# The City That Doesn't Want Children: How Building Typology Shapes Demographic Fate
Seoul's total fertility rate in 2024 was 0.58. That is not a typo. Replacement rate is 2.1. South Korea is running at 0.75 nationally and 0.58 in its capital — the lowest of any major city on the planet. Singapore sits at 0.97. Tokyo at 0.96. Hong Kong at 0.84.
These are not marginal deviations from replacement. They are civilizational numbers. A society at 0.58 loses more than half its population every generation, geometrically, without any mechanism to stop it once it starts.
The policy response has been to throw money at the problem. South Korea has spent more than $200 billion on pro-natalist subsidies since 2006. Singapore has had cash bonuses, housing priority, and extended parental leave since 1987. Forty years and enormous resources later, Singapore's TFR fell below 1.0 for the first time in 2023 and stayed there in 2024. Hungary raised its TFR from 1.25 in 2010 to a peak of 1.61 in 2021 through one of the most comprehensive family support programs in the OECD — and by 2024 it had declined back to 1.39.
No country has demonstrated a durable policy-driven fertility increase. Not one.
This suggests the problem is structural, not incentivizable. The fertility collapse is not happening because people lack information about tax credits. It is happening because the built environment has made having children physically and economically irrational.
The Bedroom Equation
In 2024, researcher Lyman Stone analyzed approximately 12 million women of reproductive age from American Community Survey data spanning 2006 to 2022. His finding is stark:
Women in one- or two-bedroom apartments average a TFR of 1.2 to 1.35. Women in three-bedroom homes average 1.9 to 2.0, near replacement.
The difference between sub-Korean and near-replacement fertility is one bedroom.
You cannot raise three children in a studio apartment. So people don't. The architectural constraint becomes a demographic outcome, and the demographic outcome compounds geometrically through Lutz, Skirbekk, and Testa's "low-fertility trap" mechanism: fewer mothers in each generation means fewer births in the next, while smaller families become the cultural norm, while housing costs price out further formation. Once below 1.5, the trap tends to hold.
What Towers Actually Deliver
The connection between skyscrapers and small apartments is not accidental. It is economic. When land is expensive, developers build tall to amortize land cost over more units. When selling expensive-land units, they optimize for price per square foot, which means smaller rooms. A three-bedroom unit in central Tokyo or Hong Kong costs what most families in those cities cannot afford; the affordable unit is studio or one-bedroom.
Hong Kong residents live with a median 16 square meters of floor space per person. The average Tokyo condominium in the 23 wards is 66.5 square meters total. The US median new single-family home is 2,210 square feet, or roughly 205 square meters. These are not comparable living environments for family formation.
The Fire Code Accident
Here is something almost nobody mentions: the American double-loaded corridor apartment building, with its long anonymous hallways serving dozens of units on each floor, is not the product of density logic or economics. It is an artifact of an early 20th-century fire code requiring two staircases in residential buildings. That code forced a wide building footprint with a central corridor — the hotel layout applied to housing.
Europe never adopted this requirement. European apartment buildings continued using the point-access stairwell model: one stair serving four to six units per landing. A landing with four neighbors becomes a functional micro-neighborhood. You know who lives next to you. Children play on the landing. The social topology is fundamentally different from a hallway serving forty.
Oscar Newman documented the consequences in Defensible Space (1972): crime rates in high-rise public housing corridors were dramatically higher than in comparable low-rise buildings. More families sharing an anonymous circulation space reduced the capacity for the informal social monitoring that makes residential environments feel safe and function as community.
Several American cities — Denver, Culver City, San José — have already changed their fire codes to permit single-stair residential buildings. The corridor was a regulatory accident. It is recoverable.
The City Was Always a Sink
Pre-industrial cities were demographic sinks in a more literal sense. For most of the 17th and 18th centuries, European cities had more deaths than births. London accumulated approximately half a million more deaths than births over the course of the 17th century. Cities like Paris had infant mortality around 300 to 400 per 1,000 births when the national average was roughly 180. Pre-modern cities could not sustain their populations through natural increase. They required constant rural in-migration just to maintain size.
The mechanism was disease: typhus, plague, cholera, contaminated water — all products of density without sanitation infrastructure. Once the Victorians built proper sewage systems and clean water supplies, cities became capable of natural population growth for the first time.
The modern megacity has inverted the mechanism but reproduced the outcome. Tokyo, Seoul, and Singapore are not killing their residents. They are preventing the formation of the next generation, through cost, space constraints, and the social topology of vertical living. The pre-modern city was a biological trap. The modern city is an economic and architectural one.
The Typology That Works
Jane Jacobs spent her career opposing towers. She was right, and the fertility data now provides a quantitative mechanism she did not have: unit size is the binding constraint, and the building typology she defended — three- to six-story mixed-use street buildings, varied ages, frequent intersections, ground-floor commerce — is precisely the typology that can deliver three-bedroom units at urban densities without requiring the land costs that drive tower economics.
Leon Krier's urbanism prescription pointed to the same form: quarters of at most 35 hectares, crossable in ten minutes on foot, three to five story buildings maximum, all daily functions mixed within the quarter. The perimeter block (buildings enclosing a shared courtyard, ground-floor commercial uses, residential above) delivers street life, semi-private outdoor space, and adequate unit sizes simultaneously.
The Finnish housing register data confirms this directionally: couples in row or terraced houses show a 36% higher risk of first conception than apartment dwellers, and those in single-family houses 53% higher — though the authors note this partly reflects selection, as families planning children tend to move to larger housing first.
Architecture Is Demographic Policy
The US national TFR hit 1.599 in 2024, a record low. A University of Toronto analysis found that rising housing costs since 1990 are responsible for approximately 51% of the total US fertility decline over the following decades. Had housing costs remained stable, an estimated 13 million more children would have been born.
We have been running demographic policy through the built environment without acknowledging it as such. Zoning codes that prohibit three-bedroom apartments at urban densities are fertility policy. Fire codes that mandate the double-loaded corridor are fertility policy. Land use regulations that make the rowhouse and the courtyard apartment illegal in most American cities are fertility policy. They just don't appear in the budget line next to the baby bonus.
South Korea spent $200 billion trying to fix with money what it built into its cities. Singapore has been trying for forty years. Neither can overcome the structural constraint that their built environments impose on family formation.
The fertility crisis is, in part, an architecture crisis. The intervention that works is not a cash transfer. It is a three-bedroom unit within walking distance of a school, at a price that doesn't require choosing between a family and a mortgage.
That is a building typology problem. And building typology is something we actually know how to change.