Wednesday, April 05, 2006

Sidebar Matters II

I've just put a new link in the Population Statistics section to the New Zealand government's Population and Sustainable Development site (the link was kindly sent by a reader in New Zealand, thanks Robert).

There are a number of pages with discussion on key topics (such as population ageing) as well as some thoughts on topical demographic features (e.g. this page on the baby boom). The site also provides links to other sites where demographic and other data may be obtained.

44 comments:

S.M. Stirling said...

Speaking of new data, the CIA World Factbook has come out with their updated (2006) estimates for many (tho' not all) countries.

Some interesting points:

Population decline is picking up speed in southern Africa -- up to 0.4% annually in South Africa, and nearly that in Botswana. The confluence of falling TFR's (now down to 2.2 in SA) and increasing mortality from HIV is ripping through the area unabated. Even in Lesotho and Swaziland (TFR's of 3.23 and 3.7 respectively) growth has stopped.

American fertility is stable -- up marginally from 2.08 to 2.09

Not much change in Europe -- French TFR's at 1.85 and 1.84, statistically not significant.

Mahgrebi and Middle Eastern TFR's continue to fall fairly quickly -- Algeria down from 1.92 to 1.89, Iran from 1.82 to 1.8, Morocco from 2.74 to 2.68, Jordan from 2.7 to 2.63, Lebanon from 1.92 to 1.9.

I'd bet that the Arab countries which have begun to fall seriously are headed for "lowest-low" rates, similar to Spain or Italy -- 1.2 or so.

Edward Hugh said...

"Speaking of new data, the CIA World Factbook has come out with their updated (2006) estimates for many (tho' not all) countries."

OK, thanks for this.

"Population decline is picking up speed in southern Africa"

Yep, this is important. These are (with some former members of the USSR) what the UN calls the demographic outliers (18 of them) with lowest-low (some of the ex USSR group) or declining (the rest) fertility (although there is a fertility decline 'stall' in some African countries) and declining life expectancy. All very preoccupying.

"Not much change in Europe"

My guess is that this is what to expect. France is at one extreme, the troublesome isssue are those around or below 1.3, and whether they are stuck in a fertility trap.

"Mahgrebi and Middle Eastern TFR's continue to fall fairly quickly"

Not in Iraq, Saudi and Palestine they don't. I think we need to be pretty discriminating here. But the Mahgrebi trend (where the population has, remember, a large Berber component) is very interesting: Algeria, Tunisia, Morocco are definately now on the development ramp. What the EU needs to do is 'rope em in'. as they have been doing with Eastern Europe. Probably not direct accession, but some form of partnership along the lines already mapped out in EuroMed.

"I'd bet that the Arab countries which have begun to fall seriously are headed for "lowest-low" rates"

Look, I think this is all hard to see becuase we still lack an adequate theoretical framework for understanding all of this, but we do have some pointers.

I think, as we have been saying, we need to leave the US to one side, since it is quite unique.

With that provisio the rule seems quite clear, the later the transition starts the quicker it proceeds and the lower the 'floor' which TRFs reach.

So I seriously think we might expect to see the 1.2 number breached downwards in the next wave. Also remember that the process of closing on the technology frontier is also quicker, so I think we might see rates of economic growth at some point which breach the China 10% per annum existing ceiling. I think these two things are connected.

Basically the tendency to ever lower TFR 'floors' is a product of ever more rapid birth postponement and ever higher mean ages at first birth. Basically the bigger the leap in ages, and the smaller the fraction of a generation in which this takes place the lower the TFR floor.

If you don't know it you might find the papers at this conference helpful:

http://www.oeaw.ac.at/vid/meeting_postponement_prog.shtml

I seriously think you need to be aware of what could happen to US TFRs if the partr of your population with a third world fertility profile begins a rapid postponement transition at some stage. And if it doesn't I really don't see how you are going to move significant parts of your population up the value chain in terms of the kinds of activities they can engage in.

On countries like Saudi, Iraq, Cambodia, Haiti, Pakistan, Laos, Kenya,Cameroon,
Iraq, Senegal, Sudan, Madagascar, Tanzania, Zambia, Nigeria, Ethiopia
(Etc, etc) with TFRs in the 4.5 - 6.0 range we really don't understand very well what the proximate causes which will set off the badly needed transition really are, only time will tell, but obviously we could conjecture that when they start the transition will be ultra rapid, and the crash in TFRs will really be a crash.

Edward Hugh said...

Incidentally, looking at the factbook, the Turkey enigma continues.The CIA give a number of 1.92 (so it looks like the other data point wasn't simply a mistake).

OTOH the Turkish government itself gives a figure of 2.21 for 2004 (the latest year for which data is available).

