[Someone continues to beat a drum and “justify” their previous claim that a difference of 5% in global population by 2050 is accurately characterised as “a lot” of a difference, and an indication that everything is alright now that “trendy doomsayers” have been proved completely wrong. I understand this argument may partly stem from, although not explicitly, items such as NYT’s “The Population Explosion is Over” and the WSJ’s “Population Implosion” announcement, and other “rationales” for cutting population funding to US NGO’s].

Oh, oh. Some relevant context has been “accidentally” deleted; let me re-insert it:

$ David S. Maddison <> wrote: $$ And BTW, all those trendy doomsayers turned out to be wrong when the $$ UN announced that the future population was going to be a lot less $$ than expected didn’t they?

Hmmm. As we’ve seen, this hype claim seems to derive from the use of the word “significant” attached to “growth rate estimates” in a press release, rather than one relevant to “future populations”.

In any case, what amounts to a propaganda cliche of “a lot less” is associated with a change of around 6% in estimates of growth rates, and a median “future population” diff of an estimated 5% by 2050, or (equivalently) a delay in population projections c2050 by a matter of no more than 5 years.

This is a “significant” cause for celebration?

>Well, the UN agrees that the new estimate is “significantly lower”, >which most people would consider to be equivalent to a “lot less”. And

[See e.g. http://www.undp.org/popin/wdtrends/wpgrow.htm, which appears to be a related UN press release to the one posted].

It was clearly the “future population” being “a lot less” and the predictions of “trendy domesdayers” that were the question before this latest mis-direction.

And the “UN agrees” to no such thing. You are simply parroting uninterpreted and out of context part of one paragraph of one press release. Your (now) *claim* that is related to an alleged consensus position of “the [members of the] UN” is (at best, and ignoring the argument by assertion aspect) an “appeal to authority”, (also) a fallacy. (For alternative views, see e.g. “85 Nations Believe Birth Rates Too High”, The Population Institute, Sep/Oct 1997).

That the “UN position” represents a “consensus position” at all, is another fallacy. E.g. the Washington-based Population Institute, funded by the UN, appears to argue its 95/6 “world population doubling in about 40 years” estimate is not significantly changed by the latest Revision — see http://www.pathfinder.com/money/latest/press/PW/1997Dec30/696.html or their 1997 report at http://www.populationinstitute.org/overview97.html (“Whether the tidal wave is 80 feet or 100 feet high, the impact will be similar”, said Fornos — but skip to the section “The Problem is Far >From Over”; see also “Developing World’s Quality of Life is Declining”, and “Future Population May Overwhelm Food Supply”, TPI Nov/Dec 1997).

And even “other” UN documents, independent of considerations relevant to recent Revisions — e.g. “Long Range World Population Projections: Two Centuries of Population Growth, 1950-2150”, NY, UN (1992) — show that even had global fertility suddenly reduced to replacement levels (a fall of around 33%) in c1990, the global population would still reach almost 8.4 billion over the next several decades (an “age structure” thing, see below).

There is a “significant” and, so far un-remarked by yourself, distinction between a 5 year *rate *estimate* (latest ploy) and the “future population” (original confusion, and underlined above).

While a recent growth *rate* has been adjusted downward by around 6% (a “whole lot”?) the resulting median estimate of the “future global population” (say, arbitrarily at mid C21) is affected by no more than 5%. After all — the previous c20 years of growth estimates (especially relevant, as will be indicated below) haven’t been similarly revised (in the past, and unbeknownst to some, some of these have been sometimes revised UPWARD by similar amounts). Additionally, the 5% is seen to be rather less than the *spread* of UN population estimates (i.e. the difference between the “low”, “median” and “high” estimates).

How can a 5% revision be taken seriously as a “significant change” relevant to “future populations” (i.e. as at c2050) if said population estimate ITSELF is not known more accurately that (say) “somewhere between 7.7 and 11.1 bn, most likely near 10 bn” (characterised by one UNFPA release as “this vast 3.4 billion range” — see http://www.unfpa.org/swp97e/box1.html)? I.e. the “a lot” 5% adjustment is related to a quantity with an error bound of (charitably) +-10%.

It should be apparent to anyone perusing “real world” data, that changes in demographics of 5-10% over 1 year is “not unusual”. Again, changes in “growth rates” are also seen to fluctuate for various reasons. Finally, changes in *estimates* should be expected, and are observed to change, as (estimated or re-crunched 😉 “actual” data becomes routinely substituted for earlier (almost completely guesstimated) “data”. Quarterly or annual economic or demographic data are all typical, e.g., in these regards, and are printed (but perhaps not observed) almost weekly in newspapers, etc.

