Is risk a harm?

• Author: Finkelstein, Claire
• Position: Preferences and Rational Choice: New Perspectives and Legal Implications

INTRODUCTION

A person drives at breakneck speed down a small residential street, risking the lives of children playing. A doctor performs a dangerous operation under the influence of alcohol, imposing a significant risk of harming the patient. A sleep-deprived commercial pilot flies a plane full of passengers, thereby increasing the risk of an accident. In such cases, if the risk the agent creates eventuates, he may be liable for damages or subject to punishment, assuming that he has no legal justification for imposing the risk on others. But how should we think of the situation if the feared result does not eventuate? Should the actor who wrongfully endangers those around him without ultimately harming them be liable or punishable for the unjustified risk he creates?

Some would answer affirmatively, on the grounds that the driver, the doctor, and the pilot are as blameworthy for their wrongful behavior as they would be if the risk they created resulted in actual injury. They would say that an actor's blameworthiness is a function of what he takes himself to be doing when he acts, rather than of the resulting state of affairs his acts produce. (1) According to this argument, two individuals who impose precisely the same unjustified risk of harm are in equivalent moral positions, even if one ends up causing a person's death and the other does not. The explanation offered for this is that the difference between them is entirely a matter of luck. And, the explanation continues, since human beings are responsible only for the things they control, and since luck's effects are beyond their control, actors who make the same choices, but produce different results, are morally equivalent. (2)

Whatever the merits of this suggestion in the moral arena, its significance for the law is doubtful. (3) For blameworthiness plays a more minor role in judgments of legal responsibility than the proponents of the foregoing account would countenance. In general there is no compensation for blameworthy conduct that produces no damages. (4) Tort law does require the defendant's conduct to have been "wrongful," meaning that the defendant must have breached a duty to the plaintiff. But the duties in question need not be moral in nature, and thus we cannot assume that the wrongful actor is also a morally blameworthy actor. And in criminal law, where moral wrongdoing is obviously of greater significance than it is in tort law, blameworthiness is neither necessary nor sufficient for criminal liability. (5) In both tort and criminal law, then, blameworthiness is of questionable importance for determining liability.

There is, however, a different concept that is arguably of greater importance in both areas: the concept of harm. While both tort and criminal law recognize that there are harms that should not generate liability, harm appears to be at least a necessary condition for liability in both areas. (6) In tort, there can be no damages if no one has been harmed, since there is no basis for a civil plaintiff to sue a defendant for wrongdoing alone. (7) In criminal law the point arguably holds as well, despite the existence of a class of so-called "victimless crimes," which appear to prohibit conduct on the basis of pure moral wrongdoing. But the most defensible instances of victimless crimes are ones in which someone is in danger of suffering either harm or exploitaiton, despite the fact that he has consented to the activity in question. (8)

Given the importance of harm in civil and criminal law, we might ask a different question than the one about wrongdoing on which commentators normally focus: Is the imposition of a risk a harm to the person on whom it is inflicted? In this Article, I attempt to defend an affirmative answer to this question. While some legal scholars have found this position intuitively plausible, (9) none has to my knowledge attempted to defend it in any detail. My argument does not, by itself, entail a position on whether defendants who expose others to unjustified risks should be held liable or punished for doing so. (10) To say that risk is a harm is not to suggest that it is a harm the law will countenance--that would be a further claim. (11) It does, however, suggest that it would not be incoherent for the law to mandate compensation or punishment for mere risk infliction, assuming, as I suggested above, that harm is a necessary condition for liability.

The heart of my argument is that agents have a legitimate interest in avoiding unwanted risks. (12) A person who inflicts a risk of harm on another damages that interest, thus lowering the victim's baseline welfare. This account distinguishes between risk-based harm, which I call "risk harm," and ordinary, tangible harm, which I call "outcome harm." I claim that risk harm is a form of harm that is independent of outcome harm, on the grounds that minimizing one's risk exposure is an element of an agent's basic welfare. "Real" harm, in other words, is not limited to outcome harm.

