Against the Smartphone Theory of Everything

The phones are not the only problem

(Illustration by Isabella Pereira/TheArgument, photos by Sefa Karacan/Anadolu Agency via Getty Image and Neville Elder/Corbis via Getty Images)

In 2018, Stanford economist Matthew Gentzkow and his colleagues paid about 1,700 Americans to deactivate Facebook for four weeks before the midterm elections. Those who logged off were happier, less anxious, and less politically polarized.

Findings like these have trickled through academia into mainstream media and the public discourse, where parents, teachers, depressed youths, and politicians are well-primed to see smartphones and social media as The Problem^TM.

But here’s what nobody talks about: The people who deactivated also knew less in general. They came back measurably more ignorant about what was happening in their own country. And in a second, larger study of more than 35,000 Facebook and Instagram users, the same thing happened.

The strongest, most reliable finding from years of randomized social media research is not that Facebook makes you crazy; it’s that, for better or worse, Facebook is how you find out what’s going on in the world.

This finding is a Rosetta Stone for the smartphone debate, because it reveals something both sides keep getting wrong: The smartphone is not a cigarette. It isn’t a toxin that we can isolate and test and ban. It is an information-delivery system — a relentless, inescapable IV drip of news, connection, outrage, friendship, conspiracy, solidarity, and garbage — whose effect on any individual depends entirely on what’s in the drip.

Today, billions of people look at their phones and see the whole world. But some theorists look at the whole world and see only phones.

The debate has calcified into two camps: The Jonathan Haidts of the world, who blame phones for basically every modern malady, and the skeptics who say the evidence is thin and the panic overblown.

NYU Professor Arpit Gupta coined the term “Smartphone Theory of Everything” (STOE), which, in his telling, seems to explain every modern problem there is. The rise of youth anxiety? It’s the phones. The rise of global populism? The phones, again. The surge in attention disorders in the U.S.? The global decline in literacy? The scourge of political polarization? Phones, phones, and more phones.

I have been reporting on this space for several years and my conclusion is that the “Smartphone Theory of Everything” is wrong about “Everything” … but it isn’t wrong about everything.

Throughout my reporting, I’ve relied on several experts across psychology, economics, and political science. Rather than base my analysis on individual correlational studies, I leaned on randomized trials, meta-analyses that evaluated hundreds of studies, and a “consensus” survey that asked over 100 academics what they thought about claims about smartphone and social media use on personal and mental health.

My argument, in short, is this: The smartphone is not a poison, it’s a displacement machine.

When you stare into your phone, you are displacing some other activity — sleeping, socializing, playing outside, paying attention in class, watching TV…

In the countries where the effects are worst — the English-speaking West, especially among vulnerable teenagers — something about the combination of what phones deliver and what they displace is producing uniquely bad outcomes.

I have hypotheses about why, but I want to be honest: This is the part of the argument where the evidence thins out. (My guess is that the information flowing through the IV drip is the most anxious and polarizing and the displaced activities are the most beneficial for these populations.)

The phone is a device that interacts with a society’s preexisting pathologies. In the English-speaking West, they are revealing, not creating, a deeper rot.

The strongest proponents of the STOE often ignore the stubborn fact that phones are practically global, but their worst effects are strangely concentrated in the richest English-speaking countries. But the fiercest critics of the STOE often ignore the findings of randomized trials and real-world experiments, such as phone bans in schools, which have mostly shown that taking phones away from people makes them a little happier and better at focusing.

The puzzles neither side of the STOE debate can explain

Puzzle 1: If phones are global, why is the damage local?

One of the most interesting wrinkles in the Smartphone Theory of Everything is that while phones are everywhere, the problems that they cause are often rising fastest and first in the richest countries — especially in the U.S.

Take, for example, the theory that smartphones make people sad. According to the latest World Happiness Report, happiness among young people has plummeted most severely in Western developed countries that speak English, such as the United States, U.K., Canada, Australia, and New Zealand.

