That demographics has a bearing on world affairs, the global configuration of power, and even the conduct of international relations may seem entirely self-evident. We hardly need demographers to impart such an insight to us. Other things being equal, a country with a large population will be capable of mobilizing greater resources to champion its foreign policy objectives than a smaller country. Other things being equal, a shift in the regional distribution of population will affect the regional balance of power.[1] It is obvious, moreover, that population change--by its tempo and its composition--can affect the vigor and efficacy with which a presiding government may pursue its chosen policies, both foreign and domestic; in the extreme, a country’s internal population dynamics might even destabilize an established political order or topple a sitting regime. Yet if all this is clear in the abstract, any systematic effort to assess the actual impact of population on power politics or international security confronts a formidable array of inconvenient, untidy, and sometimes bewildering particulars. For other things are not always equal in the interplay between population and world politics. In fact, they almost never are, and the relative influence of the diverse factors that shape outcomes in the global arena are such that demographic trends may seldom be afforded a decisive role. What appears “obvious” to many observers about some impact of population on international affairs, for its part, often proves to be either impossible to substantiate or demonstrably false.

For reasons that need not detain us here, the subject of “population” seems to evoke stark images for, and to elicit strong preconceived opinions from, the modern intellectual and internationalist. Despite all the valuable research that has to date come from them, however, the several disciplines subsidiary to “population studies” have as yet been able to offer only a limited explication of the basic causes and consequences of population change; the predictive powers of the demographic sciences are even more circumscribed. The unsettling result of this mismatch of “knowledge” and understanding is that when it comes to population matters, reach is often greater than grasp.

Measured understanding and nuanced appreciation of the complex relationships between population change, socioeconomic change, and political developments would seem to be particularly important for a “discriminative policy toward the developing world,” the approach Paul Kennedy and his coauthors of the “pivotal states strategy” urges Washington to adopt.[2] It is not only the exceptionally rapid pace of demographic growth in low-income regions today[3] that recommends such attention. For the “pivotal states” approach itself posits that demographic forces will play a major role in the political and security prospects of developing countries.

“Overpopulation,” “population surge,” “all the familiar pressures of rapid population growth,” and “population problems . . . straining the country’s natural resources,”[4] in this reading of current events, contributes significantly to a clear and present danger of “social and political collapse” in strategic Third World states--collapse that could in turn trigger further “transboundary mayhem.”[5] Although the authors do not explicitly detail the mechanisms and linkages they envision, their Weltanschauung (worldview) seems to track with the outlook expressed in a corpus of scholarly and popular texts that have attracted considerable attention during the 1990s.[6]

This paper will offer a contrasting evaluation of the role of population in international affairs. Many current analyses, which focus on environmental and demographic factors as a cause of conflict, seem to presume an unwarranted influence for demographic factors in determining the course of social, economic, and political events in developing regions. The actual impact of population change in developing regions appears to be less clear-cut and less categorical than many contemporary colleagues suppose. Depending upon the specific circumstances, rapid population growth can create opportunities as well as constraints, can generate benefits as well as costs, and can arguably be conducive to political stability as well as turmoil. Finally, mounting evidence suggests that the scope for affecting demographic trends in low-income countries through voluntary population policies is very much more limited than many politically-minded proponents of international family planning efforts seem to recognize. 

The Definitional Ambiguity of "Overpopulation"

The notion of “overpopulation” figures prominently in the present exegesis of the “pivotal states” approach. The authors are hardly alone in forwarding the concept. “Overpopulation” is regularly discussed and debated in the popular media and in scholarly journals; for the United States and other governments, it remains a term of art in official policy deliberations.[7] Yet there is one serious problem with the concept of overpopulation: population trends and demographic data provide no basis for defining it.[8] Familiar as the term may be, the fact remains that demographic indicators cannot describe overpopulation consistently and unambiguously.

