Demography And Leaping To Conclusions

But first, a note about those two articles published here this morning: As I hope many (if not all) of you guessed, it was a scheduling accident. In particular, as regards the piece “Russia’s Economy Is Now Europe’s Largest,” this is what I expected to see once the World Bank released its PPP-adjusted GNI figures for 2012 (it always does this in April, but for whatever reason it has been late this year). Russia is already very close to Germany as of 2011, and due to a planned harmonization of its GDP counting methodology with international standards – i.e. the introduction of imputed rent – it is projected to automatically increase by 5% in addition to its normal growth. Hence the title: It’s something that’s likely to happen, hence I wrote that title with “see data” for text and scheduled it for publication on April 30th, on the assumption that the World Bank will have released its new data by then and I’d have time to write the actual post. But the World Bank dallied and I forgot about the scheduled publication date.

Now, onto the demography. In contrast to the first 3 months of 2012, when both mortality and birth rates saw big improvements, they have now gone into reverse for the same period in 2013. Namely, births have decreased by 0.8%, deaths have also risen by 0.8%, and the rate of natural decrease widened from 35,000 in 2012 to 43,000 in 2013. Mark Adomanis notes that this is a “pretty worrying development,” with “2013 is shaping up to be the year in which Russia’s streak of improving demography comes to an end.”

That is true enough if current trends continue. The operative word being “if.” But as I cautioned in my last post, extrapolating from a month or even a few months is a risky proposition, given the sharp seasonal swings in mortality. This is particularly true for extrapolations from winter months, in which there is an additional strong independent factor in the form of the influenza/pneumonia situation (which, on average, is worse during colder winters) and climatic factors (e.g. all else equal, Russians tend to drink more during colder winters). Now we know that this year’s has been exceptionally severe in Russia, so let’s look at the detailed breakdown for causes of mortality: -3% for infectious diseases, -1% for cancers and heart/CVD diseases, +18% for pulmonary diseases (e.g. pneumonia), no change in digestive tract diseases, -5% for deaths from external causes (but a +1% rise in alcohol poisonings), and +14% in deaths from sundry causes.

Now on the surface this now looks quite a lot better. Despite a harsh winter, deaths from heart disease (the biggest killer in Russia) continued to decline, as did deaths from external causes (which disproportionately affect the young and thus have an especially negative impact on life expectancy). The additional 3,000 increase in deaths from pulmonary diseases will have mostly accrued to elderly deaths from pneumonia, most likely due to the season swings typical of its epidemiology. The major, overriding question is this: What are “deaths from sundry causes” (прочие причины смерти)? I don’t know. But we can speculate. In old age, it is frequently unclear which of a panoply of ailments finally do someone in. And harsh winters are associated with mortality rises, especially among the elderly. Perhaps a large part of that 7,000 rise in deaths from “sundry causes” were simply a result of more elderly dying due to general winter causes and not being precisely classified.

In any case, I submit that three months is too short a time period to make any meaningful conclusions as to the final trajectory for the year. Again, I refer to Sergey Zhuravlev’s graph as a graphic demonstration of why that is so. Russia’s improvements in mortality are not a steady process, to the contrary they look like a series of intermittent sharp declines followed by steady periods of up to a year’s duration.

Then there is the decline in birth rates. To be fair, it is increasingly unrealistic to expect further rises in crude birth rates, because the “echo effect” is real (if often overstated, because of a failure to adjust for birth postponement/rising age of mother at childbirth). Russia’s total fertility rate started plummeting around 1991; the girls born then are now in their early 20’s. The average age of the mother at childbirth is now about 27 and rising, and up by 2 years since the 1990’s. Nonetheless, despite that counteracting effect, the fact is that the demographic “chasm” of the 1990’s continues to gain on women at their peak fertility (even if the age of peak fertility continues to increase) and it is a deep chasm, with women of the age of 5-19 making up just 61% of women of the age of 20-34 in 2012. So as there will be accumulating downwards pressure from changes in the age structure until the late 2020’s/early 2030’s we can now expect crude birth rates to start consistently falling from year to year.

  • Dear Anatoly,

    Though I am no expert on these questions I am fairly sure that the long, cold winter is probably what caused what looks to me like a blip. The point is that the underlying causes of Russia’s high mortality have been continuing to improve. Putting it I am afraid rather crudely I suspect that more people (especially old people) died from the cold.

    I suspect by the way that if the anti smoking campaign takes hold over time that will make a big difference though presumably that is not coming through in the figures yet..

    Incidentally didn’t a very hot summer a few years ago also lead to an upward blip in mortality figures or am I remembering this wrongly or perhaps this is not comparable?

    • AK

      No, Alex, you are entirely correct. Excess mortality for the 2010 heatwave was 56,000 as I recall. This is not solely a “poor Russia” problem either: There were 20,000 excess deaths in France and Italy each during the 2003 heatwave (I was unfortunate enough to be in Paris at that time so I remember it well).

      This is the reason that one really has to wait for more than a few months before beginning to speak of a trajectory. Had Adomanis, say, used July-August 2010 as his base, he might have concluded Russia was headed back to the mortality profile of the 90’s and early 2000’s.

