Climate obstruction and Facebook advertising: how a sample of climate obstruction organizations use social media to disseminate discourses of delay

In this article, we present the results of a content analysis of a sample of fossil fuel corporations, industry associations, and advocacy groups’ ads using Facebook ad services during the election year from January 2020 to January 2021. Creating a messaging typology, we analyzed 30,100 ads on Facebook and found that different organizations used these messages to reinforce pre-existing beliefs on the importance of fossil fuels. In doing so, these organizations can use Facebook to reassert their interests in the public voice and support standards of behavior that rely on fossil fuels, protecting their industries. Additionally, we found differences between the types of messaging and ads targeted to particular users by age, gender, and state. These ads represent the evolving misinformation playbook from climate obstruction organizations that require further attention and consideration, particularly via social media platforms that may have limited or unclear regulations.


Introduction
Supported by public relations (PR) campaigns, the fossil industry and related climate obstruction organizations have managed a campaign to undermine climate mitigation for the past 30 years (Brulle and Werthman, 2021).This article adds to the existing literature on PR and climate obstruction, presenting the results of a content analysis of a sample of fossil fuel corporations, trade associations, and advocacy groups ads using Facebook ad services during the US election year from January 2020 to January 2021.We contend that climate obstruction organizations use Facebook ad services to deploy delaying discourses to legitimize and greenwash these industries.In turn, this reinforces a social license to operate in response to the climate crisis using narrative techniques that tap into societal values and moral judgments to continue using fossil fuels despite its negative evidential impacts.To this end, our paper answers the following research questions: 1) what are the different narrative strategies utilized by obstruction actors online during this period ? 2) what are the differences between the deployment of narratives by the three groups of obstruction actors related to location, age, gender, and nance?And 3) what is the role played by social media platforms in facilitating (or not facilitating) critical debate?2. Literature Review

Climate Obstruction and Public Relations
The climate obstruction movement comprises multiple organizations and individuals who have engaged in an organized campaign to delay and obstruct action on climate change.These include fossil fuel corporations, conservative, partisan, and neoliberal think tanks, conservative foundations and donors, trade associations, advocacy groups or Astroturf groups, professional associations, research institutes, public relations (PR) rms, nance rms, individuals, and supportive politicians and alliances (Dunlap and Brulle, 2020).In turn, they have successfully undermined domestic, bilateral, and multilateral climate legislation by using monetary resources, political in uence, and successful media campaigns.
Understanding PR campaigns connected to fossil industry interests and their impact on delaying climate action is a growing area of research (Almirion and Xifra, 2020).For example, Brulle and Aronczyk (2019, 224) noted that "the US corporate community has integrated PR and lobbying into its business strategy." This extends to fossil fuel corporations using PR rms to boost the legitimacy of their organization and, in turn, their climate-related policy motives.Similarly, Brulle and Werthman (2021) note that 'PR rms are key organizational actors in climate change politics.Various sectors, including oil companies and electrical utility suppliers, have used PR rms to elevate their public messaging and attempt to in uence public support and legitimacy for their industries.Likewise, Brulle et al. (2021) examined the advertising expenditure of ve major oil and gasoline companies between 1986-and 2015.They identi ed that PR campaigns appear central to elevating the interests of these oil and gas companies.
Michaels and Ainger (2020: 160) draw attention to several European PR campaigns by gas-related industries that pay "lip service to the need to address climate change" but intend to undermine the legitimacy of support for climate legislation and greenwash businesses.Drawing attention to the European Gas Industry, the industry spent millions in lobbying activities and PR campaigns, framing gas as a partner and non-solution that should accompany renewable energies to meet climate goals.Supran and Oreskes' (2021) analysis of ExxonMobil's public discourse illustrates that the organization framed its business as a savior, downplaying the seriousness of climate change and shifting the responsibility of addressing the crisis onto individual consumers.These PR campaigns have historically taken place through traditional media outlets such as TV, radio, and newspapers.However, given social media's global and instantaneous nature, this platform provided lucrative advertising opportunities for climate obstruction organizations to expand their operations and disseminate discourses of delay that may in uence public perceptions on climate change and voter behavior.

