Mental health
The crisis that keeps getting redefined
Description
There is a mental health crisis. Or there isn't one. Or there is one but it's smaller than advertised. Or it's larger than the diagnostic categories capture. Depending on which source you read, the mental health situation among young people in developed countries is either a historic emergency requiring immediate policy response or a moral panic produced by therapy culture, social media, and specific measurement artifacts. The confusion is not accidental. Mental health is uniquely difficult to measure, the diagnostic categories are themselves contested, the boundary between ordinary human difficulty and clinical pathology is genuinely unclear, and the incentives of the actors involved — clinicians, pharmaceutical companies, advocacy groups, insurance, media — pull in multiple directions at once.
What is clear is that something has changed. Rates of reported depression, anxiety, and specific mental-health conditions have risen substantially in developed countries over the past decade, particularly among adolescents and young adults, particularly among girls and young women. Suicide rates have risen in some demographics. Disability claims for mental-health conditions have increased. Demand for therapy has exploded. The use of psychiatric medications has expanded dramatically. Whatever is going on, it is producing real numbers that require explanation, even when the explanations remain contested.
What is less clear is what these numbers mean, what is driving them, and what the appropriate response is. Some researchers argue the rise in reported mental health problems primarily reflects increased willingness to seek help and more expansive diagnostic categories — that is, it's mostly a measurement phenomenon. Others argue it reflects a real deterioration in mental health, driven by specific factors (smartphones, social media, economic precarity, specific cultural shifts). The evidence supports both views in part, and sorting out what is real underlying pathology versus what is measurement artifact is one of the more consequential and unresolved questions in contemporary public health.
● The question we're asking: what is actually happening with mental health, and how should we think about the current crisis?
● What we'll see: the measurement challenge, the actual trends, the specific drivers being proposed, and the specific responses being tried.
Table of contents
01The measurement challenge
Mental health statistics depend on diagnostic categories that have shifted over time. The Diagnostic and Statistical Manual (DSM), now in its fifth edition, has expanded the specific conditions it includes and broadened the criteria for many of them across successive editions. Conditions that once required specific severe symptoms can now be diagnosed with milder presentations. What would have been ordinary sadness or anxiety in one era becomes diagnosable depression or generalized anxiety disorder in another. This is not necessarily wrong — the expansions have reflected genuine clinical insights — but it does mean that comparing rates across eras is genuinely difficult.
Self-report surveys are the main source of population mental-health data, and they have specific limitations. People's willingness to report mental-health symptoms has changed substantially over decades, reflecting reduced stigma and increased familiarity with specific vocabulary. Someone describing themselves as 'depressed' in 2024 and someone describing themselves with the same word in 1980 may be reporting different internal states with substantially different intensities. The increase in reported rates thus partly reflects real changes and partly reflects willingness-to-report changes, and separating the two requires specific methodological effort that most popular coverage does not bother with.
02The actual trends
Several trends are reasonably well established despite the measurement challenges. Depression and anxiety rates among adolescents and young adults have risen substantially over the past decade in multiple developed countries. The rise is particularly steep for girls and young women. Suicide rates among adolescent girls in the US doubled between 2007 and 2019, though they remain lower than suicide rates in older adults. Emergency department visits for self-harm among adolescents have risen. Psychiatric medication prescriptions to young people have increased. These trends are consistent across multiple data sources and multiple countries, which makes them harder to explain away as purely measurement artifacts.
The gender pattern is striking and requires explanation. Rates of depression and anxiety have always been higher in women than men, but the gender gap has widened in recent cohorts. Boys have experienced more modest increases, girls dramatically larger ones. The specific explanations proposed include differential social media impact (girls are more affected by social comparison and appearance-focused platforms), differential pressure around achievement, and specific cultural messaging that affects girls more than boys. No single explanation has been definitively established, but the pattern is real.
03The specific drivers
Smartphones and social media are the most-discussed driver. The evidence is real but contested. Correlational studies consistently find that adolescents with heavy social media use have worse mental health than light users. The effect sizes are modest (typically explaining a few percent of variance in depression or anxiety scores), which has led some researchers to argue the relationship is too small to explain the observed increases. Others argue the effect size is still substantial at population scale, and that specific features of social media (comparison, feedback loops, fear of missing out, sleep disruption) produce specific harms that the aggregate measures underestimate. The evidence probably supports a moderate causal role, with specific subpopulations affected more than others.
Economic and structural pressures have increased. Housing costs have risen faster than wages. Young adults face worse financial prospects relative to their parents than previous generations did. The labor market for entry-level workers has become more precarious. Academic pressure has intensified. Climate anxiety, political anxiety, and specific societal uncertainty contribute to background stress. These structural factors probably contribute to the mental-health declines, though their specific contribution is hard to quantify. They do not explain the specific concentration in adolescent girls, which suggests multiple drivers are operating simultaneously.
04The responses
The therapeutic response has expanded enormously. Therapy is more culturally normalized than at any previous point. Apps and online platforms have made therapy-like interactions accessible to populations that previously lacked access. Evidence-based therapies (CBT, specifically) have well-documented effects for specific conditions. But the expansion has produced specific problems. The therapist supply has not kept up with demand in many places. Insurance coverage for mental-health services remains inconsistent. Quality varies enormously across providers. Much of the therapeutic activity is happening through apps and services with weaker evidence bases than the research trials.
The pharmacological response has similarly expanded. Antidepressant prescriptions to young people have doubled or tripled in several countries. Anxiolytics, mood stabilizers, and stimulants have all increased. The efficacy of these medications is real but modest — for mild-to-moderate depression, therapy produces effects comparable to medication, and the combination often works better than either alone. The increase in prescribing to adolescents, whose developing brains may be affected differently than adults, is a specific concern research is still addressing.
05Conclusion
Mental health is one of the hardest topics to discuss clearly because the measurement is uncertain, the causes are contested, and the incentives of the actors involved pull in different directions. What can be said is that real changes in real outcomes have occurred among specific populations, that multiple drivers are probably involved, and that the responses currently being tried are probably insufficient to reverse the trends. The crisis is real, in specific demographics on specific measures, but it is also partially measurement-driven in ways that complicate the policy conversation.