are only liable for property that you own to the extent that you have committed an illegal act with it (or voluntarily assumed the responsibility for acts of another with the said property).8
The exceptions to the requirement of an intervening act of will are likely to be as follows:
- Agreements between two or more parties to commit crimes (i.e. contract killings and joint enterprises);
- The use of innocent, unwitting intermediaries (including the victim himself) to cause an act of aggression;
- Incitements to commit acts of violence.
The first is uncontroversial. Where two or more parties agree to commit a crime then it would be a travesty of justice for only those who commit the actual portion of the crime that is an act of aggression to be held liable. For instance, it would make no sense if a gang committed a bank robbery and only those who actually burst into the bank guns blazing were liable for the theft while the person who drove the getaway van (or the mastermind still at home) did not. Instead, the entire enterprise should properly be regarded as one, long act of aggression so that each party becomes liable for the whole, at least to the extent that the specific acts of aggression do not proceed beyond what was agreed.
The second instance, the use of innocent intermediaries, is also not difficult to understand. For instance, B poisons a bowl of soup that is prepared in a restaurant kitchen that is then delivered by a waiter to C; C consumes the soup and subsequently dies. The action of B was not the final, intervening act of will that led to the death of C. The waiter had to first deliver the soup to the table and C had to voluntarily consume the soup, yet B is clearly the criminal who should be held liable. As we shall see in more detail below when discussing causation, where the actions of others are part of the natural and reasonably expected progression of events – i.e. a waiter can be expected to deliver the poisoned soup to the table and C can be expected to consume it – the liability of the perpetrator cannot be absolved. For here, the actions of others simply become part of the means of B causing the death of C. The fact that the waiter and C himself were unwittingly complicit through their actions is irrelevant. If, on the other hand, the waiter had thrown the bowl of hot soup into C’s face then the waiter would be fully liable for the scalding injuries to C as this was a new, and reasonably unexpected intervening act of will on the part of the waiter that had nothing to do with B’s action. Another example is if B plants a bomb in C’s car which will detonate when C starts the ignition. Even though C’s act of turning the key in the ignition is the final, intervening act of will that causes the explosion, it is part of the reasonably expected course of events, and B should be wholly liable.
The third exception – incitements to violence – is more controversial given that the legal prohibition of such incitements appears to be a violation of free speech, and that willing acts of aggression are committed by the incited party. Nevertheless, there is a case for an individual to be held liable where he demonstrates objectively an intention that the incitement should be acted upon.9 For in this situation the perpetrator is merely using other people as a tool to obtain his ends violently instead of doing it himself. His objective is the same as anyone else who commits violence personally – it is merely the means that are different, and you cannot escape liability simply because your means consist of other people. It could be said also that one does not enjoy the right of free speech (i.e. self-ownership of one’s mouth and vocal chords) in order to persuade people to commit acts of violence in the same way that one does not enjoy the right to self-ownership of one’s fist in order to punch someone in the face. Moreover, in many cases the difference between incitement to commit violence and agreement to do so may not be entirely clear. On the other hand, genuine discussions about the morality of using violence in hypothetical situations would not attract liability, and to that extent the right of free speech would be protected. However, if such liability is found to exist within a libertarian legal system then it is likely to be applied only in the most narrow and specific of situations where both the intended perpetrators of the violence and the victim(s) are clear. Vague and general statements, or abstract, artistic depictions of violence are unlikely to suffice, however influential they may prove to be. Prohibitions of incitement are there to protect victims against people who are too cowardly to do their own dirty work; they cannot be used as vehicles for criminals to offload their personal responsibility onto someone else who happened to “influence” them in some vague manner.
This brings us to the matter of causation and the question of whether or not the wilful initiation of an act by the defendant caused a physical invasion of the person or property of another. It is necessary to discuss this aspect at length as Austro-libertarianism is uniquely placed to approach the matter in a sound and coherent manner.
The concept of causation is a confused and problematic area in contemporary legal systems. Simple, factual analyses of causation are insufficient for the reason that anything and everything could be held to be an operative cause of what followed. If we are talking about injuries from a car accident, then there are a variety of causes – the fact that the victim was driving on the road at the time; the building of the road in the first place; the fact that the car was constructed. Indeed, the creation of the Earth is a cause of all accidents that have ever happened throughout human history. Every aggressive action is also, in some way, aided by the laws of physics. If B pushes C off a cliff then, technically, it is gravity that causes C to fall to his death. Why is B’s act of pushing a more important cause than the phenomenon of gravity? Moreover, strictly speaking, we must recognise that there is no such thing as a watertight “factual” analysis of what caused what. Factual determinations of causation result only from controlled experiments where we have the ability to repeat the situation and hold all variables that are not under consideration constant. Hence, by altering the input of the variable being tested, we can reasonably conclude that any change in output is caused by the alteration of that input. Such a method is not possible when examining the facts of legal cases. The aggressive act has happened and the victim is left with the harm. We do not have the luxury of undoing it, recreating the situation and seeing what results if we vary a single factor of input. All enquiries of causation, therefore, require hypothetical projections of what would have resulted had the circumstances varied.
Every legal system therefore needs to employ some kind of mechanism that sifts mere factual causes into legal causes. In our contemporary legal systems, these tend to include concepts such as “proximate cause” or “remoteness of damage” in order to narrow the enquiry to the “relevant” cause and the extent of the damage for which the defendant will be liable. However, possessing no clear rationale as to what should generate legal liability, vague and malleable concepts such as what is “reasonable” to hold the defendant liable for wade into the fray, particularly when we consider situations where the number of resulting property invasions seem disproportionate to the initiating action. In other words, like a butterfly flapping its wings and causing a hurricane, the damage resulting is more widespread than would normally be the case, sometimes by setting off a chain of events – e.g. causing a spark that initiates a widespread fire, or knocking over a large object that falls into a wall that collapses onto a crowd. It may well be unfair to hold a defendant liable when the aggressive result has been completely out of proportion to anything imaginable when you consider his original action. However, legal concepts such as “remoteness of damage” often boil down to being little more than arbitrary policy vehicles for determining who should bear the loss (indeed, the famous English judge Lord Denning more or less admitted as much when ruling in Lamb v Camden LBC10). This approach of contemporary legal systems towards causation, through a mixture of factual and policy applications, can therefore result in everything and then, suddenly, nothing being blamed for the damage to the victim.
