Demography, Migration, and Urban Futures in 2035: Scenarios, Signals, and Falsifiable Predictions

Fertility keeps falling almost everywhere, aging keeps compounding, and climate mobility keeps concentrating in a handful of places — here is what would have to be true, and what would prove it wrong, by 2035.

National population registers are the raw material behind every demographic scenario — and behind every one of them being wrong at the margins. — Image prompt and art direction by Brecht Corbeel; image generated to that direction.

Abstract

Three trends are treated separately in most demographic commentary and are in fact one system: falling fertility, accelerating population aging, and migration reshaped by both labor demand and climate hazard. This article documents the current trajectory of each using UN, OECD, Eurostat, and national statistics data, then constructs three explicit 2035 scenarios — a fertility reversal, a climate-displacement surge in a named region, and an aging-driven policy shift in a named country — each with a stated horizon, assumptions, observable indicators, and a disconfirmation condition that would prove it wrong. It separates verified fact from vendor-style projection, analytic inference, and outright prediction throughout.

Every demographic forecast for 2035 is built from three moving parts that are usually discussed separately and should not be: how many children are born, how long people live once born, and where people go when a place stops being livable or stops offering work. Fertility, aging, and migration are one coupled system, and the countries running out of runway on any one of them are almost always running short on the other two at the same time.

This article does three things. First, it lays out the documented, currently-verifiable trajectory of each part of the system using United Nations, OECD, Eurostat, WHO, and national statistics data. Second, it separates that documented trajectory from the analytic inferences drawn from it and from the vendor- and government-style claims made about interventions meant to change it. Third, it converts the most contested question in each domain — will fertility measurably recover anywhere by 2035, will climate displacement produce a documented surge in a specific region, will aging force a specific government’s hand — into an explicit, falsifiable scenario: a stated horizon, the assumptions the scenario depends on, the indicators that would confirm it is unfolding, and the condition that would prove it wrong.

Fact: the fertility decline is close to universal, and it is not new

The global total fertility rate has fallen from roughly 4.8 births per woman in 1970 to about 2.2 in 2024, according to the UN Population Division’s 2024 revision of World Population Prospects, which incorporates data from 1,910 national censuses conducted between 1950 and 2023 alongside vital registration and survey data [1]. More than half of all countries and areas now sit below the 2.1 replacement threshold — the level at which, absent net migration, a population’s size holds constant over the long run [2]. This is not a recent shock. It is the continuation of a decline that has been running for over half a century and has, at the country level, rarely reversed for more than a few years at a time.

The regional detail matters more than the global average. The European Union’s total fertility rate fell to 1.34 births per woman in 2024, described by Eurostat as a new low for the bloc [3]. Across the OECD, fertility sits below the 2.1 replacement level in every member country except Israel, and the OECD’s own tracking shows the modest 2000s rebound in many high-income countries has given way to a resumed downward trend since [4]. South Korea remains the clearest extreme case: its rate fell to a record low of 0.72 in 2023 before edging up to roughly 0.74–0.75 in 2024, still far below replacement and still the lowest figure among OECD members, even as marriage numbers rebounded from pandemic-era delays [7].

That last point deserves emphasis because it is the one place in this data where a trend line bent rather than simply continuing. South Korea’s 2024 uptick is a fact, not a projection — births rose year over year for the first time in nearly a decade. Whether it is the start of a genuine reversal or a one-year artifact of deferred marriages catching up is exactly the kind of question a falsifiable scenario, not a headline, should be built to answer, and one follows below.

Fact: population aging is compounding, and the shift in where the old live is as important as the shift in how many there are

The World Health Organization states that the share of the world’s population over 60 will nearly double between 2015 and 2050, from 12 percent to 22 percent, with the 60-plus cohort growing from 1 billion in 2020 to a projected 1.4 billion by 2030 and 2.1 billion by 2050; the 80-plus cohort is expected to roughly triple over the same span, reaching 426 million [5]. In 2020, people aged 60 and over outnumbered children under 5 for the first time in recorded history — a genuine demographic crossover, not an extrapolation [5].

The geographic distribution of this aging is the fact most often left out of the headline number. High-income countries aged first — Japan already has roughly 30 percent of its population over 60 — but the WHO projects that by 2050 two-thirds of the global 60-plus population will live in low- and middle-income countries, which are aging on a compressed timetable without the income levels high-income countries had when they crossed the same threshold [5]. Japan’s own Statistics Bureau publishes monthly population estimates that track this directly through its national population register [10], and Japan is consequently the clearest live test case for how a government responds once the fiscal arithmetic of aging becomes unavoidable, discussed as a scenario below.

