What Is Replacement-Level Fertility? Why the Rate Is 2.1

Replacement-level fertility is the average number of births per woman needed for one generation to replace itself in the long run, assuming no migration. In many low-mortality populations it is approximately 2.1 births per woman.

The decimal matters, but the benchmark is often used too rigidly. It is not a universal constant, it does not describe personal choices, and crossing it does not switch population growth on or off immediately.

Why the benchmark is above two

Two children per woman might appear sufficient to replace two parents. At population scale, however, not every child survives through the reproductive years, and the number of girls and boys born is not exactly equal. A small additional amount is therefore needed.

Where mortality before and during reproductive ages is higher, replacement fertility is higher than 2.1. As survival improves, the replacement level falls closer to two.

The arithmetic is simple. About 105 boys are born for every 100 girls, so roughly 2.05 births are needed to produce one daughter. If some of those daughters die before or during their childbearing years, the number of births must rise further to compensate: with 98% of girls surviving to the mean age of childbearing, as in most high-income countries, the result is about 2.1. Where only 85% survive, as in parts of sub-Saharan Africa a few decades ago, replacement requires about 2.4 births per woman.

Net reproduction rate: replacement measured directly

Demographers avoid the rule of thumb by using the net reproduction rate (NRR): the average number of daughters a woman would have who themselves survive to her age at childbearing, given current fertility and mortality. An NRR of exactly 1 means each generation of mothers is replaced by one of the same size, whatever the survival conditions. An NRR of 0.7 means each generation is about 30% smaller than the one before.

For the world as a whole, the UN projects an NRR of about 1.03 in 2026, just above replacement. The figures for the largest countries show how far apart national paths have moved since 1950.

A net reproduction rate of 1 is exact replacement; below 1, each generation of women is smaller than the one before. Source: UN World Population Prospects 2024; years after 2023 are medium-scenario projections.

India and Indonesia now have net reproduction rates a little below 1, yet both are projected to keep growing for decades; China and Brazil are far below it. The difference in timing is the subject of the next sections.

How many countries are below replacement

In 1950 only a handful of countries and areas had fertility below 2.1. Today most do, and in the UN medium projection almost every country is below that level by 2100.

Total fertility rate in births per woman. A country counts as below replacement when its rate is under 2.1 in that year. Source: UN World Population Prospects 2024; years after 2023 are medium-scenario projections.

Because the largest populations, China, India and the United States, are all below replacement, the share of humanity living in below-replacement countries is larger than the count of countries alone suggests. Countries still well above 2.1 are concentrated in sub-Saharan Africa, together with Afghanistan, Pakistan and a few other countries in Asia and Oceania; see the full fertility rate by country ranking.

Total fertility rate is a period measure

The total fertility rate (TFR) summarizes the age-specific fertility rates observed or projected in one period. It answers a hypothetical question: how many births would a woman have if she experienced those rates through her reproductive life?

It is not necessarily the completed family size of a real cohort. Birth timing can shift the period rate downward when people postpone births and upward when delayed births occur later.

Why below replacement does not mean immediate decline

Three forces can keep total population growing:

1. Population momentum. A large generation entering reproductive ages can produce many births even at a lower rate per woman.
2. Longer survival. Fewer deaths and longer lives increase the number of people alive.
3. Net migration. More arrivals than departures can offset natural decrease.

That is why a population pyramid and growth series add essential context to the fertility rate.

Vietnam is a clear example. Its total fertility rate was 2.03 in 2000 and is projected at 1.86 in 2026, yet its population is projected to keep growing from about 102 million in 2026 to a peak of about 110 million around 2050. The large cohorts born in the 1980s and 1990s are now in their main childbearing years.

Vietnam, 2026

Why returning to 2.1 may not stop decline immediately

The reverse also applies. After fertility has remained low for decades, fewer people enter reproductive ages. Births may remain below deaths even if TFR returns to replacement level because the potential-parent cohort is small and the population has many older adults.

Italy shows the structure behind this. Its total fertility rate has been below 1.5 since about 1990, and the projected 2026 pyramid has far fewer children than adults in their fifties. Even with the UN medium scenario’s modest rise in fertility to 1.35 by 2050, the population is projected to fall from about 58.9 million in 2026 to 51.9 million.

Italy, 2026

Stabilization is a long-run property under a set of assumptions, not a one-year outcome.

Reading the fertility ranking

Our lowest fertility ranking and the full fertility rate by country table use the projected 2026 TFR from UN WPP 2024. Use it to compare the period rate consistently, then open each profile to check:

• whether the base has narrowed over several decades;
• the size of cohorts in reproductive ages;
• projected total change to 2050;
• median age and the older-population share;
• migration evidence from an appropriate source.

What the number does not tell you

TFR does not explain why people have the number of children they do. Housing, employment, care, health, partnership, education, access to contraception, social expectations, and policy can all matter. National averages also conceal major variation among groups and regions.

Avoid calling any single fertility level “good” or “bad.” The demographic consequences depend on context, speed of change, age structure, and how institutions adapt.

Sources

• United Nations World Population Prospects 2024
• UN WPP 2024 Summary of Results
• World Bank metadata: total fertility rate

Frequently asked questions