Ethics Book Review 130 Harv. L. Rev. 659

Law and Moral Dilemmas


A runaway trolley rushes toward five people standing on the tracks, and it will surely kill them all. Fortunately, you can reach a switch that will turn the trolley onto a side track — but then you notice that one other person is standing there. Is it morally permissible for you to turn the trolley to that side track, where it will kill one person instead of five? Is it not only morally permissible, but even morally required? This classic thought experiment is a mainstay in the repertoire of law school hypotheticals, often raised alongside cases about cannibalism at sea, tossing people from overcrowded lifeboats, or destroying buildings to save a city from fire.

The liturgy of the trolley problem is one we all know. It is a call-and-response, suitably Socratic in style. The question above is asked. The unsuspecting subject will firmly reply: “Yes, it must be morally permissible to avoid five deaths, when the alternative is a single death.” The confidence of this reply gives the next question its zing: “What if you cannot divert the trolley, but you can stop it by pushing someone in front of it? No? Yet isn’t this also sacrificing one to save five?”

This venerable vignette and its still darker variations have continued to capture our imagination beyond the classroom as well — informing public thought on subjects ranging from climate change, to abortion, to bioethics, to capital punishment, to takings and public necessity, to killing in wartime, and to torture. At a higher level of abstraction, it is invoked in controversies over how to mark the ethical boundaries of contested modes of regulation: When should consequentialist approaches to law or policy (such as cost-benefit analysis, optimal deterrence, and social-welfare aggregation) be trumped by deontological commands (such as “do not kill”), and vice versa?

Professor Frances Myrna Kamm, a moral philosopher who for de-cades has been a leading analyst of this thought experiment, has now published her richly stimulating Tanner Lectures. Joining her lectures as chapters in the book are a trio of rigorous and unrelenting responses from a panel of philosophers — Professors Judith Jarvis Thomson, Thomas Hurka, and Shelly Kagan — as well as an introduction by legal scholar Professor Eric Rakowski.

The book captures a moment of intellectual unrest arising from a sudden about-face by Thomson — a “founding mother” of the trolley problem (p. 21) — concerning one of the core premises underlying a generation of trolley debates. In short, Thomson, who first established the standard view that it is morally permissible for a bystander to turn the trolley, now argues that it is not. But if turning the trolley is just as verboten as pushing someone in front of it, there is no curious contrast to reconcile — so what remains of the trolley problem?

The beating heart of this book is a fresh, and often raw, analytical quarrel between Kamm and Thomson about this reversal — and about how much of the trolley problem continues to exist. As Rakowski observes, “[t]his is a truly remarkable exchange between the two leading contributors to this moral philosophical debate”; no doubt “these lectures, commentaries, and replies [will be] absolutely invalu-able for future work on the trolley problem” (p. 5).

The book’s timing is impeccable. It arrives just as the trolley problem is gaining newfound attention in the public discourse, due in part to its uncanny resemblance to emergent questions about how to program autonomous vehicles — such as military drones or driverless cars — to act ethically. If your self-driving Volvo may have to decide whether to swerve hard into a wall, killing you, in order to avoid an accident that will kill five others, what should it do?

At this moment of special vitality for the classic thought experiment, I wish to broach a topic that the current trolley debates have tended to neglect: Can our intuitions about moral dilemmas be influenced by the presence of the law?

This gap in the philosophical discourse is set in sharp relief by the ample attention paid by lawyers — and by the public — to questions of how and when the law can influence moral beliefs and social norms. Think of Prohibition, Brown v. Board of Education, or same-sex marriage. Not surprisingly, legal scholars have long theorized about the law’s role as a source of moral or ethical knowledge. Some empirical research on the potential for law to directly influence moral beliefs has also been done — but it appears to be scarce in the context of moral dilemmas such as the trolley problem.

By conducting a set of randomized survey experiments, I hope to jumpstart this line of inquiry. Moral dilemmas are an especially intriguing domain for the study of law’s potential influence. They are pervasive in the real world in the form of tragic choices or other harm-harm tradeoffs and are often regulated by law or policy. Moreover, they have a peculiar structure: they pose a contest between deeply felt moral commands. We must save the five. We must not kill the one. These are not mild suggestions. If the law does move the moral needle in such a case, it would be doing so where potent moral intuitions normally govern — a possibility these new experiments aim to test.

Part I elaborates on the exchanges between Kamm and her commentators, highlighting questions raised by their debates about the stability of moral intuitions. Part II explains the design of two experiments I conducted, in which survey subjects are asked their moral intuitions about trolley scenarios that include randomly assigned information about what the law says. The first experiment follows the standard trolley scenario, with a bystander standing at the switch; the law variously criminalizes or justifies turning the trolley. The second replaces the bystander with a railroad engineer, whose official role includes duties regulated by the law. By varying what the law says, these experiments together explore the role of law — and in particular, the law of roles — in shaping our moral intuitions.

Part III reports the findings, which offer evidence that the presence of law can influence our intuitions about the trolley dilemma. Telling subjects that turning the trolley constitutes criminal homicide increases the number who say that doing so is morally prohibited. In comparison, telling subjects that the law requires the engineer to minimize casualties reduces the number who say that turning the trolley is morally prohibited; it also increases the number who say that doing so is morally required. Moreover, the content of the law appears able to exert some influence even when subjects are told that the law will not be enforced.

Part III also suggests lines of questioning — for the trolley debates, for empirical research, and for legal design — raised by the following concern: What if the moral intuitions we can observe have already been shaped by people’s impressions, however vague or subconscious, about what the law expects?

More pragmatic questions arise, as well. In the context of self-driving cars, for example, should regulators hasten to announce that the role of the engineers who program such cars is to minimize casualties — before the public’s moral intuitions start to be forged by a series of tragic accidents? The Conclusion points to questions for further research concerning the possible mechanisms of law’s influence.

* Professor of Law, Columbia Law School. I wish to thank Jessica Bulman-Pozen, Glenn Cohen, Elizabeth Emens, Shannon Fanning, Jeffrey Gordon, Scott Hemphill, Kathryn Judge, Frances Kamm, Jeremy Kessler, Benjamin Liebman, Katerina Linos, David Pozen, Eric Rakowski, Frederick Schauer, and Barbara Spellman for insightful conversations and comments on drafts. Arisa Akashi, Bram Schumer, and Yuko Sin provided excellent research assistance, and I am grateful for what they have taught me. This project has been funded by a Global Innovation Award from Columbia Law School, and I also thank the grant committee for their