Pre-emption strategies to block taxes on sugar-sweetened beverages: A framing analysis of Facebook advertising in support of Washington state initiative-1634

Abstract

In 2018, the sugar-sweetened beverage (SSB) industry introduced a ballot measure (I-1634) in Washington State of the United States to prevent further local taxes on groceries. The measure, which passed, is emblematic of new pre-emptive legislative strategies by the SSB industry to block soda taxes and conceal those strategies under the guise of preventing burdensome ‘grocery taxes’. This paper uses qualitative framing analysis to examine a public archive of 1218 Facebook advertisements to understand how I-1634 proponents shaped public discourse and engaged in misinformation efforts online during the lead up to the passage of I-1634. Coding strategies identified 7 compelling and inter-related framing strategies used by the campaign. These included strategies that misinformed the public about the threat of grocery taxation and the economic impacts it would have on the region. Strategies to conceal the true intent of the ballot measure and the sponsors of the campaign were aided by Facebook’s advertising platform, which does not moderate misinformation in advertising and allows advertisers to conceal their sponsors. We urge public health researchers and advocates to pay more attention to how Facebook and other social media platforms can be used by industries to target voters, misinform publics, and misconstrue industry support.

Background

Excise taxes or levies on sugar-sweetened beverages (SSBs) – understood as any beverage with a form of added sugar (CDC, 2021) – are an effective public health measure to lower consumption of SSBs at national, regional, and local levels (Teng et al., 2019). SSBs are a leading source of added sugar intake globally (World Health Organization, 2017) and their consumption is associated with the development of cancers (Chazelas et al., 2019), overall mortality (Collin et al., 2019), and cardiovascular diseases (Malik et al., 2019). As of October 2020, over 40 countries have implemented SSB taxes, including Mexico, Bermuda, Chile, Finland, India, Latvia, the United Kingdom, and Sri Lanka (Falbe, 2019). In the United States (US) city of Berkeley, California, SSB consumption dropped 21% four months after a local-level $0.01-per-ounce excise tax was introduced, compared to a 4% increase in surrounding cities without similar taxes (Falbe et al., 2016). In Barbados, average weekly sales of SSBs dropped 4.3% after a 10% value-based tax, while sales of other non-carbonated beverages such as water increased 7.5% (Alvarado et al., 2019). Chile observed a 21.6% decrease in monthly SSB purchasing after introducing a sliding scale tax (Nakamura et al., 2018). In addition to lowering rates of SSB consumption, SSB taxes are cost-effective. In the US, it is estimated that SSB taxes would generate up to $142 billion in revenue, prevent 850,000 cases of cardiovascular diseases and 269,000 cases of diabetes, and save up to $53.2 billion in healthcare costs arising from SSB consumption (Lee et al., 2020).

The beverage industry has strongly opposed SSB taxation, often leveraging enormous resources to oppose municipal- and state-level measures. In Philadelphia, Pennsylvania, a proposed beverage tax was met with a $1.5 million opposition campaign led by the American Beverage Association (ABA) (McCrystal, 2020). Across four cities considering a SSB tax in 2016, lobbying figures by the ABA and prominent SSB companies exceeded $37 million (Belluz, 2016). Leaked industry documents reveal a coordinated attempt to block the introduction of new taxes, using strategies such as coalition building, lobbying policymakers, and promoting messages that taxes are discriminatory (Pfister, 2016). The SSB industry has a lengthy history in undermining public health measures to reduce the consumption of their products. In 2018, the Canadian Beverage Association (CBA) lobbied to prevent a policy measure that restricted the content of advertising to children (Weeks, 2018). Studies funded by the SSB industry are less likely to report obesity from SSB consumption compared to non-industry funded studies, thus causing controversy about the contributing causes of obesity and how to design obesity-related public health initiatives or interventions (Schillinger et al., 2016). A 2015 New York Times report describes how Coca-Cola has spent millions of funding scientists to conduct research that shifts blame for obesity onto non-diet related causes such as physical inactivity (O’Connor, 2015).

The SSB industry is increasingly utilising concerning strategies of pre-emption to block the introduction of new SSB taxes (Crosbie et al., 2021). Pre-emption occurs ‘when a higher level of government limits the authority of a lower level to enact new policies’ (Crosbie et al., 2019). In other words, pre-emption policy is introduced at the state level to prohibit cities from introducing local SSB taxes. The use of pre-emptive strategies by the SSB industry mimics that of the tobacco industry, which starting in the 1980s achieved protections against tobacco control methods such as advertising restrictions and banning indoor smoking, some of which still exist (Crosbie & Schmidt, 2020). In the US, pre-emption SSB tax-related laws were first introduced in the states of Arizona and Michigan (O’Connor & Sanger-Katz, 2018). In 2018, California, where four municipal SSB taxes are enacted, pre-emptive bans on new SSB taxes were introduced in exchange for the ABA and other groups removing another ballot measure that would have implemented barriers on municipalities introducing any new taxes (Dewey, 2018). Also in 2018, pre-emptive bans on the introduction of any new groceries taxes, largely considered to be directed at SSBs, were voted on in Washington state via ballot measure, and ultimately passed (Camden, 2018).

