Are social media and a safety-first culture harmful to children?
The Anxious Generation, by Jonathan Haidt. 400pp. Allen Lane, £25; Penguin Random House, $30, March 26, 2024.
Review of The Anxious Generation: How the great rewiring of childhood is causing an epidemic of mental illness, by Jonathan Haidt. 400pp. Allen Lane, £25; Penguin Random House, $30, March 26, 2024.
There is nothing more alluring in polarized times than straightforward solutions to complicated questions. Much of the political and social turmoil of our time arises from the inclination to believe that overwhelming troubles can be resolved by partisanship. Authoritarianism reassures people who cannot figure out what to do and are relieved when someone else tells them. Such thinking has crept beyond the political to infuse daily life. Nuance entails uncertainty; in a confusing world it is easy to fall prey to almost any form of clarity.
Jonathan Haidt’s The Anxious Generation is a compendium of important and profound insights about contemporary childhood embedded in such wishful lucidity. His twinned basic propositions—that children should have less supervision and more free play, and that they should have less access to social media and some other parts of the internet—have a strong basis. It is likely that his sweeping simplifications will help to move forward much-needed social change; it is unfortunate that the impetus for that change is often grandiose and misleading statements, an endless succession of graphs and footnotes notwithstanding. The word sometimes seems not to be in his vocabulary; the key associated with the question mark seems not to work on his computer. He never lapses into the rhetoric of uncertainty that would serve truth. Nowhere does he refer to the incomprehensibility of social decay. Never does he express uncertainty that it is possible to know the causes of something as complex as the fluctuations in youth mental health, so his remarks allow for almost no contemplation of the exceptions to his propositions. Usually exceptions in aggregate outnumber the dominant narrative.
“Thrilling experiences have anti-phobic effects” is sometimes true, but there are many children who are traumatized by experiences that thrill others and whose overall sense of fear is heightened by rollercoasters rather than assuaged by them. “Like young trees exposed to wind, children who are routinely exposed to small risks grow up to become adults who can handle much larger risks without panicking.” Many children, exposed to risk or otherwise, do not grow into adults who can handle large risks without panicking. “Children’s minds are ‘expecting’ the challenges of the real world, which is embodied, synchronous, and one-to-one or one-to-several, within communities that endure.” If that is what children’s minds are expecting, they must be ill prepared for reality, because life has never been synchronous and communities do not tend to endure. “Those after-school hours were probably more valuable for social development and mental health than anything that happened in school (other than recess).” After-school play and recess are certainly important in children’s development, but are they more important than the socialization that is taught in the classroom or that emerges through the opening up of minds to the vast worlds of literature and mathematics?
Haidt’s book might have been called, “Make Childhood Great Again”.
Haidt’s arguments would be much stronger if they were not absolute. Some children who are brutalized during after-school play learn a sense of inferiority and self-hatred. The “all-encompassing, play-crushing power of safetyism” is evident in one elementary school because “administrators … don’t trust their students to play tag without adult guidance”. He has cherry-picked an extreme example, but children find ways to introduce play despite all the rules around them; indeed, they introduce bullying despite “safetyism”; and neither joy nor horror is “crushed”. It’s not that I question whether this movement has gone too far in many instances, but the sweeping generalizations erase reality. Haidt wants to convince us that the world is not so dangerous, so we should let our kids roam free, but he does not allow room for the argument that what he calls “safetyism” has its benefits, and that a centrist perspective might make more sense than a polarizing attack on flawed policies that should not be disposed of entirely. “All children are antifragile,” Haidt writes. All children aren’t anything, ever; many children are terribly fragile and do best when protected. He then comes down on the “worship of ‘safety’ above all else”—but most highly protective parents “worship” a great many other things besides. All these quotations are from a single chapter, but the same pattern occurs throughout the book.
The author’s work is reactionary, embedded in romanticization of the good old days and dismay at what modernity has wrought. This tendency—the defining characteristic of conservatism—which began undermining long-held ideas of social justice in the Reagan-Thatcher era, has since escalated into “Make America Great Again”, Donald Trump’s most successful slogan. Haidt’s book might have been called, “Make Childhood Great Again”. It nods to wokeism, but neglects to acknowledge that children themselves are enormously diverse, and that a solution to the “youth mental health crisis”, the very existence of which is hotly debated, is as chimerical as a “cure for cancer”. We have not actually shifted from “discover mode” to “defend mode”. We have shifted the balance in certain contexts, but we all move constantly from one to the other. As for Haidt’s reflections on matters such as the Apache Sunrise Ceremony, he acknowledges so little about the culture in which that ceremony arose that the phrase “noble savage” echoes.
