The social conservatives' idea that the recent US drop in fertility is the product of "liberalism" makes no sense. There was a far greater decline in the period 1962-67 (prior to the rise of feminism) and then a second one in 1971-76 (during first-wave feminism), but then the rate rebounded and stabilized during the period 1977-2007. The period of decline we're in begins with the Great Recession and reaches bottom during the pandemic. (I'm using CDC data.)
Only one of these drops could reasonably be attributed to social liberalism. The first marks the end of the anomalous Baby Boom and the first wave of easy access to contraception. (A comparably large earlier drop from 1910-1930 may reflect sharp drops in childhood mortality.) To account for the current drop we need to examine the very substantial social changes occurring from 2008 on, including economic uncertainty, rising expectations for parenting, and the spread of virtual culture, including the last's correlation social isolation and delays in dating, sex, and marriage (among many other factors).
The current drop is part of a global drop in fertility rates that includes countries like Russia, Hungary, Kenya, Pakistan, Indonesia, etc. (I'm relying on a site called Macrotrends.) It is clearly not generated by political or social ideology. The likely causes range from drops in childhood mortality together with more available contraception in less developed countries to rises in living standards and virtual culture in more developed ones.
I think the NYTimes piece is clearly superficial, as Ms. Henkel indicates, but the problem is real. From the US perspective, commenter Austin L's point is the key one. After 2030 the US is going to be entirely dependent on immigration to maintain (much less grow) its total population. Working-age population is tightly linked to economic output, and if we allow population to fall we are either betting on AI to replace people *without* causing chaotic social disruption or accepting a future far less opulent than our present (while US declines help buffer negative trends in other countries).
I agree with a lot of this, except for the paragraph that starts:
> To account for the current drop we need to examine the very substantial social changes occurring from 2008 on ...
Why not accept the simplest explanation, that it's all downstream of developments in birth control? It just takes a while for technological and cultural changes to work their way through society.
Contraception has been getting cheaper, more reliable, and more accessible for a couple centuries now, (improving across each of those dimensions at different rates), and I don't see any reason to think fertility decline is more complicated than that. (Or, if you prefer, it's way more complicated than that, because there are infinite reasons people do not choose to have kids.)
We're not going to outlaw contraception, so i hope we can find a way to inspire more people to have families. But I think the search for discrete causes is futile: now that we can choose, we just want to do other stuff.
Hi Deadpan. I agree with you that we shouldn't complicate a problem by ignoring the dominance of one factor. However, I don't think contraceptive technology can account for what we're seeing in the US. (It may be the dominant factor in poorer countries that have only more recently made contraception readily available.)
I think ease of contraception had a lot to do with the decline in the 1960s and early 1970s (just based on experience), but I think the argument that contraceptive technology and associated cultural changes required an additional forty years to "work their way through society" has no evidence to back it. For example, the US fertility rate *rose* every year between 2001 through 2007 -- altogether a rise of about 7% -- but then abruptly changed direction: from 2008-2020 it dropped annually (one year it stayed flat), a total of about 23% over that period. (It has since rebounded a bit and is now close to 2016 levels, still about 16% below 2007 levels). (I should note that I'm using CDC data for 2001-2007, but Macrotrends for the later period because the CDC stopped collecting data after 2018!).
I know of no contraceptive breakthroughs leading up to 2008, and I have no idea why we'd think 2008 was a watershed year for long existing technologies to have maximum impact. So while it may be that there is a single dominant cause I strongly doubt contraception is it. I think it's actually much more likely a combination of cultural changes that have accelerated social isolation and risk aversion. I could speculate on what those have been (and did in my original post), but I have no expertise.
I hear you, but I expect the search for evidence of the type you want is going to run aground, due to endless confounders, contradictory indications, and poorly-conceived timelines. You're going to find the same kind of red flags for any theory you propose: the cultural, technical, and political factors at work are too dense and complex to observe in obvious metrics. Those forces create strange loops and feed back on each other.
I think it's like obesity: technological developments have given us unprecedented access to calories (good), but also endangered our health (bad). It's unfolded unevenly, and different populations are affected unequally. But looking for any other cause (soda, tv, decline in youth sports, etc ) is doomed to failure.
I don't think that all the other lines of evidence are messy. Marriage rates are also declining worldwide. Don't you think that has something to do with it? It doesn't seem plausible to me that access to contraception is causing a decline in marriage rates, so it seems like an independent factor.
Well, then we agree! But that means opening the door to cultural factors that aren't solely driven by specific technological advancements. In other words, it's not all just access to birth control. If it was, then I don't think we can or ought to reverse the decline. If it also has to do with people's attitudes and worldview, perhaps there's more reason to believe change is possible.
As an addendum to my sibling comment, it probably sounds like I'm promoting epistemic nihilism, and that's because I basically am. I don't think we can really get to the bottom of this by looking at legible data sets. I think my hypothesis lines up pretty well with history, if you make allowances for local fluctuations, but I also know that's a blank check that allows too many degrees of freedom.
My story is fundamentally hand-waving, but rooted in powerful intuition pumps. First, an appreciation of how contraception interferes with evolved mechanisms that guaranteed successful reproduction (and adaptation). Second, consideration of the counterfactual "what would happen if birth control were no longer available?" (Not a recommendation!.)
I think the rise of intensive parenting and the extended period of “emerging adulthood” where young adults still depend on their parents is an underrated factor. I know that intensive parenting is largely a US/Western phenomenon, but, *worldwide*, children are no longer put to work; increasingly they are in school until at least 16 or 18. Which means parents are on the hook for feeding, housing, clothing, etc. them until they do get jobs or move out. I’m not saying “let’s return to the bad old days of children working in the mines at 7 years old!” I am saying that children are economic liabilities rather than assets and this is true even in poorer countries. So parents will have fewer of them.
I'm not sure it's wise to draw simple chronological inferences like that. Since TFR decline is probably multifactorial and noisy, the presence of flat periods doesn't disproof the notion that any particular factor is exerting continuous downward pressure.
Consider that when you average a larger pool of countries (cf: the Our World in Data page on this), the decline looks pretty steady. To me this suggests noise may be playing a role in the idiosyncrasies of the US-specific curve.
