Below is a lightly edited transcript. For the article that inspired this one, see It’s Not Social Media, Life Is Just Worse.
Welcome
Welcome to the btrmt Lectures. My name is Dr Dorian Minors, and if there’s one thing I’ve learned as a brain scientist, it’s that there’s no instruction manual for this device in our head. But there are patterns. Patterns of thought, patterns of feeling, patterns of action— because that’s what brains do. So let me teach you about them. One pattern, one podcast, and you see if it works for you.
The last few episodes have been a kind of pop-neuroscience debunking run. I covered the amygdala, I covered cognitive bias, I covered how people use neuroscience to over-engineer calming down. So I wanted to do something a little different from the brain science today. I wanted to talk about something more social, more structural—something I think a lot of people are getting wrong in a way that’s starting to have real consequences and no good solutions.
Before I was at Sandhurst, before I was a brain scientist even, I was a clinician. A crisis counsellor. And one of the things I dealt with a lot was people’s digital hygiene—technostress, as it seems to be getting called these days. The idea that people are struggling with their digital habits, their screen time management, doomscrolling, whatever you want to call it.
I think we all have this distinct sense that the digital world is a perilous place. This economy of shame, outrage, yearning, and terror that floods our media streams. Algorithms designed to suck us in and drain us dry. Algorithms that, famously, the creators won’t use themselves or let their kids use. And I want to talk about it, because I think the social media problem is actually a bit of a distraction.
So let’s get into it.
The story everyone’s heard
Here’s the story you’ve probably heard. Mental health problems are on the rise, particularly in young people. Social media use is also on the rise. So, probably, social media is causing the mental health crisis.
There are some fun graphs around this. Social media use starts ticking upward around 2010, and around the same time you have this spike in mental health concerns. Ipso facto, et cetera.
It’s hard not to draw that conclusion, because it feels right. You’ve probably felt it yourself. Doomscrolling at midnight, closing the app feeling worse than when you opened it. I remember distinctly, in the final weeks of my PhD, finishing a day of writing at four in the morning and being perversely addicted to watching car crash reels on Instagram until the sun rose. Try getting a good night’s sleep for the next day of writing after that.
Or watching your kids—or someone else’s kids—glued to their phones or their iPads in a way that just seems unhealthy. There was this Tumblr post that said if aliens came to Earth, they’d think we were the slaves of the little black boxes we carried around in our hands while sitting, or walking, or even driving our cars.
And there’s a figurehead for this concern now. Jonathan Haidt, a social psychologist, who pivoted through some excellent work on moral psychology in the early 2000s— stuff I actually teach in my classes today—into some slightly more questionable complaints about how emotionally fragile our society is becoming, and most recently into a book called The Anxious Generation. It sold over two million copies. Spent a year on the bestseller list. The argument: smartphones have rewired children’s brains. We evolved for a play-based childhood and we’ve replaced it with screens. We can’t handle this new world of social media.
Off the back of this book—and the broader anxiety it represents—thirty-five US states have passed phone restriction legislation. Australia, dearer to my heart, has banned social media for under-16s. Other countries are now considering the same thing. This very mainstream craze around forcing people to do less social media.
And I should say, the concern is very understandable. Social media companies aren’t really benevolent creatures. Algorithms are explicitly designed to hold your attention. It’s very easy to look at TikTok or Instagram and think, yeah, that can’t be good for a fourteen-year-old.
But understandable isn’t the same as correct. I’ve been writing about this for a while, pointing out that the research doesn’t really support the story. And more pressingly, everyone has been writing and podcasting about how Haidt seems to have got himself caught up in increasingly shrill moral panicking. What’s good is that since I wrote my article, the research has moved on quite a bit. So today, let me do a little update.
What the research actually says
Let’s start with what the research actually says, because the gap between the public narrative and the scientific evidence is kind of troubling.
In 2024, Christopher Ferguson published a meta-analysis of 46 studies looking at youth social media use and mental health. A meta-analysis is an analysis that collects studies examining a phenomenon and then looks at trends across those studies. He took 46 of them, looked across them, and what he found was that the effect of social media use on mental health was “not statistically different from zero.” That’s not the same thing as “small.” That’s academic speak for nothing. And there are two forms of nothing academics can talk about: no evidence because we haven’t looked yet, and no evidence because we’ve looked and couldn’t find any. Here, Ferguson is saying the latter. He looked across 46 studies and didn’t find any evidence.
He then explicitly warned that policymakers and professional guilds—psychology bodies, medical associations—are misleading parents with unsupportable claims about the harms of social media.
That same year, Amy Orben’s group— she runs a lab at my old department at the University of Cambridge that basically does nothing but study this exact question—published the biggest meta-analysis yet. 143 studies. Over a million adolescents. The correlation between social media use and mental health problems? 0.08. For context, a correlation of 0.08 is almost indistinguishable from zero. If you put it in percentage points, where a correlation coefficient of one is 100%, it means social media use explains less than 1%—0.8%—of the variation in mental health outcomes. Less than a per cent.
