Smartphones and the Global Baby Bust

The numbers that opened John Burn-Murdoch's recent Financial Times piece are the kind that stop you mid-scroll. South Korea's fertility rate has fallen below 0.8 — less than half the 2.1 replacement threshold that keeps a population stable. The US, the UK, Mexico, Indonesia, Turkey, Iran: all declining, all on roughly the same downward trajectory, all bending at roughly the same moment in history. Around 2008, give or take.

That last detail is the one worth sitting with. Demographers have long documented how fertility falls as societies develop — lower infant mortality, rising female education, urbanization. These transitions follow a recognizable pattern and a recognizable timeline tied to a country's stage of economic development. What Burn-Murdoch argues, in a conversation with guest host Christine Emba on Vox's The Gray Area, is that the current decline doesn't fit that pattern. Mexico's fertility is now lower than the US's — and you cannot explain that through female education levels or income or urbanization rates. Something else is happening, and it's happening everywhere at once.

The something else, Burn-Murdoch suggests, is the smartphone.

The externality nobody priced in

I want to be precise about the mechanism here, because it matters — and because the mechanism is one I recognize from a different beat entirely.

Burn-Murdoch is not arguing that phones emit radiation that affects reproductive biology. The hypothesis is behavioral and structural. His core claim: smartphones, and the engagement-optimized platforms they carry, extract unstructured social time from users at scale — time that would otherwise be spent in the face-to-face interactions from which relationships, and eventually families, emerge.

Research cited by Burn-Murdoch from WBUR's Here & Now shows that Americans' face-to-face socializing has declined measurably over the same roughly 15-year window in which screen time climbed. The correlation isn't proof of causation, but the researchers Nathan Hudson and Hernan Moscoso Boedo, both at the University of Cincinnati according to UC's own news office, found a more compelling signal: areas in the US and UK where high-speed 4G mobile networks rolled out earlier saw birth rates — particularly among younger cohorts — fall faster than areas where that connectivity arrived later. The rollout was largely random with respect to income and demographics. It just happened to hit urban areas before rural and mountainous ones. That quasi-randomness is exactly what makes the finding worth taking seriously.

What Burn-Murdoch describes is, structurally, an externality. Engagement-optimized platforms are designed to maximize time on screen. That's rational behavior from a platform's perspective, and it's individually rational behavior from a user's perspective in any given moment. But the cumulative effect — the aggregate extraction of unstructured social time across millions of people, over years — produces a harm that no individual user chose and no individual platform intended to cause. The same logic applies to carbon emissions: individually rational, collectively catastrophic, and almost entirely invisible to the market mechanisms meant to correct it.

This is not a metaphor deployed for rhetorical effect. It's a diagnostic frame. And it matters because it tells you something important about why the policy responses to date have failed so completely.

The mismatch between tools and problem

Governments have not been idle. China reversed its one-child policy and introduced tax disincentives on contraceptives. South Korea has spent years on pronatalist subsidies. European governments have experimented with baby bonuses. The results, Burn-Murdoch says, are uniform and underwhelming: "They tend to be sort of taking a steeply downward-sloping line and making it slope slightly less steeply downwards."

That's not nothing, but it's also not a solution. And the reason it isn't is precisely the externality logic. A baby bonus works on the financial margin — it moves the needle for couples who are already stable, already housed, already in a long-term relationship, and simply doing the math on whether they can afford another child. It does nothing for the upstream problem, which is that fewer people are forming those stable relationships in the first place.

Burn-Murdoch is direct about where the fertility decline is actually concentrated: not among highly educated, high-earning women who are delaying parenthood for career reasons — the narrative the political right tends to reach for. "What we're primarily looking at," he says, "is a decline among those in the bottom half or sort of bottom third of society. Young men and women who generally don't have a bachelor's degree, have lower incomes." This group is having fewer children, but more fundamentally, they are coupling up less. The people who do form stable, secure relationships are still having children. The friction has moved upstream of that.

