Tuesday, February 16, 2010

More on Russia's population trend

To follow up on Randy's discussion, here is a chart of Russia's age structure


It is fairly close to an inverted pyramid. And since the total fertility rate is currently below the replacement rate (1.4), the peak in cohort size between 20 and 30 yrs will show up as a smaller peak in the 10 and under cohort going forward. As the next chart shows, the population has stabilized temporarily. Since a sizeable portion of the population is now 50 and older, that group will shrink rapidly over time.



A rough calculation assuming a rate of population decline of .5% per year would result in a decrease per year of 700,000, and a decline in population to 128 million in twenty years’ time. Looking at the chart above, that would represent a serious decline.

19 comments:

akarlin said...

The rapid population fall you describe presupposes that both fertility rates and life expectancy remain stuck at the 1994-2006 "transition shock" levels, but there are grounds for believing that both are rather unlikely.

For instance, projecting a TFR of 1.5 (i.e., the same as in 2008, when it was 1.49, so discounting any further increases) and assuming a very modest degree of LE rise (74 years by 2025 - it is already close to 70), and 300k annual migration (currently around 200-250k), "the population size will remain basically stagnant, going from 142mn to 143mn by 2023 before slowly slipping down to 138mn by 2050".

Of course it is also entirely possible that Russia's LE will converge to developed-country levels quicker and that the TFR will stabilize at 1.7-1.8, in which case its population may grow back to around 150mn by 2025.

Scott said...

A very reasonable scenario. Increased longevity would be due to either lifestyle changes in the population or changes in health care.

CB said...

Mathematically, the rate of change with a 1.5 TFR is still -1,1% every year without migration. 1,8 is ca. -0,5% pa assuming a generation length of 30 years. shorter generations create faster rates of shrinking.

Randy McDonald said...

@ Sublime Oblivion: Are we talking about cohort fertility as opposed to period fertility?

Randy McDonald said...

Hmm. The problem with life expectancy increases, as I see it, isn't so much that the average life expectancy isn't low overall, more rather that while female lfie expectancies are generally comparable to those you'd find in the developed world male life expectancies are quite low. Moreover, this isn't a one-time shock so much as it is the fulfillment of a sustained trend.

I've noticed that elsewhere you've argued that relatively high desired fertility in Russia compared to Europe will necessarily lead to relatively high fertility compared to Europe. This isn't necessarily the case, since women in Spain and Italy want their completed fertility to be well above replacement levels even though actual fertility is very much lowest-low. What's more, after sustained periods of low fertility desired fertility also drops, as in German-speaking Europe. After a decade and a half of low fertility, is Russia going to be much different?

One factor that also needs to be raised, even though it's quite speculative, is that of migration. Will Russia remain a land marked by immigration, or will emigration pick up? I've blogged earlier about the beginnings of immigration by Russia from the Far east to Heilongjiang, and despite Russia's recent economic growth gaps in living standards are still noticeable.

Nice stuff, BTW.

john said...

Hard to see 300,000 net immigrants to Russia per annum over the next forty years. The ethnic Russian diaspora in Central Asia is pretty much tapped out already (OK, there's eastern Ukraine and Belarus, but - correct me if I'm wrong - those populations aren't emigrating in high numbers, and they're facing demographic collapse as well), even Central Asia mostly has low fertility, and I can't see Russia wanting or allowing such high levels of sustained 'non-white' immigration (yes, I do think Russia is rife with xenophobia).

Sure, Russian TFR could stabilize at 1.8, and sure, health indicators could improve. But those are big assumptions, the assumptions for immigration are even bigger, and, as Randy points out, there's always the possibility that Russia produces emigrants as well (esp. I would predict, to fill labor shortages in aging Eu countries...)

CB said...

They already do, in Germany there's a fast growing Russian community in addition to the Germans who came from Russia in the 90s. Especially in Berlin you can watch it.

akarlin said...

