[UNOFFICIAL TEXT ONLY—Please do not quote or cite.
See forthcoming Law and Philosophy (2008) for Official Text]
What Is the Wrong of Wrongful Disability?
From Chance to Choice to Harms to Persons
M. A. Roberts
Abstract: The issue of wrongful disability arises when parents face the choice whether to produce a child whose life will be unavoidably flawed by a serious disease or disorder (Down syndrome, for example, or Huntington’s disease) yet clearly worth living. The authors of From Chance to Choice claim, with certain restrictions, that the choice to produce such a child is morally wrong. They then argue that an intuitive moral approach—a “person-affecting” approach that pins wrongdoing to the harming of some existing or future person—cannot account for that wrong since the choice to produce such a child cannot, under the logic of the nonidentity problem, harm that child. The authors propose that we supplement the person-affecting approach with an “impersonal” principle that takes the form of their well-known principle N. In this paper, I argue that the authors are mistaken to suppose that a plausibly articulated person-affecting approach cannot account for the wrong of wrongful disability. We can retain an intuitive, comparative, “worse for” account of harm and still identify serious harms imposed by the choice of wrongful disability. In particular, I argue that harm, both to the impaired child and to others, comes not in the form of that procreative choice’s procreative effect but rather in the form of its many distributive effects. I also argue that the rare, residual case in which a person-affecting approach would approve of the choice of wrongful disability does not function as a counterexample to that approach.
As a separate matter, I address legal claims for wrongful disability, which are closely akin to claims for wrongful life. The legal claim is brought by the impaired child, not against the parents, but rather against health care providers whose negligent failure to diagnose or inform parents of an increased risk of a genetic or congenital impairment results in the birth of the impaired child. The authors’ treatment of the moral wrong that is done as impersonal in nature suggests that courts are correct to dismiss any such claim. Once we identify harm, however, the person-affecting approach can identify a clear foundation in the law for the wrongful disability claim.
In From Chance to Choice, Buchanan, Brock, Daniels and Wikler sensibly argue that freedom of choice exercised in the procreative context can only be taken so far. Trying to
define just how far that is, the authors begin by examining the person-affecting approach in ethics, which they say “underlies common moral views” as well as “legal requirements” governing parents’ responsibilities with respect to their offspring. A central tenet of the person-affecting approach is that an act can be morally wrong only if that act makes things worse for, or harms, some existing or future person; a “bad” act, at least eventually, must be “bad for” someone. More generally, according to the person-affecting approach, the claim that an act is morally wrong must be grounded in facts about how that act affects existing and future persons. Does it make them better off? Does it create additional wellbeing for them? Does it make them worse off? Does it create less wellbeing for them when the agent could have created more? Does it harm them?
The person-affecting approach offers a compelling picture of morality. The authors believe, however, that difficulties emerge for that approach when we apply it to the problem of wrongful disability. Wrongful disability is the choice to bring a child into existence who will experience “serious suffering” or a “serious loss of happiness or good” as a result of a genetic or chromosomal defect or other congenital disease or disorder that, at present, cannot be cured or substantially ameliorated. Critical to the concept of wrongful disability is that the child’s suffering is not so grave that the existence itself is less than one worth having. The claim of the genuinely wrongful life, in contrast, is generally understood to assert that the impaired child would have been better off never having existed at all. The wrongful disability claim concedes that the value of the particular existence falls within the positive range—that the child’s overall, lifetime wellbeing level is not less than the zero level that can be assigned to the alternative of never having existed at all. Yet the disease or disorder that that child suffers is, by definition, serious. The choice to impose such an existence on anyone seems troubling, however confident we are that the life itself will be worth living.
Buchanan et al. believe, with restrictions, that the parents’ choice to bring the seriously impaired child into existence, when they had the alternative of producing a healthier but distinct child instead, is wrong. According to the person-affecting approach, that means that the parents’ choice must itself harm some existing or future person. The obstacle has been to identify just who that person might be. By the logic of what is called the “nonidentity problem,” it seems that it cannot be the impaired child. After all, that child could not have existed and not suffered the impairment, and, by hypothesis, the child’s existence is, though flawed, worth having. Nor would the parents seem to have been harmed by the choice they themselves have made. Still other persons may appear to have no clear interest in whether the parents make the choice of wrongful disability or not. If such persons are inconvenienced, or even harmed, by the parents’ choice, the moral significance of that fact might seem quite obscure.
The choice of wrongful disability, according to the authors, thus constitutes an example of a choice that is wrong but harms no one. If their argument succeeds, then no purely person-affecting approach can be accepted as correct. At a minimum, we will need to reject the core person-affecting tenet that wrongdoing can always be pinned to the harming of some person or another. We can retain the idea that many wrongs are wrong in virtue of the harm they impose on persons. But our theory will need to be expanded to include the idea that some wrongs are wrong in virtue of their impersonal effects. A conception of free-floating wrongs is thus endorsed—a picture of wrongs that can exist untethered from how persons, whether existing or future, are made to fare by the choices that we, as moral agents, make.
I do not believe, however, that wrongful disability seriously challenges the person-affecting approach. I concede that the choice to have the impaired child is, in the vast majority of cases, wrong. What I doubt is the authors’ view that that choice represents a case of “no harm done.” Discerning harm to offspring in the wrongful disability or any other procreative context has been, of course, a tricky matter. I believe, however, that a clearer view of harm can be achieved if we look beyond the peculiar procreative effect of the choice of wrongful disability to the full range of harms that may be imposed by that choice. Thus, the choice of wrongful disability obviously has a certain procreative effect. It brings an impaired child into existence. But it also has numerous distributive effects—effects that fall on the impaired child as well as persons other than the impaired child. Once we recognize those distributive effects, discerning harm in the wrongful disability context will seem less a struggle with an irrefutable logic and more an ordinary exercise in practical reasoning.
Having discerned harm—and, as we shall see, a morally significant, or unjustifiable, form of harm at that—we can then provide a person-affecting account of the wrong of wrongful disability. After all, the authors nowhere challenge the idea that the person-affecting approach has the resources to find wrongdoing, in appropriate cases, on the basis of harms we are able to discern. The stumbling block to the finding of wrongdoing, rather, has been that the harms themselves have seemed so elusive.
That the wrong of wrongful disability can be accounted for within the confines of the person-affecting approach is a valuable result. For one thing, it means that we are not compelled to adopt the authors’ conclusion that an adequate moral theory must be “hybrid” in nature—that it must include, that is, both impersonal and person-affecting elements. Impersonal values seem inherently at odds with person-affecting values. We thus cannot have any great confidence in the hybrid view until we understand how its internal tensions will be resolved within a more complete theory that is itself plausible and consistent.
Moreover, a person-affecting account of wrongful disability is nicely positioned to help resolve the controversy of whether the impaired child’s legal claim should be recognized as valid. Such a claim would be brought not against the child’s parents—for reasons of efficiency, we want to maintain doctrines of intra-family immunity—but rather against a health care provider—an obstetrician, for example, or a genetics testing laboratory—who has negligently failed to diagnose or inform the parents of their increased risk of bearing a child with a serious genetic or congenital disease or disorder.
At least at a theoretical level, the legal claim for what is called “wrongful life” is normally taken to include the assertion that the impaired child’s existence is so badly flawed that it is less than one worth living. The nomenclature remains the same, however, even when that assertion is clearly at odds with the facts of a particular child’s life—when the impairment is, e.g., due to hereditary deafness or Down syndrome. In other words, claims titled “wrongful life” are often, in practice, tantamount to claims for wrongful disability where, by definition, the flawed existence is worth having.
But whether the underlying facts suggest the genuinely wrongful life, or instead wrongful disability, the impaired child’s claim is not considered legally valid in most jurisdictions. At least in part for the same reason the authors consider wrongful disability to involve “no harm done,” most courts are skeptical that the impaired child has been harmed by the provider’s negligence. Courts, in addition, have the sense that the damages owed to the child cannot be calculated on any reasonable basis.
The authors’ hybrid approach accounts for the moral wrong of wrongful disability. However, it does nothing to support the notion that the corresponding legal claim should be recognized as valid. In fact, the authors’ treatment of wrongful disability strongly suggests that courts are correct to dismiss the impaired child’s claim in any case in which the existence is not so badly flawed that it is less than one worth having.
In contrast, the person-affecting approach I describe in what follows tethers its finding of moral wrongdoing to particular facts about harm—harm in an ordinary, comparative, “worse for” sense; not in a sense the law itself will be reluctant to recognize. If I am correct that harm is in fact prevalent in the wrongful disability context, then it will become appropriate for courts to recognize wrongful disability as a valid legal claim.
Wrongful disability is a good example of the more general problem of indeterminacy in the law. Our need to resolve hard legal cases in respect of which the law itself is unsettled makes it is highly desirable that we identify a person-affecting account of the wrong of wrongful disability (and many other troubling varieties of procreative choice). After all, the relevant branches of the law are intransigently person-affecting in nature. That means that, in the absence of a radical restructuring of the law, an impersonal approach will generate results—wrongs without harms and without claimants—that seem legally beside the point. In contrast, the implications of a person-affecting approach will remain clearly salient within the law.
* * *
Part 1 below describes the person-affecting approach and outlines just why the authors and many others consider the choice of wrongful disability a case of “no harm done.” Part 2 disputes that view. I argue that the harm of wrongful disability comes not in the form of the procreative choice’s procreative effect but rather in the form of its many entirely predictable distributive effects. Part 3 addresses the impaired child’s legal claims for wrongful disability and for the genuinely wrongful life. I argue that a person-affecting approach provides a foundation in the law for both those claims. Part 4 discusses the person-affecting approach in the context of the problem of indeterminacy in the law.
1. Person-affecting morality and the wrongful disability challenge
1.1 The person-affecting approach; the authors’ “principle M.” Procreation is properly considered a highly personal matter. We understand that a procreative choice—the choice, for example, to continue with a pregnancy or to have an abortion; to have two children or to have ten—that is right for one agent may not be right for another. Yet it seems obvious that morality’s accommodation of personal choice in the procreative context cannot be stretched indefinitely. Not all procreative choices can properly be deemed morally permissible.
The issue under consideration here is whether commonly accepted, intuitive, “person-affecting” principles of morality can define those limits. At first glance it seems that they have the power to do just that. The person-affecting approach tethers wrongdoing to how persons—existing or future—are affected by the act under scrutiny. It tells us that, if an agent has created the most wellbeing that he or she can for each such existing or future person (or, more narrowly, for each such existing or future person for whom the agent is, in the authors’ locution, “responsible”), then that agent has done all he or she is morally required to do. So, an act can be morally wrong, according to the person-affecting approach, only if that act creates less wellbeing for some existing or future person when the agent had the alternative of creating more; an act is wrong only if it harms an existing or future person. That necessary condition on wrongdoing constitutes a core tenet of the person-affecting approach. But the person-affecting approach includes certain sufficient conditions on wrongdoing as well. Thus, an act will be considered wrong when the agent could have created more wellbeing for a particular person without creating less for any other existing or future person and fails to do so. What the agent does is wrong if it harms without reason—if, that is, the agent harms one person and could have avoided harming that one person without harming anyone else. Moreover, person-affecting principles are capable of addressing tradeoff scenarios—cases in which an agent can create more wellbeing for one person only by creating less wellbeing for another. If the agent distributes wellbeing between those competing persons in an appropriate way, the person-affecting approach will not consider what the agent has done wrong. We will, instead, recognize that, while the agent has imposed a harm on one person or the other, that harm is justifiable and hence that the act itself is morally permissible, that is, not wrong.
Immediate implications of the person-affecting approach include the following. (1) We are not required to produce ever more children simply on the grounds that those children themselves would have lives worth living. (2) We have stringent obligations in respect of any child we do or will happen to produce. And (3) whether a choice that involves a tradeoff in wellbeing between individuals is wrong will depend on whether the agent has distributed wellbeing in an appropriate way between the competing individuals.
The authors sum up some of the important elements of a person-affecting approach in their own principle M. Narrowly tailored to govern the relationship between the agent and those “dependent” persons for whom the agent is “responsible,” principle M charges the agent with the duty to avoid letting a child or other person suffer any “serious harm or disability.” That statement is, however, qualified by a provision that recognizes that agents often face tradeoff scenarios. The authors suggest that such scenarios are to be analyzed, under a person-affecting approach, by a “weighing of harms prevented against benefits secured.” Thus principle M, rightly, avoids the implication that agents are obligated not to harm the child in the case where their only means of achieving that result would be by imposing “substantial burdens or costs or loss of benefits . . . on themselves or others.”
