By 2030, according to a 2017 study in The Lancet, the women of South Korea are on track to become the first national population on record whose average life expectancy at birth passes 90 years.
That is the finding that travelled around the world when the paper appeared, and it repays care before the number hardens into folklore. What the study reported was a probability. In its own terms, there was a better than even chance, put at 57 per cent, that female life expectancy in South Korea would cross the 90-year mark by 2030. The same models gave a 90 per cent probability that it would clear 86.7 years, the highest national figure recorded anywhere as of 2012.
The paper, “Future life expectancy in 35 industrialised countries: projections with a Bayesian model ensemble,” was led by Vasilis Kontis and Majid Ezzati at Imperial College London, working with the World Health Organization. It is one modelling study, built on one method, and the number at its heart is a forecast rather than a measurement. The finding is worth taking seriously, but it should not be read as the final word.
What the study actually did
Forecasting mortality is harder than it sounds, because any single statistical model carries its own assumptions about how death rates will behave decades out. The Kontis and Ezzati team tried to reduce that fragility by not relying on one model at all. They combined 21 separate forecasting models into what they called a Bayesian model ensemble, weighting each according to how well it had tracked historical mortality data. The approach is borrowed in spirit from weather forecasting, where several models run together and their disagreements become part of the output rather than something hidden.
Their result placed South Korean women at roughly 90.8 years by 2030, ahead of France, Spain and Japan, which clustered nearer 88. The gap between South Korea and the rest was not a rounding artefact. Something specific about the country, in these models, kept pulling its projection higher than any other’s.
The ceiling that was supposed to hold
The reason the figure landed as more than a routine demographic update has to do with an older belief it contradicted. For much of the twentieth century, a number of demographers treated an average national life expectancy of 90 as close to a natural limit, a point where biology would cap further gains.
Ezzati addressed this directly in remarks reported by AFP at the time. As recently as the turn of the century, he said, many researchers believed life expectancy would never pass 90. In comments to Reuters he went further, saying he did not think humanity was anywhere near an upper limit, if one existed at all. The Lancet paper described the 90-year barrier as something several forecasters had deemed unattainable only a decade or two earlier.
Whether they were wrong remains an open question, since the year 2030 has not yet arrived.
What the paper attributed the lead to
The study and its accompanying materials were unusually specific about why South Korea outperformed.
Near-universal healthcare through the National Health Insurance Service, which has covered effectively the whole population since the late 1980s, came first. The team also pointed to strong management of cardiovascular risk, low rates of smoking among women, low blood pressure across the population, and good childhood nutrition. Ezzati told the BBC that South Korea had got a lot of things right, describing a comparatively equal society in which improvements to education and nutrition had reached most people, and pointing to the country’s handling of hypertension and its low obesity rates.
That last point does more work than it appears to. What mattered was the breadth of access across the population, alongside the quality of care at the top. A country can raise its top-end medicine a great deal without moving its average much. South Korea’s story, in this paper, is about the floor rising.
Where the numbers stand now
Eight years on, the trajectory is broadly intact, though uneven. South Korean female life expectancy at birth sat at roughly 86 to 87 years in 2023, depending on the source, with Statistics Korea reporting the higher end of that band. National life expectancy overall dipped in 2022, in line with the pandemic’s effect on mortality almost everywhere, then resumed climbing.
Reaching 90 by 2030 from that starting point would require the gains to continue at a fair pace over a short window. The projection never assumed a straight line, and a single disrupted year does not sink a forecast built to 2030. But the headline figure was always a central estimate sitting inside a range, and it reads best that way.
What the projection does not settle
There is a tension in the South Korean case that the longevity headline tends to bury. The same country projected to lead the world in how long its women live also has, by a wide margin, the lowest fertility rate of any country, well under one child per woman. A population can be very good at keeping people alive into great age while producing very few new people. South Korea is doing both at once, and the second fact will shape the country far more visibly over the coming decades than the first.
This is the part the Kontis and Ezzati paper handled with care and much of the coverage did not. The authors framed their conclusion as a caution rather than a milestone: health systems, pensions and social care would need serious planning for populations older than any before them. A long-lived population is a cost structure as much as an accomplishment.
Whether South Korean women cross 90 on average in 2030 will not be knowable until the mortality data for that year is in. Until then the sensible thing is to hold the figure the way the paper offered it: as a probability, attached to a method, sitting next to the demographic question that will end up mattering more.
