A theoretical model of the demographic transition

4Table 1 will lead us to qualify these observations. It is a theoretical model which describes what happens in a population passing through the different stages we have just mentioned in our description of population growth in the industrialized countries. This table presents a series of stable populations corresponding to these different stages, taken from the collection of stable populations calculated by Coale and Demeny at the Office of Population Research of Princeton University [2]. This collection consists of four families – the North, South, East and West models – with a series for males and females in each family. Table 1 was constructed using the female series of the West model. The first line at the top of the table gives the gross reproduction rate per woman and the successive lines correspond to decreasing death rates.

5The starting point is the square corresponding to a gross reproduction rate of 3 daughters per woman and an expectation of life at birth of 20 years.

6Mortality then declines, while fertility remains stable. Keeping to the column where the gross reproduction rate is equal to 3.00, we thus proceed down the squares where expectation of life at birth is successively 30, 40, 50 and 60 years. When it reaches 50 or 60 years, fertility begins to decrease. We consequently change to the columns corresponding to a gross reproduction index of 2.00 and then of 1.00, while at the same time proceeding towards higher and higher expectations of life. The last stable population calculated by Coale and Demeny corresponds to a gross reproduction rate of 0.8 and an expectation of life at birth of 77.5 years (the last square at the bottom right of Table 1). A square has been added at the top right, outside the framework of the table, which corresponds roughly to the fertility and mortality conditions prevailing in the Federal German Republic : gross reproduction rate R = 0.57 and expectation of life at birth for women e₀ = 79.5 years.

Table 1

Stable populations, females : Model West, Coale and Demeny. Composition by broad age groups : 0-29, 30-59, 60 +. Mean age at death (years). Growth rate (p. 1000)

Age groups (years) Federal German Republic R = 0.57 Age structure (0-29) + (0-29) (60 +) (60 +) 0-29 19.86 55.96 30-59 35.49 100.00 181.77 0.44 60 +. 44.65 125.81 Growth rate (p. 1000) r = – 20 R = 3.00 R = 2.00 R = 1.00 R = 0.80 Composition by broad age groups 0-29, 30-59, 60 + e0 Age groups (years) Age structure (0-29) + (0-29) Age structure (0-29) + (0-29) Age structure (0-29) + (0-29) Age structure (0-29) + (0-29) (60+) (60+) (60+) (60+) (60+) (60+) (60+) (60+) 20 0-29 61.80 190.62 30-59 32.41 100.00 208.44 10.69 60 + 5.78 17.82 All ages 99.99 30 0-29 66.16 231.25 30-59 28.61 100.00 249.53 12.65 60 + 5.23 18.28 All ages 100.00 40 0-29 68.75 261.41 30-59 26.30 100.00 280.23 13.88 60 + 4.95 18.82 All ages 100.00 50 0-29 70.52 285.16 59.06 184.91 30-59 24.73 100.00 304.37 14.84 31.94 100.00 213.09 6.56 60 + 4.75 19.21 9.00 28.18 All ages 100.00 100.00 60 0-29 71.90 305.44 60.58 197.65 38.46 96.08 30-59 23.54 100.00 324.85 15.74 30.65 100.00 226.30 6.90 40.03 100.00 149.86 1.79 60 + 4.57 19.41 8.78 28.65 21.53 53.78 All ages 100.01 100.01 100.02 70 0-29 38.85 100.00 31.89 79.65 30-59 38.85 100.00 157.37 1.74 40.04 100.00 149.78 1.14 60 + 22.29 57.37 28.08 70.13 All ages 99.99 100.01 77.5 0-29 30.66 80.11 30-59 38.27 100.00 161.32 0.99 60 + 31.08 81.21 All ages 100.01 Mean age at death Mean age at death Mean age at death Mean age at death Mean age at death (years) and growth rate (p. 1000) Expectation of life at birth All ages 5 + 30 + Growth rate (p. 1000) All ages 5 + 30 + Growth rate (p. 1000) All ages 5 + 30 + Growth rate (p. 1000) All 5 + 30 + Growth rate (p. 1000) 20 21.2 42.4 54.1 – 1.89 30 21.0 42.5 55.5 11.98 40 21.9 43.7 57.3 21.11 50 24.2 46.3 59.6 27.74 37.5 54.5 13.14 60 29.2 51.0 61.8 32.96 44.0 59.0 18.30 63.8 68.6 – 6.00 70 70.8 73.2 – 2.20 73.3 74.7 75.5 – 9.84 77.5 78.8 79.2 79.4 – 8.18

Stable populations, females : Model West, Coale and Demeny. Composition by broad age groups : 0-29, 30-59, 60 +. Mean age at death (years). Growth rate (p. 1000)

e₀ = expectation of life at birth. R = gross reproduction rate per woman

7The following demographic indices which seemed useful for interpreting the table have been added at the bottom : the mean age at death in the different squares for three broad age groups (deaths of persons of all ages, aged 5 and over, aged 30 and over). The population growth rate is shown next to the mean age at death.

