Population Policy: Ideology as Science

In recent years the poorer regions of the earth have been swept by a “population revolution” which, though it has attracted comparatively little attention, is nevertheless both unprecedented and pregnant with consequences for the peoples of the countries affected. This “revolution” has been taking place neither in the bedroom nor in the health clinic, but rather in the corridors of government. Among otherwise diverse countries in Africa, Asia, and Latin America, a single idea has rapidly gained currency: that a modern government should have a population “policy”—an array of laws and measures specifically aimed at shaping the composition, size, and rates of change of the national population.

India's was the first Third World government to endorse the principle of an active population policy. It took that step in December 1952, with the adoption of its first Five Year Plan, a document establishing as a long-run governmental objective the direction of the country's population toward “a level consistent with the requirements of the national economy.” By the mid-1980s, over thirty governments in the less-developed regions of the world were following suit. This group included the governments of six of the world's ten most populous countries. By the late 1980s, more than seventy of the world's governments reported that they viewed their national fertility or population growth rates as “unsatisfactory,” and that they considered policy interventions to alter these rates to be “appropriate.” As of 1990, more than three billion people were living under such governments: over four-fifths of the population of the less-developed regions of the earth at the time, and nearly two-thirds of the population of the entire globe.

The adoption of comprehensive population policy marks an eventful change in the conception of the role of government in civic life. In the past, governments were often called upon to perform duties with demonstrable demographic consequences—the regulation of immigration, for example, or the eradication of communicable disease. The demographic impact of such programs, however, was typically subsidiary to their intended purpose (the preservation of national sovereignty, the promotion of public health, etc.). The idea of harnessing state power to the goal of altering the demographic rhythms of society per se suggests—and indeed almost seems to require—a new sort of relationship between the state and the citizen.

The nature of this new relationship is indicated by some of the targets that have been set by contemporary population policies. The government of Bangladesh, for example, has committed itself to the goal of a total fertility rate of 2.34 births per family by the year 2000; women in Bangladesh today, however, are thought to have an average of just under five children. The government of Ghana, for its part, has determined that its national fertility rate at the start of the twenty-first century should be 3.3 births per family; yet Ghana's parents are currently guessed to be having an average of about six children per family. If Bangladesh and Ghana are to meet their established targets, both governments must apparently oversee a reduction in their people's fertility by roughly 50 percent in the next six years—and in Ghana there are as yet no indications of sustained fertility decline. Just how such a radical alteration of personal behavior in so intimate a sphere of life is to be achieved is not clear to outside observers; it may not even be apparent to the planners who envision these targets. But if such reshaping of national fertility patterns is to be set in motion by government action, there would obviously have to be direct, far-reaching, and even forcible state interventions into the daily lives of the overwhelming majority of the citizens of these two countries—as well as others with similar “targets.”

What accounts for the rise of population policy? “Population planning” as currently practiced in China—with its “birth quotas” and its manifold pressures and penalties to “convince” parents to have but a single child—may seem particularly consonant with the philosophical underpinnings of a Communist dictatorship. Yet a stringent and encompassing population policy, only somewhat less ambitious than Beijing's, has also been executed in Singapore—a society with a nominally democratic government. Indeed, the list of Third World governments committed to the goal of shaping the demographic pattern of their societies seems to cover the political spectrum, including not only dictatorships and one-party states (Haiti, Indonesia), but also monarchies (Morocco, Nepal) and a number of genuine constitutional democracies (such as Barbados and Botswana).

The revolution in government presaged by an activist population policy would thus seem to be based less in politics per se than in “science,” for in the final analysis it is the field of learning known as “population studies” that provides population policy with its raison d'etre. This field, it is widely believed, has advanced sufficiently to permit meaningful predictions of the impact of population changes on the social and economic development of both rich and poor societies. Insofar as modern governance is predicated on the idea that national directorates can and should act to improve the material well-being of their subjects, and insofar as a useful understanding of the specific economic and social consequences of population change appears to have been developed, it should not be surprising that a growing number of governments have seen merit in the prospect of shaping the demographic contours of their country so that national welfare and social prosperity might be “scientifically” advanced.