Habermas' Theory of the Public Sphere
To better understand climate obstruction Facebook-based PR campaigns, we draw on two areas of scholarship: rst, on Habermas's (1989) Theory of the Public Sphere, second, existing contributions to businesses and organizational studies.In short, Habermas (1989) argues that the public sphere is integral to political democracy.The public sphere offers a space for critical and rational debate on issues contributing to deliberative democracy.Historically, the public sphere was used to 'inform citizens' and align public views with state interests.However, changes in the political economy from the 18th and 19th centuries meant that the public sphere transformed, supposedly providing a platform to exchange knowledge without state interests.But then, under neoliberal changes in the political economy, particularly in the West, the public sphere regressed to reinforce state and private/corporate interests that become asserted in the public voice (Habermas, 1992).
Similarly, Bernstein (2012) notes that the public sphere was replaced by the state and corporate interests to advance policies that centralized market forces to prevent necessary public intervention, whereby "public opinion has degenerated into the sophisticated manipulation of opinion by special interests" (p.767-768).Bernstein (2012) extends that because the public sphere has a normative core, the public sphere is used to reinforce standards of behavior that align with corporate and state interests.In other words, norms can help evaluate behavior, which leads to the possibility of preserving something such as fossil fuel use despite its documented negative implications.For the obstruction industry, this normative core is used to create what appear as meaningful conversation-like interactions on platforms such as Facebook, with the public viewing ads like a stakeholder.In addition, businesses convey messages about the company to convince the public that their industries are good, productive, and vital to society.Thus, messages are conveyed in the public sphere to promote and strengthen behavioral practices that cannot be divorced from state or corporate interests.
Researchers have applied Habermas' concept of the public sphere to social media.For instance, Rodriquez (2021) suggests that Facebook groups and reactions optimize what it means to have a public sphere for deliberative democracy.Similarly, Beers (2006) argues that social media and independent news media online are closer to the ideal of Habermas' public sphere.Likewise, Sani (2014) argues that the social media PR campaigns used during the 2013 Malaysian General Election played an essential role in positively affecting democracy in Malaysia.In short, these authors suggest in part that social media platforms are a space that allows the interaction of critical discussions away from political and economic forces.
In contrast, others have questioned if social media platforms can function as an ideal public sphere.For instance, Vaidhyanathan (2018) contends that Facebook segments users into market groups and uses precise and targeted communication that undermines democracy.Therefore, Facebook does not provide a public forum to deliberate and instead reproduces ideological echo chambers.This means that political and commodity advertisements only reinforce ideological currents limiting a user's critical engagement and independent thought.Using the case studies of two climate skeptic advocacy groups on Facebook, Bloom eld and Tillery (2019; 30) note that these organizations use rhetorical devices and practices that "challenge argument standards for public spaces to be reasonable, rational, and deliberative."This means that the marketing practices of these speci c groups related to fossil fuel interests are used to convince the public on a social media platform that these values and opinions are their own, and not in the interests of the state and corporations.Harper (2017) documents how big data can be aggregated and used to tailor ad distribution to speci c user pro les.This activity assumes that these messages are more likely to be favorably received because they reinforce pre-existing values and beliefs.This can have political implications.For instance, Karppi and Nieborg (2020) describe the Cambridge Analytical scandal, demonstrating how Facebook combined big data and behavioral media research to manipulate outputs that targeted certain users had 'unintended' consequences in the 2016 US election (See also Bennett andLyon, 2019, Konitzer et al., 2019).
Facebook employs a series of policies related to advertising by the private sector or related organizations that, on the surface, could provide an opportunity for free knowledge exchange because users of the platform are informed that advertisements are 'paid for' and thus have interested parties.Two of these policies focus on 1) information related to social, political, and economic issues that could be considered 'politically' oriented and 2) misinformation.For example, on social issues, elections, and politics, Facebook outlines that the advertiser must comply with applicable laws, and it may restrict access to the ads that fall into "the categories of "civil and social rights, crime, the economy, education, environmental politics, guns, health, immigration, political values and governance, security and foreign policies" (np).On misinformation, Facebook prohibits ads that include claims debunked by third-party fact-checkers or, in certain circumstances, claims debunked by organizations with expertise on the subject.In addition, advertisers that repeatedly post information deemed false may be restricted from posting ads for a time.
A caveat is that Facebook can employ discretion in this decision-making process.This discretion requires transparency.While the concept of transparency has varying de nitions (Albu and Flyverbom, 2016), here, it refers to the understanding that an organization is based on the increased disclosure of information, indicating accuracy (Wehmeier and Raaz, 2012).Transparency is essential for Facebook to build trust between the public and the company (Berggren and Bernshteyn, 2019).Whether the content itself is accurate, achieving higher levels of transparency can help reduce questions about the credibility of information Facebook promotes.In turn, this offers a layer of accountability to the company decreasing the likelihood of questioning sources and enhancing the organization's reputation.Moreover, if corporate actors use Facebook to reinforce or shape societal values, organizational transparency and accountability of Facebook must be a priority.This is because the perception of doing good and trusted information from fossil industries are intertwined with Facebook.As such, corporations not only use social media platforms to market and communicate with potential buyers (e.g., Johannessen et al., 2016), they also speak to public behavior and political debates that suggest a necessary transformation in behaviors to -in this case -address the climate crisis.
In the case of 'environmentally friendly' products, these practices can be thought of as greenwashing.
While there are different de nitions of 'greenwash,' public understandings of the term have typically centered around communication that "misleads people into holding overly positive beliefs about an organization's environmental performance, practices, or products" (Lyon and Montgomery, 2015: 225).Through their review of a wide range of literature on greenwash, Lyon and Montgomery identify notable methods of greenwash, including selective disclosure, empty claims, misleading narratives, and misleading imagery.Lyon and Montgomery also point to the marketing literature, where the de nition of greenwash or misleading communication depends on how individuals interpret claims (Oswald, 2011).As such, greenwash is 'in the eye of the beholder.'These ads are essential for branding and re-labeling fossil industries as 'green' and serve a political purpose for the obstruction industry.More speci cally, it is to extend their outreach to public perceptions of their industries to reinforce traditional ways of thinking on the consumption of fossil fuels; for example, it is a necessary component of the energy transition.In addition, they can connect individuals' everyday lives and consumption choices to political proposals for climate action.By tuning into these values, the outcome can shape voter choices, choosing a political option that promotes policies that align with their own beliefs.For example, Li et al. (2022) examined the clean energy claims of BP, Chevron, ExxonMobil, and Shell.They note that between 2009-2020 there had been a substantial increase in discourses that intended to present their businesses as 'clean' and 'low carbon' options but demonstrated little action and a continuous business model reliant on fossil fuels.Thus, "that accusations of greenwashing by oil majors are well-founded" (p.19).
In addition to greenwashing, Facebook becomes a vehicle to promote and generate support for a social license to operate.Research across various disciplines demonstrates that industries invest time and resources to work with communities and stakeholders to outline the costs and bene ts of industrial development (Hall et al., 2015).In turn, it legitimizes their industries, successfully establishing a sense of social acceptability for development.For example, Gaither and Gaither (2016) examine the advertisements of the American Coalition of Clean Coal Electricity and the American Petroleum Institute.
They conclude that these two organizations tap into US cultural values that help shift the dialogue.Rather than focusing on climate change itself, they turn attention to how environmental policies impact American citizens' daily lives.
While not necessarily in a local context in our study, paid-for ads by corporations on these social media platforms offer another potential space to convince the public that the bene ts of the industry outweigh that of the climate crisis.While this may not be a consultation practice within a local community generally, the ability to send targeted ads to speci c geographic locations provides the opportunity to impact public opinion on voting issues.For instance, Cole et al. (2021) examined the Coloradan Fossil fuel Misinformation playbook.They highlight the example of tactics used by different actors in the obstruction movement closely related to the fossil industry in Colorado.They draw attention speci cally to two front groups that used the platform Facebook to greenwash.They note that 'the ads go to great lengths to highlight the positive contributions fossil fuels make to the state, leaving out the damage and harm that comes from the industry's operations" (p.16).As a result, the positive framing of fossil fuel industries to a region rea rms its acceptability even during the climate crisis.Furthermore, given that climate change was a central issue in the 2020 General Election (Tyson, 2020), tapping into the values of the public is vital to appeal to individual responsibility as the solution to the problem and nd support for minimal change.

