Population Collapse

• Births rose 6.8% to 254,500 in 2025, the largest annual rise in 15 years, and the TFR rose to 0.80 from 0.75 (Ministry of Data and Statistics, reported by Korea JoongAng Daily, 25 Feb 2026). A later final release was reported at 254,300 births (Korea Times, Aug 2026).
• Deaths were 363,400, so the population still shrank. Reports attribute part of the rise to the 'echo boom' cohort and a marriage rebound; 37.3% of births were to mothers aged 35 and over (press reports of the final release, Aug 2026).
• Seoul's TFR was 0.63 in 2025. The 0.72 of 2023 remains the record low.

China's National Bureau of Statistics reported 7.92 million births in 2025 (down from 9.54 million), 11.31 million deaths, and a population decline of 3.39 million to about 1.405 billion (NBS figures as reported by press, 19 Jan 2026). The birth rate of 5.63 per 1,000 is the lowest since 1949, and people aged 60+ were 323 million, 23% of the population. A national childcare subsidy of 3,600 yuan per child under three was introduced in 2025.

Japan's health ministry reported 671,236 births to Japanese nationals in 2025 (down 14,937), a TFR of 1.14, and 1,589,489 deaths, a natural decrease of 918,253, the 19th straight year of decline (Ministry of Health, Labour and Welfare, 3 Jun 2026, as reported by Nippon.com). The 2025 census preliminary count was about 123 million, a record 2.5% fall in five years.

• Italy: about 355,000 births in 2025 (down 3.9%), a TFR of 1.14 (from 1.18) and about 652,000 deaths; 25.1% of residents were aged 65+ on 1 January 2026 (ISTAT, 31 Mar 2026).
• France: INSEE reported about 645,000 births against 651,000 deaths in 2025, the first natural decrease since the Second World War; net migration of about 176,000 kept the population growing (INSEE, Jan 2026, as reported by press).

The CDC's National Center for Health Statistics provisional report counted 3,606,400 births in 2025, down 1%, with a general fertility rate of 53.1 births per 1,000 women aged 15-44 (NCHS, 9 Apr 2026). A later NCHS Data Brief (No. 571, Oct 2026) was reported to give a revised 3,612,012 births and a rate of 53.2.

No new UN World Population Prospects edition has appeared since 2024; its baseline still puts the world TFR near 2.23 in 2026 and the population peak near 10.29 billion around 2084. An independent estimate using newer birth data puts the 2026 world rate at about 2.19 or lower, which suggests the world is at or just under replacement.

Not refreshed: the US total fertility rate (the 1.62 shown is from earlier data; one report cites 1.57 for 2025 but it could not be checked against the NCHS tables), France's TFR (1.66 shown; a press summary cites 1.56), China's TFR, India's TFR, Hungary's and Singapore's fertility, old-age dependency ratios, and 65+ shares, which remain UN WPP 2024 figures.

• Follow the rate, not the headcount. Population keeps growing for decades after fertility falls; momentum hides the turn.
• 2.1 is the line. Below ~2.1 children per woman, each generation is smaller than the last.
• Sober, not alarmist. Depopulation is treated as a problem to manage, not a doom to fear.
• Every figure is dated and sourced. Rates, projections and ratios from the UN, World Bank, OECD and national statistics.

Source: UN WPP 2024; The Lancet / GBD 2021 (2024).

At the record low of 0.72 children per woman (2023), and still only 0.80 in 2025 after two years of modest rebound, each Korean generation is well under half the size of the one before. Seoul is the demographic frontier the rest of the rich world is walking toward.

Source: World Bank; UN (2023); national statistics (2025 for Korea, Japan and Italy).

• Built-in delay. A large generation of young adults keeps births up even after family size shrinks.
• Longer lives mask it. People live ~73 years, so a shrinking birth cohort takes decades to shrink the whole population.
• The half-life. At a TFR of 0.8, annual births halve roughly every 22 years, compounding quietly.
• Peak, then fall. The world still grows to ~10.3B in the 2080s, then declines: the first sustained fall since the Black Death.

Source: UN WPP 2024.

Nobody decided to end the family. Hundreds of millions of people, weighing cost, time and opportunity, simply chose fewer children, later, and the arithmetic did the rest.

