Journal Issue

Population Growth and Economic Development

International Monetary Fund. External Relations Dept.
Published Date:
March 1969
  • ShareShare
Show Summary Details

George C. Zaidan

IT TOOK THE WORLD over 18 centuries to increase its population from a quarter to one billion persons. Today one billion persons are being added every 15 years, and the world population is growing at a rate that is 30 times as high as the average rate of growth between the first century A.D. and 1650. In less developed countries that rate is 40 times as high.

Why are these rates so high? What are their economic implications? What are the prospects of a deceleration? These are large and important questions which need to be taken one at a time; in this article I am concerned only with the second.

In attempting to indicate what we know today about population and economic development, there is one important preliminary point. Many people have been alarmed by the population explosion and some writers have made forecasts of impending famine. However, although the population of the world today is larger than ever, the standard of living of a large proportion of mankind is also much higher than at any time in recorded human history. Equally, looking into the future, it appears that the potential for economic growth is far greater than the potential for population growth. For instance, new strains of wheat, rice, and other foods have been discovered that could increase yields by two to five times over short periods of time. In contrast, the world population would only double over 35 years or so. Of course to exploit this potential would require large accompanying social changes, but the potential is there and it is immense.

But while the potential for economic growth is larger than the potential for population growth, this is not strictly relevant to the question that is being considered here. The crux of the matter is not whether the world can adjust to the present high rates of population growth but rather how much better the prospects for development would be if these high rates could be reduced. Here there are some impressive statistics. A representative sample of less developed countries1 shows that over 65 per cent of total investments are devoted to maintaining the level of per capita income at a constant level, whereas the corresponding figure for a sample of developed countries 2 was less than 25 per cent. To reduce this gap, a developing country can either increase its national income and/or reduce its population. In terms of the resources required to achieve a given increase in living standards, these will be many times lower if they are used to reduce fertility than in any alternative investment. This is ultimately the basis of the need for population control. Furthermore, the total costs of such a program are a small proportion of the total expenditures on development (3 per cent as a maximum), so that investment in population control can never hope to replace expenditures on economic development, but is rather a complement to the latter.

The Economic-Demographic Model

Since death rates can be expected to decline further, and since emigration can rarely be influenced to a significant extent, economists have looked with particular interest at the effect of reduced fertility on economic development. More particularly the question was put: how would a linear reduction of 50 per cent in the birth rate over a period of 25 years (and constant thereafter) affect the standard of living in a developing country? The first model of this sort was applied to India3 (and Mexico) but several other studies have been carried out since, and they have all confirmed the numerical magnitude observed in the first study. Before discussing the results of this study it may be instructive to spell out why it is that large benefits have been observed. Chart I attempts to reproduce in simplified form the mechanism through which this model operates. As can be seen from the Chart, there are three different paths through which a reduction in birth rates affect the rate of growth of per capita income. The first occurs because the reduction in fertility leads to a smaller total population. Table 1 shows differences in the total population (and its age distribution) with and without a reduction in fertility. In both examples a similar reduction in mortality is assumed, and the initial population is assumed to be one million.



Group ↓ Year →
Projection A
Projection B
Projection A—fertility remains unchanged. Projection B—the birth rate is lowered by 50 per cent over a period of twenty five years and constant thereafter. In both cases mortality is constantly improving. Figures are in thousands (initial population = 1 million). Figures in brackets are percentages.
Projection A—fertility remains unchanged. Projection B—the birth rate is lowered by 50 per cent over a period of twenty five years and constant thereafter. In both cases mortality is constantly improving. Figures are in thousands (initial population = 1 million). Figures in brackets are percentages.

The result of this smaller population is that the national income is shared by fewer persons. The national income itself is not reduced and may well be increased in the short run, since the reduction in fertility does not damage the productive capacity of the economy; this depends on the accumulation of capital, the quantity and quality of the labor force, technology, and the amount of natural resources. The reduction in fertility is assumed to have no effect on any of these “factors of production” except the accumulation of capital and the quantity of the labor force. The first effect is positive and immediate in the sense that the savings potential of an economy increases with reduced fertility, whereas the second is negative but occurs with a time lag of 15 years or so, since the quantity of the labor force would be smaller than otherwise. These two other effects, depicted in Chart I, are discussed at some length below.

The second path is via the accumulation of capital. Capital is assumed to increase because a reduction in fertility will lead to a much smaller proportion of children, as can be seen in Table 1. This in turn means that the ability of a country to save is much greater, as families have fewer children, and can therefore “divert” into increasing their savings part of the income that these children would have consumed had there been no reduction in fertility. This “diversion” can either occur voluntarily through private savings, or because of the ability of the government to raise more taxes, at a given level of sacrifice. These increased savings represent a potential source for the faster growth of total income, but the question arises whether parents will in fact increase them, or whether instead they will consume all their increased (per capita) income. Alternatively, if governments raise their revenues the question arises whether this increase will go into government consumption or government savings. This question needs to be studied, but a recent investigation suggests that the “burden of dependency” is an important explanation of the high observed differences in saving ratios among the countries of the world,4 in addition to other variables that are normally considered to “determine” savings proportions.

