Originally published in January 2018 and comprehensively updated/edited in June 2026.
Why did so many people feel anxious and unsafe in the late 2010s? This essay explores the neuroscience, psychology, and social forces behind a widespread but hard-to-name unease.
By 2018, millions of people were waking up, reaching for their phones before they had gotten out of bed, and encountering conflict before they had eaten breakfast. The first scroll of the morning might bring a political argument, a breaking news alert, an outrage cycle already days old but still burning, or a reminder that someone they used to trust — a politician, a journalist, an institution — had apparently been lying. Then they would get up, go to work, and feel vaguely terrible without being entirely sure why.
That experience — the low-grade dread, the sense of being permanently on edge — was not unique to any one demographic or political tribe. It was, by most available measures, widespread. And it was peculiar, because by the conventional metrics societies use to assess wellbeing, things looked manageable for many people. The economy, by conventional measures, was functioning. Crime, by historical standards, was not at a peak. No major war was consuming the countries where this anxiety was most acutely felt. And yet conversations at dinner tables turned hostile. Trust in institutions fell to historic lows. Something felt profoundly wrong, and most people couldn’t name it with any precision.
Something was happening beneath the surface of social life — something that science was only beginning to understand and that most people had no language to describe. The human nervous system, shaped over millions of years to detect danger in the physical world, was struggling to process a new kind of environment. The threats were real, but they were abstract, diffuse, and relentless. And the very tools people were using to connect with each other were, in many cases, making things worse.
Understanding why humans felt unsafe without understanding why requires looking at several interlocking forces: the neuroscience of threat detection, the sociology of eroding trust, the psychology of social belonging, and the specific technological conditions that came to define life in the digital age.
The Biology of a Nervous System Under Pressure
To understand what was happening to people emotionally and psychologically in this period, it helps to start with some basic biology. The human brain did not evolve to process the information environment of the early 21st century. It evolved to navigate small social groups, physical landscapes, and threats that were concrete and immediate.
How the Brain Processes Threat
The amygdala — a small, almond-shaped structure deep in the brain — functions as something like an alarm system. It continuously scans the environment for signals of danger and, when it finds them, triggers a cascade of physiological responses: elevated heart rate, heightened alertness, the release of stress hormones like cortisol and adrenaline. This system is extraordinarily fast and operates largely outside conscious awareness. By the time you’re consciously aware that something startled you, your amygdala has already sent the signal.
The problem is that this system was calibrated for a world that no longer exists for most people in industrialized societies. In the ancestral environment, a threat detected was usually a threat nearby — a predator, a rival, a storm. The stress response was adaptive because it prepared the body to act. The threat would resolve, one way or another, and the nervous system would return to baseline.
In the modern world, and especially in the social media age, threats are rarely this simple. They are abstract (financial instability, geopolitical uncertainty), statistically distant (crime rates you read about but may never experience), and — critically — unresolvable in the immediate sense. You cannot outrun a news cycle or fight a tweet. The body, however, doesn’t fully distinguish between types of threats. Research by neuroscientist Joseph LeDoux and others has shown that the amygdala responds to symbolic and social threats in ways that parallel its response to physical ones. Social rejection, status loss, and perceived exclusion activate many of the same neural circuits as physical danger.
This meant that by 2018, millions of people were walking around with nervous systems that were chronically activated — not by predators, but by smartphones. The constant stream of alarming news, social comparison, and interpersonal conflict that characterized social media use was keeping threat-detection systems in a low-grade state of alert. The body experienced this as stress, unease, and a persistent sense that something was wrong. The mind, searching for an explanation, often found one — but not always the right one.
Persistent anxiety without a clear source rarely stays that way for long. The mind is an explanation-generating machine — it searches for causes, patterns, and narratives that can convert diffuse unease into something concrete and actionable. Sometimes those explanations are accurate. Often they are not. But psychologically, a wrong explanation can still feel preferable to no explanation at all. A named enemy, however imperfectly identified, feels easier to confront than a threat that cannot be clearly seen. Researchers studying radicalization and conspiracy thinking have found that susceptibility is not primarily a function of low intelligence or gullibility. It is more closely tied to something far more basic: a prior state of anxiety and distrust that makes explanatory frameworks feel urgently necessary. Conspiracy theories, culture-war narratives, and moral panics all offer the same psychological service — they convert vague, formless dread into a specific story with a villain. The anxiety doesn’t disappear, but it becomes legible. And legible fear is, in a strange way, easier to live with than fear without a name.