I agree that it doesn't make sense to attribute the decline to very specific social moments like first wave feminism, for the reasons I just articulated. Nor would I blame specific political ideologies. But I also don't feel comfortable rejecting a broad cultural thesis outright on the basis of the minute details of the shape of these TFR curves. Certainly globalism has been a major factor in the postwar period, making a cultural thesis more plausible. Surely there is a strong link between changes in attitudes to childrearing and technological advance, but we aren't obligated to believe that link is inevitable. Some people still have large families without joining the Amish. Point being, culture matters.
The best theory I've heard about the decline in birth rates in developed countries I found in a tweet from Aella: "I suspect it's that civilizational advancements improved the quality of childfree life much more than it improved the quality of life with children. Before, life without kids might have been nearly as hard as life with kids, so you weren't actually losing much." I think about this often and am surprised that I haven't seen this specific angle discussed more.
Agreed! Changing opportunity costs is a huge part of this story. Not having kids, especially for women, used to mean sentencing yourself to life as an outcast. No longer.
And at the same time, it could be argued that the intensification of parenting culture (at least for middle class and up), along with government support not keeping up with the realities of dual-income households, has made life with kids more onerous.
Totally. Finding ways to facilitate low-intensity parenting (including for women, not just men) should be top of mind for anyone who cares about the birth rate. There is a good line from this 2022 Atlantic piece about modern parenting norms: "Moving away from intensive parenting will also require a culture in which parents’ needs outweigh child optimizations" --- being pro adult is often being pro child. https://www.theatlantic.com/family/archive/2022/05/intensive-helicopter-parent-anxiety/629813/
How we raise kids has changed quite a lot. Social conservatives today and people that hype trad-wife-ism have a totally ahistorical view of how mothers raised children in the past.
Parents just spent way less time around their kids unless they were working together (such as in the days of peasant farming) or later the kids went to school, and otherwise the parents didn't really pay much attention to them otherwise. The rich had people to pay attention to them. The idea of a mother doting over her kids addressing their every need and following them around is a very modern invention.
So parenting today is a much, much more involved process, and that's not always a bad thing! That does mean a parents ability to work, while giving kids the attention they are expected to in modern society can be difficult. Which is why I think so many conservatives want a return to the single income nuclear household, it just fits easier with a high parent-child association time view of the world.
Personally I think that the solution here is actually one that's more of a return to the mean of human existence. Jobs closer to home (including WFH, which is as old as farming lol), rather than trying to force the return of the single income house.
I think calling it a "modern" invention even undersells it a bit! From my vantage point it seems to be something that has come to be in the last 10-15 years. I'm an elder millennial and while I think it's true that my mom was more involved with me growing up than her mom was with her, it definitely was not to the level that it seems like is expected today.
As with pretty much all things, I blame social media.
> "The opportunity cost hypothesis: Spending time on parenting means giving up something. Because the world has improved around us, that “something” is better than it used to be."
I have a hard time believing that the birth rate with any generation is ever going to equal what it was before the 2000s. That is why not having a net decrease in immigration is so important. It might not happen right away or for several generations, but if America closes its doors to the rest of the world, especially those places where young working-age men and women are immigrating from, it will surely suffer.
There will be more suffering in the countries these working-age people are leaving. Sending countries also have a declining birth rate (and a lot less wealth).
Mexico has a lower birth rate than the US! (and that is just one example) There are going to be very few countries, going forward, with a large supply of young people eager to up stakes and live in another country. For one thing, as you note, these countries will actually need some working age people themselves. For another, as families get smaller, there are fewer “spares.” You can’t “spare” two or three children for emigration and still have two or three at home to look after you (in countries without old-age pensions and the like, or just for company). Families with one or two children are not going to want to see those kids leave and possibly never return.
There was an article years ago in the Wall Street Journal (IIRC) about a Mexican couple, both of whom were from very large (10 or so kids!) farm families, who had clawed their way up to clerical jobs in Tijuana. They had ONE son, who they were planning to send to college, and that son said he looked forward to having a well-paying middle-class job and staying in Mexico.
So, tl;dr you are right, there is no more of the vast labor pool of young people willing to immigrate to a strange and often hostile country in search of opportunity. Almost all the world’s polities are aging.
My biggest question that I feel this debate often ignores to me is why do we assume either trend will stay in perpetuity? As far as I can tell, in some ways we are returning to sort of pre-baby boom numbers. There were far fewer humans on the planet then.
It seems to me entirely possible that we had a few big generations and our species reacted collectively by having fewer kids. In a few generations, perhaps the kids of Gen Z, will have larger ones. It seems like people act like whatever trend is currently happening will happen forever and I don’t know why. It wasn’t true with population growth re the population bomb, so it seems a little absurd to assume the reverse will be true. “Just following the trend line” makes less sense when things don’t always happen on a trend line.
There are several theories about self-correcting fertility out there, most of them have to do with economic development and the labor market, i.e. the idea that societies move through predictable stages of high fertility/ high mortality, eventually falling to an equilibrium of low fertility/low mortality. But what comes after is more of an unknown.
I am also optimistic about post-industrial societies being able to course correct. You're right that a wave of nostalgia for big families might ripple across Gen Alpha. The only trouble is that once the infrastructure of a society is built around low fertility (pension systems, parental leave etc, urban planning etc.) it makes it hard for people to break with the family planning patterns of earlier generations.
I mean sure, but I think basically infrastructure adjusts when it has to. No one was “prepared” for the baby boom, and they have reshaped infrastructure around them as they’ve aged through it (including to elderly care as it turns out). I just think this is something we should collectively get less neurotic about, it will likely stabilize over generations and to me it always feels like a cudgel against women. We’re either foolishly having too many children and draining resources or cold prudes who refuse to bring warmth and children into the world. It seems to me altogether unlikely that we will ever have some platonic ideal of a perfect TFR and likely not for any length of time even if we could get it. What a perfectly delightful eternal crisis point.
