Friday, April 13, 2007

Evolution of Fertility in the OECD 1970-2004

As Edward noted just below we have been taking a long break here at DM as of late. However, we have not been sitting idle and in this entry I would like to present the stylized facts of fertility trends in the OECD. Essentially, this entry is a re-post of an entry over at Alpha.Sources which is also, as it were, very relevant to DM. The data are taken from the recent 2007 OECD Factbook and more specifically this raw OECD excel data sheet. I realize that the tables are a bit messy in terms of distinguishing between the countries but I hope it is useful anyway. You can always consult this excel file which contains the graphs in plain design as well as graphs for individual countries.

The review begins with the overall evolution of fertility in the OECD; enjoy!

You do not get a much better picture of the demographic transition in the developed world than this I imagine but as we shall see there are notable differences between countries and regions which of course make the macroeconomic analysis and conceptualization more difficult. Nobody of course knows how far fertility can and will drift down but there does seem to be some kind of stabilization at 1.6 children per woman aged 15-49 in the aggregate OECD figure from 2000 and onwards. Note of course also that this figure does not take into account the different sizes of the countries and thus different contribution to the average population level in the OECD.

Moving back to the TFR rates however there are as mentioned differences and they are very important to pay attention to. Let us first look at the Anglo-Saxon (English speaking) countries which are doing a good job keeping the OECD aggregate TFR level afloat.

Here of course the TFR rate in the US just about at replacement levels should be noted. The convergence process from the 1970s and onwards is also interesting to note and especially that this convergence process coincides with notable divergence at the end point. Lastly, the convergence process itself shows that countries move through the demographic transition in quite different tempi and also as noted with different end points. Next, let us have a look at continental Europe where the overall convergence process is even less marked and thus also hides important divergences; note for example the difference between France and Germany. France, Belgium and the Netherlands are all above the 1.5 level with France close to 2.0. Conversely, Germany, Austria and Switzerland have converged close to the dreaded level of 1.3 which signifies lowest-low fertility (i.e. the fertility trap).

Moving on to southern Europe we see a distinct process of convergence towards lowest-low fertility. In terms of macroeconomic fundamentals immigration is very important to take into account here especially in Spain where the pickup in fertility rates from 1998 and onwards perhaps can be attributed to a positive fertility component from immigration. However overall, the fertility situation looks fairly dire in the southern parts of Europe.

And while we are in Europe why don't we take a look at Eastern Europe. The picture is of course incomplete but given the general level of development and their place in the global value chain the general fertility levels in this region should be the cause of much worry amongst the policymakers. It is thus most worrying to see these countries move through the demographic transition at a pace which is clearly not in sync with the pace of economic development.

On the second last graph we have the Nordic countries where fertility levels are generally higher than in the rest of Europe and close to replacement leves albeit somewhat (1.7+) lower for all countries except Iceland. Perhaps most interestingly the evolution in fertility is considerably more volatile than in other OECD countries. Note for example Denmark's pick-up from 1980s to the current level of TFR at about 1.8 and also Sweden has seen a pick-up from 1998.

Lastly, I am featuring a graph of a comparison I have often made between the three oldest countries in the world (measured by median age), Germany, Italy and Japan. As we can see from the graph all three countries are firmly set with fertility levels below 1.5 and close to lowest low fertility of 1.3.


Fertility levels can tell us a lot but by no means everything. Especially, the net migration rate is an important input to the full comprehension of the overall population dynamics in the individual countries. Moreover, the way fertility is treated above also does not take into account to cohort perspective. As such, it is important task to divide the fertility levels into more specific periods to really get a full picture. For example the periods 1970-80, 1980-1990, and 1990-2004 reveal notable differences in average fertility levels and as such y-o-y fluctuations in fertility rates are not very useful in themselves. I have begun the calculations on these periods in the .excel document linked to above. In the end however I hope that you have found this useful anyways and pleaes do feel free to use the data and graphs fore further study. More specifically I recommend you to download the .excel file and have a look at the data set. You never know when a paper comes along which demands the use of one of the fertility tables above. Of course you will also be able to customize the table design as you see fit from within the .excel file linked above.


Assistant Village Idiot said...

Months ago I noted over at neo-neocon that the aggressor nations of WWII seemed to have a lower birthrate than the victim and neutral nations, and that the Cold War aggressors were showing the same pattern. That seems to still be the case.

Admittedly, in some cases it is hard to make a sharp distinction between aggressor/victim/neutral.

It does lead to speculation that collective feelings of guilt might decrease a culture's willingness to reproduce itself (which on an individual level, was the point of neo's post).

