The Future Has No Data: The Case of Population Forecasts

Why long-term demographic projections are ideological fictions that tell us more about their authors' hangups than they do about the future

Every few years, a new round of demographic projections makes headlines. Thus we have heard in recent years that Nigeria will surpass China in population by 2100. That Europe is being “replaced.” Or that falling birth rates are a “civilizational crisis.” Invariably phrased in apodictic and portentous ways, these pronouncements arrive draped in the authority of science — complete with confidence intervals, regression models, and the institutional prestige of the United Nations Population Division or the Pew Research Center. They’re treated as forecasts in the same sense that next week’s weather is a forecast: uncertain, perhaps, but grounded in data and rigorous method.

They are not. Long-term demographic projections — by which I mean projections extending beyond a generation or so — aren’t scientific forecasts. What they are, is ideological documents. And the sooner we learn to read them as such, the better equipped we will be to understand what sort of epistemic work they’re actually doing.

The Unborn Women Problem

The core of the matter, stated as plainly as possible: any demographic projection extending more than a few decades into the future must make assumptions about the fertility decisions of women who have yet to be born — and such assumptions are, as Betsy Hartmann argued already thirty years ago, always ideological projections.¹

This isn’t some technical limitation that better methodology may someday overcome. Anyone projecting population in 2100 is necessarily projecting the fertility choices of women who will be in their prime reproductive years in the 2060s, 2070s, and 2080s — women who are, today, entirely hypothetical, for the main and simple reason that these women don’t exist yet. They themselves have yet to be born. They have no fertility preferences to survey, no behavior to observe, no data to generate.

So what fills this void for the demographic modeler? The demographer must assume something. And that assumption cannot be derived from evidence, because there is no empirical evidence to derive it from. It must come from somewhere else: from a theory of how fertility decisions are made, from a model of how societies change over time, from a vision — hopeful or anxious — of what the future holds.

That isn’t empirics; that’s ideology.

What I learned from minoring in demography in college (which is when I first became fascinated by the deep weirdness of demographic projections) is that this ideological aspect of demographic projections isn’t some minor caveat appropriately buried in a methodology appendix. It is in fact the load-bearing wall of the entire long-range forecasting exercise. Short-term projections (five, ten, maybe twenty years) can credibly extrapolate from current trends, current cohorts, and observed behavior. But the further out the demographer projects, the larger the share of their forecast will rest not on what actually existing women are doing or intend to do but on what the modeler expects — or fears, or hopes — that those women’s as-yet-unborn granddaughters will do. By the time demographers are projecting a century out, they’re no longer at all in the realm of science, but instead are writing a “just-so story” about the future.

Which is to say: long term demographic projections are also “projections” in the psychological sense, that is, unconscious defense mechanisms through which a person attributes their own unacceptable thoughts, feelings, or traits onto someone else. Instead of confronting internal discomfort — such as jealousy, insecurity, or fear — the individual deflects it outward.

This is what a multi-generational demographic projection really is. The assumptions demographers make about future fertility are, without exception, expressions of beliefs about gender, culture, development, and social organization. Those beliefs have histories, constituencies, and political valences. In the third section of this piece, I will argue that they are almost always also racialized. But first, let us look at the track record.

A History of Being Wrong, Spectacularly and Repeatedly

The social-scientific study of population was born in anxiety. France in the nineteenth century had experienced a puzzling demographic divergence from its European neighbors: French fertility had declined sharply, for reasons that historians still debate (explanations range from the influence of Revolutionary-era inheritance law to distinctive patterns of Catholic practice to early urbanization). The result was a France that was falling behind Britain and Germany in population — and therefore, French elites feared, in military and economic power.²

This elite anxiety funded the emergence of demography as a formal discipline. The question driving the research was urgent and political: why weren’t French women having enough babies, and what could the French government do about it? The field was born, in other words, in the service of a particular political project.³

The founding figures of the field illustrate the point. The word démographie itself was coined by a Frenchman, Achille Guillard, in his 1855 book Éléments de statistique humaine, ou démographie comparée — the founding text of the discipline as a named science. Guillard defined demography as “knowledge, derived from observation, of the laws by which populations form, sustain themselves, renew themselves, and succeed one another.” His son-in-law, Louis-Adolphe Bertillon, institutionalized it: he held the first professorship in demography at the School of Anthropology in Paris and taught the first demography course at the Paris Medical School in 1875. The discipline was, from the outset, a French family enterprise.⁴

The third generation shows where that enterprise was always heading. Jacques Bertillon — Guillard’s grandson, Louis-Adolphe’s son — was by the turn of the century France’s most prominent demographer and its most tireless pronatalist agitator. He argued that France’s low birth rate relative to Germany’s was sapping the nation’s economic and military strength, warned explicitly of “race suicide,” and in 1896 founded the Alliance nationale pour l’accroissement de la population française to lobby for state intervention in French reproductive life. His recommended remedies included tax incentives for large families and legal bans on contraception and abortion — positions that directly shaped the repressive 1920 law restricting both. By 1900, more than 130 French senators, mobilized by Bertillon’s agitation, were demanding an official parliamentary commission on depopulation; by 1902, the Interior Ministry had convened one.⁵

