CN-G-02: China's Demographic Crisis — From the One-Child Policy to Population Decline (1979–2026)

Status: DRAFTWords: 9,377

1. Key Takeaways

The one-child policy was not the beginning of China's fertility decline but an intensification of a decline already well advanced by 1979. The total fertility rate (TFR) had already fallen from roughly six children per woman in the late 1960s to under three by the late 1970s — a collapse driven by the wan xi shao ("later, longer, fewer") campaign of 1973 onward, rising female education and labour-force participation, and urbanisation. [TBD-VERIFY: the TFR is widely cited as falling from ~5.8 in 1970 to ~2.7–2.9 by 1978–1979; precise annual figures vary by source and reconstruction method.] This timing is the empirical foundation of the central revisionist argument (Whyte, Wang Feng, Cai Yong; Greenhalgh) that the most coercive birth limits were imposed after the steepest decline had occurred, and therefore did less demographic work — and inflicted more avoidable harm — than the official "averted 400 million births" narrative claims.
The decision for one child rather than a two-child or later-marriage design was a contingent product of a particular kind of expertise, not an inevitability. Susan Greenhalgh's Just One Child (2008) documents that the one-child target emerged from cybernetic-control and missile-trajectory modelling by defence scientists — above all Song Jian — who applied Club-of-Rome "limits to growth" logic and systems-engineering optimisation to population as a controllable variable. Social scientists who favoured a softer two-child approach were marginalised. The policy was thus, in Greenhalgh's reading, "made by scientists" of the wrong discipline, and its severity reflected the authority of natural-science expertise in early-reform China rather than any demographic necessity for the one-child threshold specifically.
Enforcement was decentralised, quota-driven, and ranged from administrative permits and fines to documented coercion including forced sterilisations and late-term abortions. The system ran on birth permits, cadre birth quotas tied to local-official performance evaluation, and the "social maintenance fee" (社会抚养费) — substantial fines for out-of-quota births. At its harshest, particularly during campaign peaks in the early 1980s and again in periodic "strike-hard" birth-planning drives, enforcement included compulsory IUD insertion after a first birth, mandatory sterilisation after a second, and coerced abortions. [TBD-VERIFY: aggregate national figures for sterilisations and abortions attributable specifically to coercion are not reliably established; official statistics record very large numbers of sterilisations and IUD insertions — e.g., tens of millions in peak years such as 1983 — but the share that was coerced versus accepted cannot be cleanly disaggregated from the record.]
The policy produced one of the most skewed sex ratios at birth ever recorded in a large population — the "missing girls." Sex-selective abortion (after ultrasound spread in the late 1980s), under-reporting of female births, and in earlier decades female infanticide and abandonment pushed the sex ratio at birth to a peak of roughly 118–121 boys per 100 girls in the early-to-mid 2000s, against a biological norm of ~105. [TBD-VERIFY: the peak national sex ratio at birth is variously given as ~118 (2005 1% survey) to ~121 (2008–2010); figures differ by source and year.] The cumulative result is a "surplus" of tens of millions of men of marriageable age — a structural feature with implications for marriage markets, migration, trafficking, and social stability that will persist for decades regardless of current policy.
The "4-2-1 family" — one child supporting two parents and four grandparents — concentrated eldercare and old-age-security obligations onto a shrinking base. With the one-child cohort now in middle age and their parents entering deep old age, the inverted family pyramid has become a lived fiscal and care reality. It also compounds the fragility exposed by events such as the 2008 Sichuan earthquake, in which parents who lost their only child (the shidu 失独 families) had no other children — a category Mei Fong (One Child, 2016) uses to humanise the policy's tail risk.
China's population began an outright decline in 2022 — the first sustained fall since the Great Leap Forward famine — and India overtook China as the world's most populous country in 2023. The NBS reported (17 January 2023) that the population fell in 2022 for the first time since 1961, and declines continued in 2023 and 2024. [TBD-VERIFY: the NBS reported the end-2022 population at ~1.41175 billion, down ~850,000 from 2021; end-2023 down ~2.08 million; end-2024 figures should be confirmed against the NBS release. Exact decrement figures vary slightly across releases and revisions.] The UN's World Population Prospects projected India's overtaking for 2023. These were not surprises to demographers — Yi Fuxian and others had argued the turning point was arriving earlier than official projections assumed — but they crystallised the shift from a population-control to a population-decline problem.
The total fertility rate has collapsed to among the lowest in the world — roughly 1.0–1.1 — far below the ~2.1 replacement level and below even the official "low-fertility trap" thresholds. [TBD-VERIFY: China's TFR is variously estimated at ~1.0–1.1 for 2022–2023; some estimates put 2023 near 1.0 or below, others slightly higher; there is no consensus single figure and the NBS does not publish a definitive annual TFR.] At such a rate, each generation is roughly half the size of its predecessor — a trajectory that no pronatalist policy has yet reversed in any comparable East Asian society (South Korea, Taiwan, Singapore, Japan), which is the empirical heart of the reversibility debate.
The policy reversal was incremental and arguably too slow: selective two-child (2013), universal two-child (2016), three-child (2021), and then a turn to active pronatalism. Each loosening produced a smaller-than-projected birth response, suggesting that by the 2010s the binding constraint had shifted from policy permission to the cost of raising children — housing, education, and the neijuan ("involution") of hyper-competitive child-rearing — and to changing marriage and gender norms. The 2021 three-child decision was explicitly paired with measures to lower the cost of childbearing; by 2025 the centre had moved to direct childcare subsidies. [TBD-VERIFY: the specific design and amounts of the 2025 national childcare-subsidy measures should be confirmed against the relevant State Council documents.]
The state is simultaneously managing a shrinking workforce, a rising old-age dependency ratio, and a strained pension system — and in 2024 took the politically sensitive step of raising the statutory retirement age for the first time in decades. The NPC Standing Committee decision of 13 September 2024 (effective 1 January 2025) raises retirement ages gradually over fifteen years — for men from 60 to 63, and for women from 50/55 to 55/58 depending on job category. [TBD-VERIFY: the precise end-state ages and the phase-in schedule should be confirmed against the 13 September 2024 decision text.] This is paired with an eldercare-services build-out under the "silver economy" rubric.
The animating thesis of the economic-implications literature is that China risks "getting old before getting rich." Nicholas Eberstadt and others argue that, unlike Japan, South Korea, or the West, China's society is ageing at a per-capita income level far below that at which those societies aged — meaning the fiscal and care burden of a super-aged society falls on an economy that has not completed the climb to high-income status, with the demographic dividend (the favourable worker-to-dependent ratio that powered the growth miracle) now reversed into a demographic drag.
The corpus treats three questions as genuinely contested rather than settled: whether the one-child policy was necessary, what causes today's ultra-low fertility, and whether the state can reverse it. On necessity, the official "growth-enabling, hundreds-of-millions-of-births-averted" account contends with the revisionist "the decline was already underway and the coercion was avoidable" reading and the "modernisation would have lowered fertility regardless" position. On causation, policy legacy contends with cost-of-living/involution and with changing gender/marriage norms. On reversibility, the pronatalist-optimist position contends with the comparative-East-Asian pessimism that no society at China's fertility level has recovered. These are the document's three three-account axes.