Yes, I actually really appreciate this, and generally find the fixation on women in the birth rate debate very revealing. Men are also postponing having children, despite also experiencing declining fertility with age (albeit less dramatically). That should get more attention. I enjoyed Matthew Yglesias' take on this recently: https://www.slowboring.com/p/yelling-at-ambitious-young-women
I largely agree. There is an equilibrium. Alarmism is usually not the answer.
I don't think humanity is just going to complacently extinguish itself because they can't be bothered to have kids. As the demographic pressure of the elderly increases, people will adapt and respond.
That doesn't mean we shouldn't be worried at all, though. Just because we'll eventually course-correct doesn't mean we wouldn't have been better off doing it earlier.
At some point, both sides have to acknowledge what the issue actually is though. Liberals and conservatives both miss the mark because both focus on college educated women when the data just keeps showing us over and over again that working class women are the ones with more dramatic drops including not having children at all. Probably because college educated women can afford more fertility treatments!
But conservatives hate that women are college educated and liberals are the college educated ones so both just miss the problem completely. Worth exploring what solutions look like if we actually focused on the problem!
Uh, I think assuming nobody wants unplanned teen pregnancies is pretty generous. It's no coincidence that the same people who oppose women having carreer opportunities tend to support ineffective 'abstinance only' sex ed that directly leads to unplanned teen pregnancies.
I am just one person, but having spent a lot of time among people in favor of stay-at-home mothers and marriage and against contraception, I am very confident that very few of them are secretly in favor of unmarried teenage girls having children.
My experience is they are definitely in favor of 18 & 19 year olds getting locked into stay at home parenting as long as it comes with a wedding too, after conception is not preferred but still ok as it meets the goals. This is the preferred course for young people in most fundamentalist religions as it locks them into a lifestyle that goes with the religion by taking away wider opportunities in life.
We must know very different people. None of the religious conservatives I know would identify with what you're saying. They aren't fools. You can't trick an 18-year-old into living a conservative lifestyle -- divorce and single parenthood are the likeliest outcomes, both of which are considered great evils by these folks.
Okey dokey we must know different religious people indeed. You can't trick them but you can TRAP them by making sure they have young kids before they are old enough to really take stock and realize they would have preferred to go a different way. This is a big reason that young marriage with early kids is a hallmark of multiple fundamentalist sects in the US and around the world. It's promoted in Christian colleges across the US for sure, and in Mormon ones too.
Fair enough! I guess I'm trying to convince you that this phenomenon can be explained without malice or deception. If a religious group is convinced that getting married and having a lot of children is a good way for young people to go about life, and they are able to genuinely convince the young people in their orbit of the same, then it makes sense that this would be the result.
If, on the other hand, this religious group was trying to get young women only marginally affiliated with them to have unprotected sex and accidentally conceive children because they thought this would somehow lead those women to morally upright lifestyles, that would be very foolish and bad. But I don't think this is happening.
Unless you're radically committed to self-determination as the ultimate moral good, it doesn't make a lot of sense to believe you know what's best for your child, but not encourage them to do it out of fear that at some point in the future they'll change their mind and wish they had pursued something which you, the parent, believe is a bad thing to pursue. What kind of parent would you be if you didn't try to influence your child's understanding of right and wrong?
To take an extreme example, I would rather my children choose to work as a manual laborer and make very little money then make it big by climbing the ladder in a drug cartel. Whether they're likely to later regret their choice and wish they had joined the gang does not enter as a consideration.
Point being, there is a non-malicious logic that would explain the pattern that you are observing.
The bigger error, to me, is to not recognize that for population level effects, the age when mothers have babies matters. Imagine the extremes. In one scenario, every woman still alive except , say, 5 percent that are infertile, has babies only at age 19 and age 21, and then those children have children at the same age, etc. Then imagine those same 95 percent of women have babies only at age 44 and age 45 (if they're still alive), and their children follow the same pattern. After any period of time, the first scenario will have a lot more people than the second. The doom scenarios of population shrinking can happen even if the vast majority of woman have replacement level fertility but they're all waiting until very late to have children. The timing actually does matter for population level trends.
If everyone has precisely two children in their life, then a change in that timing can change the total population. But that change is only a finite scaling of the population. If the population is currently stable and you double the age at which parents have children on average, then in the long run the population will stabilize at half the current population.
But if the total number of children people have in their life is less than 2, then population will never stabilize, and will fall to zero, whereas if the total number of children people have in their life is more than 2, then the population will never stabilize and will eventually grow towards infinity.
If total fertility rate goes up while age at birth also goes up, you may have a short period of population decrease while the constant scaling of the age increase phases in, but the exponential growth of the higher fertility rate eventually wins out, and the same for the reverse.
Age at birth does matter for the population over the next few generations, but there’s no significant threat from a one-time halving of the population - it’s only if the population is expected to *keep* halving every generation or two that it gets threatening.
Good points. I added the "still alive" and "modest percent infertile" so that the total replacement would be close to but lower than 2. I also was not intending to validate the doom scenario of the population falling to literally zero. I think that extinction scenario is kind of silly to think about when the population having a relatively quick one-time halving could be pretty darn catastrophic on its own. At least in our and our great-grandchildren's lifetimes, how fast demographic change happens is incredibly important, and that rate of change is heavily influenced not only by average fertility rate, but also the average age of mothers when they have children.
Lol NYT is what you get when a society glorifies being “bad at math” and then proudly taking high level jobs like in journalism, without any real math skills.
This isn't what happened at all, the demographer who helped with that article certainly isn't bad at math and wanted that chart to show the thing it showed because he thinks that's the important message.
To me it seems like the drop is primarily due to a drop in unplanned pregnancies, since the drop is so concentrated among young people. That is almost certainly due to an increase use in contraceptives and lower rates of sex among young people.
Social conservatives blame the drop on feminism and liberalism, but women advancing in their careers has no relation to being a teen mom. I don't think anyone seriously wants to increase the rate of totally unplanned and unwelcome pregnancies.
What we need to do is increase the number of *planned* pregnancies to unprecedented levels. There aren't many lessons to learn from the past, because the reason why we had more pregnancies in the past was due to more *unplanned* pregnancies.
