Opportunity Knocks #149: The problem with just taking the phones away
Every week I share reflections, ideas, questions, and content suggestions focused on championing, building, and accelerating opportunity for children.
Jonathan Haidt gave a TED Talk in April in which he argued for “techno-skepticism”: a framework that would require technology companies to prove their products are safe for kids before releasing them and hold them accountable when they don’t.
Haidt has done important work building the evidentiary case of why this perspective may be necessary. His three underlying rules informing this mindset shift: 1) protect brain development through puberty; 2) prioritize people and books over screens; and 3) beware of artificial relationships for minors, are reasonable.
But as Haidt’s argument moves from books and talks into law, policy, and parenting culture, I think two questions are left open. To be clear, Haidt himself is more nuanced than many of the policies and parenting norms now moving under his banner. My concern is how individuals and institutions translate and adopt techno-skepticism.
The first question concerns the remedy itself. What happens if techno-skepticism degrades into a single action: removing technology from childhood, as society increasingly operates within ideological, cultural, and policy binaries? If that happens, I can’t help but wonder whether it is the same mistake in one specific respect—acting at scale before the evidence was in—that we made when we handed young people smartphones in the first place.
The second question concerns the target: techno-skepticism focuses on what companies must prove before children gain access to technology. That is necessary, but what do children need after access is restricted, delayed, or redesigned?
Haidt is absolutely right that we need a better path.
I think that path is developmental stewardship: restriction where necessary, graduated exposure where possible, hostile systems kept unreachable, scaffolding built into devices, and a much higher burden of proof for any technology company seeking access to childhood.
Let me concede the strongest case for removal up front, which is why I think it is gaining so much traction: removing a known harm with no developmental upside is justified and does not require a fully specified replacement.
The challenge is that the problematic (harmful in too many instances) thing is the same thing kids will eventually have to master. The medium through which adult social and economic life is now conducted is technology, and I don’t think the incentives are strong enough for us to go backward. This fundamental constraint—harm-entangled-with-a-required-competence—is perhaps the most difficult challenge in this entire conversation and the crux of many modern parenting dilemmas.
A skeptic will object that I am conflating the genuinely necessary (browsers, messaging, search, the literacy adult life actually demands) and the genuinely hostile (the engagement-maximizing feed, the variable-reward loop, the companion bot). This is fair. Though necessary isn’t the same as safe, a group chat can be the place where a child exercises judgment and where cruelty travels. What’s clear is that a child does not need an early relationship with hostile elements, and these kind of relationships should be deferred. But the required and the hostile now ship on the same device, through the same interfaces, often inside the same apps, and that entanglement, not the necessity of any single feature, is the thing removal cannot cleanly cut because entanglement is the business model. What could change that is a different kind of builder: public-benefit, kid-first, with no incentive to keep a child scrolling or chatting. I won’t pretend that the builder exists at scale yet; absent one, this likely takes significant procurement pressure, regulation, or a high-net-worth individual, funder, and nonprofit willing to lose money.
Haidt’s essential case against technology—specifically smartphones and social media, and increasingly specific forms of AI, including artificial companions—is that it replaces the inputs children need to thrive: the in-person bonding, the shared meals, the play, the real social life that, as humans, we run on. That’s a displacement argument, and I agree with it. But it cuts against removal as a complete strategy, not for it. Removal can clear the space; it doesn’t fill it.
Haidt himself recognizes this. His advocacy for a “play-based childhood” through initiatives like Let Grow is built on the premise that we must actively restore independent play, not just subtract screens.
A fair reader should grant that removal is a lower-risk move: undoing an intervention that has caused some harm is not the same gamble as launching one in which the harm is unknown.
But the defense of removal that society seems to be running with rests on a quiet assumption: that taking the phones away automatically returns kids to a healthy baseline.
Unfortunately, the baseline itself is part myth. American childhood was troubled long before smartphones; driven by decades of changing parenting patterns, hyper-scheduling, and the loss of accessible public spaces, independent play, and unsupervised mobility. Anxiety and obesity were already climbing. The pre-smartphone world we’d be reverting to wasn’t well; it was just differently unwell.
