The Decimal That Decides the Century
For most of living memory the worry about human numbers was that they would never stop climbing. That worry has quietly inverted. Demographers, the people who count and forecast populations, now agree that the world total will peak this century and then fall, and the fight among them is no longer whether but when, and how far.[1] The gap between the leading forecasts by the year 2100 is well over a billion people, and almost all of it traces to a single unresolved question about whether very low birth rates ever climb back.[8] The answer decides how the century’s pensions, labour forces and even its climate scenarios are drawn.
The field in brief
A population forecast is not a guess at a total. It is bookkeeping, carried forward year by year. The standard tool, the cohort-component method, slices a population into groups by age and sex, then each year ages everyone forward, removes a share according to death rates, adds newborns according to birth rates, and adds or subtracts migrants.[7] Every forecasting shop uses this same accounting skeleton, so their famous disagreements come entirely from the assumptions they feed it.
The assumption that matters most is about births. The total fertility rate is the average number of children a woman would have over her life at current age-specific rates, a snapshot figure rather than a prophecy.[7] Roughly 2.1 is replacement, the rate at which a population just replaces itself over the long run, and more than half the world’s countries are now below it.[1] When a country falls far under replacement, nobody knows what happens next, because no society has yet spent several generations there. Either the rate hits a floor and drifts back up as families adapt, or it stays low. That one behavioural question, applied to dozens of countries and compounded across eighty years, is the whole ballgame; a difference of two-tenths of a child, sustained, moves the world total by more than a billion.[9]
The fight
Three institutions carry the argument. The United Nations Population Division in New York publishes the official world numbers, guided by figures such as Patrick Gerland and, until recently, John Wilmoth.[1] The Institute for Health Metrics and Evaluation in Seattle, led on this work by Stein Emil Vollset within Christopher Murray’s global-health enterprise, published a rival forecast in 2020.[3] The Wittgenstein Centre in Vienna, built around Wolfgang Lutz and Samir KC, offers a third.[5] Their headline numbers sit surprisingly far apart.
| Forecast | Peak population | Peak year | 2100 population |
|---|---|---|---|
| UN, 2024 | ~10.3 bn | mid-2080s | ~10.2 bn |
| IHME, 2020 | ~9.7 bn | ~2064 | ~8.8 bn |
| Wittgenstein, 2024 | ~10.1 bn | ~2080 | ~9.9 bn |
The whole spread comes from one modelling choice. The UN runs a statistical model that treats each country’s fertility as passing through a transition and then converging back toward a long-run rate near 1.75, so recovery is built into the machine.[8] The Seattle team rejected that. It ties future births to two measured things, how far girls go in school and how many women who want to avoid pregnancy can, and when those trends are run forward the rate keeps falling and does not bounce, which is why its curve peaks earlier and drops harder.[8] The Vienna group sits between them methodologically, also building fertility up from education and human capital but setting its assumptions through structured expert argument rather than a purely statistical fit.[5] Its own 2024 round has drifted upward, close to the UN’s figures.[4]
The disagreement is loudest over sub-Saharan Africa, the region that will supply most of the century’s population growth. There the three forecasts for 2100 fan out from 3.8 billion in the UN’s to 3.1 billion in Seattle’s to 2.6 billion in Vienna’s, because the education-and-contraception link that the low forecasts lean on is empirically weaker in the region than elsewhere.[7] The 2020 Seattle forecast drew a pointed published critique from Stuart Gietel-Basten and Tomáš Sobotka, who argued it dressed a deeply uncertain future in too much confidence.[6] What keeps the fight live is that reality keeps breaking toward the pessimists: the UN itself has revised its long-run numbers downward and now puts the odds of a peak this century at about four in five,[2] and fertility in several countries has fallen below one child per woman faster than any model expected, with no sign yet of the rebound the official projection assumes.[8]
What the fight reveals
A quarrel over a decimal turns out to sit under some of the century’s largest plans. Pay-as-you-go pensions and health systems depend on the ratio of workers to retirees, so whether the rich world stabilises or keeps shrinking is the difference between manageable ageing and a slow fiscal emergency.[3] If domestic births stay low, shrinking-workforce countries face a stark choice between decline and large-scale immigration, and the world’s future workers increasingly come from a few high-fertility regions.[3] The reach runs even into climate science, because the standard scenarios that feed long-run warming models take their population and schooling paths straight from the Vienna projections, so a disagreement about the fertility floor propagates into projected emissions.[5] On the central question the evidence does now lean one way, toward lower and later-rebounding fertility than the field long assumed, and the UN’s own downward revisions concede as much.[8] But the crux stays genuinely open, because the one fact that would settle it, how a society behaves after generations below replacement, does not yet exist anywhere on Earth.[9]
- United Nations — ‘UN projects world population to peak within this century’ (World Population Prospects 2024): 8.2 bn in 2024, peak ~10.3 bn in the mid-2080s, more than half of countries below replacement. Primary.
- UN, World Population Prospects 2024, Summary of Results — the stated 80% probability that the peak occurs this century and the ~10.2 bn figure for 2100. Primary.
- Vollset et al., ‘Fertility, mortality, migration, and population scenarios for 195 countries and territories from 2017 to 2100,’ The Lancet (2020) — the IHME forecast: peak ~9.7 bn near 2064, ~8.8 bn by 2100. Primary.
- IIASA / Wittgenstein Centre — ‘Population and human capital projections to 2100’ (Feb 2024), the latest Vienna numbers (SSP2 peak ~10.1 bn ~2080, ~9.9 bn in 2100). Primary.
- IIASA — ‘9 billion or 11 billion? The research behind new population projections’ (2014): the Vienna education-and-expert-argument method and its historically lower, earlier peak. Field source.
- Gietel-Basten & Sobotka, ‘Uncertain population futures: Critical reflections on the IHME scenarios’ (2020) — the named published critique of the 2020 forecast’s certainty. Commentary.
- Population Reference Bureau — ‘Understanding and Comparing Population Projections in Sub-Saharan Africa’: side-by-side 2100 totals (UN 3.8 bn, IHME 3.1 bn, IIASA 2.6 bn) and why the education/contraception link is weaker there. Comparison source.
- Oxford Institute of Population Ageing — ‘A new approach to fertility projections’: the UN’s assumed convergence toward ~1.75 and the absence of clear rebound signals. Commentary on the crux.
- Spears, Vyas, Weston & Geruso, ‘Long-term population projections: scenarios of low or rebounding fertility,’ PLOS One (2024) — shows the rebound-timing assumption swings the eventual total by billions. Primary.