What, after all, would be the criteria by which to judge a country “overpopulated”? Population density is one possibility that comes to mind. By this measure, Bangladesh would indeed be one of the contemporary world’s most overpopulated countries--but as it happens, Bangladesh would not be as overpopulated as Bermuda. By the same criterion, the United States would be more overpopulated than the continent of Africa, Germany would be more overpopulated than the Philippines, Italy would be more overpopulated than Pakistan, Malta would be more overpopulated than Haiti, and virtually the most overpopulated spot on the globe would be the kingdom of Monaco![9]

Rates of population growth offer scarcely more reliable guidance for the concept of overpopulation. In the contemporary world, Africa’s rates of increase are the very highest. Yet rates of population growth were even higher in North America in the second half of the eighteenth century.[10] Would anyone seriously suggest that frontier America suffered from overpopulation?

What holds for density and rates of growth obtains for all other demographic variables as well: birth rates, “dependency ratios” (the proportion of children and elderly in relation to working age groups), and the like. If overpopulation is a demographic problem, why can’t it be described unambiguously in terms of population characteristics?

For a simple reason: overpopulation is a problem that has been misidentified and misdefined. The visions evoked by the term overpopulation--hungry families; squalid, overcrowded living conditions; early death--are real enough in the modern world. But these are properly described as problems of poverty. As a human condition, poverty, like all other possible human attributes, is represented in individual members of a population. But it is an elementary lapse in logic to conclude that poverty is a “population problem” simply because it is manifest in human populations. 

Population Growth, Poverty, and Economic Development: The State of the Literature

The relationship between population growth and material poverty in low-income areas, for its part, appears to be rather more complex than is commonly supposed today. Surprising as this may sound, robust evidence to support the widely held presumption that rapid population growth exerts a general, significant, and adverse influence upon poverty levels and development prospects in Third World countries has yet to be agreed upon by specialists on the subject.

An authoritative assessment of the available economic research on the relationship between rapid population growth and poverty in developing countries was offered in 1994 by Professor Dennis Ahlburg, an economic demographer with the University of Minnesota, in a study prepared for the Overseas Development Council (ODC). His words merit careful reading:

Poverty is the result of many factors including population growth. However, the importance of population growth’s contribution to poverty is far from clear. At the level of the household, it appears that additional children reduce the education and health of other children in the family but estimates of the size of these effects vary considerably: sometimes they are positive, sometimes insignificant, and more often than not negative and fairly small. At the national level, it is even harder to establish the effect of population growth on poverty. Net negative effects have been found on variables related to poverty, such as income, education, and health, but there is little direct evidence of a significant negative effect on poverty itself . . . [It] is not clear whether population growth causes poverty in the long run or not.[11]

What holds for population growth and poverty obtains for population growth and development as well. In the same ODC project, Professor Robert Cassen of Oxford University--an enthusiastic proponent of international family planning efforts and an advocate of slowing Third World population growth--summarizes the state of the literature in the following terms:

[The] available evidence from empirical studies does not clearly show that population growth exerts a negative influence on development . . . The issue of whether per capita economic growth is reduced by population growth remains unsettled. Attempts to demonstrate such an effect empirically have produced no significant and reliable results, although most of these studies have examined developing countries in the aggregate rather than individually.[12]

These recent judgments do not betoken a sudden shift in expert thinking. A 1986 report on population growth and economic development by the National Academy of Sciences’ National Research Council (NRC) arrived at broadly similar findings:

[It] seems unlikely that slower population growth will allow a larger number of people, over future generations, to enjoy a given standard of living thanks to lower natural resource prices . . . We do not find convincing evidence that lower fertility will result in faster growth in [educational] enrollment ratios . . . We have found little evidence that the aggregate savings rate depends upon growth rates or the age structure of a population . . . A reduced rate of urban labor force growth in developing countries (most of which is a product of natural increase among the urban population) is not likely to be systematically accompanied by corresponding reductions in joblessness.[13]