8Before commenting on the table, a word should be said about the age groups into which each population is broken down. It is the group aged 30-59 which determined the choice of the other two. The idea is that this age group corresponds to the people responsible for the way a society is run : it will often be referred to as the group of "decision-makers". The group aged 60 and over corresponds to the elderly, while the 0-29s are "youths" in the broadest sense. The latter group includes individuals already involved in working life, but who do not yet hold positions of responsibility.

9These items of information will cast more light on the different stages of the demographic transition examined in Table 1.

10During the first stage, characterized by high mortality and fertility, the proportion of "youths" is high, while that of elderly is low : there are almost 11 times more young than elderly persons. The growth rate is very low [3]. The main concern of the 30-59 year old "decision-makers" is, therefore, the younger generation. For 100 decision-makers, there are 200 young people and only 18 elderly. The latter consequently pose no problem ; being in a very small minority, they can even be treated generously, as this will have little impact on the economy.

11When mortality starts to decline, the basic demographic "environment" does not change. Its main constituent, the younger generation, increases. Whereas the number of elderly people per 100 decision-makers increases only very slightly (18 for an expectation of life of 20 years and 19 for 60 years), the number of young people rises sharply (from 200 to 300 for a corresponding increase in expectation of life). At the end of this phase, there are 16 times more young than elderly persons.

12Throughout this stage, the younger generation thus remains the primary concern of the decision-makers. But a factor appears which will greatly complicate matters for them : population growth rises substantially. Initially very low, it exceeds 3 % per year when life expectancy reaches 60 years.

13The decision-makers stagger when faced with such numbers. It is at this time that the fertility decline gets under way, while mortality continues its downward trend. Columns R = 2.00, R = 1.00 and R = 0.80 in Table 1 show how the pattern changes as both fertility and mortality decrease.

14Before discussing the changes which appear, mention should be made of a phenomenon which was already present during the first stage, and gathers momentum during the second : the increase in mean age at death. In column R = 3.00 which describes the first stage, mean age at death is seen to increase for deaths at all ages, at ages 5 and over, and 30 and over, but only moderately. The elderly have what the decision-makers consider a deplorable tendency to hang on to their capital for longer and longer, but this prolongation is slight and apparently does not greatly affect the economy. The same is not true of the next stage.

15Let us now examine in detail how this stage progresses, as described in columns R = 2.00, R = 1.00 and R = 0.80.

16When fertility declines, the "burden" represented by the younger generation naturally decreases, but at the beginning of this stage, it is still high. It is roughly the same as at the beginning of the previous stage : almost 200 young people per 100 decision-makers. The proportion of the elderly increases : it is now only one-seventh that of the youths (against one-sixteenth at the end of the previous stage). The situation cannot be said to have changed fundamentally : the decision-makers still focus their attention on the younger generation. But what is very different is the decrease in the rate of population growth. The decision-makers can breathe again, the pressure of numbers eases off. On the other hand, the increase in mean age at death, which was already visible during the previous stage, accelerates. Capital transfer occurs later and later. The factors which will characterize the end of the fertility decline stage are beginning to appear.

17Columns R = 1.00 and R = 0.80 and the square corresponding to the Federal German Republic (R = 0.57) describe this terminal stage, characterized by the following phenomena :

18• The proportion of elderly increases while that of young people decreases, so that they become practically equal. Their relative positions are even reversed in the Federal German Republic, where there are twice as many elderly as there are young people. This is a new situation, corresponding to a society which is completely different from that described so far. In this context, the decision-makers are faced with a dual concern, both the young and the elderly, two groups which are at variance in the economy.

19• Mean age at death increases to an extent which is alarming for the transfer of capital, which remains in the hands of the elderly.

20These are the two major factors that characterize this stage. Many others exist, but none has comparable impact. Among these other concomitant factors, a positive one is often mentioned : the drop in the combined proportion of young people and elderly, who together represent only 150 individuals per 100 decision-makers, against 300 during the first stage. The decision-makers, therefore, have to cope with only half the total burden of dependency that their grandparents had to support, a remark which is often made to stop them complaining. But in reality, they care little about the situation their grandparents knew : these are two different generations of decision-makers with different experiences. Those who had to cope with a burden of 300 individuals per 100 decision-makers have long since died, and the new generation has no desire to compare today’s situation with that in the near or distant past. What matters for them is the present, and what they experience is equal proportions of young people and elderly, with in the near future, in the demographic context of the Federal German Republic, twice as many elderly as young.

21Another factor intervenes at the end of the fertility-decline stage, that is favourable at first, but then rapidly becomes unfavourable : the drop in population growth. This is beneficial to the economy, which has to deal with fewer new arrivals. However, if fertility declines sufficiently, the growth rate becomes negative and the population starts to decrease. This does not bother the decision-makers ; only demographers, who are constantly preoccupied with the long run, are disturbed. Indeed, a population which decreases will inevitably become a population which disappears, a situation which is worrying to say the least. They wonder whether, in the future, fertility trends will recover sufficiently to avoid this extinction.

22The decision-makers not only decide how a society is to be run, they also set its fertility level. The fact that they are faced with a situation in which there are as many elderly as young people, or even more, is hardly favourable to a rise in fertility. The decision-makers have no influence over the numbers of the elderly : they are there, one has to accept that fact and provide them with the means to live. But they can influence the number of young people. The only means of lessening the overall burden on society is to lower fertility so as to reduce the proportion of the young, which will simultaneously increase the proportion of elderly, thus further reducing fertility. All the requirements for a snowballing fertility decline are fulfilled.

23The aspects investigated in the present paper are purely demographic. A complete study would, of course, consider savings, and particularly investment, which are in competition with the demands of the elderly and the young. It should also take into account social progress which, all other things being equal, increases the consumption of these two age groups. But the "ceteris paribus" assumption is completely unrealistic. It is well known that, fortunately, advances in productivity lead to other changes. Finally, the voting power of the elderly is incomparably greater than that of the younger generation, and politicians are obliged to take this into consideration.

24P. Bourcier de Carbon, a demographer at INED, has attempted to integrate all these different factors by constructing a demo-economic model [4] which is obviously more complex than that presented in Table 1. However, both models reach the same conclusion : they lead to replacement of the last stage of the demographic transition – a stabilization of the logistic type – by one of "demographic implosion", where fertility continues its decline and never returns to replacement level. In conclusion, this analysis suggests that other scenarios may exist in addition to the classic model of demographic transition. The rest of this chapter will be devoted to examining these different scenarios.

The classic scenario of demographic transition (no.1)

25We begin with the classic scenario of demographic transition, adopted by the United Nations for its population projections. It assumes that all countries in the world will eventually reach the replacement level of 2.1 children per woman. This level will be reached at different dates, but the world population will ultimately stabilize.

26In the less developed countries, fertility has been falling for the last fifteen years, but is still a long way from replacement level, and continuing the decrease down to 2.1 children per woman will present no problem.

27For the industrialized countries, the problem is more complex. Instead of stopping at 2.1 children per woman, fertility in these countries has dropped below replacement level and seems determined to stay there. The demographic momentum accumulated in the age structure means that this persistently below-replacement fertility has not yet resulted in any significant decrease in total population. The industrialized countries will not experience substantial population decline until after the year 2000. The Population Division then proposes the following assumption : faced with this decline, couples will change their behaviour ; to avoid the ultimate disappearance of their country, some of them will public-spiritedly respond by increasing their fertility so that overall fertility will reach 2.1 children per woman. This assumption is of course gratuitous, but does not seem absurd. Once this is accepted, the populations of the different countries in the world can be estimated up to the point of stabilization. These calculations have been made by the United Nations for broad regions and by the World Bank for individual countries [5].

28The results are presented in Table 2. In 1800, North America, Latin America, Africa and particularly North Africa were strikingly underpopulated [6]. When completed, the main outcome of the demographic transition will have been the peopling of these regions which were practically deserted in 1800. (The population multipliers since 1800, in the right-hand column of Table 2, show the extent of this change.) These are followed by Asia, excepting China and Japan. In this region, the population multipliers will be lower, but remain substantial, and above all, they apply to populations which were already considerable in 1800. Finally, for the other regions of the world (Europe, China and Japan), population increase will have been much more modest.

Table 2

Population growth in broad regions of the world, observed (1800-1985) and projected (1985-2100) (millions of inhabitants)

Years Continents regions countries 1800 1939 1980 1985 2000 2025 U.N. limit c. 2100 Multiplier from 1800 to 2100 World 954 2 195 4 453 4 842 6 127 8 177 11 011 11.5 China 330 455 1 003 1 063 1 256 1 460 1 481 4.5 Japan 25 72 117 120 128 128 128 5.1 India, Pakistan, Bangladesh 180 381 864 964 1 250 1 621 2 538 14.1 Rest of Asia 96 254 607 677 910 1 258 1 793 18.7 Asia 631 1 162 2 591 2 824 3 544 4 467 5 940 9.4 Europe 146 403 484 492 513 527 553 3.8 USSR 49 170 265 278 315 367 377 7.7 North America 5 143 252 264 298 347 325 65.0 Oceania 2 11 23 25 30 40 40 20.0 Population of European origin 202 727 1 024 1 059 1 156 1 281 1 295 6.4 North Africa 10 49 108 125 185 295 460 46.0 Rest of Africa 92 126 368 428 692 1 348 2 376 25.8 Africa 102 175 476 553 877 1 643 2 836 27.8 Latin America 19 131 362 406 550 786 940 49.5

Population growth in broad regions of the world, observed (1800-1985) and projected (1985-2100) (millions of inhabitants)

Figure 2

Distribution of the world population among five broad regions, observed (1800-1985) and projected (1985-2100)

Distribution of the world population among five broad regions, observed (1800-1985) and projected (1985-2100)

Table 3

Distribution of world population among broad regions, observed (1800-1985) and projected (1985-2100)

Continents regions countries Years 1800 1939 1980 1985 2000 2025 U.N. limit c. 2100 World 100.0 100.0 100.0 100.0 100.0 100.0 100.0 China 34.6 20.7 22.5 22.0 20.5 17.8 13.4 Japon 2.6 3.3 2.6 2.5 2.1 1.6 1.2 India, Pakistan, Bangladesh 18.9 17.3 19.4 19.9 20.4 19.8 23.0 Rest of Asia 10.0 11.6 13.7 14.0 14.8 15.4 16.3 Asia 66.1 52.9 58.2 58.3 57.8 54.6 53.9 Europe 15.3 18.4 10.9 10.2 8.4 6.5 5.0 URSS 5.2 7.7 5.9 5.7 5.1 4.5 3.4 North America 0.5 6.5 5.7 5.5 4.9 4.2 3.0 Oceania 0.2 0.5 0.5 0.5 0.5 0.5 0.4 Population of European origin 21.2 33.1 23.0 21.9 18.9 15.7 11.8 North Africa 1.1 2.2 2.4 2.6 3.0 3.6 4.2 Rest of Africa 9.6 5.8 8.3 8.8 11.3 16.5 21.6 Africa 10.7 8.0 10.7 11.4 14.3 20.1 25.8 Latin America 2.0 6.0 8.1 8.4 9.0 9.6 8.5

Distribution of world population among broad regions, observed (1800-1985) and projected (1985-2100)

29In Table 3 and Figure 2, we show the populations of these different regions as percentages of the world population. The proportion represented by industrialized countries of European origin, that is, excluding Japan, reaches a maximum in 1939 (33.1 %) and then steadily drops to reach 11.8 % in 2100.

30In Tables 4 and 5 and Figure 3, the populations have been grouped according to four major civilizations :

1. Christendom, represented by the industrialized countries (minus Japan) and Latin America ;
2. China ;
3. Islam ;
4. the rest of the world.

31The last group is very heterogeneous and a more detailed level of analysis would be necessary to describe it. However, this classification into four groups is adequate to reveal remarkably divergent trends. In Figure 3, we show how Islam has boomed, while Christendom, thanks to Latin America, just manages to "save the situation".

Figure 3

Distribution of the world population according to civilization, observed (1980 and 1985) and projected (1985-2100)

Distribution of the world population according to civilization, observed (1980 and 1985) and projected (1985-2100)

The point of view of the gerontologists

32There are two questionable points in the United Nations projections, the rise of fertility in the industrialized countries to recover replacement level, and the assumption concerning future mortality trends. The projections assume that mortality will tend to the same limit in all countries, corresponding to an expectation of life at birth of 75 years for both sexes combined. This level seems somewhat low for a mortality projection up to the end of the 21^st century. It has already been reached in many countries and even exceeded in some. Is it feasible that mortality decline will stop at the level already reached in 1987 in the industrialized countries, and that in 2100 people will live no longer than at the present time ?

Table 4

Evolution of the world population according to civilization, observed (1980 and 1985) and projected (1985-2100) (millions of inhabitants)

Civilization Years Multiplier from 1980 to the U.N. limit 1980 1985 2000 2025 U.N. limit c. 2100 World 4 453 4 842 6 127 8 177 11 011 2.5 Christendom 1 385 1 463 1 703 2 061 2 228 1.6 Islam 800* 857 1 429 2 503 4 412 5.5 China 1 003 1 063 1 256 1 460 1 481 1.5 Rest of world 1 265 1 459 1 739 2 153 2 890 2.3

Evolution of the world population according to civilization, observed (1980 and 1985) and projected (1985-2100) (millions of inhabitants)

^* The estimate of 800 million followers of Islam is given by Nafis Sadik in "Moslem Women Today", Populi, XII, 1, 1985, p. 38.

Table 5

Distribution of the world population according to civilization, observed (1980 and 1985) and projected (1985-2100)

Civilization Years 1980 1985 2000 2025 U.N. limit c. 2100 Christendom 31.1 30.2 27.8 25.2 20.2 Islam 18.0 17.7 23.3 30.6 40.1 China 22.5 22.0 20.5 17.9 13.5 Rest of world 28.4 30.1 28.4 26.3 26.2 World 100.0 100.0 100.0 100.0 100.0

Distribution of the world population according to civilization, observed (1980 and 1985) and projected (1985-2100)

33Gerontologists claim that their science is on the eve of achieving considerable progress in the fight against individual ageing (or senescence). In a recent publication [7], Professor Roy L. Walford of the USA described the situation very clearly. According to this author, the science of ageing is in a pre-revolutionary phase, by which he means a situation where there are several explanatory models : a given model explains almost everything, but not certain facts that other models explain, while these other models in their turn do not explain certain facts that the first model explains. When this type of situation arises in science, a fundamental mutation is about to occur. In the near future, all these models will be merged into a single one, which will provide us with the means to act on the ageing of our organism. During a first stage, expectation of life would approach the limit of human life, which in homo sapiens is around 115 years, a figure which has scarcely altered in the course of time. The gerontologists assert that this limit could be pushed back during a second stage and that an expectation of life of 150 years no longer seems implausible for the end of the 21^st century : indeed, it would be the norm. The drawbacks of old age would be the same and would appear, as now, during the last ten years of life. Man could thus look forward to 140 years of healthy life. Such progress would naturally affect the very foundations of our society. It would also modify world population size, which leads us to our second scenario.

Mastering senescence (scenario no.2)

34This new scenario is similar to that adopted by the United Nations. We continue to assume that fertility in all countries will stabilize at replacement level, that is, 2.1 children per woman. We will, therefore, accept the assumption that a certain proportion of couples in the industrialized countries will modify their reproductive behaviour upon seeing their country’s population decline, and that this reaction will bring fertility back to a mean level of 2.1 children per woman. But this time the mean expectation of life at birth will be assumed to tend towards a limit (100 years) much higher than that adopted by the United Nations (75 years).

35As a first stage, we will examine what the pattern of population growth in Europe would be if the fertility level in all countries immediately dropped to a value close to that experienced in the Federal German Republic and Italy (1.5 children per woman) and if expectation of life at birth started to increase after the year 2000, reaching a value of 100 years in 2050 and then remaining at this level [8]. (This assumption is well below the optimistic forecasts of Professor Walford.) According to these calculations, the population of Europe would not start declining until after 2050. The reaction of couples to this decline would, therefore, only occur after this date, much later than in the United Nations scenario. Furthermore, after maintaining a very low fertility level for a long time, the European countries would in 2050 have accumulated such a potential for negative growth that their populations would continue to decline for some time, even if replacement level fertility were achieved. Finally, the calculations show that the population of Europe would stabilize around 2125 at the size reached in 1939. For our scenario, we assume that these results apply to all presently industrialized countries.

36The less developed countries would probably also gain by the reduction of mortality and would also reach an expectation of life at birth of 100 years. This would not affect their progress towards a fertility of 2.1 children per woman. Only the limit of their population size would be changed. This can be roughly estimated as being proportional to the expectation of life, in other words, when expectation of life increases from 75 to 100 years, this limit is raised by a third.

37This leads us to Tables 6 and 7, which show, for the different regions of the world, the population limits under the second scenario ; the corresponding limits under the United Nations scenario are presented for comparison. The pattern observed in the United Nations scenario is accentuated in the second. In particular, the proportion represented by the industrialized world further decreases, to 5.5 % against 11.8 % in the first scenario. Islam strengthens its position (43.3 %), while Christendom accounts for a mere 14.6 % ; in 1985, the proportions were 17.7 % and 30.2 % respectively. Such changes would certainly create many problems.

Table 6

Limits of stabilization of the world population according to two scenarios (millions of inhabitants)

Continents regions countries U.N. limit eo = 75 years Limit with eo = 100 years Distribution (%) eo = 75 years eo = 100 years World 11 011 13 583 100.0 100.0 China 1 481 1 975 13.4 14.6 Japan 128 72 1.2 0.5 India, Pakistan, Bangladesh 2 538 3 384 23.0 24.9 Rest of Asia 1793 2 391 16.3 17.6 Asia 5 940 7 822 53.9 57.6 Europe 553 403 5.0 3.0 USSR 377 170 3.4 1.3 North America 325 143 3.0 1.1 Oceania 40 11 0.4 0.1 Industrialized countries of European origin 1 295 727 11.8 5.5 North Africa 460 613 4.2 4.5 Rest of Africa 2 376 3 168 21.6 23.3 Africa 2 836 3 781 25.8 27.8 Latin America 940 1 253 8.5 9.2

Limits of stabilization of the world population according to two scenarios (millions of inhabitants)

Note : For the industrialized countries the limit with e_o = 100 years corresponds to population size in these countries in 1939. This limit is certainly too low for North America and Oceania, which in 1939 had not yet experienced the immigration which appeared after world war II.
For the less developed countries the limit with e_o = 100 years is equal to the limit with e_o = 75 years increased by a third.

Table 7

Limits of stabilization of the world population, by civilization, according to two scenarios (millions of inhabitants)

Civilization U.N. limit eo = 75 years Limit with eo = 100 years Distribution (%) eo = 75 years eo = 100 years World 11 011 13 583 100.0 100.0 Christendom 2 228 1 980 20.2 14.6 Islam 4 412 5 883 40.1 43.3 China 1 481 1 975 13.5 14.5 Rest of world 2 890 3 745 26.2 27.6

Limits of stabilization of the world population, by civilization, according to two scenarios (millions of inhabitants)

38The transformation of the Mediterranean basin area is absolutely explosive (see Table 8). Within the space of three centuries, the respective weight of the South-and-East and North-and-West regions is reversed and the global population of the Mediterranean basin increases 11-12 fold. Whereas in 1800, there were 20 Moslems for 80 Christians, at the end of the projection there are 70-84 Moslems for 30-16 Christians. The need for a collaboration, indeed a reconciliation, between the two religions has never been so clearly spelt out. This may, and no doubt must, pass through a process of economic cooperation. The States concerned have 150 years to come up with the necessary solutions.

The demographic implosion (scenario no.3)

39The next scenario is the "demographic implosion" suggested by Table 1. Fertility continues to decline in the industrialized world, reaching very low levels, below replacement, which is never recovered. The less developed countries follow the same pattern, but with a certain delay. The result is obviously catastrophe : sooner or later, the human race will disappear. To quantify these somewhat vague assumptions, the following suppositions were made : the situation in all countries is assumed to tend towards that in which fertility stabilizes at the level current in the Federal German Republic, that is, a total fertility in the region of 1.2 (gross reproduction rate slightly below 0.6). With current mortality, this fertility level would lead to an intrinsic rate of natural increase of – 2 % per year (with a crude birth rate of 4.7 p. 1,000 and a crude death rate of 24.7 p. 1,000). Once these assumptions were put, the projections could have been calculated by following arbitrary courses towards this limit. It was considered more plausible to adopt the courses already observed.

Table 8

Mediterranean population growth since 1800 (millions)

Region 1800 1939 1985 2000 2025 Limits Multipliers from 1800 to the limits eo = 75 years eo = 100 years eo = 75 years eo = 100 years South and East a) 13.4 56.8 168.0 234.4 340.8 510.0 678.0 38.1 50.6 North and West b) 51.9 133.0 183.5 190.1 187.0 211.0 133.0 4.1 2.6 Total 65.3 189.8 351.5 424.5 527.8 721.0 811.0 11.0 12.4 Distribution (%) South and East 20.5 29.9 47.8 55.2 64.6 70.7 83.6 North and West 79.5 70.2 52.2 44.8 35.4 29.3 16.4 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0

Mediterranean population growth since 1800 (millions)

a) Algeria, Egypt, Libya, Morocco, Tunisia, Lebanon, Israel, Turkey
b) France, Spain, Italy, Greece, Yugoslavia
Notes :
1) For 1985, 2000 and 2025, the figures are taken from World Population Prospects. Estimates and Projections as Assessed in 1984. United Nations Publication no. E 86-XIII-3. The medium variant was adopted for the South-and-East region, the low one for the North-and-West region.
2) For 1939, the figures are taken from the United Nations Demographic Yearbook, 1948. United Nations Publication no. 1949-XIII-1. For the South-and-East region, data are estimates, as certain countries are lacking in the 1948 Yearbook.
3) For 1800, the figures are estimates calculated as follows from J.-N. Biraben, "Evolution du nombre des hommes", Population, 1, 1979, p. 16 :
a) For 1950, J.-N. Biraben gives a figure of 52 million inhabitants for North Africa, and the United Nations give 69.5 million for the South-and-East region
We calculated the ratio :
which we applied to the population of North Africa given by Biraben for 1800 : 10 × 1.337 = approx. 13.4 million.
b) For 1950, J.-N. Biraben gives 395 million for Europe excluding the USSR and the United Nations give 140.4 million for the North-and-West region
We calculated the ratio :
which we applied to the population of Europe excluding the USSR given by Biraben for 1800 : 146 × 0.3552 = approx. 51.9 million.

40Figure 4 explains the method used. It shows the mean annual rates of growth by five-year periods between 1950 and 2025, for various regions and countries. From 1950-1955 to 1975-1980, these rates are those actually observed. Between 1980-1985 and 2020-2025, they are United Nations projections. The low variant has been adopted for the industrialized countries and the medium one for the less developed countries. We observe that the former set of countries – represented by curve B – is at the same level in 2020-2025 as Western Europe was in 1980-1985. We thus assume that after 2025, it will follow the same trend as Western Europe after 1985. This brings it by 2060-2065 to roughly the same level as the Federal German Republic in 2010-2015. We then suppose that curve B will follow the same trend as in the FGR, and project it accordingly to 2075-2080. At this date, the growth rate is – 1.13 %, fairly close to the limit of – 2 %. Curve B was extended freehand to reach this rate of – 2 % in 2105-2110.

Figure 4

United Nations. World population prospects (Estimates and projections as assessed in 1984). Projections extended to 2075-2080. Rates of population growth by five-year periods (%)

United Nations. World population prospects (Estimates and projections as assessed in 1984). Projections extended to 2075-2080. Rates of population growth by five-year periods (%)

41In Figure 4, the hatched zone indicates corrections made after the year 2000 for Western Europe and the Federal German Republic. These are explained in Figure 5, which is identical to Figure 4, except that it shows the birth rate instead of the growth rate. Since in the United Nations projections, the reaction of couples in Western Europe and the Federal German Republic to low fertility is expected around 2000, when population decline becomes substantial, fertility is made to rise immediately after 2000. This increase naturally disappears from our third scenario, where the birth rates continue to decrease after 2000. We thus extended the downward trend after the year 2000, obtaining a corrected birth rate curve [9]. The corresponding correction was made to the growth rate.

Figure 5

United Nations. World population prospects (Estimates and projections as assessed in 1984). Projections extended to 2075-2080. Crude birth rates by five-year periods (p. 1,000)

United Nations. World population prospects (Estimates and projections as assessed in 1984). Projections extended to 2075-2080. Crude birth rates by five-year periods (p. 1,000)

42Let us now consider the less developed countries. At the end of the United Nations projections, in 2020-2025, the growth rate in these countries is 1.1 p. 1,000. When the trend is extrapolated freehand to 2025-2030, the rate is in the region of 0.9. This is the rate experienced in the industrialized countries 45 years earlier, in 1970-1975. We therefore assumed that the less developed countries would follow the same trend as the industrialized ones, 45 years later.

43The results presented in Figure 6 show that, although both sets will eventually meet the same fate, the industrialized countries will become extinct long before the less developed ones. The growth potential of the latter is adequate to continue their upward trend until 2080, when the population peaks at 8,400 million. The population of the industrialized countries peaks at 1,400 million in 2020, and the world population at 9,400 million in 2070.

Figure 6

World population growth in the scenario where fertility drops to the F.G.R. level then stabilizes

World population growth in the scenario where fertility drops to the F.G.R. level then stabilizes

44Figure 7 shows how the proportion of world population in each group changes. In 2150, the industrialized countries represent only 5 % of the total world population ; stability having been reached, this proportion then remains constant.

Figure 7

Distribution of the world population between industrialized and less developed countries in the situation described in Figure 6

Distribution of the world population between industrialized and less developed countries in the situation described in Figure 6

45The third scenario, using the United Nations population projections, implicitly supposes that the expectation of life at birth will reach a ceiling at 75 years. What would happen if this were increased to 100 years, as assumed in the second scenario ? In this case, the general aspect would not change. The mortality decline would delay the decrease in population size, but not arrest it. The industrialized countries’ peak would be replaced by a plateau, while that in the less developed countries would be higher. But the final outcome would still be the extinction of the human race.

46In an article entitled "Essai sur l’évolution du nombre des hommes" [10], Jean-Noël Biraben has attempted to reconstruct world population growth since the very beginnings of our species. He proposes estimates of total world population between 40,000 BC and 1970, and regional estimates for 400 BC to 1970. It is interesting to examine how the trends constructed by Biraben are extended by our three scenarios. Figures 8 and 9 show that, with the "catastrophic" scenario, they describe a phenomenon which is very similar to the life of a star. After glowing modestly for millions of years, its brilliance is suddenly magnified on a gigantic scale : this is what astronomers term a supernova. But this does not last for long before the star fades rapidly and disintegrates. What happens during this last stage is a mystery : for astronomers, the star has simply disappeared, they can no longer trace it on their photographs. In the language of astrophysics, the world population would be about to reach its maximum brilliance (9,400 million in 2070) before disappearing.

Figure 8

Estimation of world population since 1600 BC

Estimation of world population since 1600 BC

Figure 9

Estimation of the evolution of world population since 40,000 BC

Estimation of the evolution of world population since 40,000 BC

The successive demographic transitions

47On the growth curves proposed by Biraben, the successive demographic transitions experienced by mankind are clearly visible. The first of these occurred between 40,000 and 35,000 BC. The world population then increased tenfold, from 500,000 to 5 million, and remained at this level for over 20,000 years. Experts in paleo-demography continue to puzzle over what caused this demographic transition during the Higher Paleolithic. A change in climatic conditions which made more food available is one possible explanation. Others have been put forward : Alain Testart [11] recently demonstrated that before turning to agriculture, at least some of the Higher Paleolithic populations had discovered how to preserve food. This represented a decisive progress for man, as it freed him of his dependence on seasonal fluctuations in food supplies. This might explain the tenfold population increase [12]. Moreover, preservation paved the way for farming : it required at least the beginnings of social life. The next step was to cultivate foodstuffs for preservation and to tether animals instead of hunting them and… behold the sedentary farmer. This is the second demographic transition shown by Biraben, which appeared between 10,000 and 5,000 BC. Noah symbolizes this change, stepping off the ark to plant a vine : a hunter-gatherer boarded the ark, a farmer landed. This transition continues up to the second half of the 18^th century. The world population grew from 5 million to almost 1,000 million (771 million in 1750 and 954 million in 1800), in other words, it was multiplied by 200 during the second transition, compared with 10 during the first.

The industrial revolution and human rights

48It was in the second half of the 18^th century that the third demographic transition, the final stages of which we have just experienced, made an appearance in Europe. Like the earlier ones, it was related to a technical progress : in this case, the discovery of how to harness energy. In France, per capita consumption of coal-equivalent soared from 70 kg in 1830 to 5,000 in 1987. This new power allowed farmworkers to increase their productivity, and consequently some of them became available for work in industry.

49None of the major demographic phenomena described to characterize what is termed the industrial revolution could have occurred if energy resources had remained at their pre-1800 level. The industrial revolution also marked a revolution in human rights. The philosophers of the Enlightenment laid down the principles of a new society, which continue to inspire the industrialized states today. But to adopt the principle that every human being is free to rule his own destiny, is to run the risk that the human race will one day disappear. Technical progress has made it possible to determine the size of one’s family and, if man is free, he has the right to choose not to survive.

50This brings us to the fourth demographic transition, now known as the post-industrial revolution, which our three scenarios have attempted to describe. In the third scenario, mankind starts from nothing, in 600,000 BC, and returns to nothing, around the year 2400.

The total number of human beings in the "catastrophic" scenario

51It is tempting to try and estimate the number of human beings ever born at the time when the last will die. This is very simple when both population estimates and birth rates are available : the former are multiplied by the latter and the results are summed. From 40,000 BC to 1750, we have the estimates proposed by J.-N. Biraben. For this period of 41,750 years, we assumed a birth rate of 40 p. 1,000. After 1750, we took the United Nations estimates of populations and crude birth rates for industrialized and less developed countries separately [13]. This provided a series up to 1987, which we have completed with the estimates of our third scenario.

52From 40,000 BC to 1987, the results are the following [14] :

53

54From 1987 to the last man on earth, we have estimated the number at 16,800 million.

55A link is missing in the chain : the whole of the period between the birth of mankind and 40,000 BC. The beginning of this period can be put at c. 600,000 BC. The world population in 40,000 BC was around 500,000, and in 600,000 BC, only a handful of individuals. Assuming exponential growth, the population between 600,000 and 40,000 BC would be estimated at between 0,900 and 1,800 million, depending on whether the population in 600,000 BC is taken to be 1, 50 or 500 couples. But it is risky to suppose that growth was exponential. The growth curve presented by Biraben shows sudden rises corresponding to technical progress, separated by long periods of stagnation. There is no reason to suppose that the pattern was any different before 40,000 BC. A technical progress of primary importance occurred between 600,000 and 40,000 BC, the discovery of fire. This no doubt had similar demographic effects to such innovations as preservation and agriculture. As it occurred c. 200,000 BC, we can assume that the world population varied little between 200,000 and 40,000 BC, in the region of 500,000 individuals. Between 600,000 and 200,000 BC, in the pre-fire period, we can suppose that it was as low as 40,000. With these assumptions, the summed total can be estimated at 3,840 million births between 600,000 and 40,000 BC. From the origins of man to 1987, the number of individuals would thus be :

5676,510 + 3,840 = 80,350 million.

57If we add the 16,800 million after 1987, we obtain a total of 97,150 million individuals from the beginning to the end of mankind.

58The United Nations Organization recently invited the world to celebrate the 5,000 millionth citizen of the world, and chose 11 July 1987 to mark this occasion. It would have been more accurate to say that, in 1987, the 80,000 millionth human being would be born and the 75,000 millionth would die, leaving a population still alive of 5,000 million.

Professor Winkler’s error

59Professor Wilhem Winkler unfortunately succumbed to the temptation of assuming exponential growth for the world population. At the International Population Conference held in Vienna (Austria) in 1959, he presented a paper entitled "How many human beings have lived in the world up to the present day ?". He assumed that the world population had followed an exponential growth pattern from a handful of individuals in 600,000 BC to the population in 1959, estimated at the time at 2,750 million. He presented the population growth curve as a straight line on a semi-logarithmic graph : in fact, three straight lines, depending on whether the initial population in 600,000 BC was taken to be 1, 50 or 500 couples (Figures 8 and 9). The three lines, which on this scale practically merge into one, are well above the growth curve proposed by Biraben.

60Professor Winkler [15] thus obtained a number of human beings born between 600,000 BC and 1959 which was substantially higher than the estimate obtained by using Biraben’s figures :

61

62Compared with our calculations which give a figure of 77,950 million between 600,000 and 1959, the error introduced by assuming exponential growth is impressive. In an article published in the Population Bulletin of February 1962 by the Population Reference Bureau of Washington DC (USA), Flechter Wellemeyer and Frank Lorimer estimated the number of human beings born between 600,000 BC and 1962 at 77,000 million, a figure which is very close to our own.

63The 97,150 million we calculated for the whole of mankind from its origins to its extinction were obtained by assuming a constant birth rate of 40 p. 1,000 between 600,000 BC and 1750. With a rate of 50 p. 1,000, we would have obtained 113,520 million. The total number of human beings ever born in the third scenario is therefore in the region of 100,000 million [16]. This is relatively low : other species have been much more prolific.

The yoke of the menopause

64Must the third scenario necessarily lead to the extinction of the human race ? Could not the human brain, so inventive, find a means of saving the species ? Concerning Professor Walford’s hopes that progress in the field of senescence would extend expectation of life to 150 years, we said that this would not fundamentally change scenario 3, but only delay extinction.

65However, mastering senescence means progress in another field : the ageing of women’s reproductive capacity. Out of 500,000 potential ova at birth, only 500 reach maturity, all the others being lost. It is an ageing process which occurs early in life, extending from puberty to the menopause. It is no doubt because of this that it has been neglected by gerontologists.

66But if the latter succeed in extending life as far as they hope, their discoveries could have some repercussions on ovarian ageing and it may become possible to throw off the yoke of the menopause [17]. In this case, the conditions of human reproduction will be totally changed. With 140 years of good health and the possibility of childbearing up to age 100, even with barely more than one child per couple, the extinction of the human race can be avoided. If each woman contracts two unions and gives birth to 1.2 children in each – which corresponds to current fertility in the Federal German Republic – a total fertility rate of 2.4 is reached, well above the replacement level of 2.1 children per woman.

67This could be the essence of the post-industrial demographic revolution. After the discoveries of fire, the preservation of food, agriculture, harnessing energy : the mastering of biology. Five points of no return for mankind [18]. The next stage could be to people the Universe ; but this is the realm of science fiction…