Social Norms to Influence Behavior

A guide to understanding nuances in social norms interventions

Social norms are collective beliefs about what kind of behaviours are expected (or appropriate) in a given situation. No matter how much you think you are your own person, we are all influenced by them. But being affected by what other people think or do just means that you’re a human being.

In this article I will discuss some of the ways in which social norms affect our behaviour and how simply telling someone what others are doing may not be enough.

Imagine you’re on a walk with a friend when you decide to drop by the library to pay a visit to another friend who works at the reception. The second you set foot in the library you will likely quiet down and start whispering to each other, instead of speaking at a normal volume, just because that’s what everyone does in a library.

As social animals, we tend to follow social norms to feel a sense of belonging to the group and because we internalise social norms and believe they are “the right thing to do”, even though these beliefs can be misguided sometimes. We care about how we are perceived and treated by others, so following social norms protects us from the negative consequences that can come with transgression (e.g., being mocked or excluded from the group).

As such, telling people what others are doing is usually a good way of influencing behaviour. Take a behaviour like wearing your seatbelt. Nowadays, most of us would agree that wearing a seatbelt is important for safety and would consider it a “normal” thing to do when getting in a car. However, this wasn’t always the case. A few decades ago people used to underestimate how important and widespread the use of seatbelts was. Social norm campaigns helped people realise that the use of the seatbelt was actually more popular than they thought, correcting that misperception and shifting both actual behaviour and beliefs about what’s appropriate in a car [1].

Social norms are a well-researched area in behavioural science [2]. When researchers talk about norms, they often make a distinction between:

• Descriptive norms - what most people do
• Injunctive norms - what our reference group think is the right thing to do 
• ‍

Although you may have read about them with different names. For example, Cristina Bicchieri, a leading authority in the study of social norms, refers to these as empirical expectations and normative expectations, respectively. Just like the seatbelt example illustrates, norms are not static. They are dynamic in nature and are constantly reviewed and reassessed, so that beliefs that are normative today may not be normative in the future. 

Social norms have often been used to steer people towards more pro-social and pro-environmental behaviours. For example, a well known study by Goldstein and colleagues [3] showed how displaying descriptive normative messages in hotel rooms led to a 44% rate of towel reuse compared to 35% when standard environmental messages were displayed.

Whatsmore, they showed how making the norm as local and relevant to the target group as possible increased the efficacy of the social norm messaging, with towel reuse increasing to up to 49% when the message referred to the behaviour displayed by previous guests staying exactly in that hotel room. This shows how we care about in-group norms in particular.

The more similar we are (or feel) to the reference group, the more likely we are to adhere to the group’s social norms. We can relate to other members of the group through personal similarities (e.g., in gender, ethnicity, age, values, beliefs…) or through situational similarities (e.g., being in the same place, like the same hotel room), even when these similarities are artificially created by, for example, randomly assigning people to specific groups. Similarly, normative messaging has been found to reduce alcohol consumption [4] or help save energy [5].

Social norms are powerful influences on behaviour, so much so that sometimes social norms interventions are presented as ‘the Holy Grail’ for behaviour change. However, research shows that these interventions can also fail and sometimes have unintended results. In this article I will go through 2 considerations when thinking about using social norms to influence behaviour.

Social norms can backfire

Telling someone what others are doing can be a powerful way to influence their behaviour. Social norms are meant to challenge individuals’ perceptions of a behaviour and guide their behavioural patterns and intentions. However, research shows that by presenting information in a way that suggests that the behaviour is common, you might end up unintentionally increasing the undesired behaviour: the boomerang effect. 

For example, Schultz and colleagues [6] conducted a field experiment to test the efficacy of descriptive normative messages to promote household energy conservation. They manipulated the information that households received about their energy usage, including information about the average energy consumption of other neighbouring households (descriptive social norm). They found that households that were using more energy than their neighbours reduced their energy consumption after receiving the descriptive norm information.

However, they noticed that households that had been using less energy than their neighbours increased their energy consumption after receiving the normative messages, as if using less energy than the neighbours had given them “permission” to use more. This boomerang effect has been found across different behaviours. From research showing that descriptive normative messages related to alcohol consumption amongst students can increase drinking [7] to evidence showing that descriptive norms putting an emphasis on how many people are vandalising a forest by stealing pieces of wood can lead to more stealing [8].

The boomerang effect may occur when social norms suggest that the undesired behaviour is what most people are doing (“Many students are drinking alcohol on campus so it’s OK if I also drink”) or because they make the undesired behaviour more salient, which may in itself lead to an increase in the behaviour that we are aiming to reduce.

💡 How to avoid the boomerang effect

To avoid the boomerang effect, you should avoid presenting the descriptive norm information in a way that suggests that the undesired behaviour is the most common. Instead of just highlighting the current state of a behaviour (X% of people are doing it), research suggests that mixing descriptive and injunctive norms is an effective way to avoid the boomerang effect. In their study on social norms and energy consumption, Schultz and colleagues reported that mixing descriptive norms (information on your neighbours’ energy consumption) with injunctive norms (a smiley face to indicate that consuming less energy was positive or a sad face to indicate that consuming more energy was negative) eliminated the boomerang effect observed when only descriptive normative information was displayed. So thinking that most people are doing something AND thinking that doing the same is the right thing to do leads to a more powerful social norm. 

Being in the majority is not always possible

Interventions that make use of social norms tend to assume that the most effective reference groups for people to compare their behaviour are majorities (what most people are doing), since modelling an unusual or atypical behaviour seems not to make as much sense socially speaking.

However, describing a desired behaviour as something that most people are doing is not always possible. For example, data suggests that some recommended health behaviours such as eating 5 or more portions of fruit and vegetables a day or regularly testing for STIs are done only by a minority of people [9,10]. Research on minority social norms interventions shows mixed results.

Some studies have found that minority groups can still influence behaviour in the desired direction. For example, Aronson and O’Leary [11] demonstrated the influence of minority reference groups in a classic study to increase water conservation by encouraging people to turn off the water while soaping up. In their field study, they showed how one single person modelling the desired behaviour (turning off the water while soaping up) led to 49% of students complying with the behaviour, compared to a 6% of students when prompts encouraging them to save water were used without any modelling.

However, studies have also shown that minority norms can have detrimental effects on behaviour. Stock and colleagues [12] investigated the effects of majority vs minority descriptive social norms on intentions to eat and actual fruit intake. They found that minority social norms impacted both negatively. However, this negative influence of the minority social norm was mediated by how strongly participants identified with the minority reference group. Participants who strongly identified with the reference group showed a decrease in the intentions to eat fruit and in their actual fruit intake during the study. Participants who did not strongly identify with the group had a similar fruit intake to those who received majority social norm messages instead (this again highlights how important it is to relate to the reference group).

The strength of the identification with the reference group has also been found to be an important factor for social norm influence in other studies [13]. Given how sometimes referencing the minority may be unavoidable, it is important to understand these nuances in the research.

💡 Considerations when presenting minority reference groups

When presenting minority social norms, some evidence suggests that a way to avoid negative influences on behaviour is to emphasise the increasingly changing nature of the norm in the desired direction, even if the majority of people are not behaving in the desired way yet. For example, telling customers in a restaurant that an increasing number of people are reducing their meat intake, even if vegetarian or vegan diets are still the minority. That is, using dynamic (or trending) social norms.

In a pre-registered field study, Loschelder and colleagues [14] aimed to increase the use of reusable cups in a university cafe by placing signs by the coffee machines. In the dynamic social norm condition, signs told people how “customers are changing their behaviour. More and more are switching from the to-go cup to a more sustainable alternative. Take part in this, choose a sustainable cup”. The results from the study showed that presenting a dynamic norm where more people were joining the movement (even if they may still be the minority) had the strongest effect on sustainable choice compared to the control, descriptive social norm and mix of descriptive and injunctive social norm conditions.

These findings have been replicated in other sustainability-related studies as well [15]. Dynamic social norms may be more effective at increasing conformity when the behaviour is not (yet) performed by a majority because people predict that the norm will be followed by the majority in the future or because it may be more motivating to join a growing movement.