The Future of Global Population

Nicholas Eberstadt

Winter 2022

Homo sapiens have probably always instinctively worried about the survival of their kind. And since the beginning of recorded history, civilized peoples have been considering the future of human numbers intellectually. In the book of Genesis, the Lord tells Abraham that his descendants shall be as numerous as the stars in the sky and the grains of sand on the seashore — very good news, of course, but a sum beyond reckoning at the time. Since then, mortal man has been busy developing quantitative techniques to enumerate populations he cares about, and to calculate the hypothetical world population of humans — their quantity, their geographic distribution, and their profile — far into the future.

Population enumerations known as "censuses" originated in the imperial decrees of standing empires in East Asia and the Mediterranean — including one that famously involved the town of Bethlehem — thousands of years ago. But government capabilities and priorities have really only allowed for a relatively accurate estimation of the size of the global human family since the 1980s, when the post-Mao Chinese Communist Party divulged to the rest of the world the tabulated contours of China's enormous population.

Of course, the tools for computing future trajectories of world population were in hand long before the China reveal. Those came out of simple arithmetic and algebra, and were honed into a particular sort of applied mathematics in the field of inquiry known as "demography," which took that name in the mid-19th century. Although estimating and projecting national and global populations might seem like a negligible task from the lofty vantage point of pure mathematics or theoretical physics, the practical instruments required for these undertakings were not self-evident — they required some effort to design and improve. Gradually, demographers assembled the foundations for today's quantitative approach to analysis and projection of national and global population trends: life tables, stable-population theory, the cohort-component analysis method, and, most recently, stochastic projections.

Taken together, these and other allied demographic tools now permit a formal, detailed, and, in some ways, elegant quantitative assessment of the prospective evolution of any population's size and structure. The project is critically abetted by the convenient fact that, since extraplanetary travel and interspecies mutation are off the table (at least for the time being), the human population is a closed system: The initial population net of births, deaths, and migration must equal the population of the end state, with everyone aged accordingly. In short, everything must add up from one period to the next.

Such tools afford precise and exacting calculation of the coming population of the world and its subsidiary political groupings. But there is a catch: To use these modeling methods, one must "input" assigned values for future mortality, fertility, migration, and more. And at this portion of the program, we find today's demographers almost completely at sea.

Though they do well enough with long-term survival trends under non-catastrophic conditions (their actuarial insights underpin the life-insurance industry, which has not yet gone out of business), population experts have nothing reliable to offer about the longer-term outlook for fertility or migration — the other indispensable elements of population arithmetic. So it is that almost all population projections issued nowadays by authoritative and expertly staffed institutions are cautious in nature — incrementally inching fertility assumptions upward or downward from today's registered levels, guessing that tomorrow's international migration flows will be the same as today's plus or minus a couple of points.

A field of study in this condition is asking for a shock. And this was the state of play in the demography business when Max Singer — the late polymath, iconoclast, and co-founder of the Hudson Institute — intruded with his 2011 opus, History of the Future.

Singer was never afraid to make bold, sweeping, and confident predictions. But he also had the uncanny capacity to make those predictions in ways that poked at the weaknesses of confident professionals in various arenas. And this was certainly the case with that still underappreciated work.

Early in the book, Singer describes the characteristics and dynamics of a future world of modern, free, developed societies with some specificity. When it comes to population, however, he opines that

the method used in most of this book doesn't work; since modern demography has not had enough time to reach a steady state it is quite possible that the demography of already-modern countries does not reveal how modern countries will behave in the future. Also, the future in this as in many other areas depends on future human choices and values, not on the inherent features of modernity.

The future global-population prospect, he observes, will turn mainly on childbearing patterns in the years ahead — but significantly, in his judgment, "we cannot say that below-replacement-level fertility is an inevitable result of modernization....Fertility rates are independent of death rates; they depend on people's desire for children."

This brings Singer to a consequential conclusion about the future demographic outlook of our planet. In his words:

Demographers have long proposed that...birth rates would eventually come into balance with death rates, and world population would level off....In fact, the experts are importantly wrong. There is no justification for the comforting theory that in the long run, once we have adjusted to new living conditions and low death rates, birth and death rates will be more or less the same and world population will be reasonably stable.

