Here’s an interesting sociological fact. It’s mathematically trivial but is hugely important for how we think about the world.
For a long time, a large proportion of the population have been kids and young families. That’s because, until relatively recently, average lifespans were short: children often did not live to adulthood, and the steady state required a continuing supply of replacement births. Then, in the past century, death rates went down, but people were still in the habit of starting parenthood young and having many kids. There were changes in all sorts of things relating to vital statistics, but the population was still predominantly young. Long-term, though, we should expect children and young families to be a smaller proportion of the population.
The steady-state balance between life expectancy and age distribution is clear enough. For example, if everybody lives to be exactly 80, then in steady state the average age will be 40, only 20% of people will be under the age of 15, etc. This is just an approximation, but it gives the general sense of things.
Nonetheless, given centuries of recent experience, it has just seemed like the natural order of things to have lots of children underfoot, to the extent that when a society is low on babies compared to historical expectations, it feels wrong.
Typically this has been framed in terms of the end of exponential population increase, or as a surfeit of old people relative to the number of working-age people to support them, but a lower proportion of children and young family is part of the story.
To this point, sociologist Philip Cohen recommends a demography article by Ansley Coale from 1964: “How a Population Ages or Grows Younger.” Here’s Cohen:
It makes some key observations that are counterintuitive at first and serve as a great introduction to demographic thinking. Most important (according to him in this 1987 interview) is that reducing mortality in populations with high mortality often leads to a younger population — because more children survive. And then he explains that if we want to survive as a species . . . we are going to have to reduce fertility rates drastically, or else live with very high mortality rates. Fortunately, that’s what we’re doing [reducing fertility rates]. But then he’s also sad that the future will be much older, and less vibrant, than the past.
The point is obvious but often seems to get lost in the details. For example, just today this article appeared in the business section of the New York Times:
China Told Women to Have Babies, but Its Population Shrank Again
Faced with falling births, China’s efforts to stabilize a shrinking population and maintain economic growth are failing. . . .
Chinese women have been shunning marriage and babies at such a rapid pace that China’s population in 2023 shrank for the second straight year, accelerating the government’s sense of crisis over the country’s rapidly aging population and its economic future. . . .
9.02 million babies were born in 2023, down from 9.56 million in 2022 and the seventh year in a row that the number has fallen. Taken together with the number of people who died during the year — 11.1 million — China has more older people than anywhere else in the world . . .
The shrinking and aging population worries Beijing because it is draining China of the working-age people it needs to power the economy. . . .
I like this article a lot more than many articles on the topic, in that it doesn’t just treat the birth rate as some sort of policy knob to be turned; it focuses on women’s choices. I guess that men’s preferences make a difference here too, but I’ll defer to the sociologists and demographers on this one.
One thing that does bother me a bit about this news article is that it’s so focused on a single country. I mean, sure, yeah, I get it: it’s an article about China so they should be talking about China. I just think they should also mention that birth rates are falling around the world. For example, in the above graph it would be good to see the corresponding lines for some other countries.
There’s also this thing where they seem to be attributing the falling birth rates in part to women’s equality and in part to unequal conditions for women. There seems to be some incoherence in the story, in part because they’re explaining a single trend using multiple predictors. Anyway, I’m not trying to slam the news article—I like it!—; I’m just thinking of ways it could be better.
To return to the Cohen’s post . . . He did this cool thing where he downloaded Coale’s article from 2004, entered it into a word processor, and edited it. He created a scholarly remix, and details are at the end of his post. Here’s Cohen:
It’s a good article for teaching, but it was written before he even knew the Baby Boom was ending, and before fertility fell all over the world, and so on. If you don’t know that history it can be confusing to read, and if you do know the history it is still distracting and you want to keep looking things up to see what’s going on today.
So I [Cohen] updated it . . . The trickiest part was the discussion of global growth rates over the last 2000 years. He took the world from 250 million people in year 0 to 3 billion people in 1960. I wanted to go to 8 billion in 2024. . . .