The difficulty is getting the political support for the kinds of solutions that we'd need to increase planned pregnancies. A large portion of the liberal public still believes in the "population bomb" and thinks fewer births is a good thing. And conservatives usually recoil at the idea of paying women to have children with fears about "welfare queens".
We could do substantial income transfers to parents, and especially parents with children under 5. But for that to be effective would be extremely expensive and need to target parents of all income levels. Paying for that would be difficult, and it is hard to imagine that it will be the top priority anytime soon. Ideally we would transfer some money from payments to seniors to this, but as society ages that will be even harder to do.
The NYT article is quite misleading. Several of the charts seem to contradict each other, and it uses weasel words ("most women will eventually become mothers" – this implies a minimum TFR of 0.25, if 50.1% of women have one child) to avoid taking a stand, instead addressing nebulous "anxieties" around motherhood. What's the point of data journalism if you're just muddying the waters?
Regardless of the cause, fairness, or politics of the fertility drop, we should celebrate the drop itself.
Climate change is reducing available cropland. Sea level rise will force huge coastal populations to squeeze into the remaining livable areas. Either could easily lead to wars when the required land is on the other side of a national or cultural border. We could all cooperate to help the affected populations, but history says that will not be the chosen solution everywhere.
For that matter, what drives climate change? Adding a whole human to the population far outweighs all the electric cars, heat pumps, and low-on-the-food-chain eating you can handle.
AI's affect on society? In the long run we may have happy, easy lives as AI delivers what we need with less human effort. In the coming decades it is more likely to lead to painful economic disruptions and young workers will be hit the hardest.
Homelessness? We can replace parks and farmland with high-density housing, or perhaps not have so many people to house. To be clear, I don't take homelessness lightly. In any scenario there will be many people who need help and should get it.
And that thing about the human population declining to 4-digit levels. Statistical nonsense! Extrapolating a trend based on a few years or decades of data to many thousands of years is meaningless. Fertility can turn around in a generation.
Good explanation. But note it’s not GenZ that’s really concerning here. The first GenZ cohort is in their late 20s and shows similar fertility to the last cohort of millennials. The big drop of from the first millennial cohort, born in the early 1980s, to the second and third, who are currently in their 30s and show no sign of catching up
A practical example of women having children later, but not having less of them overall could be seen in Germany in the last couple of years: Women born 1965 already had 0.93 children at age 30, while women born 1975 had only 0.76. In the end however, the final fertility rate of the 1965 cohort was a bit lower (1.55) than that of the 1975 cohort (1.58), as can be seen in the data of the German Federal Statistical Office (sadly in german: https://www.destatis.de/DE/Themen/Querschnitt/Demografischer-Wandel/_Grafik/_Interaktiv/kohortenfertilitaet-kinder-je-frau.html). So it is possible (though final fertility rates of about 1.6 are still quite low). The steep drop in fertility rates in the US doesn't make it seem very likely though.
Demographer here-- while I think it's a fair point that a percentage change figure gives the wrong impression about how these age-specific trends balance out, I don't think it’s fair to say that the Upshot piece itself overstated its conclusions or was "wrong" about birth rates.
The title was "Women in Their 20s May Not Be Having Babies, but by 45 Most Probably Will", which referred to the share of women who ever have any child by age 45 (the 4th figure), and is still around 88% for the most recent cohorts. So "most" is reasonable language. The demographers quoted in the piece also state clearly they are not suggesting full recuperation due to postponement: "it's unlikely that fertility will fully catch up to the level of earlier decades."
The titles of the figures in the Upshot piece are perhaps the most guilty of lacking nuance...they really should say "most women *up until now*" have still had on average of two children by age 45. But the visualization itself is accurate- and it is the closest to ground truth we have, an observed fact about what women's fertility has actually looked like, which is really important context for the falling period rates-- I don't think this could be considered "misleading."
The revised absolute-numbers chart from Novasad is a useful counter to the percentage change framing, but it makes an error in the opposite direction. It uses period data to make a claim about a cohort process that hasn't finished yet. The women who are not having babies in their 20s today have not yet lived through their 30s, so their postponed births haven’t show up in this chart yet. And the use of absolute birth counts rather than rates misses changes in the size of each age cohort over this time, so it doesn't make sense to estimate potential "recuperation" this way, regardless.
Women already aged 35 (born 1990) are closer to their completed fertility, and they are still projected to have around 1.9 children on average by the end of their reproductive years based on the most recent data, so lower, but not a dramatic drop. Their birth rates for ages 35-39 would have to increase around *30%* from current levels to reach completed fertility around 2.0...this is still a sizeable increase, but not as dramatic as the 10X back of the envelope number quoted here based on period counts. And what the relative change figure *does* convey is the direction of change of these rates-- birth rates are growing at older ages. For the 1995 cohorts and younger, the range of plausible outcomes for completed fertility is wider and more uncertain because their future has not been written. But the pessimistic scenario of a dramatic break from prior cohort patterns also requires a strong assumptions. The truth is almost certainly in between the "pure delay" and "pure decline" extreme scenarios.
In reality, there is no perfect way to visualize these trends because of the need to project period trends on uncompleted cohorts: Age-specific fertility rates are in flux, with births shifting to later ages. Younger women simply have not yet lived the years that will answer the question of where they will end up. Period measures during a postponement phase will always make fertility drops look worse than it turns out to be, which is fair to point out given the emphasis on over-interpreting period rates.....that is a mathematical feature of the calculation, not a misleading or “wrong” conclusion. There are trade-offs to any visualization of the data because we simply can't fill in those future unknown numbers (and rolling forward current period rates when postponent is happening doesn't predict anything at "90%", whatever that means).
The social conservatives' idea that the recent US drop in fertility is the product of "liberalism" makes no sense. There was a far greater decline in the period 1962-67 (prior to the rise of feminism) and then a second one in 1971-76 (during first-wave feminism), but then the rate rebounded and stabilized during the period 1977-2007. The period of decline we're in begins with the Great Recession and reaches bottom during the pandemic. (I'm using CDC data.)