Under conditions of modernity, he argues, a society's long-term fertility level could be either below replacement level or somewhat above it, depending on attitudes about the family and other key values that are not going to be predetermined by material conditions or other predictable factors. This means that the future population of our planet could be vastly different from today's in either direction. "In short," he writes, "one of the things we don't know about the modern world of the future is how many people will live in it." He posits the fantastically useful concept of a Columbus unit — a period of about half a millennium, the time between Christopher Columbus's arrival in America and our own day — and suggests that the world's population a Columbus unit from now might be as low as 250 million or as high as 250 billion. Either trajectory, he insists, could be "both feasible and plausible."

Singer's prognoses were not taken seriously by most demographers when History of the Future was released 10 years ago, and they would perhaps enjoy only a slightly less frosty reception from the profession today. Singer is an untrained interloper and a disciplinary trespasser; he does not use the special language of the guild, know its secret handshakes, or acknowledge its totems. And his exposition makes a few small errors that, while ultimately trivial, provide pedants and small-minded know-it-alls an excuse for wholesale dismissal of his work.

Yet if we dare to look at the big picture, as Singer invariably did over his lifetime, we can see that he grasped much about the population question that others, including many ostensible demographic experts, did not understand back then and still do not understand today. There is an impressive wisdom, nuance, and analytical power in his demographic disquisition, although his simple, unassuming style of reasoning and writing may lead some readers to breeze through and underestimate a text worth pondering. Instead of following suit, we should take up the challenge Singer's boldness proposes and apply the tools of demography beyond their usual medium-term boundaries.


Singer focuses on fertility, recognizing what demographers have long known: that under conditions of orderly progress, fertility is the dominant driver of population growth and structure — the most important factor in any regularly convened population's size and composition.

Over the past century and a half, social science has striven to understand the forces accounting for differences in fertility within and among nations, and for changes in fertility patterns and trends over time. A vast compendium of studies has examined all manner of factors — material, cultural, psychological, technological — that might be determinants of fertility or fertility change. Income, education, urbanization, infant mortality, the status of women, and the availability of modern contraceptives are just a few of the preferred theoretical influences on fertility that the past generation or so of researchers has been examining. But as social historian Charles Tilly pointed out in the 1970s, "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." This unsatisfactory state of affairs has far-reaching implications, for if scholars cannot robustly explain fertility change in the past, they certainly cannot predict it in the future.

In the 1990s, Lant Pritchett (along with Lawrence Summers) provided a major, if seemingly simple, advance in the study of fertility's determinants: They showed that wanted fertility — the number of children women say they desire — is a persuasively accurate predictor of fertility, both across countries and over time, even after statistical controls. Although this finding was a reassuring reality check of sorts — confirming human agency in family formation, as opposed to heedless animal procreation — it was not welcomed by all quarters of the population-research community, least so by would-be population planners who asserted a sizeable "unmet demand" for family planning in low-income countries purportedly beset by "unwanted excess fertility."

Singer is undistracted by these little controversies, proceeding instead as if it is plainly obvious that fertility is a matter of desired family size. "The world we get," he argues, "will depend on how many children people want to have and to raise." In so saying, he dismisses the materialist and developmentalist temptation to think about fertility patterns in so-called "structural" terms.

It's an assumption that suggests we have been living through a revolution in human desires. Today, the global fertility rate is down by over half since the 1960s. With India and Bangladesh as new members to the club, over two-thirds of humanity may be living in sub-replacement-level countries and territories. All around the world, people have more of almost everything, from income and education to housing, appliances, and free time — almost everything, that is, except children. Whether this great worldwide turn to smaller families is due to the rise of some common, newly shared outlook, or to meaningfully different, localized viewpoints that happen to push childbearing trends in a common downward direction, is an as-yet unanswered question.

In the 1980s, European demographers came up with the theory of the "Second Demographic Transition" to explain their region's slide into a seemingly permanent sub-replacement fertility regimen. They pointed to the fragmentation of the traditional family, the rise in single-parent homes, and the new normal of co-habitation and unmarried motherhood as the immediate culprits. But the driving engine, according to this exegesis, was an overarching shift in values among the rising generation of prospective European parents — an almost heliotropic turn toward what the theorists called "self-actualization."

