WARS, unemployment, poverty, and many other evils under which humanity suffers are usually attributed to over-population. If there were less people there would be no need for territorial expansion, and everyone might get a larger share of the social product. But mankind seems obsessed by the desire to propagate itself beyond any reasonable limits. It grows more and more rapidly. At its present rate of increase, we are told, the world's population will have reached 5,000 millions before the end of this century, 5,000 millions being at the same time the absolute maximum which the planet can sustain.

There are indeed plenty of authorities to corroborate this belief. In his recent book "Standing Room Only?" Edward Alsworth Ross could point to the fact that the International Statistical Institute reported that from 1920 to 1924 the population of the world had shown "an increase of 103,378,000, or 5.77 percent -- a rate which would double mankind in a half-century." The Statistical Institute three years ago estimated that the world's population in 1920 was 1,791,496,000, and that in 1924 it was 1,894,874,000. But it has since raised its estimate for the year 1920 to 1,811,012,000, and has made a new estimate of 1,879,595,000, for 1926; this is smaller than the one given for 1924, so that it now shows the six years' increase to have been 68,583,000, or 3.79 per cent only -- a rate which would double mankind in about 110 years.

The maximum of 5,000 millions which the world can sustain is the estimate of as high an authority as E. M. East, who, in a report submitted to the World Population Conference in Geneva, declared: "The effort of the human race to expand its numbers is limited to the produce of about 13,000 million acres of tillable soil, two-fifths of which is now under cultivation. And since it takes at least two and a half acres to support each individual under the present standards of agricultural efficiency, it is clear that the world can sustain only 5,000 million people, unless unforeseen radical discoveries in science bring about revolutionary changes in our economic system." As a matter of fact, there are now a little over two and a half acres under cultivation for each individual, but can this ratio actually not be reduced unless "radical discoveries" bring about "revolutionary changes?" Is it not a fact that a country like France, with a rather mediocre standard of agricultural efficiency, has one and one-half acres only under cultivation per individual, and that the ratio is still lower in Germany? Even granting that there be not more than 13 billion acres of tillable soil, the maximum population would not have to be placed at less than 10 billions. And with a rate of increase such as that ascertained by the International Statistical Institute for 1920-1926, this limit would not be reached before the year 2200.

But, whether in 2000 or in 2200, what would be the outcome? Professor Ross gives a plausible answer: "Nature, to be sure, offers her solution -- simple, ruthless, effective. When Food can no longer keep up with Population, privation and toil will raise the death-rate, as they have raised it a thousand times in the past, until Life and Death are once more in balance." He also gives the remedy: "If posterity recoils from this gloomy prospect, then the voluntary restriction of increase must become general. Why wait till there is little left to salvage? If eventually, why not now?"


The logic seems perfect. Since, with the limited possibilities of food supply, the present rate of increase cannot go on forever, there seem only two alternatives: a rise in mortality or a reduction of fertility until life and death once more balance. There is only just one gap in this argumentation. It has not yet been proved that with present fertility and present mortality life and death are out of balance. Some readers may assume, like Prof. Ross, and Prof. East, that more proof is unnecessary. Even if the world's population has increased from 1920 to 1926 by 68,583,000 only, there must have been an average yearly excess of births over deaths 11,430,000, that is to say of 0.62 percent of the population. Does this not prove beyond doubt that fertility far exceeds mortality?

It proves, indeed, that births amply keep up with deaths. But this does not imply that the reproduction of the world's population is still ample. If the newly born were merely to replace the dead all that would be necessary would be that births equal deaths, and if no death occurred no birth would be needed. This consideration in itself shows that something must be wrong with the usual comparison of births and deaths. For if in a given population no death occurred and no birth, this population would ever grow older and after fifty years there would be no more women of child-bearing age and no more men with full physical working capacity. The total population would be as large as fifty years earlier, but in those fifty years it would not have done anything towards its reproduction and it would have lost any future chance of reproduction.

Let us assume that the yearly excess of births (11,430,000) came about by 81,430,000 births and 70,000,000 deaths. The reaction of public opinion probably would be that this means an enormous waste of human energy and that it would be far better if both births and deaths were considerably less. News that there actually had been only 31,430,000 births and 20,000,000 deaths would be heartily welcomed. People would not realize that with so small a number of births the population of the world would be bound to die out.

