Why birth rates are falling everywhere at once
Across rich and poor countries alike, fertility began sliding within a few years of each other — clustering around the moment smartphones and fast mobile internet arrived. Here’s the data behind the debate, rebuilt from UN, World Bank, Eurostat and other public figures.
The long view: this decline is decades old
Using the UN’s own numbers back to 1950, the shaded “smartphone era” shrinks to a thin sliver at the right-hand edge. In Korea, Mexico, Indonesia and Brazil, roughly 90% of the total fall in fertility had already happened before smartphones existed. Whatever phones did, they arrived late to a decline that was already most of the way done — which is why a single time-series can’t tell you smartphones caused it.
Fertility rate since 1950 — UN World Population Prospects
High-fertility countries fell toward (or below) replacement decades before the smartphone era began.
Do births and marriages move together?
Each panel overlays a country’s fertility rate (purple, left axis) on its crude marriage rate (blue, right). In Korea the two fall almost in lockstep — powerful support for “fewer marriages, not fewer kids per marriage.” In the US and France the marriage line drops but fertility holds up better, because so many births there now happen outside marriage. Same shift, different plumbing.
Births vs marriages, per country
Fertility rate — births per woman (left)Crude marriage rate — per 1,000 (right)
The direct evidence: fewer people are marrying
Here the “fewer couples” claim stops being an assertion. The crude marriage rate — marriages per 1,000 people each year — has fallen across the board. Korea is the clearest case: its rate more than halved (≈9 to under 4), and because births there still overwhelmingly happen inside marriage, its fertility collapse essentially is a marriage collapse. People are also marrying much later (US women’s median first-marriage age rose ~25→29, France ~28→33 in just two decades).
Crude marriage rate (marriages per 1,000 people)
Fewer marriages per year — the partnership-formation channel behind “fewer couples.”
Coverage caveat: this series is OECD / rich-country weighted — Brazil, Indonesia and Nigeria aren’t covered — so read it as evidence about high-income and East-Asian countries, not the whole world.
Fewer kids per woman — or just fewer couples?
The honest test. This is completed fertility — the number of children women actually ended up with, plotted by the year they were born (not a single year’s snapshot, which delayed births distort). If the collapse were purely “fewer couples,” women who do have kids should still end up with about as many as before. The answer is split: in the US, UK and France completed fertility held remarkably close to ~2, so the scary single-year figures overstate the real per-woman decline. But in Italy, Spain and Japan women genuinely ended up with far fewer (~1.4). So it’s partly fewer couples, partly a real drop in kids per woman — and it depends where.
Completed fertility by woman’s birth cohort
X-axis = the year the woman was born; the line is how many children she had by the end of her childbearing years. Data ends with women born ~1974.
Caveats: this is per woman, not per married couple — no ready-made cross-country marital-fertility series exists (even the Human Fertility Database doesn’t carry it yet). It also only runs to women born ~1974, so it can’t yet judge the post-2010 dip; and Korea, the cleanest “marriage-collapse” case, isn’t in the cohort database.
Is fertility falling among married women?
The cleanest test — births per married woman — isn’t published as a ready-made cross-country series (even the Human Fertility Database doesn’t carry it yet; it can be computed for EU countries from Eurostat). So here’s the honest workaround using what’s open — two things we can measure directly: the share of women aged 15–49 who are married has fallen sharply (US 66% → 52%), while completed fertility per woman (chart above) held close to ~2 in the US, UK and France. Together that points to fewer married women far more than married women choosing fewer kids — the “fewer couples” reading. But where fertility fell hardest (Italy, Spain, Japan), completed fertility also dropped, so married-couple fertility isn’t perfectly flat there either.
Share of women aged 15–49 who are married or in a union
Fewer women are partnered at childbearing age — the compositional side of the decline.
Honest note: this is an indirect answer. A true “births per married woman” rate isn’t published as a ready-made cross-country series — though it can be computed for EU countries from Eurostat’s births-by-marital-status data (a possible next step). Share-married falling, plus roughly flat completed per-woman fertility, is consistent with married-couple fertility being fairly stable in the West — but it doesn’t measure it directly.
The direct measure: do married women have fewer kids? (EU)
And here’s the thing we couldn’t get before — a real marital fertility rate, computed from Eurostat: the share of married women (aged 18–49) who give birth in a given year. The answer is striking — among married women, fertility is flat or even rising (Netherlands 7.8% → 9.0%, Germany 5.4% → 7.8%) even as total fertility fell. So the birth-rate collapse is almost entirely about fewer people marrying — not married couples choosing fewer kids.
(Why a % and not “children per married woman”? A lifetime children-per-married-woman figure is badly distorted — young married women are a tiny, self-selected group — so the honest measure is the annual rate. Children-per-woman for the whole population is the fertility-rate chart near the top.)
% of married women (18–49) who give birth each year
Computed from Eurostat. Roughly flat-to-rising while total fertility fell — i.e. the decline is compositional.