  • moscowexile

    re: the wrongly scheduled article on Russia reaching №1 position in the European economies

    I did a check on the available data after reading a comment demanding that AK give links to data supporting his case. For what it’s worth, I agree with AK’s oncoming announcement: the figures available now indicate that, much to the ridicule of many, no doubt, the Evil Empire might well be heading for the top position.

    I was going to post what data I had gleaned in reply to the commentator’s demand that such data be presented, but the article had been removed by then.

  • Sergey

    In fertility, the “good” news is that the seasonal pattern now looks like the one from 2008 and earlier, when February/March daily births were not much larger or smaller than in January. In the last couple years, February and March became much more fertile than January.

    If we are just observing the seasonal pattern shift, chances are total births for 2013 won’t be much different than in 2012. After all, there were 89 days in 2013Q1 versus 90 days in 2012Q1, or about 1.2% less. A much more sinister alternative is that this is a start of a negative wave, a counterpart to the positive bulge that started in 2011H2 and is now at its end. Another quarter of data might clarify what’s going on.

    With a leap year adjustment, deaths are actually very high this year. One positive is that after disastrous start (January deaths worse than in 2010-12), both February and March were the best months in recent record in terms of deaths per day statistics. If the summer is OK, we might still get overall drop of total deaths in 2013.

    What is unambiguously good is that infant mortality has started to fall already: 8.3 promille vs 8.4 last year. In 2012, the switch to the wider definition of birth didn’t happen on Jan 1 in all regions, and the infant mortality was still under-counted somewhat (IIRC, from Apr 1, 2012 all regions switched to the new definition). Whatever happens to the total deaths, infant mortality will be lower this year.

    • AK

      After all, there were 89 days in 2013Q1 versus 90 days in 2012Q1, or about 1.2% less.

      Good point. So in adjusted terms, the birth rate has actually remained virtually the same from 2012. Though by the same token, the death rate increased more significantly…

      Incidentally, do you have any idea about what constitutes “other causes”? I had a guess at it in the post, but it’s really just that – a guess.

      • moscowexile

        I should think that “death by other cause” is simply a death whose cause is not of high enough recordable frequency to be included in a list such as this.

        For example, my great-grandmother died as the result of bleeding to death in front of an open fire whilst asleep. She was very old, I was told, and in her sleep she had apparently cut her ankle with the sharp edge of a clog-iron. Clogs were the usual footwear for the working classes in my old neck of the woods.

        No doubt, on her death certificate cause of death was given as “exsanguination”; I doubt that her cause of death was given as “bled to death after cutting leg on clog-iron”.

        In any case, I should imagine that in any statistic alanalysis, my great-grandmother’s cause of death would be classified as “other”.

      • Sergey

        Demographic Annual for 2010,, is probably the quickest source. From which we learn that out of about 2 mil total deaths in 2009, causes of death that aren’t usually detailed in demographic reports were responsible for about 141.1 thousand. The causes with >1,000 deaths are listed, roughly rounded to avoid cluttering:

        diseases of urogenital systems: 11.5 thousand

        endocrine, nutritional and metabolic diseases (mostly diabetes): 10.5

        mental and behavioral disorders (most alcohol-related): 6.1

        diseases of osteomuscular systems and connective tissues: 1.9

        diseases of the blood and blood-forming organs and certain disorders involving the immune mechanism (anemia the largest killer): 1.2

        diseases of skin and subcutaneous tissue: 1.8

        symptoms, signs and abnormal clinical and laboratory findings, not elsewhere classified (55% of that is senility): 92.3

        Overall, about 37% of “other reasons” deaths are deaths from old age, most of unclassified ones are probably old age deaths as well.

  • Latyshka

    The number of fertile women will decrease in the whole of Eastern Europe, due to the 1990’s fall in births, today the girls born in 1980s (when there was a baby boom) are still having kids. In about 5-10 years, the real problems will start all over EE. It will be interesting to see how that will affect social trends, labour, family formation, etc. Seems like a really serious trend on a big scale. However, we can hope that the fertility quotient will rise (even if the actual number of births will be smaller). If Russia could push the fertility rate from 1.6-1.7 to 1.8, that would be nice.

    Btw, according to the latest polls, 77% of Russians consider themselves subjectively happy – seems like a rise from 2011, when it was 67%:

  • Yes it appears that Germany and Russia like two great magnets will draw the best and brightest from (in the former country): Poland, Hungary, Italy and Spain (and in the latter thanks to common religion/alphabet): Greece, Bulgaria. Romania and Baltics might split since older Balts in late 30s or 40s still speak Russian but younger generations will go West in what the American geo-politician Joel Kotkin calls the ‘New Hansa’ — Germany, Netherlands, Baltics and Nordic countries. So despite all London bankster schemes like Cyprus to get Germany and Russia again hostile towards eachother they are going to get together with China and form a new Eurasian economic and currency bloc.

    This is why the WSJ’s latest is rubbish — Russian demography does not depend solely on the generational cohort born in then USSR during mid to late 1980s. And did you notice they had to publish the piece on Victory Day? That’s just smarmy.