Honestly, it's probably the phones
The most plausible explanation for teenage unhappiness.
“Blow up your TV/ Throw away your paper/ Go to the country/ Build you a home” — John Prine
There’s been an interesting dialogue going on about teen unhappiness in the United States. It was kicked off by a CDC survey that showed feelings of sadness and hopelessness are on the rise among American high schoolers, especially among girls. It’s not just people giving different responses to surveys, either — teen suicide is way up, as are symptoms of anxiety and depression. Things started getting worse around 2012 or 2013:
One possible reason, suggested by Taylor Lorenz, is that between climate change, inequality, precarity, and Covid, the world is just a much worse place than it was in 2011. But as I pointed out in a post last week, most of these things (except for Covid, obviously) were looking worse a decade ago:
Eric Levitz points out that life in the U.S. is much better, in terms of material standards, than it used to be back in the days when teen suicide rates were much lower. In fact, if anything, wealth seems to make teens less happy; a new paper by Rudolf & Bethmann finds that although rich countries tend to have happier adults, their adolescents tend to have lower life satisfaction.
In another recent post, Matt Yglesias suggests that progressive politics might deserve some of the blame. He notes that the rise in unhappiness seems to have started earlier and been a bit more pronounced among liberal teens:
Yglesias argues that the progressive politics of the 2010s encouraged progressives to think of everything in catastrophic terms, making them less happy. In essence, he sees teenagers as having been prodded by political and media figures into adopting the same kind of doomer worldview espoused by people like Taylor Lorenz.
I’ve long argued — as Yglesias does — that a negativistic, despairing tone is bad for long-term political motivation. So we agree that progressives should make an effort to shift from rage to determined optimism. But that being said, I’m a bit leery of Yglesias’ hypothesis. It seems just as likely to be a case of reverse causality — maybe progressive politics has a negative tone because people are unhappy, rather than vice versa. Maybe Yglesias is right, but I think there’s another much more obvious explanation that should be our mental starting point (or, if you prefer, our Bayesian prior).
That explanation is smartphones.
Why “it’s the phones” should be our prior
The first reason smartphones should be our prior is that the timing just lines up really well. The smartphone was invented in 2007, but it didn’t really become commonplace until the 2010s, exactly when teen happiness fell off a cliff:
Younger Americans adopted the technology more quickly than older ones; 2010-11 seems to have been an especially important moment. And of course the “killer app” for smartphones was social media. When you had to go to a computer to check Facebook or Twitter, you could only experience it intermittently; now, with a smartphone in your pocket and notifications enabled, you were on every app all the time.
Why would that make us unhappy? There’s an obvious reason: social isolation.
Pretty much everyone knows that social isolation makes people less happy, and research strongly backs this up. It’s known to be a suicide risk. The worst punishment in a prison is solitary confinement, which some view as a form of torture. In case you doubt that the relationship between social isolation and unhappiness is causal, you should recall that we recently ran a gigantic natural experiment on much of society in the form of Covid, and the results were clearly negative.
But why would devices that make people more connected lead to social isolation? Isn’t that backwards? Doesn’t having access to all of their friends and acquaintances at all times via a device in their pockets mean that kids are less isolated than before?
Well, no. As the natural experiment of the pandemic demonstrated, physical interaction is important. Text is a highly attenuated medium — it’s slow and cumbersome, and an ocean of nuance and tone and emotion is lost. Even video chat is a highly incomplete substitute for physical interaction. A phone doesn’t allow you to experience the nearby physical presence of another living, breathing body — something that we spent untold eons evolving to be accustomed to. And of course that’s even before mentioning activities like sex that are far better when physical contact is involved.
Of course, smartphones, by themselves, don’t force you to stop hanging out in person. But there are several reasons they reduce it. First, they’re a distraction — the rise of smartphones was also the rise of “phubbing”, i.e. when people go on their phones instead of paying attention to the people around them. Second, phones provide a behavioral “nudge”, like a pantry stocked with junk food — when your phone is right there in your pocket, it’s easier to just text a friend instead of going and hanging out, even if the latter would be less fulfilling. And third, in-person interaction is a network effect. If 20% of people would rather be on their phones, that reduces everyone else’s options for in-person hangouts by 20%.
