Background
On 6th April 2018, the United Kingdom (UK) Government introduced a Soft Drinks Industry Levy (SDIL) commonly referred to as the “sugar tax”. The levy was intended to reduce sugar consumption, primarily through reformulation by soft drinks manufacturers to reduce sugar content and avoid paying the levy [
1,
2]. Excess consumption of free sugars is a contributory factor in the rising prevalence of non-communicable diseases (NCDs), both as a direct cause [
3,
4] and through the contribution to energy imbalance resulting in obesity [
5]. In 2015 the World Health Organisation (WHO) issued guidance on sugars consumption for adults and children, recommending reducing intake of free sugars to less than 10%, and ideally less than 5%, of total energy intake [
6]. In the lead up to UK Government’s introduction of the SDIL, there was much policy debate about how to achieve this goal [
7,
8], with evidence suggesting the need for upstream policy intervention [
9‐
11].
Products containing tobacco and alcohol have long been subjected to taxation, which was traditionally seen as a way of raising revenue for public spending [
12]. More recently, there is increased emphasis on taxation of these products to promote population health, given the inverse relationship between price and consumption [
13‐
16]. However, the role of price increases through taxation in food and nutrition policy is not clear-cut [
12,
17,
18]. Sassi et al. note that, in the case of foods, the value of using taxes depends upon their design and the context in which they are applied [
12]. Specifically, if people are aware that a product is taxed for public health reasons, rather than to raise revenue, they may be more likely to change their consumption [
18]. A number of governments around the world have adopted taxation of selected energy-dense foods and drinks [
19]. A recent policy analysis describing 13 international case studies concluded that taxation seemed to have the desired effects on prices and consumption of energy-dense products [
19].
The successful introduction of upstream food and nutrition policies is a highly political enterprise with multiple vested interests [
20]. Stakeholders include politicians, the commercial sector, public health professionals, academics, non-government organisations, government advisors, journalists, public figures such as celebrities, and grassroots organisations [
21]. The example of the Danish ‘fat tax’, implemented in 2011 and abolished after only 15 months despite recent evidence of its success [
22], highlights tensions between stakeholders. Bødker et al. note that active industry lobbying and judicial actions undermined policy support [
23]. This illustrates the highly politicised nature of introducing new policies when the evidence base is limited and policies are opposed by those with vested corporate interests [
20]. Research suggests that one reason governments do not pursue upstream approaches to food and nutrition policy is the power and influence of the food industry [
20,
24‐
26]. Understanding industry influence in such public policy debates is challenging given the complex network of stakeholders involved. As Smith et al. note, understanding this network of stakeholders, their relationships, and interactions is necessary for elucidating advocacy strategies to counter industry and protect public health [
27].
One theory of policy change or stability, the Advocacy Coalition Framework, suggests that policy subsystems are formed around competing advocacy coalitions, which are based on shared ideological belief systems. Changes in core policy preferences by political actors result in changes in the balance of coalitions and hence policy shifts [
28,
29]. Leifeld suggests that the articulation of policy beliefs by interest groups in the media and other arenas, in the form of discourses, reveals their policy preferences and encourages other stakeholders in the policy debate to support them or reveal their opposition [
30]. Mapping of such agreement or disagreement in discursive forums can produce representations of advocacy coalitions and policy networks [
30]. Thus, network analysis can be used to explore the complex interactions and alliances that stakeholders form in attempting to influence government policy [
21,
31‐
33].
Social network analysis has been used in tobacco control policy debates to highlight the polarisation of opposing coalitions and draw attention to the complex processes of consensus-seeking, alliance-building, and strategic action, which are integral to the development of policy [
33]. In the case of Minimum Unit Pricing (MUP) for alcohol, discourse network analysis (DNA) [
34,
35], a combination of category-based content analysis and social network analysis, has been used to provide insights into the formation of discourse coalitions and cast light on the complexity of alignments between stakeholders engaged in the MUP debate as cited in UK newspapers [
36]. Whilst Hilton et al. found that both proponents and opponents of the SDIL actively engaged with the news media to promote framings that would advance their interests [
37], there is little evidence on how the complex network of coalitions and alliances of stakeholders formed and changed during the SDIL policy debate, and how those networks may have impacted the Conservative Party’s shift in policy position and adoption of the regulation.