The authors consider the person-affecting approach adequate to address “typical cases of harm where a distinct individual is made worse off.” That is so, they believe, even when the person whose wellbeing is at stake does not yet exist. Thus, the person-affecting approach, as they construe it, can explain why it would be morally wrong for agents to negligently dispose of “waste known to be hazardous to a developing fetus.” The authors also appeal to a person-affecting approach to explain how it can happen that abortion is morally permissible in a particular case yet that the woman who chooses to continue the pregnancy is obligated to refrain from imposing a “much less serious harm” than nonexistence on the fetus. Similarly, the authors’ account of the genuinely wrongful life claim—the claim that is typically taken to assert that the child’s life is less than worth living—is itself person-affecting in nature. The child suffers a life that is “only a burden and torment” from the child’s own point of view. If we also think (as I do) that it makes sense in such cases to say that the child is made worse off by having been brought into the badly flawed existence than by having never been brought into existence at all, we can take the authors’ analysis a step further and use the term “harm” without any qualm to describe the effect of the provider’s negligence on the impaired child. Whether or not we take that further step, however, the authors clearly consider the person-affecting approach capable of finding that a wrong has been done in such cases.
The person-affecting approach thus seems to offer a plausible picture of the moral obligations that we have in respect of future persons. The fact that the person-affecting approach seems to be at work in the law as well as morality does not, of course, prove the validity of that approach. But it does show how deeply woven into the fabric of our lives person-affecting principles in fact are, and, if we think that the law, forged over long centuries in courtrooms and legal offices and not short decades in the isolated studies of philosophy professors, might contain some basic moral insights, it may provide evidence of a sort for the person-affecting approach in ethics. It is undeniable that the branches of the law that most closely concern procreation—tort law, family law and constitutional privacy law—are intransigently person-affecting in nature. They protect persons from harms and from risks of harms; they do not recognize wrongdoing when there is and can be no claimant. And they protect not just persons who exist at the time the act under scrutiny is performed but future persons as well. Agents are not, in other words, immunized from legal liability and all state regulation by virtue of the fact that their victims do not happen to come into existence until after the act under scrutiny is performed. At the same time, however, once it is established that the conduct under scrutiny imposes neither harm nor risk of harm—once it is shown that that conduct does not involve a loss of wellbeing for any person, existing or future—the agent becomes legally invulnerable. Thus, the doctor who has been shown to have done no harm will not be deemed to have engaged in an intentional tort or in negligence. Parents who have been shown to have imposed no harm—and no risk of harm—on their children will not be vulnerable to losing custody or control of those children. Finally, choices that affect the most central and intimate aspects of our lives and can be shown to harm no one are just the choices that most plausibly garner the protection of the Fourteenth Amendment’s due process clause.
The person-affecting approach thus has many obvious plusses. But can it adequately account for all of our moral obligations? Or is wrongful disability an example of a case in which an act is considered permissible under a person-affecting approach yet seems to us to be clearly wrong?
1.2 The problem of wrongful disability. The particular instances of wrongful disability under scrutiny by the authors are just those that satisfy the following condition A:
A. The problem of wrongful disability: An agent chooses to bring a severely impaired child into existence, or not to prevent a severely impaired child from coming into existence, whose life will be worth living but who will, as a result of a genetic or congenital disease or disorder that cannot today be cured or ameliorated, experience “serious suffering or limited opportunity or serious loss of happiness or good,” when (i) the agents had the alternative of bringing an unimpaired though distinct (“nonidentical”) child into existence in place of the impaired child and (ii) the choice of that alternative would have imposed no substantial burden on the agents themselves or on others.
The choice to bring a Down syndrome child into existence is a good example of the kind of choice that condition A contemplates. An untreatable chromosomal disorder, Down syndrome does not render a child’s life less than worth living. But it does cause “serious suffering or limited opportunity.” According to Buchanan et al., the choice to produce such a child, when the agent had the alternative of producing a healthier though distinct child instead and condition A is otherwise satisfied, is morally wrong.
Yet the choice seems to the authors to be one that harms no one. The impaired child is, of course, the very person who is most obviously caused to suffer by that choice. In deciding that child has not been harmed, the authors explicitly rely on the logic of the nonidentity problem. The choice of wrongful disability has served to bring the child into existence, and that existence, though flawed, is worth having. Moreover, since the flaw reflects a disorder, Down syndrome, that we at present have no way to repair, an unflawed existence is not an alternative for that child. A healthier child could have been brought into existence in place of the impaired child. But that healthier child would have been a distinct child, “nonidentical” to the original. From those facts, the authors think we are compelled to conclude that the impaired child has not been harmed by the parents’ choice. How can the child have been harmed, when there is nothing the parents, or anyone else, could have done that would have improved that child’s plight? How can that child have been harmed when no alternative act exists for the parents by which they could have created more wellbeing for that child than they in fact have?
Before we can conclude “no harm done,” however, we must also consider how the choice under scrutiny affects persons other than the impaired child. Here, too, harm seems elusive. The parents themselves seem not to have been harmed. Presumptively, they do not harm themselves when it is part of the case that they have chosen to have the impaired child in place of a healthier child. And the notion that persons other than the impaired child and the parents have been harmed may seem less than obvious. If other persons are inconvenienced, or even harmed, by the parents’ choice, the moral significance of that fact, especially in the case where the parents distribute time, energy and money in an appropriate way between those persons and the impaired child, will seem obscure. Any harm that is done to persons other than the impaired child and the parents may thus seem justifiable.
According to the authors, then, we are left with a choice that is wrong but harms no one at all—no existing and no future person. If the authors are correct on that point, then we must recognize that at least some acts are wrong not in virtue of how they affect persons but rather in virtue of certain impersonal effects that they have.
1.3 Principle N. The authors believe that wrongful disability shows that morality must be understood to include not just person-affecting elements but impersonal elements as well. While many theorists agree with the authors on that point, there is little consensus about precisely what form that impersonal element will take. For their part, Buchanan et al. propose to supplement the person-affecting principle M with the impersonal principle N.
According to principle N, agents must take steps to insure that any child they produce will have an existence that avoids “serious suffering or limited opportunity or serious loss of happiness or good” in any case in which the agents could have instead produced a better-off, though distinct child without imposing “substantial burdens” on others. Agents have that obligation, according to principle N, even if the child’s flawed existence is unambiguously worth having and an unflawed existence is not an alternative for that child. Principle N easily deems many choices of wrongful disability—all, indeed, that satisfy condition A—morally wrong.
Rakowski describes the authors’ combination of principles M and N as a “hybrid view.” That view is attractive in that it appears to chart a course between two of the most challenging problems in population ethics. The nonidentity problem is one of those problems. What Parfit calls the “repugnant conclusion” is the other. Principle N, by taking an impersonal approach, can declare the choice of wrongful disability wrong—that no one person has been harmed, or made worse off, is immaterial. At the same time, principle N avoids any explicit commitment to what is perhaps the most historically weighty of any impersonal approach in ethics—that is, aggregative consequentialism. According to aggregative consequentialism, the permissibility of a given act is determined by whether that act maximizes the sum of all individual wellbeing levels, present and future—whether, that is, wellbeing in the aggregate has been maximized. Aggregative consequentialism coincides with principle N on the issue of wrongful disability. But it also seems to imply that it is better, indeed obligatory, for agents to produce a vast population of persons whose lives are only barely worth living but whose individual wellbeing levels sum to a very larger number X, when agents might instead have produced a somewhat smaller (though still very large) population of very well-off persons whose individual wellbeing levels sum to a number that is less than X. That result is considered “repugnant” by many theorists. Principle N avoids that result by simply declining to assess any choice between bringing populations of distinct sizes into existence (any “different number” choices). It does not tell us that the repugnant choice is wrong. But at least it does not tell us that that choice is permissible.
At the same time, the hybrid view is not a thing of grace and beauty. In particular, the impersonal values that it recognizes seem at odds in important respects with its person-affecting values. Until we see how those tensions are resolved within a more complete theory that is itself plausible and consistent, we cannot be confident that the hybrid view is itself correct.
Rakowski puts the critical point this way:
[The hybrid thesis] is a novel idea, because it combines moral principles that commonly are regarded as rivals, each excluding the other within the relevant moral space. Unfortunately, it is unclear how these parts mesh or what morality decrees when person-affecting and non-person-affecting components pull in different directions. Perhaps such an account is possible and perhaps it can rest on something more than intuitive responses to a range of cases in which we are asked to judge the comparative wrongness of actions without relying on some theory to supply guidance. But given the ingenuity already lavished on the problems that Parfit poses, there is reason to remain skeptical until that account is offered.
Buchanan et al. look at things, however, quite differently. They argue, in effect, that we have no choice but to embrace principle N. “Our argument for the non-person-affecting principle N is only that it is needed, in addition to the person-affecting principle M . . . . Wrongful disability proves that an impersonal principle is needed, as do, according to the authors, many like instances of the nonidentity problem.
By the authors’ own account, then, our reason for accepting principle N would evaporate should it turn out that we can explain the wrong of wrongful disability in terms that are purely person-affecting. I will argue in what follows that we can do just that. In making that argument, I will concede that the choice of wrongful disability is wrong in the vast majority of cases. Agents ought to avoid making the choices condition A describes. When would-be parents have the alternative of bringing a healthier child into existence, they generally ought not choose to bring a child into existence who will unavoidably experience “serious suffering” or a “serious loss of happiness or good.” That is so, even when that existence, though flawed, is unambiguously worth having. Similarly, health care providers do the wrong thing in cases in which their own negligent omission—their failure, for example, to recommend genetic testing to the would-be parents in appropriate cases—results in the coming into existence of any such child.
My main quarrel, then, is not with the authors’ assessment of wrongful disability as wrong but rather with their view that that choice represents a case of “no harm done.” Accordingly, my goal is to dismantle the “no harm done” assessment that the authors and many others have unwaveringly adopted in the context of wrongful disability. I believe that goal can be achieved by widening the scope of our scrutiny to include the full range of harms plausibly imposed by that choice.
In recent years, a number of theorists have similarly proposed a widening of the scope of possible harms in the procreative context. But many such accounts seem clearly stipulative in nature. If we say that by definition any act that imposes serious suffering—the rescue attempt that results in a broken leg; the open-heart surgery that requires weeks of recuperation—imposes “harm,” we shall find harm prevalent both in the wrongful disability context and beyond.
I want to argue, however, that we need not go down that path. Instead, we can adopt an ordinary, comparative, “worse for” sense of harm, according to which agents harm a person just in case they could have created more wellbeing for that person and instead create less. We can thus maintain the important distinction between making a person suffer, however seriously, in a way that is required to secure a certain level of wellbeing for that person and harming that person. We recognize that the fact that we each are doomed to lives that include substantial quantities of psychological and physical pain will not automatically mean that each of us has been harmed by the acts that have brought us into existence. And we can retain the intuitive idea that both failing to benefit when one might have and burdening when one need not have will constitute harms; and we will have no reason to think that a “harm” that is necessary to secure a benefit truly is a harm, while a “harm” that prevents a still greater harm is not. We can, instead, take the position that neither such adversity is genuinely a harm at all.
2. The many harms and victims of wrongful disability
2.1 Two bad inferences to “no harm done.” We think of wrongful disability as one variety of procreative choice. And it is. But thinking of wrongful disability in that way may lead us to focus on just one sort of effect that wrongful disability might have—its procreative effect. In fact, however, wrongful disability may also have profound distributive effects. Plausibly, the wrongful disability choice that leads to the child’s unavoidably flawed existence does not harm the child in virtue of the procreative effect it has on the child. The logic of the nonidentity problem perhaps tells us that much. But it is a mistake to think that the wrongful disability choice cannot also harm the impaired child by virtue of its various distributive effects. And it is still another mistake to think that those distributive harms—whether they fall on the impaired child or persons other than the impaired child—somehow are not morally significant or that they can somehow be justified. The better view is that the choice of wrongful disability imposes certain distributive effects as well as procreative effects, and that those distributive effects can themselves constitute unjustifiable harms to (1) the impaired child, (2) persons other than the impaired child or (3) both.
Most theorists have overlooked the distributive harms that the choice of wrongful disability may impose on the impaired child. Having been persuaded by the logic of the nonidentity problem that the procreative effect of that choice does not constitute harm to the impaired child, theorists, including Buchanan et al., seem to have leaped without further ado to the conclusion that that choice imposes no harm at all on that child. But that is a bad inference. There is no basis for thinking that distributive harms—and unjustifiable distributive harms, at that—cannot befall the impaired child.