Only one of these drops could reasonably be attributed to social liberalism. The first marks the end of the anomalous Baby Boom and the first wave of easy access to contraception. (A comparably large earlier drop from 1910-1930 may reflect sharp drops in childhood mortality.) To account for the current drop we need to examine the very substantial social changes occurring from 2008 on, including economic uncertainty, rising expectations for parenting, and the spread of virtual culture, including the last's correlation social isolation and delays in dating, sex, and marriage (among many other factors).
The current drop is part of a global drop in fertility rates that includes countries like Russia, Hungary, Kenya, Pakistan, Indonesia, etc. (I'm relying on a site called Macrotrends.) It is clearly not generated by political or social ideology. The likely causes range from drops in childhood mortality together with more available contraception in less developed countries to rises in living standards and virtual culture in more developed ones.
I think the NYTimes piece is clearly superficial, as Ms. Henkel indicates, but the problem is real. From the US perspective, commenter Austin L's point is the key one. After 2030 the US is going to be entirely dependent on immigration to maintain (much less grow) its total population. Working-age population is tightly linked to economic output, and if we allow population to fall we are either betting on AI to replace people *without* causing chaotic social disruption or accepting a future far less opulent than our present (while US declines help buffer negative trends in other countries).
I agree with a lot of this, except for the paragraph that starts:
> To account for the current drop we need to examine the very substantial social changes occurring from 2008 on ...
Why not accept the simplest explanation, that it's all downstream of developments in birth control? It just takes a while for technological and cultural changes to work their way through society.
Contraception has been getting cheaper, more reliable, and more accessible for a couple centuries now, (improving across each of those dimensions at different rates), and I don't see any reason to think fertility decline is more complicated than that. (Or, if you prefer, it's way more complicated than that, because there are infinite reasons people do not choose to have kids.)
We're not going to outlaw contraception, so i hope we can find a way to inspire more people to have families. But I think the search for discrete causes is futile: now that we can choose, we just want to do other stuff.
Hi Deadpan. I agree with you that we shouldn't complicate a problem by ignoring the dominance of one factor. However, I don't think contraceptive technology can account for what we're seeing in the US. (It may be the dominant factor in poorer countries that have only more recently made contraception readily available.)
I think ease of contraception had a lot to do with the decline in the 1960s and early 1970s (just based on experience), but I think the argument that contraceptive technology and associated cultural changes required an additional forty years to "work their way through society" has no evidence to back it. For example, the US fertility rate *rose* every year between 2001 through 2007 -- altogether a rise of about 7% -- but then abruptly changed direction: from 2008-2020 it dropped annually (one year it stayed flat), a total of about 23% over that period. (It has since rebounded a bit and is now close to 2016 levels, still about 16% below 2007 levels). (I should note that I'm using CDC data for 2001-2007, but Macrotrends for the later period because the CDC stopped collecting data after 2018!).
I know of no contraceptive breakthroughs leading up to 2008, and I have no idea why we'd think 2008 was a watershed year for long existing technologies to have maximum impact. So while it may be that there is a single dominant cause I strongly doubt contraception is it. I think it's actually much more likely a combination of cultural changes that have accelerated social isolation and risk aversion. I could speculate on what those have been (and did in my original post), but I have no expertise.
I hear you, but I expect the search for evidence of the type you want is going to run aground, due to endless confounders, contradictory indications, and poorly-conceived timelines. You're going to find the same kind of red flags for any theory you propose: the cultural, technical, and political factors at work are too dense and complex to observe in obvious metrics. Those forces create strange loops and feed back on each other.
I think it's like obesity: technological developments have given us unprecedented access to calories (good), but also endangered our health (bad). It's unfolded unevenly, and different populations are affected unequally. But looking for any other cause (soda, tv, decline in youth sports, etc ) is doomed to failure.
I don't think that all the other lines of evidence are messy. Marriage rates are also declining worldwide. Don't you think that has something to do with it? It doesn't seem plausible to me that access to contraception is causing a decline in marriage rates, so it seems like an independent factor.
I think both phenomena are part of a dynamic equilibrium, and mutually amplify each other.
Well, then we agree! But that means opening the door to cultural factors that aren't solely driven by specific technological advancements. In other words, it's not all just access to birth control. If it was, then I don't think we can or ought to reverse the decline. If it also has to do with people's attitudes and worldview, perhaps there's more reason to believe change is possible.
As an addendum to my sibling comment, it probably sounds like I'm promoting epistemic nihilism, and that's because I basically am. I don't think we can really get to the bottom of this by looking at legible data sets. I think my hypothesis lines up pretty well with history, if you make allowances for local fluctuations, but I also know that's a blank check that allows too many degrees of freedom.
My story is fundamentally hand-waving, but rooted in powerful intuition pumps. First, an appreciation of how contraception interferes with evolved mechanisms that guaranteed successful reproduction (and adaptation). Second, consideration of the counterfactual "what would happen if birth control were no longer available?" (Not a recommendation!.)
I think the rise of intensive parenting and the extended period of “emerging adulthood” where young adults still depend on their parents is an underrated factor. I know that intensive parenting is largely a US/Western phenomenon, but, *worldwide*, children are no longer put to work; increasingly they are in school until at least 16 or 18. Which means parents are on the hook for feeding, housing, clothing, etc. them until they do get jobs or move out. I’m not saying “let’s return to the bad old days of children working in the mines at 7 years old!” I am saying that children are economic liabilities rather than assets and this is true even in poorer countries. So parents will have fewer of them.
I'm not sure it's wise to draw simple chronological inferences like that. Since TFR decline is probably multifactorial and noisy, the presence of flat periods doesn't disproof the notion that any particular factor is exerting continuous downward pressure.
Consider that when you average a larger pool of countries (cf: the Our World in Data page on this), the decline looks pretty steady. To me this suggests noise may be playing a role in the idiosyncrasies of the US-specific curve.
I agree that it doesn't make sense to attribute the decline to very specific social moments like first wave feminism, for the reasons I just articulated. Nor would I blame specific political ideologies. But I also don't feel comfortable rejecting a broad cultural thesis outright on the basis of the minute details of the shape of these TFR curves. Certainly globalism has been a major factor in the postwar period, making a cultural thesis more plausible. Surely there is a strong link between changes in attitudes to childrearing and technological advance, but we aren't obligated to believe that link is inevitable. Some people still have large families without joining the Amish. Point being, culture matters.