Unfortunately for all involved, contemporary population policies have in large part been promoted and adopted on the basis of a serious misconception. The relationship between population change and economic development is as yet rather poorly understood, the assertions of certain bold professionals to the contrary notwithstanding. For many of the relationships between the two that have been suggested are at best highly tentative, and in any event cannot be construed to imply causation. Still others are characterized by false precision or misplaced specificity. And despite the authority that population “scientists” today lend to the worldwide effort to promote birth control, convincing evidence that voluntaryfamily planning programs have resulted in sustained changes in fertility norms is still lacking. None of this augurs well for the populations upon which far-reaching and purportedly scientific population policies are to be applied.

The relationship between population change and social or economic change is a topic awesome in its complexity. Many of the issues raised by the “population question,” moreover, involve matters of the deepest personal conviction or preference. Any discourse on population quickly comes to touch, if only implicitly, on such questions as the nature of free will; the equality of man; the rights of the living and the unborn; the obligation of the individual to his group, his society, or his God; the sanctity of the family; society's duties to the poor; the destiny of one's nation or one's race; the prospects for mankind; and the value of life. Such questions are best, or at least most comfortably, addressed in consultation with one's conscience, creed, or ideology. Thus, in any analysis of so-called population problems, the element of faith, whether directly expressed or inadvertently demonstrated, is unavoidable. Although ostensibly secular in nature, population studies often exhibit many of the trappings most commonly associated with religious movements.

In recent decades much of the most influential thinking on the “population question” has acquired a certain messianic tone. It has not been uncommon for respected authorities to evoke the image of an apocalypse brought on by adverse population trends—and to justify programs requiring great sacrifice or exertion by the salvation from hideous alteratives that they alone would offer. In 1992, for example, the UN Population Fund announced that “a sustained and concerted program starting immediately” to curb worldwide population growth was essential, since current trends, in the words of Nafis Sadik, UNFPD's director general, create a “crisis [that] heightens the risk of future economic and ecological catastrophes.”

Such warnings have been issued before. Stanford University's Paul Ehrlich began his 1968 bestseller, The Population Bomb, with a prophecy: “The battle to feed all of humanity is over. In the 1970s the world will undergo famines—hundreds of millions of people will starve to death in spite of any crash programs embarked upon now.” The most optimistic of Professor Ehrlich's “scenarios,” involving a radical worldwide program of population control and resource conservation to reduce the world's total population to 1.5 billion (less than a third of its present level) in some future century, envisioned the death by starvation of about a fifth of the people alive in the world in 1968. By the same token, The Limits to Growth, the 1972 million-selling computer modeling study sponsored by the Club of Rome, produced simulation of future trends that suggested an impending “collapse” of global population—proportionately and absolutely more devastating than was the Black Death in medieval Europe—in the absence of a balancing of world birth and death rates by 1975, cutbacks in pollution, and substantial increases in industrial and agricultural efficiency.

Belief in such visions was not always diminished when predictions predicated upon them were decisively disproven by events. Instead, the demographic “true believers,” in the fashion of millenarians, often simply reformulated their prophesies so that they could not be empirically disproven.

Rapid population growth is not the only demographic phenomenon that has prompted observers to invoke specters of catastrophe. Dire consequences have with equal assurance been ascribed to population slowdown or decline. During the 1930s, noted economists such as John Maynard Keynes and Gunnar Myrdal warned that the failure of Western populations to reproduce themselves would contribute to unemployment, insufficiency of investment, agricultural crisis, and low living standards—precisely the problems that a more recent generation of population experts has described as consequences of rapid population growth!

The tendency to invest in population theories with an almost religious zeal—as well as to harness them to the service of political movements buoyed by public hysteria—might be considerably reduced if there actually were a body of knowledge demonstrably capable of explaining population change, or of connecting this change predictively with various determinants or consequences. Unfortunately, such an understanding of the process of population change does not exist. Although the postwar field of “population studies” has occasioned wide-ranging, detailed, and often ingenious academic investigations of the interplay between population, economy, and society, nothing like a generalized understanding of the socioeconomic causes or effects of population change can be found today—or even looks to be in the offing.

One is, of course, asking too much of demographers to expect them to provide any overarching “laws of population.” No one would demand of an historian that he provide a unified “theory of history.” For all the mathematical rigor required in some of its investigations, the study of human population is a field of social inquiry, not a natural science. Such relationships as it may uncover at given times in diverse places are, like all social phenomena, shaped by the factors of human values and volition, quantities that may seldom be reduced to fixed parameters.

