Navigation
Newsletters Index (06:1)
apaOnline Home Page

APA Newsletters

Fall 2006
Volume 06, Number 1

Newsletter on Philosophy and Law

Articles

Harm and Homicide

Christopher Knapp
Binghamton University, State University of New York

No act seems more deserving of its criminal status than homicide.1 Explaining what makes it deserving of that status, however, turns out to be a complicated matter. Accounting for the evil of its essential component—the victim’s death—has troubled philosophers since Epicurus. And settling on what distinguishes those killings that deserve criminal sanction and those that do not is at the heart of some of the most pressing issues in law and policy—the legal status of abortion, euthanasia, the death penalty, innocent bystanders to war, and members of other species. Among the very best places to turn for insight into these matters is the work of Joel Feinberg. The breadth, insight, clarity, rigor, and balance of his writing on the legal issues surrounding life and death has few rivals.

There is a theoretically and practically important issue concerning homicide that Feinberg’s work bears on, but that Feinberg himself never wrote about explicitly. In order to see the issue, consider the following example. Imagine two men, each of whom is convicted of killing one of his siblings so as to gain a greater share of an inheritance. One of them killed his twenty-year-old ambitious and talented brother. The other killed his fifty-year-old unmotivated brother, who is of decidedly ordinary abilities. Otherwise, their crimes are similar—both men killed using the same means, inflicted the same amount of pain in the process of killing, premeditated the killing to the same degree, stood to gain the same amount from the death of their sibling, regret their actions equally, and so on. It seems fitting to say that, whatever punishment these killers deserve, it should be the same. More particularly, it seems that their punishment should not differ in virtue of the characteristics of the differences in the age, ambitions, or talents of their victim. We would find it both odd and unpersuasive, for instance, if one of their lawyers were to argue that her client’s sentence should be lightened relative to the other’s in virtue of whom her client killed.

Jeff McMahan has called attention to a feature of commonsense morality that provides a straightforward explanation of these intuitions. Commonsense morality, McMahan claims, holds that "the extent to which it is wrong to kill a person is unaffected by facts about the victim other than those that may make him or her liable to be killed, or deserve to be killed, if indeed it is possible for a person to deserve to be killed—that is, facts other than those that may make him non-innocent in the relevant sense."2 If this is right, we can ground our intuitive sense that our fratricidal murderers deserve the same punishment in our moral commitments. Since their killings differed only with respect to age, talents, and ambitions of their victims, their killings were equally wrong. Further, since these characteristics of the victims are not plausibly relevant to the criminals’ culpability for their actions, their killings are equally blameworthy. Finally, since wrongdoers deserve punishment proportionate to the wrongness and/or blameworthiness of their actions, the killers deserve the same punishment.

This is hard to square with harm-based theories of criminal punishment, however. According to such theories, the appropriate punishment for a crime is at least partly a function of the amount of harm the criminal caused or intended to cause in committing the crime; the greater the harm caused or intended, the stiffer the punishment should be. But it is natural to say that some deaths are worse for their victims, and hence harm them more, than others. All else being equal, dying at twenty is a worse fate than dying at fifty. Losing the opportunity to complete projects from which one derives meaning is worse than losing years of indolence. Being denied a life promising success seems worse than being denied a life that promises failure. Consequently, if we hold all else equal, murdering one’s older and unmotivated sibling of ordinary abilities causes less harm than murdering one’s younger, more ambitious, and talented sibling. On harm-based theories of punishment, then, the former murder warrants less punishment than the latter.

So, given some plausible assumptions, harm-based theories of punishment conflict with common sense regarding the punishment our fratricidal murderers are due. The issue, then, is simply who is right. Is it appropriate to give some killers more or less punishment in virtue of how much the victim stood to gain from the rest of her life?

The remainder of this essay develops an answer to this question. I begin by showing how Feinberg’s account of the harm of death and his theory of punishment lead to the recommendation that killers be punished in proportion to the harm they cause their victims in virtue of ending their lives. I then turn to the question of whether the commonsense rejection of that recommendation can be justified. Here I focus on McMahan’s recent proposal that the commonsense view is a consequence of two commitments: first, that killings are wrong in virtue of failing to respect the worth of the victim; and second, that the worth of all people is equal. The prospects of justifying the commonsense view by appeal to these commitments, I argue, are dim. But McMahan’s proposal does suggest a psychological explanation of the intuitions in conflict with Feinberg’s theory. Moreover, this plausible explanation reveals these intuitions to be understandable mistakes. Given this error theory, and the difficulties of constructing a theoretical defense of the commonsense position, I conclude that we have good reason to follow the implications of Feinberg’s theories and make the punishment for homicide sensitive to those characteristics of the victim that bear on how much she loses in virtue of dying. I end with some suggestions of how we might incorporate this recommendation into current sentencing law.