Discourses of Delay and Narrative Strategies
Researchers have advanced our understanding of the delaying messages employed by climate obstruction actors (For an overview, see Lamb et al., 2021).We extend the analysis of delaying narratives, applying a re-developed typology from Miller and Lellis (2016).We re-developed Miller and Lellis' categories for this study (See Table 1).The rst category is Community.Participants in Miller and Lellis' study identi ed the value of community in ads that highlighted the 'positive' economic impact of the fossil fuel industry on local communities (i.e., provides jobs).In the context of this study, we included a broader range of nancial bene ts to the community from the oil and gas industry, including the generation of tax revenue and industry donations to local programs.
On the second category, we adapt this from the category Innovation.Participants in Miller and Lellis' study identi ed fossil fuel company messaging that presented their businesses as innovative and providing solutions that create positive change, whether through modernization, developing new technologies, or raising the argument that the company is progressive.For this study, we isolated the category of innovation to concern climate solutions.This category is designed to capture narratives on how the oil and gas industry provides solutions to climate change.The third category developed by Miller and Lellis was Patriotism, one of the most explicit and consistent categories that emerged in their data because these messages conveyed pro-American values.As such, participants identi ed the central role of 'US' industries as world leaders, connected to the identity of US consumers.Here, pro-American values include traditional messages associated with fossil fuels and the American ways of life.
Miller and Lellis' fourth category was Resilience.In their study, participants identi ed resilience with the idea of a " ghting spirit" and "the underdog."The other aspect of resilience participants recognized was the economic bene t of having the industry around contributed to the nancial resilience of people "who had been hard hit by the economy."Concerning the sample of ads used in our study, the theme 'resilience' was more aligned to both the categories 'patriotism' and 'community' rather than be distinguishable and separate.As such, we did not generate a new category for 'Resilience' and instead combined these under the two categories of patriotism and community.However, Miller and Lellis' study identi ed another value: Pragmatism.The value of pragmatism was consistently highlighted by Miller and Lellis' study participants, expressed as a need to balance environmental and economic concerns rather than 'go green all at once.' Additionally, one participant commented: 'What they're trying to point out is that it [the industry] is a necessary thing, and it's not as bad as what people perceive it as.'These ideas around the necessity of fossil fuels and the need to remember their bene ts when considering emissions reductions were used to create our nal category of 'a Pragmatic Energy Mix.' Notably, Miller and Lellis use Habermas (1989) to understand their ndings.They argue that corporate advocacy campaigns can in uence legislative or policy debates through a form of manipulated publicity 'designed to ensure public acquiescence of the positions of powerful corporate actor's" (p.250).Corporations use PR campaigns to promote their commodities and speak to these political and legislative debates central to protect their interests.Furthermore, corporate campaigns use targeted ads that engage with 'values' to manipulate the public interests in line with their own.They do this by appealing to commonly held societal values as an a rmative method to legitimize their industry.In the case of fossil fuel corporations, a rmative messages speak to the public interests and values that centralize fossil fuel use in daily lives.The impact is the successful legitimization of that harmful business practices, public opinions and potential voter choices based on support or non-support for environmental and climaterelated policies.

