INSIDER TAKE
Japan's Two-Speed Country: Why Tokyo Booms While the Map Empties
Japan isn't uniformly shrinking — it's sorting. Depopulation is funneling people, jobs, and capital into Tokyo while rural towns hollow out.
On this page 7
TL;DR: Japan is losing people, but the losses aren’t spread evenly. They’re pooling in rural towns while Tokyo and a handful of other metros absorb the survivors. Depopulation is acting as a sorting machine — and if you’re looking at Japanese real estate without understanding which side of the sort you’re on, you’re flying blind.
The Headline Everybody Gets Half Right
Open any newspaper in 2024 and you’ll find a story about Japan’s population crisis. The birth rate is near record lows. The total population has been falling since roughly 2008. There are somewhere around nine million empty houses — akiya — scattered across the country. Ghost towns are real, not metaphorical.
All of that is true. None of it tells you what’s actually happening to Tokyo.
Greater Tokyo — the prefecture plus the three surrounding ones — is home to around 36 million people. That is more than the entire population of Canada, packed into an area smaller than Connecticut. And while Japan as a whole is contracting, this mega-region has been remarkably sticky. Population in Tokyo prefecture itself was still growing in the early 2020s, sustained partly by internal migration from the countryside and partly by a slow but steady opening to foreign residents.
The mental model most people carry — “Japan is shrinking, therefore Japanese property is a melting ice cube” — is simply wrong. The correct model is messier and more interesting: Japan is sorting.
What a Sorting Machine Actually Does
Think about what happens when a country stops growing. Young people still need jobs. Jobs concentrate where other jobs are. So people leave Akita, leave Shimane, leave the fishing villages of Kochi, and move to where the work is. Tokyo. Osaka. Nagoya. Fukuoka, maybe.
The places they leave don’t gently deflate like a slow tire. They collapse asymmetrically. Shops close. The tax base shrinks. The school closes, which means families with children stop moving there, which shrinks the tax base further. The local hospital reduces hours. More people leave. This is a feedback loop, not a gentle trend.
Meanwhile, the receiving cities get denser, more expensive, and more economically productive. Tokyo’s land prices in central wards have roughly doubled over the past decade in yen terms. In dollar or euro terms, the gains look even wilder, depending on when you converted.
The sorting machine takes the same country and produces two completely different real estate markets inside it. One is compounding. The other is evaporating. They happen to share a flag.
Akiya: The Visible Symptom of the Wrong Side
The akiya phenomenon — nine-million-plus vacant homes — is the most photogenic evidence of the sorting dynamic. These aren’t abandoned overnight. The typical akiya story: elderly owner dies, adult children long since moved to Tokyo, no buyer in a town with no jobs, property sits and slowly returns to the ground.
Some of these properties are being listed for one million yen or less. Sometimes one yen. This generates enormous Western media coverage — “Buy a house in Japan for the price of a used car!” — and enormous confusion about what is actually on offer.
What’s on offer, in most cases, is a wooden structure in a municipality that is administratively insolvent or close to it, in a region where land values are negative in the sense that demolition costs exceed what anyone will pay. The municipalities want these gone. They’ve created akiya banks — registries of vacant properties — precisely because they’re desperate for anyone to do something with them.
The akiya market is real. For a certain type of buyer — someone who wants to restore a traditional farmhouse, live cheaply in the Japanese countryside, pursue a lifestyle project — it can be genuinely wonderful. But as an investment thesis, buying cheap-and-emptying assets in structurally declining regions is not contrarian cleverness. It’s fighting the current. The current is strong.
Why Concentration Accelerates (And Doesn’t Reverse)
Here’s the part that surprises people: this process tends to speed up rather than stabilize.
Japan’s rural depopulation has been discussed for forty years. Policy has tried to reverse it — regional revitalization programs, subsidies, remote-work incentives pushed hard after Covid. The results have been modest at best. A handful of towns in Okayama and Tokushima attracted some telecommuters. The overall trend didn’t budge.
