Good Ethics and Bad Choices: The Relevance of Behavior Economics for Medical Ethics. Jennifer S. Blumenthal-Barby. The MIT Press. Cambridge, MA. 2021, 251 pages, $45 paperback.

Publisher webpage:

Review by S. Scott Graham (he/him)

University of Texas at Austin

Date posted: December 2023

Read below or download here.

Many students are introduced to rhetoric through Plato’s Gorgias and Phaedrus. The dialogues offer two contrasting visions of the ethics of rhetoric, neither entirely comfortable for the rhetorician. In short, Gorgias suggests that rhetoric is an inherently manipulative art used to make the weaker argument the stronger. Phaedrus, on the other hand, concedes that there is a role for rhetoric; philosophers may ethically use rhetoric to persuade the ignorant masses to behave in their own best interests according to dialectically established truth. It is essentially these very same issues, as applied to medicine, that play out across the pages of Jennifer S. Blumenthal-Barby’s Good Ethics and Bad Choices: The Relevance of Behavior Economics for Medical Ethics. A philosopher and bioethicist by training and practice, Blumenthal-Barby is deeply invested in exploring the ethical foundations of persuasion in clinical care. In her recent book, she tackles questions about if, and when, it is ethically permissible for doctors to persuade their patients.
As the subtitle suggests, the particular focus of the text is behavioral economics. In recent years economists, psychologists, and marketing researchers have developed an extra-rhetorical framework to persuasion grounded in “nudges,” “boosts,” and “choice architecture.” These persuasive tactics are designed to leverage subconscious decisional heuristics and cognitive biases to secure an audience’s assent without engaging their rational consideration or deliberation. As Blumenthal-Barby argues, the availability of these techniques raises the ethical stakes for physicians given that these techniques may well threaten patient autonomy. This threat notwithstanding, Good Ethics and Bad Choices ultimately argues that there are some cases where the persuasive tactics of behavioral economics are morally permissible in clinical contexts. Specifically, Blumenthal-Barby argues that nudges may be ethically permissible when none of the following obtain:
(1) The patient’s decision is well thought out and in line with her goals, (2) there is uncertainty about what choice would be best for the patient (because of either lack of evidence regarding clinical efficacy or effectiveness of proposed options, lack of information about the patient’s interest and values, or unchecked distorting biases in the physician’s own judgments, or (3) the patient would object to any attempts to shape choice, resulting in damage to the therapeutic alliance. (p. 143)

Blumenthal-Barby’s path to this position is a complicated one, which is appropriate given the stakes involved. In making her arguments, the introduction begins by outlining the scope of the book and moves efficiently onto chapter one which details the basic psychology of decision-making. Blumenthal-Barby draws readers’ attention to the psychological distinction between “System 1” and “System 2” reasoning. Metaphorically, System 1 is your gut. It’s quick, intuitive judgments. In contrast, System 2 is characterized by rational, methodological, and overt deliberation. As Good Ethics, Bad Choices details, System 1 decision-making is where cognitive biases can run amok. Decisional heuristics, which include cognitive biases, provide the framework for rapid decision-making and essentially shortcut System 2 thought processes. The rapidity of System 1 decision-making can be an asset in emergent risk situations. I’m not going to activate System 2 to carefully identify the snake I just stepped on and assess its envenomation potential, I’m just going to jump and scream and run. Of course, the cognitive biases of System 1 often lead to all sorts of problems, including everything from self-serving and optimism biases to implicit biases like racism and ableism. In Good Ethics, Bad Choices, Blumenthal-Barby pays particular attention to the kinds of cognitive biases that can lead patients into suboptimal medical decision-making. With the health threats of cognitive biases established, chapter two turns to the possible benefits of behavioral economics in clinical contexts. In short, behavioral economics is an approach to persuasion that relies on intentionally activating System 1. Through leveraging what’s known about cognitive biases, choice architects encourage certain decisional outcomes without activating System 2. These techniques, as Blumenthal-Barby readily acknowledges, raise significant questions about patient autonomy. Indeed, many would argue that intentional effort to avoid System 2 and direct patients toward specific choices is a violation of respect for autonomy.