It is instructive to consider some of the limits of today's “population science.” Despite the sophistication of population mathematics, the rate of growth of human populations cannot be predicted with accuracy over any extended period of time. Population projections from the fairly recent past highlight the problem. In the 1920s Raymond Pearl, then one of America's leading population biologists, predicted that the United States would reach a population of 200 million around the start of the twenty-second century; in actuality, America passed the 200 million mark in the 1960s. In the 1930s, France's foremost demographers agreed that French population was certain to fall; various projections indicated a drop of between two and twelve million people—that is, from 5 to 30 percent—between 1930 and 1980. In actuality, despite the losses it sustained in the Second World War, France's population rose by about 30 percent over that period. More recent refinements of techniques have added little to the precision of demographic forecasts. In 1959, for example, the United Nations' medium variant projection put India's 1981 population at 603 million—an estimate which, twenty-two years later, turned out to be off by about 100 million people. Long-term population projects are not always wrong, of course, but they can be right only by chance, for there is no scientific method for predicting either death rates or birthrates.

There is reason, moreover, to expect population projections to become even less accurate in the future. Advances in the fields of public health and public administration now make it possible for governments in low-income countries to reduce national mortality at an increasingly rapid pace—if these governments are resolved to pursue the necessary policies for doing so—even in the absence of more general social improvement. At the same time, the speed with which fertility change may occur has been increasing. In England and Wales, it took almost eighty years in the nineteenth and early twentieth century for the birth rate to fall by 15 points. In the People's Republic of China in the 1970s, a reduction in the national birthrate of about 20 points was accomplished—by whatever means—in a single decade. Even without aggressively anti-natalist measures, a drop in birth rates of over 15 points occurred in postwar Japan in barely ten years (1948–1958). With a growing possibility of the rapid alteration of demographic trends, the horizon of accuracy on population projections, far from extending, may be drawing ever closer to the present.

One of the reasons that long and even medium-range population projections are of such limited accuracy is that there exists no method for predicting fertility change in contemporary societies. There is, in fact, no method for predicting even the onset of fertility decline in those societies where birth rates are high, and seemingly stable. The search for social or economic determinants or preconditions for changes in fertility has in large measure proved frustrating, for societies of the modern world and the recorded past exhibit a breathtaking diversity of relationships between demographic, economic, and social conditions.

Low fertility, for example, is commonly thought to be associated with high levels of health; yet life expectancy in contemporary Kenya, where the total number of births per woman is currently believed to hover around 6.5, appears to be similar to that of Germany in the mid-1920s, where the total fertility rate was only 2.3. Fertility decline in nineteenth-century France proceeded even when levels of national mortality were considerably higher than those prevailing in Bangladesh today; by contrast, there is as yet no indication of secular fertility decline in Oman, even though its birthrate is thought to be 30 percent higher, and its life expectancy nearly thirty years greater, than that of France in 1830. It is commonly said that fertility and income are negatively correlated, but the limits of such generalizations are suggested by the World Bank's latest World Development Report, in which Tajikistan's level of per capita output and its birthrate are both depicted as being about twice as high as the corresponding figures for Sri Lanka. The great diversity of relationships that may be seen between demographic and social or economic conditions makes virtually any simple generalization about populations and development hazardous, for there is almost always at least one example that can be found to call the generalized relationship into question.

If there are broad difficulties with theories of population, there are also a number of serious problems in the study of population and development that are entirely practical. Foremost among these is the problem of false precision. In the statistical accounts used most frequently for the analysis of world demographic and economic conditions, numbers and trends are often presented with a degree of implied specificity that is entirely unwarranted by the actual margins of error surrounding them. Not surprisingly, this has often led scholars to erroneous or untenable conclusions.

The nature of the problem is suggested by the 1985 edition of the World Bank's World Development Report, the most widely circulated annual publication on development issues. In its statistical appendix, it presents an estimate for the population of Somalia as 5.1 million—implying a margin of error of 100,000, or about 2 percent. The same appendix gives Somalia a birthrate of fifty per thousand for both 1965 and 1983–again implying a 2 percent margin of error. It might therefore come as a surprise to learn that Somalia has no system whatsoever for registering births, and as of 1985 had never conducted a national census. Those figures for Somalia were essentially invented: guesses dignified with decimal points.