The best theory I've heard about the decline in birth rates in developed countries I found in a tweet from Aella: "I suspect it's that civilizational advancements improved the quality of childfree life much more than it improved the quality of life with children. Before, life without kids might have been nearly as hard as life with kids, so you weren't actually losing much." I think about this often and am surprised that I haven't seen this specific angle discussed more.
Agreed! Changing opportunity costs is a huge part of this story. Not having kids, especially for women, used to mean sentencing yourself to life as an outcast. No longer.
And at the same time, it could be argued that the intensification of parenting culture (at least for middle class and up), along with government support not keeping up with the realities of dual-income households, has made life with kids more onerous.
Totally. Finding ways to facilitate low-intensity parenting (including for women, not just men) should be top of mind for anyone who cares about the birth rate. There is a good line from this 2022 Atlantic piece about modern parenting norms: "Moving away from intensive parenting will also require a culture in which parents’ needs outweigh child optimizations" --- being pro adult is often being pro child. https://www.theatlantic.com/family/archive/2022/05/intensive-helicopter-parent-anxiety/629813/
How we raise kids has changed quite a lot. Social conservatives today and people that hype trad-wife-ism have a totally ahistorical view of how mothers raised children in the past.
Parents just spent way less time around their kids unless they were working together (such as in the days of peasant farming) or later the kids went to school, and otherwise the parents didn't really pay much attention to them otherwise. The rich had people to pay attention to them. The idea of a mother doting over her kids addressing their every need and following them around is a very modern invention.
So parenting today is a much, much more involved process, and that's not always a bad thing! That does mean a parents ability to work, while giving kids the attention they are expected to in modern society can be difficult. Which is why I think so many conservatives want a return to the single income nuclear household, it just fits easier with a high parent-child association time view of the world.
Personally I think that the solution here is actually one that's more of a return to the mean of human existence. Jobs closer to home (including WFH, which is as old as farming lol), rather than trying to force the return of the single income house.
I think calling it a "modern" invention even undersells it a bit! From my vantage point it seems to be something that has come to be in the last 10-15 years. I'm an elder millennial and while I think it's true that my mom was more involved with me growing up than her mom was with her, it definitely was not to the level that it seems like is expected today.
As with pretty much all things, I blame social media.
fyi, this is also the explanation offered by demographers Spears and Geruso in their (excellent) book, *After the Spike*. I discuss this a bit in my book review: https://www.goodthoughts.blog/p/a-human-abundance-agenda
> "The opportunity cost hypothesis: Spending time on parenting means giving up something. Because the world has improved around us, that “something” is better than it used to be."
I have a hard time believing that the birth rate with any generation is ever going to equal what it was before the 2000s. That is why not having a net decrease in immigration is so important. It might not happen right away or for several generations, but if America closes its doors to the rest of the world, especially those places where young working-age men and women are immigrating from, it will surely suffer.
There will be more suffering in the countries these working-age people are leaving. Sending countries also have a declining birth rate (and a lot less wealth).
Mexico has a lower birth rate than the US! (and that is just one example) There are going to be very few countries, going forward, with a large supply of young people eager to up stakes and live in another country. For one thing, as you note, these countries will actually need some working age people themselves. For another, as families get smaller, there are fewer “spares.” You can’t “spare” two or three children for emigration and still have two or three at home to look after you (in countries without old-age pensions and the like, or just for company). Families with one or two children are not going to want to see those kids leave and possibly never return.
There was an article years ago in the Wall Street Journal (IIRC) about a Mexican couple, both of whom were from very large (10 or so kids!) farm families, who had clawed their way up to clerical jobs in Tijuana. They had ONE son, who they were planning to send to college, and that son said he looked forward to having a well-paying middle-class job and staying in Mexico.
So, tl;dr you are right, there is no more of the vast labor pool of young people willing to immigrate to a strange and often hostile country in search of opportunity. Almost all the world’s polities are aging.
My biggest question that I feel this debate often ignores to me is why do we assume either trend will stay in perpetuity? As far as I can tell, in some ways we are returning to sort of pre-baby boom numbers. There were far fewer humans on the planet then.
It seems to me entirely possible that we had a few big generations and our species reacted collectively by having fewer kids. In a few generations, perhaps the kids of Gen Z, will have larger ones. It seems like people act like whatever trend is currently happening will happen forever and I don’t know why. It wasn’t true with population growth re the population bomb, so it seems a little absurd to assume the reverse will be true. “Just following the trend line” makes less sense when things don’t always happen on a trend line.
There are several theories about self-correcting fertility out there, most of them have to do with economic development and the labor market, i.e. the idea that societies move through predictable stages of high fertility/ high mortality, eventually falling to an equilibrium of low fertility/low mortality. But what comes after is more of an unknown.
I am also optimistic about post-industrial societies being able to course correct. You're right that a wave of nostalgia for big families might ripple across Gen Alpha. The only trouble is that once the infrastructure of a society is built around low fertility (pension systems, parental leave etc, urban planning etc.) it makes it hard for people to break with the family planning patterns of earlier generations.
I mean sure, but I think basically infrastructure adjusts when it has to. No one was “prepared” for the baby boom, and they have reshaped infrastructure around them as they’ve aged through it (including to elderly care as it turns out). I just think this is something we should collectively get less neurotic about, it will likely stabilize over generations and to me it always feels like a cudgel against women. We’re either foolishly having too many children and draining resources or cold prudes who refuse to bring warmth and children into the world. It seems to me altogether unlikely that we will ever have some platonic ideal of a perfect TFR and likely not for any length of time even if we could get it. What a perfectly delightful eternal crisis point.
Yes, I actually really appreciate this, and generally find the fixation on women in the birth rate debate very revealing. Men are also postponing having children, despite also experiencing declining fertility with age (albeit less dramatically). That should get more attention. I enjoyed Matthew Yglesias' take on this recently: https://www.slowboring.com/p/yelling-at-ambitious-young-women
I largely agree. There is an equilibrium. Alarmism is usually not the answer.
I don't think humanity is just going to complacently extinguish itself because they can't be bothered to have kids. As the demographic pressure of the elderly increases, people will adapt and respond.