The Washington state ballot measure – known as Initiative 1634 (I-1634) – was put forward by a coalition called ‘Yes! To Affordable Groceries’ which received over $20 million in funding from the ABA, Coca-Cola, PepsiCo, Keurig Dr Pepper, and Redbull North America (O’Sullivan, 2018). The measure was described to the public as preventing the introduction of new grocery taxes and closing a ‘loophole’ that allows local governments to implement taxes (Brand, 2018; Romano, 2018a). However, articles from reputable news outlets, such as the New York Times, argued that SSB tax opposition motivated the measure (Jacobs, 2018). Critics of the campaign argued that the coalition deliberately misled voters about the intentions of the campaign. Laura MacCleery of the Centre for Science in the Public Interest summarised: ‘They’re calling it a grocery tax measure when nobody in the public thinks of soda when they think of groceries’ (Bratskeir, 2018). Additionally – there were no plans to tax grocery items and Washington law already prohibited taxes on essential groceries (Jacobs, 2018; Romano, 2018b) The primary funders of the measure concealed the industry sponsors, endeavouring to portray the initiative as locally-led (Evich, 2018).

The passage of I-1634 and other pre-emptive bans on SSB taxes poses a threat to the adoption and implementation of an evidence-based public health tool in areas where a tax currently does not exist. The highly coordinated SSB industry is likely to learn from I-1634, and introduce similar measures in other states (Crosbie et al., 2019). Crosbie and colleagues examined tactics by the SSB industry to achieve state-level pre-emption by reviewing website content and other public materials, documenting such risk (Crosbie et al., 2021). Pre-emption as a strategy relies on fairly intensive messaging efforts, and, as we argue here, framing strategies to justify these measures to the public using moral, economic, and sociocultural claims and symbols. Research is needed to better understand the communication tactics used in advertising campaigns such as I-1634 to inform advocacy efforts to prevent future bans. The I-1634 campaign also relied heavily on social media platform advertising to project and control these narratives. Industries have capitalised on minimal safeguards on social media platforms regarding misinformation or misleading statements to heavily influence public opinion. Following calls for public health researchers to pay more attention to how digital technologies shape determinants of health, we leverage a recent public archive of Facebook and Instagram advertisements to study how the ‘Yes! To Affordable Groceries’ campaign shaped public discourse and engaged in misinformation efforts online during the lead up to the passage of I-1634 (Azzopardi-Muscat & Sørensen, 2019; Baum et al., 2014; University of Oslo, 2019). As Facebook is both a primary means of connecting to information and the internet for many in low-income countries (Nothias, 2020) and a primary, though unreliable, source of news for many users worldwide (Boukes, 2019), particular attention must be paid to how food and beverage industries are leveraging the platform to misinform consumers and influence regulatory processes.

Materials and methods

Data collection

Paid advertisements by the ‘Yes! To Affordable Groceries’ coalition Facebook page were retrieved through the Facebook Ad Library, resulting in the identification of 1218 paid advertisements (Facebook, 2020). The first author manually recorded the unique advertisement link and copied the URLs into a spreadsheet. Then, the first author developed an automatic internet scraper using Data-Miner to record advertisement details from the previously collected URLs. The advertisement details collected included: advertisement identification number, date range advertisement ran, low and high spend estimate, low and high viewer impressions estimate, advertisement description, and advertisement image or video. Advertisement data was then uploaded to Dedoose – a qualitative and mixed methods research platform – for analysis. To facilitate coding, only unique advertisement media and descriptions were uploaded. Each unique media file – image and video – was assigned a file number, whereas advertisement descriptions were already uniquely identifiable. Each unique descriptor would be coded on Dedoose and subsequent coding would be updated to applicable advertisements in the main project file.

Analysis

The first and second authors independently reviewed advertisement data to develop a qualitative coding framework and analysis plan. After review, the authors met and agreed to use framing analysis – a qualitative method to appraise how certain information is conveyed to relay particular meanings, to offer narratives about cause and effect, and to appeal to and coalesce political groups – to assess the ‘Yes! To Affordable Groceries’ advertisements (Holton et al., 2014; Koopmans & Statham, 1999). While there is wide and divergent use of framing analysis across a number of different fields, here we follow the work of communications and policy scholars who identify framing as amalgamating a complex set of symbols, moral claims, information selection, categorisation, narration, and story-telling (Gamson, 1992; van Hulst & Yanow, 2016). Frames are also understood to be politically sensitive and mediate a relationship between stakeholder identities and public communications (van Hulst & Yanow, 2016).

