Work in psychological science around the creation, communication of maintenance of social norms to change behaviour has proven very influential. Due to both its conceptual simplicity and its ability to be seemingly widely applied with stunning effects, social norm theory has ended-up being central to a whole host of behavioural interventions. From littering to stealing, from missing doctors’ appointments to completing tax returns on time, and from speeding to saving energy. The latter allowed energy saving behaviour company Opower (now part of Oracle) to secure a valuation of $750million. Not too bad for finding a way to implement descriptive and injunctive norms.
When it comes to designing behavioural interventions to change behaviour for the better, using social norms has become….a norm (ahem).
But there is a challenge to using norms — descriptive norms, specifically — to modify behaviour. If the descriptive norm does not align with the desired behaviour, then the effect it may have on the wider population is to move behaviour away from the target, rather than towards it.
This was famously seen in Schultz et al.’s research looking at the interplay between descriptive norms and injunctive norms when it comes to energy saving behaviour at home, and was used to great effect in Opower’s programme design. When home owners received energy bills that showed they were using less energy than the average in their street, in order to stop the negative effect of those homeowners increasing their consumption, a social injunctive norm was primed through printing a smiley face on the bill. This way, those above the average lowered their usage, and those below the average tended to stay below the average.
Seems a smiley face has proven to be extremely effective at priming an injunctive norm — most of us are familiar with speed warning devices by the side of the road that move from a smiley to frowning face as we click over the limit. And for all those who have had that experience, the feeling is quite profound as you tip from being socially accepted to being socially in the wrong.
But we also take our cue from when norms tell us not what’s an acceptable level of behaviour, but how many people are doing it.
Communicating proportions of a population who engage in a (constructive) behaviour can be a highly effective use of social norms. Filling-in tax returns on time (and honestly) is a socially expensive behaviour which has been dramatically reduced, simply by communicating the proportion of the population who do not engage in the bad behaviour. Which is fine, until the behaviour you want to encourage is very much in the minority. If only 30% of people are doing the right thing, that means more than twice as many are not. In other words, the wrong thing is the norm. In these instances, using a norm to communicate a wish to change behaviour can have dramatic effects — in the wrong direction. Robert Cialdini — arguably the chief architect of finding ways to use normative influences for better — points to the US campaign to reduce littering in the 1970s as the perfect example of this rebound effect (watch the original PSA film and you can see immediately why the normative message would work in the opposite direction as intended — you can see it here).
Fast forward almost 50 years, and this challenge with using normative effects is particularly relevant when it comes to key behaviours that need to change today— for example driving an EV, removing meat from our diet, or saving energy at home. All of these behaviours are very much in the minority i.e. they are not the norm.
However, it would seem that social norm theory is the gift from social psychology that keeps on giving.
Recent work by Gregg Sparkman and Gregory Walton at Stanford has explored the novel concept of a dynamic social norm. Specifically, this describes a behaviour that is expressed both in terms of the proportion of a population that is currently doing it, along with a reference to suggest it is likely to grow. In other words, it’s a behaviour expressed as an emerging norm; a behaviour with momentum.
Here’s a concrete example from the research:
The researchers approached students and staff at a US university who were queueing for their food in one of the cafeterias. Alongside the control group, those in the queue were asked to read a statement relating to the proportion of US residents who were cutting meat out of their diet. For one group, the statement read as follows:
“Some people limit how much meat they eat. This is true both nationally and here at Stanford. Specifically, recent research has shown that 30% of Americans make an effort to limit their meat consumption. That means that 3 in 10 people eat less meat than they otherwise would.”
This was the standard, or what the researchers call, a static norm.
And for the second group, it read as follows:
“Some people are starting to limit how much meat they eat. This is true both nationally and here at Stanford. Specifically, recent research has shown that, over the last 5 years, 30% of Americans have started to make an effort to limit their meat consumption. That means that, in recent years, 3 in 10 people have changed their behavior and begun to eat less meat than they otherwise would.”
This was the dynamic norm, in that it (subtly) indicates a direction of travel for the effect (‘have changed’ and ‘begun to).
The researchers then recorded the lunch the respondent then went and ordered (via giving them a voucher for taking part). The results are striking. For those in the static norm condition, 17% of those in the group went on to choose a red-meet free option. In the dynamic norm condition, that number increased to 33% — almost double. What’s more, under the conventional (static) norm condition, the number of people choosing to go meet-free was essentially no different than in the control condition, where the information had not even referred to meat eating behaviour.
And it’s important to recognise that the study described above involved not gauging people’s reactions or intentions to do something, but their actual behaviours (in this case, choosing a meal).
When it comes to convincing consumers to make energy-smart buying decisions, we face this dilemma constantly: the proportion who are already engaged in the behaviour is damagingly low, potentially destroying the value of conventional static norms.
Take electric vehicle (EV) adoption. Currently across the US, EV adoption counts for less than 2% of the total car buying market. So the potential to use traditional normative cues to drive behaviour is likely zero. But a dynamic norm? 2% of changed their car choices in recent years and have begun to think differently about mobility? That could represent a real opportunity to influence in the decision process.
If you agree, you’re part of the small but growing group who now believe dynamic norms have the potential to drive more energy-smart behaviour, and who are rethinking how we can use new forms of norm for behaviour change.