The global fertility crisis is already here

For anyone tempted to try to predict humanity’s future, Paul Ehrlich’s 1968 book The Population Bomb is a cautionary tale. Feeding on the then popular Malthusian belief that the world was doomed by high lbirth rates, Ehrlich predicted: “In the 1970s hundreds of millions of people will starve to death.” He came up with drastic solutions, including adding chemicals to drinking water to sterilize the population.

Ehrlich, like many others, got it wrong. What he needed to worry about was declining birth rates and population collapse. Nearly sixty years on, many predict the world will soon reproduce at less than the replacement rate.

But by my calculations, we’re already there. Largely unnoticed, last year was a landmark one in history. For the first time, humans aren’t producing enough babies to sustain the population. If you’re fifty-five or younger, you’re likely to witness something humans haven’t seen for 60,000 years, not during wars or pandemics: a sustained decrease in the world population.

A society’s reproduction level is measured by the fertility rate — the average number of children a woman has. The replacement level is accepted as 2.1: any higher and the population grows; any lower and it falls. Like the R number in epidemiology (which we heard so much about during the pandemic), the replacement level is a critical figure. Either side of it leads to dramatically different outcomes. The replacement level is put at a little over two to take account of the slight imbalance in male and female births — slightly more of the former are born. Also, not all girls survive until reproductive age.

According to the UN World Population Prospects, the global total fertility rate last year was 2.25 — a little above the replacement rate. But the UN was wrong. It’s not easy to calculate the figure because there’s a lack of statistics in many countries. In others, political constraints bind the organization. For many places with reliable records, last year’s birth numbers were between 10 percent and 20 percent lower than UN estimates. In Colombia, the UN estimate was 705,000 births. Yet its national statistical agency counted 510,000.

For the first time, humans aren’t producing enough babies to sustain the population

There’s another reason to be skeptical of the UN figures — the replacement fertility level of 2.1 is valid for the UK, not universally. We get the 2.1 figure using a calculation: 1.06 boys are born for every girl in Britain. To ensure an average of one girl born, we need 2.06 children overall to be born. We then look at the probability a woman lives to reach her reproductive years, which in Britain is 0.98. To get the reproductive rate, we divide 2.06 by 0.98 — which equals 2.1.

However, in many developing countries fewer women survive to a reproductive age. Globally, the figure drops to 0.94. So the replacement fertility level needed worldwide is more than 2.1.

Many countries also have a higher male-to-female ratio, often due to selective abortion. In China, it’s around 1.15; in India, 1.1. An estimate for the sex ratio globally is 2.08. To estimate a global replacement fertility rate we divide 2.08 by 0.94, which comes out at 2.21 children per woman.

By adjusting the UN’s figures to account for the lower births in many countries, I estimate the global fertility rate last year was 2.18, i.e. below the 2.21 replacement threshold. It could be even lower than that, as it’s likely that the birth rate in many African countries saw a larger fall than the UN estimated.

This doesn’t mean the global population is already falling. “Demographic momentum” means that women born in the 1990s and 2000s are currently having children, while their parents’ generations haven’t yet died. Longevity, meanwhile, is increasing. So although global births are falling, they still exceed deaths. At present rates the human population will peak in around thirty years. Then start plummeting.

Economists have long predicted fertility rates would decline as countries become wealthier. But the fall over the past decade has happened in rich, middle-income and poor countries. It has also been faster than anyone predicted.

South Korea is the most extreme case. The fertility rate last year was 0.72 — roughly one-third of the replacement rate. In 2015, it was 1.24. In less than a decade, South Korea has transitioned from very low fertility to astonishingly low. And there’s no sign of this decline slowing. The same trend can be seen across Asia (China, Vietnam, Taiwan, Thailand, the Philippines and Japan).

But it isn’t unique to Asia. Turkey’s fertility rate plummeted from 3.11 in 1990 to 1.51 in 2023. The UK’s was 1.83 in 1990, 1.49 in 2022. The situation in Latin America is striking too: Chile and Colombia had rates of 1.2 last year, Argentina and Brazil were at 1.44 — all well below the UK. Each of them had high fertility rates three decades ago.