2. The Record in Brief — A Policy and Its Reversal

China's population governance over the half-century from 1979 to 2026 describes an almost perfect arc of intention reversed. The state that in 1979–1980 mobilised one of the most ambitious and coercive birth-restriction campaigns in human history is, by the mid-2020s, mobilising an equally state-driven campaign in the opposite direction — to persuade citizens to have more children, with diminishing success. The same instruments of the party-state — propaganda, cadre evaluation, financial incentives and penalties, and the family-planning bureaucracy reconstituted as a "birth-encouragement" apparatus — have been turned 180 degrees. Few episodes in modern governance so cleanly illustrate both the reach of a high-capacity authoritarian state into the intimate sphere of reproduction and the limits of that reach once social and economic forces run against the state's preference.

Several propositions can be stated as broadly established before the document arbitrates the genuinely contested questions in §12. First, China's fertility decline began well before 1979: the steepest fall in the total fertility rate occurred during the 1970s, under the voluntary-but-pressured wan xi shao campaign, not under the one-child policy. Second, the one-child policy — formally launched in 1979–1980 and codified over the following years — was real, was enforced with coercion that at its peaks included compulsory sterilisation and abortion, and varied enormously across regions, ethnic groups, and the rural-urban divide. Third, the policy produced large and durable demographic distortions: a severely skewed sex ratio at birth, a rapidly inverting age structure, and the "4-2-1" family form. Fourth, the policy was relaxed in stages from 2013 and abolished as a one-child rule by 2016, replaced first by a two-child and then by a three-child limit, and finally by active pronatalism — and the birth response to each relaxation was smaller than the government projected. Fifth, the population began to decline in 2022 and the TFR has fallen to among the lowest recorded for any large country.

What remains contested — and what the corpus's three-account discipline is designed to hold open rather than resolve — is the counterfactual and the causal questions. How many births did the one-child policy actually avert, over and above what modernisation and the pre-1979 decline would have produced anyway? Was the coercion necessary to China's growth miracle, or was it a costly error layered onto a decline already in train? What is now driving fertility to collapse below the level of even China's permission-free neighbours — the lingering norm of the small family bequeathed by the policy, the crushing cost structure of contemporary Chinese child-rearing, or a deeper transformation in how young Chinese women and men regard marriage and parenthood? And can a state that succeeded (at great cost) in driving fertility down succeed in driving it back up — or is ultra-low fertility, once entrenched, beyond the reach of any policy? These questions structure the analysis that follows.

3. Origins — The Fertility Transition Before 1979 and the Making of the One-Child Decision

The single most important fact for understanding the one-child policy is that China's fertility had already fallen dramatically before the policy began. In the early 1950s, the People's Republic had pursued a broadly pronatalist line consistent with Maoist confidence that "of all things in the world, people are the most precious" and a suspicion of Malthusianism as bourgeois ideology. Population grew rapidly. By the late 1960s the total fertility rate stood at roughly six children per woman. [TBD-VERIFY: late-1960s TFR is commonly cited at ~5.7–6.0; the Great Leap famine years (1959–1961) saw a sharp temporary collapse and rebound that complicate the trend line.] The decisive policy turn came not in 1979 but in 1973, with the launch of the wan xi shao (晚稀少, "later [marriage], longer [spacing], fewer [children]") campaign — a vigorous but still largely persuasion-and-pressure-based family-planning drive. Under wan xi shao, the TFR fell from roughly 5.8 around 1970 to under 3 by the late 1970s — the steepest sustained fertility decline in the country's history, accomplished before the one-child rule existed. [TBD-VERIFY: the magnitude of the 1970s decline is well documented in the demographic literature (Banister 1987; Scharping 2003), but precise annual TFR figures vary by reconstruction.]

This timing is the empirical bedrock of the revisionist scholarship (§12). Martin King Whyte, Wang Feng, and Yong Cai, in "Challenging Myths About China's One-Child Policy" (The China Journal, 2015), argue that because the largest part of the decline had already occurred under the softer 1970s campaign, the marginal demographic contribution of the post-1979 coercion was modest relative to its claimed "400 million births averted." Yong Cai's earlier decomposition (2010) and the Wang–Cai–Gu analysis (2013) make the related point that cross-national comparison with other developing societies that underwent rapid modernisation without one-child coercion suggests fertility would have continued falling toward replacement regardless.

Why, then, did China adopt a one-child rule in 1979–1980 rather than continuing the wan xi shao two-child approach? Susan Greenhalgh's Just One Child (2008) provides the authoritative account, and it is a story about the sociology of expertise. In the late 1970s, as Deng Xiaoping's reform leadership made economic modernisation the overriding priority and treated population as a denominator that had to be controlled to raise per-capita output, a group of defence and missile scientists — most prominently the cyberneticist and control-systems engineer Song Jian — applied systems-engineering and "limits to growth" modelling (drawing explicitly on the Club of Rome's The Limits to Growth and on European projections Song encountered abroad) to the population question. They produced projections treating population as a controllable variable to be optimised toward a target (often cited as an "ideal" of around 700 million by 2080), and concluded that only a one-child norm could achieve it. Social scientists and demographers who favoured a less drastic two-child design were sidelined. Greenhalgh's argument is that the one-child target was thus the product of natural-scientists' authority in a moment when that authority was ascendant, and that a less coercive path was available and advocated but not taken. The policy was formally signalled in 1979, set out in an Open Letter from the CCP Central Committee in September 1980 urging members to limit themselves to one child, and built out through provincial regulations over the early 1980s.

