"Japan Is Shrinking" Is the Wrong Chart — Tokyo Is Getting Denser While the Countryside Empties

TL;DR: “Japan’s population is shrinking” is true at the national level and badly misleading at the level you actually buy at — the city. Tokyo’s 23 special wards keep hitting record population with rental occupancy around 96.6%, while roughly 40 of 47 prefectures lose people every year (directional, as of writing). You are not buying “Japan.” You are buying the one drain the entire country is flowing into.

---

The headline you read is the wrong unit of analysis

Almost every foreigner who looks at Japanese property hits the same wall first: a chart of the national population sloping down, and a caption that says the country is dying. The conclusion writes itself — why buy real estate in a place losing people?

Here is the problem. A country is not a market. You cannot buy “Japan.” You buy a building, on a street, in a ward, in a city. And the chart that matters is the one for that city — not the one for the 38 million people spread across rural prefectures you will never own a square meter in.

When you switch the unit of analysis from country to city, the story inverts. Greater Tokyo holds roughly 37 million people — about 30% of Japan’s entire population — packed into one metro area (directional, 2025/26). That is not a country dispersing. That is a country concentrating. The national decline and the Tokyo boom are the same event seen from two ends: people are not vanishing, they are moving, and they are moving toward you if you own in the center.

Honest caveat: this thesis is about Tokyo and a handful of other magnet cities. It is emphatically not a defense of Japanese property in general. Most of the map is genuinely emptying.

From the desk — The conversations I have over and over start with a buyer quoting the national-decline chart at me, and they always go quiet when I pull the ward-level inflow and occupancy numbers for the building they are standing in. In years of leasing central units, the real bottleneck I run into is finding inventory, not finding tenants — that 96-handle occupancy is not a brochure line, it is what makes my phone ring before a unit even lists.

Japan is two countries: a magnet and a drain

Look at where people actually went last year. In 2025, Tokyo’s 23 special wards logged a net inflow of around 39,000 people, and the broader Tokyo area pulled in roughly 65,000 (as of writing). Six prefectures gained population — Tokyo plus Saitama, Chiba, Kanagawa, Osaka, and Fukuoka, which are the cores of the country’s biggest urban economies.

The other 40 prefectures lost people.

Sit with that split. It means the “Japan is shrinking” chart is an average of two opposite trends crushed into one line — a small number of cities inhaling population and a large number of regions exhaling it. Averages hide everything that matters. The national number tells you Japan is shrinking the same way the average depth of a river tells you it is safe to walk across.

For a buyer, the takeaway is brutally simple: domestic migration is doing your demand forecasting for you. Japanese people, with full information and no language barrier, are voting with their moving trucks. They are leaving Wakayama and Tochigi and Akita, and they are crowding into the wards. You want to own where the trucks are arriving, not where they are leaving.

If you want to see the magnet up close, our ward-by-ward breakdown shows which parts of central Tokyo are absorbing the most people and capital.

The akiya horror stories are a countryside story

The single most effective scare used against Tokyo buyers is the akiya — the “abandoned house.” The number is real and it is large: Japan now has roughly 9.0 million vacant homes, about 13.8% of all housing stock (directional). Foreigners read that and picture empty buildings rotting in central Tokyo.

That is not where the akiya are. They cluster in rural prefectures — places like Wakayama and Tokushima run vacancy rates around 21%. These are inherited family homes in towns with no jobs, no young people, and no buyers, sometimes listed for the price of a used car because nobody wants them at any price. They are a symptom of the drain, not the magnet. The akiya crisis and the Tokyo squeeze are, again, the same migration viewed from opposite ends.

“But Tokyo’s own vacancy is around 10–11%,” someone will tell you. True, and also misleading. Much of that headline figure is unrentable old or inherited stock and normal frictional turnover — units between tenants, units waiting on probate, units in buildings nobody would rent in their current state. It is not a pile of empty modern apartments in prime wards going begging. Strip out the dead stock and the buyable, liquid segment — the modern units in central wards that you would actually purchase and rent — is far tighter than the citywide number implies.

The proof is in the occupancy: ward-level rental occupancy runs around 96.6% (directional). You do not get a 96-handle occupancy rate in a market drowning in empty homes. Tight occupancy is exactly what underwrites your rent roll and your resale liquidity — it means a tenant when you need one and a buyer when you want out.

Falling supply into rising demand

Now add the supply side, because this is where the squeeze gets serious.

New condominium supply in Tokyo is forecast to fall toward 50-year lows around 2026, even as population keeps concentrating into the same wards (directional). Read that twice. Construction is constrained by land scarcity, building costs, and labor shortages, while demand is structurally rising. Falling supply meeting rising urban demand is not the setup for a glut. It is the setup for a price floor that keeps lifting.

The market is already confirming it. Tokyo 23-ward resale condo prices, on the standard 70-square-meter basis, crossed roughly 100 million yen for the first time in 2025 — up about a third year-on-year (directional). Prices are not waiting for the thesis to be proven; they are proving it in real time.

Honest caveat: a third in a single year is hot, and hot moves can pause or give back. Currency, interest rates, and global appetite all swing. This is a long-run structural argument, not a promise that 2026 prints another 30%. Buy the concentration trend, not the last twelve months of momentum.

What this means for you, and how to move

Strip it all back and the decision is cleaner than the headlines make it feel. The “melting ice cube” framing applies to rural Japan — to the akiya towns the migration data shows people fleeing. It does not describe the central wards, where population sets records, occupancy sits near 96%, new supply is shrinking toward generational lows, and prices just broke 100 million yen. National decline and Tokyo concentration are the same force. You get to choose which side of it you own.

So move like someone who understands the split:

• Buy the magnet, not the country. Stay inside the wards and the gaining prefectures the migration data points to. Skip the cheap rural “bargains” — that low price is the market telling you the truth.
• Underwrite on the liquid segment. Ignore the scary citywide vacancy figure. What matters is occupancy and turnover for modern units in the specific ward you are buying. Near-96% occupancy is your margin of safety.
• Compare wards before you fall for a building. Demand is not uniform across the 23. Use our ward profiles and the comparison tools to see where inflow, occupancy, and yield actually line up.
• Learn the vocabulary before you sign. Reikin (non-refundable “key money” paid to a landlord), minpaku (licensed short-term rental), and the rest of the closing terms are in our glossary — knowing them is how you avoid overpaying at the table.

The “Japan is shrinking” chart will keep scaring off buyers who never zoom in past the country level. That is your edge. While they argue with a national average, you can underwrite the city the whole country is moving into — and act before the next record print makes it obvious to everyone else. Pick two or three target wards this week, pull the occupancy and price data, and start building a shortlist. The migration is not waiting for you to decide.