On the other hand, “happiness at every age has risen sharply in Central and Eastern Europe,” the 2024 report said. In East Asia, happiness is increasing “at every age.”

The same is true for suicide. Emergency room visits for suicide attempts among young women soared across the Anglosphere in the last few years. But, as I’ve reported, the suicide rate among people ages 15 to 19 fell in most European countries in the last decade.

Or take attention deficit disorders. The surge in ADHD cases seems to be another U.S.-heavy phenomenon, with U.S. diagnoses vastly surpassing that of European and OECD countries. To understand rising anxiety and attention disorders in the U.S., we have to recognize the phenomenon of diagnostic inflation — i.e., medical providers expanding the definition of anxiety and ADHD to treat more cases.

In politics, the U.S. also seems to be an outlier in some maladies that are associated with smartphones. A 2020 paper on polarization in the West found that “affective polarization” — the ire that people feel for the party they oppose — rose fastest and first in the U.S., with most of the increase predating the smartphone age. The researchers wrote that polarization took off in the 1990s, right around the introduction of Fox News, which was nearly 20 years before the smartphone revolution took off.

As for smartphones creating populism and distrust, this also appears to be a disproportionately Western phenomenon. A 2022 Nature Human Behaviour review of 496 articles found that digital media is most strongly associated with declining political trust and growing populism in developed democracies, such as the United States. In autocracies and developing democracies, by contrast, the largest effect of digital media on politics seems to be that it increases political participation.

What do we make of all this? If you’re anti-smartphone, this pattern is a real problem for your theory; you can’t explain a localized epidemic with a global cause. But the smartphone defenders aren’t off the hook because the epidemic is real, and phones are clearly implicated in the mechanism. I think the answer is that smartphones are interacting with other phenomena that are distinctly Western or American and creating berserk local effects.

Puzzle 2: Is it really about the screens?

My favorite Jonathan Haidt argument is that phones replace play-based adolescence with phone-based adolescence. That is, the most important thing about phones isn’t what’s on the screen, but rather everything that’s off the screen when you’re lost gazing into your pocket device. I’d broaden the point beyond adolescence: for people of every age, the smartphone’s deepest effect may not be what it delivers but what it displaces.

In 2025, researchers found that randomly removing internet access from smartphones produced a range of benefits, including improved mental health, subjective well-being, and the ability to sustain attention. More than 90% of the nearly 500 participants experienced at least one benefit. As best as the researchers could tell, the most significant reason for improved mental health and subjective well-being came from participants spending more time “socializing in person, exercising, and being in nature.”

Another randomized trial that paid people to deactivate Facebook before the 2018 midterm elections also found that people spent more time with friends and family. From that paper:

The 60 minutes freed up by not using Facebook … were allocated to both solitary and social activities offline. Solitary television watching increases by 0.17 points on our scale; other solitary offline activities increase by 0.23 points, and time devoted to spending time with friends and family increases by 0.14 points.

It’s very hard for researchers to control what participants are doing with their phones. But it’s not hard for researchers to see what people are doing when they’re not on their phones. They sleep more! They socialize more! They go outside more! (And, yes, they watch TV more.)

And yet, they also learn less about the world. There is a widespread fear that phones and social media are a source of misinformation and conspiracy theories. But the most careful randomized studies have found something close to the opposite.

“When we pay people to stay off of social media, the strongest finding is that they get less information,” said Matthew Gentzkow, a Stanford economist.

Across the two studies that randomized people to deactivate Facebook and Instagram, Gentzkow and his colleagues found that going off social media had minuscule effects on political polarization, but it also left people measurably less informed about current events. The first study paid nearly 1,700 Americans to deactivate Facebook for four weeks before the 2018 midterm elections. People who deactivated were happier, less anxious, and less politically polarized. But they also knew less about the news.