Although the NRC study did “reach the qualitative conclusion that slower population growth would be beneficial to economic development for most developing countries,”[14] it advised that “many of the initial [negative] effects of population growth” could be ameliorated or even reversed in the long run if institutional adjustment mechanisms are in place. Among the most important of such mechanisms are property rights in land and properly functioning markets for labor, capital, and goods.[15]

In the late 1950s and the early 1960s, it is true, a number of academic exercises in economic modeling produced simulations indicating that rapid population growth would severely compromise the efforts at economic development in low-income regions.[16] (It was on the basis of one such study that President Lyndon Johnson in 1965 asserted “the fact” that “less than five dollars invested in population control is worth a hundred dollars invested in economic growth.”[17])

Although these econometric constructs were surely innovative in their day and at one time may have seemed to some observers to furnish imposing “proof” of the negative relationship between population growth and development, closer inspection of the models reveals that their results were actually predetermined by the assumptions that had been embodied in them.

By and large, the critical assumptions in these models were not only crude, but also quite unrealistic for a world inhabited by human beings. They posited, for example, that the ratio of capital to output would remain fixed and relatively high in a period of rapid labor force growth, that rates of return to labor and capital would not be affected by the relative scarcity or abundance of these factors of production, that economies of scale would not affect economic growth, and that increased consumption by impoverished populations would leave their productivity unaffected. By relaxing this straightjacket of constraints, the projections generated by these models could be altered substantially. In fact, a well-known population economist has demonstrated that, on the basis of alternative and hardly more unreasonable assumptions, these very models could be made to “prove” that more rapid population growth would systematically accelerate economic development![18] The latter presumption, incidentally, would have been taken as utterly unobjectionable by some eminent economists in the fairly recent past; during the Great Depression, for example, both John Maynard Keynes and Gunnar Myrdal argued that low fertility and slow population growth were impediments to economic progress.[19]

Suffice it to say that the inductive, “it stands to reason” school of argumentation can only do so much to enhance our understanding of the economic consequences of population growth in low-income settings. For as Allen C. Kelley of Duke University and William P. McGreevey of the World Bank have observed,

The contradictions in the mind of Malthus still confront many analysts and policymakers in the population field, whose feelings, impressions, and personal certainties about the negative effects of population growth are not supported by a careful sifting of the facts.[20] 

Global Trends in Population and Output in the Twentieth Century

We are thus perhaps better advised to examine the empirical record. A thumbnail sketch of the broad demographic and economic contours of the past century is instructive in this regard. At the very least, we know for a fact that rapid and sustained population growth does not preclude rapid and sustained economic and social advance. If it did, the vast material transformation we have already witnessed in the twentieth century could not have occurred.

Since the beginning of this century, according to the best available estimates, the world’s population has more than tripled. As already mentioned, nothing like this had ever taken place before--and although the tempo of global population growth appears to have peaked and to be declining, it is still proceeding with extraordinary speed by historical perspective. This unprecedented demographic explosion, however, did not consign humanity to Malthusian destitution. Just the opposite: a worldwide explosion of prosperity accompanied it. The magnitude of the economic leap that has taken place in our century is not always appreciated. According to Angus Maddison, the eminent economic historian now at the University of Groeningen, the world’s per capita GDP quadrupled between the turn of the century and the early 1990s.[21] In Latin America and the Caribbean, per capita GDP, by his estimates, has more than quadrupled this century; in Asia and the Pacific, it has more than quintupled; and even in tormented Africa, it may have more than doubled.

Maddison has produced long-term estimates of per capita output trends for eight of the nine countries designated as “pivotal states.” In each of these eight countries, substantial increases in per capita GDP have been registered over the course of the twentieth century. Between 1900 and 1992, by his estimates, per capita GDP sextupled in Brazil; it more than quadrupled in Mexico and Turkey; it more than tripled in Indonesia and Egypt; it more than doubled in long-troubled South Africa; and it more than doubled in both Pakistan and India--locales that even specialists sometimes assume to be strangers to economic progress. Although Maddison’s calculations cannot be exact, they do convey a sense of the economic changes these countries have already undergone.

