Background
Despite the considerable effort to address the burden of non-communicable chronic diseases (NCDs) such as diabetes and cardiovascular disease, the prevalence of risk factors for NCDs, with the possible exception of tobacco use, have changed little over the past few decades and remain high [
1]. Overweight and obesity and their health consequences pose a particular challenge contributing 8.4% to the burden of disease in Australia [
1]. Although the majority of strategies designed to reduce the lifestyle-related risks leading to NCDs, including overweight/obesity, have been directed towards the individual behavior change, many health promotion advocates [
2,
3] and lead health agencies [
4,
5] have emphasized the important role of policy and environmental-level change for greater impact. Such actions include introducing fiscal interventions, such as taxes on unhealthy products and/or providing subsidies for healthier options; restricting advertising and promotions; reducing the availability of certain products; changing built environments to support physical activity; and mandatory reformulations of the food supply. Marteau et al. (2019) argue that these more upstream approaches, targeting the physical, economic, social and commercial systems driving behaviors, demonstrate the biggest health gains and should be prioritized by government [
6]. A recent Australian consensus statement from leading experts in chronic disease prevention has recommended 11 policy actions which should be prioritized based on previous evidence [
7]. For example, the report recommends actions such as a volumetric tax on alcohol as price increases through taxation for alcohol have been shown to have the largest effect in reducing alcohol-related morbidity and mortality. Greater investment in active travel infrastructure and a national physical activity plan is also prioritized as changes in urban design which support modal shift from motorized transport to public and active transport have been demonstrated to significantly increase walking [
7]. Other groups such as the World Health Organization have recommended similar interventions, which they argue are cost-effective ‘Best Buys’ for chronic disease prevention [
8]. However, calls to make these upstream changes through regulation and structural change can be met with community suspicion about the effectiveness and true purpose of such interventions [
9,
10] and reluctance on the part of policymakers due to concerns of public acceptability, amongst other considerations [
11,
12]. Yet the evidence base in public health strongly favors descriptive correlational studies, rather than studies examining intervention effectiveness, sustainability and public support – evidence which could assist policymakers to build a case for population-level interventions [
13].
According to previous research, the likelihood a policy or intervention will be implemented is the result of a range of intersecting contextual and structural factors [
11,
14,
15]. For example, Kingdon (2003) [
16] argues that policy action requires the alignment of three “streams”: a clear definition of the (health) problem; a proposal for policy change; and a favorable political context. Health promotion advocates and researchers need to be sensitive to these “policy windows” to capitalize on favorable conditions and implement policy change. Being able to detect such moments is challenging; combining close relationships between researchers and policymakers [
15,
17] with good evidence for the likely acceptability to the general public of proposed interventions may address political sensitivity to popular sentiment [
9,
18].
Monitoring trends in community opinion can help researchers, advocates and policymakers identify “policy windows” or critical opportunities to build their case for policy change. Analyzing these trends may also shed light on the social and political contextual factors which may influence public opinion. However, few studies have examined changes in opinion towards government intervention for prevention of lifestyle-related disease over time. The majority which do so are based on tobacco control surveys in North America and have identified trends such as increased support for government policies and interventions regulating access to, or sales of, tobacco [
19‐
23]. Similar trends have been found for public support for plain packaging of tobacco products in cohort studies in Australia [
24] and the UK [
25], removing point-of-sale tobacco advertising and displays in Canada [
26] and smoke-free bars among American young people [
27], demonstrating reasonable evidence for a causal effect of implementation on support for policy change.
In contrast to what appears to be reasonably uniform trends for tobacco control policy, published research on support for alcohol regulations is mixed. In Canada from 1996 to 2011 [
28] and Ireland from 2002 to 2010, [
29] support has been shown to depend on the policy mechanism and the degree to which policies on price and availability were already in place. Evidence from Australia on attitudes towards alcohol control policies between 1995 and 2010 showed increasing support beginning in 2004 for policies which restrict availability and accessibility of alcohol [
30]. However another study encompassing a later period (2001–2016) [
31] showed a drop of 10% in support between 2013 and 2016 following the introduction of restrictions for late night service of alcohol in the capital city of one state (Sydney, New South Wales). Other research suggests that advocacy campaigns may be important interventions for increasing public acceptability of policy changes to reduce risk factors such as sugary drink consumption [
32].
Together, these studies have shown that there is complexity in how the community perceives whether government intervention for prevention is appropriate, with community perceptions influenced by degree of implementation in addition to strategy type, target behavior and demographic group, as noted with cross-sectional studies [
10,
33,
34]. However, to our knowledge, no published study has reported on changes over time in perceptions about the role of government in promoting health and preventing NCDs more broadly or whether those changes differ by demographic subpopulation. This study will draw on the data collected in the national AUStralian Perceptions Of Prevention Survey (“AUSPOPS”) over two time periods (2016 and 2018) to examine change and modifiers of change in general attitudes to government intervention in addition to specific interventions.
Discussion
Despite the short intervening period, our two surveys of community perceptions showed significant change both in terms of level of support for government intervention, as well as the degree of polarization of opinions. We observed differences in support for intervention in general and for specific interventions, including for interventions already in place (e.g., bans on smoking in cars with children) and those yet to be implemented (e.g., sugar tax). Our analyses also showed that change was not always uniform across demographic subgroups, with the degree or direction of change varying by gender, age and education for a number of measures. The analyses point to a multilayered picture of increasing appetite for government leadership on prevention and greater disparity between subpopulations in opinion. These findings are discussed in greater detail below in the context of previous research along with implications for public health policy practice and research.