That doesn't mean we shouldn't be worried at all, though. Just because we'll eventually course-correct doesn't mean we wouldn't have been better off doing it earlier.
Both liberals and conservatives should embrace greater access to assisted reproductive technologies to help close the desire-intention gap.
At some point, both sides have to acknowledge what the issue actually is though. Liberals and conservatives both miss the mark because both focus on college educated women when the data just keeps showing us over and over again that working class women are the ones with more dramatic drops including not having children at all. Probably because college educated women can afford more fertility treatments!
But conservatives hate that women are college educated and liberals are the college educated ones so both just miss the problem completely. Worth exploring what solutions look like if we actually focused on the problem!
Uh, I think assuming nobody wants unplanned teen pregnancies is pretty generous. It's no coincidence that the same people who oppose women having carreer opportunities tend to support ineffective 'abstinance only' sex ed that directly leads to unplanned teen pregnancies.
I am just one person, but having spent a lot of time among people in favor of stay-at-home mothers and marriage and against contraception, I am very confident that very few of them are secretly in favor of unmarried teenage girls having children.
My experience is they are definitely in favor of 18 & 19 year olds getting locked into stay at home parenting as long as it comes with a wedding too, after conception is not preferred but still ok as it meets the goals. This is the preferred course for young people in most fundamentalist religions as it locks them into a lifestyle that goes with the religion by taking away wider opportunities in life.
We must know very different people. None of the religious conservatives I know would identify with what you're saying. They aren't fools. You can't trick an 18-year-old into living a conservative lifestyle -- divorce and single parenthood are the likeliest outcomes, both of which are considered great evils by these folks.
Okey dokey we must know different religious people indeed. You can't trick them but you can TRAP them by making sure they have young kids before they are old enough to really take stock and realize they would have preferred to go a different way. This is a big reason that young marriage with early kids is a hallmark of multiple fundamentalist sects in the US and around the world. It's promoted in Christian colleges across the US for sure, and in Mormon ones too.
Fair enough! I guess I'm trying to convince you that this phenomenon can be explained without malice or deception. If a religious group is convinced that getting married and having a lot of children is a good way for young people to go about life, and they are able to genuinely convince the young people in their orbit of the same, then it makes sense that this would be the result.
If, on the other hand, this religious group was trying to get young women only marginally affiliated with them to have unprotected sex and accidentally conceive children because they thought this would somehow lead those women to morally upright lifestyles, that would be very foolish and bad. But I don't think this is happening.
Unless you're radically committed to self-determination as the ultimate moral good, it doesn't make a lot of sense to believe you know what's best for your child, but not encourage them to do it out of fear that at some point in the future they'll change their mind and wish they had pursued something which you, the parent, believe is a bad thing to pursue. What kind of parent would you be if you didn't try to influence your child's understanding of right and wrong?
To take an extreme example, I would rather my children choose to work as a manual laborer and make very little money then make it big by climbing the ladder in a drug cartel. Whether they're likely to later regret their choice and wish they had joined the gang does not enter as a consideration.
Point being, there is a non-malicious logic that would explain the pattern that you are observing.
Very clear layout of the statistics. The solution is immigration!
The bigger error, to me, is to not recognize that for population level effects, the age when mothers have babies matters. Imagine the extremes. In one scenario, every woman still alive except , say, 5 percent that are infertile, has babies only at age 19 and age 21, and then those children have children at the same age, etc. Then imagine those same 95 percent of women have babies only at age 44 and age 45 (if they're still alive), and their children follow the same pattern. After any period of time, the first scenario will have a lot more people than the second. The doom scenarios of population shrinking can happen even if the vast majority of woman have replacement level fertility but they're all waiting until very late to have children. The timing actually does matter for population level trends.
If everyone has precisely two children in their life, then a change in that timing can change the total population. But that change is only a finite scaling of the population. If the population is currently stable and you double the age at which parents have children on average, then in the long run the population will stabilize at half the current population.
But if the total number of children people have in their life is less than 2, then population will never stabilize, and will fall to zero, whereas if the total number of children people have in their life is more than 2, then the population will never stabilize and will eventually grow towards infinity.
If total fertility rate goes up while age at birth also goes up, you may have a short period of population decrease while the constant scaling of the age increase phases in, but the exponential growth of the higher fertility rate eventually wins out, and the same for the reverse.
Age at birth does matter for the population over the next few generations, but there’s no significant threat from a one-time halving of the population - it’s only if the population is expected to *keep* halving every generation or two that it gets threatening.
Good points. I added the "still alive" and "modest percent infertile" so that the total replacement would be close to but lower than 2. I also was not intending to validate the doom scenario of the population falling to literally zero. I think that extinction scenario is kind of silly to think about when the population having a relatively quick one-time halving could be pretty darn catastrophic on its own. At least in our and our great-grandchildren's lifetimes, how fast demographic change happens is incredibly important, and that rate of change is heavily influenced not only by average fertility rate, but also the average age of mothers when they have children.
Lol NYT is what you get when a society glorifies being “bad at math” and then proudly taking high level jobs like in journalism, without any real math skills.
This isn't what happened at all, the demographer who helped with that article certainly isn't bad at math and wanted that chart to show the thing it showed because he thinks that's the important message.
I think that’s a naive read of naive people.
Is your claim that @familyunequal doesn't know the relevant math? Or that Cain Miller tricked him into presenting his data that way?
Lol it’s in the article right here. They made a “statistically dubious claim”. Duh.
I agree that the claim is statistically dubious. But that's not because the relevant people don't know math.
It was a clear mathematical error that the journalist didn’t pick up on.
To me it seems like the drop is primarily due to a drop in unplanned pregnancies, since the drop is so concentrated among young people. That is almost certainly due to an increase use in contraceptives and lower rates of sex among young people.
Social conservatives blame the drop on feminism and liberalism, but women advancing in their careers has no relation to being a teen mom. I don't think anyone seriously wants to increase the rate of totally unplanned and unwelcome pregnancies.
What we need to do is increase the number of *planned* pregnancies to unprecedented levels. There aren't many lessons to learn from the past, because the reason why we had more pregnancies in the past was due to more *unplanned* pregnancies.