In looking back, we often are surprised at the lack of understanding displayed by our forefathers when dealing with certain problems which to us seem very easy to solve. If there is one thing certain it is that future generations will be amazed by the utter helplessness with which in the second quarter of the twentieth century people still attacked the population problem. They will hardly be made to believe that our generation did not know that, however low may be the number of deaths, there must be a definite number of births in order to insure the reproduction of the population. Everyone with common sense, they will say, might have realized that on the average each woman must have two children who should become parents of two children, etc., if the population is not sooner or later to decrease.

Unfortunately, we know very little about the yearly number of births in the world. We do not know whether it is nearer to 81,430,000 or to 31,430,000. Even the total number of inhabitants is unknown. There exist, it is true, estimates for 1926 by the three highest authorities in this matter, and these do not differ much between themselves. They are as follows: International Statistical Institute 1,879,595,000; International Institute of Agriculture 1,894,979,723; League of Nations, Economic and Financial Section 1,926,715,000. But these three agencies assume the population of China to be 433 or 450 millions, while according to a careful study by Walter F. Willcox it does not exceed 300 millions. The world's population, then, is perhaps not higher than about 1,750 millions.

The available data are thus not sufficient to allow a judgment upon the vitality of the world's population as a whole. All we can say is that the yearly excess of births over deaths, amounting to 0.62 percent, does not prove anything either way.

Fortunately, there is one section of the world, namely Western and Northern Europe, for which the necessary data are available and have been adequately analyzed.[i] This area[ii] in 1926 had a population of 188,267,000, that is, 10 or 11 percent of the world's population, and had an excess of births over deaths amounting to 0.62 percent, that is, exactly the same rate as the International Statistical Institute set for the whole world.


Around 1850, 3½ million children a year were born in Western and Northern Europe. Fifty years later the annual number of births was 4¾ millions. At present it is again about 3½ millions. In the meantime the population has considerably increased -- from 113 millions in the middle of the nineteenth century to 189 millions in 1927. The birth rate (that is, the number of births per 1,000 inhabitants) therefore has considerably decreased: from 1841 to 1885, it averaged about 32; by 1913 it had fallen to 24; in 1927 it was 18 only. The decrease of the birth rate began in France but involved sooner or later every country so that there is no longer any marked difference between the birth rates of the various parts of Western and Northern Europe.

The birth rate shows the percentage by which a population increases through the births of children, but since it is calculated without regard to the sex and age composition of the population it does not afford an adequate gauge for the measurement of fertility. If in a given population the percentage of women of 20-45 years of age is large, the birth rate is likely to be high even if fertility is small, while if the percentage of women in those age groups is small, the birth rate is likely to be low even if fertility is large. It then becomes necessary to compute fertility rates by the ages of mothers and on this basis to derive the total number of children born to each woman.


Years Denmark Norway Sweden Finland England France Germany
1871-1875 .... } 4.7 4.4 4.9 .... .... ....
1876-1880 } 4.6 4.4 5.0 .... .... ....
1881-1885 .... 4.3 4.9 .... .... } 5.1
1886-1890 } 4.4 } 4.4 4.2 4.9 .... ....
1891-1895 4.0 } 4.7 .... 3.0 } 4.9
1896-1900 4.2 } 4.3 4.0 .... } 2.9
1901-1905 4.0 3.9 } 4.4 .... } 4.4
1906-1910 3.8 } 3.8 3.7 .... 2.7
1911-1913 } 3.4 } 3.3 } 3.6 .... 2.5 ....
1914-1915 .... .... } 1.6 ....
1916-1919 } 3.1 } 3.4 } 2.9 .... ....
1920 .... } 2.6 ....
1921 } 2.9 .... 2.8 } 3.2 2.7 ....
1922-1924 .... .... .... } 2.4 ....
1925 .... .... 2.2 2.3
1926 2.6 2.7 2.3 2.9 2.2 2.4 2.2
1927 2.5 2.6 .... .... 2.0 2.3 2.1

Forty or fifty years ago the average number of children born to each woman (married and unmarried) was four or five in all countries of Western and Northern Europe, with the exception of France and Ireland where it was about three. In 1927, it was practically everywhere below three and averaged about 2.2. In England it was only two. Yet the number of births in England was not less than 655,000 as compared with 485,000 deaths. It will certainly seem at first sight that an excess of 170,000 births is a proof of considerable vitality. As matters stand, those 655,000 births mean that on the average each woman during her life-time gives birth to but two children, and of the children thus born not one may die before reaching marrying age, if the population is to hold its own. But the assumption that no child would die before reaching marrying age is obviously wrong.


