Cass R. Sunstein
Summary. It is important for people to make good choices about important matters, such as health insurance or retirement plans. Sometimes it is best to ask people to make active choices. But in some contexts, people are busy or aware of their own lack of knowledge, and providing default options is best for choosers. If people elect not to choose or would do so if allowed, they should have that alternative. A simple framework, which assesses the costs of decisions and the costs of errors, can help policymakers decide whether active choosing or default options are more appropriate.
Consider the following problems:
- Public officials are deciding whether to require people, as a condition for obtaining a driver’s license, to choose whether to become organ donors. The alternatives are to continue with the existing opt-in system, in which people become organ donors only if they affirmatively indicate their consent, or to switch to an opt-out system, in which consent is presumed.
- A public university is weighing three options: to enroll people automatically in a health insurance plan; to make them opt in if they want to enroll; or, as a condition for starting work, to require them to indicate whether they want health insurance and, if so, which plan they want.
- A utility company is deciding which is best: a “green default,” with a somewhat more expensive but environmentally favorable energy source, or a “gray default,” with a somewhat less expensive but environmentally less favorable energy source. Or should the utility ask consumers which energy source they prefer?
- A social media site is deciding whether to adopt a system of default settings for privacy or to require first-time users to identify, as a condition for access, what privacy settings they want. Public officials are monitoring the decision and are considering regulatory intervention if the decision does not serve users’ interests.
In these cases and countless others, policymakers are evaluating whether to use or promote a default rule, meaning a rule that establishes what happens if people do not actively choose a different option. A great deal of research has shown that for identifiable reasons, default rules have significant effects on outcomes; they tend to “stick” or persist over time.1 For those who prize freedom of choice, active choosing might seem far preferable to any kind of default rule.
My goal here is to defend two claims. The first is that in many contexts, an insistence on active choosing is a form of paternalism, not an alternative to it. The reason is that people often choose not to choose, for excellent reasons. In general, policymakers should not force people to choose when they prefer not to do so (or would express that preference if asked).
The second claim is that when policymakers decide between active choosing and a default rule, they should focus on two factors. The first is the costs of making decisions. If active choosing is required, are people forced to incur large costs or small ones? The second is the costs of errors: Would the number and magnitude of mistakes be higher or lower with active choosing than with default rules?
These questions lead to some simple rules of thumb. When the situation is complex, technical, and unfamiliar, active choosing may impose high costs on choosers, and they might ultimately err. In such cases, there is a strong argument for a default rule rather than for active choosing. But if the area is one that choosers understand well, if their situations (and needs) are diverse, and if policymakers lack the means to devise accurate defaults, then active choosing would be best.
This framework can help orient a wide range of policy questions. In the future, it may be feasible to personalize default rules and tailor them to particular groups or people. This may avoid current problems associated with both active choosing and defaults designed for very large groups of people.2
Active Choosing Can Be Paternalistic
With the help of modern technologies, policymakers are in an unprecedented position to ask people this question: What do you choose? Whether the issue involves organ donation, health insurance, retirement plans, energy, privacy, or nearly anything else, it is simple to pose that question (and, in fact, to do so repeatedly and in real time, thus allowing people to signal new tastes and values). Those who reject paternalism and want to allow people more autonomy tend to favor active choosing. Indeed, there is empirical evidence that in some contexts, ordinary people will pay a premium to be able to choose as they wish.3,4 (Compare the related phenomenon of reactance, which suggests a negative reaction to coercive efforts, produced in part by the desire to assert autonomy.5) In other cases, people will pay a premium to be relieved of that very obligation.
There are several reasons why people might choose not to choose. They might fear that they will err. They might not enjoy choosing. They might be too busy. They might lack sufficient information or bandwidth.6 They might not want to take responsibility for potentially bad outcomes for themselves (and at least indirectly for others).7,8 They might find the underlying questions confusing, difficult, painful, and troublesome—empirically, morally, or otherwise. They might anticipate their own regret and seek to avoid it. They might be keenly aware of their own lack of information or perhaps even of their own behavioral biases (such as unrealistic optimism or present bias, understood as an undue focus on the near term). In the area of retirement savings or health insurance, many employees might welcome a default option, especially if they trust the person or institution selecting the default.