Why has the most rapid period of population growth in history coincided with the most dramatic economic expansion in the human experience? Part of the answer may lie in the “population explosion” itself--or more precisely, in its proximate causes. The modern population explosion was sparked not by an upsurge in fertility, but rather by a widespread, profound, and still continuing drop in mortality levels. Since the start of our century, the average life expectancy at birth for a human being has probably doubled--it may have more than doubled.[22] Every corner of the earth has joined in this health revolution--and on the whole, incidentally, health progress in our century has been more dramatic in the less developed regions than the more affluent ones.

Between the early 1950s and the early 1990s, for example, life expectation at birth in what the United Nations (UN) labels the “more developed regions” rose by about 9 years. In the “less developed regions,” by contrast, life expectancy increased by an estimated 21 years. Even in the seemingly stagnant and often politically unstable countries classified by the UN as “least developed,” life expectancy at birth is believed to have risen by over 14 years during this same period. Although we are by now used to hearing about the “widening gap” between the world’s rich and poor, these numbers unambiguously point to a steady and substantial closing of a highly meaningful difference in life chances around the world.[23]

All of the “pivotal states” have recorded major increases in life expectancy during the postwar era. (See Table 1.) According to the latest UN estimates, every pivotal state but South Africa witnessed an increase in life expectancy at birth of 21 to 25 years between the early 1950s and the early 1990s. In South Africa, the average estimated increase was 18 years; this average, however, slightly understates the progress experienced by so-called “nonwhite” population groups within the country. 

Table 1: Current Population Projection for "Pivotal States"

1995 versus 2050 (in millions)

Note: Population rounded to nearest hundred thousand.

Source: United Nations, World Population Prospects: The 1996 Revision (New York: UN Population Division, forthcoming), Annexes II and III, pp. 78, 116, 172, 228, 230, 292, 324, 374, 404.

There is an economic significance to the ongoing mortality revolution in the Third World. Improvements in health mean improvements in productivity. It is not just that healthier populations are able to work harder; better health and reductions in mortality enhance the potential of what economists call “human capital”: education, training, skills, and the like.

Populations with enhanced productivity potential, of course, may not be able to capitalize upon their new capabilities for a variety of reasons. An adverse business climate or policy environment can reduce rates of return on human capital just as surely as on physical capital. Even so, the fact is that the same forces that have powered the modern population explosion have simultaneously eased the constraints against attaining higher levels of per capita output in contemporary Africa, Asia, and Latin America. 

Population Growth and Resource Scarcity

The intuitive inference that population increases should result in an intensifying resource scarcity (and perhaps to attendant social stresses or political conflicts) seems only natural and reasonable on its face. A review of the empirical record, however, should prompt us to reexamine and qualify our expectations.

Over the course of the twentieth century, there has been a tremendous surge in the demand for all manner of goods and services, including natural resources. The increase in demand has been driven not only by the tremendous increase in population, but also by widespread and notable increases in per capita incomes. Maddison, for example, estimates that global GDP was over 14 times higher in 1992 than it had been in 1900.[24] Despite this explosion of demand, however, real prices for natural resources have not moved appreciably upwards. Just the opposite: long-term prices for primary commodities--foodstuffs, agricultural goods, timber, minerals, energy products, and the like--have fallen sharply and unmistakably. According to one careful study, the inflation-adjusted price of primary commodities (energy products included in the mix) fell by over a third between 1900 and the early 1980s. The long-term rate of decline in real prices for these natural resources, according to these estimates, was proceeding at about 0.4 percent per year.[25] In the 1990s, real prices of primary commodities have continued on this secular downward trend.

For obvious reasons, food is considered a particular vital primary commodity. Interestingly enough, the real price of cereals in the international marketplace has declined even more rapidly than the average real prices for primary commodities on the whole. Between 1960 and 1990, for example, the real price for cereals fell by more than half--this despite an increase in gross global cereal imports from less than 60 million metric tons to over 160 million tons.[26]

The paradox here is that prices are meant to measure scarcity. Other things being equal, scarce items are supposed to be more dear; plentiful items, more cheap. By the information that prices are intended to convey, it would appear that the natural resources put to use commercially by human populations have been becoming less scarce--not more--during the era of rapid global population growth.