The early implementation was harsh. A campaign peak in 1983 saw very large numbers of sterilisations and IUD insertions; the period is associated in the literature with the most intense coercion. [TBD-VERIFY: official statistics record on the order of 20 million sterilisations and over 14 million abortions in 1983 alone, with mass IUD insertions; these aggregate figures are cited in Scharping (2003) and Banister (1987), though the coerced share is not separable.] Internal recognition that the rigid one-child rule was generating rural resistance and enforcement crises led, by the mid-1980s, to a series of accommodations — most importantly the "1.5-child" rule in much of the countryside (permitting a second child if the first was a girl) — producing the patchwork that characterised the policy for most of its life (§5).

The one-child policy was never a single uniform law; it was a system of birth-planning administration that fused national directives, provincial population-and-family-planning regulations, and intensely local enforcement. Its operating logic rested on four mechanisms.

The first was the birth permit. A couple wishing to have a child required authorisation — a birth-planning certificate — and births outside the plan ("out-of-quota" or "over-quota" births) were unauthorised, with cascading administrative consequences. The permit system made reproduction a licensed activity administered by work units (in the Mao-era and early-reform urban economy) and by village and township family-planning cadres.

The second was the quota tied to cadre evaluation. Birth-planning targets were assigned downward through the administrative hierarchy and incorporated into the cadre-responsibility evaluation system, frequently as a "one-vote veto" (一票否决) item — meaning a local official who missed birth targets could see an otherwise strong performance record nullified. This created powerful incentives for local cadres to meet quotas by whatever means, and is the structural source of much of the documented coercion: the most aggressive enforcement was often a local-cadre response to top-down targets rather than a centrally ordered atrocity, even as the centre set the targets and rewarded their fulfilment.

The third was the financial penalty, formalised over time as the "social maintenance fee" (社会抚养费, shehui fuyang fei). An out-of-quota birth incurred a fee, often calculated as a multiple of local average income and capable of reaching very large sums relative to a family's resources. The fee was both a deterrent and, for some local governments, a revenue stream — a dynamic that drew criticism for creating perverse incentives. The fees could be ruinous; failure to pay could block the hukou registration of the child (§5), with lifelong consequences. [TBD-VERIFY: aggregate national revenue from social maintenance fees has been estimated in the tens of billions of yuan annually at points in the 2000s–2010s, but consolidated official totals were rarely published and estimates vary.]

The fourth, and most contested, was direct physical coercion: compulsory IUD insertion (commonly required after a first birth), mandatory sterilisation (commonly required after a second), and coerced abortion of unauthorised pregnancies, including, in documented cases, late-term abortions. The intensity of coercion varied across time and place — peaking during campaign drives and easing between them — and the question of how much of the recorded sterilisation and abortion volume was coerced versus accepted is genuinely difficult to answer from the statistics, which record procedures but not consent. [TBD-VERIFY: there is no reliable national figure for the number of coerced sterilisations or abortions; the documented record establishes that coercion occurred, was at times systematic at the local level, and was periodically condemned even in official channels when it produced scandal, but a defensible aggregate is not available.] Mei Fong's One Child (2016) and the medical-demographic survey by Hesketh, Li, and Zhu (NEJM, 2005) both document the human consequences — the abortions, the sterilisations, the under-reported and hidden children — while noting the impossibility of a clean coerced-versus-voluntary accounting. Thomas Scharping's Birth Control in China 1949–2000 (2003) provides the most systematic institutional history of the enforcement apparatus across its first half-century.

The system also generated its own evasions and distortions, which matter for the later data debate. Families hid out-of-quota births (the so-called "black children," hei haizi 黑孩子, who lacked hukou registration); local cadres under-reported births to meet targets; and the gap between recorded and actual births became a source of the dispute, advanced most forcefully by Yi Fuxian (Big Country with an Empty Nest), over whether official population totals were systematically overstated.

A point of analytical importance, emphasised by Greenhalgh and Winckler in Governing China's Population (2005), is that the relationship between the centre and the enforcing locality was neither simple top-down command nor local improvisation but a dynamic in which the centre set targets and tolerated — or periodically disowned — the coercion that meeting those targets required. When local coercion produced scandal (a forced late-term abortion that drew public attention, a village campaign that turned violent), the centre could and did condemn "crude" or "illegal" methods, issuing directives against coerced abortion and sterilisation while retaining the quota structure that generated the pressure for them. This pattern — central targets, local coercion, periodic central disavowal of the means — is characteristic of how the high-capacity Chinese state implements intrusive policy, and it complicates any attempt to assign responsibility cleanly between a centre that "ordered" coercion and localities that "exceeded" instructions. The enforcement intensity also moved in waves rather than holding constant: campaign peaks (notably the 1983 drive) alternated with periods of relative relaxation, often in response to rural resistance or to leadership concern that over-zealous enforcement was damaging the Party's standing in the countryside. The result, across the policy's life, was an enforcement record that was simultaneously severe and uneven — capable of great coercion in a given place and year, and substantially evaded in another.

5. The Hukou Interaction and the Geography of Enforcement

The one-child policy cannot be understood apart from the hukou (户口) household-registration system, with which it was deeply entangled (the hukou system's broader governance role is treated in CN-E-04, when written). Three interactions were central.

First, the policy was never uniform across the rural-urban divide or across ethnic lines — and hukou status was the principal axis of that variation. Urban-registered households were generally held to a strict one-child standard. Rural households, after the mid-1980s accommodations, were widely permitted a second child under the "1.5-child" rule if the first was a girl (an exception that itself encoded and reinforced son preference; see §6). Ethnic-minority households were, in most regions, subject to looser limits or exempted altogether — a policy distinction that became politically charged in the 2010s when birth-control enforcement against Uyghurs and other minorities in Xinjiang intensified in the opposite direction (a development treated in CN-G-01 as part of the Xinjiang apparatus, and one of the pillars of the contested "genocide" characterisation). The result was a graduated map of fertility entitlement keyed to registration category, geography, and ethnicity rather than a single national one-child rule.