Why theorists have seemed not to focus closely on the distributive harms imposed by wrongful disability on persons other than the impaired child is harder to explain. However, the thinking may have been that, provided that the parents have distributed scarce family resources between the impaired child and persons other than the impaired child—e.g. that child’s older or younger siblings—in an appropriate way, any loss of wellbeing—any harm—suffered by such persons would be perfectly justifiable within a person-affecting approach. But that, also, is a bad inference. That the parents, having chosen to produce the impaired child, then make the further choice to distribute wellbeing appropriately between that child and others—and, hence, not to unjustifiably harm those others by that further choice—does not imply that the harm imposed by the original choice can itself be justified. Those are two distinct choices—and two distinct sets of distributive effects.
This is just an outline of how we can trace distributive, unjustifiable harms imposed on both the impaired child and others back to the choice of wrongful disability. In what follows, I consider the issue in more detail, first, in connection with persons other than the impaired child and, second, in connection with the impaired child.
2.2 Unjustifiable harms to persons other than the impaired child. A person-affecting approach, of course, will be concerned with the plights of persons other than the impaired child. The authors’ principle M itself makes that point. By its own terms, principle M protects not just the impaired child but rather all persons for whom the agents are “responsible.”
For purposes here, let’s assume that the parents themselves have not been harmed by their choice to produce the impaired child. Nor would they have been particularly burdened had they chosen to bring a healthier child into existence instead, so long as we restrict the cases under consideration here to cases that satisfy condition A. The choice is, then, one that is neutral from the parents’ point of view. But that result does not address the situation of persons other than the parents—persons, as the authors put it, for whom the parents are “responsible.” And that class of persons is an important one. Nor is it a particularly narrow one. The concept of responsibility at play here is, after all, not a legal concept but rather a moral concept. That means that any plausibly articulated person-affecting approach will be concerned not just with the impaired child and that child’s parents but also with—for example—any earlier and later born offspring, the parents’ own parents and potentially many other persons as well.
The theoretical point is this. A particular case of wrongful disability—a case that satisfies condition A—can challenge the person-affecting approach only if we are confident that the wrongness of the choice cannot be explained by reference to facts that are themselves person-affecting in nature. If the assessment of wrong can be explained, for example, by reference to the harm that has been done to a particular person for whom the agents are responsible, then the case cannot count against the person-affecting approach. There is, however, an important caveat here. We cannot explain why a particular choice is wrong, under a person-affecting approach, unless that choice harms some person or another. But nor can we explain why a particular choice is wrong, under a person-affecting approach, in a case in which any harm that choice does impose is a harm that can be justified within the person-affecting approach itself. As noted earlier, agents routinely face tradeoff scenarios, and such scenarios inevitably involve harm to one person or another. In any tradeoff scenario, the agent can create more wellbeing for one person only by creating less wellbeing for another—only, that is, by harming another. However, we do not think that such harms themselves have any particular moral significance—we do not, that is, consider them to be instances of wrongdoing—so long as the agent has distributed wellbeing between the two competing individuals in an appropriate way. When the agent has done that, we will then say that the harm is justifiable, and that no wrong has been done.
The upshot is that the sorts of cases that can potentially challenge the person-affecting approach will need to satisfy, not just condition A, but condition B as well:
B. No unjustifiable harm to persons other than the impaired child: The procreative choice fully takes into account the wants, needs and interests of existing and future persons (other than the impaired child) for whom those agents are morally responsible; i.e., the choice confers on each such person either (i) a maximally high level of wellbeing, thus negating the claim that that act harms that person, or (ii) an appropriately high level of wellbeing, that is, a level of wellbeing that reflects an appropriate distribution of wellbeing between that person and each other existing and future person for whom the agents are morally responsible, thus negating the claim that that act imposes an unjustifiable harm on that person.
A person other than the impaired child whose wellbeing is maximized by the choice under scrutiny—who has not, that is, been harmed by that choice—will satisfy clause (i) of condition B. A person other than the impaired child who has been harmed by that choice in a way that is nonetheless justifiable will satisfy clause (ii) of condition B. Thus, any case that satisfies condition B will have the following feature: facts about how the choice under scrutiny will affect persons other than the impaired child cannot ground a person-affecting account of why that choice is wrong. If those are, in other words, the only facts the person-affecting approach has to work with, then the person-affecting approach will be forced to say that the choice under scrutiny is perfectly permissible.
The difficulty for the objection against the person-affecting approach is that, in a very large number of cases, satisfying condition A means failing condition B. It is true that the choice to produce the impaired child will not have any procreative effect on persons other than the impaired child. But it will have a predictable array of distributive effects on those persons. It will require that parents stretch their scarce resources considerably farther than they would have had they chosen to produce a healthier child, or no child at all, instead. Parents will be forced to do less for any earlier or later born offspring in order to do what we think they clearly must for their impaired child. Why, then, do the authors seem to suppose, notwithstanding the obvious adverse distributive effects generated by the choice of wrongful disability, that the earlier and later born offspring are not harmed? The procreative effect of wrongful disability on the impaired child is unavoidable if that child is to exist at all. But the distributive effects on persons other than the impaired child are a different story entirely.
It seems that the authors cannot realistically deny that, in a very large number of cases, wrongful disability will impose distributive harms on persons other than the impaired child. Why do they not consider that particular fact highly relevant to the person-affecting analysis of wrongful disability?
Plausibly, the authors are here relying on a certain tacit assumption—the assumption that those harms can be justified. Any mundane case may seem to support that particular assumption. Consider the case of parents who had every intention of giving an already-existing child p an excellent, expensive college education entirely appropriate for p in light of p’s tastes and talents. Suppose that the parents have saved accordingly. When they then opt to produce a severely impaired second child q, they will need to rebalance their resources. They will otherwise be unable to discharge the substantial obligations they have now incurred—obligations that they are morally and legally required to discharge—with respect to q. As good parents, they will not both produce q and then proceed through life as though q has no special needs. They ought not treat p just the way they would have treated p had they not opted to produce q as well. Thus, they shift their savings into a trust for q to insure that q will receive the medical, educational and social support q will need both as a child and as an impaired adult. p ends up at a less expensive and less appropriate college. But resources and wellbeing often come together. Parental resources are normally limited; well-deployed additional resources will normally mean additional wellbeing. The less expensive college education means, in light of p’s tastes and talents, a loss of wellbeing for p, that is, a harm to p. It is not the gravest of harms, to be sure, but it is a harm.
Yet it is a harm p is made to suffer for a very good reason. It is a harm imposed on p so that a still graver harm will not be imposed on q. Let’s even make the plausible assumption that the parents have valiantly sought to balance wellbeing between p and q in an appropriate way and that the balance that they have achieved is itself both wise and fair. On those facts, it might seem that the harm imposed on p is perfectly justified—and thus that the choice under scrutiny will surely satisfy condition B. After all, clause (ii) of condition B requires simply that p’s wellbeing be made appropriately high—it need not be maximized. But if this fairly mundane instance of a case that satisfies condition A also satisfies condition B, then surely most instances of wrongful disability that satisfy condition A will also satisfy condition B.
In fact, however, this inference is a mistake. It assumes an implausibly weak test for determining when harm can been justified. On that test, the class of persons with respect to whom wellbeing must be appropriately distributed includes q. On that test, any sacrifice that p is forced to make on q’s behalf—any diversion of resources and wellbeing from p for the benefit of q—will constitute a harm to p that is justifiable as long as the agents have not gone too far to accommodate q’s special needs at p’s expense.
But that inclusive reading of the harm justification test gives agents far too much latitude in making the wrongful disability choice—or, indeed, any other procreative choice. All the agents need to do to insure that the harm imposed by their choice is justifiable—to insure, that is, that their choice satisfies condition B—is to distribute wellbeing in an appropriate way between the person or persons they have chosen to bring into existence and any other persons who do or will otherwise exist. But that cannot be right. That way of looking at things leads to still another form of the repugnant conclusion. Suppose, for example, that the agents choose to bring ten or twenty or a hundred or a million severely impaired—or, for that matter, perfectly able!—children q1, q2, . . . , qn into existence. They then make the further choice—quite naturally, and in a way that may be morally and legally required—to divert resources and wellbeing away from p and in favor of q1, q2, . . . , qn. And suppose that they make that choice in a way that distributes wellbeing among p and q1, q2, . . . , qn in an appropriate way. Under the weak, inclusive test, the harm that is then imposed on p must be considered justified. However badly and avoidably p suffers, that harm cannot ground, under a person-affecting approach, a finding of wrongdoing.
In fact, however, any plausible test for harm justification will be significantly more stringent than what we have just described. The choice under scrutiny is the procreative choice whether to bring q into existence or not. It is not the separate and independent choice (the “distributive” choice) to distribute wellbeing between p and q in an appropriate way on the assumption that q does or will exist. In assessing that procreative choice, a person-affecting approach must compare (1) the effects of that choice on each person, including p, against (2) the effects of each alternative choice, including those that exclude bringing q into existence to begin with. When we do that comparison, we find that when the agents choose wrongful disability they create less wellbeing for p when they had the alternative of creating more. That constitutes harm to p. Moreover, since at least one such alternative—an alternative that leaves q out of existence altogether—both creates additional wellbeing for p and distributes wellbeing between p and each other person who does or will exist under that alternative in an appropriate way, that harm to p is unjustifiable.
We have here taken the harm justification test to be exclusive in nature. And that seems right. It is surely correct that it is not the case that (1) harm to p imposed by the choice of wrongful disability is necessarily justified in virtue of the fact that (2) if q were to come into existence, then the agents’ choice to divert resources and wellbeing away from p and in favor of q would be justified. That test for harm justification is not stringent enough. Instead, for purposes of determining whether the choice to bring q into existence harms p in a way that is unjustifiable, we must take into account that the alternative exists for the agents of not bringing q into existence to begin with. Under that alternative, the class of persons in respect of which wellbeing must then be distributed in an appropriate way will be one that excludes q.
2.3 Does this account devalue the impaired child? Three objections should be considered at this point. (1) It might seem that the preceding account of wrongful disability—including the more stringent, exclusive test for harm justification—involves us in the shady business of ignoring q’s wants and needs should q exist. It may seem to require us to accept the heinous result that no diversion of resources and wellbeing away from p and in favor of q—that no harm to p—could ever be justified, even when it is a given that q does or will exist—even, that is, when the agents no longer have the alternative of not bringing q into existence to begin with.
The person-affecting approach is not, however, committed to that result. After all, the person-affecting approach is not limited to the assessment of procreative choice. It will apply to other sorts of choices as well—including the distributive choices that agents make independently of their procreative choice. Thus, to decide the procreative choice whether to bring q into existence, we exclude q from the balancing test. But in any case where q does or will exist—in any case in which the choice has been or will be made to bring q into existence—q must emphatically be included in the class of persons in respect of whom wellbeing must be appropriately balanced. If the parents make the moral error of bringing q into existence to begin with, they ought not, in other words, compound their error by then doing less than they ought on q’s behalf.
(2) It seems that agents had still another alternative for insuring that p suffers no harm—or, at least, no morally significant harm—short of refraining from bringing q into existence. They could have simply left p—and each other person for whom they are responsible—out of existence altogether. If the parents are not responsible for anyone other than q—if, e.g., they choose not to produce p or anyone else for whom they would then be responsible—their own scarce resources can be devoted to q’s needs alone. But if that is correct, then the person-affecting account will not dictate that it is wrong for the parents to choose to bring q into existence. Instead, it will say that parents ought either choose not to bring q into existence or choose not to bring p (or still others) into existence.
I think this point is in fact correct—putting aside, for the moment, that we may well be “responsible” for many, many persons other than those we choose to bring into existence (after all, the concept of responsibility that is invoked here is a moral concept, not a legal concept). But it is a caveat, not a refutation. We take it as a given that persons other than the impaired child do or will exist for whom the parents are responsible. The choice under scrutiny, then, is whether it is permissible for the parents to bring the impaired child into existence as well. We can certainly revise that underlying supposition if we want to do so. We can make it a given that the impaired child does or will exist, and then assess whether it is permissible for the parents to bring another child—perhaps a non-impaired child—into existence as well. But that would be another case.
