Social norms are often used in public information campaigns by government, however, behavioural change is not always adequately measured. Social science can help marketers, decision-makers, and businesses to effectively use social norms to change behaviour. Below are two case studies on the environment, health, and the use of humour to target damaging behaviour.

Summary

• Many public information campaigns rely on social norms, but the behavioural impact of these advertisements are not always tested. This sometimes means that we don’t know if the messages are having a positive result. It might also lead to unintended outcomes.
• Social science can ensure messages are effective. We can target the most relevant social norm to the desired behaviour, and then measure behavioural change.
• An environmental study shows that injunctive norms can stop people from doing the wrong thing, but only when the bad behaviour is described negatively. In this case, telling visitors “don’t take the wood” reduced stealing from a national park. Telling visitors that many people have stolen wood in the past and caused damage to the park led to more theft. This made it seem as if stealing was normal
• This study shows it’s important to test communications. Reminding people about what is commonplace can backfire, even when the behaviour is bad. If everyone else is stealing, why shouldn’t I?
• A health study shows that injunctive norms impact binge drinking. Young people who personally approve of excessive drinking, and who believe their friends and parents would approve, are more likely to drink heavily. They were not influenced by other students or others of the same gender.
• This study shows that understanding the most influential people of specific groups requires testing. Many advertisements target parents as role models, and same-age peers, but the focus of these campaigns may not always be persuasive if the individual believes those closest to them are supportive of their behaviour.
• Social science can help advertisers, policy-makers, and businesses to choose the correct social norm, and track behaviour change, using scientific theories and methods.

Social norms are often untested in public campaigns

Last time, we learned that social norms are the shared, informal, and unwritten rules that shape our understanding of acceptable behaviour. Norms are culturally and historically shaped, and fundamentally social (that is, collectively enforced).

Descriptive norms are the typical patterns of behaviour that we come to expect, based on our observations of other people’s overt actions. This relies on our perception of direct and indirect communication, and our desire to be accepted (i.e. one of the crowd).

Injunctive norms are the prescriptive moral rules of what is acceptable, especially in relation to reference groups. That is, people like ourselves, or those whom we’re close to (e.g. family, friends). Injunctive norms reflect the morality of what we think we ought to do, based on what we think others expect from us, as well as our need for approval (i.e. doing the right thing).

Society relies on norms to encourage compliance with rules and laws. We are conditioned to follow desired behaviours through education, communication campaigns, signs, and other public reminders. But do these campaigns actually work to change poor behaviour?

Public campaigns often focus on emotion, with a social norm message, but the outcome is not always measured. This can sometimes lead to weak outcomes.

For positive social messages, campaigns often focus on prosocial messaging (selfless actions that improve the wellbeing of others in our communities, such volunteering). For example, the 2021 Australian Government campaign, “Do the right thing” (below), encourages the public to take care of our environment. The advertising uses positive images of happy families on a picnic, a cheery and familiar jingle (used in the 1970s, and 1990s), coupled with a prosocial social norm message (“change what you do, and others will too”).

While the Keep Australia Council reports it is the most successful anti-litter campaign, it last measured success in 2020, through its own biannual litter survey across 983 undisclosed national sites. Anecdotally, this campaign has high public recognition and trust, but the evidence does not illustrate whether this campaign directly impacts on individual behaviour. Other studies find that concurrent strategies lead to reduced waste, such as community education on recycling alongside and providing recycling bin and kerbside collection service. In sum, there is no evidence that social norms alone are being used as effectively as they might be in this well-known environmental campaign. Instead, social norms are bolstering other prosocial behaviours (e.g. recycling).

When deterring negative behaviours, public campaigns often focus on negative emotions, tacking on a social norm message. Negative experiences often leave a higher impact on our memories, even in comparison to positive events of equal importance (negativity bias). This is not always a good thing, as we pay inordinate attention to negative comments and make quick, often incorrect assumptions. Social marketing drawing on negative appeals, such as fear, guilt, and shame, often lead vulnerable people to tune out the message.