Second, the registration of an out-of-quota child was a key enforcement lever. A child born outside the plan, or for whom the social maintenance fee had not been paid, could be denied hukou registration — and without hukou, a person historically lacked access to public schooling, formal employment, healthcare, and identity documents. This made the threat of non-registration a powerful coercive instrument and produced the population of undocumented "black children." A significant administrative reform came in January 2016, when the State Council ordered that all unregistered persons — including those previously denied hukou for birth-planning reasons — be granted registration, formally severing the link between birth-planning compliance and the right to exist administratively. [TBD-VERIFY: the January 2016 State Council opinion on registering persons without hukou is documented; estimates of the number of previously unregistered persons it covered (commonly cited around 13 million from the 2010 census) should be confirmed.]

Third, the policy interacted with the great internal migration that hukou governs. As hundreds of millions moved from countryside to city for work over the reform era, enforcement followed registration rather than residence in complex ways: a rural-registered migrant couple in a city remained subject to their home jurisdiction's birth rules in principle, while the practical capacity of either jurisdiction to monitor them weakened. The mobility unleashed by reform thus partly eroded the village-level surveillance on which early enforcement had depended, even as the hukou system preserved the formal architecture of differentiated entitlement. These interactions matter for the demographic-data debate (§9) because both under-registration in the countryside and the mobility of migrants complicated the counting of births and people — the raw material of the dispute over how large China's population and how low its fertility actually were.

6. The Demographic Consequences I — The "Missing Girls" and the Sex-Ratio Imbalance

The most internationally notorious consequence of the one-child policy is the severe imbalance in the sex ratio at birth — the phenomenon of the "missing girls." In a population without intervention, roughly 105 boys are born for every 100 girls; this is the biological norm. China's sex ratio at birth rose steadily from the mid-1980s, reaching a peak in the early-to-mid 2000s. [TBD-VERIFY: the national sex ratio at birth is commonly reported as rising from ~108–111 in the late 1980s to a peak of ~118 (2005 1% population survey) or as high as ~121 in some 2008–2010 estimates, before declining toward ~111–112 by the late 2010s; precise figures vary by source and year.] In some provinces and birth orders the ratio was far higher still — second-order births in parts of the rural south and centre showed ratios well above 130. [TBD-VERIFY: province- and birth-order-specific peaks above 130 are reported in the literature (Attané 2013; Zhang 2017) but specific figures should be confirmed.]

The mechanism combined three factors, weighted differently across eras. In earlier decades, female infanticide and the abandonment of girls — practices with deep roots in the patrilineal, patrilocal family system in which sons carry the lineage and support parents in old age while daughters "marry out" — contributed to the imbalance. From the late 1980s, the spread of cheap ultrasound technology enabled prenatal sex determination and sex-selective abortion, which became the dominant driver of the rising ratio even though sex-selective abortion was formally illegal. The third factor was the under-reporting and concealment of female births — daughters hidden from registration to allow a further attempt at a son, which inflated the apparent imbalance and means some "missing girls" were missing from the statistics rather than from life. Isabelle Attané's The Demographic Masculinization of China (2013) and Junsen Zhang's survey (2017) treat the disentangling of these mechanisms as a central methodological problem.

Crucially, the sex-ratio distortion is best understood as the interaction of the birth limit with pre-existing son preference, not as a product of the policy alone. The "1.5-child" rural rule (§5) is the clearest illustration: by permitting a second child specifically when the first was a girl, the state's own enforcement design encoded the assumption that a son was the desired outcome and effectively channelled families toward sex selection on the second birth. Where fertility falls sharply but son preference persists, parents face acute pressure to ensure that their one or two permitted children include a boy — and technology gave them the means. This is why the imbalance was milder among ethnic minorities and in some urban populations with weaker son preference, and why other East Asian societies (South Korea most notably) experienced their own sex-ratio spikes during rapid fertility decline without one-child coercion, then saw the ratio normalise as son preference weakened. China's ratio likewise began to decline in the 2010s, suggesting a similar, if slower, attenuation of son preference alongside the policy's relaxation.

The durable consequence is a cohort of "surplus" men. The cumulative excess of male over female births across the high-imbalance decades has produced a marriage-market distortion in which tens of millions of men of marriageable age will be unable to find a spouse — the "bare branches" (光棍) of the demographic literature. [TBD-VERIFY: estimates of the cumulative male "surplus" range widely, commonly cited around 30–35 million excess males in the 2010s–2020s; the figure depends on age bands and assumptions and should be treated as an order-of-magnitude estimate.] The implications — for marriage migration (including cross-border bride sourcing and associated trafficking concerns), for the savings behaviour of families competing in a tight marriage market, for old-age security among never-married men, and for social stability — are the subject of a substantial economics literature (Zhang 2017 surveys the savings and human-capital channels). This is a structural feature already locked into the age pyramid; it will shape Chinese society for decades regardless of any change in current fertility.

7. The Demographic Consequences II — The 4-2-1 Family and the Ageing Society

The second great structural consequence is the inversion of the age pyramid and the family form that accompanies it: the "4-2-1" family, in which a single child, married to another single child, bears responsibility — financial and caregiving — for two parents and, as they survive into deeper old age, four grandparents. Where the traditional Chinese old-age-security model rested on multiple children sharing the support of elderly parents (and on the lineage-continuation logic that underpinned son preference), the one-child cohort confronts that obligation alone, often while raising a child of their own and bearing the cost structure described in §10.