Blumenthal-Barby argues in favor of behavioral economics on the basis of three key claims:
[B]ehavioral economics shows that patients’ decision-making is at risk for being (1) nonautonomous or autonomy-impaired, (2) poor quality (as defined and measured by decision scientists), and (3) harmful to patients and their interests (using many different philosophical accounts of well-being and harm). (p. 59)
Essentially, if patients are already making decisions largely driven by cognitive biases, decisions that often prevent optimal health outcomes and frequently lead to decisional regret and diminished well-being, then clinical use of behavioral economics adds no additional harm and may improve health outcomes.
Chapter three builds on chapter two to mount a proactive argument in favor of choice architecture. Here Blumenthal-Barby deploys a bit of her own choice architecture, shifting from the alienating language of “behavioral economics” to the potentially more palatable “nudge.” In so doing, Good Ethics, Bad Choices invokes Richard Thaler and Cass Sunstein’s (2009) famous Nudge. Despite the colloquial character of “nudge,” it is a technical term. Blumenthal-Barby echoes Thaler and Sunstein to define it as “‘any aspect of choice architecture that alters people’s behavior in a predictable way without forbidding any options or significantly changing economic incentives’” (p. 63). Essentially, nudge theory presupposes that it is possible to walk a fine line between activating System 1 and manipulation. The theory is underwritten by Thaler and Sunstein’s notion of “libertarian paternalism,” (or as Blumenthal-Barby re-describes it, “justified soft paternalism”) which they endorse as ethically permissible because it indexes value judgements to the perspective of those who are nudged.

Of particular interest to rhetoricians, this chapter also includes a rumination on the ethical permissibility of any clinical persuasion. This begins with a reflection on Plato’s moral arguments about rhetoric in The Republic, Gorgias, and the Phaedrus. Drawing on the caricature of rhetoric in Gorgias, Blumenthal-Barby implies that rhetoric and nudging are nearly coextensive (see below for my concerns about this partial conflation) given their twin willingness to leverage emotional argument as opposed to purely rational persuasion. Then through invoking the Phaedrus’s insight that there cannot be speech without rhetoric, Blumenthal-Barby argues there can likewise be no speech without nudging, and therefore purely rational persuasion is unattainable. The corollary here is that if rhetoric/nudging is unavoidable, clinicians have a duty to practice it responsibly.

Chapter four tackles three primary lines of argument: (1) It describes a taxonomy of nudge types to explore if all nudges are morally permissible; (2) It argues that traditional bioethical conceptions of “manipulation” are overly broad; and (3) It interrogates if and how nudging can fit into the shared decision-making. I particularly recommend to the attention of rhetoricians, the schematic of influence types offered in figure 4.1 (p. 119). The diagram provides a thoughtful distillation of different nudge tactics and fits well with rhetoricians’ penchant for taxonomizing persuasive techniques. The hierarchical schematic identifies five primary persuasive approaches (reason and argument; nonargumentative influence (nudges); incentives; omission; and force or threats). The second level divides nudges into transparent and nontransparent nudges, thereby separating overt techniques like explicating, inducing affective states, playing on desires, and peer pressure from more covert techniques like framing, priming, and environmental arrangement. In comparing nudge types, Blumenthal-Barby argues that the morality of any given nudge should be determined contextually. For her, any given nudge is likely to be morally permissible if it prompts patients toward well-informed considerations about true benefits and in so doing does not inhibit the free choice of patients to select among alternatives. Subsequently chapter five is devoted to exploring detailed case studies where nudging may or may not be morally permissible.

Ultimately, Good Ethics, Bad Choices provides a thorough argument for clinical behavioral economics. Nevertheless, that argument makes me uncomfortable. Certainly, this is partially due to the strong anti-paternalist commitments of rhetoric of health and medicine. My own System 1 thinking makes me a priori resistant to any level of clinical paternalism. That said, a System 2 analysis of Blumenthal-Barby’s arguments is harder to reject. It fits well with the general rhetorical commitment that ethics are highly contextual. Additionally, I am persuaded by the argument that choice architecture is pervasive and therefore requires serious ethical attention. However, my largest concern about the argument in the book centers around the very pervasiveness of System 1 thinking. In some respects, the book assumes that clinicians can be entirely System 2 about System 1. Ethical choice architecture is supposed to be a product of careful elicitation of patient perspectives and meticulous calibration of nudge techniques to those perspectives. The idea that this mode of behavior is attainable is cast in doubt by the very arguments of behavioral economics and the data on implicit biases in clinical decision-making.