Many parents are overprotective because, in a certain middle-class world, such conduct constitutes the best proof of love. Expressions of love change: my grandparents stuck to “spare the rod and spoil the child”; my parents and everyone they knew believed that children who never had a spanking were going to behave badly and would be unprepared for life; I do not hit my children. If you are out of step with your times, you may feel insecure about the depth of your love for your children—and your children may question it too. My grandparents’ treatment of my parents would seem harsh were it undertaken now, but today’s dominant mode of child-rearing would have marked a child as a peculiar outcast in 1930. Children perceive attitude. Abuse is always poisonous, but if children perceive their punishment as just they will accept it, and what appears to be just is a generational question. Furthermore, many children today are underprotected, especially those whose parents are struggling to support them and cannot afford the time that overprotection demands. As Haidt himself points out, children from such backgrounds tend to have worse mental health than their more prosperous, and usually whiter, overprotected peers.
The problems of our time—and the problems of our youth—are without doubt tied to technology, but the connections are often indirect. Instagram is out of control and should be regulated, but it is not lead in the water. Haidt announces rhetorically that his is the only signal correlation: the time when most American children had smartphones, and when those smartphones included reverse cameras that allowed for flattering selfies, is the time when adolescent mental health plummeted. The period of which Haidt writes is also the time when, in the US, Trump began campaigning for the presidency and civility diminished. It is the time when the average sperm count was found to have declined by nearly half. It is the time when many toxins and microplastics proliferated in the food chain. In the same period the number of young people who identify as queer has escalated dramatically. So has the rate of autism diagnoses. Climate change has become increasingly evident and social mobility less accessible. Young people today are aware that whatever they are training for may well be taken over by AI. International travel has become increasingly commonplace. Guns have proliferated in North America. Gluten sensitivity and nut allergies have become astonishingly frequent. Cyberterrorism has set in; privacy has largely disappeared. The age of first sexual encounter has come down significantly. Some of these changes may be owed to social media, but others are not.
Even though social media are clearly harmful in significant ways, they must be considered in the broader context of an ever-changing world. ADHD and dyslexia diagnoses are more prevalent than they were a generation ago. Single parenting and divorce are far more common. The disappearance of civility in public discourse can be tied in part to internet culture, and that brutality has affected children, but the author is looking more narrowly, and he never recognizes that what appears to be a sudden cultural shift is often the expression of multiple cumulative factors rather than of a single correlation.
Haidt describes sinister decisions at the highest levels of social media companies aimed at courting ever-younger cohorts, and delves into the combination of adolescent psychology and casino tactics that has made social media almost irresistible to young people. He is adept at describing both the malign capitalism that drives Meta, TikTok and their ilk, and the shocking absence of regulation in the US—a situation that has been partially rectified in the UK through the Online Safety Act 2023. In a simile for the proliferation of smartphones, he writes: “It was as if the US government suddenly opened up the entire state of Alaska for drilling and oil companies competed fiercely to stake out the best territories and start sinking wells.” Haidt’s proclivity for histrionic figures of speech works better for describing corporate recklessness than for describing the children targeted. He argues that “an increasing number of heterosexual men” (I have no idea why non-heterosexuals are excluded) will find that a socially isolated lifestyle “with a programmable mechanical girlfriend is preferable to … the social risk of approaching a girl or woman in real life and asking her out on a date.” The number of heterosexual men inclined in this direction may well increase (or it may not), but the tenor of the writing suggests a phenomenon that remains highly unlikely.
It is worth making the changes he proposes. Work hard; organize in your community; let your kids discover the world for themselves; ensure that they engage in social interaction that is not online; get phones out of schools; don’t start with a smartphone; steer your child away from social comparison; organize those around you so that your child does not feel punished by your anti-tech policies.
Predicting the future is a fool’s errand; few and far between are the social scientists whose predictions have proved accurate. Many boys and girls are struggling, but whether “boys and girls … have ended up in the same pit, where they are drowning in anomie and despair” is open to question. In writing about community and spiritual elevation Haidt remarks: “It is a thrill to be one of thousands of fans in a stadium, all singing and stomping in unison after each goal”. This is hardly true if the goal has been scored by the opposing team—but in any case, I have always found that experience unpleasant no matter which team is triumphing. His explanation that this ostensible universal thrill relates directly to how “humans evolved to be religious” is likewise free association. Do “screens lead us to forget that our physical bodies matter”? Few people who eat or sleep or stretch or have sex (even with themselves) have forgotten their physical bodies. Regarding religion, Haidt advises that we separate from others on spiritual retreats—which is striking, given his insistence that separation has poisoned the Anxious Generation.
Toward the end of his book Haidt dwells on the Polynesian expression “standing on a whale, fishing for minnows”—ignoring the gigantic problem while you search for little hints—and proposes that his conclusions are the whales, while the rest is minnows. Marine metaphors can become wearisome, but I would suggest that he is standing on a marlin and fishing for tuna—that the difference in importance between the real problems he is identifying and the other problems of contemporary life is not so singular or dramatic. He describes experiments in which students deprived of social media felt better—but what about an experiment in which kids were told that climate change had been thwarted, that the world’s wars were over, that education was going to be free, that they had a good chance of living better than their parents, that racial inequality was all worked out, that opioids no longer existed and that their families would remain committed and loving until they died?
In his Substack writings Haidt has consistently championed his chosen causes of our problems, with the apparent presumption that we are in a zero-sum game and that causes are quantifiable. It is worth making the changes he proposes. Work hard; organize in your community; let your kids discover the world for themselves; ensure that they engage in social interaction that is not online; get phones out of schools; don’t start with a smartphone; steer your child away from social comparison; organize those around you so that your child does not feel punished by your anti-tech policies. Haidt urges us to be careful but trusting parents and to be aware of what our children are doing. All that makes sense for individual children—but helping a child and helping a society are two different objectives.
If we expect the results that Haidt proposes from those changes alone, disappointment beckons. His enthusiastic universalization ignores the reality that almost everything has a direct effect on some people and a paradoxical effect on others. He quotes many sources, including Zen Buddhism, about the path to happiness, and much of what he writes or references is profound and true. But happiness is not reached by a shared path; people find their different ways to it. “The play-based childhood is over”, he writes, a point at odds with the possibility that a childhood can be play-based even with phones around.
Children tend to establish a pecking order, and while Haidt’s proposals work well for those at the top and even in the middle, they can be catastrophic for those at the bottom, despite his surprising insistence that overall rates of bullying decline for unsupervised children.
Haidt rails against parenting books despite having written one. Dr (Benjamin) Spock suggested in 1946 that parents rely on their instincts. Haidt pushes for much the same thing, though the instincts in question are slightly different ones. He correctly observes that many parents don’t want to overprotect their children, nor to let them spend all day on their phones, and proposes that they should follow that intuitive wisdom. Social isolation is bad for many children, though some children find interacting with others painful in ways that are not constructive. Disappearing into technology in contexts where there might otherwise be warm conversation is damaging in most cases, but every advancement instigates a backlash and, while overuse of phones or other mobile devices can have devastating consequences, social media also has benefits. In both instances Haidt gives almost no time to those who are at the margins; his ideal of children at unsupervised play seems unacquainted with Lord of the Flies and tales of Victorian boarding schools, in which the extraordinary cruelty that unsupervised children visit on one another is delineated. Unsupervised play is generally not very good for kids who are autistic, or queer, or obese, or disabled, or who belong to a race underrepresented in their region. Those children may find community with their helicoptering parents or online with others like them. Children tend to establish a pecking order, and while Haidt’s proposals work well for those at the top and even in the middle, they can be catastrophic for those at the bottom, despite his surprising insistence that overall rates of bullying decline for unsupervised children.
