Thread by @lymanstoneky

Good question!

Mongolia's demography is absolutely fascinating and it's worth taking a closer look in regional comparison!

https://twitter.com/RichardHanania/status/1480890785918636037

To start with, this is not my first foray into Mongolia commentary. It's an amazing country I've tweeted about a lot because I'm lowkey obsessed with it. It's basically the Kentucky of Asia.

https://twitter.com/lymanstoneky/status/1350087827724390402?s=20

You can read some of my takes on the plight of Mongols in China here as well. The cultural commonality and difference between ethnic-Mongols in China and Mongolians is a fascinating divide without clear parallel anywhere in the world!

https://twitter.com/lymanstoneky/status/1301880906681069568?s=20

But okay, on to fertility!

What's going on with Mongolia's fertility rates?

On face value, they're way up. Here I use Mongolia's official vital stats instead of World Bank, because their vital registry system is generally regarded as being complete.

The first huge elephant in the room with Mongolian demography then is understanding the *pre-1990* fertility norm. Before the fall of Communism, Mongolian women were having like 4.5 children each.

Because the fall of communism created massive chaos everywhere....

.... my view is that the "correct" baseline for assessing post-Soviet fertility is something like "fertility in 1988." I've done some extensive analysis along this, specifically in reference to Russia's remarkable fertility rebound, here:

https://twitter.com/lymanstoneky/status/1467873840474771459?s=20

So by that standard, even at a TFR of 3, Mongolia has still experienced a massive decline in TFR.

You might think this is because Mongolia has had extremely rapid economic growth leading to "modernization," but that's not actually a super likely explanation.

Why?

Well, we can look at another post-Soviet Central Asian resource-rich country which managed to avoid civil war after Communism and had massive growth and high pre-1990s TFR: Kazakhstan.

Mongolia's real PPP per capita has risen about 127% since 1988. Kazakhstan's has risen about 88%. But Mongolia's fertility remains 40% below pre-1989 levels, while Kazakhstan's has returned to those levels.

MAYBE the difference between 88% real income growth and 127% real income growth explains that difference.... but I'm pretty skeptical, especially since Kazakhstan's total income level is a lot higher.

So what gives?

Well, without getting too much into the sociology of sedentarization, nomadism, and the fertility transition, I think we can just say that Mongolia underwent social transformations far beyond what is measured in PPP per capita.

Those transformations mean that Mongolia's TFR is actually one of the *lowest* in the post-Soviet world, as compared to its pre-1989 fertility rates, even though it is one of the highest absolute TFRs in the former Communist countries, or even in Asia!

Okay, so that's all interesting. But we can compare Mongolia across a lot more countries at once. Here, I compare Mongolia to every other Buddhist-majority country in the world + Kazakhstan and Kyrgyzstan, showing TFR vs. real PPP per capita.

What you will notice is that even controlling for real income levels, Mongolia had MUCH higher TFR than these comparable countries early in the data. And shockingly, Mongolia transitioned from 6 children per woman to 2 children with ZERO ECONOMIC GROWTH.

Folks this is nuts!

Mongolia experienced its entire fertility transition without *any* increase in standards of living measured in "consumption capacity" terms! It's an incredible example and really calls into question purely-economic theories of transition.

The "standard" economic-transition experience can be seen in countries like Sri Lanka, Bhutan, Thailand or Laos: as income rises, TFR falls. But Mongolia is wacky. And it's not JUST the post-Soviet period driving this: it was a nearly vertical gradient in the 1980s!

The other thing to notice is the fishhook-shape apparent in all the 3 post-Soviet countries: Kazakh, Kyrgyz, and Mongolia. They all saw TFR and income fall at the same time, then recover at the same time.

This results in a kind of wacky situation where to first approximation "economic modernization" is associated with "rising fertility" in these countries, though really what's going on is just return-to-normalcy.

Okay but back to Mongolia: before 1990, Mongolia had higher TFR for its income than any of these plausible comparison countries. This has been true once again since about 2005-210 sometime. Mongolia's TFR *is* unusually high for its income levels!

So @RichardHanania is not wrong to flag Mongolia for interest: it really is an oddity! A zero-economic-growth fertility transition, followed up by anomalously high TFR for its income.

Okay, so how do we explain it?

Well, @devarbol helpfully noted this recent paper, which decomposes Mongolian TFR into the "quantum" and "tempo" components. Basically "implied ultimate family size" and "spacing/timing of births."

Sorry, this paper:
researchgate.net/publication/34…

What the paper shows is a lot of the Mongolian trend is driven by tempo effects. During the 1990s and early 2000s Mongolian women paused childbearing, and since then they've resumed, without necessarily changing the mathematically implied total family size.

But while this is a lot of the story, it's not the whole story: since about 2009 there's a clear turn upwards in implied total family size as well, and from an already quite high baseline given Mongolia's income levels!

One thing always worth checking is, "How many children do women say they WANT to have?"

Here's results from the most recent survey I could ID for each country vs. 2019 TFR. Very high correlation!

But.... correlation may not mean anything! Luckily, we have some change-over-time in fertility preferences as well! Here's what we have for these countries:

Right of the bat, Mongolia is interesting. Mongolian fertility preferences are NOT high compared to historic comparison countries. They BECAME relatively high simply by being stable while preferences changed more in other countries.

We can do better than this. We can plug PPP per capita and fertility preferences into a panel model of these countries. Spoiler: in a panel model of *all* countries with this kind of data and other controls available, preferences are EXTREMELY predictive.

Looking at just these countries, changes in desired TFR are extremely predictive: when preferences change, TFR changes in pretty lockstep. PPP has a much smaller effect.

Of course, PPP and preferences are themselves correlated: higher income is associated with lower fertility preferences!

But where both variables are included, preferences are the dominant explanatory factor.

So, Mongolia has unusually high TFR compared to its income levels. Why? Most likely: because Mongolians tend to have fairly high fertility preferences. It's not common you get countries as well-off as Mongolia with desired family size of >3!

Pronatal policy could be at work as well on various margins. But the key point here is that Mongolian *culture* is more pro-natal.

So then the really interesting question becomes: why are Mongolian fertility preferences so high?

Here it's not clear. But what we can say is that Mongolia's fertility preferences were not *always* unusual. Here's preferences vs. income for these countries.

Compared to income, Mongolia's preferences *were* typical. They've *become* atypical as incomes rise.

So the distinctive trait of the Mongolian "demographic regime" is "fertility preferences which are not declining despite rising economic development." And that is very interesting and unusual! And it's the thing to explain.

"But Lyman, if it's all tempo, why do we care?"

1) Again, it's not all tempo
2) Even aside from the recent recovery, TFRs were high
3) Explaining why Mongolian tempo suddenly rocketed upwards is interesting too!

Finally, one of the better discussions of "deep cultural roots" of Mongolian fertility is here: demographic-research.org/volumes/vol33/…

It's like a greatest hits of weird-historic-demography: 44% of men being celibate monks! Endemic widespread venereal disease! 5-year-breastfeeding-period!