What those nineteenth century French demographers didn’t foresee — in fact couldn’t foresee — was that France, and then nearly every other industrialized country, would soon fall below replacement-level fertility (generally deemed to be 2.1 lifetime births per woman). That threshold was crossed during the First World War, when the catastrophic death toll was accompanied by a collapse in births. Fair enough: wartime is unusual. But then fertility stayed low throughout the interwar period. This too was unexpected. The models and theories of the time had no framework for sustained sub-replacement fertility in peacetime. Indeed, demographic anxieties were a defining feature of interwar North Atlantic politics, and one deeply entwined with the catastrophe of Nazism.⁶

Then, after the Second World War, come something no one predicted at all: the Baby Boom. In 1947 the U.S. Census bureau sponsored a report, written by the country’s leading demographers, entitled, “Forecasts of the Population of the United States 1945-75,” which contained not a hint of the fertility boom which by then was already underway: across the United States, Canada, Australia, and much of Western Europe, fertility surged. Families that the Depression and the war had postponed were suddenly being formed all at once, and then kept forming. As this became increasingly apparent, the demographic establishment scrambled to update their models — often gaining new grants to try to understand why their models built with their old grants had failed so conspicuously.⁷

It updated them wrong. In the middle of the Baby Boom, with fertility rates elevated and every projection pointing upward, demographers and public intellectuals became seized by alarm about overpopulation. Most infamously, Paul Ehrlich’s The Population Bomb, published in 1968, argued that mass starvation was imminent and unavoidable. Hundreds of millions would die. The catastrophe was deemed unavoidable, the only question was how bad it would be. The projections became the basis for panicked advocacy in favor of coercive population control policies.⁸

Ironically, the baby bust had already begun when Ehrlich was writing, though he didn’t see it. Fertility in the United States peaked around 1957 and by 1968 had been declining for a decade. Within a generation, the worry had inverted entirely: not too many babies but too few.

The theoretical apparatus that was supposed to make sense of all this was the “demographic transition” model — the idea, systematized in academic demography through the mid-twentieth century, that societies move through predictable stages from high fertility and high mortality to low mortality and eventually low fertility. The model’s appealing implication was convergence: all societies would eventually stabilize at around replacement-level fertility (roughly 2.1 children per woman), and population would become “stationary” — the latter itself indicative of the ideological belief that “stability” would be the final outcome of the process of modernization.

The historian Simon Szreter, in his foundational 1993 article in Population and Development Review, subjected this theory to a devastating critique.⁹ The demographic transition model was not an empirical generalization from the historical record but a theoretical imposition upon it — one that conveniently aligned with mid-century modernization theory and its assumption that all societies were converging on a Western liberal endpoint. As Szreter showed, the actual historical data on European fertility transitions showed enormous variation in timing, pace, and pattern that the tidy model couldn’t explain. (As an aside: this was the article that more than any other inspired me to write my dissertation on the intellectual history of modernization theory, which became my first book, Mandarins of the Future.)

And then reality administered the final refutation: nowhere in the world fertility has fertility in fact “converged on replacement.” Over the last couple of decades, it has blown through it over huge parts of the world. Today, South Korea has a total fertility rate of approximately 0.75 (something that would have been unimaginable to anyone observing Korea in the 1940s or 1950s, when the average woman would have 5-6 children and half the population was under the age of 18). Likewise, Italy, Spain, Japan, and dozens of other countries are well below replacement. Meanwhile, sub-Saharan Africa retains relatively high fertility rates, though these have also been declining fast enough to embarrass the most-cited long-range projections.

Which brings us to today’s spectacular projections. Twenty-five years ago, the UN’s median projection had Nigeria’s population reaching 800m by 2100, which was expected to be higher than China’s. This number is cited constantly in discussions of “Africa’s demographic dividend” and in anxious racialized Western commentary about migration, stability, and resource competition. It is treated as a fact about the future.

Since then, there have been repeated revisions to these projections. Today, the UN’s median projection forecasts Nigeria’s population will be 477m by 2100 — nearly a third less than they were anticipating a generation ago. In each case, these forecasts are not rooted in empirical data, but rather in assumptions about how Nigerian women who are currently in primary school will decide, over the next seven decades, to organize their reproductive lives. These aren’t facts, they’re stories.

What the Stories Are Actually About

So why do people keep telling these stories, with such apparent confidence, and why do they receive such credulous attention?

Because demographic projections aren’t really about the future. They’re about the present — about present fears and present aspirations, projected forward and dressed in the language of science to give them authority they could not otherwise claim.¹⁰

The anxieties that drive demographic alarmism have been, with great consistency across the history of the field, anxieties about race and gender. Not always explicitly. But follow the arguments about these projections far enough, and you always find them there.