The macro-demographic counterpart is one of the most rapid population-ageing processes ever recorded. China is ageing faster than the societies that aged before it, because the fertility collapse was both steep and partly policy-compressed, while life expectancy rose dramatically over the same period. The share of the population aged 60 and over, and 65 and over, has climbed steeply and is projected to continue rising sharply through the 2030s and 2040s as the large cohorts born before and during the early reform era enter old age while the cohorts behind them shrink. [TBD-VERIFY: the Seventh National Population Census (released 11 May 2021) reported the population aged 60+ at ~264 million (~18.7%) and 65+ at ~190.6 million (~13.5%); projections of the 60+ share approaching or exceeding 30% by around 2035 are widely cited and should be confirmed against the specific source.] The Seventh Census of 2021 was a key data event: it confirmed the rapid ageing, a continued low birth rate, and a population still — at that point — fractionally growing, and its release was accompanied by official acknowledgement that the demographic situation had become a strategic concern.

Mei Fong's One Child (2016) gives the 4-2-1 structure its most vivid human form through the shidu (失独) families — parents who lost their only child and, past childbearing age, have no other. The 2008 Sichuan earthquake, in which the collapse of school buildings killed large numbers of children, created a wave of such families and exposed the tail risk of the single-child design in a way that resonated publicly: a policy premised on each family's one child became, for those families, the loss of the entire next generation. The shidu phenomenon generated demands for state compensation and support and stands as the policy's most poignant emblem.

The ageing of the population is also the hinge between the demographic story and the economic one. For roughly three decades, China enjoyed a demographic dividend: a falling dependency ratio as the large cohorts born before the fertility decline moved into working age while the cohorts behind them were small, producing an unusually favourable ratio of workers to dependents that economists credit with a meaningful share of the growth miracle. That dividend has now reversed. The working-age population (commonly defined as ages 15–59 or 16–59) peaked around 2011–2012 and has been declining since, and the old-age dependency ratio is rising steeply. [TBD-VERIFY: the working-age-population peak is commonly dated to ~2011–2012; the exact year and definition (15–59 vs 16–64) vary by source.] The shift from dividend to drag is the demographic substance behind the "getting old before rich" thesis treated in §11.

8. The Slow Reversal — Selective Two-Child (2013), Universal Two-Child (2016), Three-Child (2021)

The dismantling of the one-child policy proceeded in cautious stages over the 2010s, and the pacing — widely judged in retrospect as too slow — is itself a governance datum about how difficult it is for a high-commitment policy apparatus to reverse course.

The first formal relaxation came at the Third Plenum of the 18th Central Committee in November 2013, which authorised the "selective two-child" policy (单独二孩, dandu erhai): couples in which at least one spouse was an only child could have a second child. The reform was framed cautiously and rolled out province by province through 2014. The response was strikingly muted: far fewer eligible couples applied for a second child than the planners had projected, an early and underappreciated signal that the desire for additional children had fallen below the level the policy assumed. [TBD-VERIFY: of an estimated ~11 million eligible couples, only on the order of 1 million applied in the first year or so; specific figures should be confirmed against the National Health and Family Planning Commission reporting of 2014–2015.]

The second and decisive step came at the end of 2015. The Fifth Plenum of the 18th Central Committee (October 2015) announced, and the NPC Standing Committee enacted in December 2015, a universal two-child policy (全面二孩, quanmian erhai), effective 1 January 2016, ending the one-child rule for the entire population. There was a modest, temporary uptick in births in 2016 — partly a "catch-up" of previously constrained second births — followed by a renewed decline as the catch-up exhausted itself and underlying fertility reasserted its downward trend. [TBD-VERIFY: births rose to ~17.9 million in 2016 from ~16.5 million in 2015, then fell in 2017 and every year after; figures vary slightly by source/revision.]

The third step, the three-child policy, was announced by the Politburo in May 2021 and formalised in the State Council's July 2021 Decision on Optimizing the Family Planning Policy and Promoting Long-Term Balanced Population Development, with the Population and Family Planning Law amended in August 2021. By this point the framing had shifted fundamentally: the document paired the higher birth limit with a suite of "supportive measures" intended to reduce the cost of childbearing — references to housing, education, employment protection for mothers, and the reduction of the burden imposed by after-school tutoring (linking the demographic policy to the "Double Reduction" education campaign of the same summer; see CN-G-03, when written). The three-child limit was less a binding constraint — almost no one was being prevented from having a third child by the limit — than a signal that the entire posture had inverted: the state now wanted more births and was beginning to recognise that permission alone would not produce them. In practice the three-child policy was rapidly overtaken by a more general pronatalism in which formal birth limits became largely irrelevant and the live question became incentives (§10).

The slowness of the reversal is itself analysed in the scholarship. Wang Feng, Cai Yong, and Baochang Gu, who had argued for years before 2013 that the policy should be abolished, attribute the delay partly to the institutional entrenchment of the family-planning bureaucracy (the National Population and Family Planning Commission, with its large staff and revenue from social maintenance fees, had an organisational interest in the policy's continuation) and partly to a persistent official fear — rooted in the original 1980s logic — that relaxation would produce a fertility rebound that never came. The bureaucracy was reorganised: the family-planning commission was merged into a new National Health Commission in 2018, symbolically downgrading birth control from a standalone mission to one health function among many.

9. The Turning Point — The 2022 Population Decline, India's Overtaking, and the Collapsing TFR

The demographic turning point arrived, by official reckoning, at the end of 2022. On 17 January 2023, the National Bureau of Statistics reported that China's population had declined in 2022 — the first annual decline since 1961, in the aftermath of the Great Leap Forward famine. [TBD-VERIFY: the NBS reported the end-2022 population at ~1.41175 billion, a decline of ~850,000 from end-2021, with ~9.56 million births and ~10.41 million deaths; subsequent releases reported a larger decline of ~2.08 million for 2023; end-2024 figures should be confirmed against the relevant NBS release.] The crossing from growth to decline was symbolically momentous for a state whose entire population-governance apparatus had been built around the premise that the central demographic problem was too many people.

Close behind came the second milestone: India's overtaking of China as the world's most populous country. The United Nations Population Fund and the UN Population Division's World Population Prospects projected that India would surpass China during 2023, a transition widely reported in April 2023. [TBD-VERIFY: the UN projected India's population would reach ~1.4286 billion against China's ~1.4257 billion around mid-2023; exact figures are UN estimates, as India's own census had been delayed.] The handover of the "most populous nation" title — held by China (and its predecessor states) for centuries — to India crystallised the shift in the global demographic balance and fed directly into the strategic-competition discourse about the relative trajectories of the two giants.