Why? Because the agglomeration effects that make Tokyo valuable are self-reinforcing. Employers want to be where the talent pool is. Workers want to be where the employers are. Universities, hospitals, entertainment, social networks — they all densify in the center. Each person who moves to Tokyo makes Tokyo slightly more attractive to the next person and slightly less attractive to stay where they came from.
There’s also a generational math problem. Rural Japan is disproportionately old. A town that is 40% over 65 today will look very different in twenty years. The young people left a generation ago. The replacement simply isn’t coming.
Japan’s government knows this. The phrase used in policy circles is “compact cities” — explicitly accepting that some towns will not survive and concentrating services into viable nodes. That’s not a rescue plan for the countryside. That’s an orderly retreat.
What This Means for Where Value Compounds
Tokyo’s property market has several features that look strange until the sorting framework makes them sensible.
Vacancy rates in central Tokyo wards are among the lowest in any major global city — often under 3% for well-located apartments. Rents have been rising. The city is still building, but construction costs have risen sharply (labor shortage plus materials inflation), which puts a floor under new-build prices and pulls up resale values behind them.
Foreign buyers are increasingly present — particularly buyers from other parts of Asia who view Tokyo as a stable, relatively cheap (compared to Hong Kong or Singapore or Sydney) global city. The yen’s weakness through 2022-2024 supercharged this. Even with some yen recovery, Tokyo residential property in USD terms is still meaningfully cheaper than it was fifteen years ago in many segments.
None of this means Tokyo is without risk. It doesn’t. The risks are real and worth understanding (earthquake exposure, lease structure peculiarities, the management fee trap in older condos). But the directional case — that a shrinking Japan is still producing a strengthening Tokyo — is not a paradox. It’s the sorting machine at work.
The trap for outside observers is applying the national narrative to a local market. Japan is shrinking. Tokyo is not (yet, meaningfully). These are both true. Treating them as the same fact is the error.
Where This Could Be Wrong
I believe the sorting thesis is correct directionally. But let me tell you where it could break.
Remote work changes the math. If Japan’s corporate culture genuinely liberalizes around remote work — and it moves slowly, but it does move — the agglomeration advantage weakens. Young workers could choose Kyoto over Tokyo without career penalty. This hasn’t happened at scale yet. It could.
Tokyo’s aging is coming. Tokyo is younger than rural Japan, but its population will age too. The question is whether immigration policy evolves fast enough to replenish the workforce. Japan has been loosening immigration rules incrementally, but “incrementally” is the operative word.
Policy can distort. Japan’s government has tools. BOJ rate policy affects mortgage rates. Fiscal stimulus flows where politicians direct it. The sorting trend is structural, but it’s not immune to distortion.
A severe earthquake changes everything. Tokyo is overdue by historical averages. This is a real tail risk, not a theoretical one. Every serious Tokyo investor should understand it and decide how much exposure they’re comfortable with.
FAQ
Is all of rural Japan a bad investment? No. Kyoto is rural-adjacent and fine. Well-connected second-tier cities — Fukuoka, Sapporo, parts of Osaka — have their own agglomeration dynamics. The danger zone is small towns in prefectures with accelerating population loss and no economic anchor. Know which category you’re in before you buy.
How do I tell a sorting “winner” from a “loser” location? Population trend over ten years, employment base, age distribution, and distance from a Shinkansen or major rail node. A town losing 2% of its population per year compounding is in a different universe from a central Tokyo ward.
Are akiya worth buying at all? For lifestyle purposes, sometimes yes. As a primary investment with return expectations, I’m skeptical outside of very specific situations — tourist areas, near ski resorts, towns with demonstrated revitalization traction. The cheap price is usually cheap for a reason.
What’s the practical takeaway for a foreign investor? Focus on central Tokyo wards (specifically within 30 minutes of a major station), understand the leasehold vs freehold distinction, and price in management costs honestly. The thesis works if you’re in the right location. The same capital in the wrong prefecture is a very different story.
Why hasn’t the Tokyo market crashed if Japan is in decline? Because national decline and urban concentration can coexist. Detroit cratered while New York didn’t, during the same decades of American industrial restructuring. The unit of analysis matters. Japan-as-a-whole is the wrong unit for evaluating Tokyo real estate.