Fact: internal displacement, driven overwhelmingly by disasters, reached a new global high in 2024

The Internal Displacement Monitoring Centre’s 2025 Global Report on Internal Displacement recorded 45.8 million internal displacements in 2024 — the highest total since IDMC records began in 2008, and more than double the annual average of the preceding decade [6]. Disasters, not conflict, drove the majority of these movements: 99.5 percent of the year’s disaster-linked displacements were tied to climate-related extreme weather. Twenty-nine countries and territories recorded their highest disaster-displacement figures on record in a single year, and cyclones alone accounted for 54 percent of all disaster displacements [6].

The regional concentration is the load-bearing fact for the scenario below. The Americas recorded 14.5 million disaster-related internal displacements in 2024 — more than the previous five years combined — with the United States alone accounting for roughly 11 million movements, nearly a quarter of the global total. South Asia’s disaster displacement nearly tripled year over year to 9.2 million, following drier El Niño-linked conditions in 2023 [6]. Both of these are documented 2024 outcomes, not projections, and both are the kind of regional concentration a 2035 scenario needs a named place to test against.

Analysis: why these three trends function as one system

None of fertility, aging, or migration is intelligible as a standalone number for city planning, pension solvency, or housing demand. A country with low fertility and rising longevity faces a shrinking working-age base supporting a growing retired one — the core mechanism behind every pension-reform debate in an aging high-income country. That shrinking working-age base is frequently offset, where it is offset at all, by immigration: workers arriving to do jobs a smaller domestic cohort cannot fill. Japan absorbed over 340,000 foreign workers in 2024, roughly twice the pace the government and the UN had forecast, even though a broad political shift toward treating large-scale immigration as ordinary and permanent remains, by most current accounts, unlikely given recent nationalist electoral gains [8]. That is an analytic judgment about political trajectory, not a documented fact, and it is stated as such here.

Figure 1. Falling fertility reshapes housing demand a generation before it shows up in a census — planners model it in miniature first. — Image prompt and art direction by Brecht Corbeel; image generated to that direction.

Housing and urban form follow the same logic on a longer lag. A generation of below-replacement births shows up first in shrinking primary-school cohorts, then in a smaller pool of new household formation two decades later, and only then in aggregate housing-demand figures that planners and investors treat as leading indicators. Migration — both the labor-driven kind and the disaster-driven kind documented above — can mask, delay, or regionally concentrate that effect: a city losing native-born young adults to lower fertility upstream can still see rising housing pressure if it is simultaneously a destination for displaced or economic migrants. Any single-cause account of housing pressure in an aging, low-fertility, migration-receiving city is very likely incomplete for this reason, which is an analytic conclusion drawn from the facts above rather than a fact in its own right.

Figure 2. Vital registration systems, not surveys, are what let a country claim its fertility numbers are exact rather than estimated. — Image prompt and art direction by Brecht Corbeel; image generated to that direction.

Vendor and government claims, marked as such

South Korea has committed roughly 270 billion United States dollars in cumulative spending on pro-natal incentives over roughly two decades, according to widely repeated reporting — a government claim about input, not a demonstrated claim about outcome, since the fertility rate it was meant to raise continued falling for most of that period before its small 2024 uptick. Framing this spending figure as evidence that “incentives work” or “incentives fail” would overstate what a single aggregate spending number, uncorrelated in this article to a rigorous causal study, can support; it is presented here only as a government input claim, not a validated outcome.

Similarly, Japan’s April 2025 mandate requiring employers to secure work opportunities for staff up to age 65 is a documented government policy claim — East Asia Forum’s account of the measure is treated here as reporting on an enacted legal requirement, not as a demonstrated solution to Japan’s labor-supply shortfall [8]. Whether it measurably changes the employment rate of Japanese workers in their early sixties, versus simply formalizing practices many employers already followed, is an open empirical question that current reporting does not settle.

Figure 3. Climate-driven internal displacement is tracked one household case file at a time, long before it becomes a regional statistic. — Image prompt and art direction by Brecht Corbeel; image generated to that direction.

Scenario one: does any major economy’s total fertility rate measurably reverse its documented decline by 2035?

Horizon. Calendar year 2035, using each country’s most recent full-year vital-registration total fertility rate as reported by its national statistics agency or by the UN Population Division.

Current trajectory (fact). Every OECD member except Israel sits below the 2.1 replacement threshold [4]; the EU average is 1.34 [3]; South Korea sits near 0.74–0.75 after a single-year uptick from a 2023 low of 0.72 [7]. No high-income country has sustained a multi-year fertility recovery back toward replacement level in the past three decades.

Assumptions. No major war, pandemic, or currency crisis materially disrupts fertility behavior in the countries tracked; national statistics agencies continue publishing comparable annual total-fertility figures through 2035; “major economy” here means any G20 member.