A simple logical consideration has shown us that the balance of life and death cannot be established by comparing the yearly numbers of births and deaths. England is one specific case of how misleading such a comparison is. England in 1927 had 655,000 births and 485,000 deaths. It is very likely that further improvements in public health may still further reduce mortality. But no matter how low it may be reduced, the population of England is bound to die out if fertility remains as it is.

The pertinent question is not: "Is there an excess of births over deaths?" -- Rather is it: "Are fertility and mortality such that a generation which would be permanently subject to them would during its lifetime, that is, until it has died out, produce sufficient children to replace that generation?" If, for instance, 1,000 newly born produce in the course of their lives exactly 1,000 children, the population after the death of the older 1,000 will remain unaltered. If fertility and mortality continue to be what they were, the 1,000 children will in the course of their lives again produce 1,000 children, and if fertility and mortality remain permanently the same, the population will always exactly hold its own. If more than 1,000 children are produced by a generation of 1,000 newly born, the population will increase; if less than 1,000 are produced, the population will decrease and finally die out.

Since we are concerned here with birth-giving only, it suffices to take into account the female population. The pertinent question then is: "Are fertility and mortality such that 1,000 newly born girls will in the course of their lives give birth to 1,000 daughters?" If this is the case, the first generation of 1,000 females will at its death have been fully replaced by the daughters they have borne, and the population will remain constant; otherwise, it will in the long run increase or decrease.

The table on the next page shows the net reproduction rate, that is, the number of daughters borne by each newly born girl during her lifetime or -- to put it perhaps more simply -- the number of future mothers born to each woman.

Up to about 1910 the decrease of fertility was offset by the decrease of mortality; the number of girls born to each woman, and who became mothers in their turn, was about 1.4 or 1.5 in all countries of Western and Northern Europe, with the exception of France and Ireland where it was about one. But since 1910 the net reproduction rate has decreased at a terrific speed. In 1926 and 1927 it was still above one in some of the smaller countries, like Holland and Denmark, but it was below one in all the larger countries -- in France, and especially in England and Germany. In all Western and Northern Europe the net reproduction rate was about 0.93 in 1926 and about 0.87 in 1927.

If each woman has one daughter who becomes a mother in her turn, the population will hold its own. If she has 1.5 such daughters, the population will increase by one-half in one generation. If she has less than one such daughter, the population will sooner


Years Denmark Sweden Finland England France Germany
1881-1884 .... } 1.455 } 1.485 .... .... } 1.448
1885-1890 } 1.463 .... ....
1891-1894 } 1.433 .... .... .... } 1.512
1895-1897 } 1.509 .... .... ....
1898-1900 .... .... } 0.979
1901-1903 } 1.524 } 1.427 } 1.433 .... } 1.480
1904-1905 .... ....
1906-1907 } 1.486 .... ....
1908-1910 .... } 0.930 ....
1911-1913 } 1.372 } 1.288 } 1.161 .... ....
1914-1915 .... .... ....
1916-1919 } 1.228 } 1.111 .... .... ....
1920 } 0.994 ....
1921 } 1.192 } 1.137 } 1.146 1.087 ....
1922 .... } 0.937 ....
1923-1924 .... .... ....
1925 .... .... 0.937
1926 1.097 0.95* 1.09* 0.88* 0.94* 0.89*
1927 1.03* .... .... 0.82* 0.91* 0.83*
  * Preliminary figures.

or later decrease. With a fertility and a mortality such as prevailed up to about 1910, the population then was to increase by about one-half per generation in all countries of Western and Northern Europe, with the exception of France and Ireland where the population about held its own. With a fertility and a mortality such as prevail at present, the population of some smaller countries still shows a real surplus, but the population of the larger countries is doomed to die out.