Falling fertility is not a failure of will. It is the rational response to a world that made children expensive and the alternatives abundant.

A scatter of fertility rate (TFR) against GDP per capita shows Niger (poor, high fertility), India (middle), Israel (rich, high fertility, the exception), and France, the United States, Japan and South Korea (rich, low fertility). Source: World Bank; UN (2023); national statistics (2025 for Korea, Japan and Italy).

• The cost of a child. Raising one to 17 runs ~$310,000 in a US middle-income home, before college.
• Housing first. A home that cost ~3x income in 1980 now costs ~5x; family formation waits on a mortgage.
• The opportunity cost. A woman's peak fertility overlaps her peak education and career years: the timing trade-off.
• Cities, not farms. Urban children are dependents, not labour; the old economic reason for many children vanished.
• The mimetic shift. We want what our peers want, and status now flows to careers and autonomy, not big families.
• Technology of choice. Contraception, later marriage and abundant substitutes decoupled sex, partnership and childbirth.

Source: USDA; Brookings; U.S. Census.

Source: Economic History Association; USDA.

1. Fewer births
2. An older population
3. Higher taxes, thinner family support
4. Children cost more, feel riskier
5. Births delayed and foregone, which feeds back into fewer births

1. 01 High-highMany births, many deaths; population stable and young.
2. 02 Mortality fallsMedicine and food cut deaths; population booms.
3. 03 Fertility fallsUrban, educated families choose fewer children.
4. 04 Low-lowBirths and deaths both low; growth stalls, population ages.
5. 05 Below replacementFertility stays under 2.1; the population begins to shrink.

Source: Demographic transition model.

Source: UN WPP 2024.

Source: UN WPP 2024; UN DESA (2023).

• Median age soars. Japan's is ~49; Korea and Italy follow; the US ~39 and rising.
• The base narrows first. Schools empty before workplaces do; Korea is closing them by the hundred.
• Deaths overtake births. In Japan, Italy, China and much of Europe, more people now die each year than are born.
• It compounds. A smaller young generation means fewer future parents, deepening the next fall.

Source: UN WPP 2024; national statistics.

An economy grows by adding workers or making each one more productive. Take away the workers, and everything rests on productivity, while the bill for the old keeps rising.

Depopulation is not automatically poverty. But it is a permanent headwind, and it lands hardest on the promises made to the retired.

A dumbbell chart compares the number of working-age people per person aged 65+ in 1970 and in 2050 (projected) for South Korea, Japan, China, Europe and the United States. In every case the 2050 figure sits far below the 1970 figure, with South Korea showing the steepest fall and the United States and Europe the smallest. Source: UN WPP 2024 (working-age per person 65+).

• Shrinking workforce. Fewer workers caps output growth unless productivity accelerates to compensate.
• Pension strain. Pay-as-you-go promises buckle as contributors fall and claimants rise.
• Healthcare and eldercare. The over-80s cost health systems most, and there are far more of them.
• Slower growth. Fewer workers and older consumers mean structurally lower GDP growth.
• Deflationary drift. Ageing societies save more and spend less: Japan's low-growth, low-rate template.
• The great transfer. Wealth concentrates in the old; the young inherit later and buy homes later.

Source: IMF; OECD; Bank of Japan.

Source: OECD Pensions at a Glance; IMF Fiscal Monitor.

• State pensions (05): the largest promise, indexed, universal, pay-as-you-go.
• Healthcare (04): demand rises steeply with age.
• Long-term and elder care (03): the fastest-growing line, and often unfunded.
• Public debt service (02): heavier as deficits widen and growth slows.
• Everything else (01): defence, education, investment, squeezed last.

Source: IMF; OECD.

Source: Bank of Japan; Statistics Bureau of Japan.

1. 1966Romania bans abortion. Decree 770 doubles TFR to ~3.7, then it slides back, at terrible human cost.
2. 1979China's One-Child Policy prevents ~400M births; leaves ~30M 'missing' women and a ~1.0 TFR.
3. 2006Korea opens the taps: begins two decades and $270B+ of incentives; TFR still falls to 0.72 by 2023, before a modest rebound to 0.80 in 2025.
4. 2011Hungary goes all-in: spends ~5% of GDP on families; TFR rises modestly, then stalls.
5. 2024Korea declares emergency: creates the world's first Ministry for Population Strategy.

