When Facebook expanded beyond college campuses in 2006 and Twitter opened to the public the same year, the dominant narrative was utopian. Technology would flatten hierarchies, democratise information, and let ordinary people build genuine global communities. In some ways, this happened. Activists organised protests that toppled governments. Long-lost relatives reconnected. Communities of people with rare diseases, obscure interests, and minority identities found each other across continents. The Pew Research Center consistently finds that majorities of social media users report staying in touch with friends and family as a primary benefit.
But the evidence accumulated over the past decade is more complicated than the utopian framing suggested — and more complicated than the dystopian counter-narrative suggests too. Social media has not simply expanded human connection. It has restructured it, introducing dynamics that were genuinely novel in human social life. Understanding those dynamics clearly, without either technophobic panic or techno-optimism, is essential for navigating platforms that now mediate a significant portion of daily social life for billions of people.
What the Research Actually Shows About Wellbeing
The relationship between social media use and psychological wellbeing is one of the most studied questions in contemporary psychology — and one of the most contentious. Headlines alternate between "social media causes depression" and "social media effects are minimal." The honest answer is more nuanced than either.
The correlation versus causation problem. Most studies showing negative relationships between social media and wellbeing are correlational, not causal. People who are already depressed, lonely, or anxious may use social media more — or differently — than people who are not. A landmark 2019 pre-registered study by Orben, Dienlin, and Przybylski published in PNAS, using large-scale representative panel data, found that social media's effect on adolescent wellbeing was statistically real but small, and that the effects were "nuanced, small at best, reciprocal over time, gender specific, and contingent on analytic methods." A companion study in Nature Human Behaviour found the effect size comparable to wearing glasses — present but not large.
The crucial distinction in this research is between passive and active use. Passive scrolling — consuming others' content without posting, commenting, or messaging — is more consistently associated with lower wellbeing than active communication and creation. This distinction runs through much of the research and has important practical implications that most public discourse ignores.
Where effects are more robust. Despite the methodological debates, some findings replicate more consistently. Upward social comparison — comparing yourself unfavourably to others' curated highlights — is reliably associated with lower self-esteem and increased anxiety. Platforms that optimise for visual content and engagement create environments that differ fundamentally from offline social contexts where comparison is more naturally balanced.
The American Psychological Association's 2023 health advisory on adolescent social media use identified social comparison and cyberbullying as the two mechanisms with the strongest evidence for harm, particularly among girls aged 11–13. Psychologist Jonathan Haidt, in his 2024 book The Anxious Generation, argued more forcefully that smartphone-mediated social life has caused a genuine mental health crisis among adolescents — a claim that has been both supported and contested within the research community, with the debate centring on causation versus correlation rather than on whether something has gone wrong.
What is not seriously disputed is that adolescent mental health indicators — anxiety, depression, loneliness — have deteriorated in the smartphone era across most English-speaking countries, and that the deterioration has been more severe among girls. The scientific debate is about how much of this deterioration is attributable to social media specifically, and through which mechanisms.
The Transformation of Friendship
Social media has changed not just how many connections people have but the texture of those connections. The sociologist Robin Dunbar's research on social networks suggests humans have cognitive capacity for approximately 150 meaningful relationships — his famous Dunbar's number. Social media enables managing a vastly larger network of weak ties, but the question of whether those weak ties enrich or dilute the experience of close friendship is genuinely contested.
The weak tie advantage. Weak ties — acquaintances, former colleagues, distant relatives — have long been understood to have economic and social value because they expose you to information and opportunities outside your immediate circle. A landmark 2022 study published in Science, analysing 20 million LinkedIn users, found that weaker professional ties were more valuable for job mobility than strong ties, because they provided access to more diverse information networks. Maintaining these networks at scale, which social media enables, provides genuine value of a kind that was practically impossible before the internet.
The maintenance paradox. However, maintaining large networks on social media may come at a cost to deep friendship. Dunbar's research has consistently found that online interaction does not substitute for the face-to-face time that sustains close relationships. Social media can maintain weak ties effectively but is less effective at strengthening or maintaining the inner circle of intimate friendships, which require shared physical presence, accumulated in-person experience, and the kind of unstructured time that is difficult to replicate on a platform optimised for brief, broadcast-style interactions.
The maintenance paradox is that social media makes it easier to manage many relationships while potentially crowding out the time investment needed to deepen a few of them. You can know what 400 people had for lunch. You may have less time to eat lunch with the people you actually care about.
Identity and Self-Presentation in the Platform Age
Social media has introduced a new dimension to human identity development: the continuous public performance of self. Sociologist Erving Goffman's 1959 theory of self-presentation — the idea that social life involves managing impressions through "front stage" and "back stage" behaviour — took on new meaning in the era of Instagram, TikTok, and LinkedIn. On these platforms, front-stage performance is persistent, searchable, and quantified through likes and follower counts in ways that were impossible in offline life.
The quantification of social approval. Research on Instagram specifically has documented a phenomenon researchers call "self-objectification" — viewing and evaluating oneself from an external perspective, as if through an audience's eyes. Studies published in Computers in Human Behavior have found that heavy Instagram use correlates with increased self-objectification and body image concerns, particularly among young women. The visual nature of the platform and its emphasis on appearance make it distinctively risky in this respect compared to text-based platforms.
