Are you not considered an organ donor unless you opt-in, or are you considered an organ donor unless you opt-out? Although the choice remains the same in both cases, people are more likely to be organ donors when organ donation is the default than when it is not 1,2.
Setting default options is a well-known example of a simple and effective “nudge” 3. Nudges are conventionally defined as “any aspect of the choice architecture that alters people’s behavior in a predictable way without forbidding any options or significantly changing their economic incentives” 4. Over the years, policymakers and organizations have employed nudges to steer decision makers toward desired outcomes in numerous domains beyond organ donation, such as retirement saving 5 and green energy use 6.
Ethical concerns with nudges
Despite their apparent success and increasing popularity, nudges have raised serious ethical concerns 7.
Autonomy. Nudges often rely on mechanisms that bypass people’s deliberative processes, undermining their agency and opportunity to think for themselves.
Reversibility. When nudges operate outside people’s conscious cognitive systems, people may be unaware of how nudges affect their behavior.This makes reversing the effects of nudges extremely difficult, if not impossible.
Preference identification. The nudge approach makes two assumptions: 1) choice architects know what is best for decision makers and 2) decision makers do not mind others interfering with their choices. In the real world, these assumptions are often unwarranted, and nudges can become paternalistic interferences in one’s decision-making process.
Public vs. Private sphere. Even if the above concerns were addressed, the benefits of nudges are limited to the public sphere. Nudges rely on external modifications to the choice architecture, and therefore cannot address many self-control related issues in people’s private lives
Self-nudges
To overcome these ethical concerns, Reijula and Hertwig (2020) proposed the concept of “self-nudges”: nudges designed by decision makers themselves rather than outsiders. A typical example of a sefl-nudge is a grocery list that includes only the healthy food that you want to buy, and helps you resist buying snacks that you want to avoid. By allowing decision makers to be their own choice architects, self-nudges are by definition transparent, and resolves the root problem of all the ethical concerns of traditional nudges. Decision makers can preserve their autonomy to choose as they actively participate in designing their own choice environments. As a result, they are more likely to recognize the effects of nudges, and can reverse these effects if desired. Furthermore, self-nudges should naturally reflect decision makers’ true preferences, reducing the risk of paternalistic interreference in their decision-making processes. Finally, decision makers can easily implement self-nudges in the private domains of their lives.
However, there is little empirical evidence for whether self-nudges are as effective as traditional nudges in influencing people’s behaviors. There are at least two reasons why self-nudges may not be as effective. First, self-nudges require decision makers to know the intended effects of nudges, but nudges may become less effective when their effects are made transparent. Second, relative to experts, laypeople may not know how to design optimal choice architectures to influence their own behaviors. If people are poor choice architects for themselves, then self-nudges would be less effective than traditional nudges.
In our study, we address this research gap by providing empirical evidence for the effectiveness of self-nudges compared to traditional nudges.
Our study
Participants
Participants were recruited from the University of California San Diego (UCSD) (N = 383). All participants completed the study via a Qualtrics survey, and received course credit for their participation.
Procedure
Participants were randomly assigned to one of four conditions: a control condition, two traditional nudge conditions, and a self-nudge condition. In the control condition, neither video was preselected as the default option. In each traditional nudge condition, one of the two videos was preselected as the default option, as in typical default experiments. In contrast, in the self-nudge condition, the participants were asked to select a video as the default option for themselves before making the choice.
The study consisted of two parts.
In Part 1, participants learned about two videos and were told that they would choose one video to watch later in Part 2. This is what they learned:
| Plastic Pollution Viedo |
Life Hack Video |
| Education video on the impact of excess plastics |
Entertainment video making fun of silly life hacks |
| Length: 1 minute 7 seconds |
Length: 1 minute 7 seconds |
They then saw a preview of how the two videos would be presented to them–this differs by condition. In the control condition,they saw the two videos and either was preselected; in the two traditional nudge conditions, they saw the two videos and one was preselected. In addition, participants in these three conditions were told to only think about their choice options and to not take any action yet. In contrast, in the self-nudge condition, participants saw the two videos and learned that either video can be preselected, and were asked to preselect a video for their own choice later in Part 2. In the self-nudge condition and the two tradition nudge conditions, participants further learned that people are more likely to select options that are preselected for them. This is what they saw in the control condition, one of the traditional nudge conditions, and the self-nudge condition.
In Part 2, the two videos were presented as shown in the preview or as selected by the participants themselves, and participants then made thier choice and watched the video.
Findings
We first noticed that participants had a baseline preference for the Life Hack Video.
We then compared participants’ choices in the two traditional nudge conditions in order to test for traditional default effects. A Chi-square test of independence showed that participants’ choice did not significantly vary by condition (χ2 (1) = 0.605, p = 0.437). These results suggest that participants’ choice of video was not influenced by the default option. In other words, there was no traditional default effect.
In order to test whether self-nudges are more effective than traditional nudges, we compared the percentage of participants who stayed with the default—watched the preselected video, both in the two traditional nudge conditions and in the self-nudge condition. Two-sample proportion tests between each traditional nudge condition and the self-nudge condition indicate that **a significantly larger proportion of participants stay with the default when they chose the default themselves than when the experimenter chose the default for them. **
References
- Johnson, E. J., & Goldstein, D. (2003). Do Defaults Save Lives? Science, 302(5649), 1338–1339. https://doi.org/10.1126/science.1091721
- Steffel, M., Williams, E., & Pogacar, R. (2016). Ethically deployed defaults: Transparency and consumer protection through disclosure and preference articulation. Journal of Marketing Research, 53. https://doi.org/10.1509/jmr.14.0421
- Jachimowicz, J., Duncan, S., Weber, E., & Johnson, E. (2019). When and why defaults influence decisions: A meta-analysis of default effects. Behavioural Public Policy, 3, 1–28. https://doi.org/10.1017/bpp.2018.43
- Thaler, R. H., & Sunstein, C. R. (2008). Nudge: Improving decisions about health, wealth, and happiness. New York: Penguin.
- Madrian, B. C., & Shea, D. F. (2001). The Power of Suggestion: Inertia in 401(k) Participation and Savings Behavior. The Quarterly Journal of Economics, 116(4), 1149–1187.
- Ebeling, F., & Lotz, S. (2015). Domestic uptake of green energy promoted by opt-out tariffs. Nature Climate Change, 5(9), 868–871. https://doi.org/10.1038/nclimate2681
- Reijula, S., & Hertwig, R. (2020). Self-nudging and the citizen choice architect. Behavioural Public Policy, 1–31. https://doi.org/10.1017/bpp.2020.5