Data
We generated a list of all advertisers running ads related to climate change in the US between 1 st January 2020 and 1st January 2021.To create this list, we used Facebook library API to obtain climate-related ads that ran throughout 2020.We then aggregated Ad level data to get a list of advertisers and how much they spent on the climate-related ads.From this list, we created a sample of the top ten fossil fuel companies by spending on Facebook ads.We then added to the sample the top ve industry associations (by the same measure) representing the oil and gas sector and ten advocacy groups that spent over $5000 which claimed to represent the fossil fuel industry in their own material or had proven nancial ties to fossil fuel companies or industry associations.These links are well documented in media reporting and the climate obstruction literature.In total, our sample included 25 entities (A complete list can be found in Appendix A).Because ad campaigns often run multiple adverts with identical or very similar text, we used an algorithm to "compress" the dataset by matching ads where the ad's text shared at least 80% of the same characters.Across all 25 entities, 30,100 unique ads were running on Facebook's US platforms during the period of data collection.

Coding Scheme and Analytical Approach
We created a coding scheme that included the following categories: 'Community & Economy,' 'Climate Solutions,' 'Pragmatic Energy Mix,' and 'Patriotic Energy Mix.' Table 1 identi es each of these categories, an explanation, and a list of key narratives.Ads that did not t these categories were excluded.Next, we assessed each ad using qualitative content analysis to determine if the ad contained one of the narratives identi ed in the typology.Initially, we had started with a larger team of coders and 25 subcategories under the four broad categories based on Miller and Lellis' work.To narrow this down, we performed three rounds of intercoder reliability testing to obtain an eventual Fleiss-Kappa score of 0.78.Each iteration of intercoder reliability testing offered us insights into the utility of the category we had generated.After these three rounds, we ended with the above typology, which contained four broad themes, each with three subcategories.On reporting the results, we use the four broad themes rather than subcategories, much like Miller and Lellis' analysis.

Methodological Limitations
First, while the Facebook Ad Library covers over 125 countries and territories, at the time the research was conducted, the Ad Library API, which allows researchers to download the ad data, only extended to the UK, the USA, and Brazil at the time of data collection.The Ad Library has since expanded in coverage, which would, in theory, allow us to repeat the study in different geographical regions.Second, we limited the project to English-speaking ads only; therefore, we could not analyze ads in alternate languages.Notably, election ads in the USA will target different communities, including multilingual communities.As such, we were unable to examine these advertising practices.Third, the Ad Library only stores ads that have been disclosed as social issues or political ads.The advertiser is meant to be disclosed and then checked by Facebook.However, as discussed in the results section, this disclosure mechanism was ineffective.Ads about social and political issues that have not been disclosed are not stored on the Ad Library and are inaccessible to the researcher past the ad's run-time.Therefore, it is impossible to know how many ads containing content about social and political issues are stored on the Ad Library.Therefore, the absolute number of ads could be signi cantly higher.