The psychologist Jean Twenge, the leading proponent of the theory that phones cause unhappiness, has a great run-down of these various mechanisms.
In any case, the data clearly shows that isolation is increasing. Teens had been getting gradually more isolated through the decades — perhaps as a result of larger houses and better entertainment options at home. But face-to-face interaction really plummeted right after — you guessed it! — 2010.
(Side note: Interestingly, teen loneliness briefly went way down from about 1996 to about 2006. That’s consistent with the idea that the old computer-based internet was a complement to in-person interaction — a place you could build an alternative social circle — instead of a substitute for it. Like I always say, the internet used to be an escape from the real world; now, the real world is an escape from the internet.)
Most of the other main explanations are also about phones
Anyway, the final reason that “it’s the phones” should be our prior is that most of the other explanations are actually just dependent on phone use. For example, there’s a common idea that social media apps like Facebook and Instagram make some people — especially teen girls — feel overly judged. Here’s the always-excellent Derek Thompson summing up that thesis in April 2022:
[Instagram’s] internal research from 2020 found that…one-third of teen girls said “Instagram made them feel worse,” even though these girls “feel unable to stop themselves” from logging on…[There’s also] a big new study from Cambridge University, in which researchers looked at 84,000 people…and found that social media was strongly associated with worse mental health during certain sensitive life periods, including for girls ages 11 to 13…One explanation is that teenagers (and teenage girls in particular) are uniquely sensitive to the judgment of friends, teachers, and the digital crowd.
Really, though, this is just a story about phones. Without phones, you’re forced to be away from Instagram for a lot of the day; with a phone in your pocket, the ever-present judgment of your peers is always just a thumb-swipe away.
Then there’s the idea that young people are overwhelmed by negative news. Here’s Thompson again:
[T]eenagers’ perception of the world seems to be causing them more stress…“In the last decade teenagers have become increasingly stressed by concerns about gun violence, climate change, and the political environment,” [clinical psychologist Lisa Damour] wrote…“We’re coming out of the pandemic, and then suddenly Russia goes to war. Every day, it feels like there’s something else. It creates a very gloomy narrative about the world.”…
This sense of doom…comes from us, the news media, and from the social-media channels through which our work is distributed. News sources have never been more abundant, or more accessible. But journalism also has a famous bad-news bias, which flows from an unfortunate but accurate understanding that negativity generally gets more attention. When we plug our brain into a news feed, we are usually choosing to deluge ourselves with negative representations of reality…We cannot rule out the possibility that teens are sad about the world, not only because the world contains sadness, but also because young people have 24/7 access to sites that are constantly telling them they should be depressed about it.
This is just the bad-news bias plus the availability heuristic. News organizations and social media shouters have an incentive to show you bad news, because that’s what gets them attention and clout and/or money. And when you see bad news all day, then like Taylor Lorenz, you will think the world is going to hell in a handbasket.
But this, too, is really a story about phones. Before smartphones, people were usually disconnected from the news — now we’re checking it constantly. And phone-enabled social media sites like Twitter didn’t just increase news consumption — it also increased the supply of people shouting about disasters and doom for clout.
Even Yglesias’ politics-based thesis is really a story about phones, I think. Progressive politicians haven’t become any more inclined to rage and despair than in the past; Bernie Sanders and Alexandria Ocasio-Cortez can even sound downright utopian. But phones put young progressives in constant contact with progressive social media shouters, who are constantly bombarding them with tales of the death of democracy, spiraling poverty and inequality, oligarch-dominated politics, unstoppable climate change, ubiquitous hegemonic racism and cisheteropatriarchy, and so on and so forth.
This is not to say that every possible explanation for rising unhappiness can be traced to the rise of smartphones, of course. For example, there’s the idea that suburbanization is responsible for social isolation. That idea is contradicted by a lot of empirical evidence — the suburbs are not really more isolating than the city. But anyway, there are a few explanations out there that don’t depend on phones.
But almost all the leading hypothesis can be connected directly to the rise of smartphone use. And that means that if we’re looking for one big “silver bullet” or “grand unified theory” of modern teenage unhappiness, phones are probably the place to start looking.