So removing technology doesn’t necessarily restore kids’ wellbeing. It might restore the prior set of deficits, now minus a phone. Maybe that would be an improvement. But less screen time is not the same as more of what children need, and a policy that delivers the first while assuming the second is making exactly the move it accuses the tech companies of: acting at scale on a hopeful theory before the evidence is in.
This isn’t what Haidt argues, but the broader cultural movement that lags behind him is starting to treat removal itself as the complete theory of child development. And we are starting to see the first major real-world test of this removal-as-strategy momentum play out in our schools.
Most states have now banned or restricted the use of phones in schools. The policy momentum is extraordinary. The evidence base is emerging, but it is nowhere near as settled as the policy momentum suggests.
Again, this isn’t Haidt’s case, but it is the test society is actually running, so the results matter. The largest study ever of school cellphone bans, covering roughly 4,600 schools, found mixed results of device removal. Teachers reported fewer distractions, but little evidence of improved academic achievement or better behavior. The pattern repeats across the wider literature: The first NBER study to examine the causal effects of school smartphone bans specifically on mental health found no clear evidence that ban policies reduced screentime or improved psychological wellbeing (a caveat worth stating: the study drew from data in three states, with limited post-ban data); a review of 22 studies across 12 countries was inconclusive; a Florida study is the partial exception, as it found a short-term suspension spike concentrated among Black students, but second-year gains in test scores and attendance once that period passed.
One reason these bans look weak in the data is that most are narrow. Kids drop their phones in a Yondr pouch when they enter the school building and binge after the final bell, which is probably why studies keep finding no drop in total screen time. But the deeper point holds even where bans work as intended. School bans may improve the school day, yet they do not, by themselves, rebuild the missing developmental inputs like play, belonging, sleep, friendship, adult attention, and practiced judgment.
Let me be unambiguous, because this is where I am most likely to be misread. The foundation of childhood has to be offline. No child needs digital fluency more than sleep, friendship, movement, books, play, boredom, chores, and adults who know and care about them. My argument is not that digital life belongs at the center of childhood. It is that, before full autonomy, children need a supervised on-ramp to digital life built on what we know about healthy child development.
To be fair, some of the strongest arguments for school phone bans are hard for any study to capture. You can’t easily quantify the cultural shift of a hallway or cafeteria where kids are looking each other in the eye, joking, flirting, and simply being together instead of staring into glowing rectangles. Those gains are real, and they matter deeply to a school’s soul.
But I think we have to be careful not to mistake a localized truce for a systemic victory. A smartphone ban may improve school culture—a worthy goal—but it does nothing to prepare a child for the 3:00 PM to 10:00 PM digital wilderness at home, or the total autonomy awaiting them at graduation.
And there are likely unintended consequences.
Blanket removal policies will hit hardest in communities where families have less time and fewer alternatives. The wealthy family can manage nuanced use, fund structured activities, and enforce boundaries at home. A low-income family may get the ban without the replacement infrastructure. Again, less screen time is not the same thing as more play, more belonging, more adult attention, or more opportunity.
The only equity-robust answer to protect the kids hurt worst by the status quo is one where the scaffolding is structural: built into the device, the defaults, and the school curriculum, rather than dependent on a parent with time.
If techno-skepticism only leads us to removal, we risk repeating the same mistake in reverse: acting at massive scale before we know what actually helps children thrive. The asymmetry is real: too much access too early means a child spends vulnerable years inside the harm; too little means a child learns later, probably with less gudiance. Those are not the same risks.
The mistake was not only the technology itself, though some technology is plainly harmful. The mistake was absolutism: the belief that a single binary action—give everyone a phone, take everyone’s phone away—can substitute for the messy, time-consuming, deeply human work of helping children navigate a complicated world.
The better frame is developmental stewardship: restriction where necessary, graduated exposure where possible, hostile systems kept unreachable, scaffolding built into devices and schools, and a much higher burden of proof for any technology seeking access to childhood.
We got it wrong when we handed every twelve-year-old a smartphone and called it progress. We should be careful not to get it wrong again by ripping it away and calling that progress, too.
The goal is not more technology or less technology. The goal is to organize childhood around judgment, friendship, play, belonging, competence, and care.
Until next week, be calm and be kind,
Andrew