An iterative process was used to categorise and analyse framing strategies in the I-1634 campaign. The first round of open coding was used to develop initial framing categories. These were then refined through discussion, cross-referencing with the data and existing literature, and analytic note-taking. This process yielded seven categories of framing strategies, which focused on both the narrative and strategic approaches of campaign messages: (1) grocery taxes negatively impacting certain groups; (2) I-1634 supported by specific community groups; (3) economic impacts of grocery taxes; (4) legislative threat; (5) misleading statements that stress the hypothetical; (6) adopting progressive messaging; and (7) standing up for the greater good. In addition to these frames, advertisements were coded for: the demographics of people portrayed in visual and text data from advertisements (age, sex, and race); information about groups most impacted by grocery taxes (agricultural industry, customers, families, fixed income community, low and middle-income households, regional businesses, seniors, working people, employees, and people of colour); organisations mentioned as supporting I-1634 by type (news organisations, trade groups, restaurants, stores, and other); and the food items mentioned or shown in advertisements by type. The first author coded each advertisement and the second author reviewed 20% of advertisements to ensure consistency. Disagreements in coding were resolved through discussion and recoding was conducted as necessary.

Results

Advertisement spend, reach, and demographics

The ‘Yes! To Affordable Groceries’ campaign spent between $119,400 and $594,502 on advertisements, receiving between 10,547,000 and 29,362,000 impressions. The campaign relied on personalised narratives from people representing diverse coalitions across Washington State. Given the divisive politics of other SSB taxation campaigns in the US, and previous industry efforts to frame taxation as classist or racially discriminatory, we paid close attention to how the campaign appealed to specific identity groups. Visual elements from campaigns frequently pictured individuals or small groups endorsing the campaigns. Among those depicted by the advertisements, 75.7% (n = 1221) were adults, 15% (n = 241) were children, and 9.3% (n = 150) were seniors; 55.4% (n = 893) were females and 44.6% (n = 719) were males. The majority of persons shown were White (66.1%, n = 1066), followed by Black or African American (14.8%, n = 238), Asian (7.5%, n = 121), Hispanic or Latino (5.8%, n = 94), South Asian descent (4.7%, n = 75), and American Indian or Alaska Native (1.1%, n = 18).

Grocery types mentioned and shown

Campaign advertisements also frequently featured imagery and mentions of different foods. Approximately 19.1% of advertisements mentioned or displayed SSBs (n = 233), and approximately three times as many advertisements featured other types of foods (n = 639) such as meats and proteins (n = 210), dairy products (n = 136), vegetables and fruits (n = 128), and general or unspecified grocery and food items (n = 165).

Negative impacts on specific groups

The first framing theme in the campaign emphasised negative impacts from potential ‘grocery taxes’ on specific demographic groups in the state. Advertisements routinely focused on unique threats to different population groups. These groupings often focused on categories of ‘everyday people’ who constituted large sectors of the state population, and narratives emphasised the ways taxes would make life even harder for groups considered struggling, hard-working, and deserving. These groups included: families (n = 616), businesses (n = 600), working people (n = 203), customers (n = 144), agricultural communities (n = 141), seniors (n = 130), those on fixed incomes (n = 120), low and middle-income households (n = 76), employees (n = 42), and regular people (n = 21). All group names were retrieved directly from advertisements. Table 1 shares illustrative quotes used to describe each group. Many advertisements made appeals across these groups as well, as can be seen in some of the quotations below.

Table 1.

Impacted groups – excerpts.

Descriptor	Excerpt
Families	‘New taxes on groceries like meat or dairy would make it more challenging and expensive for me to provide my kids with quality and nutritious meals. We need to protect working families, small businesses and the fixed income community from harmful taxes on our groceries!’
Businesses	‘It’s already expensive enough to own and manage a business in Washington. Additional taxes on groceries would just drive customers away, lead to job losses and cut sales for small businesses operating on thin margins’
Working people	‘A grocery tax would just put the heaviest burden on the people who can least afford it: seniors, working people, and those living on a fixed income’
Customers	‘Any other tax increases, especially on groceries, would not only affect our costs of certain items, but would increase the prices that the people in our community pay every day for groceries’
Agriculture	‘Agriculture is a major part of our state’s economy and an important job creator. But family farmers are already operating on thin margins, and local taxes on food will have a significant impact on how our farms operate, as those costs will make it more expensive to produce our goods’
Seniors	‘I am 76 years old and live on a fixed income. My doctor says I need lean protein to keep myself going strong for another 20 years. Taxes on groceries directly impact the quality of the life I deserve’
Fixed income	‘For people living on a fixed income, from hardworking parents to seniors, grocery taxes can make the necessities of life unaffordable’
Low and middle-income households	‘We punish all city residents especially those with low incomes with the most regressive tax structure of any city in the state, according to a recent report by the Economic Opportunity Institute’
Employees	‘We have 39,500 farms and ranches across Washington that provide 160,000 jobs in our state. New taxes on everyday groceries would threaten the jobs of the people who work these farms and ranches’
Regular people	‘We keep choosing to tax things that make it harder and harder for regular people to get by and keep dinner on the table’