It is quite usual for (more specifically) population growth rates to fluctuate from decade to decade. See, e.g., “Australian Political Facts” (op cit) for tables of Aussie demographics from c1900-c1980, or other relevant stats.

*Global* population growth *rates* also vary “quite significantly” (if 5% is claimed to be “a lot”). Published data for C20 have shown both increase and decrease (see, e.g., the Cambridge encyclopedia under “world population”). It is reasonable to expect that such variations are at *least* partly due to various national and international policies and procedures, changes to food production distribution and storage, public health, wars, one child families, “encouraged” sterilisation or contraception, improved education and standard-of-living (usually associated with lower birth rates), etc.

But *some* variation quite reasonably also results from completely unrelated “ringing” that occurs in population demographics. I.e. the age structure of a population can have comparable or conceivably even greater short-term impact on growth rates than other factors.

The typical scenario in a population (even in the absence of technological change that may benefit, say, 5% of the world’s population at somewhat the “expense” of a significant part of the other 95%) is that at some time X, a significant fraction of that population is (say) 15-25. (That a population can “usually” be expected to reach such a situation is related to stochastic processes, Markov theory, laws of big numbers, and is somewhat beyond the scope of this discussion). This scenario results in an increasing number of children as the 15-25s at X enter their child-bearing/rearing years over decade X to X+10. But for some time thereafter, the growth rate downturns, as the somewhat lesser number of 5-15 yo’s at X become 15-25 yo’s. I.e. after X+10 there is a somewhat superficial (and periodic) affect of overall “population aging”.

It takes c 2 decades for the children 0-5 at X (i.e. those born to “baby boomers” [obviously not to be confused with the term used for the relatively small number of Westerners born c1945, or *their* children born c1970] at X) to mature into *their* child-bearing years, and then to repeat another up/downswing cycle.

The long-term pattern is then one of (apparently) “significant” fluctuations in population growth, with both “significant” up-swings and down-turns over a couple of decades (aka “ringing”).

As to the *objective* significance of a 1.57% vs 1.48% growth rate (rather than what poor comprehension might simply infer from a first paragraph somewhere): the former number indicates a doubling of global population (surely, something that “anyone” — if they would argue 5% is a “lot” — would call a “very very very great lot”) takes less than 45 years, while the latter indicates doubling must “wait” for around 47.2 years. (For future reference: a 1% growth rate doubles in around 70 years; 0.5% in around 140 years). Apparently the difference between 45 and 47.2 is “a whole lot of difference” in the context — and a cause for loud rejoicing in non-“doomsayers” (aka cloud cuckoo) country.

But considering the uncertainty in the estimates, it’s wise not to count eggs until hatched. The global population consists of at least 2 disparate segments: in the so-called developed countries, 20% of the population is growing at a rate of 0.4% pa; but in the “less developed regions” is growing at 1.8%. Some researchers (see Lutz) indicate that it’s far from likely the latter growth rate will decline to significantly under 1%, as proposed by some models from the Population Division of the UN’s Dept of Econ and Social Information and Policy Analysis, particularly as one of the so-called “certainties” of the C21 is that “developed” countries with and because of their lower growth rates will form a smaller and smaller segment of the global population (i.e. the global growth rate will tend to converge to the *larger* of the 2 growth rates — a dicey prospect, falling under the heading of “order statistics”).

And the posited reasons for the decline in the estimates include some “one offs”. E.g. wars in Africa and the middle East, and the spread of AIDS has reduced life expectancy below previous estimates. In East Africa life expectancy is (now) estimated at 46.7 years — 3.9 less than previous Revisions. In E Europe it unexpectedly declined to 68 years in 90/95 from 70 years in 85/90. Also, improvements in gender equality (gads! PC can *achieve* something?) and women’s health have contributed to some significant reductions in “family size” in some countries (See, “In Latin America, Fertility halved since 1960s”, The Population Inst, Nov/Dec 1997).

As to the *historical* relevance of the cited growth “revision” and it’s otherwise counter-claimed (by some 😉 comparative normalcy, I find in a c1960 Rigby encyclopedia (published in Adelaide, of all places) that nothing much appears to have changed over the past c50 years, prediction wise.

Under “world population” in the relevant volume (I presume all other reasonable encyclopedia of the period will show similar) the article indicates the global population c 1950 was 2.4 bn (the UN estimate 1995 for 1950 was 2.52 bn, the USBC estimated it as 2.556 bn — see http://www.census.gov/ftp/pub/ipc/www/worldhis.html), and this was (then) expected to double by the turn of the century, with another doubling by the mid 21st century. I.e. the “population doubling” frequency c1960 was said to be around 50 years at that time.