This study uses DNA to offer the first visual representation of the network of stakeholders and advocacy coalitions apparent in UK newspaper content in the lead-up to, and following, the announcement of the SDIL in the UK. Specifically we aim to: (i) determine the membership of coalitions active in the debate and how these developed over the period of policy debate (May 2015 to November 2016); (ii) explore alliances and cleavages across different sectors/interest groups; and (iii) generate network insights on industry lobbying, campaigning, and policy advocacy.
Discussion
This study highlights the complexity of the network of stakeholders and their involvement in the debate on sugar tax and the SDIL. Our analysis identified the involvement of a large number of stakeholders, and some apparent divisions within the food and drink industry and commonalities between some industry segments and public health advocates. The coalitions changed over time, with peaks in bipolarisation coinciding with publication of evidence on the health harms of excess sugar consumption and policy announcements. In the first time-period, polarisation appears to arise primarily from ideological positioning (whether or not taxation of any kind is an appropriate measure), whereas later it comes from contradictory positions on whether or not SSBs are the appropriate target of policy regulation. The impact of the unexpected policy announcement in early 2016 may have contributed to the increased bipolarisation of the network and alignment within both supportive and sceptical coalitions, as stakeholders united to strengthen their positions in response to a specific policy.
Corporate social responsibility strategies represent an important mechanism by which controversial, or potentially socially harmful, industries seek to mitigate the level of controversy arising from their business activities [
58]. Fooks et al. suggest that corporate social responsibility activities allow corporate stakeholders, such as those in the tobacco industry, to justify ethically problematic market actions that promote economic interests over public health concerns [
59]. One example observed in the alcohol industry is the dissemination of health information to the public while misrepresenting the evidence of health harms associated with their products [
60,
61]. In the case of the SSB industry, manufacturers have sought to emphasise the importance of physical activity over calorie restriction in dealing with obesity; as exemplified by Coca-Cola’s investment in the Global Energy Balance Network [
62]. Our analysis of discourse networks suggests a more complex interplay between protecting profitability and corporate social responsibility strategies in the case of the SDIL debate. Different sectors of the food and drink industry present different views, resulting in unanticipated commonality between some industry sectors and public health campaigners, and cleavages between industry segments. Retailers and retail associations, manufacturers and restaurants were not entirely aligned in their media statements. Parts of the retail sector were situated outside the sceptical coalition, including Sainsbury’s, the British Retail Consortium, and Abokado (a retailer and manufacturer with a mission to “lead happier and healthier lives”), reinforcing their position as being “part of the solution” [
63]. In contrast, soft drinks manufacturers appeared in the sceptical coalition alongside think tanks and economic analysts, drawn together by similar statements characterising the policy as unfair, an inappropriate intervention in the market and too simplistic. This set of arguments is familiar from alcohol industry opposition to the Minimum Unit Pricing policy [
64], and other so-called unhealthy commodity industries [
65].
The prominence of public health advocates and campaigner nodes in the supportive coalition perhaps reflects intense lobbying by these stakeholders prior to the announcement of the policy. A recent review of public health advocacy to reduce health inequalities revealed inconsistencies in framing of policy and a lack of coherence between theory and action, which resulted in multiple barriers to consistent public health advocacy [
66]. In contrast, the networks revealed by this study of the public debate on the SDIL suggest unity among public health advocates on the scale of the problem and the importance of regulatory action. Two factors that may have facilitated alliances in support of the SDIL may have been that, firstly, the levy was designed to both encourage industry reformulation and reduce individual consumption, and secondly, the levy targeted a commodity that can be linked directly to health harms with no nutritional benefit.
