Promotion of Behavioural Change for Health in a Heterogeneous Population

"A considerable amount of research has demonstrated the importance of social norms and conformity for vaccine delivery..."

Public health policy objectives often conflict with local culture. Thus, to avoid a backlash when trying to change people's behaviour, policymakers may work with public health professionals to design interventions aimed at changing behaviour that do not restrict freedom of choice. Such interventions can have two main effects: (i) a direct effect on people initially sought to be reached by the intervention; and (ii) an indirect effect mediated by social influence and by the observation of other people's behaviour. However, behavioural spillover can be influenced by the interaction between: (i) the variation in people's tendency to make a socially beneficial behavioural choice based on how common this choice is among their peers; and (ii) the variation in perceptions of the costs and benefits associated with a behavioural choice. This paper illustrates the key principles of this interaction by examining two areas of public health policy: tobacco smoking and vaccination.

The example of antismoking campaigns shows when and how public health professionals can amplify the effects of a behaviour change intervention by taking advantage of the indirect pathway. Some people in the population will change their behaviour after being directly exposed to the media campaign (i.e., the direct effect). The remainder will not have changed their behaviour, either because they did not respond after being exposed to the campaign or because they were not exposed to it. This second part of the population may still change their behaviour by stopping smoking after observing others doing so. However, some people may stop smoking only after observing 85% (or 50%, or 20%) of people they know doing so.

Intuition might suggest prioritising those more amenable to change when designing a communication campaign. But the picture is more complex. As explained here through various models, when the probability of responding to an intervention by changing behaviour is high for everyone except those most resistant to change, all individuals but the most resistant will probably change their behaviour and stop smoking if exposed to the media campaign. In this case, reaching out to individuals amenable to change would yield only a small increase in the direct effect of the intervention. Behavioural spillover and, by extension, the total effect of the intervention, are much larger when either randomly selected or resistant people are the focus. (Other modeling here examines what would happen when the probability of responding to an intervention by changing behaviour is very low for everyone, as well as the benefit of just reaching out to randomly selected individuals.)

The example of vaccination campaigns illustrates how underlying incentive structures can interfere with the indirect effect of an intervention and stall efforts to scale up its implementation. In short, public health professionals generally aim to increase the proportion of vaccinated individuals in the population beyond the threshold for herd immunity; social influence may support them in this endeavour. However, once a certain proportion has been vaccinated, anticoordination incentives can favour opting out of vaccination. Namely, the perceived value of vaccination may decrease as the number of people vaccinated increases (the risk an unvaccinated person will catch the disease declines as vaccination becomes more widespread). The perceived cost of contracting the disease when unvaccinated must be balanced against the possible fixed costs of vaccination, including fear. Once enough people have been vaccinated to ensure that, for the unvaccinated, the perceived value of not getting the vaccine exceeds the perceived value of vaccination, the proportion of people vaccinated will reach a stable equilibrium.

As outlined here, public health professionals could use different strategies to counter anticoordination incentives (a combination of these strategies is recommended):

• Make vaccination rewarding: Reduce the monetary and time cost of vaccination or provide financial incentives;
• Make vaccination appealing: Communicate effectively about the low risk of vaccination side effects and the health benefits of the vaccine and counter people's fears, which may originate from or be exacerbated by misinformation and/or antivaccine movements; and
• Make vaccination easy: Use behavioural nudges - for example, in the form of reminders, prompts, or default options.

Recommendations are presented on how public health professionals can maximise the total effect of behaviour change interventions in heterogeneous populations based on these concepts and examples:

1. Ask: Are the resources and infrastructure available to reach the whole population with a behaviour change intervention?
   • If yes: Behavioural spillover can be cost efficient but may not be necessary to achieve the policy goal.
   • If no: Behavioural spillover might help indirectly reach those who cannot be reached directly.
2. Ask: Is detailed information available about the heterogeneity of attitudes in the population?
   • If yes: Knowledge about attitudes can be used to maximise the total effect of the intervention, not just the direct effect.
   • If no: There are different solutions: (i) develop the methods and capacity needed to assess existing attitudes; (ii) use available sociodemographic data as a proxy for attitudes; or (iii) choose to reach out to randomly selected individuals with the intervention.
3. Ask: How effective is the intervention in changing behaviour among different individuals in the population?
   • If it is effective among everyone except the most resistant, it is advisable to avoid reaching out to the most amenable.
   • If it is effective among only the most amenable, the campaign is not particularly effective overall. Solutions: (i) improve the effectiveness of the intervention; (ii) selectively reach out to only the small amenable population using low-cost measures; (iii) directly reach out to a large subset of the population with an intervention; or (iv) design an intervention that decouples group identity from the target behaviour if resistance to the behaviour change is related to deeply rooted group identities or traditions.
4. Ask: Does the target behaviour involve coordination incentives (i.e., individuals will prefer behaviour A to behaviour B when everyone else is exhibiting behaviour A and will prefer behaviour B to behaviour A when nobody else is exhibiting behaviour A)?
   • If yes: The focus should be on triggering behavioural spillover because conformity and coordination incentives will both support the policy objective once the desired behaviour is sufficiently common.
   • If no: It is possible that anticoordination incentives may be encouraging people to do the opposite to others. These can be attenuated by: (i) increasing the value of the target behaviour; (ii) providing information; or (iii) employing behavioural nudges.
5. Evaluate the impact of the behavioural change intervention: Even if an intervention seems unsuccessful at first, it may create large-scale behavioural change through spillover. Consequently, the effect of the intervention should be evaluated in a sample of people who were not the focus. Repeated evaluations are advisable, as a single evaluation immediately after the intervention ends could miss indirect effects that unfold over time.

In conclusion: "Ideally, public health professionals should gain a better understanding of both the direct and indirect effects of their interventions by assessing the different attitudes and beliefs among people in their target population. The strategies we describe could help public health professionals take their first steps towards effectively managing behavioural spillover when designing public health interventions."