Monday, May 28, 2007

Childbearing in Europe

As I mentioned in this post, the Population Association of America annual meeting 2007 played host to quite a large number of very useful and enlightening papers. From time to time I will try to highlight some of the more interesting ones among them here on Demography Matters.

In this vein, I would like to touch in this post on some of the questions raised in the paper Childbearing Trends and Policies in Europe: Is a new Demographic Disequilibrium Emerging? presented by Tomas Frejka, Jan Hoem, Tomáš Sobotka and Laurent Toulemon.

As the authors explain the paper is in the form of a report on "work in progress" and presents selected findings and conclusions from an international comparative research project which started in October 2005 - sponsored by the Max Planck Institute for Demographic Research - and involves the study of 18 countries: Albania, Austria, Bulgaria, Czech Republic, France, Germany, Hungary, Italy, Lithuania, Netherlands, Poland, Romania, Russian Federation, Slovak Republic, Slovenia, Spain, Sweden and Ukraine. As such the findings of the research - which will be finally presented in book form - represent an important source of data and analysis.

Now the authors define their project in the following rather radical terms:

In 1933 Landry concluded in his theory of the la révolution démographique that in a third pattern “there is no longer an equilibrium.” In 2001 Chesnais reminded us that Landry “envisaged a scenario of ‘permanent disequilibrium’ ”… and stated that “(T)here are strong arguments in favor of the eventual globalization of the birth deficit” … in which “(T)he long-term downward trend seems irreversible.” In 2003 Van de Kaa reflected that “(P)erhaps it is now time for someone to start thinking about writing a … volume entitled: ‘The Postmodern Decline of Population’. For Europe’s demographic future appears to be a thing of the past.” This volume is the initial attempt to write such a book. It is an assessment of whether humanity is facing a new demographic disequilibrium.

Now talk of an "irreversible" demographic disequilibrium may seem like strong language, but a long hard look at the available evidence may make the use of such terminology not so extreme as it seems at first sight. In most advanced countries fertility has long dropped below replacement rate, and in many cases well below this level, and trends in the developing world now give clear signs that this pattern may well continue to repeat itself. Obviously the idea of a planet with a somewhat smaller population level is not necessarily a bad thing in and of itself, but this idea of "disequilibrium" is perhaps rather more disturbing, since it does raise the question of just how, and when, the present trend may be reversed.

Turning to the European context, as the paper states:

a) With the possible exception of Albania there was not a single European country in 2004 with fertility at or above the replacement level (Council of Europe 2006).

b) As of 2004, 17 out of 38 European countries with over 100,000 inhabitants, 45 percent, had negative rates of population growth. This is a very recent phenomenon. In 1990 there were only three such countries. In some countries the negative rate of natural increase is offset by immigration. Nevertheless, 14 of the 38 countries had negative rates of population growth.

c) In 2002, 16 out of 39 European countries recorded period total fertility rates (TFR) below 1.3 and 25 countries recorded period TFRs below 1.5. The ‘lowest-low’ fertility spread throughout Europe affecting countries with more than one half of the European population by 2002, up from nil in the early 1990s. This proportion subsequently declined as the TFR in several large countries reverted above the threshold of 1.3. The spread of such low period TFR levels has been closely linked to the postponement of childbearing taking place in all regions of Europe. Almost three quarters of Europeans currently live in societies with the TFR below 1.5.

The key point to grasp is perhaps that the principle driver of the lowest-low reading which is being registered on the TFRs has been the phenomenon known as birth postponement. This impact is a result of the fact that during a long period of time - as average first birth ages move steadily upwards - comparatively few children are actually born, and this gives rise to a rather distorted TFR reading, since as some of the women who have remained childless during their 20s move up the age ranges they may (and do) start to have children - the so-called recovered births - and hence a certain recovery in final fertility rates, or completed cohort fertility rates, may occur. (The reasoning which lies behind such thinking is explained further in this post here)

Now there is one proviso to be made at this point, and that is, as Claus and I have been recently explaining (and here)the real possibility exists that some of these countries with lowest-low fertility are caught in some kind of low fertility trap. This may not be a permanent trap, but it is a process which could last for several decades or even generations, and if this were to prove to be the case the issue would be a fairly serious one since the long term structural impact on national demography (and probably economic growth and living standards) would be important. So even if the "recovered births" may finally make their presence felt, if this were only to happen in several decades time, the population structure (and of course the reproductive generational base) would be seriously affected, and with long lasting consequences. As they say, demographic processes tend to cast a long shadow.