A number of objections may of course be lodged against this reading of events. The prices mentioned above are set in international marketplaces and may differ sharply from local prices in domestic markets; those without money cannot vote on the cost of these commodities; and in any case markets possess no special knowledge that could permit them to divine the “true” cost of anything. Even so, given the worldwide increase in per capita purchasing power, progressive globalization of markets, and the duration and magnitude of the long-term decline in real prices for primary commodities, arguments that price trends systematically misrepresent economic scarcity for commercially utilized natural resources must become increasingly acrobatic.

Population dynamics cannot explain the direction that primary commodity prices have taken over the twentieth century. On the other hand, the workings of a complex global economic system--in which substitutability of inputs in the production process is possible, technological innovation is regular, and market mechanisms grow stronger and more sophisticated--can provide an adequate explanation for the paradoxical discrepancy between trends in long-term demand and long-term resource prices. In areas where the local economic system is not functioning effectively--or in areas where the populace is for some reason cut off from the international economy--burgeoning population growth may indeed provoke resource scarcities that would lower living standards and lead to social or political conflict. But such situations constitute the exception rather than the rule in the world today and intrinsically beg questions about local economic arrangements rather than local demographic trends. 

Fertility Change and the Accuracy of Demographic Forecasting

Long-term political strategies rely on long-term projections about real-world conditions. More than any of the other social sciences, demography makes a practice of offering projections for the distant future. The latest World Population Projections by the UN Population Division, for example, provide detailed calculations on possible demographic conditions in all the countries of the world in the year 2050.

The calculations themselves are flawless--internally consistent computer algorithms generate them. The reliability of these numbers is quite another matter. The point is not always appreciated outside the confines of the discipline, but population specialists have absolutely no methodological basis for devising accurate long-term population forecasts.

Under noncatastrophic conditions, the pace of natural increase and the composition of a convened human population will be determined mainly by its fertility trends. Demographers, unfortunately, lack any general theory for explaining and predicting fertility change. Not that there is any shortage of theories here; rather, as historian Charles Tilly of the New School of Social Research once observed,

The problem is that we have too many explanations which are individually plausible in general terms, which contradict each other to some degree, and which fail to fit some significant part of the facts.[27]

Puzzling uncertainties have surrounded the phenomenon of secular fertility decline literally from the beginning. The first country in the world to embark upon long-term fertility decline did so in the late eighteenth century. That country, however, was not industrializing England, as modernization theories would lead us to expect, but rather France--then impoverished, overwhelmingly agrarian, predominantly illiterate, and devoutly Catholic. More modern events have proved no less nettlesome. Despite the almost overwhelming availability of information on conditions in contemporary industrial societies, demographers were unable to anticipate either the transnational Western postwar baby boom, or the subsequent OECD-wide shift to a below-replacement fertility regimen. For developing countries, they have been unable to forecast either the onset of fertility decline or the trajectory that fertility change follows once it begins.

Lacking as they do any sound compass that could guide them, projections of demographic trends can veer off course remarkably rapidly. Relatively recent editions of World Population Prospects--arguably the best of the many efforts at demographic projections regularly undertaken--are illustrative. In 1977, the medium variant projection for Bangladesh anticipated a birth rate of 41 per 1,000 in the early 1990s. According to the most recent estimates, though, Bangladesh’s birth rate in the early 1990s was actually about 27 per 1,000--fully a third lower than projected only 15 years earlier. Conversely, Afghanistan’s medium variant birth rate for the early 1990s was projected to be 40 per 1,000; in the event, birth rates in Afghanistan apparently averaged about 50 per 1,000 in the early 1990s.[28]