A non-exhaustive list of countries where the rate isn’t only below replacement but falling quickly includes India, the US, Canada, Mexico, Bangladesh, Iran and all of Europe. We know less about Africa because of poor quality data. The available evidence, however, suggests it’s undergoing a rapid decline: where we do have more reliable information — Egypt, Tunisia and Kenya — it shows fertility rates plummeting at an unprecedented pace. The only countries where fertility isn’t falling are the former Soviet Central Asia republics, and they are too small to make much difference.

Whenever I raise the issue of falling birth rates during lectures, I’m always met with three questions. The first is: won’t a falling population benefit the environment? This is misguided. A gently falling population could be good for sustainability, but we’re facing population collapse and economic turmoil. Environmental concern is a “luxury good”: we do it more when prosperous. Voters in 2050 in a country with acute budgetary problems caused by an ageing population will care a lot less about global warming.

The second question is: can’t we bring in more immigrants? But the falling population is for the planet, not one country. Every Argentinian who moves to Spain alleviates Spain’s demographic woes but aggravates Argentina’s. This argument also ignores the huge number of immigrants needed to keep the population constant in countries such as South Korea. By 2080, 80 percent of people living there would need to be immigrants or the children of immigrants. Can any society absorb so many without social unrest?

It’s not clear either that immigration fixes pensions or healthcare costs. When immigrants are young, they pay taxes; as they grow old, they draw pensions and use health services. The same is true for first- and second-generation immigrant children.

The third question is: won’t AI make a population collapse immaterial by doing all the work for us? This is wishful thinking. AI’s effect on productivity won’t match the hype. Daron Acemoglu, a leading expert on the macroeconomics of AI, estimates it will increase productivity by 0.66 percent over the next decade. Even multiplying his estimate by ten, the figure would be much lower than what’s needed to overcome the declining labor force. The gulf between what the McKinseys of the world think and what the real experts think is vast.

Then there’s the fact that AI can’t deliver the services actually needed. It’s easier to teach a machine to read financial statements than to empty bedpans. The problems caused by population collapse, such as empty rural areas and unbalanced family networks, cannot be fixed by AI.

Countries from France to South Korea have introduced policies such as extended parental leaves and generous child tax credits. These have had limited success in reversing the decline. Raising a child is an eighteen-year commitment; extending parental leave from two to six months offers marginal relief. Ditto tax relief schemes.

Fertility rates have fallen faster in large metropolitan areas than in rural areas, probably because of housing costs. Take Bogota, Colombia. Last year its fertility rate was 0.9, far lower than in rural Colombia. The same is true in Mexico. In Mexico City, the fertility rate last year was 0.95, much lower than in rural Mexico. Both cities are very expensive. Extra-low fertility rates in South Korea are most likely driven more by the astronomical real-estate prices in Seoul than by any other variable. Small, expensive homes deter fertility.

Our societal structures have also become deeply unwelcoming to large families. Child car seats are a good example. In the UK, children must use a child car seat until they’re twelve. There’s evidence this lowers birth rates as it makes it harder to fit more than two children in a car. When I was young my parents put four of us in the back seat. This isn’t to argue for repealing car-seat regulations, but it’s an instance of government policies having unintended consequences.

Another issue is that social norms have shifted: raising children isn’t a priority for many, not in the more conservative societies of East Asia or the more progressive ones of northern Europe. In 2016, China abandoned its one-child policy and allowed couples to have two children. The fertility rate increased from 1.57 in 2015 to 1.7 in 2016. By 2018, the effect had disappeared: it fell to 1.55 — lower than before the restriction was lifted.

If the UK government were to devise a strategy to increase the fertility rate from 1.49 to around 1.8 — still below the replacement rate but much closer to a sustainable level — it would need to address a mix of economic factors and societal support for large families.

Societal support could include making it easier for the young to marry. Safer streets would allow children to spend more time unsupervised and travel to school on their own, easing the burden on parents. School holidays could be organized in ways that don’t disrupt parents’ work.

Creating the conditions for large families to flourish is the only way to reverse the trend in fertility rates. If we fail to do so, then the coming demographic winter will be far harsher than anyone cares to admit.

This article was originally published in The Spectator’s UK magazine. Subscribe to the World edition here.