Below-replacement fertility is usually described as a gap from about 2.1. The direct daughter-replacement measure is the net reproduction rate (NRR), but the familiar total fertility rate (TFR) benchmark can support a simpler illustration. A modest gap repeated many times becomes a very different condition from a short-lived decline.
Consider a normalized, nonoverlapping-generation stress test. Start with a daughter generation whose size is set to 100. Hold the period TFR constant, divide it by a fixed replacement benchmark of 2.1, and apply that ratio once per generation:
normalized generation n = (TFR / 2.1)ⁿ
The assumptions are intentionally stark: no migration; a constant TFR; fixed low mortality and the typical male-heavy sex ratio already embedded in the 2.1 benchmark; generations treated as nonoverlapping and lasting 25 to 30 years; and no change in the timing of births. This is a normalized replacement-ratio proxy, not an exact calculation of NRR and not a population projection.
| Constant period TFR | Ratio after one generation | After 10 generations | After 100 generations |
|---|---|---|---|
| 1.8 | 85.7 percent | 21.4 percent | 2.0 × 10⁻⁷ of the starting size |
| 1.5 | 71.4 percent | 3.46 percent | 2.4 × 10⁻¹⁵ of the starting size |
| 1.2 | 57.1 percent | 0.371 percent | 5.0 × 10⁻²⁵ of the starting size |
Ten generations is roughly 250 to 300 years. At 1.8, the tenth descendant generation in the illustration remains about 21 percent as large as the starting generation; at 1.5, it is about 3.5 percent; at 1.2, about 0.37 percent.
One hundred generations is roughly 2,500 to 3,000 years. The three ratios are then lower than the starting value by roughly seven, 15, and 24 orders of magnitude. I don’t expect any fertility rate, mortality regime, political order, or migration pattern to remain fixed for that long. The numbers are not forecasts and soon stop resembling any plausible society. The stress test has one purpose: a permanently low replacement ratio cannot produce a stable, permanently smaller population. It must eventually change, be offset, or keep compounding.
That conclusion does not tell us when fertility will change. It does not tell us the right population for a country, and it cannot measure the worth of any family. It separates two ideas that are often blurred together. A population can shrink once and stabilize. A population cannot stabilize while every generation remains smaller than the one before it.
From stress test to projection
The stress test asks what repeated multiplication contains under fixed assumptions. A demographic projection asks a different question. It begins with the population by age and sex, ages each cohort forward, and applies assumptions about age-specific fertility, mortality, and migration. The United Nations uses probabilistic methods to express uncertainty around future fertility and mortality, then publishes scenarios and intervals rather than one guaranteed result.
The starting population gives near-term projections unusual structure. Most people who will be age fifty in 2050 are already alive. The size of the cohort that could become parents in the early 2030s is largely known. Farther out, each assumption passes through children who have not yet been born, and uncertainty widens.
A population remade by successive cohorts Japan’s population by age group, with later values projected. The shape reflects decades of births, deaths, and migration rather than one year’s fertility rate. via Wikimedia Commons
Projections are conditional statements, not promises or prophecies. Migration can shift national totals quickly. Longevity can improve faster or slower than expected. Period fertility can rebound as births move to later ages, or a recovery can persist and enlarge later cohorts. None of those moving parts appears in the normalized stress test above.
Uncertainty does not make every path equally plausible. Age structure constrains the next few decades, and arithmetic constrains any scenario that holds fertility below replacement indefinitely. The lower the rate and the longer it lasts, the more a later recovery must work through the missing cohorts left by earlier years.
This chapter has described the turn: mortality fell, population surged, fertility followed downward, and momentum delayed the visible result. The demographic mechanism is clear enough. The causes are not. Why did so many societies move below replacement—and why did family size keep falling after the population explosion was already over?
Citations
- United Nations, Department of Economic and Social Affairs, Population Division, World Population Prospects 2024: Methodology Report, 2024.
- United Nations, Department of Economic and Social Affairs, Population Division, World Population Prospects 2024: Summary of Results, 2024.