What is clear is that fertility in Turkey is falling (especially if we discount the Kurdish component it is falling rapidly), and it will of course be interesting to follow just how far down TFRs actually go.

You can find official Turkish data here:

http://nkg.die.gov.tr/en/goster.asp?aile=1

Interestingly many of the new EU accession countries (or would be ones like Turkey) have more accesible and better organised statistic sites than older members like Spain, Greece, Portugal and Italy. So the reform process must count for something, even if the 'Chapters' on statistical transparency may be among the easier ones to comply with.

Robert said...

Basically the tendency to ever lower TFR 'floors' is a product of ever more rapid birth postponement and ever higher mean ages at first birth. Basically the bigger the leap in ages, and the smaller the fraction of a generation in which this takes place the lower the TFR floor.

Yes, certainly. Now what I'm about to suggest is just a hypothesis. It's probably a testable one, but at the moment I'm not sure what the right test would be ... anyway, let's suppose that building up educational infrastructure takes time (although not as much time as it used to) but that the social change from women not partaking of (especially higher) education, to women taking just as much if not more education than men can be accomplished fairly rapidly, especially in a youth-dominated society.

Now, taking the strong correlation between women taking education and women delaying motherhood for granted, we might reason that in societies that already have reasonable gender equity before they begin ramping up their educational institutions, the average level of women's education, and with it the pace of birth postponement, will advance no more quickly than the overall level of education does, that is, gradually. But those societies that have a significant investment in education, but lag in gender equity, are at risk for a rapid increase in women's education, and with it a rapid entry into demographic transition.

Edward Hugh said...

"But those societies that have a significant investment in education, but lag in gender equity, are at risk for a rapid increase in women's education, and with it a rapid entry into demographic transition."

Definitely. This sounds like a very reasonable hypothesis. The thing is it is hard to think of any societies have "reasonable gender equity before they begin ramping up their educational institutions".

In a curious sense the US might just be a case, but then you need to distinguish between formal gender equality, and gender equality inside the home. Thos parts of US society that still haven't ramped up educationally may well have severe deficiencies in gender equality inside the home.

S.M. Stirling said...

"Not in Iraq, Saudi and Palestine they don't."

-- actually, yeah, as far as Saudi Arabia and Palestine they _do_.

Saudi Arabia is down from 4.5 (2005 estimate) to 4 (2006 estimate). That means that on current trends it'll be 3.5 in 2007, 3 in 2008, 2.5 in 2009, and 2 in 2010.

It probably won't be _quite_ that rapid, but I'd bet it will be close.

And this is down from around 7 only two decades ago. The real decline only began in the 90's and it's been very quick, though not quite as quick as say, Syria.

Likewise, most of the Gulf states are now down below 3 and Syria has gone from 7 to 3.4 in less than 20 years. I'd wouldn't be very surprised if Syria were down to below 2.1 before the end of the decade.

As for Gaza and the West Bank, the Palestinian Authority figures most people use are notoriously bad -- not just incompetently gathered, but deliberately corrupted, to the extent that they overcount the total population by 1,500,000 and have TFR's too high by over a third.

(see http://www.pademographics.com/)

This makes sense; it was intensely unlikely that Palestinians in the West Bank had a TFR of around 5 when their close relatives -- literally the same families -- in Jordan were at 2.63

S.M. Stirling said...

"With that provisio the rule seems quite clear, the later the transition starts the quicker it proceeds and the lower the 'floor' which TRFs reach."

-- yeah, but that's probably an observational artifact.

Add in that, say, Norway, Sicily, and Tunisia have had different demographic regimes for a very very very long time, essentially forever -- and that Sicily has historically been closer to Tunisia than to Norway in most respects of family formation and so forth.

Thus when you hit the three countries mentioned with the same stimuli (for example, increased female education), they react somewhat differently.

That's why I'm betting Tunisia will eventually end up closer to Italy than Norway. In fact it _already_ has a lower TFR than Norway, and the very swift fall shows no sign of stopping.

"Basically the tendency to ever lower TFR 'floors' is a product of ever more rapid birth postponement and ever higher mean ages at first birth."

-- this is self-limiting. Female fertility starts to decline steeply a good decade or more before menopause. Births delayed are quite likely to be births that never happen, as women trying to have children when past 35 are finding.

There might be technological answers to this eventually, but they're not around now.

S.M. Stirling said...

"I seriously think you need to be aware of what could happen to US TFRs if the part of your population with a third world fertility profile begins a rapid postponement transition at some stage."

-- well, that isn't a significant factor. The major reason for the drift upward in US birth-rates since the 70's, despite the continued fall in minority TFR's, is the rise in non-Hispanic white fertility.