The third path is via the growth of the labor force. Here, it is important to make a distinction between the short term (defined as the time between birth and the average age at which persons enter the labor force, or about 15 years) and the long term. In the short term a reduction in fertility has no effect on the size of the labor force, since all the persons who will be entering the labor force over the next 15 years or so have already been born today. In the long term, however, there is a negative effect, in that the labor force would be smaller than otherwise. The impact of this on the economy will depend on the particular country that is being studied: in conditions of “surplus labor” or economies where unemployment is high, such a reduction will not affect the national income of an economy, at least initially. Elsewhere the negative effect will be more substantial and immediate.

Combining these three effects, we can see how a reduction in fertility affects favorably the standard of living. On the one hand there are fewer people; on the other, total output is larger in the short run under conditions of reduced fertility. After a period of 15 years, there is some doubt as to whether total income would be larger than otherwise, since this depends on the relative magnitude of the second positive effect, as compared to the last negative effect.

Such then is the mechanism of the model. The results derived from it are striking. Table 2 shows that the standard of living in a country which reduced its fertility would be 40 per cent higher5 than otherwise at the end of 30 years, and over twice as high at the end of 60 years.

Years =0102030405060

By this measure the total population is weighted by the number of infants in it, and converted to a number of “equivalent adult consumers.” The assumptions made are that a child’s consumption is 0.5 that of a male adult, and that an adult female consumes 0.9 times the consumption of an adult male.

By this measure the total population is weighted by the number of infants in it, and converted to a number of “equivalent adult consumers.” The assumptions made are that a child’s consumption is 0.5 that of a male adult, and that an adult female consumes 0.9 times the consumption of an adult male.

Some Other Implications of Population Growth

Though the model outlined above shows the major economic benefits of reduced fertility, it omits some important effects, most of which would increase the observed benefits. These effects are omitted because they pose difficult problems of measurement,6 but they may be no less important than the effects that have already been discussed. Some of these omitted effects which are shown through dotted lines in Chart I are particularly interesting.

The age structure of a population is determined primarily by the birth rate. Thus less developed countries—with birth rates of 40 per 1000 or more—typically have a proportion of children (aged 15 years or less) of 40—45 per cent of the total population, whereas in developed countries—with birth rates below 20 per 1000—the corresponding proportion is about 25 per cent. The death rate also has an effect on the age structure, but the quantitative effect is small. This can be seen intuitively, because a decrease in death rates leads to more people surviving at all ages. If the death rate in all ages is reduced by the same percentage amount, then the age structure remains unchanged. Hence it is not the size of the decline in the death rate that influences the age structure, but its differential impact in different age groups. Usually the proportionate decline is larger among young infants, so that a decline in the death rate is associated with a younger age structure. By contrast a decline in birth rates always leads to a much more marked and older age structure, since it is only the numbers of young people that is reduced.

A reduction in fertility would, through better nutrition, health, and education, affect positively the quality of the labor force. This would in turn lead to a larger total income than otherwise. There are two reasons why nutrition and health would improve. On the one hand the total population would be smaller, while on the other, total income and hence total expenditure on food consumption and health may be larger. For both these reasons, each member of the labor force would be better nourished and healthier. Total expenditures on education may be larger and, after some time lag (the time between the start of the reduction in fertility and the age of entry to schools) the school age population would be smaller. Eventually, as a higher proportion of educated persons entered the labor force, the labor force would be more skilled.

A second set of relationships that have been omitted concern the impact of economic development itself on the rate of population growth. The rise in per capita incomes encompasses vast changes (such as education, urbanization, and the employment of women) that reinforce the initial decline in fertility. At the same time a faster rate of economic development may accelerate the decline in mortality. But again this decline in mortality, which is concentrated among young children, sometimes leads to a decline in fertility, as parents perceive that a larger number of their children survive to the adult ages.

Characteristics of the Model

The model has several interesting characteristics and implications. One concerns the effects of delaying this decline in fertility, or the “costs of waiting.” The argument is often heard that since every country that has become highly developed has experienced a reduction in fertility of 50 per cent or more, public policy should be directed at accelerating economic development to the maximum possible extent, and fertility decline will take care of itself. While it is true that all countries that are today highly developed have experienced at least a 50 per cent decline in their fertility, it is not clear when such a decline “normally” starts. A leading authority estimates that for many less developed countries of today with low per capita incomes, it would require 30 to 60 years for these countries to reach a level of industrialization where an “autonomous” decline in fertility would set in. The costs of such a delay would be large. Table 3 shows the ratio of income per consumer in a population where the decline starts immediately to one in which there is a 30-year delay (in both, fertility is assumed to be halved over 25 years, and is constant thereafter). It is seen that to wait for an automatic decline in fertility is to forego the relative gains shown in Table 3. These gains rise to 63 per cent after 50 years, and then settle to 40 per cent, which is the gain previously noted at the end of 30 years.

Years =0102030405060100

Another characteristic of the model is that for at least 15 years, and possibly for a much longer period of time (especially in those countries that suffer from “surplus labor”), total income would be larger under conditions of reduced fertility. This point is particularly relevant with respect to two arguments that are often made for having a larger population. The first argument is that a larger population is beneficial because it leads to a larger total consumption. This larger market allows the economy of a country to exploit to the maximum the economies of scale of this larger market. There are at least two reasons why this argument is unconvincing. First, it holds only in a closed economy that has no trade. Once trade is possible, then specialization would permit the exploitation of some of these economies. Second, the argument confuses demand based on “need”—which would be larger with a bigger population—with demand backed by purchasing power. Since total income would be larger with reduced fertility for at least 15 years and possibly longer, demand in the sense of the ability to purchase a larger total amount of goods would be greater with reduced fertility.

A second argument that is made for a larger population is that of military power—that country A cannot afford to limit its population while country B, across the border, is growing at such a rapid pace. The previous discussion suggests several reasons why such an argument is untenable. First, a program of fertility reduction would not affect the growth of the persons eligible for military service for a long period of time—between the onset of the decline in fertility and the average age of conscription, say 15 years or more. Even after this period it is worth noting the large absolute increases that occur under conditions of reduced fertility. Table 1 shows that the 15-64 age bracket would be 240 per cent higher at the end of 30 years, even if fertility is halved. Second, military power is not only a matter of numbers, but especially of the quality of the people and the armaments at their disposal. A population in which fertility was being reduced would be healthier, better fed, and better educated. In addition, because of the larger total income in the short run, it would be able to afford more military equipment at the same level of sacrifice (where “sacrifice” is defined to be a given proportion of the national income going into defense expenditures).

Benefits Independent of Economic and Demographic Conditions

A third characteristic of the model is that the magnitude of the benefits is determined by the rapidity of the decline in fertility and—a striking fact—is not affected to a significant extent by the initial economic and demographic conditions prevailing in a given country. Thus the model was worked out with the detailed data for both India and Mexico and in spite of the different conditions (Mexico has higher fertility, lower mortality, and a higher per capita income than India), the results were essentially the same. Subsequent applications to other countries have tended to confirm this.

The fact that these models are applicable to all countries may be a source of some puzzlement in view of the variety of conditions prevailing in the less developed world. This feature is best explained because of the fact that benefits are measured in terms of relative rates of growth of per capita income, together with the fact that the main source of benefits in the short run arise because the production that unborn children would otherwise have consumed is available for distribution among the “remaining” population. For instance, assume two countries A and B with per capita incomes of $100 and $400, respectively. Now assume the reduction in fertility reduces family size from eight to six persons. Then in both countries the relative gain of the family in per capita income 7 is the same—33 per cent. If the rate of growth of per capita income would have been zero in the absence of a reduction of fertility, then country A’s per capita income would have been $133 and country B’s $533.

A Zero Birth Bate?

Another characteristic of the model occasionally prompts nonspecialists to pursue the reductio ad absurdum and to ask whether, since the greater the reduction in fertility, the larger the benefits this implies, a zero birth rate would in some sense be economically “the best”? This question is theoretical for at least the next decade or two, since the halving of the birthrate over such a period would represent a considerable achievement. The answer to it then depends on what we mean by “economically the best.” If we are looking at the rate of growth of per capita income as our sole measure of welfare, then the answer would be yes. If we look at total income then in the short run the answer would still be yes, though a time eventually comes when this is no longer so. Finally, an important point to make is that all the argument is in terms of trying to achieve the largest growth of national income—either in total or per capita terms. However, the measurement of national income does not include the joy that parents get from having children. It is difficult to see how this can be measured, but beyond a point there is obviously a clash between individual and social welfare. Both for this reason and because of the fact that few would be prepared to accept per capita income as the only measure to be maximized, a point will be reached beyond which a further reduction in the birthrate is no longer desirable. However, it seems to me that as long as the gap between developed and developing countries is increasing, it is legitimate to look at per capita income as the main measure to be maximized and to attach only minor importance to other factors. Translated into policy terms, this means that developing countries might aim at bringing down their birthrates approximately to the level prevailing in developed countries as quickly as possible provided this remains an economically efficient proposition. Beyond this point there may be room for arguing over the desirability of further declines, but this is not a question over which one needs to worry in the near future.

