Claims about the harms of social media are everywhere. Many are taken as fact. In the face of this certainty, it is useful to remember that it was once taken for granted that serial dramas on the radio, comic books, going to the cinema, and arcade games were all once considered to be undeniably harmful, particularly for the youth. This is not to discount the legitimate interests of the public and policy makers to understand whether and how social media influence well-being. Locating the latest and most accurate research information is challenging. Evidence on the topic is endlessly updated, reviewed, and debated by researchers from fields ranging from psychology and communication through education and computer science to neuroscience and economics.
To understand this new form of an old controversy and to summarize complex and scattered academic evidence, this paper will use a myth versus warranted claim structure. This paper will present 10 myths about the harms of social media. Each myth is directly quoted or paraphrased from public discourse, including op-eds, podcasts, claims of politicians, and news headlines. In response to each myth, a warranted claim supported by peer-reviewed research will be offered. Meta-analyses, studies with large or representative samples, and studies using longitudinal methods will be prioritized as sources for supporting and justifying each warranted claim.
Social media are platforms that enable searchable and publicly distributed content, including text, photos, memes, and videos. Social media are often identified by a branded platform name (eg, Instagram). Users create content and this content is transmitted in a decentralized fashion. Users—to varying degrees depending on the platform—decide what is shared and reshared. Platforms vary considerably as to whether it is normative for users to share content, or whether the content is typically distributed by prominent accounts or users (eg, influencers or content creators). Meta-analyses suggest that social network sites, particularly Facebook, are the most studied social media. Thus, most meta-analytic claims are drawn from evidence from decades of research on Facebook use, primarily among American college students . Social media platforms that distribute video content (eg, TikTok or YouTube) have recently become quite popular, but are less well-represented in the research on well-being. Whenever possible, this viewpoint will focus on time using social media, rather than screen time generally, which often includes TV, gaming, and internet use.
In this paper, well-being will be defined as including both eudaemonic and hedonic well-being. While the former focuses on meaning, connection, and life purpose, the latter focuses on pleasure, enjoyment, and entertainment. Both well-being and ill-being will be examined. Well-being is typically measured in social media research as life satisfaction, positive emotions, and the absence of negative emotions, while ill-being is typically measured as the presence of mental health symptomology (eg, anxiety, depression, or loneliness).
This paper will focus on the effects of social media on the well-being of users as an aggregate. Social science research reports effect sizes, typically expressed as a correlation or mean difference. By nature, the goal of quantitative social science is to document the association between 2 (or more) variables, not to document the singular experience of each participant in the study. Thus, trends in the data, reported as aggregate effects, are not certainties that equally apply to all people. No association between 2 variables is equally true for every single person in the sample. Thus, it can be true that the warranted claims are accurate for the whole and that they do not perfectly align with every individual’s experience.