All of these problems largely disappear when we remember that what we, as libertarians, are interested in is the action of the defendant – action which, as we know from our understanding of praxeology, uses means in order to obtain definitive ends. We can determine what is and is not a culpable cause by using this praxeological framework, because what we are interested in is whether, from the point of view of the acting defendant, the means utilised by his action could be expected to create the circumstances that caused the invasion to the plaintiff. In other words, a libertarian court is trying to determine what an acting individual would have to do in order to bring about the invasive act – what we might term the praxeological cause. Let us now explore this in detail.
In the first place, it is suggested that the question of a causal connection between the wilful act of the defendant and the physical invasion of the victim should be considered first, separately from the effects of the said invasion which should be left until the second stage we outlined above (the extent of the liability). In the present stage we are trying to establish whether the defendant’s action invaded the person or property of another, i.e. whether he is liable. The extent of the loss or damage caused by the invasion concerns how much he is liable. The two questions are, of course, closely related as all actions are designed (or can be expected) to create effects of a certain extent, and the more extensive the effects of an invasive action the more likely it is that a typical plaintiff will regard the act as aggressive. Indeed, libertarian courts may not maintain an active distinction between the two questions (in the same way that “remoteness” is discussed in terms of extent of damage). There are, however, some important considerations concerning the extent of the damage caused to the plaintiff that, for our purposes here, are best left for a separate analysis. Here we will concern ourselves only with whether the action of the defendant caused an invasion of the person or property of the plaintiff.
The three-stage test that should be applied is that the action of the defendant must be:
- A factual cause;
- A praxeologically necessary cause; and
- A praxeologically sufficient cause
of the invasion to the plaintiff’s person or property.
The first stage, “factual cause”, is essentially the “but for” test of causation – whether the invasive act would have occurred “but for” the act of the defendant. Obviously if the invasion to the plaintiff would have occurred regardless of the defendant’s action then the latter cannot be held liable as any action of the defendant merely correlates with and is coincidental to the invasion to the plaintiff (although the defendant may be held liable for an attempt, as attempts are simply acts of aggression that have been initiated but have subsequently failed to achieve their objective – perhaps, for example, as the result of an act of self defence by the intended victim).
A defendant will not be absolved on this ground if he attempts to demonstrate that his act of aggression merely superseded another act of aggression that would have occurred anyway. For instance, let’s say that A and B (independently) want to poison C. A poisons C’s dinner while B poison’s C’s glass of wine. If C takes a mouthful of food first and promptly dies before having had an opportunity to touch the wine, it is no defence for A to say that, had he not poisoned the food, C would have died anyway from B’s poisoning of the wine. It is the causes of the actual aggression that occurred that are relevant.
Once this factual test is satisfied we move onto the second stage which is to determine the praxeological necessity of the cause. Here, the court needs to use its introspective knowledge of actions and how actions utilise the regularity of cause and effect in order to bring about certain outcomes. The question the courts will ask, therefore, is if a person in the defendant’s position aimed at bringing about the invasion to the plaintiff, would he regard the action taken as a necessary step in order to achieve that outcome? What we mean by this is if the defendant wanted to divert the course of history from one in which the invasion of the plaintiff could not be expected to occur to one in which it did occur, would he regard the action that he took as being a necessary step to create this diversion? By “necessary” we do not mean that it was the only action that the defendant could have to taken in order to bring about the aggressive result, nor does it have to have been the most optimal or efficient – merely that it was one of a range of possible actions which the defendant could have used in order to divert the course of events from a path on which the invasion to the plaintiff could not be expected to happen onto a path on which it did happen.
For instance, if B pulls a gun on C and fires it into C’s chest, then it is clear that any person in B’s position would reasonably regard the act of raising and firing the gun as a necessary act in order to expect the outcome of the bullet striking C’s chest to occur. We can ascribe to B the expectation that the bullet would not have entered C if he had not acted the way in which he did. If, on the other hand, B had raised a banana at C pretending it was a gun, and then, to everyone’s surprise, a bullet shot out of the banana and struck C, B’s raising of the banana was a necessary, factual cause of the bullet entering C. And yet so outside of the regularity of cause and effect was this outcome that it would be ludicrous to suggest that anyone in B’s position would regard raising a banana as a praxeologically necessary cause. Nobody who was intent upon actually killing C would raise a banana at him in order to accomplish that end.
Similarly, if a workman knocks a tin of paint off a scaffold and it strikes a passer by below then, once more, any person in the workman’s position can reasonably expect that the worker’s intervention was necessary to cause the tin to fall from the scaffold. If, on the other hand, he was whistling while he worked, and this whistling disturbed a bird which knocked the tin of paint off the scaffold then (ignoring the question of whether he made his workplace sufficiently safe), once again, although factual causation is satisfied, no person in the workman’s position who was intent upon causing the paint to fall off would regard the employment of the means of whistling at a bird as a necessary step in achieving that end.
Why is it just to acquit the defendant when his action was a factually necessary cause but not a praxeologically necessary one? The reason is because the law is there to proscribe actions, actions that have been freely chosen by an actor as a moral agent. But such actions are chosen only because there is a regularity of cause and effect that allows us to ascribe an expectation to the defendant that the said action will produce a certain result. Where the effect of an action is wholly outside the regular nexus of cause and effect then the defendant cannot be said to have exercised any moral choice vis-à-vis the result of the action, and so the resulting invasion to the plaintiff’s property should properly be regarded as an accident of nature.11 Indeed, this is precisely the reason why we do not hold children responsible for all of their actions even though those actions are freely chosen; for children have not yet gained the experience that gives them a firm enough grasp of the consequences of their actions.