Community group endorsements

Campaign advertisements strove to demonstrate broad-based community support and endorsements from prominent groups, organisations, and businesses operating in Washington state. Organisations mentioned as supporters or having released content in favour of the ‘Yes! To Affordable Groceries’ campaign were trade or advocacy groups (n = 36), restaurants or cafes (n = 19), news organisations (n = 12), stores (n = 8), and other (n = 3). Though these groups and organisations were relatively small in number (totaling 78 unique entities), they are mentioned 1079 times in the advertisements. The organisations varied in their location, with certain advocacy or trade groups being national, regional, or local. A list of organisations identified are in Appendix.

Legislative threat

A third framing theme in the advertisements was that of potential legislation, and the reasons to introduce such legislation (n = 234). Advertisements portrayed grocery taxes as an imminent threat outside of consumer and citizen control. Narratives included: portraying SSB taxes as an excuse to introduce other taxes; invoking irresponsible spending by politicians; discussions of legislative loopholes; portrayals of politicians trying to fill budget gaps; calling legislation a deceptive public health measure; and referring to politicians imposing their personal will upon the public. Table 2 provides illustrative quotes for advertisements under this framing theme.

Table 2.

Legislation threat – excerpts.

Descriptor	Excerpts
Excuse to impose other taxes	‘Seattle’s beverage tax was just the beginning. With a loophole in the law, local governments can impose taxes on everyday groceries, including meat and poultry’
Exploiting legislative loopholes	‘I-1634 will close a loophole in Washington law and will prohibit politicians from taxing our groceries. Recently, Seattle exploited this loophole to pass a huge new tax on everyday beverages’
Irresponsible spending by politicians	‘Enough is enough. Politicians are nickel and diming us at every turn with new taxes. I-1634 will prohibit local politicians from taxing groceries in Washington’
Deceptive public health measure	‘There is no public health benefit to making people a dollar poorer. Rather, regressive taxes pose a threat to public health among our most vulnerable citizens’
Politicians imposing personal will	‘People have the right to choose what items go in their grocery carts. Don’t let local politicians impose their personal will on Washingtonians’
Filling budget gaps	‘There are other ways for governments to fill budget gaps besides reaching into the pockets of hardworking families and neighbourhood businesses’

Emphasising negative economic impacts

Another prominent framing theme portrayed the negative economic impacts of potential grocery taxes on businesses (n = 179). Campaigns invoked significant economic threats if such legislation should be implemented. These threats included: the loss of jobs; businesses closing; loss of income; higher business overhead and costs; customers and businesses going to other states; business margins; and the dwindling size of food industry due to economic struggles. Table 3 provides illustrative quotes featured in the advertisements.

Table 3.

Negative economic impacts – excerpts.

Descriptor	Excerpts
Loss of jobs	‘Many workers are involved in the production and distribution of our groceries. When we target their products, we target their jobs, jobs with good pay and benefits’
Businesses closing	‘We’ve only been open a month and already, we’ve been welcomed into this community. Being the only place you can buy fresh baked pita is one thing we do that gives our community something special they can’t get at any box-chain. A tax on grocery items could threaten places like ours’
Loss of income	‘Excessive taxes have real consequences felt by every member of the community. From higher prices to lost income, taxes just make it more difficult to make a living’
Higher overhead and costs	‘Agriculture is a major part of our state’s economy and an important job creator. But family farmers are already operating on thin margins, and local taxes on food will have a significant impact on how our farms operate, as those costs will make it more expensive to produce our goods’
customers and businesses going to other states	‘We had a great conversation with folks about how to protect local businesses from grocery taxes that increase prices, drive customers across state lines and cut incomes for working people’
Business margins	‘It’s already expensive enough to own and manage a business in Washington. Additional taxes on groceries would just drive customers away, lead to job losses and cut sales for small businesses operating on thin margins’

Misleading statements that stress the hypothetical

A fifth frame, also highly inter-related to the third and fourth ones, is the use of frequently misleading statements that stressed hypothetical scenarios which could befall voters if they did not endorse I-1634 (n = 762). These were characterised by claims and/or predictions about future actions and policies that were unlikely or misrepresented. Advertisements often portrayed such actions are already scheduled or inevitable, and portrayed I-1634 as the only means to stop these scenarios from occurring. Illustrative quotes of advertisements under this framing theme are provided in Table 4.