We may wonder whether this describes what has taken place in the still unfolding birth collapse in East Asia, a region steeped in its own very different tradition. Co-habitation and premarital sex may be on the rise in many of those societies, but illegitimacy and single parenthood are, as a rule, still severely stigmatized. Instead, we see what has been termed a "flight from marriage" as well as a rise of mass childlessness in Japan — the region's demographic pioneer — leading the way to a future in which a slight majority of today's young women complete their lives without any biological grandchildren.

As to new norms about premarital sex, Japan offers the curious and contrary example of a country in which the rising generation of young people are increasingly unlikely to have ever had sexual relations with a partner and increasingly likely to express an aversion to sex altogether — a pattern so seemingly atypical of our species as to evoke thoughts about pre-extinction.

And then there is the plunge into sub-replacement underway in low-income societies. Today, the overwhelming majority of people in sub-replacement societies live outside the affluent and democratic West. The case of China might seem to be explained by Beijing's monstrous one-child policy, but that program was shelved by 2016, and available evidence suggests the mainland has marched even deeper into sub-replacement since then. In the greater Middle East, a number of Muslim-majority countries are either flirting with sub-replacement or embracing it with abandon, the poster child for the latter being the Islamic Republic of Iran — a closed society run by a repressive clerisy whose capital reports a fertility level lower than Zurich's.

Now it is true that Iran meets some criteria for a "modern society" — it is highly urbanized, for example, and many or most men and women of childbearing ages there appear to have at least a secondary education. What then to make of impoverished Myanmar/Burma — predominantly rural and still struggling with mass illiteracy — where four decades of sustained fertility decline has brought the nation to the threshold of sub-replacement, or maybe even beyond?

As a practical matter, we have no idea how low fertility can go, or for how long; we only learn by looking in the rearview mirror, after the latest demographic results have been tabulated and released. In the early 1950s, Luxembourg was the lowest of the low, with just under two births per woman; a half-century later, it was the territory of Macao, with an eye-popping fertility rate of 0.85 — almost 60% below replacement level. Macao, of course, is more of a city than a nation. But today, countries like South Korea and Taiwan are registering national fertility rates approaching 1.0; last year, South Korea's capital city of Seoul reported a rate of just 0.64 — less than a third of that required for long-term population stability without in-migration. Within living memory, most demographers would have said such patterns were impossible in times of peace and plenty. Popular behavior — mass decisions by ordinary people about the size and contours of their own families — may continue to confound the experts.

But Singer takes the fact that we have lived through a period of drastic and hard-to-explain changes in fertility to mean that things might change again — at least in the long run. A prolonged return to above-replacement fertility rates at some point in the coming generations is surely possible. The point is crucial to Singer's vision of the demographic future, and most demographers today don't rule it out. The U.N. Population Division's (UNPD) "high variant" fertility scenario envisions a contingency of steady, above-replacement fertility for the "more developed regions," though the experts are hazy about just why this should be possible.

Singer invokes the United States as an outlier with regard to the general Western disposition toward sub-replacement. Evidence was on his side at the time he was writing. Indeed, for almost a generation before History of the Future appeared, America's fertility rate was very close to replacement level; I myself wrote of "American demographic exceptionalism" thenabouts. Barely a decade on, U.S. childbearing patterns no longer look so exceptional: Fertility has been sliding steadily since the Great Recession and, in 2019, was almost 20% below replacement level. With the Covid-19 pandemic shock, U.S. fertility fell further in 2020 — and probably did so again in 2021.

But Singer also adduces the case of Israel — his own beloved second home — to make his point. There, it looks as if he is on much stronger ground, at least for now. Israel is a fully modern society — except for its fertility level, which, in 2019, was 3.01. The fertility rate of its Jewish population was 3.09 — very close to the 3.16 reported for Israeli Muslims — and has risen markedly since the early 1990s, when it hit a low of 2.62. Ultra-Orthodox communities are of course part of Israel's population arithmetic, but above-replacement fertility is also characteristic of the less observant and even secular strata of the Jewish population there. Overall fertility for Jerusalem was just under 3.9 in 2019 — and by the looks of things, quite similar for Muslims and Jews. By these figures, fertility levels in the capital of this modern society would have been two or three times as high as those in the capitals of Mexico (1.3), Turkey (1.5), and India (1.5) at the time.