Conversely, we suggest that the food and drink industry is inconsistent in their lobbying tactics. In studies of other industries’ efforts to influence policy, stakeholders are portrayed as employing consistently effective tactics to oppose upstream regulation, including the use of a playbook of succinct, well-drilled messages delivered by central spokespeople [
61,
65,
67,
68]. In contrast, this study demonstrates that the structure of the sceptical coalition pre January 2016 takes the form of two sub-coalitions representing industry sub-sectors; one emphasising public health framing of the debate (supermarkets and retailers), and the other focusing on ideological arguments (food and drink industry, politicians and think tanks). Post January 2016, the sceptical coalition appears to be more aligned. It is dominated by soft drinks manufacturers, their representatives and think tanks, with Coca-Cola and the British Soft Drinks Association at its heart. Representatives of retailers and restaurants either become peripheral to the coalition or move to the supportive coalition, along with most politicians. Stakeholders in this more cohesive sceptical coalition are united by concepts criticising SSB taxation as a regressive, unfair, punitive tax on the soft drinks industry and questioning its likely effectiveness as a policy measure, while reinforcing corporate social responsibility rhetoric. This suggests that soft drinks manufacturers were less coordinated before the SDIL announcement, perhaps believing that existing voluntary agreements with government, in the form of the Public Health Responsibility Deal [
69], would protect them from further regulation. In other words, the industry may have been caught off-guard. This suggestion is supported by the position of the UK Government and the Conservative party in the network in the early time-period, when their statements were aligned with industry representatives in the sceptics’ coalition.
Kingdon defines policy entrepreneurs as
“people who are willing to invest their resources in pushing their pet proposals or problems” [
70]. As such they can be instrumental in setting the policy agenda, highlighting solutions to problems, getting the attention of policy makers and thus facilitating policy change [
71]. More recently Pepin-Neff and Caporale have highlighted the importance of high-profile individuals in bringing about political change [
72]. The size and central position of Jamie Oliver’s node in the supportive coalition is suggestive of his role as such a celebrity policy entrepreneur in the public debate on sugar tax and highlights the increasing sophistication and importance of public health activists. Sustain, an alliance of organisations advocating for food and agriculture policies and practices that enhance health and welfare, published an analysis entitled “How the sugary drinks tax was won: 10 lessons for committed campaigners” [
73]. They also highlight the importance of working with high-profile advocates to take public health campaigns to a new level of recognition and impact, as well as facilitating alliances of medical and public health professionals, academics, journalists and politicians [
73]. Our findings support this recommendation, showing Jamie Oliver occupying a dominant central position at the heart of the supportive coalition, particularly in the period before the SDIL policy announcement.
Although our use of DNA provides a number of novel insights into the network of stakeholders active in the sugar debate, the study has some limitations. The data was limited to the statements attributed to stakeholders in UK print media. Newspaper debates are only one arena among several in which political discourse unfolds, and this research cannot comment on the parliamentary or judicial arenas, or any discussions that occur behind closed doors. However, understanding public debates in the media arena offers a useful “door opener to the backstage of politics”, as Wodak and Meyer argue [
74]. Additionally, restricting data to quotes or comments cited in newspapers that were directly attributable to stakeholders minimised any editorial impact. The application of a strength of tie threshold value of ≥0.4 removed a number of inter-stakeholder associations and allowed for a focus on only the most robust ties in the network. A final limitation arose from restricting the research to print media, which may have excluded stakeholders who operate exclusively in social media. For example, the campaign group People Against Sugar Tax did not appear in this analysis. However, the researchers’ shared experience of media research suggests the number of such organisations is small.
Conclusion
In conclusion, polarisation of stakeholders visible in the media debate on SDIL arose from differences in ideology, focus on a specific policy and statements on the weight of evidence. The food and drink industry stakeholders appeared to be caught off-guard by the specific SDIL policy announcement and seemed less cohesive as a coalition than might have been anticipated based on research on other manufacturers of unhealthy commodities. Formation of advocacy coalitions in support of upstream regulation seems dependent on alignment around a clear ideology and policy objectives. A vocal celebrity policy entrepreneur could provide an important locus for this, in the way that Jamie Oliver appeared to do in this example of sugar tax and the SDIL.
Use of DNA methods offers the first visual map of the SDIL network and allows exploration of a complex set of relationships involved in framing public health policies. Further research is needed to explore what motivated the complex dynamic in this contested policy debate, potentially through interviews with the stakeholders involved. Further DNA analyses may be conducted around broader public health problems, not focussed on a specific policy measure, but instead using a health problem as a starting point and analysing actors’ convergence on a set a policies over time to further our understandings of health policy making.
Acknowledgements
Many thanks to colleagues at the MRC/CSO Social and Public Health Sciences Unit, University of Glasgow for helpful comments on early drafts. Particularly Mr. Chris Patterson, Dr. Peter Craig, Dr. Michael Green, Dr. Hannah Hale and Dr. Lynne Rush.