Table 4.

Stressing the hypothetical – excerpts.

Descriptor	Excerpts
Future actions and policies represented as scheduled or inevitable	‘With everything we have on our plates, we shouldn’t fear a tax that will take the food right off of them’. Joseph D., Seattle. ‘We should be implementing policies that work for everyone from business owners to farmers to working families. Taxes on groceries just hold the most vulnerable members of our communities back’
	‘We keep choosing to tax things that make it harder and harder for regular people to get by and keep dinner on the table. Now, we’re talking about taxing dinner itself? No way! Even a small tax on meat or dairy could mean hundreds and hundreds of dollars a year more for groceries’
	‘Given the amount of fresh dairy and produce that we buy each day, alone, new grocery taxes would truly make it difficult for us to make our livings here. We aren’t located in such an advantageous spot as to allow us to raise our prices significantly without losing a lot of business. We are real people, working hard every day, and grocery taxes would really hurt us’

Adopting progressive messaging

Advertisements engaged in progressive messaging to appeal to broad and diverse voter coalitions (n = 554). Advertisements appealed to societal issues facing potential voters, most often referencing the cost of living, equity concerns, supporting the underserved in society, and advocating for the redistribution of wealth. Advertisements positioned a theoretical grocery tax as ‘regressive’ and against the interests of a fair society. Illustrative quotes are provided in Table 5.

Table 5.

Adopting progressive messaging – excerpts.

Descriptor	Excerpts
Cost of living	‘I loved growing up in Washington and was so happy to return here as an adult. But, when I came back, I was shocked by how expensive everything had gotten; how difficult it was to live off my paycheck. I want to see us join together and make sure all the hardworking people of Washington are able to keep enjoy living here without having to worry about what will happen when their paycheck can’t cover basic groceries’
Supporting the undeserved	‘We often reach for punitive measures to solve problems. We punish all city residents especially those with low incomes with the most regressive tax structure of any city in the state, according to a recent report by the Economic Opportunity Institute’
Equity concerns	‘This is about fairness and affordability. I-1634 is our opportunity to close a loophole that allows local lawmakers to lean directly on our families and neighbourhoods for revenue, when that is in fact the last place they should look. It’s time to create fairness in our state’s tax structure. Let’s start by proactively stopping local grocery taxes and voting #YESto1634’
Redistribution of wealth	‘Washington is already recognised as the state with the most burdensome tax structure in the country when it comes to low and middle-income households. Raising taxes on everyday groceries would not only be unfair to those who can least afford it, but would make it even more difficult for working families to earn a living or pay for groceries’

Standing up for the greater good

Lastly, and related to the previous theme, the framing theme of standing up for greater good (n = 444) can be seen throughout many of the advertisements. This includes messaging that positions the ‘Yes! To Affordable Groceries’ campaign as standing up and protecting everyone in Washington against unfair, unethical, and harmful taxes on their basic necessities. Messages like these aim to portray I-1634 as aligned with ‘the little guy’, fighting the goliath of government taxation and over-reach. This approach achieves two goals simultaneously: criticising local governments that would seek to regulate SSBs in the form of taxes, while putting SSB companies in the role of beneficent protector. Table 6 provides illustrative quotes.

Table 6.

Standing up for the greater good – excerpts.

Descriptor	Excerpts
Aligned with the ‘Little Guy’	‘With the support of small businesses like Niko’s from across Washington, we’re working with communities to protect working people and families. Join us today to say Yes! to prohibit local grocery taxes! #YesTo1634’
Protecting Washingtonians	‘Vote Yes! To Affordable Groceries to protect working people, small businesses, and communities across Washington’
Standing up for Washingtonians	‘We’re standing up for working families and small businesses across Washington. Join us to advocate for affordability and protect our communities from excessive grocery taxes’

