How Will Rich Country Fertility Ever Get Back Above Replacement?

For population to be steady or rising, the average women needs to have at least two kids. In almost every rich country- including the United States, all of Europe, and all of East Asia- this isn’t happening. In the extreme case of South Korea, where total fertility averages about one child per woman, the population will fall by half each generation. If this were to go on for 10 generations, South Korea would go from a country of 50 million people- larger than any US state- to one of 50 thousand people, far smaller than any US state. This sounds crazy and I don’t expect it will actually happen- but I can’t say what exactly will stop it from happening.

Global population growth has fallen from a peak of 2.1% per year to the current 1%, and is expected to fall to 0 by 2100. The remaining population growth will happen in poor countries, then stop for the same reasons it did in rich countries- the demographic transition from poverty, argicultural work, and high infant mortality to high incomes, high education, and low infant mortality. As the graph below shows, higher income is an incredibly strong predictor of low fertility- and so if economic growth continues, we should expect fertility to continue falling. But where does it stop?

2019 TFR from Population Reference Bureau vs 2019 PPP-adjusted GDP Per Capita fron World Bank

Some have theorized a “J-curve” relationship, where once incomes get high enough, fertility will start rising again. You can see this idea in “Stage 5” of Max Roser’s picture of the demopgraphic transition here:

This makes sense to me in theory. As countries get richer, desired fertility (the number of kids each woman wants to have) has fallen, but realized fertility (the number of kids each woman actually has) has fallen faster. In a typical rich country women would like to have 2-2.5 kids, but actually ends up having about 1.5. There are many reasons for this, but some are clearly economic- the high cost of goods and services that are desired by rich-country parents, like child care, education, and spacious housing near high-paying jobs. Perhaps in a rich enough country all these could be obtained with a single income (maybe even from a part-time job). But it seems we aren’t there yet. Even zooming in on higher-income countries, higher incomes still seem to lead to lower fertility.

TFR vs GDP Per Capita in countries with GDP Per Capita over 30k/yr

The only rich countries with fertility above replacement are Panama and the Seychelles (barely meeting my 30k/yr definition of rich), Kuwait (right at replacement with 2.2 kids per woman), and Israel- the biggest outlier, with 3 children per woman at a 42k/yr GDP. This hints that pro-fertility religious culture could be one way to stay at or above replacement. But in most countries, rising wealth seems to drive a decline in religiousity along with fertility. Will this trend eventually come to Israel? Or will it reverse in other countries, as more “pro-fertility” beliefs and cultures (religious or otherwise) get selected for?

To do one more crazy extrapolation like the disappearance of South Korea, the number of Mormons is currently growing by over 50% per generation from a base of 6 million while the rest of the US is shrinking. If these trends continue (and setting aside immigration), in at most 10 generations the US will be majority-Mormon. Again, I don’t actually expect this, but I don’t know whether it will be falling Mormon fertility, non-Mormon fertility somehow rising back above replacement, or something else entirely that changes our path.

What would a secular pro-fertility culture look like? For my generation, I see two big things that make people hold back from having kids: a desire to consume experiences like travel and nightlife that are harder with kids, and demanding careers. I see more potential for change on the career front. Remote work means that more quality jobs will be available outside of expensive city centers. Remote work, along with other technological and cultural changes, could make it easier to work part-time or to re-enter the work force after a break. Improving educational productivity so that getting better-education doesn’t have to mean more years of school would be a game-changer; in the short run I think people will spend even more time in school but I see green shoots on the horizon.

Looking within the US, we are just beginning to see what looks like the “J-curve” happening. Since about the year 2000, women with advanced degrees began to have more children than those with only undergraduate education (though still fewer than those with no college, and still below replacement):

From Hazan and Zoabi 2015, “Do Highly Educated Women Choose Smaller Families?”

We see a similar change with income. In 1980 women from richer households clearly had fewer children, but by 2010 this is no longer true:

Fertility of married white women, from Bar et al. 2018, “Why did rich families increase their fertility? Inequality and marketization of child care”

The authors of the papers that produced the two graphs above argue that this change is due to “marketization”, the increasing ability to spend money to get childcare and other goods and services that make it easier to take care of kids. If this is true, it could bode well for getting back to replacement- markets first figure out how to make more excellent daycare and kid-related gadgets, then figure out how to make them cheap enough for wide adoption.

Zachary Bartsch June 17, 2021 / 9:09 am

The ‘V’ shape in (income percentile, fertility) space may be a general equilibrium. If the primary child-rearing aid is a nanny (labor), then we should not expect the remainder of the income distribution to follow suit as incomes rise broadly.

LikeLike

Scott Buchanan June 19, 2021 / 3:08 pm

Interesting data and discussion. I wonder if the fertility vs GDP plot would tighten up if you used median instead of mean GDP. I am guessing that in some oil-rich countries like Kuwait and Saudi Arabia there are some super-rich sheiks and a lot of poorer residents (including non-citizen workers from other countries) who do the actual work. Plus in those more traditional cultures, likely lots of grandmas and aunts around to help with childcare. I suspect very few Saudi women work outside the home, compared to , say, Singapore.

I think getting married much older now than say 30 years ago is a factor in births…if a woman marries at say 31 instead of 24, waits 2 years to let the marriage settle and job firm up, then tries to have kids, there is a much higher risk of infertility. A 30-something woman struggling with that problem told me, “I spend my best child-bearing years in graduate school”. Though, if you are wealthy enough, there are in-vitro technologies that can often get you a child in your late thirties.

LikeLike