The authenticity counter-movement. Platforms and users have responded with authenticity movements — "real" posts, unfiltered selfies, vulnerability sharing. BeReal, acquired by French mobile gaming company Voodoo in 2024, built its entire product proposition around notifying users simultaneously to share unfiltered images, precisely as a counter-reaction to Instagram's curated aesthetic. Whether authenticity performance is genuinely different from curation performance is a philosophical question the platforms have not resolved. What is clear is that users are sophisticated enough to recognise the tension and are actively seeking formats that feel less constructed.
What Social Media Does Better Than Anything Else
Despite all the documented risks, social media's capacity for community formation remains one of its most genuinely valuable and underappreciated contributions to human culture.
Rare disease and health communities. For people with uncommon medical conditions, disabilities, or stigmatised experiences, online communities provide social support that is simply not available locally. Pew Research on health information online has consistently found that peer health communities are a primary information source for people managing chronic conditions. The social support these communities provide — practical, emotional, and informational — has measurable health benefits for members. Someone with a rare autoimmune condition living in a small town may have no local peers who understand their experience. Online, they can find thousands.
Cultural preservation and diaspora connection. Social media has enabled speakers of endangered languages, practitioners of traditional crafts, and members of diaspora communities to find each other and sustain cultural practices that would otherwise be lost. Indigenous language communities, traditional music traditions, and folk craft networks have all benefited from the ability to connect practitioners globally and attract younger members who might not have encountered these traditions in their immediate environment.
Political organisation and accountability. Social media has enabled grassroots political organisation at scales and speeds previously impossible, from the Arab Spring to climate activism to mutual aid networks during crises. Journalists use social media to surface stories, amplify sources, and distribute accountability reporting beyond traditional institutional channels. These are genuine contributions to public life that the technology's critics sometimes underweight.
The Attention Economy and Its Costs
The dominant business model of social media platforms — monetising attention through advertising — has cultural consequences that extend beyond individual wellbeing. Platforms optimise for engagement, and the content that generates the most engagement is often emotionally activating, outrage-inducing, and divisive rather than calm, nuanced, or prosocial.
The Facebook Files, published by the Wall Street Journal in 2021, revealed internal research showing that Facebook's own teams had identified serious harms from their algorithms — including amplification of misinformation and content that triggered anger — and that these findings did not consistently drive product changes. The gap between what platforms know about algorithmic harm and what they choose to do about it remains one of the central unresolved tensions in the social media era.
This is not an individual user problem. It is a structural product design problem. No amount of personal discipline or media literacy fully compensates for platforms that are engineered to capture and exploit attention. The incentive structure that produces these outcomes cannot be resolved at the individual level.
What We Can Do
The research on social media is not uniformly negative, and the appropriate response is not abstention but intentionality. Several evidence-based strategies have emerged from the research.
Shift from passive to active use. The single most consistent finding across studies of social media and wellbeing is that passive scrolling is more harmful than active communication. Deliberately shifting time from consuming others' content toward posting, messaging, commenting, and creating uses the platforms in ways closer to their genuine social function — and produces measurably different outcomes in most studies.
Curate your feed deliberately. Algorithmic feeds respond to the engagement signals you provide, often unconsciously. Deliberately unfollowing or muting accounts that reliably trigger comparison or negative emotion — even when the content is objectively high-quality — changes your information environment. Research from the Stanford Human-Computer Interaction Group has shown that even modest changes to feed composition can shift the emotional tone of social media experiences significantly.
Protect time for in-person relationships. If Dunbar's research is correct that online interaction does not substitute for physical presence in sustaining close relationships, then the practical implication is that social media time should not consistently displace in-person time. The relationships that matter most require the investment that screens cannot fully provide.
At the policy level: Individual strategies help with individual experiences, but the structural problems of social media — algorithmic amplification of outrage, inadequate protection for minors, data exploitation — require regulatory responses. The EU's Digital Services Act (2022) and the UK's Online Safety Act (2023) represent early attempts to create accountability frameworks for platform harms. Neither is sufficient on its own, and the US has lagged in enacting comparable legislation despite extensive congressional hearings. Effective regulation will need to balance free expression, innovation, and genuine harm prevention — a balance that no jurisdiction has yet achieved.
The Honest Balance Sheet
Social media is neither the connection utopia its early proponents promised nor the unmitigated harm its fiercest critics claim. It is a powerful communication technology whose effects depend heavily on how it is designed, how it is used, and what governance structures surround it.
The gains are real: community for the isolated, information for the curious, platform for voices that would otherwise go unheard, connection across distance and difference. The costs are also real: comparison for the vulnerable, outrage amplification for the politically engaged, structural incentives that prioritise engagement over truth, and documented harms to adolescent mental health through mechanisms that researchers are still fully working out.
Navigating it wisely requires understanding both sides of that ledger with honesty — and demanding that platforms, regulators, and individual users each take their portion of responsibility for the culture these technologies are actively building.
How has social media changed your experience of friendship and community — for better or worse? Share your perspective in the comments below.