What is a “Population Problem”?

Another implication of the fact that the benefits are more or less the same for all countries is that it can be argued that it is equally urgent to start a population policy in all countries. This would indeed be so if our criterion for adopting a population policy was based solely on the relative percentage increases in per capita income that would result from such policy.

Although this is one possible criterion, several other alternatives are possible. One possibility would be the extent to which an economy has to invest (as a proportion of its income) in order to keep per capita income at a constant level. This would give a measure of the resources that are “wasted” to absorb the growth of population. Table 4 gives these figures for some 33 countries. Something that stands out from this table is that for all developed countries that are shown in the table the proportion that is “wasted” is less than 5 per cent. Another point worth noting is that this proportion is not related to the density of the population in a given country. This is especially worth emphasizing in view of the fact that in the popular image, density is often associated with the gravity of a population problem. Columbia, for instance, has large tracts of vacant and fertile lands but nevertheless it has to invest 14 per cent of its national income in order to absorb the increase in its population. Alternatively, one notes that a country such as Pakistan has less of a “population problem.” This seems to run counter to the popular image which associates the population problem with the teeming millions of Asia. It is not that this image is mistaken but that it does not follow from this particular criterion.

Over 10%Colombia, India, Morocco, Brazil, Ghana, Tunisia
7.5-10%Malaysia, Peru, the United Arab Republic, Thailand, Mexico, Philippines, Turkey
5-7.5%The Sudan, Pakistan, Nigeria, Indonesia, South Korea, Chile, Ethiopia
Less than

United States, Norway, France, Sweden, Denmark, Finland, West Germany, Italy, the United Kingdom, Belgium, Austria, Greece, Portugal
Note: Population growth figures are for 1967. The incremental capital-output ratio is for the period 1960-65. Countries included are those which had the necessary data, and they are classified in descending order.
Note: Population growth figures are for 1967. The incremental capital-output ratio is for the period 1960-65. Countries included are those which had the necessary data, and they are classified in descending order.

An alternative criterion might be the absolute size (and not the proportion) of investments that go into absorbing population increase. This may be of interest to donors of aid, who may be more concerned with this absolute rather than relative size. Thus countries with large populations would come on top of the list, as can be seen in Table 5.

Malaysia, Morocco250
Tunisia, the Sudan90
Note: The investment proportions required to keep per capita income at a constant level are those of Table 4. The GDP estimates are for 1966 at 1964 prices.
Note: The investment proportions required to keep per capita income at a constant level are those of Table 4. The GDP estimates are for 1966 at 1964 prices.

The previous various criteria of what constitutes a population problem emphasize that there are different ways of looking at the same phenomenon; depending on the way we look at this phenomenon, a different classification is obtained. However, more important than the particular classification one chooses is the point that by historical standards the problem is urgent in all less developed countries. Furthermore, as death rates in these countries fall further, the problem is likely to increase, the more so among countries whose problem is at present “mild.” For these reasons, it may be more relevant to emphasize the much larger difference between developed and less developed countries, rather than the differences among the latter group.


The sample included the following 22 developing countries: Brazil, Chile, Colombia, Ethiopia, Ghana, India, Indonesia, Kenya, South Korea, Malaysia, Mexico, Morocco, Nigeria, Pakistan, Peru, the Philippines, the Sudan, Thailand, Tunisia, Turkey, Uganda, and the United Arab Republic.


The United States, Canada, Australia, and 16 West European countries.


The discussion of this whole section, is based on the work of A.J. Coale and E. Hoover, “Population Growth and Economic Development in Low Income Countries: A Case Study of India’s Prospects” (Princeton, N.J., 1958). See also A.J. Coale, “Population and Economic Development in the Population Dilemma” (Prentice-Hall, N.J., 1963), edited by P. M. Hauser. Tables 1, 2, and 3 of this article are taken from this paper.


N.H. Left, “Population Growth and Savings Potential,” unpublished preliminary report of the Office of Program Coordination, U.S. Agency for International Development.


It may be important to note that the whole measurement of these economic benefits is from the standpoint of a country. At the family level the calculus of benefits and costs of an additional child may be very different. This is a crucial distinction in considering policies designed to reduce fertility.


In his latest book, Asian Drama, G. Myrdal is critical of these omissions and of the Coale-Hoover approach in general. His criticisms, however, are directed at the accuracy of the model, and in particular the narrow range of its results. But Myrdal accepts the conclusions that the economic effects of a fertility reduction are “favorable … and that these effects are very considerable and cumulative, gaining momentum over the years” (Page 1467).


In terms of income “per equivalent adult consumer,” the gain would be smaller since a family of six would have a smaller proportion of children.

Other Resources Citing This Publication