(3) The preceding account may seem to fail to acknowledge the many benefits that will accrue to p as a result of the impaired q’s being brought into existence—benefits that will make up for any diversion of resources and wellbeing from p and favor of q. Those benefits will be plentiful. After all, q will love and be loved, and the affectional relationship between p and q will be good for both individuals. Moreover, even as p makes sacrifices so that the needs of q can be met, p will be accruing gains in maturity and character.
I do not dispute any of these claims. At the same time, we would not gloss over the loss to an older sibling imposed by a diversion of resources and wellbeing from that child to an impaired younger sibling if the impairment were due, not to a genetic disease, but rather to an intentional assault, medical negligence or drunk driving. Nor would we sentimentalize things with the idea that it good to learn early in life the virtue of sacrifice. It is; but we would not leap from that fact to the idea that, on balance, no harm has been done. We would instead say that there are other ways to learn early in life the virtue of sacrifice, and that siblings can have affectional relationships whether one of them is impaired or not. Why would things be different in the context of wrongful disability?
2.4 Why an adequate account of wrongful disability must go beyond harms to persons other than the impaired child. We have just seen that we cannot blithely assume that the choice of wrongful disability will never, or only rarely, impose harm on persons other than the impaired child, or that any harm that it does impose can be justified. Rather, in a very large number of cases, choices that satisfy condition A will fail condition B. In any such case, the person-affecting approach will imply that the choice of wrongful disability is, in fact, wrong.
That result is, however, qualified. It asserts, not a categorical truth that cannot be escaped no matter how exceptional the facts, but rather a claim that plausibly holds for “a very large number of cases.” The possibility is thus left open that, in still other cases, the choice of wrongful disability will not impose unjustifiable harms on persons other than the impaired child. Some cases will, in other words, satisfy both condition A and condition B.
In respect of that latter, smaller class of cases, it seems that the authors would nonetheless take the position that the choice of wrongful disability is nonetheless wrong. I do, too. I believe that the authors’ position is at least approximately correct.
A second, related point should be noted as well. Any account that explains the wrong of wrongful disability solely by reference to harms to persons other than the impaired child will seem to miss an important point. Such an account, as we have just seen, generates, for a very large number of cases, the intuitively correct result. However, in failing to locate the wrong in the vicinity of the impaired child, it seems to generate the right result for the wrong reasons. If the choice of wrongful disability is wrong, it would seem to be wrong, at least in large part, because of what it does to that child and not what it does to that child’s relatively healthy older and younger siblings. Misidentifying the basis for the wrong is not simply troubling in some vague way. It would also yield some very concrete results that we know cannot be correct. In particular, if our analysis of wrongful disability ends with the appeal to harms to persons other than the impaired child, it would be all too easy for parents to escape the wrong of wrongful disability. They would need simply to take steps to insure that any harm their choice happens to impose will be confined to (1) themselves and (2) their impaired child. As long as persons other than the impaired child for whom the parents are responsible are not made to suffer any particular loss of wellbeing—as long as those persons are shielded from the distributive harms of the parents’ choice—then the wrong of wrongful disability would seem simply to evaporate. But surely it does not. In fact, in the case we are now imagining, the wrong of wrongful disability may seem more transparent than ever.
2.5 Unjustifiable harms to the impaired child. These points suggest that an account of wrongful disability that appeals solely to harms to persons other than the impaired child would be problematic. But there is no reason to think that the person-affecting account would make that mistake. There is no reason, in other words, to think that a person-affecting account of harm would not itself look beyond harms to persons other than the impaired child and take into account, as well, harms to the impaired child.
However, when we begin to examine the plight of the impaired child, it may seem that we again face a dearth of harm. After all, the logic of the nonidentity problem convinces us that it cannot be in virtue of the choice’s procreative effect on the impaired child that that child has been harmed. Moreover, we can carry forward the earlier assumption that the parents themselves are not harmed—or at least not unjustifiably harmed—by the choice they themselves have made. Finally, condition B insures that the choice does not unjustifiably harm anyone other than the impaired child.
But here, too, our ability to discern harm will depend on the scope of our scrutiny. The cases at issue here leave the impaired child as the only potential victim of the wrongful disability choice. Obviously, any adequately formulated person-affecting approach will be concerned with harms to that child. Just as obviously, any adequately formulated person-affecting approach will be concerned with harms imposed on that child, not just by virtue of the procreative effect of the wrongful disability choice on that child, but also by virtue of the distributive effect of that choice. The logic of the nonidentity problem shows that that procreative effect does not harm the impaired child. But that logic says nothing whatsoever about the distributive effect of that choice on that child.
To insure, then, in a general way that the case under scrutiny is one in which no harm–or, at least, no unjustifiable harm—has been imposed on the impaired child, we must do more than rely on the logic of the nonidentity problem. We must instead insist that the case under scrutiny satisfies condition C.
C. No unjustifiable harm to the impaired child: The procreative choice fully takes into account the wants, needs and interests of the impaired child; i.e., the choice confers on that child either (i) a maximally high level of wellbeing, thus negating the claim that that choice harms that child, or (ii) an appropriately high level of wellbeing (a level of wellbeing that reflects an appropriate distribution of wellbeing between that child and each other existing or future person for whom the agents are morally responsible), thus negating the claim that that choice imposes an unjustifiable harm on that child.
About those cases that satisfy both condition A and condition B, then, we say the following. For any such case that fails condition C, the wrong of wrongful disability can be grounded in the harm—that is, the unjustifiable harm—that choice imposes on the impaired child. The problem class of cases, then, consists of those that satisfy condition C. It is those cases, in other words, that challenge the person-affecting approach. As to those cases, we can be quite sure that the choice of wrongful disability harms—that is, unjustifiably harms—no one at all. Yet it may seem that the choice itself is clearly wrong.
What I now want simply to point out here is just how rare would be the case that manages to satisfy conditions A, B and C. Almost any case, in other words, that satisfies condition B will fail condition C.
Let’s start with a clear picture of what a case that satisfies condition B will look like. Having made the choice to bring the impaired child q into existence, the parents can avoid imposing unjustifiable harm on persons other than q—any earlier or later born offspring or others for whom the parents are responsible—only by not diverting resources and wellbeing from those other persons to q. The parents can, in other words, avoid imposing unjustified harm on p—for example—only by making the tradeoff that they now confront in a certain way. But the needs of q predictably will be substantial. Not diverting resources and wellbeing from p to q in order to meet q’s needs will clearly harm q. After all, the loss of wellbeing q will suffer is perfectly avoidable. q could have existed and not suffered that particular distributive effect. In that respect, the distributive effect on q is quite unlike the procreative effect. Moreover, the harm to q is unjustified. The tradeoff that the agents are forced to make has been made badly. A distribution of wellbeing that vastly favors the less needy p and badly deprives the impaired q will clearly not constitute an appropriate distribution of wellbeing between p and q.
We can put this point in more general terms. We can certainly imagine a case in which the parents take steps to shield p and other persons for whom they are responsible from the adverse effects of the parents’ choice—take steps, that is, to insure that no such person is unjustifiably harmed by that choice. Such a case will satisfy condition B. However, in a world where resources are limited for the vast majority of parents, where do the parents find the resources necessary to protect p and all those other persons for whom the parents are responsible? They do not rob a bank. Rather, they take resources, and wellbeing, away from the supply that would otherwise be available to meet q’s special needs. They do less for q so that they can do more for p. But no one thinks that that is an appropriate way for the parents to balance wellbeing between p and q. The harm done to the impaired q thus remains unjustified.
Any case that satisfies condition B, then, will be very, very likely to fail condition C. Once we establish, however, that the case fails condition C—that an unjustifiable harm has been done to the impaired q—we will then be able to provide a person-affecting account of the wrong of wrongful disability.
2.6 Is the impaired child really harmed? The netting problem. The claim that the impaired child q has been harmed may seem to fail to take into account that the various effects on q need to be netted against each other to determine whether the choice of wrongful disability imposes any genuine harm on q. The distributive effect that choice has on q when resources and wellbeing are diverted from q to p, in other words, must be netted against the procreative effect that choice has on q when it brings q into an existence that is, overall, worth having. It is a minus for q that q is made to suffer an adverse distributive effect. But even in the case where condition B is satisfied and condition C is not—even in the case, that is, where the parents protect p and leave q to fend for himself—the net value of that existence (more precisely, the overall, lifetime level of wellbeing that accrues to q) would seem on balance to constitute a plus for q, or at least not a minus. Surely, when we net the minus of having one’s special needs left unmet against the many plusses of existence, we find that the choice of wrongful disability does not, after all, harm q. If anything, it benefits q.
We should agree that the values of the various effects for q should be netted against one another. The fact, however, that that operation generates an overall positive (or at least a neutral) value for q does not imply that q has not been harmed. An act may create more plusses for a person than it does minuses and still constitute harm. Consider, for example, the cardiac surgeon who meticulously performs an open heart procedure on a very ill patient. Suppose that the surgeon is well aware that medication would have been an equally effective treatment for that particular patient without subjecting the patient to pain, disability and weeks of recuperation. Suppose, too, that the surgeon wanted to operate solely for the purpose of increasing his skill level in a difficult procedure, and that, had he not operated, he would have walked out of the hospital in a pique without administering the critical medication. Had the surgeon not done just as he did, then, the patient would have died. On those facts, it is true that what the surgeon has done creates more wellbeing for the patient than it destroys. But he nonetheless has clearly harmed the patient. He has created less wellbeing for the patient when he could have created more.
The same analysis holds for wrongful disability and procreation generally. That the parents’ choice of wrongful disability benefits the impaired q on a net basis—that the overall level of wellbeing that has been conferred on q remains in the positive range—does not imply that that choice does not also harm q. The bare fact that the maltreated and impaired child is accorded an overall level of wellbeing that remains in the positive range does not imply that no harm has been done to that child.
2.7 The moral bind of wrongful disability. Of course, we can imagine still other cases in which condition C is satisfied but condition B is not. Imagine, for example, that morally misguided parents divert all their available resources for the benefit of their impaired child and leave all persons other than the impaired child for whom they are responsible to fend for themselves. The parents thus insure that the impaired child does not suffer any unjustifiable harm. But now we face a case in which persons other than the impaired child suffer harms that are themselves unjustifiable—just the kind of case that fails condition B.
The upshot is that parents who choose wrongful disability are placed in a moral bind. Cases that satisfy condition C are very likely to fail condition B. Cases that satisfy condition B are very likely to fail condition C. At least in the vast majority of cases, then, the choice of wrongful disability will impose unjustifiable harms on the impaired child q or on a person p other than the impaired child or (most likely) on both q and p. In the vast majority of cases, it is just a matter of who will suffer unjustifiable harm as a result of the parents’ choice to produce the impaired child. But the moral bind is one of the parents’ own making. It is, by hypothesis, a choice, something they need not have done. They had the alternative of not imposing unjustifiable harm on anyone at all since they had the alternative of not producing the impaired child to begin with. But to say that is to say that their choice of wrongful disability is itself wrong.
Where the parents harm p by diverting resources and wellbeing from p to q, it is the alternative of not producing q to begin with that shows that what they have done unjustifiably harms p. When the parents, instead, harm q by diverting resources and wellbeing from q to p, it is the alternative of distributing wellbeing more appropriately between p and q that shows that q is unjustifiably harmed. Provided, then, that there do or will exist persons other than the impaired child q for whom the parents are responsible, the only way out of the moral bind is for the parents not to produce q at all.
According to the person-affecting approach, then, what is morally objectionable about the choice of wrongful disability is the moral bind that it creates—a moral bind that adversely and unjustifiably affects the impaired child, persons other than the impaired child or (most likely) both. Thinking of wrongful disability in this way underlines that it is the impaired child who is most clearly placed at risk by the parents’ choice of wrongful disability. After all, the healthy older sibling faces the loss of a wonderful college education, while the impaired child faces growing old without adequate resources. In this way, the person-affecting account locates the wrong of wrongful disability at least roughly in the plight of the impaired child. In contrast, under principle N, the wrong of wrongful disability is to be located entirely in a certain impersonal effect—not in what has been or will be done to the impaired child, but rather in the bare fact that that child was not left out of existence altogether in favor of the healthier child. If that procreative effect on the impaired child is what is supposed to explain the wrong of wrongful disability, it does not seem to do a very good job. For that procreative effect does not, in itself, seem to constitute a harm. Moreover, if it is the purely impersonal effect that is intended to explain that wrong—as it seems to be—then the person-affecting account does a significantly better job in making a connection between the wrong that has been done and the plight of the impaired child.
2.8 Rare, residual cases of “no harm done.” To complete the person-affecting account of the wrong of wrongful disability, we need to consider one last class of cases. That class consists of cases that satisfy all three conditions A, B and C. In any such case, the person-affecting approach will consider the choice of wrongful disability permissible.
Such cases might arise when parental resources—time, energy and money—are vast. However, under any reasonably broad account of just what persons the parents are themselves responsible for—morally, not just legally—it should be clear that only in a very few cases, and perhaps in no actual cases, will family resources be so ample that parents can sidestep issues of unjustifiable harm altogether.
What do we say about those rare, residual cases of wrongful disability? No unjustifiable harm of any sort having been done to any person who does or will exist, the person-affecting approach implies that the choice of wrongful disability is not wrong. A theory is shown to be false whenever a single bona fide counterexample to that theory is put forward—a single actual or possible case, that is, that demonstrates clearly that the theory generates a false result. But it is difficult to see that the case at hand really is such a case since it is implausible that the parents’ choice to have the impaired child, in this highly unusual scenario, is clearly morally wrong. We may feel unsure of the choice’s moral status—perhaps because we are unsure that it really is a case in which no unjustifiable harm has or will be done to any person—but queasiness is not a counterexample.
If we do accept the idea that, in rare, residual cases, the choice of wrongful disability is permissible—or, at least, not clearly impermissible—genetics counselors and others who assume the role of moral advisor will have a more important role to play than ever. In particular, they will need both (1) to acknowledge that, at least in theory, the would-be parents’ choice of wrongful disability is permissible and (2) to convey an accurate understanding of just how rare the “no harm done” scenario will be in the real world. The bare recognition by the genetics counselor that the choice of wrongful disability is, if rarely, permissible will soften the moral message, which I think undeniably sounds harsh. But it need not obscure that message. Are the parents’ resources really so ample that they can meet the impaired child’s special needs without unjustifiably harming any earlier or later born offspring—or anyone else for whom they are or may become “responsible”? Will it really make no difference to any of those persons if the parents place a very large portion of their assets in a trust for their impaired child—assets sufficient to provide that child with the care he or she needs and deserves both in the immediate future and beyond—and contribute, on an ongoing basis, an equally large portion of their income to that trust?
At the same time, by focusing our attention on just how and what sort of harm is likely to arise out of the choice of wrongful disability, the person-affecting approach can help us avoid dramatizing the case against it. Thus, Buchanan et al. sensibly reject the “nondirectiveness norm in genetic counseling.” Why understand clearly that a choice is very, very likely to be morally wrong and not bother mentioning that fact? But the authors’ own condemnation of wrongful disability at some points—though not all—in their discussion seems exaggerated in a way that may (1) undermine the credibility of the genetics counselor and (2) lead physicians, medical insurers and even the parents themselves to misunderstand and mischaracterize the wrong of wrongful disability. The authors thus write:
Although reasonable disagreement exists about whether some failures to prevent harmful genetically transmitted conditions would be morally wrong . . . in other cases, such failures would be as clearly and uncontroversially wrong as cases of child abuse and neglect whose wrongness is not seriously questioned.
The authors’ focus on the procreative effect of wrongful disability may be what leads them to place wrongful disability and child abuse in the same moral category. It is in virtue of that impersonal, procreative effect—a child who is seriously impaired as a result of a genetic or other congenital disease or disorder—that principle N deems the choice of wrongful disability itself wrong. And that effect might be perfectly replicated in a child who is left seriously impaired as a result of a parental beating. Looking at things on the basis of those twin effects alone, we might well consider the acts that generated them to be morally on par.
According to a person-affecting approach, however, the wrong of wrongful disability is not grounded in its procreative effect—the procreative effect is not, in itself, a harm—but rather in the distributive effects that arise when scarce resources, and ultimately wellbeing, are diverted from the impaired child to other persons for whom the parents are responsible or from those persons to the impaired child. The wrong of wrongful disability is thus more akin to the wrong of denying the impaired child superb medical care in favor of run of the mill medical care—or the wrong of arbitrarily denying the older or younger sibling the excellent college education in favor of the run of the mill college education—than to the wrong of child abuse. It is, in the vast majority of cases, a bad thing, and, I have argued, a wrong, but it is not evil epitomized.
3. The impaired child’s legal claims for wrongful disability and wrongful life
3.1 Two Classes of Impairments and Two Claims Against Health Care Providers . The preceding discussion focuses on cases in which parents face the choice whether to bring an impaired child into an unavoidably flawed but worthwhile existence in place of a healthier child. In other cases, the coming into existence of the impaired child can be traced back not to parental choice but rather to the negligence of a health care provider. One example of physician negligence would be an obstetrician’s failure to inform a couple that they face an elevated risk of producing a child with a serious genetic or congenital disease or disorder or to recommend appropriate testing. Another example would a laboratory’s negligence in the performance of genetic testing or its failure accurately to report its results. The array of methods available for assessing the risk of impairment both pre- and post-conception includes the straightforward—taking a family history—and the very sophisticated—the genetic testing of the couple. Post-conception, procedures include pre-implantation genetic diagnosis (PGD), chorionic villus sampling (CVS), sonogram and amniocentesis. New noninvasive technologies for genetic testing may well be on the horizon in view of research indicating that fetal cells cross over into the woman’s bloodstream early in pregnancy.
In this part, I focus on the question of whether the impaired child has a valid legal claim in negligence against the health care provider in the case where accurate information very probably would have triggered the non-conception or abortion of that child.
Skepticism regarding harm and a sense that the calculation of damages would have no reasonable basis have led most courts to dismiss the impaired child’s legal claim. Such claims must, like their moral counterparts, somehow address the issue of harm. The authors’ analysis of the parents’ choice to bring the impaired child into existence suggests that, on their view, the assessment that what the health care provider has done is morally wrong can survive even if no one has been harmed. They can take that position because they accept a hybrid moral view, which includes, alongside the person-affecting principle M, the impersonal principle N. For legal purposes, however, things are quite different. The impaired child’s legal claim against the provider can survive only if we can articulate just how the impaired child, who owes his or her very existence to the provider’s negligence, can cogently claim to have been harmed by that negligence.
What we want ultimately to do is apply the work on harm from part 2 above to the issue of the validity of the child’s legal claim. Before we can do that, however, we must first make a distinction between two varieties of legal claims that impaired offspring might make—roughly, a distinction between two classes of impairments. The first claim includes the impaired child’s assertion that he or she would have been better off never having existed at all—that is, that the existence itself is less than one worth having and that the child’s being forced to endure it—that procreative effect—constitutes a harm to the child. The authors’ own examples of disorders that potentially fall into that class include Lesch-Nyhan disease and Tay-Sachs disease. The second claim includes precisely the same sorts of assertions we examined in part 2 in connection with wrongful disability except that the child here asserts that he or she is harmed as a result of provider negligence rather than parental choice. Thus, the claim does not include the assertion that the impaired child would have been better off never having existed at all. Examples of disorders that fall into this second class would be Down syndrome, Huntington’s disease and hereditary deafness.
Now, within the law, the impaired child’s claim, whether or not the facts plausibly suggest that the impairment is so severe that the existence is less than one worth having, is standardly referred to as a claim for “wrongful life.” Moreover, courts typically understand the child’s claim to assert that existence itself constitutes a harm—again, whether or not the facts plausibly suggest that the existence is less than one worth having. This blurring of the two varieties of “wrongful life” claims evolved, no doubt, from the need to distinguish the case where the impaired child’s claim against the provider asserts ordinary medical malpractice—the case where, but for the provider’s negligence, an injured child would have been born whole—from the peculiar claim that the provider’s negligence has caused, not the underlying genetic or other chromosomal disorder, but rather the child’s very existence.
Having made that distinction, however, we now require the further, more refined distinction. We have, after all, no basis for thinking that the two varieties of “wrongful life” claims cannot be rooted in two distinct assertions regarding the harm that has been done. We thus examine the question of harm for both varieties of claims, one at a time.
3.2 The Genuinely Wrongful Life Claim . Where the child claims that existence is less than worth having, the claim of harm is not challenged by the logic of the nonidentity problem. The obvious procreative effect of the provider’s negligence is itself asserted to constitute harm, and we have no need to scrutinize the facts of the case for the possibly less obvious distributive effects of that negligence. In this context, the procreative effect consists in the coming into existence of a child for whom it would have been better never to have existed at all. The authors’ own assessment is that existence itself in such cases constitutes a “torment” and a “burden.” Another way to put the point is to say that the genetic disease or disorder is accompanied by such a horrific level of misery and suffering and so few moments of peace or pleasure that the child’s lifetime wellbeing level will fall into the negative range. When we compare that to the zero level of wellbeing that would accrue to that child had he or she never existed at all, we find harm (harm in an ordinary, comparative, “worse for” sense). In the authors’ view, the person-affecting approach is adequate to account for the wrong that has been done. That seems correct. And that person-affecting moral picture—which includes the recognition that the impaired child has been harmed and that that harm grounds a finding of wrongdoing—supports the conclusion that the child’s claim of the genuinely wrongful life should be recognized as a legally valid claim in negligence.
3.3 The Wrongful Disability Claim . The court’s dismissal of the impaired child’s claim when the child’s existence is not plausibly less than one worth having may seem to occupy firmer ground. Thus the authors themselves, having accepted the logic of the nonidentity problem, make the point that the worthwhile existence cannot constitute a harm to the impaired child. “[W]e believe the common intuition that in the [context of wrongful disability] the woman wrongs her child should, on reflection, be abandoned . . . . there is no victim with a special moral complaint from having been made worse off as a result of the woman’s action . . . .” What applies when the impaired child’s existence is the result of parental choice would seem to apply as well, for the authors, when that existence is the result of provider negligence. If the suffering the child endures does not make the child a “victim” in the one case, it will not make the child a “victim” in the other. Of course, the authors can provide, by reference to the hybrid view, a moral basis for the impaired child’s claim against the provider. The impersonal principle N seems capable (possibly with minor adjustments) of deeming the provider’s negligence in bringing the impaired child into existence a moral wrong. But, in the absence of a finding of harm, that result will be legally beside the point.
In fact, however, this “no harm done” result comes much too quickly. Whether the case involves parental choice or provider negligence, we cannot validly leap from the fact that the child has not been harmed by virtue of the peculiar procreative effect of the provider’s negligence to the result that that child has not been harmed in any respect at all. The analysis of harm that is appropriate in evaluating the moral status of the parental choice applies just as clearly when we turn to evaluate the legal status of the provider’s negligence. In particular, while the procreative effect of the provider’s negligence on the impaired child plausibly does not constitute harm to that child, the many distributive effects of the provider’s negligence plausible do constitute harms—harms to the impaired child or that child’s older or younger siblings or both. Nor can those harms be dismissed as morally, and hence arguably as legally, insignificant on the grounds that they can be justified. As we have seen, under an appropriately stringent test of harm justification—an exclusive rather than an inclusive test—they can’t. Nor can we plausibly argue that netting burdens against benefits will necessarily establish that the impaired child (at least) has not suffered harm. As we have seen, that the maltreated and impaired child is accorded an overall level of wellbeing that falls within the positive range does not imply that no harm has been done to that child.
We have here simply widened the scope of our scrutiny beyond the procreative effect of the provider’s negligence to include its various distributive effects. Thus, we claim that the provider’s negligence harms the impaired child not by bringing that child into an existence of a certain sort—an existence that is, e.g., less than one worth living—but rather by necessitating a diversion of resources and wellbeing away from persons other than the impaired child in favor of the impaired child or from the impaired child in favor of persons other than the impaired child. The harm to the impaired child thus lies not in the procreative effect of the provider’s negligence but rather its adverse distributive effects. This fact does not, of course, ground the impaired child’s claim in isolation. Rather, it grounds a disjunctive claim of the following sort: the impaired child or that child’s older sibling or that child’s younger sibling or any two of those persons or all three have suffered serious and unjustifiable distributive harms as a result of the provider’s negligence.
Of course, in a particular case, it may happen that the parents have dedicated a very large portion of their limited resources—time, attention and money—to care for the special medical, educational and personal needs of their impaired child. It may then seem that the parents have therefore shielded the impaired child from any distributive harm, or at least from any unjustifiable distributive harm, that would otherwise have arisen from the provider’s negligence. But the provider’s moral problem does not go away in that case—any more than the parents’ does in the context of wrongful disability. Nor should the provider’s legal problem. For the parents’ distribution of resources and wellbeing in favor of the impaired child can be accomplished only by their diversion of resources and wellbeing away from persons other than the impaired child for whom the parents are responsible. Now, for legal purposes, the class of persons for whom the parents are “responsible” is, plausibly, not quite so extensive as it is for moral purposes. Nonetheless, it clearly includes the parents’ earlier or later born offspring. The upshot is that the parents shield the impaired child from unjustifiable harm only by imposing unjustifiable harm on their other children—their own earlier and later-born offspring.
In the end, then, the provider’s negligence will function, in the vast majority of cases, to create an impossible bind. Given their limited resources, all the parents can do is choose which child will be unjustifiably harmed. But here the bind is not the parents’ doing; it is the provider’s. The function of the damage award, on behalf of the impaired child and his or her older and younger siblings, would be to insure that no such unjustifiable harm would be suffered by any member of that collection.
It seems, then, that the impaired child’s wrongful disability claim, as well as the child’s claim of the genuinely wrongful life, should be recognized as a legally valid claim in negligence. Appropriate damages would reflect the difference between ordinary child-rearing expenses and the cost of rearing the impaired child, calculated in light of the nature of the impairment and including any extraordinary expenses for medical, education and personal care.
3.4 Overlapping claims. I argued above that where the child’s claim is that he or she would have been better off never having existed at all—that is, where the child’s claim is for the genuinely wrongful life—the procreative effect of the provider’s negligence constitutes a harm. I simply note here that that result does not imply that the distributive effect of the provider’s negligence does not also constitute harm. Distributive harm becomes the critical issue in the discussion of the wrongful disability claim—the claim that is made on the basis of facts that do not plausibly support the assertion that the child would have been better off never having existed at all. However, if that analysis is correct, then it applies, possibly with still greater force, in the context of the genuinely wrongful life claim as well. If distributive harm is imposed on the impaired child or his or her older or younger siblings when the child’s impairment is less severe, then still more acute forms of that harm may be imposed when the child’s impairment is more severe and additional family resources are required for the treatment and care of that child.
3.5 A final word on terminology. Making the distinction in an explicit way between the two varieties of the impaired child’s claim would have certain advantages. We could reserve “wrongful life” for the legal claim of the genuinely wrongful life—the claim that the flawed existence is less than worth having. And we could, following the authors, adopt the phrase “wrongful disability” for the legal claim that asserts harm—distributive harm—to the impaired child or his or her older and younger siblings or both but can do without the claim that the flawed existence is less than worth having. That latter assertion, after all, has struck many courts as puzzling and provocative; and, as noted, it is often made on the basis of facts that do not seem plausibly to support it. The adversarial argument, then, could focus on the far less mysterious proposition that provider negligence resulting in the existence of a seriously impaired child in place of a healthier child has clearly demonstrable effects on the family budget—effects that mean that needs that would, or should, otherwise be met are not.
4. Moral theory and the problem of indeterminacy in the law
If the authors are correct that morality has a hybrid structure and includes both person-affecting and impersonal elements, then we must rethink some accepted wisdom about the relationship between morality and the law.
In recent years, we have become accustomed to the naturalist idea that moral theory may have something useful to say about certain hard cases—including just about any future person case we can imagine, wrongful disability being a prime example—with respect to which the text of the law itself is indeterminate. But moral theories that can be applied in this way must have certain features. They must be relatively clear and not so hard to fathom that we cannot understand and apply them, and they must be complete enough that they can be tested across a reasonably broad collection of cases. We must, in other words, be prepared to endorse and defend any theory that we want a court to take seriously. It is unclear, however, that the authors’ hybrid thesis will meet that test. Until we see how that thesis is to be situated within a theory that is somewhat more complete, we will be in no position to see whether its internal tensions can be resolved or not.
There is a still more basic concern. The branches of the law that most closely bear on future person problems—tort law, family law and constitutional privacy law—structure their analyses of claims in terms of whether there exists—or will exist—someone or another who qualifies as a harmed person, or victim. On the tort law side, the law requires a plaintiff who claims to have been harmed by the defendant’s negligent or intentional conduct. On the constitutional side, if state regulation of intimate conduct central to our lives is to be judged consistent with the due process guaranties of the Fourteenth Amendment, that regulation must further a compelling aim or interest on the part of the state. It must protect a person—and, for this purpose, a future person would seem to do—from harm or from the risk of harm. If no person’s wellbeing is at stake—if the regulation functions (for example) merely to increase total, aggregate wellbeing—its constitutional status would be in serious doubt.
Because the law is structured along person-affecting lines, a moral account capable of providing any insight into hard future person cases must itself be person-affecting in nature. An impersonal principle, whether on its own or as an element of a hybrid approach, will not have a sufficiently good fit with the law to offer any guidance cognizable within the law. But that means that, in accepting the hybrid approach, we cut the lifeline between law and morality. We will be forced to accept that the law, its person-affecting structure intact, must be interpreted and applied in hard cases in a way that will be, from the moral point of view, arbitrary. The law will be looking for a potential victim, a harmed person, and morality will be denying the significance of finding any such person.
Of course, we always have the option of restructuring the law along impersonal, or hybrid, lines. There is no reason, in theory, we cannot require that damages be paid in the context of wrongful disability whether or not anyone has been harmed. (To whom? How much?) But it would be a shame to revolutionize the law in a radical way if its existing person-affecting structure is, in the end, both intuitive and adequate to account for the wrong of wrongful disability—and, perhaps most especially, if we ultimately determine that its present person-affecting structure coincides with the more plausible moral approach.
Thus, if the person-affecting account of wrongful disability I have suggested above is correct, then we are not, after all, under any pressure to adopt an approach that is impersonal even in part, and the possibility will remain open that moral theory has a significant role to play in resolving the problem of indeterminacy in the law and, in particular, hard cases involving future persons.
The argument of this paper is that a person-affecting approach can account for the wrong of wrongful disability. In that connection, three points are key. First, we must to widen the scope of our scrutiny to include the full range of harms imposed by the choice of wrongful disability. We must recognize, in other words, not just the procreative choice’s procreative effects, but its distributive effects as well. Second, we must recognize and reject the weak, inclusive test for harm justification. To justify harms to persons other than the impaired child, it is not sufficient that wellbeing be balanced, however appropriately, between those persons and the impaired child. Any harm imposed by the choice of wrongful disability needs to be justified, instead, under a more stringent, exclusive test—a test that requires us to take into account alternatives that exclude as well as those that include the impaired child’s coming into existence. Finally, we must realize that the fact that the impaired child’s life is itself worth living is not sufficient to show that that child has not been harmed. Here, again, it is the distributive effect of the choice of wrongful disability that is telling, and not the procreative effect. The impaired child whose special needs are left unmet—perhaps so that that child’s older or younger siblings can themselves be spared—is harmed, even if that child’s existence remains worth having.
It may seem that we have gone to a lot of trouble to provide a person-affecting account of wrongful disability. It may seem easier in some ways just to adopt an impersonal approach in whole or in part. The practical implications, however, of retaining a purely person-affecting approach make the effort worthwhile. The authors’ hybrid view—along with other approaches that attempt to integrate person-affecting and impersonal elements—creates its own set of difficulties. Moreover, when we adopt a person-affecting approach, we avoid creating a great chasm between moral theory and the law. Instead, we insure the good fit that will allow moral theory to be of some use in resolving hard future person cases in the law. Serious wrongs can thus be made subject to the law, on a person-affecting approach, without the lawyer’s having to litigate on behalf of a plaintiff who has been neither harmed nor wronged and who has had all his or her needs, wants and interests fully and completely addressed. Of course, all a court would need to do to recognize that impersonal “claim” would be to revolutionize the law thoroughly along impersonal lines. But that is no small task.
. Allen Buchanan, Dan W. Brock, Norman Daniels and Daniel Wikler, From Chance to Choice: Genetics and Justice (Cambridge University Press, 2000).
. From Chance to Choice, p. 226.
. See generally Derek Parfit, Reasons and Persons (Oxford University Press, 1987), pp. 351-79, especially p. 363.
. For purposes here, we can remain largely neutral with respect to the issue of what constitutes wellbeing but simply note that (1) wellbeing is plausibly understood to represent a complex of capabilities and (2) wellbeing levels may be increased by positive psychological experiences such as happiness and decreased by experiences of pain and suffering. For an account of wellbeing as the “capability to achieve functionings,” see Amartya Sen, Inequality Reexamined (Harvard University Press, 1992), pp. 1-55. Understood as Sen suggests, wellbeing has the sorts of concrete features that seem relevant for deciding issues of harm and, ultimately, permissibility, in both moral theory and the law.
. From Chance to Choice, p. 249.
. From Chance to Choice, p. 247.
. Eric Rakowski refers to the authors’ “hybrid view.” “Who Should Pay for Bad Genes?” 90 Calif. L. Rev. 1345, 1371 (2002).
. In contrast to the impaired child’s claim, the parents’ claim (for “wrongful birth”) against the health care provider has been widely recognized by the courts. See note 47 below.
. No person-affecting approach that fails to take into account the plights of persons who will exist as well as persons who do exist can be considered plausible. The class of persons who will exist—the class of “future persons”—varies, of course, depending on the choices agents make now and in the future. References to future persons thus must be understood to be relativized to the distinct possible contexts, or futures, or possible worlds at which the act under scrutiny is performed. Thus, the class of future persons at a world w may well be distinct from the class of future persons at a world w'. A person-affecting approach, understood, for example, as a form of maximizing consequentialism, will then require that, for each existing and future person at w, agents create at w the most wellbeing they can for that person; require that, for each existing and future person at w’, agents create at w’ the most wellbeing they can for that person; and so on. There is a caveat. Sometimes, agents cannot create more wellbeing for one person without creating less for another. In such tradeoff scenarios, the loss of wellbeing imposed on one person or another will need to be justified by appeal to additional principles (fairness, equality or, perhaps, leximin, according to which agents must first create additional wellbeing for the least well-off, then, for the next least well-off and so on). See note 12 below. See also Roberts, “A New Way of Doing the Best That We Can: Person-Based Consequentialism and the Equality Problem,” Ethics 112 (2002), pp. 315-350; and Child Versus Childmaker: Future Persons and Present Duties in Ethics and the Law (Lanham, Maryland: 1998), pp. 45-85 (“Is the Person-Affecting Intuition Inconsistent?”).
. From Chance to Choice, pp. 226 and 249.
. This principle should not be understood to imply that an agent does something wrong when that agent (1) fails to make a particular person better off but (2) could have made that person better off only by bringing one or more additional persons into existence who will then be made to suffer in an avoidable way. When deciding whether to create additional wellbeing for a particular person—by, for example, bringing a slave population into existence who will serve that person’s every need—we must consider, not just the effects of that choice on those who do or will exist if the choice is not implemented, but also the effects of that choice on those who do or will exist if the choice is implemented. Because those facts are themselves highly person-affecting in nature, a person-affecting approach is capable of taking them into account in its analysis. See Child versus Childmaker, p. 63, and Roberts, “Do the Merely Possible Matter Morally? A Modal Interpretation of the Person-Affecting Intuition” (unpublished manuscript) (discussing Caspar Hare, “Voices from Another World: Must We Respect the Interests of People Who Do Not, and Will Never, Exist?” Ethics 117 (2007) 498-523).
. When will harm be justified? When has wellbeing been distributed in some appropriate way among the relevant individuals who are in competition for it? We recognize that, in tradeoff scenarios, the imposition of harm on one person can be justified when it is necessary to avoid the imposition of a more severe harm on another person. Similarly, we recognize that harm typically cannot be justified when the distribution of wellbeing that it involves makes things worse for an already badly-off person and makes things better for an already well-off person. A full theory of harm-justification is considerably more complicated than this. See note 9 above and Roberts, “A New Way of Doing the Best That We Can,” pp. 338-49. For purposes here, however, this rough picture will suffice.
. From Chance to Choice, p. 226.
. From Chance to Choice, p. 227.
. From Chance to Choice, p. 226.
 . From Chance to Choice, p. 250.
. From Chance to Choice, p. 236. As the authors note, the “flaw” in the reasoning that an act or omission that takes place prior to a particular person’s coming into existence cannot harm that person is “thinking that the harm to be prevented is a harm only to the fetus or, even more implausibly, only to a possible person before conception. In each case, the harm to be prevented is a harm that will come to the child that the fetus will later become, a child who will uncontroversially be a full moral person.” From Chance to Choice, p. 228.