This fascinating demographic laboratory has continued to surprise contemporary students of population, exposing the limits to our understanding of demographic science again and again. Contemporary Israel's demography is indeed unique among modern Western states, but is its example generalizable? Or does it depend instead on inherently irreproducible qualities like Zionism, the call of modern Judaism, the constant testing by hostile neighbors of the nation's will to survive and flourish?

Singer downplays these aspects of the Israeli experience, discussing instead the family- and child-friendly ethos there and implying the same could take root in modern societies elsewhere. He may well be right, but it is worth pondering exactly what sorts of intellectual, moral, or political circumstances might be required to elicit a sustained surfeit of births over deaths from modern peoples less certain of their covenant with their Creator.

Barring catastrophes of Biblical proportion (which have occurred with some regularity, and not only in Biblical times), Singer seems likely correct that, at some point in the future, humanity will mainly or entirely be comprised of "modern" societies, broadly defined — at least for a while, until something else comes along. And his prognosis for world-population trends is also generally persuasive: It is possible to imagine a total world population much smaller than today's, under conditions of orderly progress, but also one much larger. Either is more plausible than the presumption that some kind of homeostatic self-regulatory mechanism will suddenly assert itself. We have seen scant evidence of this over the past thousand years, let alone the past century. And population trends observed in other species do not provide adequate guidance for our own, given that we can purposely control and progressively alter our own mortality schedules and fertility patterns, along with the resource base that sustains our total population.

Within amazingly broad boundaries, we can also choose our own living arrangements, child-rearing traditions, and family constellations. Even the sociobiological assumption that humans have a built-in urge to continue their own lineage — what scientists call a "nurture imperative" — seems to be open to question in light of recent observed experience. A more cautious formulation might suggest that those familiar and seemingly universal traits are witnessed in humans whose lines survive, and in populations that are not replaced.

This conspicuous capacity of human beings to alter their own demographic rhythms makes the forecasting of their future population numbers even more hazardous than for other species. Demographers have attempted to finesse the conundrum most recently with the development of stochastic projections, which incorporate a probabilistic range of future trajectories based on parameters established from amassed observations of previous demographic experience.

These projections are an inventive exploration in applied math, and Singer gives a nod to them in his text. But like many demographic professionals today, he ultimately overestimates their promise. Stochastic projections are just fine for charting the likelihood of an outcome governed by fixed and immutable mathematical or physical relationships — a full house in poker, say, or a super-high tide. But the sum total of human demographic experience at the time of Columbus would have badly misled the stochastic projections of where the world ended up today. The same might well be true of our using today's experiences to guide projections about human fertility and mortality patterns a dozen or so generations from now.


You don't have to be a demographer to know that compound interest is a fearsome force. In population projections, sustained deviations from replacement fertility bring us into unfamiliar or even unimaginable new worlds within a surprisingly small measure of historical time.

For an idea of the range of alternative population futures that leading experts within the demographic profession regard as plausible for the year 2100 — about three generations from now — consider illustrative global projections from Austria's Wittgenstein Centre for Demography and Global Human Capital. Researchers there contrast several scenarios, and the result is a foretaste of what may lie in store — one that is in some respects even more daring than Singer's.

In the lower-fertility scenario, global fertility is approaching 1.3 by the end of the 21st century. The world adult population averages 14.5 years of schooling, with male and female life expectancy both approaching 100. The median age is over 60. In this vision of the future, total population peaks at 8.7 billion in 2050 (up from about 7.7 billion in 2020) and then falls to 7.2 in 2100. On that fertility path, humanity would total around 250 million by 2200 and under 100 million by 2300.

The world of the higher-fertility scenario, meanwhile, has a notional population of over 13.6 billion in 2100. With a hypothetical combined life expectancy not yet at 75 and less than 9.5 mean years of schooling for adults, this version of 2100 would pass muster as "modern" under Singer's definition. With a fertility rate of over 2.3, it would be on track to surpass a population of 100 billion by 2600 — a Columbus unit or so away from us on the other side. Getting to 250 billion by then would take more "lift," but the UNPD's "constant fertility" scenario, which holds global childbearing patterns at their 2019 levels for the rest of time, would manage to accomplish this feat a good bit more quickly. The world would hit 250 billion in this scenario even with unending sub-replacement fertility in today's more developed regions between now and then — after all, a rapidly growing world may have waning as well as waxing populations within it; over the long haul, it is only the tempo of growth in the waxing areas that sets the global pace.