Housing precarity is part of the story — Burn-Murdoch estimates it may account for as much as half of the birth rate decline in the US and UK since the 2000s. But it can't be the whole story, because Finland, which has not experienced the same housing crisis, is nonetheless seeing its fastest-growing household type be young adults living entirely alone. The floor dropped even where the ceiling stayed.

Why the television comparison breaks down

The obvious pushback — and Emba raises it directly — is that we've been here before. Television's mass adoption in the 1950s and 60s also displaced face-to-face time. Birth rates dipped. Society adapted and recovered.

Burn-Murdoch's response is measured, but I think the more important answer runs deeper than what he offers. The television analogy assumes the mechanism is simply time displacement — hours in front of a screen instead of hours in a room with other people. But engagement-optimized social platforms introduce something the television era didn't have: a feedback loop. The platforms learn from your behavior, serve you content calibrated to hold your attention longer, and in the dating app context specifically, are structurally incentivized not to facilitate the connection they nominally exist to provide. As Burn-Murdoch puts it, dating apps are "essentially screen time apps" — their revenue model depends on keeping you swiping, not on helping you find a partner and delete the app.

Television delivered a fixed signal to a passive audience. The smartphone delivers an adaptive signal to an active participant whose engagement data is continuously harvested and used to optimize for more engagement. The variable that makes smartphones categorically different from the television is not screen time per se — it's the optimization loop. It's the difference between a road and a treadmill.

That distinction matters enormously for policy. You can regulate a road. A treadmill that gets faster when you try to slow down requires a different kind of intervention.

The anti-natal feedback loop

One of the more unsettling dynamics Burn-Murdoch identifies is the doom loop that a declining birth rate creates around its own causes. Fewer people have children. Fewer people are therefore regularly exposed to children. The social norming that once happened organically — watching your peers navigate early parenthood, absorbing the idea that children are a plausible and even appealing part of life — quietly erodes. Children become less visible in restaurants, on planes, in the mental landscape of what adult life looks like. And the culture around them curdles accordingly, producing exactly the conditions that make the next cohort less likely to have children.

This is a feedback loop with no obvious individual-level intervention. The person choosing not to have children is not doing anything wrong. The aggregate outcome is still a problem.

Boris Johnson posted on X that "falling birth rates aren't a disaster, they're the best bit of global news in a long time" — and the argument has real force at certain scales. Fewer people consuming fewer resources has genuine ecological merit. But it contains a distributive blind spot. The countries now seeing the steepest fertility declines in the developing world are, as Burn-Murdoch notes, facing Japan's demographic challenges without Japan's resources, without Japan's robotics sector, and without the welfare infrastructure that rich countries built out before their fertility collapsed. The slowdown hits everyone; the consequences land hardest where the buffers are thinnest.

The structural problem with individual solutions

Burn-Murdoch's preferred near-term solution — less aimless screen time, more unstructured social interaction, perhaps a cultural shift that makes doom-scrolling as socially declassé as smoking — is genuinely appealing. It's also, on its own, insufficient in a way that should be familiar to anyone who has watched the "just drive less / just recycle more" era of climate communication come and go.

The individual-behavior frame mislocates the agency. The platforms extracting unstructured social time are not neutral infrastructure. They are optimized systems producing a collective harm that emerges from millions of individually rational choices. The appropriate policy response to that structure is not a cultural nudge toward putting the phone down — though cultural nudges can help at the margins. It's regulatory: restructuring the incentives that make engagement maximization the default.

Burn-Murdoch is cautious about prescribing exactly how we get there, and that caution is reasonable — the track record on social media regulation is not encouraging. But the framing matters. Demographic policy has been stuck in a loop of baby bonuses and pension-age adjustments precisely because it keeps treating the symptom. The mechanism is upstream, it's structural, and it's operating at a scale that individual choices, however well-intentioned, are not equipped to address alone.

The 2008 inflection point in birth rates is not a demographic curiosity. It is a compounding event — and compounding harms have a way of becoming much more expensive the longer you wait to intervene at the level where they're actually generated.

---

By Olivia Meng, Climate & Environment Correspondent