In behavioural research, it is largely assumed that people adapt their personal behaviour to match the behaviour of a reference group. A distinction is made between descriptive norms – the assumption that others behave in a certain way – and injunctive norms, where people believe that others expect them to behave in a certain way.
In health communication, this phenomenon is perceived as an opportunity to use the effectiveness of social norms in communication to guide individuals towards health-promoting behaviour: 'Book your NHS Health Check. Six million people have already taken part,' was used for a campaign in the UK. It is about confirming or falsifying perceived social norms to achieve the intended behaviour in the target group. In addition, social-norm-campaigns are relatively easy to set up and low in cost.
Meta-analysis shows no effect
In a meta-analysis recently published in the journal Nature Human Behaviour, a group of behavioural scientists from the UK and the University of Innsbruck found that it is not as simple as previously assumed in research and health communication. The analysis of 89 studies that examined the effect of communicating social norms to promote beneficial health behaviour in developed countries showed no statistically significant effect.
'There are both research-related and practical reasons for this finding,' explains co-author Sarah Flecke. Until recently, Flecke conducted research at the Institute for Banking and Finance at the University of Innsbruck. Now she is based at the University College London (UCL).
Before starting the meta-analysis, publication bias had to be considered: 'Many of the studies in this area show a null effect. As studies without statistically significant results are published less likely, a proportion of so-called 'grey literature' was included,' says Flecke. The research team collected unpublished studies via personal contacts.
Strict criteria were defined and pre-registered by the team before analysing the results: 'We were careful to extract information from a study’s primary research aim and hypothesis,' explains Flecke. If a study was aimed at the effectiveness in the overall population, but positive effects were only discovered in a subgroup, the main results – with a null effect – were still reported.
The authors point out that few of the included studies assessed whether normative information changed participant perceptions and used manipulation checks. Furthermore, there is a substantial heterogeneity across studies in terms of target behaviour, population, and intervention delivery.
To account for the heterogeneity, authors categorised the various studies included in the analysis and examined whether particular study characteristics were more or less effective at improving health behaviours.
Discrepancy between knowledge and behaviour
Social norms messaging is supposed to work via a preference for conformity with the behaviour of the group. It is motivated by the key assumption that perceptions of peers’ attitudes and behaviour is incorrect: 'In order to design targeted social norm campaigns, you need to know in advance what individuals perceive as the norm and what they do not.' Only ten studies elicited whether there is a discrepancy between the perception of the norm and the actual norm. In many cases, it may not be a lack of awareness of health-promoting behaviours, but other factors that cause people not to change their behaviour.
'Our analysis suggests that when appropriately controlling for publication bias, social norms messages are not effective at improving health behaviours,' says Flecke. 'Practitioners – e.g. policymakers and health communicators should be cautious about using social norm interventions and expecting them to work.'
'Still, this does not mean that social norms don't work as a communication strategy,' emphasises Flecke. The research team did observe some interventions that had been successful. Therefore, future research should look more into the mechanisms that explain this finding: 'Is it about the informational value of a norm? Is it the discrepancy between the individual’s perception and actual social norms? What role does the visibility of campaigns and frequency of exposure to the messages play? Many aspects can be derived from the limitations that we have seen in the analysed studies, providing opportunities for future research,' says Flecke.
Publication: Papakonstantinou, T., Flecke, S. L., et al. (2025). A systematic review and meta-analysis of the effectiveness of social norms messaging approaches for improving health behaviours in developed countries. Nature Human Behaviour. DOI: 10.1038/s41562-025-02275-6