So all of that happened in 2024, when I first wrote the article on this subject. Now, more recently, an umbrella review came out covering 72 reviews, including 20 meta-analyses. An umbrella review looks at the trends across studies and the meta-analyses of studies. What they found: general social media use showed “weak and inconsistent associations” with mental health. The only thing that was consistently associated with worse outcomes was what they called “problematic” use—essentially addiction-pattern behaviour. Not using social media, but being addicted to social media. Which is a very different claim. We’ll talk more about that in a moment, but I want to tease out what’s happening here first.
Stuart Ritchie, who wrote a fantastic book called Science Fictions, makes an excellent point about why the research looks like this. Basically, people claiming that social media is causing problems are cherry-picking. They’re counting studies that favour their hypothesis. And the problem is that in science, you can’t just count studies. Twenty tiny, low-quality studies saying X is true can’t outweigh one large, well-designed study saying it isn’t. Science isn’t a democracy where one study gets one vote. It’s the methodology that matters.
And that seems to be what’s happened here. A lot of low-quality correlational work that seems to point in one direction, drowned out by the careful studies that don’t find much of anything. Yet thirty-five US states legislated on this evidence base. And Australia banned under-16s.
If you listened to my lecture on atavism, you’ll recognise the pattern. People got scared of technology. They noticed something is going wrong. The obvious solution is to go back to the old ways—ban the modern problem, eliminate it. Panic led to policy. And the problem is left unresolved, because atavism is almost never the answer. People aren’t defenceless against the modern world. What it requires is a little bit of thought.
Life is actually getting worse
So the question becomes: if it’s not social media causing the problem—because mental health problems are going up—then what is going on?
There’s an important answer I’m not going to speak to in detail here, which is the quality of our testing for mental wellbeing. I’ll link to something on this in the show notes. But one of the things that’s definitely happening is we’re spending more time diagnosing people with mental illness and spending more effort as a society treating them with care. That accounts for part of it, but sadly not everything.
So I did an exercise for the article version of this. I walked through the top-level headings on “contributing factors to mental illness” in my first-year psychopathology textbook from about a decade ago and checked which ones have gone up. Nearly every single one has. Let me give you a few.
Income inequality and the cost of living. This one’s hard to miss. Widening income inequality is well documented across most developed countries, and there’s a clear empirical correlation between more inequality and more mental health problems—depression and anxiety, mostly. A paper in the Lancet in 2023 modelled the health effects of the cost-of-living crisis and found it threatens both physical and mental health, disproportionately for vulnerable households.
Housing specifically has got much worse. A Dutch longitudinal study found that housing unaffordability increased sharply, almost entirely within the rental sector, and that it’s clearly linked to poorer mental health—particularly in younger adults. If you’re under 35 and renting, you’re the demographic being squeezed hardest. And the mental health data shows it.
Job insecurity. The gig economy, zero-hour contracts, short-term work. It’s pretty clear that work is changing and turning ever more towards instability. A charitable interpretation is that we’re moving towards flexibility, and that’s also true—but flexibility is definitely a trade-off with stability. Just looking at the rate of striking in the UK over wages and working conditions over the last few years, you’ve got pretty good anecdotal evidence that something’s off. And the correlation between job insecurity and psychological distress is one of the most robust in the occupational health literature.
Urbanisation. More people live in cities than ever before. Several studies highlight the correlation between urban living and higher stress, social isolation, and mental health disorders. City living is just hard on people, even with all the advantages it confers over living more rurally.
Information overload. Now, this one does involve social media, but not in the way Jonathan Haidt would have you believe. It’s not that Instagram is rewiring your brain. It’s that the sheer volume of information we’re all processing is completely unprecedented—and it’s still increasing.
There’s this book by Neil Postman from the 1980s called Amusing Ourselves to Death. I think it’s still well worth a read. What he said, in short, is that in a technologically connected world, rather than getting information about your local area—people and places and events you can engage with and do something about—more and more, we’re getting information “from nowhere” and “to no one.” You can’t do anything about the war in Ukraine, but you’re fed information about it. You can’t do anything about the situation in Iran, but you’re fed information about it. Postman was talking about television, but the internet made that problem much, much worse.
Sedentary lifestyles. We’ve been agonising over this since the 70s, and it’s only getting worse.
And I could keep going. Substance misuse, climate anxiety, family dynamics—I just wrote an article on how much cultural turmoil is happening around family dynamics. Almost all of the headings from this textbook seem to be problems that have been increasing in recent years. When you lay all of this out, social media becomes this tiny thread in a much bigger tapestry.
A really comprehensive review came out just last year arguing exactly this. They call it “a paradigm shift in the global political economy which has largely passed under the radar.” Neoliberal policies producing fragmentation and inequality, and that’s what’s driving the youth mental health crisis. Social media is just one item in a long list of contributing factors—nowhere near the headline.
And the most authoritative review I’ve found on the subject, published in World Psychiatry in 2024, is a comprehensive treatment of the social determinants of mental health. Income, housing, neighbourhood, education, employment, discrimination. All structural. All far out-shadowing the effect of social media. The stuff that’s hard to legislate against in an afternoon.
