In London's Race for Space Was a Race for Houses we split the pandemic "race for space" into three factors: a house against a flat of the same size, a bigger home against a smaller one, and a home further out against one in the centre. Here we put those three into one number, the Space Race Factor, and compute it for every city we cover, from London to Tokyo (all ten are on the Compare page). Against 2018 to 2019, Busan (+9.9%), Tokyo (+6.6%), Seoul (+5.5%) and London (+5.0%) raced for space hardest. New York barely moved, and Taipei went the other way. Since 2022 the race has stalled almost everywhere.
Each of the three factors has a Premium, published on each city's Factors page (for example, London's): the % gap in price per square metre between two otherwise identical homes that differ in one stated way. Each is pointed in the direction a race for space would push it:
A factor's Factor Return is the % change in its Premium between two periods. The Space Race Factor is the average of the Factor Returns of the factors a city has. A positive score means the market now pays relatively more for houses, space and distance than it did. We average rather than add so that a city with one factor sits on the same scale as a city with three: added up, London's score would be +14.9% instead of +5.0%.
| City | Factors in the score | 2018–19 to 2021 | 2018–19 to latest | 2022 to latest |
|---|---|---|---|---|
| Busan | Size | +5.6% | +9.9% | +0.8% |
| Tokyo | Size, Further Out* | +3.7% | +6.6% | +1.4% |
| Seoul | Size | +5.8% | +5.5% | -2.8% |
| London | House vs Flat, Size, Further Out | +2.8% | +5.0% | +0.9% |
| Osaka | Size, Further Out* | +1.7% | +3.1% | +0.4% |
| New York | Size | -1.3% | -1.1% | +1.9% |
| Taipei | House vs Flat*, Size* | -2.9% | -13.9% | -9.3% |
| Paris | House vs Flat, Size, Further Out | no data | no data | +0.4% |
| Lyon | House vs Flat, Size, Further Out | no data | no data | -3.2% |
| Singapore | Size* | no data | no data | -2.3% |
Space Race Factor: the average Factor Return of the factors listed, from the average Premium in the first period to the average Premium in the second. "Latest" is each city's latest 12 monthly windows, leaving out the two most recent, which are still filling: to May 2026 for London, October 2025 for Paris and Lyon, November 2025 for Tokyo and Osaka, and between February and July 2026 elsewhere. * a stand-in for the factor, see the last table.
London is the one city where all three factors go back well before the pandemic, and it scores +5.0%. That can look small next to the raw gaps in the London post, where houses beat flats by about 13 percentage points, for three reasons. The raw gaps mix the three factors together, so each one is counted more than once; holding the others constant removes that. A Factor Return is a % change in the Premium, so a house premium rising from 16% to 27% is a gain of +8.8% for the house, not 10 points. And the score is an average: summed, London's three factors come to +14.9%, the highest total of any city with more than one factor. Most of that is the house premium: a house sold for 16.4% more per square metre than a same-size flat in 2018 to 2019 and 26.6% more in the latest 12 months, a Factor Return of +8.8%. Size added +5.1% as the discount for a larger home all but disappeared. Further Out added only +1.0%: the discount for outer London narrowed in 2021 and has since widened back. That is the London post's finding in one number: London raced for houses more than for distance. The London Trends page shows how much each factor added to the constant-quality London Index, and the London neighbourhood pages show prices by area and postcode district.
The highest scores are in East Asian apartment markets, and they come from Size. In 2018 to 2019 a Tokyo flat with twice the floor area of an otherwise identical one sold for 8.4% less per square metre. In the latest 12 months it sold for 7.7% more, a Factor Return of +17.6%, the largest single move in this post. Osaka (+9.1%), Busan (+9.9%) and Seoul (+5.5%) moved the same way: from a discount per square metre for bigger flats to a much smaller one, or a premium.