. From Chance to Choice, p. 227.
. From Chance to Choice, p. 232-35.
. See notes 33 and 51 below.
. From Chance to Choice, p. 249.
. Because Tay-Sachs and Lesch-Nyhan seem to involve so much more suffering than Down syndrome, such diseases would not be good examples of wrongful disability since those children’s lives may not seem unambiguously worth living. Claims involving those diseases thus may fall under the heading of “wrongful life” rather than “wrongful disability.” For discussion, see From Choice to Chance, pp. 233-35.
. Here the authors follow Parfit (1987, pp. 351-79). The type of nonidentity problem that includes the problem of wrongful disability can be described as a “can’t do better” problem. In that case, the agents cannot do more for the impaired child than to bring that child into the flawed existence. The genetic disease or disorder that creates the impairment is not one that the agents can cure or fully ameliorate; and the existence itself, though flawed, is worth having.
Other types of nonidentity problems arise as well. In particular, the probabilistic “can’t expect better” problem has been used with seemingly spectacular success to establish that a correct moral theory must be, at least in part, impersonal in nature, whatever we think about the problem of wrongful disability. That type of problem is exemplified by Kavka’s slave child and pleasure pill cases and Parfit’s depletion and risky policy examples. Gregory Kavka, “The Paradox of Future Individuals,” Philosophy & Public Affairs 11(2) (1981) 93-112; Parfit, (1987, pp. 351-379). It might seem that, in view of the “can’t expect better” problem, the struggle to retain a person-affecting account of wrongful disability is without point. If the “can’t expect better” problem proves that a purely person-affecting approach is unacceptable, then the result that the person-affecting approach can be made to work in the context of the “can’t do better” problem may seem unimportant. I have elsewhere argued, however, that the reasoning used to generate the “can’t expect better” problem is fallacious. See Roberts, “The Nonidentity Problem and the Two-Envelope Problem: When Is One Act Better for a Person than Another?” in Harming Future Persons, eds. Roberts and Wasserman, Springer (forthcoming 2008), and Roberts, “The Nonidentity Fallacy: Harm, Probability and Another Look at Parfit’s Depletion Example,” Utilitas 9 (2007) 267-311. For purposes here, I set the “can’t expect better” problem aside.
. From Chance to Choice, p. 249.
. The restrictions Buchanan et al. include in principle N are, of course, critical for that principle’s plausibility. Thus, if the only way for a couple to have a genetically related child is to go forward with the transfer of (say) a frozen embryo whose status with respect to Down syndrome is unknown, and if it would create a substantial psychological or emotional burden for that couple to have a genetically unrelated child (to make use of, for example, a donor egg or embryo), then principle N would not prohibit the wrongful disability choice; nor, apparently, would principle N require a woman who could easily become pregnant in the future, but is profoundly and immovably adverse to abortion, to abort a fetus found to have Down syndrome. From Chance to Choice, p. 249.
. 90 Calif. L. Rev. at 1371.
. Parfit, Reasons and Persons, p. 381-90. See also The Repugnant Conclusion: Essays in Population Ethics, eds. J. Ryberg and T. Tannsjo (Springer, 2005).
. From Chance to Choice, p. 254. More generally, Buchanan et al. are alert to Rawls’s critique of utilitarianism, which fails, Rawls writes, “to take seriously the distinction between persons.” See John Rawls, A Theory of Justice (Oxford University Press, 1972), p. 27. The authors thus identify themselves as, in part, impersonalists, but they do not want to be mistaken for aggregationists. By making it explicit that principle N applies only in “same number” cases, the authors avoid any implication that principle N applies in cases in which agents have the option of substituting in a group of two or more persons who are, in the aggregate but only in the aggregate, better off than the original. From Chance to Choice, p. 249 (citing Philip G. Peters, “Protecting the Unconceived: Nonexistence, Avoidability, and Reproductive Technology,” 31 Arizona Law Review 487-548 (1989) (suggesting that we appeal to the notion that “harm” can be avoided by the “substitution” of one person for one other person).
. 90 Calif. L. Rev. at 1378.
. From Chance to Choice, p. 254 (emphasis added).
. The specific hypothetical the authors explore is one in which “a woman is told by her physician that she should not attempt to become pregnant now because she has a condition that is highly likely to result in moderate mental retardation in her child. Her condition is easily and fully treatable by taking a quite safe medication for one month. If she takes the medication and delays becoming pregnant for the month, the risk to her child will be eliminated . . . . Because the delay would interfere with her vacation travel plans, however, she does not take the medication . . . . According to commonsense moral views”—views that in the end Buchanan et al. retain—“this woman acts wrongly.” From Chance to Choice, p. 244. The type of nonidentity problem described here is a “can’t do better” problem. It is that type of nonidentity problem that I attempt to resolve, in purely person-affecting terms, in what follows. Other types of nonidentity problem must be unraveled by a distinct treatment. See note 23 above.
. This view is not universally held. Capron, for one, raises doubts about it. Alexander Morgan Capron, “Genes and the Just Society: Punishing Reproductive Choices in the Name of Liberal Genetics,” 39 San Diego L. Rev. 683, 689 (2002)(“There are at least four reasons for doubting the conclusion that parents commit a moral wrong in these circumstances.”).
. Thus, some theorists have retained the semblance of the person-based approach while defining “harm” in a way that makes irrelevant the question of whether the act under scrutiny makes a person worse off than alternative acts do. Such a non-comparative approach to harm is perfectly able to identify “harm” in a wide range of cases. But it seems disconnected from our intuitive picture of what harm is and why it has moral import.
We might, for example, simply define “harm” to include pain, hardship or deficiency (beyond a certain threshold) even in cases in which wellbeing has been maximized and agents have no option for improving the individual’s situation. See Elizabeth Harman, “Can We Harm and Benefit in Creating?” Philosophical Perspectives 18 (2004), pp. 89-113; and Bonnie Steinbock, “The Logical Case for Wrongful Life,” Hastings Center Report 16 (1986), pp. 15-20. See also Cynthia Cohen, “Give Me Children or I Shall Die! New Reproductive Technologies and Harm to Children,” Hastings Center Report 26 (1996), pp. 19-27. A somewhat different tack proposes that a person can be wronged even in a case in which he or she has not been harmed. See Rahul Kumar, “Who Can Be Wronged?” Philosophy & Public Affairs 31 (2003), p. 112; James Woodward, “The Non-Identity Problem,” Ethics 96 (1986), pp. 804-31; and Woodward, “Reply to Parfit,” Ethics 97 (1987), pp. 805 (defining the extent of a “wrong” as the difference between a person’s actual state and an “unobtainable baseline” in which that person’s rights are “not violated”).
Such non-comparative accounts of harm help us explain the wrong of wrongful disability. But they leave us with the question of why the choice is wrong when no existing or future person will be made worse off in any respect at all by that choice. Moreover, they seem too broad. The surgeon who meticulously performs a very necessary open-heart procedure that saves a patient’s life will, under many such accounts, be deemed to have harmed that person. But this novel use of language risks disorienting our moral compass. The surgeon has, no doubt, caused the patient to suffer a certain amount of pain. But it does not seem that the surgeon has harmed the patient.
. David Benatar takes the contrary view. Better Never to Have Been: The Harm of Coming into Existence (Oxford: Clarendon Press, 2006).
. Cf. Seana Shiffrin, “Wrongful Life, Procreative Responsibility, and the Significance of Harm,” 5 Legal Theory (1999) pp. 117-58.
. If I am correct that the inference is bad, then the authors are far from alone in making the mistake. Not only does it very likely infect judicial reasoning (see part 3 below), but also may help to give rise to certain types of the nonidentity problem, in particular, the “can’t do better” problem. See note 23 above. Moreover, it appears in my own Child Versus Childmaker, pp. 135-77 (“Wrongful Life”), and “Distinguishing Wrongful From ‘Rightful’ Life,” The Jo. of Contemporary Health Law and Policy 6 (1990) 59-80. In both pieces, I argue for the validity of the child’s legal claim for wrongful life but only in the case in which the existence is less than one worth having. That argument now appears to me to be mistaken.
. I think that assumption can be contested; I see no reason why we can’t harm ourselves. Nonetheless, where the parents have chosen wrongful disability—where the case is not one in which the wrongful disability is a result of medical negligence—the assumption that they have suffered no harm or, at least, no harm that cannot be justified is a reasonable one.
. An effect, to constitute a harm, need not represent a grave loss. The claim that such harms are frequently imposed in the context of wrongful disability seems highly plausible. See, e.g., Jeffrey R. Botkin, “Line Drawing: Developing Professional Standards for Prenatal Diagnostic Services,” in Prenatal Testing and Disability Rights, Erik Parens and Adrienne Asch, eds. (2000), p. 330. David Wasserman disputes the empirical claim (“Botkin would drop the fiction that prenatal testing served any medical function for future children, but he argues that it does, or could serve, a legitimate medical function for actual parents and families—to protect them from harm arising from the birth and upkeep of children with severe impairments . . . . Botkin's ‘family welfare’ standard is an unstable hybrid, because there is only a tenuous link between the medical severity of the fetal abnormality and the welfare of the family . . . . A growing body of research suggests that families with severely impaired children do not differ significantly in stresses and burdens from families with normal children.”). Wasserman, “A Choice of Evils in Prenatal Testing,” 30 Fla. St. U.L. Rev. 295, 302-04 (2003). See also Botkin, “Prenatal Diagnosis and the Selection of Children,” 30 Fla. St. U.L. Rev. 265, 290 (“The literature tends to suggest that most families cope quite well with the demands of a impaired child and that the child is loved and supported as his or her own person and for what he or she brings to the family. I believe many or most of these parents would not consider the impaired child to have caused a negative impact on the family. So these issues are by no means straightforward. Yet, I think we can also say that successful coping with a significant disability requires an enormous investment of time, energy, money and lost opportunities.”) It seems plausible that such diversions—of “time, energy, money and . . . opportunities”—away from others for the benefit of the impaired child do impose harms to those others.
. It might be argued that it can be demonstrated that the choice of wrongful disability cannot impose harm of any form on any later born offspring, on the grounds that those offspring very probably would never had existed at all had the choice of wrongful disability not been made. That line of reason constitutes an example of a type of nonidentity problem I have called the “can’t expect better” problem. I argue, elsewhere, that that line of reasoning is fallacious. See note 23 above. For purposes here, I set the “can’t expect better” problem aside—and take for granted the intuitive notion that the choice to bring the impaired child into existence, if it can harm older siblings, can harm younger siblings as well.
. It is true that Buchanan et al. briefly refer to “the effects [of wrongful disability] on the parents and the broader society, such as the greater child rearing costs and burdens of having the moderately retarded child” instead of the unimpaired child. From Chance to Choice, p. 246. That reference, however, occurs in the context of a discussion of David Heyd’s narrow view that “genesis” choices are never adverse to the interests of future persons, including, for example, the future younger siblings of the impaired child, but only to the interests of already-existing persons. Heyd’s narrow view can be questioned.
. This is so, particularly if we adopt a capability theory of wellbeing. See note 4 above.
. The adverse distributive effects created by the choice of wrongful disability will be particularly acute in the very class of cases the authors hold up for scrutiny—those cases in which the choice not to bring the impaired child into existence does not create a substantial burden on any existing or future person other than, arguably, the impaired child. These are thus not cases in which the choice of wrongful disability offers a special benefit to anyone. The authors have excluded here the case of the older childless couple who has had difficulty conceiving and very badly wants a child, impaired or not, who is genetically their own. Particularly in such no-benefit cases, it seems clear that the choice of wrongful disability will very often harm someone or another. Because substantial resources will be required for the care of the impaired child, the wellbeing of someone or another—an already existing sibling, a future younger sibling, a grandparent, someone—will be diminished, in the vast majority of cases, by the agents’ choice of wrongful disability. Harm will be done.
. This assumes that it is a given that persons other than the impaired child do, or will, exist. When that supposition is changed, the person-affecting account will change accordingly. See part 2.3 above (objection (2)).
. From Chance to Choice, p. 255.
. From Chance to Choice, p. 256. At other points, however, the authors clarify that they do not, in fact, consider the impersonal wrong of wrongful disability to be as “seriously wrong” as certain person-affecting wrongs that involve like effects. From Chance to Choice, pp. 247 and 252. They, thus, do not accept Parfit’s “no difference” thesis. From Chance to Choice, p. 247.