It is true that default rules tend to stick, and some people distrust them for that reason. The concern is that people do not change default options out of inertia (and thus reduce the costs of effort). With an opt-in design (by which the chooser has to act to participate), there will be far less participation than with an opt-out design (by which the chooser has to act to avoid participation).1 Internet shopping sites often use an opt-out default for future e-mail correspondence: The consumer must uncheck a box to avoid being put on a mailing list. It is well established that social outcomes are decisively influenced by the choice of default in areas that include organ donation, retirement savings, environmental protection, and privacy. Policymakers who are averse to any kind of paternalism might want to avoid the appearance of influencing choice and require active choosing.9
When policymakers promote active choosing on the ground that it is good for people to choose, they are acting paternalistically. Choice-requiring paternalism might appear to be an oxymoron, but it is a form of paternalism nonetheless.
Respecting Freedom of Choice
Those who favor paternalism tend to focus on the quality of outcomes.10 They ask, “What promotes human welfare?” Those who favor libertarianism tend to focus instead on process. They ask, “Did people choose for themselves?” Some people think that libertarian paternalism is feasible and seek approaches that will promote people’s welfare while also preserving freedom of choice.11 But many committed libertarians are deeply skeptical of the attempted synthesis: They want to ensure that people actually choose.9
It is worth distinguishing between the two kinds of libertarians. For some, freedom of choice is a means. They believe that such freedom should be preserved, because choosers usually know what is best for them. At the very least, choosers know better than outsiders (especially those outsiders employed by the government) what works in their situation. Those who endorse this view might be called epistemic libertarians, because they are motivated by a judgment about who is likely to have the most knowledge. Other libertarians believe that freedom of choice is an end in itself. They think that people have a right to choose even if they will choose poorly. People who endorse this view might be called autonomy libertarians.
When people choose not to choose, both types of libertarians should be in fundamental agreement. Suppose, for example, that Jones believes that he is not likely to make a good choice about his retirement plan and that he would therefore prefer a default option, chosen by a financial planner. Or suppose that Smith is exceedingly busy and wants to focus on her most important or immediate concerns, not on which health insurance plan or computer privacy setting best suits her. Epistemic libertarians think that people are uniquely situated to know what is best for them. If so, then that very argument should support respect for people when they freely choose not to choose. Autonomy libertarians insist that it is important to respect people’s autonomy. If so, then it is also important to respect people’s decisions about whether and when to choose.
If people are required to choose even when they would prefer not to do so, active choosing becomes a form of paternalism. If, by contrast, people are asked whether they want to choose and can opt out of active choosing (in favor of, say, a default option), active choosing counts as a form of libertarian paternalism. In some cases, it is an especially attractive form. A private or public institution might ask people whether they want to choose the privacy settings on their computer or instead rely on the default, or whether they want to choose their electricity supplier or instead rely on the default.
With such an approach, people are being asked to make an active choice between the default and their own preference: In that sense, their liberty is fully preserved. Call this simplified active choosing. This approach has evident appeal, and in the future, it is likely to prove attractive to a large number of institutions, both public and private.
It is important to acknowledge that choosers’ best interests may not be served by the choice not to choose. Perhaps a person lacks important information, which would reveal that the default rule might be harmful. Or perhaps a person is myopic, being excessively influenced by the short-term costs of choosing while underestimating the long-term benefits, which might be very large. A form of present bias might infect the decision not to choose.
For those who favor freedom of choice, these kinds of concerns are usually a motivation for providing more and better information or for some kind of nudge—not for blocking people’s choices, including their choices not to choose. In light of people’s occasional tendency to be overconfident, the choice not to choose might, in fact, be the best action. That would be an argument against choice-requiring paternalism. Consider in this regard behavioral evidence that people spend too much time pursuing precisely the right choice. In many situations, people underestimate the temporal costs of choosing and exaggerate the benefits, producing “systematic mistakes in predicting the effect of having more, vs. less, choice freedom on task performance and task-induced affect.”12
If people prefer not to choose, they might favor either an opt-in or an opt-out design. In the context of both retirement plans and health insurance, for example, many people prefer opt-out options on the grounds that automatic enrollment overcomes inertia and procrastination and produces sensible outcomes for most employees. Indeed, the Affordable Care Act calls for automatic enrollment by large employers, starting in 2015. For benefits programs that are either required by law or generally in people’s interests, automatic enrollment has considerable appeal.
In the context of organ donation, by contrast, many people prefer an opt-in design on moral grounds, even though more lives would be saved with opt-out designs. If you have to opt out to avoid being an organ donor, maybe you’ll stay in the system and not bother to opt out, even if you do not really want to be an organ donor. That might seem objectionable. As the experience in several states suggests, a system of active choosing can avoid the moral objections to the opt-out design while also saving significant numbers of lives.
Are people genuinely bothered by the existence of default rules, or would they be bothered if they were made aware that such rules had been chosen for them? A full answer is not available for this question: The setting and the level of trust undoubtedly matter. In the context of end-of-life care, when it is disclosed that a default rule is in place, there is essentially no effect on what people do. (Editor’s note: See the article “Warning: You Are about to Be Nudged” in this issue.) This finding suggests that people may not be uncomfortable with defaults, even when they are made aware that choice architects have selected them to influence outcomes.13 More research on this question is highly desirable.
Weighing Decision Costs and Error Costs
The choice between active choosing and default rules cannot be made in the abstract. If welfare is the guide, policymakers need to investigate two factors: the costs of decisions and the costs of errors. In some cases, active choosing imposes high costs, because it is time-consuming and difficult to choose. For example, it can be hard to select the right health insurance plan or the right retirement plan. In other cases, the decision is relatively easy, and the associated costs are low. For most people, it is easy, to choose among ice cream flavors. Sometimes people actually enjoy making decisions, in which case decision costs turn out to be benefits.
The available information plays a role here as well. In some cases, active choosing reduces the number and magnitude of errors, because choosers have far better information about what is good for them than policymakers do. Ice cream choices are one example; choices among books and movies are another. In other cases, active choosing can increase the number and magnitude of errors, because policymakers have more relevant information than choosers do. Health insurance plans might well be an example.
With these points in mind, two propositions are clear, and they can help orient this inquiry in diverse settings. First, policymakers should prefer default rules to active choosing when the context is confusing and unfamiliar; when people would prefer not to choose; and when the population is diverse with respect to wants, values, and needs. The last point is especially important. Suppose that with respect to some benefit, such as retirement plans, one size fits all or most, in the sense that it promotes the welfare of a large percentage of the affected population. If so, active choosing might be unhelpful or unnecessary.
Second, policymakers should generally prefer active choosing to default rules when choice architects lack relevant information, when the context is familiar, when people would actually prefer to choose (and hence choosing is a benefit rather than a cost), when learning matters, and when there is relevant heterogeneity. Suppose, for example, that with respect to health insurance, people’s situations are highly diverse with regard to age, preexisting conditions, and risks for future illness, so any default rule will be ill suited to most or many. If so, there is a strong argument for active choosing.
To be sure, the development of personalized default rules, designed to fit individual circumstances, might solve or reduce the problems posed by heterogeneity.14,15 As data accumulate about what informed people choose or even about what particular individuals choose, it will become more feasible to devise default rules that fit diverse situations. With retirement plans, for example, demographic information is now used to produce different initial allocations, and travel websites are able to incorporate information about past choices to select personalized defaults (and thus offer advice on future destinations).2,14 For policymakers, the rise of personalization promises to reduce the costs of uniform defaults and to reduce the need for active choosing. At the same time, however, personalization also raises serious questions about both feasibility and privacy.
A further point is that active choosing has the advantage of promoting learning and thus the development of preferences and values. In some cases, policymakers might know that a certain outcome is in the interest of most people. But they might also believe that it is important for people to learn about underlying issues, so they can apply what was gained to future choices. In the context of decisions that involve health and retirement, the more understanding people develop, the more they will be able to choose well for themselves. Those who favor active choosing tend to emphasize this point and see it as a powerful objection to default rules. They might be right, but the context greatly matters. People’s time and attention are limited, and the question is whether it makes a great deal of sense to force them to get educated in one area when they would prefer to focus on others.
Suppose that an investigation into decision and error costs suggests that a default rule is far better than active choosing. If so, epistemic libertarians should be satisfied. Their fundamental question is whether choice architects know as much as choosers do, and the idea of error costs puts a spotlight on the question that most troubles them. If a default rule reduces those costs, they should not object.
It is true that in thinking about active choosing and default rules, autonomy libertarians have valid and distinctive concerns. Because they think that choice is important in itself, they might insist that people should be choosing even if they might err. The question is whether their concerns might be alleviated or even eliminated so long as freedom of choice is fully preserved by offering a default option. If coercion is avoided and people are allowed to go their own way, people’s autonomy is maintained.
In many contexts, the apparent opposition between active choosing and paternalism is illusory and can be considered a logical error. The reason is that some people choose not to choose, or they would do so if they were asked. If policymakers are overriding that particular choice, they may well be acting paternalistically. With certain rules of thumb, based largely on the costs of decisions and the costs of errors, policymakers can choose among active choosing and default rules in a way that best serves choosers. ♦
Sunstein, Harvard University Law School. Corresponding author’s e-mail: firstname.lastname@example.org
The author, Harvard’s Robert Walmsley University Professor, is grateful to Eric Johnson and three anonymous referees for valuable suggestions. This article draws on longer treatments of related topics, including Cass R. Sunstein, Choosing Not to Choose (Oxford University Press, 2015).
1.Johnson, E. J., & Goldstein, D. G. (2012). Decisions by default. In E. Shafir (Ed.), The behavioral foundations of policy (pp. 417–418). Princeton, NJ: Princeton University Press.
2.Goldstein, D. G., Johnson, E. J., Herrmann, A., & Heitmann, M. (2008). Nudge your customers toward better choices. Harvard Business Review, 86, 99–105.
3.Fehr, E., Herz, H., & Wilkening, T. (2013). The lure of authority: Motivation and incentive effects of power. American Economic Review, 103, 1325–1359.
4.Bartling, B., Fehr, E., & Herz, H. (2014). The intrinsic value of decision rights (Working Paper No. 120). Zurich, Switzerland: University of Zurich, Department of Economics.
5.Pavey, L., & Sparks, P. (2009). Reactance, autonomy and paths to persuasion: Examining perceptions of threats to freedom and informational value. Motivation and Emotion, 33, 277–290.
6.Mullainathan, S., & Shafir, E. (2013). Scarcity: Why having too little means so much. New York, NY: Times Books.
7.Bartling, B., & Fischbacher, U. (2012). Shifting the blame: On delegation and responsibility. Review of Economic Studies, 79, 67–87.
8.Dwengler, N., Kübler, D., & Weizsäcker, G. (2013). Flipping a coin: Theory and evidence. Unpublished manuscript. Retrieved from http://www.wiwi.hu-berlin.de/professuren/vwl/mt-anwendungen/team/flipping-a-coin
9.Rebonato, R. (2012). Taking liberties: A critique of libertarian paternalism. London, United Kingdom: Palgrave Macmillan.
10. Conly, S. (2013). Against autonomy: Justifying coercive paternalism. Cambridge, United Kingdom: Cambridge University Press.
11. Thaler, R. H., & Sunstein, C. R. (2008). Nudge: Improving decisions about wealth, health, and happiness. New York, NY: Penguin.
12. Botti, S., & Hsee, C. (2010). Dazed and confused by choice: How the temporal costs of choice freedom lead to undesirable outcomes. Organizational Behavior and Human Decision Processes, 112, 161–171.
13. Loewenstein, G., Bryce, C., Hagman, D., & Rajpal, S. (2015). Warning: You are about to be nudged. Behavioral Science & Policy, 1, 35–42.
14. Smith, N. C., Goldstein, D. G., & Johnson, E. J. (2013). Choice without awareness: Ethical and policy implications of defaults. Journal of Public Policy & Marketing, 32, 159–172.
15. Sunstein, C. R. (2013). Deciding by default. University of Pennsylvania Law Review, 162, 1–57.