The directionality problem
Here’s the bit I think is the most interesting, and it’s something I speculated about in the original article. But there’s now actual data behind it.
The question is this: is social media making people sad, or are sad people using more social media?
We should ask this, because all of the research is trying to associate increased social media use with mental health concerns. A correlation describes a relationship: as one goes up, what does the other one do? And with any study that’s just correlational, the relationship could go both ways. Social media could be doing what Jonathan Haidt thinks it’s doing—making people worse, rewiring their brains. But since the literature doesn’t really pan out on that, maybe it’s going the other way. Maybe people with mental health concerns are doing something different with social media.
Amy Orben’s group—again, my old department at Cambridge—published a paper in Nature Human Behaviour last year. This was a registered report, which means they described the methodology and had it peer reviewed before they collected the data.
The reason that matters is because recently—well, over the last fifteen years—we’ve run into what’s called the replication crisis. A lot of famous psychology studies can’t be replicated. They were done once, and when people try to do the exact same study again, they don’t find the same result. One reason for this is that researchers have an idea, collect data, run their analysis, don’t find what they hoped, and then mess about with the data until they find something that supports the conclusion they wanted. Obviously that’s difficult to replicate. It’s fraud, right?
So one of the solutions now is registering your methodology, having it peer reviewed, and then collecting the data and running the analysis exactly as planned. Much less room for messing around.
That’s what Orben’s group did. They looked at a nationally representative UK sample of over 3,000 adolescents with proper clinical assessments. And what they found is that adolescents who already have mental health conditions spend more time on social media, do more social comparison, and are more affected by feedback on their posts. The kids who are already struggling are the heaviest users.
A Spanish study found the same pattern from the other direction. Escapism as a motivation for using social media predicted more internalising symptoms—that’s what we call it when you direct your psychological distress inwards: anxiety, depression, withdrawal, somatic complaints like headaches or stomach aches with no medical cause. Contrast that with externalising, where you turn your distress outwards—aggression, rule-breaking, substance use. So, escapism predicted more internalising symptoms, which in turn predicted more social media use. A reinforcement loop, a vicious cycle. But the starting point was the distress itself, that motivation to escape—not the social media platforms.
And remember: with correlational studies, you can’t tell which is influencing the other. Does social media cause people to feel bad? Does feeling bad cause people to use more social media? Or is some third thing causing both?
The solution is a causal study, and I won’t get into the detail for time. But there’s one big UK longitudinal study I think is interesting. It found very little evidence for a causal relationship between social media use and mental health problems over two years. Most of the apparent effect was mediated by self-esteem—and when you controlled for that, the relationship washed out. Self-esteem was the thing driving both: increasing mental health concerns and increasing social media use.
So here’s how I’d put it. Social media seems to be acting as something closer to a thermometer—a measure of the problem. And smashing the thermometer isn’t really likely to do anything about the weather.
When we find sad kids on TikTok, maybe they’re on TikTok because they’re sad. Maybe social media is the most visible manifestation of people trying to cope with a world that is, in all these structural ways, getting harder to live in.
What does this mean for you
So what does all of this mean for you?
I think three things.
First, we should probably be more honest about why we feel this impulse to scapegoat screen time. The structural factors I’ve been describing—inequality, housing, job insecurity, urbanisation, information overload—these are enormous, slow-moving, politically inconvenient problems. You can’t fix income inequality in a legislative session. You can’t reverse urbanisation with a policy paper. Not least because it doesn’t really serve politicians to do so.
But you can ban phones in schools. You can pass an age restriction on social media. These things are visible, they have an impact on individuals, and it feels like we’re doing something. It’s the same pattern I talked about in the atavism lecture. Legitimate concerns about the modern world, real yearnings for an idyllic past—but because we don’t work like that, and because the past was never as uncomplicated as we like to make out, we end up with the wrong diagnosis and the wrong prescription.
Second, this isn’t a lecture about social media being harmless. It’s certainly not nothing. But the research is increasingly clear that what you do on social media matters much more than how long you spend. Passive consumption of weight-loss content is associated with body image problems. Active social use—talking to friends, maintaining relationships—can actually be protective. A really nice review in the Journal of Child Psychology and Psychiatry put it well: “social media use is not a monolith.” It’s more like a diet. The question isn’t whether you eat. It’s what you eat.
And third—and this is the bit that matters most—it’s up to you to shape your relationship with social media. Not the other way around. As far as anyone can tell from the evidence we have, social media isn’t doing something to you. You’re doing something with social media. And the thing you’re doing with it is shaped by everything else going on in your life.
If your life is good—if you’ve got stable housing, decent work, and people around you who actually care about you—social media is probably going to be just fine. But if your life is hard, social media might be where you go to cope. And it might not always be the healthiest coping strategy. But the answer to that isn’t taking away the coping strategy. The answer is to fix the thing that’s making life hard.
And that honestly seems like a pretty good place to leave it.