In the second study, more than 35,000 Facebook and Instagram users were paid to go dark for six weeks before the 2020 presidential election. Deactivation had little to no effect on polarization, views about election legitimacy, candidate favorability, or voter turnout. And again, people who left came back knowing less about what was happening in the world.

The story we tell ourselves is that smartphones and social media pump nonsense into the national bloodstream and that logging off would produce a better-informed citizenry. But these studies suggest that while smartphones and social media do spread some misinformation, they also function as a primary news delivery system.

If smartphones are making us depressed, conspiratorial, and anxious, it might be because the news is structurally becoming more depressed, conspiratorial, and anxious. But the STOE focuses our attention, policymaking efforts, and ire only at the accelerant and not the source of the flames.

In 2023, researchers at the University of Pennsylvania, Indiana University, and London Business School used AI to trace positive vs. negative words across around 1 billion newspaper articles from the 1850s to the 2020s. For more than a century, news positivity hovered around a stable average. But after the 1960s, negativity surged.

“News coverage has just gotten more and more negative every decade in the last 50 years, especially when you adjust for economic recessions,” Wharton economist J. H. van Binsbergen, a coauthor on the paper, told me.

As I wrote in an essay for The Atlantic, I suspect that as the media industry got more competitive in the past few decades, publishers desperate to command reader attention doubled down on the old clichés that “if it bleeds, it leads” and “bad is stronger than good.” In this light, smartphones didn’t independently make the news more depressing; they made it easier to read depressing news.

Smartphones as attention alcohol

People like to compare smartphones and cigarettes. That’s wrong. Honestly — and I say this for research purposes, exclusively — I wish smartphones were cigarettes.

If the phone experience were a mass-manufactured bundle of chemicals that we could test in isolation against a control group, this whole essay could be one sentence long: We did some tests on phones, and they’re giving people cancer.

Unfortunately — again, for research purposes, only — smartphones are not tobacco. Everybody’s online experience is unique, which means that everybody is effectively smoking a slightly different cigarette. No surprise, then, that observational analyses struggle to prove causality, and randomized experiments to prove causality are typically brief and limited.

You can’t assign child participants to heavy social media use for a full year, and you certainly can’t randomly assign kids to use their smartphones in a specific way for a long time. (“Hi Madison, we need you to spend your entire junior year marinating in angry left-wing Reddit posts to measure the impact of online Marxism on the teenage mind” is not a plausible design study.)

If you force participants to deactivate Facebook in a study, they might just download Twitter; in fact, that’s exactly what happened in a 2020 study. If you force them to give up their phones entirely, they’ll still need to maintain desktop web access; in fact, that’s exactly what happened in a 2025 study.

A better metaphor is “alcohol”: fun for many, dangerous for some. Similarly, smartphones might have small effects on the majority population and large effects on a minority population. In 2020, Instagram’s own analysis concluded that its product “made body image issues worse“ for one-third of teenage girls who already experienced body image issues. That’s a lot! But it also implies that for the majority of teenage girls, Instagram had a small or negligible effect. And, of course, most people are not teenage girls.

Cellphone Cassandras can exaggerate the conclusions of careful research, which often show small overall effects. But critics of the STOE often ignore practical conclusions by fixating exclusively on small overall effect sizes, despite the evidence of significant long-tail effects. (Then again, Cassandra was famously cursed to utter true prophecies that would never be believed.)

To summarize everything I’ve just said in one sentence: The smartphone is a displacement machine and an information-delivery system, whose harms are filtered through the structural weaknesses of any given culture.

But what about the specific claims? Are phones really terrible for sleep? How strong is the evidence that they cause anxiety outside of specific populations? Are they turning us into conspiracy theorists?

In Part 2 of this essay, I’ll do my best to place 10 popular claims into three buckets: Strong evidence, mixed evidence, and weak evidence. I’ll go through each of the most common claims and give my sense of what research actually shows.

Recommended reading: