Sure, Phones Drive Anxiety. But So Does the Economy.
As psychologists blame smartphones for our mental health crisis, they overlook decades of economic decline and growing inequality. Their narrow focus on tech diverts attention from the political and systemic changes needed to address a much deeper problem.
In the United States, the suicide rate has risen by 35 percent over the last two decades. (JGI / Tom Grill / Getty Images)
Owing to the work of best-selling psychologists Jean Twenge and Jonathan Haidt, what was once considered debatable has become widely acknowledged: mental health, particularly among young people, is deteriorating in many Western countries. The evidence for this trend is compelling and appears in medication rates, diagnoses, and survey results. In the United States, the suicide rate has risen by 35 percent over the last two decades. During the same time, the percentage of people rating their mental health as “excellent” has plummeted from 43 percent to 31 percent. In 2024, 43 percent of adults reported feeling more anxious than they did the year before, an increase from 37 percent in 2023 and 32 percent in 2022.
These alarming trends should prompt intense social analysis, yet the focus — both scientifically and in public discourse — has increasingly narrowed to a single phenomenon: the spread of social media. Twenge and Haidt have both contributed to the identification of this particular rationale, not least with their respective tech-focused books iGen and The Anxious Generation. The narrative has already had tangible effects, leading several European countries to implement bans on smartphones in schools. While the latter isn’t necessarily a bad thing, it’s clear that the discussion has taken a simplistic turn, downplaying the political dimensions of the decline in mental health.
Although representatives of the psychological discipline like Twenge and Haidt have been largely successful in reducing the crisis to a single-variable discussion, the ongoing debate about social media should be understood as a symptom of a deeper crisis within therapeutic culture and the dominant explanatory models of medicine and clinical psychology. To understand why, it’s worth revisiting the grandiose visions championed by mental health professionals not so long ago.
The Failed Promise of Brain Science
As I’ve engaged with Haidt’s and Twenge’s arguments, one quote in particular has repeatedly come to mind. In 2003, the same year the Human Genome Project was completed, Steven Hyman, director of the National Institute of Mental Health from 1996 to 2001, wrote in Scientific American about the bright future that seemed to lie ahead:
By combining neuroimaging with genetic studies, physicians may eventually be able to move psychiatric diagnoses out of the realm of symptom checklists and into the domain of objective medical tests. Genetic testing of patients could reveal who is at high risk for developing a disorder such as schizophrenia or depression. Doctors could then use neuroimaging on the high-risk patients to determine whether the disorder has actually set in.
Today, despite considerable resources devoted to bringing this vision to life, the future Hyman imagined seems unlikely. The “missing heritability problem” has shown that pinpointing vulnerability genes is far more difficult than enthusiasts of the Human Genome Project once anticipated, and we are nowhere near being able to diagnose a single mental health disorder using neuroimaging.
Instead, something quite different has happened. Mental health has declined dramatically, and this decline has called into question several previously dominant explanatory models, particularly those centered on the brain and genetics.
If we assume that chemical imbalances in the brain are responsible for poor mental health, the question today is what could be causing these imbalances when more and more people are experiencing them simultaneously? And if we consider that a population’s gene pool typically takes thousands of years to change, we must similarly ask: Why has mental well-being declined when the genetic makeup has remained the same? These questions point to forces at play outside the skull and the walls of our cells.
The recognition of a worsening mental state has also shaken the key premise of therapeutic culture: that the solution to our mental health problems lies in psychotherapeutic interventions. When one in eight US adults takes an antidepressant and one in five has recently received mental health care — with an increase of nearly fifteen million people in treatment since 2002 — it’s clear that the rising number of individuals in treatment has not succeeded in curbing the spread of mental illness.
At the same time, over the past decade, massive meta-analyses on the overall effectiveness of psychotherapies have revealed surprisingly modest results. In summary, most studies show that around half of all people undergoing psychotherapy experience some benefit, while about 5 percent see their problems worsen. A meta-analysis led by psychologist Pim Cuijpers confirms this estimate for depression treatment. However, his study also found that only about one-third of patients improved enough to no longer be classified as depressed. Given the high relapse rates for depression, this translates into a lot of therapy for someone experiencing depression.
In a meta-analysis involving a total of 650,000 mental health patients, John Ioannidis and his coauthors summarize their findings as follows: “After more than half a century of research, thousands of [randomized controlled trials] and millions of invested funds, the effect sizes of psychotherapies and pharmacotherapies for mental disorders are limited.” The solution is hard to discern. “A paradigm shift in research seems to be required,” they conclude.
A History Longer Than Smartphones
Psychologists shifting their analyses outward — from the internal workings of individuals to the relationships that constitute society — could very well mark the beginning of such a paradigm shift. However, so far, the most prominent theories adhere to the interventionist logic characteristic of therapeutic culture. More importantly, their arguments suffer from several methodological flaws.
Twenge and Haidt make considerable use of graphs depicting trends in depression, anxiety, suicide, and more, showing how the curves rise after 2012 — the starting point for what Haidt calls “the great rewiring,” when social media entered smartphones. This methodology has repeatedly been criticized for confusing correlation with causation, a point Haidt tries to address in The Anxious Generation. Yet in my view, a larger methodological issue is the tendency to draw sweeping conclusions from relatively short timeframes.
Haidt’s graphs typically begin around 2002 and end around 2018, offering only sixteen years of data from which to draw generalizations. While many of his graphs show a sharp rise in mental health problems during the 2010s, this limited timeframe can be misleading. For instance, when he highlights a dramatic rise in psychological distress among Nordic teens during the 2010s, it gives the impression that nothing noteworthy happened prior. Here, the narrow scope distorts the broader picture.
In Sweden, the Public Health Agency has been surveying young people about their mental health since 1986. If we look at the proportion of those who feel down almost every day, for instance, the increase has been ongoing since the 1980s.
Similarly, the rise in sleep difficulties has persisted for a long time. While the sharper increase in sleep issues and poor mood among girls during the 2010s could be interpreted as evidence of an effect tied to the spread of social media, it is still part of a broader trend that has been unfolding for decades.
This calls for further analysis. No matter which country we examine, we see that similar increases in mental health problems have been occurring over extended periods. In Norway, the pattern mirrors Sweden’s, and in the UK, reports of drastic declines in young people’s mental well-being have been documented for a long time. According to a Psychological Medicine study, between 1995 and 2014, the prevalence of long-standing mental health conditions rose dramatically among four- to twenty-four-year-olds. In England, the prevalence increased sixfold, while in Scotland it doubled over eleven years.
In the United States, Twenge herself has noted the long-term rise in mental health issues. In 2011, she observed that “almost all of the available evidence suggests a sharp rise in anxiety, depression, and mental health issues among Western youth between the early 20th century and the early 1990s.” And in 2000, she estimated that the “average American child in the 1980s reported more anxiety than child psychiatric patients in the 1950s.”
The Economic Elephant in the Room
These longer timelines are important because they reveal a negative trend that cannot be confined to isolated phenomena like social media. Twenge and Haidt have previously offered different forms of sociological Zeitdiagnose to explain what might be going on. For Twenge, it was long about “the narcissistic culture,” while for Haidt, the problem was how “safetyism” has disempowered students and made them overly sensitive. In both cases, the root problem seems to have been what Haidt calls “good intentions and bad ideas” (from the subtitle of his book The Coddling of the American Mind), particularly as they manifest in bad parenting and so-called wokeism. These analyses are far from politically neutral.
Interestingly, both Twenge and Haidt are keen to downplay economic factors. In The Anxious Generation, Haidt does so by citing the decrease in US unemployment during the 2010s following the Great Recession of 2009. If more people are employed, then economic factors cannot explain the mental health decline, right? Of course, this is not a very nuanced approach. We know that recessions can continue to impact disadvantaged groups long after they stop being visible in national averages. Meanwhile, as national inequality continues to grow globally, we know that inequality is a strong predictor of worsening mental health, further amplified by rising status anxiety.
Economic issues like these are difficult to address through therapeutic interventions or policy decisions. They call for structural reforms and demand analyses of society as a whole.
The reluctance to confront systemically created suffering is also evident when Twenge and Haidt address the risk aversion among young people. They are probably right in observing traits of Gen Z, such as drinking less alcohol and having fewer physical fights and unplanned pregnancies, as signs of growing fearfulness. However, their work completely overlooks the extensive sociological literature on how not only individuals but also states and sciences have become increasingly fixated on risk prevention — a field of research that has been expanding since Ulrich Beck coined the term “risk society” nearly forty years ago.
While Twenge and Haidt see risk aversion mainly as the result of poor parenting and overly “good intentions,” Beck and his followers have for some time demonstrated that it is the logical outcome of modern reflexivity and scientific rationality — which ironically doubles back on itself when social science tries to measure the effects of proliferating risk measures.
Reducing the problem to bad parenting should by now be impossible given the wealth of sociological research, not least in the recent works by Hartmut Rosa, describing how risk territories emerge from broad technological, economic, and institutional changes. At its core, the risk aversion reflects a conflict between organic institutions and a growing expert elite, critiqued by Ivan Illich and others as far back as the 1970s.
When we factor in other global “megatrends” that were recently highlighted in the Lancet Psychiatry to explain the decline in youth mental health — such as rising student debt, climate change, and job insecurity — it quickly becomes clear how the question of mental well-being intersects with the political sphere. This is worth keeping in mind before new variables are thrown into the debate in a perpetual cycle of whataboutism. As Herbert Marcuse once cautioned, the operationalism that reduces experience-based concepts like alienation to a series of measurable variables reinforces a technological rationality that hinders radical social critique.
One reason why social science has so far failed to provide definitive answers about the causes of declining mental well-being may be that all social problems, even those not yet identified by social science, bear on our mental health. On closer reflection, it’s a strange notion to think otherwise. Reducing mental health to the outcome of a few variables follows a technocratic logic that obscures society and drains meaning from politics itself. If political issues are not seen as shaping our well-being, why should we engage with them at all?
It is something of an achievement to trivialize the ongoing crisis in the way it has been when a more obvious response would have been a reexamination of capitalism. Even though global life expectancy is increasing, it should by now be impossible to claim, as Bill Gates’s favorite public intellectuals Steven Pinker and Hans Rosling have long done, that capitalism is an eternal success story.
According to the World Mental Health Survey — the world’s most rigorous epidemiological surveys on mental health coordinated by the World Health Organization and conducted across thirty countries to date — we see, on the contrary, that the problems are far more pronounced in capitalism’s most crystallized forms. In seventeen of eighteen mental problems, there is a consistent pattern of much higher prevalence in high-income countries compared to low- and lower-middle-income countries. This stark difference (which cannot be explained by access to social media as the surveys were conducted between 2001 and 2011) contrasts sharply with trends in physical health and raises questions about the value of limitless economic growth.
The Left’s tendency to dismiss findings like these as mere effects of increased detection and diagnosis is misguided, not least because surveys of this kind are specifically designed to measure prevalence independent of psychiatric practice. Human suffering is not an ahistorical constant. It is increasing, and it calls for our care and attention.