In other words, the same sort of people who would in the 1970's have had a TFR of around 1.6 now have one of around 1.8-2.2, depending on region.

The bulk of this would have occurred without any increase in immigration post-1960 -- and US overall TFR's would be only slightly lower, say around 2.0 or 1.98 rather than 2.09 or 2.1, if you removed all those immigrants from the calculus.

This is not a matter of "Third World" fertility; it's ordinary European-American families gradually deciding to have somewhat larger numbers of children.

Note that average age at first child has been increasing _at the same time_ as total TFR's throughout the 90's.

Births to teenage mothers in the US are currently at all-time lows, lower than they've ever been since we started collecting comprehensive statistics.

Since this trend has been consistent for over 30 years now, it's unlikely to change.

"And if it doesn't I really don't see how you are going to move significant parts of your population up the value chain in terms of the kinds of activities they can engage in."

-- well, a population with no young people is sort of severely limited in what kinds of activities it can engage in, too.

S.M. Stirling said...

"On countries like Saudi, Iraq, Cambodia, Haiti, Pakistan, Laos, Kenya,Cameroon, Iraq, Senegal, Sudan, Madagascar, Tanzania, Zambia, Nigeria, Ethiopia
(Etc, etc) with TFRs in the 4.5 - 6.0 range we really don't understand very well what the proximate causes which will set off the badly needed transition really are"

-- most of the countries you mention have already seen TFR declines of around 30%-50% in the past generation, so I really don't see your point.

Remember, they all started with "natural" fertility regimes based on universal early marriage for women and no contraception; which is to say, with TFR's in the 7-8 range.

If you examine them in detail (apart from some of the African countries) they just look a bit behind the curve. Late starters, in other words.

Eg., the percentage of married women in Pakistan using modern contraceptives has gone from 5% to just under 40% in the past 15 years.

Even Kenya had a TFR of 8 as recently as the time I lived there in the 1960's and 70's; it's now only a bit more than half that. A 50% decline in 25 years is hardly stasis.

S.M. Stirling said...

And, checking, I see Iraq has gone from around 4.4 to 4.12 in 2006.

The only countries that _aren't_ seeing rapid falls in TFR are the genuinely "extremely backward, extremely isolated" ones; Sub-saharan basket cases like Niger, or Asian backwaters like Afghanistan.

S.M. Stirling said...

Checking again, I see that Pakistan is down to a TFR of 4 for 2006, with (IIRC, working from memory) around 4.3 or 4.4 in 2005.

Edward Hugh said...

"Saudi Arabia is down from 4.5 (2005 estimate) to 4 (2006 estimate). That means that on current trends it'll be 3.5 in 2007, 3 in 2008, 2.5 in 2009, and 2 in 2010."

Nice to see you back, I wondered where you had gone to :).

I am sure Saudi fertility is going down, but I have severe doubts it can be that quick. This implies a whole mindset switch, and things aren't normally that rapid. You are extrapolating forward from one years data, and this is a very problematic thing to do.

There are measurement error issues to think about, and there may be rounding going on.

The only way you are going to get the numbers down to 2.0 and beyond is to get the kind of social revolution you have had say in Jordan off the ground. My feeling is that in Saudi this is still going to be hard work, but justifying why I think this would lead us perhaps a long way beyond the terrain of demographic matters.

Basically though I think that globally you are right, fertility will drop much more rapidly than people are currently anticipating (the technological catch up process will drive this) and we could see some very low versions of 'lowest low' fertility arriving just over the horizon.

On Palestine, you may be right, there is possibly overcount, I don't really know enough to comment. On the other hand there are a hell of a lot of children who show up in the TV news shoots, so this must mean something. Again socially the conditions in the Gaza strip etc are very different from those in Jordan, so some substantial difference would be to be expected. I doubt young Palestinian women are as emancipated as young Jordanian ones are.

Edward Hugh said...

"Norway, Sicily, and Tunisia"

Yes, but remember Norway's decline has been a lot longer, slower and smoother than the other two. This is what I'm conjecturing, the slower the decline, the less dramatic the 'missing births' posponement impact. What you have is much nearer to a pure quantum decline with some postponement effect.

What you are likely to get in Turkey, or India, or Saudi if you are anywhere near right, is an ultra rapid postponement process, and the tempo element will be huge, which is why we could see the arrival of extraordinarily low derived TRFs (much lower than the 1.2 or 1.3 range we have seen to date, and this could last some decades.


"-- this is self-limiting. Female fertility starts to decline steeply a good decade or more before menopause. Births delayed are quite likely to be births that never happen, as women trying to have children when past 35 are finding.

There might be technological answers to this eventually, but they're not around now."

Yep, I agree with this.

