Social Norms and Peer Incentives

People often look to others for clues about what to do, especially when they are unsure. In a hotel, guests may be more likely to reuse towels if they hear that most other guests do it. Try this: mention what similar people already do, and add a brief sign of approval so people who are already doing well keep it up.

In hotel studies, a message like "Most guests who stayed in this room reused their towels" illustrates a descriptive norm: it tells you what similar others do. If the hotel adds a smiley face, checkmark, or note like "Thank you for helping us conserve water," that becomes an injunctive cue: it signals that this behavior is socially approved.

These are two widely studied types of social norm information. Descriptive norms answer the question, "What are people like me doing?" Injunctive norms answer the question, "What behavior gets social approval?" Descriptive norms can guide behavior when people are uncertain, while injunctive cues matter because people care about belonging and approval.

Research by Schultz and colleagues revealed a critical design issue: when better-than-average performers learn they are doing better than the norm, they can relax their efforts and move toward the average. In an energy study, households already using less electricity actually increased consumption after learning they were below the community average, but adding a smiley face eliminated this backsliding by showing approval.

The same idea shows up outside sustainability settings. In healthcare-worker interventions, a meta-analysis of 116 trials found small average improvements in professional behavior, with social-comparison messages often working better when they came from a trusted source and used a reference group the audience cared about. In practice, use relevant reference groups, place the norm message near the decision point when the context allows, and let high performers know their behavior is noticed and valued.

Social norms interventions often produce modest effects, but effectiveness varies substantially by context, reference group, and implementation.[2][4]

• Schultz et al. (2007): In a field experiment with 290 households, descriptive-norm feedback reduced energy use among high-use households but increased it among low-use households; adding an injunctive approval cue removed that boomerang effect.[1]
• Tang et al. (2021): A meta-analysis of 116 healthcare trials found average improvements of 0.08 standard deviations in professional behavior and 0.17 in patient outcomes, with credible-source and social-comparison techniques associated with larger effects in that review.[2]
• Yamin et al. (2019): A systematic review of sustainability-focused field studies found that social-norm interventions tended to work better when the reference group and delivery context fit the target behavior than when messages used a generic audience frame.[4]

Social-norm interventions are heterogeneous: effects vary by context, behavior, and reference group, and average effects are often modest. People respond most to norms from groups they identify with, so a poorly matched comparison group can weaken or reverse the effect. Descriptive comparisons can also create a boomerang effect, where already-better performers relax toward the norm unless an injunctive approval cue is added. Much of the evidence focuses on short-term behavior change, so long-term effects remain less certain.