I. Feinberg on Death and Punishment

Feinberg held, along with commonsense morality, that the amount of harm a person suffers from death is not constant. In general, he believed, death harms the one who dies by "setting back" or frustrating his interests.

To extinguish a person’s life is, at one stroke, to defeat almost all of his self-regarding interests: to ensure that his ongoing projects and enterprises, his long-range goals, and his most earnest hopes for his own achievement and personal enjoyment, must all be dashed.3

And the number of unfulfilled projects, enterprises, and hopes for the future varies from time to time and from person to person. Consequently:

The degree of harmfulness of a person’s premature death thus depends on how premature it is, given the interests that defined his own particular good. …Almost everyone will die with some interests that will be defeated by his death. …But the person who will die at thirty is in a condition of greater harm on balance…than the person who will die at eighty. That is because death defeats fewer interests, and especially fewer important interests, of the latter than the former.4

This result should not be surprising, since Feinberg rightly treats life as being of only instrumental value to the person whose life it is—our continued life is good for us in virtue of being a precondition of pursuing the goals we set for ourselves, the achievement of which, in turn, constitutes our good—and the instrumental value of a thing typically varies from time to time and from person to person. (For those who are tempted by the thought that their life is intrinsically valuable to them, compare the following ways one’s life might end: in the first case, one dies in one’s sleep on one’s eightieth birthday; in the second case, one week before one’s eightieth birthday one falls into a coma that lasts many years, and from which one never awakens. No one, I suspect, will think that the second ending would be more in their own interest than the first, despite the fact that it involves many more years of life.) When we keep in mind that the instrumental value consists in its being a precondition for the pursuit and attainment of objectives in which one has a stake, moreover, it is clear that life is of greater value to those who have more to live for, where how much we have to live for is partly a function of the strength and number of our future-oriented goals and our prospects for achieving them. Death constitutes a greater setback to an ambitious and talented person, therefore, than to a person who has only a comparatively weak investment in the future and has little prospect for success, since the prospects for future fulfillment for the former person are greater than for the latter.

There are two ways in which this account of the harmfulness of death can be combined with Feinberg’s theory of criminalization to reach the conclusion that the penalty for murder should vary according to characteristics of the victim such as her age, ambition, or talent. This is because, for Feinberg, there are two senses in which a particular criminal penalty can be appropriate. The first sense is derived from the "ultimate justifying aim" of the criminal law: deterrence. In particular, Feinberg endorses a version of the harm principle, according to which:

It is always a good reason in support of penal legislation that it would probably be effective in preventing (eliminating, reducing) harm to persons other than the actor (the one prohibited from acting) and there is probably no other means that is equally effective at no greater cost to other values.5

This principle can ground an argument for there being good reason to pass a law that made the penalty for murder sensitive to certain intrinsic attributes of the victim, since such a law might well reduce harm. Imagine, for instance, a law mandating that the younger an adult murder victim was, the stiffer the penalty the murderer would receive.6 In those not infrequent cases in which murderers choose between a number of potential victims, the more severe potential penalty for killing younger people should incline them to choose older victims. Ceteris paribus, the harm that these murderers cause by killing would be reduced. Similar gradations in punishment could be used to track other dimensions along which the harm of death typically varies, to similar effect.
Of course, it is possible that the harm reduction such sentencing laws promise could be offset by some other effect of the law. Perhaps, it could be argued, the imagined law would give some citizens the impression that killing older people is not such a bad thing to do, and hence would lead some to kill when they would not have otherwise.

This strikes me as a stretch. Provided the minimum penalties for homicide remain very severe, it is hard to believe that anyone would infer from graded punishments that killing is not such a bad thing to do. But since Feinberg’s theory of criminalization is, at root, consequentialist, any argument in favor of such a law made from within this theory must remain speculative. When it comes to the punishment that a criminal deserves for homicide, however, things are more straightforward. The punishment deserved for violating the criminal law is a matter of fairness, and fairness, Feinberg claims, dictates that the punishment fit the criminal’s blameworthiness:

Fairness requires that relevantly dissimilar cases should be treated in appropriately dissimilar ways, and what could be more "relevant" to the degree of moral condemnation expressed by a punishment than the degree of moral blameworthiness of the one to be punished?7

And while blameworthiness is a complex notion, Feinberg clearly thinks that the amount of harm involved in a crime is a core determinant. In the course of arguing that failed attempts should be punished as severely as completed crimes, he writes:

But what, apart from the wrongness of his act, determines the degree of blameworthiness of the actor? …A sound if blurred insight is that the harm intended is much more important an indicator of an offender’s desert than the harm actually caused. Far more useful, however, than the concept of intentionality, are the four "culpability conditions" first proclaimed in the Model Penal Code—acting purposely, knowingly, recklessly, or negligently in regard to some harmful result.8

The formulation "acting…in regard to some harmful result" is rather strained, but the context makes the basic idea clear enough: for Feinberg, a criminal’s degree of blameworthiness partly depends on the harm her crime would have caused were it not for the intervention of good or bad luck. In the context of laws against homicide, a criminal’s blameworthiness for committing murder would thus depend on the amount of harm that would have been caused by her purposeful or knowing performance of a normally lethal action in the absence of luck. Similarly, a criminal’s blameworthiness for committing involuntary manslaughter would depend on the amount of harm that would have been caused by her reckless or negligent performance of a normally lethal action (or inaction) in the absence of luck. (For simplicity’s sake, I will follow Feinberg in using "harm intended" as shorthand for the notion of the harm that an action would have caused but for circumstances attributable to luck.)

