55 Comments
User's avatar
Matt C's avatar

I'm more worried about rage directed at the Amish once cities begin dwindling. It's not out of the question that our great grandchildren will find pretexts to murder them en masse, or forcibly assimilate their children into the dying dominant culture.

The religious subcultures that can preserve themselves and fertility won't "blame" us, or not for very long. They'll more pity us. We turned away from God's plan and cursed ourselves with decadence and infertility. Crumbling cities and vanished technological miracles will be warnings, not losses to lament.

Robin Hanson's avatar

I agree that the dominant civ might repress the rising insular fertile culture. In that case human extinction becomes more likely.

Matt C's avatar

Complete extinction doesn't feel very likely to me. As population declines we'll start losing features of modernity, and I expect fertility will start rising as we regress socially and technically. This is only my intuition, but it's a big world and decline will take centuries. I think we'll come out of the dive well before extinction.

Robin Hanson's avatar

I don't see why the dominant civ would regress culturally just because its population shrinks. More plausibly it continues to drift randomly in the vast space of possible cultures.

Matt C's avatar

If we consider what we have now to be a cultural peak, then we would expect big changes to pull in the direction of the global or historical mean.

Most of the world and most of history hasn't been liberal democracy that is careful with individual rights. Even now, without visible population shrinkage, we can see the USA moving away from liberal democracy towards strongman democracy.

I think it will be stressful for societies to know they're dwindling away to nothing, and they'll be more vulnerable to cultural disruptions internal and external.

Steven's avatar

I genuinely appreciate an effort to put responsibility on those currently best positioned to do something about it before it's too late (if it isn't already). Thank you.

Berder's avatar

This is just detached from reality. The UN projections are a peak around 2084, around 60 years in the future, not 30.

https://en.wikipedia.org/wiki/Human_population_projections

And economic growth is driven by technology, not population. Population decline or not, we're not going to forget how to make modern technology. Technology marches forward. Any analysis of the economy decades in the future that fails to account for AI in particular is a dream. The economy is going to get more and more automated, inevitably and obviously. How many machines and robots we have is just as relevant for productivity as how many people.

There are also far more people who want to innovate and are trained to do so, than spots available for them, looking at PhDs vs professorships. PhDs often fail to get work in their field. If the pool of people getting PhDs declines by 20%, there will still be more than enough people capable of doing all the innovation work we're currently doing - all the scientist and engineer jobs will still be filled. If it ever becomes a problem, it's a self-correcting problem, because those jobs will then be paid more, attracting more people into the PhD pipeline.

Robin Hanson's avatar

UN projections have been overly optimistic for a long time. Population matters a lot for innovation: https://www.overcomingbias.com/p/shrinking-economies-dont-innovate

Berder's avatar

On what basis do you claim UN projections are overly optimistic? I don't think they even think of it as optimism. For a long time the concern has been overpopulation, for which a higher population projection would be pessimistic.

What matters for innovation is the number of STEM PhDs doing research. X STEM PhDs doing research produce k * X amount of innovation per year. It doesn't suddenly go to 0 just because the population starts to shrink. Plus there are more PhDs who want to do research than there are available research positions, so the innovation rate won't feel the effect of decreased population for a long time. As long as the researchers are doing novel research, there's innovation happening at a rate proportional to the number of researchers.

Robin Hanson's avatar

No, most innovation isn't done by STEM PhDs.

Berder's avatar

Most/almost all new technology is made by STEM PhDs, as well as most/almost all new fundamental science that enables new technology. That's what I have in mind when you say "innovation" because it's the thing that creates lasting increase to GDP per capita and drives long-term change in society.

There are also business or marketing "innovations," but I wouldn't say those count. Many of those are net negative, e.g. cleverly deceptive marketing and enshittification of products, and those that aren't net negative tend to have narrow applicability that can't be adopted by other businesses. Arguably some such innovations could be positive, but as a whole their positive impact pales in comparison to new technology.

Robin Hanson's avatar

We have a big lit on "innovation" that causes econ growth. It just isn't mostly STEM PhDs.

Berder's avatar

Well, can you give some examples of what you mean then? In my perception, business leaders like to talk about innovation, but it's the researchers who are the ones actually doing it.

