First, the bad news: Global fertility is falling fast. The aging populations of rich countries are relying on ever fewer workers to support their economy, dooming those younger generations to a future of higher taxes, higher debt, or later retirement—or all three. Birth rates in middle-income countries are also plummeting, putting their economic development at risk. Practically the only countries set to continue growing are desperately poor.
By about 2084, according to the gold-standard United Nations “World Population Prospects,” the global population will officially begin its decline. Rich countries will all have become like Japan, stagnant and aging. And the rest of the world will have become old before it ever got the chance to become rich.
Sorry, did I say “bad news”? That was actually the good news, based on estimates that turn out to be far too rosy. Every two years, the UN’s demographers revise their population projections, and for the past 10 years, they’ve always had to revise in the same direction: down. Next year, they’ll do so again. In reality, the worldwide population decline is set to begin decades ahead of their expectations. Because global fertility trends are much worse than they, and probably you, think.
Jesús Fernández-Villaverde, a University of Pennsylvania macroeconomist, studies how poor countries develop. This development usually happens alongside a fertility transition. As people move from rural areas to cities, their economic opportunities expand, and kids become less crucial as a source of agricultural labor. Women gain access to contraception and education. They go from having about six kids, on average, to two. Fernández-Villaverde calls this “the standard modernization story,” and he’s been teaching it for decades.
Much of Fernández-Villaverde’s research focuses on Latin America, an economically middling region where one would expect middling fertility rates. In recent years, however, births in some Central and South American countries have plummeted to rates far lower than most rich countries’, in defiance of the standard modernization story. Each year, Fernández-Villaverde updates his data on Latin American birth rates, which he gathers from the countries’ official birth statistics, in preparation for a class he teaches about the region’s economic history. He first began noticing in 2019 that the UN was too optimistic, but only in the past few years did the discrepancies become downright alarming.
For 2024, the UN had projected 701,000 births in Colombia; it had put the chance of the number of births being lower than 553,000 at only 2.5 percent. In the end, Colombia saw only 445,000 births in 2024. That translates to a fertility rate of 1.06 births per woman, down more than half from 2008. Chile’s is even lower: At current rates, 100 reproductive-age Chileans can expect to have 52 children and only 27 grandchildren. (Demographers generally consider a birth rate of about 2.1 to be “replacement level,” or the point at which a society doesn’t shrink from one generation to the next.)
The discrepancies were not limited to South America. In 2024, Poland’s births were also below the 2.5-percent probability cutoff, as were Estonia’s and Cuba’s and Azerbaijan’s and Sri Lanka’s and Egypt’s. These supposed outlier results aren’t outliers at all—the world is just not having as many babies as the UN had thought it would.
Digging into the UN’s model, Fernández-Villaverde found something even stranger. For nearly every low-fertility country, the UN projects either one of two outcomes: The fertility rate will flatten, or it will rise to a number somewhere between one and two births per woman—still below replacement level, but not quite as catastrophic. The United States is in the first category. Our fertility rate has fallen steadily since the Great Recession, from 2.1 to 1.6. One might therefore expect the decline to continue. But the UN projects that the U.S. birth rate will stay flat, not just this year but also in 2026 and 2030 and 2060 and 2090, never rising above 1.7 or dipping below 1.6.
In the other category are countries such as Thailand, whose fertility rate has been falling for 72 years and has never stopped for longer than a single year. Nonetheless, there the UN projects a demographic miracle: Starting in two years, the country’s birth rate will begin to climb, first slowly and then a little more quickly, finishing out the century with a birth rate of 1.45, up from its projected 2024 low of 1.20.
Every part of that appears to be wrong. In reality, Thailand’s reported birth rate last year was 0.98, and preliminary 2025 data show the decline continuing. In a country the size of Thailand, the difference between the UN’s projection and the real fertility rate throughout the 21st century will amount to millions of people who will never be born.
All in all, as Fernández-Villaverde recently explained at a research symposium in London, humanity won’t start to shrink in 2084. It will start to shrink in 2055, if not sooner.
“There are two types of people,” Alice Evans, a British professor who studies falling fertility around the world, posted on X after reading Fernández-Villaverde’s presentation: those “not bothered about demographics” and “those who’ve read Jesus’s slides.”
The UN has a simple explanation for its optimistic projections: Fertility has rebounded in the past, so it will rebound again.
In Belarus, for example, the fertility rate in 1988 was at replacement level; it fell to an abysmal 1.22 only nine years later. But then it rebounded, all the way up to 1.73 by 2015. Australia’s birth rate fell to 1.7 in 2001, only to bounce back to 2.0 in 2008. France’s rate followed a similar trajectory during the same period, as did Italy’s and Sweden’s. “To the extent you think the ‘World Population Prospects’ are wrong, that is the extent to which you are saying, ‘This time is different,’” Lyman Stone, a Ph.D. student and birth-rate consultant, told me.
The thing is, this time really does look different. Birth rates in Australia and France and Italy and Sweden have now fallen to all-time lows (excluding during World War I, in France’s case). Belarus, a onetime redemption story, recorded a fertility rate of just 1.1 last year, lower than the lowest lows the country experienced in the 1990s. Deaths outnumbered births by nearly two to one. If a rebound is coming, there are no signs of it yet. Fernández-Villaverde estimates that the world is already below replacement fertility: The population is not just projected but guaranteed to shrink if things don’t change. That was not the case in the 1990s.
The UN’s model hasn’t adjusted to the new normal. If a country has ever experienced a fertility increase (as Australia and France and Belarus have), then its birth rate is assumed to be stable. If a country has never experienced an increase, then the model assumes that it will at some point, once fertility gets low enough. In other words, the model assumes as its end state a stable and modest number of births. This is perhaps a reflection of humanistic optimism. “There is, at some point, a minimum social capacity to adapt and eventually at least address some of the concerns or challenges that exist in that country,” Patrick Gerland, the chief author of “World Population Prospects,” told me. “The people living in those countries don’t necessarily want their country to totally disappear.”
To his point, the model comes with a hard-coded minimum: No country can ever be projected to have a fertility rate less than 0.5 children per woman. Like the rest of the model, this, too, might need to be revised. Macau (which the UN analyzes separately from mainland China) had a fertility rate of 1.2 a decade ago. Last year, it fell to 0.58, and it looks set to fall even further: In the first four months of 2025, births were down another 13 percent.
If you’re not sure why this is all so alarming, consider Japan, the canonical example of the threat that low fertility poses to a country’s economic prospects. At its peak in 1994, the Japanese economy made up 18 percent of world GDP, but eventually, the country’s demographics caught up with it. Now Japan’s median age is 50 years old, and the country’s GDP makes up just 4 percent of the global economy. Measured per hours worked, Japan’s economic growth has always been strong, but at some point, you just don’t have enough workers.
The fertility rates that doomed the Japanese economy ranged from 1.3 to 1.5. So imagine what’s in store for modern-day Colombia (1.06) and Chile (1.03). How will they grow with so few workers? How will they ever become rich if each worker is expected to provide for so many elderly people? The overly optimistic UN estimates have obscured just how urgent these questions really are. Because if the birth rate continues to drop around the world at its current pace, economic growth and workers’ retirement prospects will go the way of those projections: adjusting every few years to a smaller, sadder, poorer future.
The post The Birth-Rate Crisis Isn’t as Bad as You’ve Heard—It’s Worse appeared first on The Atlantic.