Although it would be nice to have a full analysis of what an action would cause were it not for the intervention of luck, this is not necessary to see that killers often differ with respect to how much harm they intend in causing the death of their victims. If a killer takes a characteristic to be relevant to the harmfulness of death, and takes the victim to have that characteristic, then they are culpable for causing the amount of harmfulness associated with that characteristic, since their causing that amount of harm is hardly attributable to bad luck.9 For instance, among the many murderers who believe that dying young is worse for a person than dying when one is old, those who purposely or knowingly kill victims they take to be young typically intend more harm than those who purposely or knowingly kill victims they take to be old.

It may be that the two different ways in which a punishment can be appropriate for Feinberg pull against each other. We have just seen that deserved punishment tracks blameworthiness, which depends on the amount of harm intended. Earlier, we saw the case for thinking that the harm principle justifies sentencing laws that are sensitive to the amount of harm victims suffer in being deprived of life. Here, though, making punishments relative to the harm actually caused seems preferable, as doing so would give criminals an incentive to discover which victims will suffer least from being victimized. If there is this tension in his account of punishment, however, it is one that many others share. For many theorists hold that punishment should be a function of the gravity of the offense and the criminal’s culpability for performing it. And since the gravity of an offense is typically taken to depend at least partly on the amount of harm it causes, these theorists will also be committed to balancing the amount of actual harm and the amount of harm intended to determine the appropriate punishment for a given crime.10

In any event, it is clear that Feinberg is committed to saying that our fratricidal murderers ought to be punished differently: the murderer who killed his older and unambitious sibling of ordinary abilities should be punished less severely than the murderer who killed his younger, more ambitious, and talented sibling, for the former murder both caused and intended less harm. This is a result his theory will share with many others according to which harm caused or harm intended plays a role in determining the appropriate penalty for a crime.

II. The Commonsense View and Respect for Moral Worth

On reflection, it is peculiar that commonsense is at odds with Feinberg here. To be sure, there is a certain intuitive attractiveness to the view that all wrongful killings are equally wrong regardless of the intrinsic qualities of the victim, and hence that punishments should not be sensitive to such characteristics. But as we have seen, this commits commonsense morality to the view that even when a killing causes the victim more harm, as it typically does when the victim is murdered at twenty rather than fifty, this difference in harm is thought to make no moral difference with respect to how wrong or blameworthy the killing is. If this were right, the relationship between the harm caused by wrongfully ending a person’s life, and the wrongness or blameworthiness of doing so would be very unusual.

Is there a way to vindicate this commonsense view? Although he ultimately neither endorses nor denies the view, Jeff McMahan has proposed a strategy. He believes that the commonsense view is grounded in a commitment to the equal moral worth of people. What makes wrongful killings wrong, McMahan suggests, is that they are actions that fail to respect the worth of their victims. And since all people are of equal worth despite variations in their intrinsic characteristics (such as age, ambition, or talents), such variations are irrelevant to how wrong a wrongful killing is.

The prospects of vindicating commonsense by appealing to such an account are not good, however. The first problem is that commonsense clearly holds that the amount of harm an action causes partly determines how wrong a non-lethal action is. The more harm is suffered by the victim of a wrongful assault, a wrongfully broken promise, a wrongful lie, or a wrongful betrayal, the more wrong and blameworthy the action is. When it comes to the relevance of harm to wrongness and blameworthiness, then, commonsense treats lethal actions as a special case.

This special exception for wrongful killings is difficult to justify within McMahan’s failure-of-respect proposal. The proposal is most plausibly construed as a general account of wrongness: generally speaking, actions are wrong in virtue of their incompatibility with respect for the worth of the victim. If this is right, then it is hard to see how such an account can make sense of the fact that the amount of harm a victim suffers always seems relevant to how wrong non-lethal wrongs are, but is sometimes not relevant to how wrong wrongful killings are. For while it is certainly true that killing a person is an extreme violation of her worth, it seems to be the same kind of violation as that involved in, say, non-lethal torture. In both killing and non-lethal torture, for instance, it is plausible to say that wrongdoers treat their victims as mere instruments for the promotion of some other ends, rather than as ends-in-themselves. It is also plausible to say that wrongdoers of both sorts claim an authority over the victim that they have no claim to: both kinds of wrongdoers deny their victims control over how their lives go. If we focus solely on the notion of violating a person’s worth, then, there does not seem to be anything capable of explaining why non-lethal violations are made worse to the extent that they cause more harm, but lethal wrongs are not made worse to the extent that they deprive the victim of a future that is particularly valuable to her.

In response, it might be thought that the difference between non-lethal and lethal harms is not a matter of their involving qualitatively different violations; instead, it might be that destroying a person is the maximal violation of a person’s worth. Wrongful killings are as wrong as an action can be, on this proposal, and hence the characteristics of the victim cannot affect how wrong they are.

The trouble is that, contrary to this suggestion, commonsense does not treat all wrongful killings as maximally wrong. Some wrongful killings are more wrong than others. To take only the most obvious kind of example, wrongfully killing in a way that causes the victim great pain is commonly understood to be worse than wrongfully killing a victim painlessly.

It is worth pausing to note the strange results that are generated by holding this view while denying that the harm a killing does the victim in death is relevant to its gravity. Consider, for instance, the following case:

Tony reluctantly decides that he must kill Bruce in order to avoid dishonoring his family, but resolves to do as little wrong in killing Bruce as possible. Tony has two poisons at his disposal. One poison will kill Bruce immediately and painlessly. The other will give Bruce one hour of stomach cramps, and then cause his death two years later.11

Of course, from Bruce’s perspective, it would be far better if Tony were to use the slower-acting poison: on balance, Bruce will be much better off if he experiences an hour of stomach cramps and dies two years later than if he dies immediately, albeit painlessly. But, if painful killings are worse than painless ones, and the amount of harm a victim suffers in virtue of dying is irrelevant to how wrong killing him is, Tony must give Bruce the poison that kills him now, for he will thereby do a lesser wrong. If commonsense would reject this result, then it must reject one of the two theses that generate it: either that the harm a wrongful killing causes before the victim dies is relevant to how wrong it is, or the claim we are concerned with here—namely, that the harm a wrongful killing does in death is irrelevant to how wrong it is.

Finally, one might be tempted to look at the specific harm in question—the harm of death—in order to give an explanation within the parameters of the failure-of-respect account of the difference between lethal and non-lethal wrongs regarding the relevance of harm to wrongness. After all, the harm of death is particular to lethal wrongs. The trouble, though, is that the amount of harm the victim suffers from her death is not irrelevant to all questions of wrongness, for it can play a role in determining whether killing is wrong. To see this, first consider the following case:

Martha and Esther are hiking in a remote forest when Esther is bitten by a poisonous snake. The poison immediately renders Esther unconscious, and will cause her death within five minutes. Martha has a dose of the only extant antidote. Unfortunately, the antidote will cause the Esther’s death in two years. Understanding these facts, Martha administers the antidote, which kills Esther two years later.12

And now compare Martha and Esther’s case to this one:

Pierre and Ted are hiking in a remote forest when Ted is bitten by a poisonous snake. The snake’s poison immediately renders Ted unconscious for ten minutes, and will cause his death two years later. Pierre has a dose of the only extant antidote. Unfortunately, this antidote will cause Ted’s death in five minutes. Understanding these facts, Pierre administers the antidote, which kills Ted five minutes later.

In both cases, the partner kills the victim by administering the antidote, but only in the second has the partner done something wrong. And the fundamental difference in these cases is how much life the victim loses when she is killed: Martha’s killing seems permissible in virtue of the fact that she thereby prolongs Esther’s life by two years; Pierre’s killing seems wrong in virtue of the fact that he thereby shortens Ted’s life by two years.

This pair of cases shows that the amount of harm a killing does the victim in virtue of causing her death can determine whether that killing is wrong. It is atypical, however, in that it involves a life-prolonging killing. This peculiar feature might be thought to matter, for there is a sense in which Martha does not harm Esther at all—in which case we are not comparing a more harmful killing to a less harmful killing, but a harmful killing to a beneficial one. But there are also situations, I believe, in which whether harmful killings are wrong depends on how much the victim loses in virtue of her death. Consider first the following case:

Eva has just hacked into the computer network of a terrorist organization. She sees that the terrorists have launched a guided missile, and that they have programmed it to hit a target labeled "Oprah’s Orphanage: 20 residents." The missile is to hit its target within seconds. Eva can reprogram the missile so that it hits an alternate target on the terrorist’s list. Out of concern for her favorite talk-show host and the orphans she has seen on TV, she reprograms the missile to hit a target labeled "Isolated Orphanage: 25 residents." She succeeds: seconds later the missile hits its new target, killing all 25 people inside.

And now compare Eva’s action with Harry’s:

Harry has just hacked into the computer network of a terrorist organization. He sees that the terrorists have launched two equally powerful guided missiles, only one of which is programmable. The programmable missile is aimed at a target labeled "Oprah’s Orphanage: 20 residents," and will reach it within seconds. The fixed-target missile is aimed at a target labeled "Isolated Orphanage: 25 residents," and will reach it within minutes. He realizes that "Oprah’s Orphanage" is the orphanage sponsored by his favorite talk-show host, which he has seen profiled on TV. He cannot alert any of the targets, but he can re-program the programmable missile so that it hits an alternate target on the terrorist’s list. Out of concern for his favorite talk-show host and the orphans he has seen on TV, he reprograms the programmable missile to hit "Isolated Orphanage: 25 residents." He succeeds: seconds later, the programmable missile hits its new target, killing all 25 people inside; several minutes after that, the fixed-target missile hits the target again, but since all the residents have already been killed, it causes no additional casualties.

Many will feel that Eva has done something wrong but that Harry has not. The most plausible explanation of this moral difference, however, concerns how much the missile victims lose in virtue of their deaths: Eva intentionally causes her victims to lose many years of life, whereas Harry intentionally causes his victims to lose only minutes of life.

So far, we have seen that commonsense morality holds that the amount of harm an action causes its victim typically bears on how wrong it is; it holds that the amount of harm a killing causes its victim before the victim dies bears on how wrong the killing is; and it holds that the amount of harm a killing causes in virtue of ending the victim’s life bears on whether the killing was wrong. Despite all this, commonsense morality also holds that the amount of harm a killing causes in virtue of ending the victim’s life does not bear on the degree to which a killing is wrong. We have also seen how this complex picture frustrates various strategies for vindicating commonsense morality by appeal to a failure-of-respect account of wrongness. Though suggestive, this is not a conclusive argument against the commonsense view. It remains possible that some elaboration of the failure-of-respect account could match the complicated counters of the commonsense view of the relationship between harm and wrongness. Or it could be that some other account of wrongness will be able to do so. I believe, however, that trying to vindicate commonsense morality’s position on harm and killing is a losing proposition—not only because of the peculiar specificity of the exception commonsense morality makes for the harm of death, but also because there is a ready explanation of the commonsense view according to which it is a product of an understandable confusion.

III. An Error Theory

If we read McMahan’s proposal as an explanatory rather than justificatory account of the commonsense view regarding harm and killing, it is far more compelling. On this reading, the proposal is that commonsense resists the implications of Feinberg’s theory because commonsense takes these implications to conflict with the equal moral worth of all people. If we say that those who kill twenty-year-olds warrant a stiffer penalty than those who kill fifty-year-olds, we seem to be saying that twenty-year-olds are more worth protecting than fifty-year-olds. This alone suggests that to follow Feinberg here is to deny the equal moral worth of all people, for if one thing is more worth protecting than another, it seems that it must be worth more. In addition, if we say that the killers of twenty-year-olds warrant stiffer penalties than the killers of fifty-year-olds because death inflicts a greater loss on a twenty-year-old than on a fifty-year-old, we seem to be saying that the twenty-year-old’s prospective life was of greater value than the fifty-year-old’s, and this too seems to imply that the twenty-year-old is worth more.

This strikes me as a plausible diagnosis of the intuition that we should not make the penalties for murder vary with features of the victim such as her age. But if so, the intuition rests on a confusion. The reason Feinberg must recommend that killers of younger people receive stiffer penalties than the killers of older people is that the life that younger victims are deprived of is generally worth more to them than the life that older victims are deprived of. But to say that one person’s future life is of greater value to him than another person’s future life is of value to her is not to say that the former person is, morally speaking, worth more. There is an important distinction between the value the future holds for a person, and that person’s moral worth. In the context of assessing killings, however, this is a distinction that commonsense morality seems to elide.

When we say that all people have equal moral worth, we are saying that whatever concern and respect is due to one person in virtue of being a person is due to all others in equal measure. If there are constraints against harming, breaking promises, or lying, for instance, the constraints must give all people equal protection against being harmed, having a promise made to them broken, or being lied to. But, as commonsense recognizes, all this is compatible with saying that some harmings, promise-breakings, and lies are graver than others in virtue of the magnitude of the costs they impose on their victims. There is no tension in saying that it is worse to deprive a person of something that is very valuable to them than to deprive a person of the very same worth something that is far less valuable to them. This is so even when it is some characteristic of the more aggrieved victim that explains why they suffer more. For instance, a broken promise that costs a desperately poor person $100 is commonly understood to be worse than a broken promise that costs a very rich person $100. What commonsense rules out is claiming that some such wrongs are graver in virtue of being wrongs against a person who is more worthy in the sense of being due a higher degree of respect or concern.

With respect to wrongful killings, we saw above, commonsense follows this reasoning to a point. It recognizes that holding some killings to be graver offenses in virtue of involving greater amounts of pain is perfectly compatible with the equal moral worth of all people. In this case, it is obvious that the grounds for this comparative judgment have nothing to do with comparative moral worth. Instead, we recognize that the killer that kills painlessly deprives his victim of her future life, but the killer who kills painfully deprives his victim of both her life and her freedom from pain. No one worries here that we are treating the victim who suffers more pain as being more worthy than the victim who suffers less pain.