Anyway, it doesn't matter too much. The main point is that whatever kind of innovation it is, the rate is going to be proportional to the number of people involved in that type of innovation. So it won't fall off a cliff the moment population starts to decline. It will continue at a rate proportional to the population - which means per capita GDP will continue to rise.

Phil's avatar

UN has credibility issues generally, and I don't see a clear reason to believe they're particularly more equipped to be more accurate or ethical than usual in this case.

Compsci's avatar

“…all the scientist and engineer jobs will still be filled. If it ever becomes a problem, it's a self-correcting problem, because those jobs will then be paid more, attracting more people into the PhD pipeline.”

Perhaps not as certain as you imagine. Lower overall quality—as in IQ/intelligence—of the population, means less extremes to the right side of the Bell Curve. Hence the proposition outlined in Edward Dutton’s book, “The Genius Famine”. The qualities that make for a good Engineer also are wanting, regardless of their certification as engineers. There is a current problem with DEI and the subsequent lowering of standards both at the University level *and* in industry hiring. I have a son in such industry at a high level (National Director). Qualified engineers are getting rarer and rarer.

Here we are getting into the concept of the “Smart Fraction”—just how much of the populace at what level of IQ/intelligence/education does it take to maintain/expand a 1st world technological society such as we currently have and will need in the future. That fraction has been estimated, but to me is just that, an estimate. The Smart Fraction estimate is but a smallish percentage of the population and definitely shrinking in Western world societies. Partially due to population replacement and partially due to declining intelligence—dysgenic mutation and collapse of our education system.

Warm bodies filling jobs does not equate to competence. That is the issue we now experience—both at the white collar and blue collar levels. As some have mentioned, this problem is upon us and national decline already “baked into the cake”—especially as Boomers retire.

Berder's avatar

“I am, somehow, less interested in the weight and convolutions of Einstein’s brain than in the near certainty that people of equal talent have lived and died in cotton fields and sweatshops.”

-- Stephen Jay Gould

The talent is out there, it's just a matter of giving people a chance. Many more people have the potential to be geniuses than actually are in the right circumstances to get their PhD and get hired to one of the few job openings.

Additionally, researcher IQ isn't that high. We're talking around 120 IQ for a STEM researcher. And the act of getting a PhD is one of the very few things that actually increases a person's IQ.

Compsci's avatar

Gould never believe in IQ or the group differences in intelligence we now understand. His thoughts are completely wrong. For the most part, there are few to no undiscovered genius in low level employ.

Your observation fails to take in the “Smart Fraction” theory. As the national IQ declines, the number of important support people will also decline—this includes your majors in STEM fields—and so will the nation decline as a 1st world technological society.

It is not for undiscovered genius that Africa, for example, is backward. It is for the lack of human capital.

Berder's avatar

"For the most part, there are few to no undiscovered genius in low level employ." - Completely wrong.

I recall reading about a study of gifted children in the UK that followed them longitudinally. They all had very high IQs, and many of them did end up working in low level jobs such as taxi driver.

Are you aware that IQ is uncorrelated with wealth in the US? It's moderately correlated with income (correlation 0.43) - up to a point, after which the relation between IQ and income plateaus - but there's no correlation with wealth.

Additionally, it's a false dichotomy that either a genius is in a research job or is in a low-level job. Plenty of potential geniuses are just in non-research jobs that are nonetheless well-paying. But they aren't using their genius.

Personally, I had a perfect SAT score and am unemployed.

Compsci's avatar

“Additionally, it's a false dichotomy that either a genius is in a research job or is in a low-level job. Plenty of potential geniuses are just in non-research jobs that are nonetheless well-paying. “

No one spoke about wealth. No one spoke about genius in the sense of necessary for a productive life (this is heavily explained in my book reference I cited above), only reflective of a decline or shift to the left of the Bell Curve wrt to national intelligence. This shift affects all people as well as the national income they produce. See: “IQ and the Wealth of Nations” by Richard Lynn.

Take your perfect SAT (somehow I doubt this) and read some of the references I cited for my expressed opinions. Until you do or at least cite your references, this discussion is fruitless.

Why do I doubt your veracity? A perfect 1600 on the SAT is estimated to be obtained by 300–500 students per year out of nearly 2 million test-takers — roughly 0.02%–0.03% of examinees. Some conversion tables put a 1600 SAT in the ~99.97th percentile of test-takers, which corresponds it to an IQ in the upper 140s to low 150s range. That would make you a genius in anyone’s estimation. Yet—“ I had a perfect SAT score and am unemployed”. If this be true, I’d say there is more at work here than intelligence, i.e., behavioral problems.