Underlying both milestones is the collapse of the total fertility rate to among the lowest in the world. China's TFR, having fallen below the ~2.1 replacement level around the early 1990s, dropped through the 2010s and into the 2020s to a level commonly estimated at roughly 1.0 to 1.1. [TBD-VERIFY: China's TFR for 2022–2023 is variously estimated at ~1.0–1.1; some analysts put 2023 at or below 1.0, others slightly above; the NBS does not publish an authoritative annual TFR and estimates depend on birth and base-population assumptions.] At a TFR around 1.0, each successive generation is roughly half the size of the one before — a contraction faster than even the most pessimistic projections of the 1990s and 2000s assumed.

The arrival of the turning point earlier than official projections had assumed vindicated a particular critic. Yi Fuxian, in Big Country with an Empty Nest (大国空巢, first published 2007, expanded later), had argued for years that China's official population figures were inflated — that local under-reporting offset by inflated census adjustments had produced an overcount — and that the demographic crisis was both larger and arriving sooner than the authorities acknowledged. While the precise magnitude of any overcount remains disputed, the broad direction of Yi's argument — that the turning point was closer than official optimism allowed — was borne out by the 2022 decline and the sub-1.1 TFR. The episode illustrates a recurring feature of the subject: the data themselves are contested, and the gap between recorded and actual births and deaths (the legacy of decades of birth concealment and quota-driven under-reporting) means that even the basic question of how many Chinese there are, and how fast the number is falling, carries genuine uncertainty.

The COVID-19 period (2020–2022) compounded the decline. The pandemic, the Zero-COVID disruptions (see CN-C-01, CN-D-01 context), and the economic uncertainty of those years depressed marriages and births further, and the chaotic exit from Zero-COVID in late 2022 produced a surge in deaths that contributed to the 2022 crossing into decline. Marriage registrations — a leading indicator of births in a society where childbearing remains closely tied to marriage — fell to multi-decade lows over the same period. [TBD-VERIFY: marriage registrations fell from a peak of ~13.5 million couples in 2013 to roughly half that by the early 2020s; specific annual figures should be confirmed against the Ministry of Civil Affairs data.]

10. The Pronatalist Scramble — Subsidies, the 2025 Childcare Measures, and the Efficacy Debate

By the mid-2020s the Chinese state had completed its turn from birth restriction to birth encouragement, and the policy frontier had moved decisively from permission to incentive. The recognition driving this shift is that removing the birth limit accomplished little, because the constraint that now suppresses fertility is not the law but the cost — and the broader social transformation — of having children in contemporary China.

The pronatalist toolkit assembled since 2021 spans several instruments. Local cash and benefit incentives proliferated first: cities and provinces introduced one-time childbirth payments, monthly subsidies for second and third children, extended maternity and paternity leave, and housing-purchase preferences for larger families. These were uneven, locally funded, and generally modest relative to the cost of child-rearing. Expansion of childcare became a central plank: the shortage of affordable nursery places for children under three is widely identified as a key barrier, and successive plans set targets for nursery-slot provision. Fiscal and tax measures included expanded deductions for child-rearing and education expenses under the individual income tax.

The most significant escalation came in 2025 with the move toward a national childcare subsidy, shifting from the patchwork of local measures toward a centrally framed cash benefit for families with young children. [TBD-VERIFY: the design, eligibility, and per-child amounts of the 2025 national childcare-subsidy measures — reported in 2025 — should be confirmed against the specific State Council / relevant-ministry documents; figures and scope circulating in reporting should not be cited as definitive without the primary text.] The 2025 measures, together with the "silver economy" and childcare-expansion drives, mark the consolidation of pronatalism as a standing priority of the 14th–15th Five-Year-Plan period rather than an ad hoc response.

The efficacy debate is sharp and is the third of the document's three-account axes (§12). The pronatalist-optimist position holds that sufficiently generous and sustained support — comprehensive childcare, housing relief, employment protection for mothers, and direct cash transfers at a scale that materially offsets the cost of children — can lift fertility, and points to modest positive effects from generous family policy in some European cases. The pessimist position, anchored in the comparative East Asian record, observes that South Korea, Taiwan, Singapore, Hong Kong, and Japan have all deployed pronatalist incentives — some lavish — without arresting their fertility declines, and that South Korea's TFR has fallen below 0.8, the lowest in the world, despite large expenditures. [TBD-VERIFY: South Korea's TFR is commonly cited at ~0.72–0.78 for 2023; confirm the specific year/figure.] On this reading, once a society reaches China's level of urbanisation, female education, housing cost, and competitive child-rearing ("involution"), ultra-low fertility becomes a stable equilibrium that incentives at any politically feasible scale cannot dislodge — because the drivers (the cost and opportunity-cost of children for educated women, the neijuan of the education arms race, the retreat from marriage) are structural and cultural rather than the product of a removable legal limit. The Chinese state's own behaviour — the steady escalation of measures, the linkage to the tutoring crackdown, the move to national subsidies — suggests an institution searching for the lever that will work and not yet having found it.

A distinctive Chinese feature complicating pronatalism is the tight coupling of childbearing to marriage and the simultaneous retreat from marriage among young Chinese, particularly urban, educated women. Falling and later marriage, the high cost of the marriage-market itself (bride prices, housing as a precondition for marriage), the surplus-men distortion (§6), and a documented shift in the attitudes of young women toward marriage and motherhood together depress the marriage rate that still gates most births. Pronatalist policy that subsidises a second or third child does little for a population that is increasingly not forming the first marriage or the first birth. This is why some analysts argue the binding causal factor is less "cost of children" narrowly and more a broader transformation in gender relations, work, and the meaning of family — a transformation the policy levers available to the state are poorly designed to reach.

11. The Eldercare and Pension Strain — Raising the Retirement Age and the "Getting Old Before Rich" Thesis

The demographic inversion translates directly into fiscal and institutional strain through two channels: the pension system and the eldercare-services system.