Somalia, to be sure, is an extreme example. Still, it would be unwise to assume the accuracy of most current demographic estimates. Near—complete vital registration systems cover only about a tenth of the population of the Third World—and in general tend to be least comprehensive in precisely those nations where the connection between population and development is of the greatest humanitarian concern. Indicative of the uncertainties attendant on measuring world population is the lag time between the estimated peaking of world population growth and the announcement of that event by demographers. It is now widely believed that the world rate of natural increase reached its maximum between 1960 and 1965, and has declined since then. Demographers, however, did not begin to suggest with any confidence that the world rate of population growth might have peaked until 1977–more than ten years after the presumed date of that event.

If demographic trends are beset with uncertainties, estimating economic output must be an even more tentative undertaking, involving as it does not only the measurement of populations, but also of their per capita production and consumption of goods and services—not to mention the valuation of these latter. For less-developed countries the problems of estimating levels of economic output and their rates of change can be arresting. To cite but one of many possible examples, a team lead by Irving Kravis of the University of Pennsylvania concluded that India's nominal gross domestic product per capita for 1970 would have been tripled if it had been measured not in terms of dollars at the prevailing rupee-dollar exchange rate, but rather in terms of the standard international costs of the goods and services that the “average” Indian produced. Kravis' estimates of purchasing power, however, did not apply to all poor countries equally. Measured by the regular exchange-rate method, 1970 per capita GDP in Kenya appeared to be 44 percent higher than in India, but after adjusting for differences in actual purchasing power, the Kravis team concluded that the Kenyan per capita GDP was almost 9 percent lower than India's. The effect of such adjustments on any presumed correlation between demographic and economic levels in the two countries, clearly, would be profound.

Population change itself further compounds the difficulty of measuring a society's economic welfare. Because children tend to consume less—often substantially less—than adults, standards of living at any given level of national income can be significantly affected by the age composition of a population. A rapidly growing population in which average age is comparatively low will appear to have a lower per capita level of consumption than a stationary population with a higher average age, even if at every single age people in both populations consume exactly the same amount. And changes in mortality can introduce even greater biases.

Innovations in public health during the course of this century have typically had a disproportionate impact on the mortality of children and on diseases associated with poverty. If poorer parents tend to have more children than richer parents, as is the case in many (but not all) low-income societies today, improvements in health could appear to be lowering per capita income and contributing to greater income inequalities in society as a whole, even though the families affected by these life-saving changes could regard themselves as indisputably better off.

Dan Usher, an economist at Queen's University in Canada, has attempted to impute an economic value to improvements in life expectancy in several developed and less-developed countries. Though his results are sensitive to his assumptions about interest rates and the self-perceived value of additional consumption, they are nonetheless interesting. By Usher's computations, adjustments to account for the value of lengthened lifespans to those whose lives were extended would have raised the nominal annual economic growth rate for Chile for the years 1931–1971 by over half, and would have more than doubled Sri Lanka's for 1946–1968. These may be extreme cases: health improvements in both societies were rapid during those years, and measured rates of economic growth were comparatively slow. Nevertheless, Usher's computations may suggest the sorts of factors that are ignored in conventional discussions of the impact of population growth on society and economy—and of the effect that ignoring such factors may have on the evaluation of the consequences of population change.

In the less-developed countries of Asia, Africa, and Latin America, governmental concern with population trends today focuses chiefly on the issue of rapid population growth. The problem is that “overpopulation,” a term so familiar and used so frequently as to suggest that it has a fixed and understood meaning, cannot in fact be defined unambiguously. Which is to say, there is no workable demographic definition of the idea of overpopulation.