Edward Hugh said...

"-- most of the countries you mention have already seen TFR declines of around 30%-50% in the past generation, so I really don't see your point.

Even Kenya had a TFR of 8 as recently as the time I lived there in the 1960's and 70's; it's now only a bit more than half that. A 50% decline in 25 years is hardly stasis."

Again, this is not really an area I know a lot about, but there has been what they call a fertility 'stall' in some parts of Africa. Some have tried to associate this with the arrival of aids.

This session at the last IUSSP has a representative selection of papers:

http://iussp2005.princeton.edu/sessionViewer.aspx?sessionId=301

Also this is what leads the UN to talk in the Human Development report about growing divergence.

Also this is clear economically in terms of the distinction between developing and less developed countries. I think it is pretty widely accepted among growth theorists that one group of countries are stuck in some kind of 'development trap', where growth is very low, or near backward, and living standards move away fromm and not towards the global mean.

Well all this clearly correlates in some way with fertility.

The UN talks about high fertility and intermediate fertility countries, and this more or less mirrors the LDC developing country distinction.

Incidentally, it isn't just some parts of Asia and some parts of Africa where fertility stays stubbornly high, the indigenous peoples of Latin America seem to be on a similar path, as we have seen in Bolivia, and we are now seeing (possibly, in terms of politics I mean) in Peru.

Bottom line, obviously the numbers are coming down generally, and the highest levels are not as high as they used to be, but in far too many cases they are not coming down anything like as fast as might be hoped.

Edward Hugh said...

"well, a population with no young people is sort of severely limited in what kinds of activities it can engage in, too."

Yes, but I think the point would be not having no kids, but having fewer of them, and having fewer and living longer, and having fewer children with more intellectual quality in each child, ie sometimes fewer can be better. My only real concern is to do with the inter-generational consequences of all this happening too quickly.

Also, I think the objective in science isn't to score points for the sake of scoring them, but to try and understand, and to try and understand so we can do a better job of whatever it is we happen to be doing.

Now...

Let's look a bit more closely at those US dervied TFRs. Live births 2003 (which is linked in the sidebar) show this data for non-hispanic white TFRs from the late 80s. As you say, they come and go a bit, but basically fluctuate somewhere around the 1.8 mark, although the level has risen a little in recent years, and this is interesting. But they are still not that different from the figures for similar European populations like the British or the French (or even the Swedes). And all of this is interesting to note, since all these populations started their fertility decline early, and proceeded with it comparatively slowly (ie over a couple of centuries).

Now other component of your population are either currently having (the afro american population) or about to have (the hispanic population) a much more rapid transition to below replacement fertility, and for the reasons of rapidity could actually hit at some point (assuming all those poor latinos aren't going to stay poor for ever) levels like those seen in southern europe or the asian tigers.

If this were to happen the US composite rate would then be significantly below replacement, which wouldn't be any big deal to worry about, in fact it would be a sign of some sort of normality, but I think we should be clear that this *could* happen (I would say it probably will, but I am just guessing).

Back to live births 2003 for a moment:

"The U.S. TFR was below the replacement rate for the 32d consecutive year in 2003. Whereas the TFRs for most groups were below replacement in 2003, the rate was above replacement for Hispanics overall (2,785.5), and for Mexican (2,957.5) and other Hispanic women (2,733.0)."

So with 32 years below replacement, you're not that different from your comparative reference groups really.

Denmark, Finland and Norway have what 1.8 (and comparatively little immigration) and your white non-hispanic is just a touch above this. I don't see what all the fuss is about really.




White
2003 . 1,856.5
2002 . 1,828.5
2001 . 1,843.0
2000 . 1,866.0
1999 . 1,838.5
1998 . 1,825.0
1997 . 1,785.5
1996 . 1,781.0
1995 . 1,777.5
1994 . 1,782.5
1993 . 1,786.0
1992 . 1,803.5
1991 . 1,822.5
1990 . 1,850.5
1989 . 1,770.0

Edward Hugh said...

"And, checking, I see Iraq has gone from around 4.4 to 4.12 in 2006."

Well like I say on the other thread, this isn't the prb figure for Iraq, which was 5.1 last year. It will be interesting to see what the number is next year, but this data on infant mortality is Basra (which is consistent with other info we've been receiving on health related deaths doesn't suggest that fertility will be coming down any time soon, quite the contrary:

As a result of water-borne diseases and a lack of medical supplies, infants born in the southern city of Basra are subject to abnormally high mortality rates, say officials of an international NGO devoted to child health issues.

"For weeks, there were no I.V. fluids available in the hospitals of Basra," said Marie Fernandez, spokeswoman for European aid agency Saving Children from War. "As a consequence, many children, mainly under five-years old, died after suffering from extreme cases of diarrhoea."

