Ethically Deployed Defaults
Defaults are extremely effective at influencing people’s choices. Decisions as small as the kind of milk to get in a latte and as large as whether or not to donate one’s organs can differ by 60% to 70% or more, depending on what is set as the default. This has made defaults an increasingly common form of influence—by setting an option as the default, policy makers and marketers have a simple but powerful way to sway choices. But defaults are so effective that people have raised concerns that they could be used unethically or irresponsibly. Some have suggested that informing people how defaults are intended to influence their decisions could protect people from defaults intended to benefit the default-setter at the chooser’s expense, but others have cautioned that disclosure could make well-intended defaults less effective and undermine the potential good that they could do.
Our research, presented at the inaugural Behavioral Science and Policy Association conference in June 2016 and forthcoming in the Journal of Marketing Research shows that disclosing how defaults are intended to affect behavior affects how fair and ethical people perceive them to be, but it does not change how susceptible people are to their influence. The good news is that defaults can be disclosed without reducing their benefits, but the bad news is that disclosure alone is insufficient to protect people from being unknowingly manipulated. Rather, a more active intervention is needed to enable people to resist potentially exploitative defaults.
Defaults, and the case for transparency
The use of defaults and other forms of choice architecture to prompt behavioral change is not only common and effective, it is also widely supported. Part of what makes nudges so popular is their ability to promote positive behavioral change without taking away people’s options or making certain behaviors more costly. According to a recent poll conducted by Cass Sunstein and reported in a op-ed in the New York Times, about 71% of Americans would support a federal mandate requiring large employers to automatically enroll employees in retirement plans so that employees would choose whether to actively opt out rather than whether to opt in. As another example, about 67% of Americans would support a federal law compelling large electricity providers to automatically enroll consumers in “green” energy plans. And, the more effective an intervention is, the more acceptable people perceive it to be.
However, nudges like these are not universally popular. People are less happy with nudges that more directly affect behavior, like defaults or plate sizes aimed at changing portion sizes, than they are with more informational nudges, like calorie labeling. The end goal of the default matters, too—for example, it would be surprising to find that people are as in favor of defaults meant to sign them up for a marketer’s email list as they are of defaults meant to guarantee that they have enough money for retirement. Further, research shows that trust in the party who set the default and whether the policy itself is consistent with one’s own political ideals can determine support for the use of defaults and other nudges to guide decisions.
Critics including the British House of Lords and the Electronic Privacy Information Center have argued that deploying defaults without warning decision makers about them is unethical because they can covertly alter behavior. They have also proposed that disclosing the intentions behind defaults is necessary to protect decision makers from being manipulated. However, proponents of defaults and other nudges caution that telling decision makers about defaults and how they are meant to affect behavior might render them ineffective and reduce their benefits.
This raises a challenge for policy makers: Is possible to make consumers aware of the intention behind defaults designed to deliberately sway choices without inoculating them and undermining their ability to influence people’s decisions for the better?
Can defaults be transparent and still work?
Perhaps a solution is to make sure that people are informed about what the default option is if they take no action. It seems, though, that disclosing what the default is and what its alternatives are may not necessarily make people any less susceptible to its influence. In their paper, “Warning: You Are About to be Nudged,” published in the inaugural issue of Behavioral Science and Policy, George Loewenstein and colleagues invited people to complete a hypothetical advance directive in which they chose whether to pursue medical treatment at the end of life. The default was either to receive treatment or to not receive treatment, and participants were informed about what the default was either before or after completing the directive. Then, everyone was given the opportunity to revise their choices. Informing people about the default, either before or after they initially chose, did not appreciably diminish the impact of the defaults on their choice to pursue treatment.
Even if people know what the default is, however, they may not understand what the default is intended to do. Our own research shows that disclosing how defaults are intended to affect people’s choices enhances consumers’ perceptions of the fairness and ethicality of defaults and improves impressions of default setters, but it does not make defaults any less influential. For example, in one experiment, we offered hot chocolate to passersby, either with or without whipped cream by default. Some of these hot chocolate drinkers were told that people are more likely to choose an option when it is the default, and that no-whip was the default to make them more likely to choose a healthier drink or that whipped cream was the default to nudge them toward a more indulgent drink. Regardless of whether this intent was disclosed, people for whom whipped cream was the default were much more likely to get whipped cream than people for whom no-whip was the default. In a variety of other contexts—including financial decisions, privacy permissions, and energy savings—we likewise find that people’s choices are equally influenced by defaults regardless of whether the effects and intentions behind them are made transparent.
Why doesn’t disclosure necessarily matter? The reason has to do with the fact that making an option the default leads people to focus disproportionately on reasons to choose the default. This is rooted in basic perceptual processes over which most people have little awareness or control. Disclosure doesn’t help because telling people how defaults are intended to influence their choices doesn’t give them the insight they need to counter their influence. In our own research, we have yet to find a case in which disclosing the intent and potential influence of a default meaningfully reduces its effectiveness, even when people learn that the intention behind the default was to benefit the default setter at their own expense, and even when people do not believe the default to be ethical. And although some researchers have found that a non-negligible proportion of the population may react against nudges from parties with whom they disagree, others have found that disclosure does not diminish default effects even when people’s preferences are inconsistent with the default. On one hand, this is great evidence in support of the case for transparency—after all, if defaults are resilient to disclosure, then there is no good reason not to disclose their use and nature. However, it does give policy makers another challenge—if disclosure cannot protect consumers from defaults designed to nudge them toward choices that are not in their own or society’s best interests, what else can be done?
How can people be protected from exploitation?
In our research, we instructed people to think about their preferences for the default and its alternative simultaneously, as in a forced choice, and write down what was important to them before they made a decision. When people did this, they were more likely to consider their options in a more balanced way and make less biased choices. This is good news – it suggests that encouraging people to articulate their preferences for the default and its alternative before deciding may help consumers counter the influence of defaults that aren’t in their best interest, while still allowing marketers and policy makers to effectively set defaults that benefit consumers and society.
In sum, to effectively and ethically deploy defaults, it is important to keep in mind three key insights: First, setting a recommended option as a default is a powerful way to encourage people to make choices that are in their own or society’s best interests without forcing them to do so. Second, the intent and potential influence of defaults can be disclosed without reducing their benefits. Doing so may alleviate concerns about whether it is ethical to set defaults to influence the behavior of people who are unaware of how defaults may affect them. Third is a caveat: this means that default disclosures offer insufficient protection when defaults are not in the individual or society’s best interests. Instead, a more active intervention that targets the processes underlying defaults’ effectiveness is needed. Encouraging people to articulate their preferences for the default and its alternative before choosing can help diminish the effect of defaults and give consumers the ability to make decisions in their own best interests.
About the authors:
Mary Steffel is an assistant professor of marketing at the D’Amore-McKim School of Business at Northeastern University and will serve as a fellow on the White House Social and Behavioral Sciences Team from October 2016-2017. She received her Ph.D. in psychology from Princeton University in 2009 and her Ph.D. in marketing from the University of Florida in 2012. Her research examines when people call upon others to help them make decisions, what are the barriers to accurately gauging others’ preferences and effectively choosing on their behalf, and how to help people make better decisions for themselves and others. Follow her on Twitter at @steffel_mary.
Elanor F. Williams is an assistant professor of marketing at the Kelley School of Business at Indiana University. She received her Ph.D. in social psychology from Cornell University in 2008. Her work investigates decision making in social contexts, including how people make choices with others and for others, and how people come to know themselves and other people.