China's pension architecture is fragmented and unevenly funded, combining urban-employee, urban-resident, and rural-resident schemes with very different benefit levels, and resting substantially on a pay-as-you-go basis in which current workers' contributions fund current retirees' benefits. A pay-as-you-go system is acutely sensitive to the worker-to-retiree ratio, and that ratio is deteriorating rapidly as the working-age population shrinks (§7) and the retired population swells. Official and academic projections have repeatedly warned of the depletion of the main urban-employee pension fund's accumulated balance within the following one-to-two decades absent reform. [TBD-VERIFY: a widely cited 2019 Chinese Academy of Social Sciences projection warned the main urban-employee pension fund could be depleted by ~2035; this is a projection under stated assumptions, not a forecast, and should be cited as such.] Regional imbalance compounds the problem: ageing, lower-growth provinces (especially in the north-east) run pension deficits while younger, higher-growth provinces run surpluses, prompting central efforts at cross-provincial pension pooling and transfers.

The most politically sensitive response came on 13 September 2024, when the NPC Standing Committee adopted the Decision on Implementing the Gradual Delay of the Statutory Retirement Age, effective 1 January 2025 — the first increase in China's statutory retirement ages in decades. China had retained unusually low retirement ages, set largely in the 1950s: 60 for men, and 50 or 55 for women depending on whether they were classified as blue- or white-collar workers. The 2024 decision raises these gradually over a fifteen-year phase-in — for men toward 63, and for women toward 55 or 58 depending on category — alongside increases in the minimum contribution period required for a full pension. [TBD-VERIFY: the end-state ages (men to 63; women to 55 and 58 by category) and the fifteen-year phase-in and contribution-period changes should be confirmed precisely against the 13 September 2024 decision text.] The reform's caution — a long phase-in, small annual increments — reflects the political delicacy of asking workers to retire later in a labour market with youth-unemployment pressures, and the measure generated visible public discontent online, an unusual outcome for a policy enacted through the NPC.

In parallel, the state has promoted an eldercare-services build-out under the "silver economy" (银发经济) and "active response to population ageing" rubrics: expansion of community and home-based care, nursing-home capacity, long-term-care insurance pilots, and a policy push to develop eldercare as an industry. The dominant model promoted is sometimes summarised as "9073" or "9064" — the aspiration that the large majority of the elderly age at home with community support, a smaller share in community day-care, and only a few per cent in institutional facilities. [TBD-VERIFY: the "9073"/"9064" eldercare-distribution formulations appear in Chinese policy and planning documents; the precise figures and their official status should be confirmed against the relevant Five-Year-Plan eldercare provisions.] The challenge is structural: the 4-2-1 family (§7) means the traditional model of family-provided eldercare is breaking down precisely as the elderly population peaks, forcing a transition to socialised care that China is attempting to build at speed and at a scale no society has previously managed. The geographic mismatch sharpens the problem — rural areas, drained of working-age adults by decades of migration to the cities (a movement governed by the hukou system; see CN-E-04, when written), are ageing fastest and have the thinnest care infrastructure, so that the elderly "left behind" in the countryside often lack both the children who once provided care and the institutional services that are concentrated in wealthier urban areas.

The eldercare and pension strains together illustrate why demography has become a whole-of-government concern rather than a sectoral one. The financing of pensions, the building of care capacity, the retirement-age reform, the labour-market consequences of a shrinking workforce, and the pronatalist effort to slow the decline are not separate policies but facets of a single problem — the conversion of the demographic dividend into a demographic burden — and the state's response spans the fiscal, labour, health, housing, and family-policy bureaucracies simultaneously. The "active response to population ageing" was elevated to a national strategy in the 14th Five-Year-Plan period, signalling that ageing had joined the small set of cross-cutting strategic priorities around which the planning system organises itself.

These strains are the substance of the "getting old before getting rich" thesis associated with Nicholas Eberstadt and widely echoed in the economic-demographic literature. The argument is comparative: Japan, the Western European states, and South Korea all became aged societies at high per-capita income levels, having completed the climb to wealth before the demographic burden arrived; China is becoming a super-aged society at a per-capita income level substantially lower than theirs was at the equivalent demographic stage. [TBD-VERIFY: comparisons of per-capita income at equivalent ageing thresholds vary by metric (nominal vs PPP) and source; the qualitative claim that China ages at a markedly lower income level than the precedent high-income agers is well established, but specific dollar figures should be confirmed.] The implication is that China must finance the pension, healthcare, and eldercare costs of a super-aged society out of an economy that has not finished growing rich — even as the labour force that generates the output shrinks. This intersects directly with the broader 2020s economic pivot (CN-D-01): the turn to "new quality productive forces," automation, and productivity growth is, among other things, an implicit answer to the demographic constraint — an attempt to substitute productivity for the labour and the demographic dividend that demography is withdrawing.

12. The Three Contested Accounts — Necessity, Causation, and Reversibility

The corpus's three-account discipline applies along three distinct axes for this subject, each with serious exponents on multiple sides.

Axis one — was the one-child policy necessary? Account A (official / necessity): The policy averted hundreds of millions of births — the figure "400 million births averted" is the canonical official claim — and thereby raised per-capita income, relieved pressure on resources, food, employment, and the environment, and enabled the reform-era growth miracle by holding down the denominator. On this view the policy, however harsh, was a sacrifice that made modern China's prosperity possible. Account B (revisionist / avoidable harm): Whyte, Wang Feng, and Cai Yong (2015) and Wang–Cai–Gu (2013) argue that the steepest fertility decline preceded the one-child policy (it occurred under wan xi shao in the 1970s), that the policy's marginal demographic effect over and above what a continued two-child campaign would have achieved was therefore modest, and that the "400 million averted" figure is an artefact of comparing actual births against an implausibly high counterfactual. Daniel Goodkind's careful re-estimation (2017) similarly questions the headline number while acknowledging real uncertainty. On this account the coercion — the forced sterilisations, the coerced abortions, the sex-ratio distortion, the accelerated ageing — was largely an avoidable cost layered onto a transition already underway. Account C (modernisation reading): Yong Cai (2010) and the development-versus-policy literature hold that fertility decline tracks socioeconomic development — female education, urbanisation, falling child mortality, rising costs of children — across societies regardless of coercive policy, and that China would have reached low fertility through modernisation alone, perhaps somewhat more slowly. The three accounts disagree less about what happened than about the counterfactual, which is by nature unobservable; the corpus records that the necessity claim rests on a counterfactual that the revisionist and modernisation readings have made genuinely contestable.

