Thread by @mmjukic

I think the real replacement fertility rate is not 2.1 kids per woman.

It's 5.1 kids.

A recent Swedish study found that in a generation born 1885-1899, an incredible 25% of people who had 2 kids had *zero* descendants by 2007!

For 1 kid? 50%.

A 🧵 on long-term fertility:

The 2.1 number seems intuitive and is taken as moral or life advice.

Two is good enough to sustain populations. More would dilute investment in each child or cause overpopulation.

But it is actually just a statistical artifact that varies considerably based on mortality.

Suppose you aren’t interested in playing your small part in statistically replenishing an entire population to the next generation, but rather interested in replenishing your own family dynasty or lineage over the long-term.

What’s the real replacement fertility rate then?

Early 20th century Sweden saw falling child mortality and avoided the World Wars. Yet a full 25% of parents with two kids still saw their lineages die out within a century.

This is replacement over the short-term, but doesn’t sound like replacement over the long-term.

According to the study, the probability of no descendants after ~120 years reaches near-zero not at 2 or even 3 kids, but rather at about *5 kids.*

So if you were an adult in early 20th century Sweden who wanted great-grandchildren, you should’ve aimed for five kids, not two.

How does a person with 2 kids in the early 20th century fail to have any grandchildren?

Ballparking it, looks like a roughly 30% chance of your kids dying before reproducing, plus a roughly 20% chance of childlessness without dying.

Thus a 50% chance for 1 kid, 25% for 2.

Importantly, most of the effect seems not to be poor hygiene causing infant mortality, but adult mortality and permanent childlessness.

Some traditionalists might be shocked to learn that it was normal throughout 20th century Europe for 15-25% of women to remain childless!

These numbers get much crazier if you do factor in child and young mortality:

If I'm reading this right, of all people born from 1885-1899, maybe about 57% had zero descendants by 2007.

In just over one long human lifetime, only a minority of people had any descendants at all!

Today, child mortality has fallen to negligible rates…

…but childlessness has been rising for decades: about 15-20% of post-reproductive age women in e.g. the U.S. or Germany are childless today.

Simultaneously, young Americans are increasingly dying deaths of despair.

In 30 years, it seems relatively likely that a child born today will live in a society with higher rates of adult mortality, later birth ages, and higher rates of voluntary or involuntary childlessness.

In other words, perhaps not too dissimilar from early 20th century Sweden.

If we take this Swedish study as a guide, then there is perhaps a 25% chance you will have zero descendants in a century even if you have two kids.

If you care about your lineage, you literally have a better chance of surviving Russian Roulette (16.67% chance of death).

You can control your own fertility. But you can’t control *your children’s,* let alone grandchildren’s.

In 2023, they may still die before reproducing or decide not to reproduce at all.

These in fact aren’t negligible chances, but uncomfortably large ones that pile up quickly.

Parents can do many things to increase the chance of kids having kids of their own, when it comes to upbringing, values, and care.

But statistically, perhaps the best thing to do is just have *more* kids.

If *you* don’t, then to continue the lineage *they’ll* need to.

Human demographics is not the story of well-adjusted normal people safely raising 2.1 kids who all go on to grow up comfortably and have 2.1 kids of their own, reproducing the species with perfect efficiency from generation to generation…

…rather up to half of people succumb to early death or childlessness, their deaths made up for by the rest who reproduce often far above 2.1 kids.

This is high churn; the ideal strategy is then not to be a fertility satisficer, but to be a fertility maximizer. Go for five!

If 5 kids is a 99% chance of descendants in 120 years even under harsh conditions, the interesting question is how many kids you need to nearly guarantee descendants in, say, 1000 years.

How far into the future do 10 kids get your lineage? Likely centuries longer than 2 kids...

The guardians and workhorses of the human species are high-fertility parents. It is the *additional* child who defeats death and grows population, not the first child.

And each child is a potential ancestor to hundreds or even millions of future people on a long enough timeline.

Some have asked if this changes based on class or wealth. The answer is yes, it does. Farmers were better off than "high status occupations," but everyone generally saw similarly high rates.

Further reading, now that I know this thread won't get crushed by the algo for outside links:

Original study:

On childless Europe:

On childless America by @lymanstoneky: sciencedirect.com/science/articl…
link.springer.com/chapter/10.100…
ifstudies.org/blog/the-rise-…