As a corollary to my argument for the existence of risk harm, I also argue that a chance of benefit is itself a benefit. And I distinguish the existence of what we might call "chance benefit" from "outcome benefit" by analogy to risk harm and outcome harm. I refer to my claims about the notions of harm and benefit as the "Risk Harm" and the "Chance Benefit" Theses respectively. In garnering support for my approach to harm, I draw freely on the benefit side when the intuitions in support of the thesis seem stronger. That is, I treat support for the general approach to risk and chance as available either from the harm or from the benefit side. I do not see any basis for thinking risk and chance would be any different with regard to the question at issue.

In Part I, I attempt to make the intuitive case for the Risk Harm Thesis. I also address some important difficulties with the idea, and consider the shape it must take in order to meet those difficulties. In Part II, I consider the law's treatment of compensation and punishment for risk alone. As we shall see, support exists in the case law for the proposition that a risk of harm is in and of itself a harm. For if, as I argue, the law treats harm-imposition as a necessary condition for torts and crimes, and if the law demands compensation or punishment for risk imposition, this signals a recognition that risk harm is itself a harm. In Part III, I address several important objections to the thesis that risk is itself a harm. While these objections are significant and expose a number of difficulties with the idea, I argue that the objections are ultimately unconvincing. I conclude with some observations about the implications of the Risk Harm Thesis for the law.

1. CHANCE BENEFIT AND RISK HARM

   The Risk Harm Thesis suggests that exposing someone to a risk of harm itself harms him. That is, exposure to risk entails a reduction of an agent's welfare, regardless of whether the risk eventuates in outcome harm. This is obviously not to say that outcome harm is irrelevant to addressing the harmfulness of risk. Without the possibility of outcome harm there is no risk. But it does suggest that the harm a person suffers by being exposed to risk does not evaporate the moment it is clear no outcome harm will result. This idea may seem paradoxical: How can we say that the reason a person is harmed by being exposed to risk is that he has an increased chance of suffering outcome harm, but that the harm persists even once it is clear no outcome harm will result? Nevertheless, I claim in this Part that this is indeed the case. It may help to prepare for this somewhat odd suggestion by considering the benefit side, namely the claim that the chance of benefit is itself a benefit. For the analogous claim on the benefit side strikes me as easier to accept.

   The Chance Benefit Thesis suggests that an agent's increased chance of receiving a certain outcome benefit is itself a benefit to the agent, such that the agent is better off for having had that chance than he would be if he had never had the chance at all. It is easy to see that a person regards himself as better off for having a chance of winning some outcome benefit before it is determined whether the chance has eventuated. But the question is whether the chance of winning is itself a benefit, apart from whether the agent actually wins the outcome benefit. If we accept this idea, we must also accept that if the agent does not win the outcome benefit, he must still regard himself as better off than he would have been if he had never had the chance of winning the outcome benefit in the first place. On this view, two agents, both of whom end up with ten dollars, would nevertheless be in different positions depending on the history by which they came to this state of affairs. The person who had a fifty percent chance of winning a million dollars but who did not win, ending up with ten dollars instead, would be better off than the person who merely had ten dollars, without the chance of winning more. The former received a chance benefit, the value of which persists even after the chance of winning the outcome benefit no longer does.

   But why should we think of the person who had a chance of winning a million dollars as better off than the person who had no such chance, if, in fact, he did not win? In what sense better off?. The answer is that the person exposed to a chance of winning has received an opportunity the other person did not have. That opportunity is a benefit he would have been willing to pay something to secure. Of course a rational person is only willing to pay for the chance of benefit ex ante--before the outcome has been determined. No one would pay for having been exposed to a chance of winning in the past that did not work out, just as no one would pay for the memory of having seen a great opera or eating a fine meal.

   To see that the benefit does not entirely disappear with the...