Berder's avatar

My previous response was all towards the point that there are plenty of potential geniuses who are not in research jobs. You're off on a different tangent.

matthew's avatar

“And economic growth is driven by technology, not population.”

That claim reflects a 1970s era growth framing, common in analyses such as the Rockefeller Commission work, which treated technological progress as largely separable from labor supply and demographic scale.

At the aggregate level, the arithmetic has not worked, and the economic literature has moved on. Decades of growth accounting show that productivity gains from technology and capital deepening have not been large enough to offset sustained declines in the number of workers. Raising output per worker by some percent has not compensated for losing whole workers.

Across advanced economies, decades of rising technology investment have coincided with slowing productivity growth where labor forces are shrinking. Technology has its strongest effects on individual and per-capita outcomes. At the aggregate level that matters far less. For total growth and for the cultural and long-term discussions like this, population and labor scale remain binding constraints.

Berder's avatar

> Raising output per worker by some percent has not compensated for losing whole workers.

Are you saying the number of workers in the US and their total output has declined since the 1970s? Neither of those are true. With just a couple exceptions, the workforce has increased each year since the 1960s, and so has GDP per capita and total GDP.

Jack's avatar

What if declining populations cause total GDP to fall, but on a per capita basis GDP continues to rise? Is that a bad outcome? It's understood that economists and politicians want every observable quantity to grow – that makes life easy. But reality won't conform to that and increasingly these economists and politicians will need to stop being so lazy.

Robin Hanson's avatar

If innovation grinds to a halt, GDP per person won't grow. More plausibly shrinks due to falling scale economies.

Euglossine Librarian's avatar

I believe you underestimate the complexity of the world and the value of people in maintaining that knowledge. Furthermore, technology does not march forward, people push it forward. As the number of people decline, the ability of society to support innovators and researchers will decline, while the need for many professions will remain proportional to population

Berder's avatar

" I believe you underestimate the complexity of the world and the value of people in maintaining that knowledge" - The point of my last paragraph is that we have many *more* people who trained to be top-level experts (PhDs) than jobs for them. There are plenty more in line. The vast majority of people work in jobs that could be automated with good enough AI and robots.

"while the need for many professions will remain proportional to population" - that's no problem then, if the need is proportional to population then the need will decline if the population does.

John's avatar

You're really leaning into this Thesis, Robin. But perhaps civilisation will advance just fine with a few million smart agentic people and heck load of AI to solve coordination and execution. And that's the worst case where AI doesnt radically increase net creativity.

Jonas's avatar

Artificial wombs could make a difference

Robin Hanson's avatar

"A difference" is a long way from "enough difference".

Phil's avatar

Depends how quickly and aggressively the technology is used, and whether its use is forcibly blocked by declining cultures.

If we're to avoid something that resembles bronze age collapse at all, the best bet is some novel means like this which did not exist historically and outcompetes the default. Artificial wombs clearly meet the former, but it would take a lot to normalize it enough to meet the latter.

Tim Tyler's avatar

The claim that "our civ will continue to fall and slow for centuries" seems to depend on the unstated assumption that we won't get machine superintelligence anytime soon. Much the same is true of the forecast of an "economic peak". That's not going to happen anytime soon if robot workers take over most of the work previously performed by humans: machines will take over productivity and innovation.

I think these unstated assumptions are likely to be false. However the bigger concern here is that they are unstated. That could mislead readers.

Nogglehead's avatar

And the meek shall inherit the earth.

Leon Voß's avatar

I've tried abating it with writing, like you, but no one listens, so what's the point? What is there left to do?

Catherine Caldwell-Harris's avatar

Must birth rate forever stay below replacement unless an insular fertile religious subcultures takes over? No. Why: humans **want** to have children.

For the last few decades, American adults have reported that their ideal family size is about 2.7 children on average. This number has remained stable even as the actual U.S. fertility rate has dropped to 1.6 births per woman.

This gap between the desired number of children and the actual number of children means that the declining birth rate is not due to a lack of desire for children, but rather practical, real-world challenges like finances, career demands, and lack of family support structures.