Now it is also important to bear in mind that there is below replacement fertility and "well-below" replacement fertility, so while it may be thoroughly unrealistic to think in terms of a large scale and generalized return to the magic number of 2.1TFR, fertility in the 1.7 -1.9 TFR range might be a much more manageable affair for our societies than ongoing fertility in the 1.3 - 1.5 TFR range. As the authors of the paper note:

There is a general consensus among scholars that “maintaining fertility at a level that does not fall much below a two-child average – say, around 1.7 – 1.8” (Demeny and McNicoll 2006:281) could be sustainable for a prolonged period.

So the tricky question is, just how can we nudge fertility up in those societies who might be deemed to be stuck in or around the lowest-low range?

Just how extended this lowest-low fertility issue is in Europe right now (although the problem is by no means a European one, since several Asian countries - including possibly China - may well already be affected) can be seen from the following charts presented in the paper (incidentally for a better view of the charts simply click over them).

Chart 1: Number of European countries with the period TFR below 1.7, 1.5 and 1.3
of 39 countries with population above 100,000 in 2006)

Chart 2: Proportion of Europeans living in countries with the period TFR below 1.7, 1.5 and 1.3

Now as we are noting, the core of the problem seems often to be associated with a strong postponement process and a very rapid rise in mean ages at first childbirth. The next two charts should make this clearer in a European setting.

Chart 3: Mean age of women at first childbirth in selected countries and regions of
Europe, 1960-2004
(arithmetic averages)

Chart 4: Proportion of childbearing realized by women below age 25 in selected countries and regions of Europe, 1960-2004 (arithmetic averages)

Now the core of the "recovered births" argument focuses on the idea that the artificially low reading being registered on the period fertility TFRs is masking a much healthier situation in terms of the longer run cohort fertility rates. However, if we look at the long run trends in CFRs (see charts below) we will see that matters are not quite this simple.

Chart 5 - Total cohort fertility rates, selected WestEuropean countries, birth cohorts 1915-1970

Chart 6 - Total cohort fertility rates, selected Central and East European countries, birth cohorts 1924-1973

In fact what we can see, is that in general terms long run European CFRs are in continuing decline, and the only clear examples of a change of course come from societies - like Sweden and France - with a long tradition of systematic (and relatively expensive) pro-natalist policy. As the authors note:

"Even though it is not known to what degree childbearing will be recuperated, it can be surmised but not proven conclusively that quantum declines are a part of the trends of fertility behaviour of women in the midst of their childbearing in a number, possibly a majority, of European countries."

And here perhaps precisely we hit the problem: it can be surmised but not proven conclusively. And in this gap between what is surmised and what is proven lies a space which allows public policy - at least in an EU context - to languish in neglect, since after all, if it isn't conclusively proved, what is the justification for priority action.

Well, I come back to the quote I offered at the start of this post, European populations in many cases are in "demographic disequilibrium". At this point it is not clear whether or not this process is "reversible", but - and especially given the fact that, as Claus and I keep arguing, some of the economic consequences of this disequilibrium are already being noted at the level of economic processes and imbalances - doesn't it seem rather foolhardy to wait until we know for sure (and especially if in the end, by the time we are able to decide, they should prove to be non-reversible) before we really "take the bull by the horns". Put another way, is it not now time to act, and decisively?


Robert said...

Assumably, demographic decline is reversible, because past episodes of demographic decline have been reversed.