Chronic Stress and Its Hidden Costs
Chronic activation of the stress response is not benign. Sustained elevated cortisol is associated with impaired immune function, disrupted sleep, reduced capacity for complex reasoning, and heightened emotional reactivity. Research by psychologist Sheldon Cohen at Carnegie Mellon University has demonstrated links between chronic social stress and susceptibility to illness. Work in the field of allostatic load — a measure of cumulative wear on the body from repeated stress — suggests that prolonged low-level stress takes a measurable biological toll even when no acute trauma is present.
What made this period particularly significant is that the stress was largely invisible. People weren’t being chased. They weren’t starving. By the metrics societies had always used to assess wellbeing, things looked manageable. But the internal experience told a different story, and the gap between external circumstances and internal state was itself a source of confusion and distress. People felt bad and didn’t fully understand why, which often made them feel worse.
The Collapse of Institutional Trust
Biology alone doesn’t explain the social anxiety of this era. The other major factor was a steep and sustained decline in institutional trust — in government, media, science, organized religion, corporations, and financial systems. This wasn’t simply a matter of perception. Many institutions had provided genuine cause for eroded trust: the 2008 financial crisis, in which banking failures devastated ordinary households while accountability remained limited; the documented failures of intelligence agencies in the lead-up to the Iraq War; the sexual abuse scandals that swept through the Catholic Church, elite universities, and Hollywood; and the revelation through platforms like WikiLeaks and reporting on NSA surveillance that governments were conducting far-reaching covert monitoring of their own citizens.
By the time smartphones became ubiquitous and social media became a central feature of public life, a large portion of the population in the United States and much of the Western world had become deeply skeptical of the institutions that previous generations had largely taken for granted.
Why Trust Matters More Than We Think
Social scientists have long understood that trust functions as a kind of infrastructure for social life. Francis Fukuyama’s influential work in the 1990s argued that trust between citizens, and between citizens and institutions, reduces the transaction costs of social cooperation — it makes it possible for people to do business, form relationships, and navigate conflict without needing to verify everything from scratch. When trust deteriorates, people expend far more cognitive and emotional energy on vigilance. They have to be more careful, more suspicious, more self-protective.
Social psychologist Roderick Kramer’s research on institutional trust documented one specific consequence of its erosion that is worth naming: what he calls “the sinister attribution error” — the tendency to interpret ambiguous behavior by others as hostile or manipulative. When trust is low, people see threat everywhere, even when it isn’t there. This closely mirrors what many observers noted about public discourse in 2017 and 2018: a pervasive assumption of bad faith that made ordinary disagreement feel like attack, and made good-faith disagreement nearly impossible to perform credibly.
The Gallup Organization’s tracking polls showed that American trust in major institutions — Congress, the presidency, newspapers, banks, organized religion — had been declining for decades, but reached particularly low points in this period. Similar trends were documented by the Edelman Trust Barometer, an annual survey of trust in institutions across dozens of countries, which in 2018 reported a continued erosion of trust in government, media, NGOs, and business worldwide.
Social Media and the Architecture of Outrage
If chronic stress and institutional distrust were the kindling, social media platforms were, for many researchers, something very close to a match.
The platforms that dominated public life in 2018 — Facebook, Twitter, YouTube — had been designed with engagement as their primary optimization target. Engagement, in the technical sense used by these platforms, means time spent, clicks, shares, and comments. And as researchers and former platform employees began documenting publicly, content that provoked strong negative emotions — particularly outrage and fear — reliably generated more engagement than content that was neutral or positive.
Algorithmic Amplification of Emotional Content
Psychologist Jonathan Haidt, together with technologist Jaron Lanier and others, developed what became an influential critique of social media’s effects on collective psychology. Their argument, supported by mounting research, was that the algorithmic curation of content systematically amplified emotional and divisive material, not because the platforms intended harm, but because outrage and fear were engagement gold.
A study published in the journal Science in 2018 by Soroush Vosoughi, Deb Roy, and Sinan Aral at MIT found that false news stories spread significantly faster and farther on Twitter than true ones, and that the primary driver of this effect was human sharing behavior — people sharing emotional content without necessarily verifying it first. The emotions most associated with viral false news were surprise, fear, and disgust. This finding suggested that the social media ecosystem had created conditions in which emotionally activating, often inaccurate content would systematically outcompete calmer, more accurate content for human attention.