The burden of proof
Now, just because phones are a good prior does not mean that we should conclude that phones are the culprit. Priors and conclusions are separated by a vast gulf called “the evidence”. In other words, there is a burden of proof here.
This burden of proof is especially important because almost every new technology is subject to a backlash at some point. TV was despised by whole generations of educated Americans — an “idiot box” that would shorten your attention span and rot your brain. There was a multi-decade panic about video games turning people violent. Industrial society itself was feared and hated for centuries by people who thought it would separate us from the land and commodify our existences. And yet very few of those panics were borne out in the long term; humanity adapted our lives to work with the new technologies, and we kept on doing OK.
All that having been said, however, there is quite a lot of evidence piling up in favor of a link between phones and unhappiness.
For example, there’s the famous 2019 paper by Allcott et al. which found that having people deactivate Facebook for a while made them happier, while also making them socialize more and worry less about politics:
In a randomized experiment, we find that deactivating Facebook for the four weeks before the 2018 US midterm election (i) reduced online activity, while increasing offline activities such as watching TV alone and socializing with family and friends; (ii) reduced both factual news knowledge and political polarization; (iii) increased subjective well-being; and (iv) caused a large persistent reduction in post-experiment Facebook use.
And here’s a paper by Lambert et al. (2022) with a similar experiment and similar findings for other social media platforms:
We randomly allocated 154 participants..to either stop using [social media] (Facebook, Twitter, Instagram, and TikTok) for 1 week or continue to use [social media] as usual. At a 1-week follow-up, significant between-group differences in well-being…and anxiety…in favor of the intervention group were observed…The intervention effect on depression and anxiety was partially mediated by a reduction in total weekly self-reported minutes on Twitter and TikTok, and TikTok alone, respectively. The present study shows that asking people to stop using SM for 1 week leads to significant improvements in well-being, depression, and anxiety.
(Update: Here’s an experiment by Tromholt (2016) restricting Facebook use in Denmark, with similar positive results for well-being. And here’s a similar experiment by Brailovskaia et al. (2022) restricting all social media in Germany, with similar results. And here’s a series of experiments by Sagioglou & Greitemeyer (2014) that restricts Facebook use, again with similar results. And here’s an experiment by Hunt et al. (2018) that covers Facebook, Instagram, and Snapchat, yet again with similar results.
Here’s an experiment by Yuen et al. (2019) that finds that using Facebook caused significantly worse mood relative to browsing the internet, lending a bit of support to my own hypothesis that the pre-social media “old internet” was much better for mental health. (But then again, here’s a study by Braghieri et al. (2022) that uses the uneven rollout of Facebook across universities as a natural experiment, finding that it did decrease mental health; that was before the introduction of smartphones.)
And here’s an experiment by Kushlev and Leitao (2020) that found that social media use leads to disconnection between parents and children:
In a field experiment at a science museum (Study 1), we randomly assigned parents to use their phones frequently or infrequently. Frequent phone use led parents to feel more distracted, which in turn impaired feelings of social connection and the meaning that parents derived when spending time with their children….These studies suggest that being constantly connected to the Internet may carry subtle costs for the fabric of social life.
Here’s Dwyer et al. (2018), finding that “phubbing” is real and bad:
[W]e recruited over 300 community members and students to share a meal at a restaurant with friends or family. Participants were randomly assigned to keep their phones on the table or to put their phones away during the meal. When phones were present (vs. absent), participants felt more distracted, which reduced how much they enjoyed spending time with their friends/family. We found consistent results using experience sampling in Study 2; during in-person interactions, participants felt more distracted and reported lower enjoyment if they used their phones than if they did not.
I can’t actually find a link to the paper, but this is from the abstract of Elaine Guo’s 2022 job market paper, which uses a natural experiment to show that wireless internet worsened teen girls’ mental condition:
This paper investigates the extent to which social media are harmful for teenagers, leveraging rich administrative data from the Canadian province of British Columbia and quasi-experimental variation related to the introduction of wireless internet there. I show neighbourhoods covered by highspeed wireless internet have significantly higher social media use, based on Google search volume data…I link spatial data on broadband coverage to 20 years of student records that provide detailed information about individual student health. Using this novel data linkage, I estimate a triple-difference model comparing teen girls to teen boys in terms of school-reported mental health diagnoses, before and after visual social media emerged, and across neighbourhoods with and without access to high-speed wireless internet. Estimates indicate high-speed wireless internet significantly increased teen girls’ severe mental health conditions – by 90% – relative to teen boys’ over the period when visual social media became dominant in teenage internet use. I find similar effects across all subgroups. When applying the same strategy, I find null impacts for placebo health conditions – ones for which there is no clear channel for social media to operate.