Fernandez went on to cite a number of problems facing local hospitals in Basra, which is located some 550km south of the capital, Baghdad. "Hospitals have no ventilators to help prematurely-born babies breathe," Fernandez said. "And there are very few nurses available, so hospitals often must allow family members to care for patients."

http://www.alertnet.org/thenews/newsdesk/IRIN/b9fdc6aeb83850f9749264765b65b5be.htm

According to doctors and NGOs, the primary causes of high infant mortality are unsafe water, diarrhoea, malnutrition, infectious diseases, maternal stress and poverty.

S.M. Stirling said...

As to Boliva, the Factbook has the current TFR at 2.85, which is fairly low and drastically lower than

Given that the "indigenous" (I prefer the historically more informative "Indian") share is well over 50% and always has been, this means that their fertility must be dropping too.

S.M. Stirling said...

As to Boliva, the Factbook has the current TFR at 2.85, which is fairly low and drastically lower than

Given that the "indigenous" (I prefer the historically more informative "Indian") share is well over 50% and always has been, this means that their fertility must be dropping too.

S.M. Stirling said...

"Again, this is not really an area I know a lot about, but there has been what they call a fertility 'stall' in some parts of Africa. Some have tried to associate this with the arrival of aids."

-- that sounds rather dubious, as the areas with the worst HIV problems are also those where TFR's have fallen most rapidly.

Of course, HIV itself reduces fertility, but countries like Botswana or South Africa, with TFR's down below 3, are obviously undergoing sustained deliberate fertility reduction.

(Of course, the age-selective impact of AIDS deaths in an African context also pushes the _crude_ birth rate down below what you'd expect with the TFR, since it selectively kills women in their primary reproductive years.)

South Africa is at 2.2, for example, and Botswana at around 2.85.

If Botswana weren't getting a lot of immigrants from Zimbabwe (itself a locus of high HIV+ rates) its popuation would be dropping at something like 1% a year.

South Africa dropping at nearly 0.5% a year and it's accelerating.

This is more rapid than almost all the predictions; I think people were just reluctant to accept that a catastrophe on this scale was possible.

S.M. Stirling said...

"Yes, but I think the point would be not having no kids, but having fewer of them, and having fewer and living longer, and having fewer children with more intellectual quality in each child, ie sometimes fewer can be better. "

-- these factors are not connected. Eg., my mother and father had 4, and my uncle had 6, and my father and uncle are in their 80's now.

All of their children are highly successful; senior IT specialists, novelists (myself), businessmen, etc. So far the (quite numerous) grandchildren seem to be following the same trajectory.

S.M. Stirling said...

"As you say, they come and go a bit, but basically fluctuate somewhere around the 1.8 mark, although the level has risen a little in recent years, and this is interesting."

-- actually, it would be more accurate to say that a steady, though very slow increase has occurred to the present level of 1.9-2.0

However, "non-Hispanic white" is at too broad a level of generalization to be useful.

The regional and cultural differenceds within the US produce persistent differences in TFR greater than those between the US and Europe as a whole.

S.M. Stirling said...

"But they are still not that different from the figures for similar European populations like the British or the French (or even the Swedes)."

-- In fact, states with the _lowest_ fertility in the US (for this group) are equivalent to those in Europe with the _highest_ rates.

Eg., Vermont has a TFR of 1.6, which is about the lowest in the US. That's well above the EU average, even before you added in the post-Communist countries.

There is no equivalent in Europe to states like Utah or Idaho (around 2.5 and rising) or the broad range (about 30) with TFR's over 2 for people of European descent.

Incidentally, "non-Hispanic white Americans" are predominantly of German descent; Germans outnumber all the British nationalities together, IIRC, as a constituent part of that group.

S.M. Stirling said...

As to the "lowest-low" possible rates, urban China as a whole is now down to around 1. Hong Kong is below 1.

This is, ah, "grossly unsustainable", of course; it would produce a demographic profile, if continued, in which the population would fall by 50% per generation and a majority would be over... what, 65? 70?

Urban Chinese seem to have ceased reproducing at all. A TFR below 1 must mean that something approaching half the women, or possibly more than half if a high percentage of children are to families with multiple births, aren't having _any_ children.

And China is urbanizing very rapidly; already over 40% of the population, around 500 million, are

S.M. Stirling said...

In the very long run, sub-replacement fertilty will self-correct.

More and more of each generation will be the children of people with either a very strong cultural drive to reproduce, or a very strong biological one, or both.

Eg., on current trends everyone in Canada will be a Hutterite, Amish, or an Orthodox Jew eventually.