The difficulty is getting the political support for the kinds of solutions that we'd need to increase planned pregnancies. A large portion of the liberal public still believes in the "population bomb" and thinks fewer births is a good thing. And conservatives usually recoil at the idea of paying women to have children with fears about "welfare queens".
We could do substantial income transfers to parents, and especially parents with children under 5. But for that to be effective would be extremely expensive and need to target parents of all income levels. Paying for that would be difficult, and it is hard to imagine that it will be the top priority anytime soon. Ideally we would transfer some money from payments to seniors to this, but as society ages that will be even harder to do.
The NYT article is quite misleading. Several of the charts seem to contradict each other, and it uses weasel words ("most women will eventually become mothers" – this implies a minimum TFR of 0.25, if 50.1% of women have one child) to avoid taking a stand, instead addressing nebulous "anxieties" around motherhood. What's the point of data journalism if you're just muddying the waters?
Regardless of the cause, fairness, or politics of the fertility drop, we should celebrate the drop itself.
Climate change is reducing available cropland. Sea level rise will force huge coastal populations to squeeze into the remaining livable areas. Either could easily lead to wars when the required land is on the other side of a national or cultural border. We could all cooperate to help the affected populations, but history says that will not be the chosen solution everywhere.
For that matter, what drives climate change? Adding a whole human to the population far outweighs all the electric cars, heat pumps, and low-on-the-food-chain eating you can handle.
AI's affect on society? In the long run we may have happy, easy lives as AI delivers what we need with less human effort. In the coming decades it is more likely to lead to painful economic disruptions and young workers will be hit the hardest.
Homelessness? We can replace parks and farmland with high-density housing, or perhaps not have so many people to house. To be clear, I don't take homelessness lightly. In any scenario there will be many people who need help and should get it.
And that thing about the human population declining to 4-digit levels. Statistical nonsense! Extrapolating a trend based on a few years or decades of data to many thousands of years is meaningless. Fertility can turn around in a generation.
Good explanation. But note it’s not GenZ that’s really concerning here. The first GenZ cohort is in their late 20s and shows similar fertility to the last cohort of millennials. The big drop of from the first millennial cohort, born in the early 1980s, to the second and third, who are currently in their 30s and show no sign of catching up
A practical example of women having children later, but not having less of them overall could be seen in Germany in the last couple of years: Women born 1965 already had 0.93 children at age 30, while women born 1975 had only 0.76. In the end however, the final fertility rate of the 1965 cohort was a bit lower (1.55) than that of the 1975 cohort (1.58), as can be seen in the data of the German Federal Statistical Office (sadly in german: https://www.destatis.de/DE/Themen/Querschnitt/Demografischer-Wandel/_Grafik/_Interaktiv/kohortenfertilitaet-kinder-je-frau.html). So it is possible (though final fertility rates of about 1.6 are still quite low). The steep drop in fertility rates in the US doesn't make it seem very likely though.
Demographer here-- while I think it's a fair point that a percentage change figure gives the wrong impression about how these age-specific trends balance out, I don't think it’s fair to say that the Upshot piece itself overstated its conclusions or was "wrong" about birth rates.
The title was "Women in Their 20s May Not Be Having Babies, but by 45 Most Probably Will", which referred to the share of women who ever have any child by age 45 (the 4th figure), and is still around 88% for the most recent cohorts. So "most" is reasonable language. The demographers quoted in the piece also state clearly they are not suggesting full recuperation due to postponement: "it's unlikely that fertility will fully catch up to the level of earlier decades."
The titles of the figures in the Upshot piece are perhaps the most guilty of lacking nuance...they really should say "most women *up until now*" have still had on average of two children by age 45. But the visualization itself is accurate- and it is the closest to ground truth we have, an observed fact about what women's fertility has actually looked like, which is really important context for the falling period rates-- I don't think this could be considered "misleading."
The revised absolute-numbers chart from Novasad is a useful counter to the percentage change framing, but it makes an error in the opposite direction. It uses period data to make a claim about a cohort process that hasn't finished yet. The women who are not having babies in their 20s today have not yet lived through their 30s, so their postponed births haven’t show up in this chart yet. And the use of absolute birth counts rather than rates misses changes in the size of each age cohort over this time, so it doesn't make sense to estimate potential "recuperation" this way, regardless.
Women already aged 35 (born 1990) are closer to their completed fertility, and they are still projected to have around 1.9 children on average by the end of their reproductive years based on the most recent data, so lower, but not a dramatic drop. Their birth rates for ages 35-39 would have to increase around *30%* from current levels to reach completed fertility around 2.0...this is still a sizeable increase, but not as dramatic as the 10X back of the envelope number quoted here based on period counts. And what the relative change figure *does* convey is the direction of change of these rates-- birth rates are growing at older ages. For the 1995 cohorts and younger, the range of plausible outcomes for completed fertility is wider and more uncertain because their future has not been written. But the pessimistic scenario of a dramatic break from prior cohort patterns also requires a strong assumptions. The truth is almost certainly in between the "pure delay" and "pure decline" extreme scenarios.
In reality, there is no perfect way to visualize these trends because of the need to project period trends on uncompleted cohorts: Age-specific fertility rates are in flux, with births shifting to later ages. Younger women simply have not yet lived the years that will answer the question of where they will end up. Period measures during a postponement phase will always make fertility drops look worse than it turns out to be, which is fair to point out given the emphasis on over-interpreting period rates.....that is a mathematical feature of the calculation, not a misleading or “wrong” conclusion. There are trade-offs to any visualization of the data because we simply can't fill in those future unknown numbers (and rolling forward current period rates when postponent is happening doesn't predict anything at "90%", whatever that means).
1. Fertility declines through the 20s (so the "35" line is wrong, but not necessarily for promising reasons): https://conference.nber.org/conf_papers/f179203.pdf
2. Derek Thompson argued (persuasively to me) that the Baby Boom was global, multi-factorial, “weird”, and most importantly: a temporary reversal of a long-term trend towards fewer births: https://www.derekthompson.org/p/what-caused-the-baby-boom-what-would