. Carolyn Jacobs Chachkin, “What Potent Blood: Non-Invasive Prenatal Genetic Diagnosis and the Transformation of Modern Prenatal Care,” 33 American Jo. of Law and Medicine 9 (2005) (citing Y. M. Lo et al., “Presence of Fetal DNA in Maternal Plasma and Serum,” 350 Lancet 485, 485-487 (1997), and Lee P. Shulman, “Fetal Cells in Maternal Blood,” 3 Current Women’s Health Reps. 47 (2003)).
. This part focuses on the impaired child’s claim against the health care provider. A separate claim for “wrongful birth” may be made by the parents when the impaired child’s existence can be traced back to the negligence of the provider and not parental choice. The wrongful birth claim asserts that the parents have been harmed in virtue of the fact that they have been made responsible for the extraordinary medical care and other out-of-pocket expenses arising out of their child’s impairment when, but for the provider’s negligence, their responsibility would have been limited to the usual and ordinary child-rearing expenses they would have incurred had they produced a healthier child instead. In contrast to the child’s claim, the parents’ claim for wrongful birth is well-established. Courts that decline to recognize claims for wrongful birth may argue that the parents’ claim is, in effect, that the impaired life itself—that procreative effect—constitutes an injury to the parents. The parents’ claim, instead, should be understood to be grounded in the additional burdens, including the economic burdens, they must shoulder when they bear a severely impaired child in place of a healthier child. See Grubbs v. Barbourville Family Health Center, 120 S.W.3d 682, 689 (Ky. 2003) (approving Georgia court’s “unwillingness to say that an impaired human life amounted to a legal injury” and, on those grounds, rejecting the wrongful birth claim as well as the wrongful life claim).
It may seem that there is no need for courts to recognize the child’s claim when they are willing to recognize the parents’. The statute of limitations accounts for part of that need. Thus, where the statute of limitation may bar the parents’ claim within just a few years of the birth of the impaired child, the child’s own claim will survive, under applicable tolling provisions, until that child reaches the age of eighteen, twenty-one or beyond. Moreover, the parents themselves may choose not to sue. Perhaps they do not consider themselves particularly to have been harmed—and do not see that their offspring may have been harmed when they have not. But the children, including the duly represented impaired child, should have the opportunity to make a claim once they reach the age of majority. After all, it may well be their college education that is foregone in favor of a trust for their impaired sibling or a trust for the impaired child that is itself foregone in favor of an expensive college education for an older or younger sibling.
. South Carolina, for example, recently rejected the claim for wrongful life. Willis v. Wu, 607 S.E.2d 63, 69 (S.C 2004). The wrongful life claim has been recognized, however, in three jurisdictions. See Turpin v. Sortini, 643 P.2d 954 (Cal. 1982); Procanik v. Cillo, 478 A.2d 755 (N.J. 1984); and Harbeson v. Parke-Davis, 656 P. 2d 483 (Wash. 1983). The plaintiffs in these three cases suffered, respectively, from hereditary deafness, congenital rubella syndrome and fetal hydantoin syndrome. The second of those disorders arose from the mother’s contraction of German measles during her first trimester, and the third from the mother’s taking Dilantin to control epileptic seizures during pregnancy.
. Buchanan et al. (2000, p. 233).
. From Chance to Choice, p. 235.
. “Ultimately, the plaintiff’s complaint is that he would be better off not to have been born. Man, who knows nothing of death or nothingness, cannot possibly know whether this is true.” Gleitman v. Cosgrove, 49 N.J. 22, 63, 227 A.2d 689, 711 (1967)(Weintraub, J., dissenting in part). The idea that we know “nothing” of “death or nothingness” is controversial. Perhaps we do know something about death and nothingness. Plausibly, we know that the amount of wellbeing the impaired child would have been accorded had he or she never existed at all is zero. When the evidence suggests that the impairment brings with it such a horrific level of pain and suffering that the child’s wellbeing level will fall into the negative range, the claim that that existence itself cannot constitute a harm—under an intuitive, comparative, “worse for” account of harm—seems ungrounded. For further discussion, see Roberts, "Can It Ever Have Been Better Never to Have Existed At All? Person-Based Consequentialism and a New Repugnant Conclusion," Jo. of Applied Philosophy 20,2 (2003) 159-85.
. From Chance to Choice, pp. 225 and 232-35.
. From Chance to Choice, pp. 251-52.
. See parts 2.2 and 2.4 above. As noted above, it is, of course, not just the authors who have made this inferential error. See note 36 above.
. See part 2.2 above.
. Thus, application of what is called the “benefits rule” does not apply to show that the impaired child has not been harmed. The argument that the fact that, after netting burdens against benefits, on balance the impaired child will have a worthwhile life does not, as we noted in part 2.5 above, imply “no harm done.” For discussion of the benefits rule, see Kathryn C. Vikingstad, “The Use and Abuse of the Tort Benefit Rule in Wrongful Parentage Cases,” 82 Chicago-Kent Law Review 1063 (2007).
. Since each such person would have been a minor at the time the negligence takes place, the claim itself should be understood to be governed, not by the relatively short period provided by the statute of limitations for any claim in medical malpractice the parents themselves might make against the health care provider—any claim, that is, for wrongful birth—but rather the substantially longer period provided to minors within the context of medical malpractice in general.
A handful of courts have discussed the notion that the siblings themselves may have a claim against the health care provider. “When a child requires extraordinary medical care, the financial impact is felt not just by the parents, but also by the injured child. As a practical matter, the impact may extend beyond the injured child to his brothers or sisters. Money that is spent for the health care of one child is not available for the clothes, food or college education of another child.” Moscatello v. Univ. of Med. & Dent. of N.J. 776 A.2d 874, 882 (N.J. Super. Ct. App. Div. 2001). The court in that case nonetheless declined to recognize the siblings’ claims.
Other courts that have considered siblings’ claims in the wrongful disability context have come to similar conclusions. See generally Deana A. Pollard, “Wrongful Analysis in Wrongful Life Jurisprudence,” 55 Ala. L. Rev. 327, n. 8 (2004)(citing Azzolino v. Dingfelder, 337 S.E.2d 528 (N.C. 1985)). In Azzolino, the appellate court described the issue in detail. “The siblings claim they have been damaged by their brother's wrongful birth in that they must suffer the family hardships, financial, emotional and otherwise, associated with having a Down's Syndrome child in the family; and have been and will continue to be deprived of the full measure of the society, comfort, care, and protection of the plaintiff parents because of the extraordinary demands placed on them by Michael; and that defendants should have reasonably foreseen that their negligence would result in such damage to the siblings. Our research has not disclosed any appellate court decisions, of this or any other jurisdiction, in which the precise question presented here has been addressed. However, six jurisdictions have considered and rejected similar claims brought by siblings of a normal, healthy child born after a negligently performed sterilization procedure.”). 322 S.E.2d 567, 578 (N.C. 1984). The court, nonetheless, refused to recognize the siblings’ claims on the grounds that “there was no duty owed by defendants to the siblings.” 322 S.E.2d at 579. The supreme court affirmed the dismissal of the wrongful life claim.
The court’s analysis of duty, however, seems questionable. The husband is considered a proper plaintiff in an action for wrongful birth even though he is not the patient of the health care provider who has been held negligent. Distinctions can be drawn between the father and the siblings, of course. The siblings are not, for example, legally required to divert resources and wellbeing from themselves for the benefit of the impaired child. But what the father and the siblings have in common seems more pertinent: it will be from the siblings, as well as the parents, that resources will be, both predictably and as a matter of legal and moral necessity, diverted. It is hard to see why harm to the husband counts but harm to the siblings does not.
Michael B. Kelly describes a position that accords in many ways with the view I have described here. He writes:
The family has been injured; the family can recover . . . . Recognition of the family as the proper plaintiff would abate theoretical concerns about the identity of the individuals involved. When the child must sue individually, negligent genetic counsellors can exploit the difference between the plaintiff (who arguably was not injured by being born) and the injured party (who was not born and thus lacks standing). The problem disappears if the family can sue as an entity. The entity, not the individuals, deserves recovery. By analogy, the shareholders of a corporation at the time of judgment in favor of a corporation need not be the same as those at the time injury occurred, regardless of whether those earlier shareholders received compensation for their cause of action when they sold their stock to successors. Similarly, we need not take notice of the exact constituency of the family at the time of judgment, as long as the family unit is identifiably the family injured by the defendant's negligence. “The Rightful Position in ‘Wrongful Life’ Actions,” 42 Hastings Law Jo. 505, 571 (1991).
A distinction between Kelly’s proposal and my own concerns how the parents’ claim is to be treated. While Kelly’s proposal considers the collection consisting of the children and the parents to constitute a single entity (the “family unit”), my proposal would allow the court to maintain the distinction between the parents’ claim (for wrongful birth) and children’s claim (for wrongful disability). Maintaining that distinction would avoid what may be considered an objectionable tolling of the parents’ own claim under the statute of limitations. It would also allow damages collected on behalf of the children to be secured, in the form of a trust, for the children’s use.
Rakowski makes the point that the choice of wrongful disability predictably causes harm to others. But he also notes that that point does not on its own resolve the question of who should be required to pay, under the law, the extra costs associated with bringing the impaired child into existence. 90 Calif. L. Rev. at 1382-83. He raises that question in the context of a broader discussion of resource allocation and distributive justice. 90 Calif. L. Rev. at 1352-67. Once we identify the harms that have been imposed, however, and, applying the exclusive, stringent test for harm justification, agree that those harms are not justifiable, we can, it seems, use traditional negligence principles then to make the case that the health care providers whose negligence has caused those harms are liable in damages. We need not invoke principles of distributive justice.
. For further discussion of how courts have perceived claims for wrongful life, see Darpana M. Sheth, “Better Off Unborn? An Analysis of Wrongful Birth and Wrongful Life Claims under the Americans with Disabilities Act,” 73 Tennessee Law Rev. 641, 648-53 (2006). See also Wendy Hensel, “The Disabling Impact of Wrongful Birth and Wrongful Life Actions,” 40 Harv. C.R.-C.L. Law Rev. 141 (2005). Hensel also argues that the recognition of such claims by courts may, in the end, do individuals with disabilities more harm than good:
Wrongful birth and wrongful life suits may exact a heavy price not only on the psychological well-being of individuals with disabilities, but also on the public image and acceptance of disability in society. Rather than focusing on a defendant's conduct, as in a traditional tort action, both wrongful birth and wrongful life suits ultimately focus on the plaintiff's disability, a status that is at least partially a societal construction. Juries in such actions are required to evaluate whether a particular disability is so horrible, from the nondisabled perspective, as to make plausible the choice of abortion or contraconception by the parent, or non-existence by the disabled child.
40 Harv. C.R.-C.L. Law Rev. at 144. Claims for wrongful disability avoid at least one prong of Hensel’s critique. In contrast with the wrongful life plaintiff, the wrongful disability plaintiff does not claim that his or her own life is less than one worth living; the claim is not that the plaintiff has been harmed by virtue of the procreative effect of the defendant’s negligence. The claim is rather that the plaintiff or an older or younger sibling has been harmed in virtue of certain distributive effects caused by the defendant’s negligence.
A second policy objection to the analysis of the legal claim for wrongful disability I have given is that it seems to open the door to the (1) child’s moral and legal claim against the health care provider in the case where the provider’s negligence leads to the existence of an additional, unimpaired child and (2) the child’s moral claim against the parents (doctrines of intra-family immunity protect the parents from any legal claim) when they choose to bring into existence an additional, unimpaired child. In both cases, the child’s claim would be that he or she is made to suffer a substantial reduction in available resources when additional children are added to the family as a result of provider negligence or parental choice. It seems that such claims would, in theory, be valid under a person-affecting approach—and that it is that feature of the person-affecting approach that will, in the end, enable it to avoid Parfit’s repugnant conclusion. At the same time, it is overwhelmingly likely that, when the additional child is not impaired and when not too many of them have been added to the family, the already-existing child will suffer no loss of wellbeing at all. We might, then, recognize that in rare cases the child will, indeed, have a moral claim against his or her parents or the provider. But we can also understand that cases in which the existing child is truly harmed will be rare, and that the child’s claim, in any particular case, may not survive a motion for directed verdict.
. For a clear discussion of the constitutional issues, see John A. Robertson, Children of Choice: Freedom and the New Reproductive Technologies (Princeton University Press, 1994) especially chaps. 1 and 3-6.
. I am grateful to David Wasserman and an anonymous reader for this journal for their invaluable comments on earlier drafts of this paper.