Scott said...

Assistant village idiot, you raise a very interesting point. However, having lived in Japan for two years, and studied the country intensely during my undergraduate studies, I can categorically say that guilt over WW II has nothing to do with the changes in Japanese families' attitudes towards having children. The decline in recent years in the Japanese fertility rate has to do among other things, with young Japanese women delaying marriage due to the low status wives have in Japanese households; which tend to be extended families with the mother in law living with the eldest son.

CV said...

interesting stuff this ...

I don't know what to in terms of AVI' s argument as I simply don't know anything about this potential feedback loop with a decrease in fertility.

I am of course thinking about how to potentially operationalize this for research purposes.

Regarding Scott's comments I think they are true since the drop in Japanese fertility indeed is said to have come as a result of a delay in marriage among other things of course.

Anonymous said...

Please in the future, make all the axes of the graph the same for easy comparison. Right now, the y axis keeps varying and it makes it harder to compare things.

CV said...

Good point ...

Will make sure to do that for the future!

Anonymous said...

It is not so much that they were the aggressors as the vanquished.

S.M. Stirling said...

I suspect the national aggregates are often misleading. For a country of substantial size, breaking it down internally -- by income, ethnic group, region, etc. -- may be important.

S.M. Stirling said...

Some very recent developments may be important.

For example, Poland has apparently "lost" something like 1,000,000 people through migration to Western Europe in the past 4 years, almost all of them in the 18-40 range and fairly evenly split by gender. The figure is imprecise because it's hard to detect returnees and multiple 'emigrations' by individuals, but it has to be something in that range.

This has to affect the birth-rate, both directly and by postponing marriage among emigrants who plan to return.

The East European influx into Iberia, the British Isles and (to a lesser degree) Scandinavia is assuming demographically significant size with startling speed. Their numbers in Britain are now approaching half those of the whole "New Commonwealth" immigration and its descendants, after only 3 or 4 years.

S.M. Stirling said...

With respect to Canada, you get a figure closer to the other English-speaking countries if you subtract Quebec.

In fact, Canada becomes a lot more like the (northern) US with Quebec out of the equation.

Edward Hugh said...

Hi Sterling,

and welcome back again.

"For example, Poland has apparently "lost" something like 1,000,000 people through migration to Western Europe in the past 4 years, almost all of them in the 18-40 range and fairly evenly split by gender."

I agree that the out-migration patterns from some Eastern and Central European countries against a background of already very low fertility is an important development, of uncertain long term consequences (but basically not good). Societies with such low underlying fertility have never before seen this kind of exodus.

Clearly this situation can to some extent be offset by in-migration from points further east although the rate at which this occurs will very much depend on the rate of economic growth in the would-be receiving countries, and this in turn may depend in part at least on internal market conditions (everyone can't live mainly from exports) which may, of course, be influenced by the hole that has now been created in some important age groups.

Also, as Claus has been indicating, there are the human capital implications of these flows. So really there are a lot of unknowns out there, but the situation is going to be complicated and very tricky.

"Their numbers in Britain are now approaching half those of the whole "New Commonwealth" immigration and its descendants, after only 3 or 4 years."

Without doubting the significance of what you are saying I am not sure about this last point, since (I think I am right in saying) the UK only keeps statistics on "immigrants", and the majority of the descendents of "New Commonwealth" immigrants are not themselves immigrants but British born, so they are not reflected in the numbers. This is one big difference between the data quality we have in this area from the US and Europe.

If we are simply talking about immigrants Spain may well be the global society with the highest proportion of immigrants per se (due to the comparative recency of the phenomenon, and having received close on 6 million people in 8 years), although the population stock of migrants and their descendants is much greater in France and the UK (and obviously the US).

S.M. Stirling said...

"the UK only keeps statistics on "immigrants", and the majority of the descendents of "New Commonwealth" immigrants are not themselves immigrants but British born, so they are not reflected in the numbers."

-- there are reliable estimates available; the total number is a between 2.5-3 million, if I recall correctly.

S.M. Stirling said...

The emigration from Poland and the other Eastern countries may be bad for those countries.

But it isn't bad for the recipient countries, which get human material of splendid quality and which can be easily assimilated, and who put off the evil day of rising dependency ratios.

Something like 3-4% of the Irish population is now East European -- that's demographically significant by any standard. And they're all of working age.

And it's certainly good for the individuals concerned, who triple or quadruple their incomes for the price of an airline ticket from Cracow to Dublin or Bucharest to Madrid.