When it comes to the harm of being deprived of a future life, however, commonsense balks. Why? The answer, I believe, has to do with the fact that, in many cases, the characteristics of people that make them suffer more from death are also characteristics that we admire. Many people find it natural to admire youth, ambition, and talent. These characteristics are also qualities that make a person’s future life worth more to them, and consequently are qualities that make their deaths a greater harm to them. So, when we contemplate giving stiffer punishments to those who cause their victims more harm in death, we are often contemplating giving stiffer punishments to killers who have killed those who have traits that we admire. And it is plausible that this fools us into thinking that the basis for the judgment that it is more wrong or blameworthy to kill the young, ambitious, and talented is that people with these traits are worth more. For if we are tempted to hold some people to have a greater worth than others, it will be those whom we hold in greater esteem. On our guard against this temptation, we reject a view that appears to give in to it.

On reflection, of course, Feinberg’s theory of punishment is not inegalitarian. The reason for punishing the killers of younger people more severely than the killers of older people has nothing to do with holding that young people are of greater moral worth than old people. It is, rather, because death deprives them of a future that is of greater value to them, just as a $100 loss constitutes a greater deprivation for a poor person than for a rich person. The resistance to the implications of Feinberg’s theory of punishment, according to this explanation, is the product of a sensitivity on the part of commonsense to anything that smacks of elitism. Practically speaking, this sensitivity is probably worth having. But once we recognize that it is raising a false alarm, that it is responding to a proposal that is not elitist, we can and should ignore its warnings.

This explanation also has the virtue of accounting for an exception to McMahan’s claim that, according to commonsense, "the extent to which it is wrong to kill a person is unaffected by facts about the victim other than those that … may make him non-innocent in the relevant sense." Many people, I believe, take the killing of people who are especially vulnerable to be worse and/or more blameworthy than the killing of people who are generally more able to defend themselves. For instance, killers who have preyed on those who are frail, or physically or mentally disabled, seem to have committed a graver wrong and to be more blameworthy. As I will discuss in more detail below, this judgment is reflected in the sentencing law of many jurisdictions. Here, commonsense treats some killings as worse than others in virtue of characteristics of the victims, even when these characteristics have no bearing on the victim’s innocence. Given the preceding, however, we should not be surprised that commonsense finds no conflict between this view and its commitments to moral equality. Psychologically speaking, few would be tempted to say that characteristics that make someone particularly vulnerable are also characteristics that make some people of superior worth. When it is a characteristic of adults, at least, vulnerability is often associated with something commonsense regards as a misfortune. Consequently, we are not fooled into thinking that giving stiffer punishments to those killers whose victims are particularly vulnerable is elitist.

The explanation also accounts for some of the particular contours of the commonsense view that we charted in the previous section. There we found a number of situations in which commonsense holds that the amount of harm a victim suffers in virtue of her death is relevant to the wrongness of killing her. In all these cases, though, something other than admirable qualities of the victim accounted for the differences in the amount of harm that death caused the victims: the differences were attributable to differences in poisons, antidotes, and missiles. So, again, there is nothing in these cases to fool commonsense into thinking that allowing a role for the amount of harm the victim suffers in virtue of dying is to abandon a commitment to moral equality.

The position I have extracted from Feinberg’s work starts with the common and well-motivated view that the harm of death varies, and combines it with a theory of criminal punishment that is well-motivated and, in its more general features, widely endorsed by philosophers and legal scholars. In contrast, there is no apparent justification for the commonsense exclusion of the harm of death from the determinants of how wrong a wrongful killing is. Indeed, as the case of Tony and the two poisons showed, it is doubtful that the commonsense position on harm and killing is internally consistent. Further, there is a good explanation of why commonsense holds this position according to which it arises from a mistaken triggering of our warranted sensitivity to moral elitism. We have more than sufficient reason, then, to follow Feinberg and part with commonsense on the matter of the punishment homicides warrant.

IV. Some Judicial and Legislative Implications of Feinberg’s View

The explanation of the commonsense view offered above does raise a worry about implementing the account I have attributed to Feinberg. Mistaken though it is, it may be difficult for us to shake the intuition that punishing killers in proportion to the harm their victims suffer in virtue of dying contravenes our commitment to moral equality. So, were courts to give killers the punishment that their crimes truly warrant, this might give some the misimpression that governments have abandoned their commitment to extending equal protection under the law to all their citizens, and this could have bad consequences.

If so, there may be a conflict between the governmental duty to punish justly and the desire for policies that have the best overall consequences. Whether or not there is such a conflict depends on how bad the consequences of contravening the commonsense intuition would be, and whether these costs are offset by the reduction in the harm that is likely to result from imposing harsher punishments on those killers whose victims stand to gain more from their futures. But if ignoring the amount of harm killers cause in causing their victims’ deaths is to commit only a very minor breach of justice, the possibility of a net social cost could perhaps be a sufficient reason to follow the current commonsense view as a matter of practice.

