In February 2017, a team of researchers from Imperial College London and the World Health Organization published one of the most rigorous life expectancy forecasts ever attempted.

Rather than rely on a single statistical model, the team built an ensemble of 21 different probabilistic forecasting models, weighted by how accurately each predicted withheld historical data. They applied this approach to 35 industrialised countries, projecting life expectancy at birth out to 2030. The work was published in The Lancet.

The headlines focused on the winners. South Korean women were projected to become the first population in human history with an average life expectancy exceeding 90 years. Japan, France, and Spain followed closely.

The losers got far less coverage. Buried in the same paper was a quieter, more uncomfortable finding about a country that had, until that point, considered itself among the world’s healthcare leaders.

The United States was projected to fall to near the bottom of the developed world. By 2030, the study found, the average American woman would be living about as long as the average woman in Mexico or Croatia.

What the study actually said

The 2017 Lancet paper is unusually detailed in how it ranks countries, and the US results are striking.

The study found that the United States had “some of the lowest projected life expectancy gains” among all 35 industrialised countries studied. Among nations expected to make minimal progress, the US sat alongside Sweden, Japan, Greece, Macedonia, and Serbia — but unlike those countries, it was starting from a substantially lower base.

The specific 2030 projections for the US were 83.3 years for women and 79.5 years for men. Those numbers placed America roughly on par with the Czech Republic for men, and Croatia and Mexico for women. By contrast, South Korean women were projected at over 90 years; French, Spanish, and Japanese women all above 88.

The paper was unusually direct about the causes, particularly given the typical neutrality of medical journals. It identified the United States as “the only country in the OECD without universal health coverage” and noted that the US has “the largest share of unmet health-care needs due to financial costs.” It cited high and rising health inequalities, persistent obesity and chronic disease, high homicide rates, and stagnant or declining life expectancy among some American subgroups.

The picture was of a wealthy country whose health outcomes were quietly collapsing relative to its peers, with the trend lines extrapolating forward to a future in which America’s life expectancy would no longer be in the same league as the rest of the developed world.

What’s happened since

The 2017 projection has now had eight years to interact with reality. The picture is more complicated than the original paper imagined.

Between 2017 and 2021, US life expectancy fell sharply — much faster than the Lancet projection had predicted. The COVID-19 pandemic killed more than 1.2 million Americans and pushed US life expectancy down to roughly 76.4 years in 2021, the lowest level in decades. That decline far exceeded what the 2017 study had anticipated for that period. The pandemic was the primary driver, but the drug overdose crisis, rising chronic disease, and uneven healthcare access all contributed.

Then, starting in 2022, US life expectancy began to recover. According to the CDC’s most recent figures, US life expectancy reached 79.0 years in 2024 — the highest mark ever recorded in American history. The recovery was driven by the dissipation of COVID-19 (deaths from which dropped 37% in 2024), a sharp 26% reduction in drug overdose deaths, and declining death rates from heart disease, cancer, and unintentional injuries.

So the 2017 projection that the US would fall behind by 2030 has been partly borne out — but in a more chaotic way than the original paper envisioned. The US has not been on a gradual decline. It has fallen further and faster than expected during the pandemic, then recovered some of the ground since.

What the comparison still shows

Even with the recent improvement, the US remains well below its peers.

The Peterson-KFF Health System Tracker, which monitors how US healthcare outcomes compare internationally, found that the average life expectancy across comparable developed countries in 2024 was 82.7 years — roughly 3.7 years longer than the US figure of 79.0. Japan, South Korea, Switzerland, and Australia were all sitting between 84 and 85 years. The US ranked 63rd in the world.

The basic shape of the 2017 prediction has therefore held. The US has not closed the gap with its peers. If anything, despite the recent recovery, the gap remains wide. And the underlying drivers the Lancet team identified — obesity, chronic disease, healthcare inequality, the absence of universal coverage — have not fundamentally changed.

Whether the US will, by 2030, actually be on par with Mexico and Croatia as the 2017 paper suggested depends largely on what happens in the next few years. The recovery from the pandemic has bought some time. But the structural problems that drove the original projection are still in place, and most demographers do not expect the US to close its gap with the rest of the developed world without significant policy changes.

Why the projection matters

The 2017 Lancet paper was not a piece of advocacy. It was a careful statistical exercise by some of the world’s most respected demographers, using one of the most sophisticated forecasting methodologies ever applied to mortality data.

What made it sting was not the methodology. It was what the methodology found. A country that spends roughly twice as much per capita on healthcare as its peers was, by 2030, projected to fall behind countries it had historically considered its inferiors. Croatia. Mexico. The Czech Republic.

The pandemic complicated the timeline. The recent recovery has slowed it. But the trajectory the 2017 paper identified is still essentially in place — and the question of whether America can change it remains, eight years later, genuinely open.

The 21 models are still running. The numbers are still moving. And what we will actually see in 2030 is now only a few years away from being history.