Discussion of main findings
There are several possible reasons why support for tobacco control interventions was generally high and higher than for interventions for other behavioural domains. First, the majority of people in high income countries no longer smoke [
119], so stronger support for tobacco control policies than, say, alcohol policies may reflect a preference for interventions that affect the behaviour of others. Second, there is a high level of awareness of the harm from tobacco, amongst smokers and non-smokers, which may lead to more support for interventions designed to reduce this harm. Third, there is a strong recent history for tobacco control interventions in the form of taxation increases and restrictions on where tobacco can be smoked and purchased. Public attitudes change with time and appear to often be influenced by the enacting of legislation, as shown in several studies by the increased acceptability of restrictions on smoking after the introduction of bans on smoking in public places [
58,
60]. This may be explained by a process of cognitive dissonance whereby attitudes follow behaviour (rather than vice versa, which is the more commonly assumed route to behaviour change) [
120], or by the operation of the status quo bias, a preference for the current state of affairs [
120,
121].
(b)
Intrusiveness of intervention
The greater acceptability of less intrusive interventions is illustrated in one poll in which 82% of those surveyed supported drink labelling (an intervention with no evidence of effectiveness in reducing alcohol consumption) compared with 45% who supported the setting of a minimum price per unit of alcohol (an intervention for which there is good evidence for effectiveness in reducing alcohol consumption) [
6]. Generally more restrictive policies are more effective, although not always. This finding appears consistent with the traditional economic world-view that people know best themselves what is good for them and are thus reluctant to accept public policy intervention that significantly interferes with their own decisions. Instead they tend to prefer interventions that do at best indirectly affect them (e.g. public awareness campaigns, education). The finding raises two questions: can public attitudes towards more effective interventions be changed and if not, when are governments justified in intervening regardless of public opinion?
Literatures from social psychology and moral psychology suggest two broad approaches to changing attitudes: first, targeting the beliefs that underlie attitudes; and second, activating the core values upon which acceptability judgments are based. Regarding the first of these approaches, information provided about the harmfulness of the target behaviour and the effectiveness of the proposed intervention to reduce this harm would be predicted to alter attitudes towards intervention [
122]. In keeping with this, a series of experimental studies found that the acceptability of financial incentives for stopping smoking and sustaining weight loss increased with rising effectiveness [
123]. Perceptions of the cause of unhealthy behaviour (e.g. individual decisions versus those shaped by the environment) also influence the perceived effectiveness and acceptability of interventions [
44,
124]. The second approach is based on observations in moral psychology that judgments are influenced by a series of core values (e.g. fairness). From this it would be predicted that acceptability would be increased by framing the outcomes of different interventions in terms of a core moral value such as fairness to those who are the focus of the intervention (e.g. children who are the focus of obesity prevention programmes) or those who are not the focus but who stand to gain by a change in behaviour (e.g. tax payers whose contribution towards the adverse health and criminal consequences of excessive alcohol consumption would be lessened by effective interventions). Framing consequences of interventions to target both beliefs and core values is likely to have the largest impact on acceptability.
Regarding when governments might be justified in intervening regardless of public opinion, one economic justification is when there is a ‘market failure’,
i.e. when the market, if left unchecked, produces a sub-optimal outcome from a societal perspective [
125]. Potential market failures relevant in the current context are imperfect information (e.g. people under-estimating the harm caused by certain health-related behaviours, in particular if those consequences materialise only far ahead in the future), and externalities (e.g. the costs and consequences borne by second-hand smoking). An additional justification might arise if the people are acting with “bounded rationality”, rather than in the well-informed way that traditional economic theory assumes. This approach, largely subsumed under the heading of Behavioural Economics, offers a broadened set of rationales for when governments may be justified to interfere with individual decisions in the area of health behaviours [
125].
(c)
Characteristics of respondents
Acceptability varied with the three characteristics of respondents that we focused upon: gender, age and whether or not the individual engaged in the behaviour that was the target of the intervention.