My previous research with my colleagues shows that negative emotions are ineffective in creating change. This trend bears out in many public health and safety campaigns.

For example, in 2008, the Australian Government released a series of graphic television, bus, and other advertising showing drunken teenage girls being sexually assaulted, a car accident with a teenage male driver, and boys fighting, with the tagline “Don’t turn your night out into a nightmare.” The central message used in marketing was “Binge drinking can lead to injuries and regrets.” More recent campaigns target parents. In late 2024, the Western Australian (WA) Government “We all need to say no” campaign reinforces that the majority of parents don’t allow their children to drink (“Join the 2 out of 3 parents in WA already saying ‘no’ to giving alcohol to under 18s.”) The television advertisement uses normative messaging “It’s easier to say no when we all say no.”)

While these tactics draw on social norms, the campaigns either did not have the intended outcome, or did not measure behavioural change. The 2008 “Night out” campaign was criticised by experts when it went to air. A subsequent academic study found that young people generally understood the meaning, but the message was less salient for the target demographic—young people who drink excessively (six or more alcoholic beverages). The “We all need to say no” campaign evaluation found that parents understood the message, but the evaluation did not measure whether this led to fewer parents permitting less alcohol, nor did it measure whether the campaign resulted in fewer children binge drinking.

Other studies show that social norms are effective when they are targeted to specific audiences in specific settings.

Targeting social norms to the desired behaviour leads to better outcomes

Social science research provides the theory and methods to design targeted campaigns that effectively use social norms, and measure behavioural change. Studies help us to focus on specific norms that match a desired outcome, and how to test the impact. Specifically, differentiating between descriptive and injunctive norms is key to behavioural change, especially for risky behaviours.

Case study 1: Negatively framed injunctive norms protect the environment

A study tested descriptive and injunctive social norms to stop people from taking wood from a national park.

Prior to the study, the Arizona Petrified Forest National Park displayed signs saying, “Your heritage is being vandalised every day by theft losses of petrified wood of 14 tons a year, mostly a small piece at a time.” While only 5% of visitors steal wood, over one million people visit the park annually, and the individual acts add up to major environmental damage. This message is a familiar one, where communications campaigns emphasise that bad behaviour is regrettably frequent.

To measure how much wood was stolen during the study, the researchers marked 600 pieces of wood along a path, and tested two sets of messages with the same meaning (stop stealing wood), but each had either a positive or negative framing. People’s decisions are influenced by the way information is presented, even when the meaning is equivalent (framing effect). For example, 10% unemployment has a negative framing, versus 90% employment, which is viewed positively, even though the employment rates are identical.

In the petrified wood study, some people saw the injunctive norm signs asking them to preserve the natural state of the park, but each sign was either negative or positively worded. The negatively framed injunctive norm was strongly worded:

“Please don’t remove the petrified wood from the park.”

It was accompanied by a picture of one person stealing the wood, with a red circle and bar crossing over their hand, signifying don’t do this. The picture depicts only one person, reinforcing what the individual shouldn’t do.

This sign led to the desired behaviour, reducing theft from 7% to 2% because it reminds individuals that other people disapprove of taking wood.

The positively framed injunctive norm used weaker language. This sign said, “Please leave petrified wood in the park,” with a picture of one visitor taking an admiring photo of wood. This message did not lower the rates of theft.

Conversely, the negatively framed descriptive norm message (which used strong language) led to the highest levels of theft. The sign said: “Many past visitors have removed the petrified wood from the park, changing the state of the Petrified Forest,” showing a picture of three visitors taking wood.

The positively framed (weakly worded) descriptive norm sign said, “The vast majority of past visitors have left the petrified wood in the park, preserving the natural state of the Petrified Forest,” showing three visitors admiring and photographing the wood.

These descriptive norm pictures show three people doing the desired behaviour. Popularity (showing what everyone does) is central to descriptive norms, because descriptive norms remind us what we think is normal to everybody else.