@Randy,

1. Only period fertility.

2. Re-desired fertility. I got the figures from this publication. My original line of reasoning:

"The West and Scandinavia tend to have reasonably healthy TFR's, ranging from 1.7 to 2.1, and on average desire to have 2.4-2.6 children. The Med and Visegrad countries have 1.3-1.4 children and desire 2.0-2.2 children. Although Germania has a TFR of 1.4, its desired number of children is the lowest in the region at 1.7-1.8. So on average although Europeans want about 2.1-2.3 children, their particular circumstances - frequently speculated to be excessive social obligations, high unemployment and perhaps subconscious forebodings of overpopulation - limit their fertility to a EU average of 1.4. In general, the greater the disparity between real fertility and desired fertility, the greater the perception that they have too few children and presumably, the greater the desire to close the "potential gap"."

PS. I had a look at the paper you cite and note that pp.13 basically confirms those figures. "Mean Personal Ideal and Expected Family Sizes of Younger European Women": UK / France / Sweden = 2.4-2.5; Italy / Spain = 2.0-2.1; Germany / Austria = 1.7. Since Russia's desired fertility according to opinion polls is around 2.5, it stands to reason that its expected real fertility should be at around 1.6-2.0 if we make a direct comparison-analogy to UK / Sweden / etc.

@John,

300k migrants per year is not a high level of migration compared to Russia's 142mn population. It is much lower in relative terms than for most West European nations. Second, there is no way that Central Asia "mostly has low fertility", unless your definition of normal fertility correlates with sub-Saharan African rates. AFAIK, they are all either above or well above replacement level TFR even today.

"Sure, Russian TFR could stabilize at 1.8, and sure, health indicators could improve. But those are big assumptions, the assumptions for immigration are even bigger, and, as Randy points out, there's always the possibility that Russia produces emigrants as well (esp. I would predict, to fill labor shortages in aging Eu countries...)"

I do not see how these are big assumptions. All that's needed for an essentially stable population for the next generation is the current level of migration, a slow rise in LE, and a TFR of 1.5 (that has already been exceeded as of 2009).

I think it is more reasonable for you to provide 1) a reason why Russia's LE should NOT increase in the face of a new anti-alcohol campaign and rising investment into healthcare facilities, 2) why Russians should emigrate in large numbers in the future, when A) emigration statistics have been falling since the early 2000's, B) large numbers didn't emigrate even during the very unstable and hyper-depressed 1990's, and C) even then the emigrants tended to be ethnic minorities like Jews and Volga Germans, many of whom have left by now anyway.

akarlin said...

PS.
Central Asia TFR's (World Bank 2007):
Kazakhstan - 2.36
Uzbekistan - 2.39
Kyrgyzstan - 2.7
Tajikistan - 3.45
And obviously they were a lot higher 20 years back, which is the figure that really matters since babies don't migrate, young and middle-aged adults do.

john said...

Sublime,

OK, TFR figures for central Asia vary according from source to source. The U.S. Census Bureau gives the following figures for 2010

Kaz. 1.87
Kyr. 2.64
Taj. 2.94
Tur. 2.19
Uzb. 1.92

Combined: 2.12

I see now that the U.N. world pop. prospects, 2008 revision, has 2005-10 TFR at

Kaz. 2.31
Kyr. 2.56
Taj. 3.45
Tur. 2.50
Uzb. 2.29

Combined, 2.5ish

i.e., close to your figures - which I assume were taken from the 2006 revision.

I have no guess which figures are superior. At any rate, 3 million emigrants per decade is a lot for this region (with about 55 million people) to produce. (though Afghanistan surely could!) And these immigrants would presumably concentrate in the biggest cities, esp. Moscow; I do have my doubts that Russians would be pleased to see Moscow as visibly multicultural as Toronto or Queens, New York.

I certainly hope that Russian public health improves! It's a big assumption though, because it's hard to give up the bottle, esp. when such destructive, constant use is virtually a cultural norm (it may be easier to affect the habits of today's children, but that won't show up in LE for a long time).

What I said was that TFR rising to 1.8 is a big assumption, which I think it is; I don't think any robust predictions can be made from surveys of desired fertility.

On assumptions for future im- and e- migration, I would say that surely much of Russian immigration has been from the Russian diaspora in Central Asia (which, as I said, can't continue), while my reason for positing _increasing_ emigration to the EU would be the progressive aging and shrinking of the labor force there, as well as catch-up growth in the (also aging) new EU states. This would indeed be a separate driver of emigration from the Jews and ethnic Germans who left in the 90's.

john said...