In 1926, the number of births in Western and Northern Europe (3,613,000) exceeded the number of deaths (2,449,000) by 1,164,000 or 48 percent. How is it to be explained that in spite of such a surplus of births the population did not reproduce itself but had a virtual deficit of about 7 percent? How is it to be explained that with a birth rate of 19.2 and a death rate of 13.0 the population does not hold its own? The answer to these most pertinent questions is to be found by a study of the age composition.

In the present population of Western and Northern Europe the proportion of women in child-bearing age is particularly large and the proportion of young children and old persons particularly small. If we exclude the influence of migrations, the population of a given territory is equal to the number of persons born in that territory in the last hundred years minus the number of persons who died in that territory in the last hundred years. If the number of births in those hundred years remains constant or increases, and if the number of deaths does not fluctuate conspicuously, there will be more children under 5 years than from 5 to 10 years, more children from 10 to 15 years than from 5 to 10 years, etc. In the actual population of Western and Northern Europe in 1926 there are more children of 10 to 15 years than of under 5 years or of 5 to 10 years; there are more persons of 15 to 20 years than in any lower age group; there are more persons of 20 to 30 years than under 10 years; and there are very few persons over 65 years. If we compare the age composition of the two sexes, we find a small excess of males in the lowest four age groups, while in all other age groups there is an excess of females, which is especially large in the ages of 25 to 45 and of over 60 years.

The decisive factor in shaping the age composition of the present population of Western and Northern Europe was the trend of births. The average yearly number of births, which had been 3,481,000 in 1841-45, increased to 4,210,000 in 1871-75 and to 4,686,000 in 1901-05, then dropped to 4,337,000 in 1911-14 and to 3,916,000 in 1920-26, after having been as low as 3,064,000 in 1915-19. It is evident that with such a trend in the number of births there must at present be comparatively few children under 5 years (born in 1921-25) and very few children from 5 to 10 years (born in 1916-20), while there should be very many persons of 15 to 50 years (born in 1866-1910) and comparatively few older persons (born before 1866). This is actually the case for the female sex but not for those age groups of the male sex which have been decimated through the war, that is the present age groups from 25 to 45 years.

The age composition of the population of Western and Northern Europe tends to lower the number of deaths. Mortality is everywhere very high among the youngest children and among the oldest persons, while it is practically negligible in the age from two or three until about 50 years. Since at present the number of young children and of old persons is small, the death rate necessarily must be low. It actually was not higher in 1926 than 13 per 1,000. But this low rate cannot possibly last with present mortality. The persons between 15 and 50 years, who now are so numerous, will grow older and will thereby swell those age groups where death claims most victims, while there are not sufficient children to fill up the age groups which are more or less secure against death. The present death rate of 13 per 1,000 is, therefore, taken by itself, misleading. That it cannot permanently last can moreover be easily realized by a simple logical consideration. A death rate of 13 per 1,000 means that 13/1,000 or 1/77 of the population die within a year, and if such a rate were permanent it would mean that the average length of life is 77 years. But the length of life, of course, is actually much lower in every country of Western and Northern Europe. Even in Denmark, with its exceptionally low mortality, the mean length of life in 1921-25 was 61 years. Denmark in that period had 11.3 deaths per 1,000 inhabitants, but the death rate derived from the actual mortality in the individual years of age was 1/61 or 16.4 per 1,000. The difference between the crude death rate, that is the number of deaths per 1,000 inhabitants, and the corrected death rate, that is the rate derived from the mortality in the individual years of age, is not as large in all countries of Western and Northern Europe. It is rather small, for instance, in France where the number of births has not changed very much in the course of the last forty years. But there the crude death rate is comparatively high; it amounted to 17.2 in 1921-25 and to 17.5 in 1926. The main reason is that in France the number of old people is comparatively high. In spite of the fact that mortality in France has by no means been particularly favorable, the percentage of persons over 50 years is 25 percent as compared with 20 percent in the rest of Western and Northern Europe, simply because France -- on account of the greater stability in the number of births -- has a more regular age composition. Yet even in France the proportion of old persons is smaller than it would be if the present infantile mortality had prevailed at the time when the persons now over 50 years were born. Even in France the corrected death rate is higher than the crude death rate. The statistics so far available for 1926 do not enable us to ascertain accurately the corrected death rate for all Western and Northern Europe, but they indicate that this corrected death rate must have been between 17 and 18 per 1,000.

