In this book, legal scholar and philosopher John Oberdiek offers an elegantly written, meticulously argued, and highly original account of when it is morally permissible to impose mortal risks on others. Tort scholars and theorists have long examined the permissibility of risky conduct, but, as Oberdiek observes, their efforts have usually focused more on interpreting legal doctrine than on the more fundamental question of the morality of risking. And insofar as scholars have evaluated this more fundamental question, they have often provided a simplistic and normatively questionable answer: cost-benefit analysis or utilitarian balancing is the only realistic and sensible way to distinguish legally permissible from legally impermissible risky conduct.1 This answer is also reflected in the most common characterization of the famous (or infamous) Learned Hand test of negligence: an actor is negligent if but only if (i) she failed to take a precaution and (ii) the burdens or costs of taking that precaution outweighed the precaution’s benefits (in reducing the risks of harm.2 At the same time, Oberdiek notes, moral philosophers have paid relatively little attention to the moral evaluation of risky conduct.3 in part because they usually assume the existence of idealized conditions under which the outcomes of a person’s actions are certain. Turning the trolley (or shoving a fat person into its path) will cause the death of one; not turning it will permit the death of five. Framing an innocent person will prevent a mob from killing more people. And so on.
In contrast with these unpersuasive or overly stylized approaches, Oberdiek’s book is a very welcome and invigorating breath of fresh air. Oberdiek offers a rigorous, nuanced, and novel account of the morality of risking, an account that seriously engages with the difficult challenge of explicating the concepts of risk, a right against risk, and the permissible level of risk under contractarian principles. Although some aspects of the analysis might be questioned, this philosophically sophisticated work should provoke renewed attention to a terribly important and unduly neglected topic.
Oberdiek begins with a lucid explanation of the two major accounts of probability—subjective and objective. The risk literature employs these terms in a specialized sense: subjective probability refers to the individual actor’s belief about a risk, while objective probability typically refers to the relative frequency of an event (such as D dying in a car accident next year). Oberdiek concludes that purely subjective accounts are “suitably practical but insufficiently normative,” because they cannot explain why we might fault an actor for engaging in what the actor believes to be low-risk conduct even though the actor failed to respond to evidence that the risk was high. By contrast, purely objective accounts are “suitably normative but insufficiently practical” (P. 4), largely because of the reference class problem.
The problem is this: One can specify the class that includes the relevant event in innumerable ways. If D is characterized as an American male, we get one probability of his dying in a car crash. If D is instead characterized as an Indiana-born actor, or a regular cigarette smoker, the probability of death could be significantly lower or higher. Many tort teachers and scholars will not be familiar with “the reference class problem” under that description, but they will be well acquainted with the problem itself, a problem that has long bedeviled reasonable foresight criteria in duty and proximate cause analysis. “On a clear day, you can foresee forever.”4 It might seem unforeseeable (i.e. extremely improbable) that a person on a subway platform will be injured as a result of a subway employee pushing a passenger onto a subway car a significant distance away, but it is surely foreseeable that pushing a person carrying dynamite onto a subway car will injure any person within the normal detonation range of such an explosive. How one characterizes the relevant risk makes an enormous difference both to the “objective probability” of an event and to the foreseeability of a risk. Thus, the objective risk approach suffers from the fatal defect of practical indeterminacy.
The solution? In a phrase, “epistemic contractualism.” (P. 40.) “[W]e have an epistemically-inflected moral duty to characterize the risks that we impose in a manner that is interpersonally justifiable. Morality itself…places normative pressure on our factual beliefs.” (P. 42.) Oberdiek endorses an “evidence-relative perspective” that characterizes risks in a way that is neither too demanding of agents imposing risk nor (as the subjective belief perspective would be) insufficiently demanding of agents and insufficiently protective of “patients” (or victims). To give content to this idea, Oberdiek brings to the stage that hoary legal construct, the reasonable person. One has a moral duty to exercise reasonable care in investigating causally relevant facts before acting; but because moral norms are action-guiding, this duty must be sensitive to our limited epistemic capacities. In the course of this discussion, Oberdiek powerfully and persuasively criticizes the view of philosopher Judith Jarvis Thompson that the actual facts (determined ex post) must be the decisive criterion of one’s moral duty.
I share Oberdiek’s view that a reasonable person perspective is a persuasive middle-ground between a fact-based and a belief-based account of permissible risk-imposition. Nevertheless, his account raises several questions. First, in articulating the content of the perspective, Oberdiek asserts that “risks must be given their gravest characterization” (P. 7), in order to be justifiable to potential victims. Yet it is not clear what this amounts to.5 Oberdiek’s illustrations here could be more illuminating. For example, he says that proper respect for the interests of coal miners in West Virginia requires consideration of the risks that such coal miners face in the mines, not the (lower) risks faced by anyone living in coal country. (P. 62.) Fair enough. But should we consider the (higher) risks faced by inattentive coal miners? Coal miners with preexisting health conditions? Coal miners who choose not to use safety equipment? His injunction to employ the “gravest” characterization of the relevant risk does not resolve these questions.
Second, that injunction also begs the more fundamental question whether one can reliably and legitimately draw a distinction between “patients” (or victims) who suffer risks and agents who impose them. In many of our activities, such as driving a car or even walking down the street, we occupy both roles. To be sure, some people are pedestrians but never drivers. In such a case, perhaps there is more force to Oberdiek’s claim that we must employ the “gravest” characterization of the risk, in order to make up for the fact that the nondriving pedestrian is the more vulnerable party. (P. 64.) In the end, however, these issues point to the importance of distributive justice constraints on risk, constraints that Oberdiek largely neglects in this book. If risks are reciprocal and benefits are widely shared, arguably it is unduly restrictive to condemn risk imposition as frequently as the patient-sensitive “grave” characterization would require.
Third, in emphasizing that norms about permissible risk are action-guiding, Oberdiek seems to presuppose a fault-based account. Yet he does not provide an argument for privileging fault-based responsibility over strict responsibility. In tort law, of course, many strict liability doctrines exist—e.g., abnormally dangerous activities, animals, vicarious liability, and manufacturing defects in products. In morality, too, some scholars have argued that responsibility is sometimes strict.6 Put differently, a strict liability norm can provide that an actor has a duty to compensate even if the norm does not entail that the actor’s primary conduct should have been different.
In Chapters 3 and 4, Oberdiek explores the moral significance of risking and what a right against risking would entail. The most common answer is straightforward: imposing risk is wrong because one could thereby cause physical harm (such as property damage or personal injury) or other material harms (such as fear of injury, a type of emotional harm; or disrupting the lives of those who choose to avoid the risk). And no one denies that people have a right not to suffer a material harm. Oberdiek abjures the obvious answer, however. He argues that even in cases of pure risk—cases in which none of these material harms comes to pass—a person has a right not to be subjected to an impermissible risk. Oberdiek also worries that the obvious answer conflates the distinct questions whether the agent is culpable and whether the act is impermissible or wrongful.
And why is pure risk imposition potentially wrongful? Not because the conduct might cause material harm, but because it can diminish the autonomy of those subject to the risk. For example, a drunk driver diminishes the autonomy of those he endangers, even if they are unaware that they are put at risk, because that dangerous conduct forecloses certain options that would otherwise be available to those who were endangered, thus narrowing their set of worthwhile opportunities. Imposing risk is like laying a trap: “the trap takes away the option, or more accurately renders unacceptable the exercise of the option, of stepping where the trap has been set.” (P. 86.)
This is a highly original and bracingly provocative claim. And in some circumstances, it is intuitively very attractive. Consider the following example from philosopher John Locke. A person wakes up in a room, has a conversation with another, and makes no effort to leave because he enjoys the conversation. As it turns out, he was locked in the room. Locke characterizes the actor’s decision to remain in the room as voluntary but also concludes that the actor was “not at liberty not to stay.”7 There is indeed an important sense in which an actor unaware of his lack of choice and content with his current circumstances may nevertheless lack a genuine choice.