Singer sees this point, too. But after taking us through the arithmetic that points to his vastly larger planetary population projections for some centuries hence, he commits a sort of rhetorical hit and run, leaving readers whose credulity may already be strained to take his word for it that such outcomes are indeed viable. There is no need to consider the question of how such a populous world could sustain itself, he avers, since centuries from now "there will be a lot more technology and knowledge, and the question is likely to look rather different from how it looks today." "In brief," he concludes, it's "not our problem."

For those inclined to take Singer seriously — as I am — that dismissal is flatly dissatisfying; for some readers, it might raise questions about his entire treatment of population. Nonetheless, I believe he is correct to maintain that a future world population of either 250 million or 250 billion could be in keeping with a modern society of the type he describes. Singer's mind-stretching demographic conjectures are merely outlandish, not positively insane.

The case for a prosperous modern society with 250 million people in it seems straightforward enough. Picturing a world containing only two Japans, trading extensively with each other and exploiting all the Earth's natural resources in their commerce, is not really much of a leap. Challenges and constraints would surely arise from long-term population decline, but a question about the planet's carrying capacity would not be one of them.

Imagining a world with 250 billion people — Singer's most provocative scenario — is a whole different matter. At the moment, it would be patently impossible to house, feed, and take care of 30-plus times the world's current human population even at the most miserable margin of survival. But Singer's assertion is neither hopelessly frivolous nor impossibly fanciful for a world one Columbus unit away, especially if we consider the scientific and technological leaps that separated the world of 1492 from our own.

Let us consider the carrying-capacity problem of a 250-billion-person world in the most simplistic of terms by briefly addressing three of the most obvious and unforgiving constraints: shelter, food, and energy.

Taking shelter as our starting point, let's assume our 250 billion people will reside on the land surface of the planet, currently 57.5 million square miles. This works out to almost 4,400 persons per square mile (including the Sahara, Antarctica, the Himalayas, etc.). By that metric, the entire world would be about as densely populated as current-day Malta, and over three times as densely as South Korea. This sounds pretty cozy. If we posit that each person is allotted 2,000 square feet of personal space, 250 billion people would occupy nearly one-third of the planet's land surface — about 18 million square miles, or, roughly speaking, the combined land mass of North and South America plus Australia. On the other hand, if these same 250 billion individuals inhabited a mile-high, 100-story structure with 50-plus-foot ceilings per floor, the entire edifice would fit inside the borders of Texas. A corresponding structure for "just" 100 billion people would fit within the borders of Nebraska. It may not be a way of life that many today would choose, but our nomad ancestors might feel much the same upon learning of our current (and almost immeasurably more prosperous) urban existence.

What about food? Today, we could not feed anywhere near 250 billion people. But we could feed more than our current planetary population — by some assessments, vastly more. And this has been the case for quite some time. In the 1970s, Roger Revelle — president of the American Association for the Advancement of Science and a leading American scientist of his day — published two studies in Scientific American estimating that, at the time, the world had the resource capacity to produce enough food annually to feed around 40 to 50 billion people.

To be clear, Revelle's was a logistical engineering-style assessment of potential production capacity that assumed all would be supplied with a modest diet of starchy staples and adequate nutrients for a healthy life — a command-planning view of what rationing could provide, not an economic inquiry into what paying customers would prefer for themselves. And it was hotly contested at the time: Paul Ehrlich, the specialist on insect populations and author of The Population Bomb, contended in 1971 that the world could feed no more than 1.2 billion on a sustainable basis, or roughly a third of the world population at the time. But Revelle was no amateur on the subjects of environment and sustainability — in fact, in the 1950s, he was one of the first researchers to establish the connection between atmospheric carbon-dioxide concentrations and atmospheric temperature, the foundation of current scientific work on climate change.

The question of how many mouths our planet can feed is no longer avidly and constantly investigated by scientists, and for good reason: The global food crises of the mid-1960s and early 1970s are far behind us. Indeed, for a generation or more, the discussion of a global Malthusian specter has been relegated to a fringe of quasi-religious true believers. Yes, hundreds of millions still suffer from undernutrition and food insecurity, but in a world patently awash with food and sloshing in wealth, this is clearly a purchasing-power problem increasingly characteristic of the few, not a production problem for the many.

