Companies inadvertently fund online misinformation despite consumer backlash

The proliferation of online misinformation poses significant societal risks, contributing to issues like increased COVID-19 fatalities, exacerbation of the climate crisis, and political discord. A key driver behind this is the financial motivation for creating and disseminating misinformation. Misinformation websites, often mimicking legitimate news outlets, thrive on advertising revenue. While anecdotal evidence and media reports have suggested the role of companies and digital platforms in financially supporting these outlets, a systematic understanding of the issue remains lacking. Existing research has primarily focused on user-level interventions to mitigate the demand for misinformation, neglecting the crucial supply-side aspect—the financial incentives driving misinformation production. This study aims to bridge this gap by exploring how advertising revenue sustains misinformation websites and examining the potential for interventions to curb this funding mechanism. The research uses a combination of descriptive analysis of large-scale datasets and experimental evidence from consumer and decision-maker surveys to examine the financial ecosystem that supports misinformation and to identify potential solutions that leverage information-based interventions to discourage the practice.

Prior research has largely focused on 'demand-side' interventions to counter misinformation, such as fact-checking, crowd-sourced labels, and nudges encouraging users to share accurate content. While these efforts are valuable, they have shown limited success in significantly altering news consumption patterns or reducing misperceptions. This study complements this existing literature by focusing on the 'supply-side' – the financial incentives driving the production and distribution of misinformation. Some academic work has proposed supply-side policies to address harmful content, and digital platforms have attempted to reduce advertising revenue flowing to misinformation websites. However, advertising from major companies continues to appear on these sites, raising questions about the effectiveness of current strategies and the need for a more comprehensive approach. This paper builds upon previous analyses of the misinformation ecosystem and programmatic advertising, offering a large-scale, multi-year perspective on the interplay between advertisers, digital platforms, and misinformation outlets.

The study employs a mixed-methods approach combining descriptive and experimental analyses. Descriptive analysis leverages a unique, large-scale dataset created by combining information on websites publishing misinformation with advertising activity data from Oracle's Moat Pro platform over three years (2019-2021). This dataset encompasses 5,485 websites (including 1,276 misinformation sites), 42,595 unique advertisers, and 9,539,847 instances of companies advertising on news websites. Misinformation websites were identified using data from NewsGuard, the Global Disinformation Index (GDI), and previous research. The study examines the prevalence of advertising on misinformation websites across various industries and investigates the role of digital advertising platforms in amplifying this phenomenon. To assess consumer responses, a survey experiment with a US population sample was conducted. Participants were randomly assigned to different information treatments, varying the information provided about companies' and platforms' roles in financing misinformation. Outcomes included changes in gift card choices (measuring 'exit' behavior) and clicks on online petitions (measuring 'voice' behavior). A separate survey of executives and managers (alumni of executive education programs) was also conducted to understand decision-makers' knowledge and preferences regarding their companies' advertising on misinformation websites. This survey involved an information provision experiment, informing participants about the role of digital advertising platforms and measuring changes in their demand for platform-based solutions to avoid such advertising.

The descriptive analysis reveals that a substantial portion of misinformation websites (74.5%) are monetized through advertising, with companies across numerous industries inadvertently funding them. The use of digital advertising platforms significantly amplifies this financing, with companies using these platforms being approximately ten times more likely to appear on misinformation websites. The consumer experiment demonstrates that consumers significantly reduce their demand for companies advertising on misinformation sites. This backlash persists even when consumers are aware of the platforms' role in advertising placement. The experiment also shows that providing company rankings based on advertising intensity on misinformation websites further reduces consumer demand for companies with higher rankings (i.e., more advertising on misinformation). The survey of decision-makers reveals a significant lack of awareness regarding their companies' involvement in financing misinformation. While most decision-makers prefer to avoid advertising on such sites, they often underestimate their own company's involvement and the role of platforms. However, when informed about the platforms' role, those uncertain about their company's involvement show a significantly increased demand for platform-based solutions to prevent future advertising on misinformation sites.

The findings directly address the research question by demonstrating the pervasive nature of inadvertent financing of misinformation through advertising and the significant consumer backlash this can trigger. The results highlight the substantial disconnect between decision-makers' preferences and their actions due to a lack of awareness and transparency. The study's significance lies in its identification of low-cost, scalable interventions to curb this funding mechanism. The substantial consumer response to information about company advertising on misinformation sites suggests that increased transparency could significantly reduce the financial incentives for producing misinformation. This addresses a key gap in current strategies focused solely on user-level interventions.

This study reveals the widespread and largely unintentional financing of online misinformation through advertising. Two key interventions are proposed: improving transparency for advertisers regarding their ad placement and enhancing transparency for consumers about which companies advertise on misinformation sites. These interventions, which could involve simple disclosures or company rankings, offer a cost-effective way to leverage market mechanisms to reduce the supply of online misinformation. Future research should evaluate the long-term effectiveness of these interventions in the field, examine the responses of companies under conditions of accurate beliefs, and investigate the incentives of digital advertising platforms to adopt these proposed solutions.

The study's consumer experiment focuses on a US sample, limiting the generalizability of findings to other countries or cultural contexts. The decision-maker survey relied on alumni of executive education programs, potentially biasing the sample towards companies with a higher focus on corporate social responsibility. Subsample analyses based on prior beliefs about company involvement in financing misinformation had relatively small sample sizes, which could impact the robustness of these findings. The study's time frame (2019-2021) may not fully capture the evolving landscape of misinformation and the technological advances, such as the rise of generative AI.