Observable indicators that would support a reversal scenario. A G20 country’s total fertility rate rising for three consecutive reported years, not one; the rise exceeding half a birth per woman cumulatively rather than a fractional rebound; the rise accompanied by rising marriage and first-birth rates among women under 35, ruling out a pure “deferred births catching up” effect.

Disconfirmation condition. If, by year-end 2035, every G20 country’s total fertility rate is at or below its 2024 level, or if any apparent multi-year rise is fully explained by a one-time catch-up of previously postponed births rather than a change in completed cohort fertility, this scenario is falsified. South Korea’s 2024 uptick alone does not confirm the scenario; it is the single most-watched candidate data point, and three more consecutive years of Korean data are the minimum needed either way.

Scenario two: does climate-driven internal displacement produce a documented, measurable surge in a specific vulnerable region by 2035?

Horizon. Calendar year 2035, measured against IDMC’s Global Report on Internal Displacement methodology, which already documents 2024 as a record year.

Current trajectory (fact). 45.8 million internal displacements were recorded in 2024, more than double the prior decade’s annual average, with 99.5 percent of disaster-linked displacement tied to climate-related extreme weather [6]. The Americas — led by the United States at roughly 11 million movements — and South Asia, where displacement nearly tripled year over year to 9.2 million, are the two regions with the sharpest documented 2024 concentration [6].

Assumptions. IDMC or a comparably rigorous successor continues publishing consistent annual methodology through 2035; no major methodological break redefines what counts as a displacement event; South Asia is named here as the named vulnerable region because its 2024 tripling, unlike the Americas’ single-country-driven total, reflects a broader regional pattern across multiple countries.

Observable indicators that would support a surge scenario. South Asia’s annual disaster displacement figure exceeding its 2024 level of 9.2 million in at least two of the three years 2033–2035; the excess concentrated in slow-onset drivers (drought, sea-level-linked coastal flooding) rather than single acute cyclone events, which would indicate a structural climate mobility shift rather than one bad storm season.

Disconfirmation condition. If South Asia’s annual disaster-displacement figure falls back toward its pre-2023 baseline (below roughly 5 million per year) for three consecutive years before 2035, or if any elevated later figure is attributable to a single acute event with no recurrence, this scenario is falsified.

Figure 4. Population aging shows up first as a fiscal line item and second as a fitting-room adjustment — both are visible on the same counter. — Image prompt and art direction by Brecht Corbeel; image generated to that direction.

Scenario three: does population aging’s fiscal burden trigger a documented policy response in a named country by 2035?

Horizon. Calendar year 2035, tracked against Japan specifically, chosen because it is the highest-income economy furthest along the aging curve with a national population register providing continuous monthly data [10].

Current trajectory (fact). Japan already has roughly 30 percent of its population over 60 [5]; its 2025 mandatory-employment-to-65 rule took effect for all employers, including small and medium enterprises, in April 2025 [8]; foreign worker inflows reached over 340,000 in 2024, roughly double prior official forecasts [8]; U.S. Social Security Administration tracking of international pension-policy changes documents Japan’s incremental 2024 expansion of Employees’ Pension Insurance enrollment to smaller employers’ part-time staff as part of the same fiscal response [9].

Assumptions. Japan’s political system continues to prioritize incremental labor-force and pension-eligibility adjustments over either a large-scale immigration liberalization or a retirement-age cut; no major economic shock forces an off-cycle emergency pension reform that would make this timeline moot.

Observable indicators that would support a “policy response confirmed” outcome. A further statutory increase in the pensionable age above 65, or a further expansion of mandatory-employment coverage beyond age 65, enacted and in force before 2035; foreign worker inflows sustaining or exceeding the 2024 pace across at least four of the ten years 2025–2035; publication, by Japan’s Ministry of Health, Labour and Welfare or the Statistics Bureau, of data showing the employed share of the 65-plus population rising above its already-elevated recent levels.

Disconfirmation condition. If Japan reaches 2035 with its statutory pensionable age unchanged from 2025, mandatory-employment coverage not extended past 65, and annual foreign worker inflows below the 2024 figure in at least six of the ten intervening years, this scenario is falsified — it would indicate the fiscal pressure documented above was absorbed through some other channel (automation, private savings drawdown, or simply reduced public services) rather than through the labor-supply and pension-eligibility policy changes this scenario specifies.

Figure 5. Every 2035 scenario in this article ultimately traces back to a cohort-component model run like this one — an estimate, not a fact, until the year arrives. — Image prompt and art direction by Brecht Corbeel; image generated to that direction.