Consider some of the possible demographic criteria by which to judge that a society is “overpopulated.” Is it that the rate of natural increase—the birthrate minus the death rate—is “unusually” high? If so, the United States in the first decade of its independence was almost certainly overpopulated: between 1790 and 1800 its rate of natural increase was 3 percent per year—a rate significantly higher than that ascribed to Bangladesh today, and roughly half again as high as the rates currently estimated for Haiti or India. What about birthrate itself? Again: the U.S. birthrate in the 1790s was about 55 per thousand—a rate higher than any of the 127 national birthrate “estimates” the World Bank gives in its World Development Report 1993, and at least 25 points higher than the latest World Bank estimates for India, Indonesia, or the Philippines. Perhaps, then, overpopulation should be judged by a country's population density, i.e., its ratio of people to land area. In that case, taking United Nations' figures for 1991, France would have been more overpopulated than the Indonesian archipelago, Japan would have been far more overpopulated than India, and Singapore (where government policy is now striving to raise the birthrate) would have been vastly more overpopulated than Bangladesh. But by far the most overpopulated state on earth would have been the Kingdom of Monaco! On the other hand, if the criterion were what is called the “dependence ratio”—the proportion of children under fifteen and adults over sixty-five in relation to the population of “working age” (conventionally designated as fifteen to sixty-five)—then, according to UN figures, as of 1960 Ireland and Nepal were approximately equally overpopulated, in 1980 Israel was more overpopulated than Sri Lanka, and the least overpopulated societies in the world as of 1990 were Singapore and Hong Kong! If emigration were to be taken as the measure of overpopulation, then Mozambique, Angola, and the other “front-line” states adjacent to the Republic of South Africa would be overpopulated, while South Africa, which is said to employ more than one million migrants from neighboring nations, would presumably be underpopulated. By the same token, the former East Germany, whose loss of citizens to West Germany proceeded with alacrity until the construction of the Berlin Wall, would have been an overpopulated country, its government's constant complaints about labor shortages notwithstanding. Finally, if overpopulation is indicated by unemployment (not strictly speaking a demographic measure, but one that is sometimes used as such), it would seem that the United States experienced serious overpopulation during the Depression (when fertility fell below replacement levels), and was least overpopulated in the mid-1960s (the years directly following the postwar baby boom).

“Overpopulation” does indeed point to a problem, but it is a problem mislabelled and misidentified. In a more careful discussion, the phenomena most frequently cited as proof of overpopulation would be deemed characteristics of poverty. Inadequate incomes, poor health, malnutrition, overcrowding, unemployment—it is images such as these that are conjured up by the notion of “overpopulation,” but they are unambiguously images of poverty and material deprivation.

Poverty and material deprivation, in all their various manifestations, cause great suffering, and in some cases government can do much to alleviate them—and even more, to provide an atmosphere conducive to broad social and material advance. But to mistake the great range of social and economic problems experienced by human populations for problems driven or created by demographic forces is a profound error. Rapid population growth is a pervasive fact of life in less-developed countries today—a form of social change so typical, and at the same time so profound, that it may spuriously be associated with almost any other social phenomenon of the present generation. Many of the problems typically ascribed to “population pressures,” however, turn out on closer examination to be caused by factors quite independent of demographic trends. And in the case of other such problems, the impact of demographic change is, at best, secondary. Attempting to redress such problems all too often simply compounds the difficulties of certain groups, and even of entire populations, in their efforts to maintain or improve their standards of living in the face of a hostile political or economic environment.

Yet, as noted, activist population programs have been embraced by governments presiding over something like four-fifths of the people of the Third World. Western governments and Western-funded multilateral organizations are currently spending over $1 billion a year on population programs in Third World countries. In some of these countries, such as Bangladesh, the budget for family planning by the 1980s was larger than the budget for all other health-related services combined. Thus it seems appropriate to ask, what do family planning programs, and national population policies, actually do? What is their demographic impact, and how do they affect living standards and future development prospects?

Voluntary family planning programs typically subsidize, advertize, or otherwise promote the use of modern contraceptive technologies by sexually active couples (usually but not always partners in marriage). Modern contraceptives are not necessarily more effective than traditional means of birth control; on grounds of effectiveness alone, nothing can improve upon total abstinence or infanticide. Modern contraception is, however, a much more pleasant alternative, and, used properly, can be more effective than the traditional methods of birth control (such as coitus interruptus, the rhythm method, or the local contraceptive potion). A cheap and readily available supply of simple modern contraceptives can allow parents who wish to make use of them to improve their own level of comfort, and may also (by facilitating the spacing of births) improve family health chances—even if their adoption has no ultimate effect on the size of the family. Such services would be seen as raising living standards today (albeit in ways whose benefits are not easily computed), and might even improve prospects for material progress by augmenting society's stock of “human capital.”

Government-sponsored voluntary family planning programs could under these circumstances best be justified as a public health service—one of the many activities a government may promote to reduce mortality, augment human capital, and improve the well-being of individuals and families. Unfortunately, many contemporary advocates of family planning programs for the Third World—be they decision-makers in local capitals or enthusiasts from Western countries—have been looking for much more than this. Typically, they see voluntary family planning as a technique for accelerating enforced reductions in fertility for societies where family size is large.

