Below-replacement fertility is described as a gap from 2.1. The more revealing comparison is between generations. A modest gap repeated many times can reshape a population; a large gap can do so with surprising speed.
Consider a deliberately simple illustration. Assume no migration, unchanged mortality, and a constant fertility rate. Divide each rate by a replacement benchmark of 2.1 to approximate the size of one generation relative to the one before it. This is not a population forecast. It isolates the compounding effect.
At a TFR of 1.8, the next generation is about 86 percent as large. After five generations, it is 46 percent; after ten, 21 percent. At 1.5, the corresponding figures are 71 percent for one generation, 19 percent for five, and 3 percent for ten. At 1.2, one generation is 57 percent as large, five generations 6 percent, and ten generations less than half of 1 percent. At 0.9, one generation is about 43 percent as large and five generations about 1.4 percent.
Now extend the same deliberately unrealistic line for a thousand years, roughly forty generations. Even the mildest rate in the example, 1.8, leaves a descendant generation about two-tenths of 1 percent the original size. I don’t expect any fertility rate to remain fixed for a millennium. That is exactly the point: decline must reverse somewhere along the way. A civilization that hopes to cross the stars has to think in centuries, and a species that hopes to remain among them has to think longer still.
Real populations do not obey this stripped-down arithmetic. People migrate. Mortality changes. Fertility moves. Generations overlap, and the starting age structure can delay or accelerate the result. The calculation clarifies why the difference between 1.8 and 1.2 reaches far beyond six-tenths of a child. Repeated over time, it describes very different scales of contraction.
What projections can say
Demographic projections begin with an advantage over many other long-range forecasts: most people who will be middle-aged several decades from now are already alive. This makes near-term age structure relatively visible. The UN then models future fertility and mortality, using probabilistic methods to express uncertainty around the result.
Uncertainty widens with time. A small change in assumed fertility, repeated across future cohorts, creates a large difference late in the century. Migration is volatile and often political. Longevity may improve faster or slower than expected. A projection’s central path is therefore neither a promise nor a prophecy.
Long-run scenarios work best as stress tests. They show that a temporary dip and a stable low-fertility regime are different conditions. They reveal how much later recovery must overcome negative momentum. And they make clear that stabilization at a smaller population is not the same as continuous decline: any rate can change.
The central demographic fact is more modest. Below replacement, each generation tends to be smaller than the last unless migration or changing rates offset the difference. The further fertility falls and the longer it remains there, the more that tendency compounds. That arithmetic underlies the social and economic questions in the chapters ahead.
Citations
- United Nations, Department of Economic and Social Affairs, Population Division, World Population Prospects 2024: Methodology Report.
- United Nations, Department of Economic and Social Affairs, Population Division, World Population Prospects 2024: Summary of Results.