Both descriptive norm messages described what many past visitors had done. However, the negatively framed descriptive norm inadvertently normalised the bad behaviour, by reminding people that visitors consistently take wood.

The positive framing for both the injunctive and descriptive messages were equally ineffective at stopping theft, because of their weaker wording, while the negative framing of the descriptive norm had the opposite intended outcome, encouraging more theft.

People who have a tendency to hold onto incorrect ideas will increase the intensity of their convictions, and refuse to change their behaviour, despite being shown contradictory evidence or a credible correction. This is known as the backfire effect (also known as belief perseverance, or boomerang effect). Messages backfire due to deeply held beliefs and preconceptions, repeated exposure to incorrect ideas, or structural factors (such as culture, religion, politics, or other socialisation). This is why debunking myths can be ineffective, because repeating misinformation can reinforce incorrect beliefs, through familiarity.

In this case, repeating that many people do the wrong thing only encourages more bad behaviour, while simply telling individuals what they should not do removes this muddle. Telling people that the unwanted activity is rare and widely disapproved is the best way to deter bad behaviour.

The backfire effect was magnified when the park administrators decided not to change their signs, despite the statistically significant result of the study. Visitors had anecdotally told Park Rangers that the existing signs were a deterrence, despite evidence to the contrary: the wood continued to be stolen with the original sign, and the study had now shown that telling visitors that many people steal leads to higher rates of theft. The Rangers and administrators were afraid that changing the signs might have adverse outcomes. They preferred to continue with their well-intentioned, subjective ideas, rather than trusting the scientific evidence (confirmation bias).

Information, even scientifically backed evidence, is not always enough when people have a vested interest in resisting change.

This case study shows the importance of testing. Western society often forces positive messages, but these are not always ideal in creating positive change. It also shows the uphill battle in changing minds when the outcome challenges the status quo.

Case study 2: Injunctive norms can reduce binge drinking

A study aimed to reduce binge drinking among youth. They tested two norm types: descriptive and injunctive. University students who drink heavily (five or more drinks for men, or four or more for women, on at least one session) were asked to report their weekly drinking.

The researchers measured the injunctive norm by asking participants to rank their approval of drinking every weekend, daily, after driving, and enough to pass out (“risky drinking”). They then ranked much they think others might approve of these levels of drinking (“perceived approval”), with reference to several reference groups: the typical student, a student of their same gender, friends, and parents. Those who had high personal approval of risky drinking and who perceived high approval from friends and parents were more likely to have higher rates of binge drinking. The perceived approval of other students and others of the same-gender had no impact.

The descriptive norm was established by asking the students how much they think other people drink, in reference to other students, and students of the same gender. Men drank more heavily than women, and perceived that other men were the same. However, both men and women in this sample overestimate how much other students drink, though men think other men drink more than the typical student, while women think women drink less than typical students. In other words, students and men are perceived to be heavy drinkers.

This study shows that injunctive norms have a greater impact on binge drinking. Young people who drink excessively are influenced by those closest to them, but not by other people like themselves.

In the Australian ads described above, parents are targeted about their children’s underage drinking, but only as models of behaviour. Other ads targeted young people by showing their friends (binge drinking) and other same-age peers (speeding). This study suggests that such campaigns need a further step: youth need to know their friends and parents don’t approve of heavy drinking, they need to see their parents and close peers model this behaviour, and then be reminded of the descriptive norm.

How social science helps

• Social science theories can help us to answer key questions. For example:
  • What is the desired behaviour change? Consult social science studies, and adapt the intervention to your local context.
  • Which social norm best applies? If most people already engage in the behaviour, and we want to remind people about fitting in, descriptive norms might be best. If we want to remind people about what’s acceptable, and we are confident influential groups would sway behaviour with their approval, injunctive norms might work
• Social science methods can help us determine how to measure behavioural impact. Test the outcome scientifically, don’t just rely on subjective observations.