That's a bit of a _reductio ad absurdum_ but the principle holds.

S.M. Stirling said...

Second thoughts on China: and the sexual imbalance is now starting to drive crude birth rates below the level that the TFR would produce in a normal population.

S.M. Stirling said...

As to the world TFR, it's now at 2.59.

Replacement level worldwide is probably at about 2.2, given mortality levels and trends.

At current trends (including trends in the acceleration of trends) we should hit 2.2 globally sometime within the next 10-15 years.

The big driver over the next decade or so will be South Asia, where India is already at 2.73. Breaking India down, the urban areas and big chunks of the rural south are already subreplacement, with the rural north and northeast still fairly high but beginning to drop quickly.

East Asia is almost all subreplacement now except for Indonesia, which is close. Latin America is at or near replacement level -- most South American countries are now well below 3.

Depending on how far below replacement world TFR's go, world population could top out and start falling between about 2030 and, IMHO at the latest 2050. At that point, the only large population groups still growing will be Americans and Africans.

In other words, even the UN's "low" projection is probably a bit high.

Edward Hugh said...

"As to Boliva, the Factbook has the current TFR at 2.85, which is fairly low and drastically lower than"

This is really getting to be troubling, since the data the CIA are offering seem to be seriously below the prb data virtually everywhere we look.

The prb had Bolivia at 3.8 last year, and looking at the political instability my guess is that is about right.

Do they have a methodological paper anywhere on how they make their own calculations, since they do seem to be at variance with virtually everyone else. Is this a scientific peer reviewed procedure? Certainly if the CIA were right it would be very interesting, since it would mean that global population would peak much sooner than the whole demographic community is expecting.

I obviously have no interest in exaggerating upwards fertility levels. The sooner they come down the better afaiac, but at the same time I do believe in facing up to reality, even if that reality may not be the one we would wish for.

"I prefer the historically more informative "Indian""

Well yes, I have no problem with this, I would never actually quibble about a word. The people themselves now seem to refer to themselves as los indigenos in Spanish (or indígenas in portuguese) and in just the same way as afro-americans are afro americans (rather than black americans) I accept the conventions.

Of course all this duck the really big question of 'mestizos' like Humala in Peru, but still.

"share is well over 50% and always has been, this means that their fertility must be dropping too."

Well the first point is that, since they have higher fertility, their share must be rising.

And the second one is yes, obviously, the fertility is dropping. The question is simply how rapidly. There can be stalls and negative feedback mechanisms along the road.

Obviously if fertility wasn't dropping virtually everywhere there would be no reason to imagine that global population was going to peak later this century.

I have been mentioning Bolivia since it is one of the countries where per capita income has been falling away from (rather than converging with) the global mean, and since even in Bolivia there will be some slow technological transfer which will raise levels, fall-backs normally imply fertility issues.

Edward Hugh said...

"-- that sounds rather dubious, as the areas with the worst HIV problems are also those where TFR's have fallen most rapidly."

Nope, I'm afraid you are just plain wrong here. Fertility does stall in areas where mortality rises (during the transition that is, not in post transition societies like Russia).

You should read the papers I link to, and look at Sriya Iyer's work that I have also mentioned, she gets the same finds in India, and uses an options theory model (based on desired number of surviving children) to explain this.

I will try and post something on all this.

Edward Hugh said...

"these factors are not connected. Eg., my mother and father had 4, and my uncle had 6, and my father and uncle are in their 80's now."

Well to quote my authority on this:)

"And there's a reason why "anecdotal evidence" is a polite synonym for "utterly worthless." "

And if you remember, at the time I didn't agree. I'm sure your familiy history is interesting, and valid as testimony of something at some level.

The question is, whether it can be used as a counter-example to huge global trends.

My case is different. My father was one of 14. I am one of two, and I have one biological child. My father left school at 12. I'm not sure when you could say I stopped formal education, but at least the double of that, and my son is 25, already has two first degrees, is about to become a brain surgeon, and I see no end to this in sight. I am not sure whether or not he will have children.

This is, IMHO, the more typical pattern.

Incidentally, I have mentioned that my father was a migrant in Arkansas from 1922 to 1933. (As it happens I was born in the UK, but as - in principal - a potential US citizen).

He went to the US to take a 'war hero' brother (WWI) in search of a new life. That brother also had 2 children. One of them went on to be the first woman to teach economics in a US university. She never married, and had no children. Her brother becamse a US army doctor (we are a family of doctors and economists) and had two children. One of those children has no children, the other I'm not sure.

But as you say, these are only anecdotes. The aggregate numbers tell the real story.

Edward Hugh said...

"The regional and cultural differenceds within the US produce persistent differences in TFR greater than those between the US and Europe as a whole."