Despite the careful approach of Good Ethics, Bad Choices, there are several moments in the book, itself, where implicit biases emerge and directly challenge the idea that clinical nudging can be fully indexed to patient values. Specifically, Good Ethics, Bad Choices uses stigmatizing language to refer to patients in several places around weight loss and drug use. In one case study, the reader is instructed to imagine someone “who does not comprehend the consequences of poor health choices, lacks self-control, and as result has an actual desire to spend all his time eating pizza and playing video games” (p. 83). Even though this individual, also described as “lazy” (p. 83), is used as an example of the ambiguity around “better off,” and when to limit nudging, the language of moral judgment here highlights how easy it can be for the values of any would-be choice architect to override patient choices. Similarly, soft paternalism is justified through the figure of the “drug addict” [sic] on the basis that drug use is inherently autonomy impairing, and it would therefore be appropriate for clinicians to imagine a non-autonomy impaired patient who held a different preference one could nudge toward (p. 65). Here again the stigmatizing language and the a priori construal of the impossibility of drug use by choice demonstrate how easy it can be for System 1 biases to override patient values. It also strikes me that this risk is exacerbated by cost-benefit analyses that might privilege considerations like physician time management (p. 104). If expedience is considered a desirable aim, then decisional heuristics (including implicit biases) are even more likely to activate.

These concerns notwithstanding, I urge rhetoricians to read this book. Not only does it provide important insights into how bioethicists are thinking about persuasion, but it also raises significant challenges for long-dominant conceptions of rhetoric. Specifically, reading Good Ethics, Bad Choices forced me to reflect on the ways rhetoric’s efforts to define the ideal orator have tacitly described an exclusively System 2 auditor. Behavioral economics is, in some ways, economists’ attempt to wrestle with the limitations of the appropriately maligned rational actor in rational choice theory. Likewise, it’s worth considering whether rhetorical theory presupposes a rational auditor with similar limitations. Aristotle identifies “judgment” and the “object of rhetoric” (1377b) and further articulates judgment as a function of deductive reasoning (Van Hooft, 2001, p. 135). Artful proofs, including ethos and pathos, are thus rationalized through speech (1356a; 1377b). Notably, this is also precisely why I would disagree with Blumenthal-Barby’s partial conflation of nudging and rhetoric. In classical rhetorical theory, pathetic and ethotic arguments are System 2 arguments because they are rendered explicit and rationalized in discourse. For rhetoricians, it’s critical to note how classical rhetorical theory about ethos and its overt construction in logos often fails to account for the System 1 persuasion.

If audiences are routinely using System 1, then a rhetorical theory that relies exclusively on System 2 may be insufficient to address the realities of persuasion. Good Ethics, Bad Choices provides an ideal opportunity to catalyze a conversation about the extent to which rhetoric should attempt to develop a systematic account of System 1 persuasion. On the one hand, psychological theories of cognitive biases and economic theories of choice architecture do not seem to fully account for the persuasive tactics of nudging. On the other hand, if System 1 rhetorics are inherently paternalistic, then we might choose not to pursue this area of inquiry. On the third hand, I imagine rhetoricians may be more open to nudging in the context of ethical benefits to society. Research in the discipline implies some consider it ethically permissible to persuade vaccine hesitant patients or parents against their self-articulated values when vaccination offers clear benefits to larger society (Lawrence, 2018; Campeau, 2022). So, while many rhetoricians may be uncomfortable with libertarian paternalism, they may be more open to choose architecture in contexts of “social paternalism,” which justifies autonomy limitations because of communitarian values (Ravitsky, 2016). Ultimately the dynamics of paternalism and clinical persuasion are fraught and complex topics worthy of further scrutiny in rhetoric.

Aristotle (1924). On Rhetoric. In W. D. Ross & J. A. Smith (Eds.). The Works of Aristotle: Rhetorica; De Rhetorica Ad Alexandrum; De Poetica (Vol. 11). Clarendon Press.
Blumenthal-Barby, Jennifer S. (2021). Good ethics and bad choices: the relevance of behavioral economics for medical ethics. The MIT Press.
Campeau, Kari. (2022). Vaccination double bind: A study of pregnancy and COVID-19 vaccine decision-making. Rhetoric Society Quarterly, 52(5), 494-509.
Lawrence, Heidi Y. (2018). When patients question vaccines: Considering vaccine communication through a material rhetorical approach. Rhetoric of Health & Medicine, 1(1), 161-178.
Ravitsky, Vardit. (2016). Social paternalism in a communitarian context: Enhancing individuals’ moral deliberation through a communal “moral voice”. The American Journal of Bioethics, 16(8), 20-22.
Thaler, Richard H., & Sunstein, Cass R. (2009). Nudge: Improving decisions about health, wealth, and happiness. Penguin.
Van Hooft, Stan. (2001). Judgement, decision, and integrity. Philosophical Explorations 4(2), 135-149.