Note that these are all causal studies — either experiments or natural experiments. There are a ton of correlational studies out there that try to look at the purely statistical relationship of phone use and happiness; some find a correlation, others don’t. But in general, causal studies should be our gold standard. And though I can’t claim to have done an exhaustive review of the literature, what I can say is that every causal study I can find shows a negative effect of phone-based social media on emotional well-being and/or healthy social interaction.
(Update: Here’s a much more thorough review of the literature by Richard Hanania, which includes many more causal studies, as well as trends from other countries. It turns out that there are a few causal studies that don’t find an effect, but there are a lot more that do, and the ones that do tend to have much larger sample sizes. Also, the trend toward decreasing happiness in the 2010s is present in most rich countries, though not all. I have added most of the studies noted in Hanania’s post in the paragraphs above, except for a couple I didn’t think were sufficiently relevant or were not actually causal research designs.)
That doesn’t mean it’s an open-and-shut case. To fully satisfy the burden of proof here, we’ll need a lot more than 13 causal studies. Researchers will need to conclusively establish the empirical fact that smartphones are bad for emotional well-being, and to tease out the exact mechanisms by which it does this.
And even if that all happens, and it does turn out that phones are the major — or a major — culprit behind teen unhappiness, we’ll need to decide what to do about that.
It takes time for humans to adapt to new technologies
Even if we do find that smartphones are behind the rise in teen unhappiness, that doesn’t mean they’re a bad technology overall. Every technology comes with its costs — cars and trains cut through natural habitats, electricity emits pollution, manufacturing produces toxic waste, and so on. In most cases we choose to manage and mitigate the cost instead of banning the technology. So no one should be talking about a smartphone ban here, even in the worst-case scenario (though parents may want to limit kids’ screen time).
In fact, our best move may simply be to wait for society to adapt. In the past, humans have shown a remarkable ability to change our societies, our cultures, and our institutions in order to thrive in the world created by new technology. The political and social chaos created by the printing press and the squalor and filth of the early Industrial Revolution were very real, and yet now our lives are so much better as a result of these things that few would dream of going back to a time without them. We adapted to radio and TV and video games and computers and websites, we adapted to cars and trains and airplanes, we adapted to washing machines and refrigerators and microwaves. I’m betting that we can eventually adapt to smartphones and social media as well.
The first adaptation might just be to consciously prioritize in-person interaction, instead of just waiting for it to happen like we used to. Dan Kois has an excellent article in Slate about Sheila Liming, whose new book Hanging Out: The Radical Power of Killing Time provides a much-needed alternative to phone addiction. Perhaps hangouts will go from being something that everyone just naturally did because it was the only way to get human contact to being something that people deliberately carve out time for and put on their calendars.
Another adaptation is probably to take social media less seriously. Twitter isn’t a field of combat where heroes decide the fates of nations — it’s just a silly room where people scream at each other and tell a bunch of lies. You don’t have to look fabulous on Instagram or go viral on TikTok to be cool and have friends. Perhaps Zoomers will realize these truths, and embrace a cheerful detachment that Millennials never managed to achieve.
Eventually, the kids have always been alright before, and I think they can be alright again, even if it takes some time.
Update: Dave Marcotte and Benjamin Hansen have a new paper that investigates the alarming rise in U.S. suicide rates over the past decade or so. They consider the smartphone/social media hypothesis, and find that it would fit the evidence, although it’s too complex to make a definitive judgement:
The bulk of this evidence highlights that for all the promise these new technologies offer, they may exacerbate underlying mental health conditions and could be contributing to the rise in teen suicides and worsening mental health (George, 2019). The causal impact of social media use on suicidal behavior is difficult to establish, so an important avenue for future research would be the use of credible research designs to shed light on the risks posed by the extent and patterns of social media use on suicidal behavior among teens and others.