Ignoring the amount of harm a victim suffers in virtue of dying, however, is a substantial breach of justice. To see that this is so, it is enough to recognize how drastically killings differ in the amount of harm the victim suffers in virtue of dying. For instance, on average, killers of twenty-year-olds deprive their victims of roughly twenty-eight years more life than killers of fifty-year-olds do.13 To get a handle on how much twenty-eight years of life is worth to a person, imagine how much an ordinary person would sacrifice at twenty in order to prolong their life for twenty-eight years: consider how much pain they would be willing to endure and how much of what they care about they would be willing to give up. To the extent that people are good judges of what is in their own interest, depriving a person of twenty-eight more years of life is at least as bad for them as causing them all this pain and hardship. If an assailant who caused a victim that amount of pain and hardship were given the same punishment as an assailant who merely pinched his victim, this would be a massive breach of justice. And since the difference in the harm caused by those who kill twenty-year-olds and those who kill fifty-year-olds seems comparably large, it is implausible that giving such killers the same punishment could be only a minor affront to justice.

If we are to avoid serious injustice in punishment, then, sentencing authorities ought to take the amount of harm the defendant caused or intended the victim to suffer in virtue of her death to be a determinant of the sentence imposed for homicide. This might be done in the current U.S. legal context in a number of ways. The most obvious is that judges who have discretion to adjust the sentences of those who commit homicide could take the amount of harm the killer caused or intended to cause the victim in virtue of her death as a factor in determining his sentence. There are, in addition, several legislative initiatives that are likely to be warranted by the preceding arguments. For instance, sentencing guidelines for non-capital cases could stipulate that the length and conditions of imprisonment as well as the criminal’s eligibility for parole vary according to how much of a harm the murder "intended" in causing the victim’s death. Not surprisingly, this would require a revision of the sentencing statutes in many jurisdictions. To take one important example, under the sentencing guidelines that currently inform federal court judges, the only provision for adjusting the sentence of convicted killers that refers to the victim’s characteristics allows for increasing the penalty if the victim was particularly vulnerable. Further, for first-degree murder where capital punishment is not imposed, the guidelines direct judges to give all defendants life imprisonment without the possibility of parole, irrespective of the victim’s characteristics. One way of accommodating the implications of Feinberg’s theories would be to revise these guidelines so as to allow prison terms to reflect the victim’s age, ambitions, and prospects for fulfillment (and/or the killer’s awareness of these factors).

Similarly, legislators could make the amount of harm a murderer causes or intends the victim to suffer in virtue of her death a factor that determines whether a murderer receives the death penalty. This should probably be done in two ways. First, under current constitutional law, a murderer cannot be put to death unless he or she is "death-eligible," where being death-eligible is, in part, a matter of having committed a murder that was "aggravated"—that is, a murder whose circumstances "reasonably justify the imposition of a more severe sentence on the defendant compared to others found guilty of murder."14 In order to satisfy this constitutional requirement, death penalty states and the federal government have drawn up lists of aggravating circumstances. These lists ought to contain an entry for killing a victim that stood to gain more from her continued existence than the victims of others found guilty of murder.

As it stands, they do not. The details of the lists vary across jurisdictions, and there are over forty-five factors that make at least one appearance on state or federal lists of aggravating circumstances.15 But for one, all these factors are clearly irrelevant to the amount of harm that death caused the victim. The sole exception is whether the victim was a child.

It is tempting to think that in treating the killing of a child as an aggravating circumstance, legislators, at least, have implicitly agreed that the punishment for murder should be responsive to the amount of harm death does the victim. After all, children typically lose more life from being killed than adults do. There is good reason to think, though, that the explanation for these provisions derives instead from children’s vulnerability. Children’s relative defenselessness makes them particularly easy prey, and so it may be important to protect them with especially strong deterrents. Further, we ordinarily think that anyone who preys on the defenseless must be especially vicious, and hence especially deserving of punishment. This interpretation is bolstered by the fact that in several jurisdictions the "young victim" aggravating circumstance appears alongside, or is bundled with, other "vulnerable victim" categories.16 New Hampshire, for instance, makes the vulnerability rationale explicit: there, a murder is aggravated if "the victim was particularly vulnerable due to old age, youth, or infirmity."17

Whatever these legislatures’ intentions might have been, though, it is doubtful that the "child victim" aggravating circumstance is the appropriate way to bring sentencing law into line with Feinberg’s theories of death and punishment. For according to Feinberg (and others), death typically harms children, especially young children, less than it does young adults. This is so for two reasons. First, young children do not have actual interests derived from "ongoing projects and enterprises, long-range goals, and hopes for their own achievement" because they are incapable of formulating the kinds of desires necessary to produce such objectives. Second, the interests whose frustration can constitute harm must be actual rather than potential. As he puts it, "One can grieve for an interest that might have been fulfilled, but was not; but that is quite another thing than grieving for a nonexistent interest that might have become real, but never did."18 Consequently, young children have less of a stake in their future, and are thus less harmed by being deprived of it, than typical adults.