If we were to revisit the question of how many mouths the world could feed today, we would be addressing it with today's technology. Presumably, our production possibilities are considerably greater than they were back in the 1970s, and in future centuries, those vistas could perhaps expand vastly further. Current global food production is constrained by, among other things, the realities of photosynthesis: At present, plant life converts solar energy into biological mass with an efficiency of under 5%, and the efficiency of conversion to edible matter is yet lower. In nature, a conversion rate of up to 47% is observed in certain algae, while under laboratory conditions, a photon can be converted into cellular chemical energy with an efficiency of up to 73% — this in existing, unmodified plant life. Such frontiers mean the potential planetary yield of foodstuffs could be much higher than what we imagine it could be today, even with some of the world's potentially arable land displaced for the habitation of a notional 250 billion people.

And what of the energy needs of such a populous planet? Powering a society containing 250 billion humans is, under present parameters, an unthinkably daunting prospect. As of 2019, over five-sixths of the world economy's energy consumption comes from hydrocarbons and fossil fuels. Raising global per-capita energy use to U.S. levels would imply something like a tripling of world energy consumption, and a global population of 250 billion would be over 30 times larger than the current total. Even setting aside the hardly incidental question of how to provide for more than a hundred-fold increase in commercial-fuel supplies, the steady consumption (that is, burning) of monumentally greater amounts of non-renewable energy would have huge, and very possibly calamitous, consequences for the global environment, the climate, and perhaps even human life.

Fortunately, there is an extraordinary source of renewable energy that humanity has barely begun to harness commercially. It is, for all intents and purposes, inexhaustible, as well as completely environmentally friendly. While it does cause significant warming, relying upon it to power the human economy will not cause climate change. I am referring, of course, to the sun.

The star around which our planet circles bathes it in non-stop energy. The sheer magnitude of this flow is difficult to comprehend, even when put in terms of numbers. According to estimates from Sandia National Laboratories (we could use other sources, of course; the figures are not in dispute), annual earthly sunshine has an energy value of almost 90,000 terawatts. That is nearly 5,000 times as much energy as humanity currently consumes from all commercial sources combined. Even if every person in a population of 250 billion used 10 times as much energy as the people of the world do today, this would still amount to less than a tenth of the annual solar radiation absorbed by our planet. If humanity could devise extremely efficient methods for the capture, storage, transmission, and application of such energy — each of these a gigantic "if" — it would theoretically be possible to support a mind-boggling quarter-trillion humans on much higher living standards than any national population has yet enjoyed — and to have plenty of solar energy left over for the cultivation of food, the flourishing of a planetary green canopy, the maintenance of healthy oceans and aquatic life, and so on.

But what if human ingenuity proved not up to the task of converting sunlight into commercial power on the aforementioned scale? Leaving aside the possibility of fossil fuels altogether, there would always be the option of relying on nuclear power.

Centuries from now, we could have currently unimagined possibilities for amazingly safe and efficient nuclear technology. With perfect conversion of mass into energy, the energy use of a population of 250 billion, with each person harnessing 10 times as much energy as our current planetary average, could be satisfied for an entire year on the basis of a cube of water just 100 meters in length on each edge. A cubic mile of water would contain enough energy to satisfy those same needs for over 4,000 years. Lest one fear we might run out of water this way, over 300 million cubic miles of water are found on the Earth's surface. And in any case, water would not be the only non-renewable resource that might be so converted — the mass of planetary water reserves are infinitesimal compared to its iron reserves, for example.

None of these musings is meant to suggest the desirability of that future world — I doubt many of us alive today would want to live in such a place, even if it promised more freedom and higher living standards than we ourselves enjoy today. Furthermore, I do nothing here to demonstrate that such a world is either viable or sustainable. But it could be viable and it could be sustainable given a Columbus unit of technological advance — and it could be so without suspending the laws of thermodynamics, resorting to interplanetary mass migrations, or throwing ourselves on the mercy of wormholes to portal our descendants to alternate multiverses.

This is an extended way of saying that, from our current vantage point, Singer's judgement about the range of plausible world-population alternatives some centuries from now looks about right, at least to me. The choices of future generations could take us very far in either direction, and we should not pretend to know which way those choices will point. In this regard, Singer's boldness is not excessive; indeed, in some important ways, it may fall short.