I'd like to throw out a hypothesis, as a counter-hypothesis to mechanism 2 in your previous post. If ideational family size is informed by chiefly by the family size that a woman grew up in, then yes, once ideational family size falls below replacement, this mechanism can only permit demographic recovery if something happens to cause women to have more children than they think is ideal.

But what if, on the margin, the decision by a woman of childbearing age to have a child, or to have another child, is informed more by the number of young children her peers have? There are economies of scale in child-rearing. Families are more willing to look after each others kids' for an evening if the families are of similar size and age ... school funding comes easily to a democracy, if many voters have children of school age.

This hypothesis produces a similar self-reinforcing long-term decline in family size, once the density of young children around becomes low enough that women notice and reconsider their own plans to have a(nother) child. But it also has some explanatory power for the post-war baby boom in both the U.S. and Europe. The children of returned soldiers were only a fraction of the boom generation. But the episode of household formation, and re-formation, that occurred with the peace produced an episode of concerted reproduction, of families making up for lost time by having perhaps the same number of children they would have had otherwise, but having them closer together. The resulting super-normal density of young children informed the reproductive decisions of other women, even those whose reproductive partners were not taken away by the war, in a self-reinforcing period of increasing fertility.

If this mechanism is right, then women belonging to larger birth cohorts are themselves more likely to have more children, since all other things being equal, you are more likely to have more peers with young children if you have more peers. This is testable from historical CFR records, though I have none in front of me at moment as I think of this. If true, then it says that for a society in demographic decline, the window of opportunity for reversing it happens when a largish cohort of women reaches childbearing age. Some shock to the system is still needed to encourage these women to actually have children, and have them at the same time, to set off a period of self-reinforcing fertility increase, but this is easiest when there are many potential new mothers around.

Unknown said...

Robert, your theory is testable right now in Japan. Japan had a relatively short post war baby boom that lasted only until the mid 50's. The largest age cohort in Japan is the 55-60 set that are just starting to retire. The next largest are their children. The 30-34 and 35-39 cohorts are much larger than those that come after them. (And larger even than the ones that came before.) Most Japanese women are now putting off childbearing until their 30's. This would seem to open up a window of opportunity for Japan to improve their situation. Given current tfr numbers it looks unlikely.

Edward Hugh said...

Hi to both of you,

"Assumably, demographic decline is reversible, because past episodes of demographic decline have been reversed."

Hmmm. I'm not at all sure about this. I have a line drawn in the sand - Alamo style - between the pre- and post-industrial revolution processes. Before the IR there were homeostatic mechanisms at work (the so-called carrying-load) which could serve to restore depleted populations. After the IR I just don't see how this would work or what would drive it - which doesn't mean that birth rates couldn't rise again, just that there is no clear mechanism in place.

The turning point would be the dis-connect between living standards and fertility. The sign has changed here, at least for the time being. As living standards have gone up fertility has gone down.

Bottom line: no guarantees.

Robert, you pose an interesting mechanism, but we really need to see some evidence of *rates* going up on a sustained basis after large cohorts to test something like this.

Also, it all depends on what weighting to give "peer group" pressure factors here. Same issue, of course, with Lutz's trap hypothesis.

"If ideational family size is informed by chiefly by the family size that a woman grew up in, then yes, once ideational family size falls below replacement"

Yes but many women all over the planet are currently having - even when the timing ingredient is stripped out - far less children on average than their ideal, so there are other factors at play.

"The children of returned soldiers were only a fraction of the boom generation."

Yep. This can't be the only explanation for the baby boom. But the post war years did come after a decade or more of economic hardship, so presumeably there was also some natural birth recovery from all this. I think the cohort thesis works at this level, when dealing with temporary factors.

"this is testable from historical CFR records, though I have none in front of me at moment"

Yep, but if we look at the CFR charts shown in the post, then they do trend down and down, and ominously so.


"This would seem to open up a window of opportunity for Japan to improve their situation. Given current tfr numbers it looks unlikely."

Yep. I tend to agree.