In some cases, the defendant may have specific knowledge of cause and effect that would serve to render him liable, while another, less knowledgeable person would be acquitted for the same act. It is for this reason that the court needs to look at the situation from the position of the defendant. For instance, let’s say that, in the banana example, B had special knowledge that the banana was, in fact, a gun that was disguised as a banana. If this was the case we could then ascribe a praxeologically necessary causal connection between B’s raising of the banana and the bullet entering C’s body. Another example is if D alone knows that a container marked “Salt” actually contains cyanide powder, and he serves this to E; D would be responsible for E’s subsequent death. If, however, F had served the powder to E while believing it to be salt then he would not be liable for the death of E.
The final stage in determining causation is whether the defendant’s actions were praxeologically sufficient in order to cause the invasive act. In order to determine this test, the question being asked by the court is did the defendant do all he needed to do in order to bring about a reasonable expectation that the invasion would occur? In other words, did he divert the course of history from one in which the circumstances could not be expected to cause the invasion to the plaintiff to one in which the said invasion could be expected?
There are likely to be three types of occasion on which the act of a defendant invades the person or property of the plaintiff:
- The body of A physically invades the person or property of B – e.g. A hits B;
- The body of A physically moves an object that invades the person or property of B – e.g. A shoots a gun at B; A uses a knife to stab B;
- The body of A places an object in a position from which it later invades the person or property of B – e.g. A leaves the hand brake off on his car; the car rolls into the person or property of B.12
In categories one and two the motion of the defendant’s body directly and contemporaneously results in physical invasion of the victim to an extent that completely obliterates any likely counteracting circumstance, and so it is clear that the defendant did everything he needed to do in order to bring about the invasive act. For example, if B holds a gun to C’s chest and fires then there is nothing more that he needed to do in order to create in himself the expectation that the bullet would penetrate C’s body.
Matters become more difficult in the third type of case where the aggressive action may have to rely upon intervening causes before the invasion to the plaintiff is completed. For instance, gravity, or how slippery the road is, may largely determine where my car goes if I leave its hand brake off; the flow of the river will carry my boat downstream if its moorings snap; a chemical reaction may cause an object created by me to explode but only specific circumstances. It is here where the test of sufficiency may need to be examined in more detail.
The correct rule should be that, if the action relies upon intervening causes in order to bring about the invasion to the plaintiff, the defendant’s action will be deemed praxeologically sufficient if these intervening causes are part of the natural and expected course of events subsequent to the defendant’s action. As we saw above, this includes what other human beings can be expected to do (i.e. a waiter can be expected to deliver a bowl of soup to a diner’s table, and the diner can be expected to consume it). Where, however, a necessary, intervening cause is extreme or unexpected, the test of sufficiency will not be satisfied.
For instance, B leaves his car parked with the hand brake off; the car subsequently rolls down hill and strikes C. The fact that, in order to cause the invasion to C, gravity needed to drag the car down the hill will not absolve B as it was part of the natural, expected course of events. What if, however, the car rolls down hill but, half way down, is struck by lightning, and the debris from the lightning strike hits C? Here the earlier tests of factual cause and praxeological necessity are satisfied – it is obvious to everyone that the car would not be where it is but for B’s action. However, the lightning strike to the car was so far out of the reasonable and expected course of events that B’s action of leaving his handbrake off cannot be said to be praxeologically sufficient in bringing about the invasion to C.
Another example: D places a bowl of toxic mushrooms in his kitchen fridge; his wife, thinking they are edible, subsequently cooks the mushrooms and is poisoned. D is liable because, even though his wife had to physically cook and consume the mushrooms herself, this was a natural and expected outcome of her finding the mushrooms in a place of food storage. If, on the other hand, D kept the mushrooms in a sealed container in a laboratory, and it was in this latter location that his wife found them before cooking them, then D would not be liable as his wife’s use of material from amongst laboratory chemicals for culinary purposes was not a part of the natural and expected course of events.
In some circumstances, the test of sufficiency may only partly fail if the intervening act was probable or possible, but not reasonably unexpected. Here the courts may reduce the liability of the defendant without exonerating him completely. For instance, if E’s boat snaps from its insecurely tied moorings and is carried downstream before G pushes it into the path of another boat the courts may find both E and G liable for the invasion of the latter boat.
All in all, actions which fail to satisfy the test of factual cause are not causes at all, merely coincidences; factual causes which fail the test of either praxeological necessity or sufficiency will be held to be necessary preconditions rather than legal causes. In other words, they were mere circumstances which were required in order to bring about the invasion to the plaintiff, but they were not, from a legal point of view, a “cause” of it.
It should be clear that the tests of praxeological necessity and sufficiency take care of the problem of remoteness – that is, where the resulting invasions to persons or property seem wholly disproportionate to the original act. This is because there are only so many consequences that, praxeologically, a person can be reasonably held to expect to result from his actions, and so there will always come a point when these actions will recede to being mere necessary preconditions of invasions to plaintiffs rather than causes.
To illustrate this, let’s look at how some of the leading cases on remoteness may be handled by a libertarian court.
In Re. Polemis13, the defendant’s stevedores were loading cargo into the plaintiff’s ship. One of them dropped a plank from the deck into the ship’s hold causing a spark which ignited fuel vapour in the hold. The resulting fire destroyed the entire ship. The test of factual cause is clearly satisfied, but a libertarian court would need to examine whether the defendant’s act of dropping the plank was a praxeologically necessary and sufficient act in order to cause the fire. The answer seems to be no – the unique combination of the freak spark and the unexpected presence of fuel vapour would mean that no one in the defendant’s position who was intent upon causing the fire that invaded the plaintiff’s ship would utilise the means of dropping the plank. (The defendant was, in fact, held liable as the court, in this instance, dismissed remoteness as a factor).