Then the future projections were a little tricky, too. I got it to this:
If, on the other hand, mankind can avoid nuclear war, pandemics, and population decimation because of global climate change, and bring the fruits of modern technology, including prolonged life, to all parts of the world, the human population must become an old one, because only a low birth rate is compatible in the long run with a low death rate, and a low birth rate produces an old population. In fact, if by 2090 the global expectation of life at birth increases to eighty-two – a level achieved by a few dozen countries so far – and the global number of children born per women falls from 2.3 to 1.9, the global population will peak around 10.2 billion. In that scenario, which is the United Nations current projection, the decline in fertility would make the whole world older than the high-income countries today: with 17 per cent under fifteen and 24 per cent over sixty-five (compared with 16 per cent and 19 per cent, respectively, today).
So, yeah, basic stuff. In the short and medium term there’s the possibility of balancing out demographic changes through immigration; in steady state it doesn’t make sense to expect or demand to have a population with a higher proportion of kids and a fewer proportion of oldsters.
Times have changed, and an important step in going forward is to recognize how history has led us to a misleading sense of what should feel normal.
Crazy that they think you need more people to have economic growth. Why do people think this?
David:
I don’t know that anyone thinks you need more people to have economic growth. It’s got to depend on who those “more people” are, and how you define “economic growth” (is it total output, or per-capita?). I guess there must be a lot of literature on the topic in economics and political science, but as with all such observational studies, it will be difficult or impossible to untangle causal relationships from historical comparisons.
Thanks, Andrew!
This is an interesting point: “falling birth rates in part to women’s equality and in part to unequal conditions for women.” Now that having children is (more) optional, it seems any social change leads to more people choosing something else. That’s what you would expect if at time 1 every woman was having as many children as possible — then you would expect any kind of choice to reduce fertility. It’s not that way anymore, but it still seems every social changes pushes birth rates down.
Also on the damned if you do or don’t angle: Given conditions of gender inequality, sometimes either a surplus or a shortage of females produces bad outcomes for women — not what you would expect from a simplistic supply-and-demand perspective. If there are too many women, like on US college campuses today, (straight) men are less required to treat them decently and it encourages objectifying practices. Makes sense that oversupply reduces “value.” But then when there are too few women (like in some Asian countries today) that should increase their “value,” but the people who profit from it end up being sex traffickers and mail-order-bride hustlers.
Readers of this blog might want to read this informative 2011 article which purports to explain the origin of China’s “one child policy”:
https://freakonomics.com/2011/11/the-academic-origins-of-chinas-one-child-policy/
“Misadventures in Baby Making” is the title. The article involves a Dutch mathematician and a Chinese “ballistics missiles specialist.” They “went out for beers and talked about population planning” as a problem amenable to optimum control theory. The former “thought nothing of it,” “while the latter worked out a theory of a bidirectional limit to the total fertility rate. This helped the Chinese leader to formulate the state family planning policy, particularly the ‘one couple, one child’ policy.”
Rather frequently there are news articles about a “demographic crisis” — often about Japan, Italy, China, South Korea, or Russia, but plenty of other countries get mentioned and sometimes it’s characterized as a “crisis” for the whole world — and they always make me angry…because we have an overpopulation crisis too! We have nearly emptied the oceans of fish, just about every aquifer is being drained unsustainably, the atmospheric CO2 concentration increased more last year than in any previous year, about 3 billion people don’t get a healthy diet because they can’t afford one, etc. etc. Perhaps I’d feel better about the articles if they acknowledged both sides of the coin, but instead we get one article that say a low birth rate is a crisis and another that says high population is a crisis, neither view acknowledging the other.