However, the autonomy explanation of why pure risking can be wrongful also raises significant questions. First, Oberdiek’s analogy to the laying of traps seems problematic. Much of the force of the analogy flows from the high level of culpability of the “trapper.” One who lays a trap has the intention or purpose to limit the freedom of another, precluding their physical mobility by ensnaring them within the trap. But risky conduct can be impermissible even if it is not intended to limit the autonomy of others. A speeding driver might simply intend to arrive at her destination more quickly or to enjoy the feeling of wind flowing through her hair, but it hardly follows that the risk she imposes is therefore permissible. Oberdiek’s real concern is the effect of risky conduct on others’ options, not the intention of the actor. ((Consider two examples, in both of which P is unaware of the risk.
- X lays a trap in a particular location by which he intends to confine P, and there is a 10% chance that P will fall into the trap.
- Y builds a ditch that completely encircles P and that will cause the same degree and duration of confinement if P falls into it, and there is an 80% chance that P will fall into the ditch. Unlike X, Y is entirely unaware of P’s presence. (If Y had reasonably used his epistemic faculties, he would have investigated first and would have discovered P’s presence.)
On Oberdiek’s account of autonomy as a range of options, X has restricted P’s autonomy to a much lesser extent than Y, yet X is the one who has intentionally “laid a trap,” while Y has (merely) negligently built a ditch. Laying a trap is thus a somewhat misleading illustration of his theory.))
The trap example displays another potential problem with Oberdiek’s autonomy approach: the example illustrates a restriction of locational autonomy, yet Oberdiek is also properly concerned with restrictions of many other types of autonomy. The trap, once triggered, prevents a person from moving or relocating elsewhere, but Oberdiek is concerned with the permissibility of a broad array of risks, including those posed by drunk drivers and negligent manufacturers of products. How exactly do these risks constraint autonomy? By making it unsafe for the actor to pursue certain options, Oberdiek explains. If speeding driver A is weaving in and out of traffic, a nearby driver B’s freedom of safe movement has been significantly restricted, even if B is unaware of A, since a small turn of the wheel by B could have resulted in catastrophe.
Examples such as these raise the deeper question whether Oberdiek has successfully identified the problem in pure risk cases. Is the diminished scope of B’s choices really what is at stake in the speeding driver case, or instead the diminished ability of B to lead his or her life free of risks of material harm such as death or injury? Consider a different case. D negligently conducts surgery on patient P, and P is now at much greater risk of a sudden fatal heart attack. Oberdiek is correct about one way that D has constrained P’s freedom (even if P does not know about the risk): P cannot now engage in high-exertion exercise, smoking, or prodigious eating as safely as before. But is this really why P would be upset to discover, on his deathbed, that D had greatly increased his risk of suffering a heart attack?
Or imagine a more extreme case: a surgeon accidentally leaves a sponge in a patient; the sponge could cause deadly internal bleeding at any time, but no individual choices the patient makes in later life will increase the risk of that fatal result. In one sense, this conduct surely must count as a dramatic restriction of P’s “autonomy,” just in light of the risk that the patient will die and lose the ability to make any future choices. But the conduct does not seem to restrict autonomy at all in the sense that Oberdiek emphasizes, of making the consequences of a person’s choices less safe.8
In the final portions of the book, Oberdiek offers an intriguing account of rights as the conclusion rather than the premises of normative arguments; rejects the distinction between infringing and violating rights; and explains how he would identify which risks are permissible and which are impermissible. The discussion of the last topic is tantalizing but incomplete. Relying on a contractualist framework, Oberdiek argues that a risk is permissible if no person, considering the intrapersonal benefits and disadvantages of a risky activity, could reasonably reject the principle that justifies that activity. This approach is illuminating, but I do wish that Oberdiek had been more specific in identifying the threshold of permissibility—for example, he could have more explicitly compared his approach to different conceptions of negligence in tort law.
Oberdiek recognizes that his approach might appear to be too restrictive of risky human activity. In reply, he suggests that in assessing the justifiability of specific instances of risky conduct, we must consider the implications of this assessment for “any activity with the same risk profile.” (P. 13.) On this wider view, “even those who do not participate in or reap the benefits of certain risky practices nevertheless do participate in and reap the benefits from imposing risk as such.” (Id.) Thus, even if Amish farmers reject modern transportation, they probably do not reject, and could not reasonably reject, the more general principle that activities with similar benefits and risks are acceptable or desirable. This reply is promising, but it does provoke a worry. In generalizing in this manner, is Oberdiek surreptitiously reintroducing an interpersonal aggregation criterion that he elsewhere critiques?
Let me conclude by emphasizing that this book is a creative, thought-provoking, carefully argued, and immensely clarifying analysis of the morality of risking. Tort theorists, moral philosophers interested in risk, and tort scholars engaged by reasonableness and negligence doctrines will learn a great deal from Oberdiek’s arguments, whether or not they find all of the arguments fully persuasive.
- See, e.g., Heidi Hurd, The Deontology of Negligence, 76 B.U. L. Rev. 249 (1996); Richard A. Posner, Economic Analysis of Law (9th ed. 2014). For a critique of Hurd, see Kenneth Simons, Deontology, Negligence, Tort, and Crime, 76 B.U. L. Rev. 273 (1996).
- For a different characterization of the Learned Hand test as consistent with nonconsequentialist principles, see Gregory C. Keating, Must the Hand Formula Not Be Named?, 163 U. Pa. L. Rev. 367 (2015); Kenneth Simons, The Hand Formula in the Draft Restatement (Third) of Torts: Encompassing Fairness As Well as Efficiency Values, 54 Vand. L. Rev. 901 (2001).
- However, this understates a bit the recent flurry of scholarship on the topic, See, e.g. Johann Frick, Contractualism and Social Risk, 43 Phil. & Pub. Affairs 175 (2003); Aaron James, Contractualism’s (Not So) Slippery Slope, 18 Legal Theory 263 (2012); Rahul Kumar, Risking and Wronging, 43 Phil. & Pub. Affairs 27 (2015) (all cited by Oberdiek at page 13, n. 5); Joseph Raz, Responsibility and the Negligence Standard, 30 Oxford J. of Legal Stud. 1 (2010); Seana Valentine Shiffrin, The Moral Neglect of Negligence, in Oxford Studies in Political Philosophy, Volume 3 (2017).
- Sturgeon v. Curnutt, 29 Cal. App. 4th 301, 307 (1994), citing Thing v. La Chusa 771 P.2d 814 (1989).
- In a footnote, Oberdiek states without further explanation: “I do not wholly endorse tort law’s elaboration of the reasonable person.” 48 n. 30. He does not clarify elsewhere how his moral account differs from tort law’s account.
- See , e.g., R.A. Duff, Strict Responsibility, Moral and Criminal, 43 J. Value Inquiry 295 (2009). The widespread view that moral luck justifies greater blame or punishment might also be interpreted as a partial endorsement of strict moral responsibility.
- John Locke, Essay Concerning Human Understanding, Book II, Chapter XXI (originally published 1690). Oberdiek does not discuss this example.
- Here is another way to think about the foreclosed options argument. If I am sitting in a chair and you fire a gun at me that will kill me only if I sit perfectly still, you have foreclosed only one of my options—remaining still in the chair. If instead you set up a contraption such that if I were to move the slightest bit, the contraption will kill me, you have foreclosed every single one of my options except the one I am presently exercising, of sitting still in the chair. From Oberdiek’s analysis, it seems to follow that the risk posed by the contraption has a much more serious effect on my autonomy, and is thus much more morally troublesome, than the risk posed by the aimed gun. Yet, if I really do plan to sit in my chair for a while, and you really do plan to kill me, this conclusion is implausible. The firing of the gun seems to interfere with my rights in a more serious way than does the contraption, precisely because the risk of serious harm that it poses is much greater.