For individuals, this meant that the information environment they encountered daily was skewed — not through any deliberate conspiracy, but through the mechanics of algorithmic selection — toward content that kept their nervous systems activated. For society, it meant that the shared information environment, once constructed primarily by professional journalists operating under editorial norms, had been partially replaced by a system with very different properties. The new system rewarded provocation, penalized nuance, and created what Haidt would later describe as a kind of “puberty” for democracy — an adolescent phase of heightened emotion, tribalism, and impulsive reaction.
Social Comparison and the Well-Being Gap
A separate but related problem was the role social media played in reshaping how people evaluated their own lives. Psychologist Leon Festinger’s theory of social comparison, developed in the 1950s, holds that humans have a fundamental drive to evaluate their opinions and abilities by comparing themselves to others. In small social groups, this comparison set was limited to people one actually knew — neighbors, coworkers, family members. Social media expanded this comparison set to potentially millions of people, most of whom were presenting curated, aspirational versions of their lives.
Research by psychologists Jean Twenge and Keith Campbell, among many others, documented rising rates of anxiety, depression, and loneliness among heavy social media users, particularly among adolescents and young adults. Twenge’s book iGen, published in 2017, argued that the generation born after 1995 — the first to grow up with smartphones and social media as permanent features of childhood — was showing measurably different psychological profiles than prior generations: more anxious, more depressed, less likely to take risks, and lonelier despite being more connected. These findings were contested by some researchers who argued the effect sizes were modest or that the causal direction was unclear. But the debate itself helped surface the question that would occupy psychologists and public health researchers for years: what exactly was the relationship between the digital information environment and human psychological health?
The Loneliness Epidemic and the Hunger for Belonging
Running beneath all of this was a phenomenon that had been building quietly for decades: a rise in loneliness and social isolation that by the mid-2010s was being described by researchers as an epidemic.
Former U.S. Surgeon General Vivek Murthy, writing in the Harvard Business Review in 2017, described loneliness as a growing public health crisis. He cited research indicating that loneliness and social isolation were associated with a roughly 50 percent increase in the risk of premature death — comparable in magnitude to smoking 15 cigarettes a day. Social researcher Robert Putnam had documented the decline of civic and social participation in the United States in his 2000 book Bowling Alone, tracking the erosion of the community structures — bowling leagues, church groups, local political organizations — that had historically provided social connection and a sense of belonging.
By 2018, those trends had continued. Americans were living alone at historically high rates. The average number of close confidants Americans reported had declined sharply since the 1980s. And while digital connection had expanded dramatically, there was growing evidence that online interaction was a poor substitute for the face-to-face contact that human nervous systems seem to require.
Why Belonging Is Not Optional
Neuroscientist Matthew Lieberman’s research, summarized in his 2013 book Social: Why Our Brains Are Wired to Connect, made the case that social connection is not a luxury but a biological need, as fundamental to human functioning as food or sleep. His work using neuroimaging showed that social pain — the experience of rejection or exclusion — activates the same neural circuits as physical pain. Being left out, ignored, or excluded does not merely hurt in a metaphorical sense; it registers in the brain in ways that overlap with the experience of physical injury.
This helps explain something about the emotional texture of social media interaction in this period. The platforms were environments in which people sought connection and belonging, but the dynamics of those environments — the public nature of social evaluation, the numerical display of likes and followers, the ease with which conversations could turn hostile — made rejection and exclusion constant possibilities, and made them highly visible when they occurred. People were seeking belonging in spaces that were structurally prone to producing its opposite.
The Rise of Moral Conflict Without Shared Framework
One more element made this period distinctive: a profound fragmentation in the shared moral and factual frameworks that allow disagreement to be productive rather than tribal.
Social psychologist Jonathan Haidt’s moral foundations theory, developed over the previous two decades, proposed that human morality is built on several distinct foundations — care, fairness, loyalty, authority, sanctity, and liberty — and that individuals and groups weight these differently. His research suggested that political polarization in the United States, at least partly, reflected genuine differences in moral perception: people weren’t simply disagreeing about facts or policies, but about which values should take priority, and these differences felt visceral and absolute because they were grounded in deep moral intuitions.
At the same time, the shared factual landscape that debate requires was fragmenting. The rise of partisan media, the decline of dominant news gatekeepers, and the algorithmic sorting of social media feeds meant that Americans increasingly inhabited different information worlds. Research by political scientists showed that partisan selective exposure — the tendency to consume media that confirms existing beliefs — was associated with higher levels of affective polarization: not just disagreement about policy, but genuine dislike and distrust of political outgroups.
