Combining Both Norms

A message can describe what people do or signal what people approve of, and those signals can point in different directions. A sign that emphasizes how common theft is may accidentally make stealing seem normal even while asking visitors to stop. Pairing truthful information about helpful behavior with approval of that behavior can support the intended action, but results depend on the setting. Check this: does your message make the behavior you want to discourage sound normal?

Separate two kinds of signal

A descriptive norm concerns what people do. An injunctive norm concerns what people approve or disapprove of. A message may contain both, but the part that attracts attention can influence how people respond. Aligning them is a useful design principle, not a guarantee that one wording is the strongest in every setting.

Avoid advertising the unwanted behavior

In research at Petrified Forest National Park, signs emphasizing widespread theft could encourage more theft than signs emphasizing disapproval. This was a comparison between experimental messages, not evidence that a particular historical park sign caused theft to triple over time. The lesson is to consider what behavior your message makes seem normal.

Support people already doing well

Schultz and colleagues gave households feedback comparing energy use with neighborhood averages. Among households already using relatively little, descriptive feedback alone produced an unwanted increase. Adding a signal of approval prevented that boomerang effect in the study. The feedback was part of a field intervention, not simply a universal effect of adding a smile to any energy bill.

Keep related findings distinct

Hotel research found that descriptive messages about towel reuse outperformed a standard environmental appeal, especially when the comparison group matched the immediate setting. That does not by itself isolate a universal benefit from combining two kinds of norm. Separate research on dynamic norms suggests that truthful information about a behavior becoming more common can motivate action even when it is not yet the majority behavior.

Test an honest message

For a hypothetical recycling campaign, a verified statement that most residents sort their waste could be paired with appreciation for doing so. If few residents do it, do not invent a majority; describe a genuine trend only if there is evidence for it. Specify the action and compare outcomes in the setting where the message will be used, including any unintended effects.

Norm messages can change behavior, but effects depend on the behavior, comparison group and signal made salient.

• Cialdini et al. (2006) compared descriptive and injunctive anti-theft messages at Petrified Forest National Park; emphasizing the prevalence of theft was more likely to encourage it, while emphasizing disapproval was more likely to suppress it. [1]
• Schultz et al. (2007) found that neighborhood-average energy feedback could produce a boomerang effect among low-usage households, and that adding an injunctive approval message prevented that effect in their field intervention. [2]
• Goldstein et al. (2008) found that descriptive towel-reuse appeals outperformed a standard environmental appeal, with the most effective normative comparison matching the guest's immediate room setting. [3]
• Sparkman and Walton (2017) tested dynamic norms and found that information about changing behavior could motivate sustainable actions even when those actions were not yet the prevailing behavior. [4]

These studies do not establish one universally optimal persuasive message or prove that effects persist indefinitely. Stated norms should be accurate and relevant to the audience. An undesirable behavior can be normalized by emphasizing its prevalence, but not every such message necessarily backfires. Results may differ with the setting, comparison group and behavior; test the intended outcome and possible unintended effects rather than treating a message formula as a guarantee.