Discussion

Framing offers an important way to analyse the discursive actions of SSB industry-sponsored campaigns and their impacts on public opinion. Our work builds upon previous SSB taxation framing research that consistently finds the SSB industry promoting messages of economic or market justice (Hilton et al., 2017), disseminating misleading information related to taxation effects (Asada et al., 2021), and messages related to personal freedom (Campbell et al., 2020). The messages used in the ‘Yes! To Affordable Groceries’ campaign exhibit the varying strategies the SSB industry utilises to delay or avoid SSB taxes by framing their lobbying and political actions as in the interests of the general public. While we do not presume these strategies are entirely new, their setting (Facebook) and target (pre-emptive legislation) are relatively novel, and thus require further inquiry into the narrative frames the industry is using to advance its agenda on these fronts. A primary strategy in this campaign is to shift public discourse to focus on the impacts of future, imagined legislation on consumers, and away from SSBs. This is similar to tobacco industry strategies arguing that introducing marketing restrictions on their products could lead to ‘negative unintended consequences’ (Savell et al., 2014) and that standard packaging could lead to a ‘slippery slope’ of future government action (MacKenzie et al., 2018). During I-1634, advertisements focused on introducing hypothetical scenarios where basic groceries, such as dairy and meat products, were at risk of being taxed, despite no risk existing except to SSBs. The theoretical grocery taxes were systematically portrayed as impacting specific underserved or important social groups – such as families, seniors, and small businesses, thus exploiting the hardships faced by certain groups. The SSB industry has previously targeted the financial or social situations of communities to promote their products and refute public health measures (Estes, 2013; Harris et al., 2020; Nestle, 2015; Powell, Wada et al., 2014). Grocery tax legislation is presented as government greed and budget shortfall filling and not in the best interests of the public, despite extensive evidence and recommendations for SSB taxes suggesting otherwise (Redondo et al., 2018; Teng et al., 2019). These messages are put forward by small businesses, advocacy groups, and stores such as supermarkets, which act as ‘influencers’ to advance industry interests without having the public attribute the legislation to industry actions.

What is notable in this case is that the messaging is successful even when directed at the hypothetical possibility of future taxes. Industry advertisers leaned on compelling social groups and community organisations to construct a straw figure to distract voters from the actual motivations for the legislation, similar to previous tobacco industry framing strategies that misrepresent legislation intent (Waa et al., 2017). This is a particularly powerful form of what we might call hypothetical boundary work. While other industries have successfully used boundary work in their discursive engagements with the public to shift definitions of the problem at hand (and whom it concerns) (Kenworthy et al., 2016), this campaign managed to do so regarding an entirely hypothetical, and unlikely, set of future regulations. Framing strategies are essential to such efforts. In addition, the campaign successfully leveraged diverse political alignments in the state to appeal to broad coalitions of people who might be negatively impacted by such future regulation. These appeals tapped into a politics of grievance that is particularly salient in the US, and especially powerful in convincing white voters to vote against their own interests when it comes to health and welfare (Metzl, 2019). By also targeting progressive messaging and imagery broadly defined groups, however, the advertising campaign was able to bring more diverse groups into that coalition of a perceived, hypothetical grievance. In doing so, it shifted the frames of which institutions were threats (taxes, government over-reach, greedy legislators), which were saviors (the ‘grocery’ industry, preemptive taxation bans), and how stakeholders could align (hardworking people, tax-payers, small businesses and families) across the political landscape.

Public health advocates and researchers can draw several important lessons from the factors that led to I-1634s success in order to prevent and address future attempts at pre-emptive legislation. First, the I-1634 proponents were disproportionately funded compared to initiative opponents: ‘Yes! To Affordable Groceries’ received over $22 million in funding, while those advocating against I-1634 raised approximately $113,000 (Ballotpedia, 2018a). The advantage in funding allowed proponents to produce more advertisements, events, and other campaign measures. The effect of this funding gap, its impact on the passage of I-1634, is displayed in the outcome a similar campaign. In Oregon, Measure 103 sought to ban the introduction of new grocery taxes, but was rejected on the ballot. In this campaign, the proponents raised approximately $8 million, compared to the opposition side who raised $11 million (Ballotpedia, 2018b). The opposition was able to widely disseminate counter-messages that the industry was sponsoring I-103, and promote the intentions behind the bill, thus contributing to its rejection. In political arenas related to public health, solutions are needed to address this funding gap.

Second, advertising platforms such as Facebook allow the production, targeting, and rapid dissemination of misleading political advertisements (Isaac & Kang, 2020). I-1634 advertisements often hid information about campaign sponsors: logos appeared only at the end of videos in small fonts that are hard to read; photos rarely contained a reference to the ABA or other donors. Instead, advertisements prominently featured the ‘Yes! To Affordable Groceries’ banner alongside messaging that the campaign is a local grassroots initiative. Facebook’s advertising platform gives advertisers multiple strategies for concealing where content is coming from, including creating posts that appear to come from regular people, and encouraging users to share and comment on advertisement posts, which then gets disseminated through their social networks. In addition, much of the messaging content of the campaign was highly misleading. The hypothetical grocery taxes were not at risk of being introduced, nor were there future plans to tax other products, except for SSBs. The economic impacts of SSB taxes, mentioned in several advertisements, were portrayed as having a significant impact on loss of jobs and store revenue after the introduction of the Seattle SSB tax, which is untrue (Powell, Wada, Persky et al., 2014). Advertisements did not mention that under Washington State law, basic grocery items cannot be taxed. Finally, the public health intentions behind SSB taxes were misrepresented as a strategy for government to fill budget shortfalls.