Their hopes are misplaced. As long as family planning is voluntary, it will remain a tool with which parents can attain the family size, or spacing, they desire. No research has so far suggested that the advent of modern contraceptives intrinsically alters parents' view of children, or ideals about the family. For this reason, similar patterns of contraceptive use may be associated with very different levels of parental fertility.

By World Bank estimates, 56 percent of the married women of reproductive age in Japan in 1989 used modern methods of contraception (pill, IUD, diaphragm, sterilization, etc.). Japan's total fertility rate for 1991 is put at 1.5 births per woman. In Turkey, where the rate of use is put at 63 percent, the total fertility rate is said to be 3.4 births per woman—over twice the level in Japan. That gap speaks not to the ineffectiveness of Turkish contraceptives, but to the decisive importance of the attitudes and intentions of the people who use them.

New inexpensive contraceptive technologies will, in many places, result in a decline in “unwanted” fertility by making birth control less prohibitively difficult. But it is unclear how large such a decline would be, and there is no reason to expect that it must always be substantial. The impact of reducing unwanted pregnancies, moreover, would not be as great on population increase as on fertility, since the mortality rate for “unwanted” births would be higher than average. And unless the very availability of modern contraceptives by itself stimulates a revolution in attitudes toward family size—as over a quarter of a century of family planning efforts have failed to do in Nepal, and three decades of programs have not done in rural Pakistan—the demographic impact of family planning would be a discrete and self-extinguishing adjustment, as the previously “unmet demand” of motivated users is progressively satisfied.

In many parts of the globe, an effective family planning program might actually increase the birthrate and the rate of population growth. In much of sub-Saharan Africa, for example, there is little demonstrated interest in modern contraception, but considerable concern about infertility. In those societies, the fate of a barren woman is unenviable. Helping parents attain their desired numbers of children might in such circumstances result in heightened fertility. The example of postwar Kenya, where the total fertility rate apparently rose from about six to nearly eight during a generation of substantial improvements in health, and despite nearly twenty years of family planning efforts, should make it clear that increasing parents' freedom to choose will serve the purposes of parents, whether or not these are in accordance with the preference or ideology of the government and its advisers.

The underlying thrust of most family planning efforts in less-developed countries over the past generation has been unmistakably anti-natalist. The anti-natalist sentiment behind support for family planning initiatives is explicitly expressed by the principal international institutions funding these initiatives, including the World Bank and the U.S. Agency for International Development. This sentiment has affected not only the allocation of public funds, but also the evaluations of performance.

In a number of less-developed countries, the commitment of public funds to family planning efforts is striking. In 1980, the World Bank estimates, government expenditure per current contraceptive user was $68 in Ghana and $69 in Nepal. These figures compare with World Bank numbers suggesting that the total governmental expenditure on all health programs worked out to be about $20

