Bioethics Made Rigorous

The beginning of an ambitious and important project, Jonathan Baron's Against Bioethics aims to dispense with endlessly ambiguous applications of principles and produce grounded, action-guiding advice to policy-makers, physicians, and patients. Urging us to abandon the hemming and hawing over what "autonomy," "beneficence," "nonmaleficence," or "justice" means in any particular case, Baron wants "to discuss several issues that engage bioethics today, and then show how these issues can be illuminated by applied utilitarianism, in the form of decision analysis" (p. 5).

Baron's goal is not to convince bioethicists to stop using familiar principles, but to establish that "all of the basic principles used in applied bioethics can be understood as having utilitarian roots" (p. 18), and that by situating them in a utilitarian framework, we can evaluate their application according to a common standard. We can identify better and worse applications of the principles by looking at the consequences of each application. The utilitarian roots of respect for autonomy, for example, are easily recognized: individuals tend to know what's best for themselves, with practice they will make better decisions, and some individuals value making their own decisions. Given these utilitarian roots, influencing others' decisions (in the form of advice, incentives, or coercion) is justified (or not) by evaluating the utility costs of the influence and the utility benefits of the outcome resulting from that influence.

Baron endorses a rather technical tool for applying utilitarian theory. "Utilitarian decision analysis" assigns a value to each possible outcome and multiplies each value by the probability of that outcome. Baron defines utility subjectively (p. 29). Subjects who will be affected by a possible outcome are the ones to assign values to it. Once probabilities have also been gauged for each potential outcome, the system indicates the best decision. UDA has two advantages: it is rigorous, and it requires an explicit identification of the criteria for a decision. But the system is also highly flexible. If two people value potential outcomes differently, or if the probabilities of outcomes are different in their cases, different courses of action may emerge as appropriate for them.

Baron also makes heavy use of cognitive psychology. For example, he argues that the standing distinction between active and passive euthanasia fails because it is based on "omission bias." A robust conclusion of cognitive psychology research, the omission bias amounts to taking harmful consequences that result from not acting as being less bad than equivalent (or even less harmful) consequences that result from acting.

Baron's use of cognitive psychology is refreshing and much-needed, as the field is underutilized in bioethics despite producing robust conclusions for more than thirty years. Still, one of the book's shortcomings is that even Baron's use of this literature is limited. He focuses on how policy-maker decisions are subject to bias, but showing the implications of work in cognitive psychology for patients' own medical decision-making would have been even more effective.

Another minor shortcoming is that Baron does not always give careful attention to the implications of his method. For example, he notes research which shows that having too many choices can actually lead to worse decision-making, but in his discussion of consent, he declares that "giving rational people a new option without removing any options cannot make them worse off. They don't have to take it" (p. 109). That claim is true only to a point.

Finally, Baron's recommended responses to biases in decision-making are a bit optimistic. His recommendations depend on the belief that people will readily try to correct the biases he illustrates. In fact, among physicians and policy-makers, there is likely to be widespread resistance to the conclusions of cognitive psychology applied to medical practice.

Baron proposes some solutions to particular problems, but these often remain empirically uninformed—many of the areas where his approach can make a difference require further research. For example, in his discussion of learning disabilities and potential treatments, he cannot identify exactly what strategy will produce the most utility (pp. 66-67). Until we know more about the advantages of different treatments across different segments of...