It should of course be said that much of this depends on relative economic growth in the EU, Russia, and Central Asia over the next few decades (much of which will be determined by energy prices, for sure). It would be foolish to feel too confident on this, I think.

Anonymous said...

What would be a reasonable guess for Russia's 2009 TFR?

I did not see the 2009 TFR statistics on the government site yet. The birth rate increase would seem to indicate that the number is above 1.5.

Also, the EU as whole almost certainly has higher than 1.4 average TFR. Only the lowest fertility European countries have lower than that number although some have around it.

Anonymous said...

John:

"I certainly hope that Russian public health improves! It's a big assumption though, because it's hard to give up the bottle, esp. when such destructive, constant use is virtually a cultural norm (it may be easier to affect the habits of today's children, but that won't show up in LE for a long time)."

It is not such a huge assumption when you take into account the recent improvements in life expectancy/death rates coupled with the fact that deaths by alcohol poisoning have halved over the past five years (including a 20% decline in 2009).

You're arguing against trends that are plainly evident and logical with vague theorizing of your own. Yes, giving up the bottle is hard, but it's certainly not impossible, nor is repairing a collapsed society as a whole, which is the basis of the demographic crisis (along with years of Soviet oppression of course). Wounds take time to heal, I've been arguing this for years; and this certainly isn't the biggest wound Russia has ever had to heal from.

john said...

Anonymous (Sublime?),

No, you're right, it's not impossible. And I have no problem admitting I don't know the statistics (though deaths by alcohol poisoning must be quite a low share of the early mortality caused by drink, and may not be a good proxy for the larger problem). If they have indeed been improving so dramatically that is certainly an important data point. My point was precisely that if there is no such data, predicting improvements involves a 'big assumption.'

You are certainle right that cultures can turn themselves around - witness, say, the enormous improvement in life in New York over the last twenty years. Of course these things happen - though if you had predicted NY's turnaround twenty years ago, I would have called it unwarranted, and I would have been right to!

Bob said...

John:

"Anonymous (Sublime?)"
No, just someone who doesn't have a blogger account (though I do follow his blog). I'll go by Bob.

"(though deaths by alcohol poisoning must be quite a low share of the early mortality caused by drink, and may not be a good proxy for the larger problem)."

I think deaths by alcohol poisoning is a perfect proxy to gauge general progress. Why wouldn't it be? Alcoholism effects many mortality statistics like homicide, suicide, traffic accidents etc (all of which have seen equally dramatic improvements in recent years; coincidence?), but none are as directly related. Heart failure accounts for the majority of deaths related to alcohol abuse by far, but that's far too general an indicator if just looking at it from the surface.


" If they have indeed been improving so dramatically that is certainly an important data point. My point was precisely that if there is no such data, predicting improvements involves a 'big assumption.'"

Indeed, if there were no such data there wouldn't be much basis for my argument. But there is, which is why I'm so bemused by these arguments that the recovery will magically stop and reverse without any logical/statistical basis of their own.

Randy McDonald said...

@ Sublime:

Is Russia directly comparable? The key difference explaining the difference between relatively high fertility in northwestern Europe versus southern and central Europe is the toleration/acceptance of people in the former regions of alternate families structures (cohabitation in place of marriage, day care and socialized childrearing, et cetera). Is Russia closer to northwestern Europe or central Europe in this regard?

Interesting stuff.

Scott said...

According to the Russian Health and Social Development Ministry, the average Russian drinks 18 liters of pure alcohol per year. This number is an average calculated from the consumption of all Russians, children and infants included, so the actual amount consumed by drinking adults is much higher than this. If the Russians were able to slash their drinking levels in half they would still be drinking in what The World Health Organization considers, the dangerous range.

Only 2 countries in the world, Moldova and Reunion, infamously boast higher liquor consumption, and when you account for the consumption of homemade alcohol, Russia comes in at number 2, instead of number 3.

This might make increasing life expectancy in Russia a bit difficult.

akarlin said...

@Scott,

To the contrary, I think this will make increasing LE easy because big reductions in alcohol consumption are possible with the right policies (see Finland-Karelia) and translate into immediate, substantial reductions in mortality.

Unknown said...

Nicesharing Heating and cooling Toronto