Field report №008 · Demographic change in the G8, plus one more · Sources: UN, IMF, OECD

The Wedge.

The G8 — and Spain, because a starting point is just a starting point — has been producing fewer children than it needs, and more elderly than it can staff, for about forty-five years. This dossier lays out what the United Nations' own 2024 projections say breaks next — in what order, and by how much. No adjectives, where the numbers won't reach for them.

The data ↓ Back to the bureau

9 of 9 countries below the 2.1 replacement line JP · IT · ES · RU shrinking today · DE at peak Japan: 29.6% of residents are 65+ Official horizon: 2100 (UN WPP 2024)

The mechanism

Two forces, one wedge. No reversal in the planning window.

It is not a policy failure. It is an arithmetic identity with a long fuse. You do not vote it away; you choose how to sit with it.

Fertility is stuck under replacement.

All nine countries are below the 2.1 needed to hold a population constant. Worse: the mothers currently of childbearing age form a smaller cohort, so even a full recovery to 2.1 today would not stop a shrinking population for roughly thirty years. That is population momentum — it outlives every government it happens under.

Longevity keeps climbing at the same time.

Life expectancy runs from 73.3 years in Russia to 84.9 in Japan today, and the UN projects it into the high 80s for most of the group by 2050. More survivors per birth cohort, living longer, needing more years of care each. The two movements compound; neither is close to reversing.

Hence the wedge.

A shrinking working-age base must support a growing elderly population who each live more years. Every problem in this dossier — pensions, labour, care spending, town finances, recruiting for the army, the immigration fight — is the same wedge seen from a different corner of the building.

The data

UN World Population Prospects 2024, medium variant.

Official projections, pulled straight from the UN's own World Population Prospects 2024 bulk files on 2026-08-24 — the source the projections were published in, not a mirror of them. The medium variant is optimistic by construction: it assumes fertility slowly converges back toward 2.1. Even under that assumption, five of the nine countries are smaller in 2100 than today.

Exhibit A — Population paths2024 ≡ 100, UN WPP 2024

Index of 2024 population. The two divergent stories: the shrinking club (JP, IT, ES, RU, DE) and the migration-clubbed growth club (US, CA, GB, FR). France is the only one that is essentially flat. The 2100 labels are head counts in millions.

Exhibit B — Total fertility rate, 2024vs. 2.1 replacement

All nine below replacement. France is the European best case at 1.64; Italy and Japan under 1.25. Note what the UN's own projections assume next: a slow crawl back toward 2.1 — the most charitable number in this dossier.

Exhibit C — Share of population 65+, 2023UN WPP 2024 age structure

Japan is already past one-in-three. Every country here is on the same curve with a different start date. What Japan has at 2023, Germany will be walking into through the 2030s–2040s.

Country	2024	2050	2100	Δ by 2100	TFR 24 / 50 / 100	Life exp. 2024	65+ 2023
United States	344	380	421	+22%	1.62 / 1.64 / 1.65	79.5	17.4%
Russia	145	136	126	-13%	1.46 / 1.54 / 1.59	73.3	16.6%
Japan	124	105	77	-38%	1.22 / 1.35 / 1.47	84.9	29.6%
Germany	85	78	71	-16%	1.45 / 1.54 / 1.60	81.5	22.8%
United Kingdom	69	75	74	+8%	1.55 / 1.55 / 1.60	81.4	19.2%
France	66	68	68	+3%	1.64 / 1.65 / 1.65	83.5	21.7%
Italy	59	52	35	-40%	1.21 / 1.35 / 1.48	83.9	24.2%
Spain	48	45	33	-31%	1.22 / 1.36 / 1.48	83.8	20.7%
Canada	40	46	54	+35%	1.34 / 1.39 / 1.49	82.7	19.4%

Population in millions (UN WPP 2024 medium variant). TFR column: 2024 actual / 2050 projected / 2100 projected. The projected TFR recovery is an assumption, not a finding — treat the 2100 population numbers as the optimistic end of the range. 65+ is the 2023 share, from the same UN file.

The problem catalogue

Seven concrete problems, in rough order of arrival.

Each one is the wedge, priced. "Now" means already visible; "Near" is 2026–2032; "Mid" is 2033–2045. The tags are honest, the order is not gospel.

