June 25, 2026 · 《Ming Pao》
In Hong Kong’s housing statistics, there is an unwritten equation: one unit equals one unit. 27 square metres, count one. 60 square metres, also count one. This equation has kept us from seeing what has truly happened over the past thirty years.
The latest research report from the Institute of Future Economics has compiled completion data for three types of housing in Hong Kong from 1995 to 2024. We are not asking how many units were built, but how much living space was built. The answer is sobering: the average unit size fell from 50.4 square metres in 1995 to 37.2 square metres in 2024 — a decline of 26%. In other words, today’s “one unit” is a quarter smaller than “one unit” thirty years ago.
The HOS figures best illustrate the problem. In 2000, the government built about 23,500 HOS units. In 2024, it built 14,000. The number of units is roughly 60% of that of the year 2000.
But if we look at actual living space — the area that truly lets people live — the HOS additions in 2024 amount to only 30% of the year 2000 figure.
Sixty percent of the units, thirty percent of the space. Where did the missing thirty percent go? It was never built.
The situation in public housing is similar, but one detail deserves separate mention: three-bedroom units have disappeared from all newly completed public housing since 2009. Not reduced — disappeared. The average public-housing unit size has fallen from 40.2 square metres in 1995 to 26.7 square metres today. The 10,000 public-housing units built today provide living space equivalent to only about 6,600 units in 1995.
The private market is no different. Between 1995 and 1999, private homes under 40 square metres accounted for 14% of completions. By 2020 to 2024, that share had risen to 45%. Nano-flats have become the mainstream of Hong Kong’s private residential market, not the exception.
Some may say: Hong Kong household sizes are shrinking, from 3.6 to 2.6 persons per household on average, so per capita living space may not have worsened. This argument is not unreasonable, but the causality may run in the opposite direction. Small units make marriage and child-rearing more expensive, and make living with parents harder. The reason households are smaller today may, in part, be that units are too small — it is not that Hong Kong people do not want to form families, but that even when they do, they cannot fit.
Singapore’s public housing averages 80 square metres. New homes in England average 79 square metres. Tokyo, 66 square metres. Hong Kong, 37.2 square metres. These cities all face high land prices, high property prices, and similarly limited land — but they chose to build different things.
Increasing the number of units is not unimportant. But if new units are getting smaller and less suitable for family living, then the increase in unit numbers is merely a statistical improvement, not an improvement in life.
Hong Kong needs to formally incorporate total new living space, average unit size, and the share of family-sized units as policy indicators in the Long Term Housing Strategy. The more fundamental question is: for public housing, HOS, and government land sales, minimum size standards with real binding force should be established and enforced. The problem has never been that the data is opaque — the data has always been there. The problem is that this data has never truly constrained any decision.
The purpose of housing policy has never been merely to provide a place to live. It is to let people settle down, start families, and live in this city for the long term.
A 27-square-metre unit may be liveable. But does it let people stay?
(Translation supported by AI)
Research report:
