South Korea's population is projected to more than halve by 2100, dropping from around 52 million people in 2026 to just 22 million, according to United Nations projections. A new analysis by Our World in Data quantifies exactly how much fertility, life expectancy, or immigration would have to change to stop the decline — and the numbers are stark.

The analysis, by Hannah Ritchie, Sophia Mersmann and Daniel Bachler, uses a population simulation tool to test alternative scenarios against the UN's baseline assumptions.

Under the UN's own medium projections:

• The fertility rate rebounds from its current 0.7 births per woman to 1 by mid-century and 1.3 by 2100.
• Life expectancy rises to 91 years by the end of the century.
• Net migration falls from 1.7% annually today to almost zero by 2050 and stays there.

The result is not only a much smaller country but a much older one. By 2100, there would be almost as many people of retirement age as working-age people.

What would immigration need to reach?

For migration-focused readers, this is the central number. South Korea currently records net migration of about 1.3 people per 1,000 residents each year.

To keep the population roughly constant, the model shows the country would need a net migration rate of about 0.9% — 9 immigrants per 1,000 residents annually. That is roughly seven times higher than today's rate.

How plausible is that? The analysis puts the target in context:

• Net migration in the United States and the United Kingdom has fluctuated between 3 and 6 per 1,000 people over the last 30 years.
• Canada has maintained rates between 6 and 10 per 1,000 for most of the last 20 years.
• Germany has reached similar levels over the past decade.

Rates above 0.8% are uncommon but not unheard of. What would be genuinely rare, the authors note, is sustaining such high immigration for more than 70 years — and with most other countries also shrinking, attracting those numbers would grow harder over time.

How much would fertility have to rise?

South Korea has seen one of the most dramatic fertility declines on record. In 1960, the average woman had six children. Today the figure stands at 0.75.

The model shows one path to a stable population: the fertility rate would need to reach 2.1 children per woman by 2050 and stay there. A smaller increase would slow the decline but not stop it.

The authors describe this as requiring a substantial "baby boom" in the next few decades. They call the fertility scenario the most realistic of the three — the required rate, 2.3 children per woman within a few decades, sits well below the country's historical average — but note that such a strong reversal after the global fertility collapse of the past half-century would be unprecedented.

Could longer lives do the job?

Life expectancy offers the least plausible route. To stabilise the population for several more decades, average life expectancy would need to hit 130 years by 2050 — an increase of about 50 years.

The pace required is extraordinary. The countries with the highest life expectancy have taken four years to add one year of average lifespan. At that rate, going from 83 to 130 years would take 188 years. South Korea would need to do it in 25.

Even if it worked, the outcome would be troubling. By the end of the century the median age would be 82, and with retirement fixed at 65 there would be more than two pensioners for every working person.

What does this mean?

The authors' conclusion is blunt: gains in life expectancy or migration large enough to stabilise the population seem implausible, and they consider it "very unlikely" that South Korea can maintain its current population level through this century.

The simulation tool is available on the Our World in Data site and lets users adjust fertility, life expectancy and migration assumptions for South Korea or any other country. The UN's underlying projections and demographic assumptions are also publicly explorable through the organisation's Population and Demography data explorer.