Immigration is the only lever The arithmetic works only with net inflows. Germany is beginning to feel it; the US is in the fight; Japan will need roughly four times its current foreign workforce by 2040. So the demographic problem collides head-on with the immigration fight — expect the political energy to land there first, not where the actuaries sit. Now

Pension & fiscal solvency Pay-as-you-go pensions are a straight transfer from the young to the old — and the young is shrinking. IMF (2021) found all seven G7 economies financially insolvent on a 50-year horizon at current policy. The first hard political choice is coming; it will not arrive late, it will arrive forced. Now → Near

Sectoral labour shortages The macro cliff is mid-decade, the micro shortages are now: health and nursing, trucking, construction, manufacturing, AI talent. Japan: 11 million workers short by 2040 (METI). OECD projects labour input turning negative around 2040, shaving ~0.1pp a year off growth. This shows up as wage inflation in the scarce sectors before it shows up in the GDP print. Now

Army manning America's and Russia's armies draw from the same prime-age cohorts as the labour markets. Russia's war is being fought against the demographic pipeline: TFR 1.46 and falling, male life expectancy ~68, combat losses falling on exactly the cohort that is shortening. For the US: a tighter labour market raises the cost of recruiting and retaining. Near

Care & health spending More old people, each needing more years of care, staffed by fewer workers — and the care labour is the least automatable of it all. OECD (2025): fiscal pressure rises ~6pp of GDP by 2060, over 40% of it attributable to ageing, with a third of the region needing adjustments above 9pp. IMF: health spending, not pensions, ends up as the larger long-run driver (Japan about +9pp of GDP by 2060). Mid

Local finance & real estate Shrinking towns: vacant housing, falling property values, the municipal tax base per square mile evaporating. Japan's akiya problem is the advance preview. General-government revenue drops up to 8% per capita under ageing (OECD 2022) because it leans on working-age taxes. Hospitals and fire services close → out-migration → repeat. Subnational governments get hit hardest, earliest, least publicly. Mid

Pay-as-you-go political economy The oldest cohort is the largest, most reliable, longest-held electorate in any of these countries. They block pension reform; the blocking defers the crisis until it is forced rather than chosen. The deferral is itself the problem, because compounding does not pause while the electorate reconsiders. Mid

Timing

Near term and mid term, separated.

Most demographic analysis collapses both. It should not: the first five years of the problem and the next fifteen are different problems that require different answers.

Near · 2026 – 2032

Already visible, or becoming visible

Japan, Italy, Spain, Russia are shrinking now and keep doing it — and Germany is at its peak. This is not a forecast.
Germany's working-age base starts falling around 2027. Pension parametric moves are already underway.
Sector shortages hit first: eldercare, nursing, trucking, construction, tech/AI. Wage inflation in those sectors, sticky "services inflation" in general.
The immigration fight gets hot in the US and Germany, and starts, late, in Japan.
Pensions are stretched but manageable with small, politically survivable parametric adjustments. For now.

Mid · 2033 – 2045

When the wedges stop being abstract

Dependency ratios peak. Japan's 65+ share heads toward 35–40%; Germany walks into the Japanese curve of the 2020s.
Health & LTC spending curves steepen across DE, FR, IT. Care labour becomes the binding constraint on the whole economy, not just the health budget.
Italy's pension bill peaks around 2040 as the 1960s boom generation retires. Bill ~18% of GDP at the peak.
Regional real-estate and local-finance stress becomes a visible category: shrinking cities' budgets, municipal bond risk, hospital closures.
Growth runs below trend in the shrinking club. The first outright pension "shortfall" political crisis occurs — the moment deferral runs out.

The files

Nine files, one pattern.

Each country, one page of facts and one of the bureau's opinions. The opinions are the only opinion on the page that the data does not support; the data supports the data.

File №01US

United States

Pop. 344M → 380M ’25 → 421M ’100 +22% by 2100 TFR 1.62 · life exp. 79.5 · 65+: 17.4%

Immigration is the country. Turn it off and the rest of the problem arrives at the doorstep. 79.5 life years — the shortest of the western club.

File №02GB

United Kingdom

Pop. 69M → 75M ’25 → 74M ’100 +8% by 2100 TFR 1.55 · life exp. 81.4 · 65+: 19.2%

Net migration carries the whole trajectory in the UN's own model. 65+ crosses 30% by the 2050s and the housing stock is already full of old people.