One of the strongest effects was the almost 14% absolute change from 46% in 2016 to 60% in 2018 in the proportion of people saying that government has a large or very large role in maintaining people’s health. While endorsement of a larger role for employers and private health insurers also increased significantly, the change was considerably smaller. In keeping with this result, support for three specific interventions also increased from 2016 to 2018, as did agreement that the government needs laws to stop people from harming themselves and that, in general, there was not enough regulation and policies in place to help people be healthy. Thus, there appears to be a gap between the role that the community perceives the government should be taking in prevention, and the perception that that role is being carried out. While the mechanism producing such changes is difficult to determine from our data, there is current evidence in the COVID-19 [
58] other health literature [
59] of strong public acceptance of government intervention among Australians. It also appears that the increasing and majority support for government to lead on prevention seen in our study is apparently compatible with strong (but stable) endorsement of personal responsibility for health, which was unchanged between 2016 and 2018. Concomitantly, there is little evidence of community concerns of a “nanny state”, confirming a previous analysis of the 2016 AUSPOPS data [
10].
Looking across support for specific interventions, no clear pattern emerged in the types of interventions where change was observed. There was, however, continuing strong support for greater restrictions on advertising unhealthy foods to children; support remained favorable (58%) and stable from 2016 to 2018 despite some subpopulations (i.e., women, people aged younger than 35, and those with a university education) increasingly feeling that the government had not gone far enough. In Australia, restrictions on unhealthy food advertising are currently under self-regulation [
60,
61], a policy which has been criticized by health advocates and researchers for being ineffective in reducing children’s exposure over the past decade [
62,
63]. As government regulation is yet to be implemented in Australia and is enjoying a surge of support in some groups, a policy change is likely to be met with public endorsement, thereby satisfying the political stream of Kingdon’s policy window [
16]. However, as with many other preventive health interventions [
33,
34], women are more (and increasingly) supportive of this intervention compared with men, and therefore, any accompanying advocacy would need to address the differential endorsement within subpopulations and address the concerns of groups whose support is on the wane. One solution may be to communicate the effectiveness of such policy interventions which has been shown to increase support in the past [
64]. Importantly, it is not clear whether this strategy appeals to some subpopulations over others and would be a key area to explore in future research.
The decrease between the two surveys in the proportion feeling that the government had not gone far enough for a couple of the interventions may reflect changes in the regulatory environment in this period. For example, the reduction between 2016 and 2018 in the proportion saying that compulsory immunization at school entry had not gone far enough could be a result of the introduction of amendments to the “No Jab No Pay” policy in 2016. The legislation principally restricts eligibility to some family and child care tax benefits if children are not immunized, and in 2016 the exemption for conscientious objectors was removed [
65]. Survey research conducted post-implementation demonstrated the amended legislation drew high levels of support (82%) among parents of children age younger than 5 years and immunization coverage was only 1% below the target of 95% [
65]. Our findings may therefore indicate that the community feels current action is sufficient for achieving this particular health objective.
Our survey findings of significant support for regulation on salt in processed foods was consistent with other Australian research [
66]. However, although the majority still felt in 2018 that the government had not gone far enough in setting limits on salt in processed foods, there had been a significant and uniform (across demographic subgroups) drop in support for change. Reformulation of the food supply is seen as a best buy for non-communicable disease prevention [
67]. However, while the UK has seen success from implementation of a national salt reduction strategy which includes voluntary reformulation [
68], Australia’s voluntary thresholds and actions [
69] have had minimal impact [
70] due to a lack of strong government leadership, targets and timelines, as well as lack of accountability for industry inaction [
71]. The change in public support in our sample may signal a need for public health advocates to maintain the visibility of issues yet to be translated into policy, such as salt reduction strategies, to ensure the potential for action is not lost through a decline in public support.
Finally, a novel aspect of our analysis was to examine not only shifts in central tendency but also variation over time. While we saw only one significant change in the mean score with the different conceptualizations of government intervention in general, all scores exhibited greater polarization in opinion. Thus, while community support for specific interventions continues to trend upwards for some and downwards for others, positions on the spectrum of different conceptualizations of how government should act out its role in general is balanced but becoming more strongly held in both directions. Our results confirm the capacity for dissonance between positions on intervention in general and for specific interventions, as hypothesized in previous AUSPOPS research [
10]. More broadly, research on the question of whether social opinion is becoming more polarized due to the reinforcing nature of personalized news and information, especially through social media, has shown mixed results [
72‐
74]. While we cannot test this assumption with our data, future research may benefit from gauging the extent of respondents’ use of social media and other information sources to investigate potential relationship between engagement with more personalized news streams and strength of opinion. Such analyses could also further inform policymakers how best to reach different subpopulations with health information according to issue and the attitude held.
Strengths and limitations
This study has a number of strengths. First, the samples from both years are nationally representative, were sampled in a comparable manner and are of sufficient size to allow for subgroup analyses. Second, the survey includes questions which address both positions on intervention generally as well as for specific interventions and used the same wording for the two comparison years. Limitations include some small demographic differences between the two samples on age, employment and holding private insurance, and a lack of information on non-responders meant we could not compare those who did and did not respond to the survey. However, the analyses were weighted to the Australian population and therefore are reflective of the general population. Response rates were low but typical of telephone surveys [
75,
76] and accounted for the changes in phone ownership towards greater mobile phone coverage [
36]. It was also possible that respondents did not confine their interpretation of “health” (in questions E1, E3 and E5) to NCD prevention which may introduce a level of error into our conclusions. However, as stated in the Methods, questions were cognitively tested prior to administration and found to be reliable in terms of intended meaning and therefore are likely to reflect community attitudes on the prevention of lifestyle-related chronic disease. Finally, the changes noted here may not be indicative of longer-lasting shifts in opinions and may be influenced dynamically by current events. However, there was no evidence of any relevant and major controversial issues present during the survey periods.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.