It is not entirely clear that there is a never-before-envisioned mental health crisis unfolding, or that this purported crisis pertains primarily to adolescents. To refer to it as a “tidal wave of adolescent mental illness that swept across many countries” is manipulative writing that provokes the anxiety Haidt supposedly wants to diminish. If there is a crisis, no one can be sure where it came from, and solutions are therefore guesswork. The predicament of modernity will continue: even if we could figure out the source of the crisis in 2024, we would be ill prepared for its reconfiguration in 2025 or 2026. In the 1970s, for example, when Haidt and I both grew up, the word depression was relatively obscure and the idea of seeing a therapist or counsellor was troubled, Woody Allen’s crowd and their analysts notwithstanding. I would not have described myself as depressed, but neither would any other teenager I knew. Many young people have been able to speak directly about psychological states they would once have hidden, but many young people have recognized those states as a possibility in ways they never did before, and have been infected by the availability of the terminology, which can impel the symptoms it describes. That has to do with the good work of activists recognizing the value of openness, but also with the increasingly profitable domain of medicating children. Once children feel depressed, they are depressed; it is not as though they are imagining cancers they don’t have. An illness the symptoms of which are the vagaries of adolescent emotion is going to fluctuate widely with social norms. Haidt focuses on 2012 as the critical year of transformation, but mental health has been declining since it was first measured at a population level a century ago.
Like other neoconservative thinkers he turns repeatedly to the idea of evolution: we evolved to take long walks in the woods and notice what is around us, and we did not evolve to hold little devices in our hands that suck our finite attention away from reality. This is true, but evolution is a slow process; we also did not evolve to clean ourselves with soap, to see doctors, to live in tall buildings, to own guns, to make use of orthodonture, to drive, to wear blue jeans or to do most of what constitutes contemporary life. Overuse of technology may indeed have a steep downside, and effective strategies to address it may need to be brought into play, but there is no need to drag evolution into that contention.
The use of dramatic catchphrases does not strengthen Haidt’s arguments. Children who overuse tech do not necessarily have a “phone-based childhood”; there has been no abandonment of the “play-based childhood”; many children play outdoors less, but many others do not, and some of those cannot. The phrase “Great Rewiring,” which features even in Haidt’s subtitle, and which is capitalized whenever used, suggests an alarming biological process. Any behaviour can cause a rewiring: the brains of Buddhist monks, as the author acknowledges, show changes associated with meditation. People who have had adverse childhood experiences develop visible differences in their brains. Everything you experience repeatedly changes your brain; if scanning were sufficiently advanced we could see not only post-traumatic stress, but also the long-term effects of listening to heavy metal or memorizing Shakespeare or working on an assembly line. Technology is rewiring our brains, but so did literacy. The adolescent brain is plastic; if it weren’t, education would be useless. We rely on a perpetual “Great Rewiring” of brains.
The associative studies that work in a business school such as the one where Haidt teaches do not readily carry over to the quasi-biological arena where this book claims to be rooted. The techniques of social science differ from those of cognitive science. Making his arguments persuasive would require studies based on more complex metrics than those he pursues—particularly given how heavily he relies on children’s self-reports, which are notoriously variable, inaccurate and subject to revision. Many young people who take their own lives have assured others a short time beforehand that they would never do so; suicidal behaviour, which can be highly contagious, may reflect the surfacing of dark feelings rather than a societal shift that has rewritten those feelings.
Technology is rewiring our brains, but so did literacy. The adolescent brain is plastic; if it weren’t, education would be useless. We rely on a perpetual “Great Rewiring” of brains.
Children who in fourth grade said they were happy may in seventh grade insist that fourth grade was the worst year of their life. Being asked how you feel forces you to think about how you feel, which can change how you feel. Miserable people may say they are happy and vice versa. “Know thyself” is not advice intended for teenagers. Self-report, behaviour, and inner state are independent variables; insight-oriented therapy often involves the surprising discovery of grief mistaken for anger, sadness for incompetence, competitive success for joy. Seeing yourself as successful may be more important than succeeding. The countries that come first on happiness indices often have the highest suicide rates. Accuracy of perception can be disadvantageous, which is one reason it is elusive, a level of complexity that Haidt does not pursue.
His positions, like those of many successful populists, contain a great deal of truth. He has assembled a great deal of knowledge and discovered (sometimes) or imposed (other times) order on at least one arena of frightening chaos in modern life. Many truths are most easily absorbed when they are exaggerated. Haidt’s analysis, when it remains with the topics on which he has acquired considerable expertise, is adept and penetrating; his writing, if somewhat messianic, is definitive and accessible. Even when his rhetoric crescendos beyond science it is likely to be socially useful. Black-and-white arguments spur action better than grey ones do. Calls to activism require flair, and Haidt continually builds fortifications around his intellectual arguments, making them sound unassailable. There is smugness and aggression in this behaviour: “If you don’t listen to what I’m telling you, disaster lies ahead.” It is also much more compelling than saying: “What I am proposing might sometimes be a good idea, probably.” Social transformation is seldom based on rationality.