The use of misleading information in corporate advertising contributes to broader problems of public misinformation, power hierarchies, and digital health harms on social media platforms such as Facebook (Storeng & de Bengy Puyvallée, 2019). While little research has examined issues of misinformation in corporate (rather than political) advertising on Facebook, researchers have noted several dynamics that contribute to the spread and entrenchment of misinformation on the platform. Research has found that social media fosters ‘echo chambers’ where users increasingly interact with, share, and find resonance with, content that confirms their existing beliefs (Quattrociocchi et al., 2016; Schmidt et al., 2018). This dynamic tends to reinforce, and entrench, divisive viewpoints. Advertising content from the I-1634 campaign shows efforts to appeal to specific groups with messaging that resonates with broader political and social beliefs and creates and demonises straw figures such as ‘grocery taxes’. Facebook’s advertising services also allow highly specific advertisement targeting to geographic and demographic groups, enabling misleading content to be diversified and targeted to appeal to specific political coalitions. The distinctive ‘platform power’ (Culpepper & Thelen, 2020) of Facebook creates additional opportunities for unhealthy commodities industries to excise power and leverage their interests through such advertisements. This compounds the already racially and economically inequitable targeting of traditional food and beverage advertising (via TV, billboards, and print media) to Black and Hispanic youth (Harris et al., 2019). Public health researchers must make it a priority to advocate for social media platforms such as Facebook to define misinformation, require advertisers to clearly label advertisements with the entities that paid for them, flag misleading information in warnings, and remove them when misinformation criteria is met. Further research – and data – is also needed to better understand how corporations are leveraging Facebook analytics to deliver highly targeted advertising content to marginalised groups.

There are several limitations to our study. First, the advertisements included are only paid advertisements found on the Facebook Advertisement Database. There may be other advertisements not included in our sample on other social media platforms such as Twitter or YouTube, television, or the ‘Yes! To Affordable Groceries’ website. However, our sample includes all advertisements on Facebook, which replicate advertisements found elsewhere. As a subset of all advertisements in the ‘Yes! To Affordable Groceries’ campaign, Facebook advertisements represent an important and substantial domain for examining discursive framing strategies. Our coding structure also relied on visual cues for identifying demographic information, a strategy that is commonly recognised as best practice in internet research (34) but not without its limitations. This strategy prioritises coding for demographic appearances, based on culturally-shared markers of race, gender, and ethnicity, but this subjective coding approach may not always reflect how individuals themselves would self-identify (Agency for Healthcare Research and Quality, 2018).

Conclusion

This article presents the first known study of the ‘Yes! To Affordable Groceries’ campaign advertisements in Washington State. It also contributes to a growing scholarship on how technology platforms are being leveraged as a commercial determinant of health. The SSB industry organising, framing, and funding of the ‘Yes! To Affordable Groceries’ campaign to pass I-1634 impeded future local government efforts to tax sugar-sweetened beverages, an effective public health promotion tool. By studying distinctive framing strategies used by the campaign to convince voters that I-1634 was in their best interests, we identify common discourses and strategies used online to influence voters. We also find troubling patterns of misinformation: most advertisements used misleading information and hid the industry-motivated intentions of the initiative. This raises questions about how advertising by harmful industries can be regulated in online social media platforms, where misinformation has the potential to spread rapidly through ‘echo chamber’ effects. Here, we identify key ways that the SSB industry is leveraging social media dynamics to appeal to, manipulate, and create divisions among, broad public coalitions in order to cultivate support for the legislation.

A framing approach that identifies common rhetorical tactics can also provide opportunities for public health institutions and advocates to be better prepared to counter such narratives in the future (Plec & Pettenger, 2012). Many of the strategies identified in this campaign have been used in the past by other harm industries. Public health must work proactively to strategically identify alternate discursive frames and counter-arguments that can be used to contest pre-emption initiatives in other states. Efforts to build the capacity of public health communications to respond proactively to advertising campaigns like this one must be better funded if they are to succeed in achieving the spread and dispersion necessary to reach communities, coalitions, and voters.