Similarly, in the leading case of The Wagon Mound (No.1) 14, the defendant’s ship was moored in a harbour alongside a plethora of other vessels. The defendant permitted bunker oil to leak from the ship which floated on the water around the wharf. Welders working on repairs to a nearby ship discarded hot metal into the water which subsequently ignited the bunker oil, causing a fire which consumed the wharf and the ships that were moored there. On appeal, judgment was found in favour of the defendant on grounds very similar to that which a libertarian court would find. Although the act of discarding the bunker oil was a factually necessary cause of the fire, nobody (not even the plaintiff, it seems) reasonably expected such oil to catch fire in the water. Thus, as nobody intent upon causing the fire would reasonably employ such a means, there was no praxeologically necessary causal connection between the leaking oil and the resulting fire.
Causation and Intention
This tests of praxeological necessity and sufficiency in establishing causation covers both crimes and torts. As we outlined earlier, the likely difference between the two in a libertarian legal order is that, in a crime, the invasion to the plaintiff is the end that is actively expected and/or sought. For instance, B drives his car head-on at C in order to run him over. In a tort, however, the defendant seeks a different end from invasion to the plaintiff, and may not actively expect his action to result in the aggressive outcome at the time it occurs. However, because the said aggression is a reasonably expected outcome of his chosen action, he is liable. For instance, B may close his eyes for a few seconds while driving, and his car smashes into C. The end B seeks in closing his eyes may not be to cause any harm to C whatsoever – it may just be to rest his eyes for a moment. However, because driving while blind can be reasonably expected to cause a collision on the road, B is liable. Of course, in some cases the degree of carelessness or indifference to the consequences of one’s actions can be extreme enough to result in the imposition of criminal liability anyway (for example gross negligence and recklessness).
Some theorists15, in undertaking an analysis of causation, maintain a distinction between causation (i.e. factual cause) and what they call “intention” (i.e. the praxeological dimension – meaning here both deliberation and negligence). The distinction is, however, largely illusory because both causation and intention are part of the same, praxeological chain. A person acts because he expects the said act to cause the outcome that he intends (or, conversely, he refrains from acting because he expects the action to cause an outcome that he wishes to avoid). As we do not have mind reading devices we can only judge a person’s intentions by examining the outcome of his actions. The more likely it is that a given action will produce a certain result then the more reasonable it is to ascribe to that person either a specific intention to (crime) or a careless attitude towards (tort) bringing about that outcome. Thus, any examination of intention will likely have to examine causation anyway.
Hoppe, for instance, cites several examples where “causation” (factual cause) is satisfied yet the missing element in establishing liability is “intention”:
- A drives on the road. B jumps from behind a tree onto the road and is killed. A has caused B’s death;
- A invites B into his house. The house is struck by lightning, and B is injured. A (and his property) has caused B’s injury, for without A’s invitation B would have been elsewhere;
- A, B’s superior, sends B into the woods, hoping that B will be struck by lightning. His hopes are fulfilled. Has A caused B’s death or injury?16
However, in all of these cases, we only know that A lacks the required “intention” because his actions are praxeologically insufficient to cause the resulting injury to B. In the first example, A was not driving in a manner that was praxeologically sufficient to cause the injury to B. Rather, B had to take the (to A) reasonably unexpected step of placing himself in A’s path for the collision to occur. In the two lightning examples, A does, of course, put B in the position in which the latter is later struck by lightning; but in neither case has A done all he can do in order expect that the lightning strike will follow and injure B, and so we cannot ascribe to A the “intention” to harm B; rather, the lighting strike is an unexpected, intervening cause of B’s injury.
Hoppe then goes on to show how a modification to the final example will demonstrate the required “intention” on A’s part:
A is able to calculate exactly when a particular tree will be hit by lightning. He sends B to this tree, and B is indeed hit.17
In contrast to all of the previous cases, A has now done everything that is praxeologically sufficient for him to bring about the injury to B. For now, following his calculations, the lightning strike is a natural and expected consequence of his direction to B. Consequently, it is because of this praxeologically sufficient causal connection we can also infer A’s intention to kill B.
It is submitted here that it is conceptually clearer to leave any explicit examination of “intention” to determining the question of whether the invasive act was actively sought or expected by the defendant, or was a mere unsought by-product, even though the type of analysis outlined by Hoppe (and the outcome to cases) is likely to be the same. Once it is accepted that the defendant’s action was a voluntary one then the causal examination needs to concern what the defendant could, praxeologically, expect to result from the said action. What the defendant actually wanted, intended or expected concerns whether his act is criminal or tortious.
Innocuous or “Reasonable” Action and Liability
As we indicated, once (praxeological) causation between the defendant’s action and the invasion to the plaintiff is established then the defendant is liable. The effect of this approach is to hold an individual person wholly responsible for all of the voluntary actions of his body and those of his property. In other words, in each and every person’s action the risk that the said action may result in the invasion or aggression of another is borne by the initiator of the action. It is therefore the responsibility for each and every individual to ensure that his actions do not result in physical invasion.
In most cases this will be straightforward to accomplish. We generally have a high degree of control over whether our actions will in fact invade another individual. If I walk down the street it is not difficult for me to avoid bumping into other people; if I drive my car I don’t have to break into a sweat to avoid ramming it into other cars. We can usually go about our daily lives without even having to consciously avoid the physical invasion of other people. As each person possesses this degree of control we know from economics that placing the responsibility of aggression on its initiator will result in fewer acts of aggression as people seek to avoid its cost.
The difficult question, however, arises from situations where aggression results from actions which, while voluntary, are the product of either a) “innocent” and typical moments of absent-mindedness or human weakness, or b) circumstances where the defendant had to make a difficult choice. Examples of a) include knocking over your cup of coffee and scalding someone; a moment of distraction that causes you to run over a pedestrian; kicking a football in the heat of an exciting game so that it accidentally goes astray and hits someone; or falling asleep with a lit cigarette in your hand. Examples of b) include emergencies where the defendant has to balance the resolution of the emergency with the harm such resolution may cause to other people. For instance, if a man needs to rush his wife to the hospital in order to save her life he is likely to drive less carefully and at a higher speed than he would if he was just driving to the supermarket.