Biologists refer to reproductive suppression and the "Self-Limitation" Hypothesis to describe how nonhuman animals limit reproduction when resources are low. Reproductive suppression holds for humans as well. (If you believe humans are an exception to the processes observed for nonhuman animals, then, explain why.)

Perceiving that you are at the bottom of a wealth scale inhibits the desire to have children, just as extreme caloric restriction inhibits ovulation, and how poverty caused ancestral women to abandon newborns.

People today say they can't have the children they want because of the affordability crisis. Stagnating wages and high inequality result from decades of specific governmental and corporate practices, even as worker productivity tripled. This has been documented in numerous books, e.g., Adam Cohen's Supreme Inequality; Jane Mayer's Dark Money.

The challenges of finances, career demands, and lack of family support structures will ameliorate when governments halt the diverse mechanisms that emerged beginning in the1970s to transfer wealth ever upwards.

Robin Hanson's avatar

"declining birth rate is not due to a lack of desire for children, but rather practical, real-world challenges like finances, career demands, and lack of family support structures."

Most all past eras had much less supportive conditions re these challenges, and yet had higher fertility. So those just can't be the main issue. I see no obvious forces that will reverse the "diverse govt mechanisms" you lament. These things are wandering randomly in a large space that is mostly maladaptive.

Catherine Caldwell-Harris's avatar

Not true that past eras had less supportive conditions. "lack of family support structures" -- In prior eras, people lived in extended families; your sister or neighbor (or anyone handy) nursed your infants at a whim; grandparents and aunts/uncles lived with you or next door; sibling caretaking was the norm; your 4 year old dauther could take your infant around strapped to her back. Almost every mother needs a lactation consultant today because we didn't grow up seeing every woman we know nurse all day long. Career demands -- far harder today; ancestral women could do chores with infants strapped to their backs or handoff a toddler to her other children; in premodern Europe infants were swaddled and hung on hooks when mothers shopped; one could go on and on with these examples; children as young as 3 roamed around in multi-age cohort bands (I even did so in 1969 living in Liberia); in the 1970s kids were out on the streets until Mom yelled out across a large city block, "Dinner"; professors in my dept say that in the 1980s they had playpens in their office to hold their toddlers; we can find innumerable examples of this easier parenting across every historical era.

Finances -- children were not costly in prior eras, they were economic assets due to agriculture. There was pressure to have more children, beyond what parents may have naturally desired, because of the economic benefits. In pre-agricultural times, children were also economic assets, assisting with whatever chores were needed but otherwise exploring on their own. Parents didn't do extensive parenting. They showed children how to asist as needed and otherwise, left them alone. The scholarship on this is deep and extensive. Innumerable factors involved in evolutionary mismatch make parenting more difficult today -- starting with the

'back to sleep mandate' and pediatricians forbidding co-sleeping-- yet Americans still want more children than they feel financially able to have.

Catherine Caldwell-Harris's avatar

Regarding your point "These can't be the main issue" --> agree there is a related issue: Opportunity costs. (Except this turns out to be the same as the financial point.) Women of earlier eras did not have a choice between motherhood and career. Worse, contemporary women aren't just faced with a choice between career OR children, they need a career *in order to have children.* Women want children, but must put career first, because of the financial demands of parenting, which did not exist in prior eras.

So women who want children must have a career first to have those children. In ancestral (evolutionary) times, you could get pregnant and raise your child without having to do much "work" other than what you would be doing anyway for your own subsistence. By the age when your child needed additional calories that would be a significant extra burden on you, that was the age when your child could contribute enough labor to compensate for some of those calories. Labor anthropologists have studied this, and note that other family members typically fill in the caloric gap (cooperative breeding and kin selection).

Some more material on this:

We can understand the modern fertility crisis when we look at the disruption of kin networks:

In Ancestral Times: The work of a child was distributed across a tribe. The cost was shared.

In Modern Times: We have privatized the cost of children. In a nuclear family living in a city, there are no grandmothers or siblings to provide free labor. The mother and father must provide 100% of the calories and care.

Extended Dependency: In the past, a 15-year-old might finally be breaking even on calories. Today, due to the educational arms race, a child may be a net consumer until age 25.