Axis two — what causes today's ultra-low fertility? Account A (policy legacy): Decades of one-child enforcement and propaganda normalised the small (one-child) family as the cultural default, so that even after the limit was removed, expectations and norms remained anchored at one child — a path-dependent legacy of the policy itself. Account B (cost / involution): The dominant constraint is economic — the cost of housing (and of housing as a marriage precondition), the cost and intensity of education in a hyper-competitive system (neijuan, "involution"), the opportunity cost to women's careers, and the absence until recently of affordable childcare. On this reading fertility is suppressed by the lived economics of contemporary urban China, which the tutoring crackdown and childcare subsidies are explicitly designed to address. Account C (gender and marriage norms): The deepest driver is a transformation in the attitudes of young Chinese — especially educated urban women — toward marriage and motherhood: declining and later marriage, rising singlehood, a reluctance to accept the "second shift" of childcare and housework on top of paid work, and a re-evaluation of whether marriage and children are desirable at all. Because most births remain gated by marriage, the retreat from marriage is, on this view, the proximate cause, and it is a cultural shift that pronatalist cash transfers are poorly designed to reverse. These three are not mutually exclusive — most analysts assign weight to all three — but they imply very different policy responses, and the disagreement over their relative weight is real.

Axis three — can the state reverse it? Account A (pronatalist optimism): A high-capacity state that succeeded in driving fertility down can, with sufficiently generous and comprehensive support — childcare, housing, cash, employment protection — raise it again; the lever simply has to be large and sustained enough, and China is still searching for the right scale. Account B (comparative pessimism): No East Asian society at China's fertility level has recovered despite pronatalist spending; South Korea, Taiwan, Singapore, Hong Kong, and Japan are the cautionary cases, with South Korea's sub-0.8 TFR the starkest. On this view ultra-low fertility is a stable equilibrium produced by structural and cultural forces beyond the reach of feasible incentives, and China will follow the same path. Account C (the data-and-demography caution, associated with Yi Fuxian): The question is in one sense already moot, because the age structure is locked in — the women who would bear the next generation are already born and are too few; even a sudden fertility rebound would take a generation to affect the working-age population, and the momentum of decline and ageing is now built into the pyramid regardless of policy. The corpus records all three: the optimism is not refuted, the comparative record is genuinely discouraging, and the demographic momentum is, for the next two decades, essentially fixed whatever fertility does.

13. Conclusion — Demography as the Binding Constraint on the Chinese Century

China's population governance from 1979 to 2026 is a study in the power and the limits of the high-capacity authoritarian state. The same apparatus that drove fertility down — through permits, quotas, fees, propaganda, and at its peaks through coercion that included forced sterilisation and abortion — has been turned to drive fertility back up, and is discovering that the second task is far harder than the first. The asymmetry is instructive: a state can compel people not to have children with a degree of success (at severe human cost and with durable distortions), but it cannot compel them to have children, and the inducements available fall well short of the structural and cultural forces now suppressing the birth rate.

The disciplined reading holds several propositions together. The fertility decline preceded the one-child policy, which means the policy's necessity rests on a contested counterfactual; the policy nonetheless inflicted real and documented harms — the missing girls and the resulting surplus of men, the coerced procedures, the shidu families, and an accelerated ageing that will burden the economy for decades. The reversal came late and produced ever-smaller birth responses, revealing that by the 2010s the binding constraint had migrated from the law to the cost of children and to a deeper transformation in marriage and gender. The 2022 population decline and the sub-1.1 fertility rate are not the policy's failures alone but the convergence of policy legacy, economic structure, and cultural change — and they are, for the next generation, demographically irreversible regardless of what the state now does, because the parents of the next cohort are already too few.

For the broader corpus, the demographic crisis is best understood as the binding constraint within which every other element of contemporary Chinese governance now operates. The pension and eldercare strain shapes fiscal policy; the shrinking workforce shapes the turn to automation and "new quality productive forces" (CN-D-01); the surplus of men and the retreat from marriage shape social policy and stability concerns; and the comparison with India's younger population feeds the strategic-competition discourse. The "getting old before rich" thesis frames the central wager of the Chinese economic model in the 2020s and 2030s: whether productivity growth can be made to substitute for the demographic dividend that demography is withdrawing. The forward questions for 2026 and beyond are three, mapped onto the three contested axes. Can pronatalism at any feasible scale move a TFR near 1.0 — or will China follow the East Asian path to ever-lower fertility? Will the retreat from marriage among young women deepen or stabilise? And can a super-aged society be financed and cared for out of a still-middle-income economy with a shrinking workforce? On all three, the corpus holds the question open, records the contesting accounts, and treats demography not as a settled fact but as the most consequential — and least reversible — of the constraints now shaping the Chinese state.

CN-A-01: The Foundational Arc — Mao, Deng, Jiang, Hu and the Pre-Xi Architecture of the People's Re
CN-D-02: The Third Plenum of the 20th Central Committee, the September 2024 Stimulus, and the Two S
CN-D-04: The 2026 Two Sessions, "New Productive Forces" Operationalised, and China's Trump-2 Tariff
CN-D-05: The 2026 Fourth Plenum and the 15th Five-Year Plan (2026–2030) — Post-Trump-2 Industrial P
CN-D-06: The Trump-2 Trade War and the Rare-Earths Leverage — IEEPA Tariffs, the Geneva Modus Viven
CN-G-04: The Hukou System and Migration Governance — China's Internal Border (1958–2026)
CN-H-PREM-01: Li Keqiang (1955–2023) — Premier of China 2013–2023
CN-H-PREM-02: Wen Jiabao (1942–) — Premier of China 2003–2013
CN-I-01: China's Party-State Institutional Architecture — The Machinery of Party-Led Governance fro
CN-N-01: China in International Perceptions — From Engagement Consensus to Systemic Rival (1978–202
CN-O-01: China Megatrends — The 2030s Questions
JP-G-03: Japan's Demographic Transition — Population Decline, Pension Reform, and the Caregiving St