The big caveat — selection: as marriage becomes rarer and more deliberate, the women who still marry are increasingly those who want children. So married-women fertility holding up doesn’t mean “couples are fine” — it partly means only the baby-inclined now marry. A real, direct measure, but not a clean counterfactual. EU-only (Eurostat coverage — the US, UK and Japan aren’t in this dataset).
…and are unmarried people just having the kids instead?
No. You’ll often see the “share of births outside marriage” rising and hear that childbearing just moved outside marriage — but that stat is a ratio, not a rate: it climbs whenever married births fall, and says nothing about whether unmarried people have more kids. Here’s the actual rate — the share of unmarried women (18–49) who give birth in a year. It’s low and flat-to-falling: unmarried ≈ 3–5% vs married ≈ 5–9%. So most unmarried people — cohabiting couples very much included — simply aren’t having kids. The rise in “births outside marriage” is a composition effect, not unmarried people becoming more fertile.
% of unmarried women (18–49) who give birth each year
Eurostat; unmarried = all women not currently married (cohabiting, single, divorced, widowed). Low, and not rising.
Are people more isolated? Living alone became the norm
Yes — the drift toward living alone is real and nearly universal. The share of one-person households roughly doubled across the rich world (US 13% → 28%, Japan 16% → 35%, Sweden past 40%). Fewer people share a home, which fits the picture of fewer and later partnerships. On top of this, US time-use data shows people — especially the young — spending more hours alone since around 2010, and surveys record a “sexual recession”: younger cohorts report less sex and fewer partners than earlier generations, not more.
Share of one-person households
More people living alone — consistent with fewer partnerships forming.
Honest caveat: living-alone rates are also pushed up by ageing populations (elderly people, often widowed, living alone) and by rising wealth (people can afford their own place). So this is suggestive of a retreat from partnership, not proof of it — read it alongside the marriage-rate fall, not on its own.
Did housing get too expensive?
A popular explanation: young people can’t afford the space to raise kids. Real (inflation-adjusted) house prices roughly doubled in the US and rose ~4–5× in the UK and Australia. But this can’t be the universal cause — Germany’s real prices barely moved and Italy’s actually fell, yet both have some of the lowest fertility in the world. Housing plausibly contributes where it exploded; it can’t explain the countries where it didn’t.
Real house prices (index, 2010 = 100)
Inflation-adjusted. Note Germany (flat) and Italy (falling) sit alongside equally low fertility.
Did people stop going to church?
Religiosity tracks fertility loosely — more-religious people and places tend to have more children, and frequent religious attendance has fallen across the West (US ~60% → ~39%). But the fit is loose: Korea and Japan were never highly religious yet have the lowest fertility, and secularisation is gradual while the recent fertility drop is sharp. A contributing current, not the switch.
Share attending religious services frequently
World / European Values Survey. Frequent attendance has fallen across most of the West.
The car-seat theory — clever, but far from global
One much-shared paper (Nickerson & Solomon, 2020) argues US car-seat laws — now requiring seats up to age ~8 — quietly act as contraception: a third seat won’t fit across a normal back seat, so some families stop at two. The study is US-only, but the mechanism plausibly extends across the car-dependent Anglosphere — the UK (restraints to age 12 / 135cm), Australia, Canada and New Zealand all pair strict car-seat rules with heavy car use, so it may bite there too. What it can’t be is the global driver — fertility collapsed just as hard where the mechanism barely applies: no comparable law, transit-heavy cities, or low car ownership.
Figures are each country’s latest fertility rate. Most rich countries do regulate car seats — the point isn’t that only the US does, but that the specific “no room for a third seat” mechanism needs car-dependence + strict age limits + small cars, a combination absent across much of the world that saw fertility fall anyway.
What the chart shows — and the debate
A Financial Times analysis by John Burn-Murdoch argued that fertility across very different countries began declining at strikingly similar moments — the US, UK and Australia from around 2007; France and Poland from about 2009; Mexico and Indonesia from around 2012; and several West African countries from 2013–2015. The common thread it pointed to was the spread of smartphones and social media reshaping how young people spend time, form relationships and partner up.
A supporting working paper by Nathan Hudson and Hernán Moscoso-Boedo (University of Cincinnati) tracked the rollout of 4G networks in the US and UK and found that areas wired for fast mobile internet earliest saw birth rates fall earliest and furthest.
The thesis is contested. Critics point out that fertility was already in long-run decline before smartphones, that the effect looks concentrated among teenagers (visible in the “Teen fertility” view above), and that correlation with technology rollout is not proof of cause. Switch metrics and countries above and judge the pattern for yourself.
Sources
Charts are rebuilt independently from public data; the FT article is cited only as the source of the smartphone-timing thesis. The long-view fertility chart uses UN World Population Prospects; the interactive explorer uses World Bank figures (themselves largely derived from the UN series, so the two are near-identical). The shaded 2007–2015 “smartphone era” band is an illustrative marker, not a data series. The marriage-rate and births-outside-marriage series cover mainly OECD and East-Asian countries, so they evidence the “fewer couples” story for high-income nations rather than the whole world.