In all of these cases does it not seem unfair to hold the defendant liable for something that was initiated out of a moment of human weakness, common to us all, and/or out of simple, common behaviour which, but for the invasion of the victim’s property, would have been allowable? And is it not unfair to hold someone liable for his actions when he is, with all good intention, responding to a pressing need such as in an emergency?
Contemporary legal systems have developed mechanisms and doctrines to attempt to tackle this problem. In the English tort of negligence, for example, one of these is the so-called “duty of care” which attempts to narrow the field of actions where negligence gives rise to liability. In other words, there will be some situations in which you can be as negligent as you want yet you will not be held liable as you are said to owe no “duty of care” to a potential plaintiff. Indeed even negligence itself is a vague and arbitrary concept, permitting the courts to consider practically every aspect of the situation while possessing no rationale as to which should be given weight in order to determine whether the defendant was, in fact, negligent.
The problem with all of this is that in spite of the “innocence”, commonality, or good intentions of an action, the victim – the one person who definitely did not have any input at all into the aggressive action – is still left standing with the loss. With the acquittal of the defendant forgotten is the man with the broken leg, the widow with a dead husband, or the paralysed child. To absolve the defendant does not make these losses disappear; rather it simply shifts them from the defendant to the plaintiff. In spite (or perhaps because) of some of the complex formulae and procedures that have developed in the case law, legal mechanisms such as the duty of care appear to be based on little more than questions of whether the defendant behaved “reasonably” in his conduct or whether it is “reasonable” for him to be held liable for the full extent of the damage caused by his initial act. They therefore amount to little more than political vehicles as to who should bear the risk of loss from invasive actions that you initiate. The more restricted the liability upon you, the aggressor, the more the burden of risk shifts to the victim (or potential victims). In fact, it does not take an overwhelming degree of cynicism to conclude that the entire edifice of English tort law consists of little more than policy choices as to who should bear this risk hammered into the awkward shape of something that vaguely resembles principle and logic.
The correct position is that while the “innocence”, absent-mindedness, or “reasonableness” of behaviour that results in aggressive action may relieve one of severe moral chastisement, it is also true that the cost of these actions cannot simply be offloaded onto other people. Indeed, the contrary leads to absurdities and socially undesirable consequences. When it comes to our innocuous, absent-minded, but nevertheless, voluntarily chosen actions, we bear the risk of harming both ourselves and other people. If you harm yourself you (rightly) have to cough up the bill to correct the damage, so why should you not have to pay if harm from the same action results to someone else? As we know from “Austrian” economics, absolving liability in such actions would simply lower the cost of them to the perpetrator, and so their incidence will increase. People will be more absent minded; they will make less effort to remedy their personal weaknesses; there will be more carelessness, more “accidents” and, thus, more aggression and more damage. Indeed, it may even create the incentive to ensure that your minor, innocuous actions impact other people before they affect you. Rather than watching whether our own actions will initiate aggression we will forever be on the lookout to protect ourselves from everyone else’s.
Similar considerations apply to emergencies or any situation where the action is calculated but is, nevertheless, “reasonable” behaviour given the circumstances. If a man drives dangerously in order to get his wife to hospital as quickly as possible so that her life can be saved, it’s absurd to suggest that he can, in the process, simply create more emergencies by knocking over any pedestrian that gets in his way. Absolving the defendant in such a case means that everyone else can be forced to bear the cost of his priorities and his urgently desired needs. Consider also the situation where a person is put in the position of having to damage at least some property. For example, let’s say a boulder is rolling down a hill towards you and you have to deflect it into either property A or property B. Property A costs £5,000 to repair, property B £3,000. If you own both properties then your “reasonable” behaviour would be to deflect the boulder into the property which was cheapest for you to repair – property B. You would then have to fork out for the repairs. If, however, property B was owned by someone else then your choice is now between damaging your own property (A) at a cost of £5,000 or damaging property B and having to provide compensation to the owner at the cost of £3,000. Again, your “reasonable” behaviour would be to save property A and damage property B. In both situations your behaviour is reasonable but it would be absurd to suggest that, in the second scenario, you should not have to pay simply because the damaged property is owned by someone else. The contrary would mean that it would be legitimate to destroy other people’s property in order to protect your own and, indeed, there would then be an incentive to employ defence provisions that do just that.
This does not mean that each and every one of us is left permanently exposed to the possibility of being bankrupted through causing catastrophic loss to another person in our moments of innocent, absent-mindedness. Merely that the solution to this problem should not be a legal one. There will come a point when the degree of fault is so minimal and the total, societal cost (both monetary and psychological) of mitigating that fault so high that it will exceed the total compensation payable to victims of any resulting, aggressive actions. At this point, it would be possible for insurance companies to step in and offer premiums that are lower than the direct cost of preventing the aggressive action.
The Extent of Liability
Once physical aggression or invasion is established, the second question concerns the extent of the liability incurred. In other words, what harm or loss to the victim did the act of aggression cause?
The correct rule is that the defendant is to be held liable for the full extent of the loss that his aggression caused. Space precludes us from entering a detailed discussion of, say, consequential losses and premiums that might be placed on objects of sentimental value (i.e. so that the compensation payable exceeds the market value of the lost or damaged property). Suffice it to say here that the principle we are recognising is that if you cause a broken arm, you are liable for a broken arm; if you cause a death you are liable for a death; and so on. Precisely what is required to remedy those consequences is a separate question.
Once more the effect of this is to hold an individual person wholly responsible for not only the voluntary actions of his body but also for their effects. In other words, just as you bear the risk of invading others through your actions, so too do you bear the risk of what results from that invasion. And again, just as the fact of aggression itself is, for the most, part, easily controllable, so too will any harm that is caused likely to be in proportion to the extent of the action owing to the general regularity of cause and effect. If I bump into someone accidentally he may come away with a slight bruise or nothing at all; if, on the other hand, I punch him in the face he is likely to end up with a broken jaw; and if I stab him twelve times he is likely to die. Normally, therefore, each individual can control not only his actions in order to prevent aggression in the first place but also the precise level of harm that his aggressive actions will cause.
As we mentioned earlier, problems arise when your actions – invasive as they may be – produce outcomes that appear disproportionately severe compared to the action itself. This includes situations where the victim has an unusual or peculiar sensitivity to invasive actions, such as a weak heart that could go into cardiac arrest as a result of the most minor of aggressive acts. Is it not unfair to hold the defendant liable for such wild and unforeseeable harm that results from his act?
Yet precisely the same arguments that apply to the establishment of liability for “innocuous” or “reasonable” actions apply also to the extent of that liability. However remote, unlikely or unforeseeable the harm from resulting from your actions, simply dismissing it does not make the loss disappear but merely shifts its burden to the victim, the one person whose actions had no involvement in its cause. While it is within the economic interests of potential victims to protect themselves from the effects of aggression – particularly if they possess an unusual sensitivity – there can be no legal compulsion for them to do so and defendants should take their victims as they find them, warts and all. In legal parlance, this is known as the “eggshell skull” rule.
Alert readers may notice that earlier we stated that an usual sensitivity to innocuous acts would not serve to render that act as aggressive even if it was genuinely perceived by the plaintiff as an interference with his ability to enjoy his property. We also said, in our discussion of causation, a person is not liable if the aggressive result of his action was outside of the regularity of cause and effect. Why, therefore, do similar considerations not apply to the extent of the damage caused? Shouldn’t a person be liable only to the extent of how much damage can be typically expected to result from his action?
The answer is no, because the earlier questions concerned which kinds of acts are aggressive. A person has a duty to live his life by avoiding acts that are typically viewed as aggressive, and it is unreasonable to expect him to avoid behaviour that is typically not viewed as aggressive. It would be a travesty to hold him to account simply because one, single person with whom he happens to cross paths has an atypical view as to when another person has committed aggression against him; and, similarly, it would be unjust to hold him to account when the aggressive result of an action is reasonably unexpected. When, however, it comes to question of the extent of the damage caused by an action that is acknowledged as aggressive, the individual has already committed himself to a course of action which can reasonably be expected to result in aggression to another individual. Once the defendant crosses the threshold of committing an aggressive act – i.e. he has committed himself to having to pay some compensation – he cannot turn around and complain that the loss caused to the specific victim was disproportionate.
The typical, and correct, justification for this rule is that criminals and tortfeasors should not be able to rely on the vulnerability of their victims in order to avoid liability. For criminals this is obvious, because the peculiar frailty of the victim can actually make the criminal’s aim easier to accomplish and, indeed, many victims are targeted precisely because they are weak and vulnerable. Reducing the defendant’s liability on account of the victim’s vulnerability would simply make such vulnerability a more attractive target.
However, we also have to question whether there is any such thing as a “typical” victim in the first place. The elderly, for instance, are likely to suffer from more serious injuries from, say, a car accident than a younger individual. Should we regard the elderly as having “special” sensitivity, limiting their recovery to the kind of injuries that could be expected to occur if they were middle aged? And, by the same logic, shouldn’t victims who have an unusual resilience to injury be given more than is required to compensate them for their less serious injuries? Moreover, the eggshell skull rule applies not only to the physiological effects of aggression but also to the victim’s particular economic and social circumstances. For instance, if a woman knocked down by a car is rendered unable to care for her children, then the compensation payable would have to include provision for their care, a factor which would not apply if the woman was childless. Should we regard every possible victim as having the average of 2.4 children, awarding them uniform levels of compensation accordingly? In the same vein, if you smash your car into a Bentley should we pretend that you only hit a Ford Fiesta?
Further, a seldom discussed justification for the eggshell skull rule (as well as the heterogeneity of compensation claims generally) is that it helps to reduce the inclusion of compensation for injuries as an element in cost accounting. Let’s say, for instance, that a firm is able to calculate that the total annual compensation that it will have to pay for its torts is one million pounds, whereas it would have to pay two millions pounds in order to install provisions that will prevent those torts from occurring. Clearly, the rational thing for the firm to do is to permit the torts to occur and to compensate the victims. If, however, every tortious event is heterogenous, with each victim requiring different levels of compensation, it makes this calculation more difficult, and the more cost effective option may be to prevent the torts in the first place.
The typical “free market” response to this proposition is that it would make firms more “inefficient”, raising the cost of their products and making their customers worse off (indeed, some jurisdictions limit compensation payable as a form of regulatory privilege precisely to protect firms from “excessive” tort liability). This is true, but it does not follow that there is any gain in social utility from incentivising firms to avoid forking out for preventative costs. Compensation is designed to restore a victim as close as possible to the situation he was in had the tort not occurred, but it is still a second best option. A woman widowed by an industrial accident would most likely prefer to have her husband back than all the money in the world; a man who loses his leg in a car accident would, given the choice, prefer to have not lost the leg in the first place. As we cannot make interpersonal utility comparisons we are unable to conclude that the residual loss to the victim is in any way made up for by everyone else being able to buy cheaper stuff in the shops.
Once again the question of causation arises and it is here that we consider the second part of this area. Having established that the act of the defendant caused a physical invasion to the victim’s person or property we then have to establish whether the invasive act caused the loss or damage to the victim.
In most cases this will be extremely clear as the invasive act will be intimately and undeniably bound with the loss or damage in the same physical space. A shoots B and B bleeds to death; C’s car rams into D’s house and the house collapses, etc. In certain questionable cases it may be necessary for the courts to develop standards – for example, expert medical testimony from three independent sources – that will be sufficient in order to determine cause.
The difficult area is likely to be in situations where there are multiple, independent causes, each of which is sufficient to result in the damage and it is otherwise impossible to tell precisely which was the actual cause. For example, two shooters independently fire their guns at an individual and two bullets lodge in the victim’s body and he dies. Either bullet was sufficient to cause the death of the victim but it is not known which one. In the Fairchild case18 a number of previous employers of the victim independently exposed him to asbestos at various stages of his career; only one asbestos fibre was required to cause the illness that subsequently resulted in the victim’s death but it was impossible to determine precisely which fibre from whose premises was the cause. What should the courts do in these cases?
It is likely that the courts will hold each of the sufficient causes proportionately liable, the exact proportions to be determined by the individual cases. All tortfeasors will therefore share the burden of compensation at the remedy stage. Where one of the sufficient causes was an act of nature then the total compensation recoverable by the victim will be reduced accordingly.
One limitation for the victim, however, is that the liability of the defendants should not be joint and several unless the independent causes were actively co-operating. In the Fairchild cases, therefore, if one of the defendants was no longer alive or recovery was otherwise unavailable then the plaintiff could not recover that defendant’s portion from the other defendants. Victims always bear the risk that their tortfeasor may no longer be able to furnish a remedy for the harm done; this applies in cases of multiple independent as well as single tortfeasors. This was the outcome of Barker v Corus19, which was, unfortunately, reversed by legislation, at least to the extent that it applies to asbestos. The contrary would hold that, if four of the Fairchild defendants were no longer alive yet one was, then the victim would be able to burden that sole, remaining defendant with the entirety of the loss in spite of the fact that causation of the harm by that single defendant has not been established to the relevant standard of proof. What we have suggested here is an equitable compromise between three factors:
- Physical aggression on the part of the defendant being sufficient to cause the harm to the plaintiff;
- The harm being caused to the plaintiff by the actions of others;
- The lack of a strict, causal relationship to the satisfaction of the required standard of proof between the defendant’s act and the harm caused.
If, however, two tortfeasors are actively co-operating then, as we mentioned earlier, the courts will find their liability to be joint and several, and each is liable for the whole of the harm caused. The contrary would permit the deliberate scattering of liability. For example, one could hire tens of shooters to kill a person at the same time (increasing the likelihood of death) yet burdening the plaintiff with the cost of apprehending and trying all of them in order to gain his full remedy.
Nature of Liablity
Having established the fact of liability and its extent, the third consideration for the court will be to determine the nature of the liability. This centres entirely on the intention of the defendant to commit the act of aggression and the classification of that act into a crime or a tort.
It is likely that this categorisation will be extremely broad and will not have to enter detailed examinations of various modes of mind. Rather the sole enquiry is whether the act of aggression was deliberate and intended, or was accidental. If it was deliberate then it is likely that the courts will regard the act as criminal; if it was accidental then the act will be regarded as tortious (or something roughly equivalent to “civil” liability in our contemporary legal systems).
It is important to realise that intention is not necessarily identical to the motive of causing the specific act of aggression. For instance, let’s say you wish to blow up your house in order to fake an insurance claim. Your motive is the prospect of gaining the insurance money, but this requires killing everyone inside the house in the process. The fact that the latter was not the purpose of your act would not absolve you from liability for murder as those deaths were an obvious consequence of causing the explosion. The key test is likely to be whether there is an active expectation that the aggressive outcome would result. At the opposite end of the scale some extreme degrees of recklessness or wanton indifference to the consequences of one’s actions may be sufficient to establish criminal intent even though the aggressive act was not sought deliberately by the defendant.
A further likely consideration is whether the actual resulting harm (in addition to the aggression in the first place) was also intended by the defendant. A shoots B and B dies; C shoves D out of his way and D dies of a heart attack. Both A and C intended an act of aggression and are both criminally liable but it would be extreme to suggest that C – who probably intended no or little harm at all – should be classified as a murderer just like A. It is likely, therefore, that the court will recognise gradations of liability according to intention of outcome, such as degrees of murder. C, would, however still be liable to make compensation for the full consequences of his act of shoving.
The upshot of all of this is that, at the remedial stage, a tortfeasor is likely to have to furnish compensation only and can then go about his life unmolested. A criminal, however, having been identified as a deliberate threat to the person or property of others, is likely to face further sanction, punishment, or rehabilitation before his reintegration into the community. Such considerations are beyond the scope of what we can examine here.
By retaining this classification we do not, in any way, mean to state that “crimes” are offences against the public or against the state, and at no point does the state replace the victim in the prosecutorial process. We are merely suggesting that those who seek justice are likely to demand this distinction between wilful criminal conduct on the one hand and negligent/careless behaviour on the other.
Under libertarian theory, self-defence is the physical response by a victim to an invasive act in order to render that invasive act inert; the physical nature of the response – i.e. violence – is therefore legitimised. How are libertarian legal systems likely to handle this concept?
Let us say that A is the individual invoking self-defence and B is the person alleged to be carrying out an act of aggression against A. In order for self-defence to be validated, there must be an actual, objectively identifiable act of aggression being perpetrated by B. A’s subjective misinterpretation of B’s behaviour would not suffice. If B raises his arm to brush his hair and it was objectively clear that this was his intention but A, expecting B to deliver an imminent blow to his (A’s) body, shoots B then however much in good faith was A’s misinterpretation of B’s behaviour he would not be allowed to invoke self-defence, and A would be held fully liable as an aggressor.
If this latter view should be doubted, then consider the position of B. If he is raising his hand to comb his hair and this is the objectively valid interpretation of his action but A goes and shoots him, would he (B) be entitled to defend himself from this act and shoot A back? If the answer is yes then A is the aggressor and not the defender. If, on the other hand, B pulled back his fist at A in order to punch him and A shoots in response then B would be the aggressor and A the defender.
The act of aggression must have been initiated – in other words, it must be in the process of occurring. The mere possibility or even probability of an act of aggression does not suffice. While actual, physical contact does not yet have to have occurred, its occurrence must be the imminent or expected result of that which has already occurred. My raising of a gun at you ready to shoot would warrant an intervention of self-defence, even though I have not yet shot. Moreover, words that threaten violence will suffice if they can be interpreted objectively as a sincere statement of intent (even if the perpetrator only subjectively intended to bluff). Fully anticipatory acts of “defence” – i.e. shooting someone before he has a chance to initiate an act of aggression against you – are, however, not warranted. While someone may be unnerved by the presence of a peculiar or sinister individual, the lawful response to this is to prepare oneself for an act of defence in the event that an act of aggression is, in fact, initiated. This could include installing extra security devices on one’s property or carrying a personal firearm.
What is the permissible extent of the defensive act? Once again, in our contemporary legal systems, “reasonableness” rules the roost with “reasonable force” being the seemingly operative phrase. However, at least in situations where the aggression is sudden and surprising, victims do not have the time to judge the extent of the aggressive act and precisely which response is proportionate. Indeed, any such act – a robbery, a car hurtling towards you, a raised fist – whether deliberate or accidental may result in the death of the victim either immediately or after an escalation. It is submitted, therefore, that the correct approach is likely to be that any act that the victim deems necessary to bring the invasive act to a close will be a permissible act of self-defence. This may involve killing the invading person or destroying the invading object. On the other hand, what is not permitted is for the act of defence to harm innocent bystanders or for it to continue once the aggressive act has been rendered inert. So someone could not, for example, fire indiscriminately into a crowd in order to stop a robber from running away, nor could one, having shot and immobilised an intruder, proceed to take an axe and hack him to pieces. These would all be fresh acts of aggression that would themselves be liable to legal sanction.
1This does not mean to say, of course, that extra-legal methods cannot be used to discourage “harmful” actions that fall short of a physical invasion.
2See, for instance, Frank van Dun, Against Libertarian Legalism, Journal of Libertarian Studies, Volume 17, no. 3 (Summer 2003), 63-90; Id. Natural Law and the Jurisprudence of Freedom, Journal of Libertarian Studies, Volume 18, no. 2 (Spring 2004), 31-54. For responses to van Dun’s position, see Walter Block, Reply to “Against Libertarian Legalism” by Frank van Dun, Journal of Libertarian Studies, Volume 18, no. 2 (Spring 2004), 1-30; Id. Reply to Frank van Dun’s “Natural Law and the Jurisprudence of Freedom, 65-72; N Stephan Kinsella, Reply to Van Dun: Non-Aggression and Title Transfer, Id. 55-64.
3This includes, incidentally, problems such as the copying of trademarks or other forms of faking brand identities in order to increase the value of an inferior product. Apart from the fact that customers are likely to have a means of legal redress through fraud, misrepresentation or mistake, if such activities become a serious problem then they will be met by market pressure on firms to prove their authenticity, perhaps by the examination of products by independent third parties and/or investment in authenticating technology.
4It would be no answer to this to say that such statements could be still be issued, but with a disclaimer. For surely deliberate lies could also be issued with such a disclaimer yet could still be believed? Thus, any law protecting property values against lies would be rendered ineffective if they allowed for exclusion of liability in such a way.
5This appears to be Rothbard’s approach – see Murray N Rothbard, Law, Property Rights, and Air Pollution, Cato Journal 2, no. 1 (Spring 1982): 55-99, reprinted in Economic Controversies, 367-418 at 409-12; also Id., The Ethics of Liberty, New York University Press (1998), 144.
6Indeed, one of the criticisms of the non-aggression principle – that it would outlaw practically all behaviour that results in innocuous invasions, such as fumes from a car exhaust entering the airspace of your garden – is based upon this misunderstanding.
7Even if this wasn’t the case though, and if we always held owners liable for their own property (so that C should sue B and B could sue A) we must bear in mind that not only is the purpose of the justice system to enforce rights and obligations with clarity but that also the “production” of justice is itself a costly affair that consumes resources. It is likely therefore that a libertarian system will always take the shortest route and permit C to sue A directly, with B dropping out of the picture (unless B wishes to sue A for the theft of the knife). The same considerations would be true also with accidents that involve a chain of property ownership – A bumps into B which causes B to bump into C which causes C to bump into D. D is likely to be able to sue A directly.
8Cf. Konrad Graf, Action-Based Jurisprudence: Praxeological Legal Theory in Relation to Economic Theory, Ethics and Legal Practice, Libertarian Papers, Vol. 3, No. 19 (2011), 34-8.
9Walter Block, however, rejects the possibility of liability in such a circumstance. See Walter Block, Rejoinder to Kinsella and Tinsley on Incitement, Causation, Aggression and Praxeology, Journal of Libertarian Studies, Volume 22 (2011), 647-50.
10(1981) QB 625 at 636.
11Cf. Stephan Kinsella and Patrick Tinsley, Causation and Aggression, The Quarterly Journal of Austrian Economics, Vol. 7, No. 4 (Winter 2004), 97-112, at 98.
12Cf. Richard A Epstein, A Theory of Strict Liability, Journal of Legal Studies 2 (January 1973), 151-204.
13Re. Polemis & Furness, Withy & Co Ltd., (1921) 3 KB 560.
14Overseas Tankship (UK) Ltd v Morts Dock and Engineering Co Ltd., (1961) UKPC 2.
15Hans-Hermann Hoppe, Property, Causality, and Liability, The Quarterly Journal of Austrian Economics, Vol. 7, No. 4 (Winter 2004), 87-95; Kinsella and Tinsley, Causation and Aggression; both articles cite, approvingly, Adolf Reinach, Über den Ursachenbegriff im geltenden Strafrecht, in Id., Sämtliche Werke, Vol. I. München: Philosophia.
18Fairchild v Glenhaven Funeral Services Ltd., (2002) UKHL 22.
19(2006) UKHL 20.