Appendix. Organizations mentioned or shown by frequency and type

Name	Frequency	Type
Code: Washington State Farm Bureau	57	Trade or Advocacy Group
Code: The Spokesman Review	30	News
Code: ptleader.com	4	News
Code: Washington Hospitality Association	35	Trade or Advocacy Group
Code: Vashon-Maury Island Beachcomber	2	News
Code: Nana’s Tiny Cakes	2	Restaurant, Bakery, or Café
Code: Shady Grove Farm	4	Other
Code: High Nooner Gourmet Sandwiches	21	Restaurant, Bakery, or Café
Code: Emma’s BBQ	3	Restaurant, Bakery, or Café
Code: Ethnic Chamber of Commerce	16	Trade or Advocacy Group
Code: TVW	6	News
Code: Highway Espresso	6	Restaurant, Bakery, or Café
Code: Seattle Channel	7	News
Code: Journal of Business	4	News
Code: Teamsters Local 174	54	Trade or Advocacy Group
Code: Spokane Business Journal	9	News
Code: Tacoma Weekly	3	News
Code: Whidbey News-Times	3	News
Code: Union Bulletin	3	News
Code: Omak-Okanagan County Chronicle	3	News
Code: French Hen	21	Restaurant, Bakery, or Café
Code: Buzz Pizzeria	9	Restaurant, Bakery, or Café
Code: Amazing Thai Lao’s	4	Restaurant, Bakery, or Café
Code: Rosauers	2	Store
Code: Amy’s On the Bay	2	Restaurant, Bakery, or Café
Code: Washington Cattleman’s Association	47	Trade or Advocacy Group
Code: JR’s Bar-n-Grill	2	Restaurant, Bakery, or Café
Code: Taqueria Los Toritos	16	Restaurant, Bakery, or Café
Code: Pizza Pro	20	Restaurant, Bakery, or Café
Code: Mi Ranchito	3	Store
Code: Nostalgia House Bakery	3	Restaurant, Bakery, or Café
Code: Mitapeap Restaraunt	7	Restaurant, Bakery, or Café
Code: Emerald City Fish and Chips	5	Restaurant, Bakery, or Café
Code: Hokus Pokus Mini Mart	5	Store
Code: Ballard Sunday Farmers Market	2	Store
Code: King County NAACP	6	Trade or Advocacy Group
Code: Central Area Chamber of Commerce	19	Trade or Advocacy Group
Code: Federal Way	6	Other
Code: Korean-American Grocers Association	36	Trade or Advocacy Group
Code: Better Spokane	35	Trade or Advocacy Group
Code: Greater Spokane Incorporated	31	Other
Code: Spokane Valley Chamber of Commerce	37	Trade or Advocacy Group
Code: Kate’s Greek and American Diner	6	Restaurant, Bakery, or Café
Code: Progressive Democrats of America	6	Trade or Advocacy Group
Code: Crosscut	2	News
Code: Mediterranean Oasis	2	Restaurant, Bakery, or Café
Code: Teamsters Local 117	29	Trade or Advocacy Group
Code: Chelan/Douglas County Farm Bureau	6	Trade or Advocacy Group
Code: Bene’s AM Fix	2	Restaurant, Bakery, or Café
Code: Lakewood Chamber of Commerce	2	Trade or Advocacy Group
Code: Teamsters 690	8	Trade or Advocacy Group
Code: Spokane County Farm Bureau	6	Trade or Advocacy Group
Code: Neiner Neiner Weiner	3	Restaurant, Bakery, or Café
Code: Bargain Giant	13	Store
Code: Teamsters 28	27	Trade or Advocacy Group
Code: Washington Retail Association	27	Trade or Advocacy Group
Code: Washington Food Industry Association	29	Trade or Advocacy Group
Code: Wray’s Market Fresh IGA	6	Store
Code: Greater Yakima Chamber of Commerce	33	Trade or Advocacy Group
Code: Jackson Street Market	6	Store
Code: The National Taxpayers Union	3	Trade or Advocacy Group
Code: The Washington Policy Center	3	Trade or Advocacy Group
Code: American for Tax Reform	3	Trade or Advocacy Group
Code: Council for Citizens Against Government Waste	3	Trade or Advocacy Group
Code: Tim’s Tavern	6	Restaurant, Bakery, or Café
Code: Washington State Dairy Federation	27	Trade or Advocacy Group
Code: Pasco Chamber of Commerce	25	Trade or Advocacy Group
Code: Whatcom County Farm Bureau	6	Trade or Advocacy Group
Code: Teamsters Local 839	25	Trade or Advocacy Group
Code: Machinists District 751	6	Trade or Advocacy Group
Code: Seattle’s Building Trades Union	25	Trade or Advocacy Group
Code: American Beverage Association	25	Trade or Advocacy Group
Code: Washington State Tree Fruit Association	31	Trade or Advocacy Group
Code: Washington Beverage Association	25	Trade or Advocacy Group
Code: Tri-Cities Hispanic Chamber of Commerce	25	Trade or Advocacy Group
Code: King County NAACP	1	Trade or Advocacy Group
Code: Hispanic Chamber of Commerce	31	Trade or Advocacy Group
Code: Yakima Grocery Outlet	6	Store