Simon Kuznets, the Nobel laureate and great student of modern economic development, often made the point that experts in economics and other disciplines did not have a good track record of discerning what the future might look like. In his view, science-fiction writers were better at this, and not by accident. As he explained:

Experts are usually specialists skilled in, and hence bound to, traditional views; and they are, because of their knowledge of one field, likely to be cautious and unduly conservative. Hertz, a great physicist, denied the practical importance of shortwaves, and others at the end of the 19th century reached the conclusion that little more could be done on the structure of matter. Malthus, Ricardo, and Marx, great economists, made incorrect prognoses of technological changes at the very time that the scientific bases for these changes were evolving. On the other hand, imaginative tyros like Jules Verne and H. G. Wells seemed to sense the potentialities of technological change. It is well to take cognizance of this consistently conservative bias of experts in evaluating the hypothesis of an unlimited effective increase in the stock of knowledge and in the corresponding potential of economic growth.

In his excursions into demographics, Singer was most assuredly an "imaginative tyro." Though he lacked professional demographic training, he was also wholly unencumbered by the caution and undue conservatism that training could impart. He was free as well from other biases characteristic of, and perhaps even useful for success within, that learned guild.

Singer's independent and unconventional way of thinking, seasoned by his uncommon vision and leavened with his signature intuition, made him a most welcome intruder into the domain of population studies. Peering ahead five centuries into the demographic future is just not something most properly minted demographers would dare to do. And yet, that hardly makes the venture unsound, much less worthless. It would be amazing if Singer did not turn out to be wrong in many respects in his forecasts for that distant future world, but his diagnostic approach is intrinsically valuable and instructive: Indeed, it illustrates the virtue and the benefit of exercising a trained and informed imagination — what economist Marcus Noland in another context called "rigorous speculation" — on the big questions of world affairs.

But at this point, having praised Singer's boldness at some length, it is time to chide him just a bit for his timidity. In his plunge into the demography of our descendants a Columbus unit removed from our time, he may actually have been unduly cautious and conservative himself — too steeped in the sense that the ways with which we are familiar are the natural order, and that things shall always be thus. Singer may have figured he had tested his readers' patience and trust quite enough, and so feared losing them altogether. But since he set the length of the field — one Columbus unit of time, to be precise — it seems unsporting not to play on it with gusto.

To begin, Singer's assumptions about future mortality and life expectancy might be far too conservative. On this, he asserts, "the upper limit on life not much more than eighty-five or ninety-five [years]." Singer may be right — this is a hotly contested issue in the life sciences, and some respected researchers side with him. But others do not, and the contrarians have some compelling evidence on their side.

For about 180 years — since 1840 — the world's top female life expectancy at birth has shot unrelentingly upward, and the trajectory is uncanny: an almost straight line traversing three centuries. The "best in class" populations have changed over time — from Norway to New Zealand to Japan and, most recently, Hong Kong — but the gradient has not; instead, it has held steady at an almost three-month increase for every calendar year. And for nearly a century, these data have consistently and sequentially shattered expert pronouncements on the ceiling for human life expectancy.

As of 2017, the top female life expectancy at birth — 87.6 years, in Hong Kong — was on the same slope as it was in the 1840s, and with no sign of slowing down. In modern societies, ever more people are living to once impossible ages. According to the Human Mortality Database, over 60% of Swedish women born in 1928 lived to age 80, while about 29% lived to age 90 — staggering proportions that have been steadily increasing over time. Corresponding year-of-birth data are not available for Hong Kong, but period life tables point to a conditional 80% survival rate for women to age 80 and a 50% survival rate to age 90 on current mortality schedules. Even without programmatic human enhancement, we might be on a path to a life expectancy above 100 years — and perhaps far above that amount — for most earthlings, and maybe not too far into the historic future.

An ordinary lifespan in excess of a century would invite a major rethink of where education, child-rearing, work, and extended periods of leisure fit within the ordinary life-cycle routine. There is also the question of family. With long lives, low fertility rates, and pervasive voluntary childlessness, a sizeable fraction of the future human population could inhabit societies where the extended kinship networks so familiar to the human experience become a bygone tradition. These would be societies with few siblings or children and only the odd collateral relative — where a person's closest biological kin would be his living ancestors. In such settings, we might find out exactly how sociable humans can be — how plastic and malleable the seemingly human desire for family actually is, and how serviceably the substitute of Johann Goethe's bruited notion of Wahlverwandtschaften, or "elective affinities" (i.e., chosen relations) can perform in practice. Indeed, we may have already entered a time when humans will not deem it unnatural to define "family" exactly as they wish, no matter what they might wish.