Well this is what I'm not so sure about. Certainly in Europe there are huge regional differences, what happens is that in the US they are on the upside, and in Europe they are on the downside.

In the US the fertility of NW European migrant descendants is not that different from their 'cousins' in France, the UK, Sweden and Ireland.

South and Eastern Europe is at the lowest-low pole, Hispanic US is at the 'highest-high' one.

My gut instinct is that if the 'highest-high' groups in the US come crashing down, as you are indicating they are doing everywhere else globally, then the US in 20 years could look pretty much like Europe does now. But there it is, that is just a theory, and there is no reason at this stage for you to accept it.

Of course prudence would counsel that you keep a weather eye open, just in case. In Greenspan/Bernanke speak, there is downside risk here.

Edward Hugh said...

"In the very long run, sub-replacement fertilty will self-correct."

This is the part I really don't accept. I don't see why it need to, and why we should worry, especially if we are living longer. The number of human life years may be the constant here.

Following your logic, in a finite number of generations everyone here in Spain will evantually be a gypsy, and, of course, extremely poor. I don't buy this.

Here is my latest (yesterday) thinking on the topic:

http://www.edwardhugh.net/homeostasis.html

Edward Hugh said...

"There is no equivalent in Europe to states like Utah or Idaho"

Yes, this is true. One question is, what does Utah live from? Idaho presumeably has agriculture. So to what extent are these communities subsidised by the more economically lucrative activities of their coastal co-citizens?

"Incidentally, "non-Hispanic white Americans" are predominantly of German descent; "

Oh, this I didn't know. This is interesting. But you wouldn't know that from looking at the presidents. They give the impression that the Irish are the predominant group.

Of course fertility in Ireland is now on the way down, and will possibly plummet to rock bottom.

"That's well above the EU average, even before you added in the post-Communist countries."

Yes, but you need to break the EU average down in to regions just like you do for the US. The argument that the prb use that the US is more diverse than the EU isn't quite right. The US is more diverse ethnically than individual EU countries (possibly outside France and the UK, and maybe Sweden) are.

Interestingly you may like to note that the most diverse countries (culturally) are those which started their transitions earlier. Ireland, for eg is pretty homogeneous, as was Spain until very recently.

But then if you break the US down by States, just how much actual diversity is there in some areas. This would be Frey's point I think.

Some areas are now becoming more diverse as first the Afro american population move inwards, and then the Latinos follow them.

Edward Hugh said...

"Depending on how far below replacement world TFR's go, world population could top out and start falling between about 2030 and, IMHO at the latest 2050. At that point, the only large population groups still growing will be Americans and Africans."

Strange as it may seem I wouldn't really disagree with this. I think 2030 is a bit soon, but 2050 is a real possibility IMHO. My hunch is that the rest of the tarnsition will be a lot more rapid than people imagine. But it is simply a hunch. At this time we don't have the scientific data to back our hunches, although we may do over the next decade.

S.M. Stirling said...

"Nope, I'm afraid you are just plain wrong here. Fertility does stall in areas where mortality rises (during the transition that is, not in post transition societies like Russia)."

-- not in South Africa, Botswana, Lesotho, or Swaziland. That's obvious on the facts.

The enormous increase in mortality and the drop in longevity has proceeded directly in parallel with the drop in TFR's.

S.M. Stirling said...

"Yes, this is true. One question is, what does Utah live from? Idaho presumeably has agriculture."

-- Idaho hasn't been predominantly agricultural since the 1930's. It's a center of high-tech and IT industries with one of the best educational systems in the US, and it's highly urbanized.

S.M. Stirling said...

"Oh, this I didn't know. This is interesting. But you wouldn't know that from looking at the presidents. They give the impression that the Irish are the predominant group."

-- Actually that's the Scots-Irish, starting with Andy Jackson in the 1830's; JFK was, as far as I know, the only Irish-Irish President.

Reagan, for example, was more English and Scottish than Irish by ancestry.

The Census reports people of German background as about 25% of the non-Hispanic white total, but that's generally agreed to be a gross underestimate.

The total was over 10% as early as the Revolution, even before the great 19th-century German wave, and a near-majority in Pennsylvania at the time.

Apart from a small minority of religious sectaries like the Amish, Germans blended in rather rapidly -- they were a major component in the 18th-century settlement of the Piedmont South, for example, but usually Anglicized their names and assimilated among the Scots-Irish and English.

Schmidt became Smith, and so forth. So a lot of people who usually identify as Southern or Appalachian are part-German by descent.

S.M. Stirling said...

"Yes, but you need to break the EU average down in to regions just like you do for the US."