The second aspect of the capital sentencing process in which the harm of death could be taken into account is when juries are deciding whether to give the death penalty to "death-eligible" defendants. At this stage, the defendant is given the opportunity to argue that some fact mitigates either his crime or his blameworthiness for committing it. Given that juries are unlikely to treat the fact that the murderer knew that a victim was older, lacking in ambition, or unlikely to successfully pursue her goals as a reason in favor of clemency, an explicit instruction to the effect that they take such considerations as mitigating factors may be in order.

Such changes to our current sentencing regime will not come easily. They will make many uncomfortable, and perhaps even strike some as offensive. My impression is that if I am right to attribute to Feinberg the position I have developed here, he would have responded to this resistance with charity and sympathy but, ultimately, would have been undeterred by it. In the concluding sentence of an essay in which he defends the well-motivated but counter-intuitive position that mere attempts should be punished with equal severity as completed crimes, Feinberg has this to say about one of his opponents:
Professor Fletcher’s mistaken analyses…I realize, were animated by an unusually powerful (and commendable) motive to avoid elitism. Even though, on some scale of "cognitive blameworthiness," these mistakes are not very blameworthy, they can nonetheless cause considerable mischief, intellectual or other, and in actual cases, make the worse appear the better cause.
In the face of resistance to the well-motivated but counter-intuitive recommendation that those who commit homicide should be punished in proportion to the harm their victims suffer in virtue of their deaths, precisely the same response would be fitting.19

Endnotes

1. Throughout, I use "homicide" to refer to all and only illegal killings. I use "murder" to refer to homicides in which the killer purposely or knowingly kills the victim.

2. "Challenges to Human Equality," forthcoming in Journal of Ethics. It is important for McMahan that the commonsense intuition is restricted to the wrongness of killing persons, where a person is a creature that has full moral status. When I speak of "killing" in what follows, it will be the killing of persons to which I refer.

3. Joel Feinberg. Harm to Others (New York: Oxford University Press), 82.

4. Ibid., 93.

5. Ibid., 26.

6. The qualification "adult" is needed here because, according to Feinberg’s theory, young children stand to gain less from their continued existence than do most adults. I explain why in Section IV. In the remainder of the paper, I will take the qualification as understood.

7. Joel Feinberg. Harmless Wrongdoing (New York: Oxford University Press, 1988), 150.

8. Joel Feinberg. "Equal Punishment for Failed Attempts: Some Bad but Instructive Arguments against It," Arizona Law Review, 37 (1995): 131-32.

9. Actual awareness of the presence and significance of the factors relevant to the harmfulness of death is surely sufficient for excluding these factors from the domain of luck, but is probably not necessary. Perhaps, for instance, a reasonable expectation that the killer be aware of these factors and their significance is enough.

10. A particularly clear example is Doug Husak, whose "principle of proportionate sentences" holds that "the punishment a defendant deserves should be proportionate to the seriousness of his criminal conduct," where "the seriousness of criminal conduct is a function of two variables—harm and culpability." See Douglas N. Husak, "Transferred Intent," Notre Dame Journal of Law, Ethics & Public Policy, 10 (1996): 92.

11. This case elaborates a problem McMahan raises for the respect-for-life proposal in his The Ethics of Killing: Problems at the Margins of Life (New York: Oxford University Press, 2002), 248.

12. This case is borrowed from Kasper Lippert-Rasmussen’s "Two Puzzles for Deontologists: Life-Prolonging Killings and the Moral Symmetry between Killing and Causing a Person to be Unconscious," Journal of Ethics, 5 (2001): 400.

13. Elizabeth Arias. "United States Life Tables, 2002," National Vital Statistics Reports, 53:6 (November 10, 2004): 3.

14. Zant v. Stephens, 462 U.S. at 877 (1983).

15. Jeffrey L. Kirchmeier. "Aggravating and Mitigating Factors: The Paradox of Today’s Arbitrary and Mandatory Capital Punishment Scheme," William & Mary Bill of Rights Journal, 6 (Spring 1998): 345-459.

16. As of 1998, eight of the seventeen jurisdictions that considered a child victim to be an aggravating factor also considered an elderly, handicapped, or disabled victim to be an aggravating factor. See Kirchmeier, op cit., 420-421.

17. N.H. Rev. Stat. Ann. 630:5(VII)(g) (1996).

18. Feinberg, Harm to Others, 96-97.

19. I would like to thank Doug Husak for helpful discussion of some of the issues dealt with in this paper.

Previous Article | Index | Next Article

APA Newsletters

Fall 2006 Volume 06, Number 1

Fall 2006
Volume 06, Number 1