Finally, there is the great and terrible question of future fertility trends. We take it for granted that we human beings are driven by a deep urge to replace ourselves. Learned and elevated thinkers have argued that this is a biological call, fused into our very DNA. An entire sub-discipline of science known as "sociobiology" has been examining and codifying such notions for decades.

But we come to the question from the perspective of lineages that have been continued over time, and the brute fact of the matter is that most human lineages have not done so. To my knowledge, no one has yet attempted a careful calculation of how many family lines have died out for the roughly 100 billion Homo sapiens ever born, but the answer is likely most, and possibly the overwhelming majority.

Suggestive is the inquiry into the historical demography of Iceland, an almost perfectly closed population until recently, and one with accurate records dating back centuries. By the reckoning of the deCODE Genetics project, 92% of Icelandic women born since 1972 are descended from just 22% of the nation's women from the 1848-92 period, and "just 7% of the women born in the early eighteenth century period are the ancestresses of 62% of contemporary women." Take such trends back scores of millennia, and it is not difficult to see why genetic researchers are now hypothesizing that the entire population of the planet today might be traced back to as few as 25 unique original "out of Africa" women very long ago.

We may be inclined to regard defunct lineages as a "supply" problem, with harsh conditions precluding survival of offspring and continuation of families despite parents' best efforts — and this may indeed be most of the story. But it may also be the case that humanity has suffered from a "demand" problem over time as well. The widely noted epidemic of sex aversion among the rising generation of Japanese youth is only the most acute and most curious recent example of what may be a much more general human attribute — insufficient interest in procreation to sustain kith and kin. This feature of the human condition may go back a long way: In his study of Genesis, Leon Kass reminds us that the Lord in this sacred text repeatedly enjoins mankind to "be fruitful and multiply." If volitional fecundity were the natural order of things, why the command?

How will modern societies and the states that govern them cope with the prospect of long-term population decline in the centuries ahead, given the assumption of constantly increasing scientific and technological capabilities over this same period? Singer is of course correct that desired family size is the key determinant of human family size, so long as voluntary human procreation and childbearing are the determinant of human birth patterns.

But what if that most basic of human bonds is severed? What if, in some advanced future, a shrinking and dying society should decide it prefers to forestall decline and eventual departure not through "imports from abroad" but through "domestic manufacture"?

The temptation to resort to the non-maternal production and non-familial upbringing of human creatures could characterize a world where the total number of "legacy humans" was rising — indeed, demographically declining societies in such a world might regard their relative diminution as more reason to venture into so dark a realm. Opening this door, of course, would also inescapably invite "improvement" and "enhancement" of the pre-existing legacy models — stretching or redefining the very meaning of the term "human" and setting evolution on a forced march directed by our own conceit. A Columbus unit from now — in a world that has the technological wherewithal to house and feed a few hundred billion people — the horrific notion that some fraction of the planetary population might be "non-traditional models," so to speak, is worth consideration.

It is not hard to imagine such a planet enjoying high living standards, at least by the yardsticks we currently measure them with. It is, however, nearly impossible to imagine such a world being free — and it seems highly unlikely that it could be peaceful in the sense we now mean.

Population trends may therefore pose a greater threat to the future of modern societies than Singer's work would lead us to expect. It is incumbent on us the living to devote unflinching attention to this and other currently remote possibilities, to take the future seriously, and to recognize how much of a difference our choices can make — and how important it is to help human beings learn to choose well.

Nicholas Eberstadt holds the Henry Wendt Chair in Political Economy at the American Enterprise Institute. This essay was originally presented at a Hudson Institute conference celebrating the life of Max Singer, Hudson’s co-founder.


from the


A weekly newsletter with free essays from past issues of National Affairs and The Public Interest that shed light on the week's pressing issues.


to your National Affairs subscriber account.

Already a subscriber? Activate your account.


Unlimited access to intelligent essays on the nation’s affairs.

Subscribe to National Affairs.