Distance went the other way. For Tokyo and Osaka our stand-in for Further Out is five more minutes' walk to the station, and both cities now pay less for that walk than before: the discount grew from 3.8% to 8.1% in Tokyo and from 6.4% to 9.2% in Osaka. Buyers in Japan wanted more room, but close to the station, and it pulls Tokyo's combined score down to +6.6% from a Size gain of +17.6%. Seoul and Busan have no distance factor, so their scores are Size alone and likely flatter them. Prices by ward are on the Tokyo, Osaka, Seoul and Busan neighbourhood pages.
New York's model has Size but neither houses against flats nor distance from the centre. Its Size Premium barely moved: a home with twice the floor area sold for about 22% less per square metre in 2018 to 2019 and about 23% less in the latest 12 months. On this measure New York had no race for space at all; the New York Trends page shows what did move its prices.
Taipei is the outlier. Its stand-in for House vs Flat compares a house with a same-size flat in a residential tower, and that premium fell from 21.6% in 2018 to 2019 to -4.8%, a Factor Return of -21.7%. The premium for 100 more square metres fell too. Taipei buyers moved towards modern tower flats, not away from them. Houses are a small share of sales within Taipei city, so treat the size of that move with caution, but its direction holds in every year since 2021. Prices by district are on the Taipei neighbourhood pages.
Rebase every city to 2022 and all ten can be compared, including Paris, Lyon and Singapore. No city has gained more than +1.9% since then, and four have fallen back. In Paris the house premium edged up from 31.6% to 34.2% while Size and distance were flat, for a score of +0.4%. In Lyon the house premium fell from 51.3% to 40.9% and the score is -3.2%. Seoul (-2.8%) has given back about a third of its earlier Size gain, and Singapore's premium for a large flat slipped from 8.8% to 6.2%. London (+0.9%) is still moving the same way, on the house premium alone.
| City | House vs Flat | Size | Further Out |
|---|---|---|---|
| London | +8.8% | +5.1% | +1.0% |
| Tokyo | no houses | +17.6% | -4.5%* |
| Osaka | no houses | +9.1% | -3.0%* |
| Busan | no houses | +9.9% | not in model |
| Seoul | no houses | +5.5% | not in model |
| New York | not in model | -1.1% | not in model |
| Taipei | -21.7%* | -6.0%* | not in model |
| Paris (since 2022) | +1.9% | 0.0% | -0.7% |
| Lyon (since 2022) | -6.8% | -1.5% | -1.3% |
| Singapore (since 2022) | no houses | -2.3%* | not in model |
Factor Returns from 2018–19 to the latest 12 months, or from 2022 where marked. "No houses": the city's data covers apartments only. "Not in model": the factor did not pass the city's factor selection.
The stand-ins, marked *, are the closest factor each city's model has:
"Race for space" meant different things in different cities. In London it meant houses. In Tokyo, Osaka, Busan and Seoul it meant bigger flats, but not further from the station. New York did not race, and Taipei moved towards towers. Wherever it happened, most of the move came by 2022, and in the four years since no city has added more than about 2%. Where a city has a Premium that is still moving, as with London's house premium, that factor is worth watching more than the headline price. Each city's Trends page, such as Paris or Tokyo, shows how its Premiums feed into its Index, and the London property calculator uses them to price a single home.
Every city's Index, Factor Returns and Factor Contributions, side by side.
Compare cities →Each city's model is a rolling 3-month cross-sectional regression of log price per square metre on that city's selected factors (location, size, building type and age, and others), so each Premium holds every other factor constant. Premiums are quoted at a fixed, stated exposure and published on each city's Factors page. How the model is built is on the About page. A Factor Return between two periods is (1 + average Premium in the second) / (1 + average Premium in the first) - 1. The Space Race Factor is the plain average of the Factor Returns of a city's House vs Flat, Size and Further Out factors, each pointed so that a positive value means more is paid for a house, for space or for distance. The first two monthly windows of every series (still filling) and the last two (likewise) are left out. Factors are selected separately for each city, so the factors in each score differ, as the tables show.