Analysis: education, health systems, and technological adaptation as downstream effects

Two further downstream effects follow from the facts above, and both are analytic inferences rather than independently sourced facts in their own right. First, education systems in low-fertility countries face a structural sequencing problem: primary-school enrollment falls first as smaller cohorts move through, which frees teaching capacity that is then, in many observed cases, redirected toward the growing demand for adult retraining and eldercare-adjacent vocational programs rather than closed outright — a reallocation, not simply a contraction. Japan’s mandatory extension of employment obligations to age 65 only makes sense against this backdrop: a labor market that needs older workers to stay productive needs some institutional capacity to retrain them, and school-age enrollment decline is one of the few places that capacity can plausibly come from without new public spending [8].

Second, health systems in aging populations face a cost curve that is close to mechanically certain in a way fertility trends are not, because the people who will be over 60 in 2035 are, with minor exceptions for migration, already alive today and already counted. The WHO’s projection that the 80-plus cohort will roughly triple between 2020 and 2050 is not a behavioral forecast subject to policy change the way a fertility rate is — it is closer to an accounting identity applied to a cohort whose size is already fixed [5]. This is the strongest reason to treat population aging as the most confidently projectable of the three trends in this article and fertility as the least: aging in the 2035 window is driven by people already born, while fertility in the 2035 window depends on choices not yet made by people not yet parents.

Technological adaptation — automation substituting for a shrinking working-age population, remote health monitoring substituting for in-person geriatric care capacity, algorithmic matching substituting for manual caseworker triage in resettlement processing — is frequently proposed as the variable that resolves the tension between aging, low fertility, and constrained migration without requiring politically difficult policy change on any of the three. That proposition is presented here as an analytic hypothesis under active discussion, not a documented outcome: current reporting on Japan’s labor market describes automation as a partial offset under active policy discussion, not as a demonstrated substitute for the labor-force gap documented above [8]. Whether it closes enough of the gap to meaningfully change the fiscal-burden scenario in Scenario Three by 2035 is itself one of the open questions that scenario’s indicators are designed to surface, not an assumption built into it.

Cities sit at the intersection of all three downstream effects because housing stock, school capacity, and health infrastructure are each sized against population projections made ten to twenty years earlier and are each expensive and slow to resize. A city that overbuilt housing against a fertility assumption that later fell short carries excess capacity for a generation; a city that underbuilt eldercare and health infrastructure against an aging population that arrived faster than planned faces the opposite problem on a shorter timeline, because the people driving that demand are, again, already alive and already counted in the register. This asymmetry — housing and school demand are forecastable but contingent on future births, health and eldercare demand for 2035 are close to already fixed — is the single most useful planning distinction this article’s facts support, and it is offered here as an analytic conclusion rather than as an additional fact.

What would change this analysis

Two kinds of new evidence would most directly revise the picture above. First, a fourth or fifth consecutive year of rising South Korean or Japanese births, decoupled from a marriage-catch-up explanation, would be the strongest available evidence that a high-income fertility reversal is structurally possible rather than a one-year artifact — nothing in the current record supports that yet. Second, a 2026 or 2027 IDMC report showing South Asian disaster displacement falling back toward its pre-2023 baseline would suggest 2024’s tripling was event-driven rather than trend-driven, which would meaningfully lower confidence in treating South Asia as a persistently vulnerable region rather than one that had an unusually bad two years. Both are testable against publicly reported data well before 2035, and both are exactly the kind of falsifiable check this article’s scenarios were built to invite rather than avoid.

The honest summary is not a forecast so much as a set of open bets with stated terms: fertility decline is the best-documented and least-reversed of the three trends, aging is the most mechanically certain because it is driven by cohorts already born, and climate-driven displacement is the most volatile and the one where a single additional bad year could either confirm or undermine the regional pattern documented in 2024. None of the three should be treated as settled before the indicators above are actually checked.

Sources

1. UN Department of Economic and Social Affairs, Population Division. World Population Prospects 2024: Summary of Results. United Nations (2024).
2. UN Department of Economic and Social Affairs, Population Division. World Fertility 2024. United Nations (2025).
3. Eurostat. Fertility statistics. European Commission (2025).
4. OECD. Fertility rates. OECD (2025).
5. World Health Organization. Ageing and health. World Health Organization (2025).
6. Internal Displacement Monitoring Centre. 2025 Global Report on Internal Displacement (GRID). IDMC / Norwegian Refugee Council (2025).
7. The Korea Herald staff. Childbirths in Korea rebound for 1st time in nearly a decade in 2024. The Korea Herald (2025).
8. East Asia Forum contributors. Japan's senior employment challenge. East Asia Forum (2025).
9. U.S. Social Security Administration, Office of Retirement and Disability Policy. International Update, January 2025. U.S. Social Security Administration (2025).
10. Statistics Bureau, Ministry of Internal Affairs and Communications. Population Estimates. Government of Japan (2025).

Originally published at https://absolutedigitalpublishers.com/articles/demography-migration-and-urban-futures-in-2035-scenarios-signals-and-falsifiable-predictions.