I can see how a sustained very low birth rate really would be a problem — not enough people to work; people having to do physical labor into their seventies or eighties; and so on — as well as maybe not being so much fun: not enough youthful energy in the arts and culture. But a birth rate that is slightly below replacement, leading to gradual population decline and a population in which 20% of the people are under 20 years old… that might be good or bad or in between, depending on how society handles it, but surely it isn’t _guaranteed_ to be a crisis. It’s only a crisis if society doesn’t adapt. And adapting to that seems much less problematic than the other way around: if a country’s population doubles every fifteen years, that’s a really rapid timescale on which to build roads and sewer systems and schools and hospitals, and if the country is trying to maintain food independence then availability of water and arable land can quickly become huge problems.
> I can see how a sustained very low birth rate really would be a problem — not enough people to work
No. The growth in productivity swamps the decrease in people. See https://cepr.net/do-we-need-more-people/
David,
Do you think that no matter how much fertility falls, productivity gains are guaranteed to make up for it? I don’t think that’s plausible. Consider fertility of zero. In 70 years, everyone will be over 70. It’s possible in principle that machines will have gotten so good that no human will need to work, but I don’t see any way to guarantee that.
If we agree that fertility of zero would be a problem, then how about 0.01? And so on.
I’m guessing he means that withing the range that is at all plausible for the future this won’t matter. Fertility globally is what 2.3 ish
https://ourworldindata.org/fertility-rate
We might imagine it falling to 1.7 or even 1.0 but if it falls to 0.01 we are doomed as a species and that could only happen through quite amazing levels of disaster.
So in the range say 0.7+ ish the productivity could take up the slack maybe
Yes, in the range of 0.7 productivity could make up the slack maybe. But maybe not! And it’s not like 0.7 is a floor: South Korea is lower than that, for example.
I think the prospect that global fertility is going to fall “too low” is very slim, it’s not something I worry about, whereas the fact that we already have over 8 billion people and the number is still climbing is something I do worry about.
If the claim is that global low fertility is not a crisis or even a worry, I agree; that’s part of the point of my original comment. But I don’t agree with any claim, explicit or implied, that productivity is guaranteed to continue to increase at any particular rate relative to fertility. It might, but it might not.
My position is basically as follows.
1) Economic growth is not an end in itself. It is valuable only insofar as it contributes to a freer, more comfortable life for people.
2) The extent to which economic growth achieves that value is best measured by per capita GDP, not by total GDP.
3) Sub-replacement fertility is unlikely to be a drag on per capita GDP, unless it is truly extreme.
If the world fertility rate went down to 0.7, then there might actually be a problem. For example, there might be some very narrow professional specialties that just stopped getting new people. (“The last person who knew how this thing worked three years ago.”) But I’m not seeing that as a serious problem for the foreseeable future.
People talk a lot about there being too few workers for all the retirees, but it’s better to ask about the ratio of workers to all non-working-age people. That is, you need to consider the kids as well as the retirees. And that’s a ratio that won’t be affected nearly as strongly by a fertility decrease.
Is the journalistic hand-wringing on this subject a product of ideology or self-interest? Actually, I don’t think so. Journalists at the NYT and other news outlets instinctively know not to use a headline like “Gradual change occurs that society is gradually adapting to.” Blaring about a crisis is the best way to get views. In a world with so many real problems, journalists shouldn’t have to make up new ones; but they do.
Phil,
> Do you think that no matter how much fertility falls, productivity gains
> are guaranteed to make up for it?
Did you read the article I linked to?
David, yes, I read the article. It does nothing to convince me (or, I dare say, anyone) that productivity is guaranteed to make up for low fertility, now and forever, everywhere, no matter how low the fertility rate.
I’m a big fan of low fertility! I’d love to have it go well below 2.0 worldwide for decades. But I don’t see any basis for claiming what it seems that you’re claiming.
> But I don’t see any basis for claiming what it seems that you’re claiming.
I am “claiming” exactly what the linked article is explaining: Concerns about low fertility in the real world are based on a misunderstanding of the actual economics.