per family in Ghana, and $8 in Nepal, in the same year. In such societies, where general levels of mortality are high and health problems are pressing, it would appear that family planning programs are not subject to the same strict criteria in competing for scarce funds that other health programs must meet. The exemption of family planning from the ordinary constraints of budgetary finance may be due to belief that family planning provides additional “services”—such as lowering the birthrate. The unscientific faith that proponents of family planning have placed in the ability of their programs to “work” is suggested by the lack of careful investigation into the actual effects of family planning projects in trial areas. Although billions of dollars are being expended annually on family planning services in less-developed countries, and such programs have been promoted by various governments at the national level for four decades, there are only a few studies that have attempted to measure the demographic and health impact of family planning against a “control group”—a similar area which lacks the service. This lacuna is all the more striking in that “control studies” are standard practice in the health sciences. Some proponents of family planning programs have argued that there is an enormous “unmet need” (as distinct from “unmet demand”) for family planning services in less-developed countries. By their estimates, meeting this need would require substantially greater funding for these services in the less-developed countries than they currently receive. The advocates may be right, but no thanks to the strength of their analysis. Their method of computing “unmet need” is, typically, to measure by sample questionnaire the fraction of married women in various societies who say they are not using modern means of contraception and also say that they want no more children, or that they wish to delay the birth of their next child. It is by no means clear that this method measures either the unmet demand for modern contraceptives, or the fraction of the female population exposed to unwanted pregnancy. Women not using modern contraceptives may be practicing traditional (albeit less pleasant or less effective) means of birth control. Western fertility questionnaires, moreover, have a mixed record in eliciting accurate responses from poor people in less-developed societies. It is sometimes the case that illiterate villagers, whether in deference or out of politeness, attempt to please the interrogator with their answers. At other times questions are pursued in a way that can be of only limited relevance in the society in question. Most fertility questionnaires, for example, are devoted entirely and exclusively to the responses of women, even in areas where the tradition of male dominance invests much of the power of decision with husbands, or even fathers-in-law. Still other parts of such questionnaires may probe into areas of interest and salience to the interviewer but be of limited intuitive meaning to the respondent. In the World Fertility Survey, for example, over a third of the fecund women who indicated that they wanted no more children also said that they would like to have a larger number than they presently had. Given such problems, the results produced by fixed-form, rapidly conducted questionnaires are notoriously unstable. A major survey in Indonesia, to cite another example, found that fewer than half the women interviewed gave the same figure for “ideal family size” when they were re-questioned four months later. In places like Taiwan, Korea, and Hong Kong, family planning programs have been widely credited with bringing birthrates down. But fertility decline was underway in these areas before their family planning programs had been established. In India, by contrast, national fertility declines did not become unmistakable until decades after the adoption of the national planning program—giving the advocates of the program the opportunity to ascribe all change in fertility to governmental birth control efforts. In still other countries, like Mexico and Thailand, the establishment of family planning programs coincided with the onset of rapid fertility decline. It is not clear whether the alleged success of these programs was due to government commitment to push for contraception, or from a change in national attitudes toward family size and spacing (which may have prompted the establishment of the programs in the first place). In any case, in a growing number of countries, the governments do not seem satisfied with the family choices parents have been making. The chosen governmental instruments for altering fertility have ranged from campaigns of information, propaganda, and exhortation, to the exercise of incentives and penalties to promote desired patterns of fertility, to the use of legal sanctions against parenting, to the refinement and application of techniques of force against would-be mothers and fathers. Such inducements may or may not, under the given circumstances, be successful in changing national fertility patterns, or in bringing birth trends more nearly into consonance with the numerical goals targeted by government planners. But to the extent that they are effective in enforcing involuntary adjustments in the behavior of parents, they may be seen to reduce—not raise—current standards of living, and to compromise, rather than advance, future prospects for development.

What, in the final analysis, can be said with confidence about the consequences of demographic change on the prospects for social and economic development in the less-developed regions? Among policy makers, ideology has often been substituted for judgment, and emotion has overpowered caution, in the assessment of changes in population, with all that these changes seem to imply about the fate of national, ethnic, or local populations. It may therefore be appropriate to conclude with a few cautious observations.

First, the obvious: population change is a form of social change. Though the implications of population change are profound—no less so because it involves the generation and termination of human life—it is a form of change that seems slow by comparison with many others. A rate of population increase of 4 percent is considered extremely rapid; a rate of price inflation of 4 percent a year is, in most developing countries today, considered to be fortuitously slow. Moreover, for all the uncertainties of long-term population forecasting, the likely change in size and composition of a national population can be predicted over the course of the coming calendar year with far greater certainty than can changes in the harvest, the gross national product, the unemployment rate, the foreign exchange rate, or the demand for any particular product. Population change, like other forms of social change, emphasizes the ability of individuals, communities, and governments to manage and to cope. For populations that cope poorly with change, any quickening in the pace of change—including the pace of demographic change—is likely to prove difficult and perhaps even costly to society as a whole. Yet coping with change—or as Nobel Laureate in Economics Theodore W. Schultz put it, “dealing with disequilibria”—is in itself an integral part of the process of modern economic development. It is a learning process that generates real economic returns. To the extent that population change may prompt this learning process, it can even contribute to the acceleration of material development in a given society.

Second, insofar as demographic change may assume a variety of manifestations, its form in the modern era has typically been both comparatively benign and relatively advantageous for the purposes of economic growth. Worldwide population growth has been propelled principally by falling death rates, which is to say, by rising expectation of life at birth. Rising life expectancy is itself an indication of improved general levels of health, and is suggestive of other changes in living conditions supportive of general improvements in health. Improvements in health, moreover, may figure importantly in augmenting the human capital upon which the potential productivity of different populations ultimately rests. Augmenting human capital does not assure the acceleration of material advance—that will depend on many other things, including the environment of governmental policies in which human talents are set to work. But improved health, and its handmaidens, may well make it possible for the pace of material advance, under auspicious circumstances, to be quickened.