Obviously the then-believed growth rate must have been *lower* than that subsequently observed (presumably due to skewing factors such as world wars, sketchy data for most of the world c1950, better-than-expected improvements to agriculture, lowered infant mortality over the past 50 years, etc), because 5.0 bn was *actually* chalked up ahead of the then-posited schedule — by c1986 in fact. The latest median prediction from the UN still puts c 9.4 bn around 2050 (and at 0.5% pa sees 10 bn by around 2060; and at 1% pa sees 10 bn by around 2055).

>the numbers involved represent 500 million less people. Most people

Ah. A case of a (projected) naive “absolute mindset”.

The largeness or smallness of a number depends on its context.

The largeness and smallness of quantities (i.e. numbers in association with some unit) are only relevant in their context. A quantity isn’t “large” or “small” in absolute terms. “Some” people understand this. Some of *them* are “scientists” (who also *allegedly* understand trivial properties of exponential growth, etc).

Whether 500 million is large depends on the sentence in which it’s being used.

Whether 500 million Angstrom is a large distance depends on *its* context (e.g. how far is that to drive?).

Whether 500 million is a large number of people depends on *its* context.

The context here is one of a global population of c 10 billion people — around twice that of c 1985. I.e. the context reveals 500 million to be 5% of the total in question, or represent (equivalently) a “delay” of around 5-10 years relative to global population growth schedules over the next 5 decades. (Even considering the 500 mn will be in less-developed countries, it represents around 6.6% of *that* sub-population).

500 million people less then — while obviously in the right direction (since it is due to the revision in an estimate, we can’t simplistically put it down to being “a step” that was taken in that direction 😉 — is an almost insignificant quantity that can be obliterated within “the twinkling of an eye”, all things remaining as “likely” predicted.

>[50 lines of fluffy argument relevant to 5% being “correctly” interpreted by all thinking persons as “a lot”].

Let’s review some basic ideas.

When dealing with (approximately) exponential growth, even alleged “significant” changes in “recent” growth rates (esp, in “recent” ones) — in this case one of 6% from an original estimate — are not *objectively* significant. They have little influence on “future population levels”, other than marginally changing (say) the “doubling” schedule. Provided a growth rate is “sufficiently positive” (i.e. modulo infant mortality and public health factors in this context) any given target population will *eventually* be achieved by simply waiting “slightly” more or less time. Exponential growth is “fast”. That is one of its basic properties.

To illustrate: instead of a c1960s encyclopedia being “exactly” right in predicting a global population of 10 billion by *exactly* 2050, if we adopt the median UN estimate there is “likely” to be around 9.5 billion c2050. 9.5 is not 10, but we must see (e.g., using a growth estimate of 1.48%) an additional 686 million people (called “a lot” by some) will magically appear on the scene WITHIN FIVE ADDITIONAL YEARS — i.e. by 2055. If world population growth is 1% c2050 by mid C21 an additional 459 million mouths and bums are due to be added to the global count between 2050 and 2055. According to Brian Carnell’s “overpopulation FAQ” the extrapolated growth rate (based on the declining rates post 1950) c2050 may be 0.54% pa — but even then 498 mn are added per decade to the global at that time (he cites “UN projections” as predicting the world population will “level off at about 9.8 billion people” c2050). I.e. the c1960 prediction is *likely* no more than 5-10 years off, unless something (more) changes in the next 50 years (pref sooner, rather than later, of course) to tip things toward the “lower” UN population model.

Hype along the lines of “the trendy doomsayers were wrong (yet again) — we can all breath easy, coz the problem has pretty much gone away” is simply some planting their heads up their respective arses (yet again).

For a reasonable review of the problems and implications of population modelling, albeit one that is critical of UN population modelling methodology, see (e.g.) the IIASA’s publications on the “POP” project (c1995), and/or “The Future Population of the World: What Can We Assume Today?”, W Lutz (ed).

>From an abstract on one relevant Web page:

[The following are certainties:]

(1) World population growth will continue. In as little as the next 35 years, the scenarios developed by the [POP project] predict that world population could either double to reach almost 11 billion, or increase by 50% to just over 8 billion. The ingredients for growth are already in the age structure of today’s world population and make it inevitable for tomorrow.

(2) Today’s industrialised countries will account for a lesser share of the world population. The share of the population living in industrialised regions will fall by more than 1/3 to a mere 14% of the world total by only 2030. Conversely, the share of the developing regions will increase to between 85 and 87%, with that of Africa increasing most rapidly.

(3) All regions will be confronted by population aging. As fertility declines, populations will become older. In Western Europe, the average age will go over 45 years and every 3rd person will be over the age of 60 by 2030. The growing share of elderly in areas with little social support such as North America and sub-Sahara Africa [don’t see these 2 mentioned in the same sentence too often], will create substantial problems for society at large.