Nor were Afghanistan and Bangladesh the only countries in which projections and actual trends did not match. The discrepancies for some of the “pivotal states” were equally striking. (See Tables 2-A and 2-B.) In 1990 to 1995, according to the UN’s latest estimates, total fertility rates (the number of children born per woman per lifetime) were roughly 1.5 births lower in Mexico and Turkey, and two births lower in Brazil, than the 1977 medium variant projections for those same years. With respect to population growth, the 1977 projections overshot Mexico and Brazil’s rates of change by more than 1 percent a year; for the nine “pivotal states” as a group, the projections were a third too high. For certain “pivotal states”--Pakistan, for example--projections and ultimate results were very close. But this congruence occurred simply by accident, as must any close fit between fertility or growth projections and real trends. 

Table 2-A: Demographic Projections versus Actual Outcomes

Note: UN 1977 medium variant estimates for 1990-95 for pivotal states are compared with current estimates for same period.

Source: United Nations, World Population Prospects: The 1996 Revision, Annex I: tables A-1 and A-18; and United Nations, World Population Prospects as Assessed in 1973 (New York: UN Population Division, 1977): tables 29 and 38.

Table 2-B: Demographic Projections versus Actual Outcomes

Note: UN 1977 medium variant estimates for 1990-95 for pivotal states are compared with current estimates for same period. Projected fertility rates are derived from gross reproductive rates

Source: See table 2-A.

As they stand, the most recent UN long-range population projections outline a wide range of possibilities for future population growth in the “pivotal states.” (See Table 1.) There is no reason, however, to believe that the ultimate population patterns of these countries will stay within the very broad boundaries that current “high projections” and “low projections” demarcate. Little as we may know about the determinants of fertility change, recent history attests that fertility patterns can shift with great speed. In Thailand, for example, the total fertility rate fell from over 6 births per woman in the late 1960s to under 2 birth per woman in the early 1990s.[29] Nor is there any reason to assume that fertility decline will tend to stabilize at a level near net replacement. Japan has recorded sub-replacement fertility for the past 40 years. A number of countries with relatively low levels of income, moreover, already experience sub-replacement fertility: China, Cuba, Sri Lanka, and Thailand among them.

In mentioning these facts I do not mean to offer predictions but instead caution that the range of alternative demographic futures for the pivotal states may be rather greater than we imagine today. In the absence of catastrophe, demographic techniques can estimate the number of survivors in future years from an extant population reasonably well. But they cannot provide us guidance when it comes to the as yet unborn--and thus to long-range population trends. 

The Demographic Impact of Population Policies

It is now conventional for population specialists and family planning authorities to ascribe a decisive role in contemporary fertility reductions to the new modern methods of contraception. By extension, contemporary efforts in the area of international family planning are widely viewed as instruments for bringing about significant and voluntary reductions in fertility in developing regions. Such mechanistic faith is misplaced. A careful review of past and current demographic patterns provides little basis for the expectation that voluntary family planning programs can elicit major and independent shifts in local fertility patterns.

The availability of modern methods of contraception is neither a necessary nor a sufficient condition for sustained fertility declines. Fertility in most of Europe, after all, dropped substantially during the nineteenth century--long before modern contraceptives were even imagined. Moreover, fertility in some Western countries had fallen below the replacement level before the modern contraceptive revolution.

Conversely, the fact that modern contraceptives are available does not assure their widespread use. After a quarter of a century of government-sponsored family planning efforts, for example, less than 10 percent of the couples surveyed in Pakistan in 1990 and 1991 said they were using modern contraceptive methods.[30] If modern contraceptive technology were the decisive factor in determining national fertility levels, we would expect fertility levels to track closely with modern contraceptive utilization rates. They do not. In the aforementioned case of Pakistan, the total fertility rate (TFR) around 1990 is believed to have been almost 6 births per woman. Bulgaria’s utilization rate for modern contraceptives in 1977 was reportedly just about the same as Pakistan’s today--yet in the late 1970s the Bulgarian TFR was around 2.2!

The reason fertili