Third, much of the current discourse on the population “problem” seems implicitly to assume that the elimination of poverty may be served by preventing the birth of poor people. This appears to be a fundamental error—an elementary fallacy of composition. Mass affluence is typically associated with great transformations of society, economy, and individual outlook. The role of the family, the status of women, and the patterns of fertility typically change in the course of this transformation. But the extent to which reductions in fertility themselves stimulate improvements in human productivity is far less clear, either for given families or whole societies.

Fourth, while the economic consequences of overall rates of population growth are often ambiguous or obscure, the impact of smaller groups within larger populations on the economic well-being of the whole society, or the impact of differential rates of growth within a national population on prospects for material advance, may be direct and important. The contribution of religious or ethnic minorities to innovation or economic advance, as Lord Peter Bauer of the London School of Economics has emphasized, has often been vastly greater than would be expected by their numbers on society. By the same token, conflict between rival groups within a society may be intensified by differences in their net reproduction rates; such differences may have implications for the tenor of civil life or the composition and character of the national directorate, factors that may play a decisive role in determining the climate for development.

Finally, the likelihood of accruing unambiguously positive benefits from an active population policy devoted to shaping the demographic composition of society seems extremely low. This is not only because it has as yet proved impossible to define consistently the notion of the “optimum” population upon which activist efforts to shape the demographic composition of society seem in theory to rest. In its applications to date, population policy in the less-developed countries has often attempted to alter through demography social problems whose causes can be traced to ill-advised or injurious governmental policies. It is a peculiar theory of the “second best” that would suggest that the fertility decisions of a nation's parents must be altered because the distorting and inefficient policies promoted by its government are immutable. Yet thinking very similar to this seems to have governed the formulation of some of the population policies put into effect in the years since World War II.

In principle, this problem could easily be corrected. But even if it were, difficulties would remain with the notion of an active “population policy.” Interventions in education, health, housing, regional planning, and other areas may have demographic consequences, but they are justified on other grounds. On what grounds would a policy whose primary aim is to induce demographic change be justified? The acceleration of economic growth is one possibility (although, if current patterns of fertility reveal the preferences of households within the society about current and deferred consumption, it is not immediately apparent why such preferences should be ignored or overridden).

Consider what an active policy to shape population in the service of accelerated economic growth would require. One possible justification would be “market failure”: the possibility that, through distortions and externalities in the price structure of the economy, parents are encouraged to rear the “wrong” number of children. There are good reasons for overcoming market failures—entirely apart from whatever effect such reforms may have on childbearing. Attacking distortions and externalities directly, furthermore, would seem preferable to compounding these distortions through “compensating” economic penalties or rewards for additional births. And it would be necessary to ask, in a systematic fashion, whether there would be any reason to assume that in the absence of distortions and market imperfections, parents would not be able to “price” the birth of their child correctly.

To make the economic case for an active population policy, population planners would ultimately need to center their arguments on estimates of the economic value of human life. They would have to show, in effect, what would be the “present value” of a child born today, and also to show how that present value would be changed by altering the size of the baby's cohort of peers, or the cohorts following. If attempted seriously, this would prove to be an extraordinarily difficult task. Assessing the present value of any given, innate, physical investment over the course of its projected “life” is at best a tentative and highly uncertain undertaking. Albert Hirschmann, the eminent economist at the Institute of Advanced Studies, has noted that many of the factors that affect the actual productivity of an investment project once it is in operation are not foreseen in even the most detailed exercises in planning. If estimating the present value of innate objects is difficult, how much more uncertain it must be to ascribe such a value to an entity imbued with life!

Long-term economic forecasts are notoriously inaccurate. Japan, Taiwan, and Korea are among the nations that defied all expectations of informed foreign observers in the postwar period; by the same token, Burma and Ghana, two of the countries most frequently cited as likely economic successes after independence, did an equally creditable job of disappointing the experts. It is hard to imagine how a population planner would have arrived at an accurate assessment of the present value of a birth in Korea in 1955, much less make confident assertions about the ways in which the present value of that birth would be changed through alterations in the contemporary Korean birthrate. Population planning for “development,” it seems, is easiest to envision under the assumptions of total technological stasis, absence of social change, and rigid restriction of social choices and alternatives. Yet such assumptions, in themselves, contradict the method of economic development and defeat its purpose, which is, in essence, the extension of human choice.

Nicholas Eberstadt is a fellow at the American Enterprise Institute, and a doctoral candidate in government and political economy at Harvard University.