This much is certain: social media is often harmful and should be regulated. The use of phones at schools is distracting and generally does not help students to learn. Adolescents should sleep more than they do; we have known that for a long time, but few have shifted the school day accordingly; in this regard phones have exacerbated an existing problem. Not everyone who smokes develops lung cancer, but enough people do to have shifted the culture away from tobacco use. The use of social media and the excessive use of social media are two different situations, as Haidt acknowledges. The law has to draw distinctions: you cannot declare that the age of sexual maturity for one person is fourteen and for another twenty-six, even though to do so might be more accurate. We need a framework, and any pretence that gargantuan social media companies are going to figure this out themselves is far-fetched. Age verification, as Haidt explains, is not so difficult, but although companies ask users to confirm that they are at least thirteen, there is no effort to enforce the limit. If you tick a box saying you are thirteen, you are all set.
Haidt ultimately lays out a battle plan. His habit of speaking to the reader in the second person—“It is probably what you were doing by the time you were eight”—is based on assumptions rooted in shared class and outlook. People in rural Montana and those in New York public housing developments were probably doing very different things at most ages, even in the generation of middle-aged parents to whom Haidt is addressing himself. When he quotes Kurt Hahn approvingly—“There exists within everyone a grand passion, an outlandish thirst for adventure, a desire to live boldly through the journey of life”—he again falls prey to universalizing. But there can be little question that many people are stuck in lives of quiet desperation, and that inspiring young people is a critical part of parenting and of maintaining a functional society. Some children don’t have a grand passion, but there are purposeful lives without one, and our current sociopolitical dynamics are sucking the sense of purposefulness from many young people’s lives.
Social media play a significant part in escalating youthful nihilism, and fixing social media would doubtless help to address youth mental illness. While Haidt’s suggestions are inappropriate to many people, they have merit. The “Great Rewiring” (great or otherwise) has done damage, and internet addiction is a monstrous problem. Whether childhood as it now exists is a “catastrophic failure” is subjective, but youth mental health is not in a good place. The US surgeon general recently proposed that social media carry a warning label. Such a measure was the first step in curbing tobacco use, and perhaps it will allow those whose lives are squandered on social media and its pernicious comparisons to gain control over their addictive or isolating behaviour. The Anxious Generation is not convincing as science, but as activism it is an impressive enterprise.
In 2003 researchers exposed a group of American nine-month-olds to a Mandarin speaker, who spent twelve sessions playing with them and chatting. By that age most infants have lost their receptivity to phonemes they do not hear regularly. In these infants that decline was reversed; they could distinguish nearly all the tones of Mandarin. Another group of infants the same age was shown videos of the same Mandarin speaker saying the same kinds of things. Infants are drawn to screens, and they watched attentively. After the same twelve sessions they had no reversal of their phonemic decline. What comes through on video is markedly different from what comes through face to face. Our perceptual capacities change as we grow older, and the brain of an adolescent differs from that of an infant. But Haidt’s charts and graphs ultimately illustrate a point established long before social media were commonplace.
The Anxious Generation is not convincing as science, but as activism it is an impressive enterprise.
If I need factual information I might interview someone on Zoom. If I need to know who a person is and why he is that way, I go in person. Boarding plane after plane, driving mile after mile, I wonder whether it is worth it, then I become attached to someone during an in-person interview and we talk for hours, and I remember that something ineffable lies in the space of the personal encounter. Perhaps, in my own way, I am a reactionary too. I am certain that electronic communication is inferior. It only follows that those for whom it becomes the dominant form of interaction have reduced lives. That is a tragic loss, whether or not it is the sole explanation, or even the main one, for a widely perceived decline in youth mental health. If Jonathan Haidt’s book gets some young people back to face-to-face interaction, then its mixed science and Savonarola-like extravagance may be all for a good cause.
Andrew Solomon is a Professor of Clinical Psychology at Columbia University. His books include Far from the Tree: Parents, children and the search for identity, 2012. He is currently writing a book about youth suicide.