Source: Columbia; China NBS; Georgetown JIA; KSH.

Across roughly two decades of cash bonuses, parental leave and subsidised childcare, Korea's fertility rate did not stabilise; it dropped to 0.72 in 2023, and has since recovered only to 0.80 (2025). The clearest evidence that money, on its own, does not buy babies.

Source: National statistics; OECD; Georgetown JIA.

• Suppression is easy. Bans and one-child rules cut births quickly and durably.
• Restoration is not. Once small families become the norm, they stay the norm.
• Money shifts timing, not totals. Subsidies mostly move when people have children, not how many.
• Culture is the ceiling. Policy works only where it reinforces what people already want.

Source: The Lancet; PubMed.

A stacked area chart of world population by region (Africa, Asia, Europe, Americas, Oceania) from 1950 to 2100 shows Asia swelling through the 20th century and flattening around mid-century, Europe flat, and Africa's band widening steadily to become the largest-growing share by 2100. Source: UN World Population Prospects 2024.

• Ultra-low East Asia. Korea, Taiwan, Japan, China, Singapore: the frontier of collapse, TFR 0.7-1.2 (Korea 0.80 and Japan 1.14 in 2025).
• Ageing Europe. 1.3-1.7 and shrinking; migration is the only thing holding numbers up.
• China's cliff. Peaked in 2022; a workforce falling by millions a year, old before fully rich.
• The American exception. 1.62 and falling, but migration keeps the US the youngest large rich economy.
• India past the peak. Just dipped below replacement (~2.0); a vast but brief demographic dividend.
• Africa, the last frontier. Still ~4.3; home to most of the century's growth, and half its babies by 2100.

Source: UN WPP 2024; World Bank.

China's population peaked in 2022 and, on UN medium projections, could fall toward 0.6-0.8 billion by 2100. It is the first great power to grow old before it grew fully rich: the defining demographic wildcard of the age.

Source: UN (replacement migration); IMF.

Source: IMF; OECD; UN.

The adults of 2050 are already born, so the next quarter-century is close to fixed. But demography is destiny only until culture, technology or economics decide otherwise.

History rarely moves in straight lines. The current trend is powerful, but it should not be mistaken for fate.

Source: Compiled analysis.

• Managed decline. Societies adapt (automation, later retirement, migration) and shrink gracefully. The base case for the rich world.
• The trap deepens. Ultra-low fertility persists, ageing overwhelms budgets, growth stalls. Today's Korea, exported.
• The rebound. A cultural or technological shift (cheaper childrearing, pro-family norms) lifts fertility. Unproven, not impossible.

1. Does any rich country reverse below 1.5? The first true recovery would rewrite the playbook.
2. How fast does China age? The largest, least-tested demographic shock.
3. Can productivity outrun the workforce? The central bet: automation versus shrinking labour.
4. Do pension ages actually rise? The political test every ageing democracy faces.
5. Where do the migrants go? The flows that will redraw the global labour map.

A civilisation's most consequential decision is being made quietly, one family at a time, in a billion private conversations that no ministry can join. The result is not famine or war, it is simply fewer people choosing to make more. That is what makes it, for now, so hard to reverse.

1. Below the line, and spreading. Two-thirds of humanity, heading toward nearly all.
2. A choice, not a catastrophe. The first voluntary, global fertility decline in history.
3. Ageing is the near-term cost. The pyramid inverts before the headcount falls.
4. The welfare state was built young. Pay-as-you-go promises meet a two-to-one world.
5. Policy nudges; culture decides. Manage the decline; do not expect to reverse it.

• UN: World Population Prospects 2024.
• The Lancet: Global Burden of Disease 2021 fertility study (Bhattacharjee et al., 2024).
• World Bank: World Development Indicators, fertility and population (2023).
• OECD: Pensions at a Glance; Family Database (2023-24).
• IMF: Fiscal Monitor; research on migration and fertility.
• National statistics: Statistics Korea, Japan Statistics Bureau, China NBS, Eurostat, US CDC/NCHS.
• Georgetown Journal of International Affairs: South Korea pronatal spending.
• Figures marked 'est.' are compilations of the above.