This is a good point. I’m surprised no one saw fit to discuss it further:
People talk a lot about there being too few workers for all the retirees, but it’s better to ask about the ratio of workers to all non-working-age people. That is, you need to consider the kids as well as the retirees. And that’s a ratio that won’t be affected nearly as strongly by a fertility decrease.
Daniel,
> I’m guessing he means that within the range that is at all plausible for
> the future this won’t matter.
Obviously. I assumed we were discussing what countries should really do or worry about, not what happens in a mathematical model if you change the parameters to their limits.
David, I think you linked the wrong article. The one you posted just talks about Japan.
Phil:
I agree with you. The point of the above post is that part of the problem, I think, is that for almost all of human history, a large percentage of the population has been kids and young families, so we’ve developed the intuition that this is the natural order of things. At the level of math, it’s clear that a long-lived stable population will have a lower percentage of kids and young families, but that we’ve found it hard to internalize this fact.
That’s a great observation about the cultural aspect of population change. (An interesting variation on the adage that old age and treachery will overcome youth and skill.)
There is no overpopulation crisis, on net globally. Rather than us running out of food, deaths from malnutrition are down and obesity is spreading. We haven’t bothered conserving aquifers because we haven’t needed to. We could desalinate ocean water if we actually needed more, but we haven’t bothered doing that on much of a scale because again we haven’t needed to. We could produce the energy needed for desalination via nuclear power, but people like J. Storrs Hall aren’t making the decisions. The truth is that we are nowhere near Malthusian limits. Groups like the Amish & ultra-orthodox are growing not because they figured out how to get all the food for themselves while everyone else starved, but because they want to.
Looked at on the scale of 100k years ago to 100k years from now, the extraction rate of fossil fuels looks like a delta function peak. We know that population can’t grow forever, it has doubled between 1960 and 1999. If it doubled every 40 years human mass would equal the crust of the earth in how long? I did the math on my phone so I could have done it wrong but I get 1400 years or so. Now, it’s ridiculous to get to even 1% of the mass of the earths crust so let’s say 1/100000 the of the mass of the earths crust… On that basis I calculate we must stop growing entirely in 780 years. But even this is ridiculous because the mass of food needed to feed that many people is also a really large fraction of the earths crust… So pretty quickly we can whittle this down to no more than a couple hundred years at the absolute outside. Doubling every 40 years if we grow for 200, that’d be 256B people. That might very likely be more than the total population of all mammals larger than a mouse today. So, yeah for sure we are going to see population growth stop in less than 200 years, probably much less. I believe population will decline before the end of my kids lives. Id guess 10-15B is the peak population at most.
Whether that decline is sharp or gradual is the main question. Growth at the rate we’ve seen between 1492 and now was a one time event.
Just like with the rest of stats, I dislike that fertility rates are talked about at some average level with, without acknowledging potential changes in the future or heterogeneity within the population. Many complain about the cost of raising a kid being prohibitive; that could be much lower 50 years from now. There are also religious, racial, and just value heterogeneity that leads to some subgroups having higher birth rates than others within a country; US birth-rates will rebound when there are only Mormons left.
No, Mormons aren’t insular enough.
The NYT is anti-China, it should focuse on the below data. The pot calling kettle black.
The IQ of web ACX group are almost all above IQ 100 and 87% White.
https://www.astralcodexten.com/p/acx-survey-results-2024
There is a survey with sample response size 1333 and the avg NChild=0.85, implying that the sub IQ 100 fertility is 2.47 . https://pbs.twimg.com/media/GMylKIYaAAAVwdm.png
A better estimate could be those with income data (N=4,548) and avg NChild=0.72 and implying the other fertility rate 2.60 . https://pbs.twimg.com/media/GMyrPb7aEAAJ6VY.png
The most controvsal ACX24 result is on Non Coersive Eugenics. Is it a taboo subject if only a small minority of
Old Marxist or Conservative are against that?
https://pbs.twimg.com/media/GNawtmPbsAAluyV.png
https://pbs.twimg.com/media/GNa9xUIbkAA_fF-.png