Algorithmic accountability on social media platforms in the context of alcohol‐related health behavior change
Alex M. Russell, Brandon G. Bergman, Jason B. Colditz, Philip M. Massey
Abstract
Montag et al. [1] challenged social media (SM) companies to ethically approach digital media platform design decisions and consider their capacity to promote healthy platform engagement. We agree, and here, expand on this notion as it relates to algorithmic accountability in the context of the relationship between SM engagement and harmful health behaviors. Here, we focus on drinking-related behaviors; however, our argument can be applied to numerous health behaviors, such as vaccination and tobacco use. SM use influences health behaviors, including alcohol use [2]. Exposure to alcohol-related content on SM is common, and the majority of alcohol-related posts portray drinking positively, often normalizing hazardous drinking behaviors [3-5]. Moreover, engagement with pro-drinking SM content (posting, liking, commenting, and viewing) is associated with increased alcohol consumption [2]. Taken together, data suggest online alcohol-related SM experiences are associated with real world harmful and hazardous drinking [6-8]. At the same time, many individuals who engage in hazardous drinking and who are more likely to engage with pro-drinking SM content will make a decision to quit or reduce drinking during their lifetime [9, 10]. Although algorithms used by SM platforms to determine what content is made visible to users remain opaque, these algorithms are designed to enhance user attention and prolong platform engagement, relying on past user activity as an indicator for the type of content likely to capture users' attention [11]. Although the effects of engagement with pro-alcohol content on SM companies' algorithms—and its downstream effects on likelihood of being exposed to pro-alcohol posts—are unknown, research has shown that digital algorithms impact behavior [12, 13]. Therefore, it is plausible the effects of engaging with alcohol content on subsequent drinking behavior may be mediated, in part, by algorithm-driven exposure to pro-alcohol content. Insomuch as SM content algorithms account for users' past online behaviors (e.g. engaging with pro-alcohol posts), the content users continue to see would not reflect nascent changes in users' attitudes toward alcohol use (wanting to quit or reduce drinking). This may hinder efforts to change alcohol use behavior. There is a need to design and test whether system-level strategies for decreasing exposure to and mitigating the effects of pro-drinking SM content (e.g. giving users' control to filter out specific types of content) can be of benefit for active SM users who are contemplating or making attempts to quit or reduce their drinking. Additionally, individuals in recovery from alcohol and other drug use disorders are using SM platforms for digital recovery support [14-17]. Efforts to increase exposure to this pro-recovery SM content may prove beneficial among people who are experiencing alcohol problems or contemplating change. In all, we agree with Montag et al. [1], to reduce the harmful effects of SM use on health outcomes and harness the strengths for SM to promote positive health behavior change, SM companies play a most vital role. Specifically, we encourage SM platforms to work with academic researchers and community partners to tailor algorithms to prioritize the monitoring of alcohol-related content, to foster an environment that serves to decrease rather than increase alcohol-associated risks. B.G.B. was supported by the National Institute on Alcohol Abuse and Alcoholism of the National Institutes of Health under award number K23AA025707. P.M.M. was supported by the National Cancer Institute of the National Institutes of Health under award number R01CA229324. The content of this manuscript is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. None. Alex M Russell: Conceptualization; writing original draft. Brandon Bergman: Conceptualization; reviewing/editing original draft. Jason B. Colditz: Conceptualization; reviewing/editing original draft. Philip M Massey: Conceptualization; reviewing/editing original draft.