The finding that women are more likely to support intrusive (and often more effective) interventions is intriguing. Given women in OECD countries behave more healthily than men, at least in terms of tobacco and alcohol consumption (which is what the bulk of the reviewed studies focused on), this finding may again reflect our earlier observation that people are less resistant to interventions that target the behaviour of others. Women’s preference for stronger measures may also reflect more direct experience of the adverse consequences of certain unhealthy behaviours (e.g. from alcohol-related harm experienced by and coming from their male partners). In addition, women more frequently provide informal care than do men [
126], perhaps leading to a stronger preference for interventions that prevent the need for care.
We are uncertain why the acceptability of intrusive interventions increases with age. This may reflect a greater awareness of the burden of disease with age. Alternatively it may reflect a growth in trust of government intervention with age. Given older individuals are more likely to vote, the patterning of acceptability by age is likely to be of particular interest to politicians.
The finding that people engaging in unhealthier behaviours are more likely to reject policies that aim to restrict such behaviours is consistent with the world-view that individual preferences for public interventions are determined by people’s self-interest. (The self-interest “theory” has also been invoked in the study of the determinants of public support for other types of public policies, e.g. taxation or the welfare state [
127]. This may mean that either those individuals take the rational, utility-maximising decision to engage in, say, smoking and thus consider any external interference as utility-decreasing; or while the unhealthy behaviour results from the individual pursuit of self-interest, the individual may not have been in a position to carefully weigh the costs and benefits of its actions to take a rational enough decision.
Conversely, those not engaging in the unhealthy behaviour that is the focus of intervention appear more willing to advocate interventions that restrict the behaviour of others. That may again be compatible with the self-interest interpretation, if, say, the non-smokers advocate intrusive and effective interventions because they experience the adverse external effects (e.g. in terms of second-hand smoking) resulting from the behaviour and so want to reduce those.
Strengths and limitations of the current review
The strength of this review is its novelty, being the first systematic attempt to map the evidence on public acceptability of government interventions to change health-related behaviour. We have identified emerging themes and gaps in the research evidence and ways in which future studies can be strengthened. The review was limited in several ways. While the search was systematic, in keeping with the scoping nature of the review, we did not use formal methods to synthesise the results. The literature was heterogeneous in several salient dimensions. Different questions and response options were provided, which influenced responses [
37,
128]. Different data-collecting methods were used, which also influenced the attitudes expressed [
129]. The respondents varied in age, gender, socio-economic status, ethnicity and country of residence, and different methods were applied across countries in order to account for those factors in determining public attitudes. The context of the surveys varied in time, source of funding, political and policy contexts, all of which are likely to shape attitudes but not always in a way that was revealed in the individual studies or in our narrative synthesis. Of particular note is the provenance of the studies with the majority from the USA. Future reviews that focus on particular behaviours or types of interventions, or that select designs from a narrower range can allow some of these potential influences on public acceptability to be teased out.
Implications for research on public attitudes
Revealing the determinants of public preferences towards behaviour change policies is a challenging research endeavour. Based on the existing evidence we reviewed, there is considerable scope for more research in this complex and highly policy relevant domain. Existing literature on public attitudes towards interventions to change health-related behaviours seems largely uninfluenced by the large literatures on judgment and decision-making [
131] and moral judgments [
132]. By omitting reference to pertinent social and behavioural science literatures on judgment, policy-makers and others are provided with a partial view of public opinions and of the options available to policy makers to influence acceptability. Its inclusion will provide a more valid account of public acceptability of the more intrusive interventions, which often have the most potential to change behaviour to improve population health. It will also provide the theoretical basis for intervening to assess the extent to which public opinions are conditional upon information provided about the likely impact of interventions. Finally, it will also allow for an exploration of how framing of interventions, to align with a population’s dominant core values and beliefs, could be a tool for influencing acceptability.
The stability of public attitudes towards interventions to change health-related behaviour is little studied. When governments plan regulation that has the potential to reduce sales, campaigns are invariably launched by the affected companies including alcohol, tobacco and food and beverage manufacturers [
19,
133]. The malleability of public opinion by competing messages from governments and commercial companies warrants investigation.
There is particular scope for adding to the literature in the domains of physical activity and diet, where we found very few studies. From a methodological viewpoint, while we see potential for a more extensive analysis of existing, preferably international survey data, we see particular promise in the development of discrete choice experiments that – in contrast to most opinion poll data – confront respondents with trade-offs between different policy options, providing a more valid way of assessing acceptability.