File №03FR

France

Pop. 66M → 68M ’25 → 68M ’100 +3% by 2100 TFR 1.64 · life exp. 83.5 · 65+: 21.7%

Europe's best-in-class fertility (1.64) — and it is still below 2.1. Same problem, arriving roughly a decade later, at roughly twice the fiscal headroom.

File №04DE

Germany

Pop. 85M → 78M ’25 → 71M ’100 -16% by 2100 TFR 1.45 · life exp. 81.5 · 65+: 22.8%

Working-age base starts collapsing ~2027. The reform window closed behind it; pension parametric adjustments and the 2040 bill are now the story.

File №05JP

Japan

Pop. 124M → 105M ’25 → 77M ’100 -38% by 2100 TFR 1.22 · life exp. 84.9 · 65+: 29.6%

The reference case. 11 million workers short by 2040 (METI). 29.5% of the country is 65 or older. The question who cares for the 94-year-old is unsolved.

File №06RU

Russia

Pop. 145M → 136M ’25 → 126M ’100 -13% by 2100 TFR 1.46 · life exp. 73.3 · 65+: 16.6%

A different animal: war, male mortality, 73.3 life years, TFR 1.46 and falling, and the data increasingly classified. The geopolitical compounding nobody is pricing.

File №07IT

Italy

Pop. 59M → 52M ’25 → 35M ’100 -40% by 2100 TFR 1.21 · life exp. 83.9 · 65+: 24.2%

The oldest country on earth. Pension bill peaks around 2040 when the 1960s boom retires. 2100: a 35-million-people Italy, and the regions are shrinking first.

File №08CA

Canada

Pop. 40M → 46M ’25 → 54M ’100 +35% by 2100 TFR 1.34 · life exp. 82.7 · 65+: 19.4%

The migration-led growth story: +35% population by 2100 on the UN's path. Home-grown stress point: the health system is already at capacity.

File №09ES

Spain

Pop. 48M → 45M ’25 → 33M ’100 -31% by 2100 TFR 1.22 · life exp. 83.8 · 65+: 20.7%

It arrived before the rest: Spain is already shrinking — peak 47.9M (2024), 33.2M by 2100 (−31%), below 40M by ~2068. TFR 1.22 and 83.8 years to live: the wedge, about a decade ahead of Italy.

Provenance

Where the numbers come from.

UN World Population Prospects 2024, medium variant — official bulk files — WPP2024_Demographic_Indicators_Medium.csv (total population, TFR, life expectancy, 2024–2100, 1-January basis) and WPP2024_PopulationByAge5GroupSex_Percentage_Medium.csv (age structure) from the UN's WPP download hub (population.un.org/wpp), retrieved 2026-08-24. This is the primary source for every projection on this page — the source the numbers were published in, cited directly.
Cohort note — the files, the charts, the table and the ticker are generated from a country list: the G8 (2014 composition, Russia suspended since 2014 but kept for continuity) plus Spain, added 2026-08-24 at the reader's request. Adding another country is one entry in the list and one row of the UN's file; nothing else moves.
IMF, "Intertemporal Solvency Analysis" (2021) — the G7 all-financially-insolvent-on-a-50-year-horizon finding; intertemporal net financial worth, US ≈ −392% of GDP.
OECD, Ageing and Dependency in OECD Economies (2024–2025 reports) — the ~6pp of GDP fiscal-pressure-by-2060 estimate, ageing's share of it, subnational revenue effects, and long-term-care spending trajectories.
Japan METI workforce projections — the 11-million-worker shortfall by 2040 and the AI/robotics and manufacturing gaps.
Russian demographic data — treated with extra caution: the numbers are increasingly classified since 2025, and war + mass emigration distort every observable series. Directions are robust; magnitudes are not.

Method note: the UN's medium variant assumes fertility converges back toward 2.1 by 2100 — the most optimistic assumption in this document. Every "it gets worse than shown" statement in the press about these projections is therefore true by construction, not by data. Figures rounded to whole millions; percentages to whole points. This page was built by a department that does nothing, professionally, and this is the one thing it did. Commit history is in the repo, for the file that is the thing.