What emerged was a cultural climate in which moral disagreement could not be managed through normal civic processes because the shared premises those processes require — a common factual base, a degree of mutual good faith, agreed-upon rules for resolving disputes — had eroded. People felt that the social world was dangerous, not just because of crime or economic instability, but because fundamental questions about who belonged, what was true, and what was right felt urgently contested and unresolvable.
Making Sense of Diffuse Unease
What unites all of these threads — the chronically activated nervous system, the collapse of institutional trust, the algorithmic amplification of outrage, the loneliness epidemic, the moral fragmentation — is the way they combined to produce a form of anxiety that was real but difficult to locate.
When a threat is concrete and identifiable, human beings are remarkably capable of responding to it. What is harder, psychologically, is sustained uncertainty — the feeling that something is wrong without a clear target. Psychologist Mark Schaller’s work on what he calls the behavioral immune system suggests that humans have evolved cognitive systems that, under conditions of perceived threat, become more sensitive to anomaly and more prone to avoidance and disgust. These systems can be triggered by social and symbolic threats as readily as physical ones, and once triggered, they tend to make people more suspicious, more conformist, and more intolerant of difference.
This pattern was visible in the social dynamics of 2018 in ways that cut across political lines. The closing of ranks, the suspicion of outsiders, the hair-trigger response to perceived slights — these were not exclusively the property of any one political tribe. They were a species-level response to conditions of sustained uncertainty, playing out simultaneously across different groups, each experiencing the others’ anxiety and defensiveness as confirmation of their own threat perceptions.
Understanding this doesn’t resolve the disagreements or undo the real grievances that animated them. But it does suggest that a significant portion of the social conflict of this period was being amplified — not by the underlying issues alone, but by the psychological and neurological conditions under which those issues were being processed.
What Came After
The years following 2018 bore out many of the concerns that researchers had been raising. Mental health conditions continued to rise, particularly among young people. Political polarization in most developed democracies deepened rather than abated. The COVID-19 pandemic, which struck in 2020, both intensified existing social fragmentation and made the infrastructure of human connection — the in-person contact, the shared public spaces, the face-to-face rituals that sustain belonging — suddenly unavailable for large portions of the world’s population. The loneliness that had been growing for decades became acute and undeniable.
By the mid-2020s, some governments had begun to treat loneliness as a public health issue, with the United Kingdom and Japan appointing Ministers for Loneliness. Major health organizations formally recognized social isolation as a significant risk factor for mortality. The U.S. Surgeon General’s office released an advisory in 2023 describing an “epidemic of loneliness and isolation,” citing decades of research on the health consequences and calling for structural interventions. Researchers in the emerging field of social prescribing were exploring whether physicians could address health conditions in part by connecting patients to social activities and community organizations.
The technology industry also faced growing pressure — from regulators, researchers, and former employees — to reckon with the psychological consequences of platform design. Age restrictions on social media became law in several jurisdictions. Algorithmic transparency requirements were debated in the European Union and elsewhere. Whether these interventions would prove adequate to the scale of the problem remained genuinely uncertain.
The Deeper Question
Stepping back from the specific events and dynamics of this period, something larger comes into view. For virtually all of human history, danger arrived through direct experience. You saw the flood, felt the cold, encountered the hostile stranger. Your nervous system responded to something real and present, and when the threat passed, the response wound down. The knowledge that something dangerous had happened was inseparable from the physical experience of it.
Modern life broke that connection. By 2018, most of the threats that people in wealthy countries were most afraid of — economic collapse, political violence, social dissolution, pandemic — were things they encountered primarily through information. They read about them, watched them, scrolled past them. The nervous system received the signal — danger, threat, something is wrong — without receiving the resolution that direct experience would eventually provide. The storm either hits or it doesn’t. The news cycle about whether the storm might hit never ends.
This is, in a narrow sense, what made the anxiety of this era so difficult to locate and so hard to metabolize. It wasn’t that the threats weren’t real. Many of them were. It was that the primary medium through which people experienced those threats — a curated, algorithmically shaped, emotionally amplified information feed — was calibrated for engagement rather than accuracy, for activation rather than resolution.
The nervous system got all the alarm with none of the closure.
Understanding that distinction does not make the underlying problems disappear. But it reframes what kind of problem they are. The challenge is not simply one of fixing institutions, regulating platforms, or rebuilding trust — though all of those things may matter. It is also a perceptual challenge: learning to navigate an information environment that the human mind was not designed for, and developing some capacity to distinguish between the signal and the noise. That is harder than it sounds. It may be the defining difficulty of the era that followed.