See also Ruth Mace's 2008 Science article, Reproducing in cities

Tim Tyler's avatar

There's a literature on the demographic transition. It isn't thought that lack of funds is the main issue. Rich, western countries are hit with sub-replacement fertility while multiple starving, and war-torn African nations still manage to turn out 4, 5 and 6 children per woman.

Catherine Caldwell-Harris's avatar

Access to modern birth control is essential for sub-replacement fertility. Sub-replacement fertility is not caused by *literal lack of funds* but by two things: (a) humans' innate preference for raising children when they deem themselves to have sufficient resources, and (b) the ability to *choose when to have children*, which was difficult before modern birth control. What goes into (a) -- This "deeming themselves to have sufficient resources" involves social comparison and the goal of having a certain desired standard of living. Take a look at Ruth Mace's article, Reproducing in Cities, and Jose Yong's article, When social status gets in the way of reproduction in modern settings. How does this literature strike you? Do they change your mind on this topic at all?

Kenton Krohlow's avatar

Shirtsleeves to shirtsleeves - people and societies (civilizations fall over centuries, don't they?) rise and fall. Unless you're the great one, not much one can do.

James M.'s avatar

We are discouraged from considering history, or civilizational decay. We are discouraged from discussing philosophy or ethics. We are encouraged (relentlessly) to focus on consumption and progressive orthodoxies. That is how we'll develop our civilization, according to our ruling class. Exactly how this is supposed to work is never really explained... but just go shopping. Scroll. Buy shit on Amazon. This is an information consumption regimen that is tailored to women and feminized men. Let's see how it work out.

https://jmpolemic.substack.com/p/were-getting-weaker-part-1

Dennywit Troubledoer's avatar

I have not found any cause for concern. What is it you hope to accomplish in perpetuity?

Are you rooting for a uniform?

Or for a losing roster?

Why would today’s respondents be representative of respondents in decline?

What moral weight would “blame” have in decline?

Does “civ” reduce suffering?

Can suffering be reduced?

Eméleos's avatar

Getting Haredim to become more modern is not a lost cause

matthew's avatar

If modernity is already failing, forcing Haredim into it does not prevent collapse. It converts a surviving subculture into a failing one and removes future optionality. Clean your own house first. Liquidating minority alternatives to delay decline is exactly the kind of peak era behavior that deserves blame.

Eméleos's avatar

There is a big difference between "modernity" and becoming more modern. We must try to avoid civilizational collapse, and I think getting more of the Haredim to accept science will not destroy Haredi culture and could have a huge positive effect.

matthew's avatar

In theory perhaps. In practice, most cultures that have attempted this have been pulled into the dominant trends around them. (Hanania’s post on Iranian women feels on point. This is not a gradual slope to traverse. It is a cliff.)

Some degree of discernment and decoupling is required, and most groups have struggled or failed at it. If a group can truly discern what to keep, that same capacity, subject to structural constraints, might instead allow it to help reform the failures of the majority culture. If it cannot, heavier hedging and separation make more sense by those who have longer term cares.

Most efforts to identify a stable middle path have either failed outright or remain unresolved with troubling indicators. The majority culture also has a clear self interest in encouraging absorption. Incorporating other peoples extends the timeline and pushes the costs into the future.

From this perspective, modernization is a high risk move. When promoted by the majoritorians functions as a con. The rewards are real, but for most cultures, flying this close to the sun has ended badly.

Eméleos's avatar

While obviously not as many as the Haredim, Modern Orthodox Jews in America have between 3 and 4 kids and Religious Zionists in Israel have between 4 and 5, and they both fully accept science. The Haredim will likely never become as modern as them (for better or for worse), even if they accept science. Their birthrates would possibly decline but I think it is very unlikely they have fewer than these other two groups.

Shadow Rebbe's avatar

Religious Zionists as a whole do not fully accept science. Modox in America much more, but I bet their fertility drops faster.

Surprisingly enough, you can reject all of science's fundamental assumptions and still be a software engineer.

Eméleos's avatar

What percent of Religious Zionists do u think fully accept science?

matthew's avatar

the concern is the trend 1-2 generations out, and if the majority group they are trending towards is reforming or just absorbing in an extractive fashion.

lots of other groups were in that situation a few generations back and no longer are.

Eméleos's avatar

To my understanding the religious Zionist are not declining in birth rates and are not reforming. What groups were in a similar situation?