The age composition of the population of Western and Northern Europe tends to swell the number of births. Since at present the proportion of children and of old persons is comparatively small, the number of births must be comparatively high. But the women who now are in child-bearing age will by and by pass this stage and will have to be fully replaced if, with present fertility, the number of births is not to decrease. The chances of such a replacement in the near future are easy to ascertain. In 1926 there were in Western and Northern Europe 23.67 million females under 15 years and 25.85 million females from 15 to 30 years. It is evident that even if all girls who now are under 15 years reached child-bearing age, they would by no means be able to replace those who now are between 15 and 30 years. This result, of course, is partly due to the reduction of births during the war. But even if there had lived in 1926 as many girls of under 15 years as of under 5, the girls of under 15 (who then would number 25.06 millions) would not suffice fully to replace those of 15 to 30 years.

The situation with which we are confronted can perhaps best be realized by starting from the present number of female births. The total number of female births in 1926 was 1¾ millions. The total number of women from 15 to 50 years was 52¾ millions. If the number of female births continued to be 1¾ millions and if no death occurred, there would be in fifty years from now 1¾ X 35 = 61¼ million women between 15 and 50 years, or considerably more than at present. But according to the mortality of 1926, the average number of years which the newly born girls may expect to live in the age of child-bearing is 29. If, then, the number of female births continues to be 1¾ millions, and if mortality remains what it was in 1926, the number of women between 15 and 50 years, fifty years from now, would be 1¾ X 29 = 50¾ millions only, as compared with 52¾ millions in 1926. But with present fertility the number of births is bound to decrease before that, since the number of women now between 15 and 30 years cannot be replaced by those being under 15 years of age.

A reduction of mortality, on the other hand, could not possibly change matters materially. If none of the newly-born girls died before 50 years they would all live 35 years in child-bearing age. The average number of years lived in child-bearing age must therefore always be lower than 35 years. In Germany it was only 20.2 years in 1881-1890. But by 1925 it had risen to 28.4, and in 1926 it was about 29 years for all Western and Northern Europe. This shows that there is not much margin left for further improvement. We may succeed in materially extending the lives of the women who have passed child-bearing age, but this, of course, will not in the least affect the number of births.

With a fertility and mortality such as prevailed in Western and Northern Europe up to about 1910, the population would have doubled in three generations. With a fertility and a mortality as they have now prevailed for some years, the population of Western and Northern Europe is bound to die out. This process, of course, will be rather slow. With the present age composition it would take decades until there actually would be an excess of deaths over births, and it would take centuries until the population would be half of what it is now. The process will be accelerated if emigration continues; it will not, of course, be affected by immigration, since we are concerned only with the present population and its descendants. It can be stopped by an essential change in mortality or in fertility. But the future reduction of mortality in the decisive ages, that is, those under 50 years, cannot be very great after all that has already been accomplished. The future, then, depends mainly on the trend of fertility.


Western and Northern Europe, in 1926, had an excess of births over deaths equal to 0.62 percent of the population, but the population did no longer reproduce itself. The entire world, in 1920 to 1926, apparently had likewise a yearly excess of births over deaths equal to 0.62 percent of the population. Should we draw the conclusion that the world's population no longer reproduces itself? We should, of course, do nothing of the kind. As long as we know so little about conditions in most of the countries outside of Western and Northern Europe we cannot tell whether with a yearly increase of 0.62 percent the world's population is reproducing itself very strongly or moderately or not at all. The conclusions which we should draw lie in a different field:

1. We should require sociologists, biologists, and all other population "experts," to familiarize themselves with elementary facts before writing new books on the population problem.

2. We should require all statesmen to study thoroughly the population problems of their countries before making new laws affecting emigration, immigration, agriculture, or industry.

[i] See the forthcoming publication of The Institute of Economics, "The Balance of Births and Deaths Vol. I. Western and Northern Europe," New York, The Macmillan Company.

[ii] It includes Great Britain and Northern Ireland, the Irish Free State, France, Belgium, Luxemburg, Holland, the Saar Territory, Switzerland, Germany, Denmark, Norway, Sweden, Finland.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now