Analysis
Table 2 breaks down the distribution of delaying messages by narrative, the number of ads and impressions, and total spending in US dollars.We took the median value of the upper and lower values for both the data on Impressions and Spend in US dollars and impressions rounded to 2 signi cant gures.First, while the most common narrative utilized by the sample of organizations is Climate Solutions (N = 12140, 48%), it had the second-lowest spend ($1,893,080) and the third-highest impressions (N = 122,248,437).The second highest number of ads contained under the category Pragmatic Energy Mix (N = 7749, 31%) had the highest spend ($4,354,825) and the highest number of impressions (N = 174,545,645).The third most common ad under the category Community and Economy (N = 5542, 22%) had the second-highest total spend ($1,893,080) and the second-highest number of impressions (N = 122,248,437).Lastly, the least common narrative was Patriotic Energy Mix (N = 7749, 31%) which also had the lowest spend ($1,482,998) and lowest number of impressions (N = 55,474,052).These descriptive ndings indicate that the most popular ad by impressions and spend is Pragmatic Energy Mix.The messages in this category emphasize that oil and gas can and should be part of the energy mix to support people's everyday lives.For example, Fig. 1 taken from an Energy Citizen's ad -an advocacy group run by the American Petroleum Institute -claims that "access to natural gas and oil means reliable energy Americans can count on to heat and cool their homes and power everyday life."Similarly, Fig. 2, promoted by the advocacy group Coloradans for Responsible Energy Development, connects fracking to "keeping our homes and businesses warm on chilly days… [and] leading the way to a reliable and more affordable energy future."These examples go to Miller and Lellis' assessment of marketplace advocacy, whereby PR campaigns from fossil fuel corporations reinforce fossil fuels as the primary energy source supporting everyday behaviors.
The second most successful ad category by the number of impressions is Community and Economy, which positively framed the fossil fuel industries' impact on local communities.For instance, in an ad posted by Leidy South sponsored by Williams Company Inc, they claimed that the 'project will generate $39M in economic activity to bene t Schuylkill County and $100M across PA (Pennsylvania) (Fig. 3).
This example speaks explicitly to the nancial contributions to a speci c community.Like Miller and Lellis' typology, the message may resonate with those that appear to focus on resilience building and are central to community well-being.Thus, ads were used to position fossil fuels as a community asset and value, aligned with corporate interests to the public.The outcome is further support for industrial development provided by the social license to operate and taps into the fossil fuel solutionism and appeals to well-being discourses of delay identi ed by Lamb et al. (2020).
On climate solutions, ads frequently cited fossil gas as 'green' or 'clean.' (27.0% of total ads, 6782 individual ads).This claim argues that fossil gas should be a signi cant part of the energy mix.For example, as the 'perfect partner' to renewables or that coal-to-gas switching is a type of climate solution.
For example, Fig. 4 is an American Petroleum Institute sponsored ad that claims: "We can all agree we need strong climate solutions-and with natural gas as a dominant energy source, US carbon emissions are the lowest in a generation."Similarly, Fig. 5 is an ad sponsored by the American Gas Association, which claims in the text and accompanying video that "young voters want affordable climate change solutions and recognize the affordability of #natgas.See how natural gas is critical for American Communities." These examples promote what Lamb et al. (2021:3) label as a form of fossil fuel solutionism, whereby "this discourse is at the heart of industry pushback against regulation.The American Petroleum Institute funds tens of millions of dollars' worth of advertisements that promote 'cleaner' fossil fuels."This fossil fuel solutionism symbolizes the chance for a normative assessment of fossil fuels, construing fossil fuel not as a cause of environmental harm but rather as a social good.
The category of Patriotic Energy Mix had the lowest number of impressions, despite being the most common noted by Miller and Lellis in their study.These narratives negatively frame foreign investments and positively frame the US oil and gas industry.For instance, in one ad, the US energy industry is described as "the industry has achieved record-high/record-breaking oil and gas production" or "the US is leading in oil and/or gas production."This notion of patriotic sovereignty and regional identity is documented in the literature regarding perceptions of climate change.For example, in the case of Australia, Wright et al. (2021) note that the fossil industry (coal) has tied Australia's national identity to the industry.Similarly, Gaither and Gaither (2016) describe a similar effect in the US and public relations campaigns.In the same way, these messages evoke the image of fossil fuels as central to an American's identity.Thus, this narrative may be useful as a successful advocacy campaign, despite its low number of impressions and relatively low expenditure.
Table 2 identi es the differences between the type of organization (corporations, advocacy organizations, trade associations), total spend, and the number of impressions.Over the election cycle, fossil fuel companies spent the most on ads ($5,680,115).This spending is understandable noting the amount of disposable capital and traditional spending on advertising these industries use.Additionally, these companies spent more on ads that adopted Pragmatic energy solutions (N = 4208).To protect the fossil fuel status quo, using messages that play into the arguments of a balanced energy mix during the energy transition is unsurprising and corresponds with Devauld and Green's (2011), assessment of Shell Oil that used greenwashing to push their role in green transitions whilst at the same time appearing to 'green' their industries.
While corporations took a lead role in spending, trade associations and advocacy organizational expenditure helped elevate the messages.For instance, trade associations spent $3,189,627, and advocacy groups spent $632,930 during the same cycle.Moreover, existing research demonstrates the nancial ties between these fossil fuel companies, trade associations, and advocacy groups (Brulle et al., 2021).Although to be clear, we cannot determine the direction of funding and its in uence on spreading delaying messages.Nevertheless, this data rea rms evidence of the collective effort of the obstruction machine to promote messages that advertise products and attempt to undermine climate policy through a focus on social, political, or cultural issues.
On messaging, industry associations promoted 'climate solutions' (N = 9701); a signi cantly higher percentage of their ads than other narratives (55.56%).In context, industry associations are formed to protect stakeholders and represent workers in these industries.By attributing fossil fuels as a solution to the crisis, reinforces the protectionism of the industry and the workers.Notably, these workers are also voters whose livelihoods are associated with the oil and gas industry.In the context of Norway, Tvinnereim and Invarslfaten (2016) identify that those working in the oil and gas industry are less in favor of constraining fossil fuel production, particularly when it is connected to prospects for future employment.Therefore, if these industry and worker representatives promote fossil fuels as a climate solution, it emphasizes the protection of employment; an economic consideration for workers who may then gravitate towards political choices that incorporate the industry rather than those that look to limit the fossil industry's role.
Lastly, while limited in the overall number of ads, advocacy groups were more likely to adopt the category of community and economy (55.84%).As a reminder, the narrative of community and economy is to present the bene ts of the industry as an overall part of community well-being generating support for industrial development in a local area, i.e., a social license to operate.Advocacy groups can best be described as citizen groups that seek to in uence public policy advocating on issues to promote or resist social change (Minkoff et al. 2008).Advocacy groups in the obstruction movement are well funded by the fossil industry sector (Brulle et al. 2021) to resist larger structural action on climate change.That these groups represent 'public interests' is thus questionable noting their funding sources.Nevertheless, they utilize this messaging associated with building communities and community well-being to target users based on their daily lives.This in turn reinforces the idea that they are acting on behalf of the public, yet behind the curtain their operations are largely determined by corporate interests.There were some notable differences across demographic characteristics and how particular messages were targeted.For example, 'Climate Solutions' narratives were the only category of messaging shown to more females than males.Siegrist et al. (2007) note that women tended to have a higher perception of risks and are less supportive of emerging technologies than males concerning nanotechnologies.While not climate-related, the notion of technological solutions in fossil fuels is one form of delaying discourse (Lamb et al., 2021).Promoting a 'new' non-fossil fuel based solution may reinforce resistance to these technologies particularly when the shift away from fossil fuels is combined with perceptions that new energy supplies are inferior to those existing or could be developed by the fossil industry.Patriotic messaging category was shown more to 25-34-year olds than the older categories, with a similar rate across genders.This nding is interesting for two reasons.One, patriotic messaging taps into traditional conservative values and with the older population (e.g Jylha and Hellmer, 2020).Therefore, the targeted messaging here does not align with this existing literature.Two, the lack of differences between genders does not align with some of the literature on petro-masculinities.Daggett (2018) highlights that fossil fuels represent more than pro ts for corporations and extend to identity creation that is gendered.
Combining fossil fuels with the identity of white, conservative American Men has been a core predictor of climate scepticism.Therefore, this lack of gender differences and not signally patriotic messaging to align with these gendered and political identities is surprising.The geographic distribution of the ads shows that the most targeted locations were in states with high levels of oil and gas production, such as Texas and New Mexico, and swing states, including Iowa and Ohio (see Table 4).In addition, states with speci c policy battles affecting the oil and gas sector in 2020, such as Alaska, also sustained signi cant advertising campaigns.In the case of Texas, there was the most signi cant spend compared to other states on ads (US$ 1,420,836.30)although per person based on US census data is relatively low.Price and Ronck (2017) note that Texas has a strong association with oil and gas and its heritage.This draws attention to the possibility that sustained ads in these locations tap into local-level community values, reinforcing these with potential political implications.
Similarly, a study by the Yale Program on Climate Change Communication and the Center for Climate Change Communication looked closely at the role of climate change in six swing states; Florida, Michigan, Iowa, Texas, Ohio and Georgia.They noted that "Across all six swing states, policies that would reduce greenhouse gas emissions and expand renewable energy sources are not only very popular among voters regardless of party a liation, they are clearly winning issues for political candidates" (2021, p.2).Because climate change and energy are salient topics in political races, it is unsurprising that this would increase policy battles, especially in states with high levels of support for fossil fuel industrial development.32.36% (US$2,823,455.24) of all spending on ads was targeted in just these six swing states alone.This targeted ad campaigning then is like the existing literature on climate change but also extends to the role of Facebook, which ultimately has used big data to distribute ads to speci c user pro les that reinforce pre-existing values and beliefs (Harper, 2017).Such practices could hypothetically impact policy and election battles, where salient political issues such as climate change present opportunities to reinforce existing social values, including the reliance on the fossil fuel industry.5 presents the messaging strategies across US states.In the case of Alaska with a high concentration of oil and gas sector, climate solutions was the most popular messaging strategy (94.5%) comparative to other messaging strategies in the state.Alaska is one of the most vulnerable states to the impact of climate change, leading to the reduction in potential incomes from for instance the Marine/ shing industry that will directly affect people's lives (Hollowed, 2020).Framing the industry as a climate solution then presents contradictory messaging.First, it reinforces the hegemony of the industry in the region despite its negative impacts on other industries in the state.Second, it uses this form of misinformation, that fossil fuels are a climate solution, to convince the public to support these industries in those locations that have relied heavily on this economic generating industry.In essence, it provides a social license to operate in the region despite its negative ecological impacts.The nal section of our analysis focuses on the contradictions between Facebook's policies and their implementation.Descriptively, our ndings raise three serious concerns the performance of Facebook's advertising policies.First is the inconsistency in labeling ads about climate change.The ads used in this research all had disclaimers.However, not all ads were correctly labeled from the beginning of their lifecycle.Facebook later added disclaimers or removed the ad; despite the removed ad already having impressions.Quantitatively, it is impossible to know how many ads that ran on the platform during this study were incorrectly labeled and then later picked up by Facebook.Nevertheless, this is not to say that this is an insigni cant nding and is an indicator of a more complex problem common on the platform (see also Sosnovik and Goga, 2021).
The second concern is the lack of further action against repeat offenders.According to the Ad Library, the entities included in this research had 182 ads (0.5% or 1% of the spend) removed for not being correctly labeled as ads about social issues, elections, or politics during the period of data collection.For example, Energy Citizens, run by the American Petroleum Institute, had 73 political and issue ads taken down by Facebook for not running with a disclaimer.In addition, after reviewing the data, Facebook informed the researchers on 31st July 2021 that it acted against the administrators of some pages belonging to the entities.These actions include, but are not limited to, taking down violating ads and restricting users' ability to advertise on the platform.However, Facebook did not specify which entities, what advert, and under what time frame they took these actions.Furthermore, Facebook supplies no further information as to when it will take any action beyond re-labeling ads or removing against repeat offenders of its advertising policies.This is a lack of transparency in its advertising policies and represents a marked difference to the explicit punishment for repeat offending that violates Facebook policies by a lay user.For instance, if a person posts a comment that is later marked as offensive and can potentially cause harm, this user is then blocked from posting for a time (Facebook, nd).Notably, this is not to say that fossil industries' business practices and solutions do not cause harm in the same way as offensive comments.On the contrary, McKie (2018, 2022) notes that the actions of organized obstruction actors undoubtedly contribute to delayed action on climate change, environmental harm and public health problems.Further research on the causal mechanisms between non-punishment for violations and infringements from corporations is integral to determining what factors may be impacting the lack of punishment.
The third concern is misinformation and misleading claims.Facebook's advertising policies prohibit ads containing misinformation and deceptive content.However, while Facebook may provide disclaimers for the ads posted by these obstruction organizations, they fail to apply the label and remove misleading claims.For example, the claim that gas is a green/clean/low-carbon energy source has been challenged by multiple sources (e.g., Stapczynski, et al. 2021).Furthermore, that climate change is a problem that can be xed using fossil fuels and individual consumers, contradicts the broader literature that emphasizes only a small cut in emissions can come with reshaping day-to-day consumer behavior (International Energy Agency, 2021).Given that these messages presented by the companies do not align with the science of climate change and proposed solutions, Facebook must reconsider how it de nes misinformation and misleading claims because such ads fall through the cracks keeping up to date with the evolving playbook of climate obstruction narratives.

Discussion & Conclusion
Between January 2020 and January 2021, Facebook received an estimated total of $9.60M from our sample of climate obstruction organizations.We answered three broad questions.First, we found four categories of ad messages used by the different actors during the election cycle: Community and Economy, Climate Solutions, Pragmatic Energy Mix, and Patriotic Energy Mix.These four categories represent a messaging strategy used by these obstruction actors to justify the continued use of fossil fuel on Facebook.For us, this falls into the realm of misinformation and misleading claims noting the existing evidence against fossil fuel use as part of a climate-focused energy transition.In turn, this is a violation of Facebook's policies.Thus, we add to the obstruction literature by applying this additional messaging strategy speci c to PR campaigns on social media and the importance of countering the evolving forms of misinformation and misleading claims posted online.
Second, we examined the differences in messaging strategies and targeted different audiences by location, age, and gender.Our ndings are similar to the existing literature on the receptivity to climate skepticism.For instance, ads target young males who are more likely to adopt skeptical views over their female counterparts.In addition, speci c narratives such as Pragmatic Energy Mix were more likely to be promoted by fossil fuel companies, going towards reinforcing the social norms that these industries are part of the solution.Moreover, our ndings indicate messages targeting swing states where oil and gas development are central political issues.We can argue then that these messages were targeted to speci c locations, ages, and gender to tap into existing beliefs on fossil fuel consumption and potentially shape how people respond to climate change and voter choices, e.g., choosing an electoral candidate that will resist a green energy transition.The hegemony of fossil fuels then continues to be promoted, utilizing targeted approaches to reach speci c audiences on Facebook.
Finally, we considered accountability, transparency, and their place for Facebook.Here we nd that Facebook fails to prevent posting misleading claims and misinformation from obstruction actors that violate their policies.For instance, the notion that fossil fuels are a green solution.This type of misinformation, or misleading claim, is subtler than climate denial messaging, such as 'climate change is a hoax,' and instead presents claims about climate change and climate solutions that contain signi cant omissions and misrepresentations.This is part of the evolving playbook of the discourses of delay used by climate obstruction organizations.Facebook then fails to adhere to its policies adequately.However, this may be problematic since Facebook sometimes outsource its fact-checking sources (Kennedy, 2021).
Taking these ndings together, for several interconnected reasons, we argue that climate obstruction organizations use Facebook to reinforce the naturalization of fossil capitalism and subsequent norms related to this political-economic system.First, using our typology of narratives can tap into these societal values through targeted and funded advertisements climate obstruction actors and social media platforms such as Facebook promote commodities in the pursuit of pro t: fossil fuels and advertising revenue, respectively.Fossil industry actors and Facebook are 'private property owners' invested in increasing economic growth and keeping 'business as usual.' Second, we cannot deny the impact of fossil fuel and social media owners attempting to in uence political outcomes through lobbying and political donations (e.g., Ard et al., 2018;Petrova et al., 2017).
Thus, unequal access to the political sphere and input into policy development from fossil industries and social media owners.This means policy-decision making is initially made with corporate interests in mind.Even though the public can express interest and opinions on potential policies through Facebook, targeted ads that appeal to the publics values may strengthen corporate and state interests in uencing voter choices that reinforce fossil fuel use.That is, Facebook plays a distinct role in potentially in uencing the collective political decisions and voter choices of the masses.While Facebook provides, on the surface, a space to critically exchange knowledge on social and political issues, it reproduces echo chambers using paid for targeted advertising permits allows these obstruction organizations to disseminate discourses of delay to those more likely receptive to this messaging.
Third, the failure to effectively monitor misleading claims or misinformation from obstruction actors that advertise on this platform reinforces existing dominant social norms that accelerate the ecological crisis.
However, this misinformation or misleading claims also fall under greenwashing.This can most clearly be seen in the presentation and promotion of fossil gas as a climate solution.This misinforms the public about the fossil-based nature of gas and misleads consumers about the climate actions companies take.
Instead, this effort indicates a 'climate-friendly' rebranding of business as usual that appropriates the proclimate values of the public.Moreover, these misinformed claims around fossil gas are subsequently used in direct policy lobbying to argue for an expansion on fossil gas in clear contrast with the recommendations of the IEA, again under the guise of being a 'climate solution' proposed by pro-climate organizations.
Taking these conclusions together, the optimistic assumption that social media is an ideal form of the public sphere is questionable.This is because signi cant data manipulation and targeted messaging for branding and political purposes highlights the limitations of knowledge exchange free from state and corporate interests.Moreover, while Facebook implements policies that should act as misinformation safeguards, they pro t from ad campaigns and diverting responsibility for any political outcomes to the public makes us question if Facebook is simply a space for social communication or has its own interests?The very purpose of what social media could achieve in a democracy has vanished or did not exist so far that capital forces have used it to facilitate and shape the public view on fossil fuel use.This means that obstruction actors can use messaging that appeal to people's everyday life to mitigate their concerns over climate change and transform the perceptions of these industries as supporting solutions to the problem, continuing business as usual.This practice can be considered greenwashing, where the public see images of an industry that is part of the environmentally positive solution to the climate crisis.
As greenwashing is in the eye of the beholder, the response is to potentially accept this solution and reinforces the hegemony of fossil fuels.
Ultimately, we cannot ascertain the direct impact of these Facebook PR campaigns on both support for climate action or inaction and voter decisions during the election year.Nevertheless, we must continue to monitor and closely re ect on the fossil industries evolving playbook, and the role of 'Facebook' as a favoured source of public communication which clearly violates its own policies or presents messages that may undermine appropriate action on climate change, playing a role in shaping publics views and beliefs on the solutions to the crisis.The American Petroleum Institute Ad https://www.facebook.com/ads/library/?id=223481808657538 Figure 5 Figure 1

Figure 3 Williams
Figure 3

Table 2
Categories, summary, number of ads, percentage of sample, spend, impressions, examples

Table 2
Category of narratives, by organizations, total spend, no impressions

Table 3
(Auxier and Anderson, 2021)n of ads separated by age, gender, spend and number of impressions.Overall, more males than females saw the ads across every age group.An exception to this was those in the 65 + group, with slightly more females than males viewing these ads.The most signi cant difference between views by males and females was found in the 25-34 year-old category, with more males viewing ads than women in this age bracket.This suggests the oil and gas industry and related obstructionKrange et al., 2018).Overall, these ndings bare some similarities with the existing literature, although we note the difference to that of McCright and Dunlap for instance, may re ect that the population accessing social media are more likely in the age ranges of 30-49 years old in the USA(Auxier and Anderson, 2021).
McCright and Dunlap, 2011;g climate science disinformation targets ads in this category to males between 25 and 34.The targeted messaging aligns with the existing evidence that documents the relationship between gender and climate skepticism.For instance,Ojala (2015)assessed that young men are more likely to adopt skeptic narrative techniques than their female counterparts.However, it does not align with ndings that skepticism increases with age more broadly (e.g.,McCright and Dunlap, 2011;

Table 4
Distribution of Ads by Region

Table 5
State and Messaging Categories.