-- States in the US roughly correspond to countries in Europe, in terms of population size, though the largest aren't quite as large as the biggest EU contries.

"But then if you break the US down by States, just how much actual diversity is there in some areas."

-- that would require a very fine-grained microscope.

Eg., Vermont and Utah are both very "white" (well over 95%). On a coarse level, they're even the same kind of white people -- mostly British, German and Scandinavian in the case of Utah, reflecting the 19th-century immigration of Mormon converts.

But they have, respectively, the lowest and highest TFR in the US.

S.M. Stirling said...

"In the US the fertility of NW European migrant descendants is not that different from their 'cousins' in France, the UK, Sweden and Ireland."

-- well, yeah, actually, it is; about 25-35% higher.

As I mentioned, people in Utah are of almost exclusively British, German and Scandinavian descent, and have a TFR of 2.5

But that's a generalization too. The _Mormons_ (about 60% of Utah's population) have a TFR nearer 3.

S.M. Stirling said...

"This is the part I really don't accept. I don't see why it need to, and why we should worry, especially if we are living longer. The number of human life years may be the constant here."

-- we aren't living any longer.

The reason the average lifespan is increasing is that more people are living to the biological limit, 80-100 in most cases.

But _some_ people have always lived that long.

The decrease is in mortality at early ages, in childhood and adulthood, not in any extension of the span at the other end.

If you're 90, it would still be fairly futile to start working towards a postdoctorate.

S.M. Stirling said...

The significant difference between TFR's in Europe and among people of European descent in the US is not just that the American rates are higher.

The really significant difference is the causative factor.

Until the 1970's, European and American fertility rates tracked each other quite closely -- bust in the 1930's, post-war baby boom, then another bust in the 60's.

Then they started to diverge, with Americans showing a slow steady increase from a low around 1.6 or a bit more.

This trend is apparently being sustained; there are year-to-year fluctuations but the overall trend is up.

In Europe, you also get year-to-year fluctuations but the rate is _down_.

S.M. Stirling said...

It's important to note that American fertility rates are not only the highest in the developed world.

They're higher than many of the _underdeveloped_ countries. Higher than the whole of East and most of SE Asia, higher than most of Latin America, higher than most of the Mahgreb and the Middle East, and even higher than parts of Africa.

In fact, on present trends they'll be higher than the world average quite soon -- within 10-15 years.

S.M. Stirling said...

Note that virtually everyone in Argentina is of either Spanish or Italian descent -- all people of mixed or non-European descent together are barely 3% of the total.

And while Argentina has not been as successful economically as either Italy or Spain lately, it's still a predominantly urbanized country with high rates of literacy and so forth.

Yet Argentina's TFR is 2.16, while Spain and Italy are around 1.2.

Edward said...

"-- well, yeah, actually, it is; about 25-35% higher.

As I mentioned, people in Utah are of almost exclusively British, German and Scandinavian descent, and have a TFR of 2.5"

I don't see where you get the 25-35 % from. The non-hispanic white (which I agree is a problematic category) is around 1.8. Fertility in the countries I mention is either slightly below, or slightly above (France and Ireland) it. Frnace obviously doesn't produce ethnic data, but there is no good reason to assume they have the extremely low fertility of there German cousins. 25 - 35% is a big difference, and Utah, which you are obviously right about, just isn't typical.

Even in population terms Utah seems miniscule, just over 2 million. Idaho is even smaller, 1.4 million 2004. This is about the scale of the Basque country in Spain, or the UK province of Northern Ireland, both of which probably have a-typical fertility without changing anything important about Europe.

Edward said...

"The reason the average lifespan is increasing is that more people are living to the biological limit, 80-100 in most cases."

This is certainly the case, although it isn't clear that there is an actual biological limit.

"The decrease is in mortality at early ages, in childhood and adulthood, not in any extension of the span at the other end. "

Well the first part of this used to be the case, but I think in the OECD world the days of major reductions in child and infant mortality are now well behind us for the simple reason that there is now relatively little of it.

So the increase in life expectancy is driven by old people increasingly reaching the older old ages. What I have been calling the 'rectangularisation' of mortality.

But it is not the case that there is "not in any extension of the span at the other end." or that:

"-- we aren't living any longer".

According to James Vaupel, in this study:

http://user.demogr.mpg.de/jwv/pdf/International%20Database%20on%20Longevity.pdf


The emergence of the supercentenarians was accompanied by a significant increase in the Maximum Reported Age at Death (MRAD). Some have discussed a handful of historical cases of individuals alleged to be dead at age 110 or age 111 before 1960, but no cases of death at age 112 or over have been proven to exist before 1973. Since that time, however, deaths have been recorded at age 114 in 1985, age 117 in 1993 and age 122 in 1997"