Consider the toxic discourse around “The Great Replacement” — the claim, now prevalent in far-right politics across Europe and North America, that immigration and differential birth rates are systematically replacing white populations. This is the garish end of conspiracy-theory demography, sometimes veering into Protocols-of-the-Elders-of-Zion level nonsense about how various malignant forces are deliberating engineering a slow-motion genocide of Caucasians. Even when not explicitly racist, the same underlying anxiety — that “our” population is shrinking relative to “theirs” — structures enormous amounts of mainstream commentary about birth rates, immigration, and national identity. When a European politician warns about demographic decline, and when a radioactive Tucker Carlson monologue warns about replacement, they’re drawing on the same reservoir of racialized dread. The only difference is in the explicitness.¹¹

The racial coding is more than incidental. Demography as a discipline emerged, as we have seen, from elite anxiety about national power — which was, in the nineteenth and early twentieth centuries, inseparable from racial ideology. The eugenics movement drew heavily on demographic data and demographic anxieties.¹² As Matt Connelly has shown, the population control movements of the mid-twentieth century were directed overwhelmingly at the bodies of non-white women in the Global South, not at the declining fertility of Europeans. The contemporary “demographic dividend” framing, which presents African population growth simultaneously as economic opportunity and as threat, inherits this history.

And underneath every demographic anxiety about birth rates, there is a gender politics that almost never gets stated directly but that structures everything. Falling fertility rates are, definitionally, the result of women choosing — both individually and in aggregate — to have fewer children. When states or pundits or demographers express alarm about this, they are expressing the view that women are making “wrong” choices. The entire apparatus of pro-natalist policy — tax incentives, parental leave top-ups, rhetorical appeals to motherhood — is premised on the idea that women’s fertility decisions should be steered toward outcomes that (usually male) political actors have decided are in the national interest.

“Demographic crisis” language almost always implies a subject who is worried and an object who needs to change her behavior. The worried subject is coded male, national, and rational. The object — the woman who is not having enough babies, or having too many, or having them with the wrong partners — is the problem to be solved. Demography’s long history of treating women as variables in population equations rather than as agents making choices for reasons of their own is hardly a bug in the discipline. It is, arguably, its founding premise.

None of this means that fertility rates are uninteresting or that demographic change has no consequences. Of course it does — for urban planning, for health care and education investments, for labor and consumer markets, for pension systems, for the structure of electorates, and yes, even for geopolitical power. Short- and medium-range demographic analysis, grounded in current data about living people, can be genuinely useful for designing sustainable policies.

But the next time someone presents you with a confident projection about what the world’s population will look like more than about 25 years from now, I invite you to ask a simple question: what assumptions are they making about the fertility choices of women who are not yet born? And then ask a tougher one: where did those assumptions come from?

The answer to the second question will tell you more about the projection’s author than about the demographic future. It will reveal what they fear, what they value, and whose reproductive choices they think need correcting.

1. Betsy Hartmann, “Population control I: Birth of an ideology” and “Population control II: The population establishment today,” International journal of health services 27.3 (1997).

2. There’s a HUGE literature on this. See for example Richard Tomlinson, “Social Policy in Europe: The French Population Debate,” The Public Interest 76 (1984); Carol Blum, Strength in numbers: Population, reproduction, and power in eighteenth-century France (JHU Press, 2002); Henri Leridon, “The development of fertility theories: A multidisciplinary endeavour,” Population 70.2 (2015); Sean M. Quinlan, The great nation in decline: sex, modernity and health crises in revolutionary France c. 1750–1850 (Routledge, 2016).

3. Paul-André Rosental, “The novelty of an old genre: Louis Henry and the founding of historical demography,” Population 58.1 (2003): 97-130.

4. Michel Dupâquier, “La famille Bertillon et la naissance d’une nouvelle science sociale: la démographie,” in Annales de démographie historique (Société de démographie historique, 1983.)

5. Virginie De Luca Barrusse, “Des liaisons avantageuses: l’Alliance nationale pour l’accroissement de la population française et les fonctionnaires (1890-1914),” Annales de démographie historique 116:2 (Belin, 2008).

6. See the special issue of Continuity and Change 5.2 (1990), on the legal and social political responses to fertility decline in Europe during the first half of the twentieth century.

7. Jan Van Bavel and David S. Reher, “The baby boom and its causes: What we know and what we need to know,” Population and development review 39.2 (2013); Dan Bouk, “Generation crisis: How population research defined the Baby Boomers,” Modern American History 1.3 (2018).

8. Matthew Connelly, Fatal misconception: The struggle to control world population (Harvard University Press, 2010).

9. Simon Szreter, “The idea of demographic transition and the study of fertility change: a critical intellectual history,” Population and development review (1993).

10. J. M. Winter, “The Fear of Population Decline in Western Europe, 1870–1940,” in R. W. Hiorns, ed., Demographic patterns in developed societies (Routledge, 2023 [1980]); Michael S. Teitelbaum, The fear of population decline (Academic Press, 2013).

11. Mark Sedgwick, “The great replacement narrative: fear, anxiety and loathing across the West,” Politics, Religion & Ideology 25.4 (2024).

12. Frank Dikötter, “Race culture: Recent perspectives on the history of eugenics,” The American Historical Review 103.2 (1998).