Read Rental Demand by Households, Not Population: From 2030 to 2050 Tokyo +3%, Osaka ▲11%, Akita ▲26% — Even a 9% Yield Loses About 40% of NOI in 20 Years (Estimate)
Using IPSS's 2024 household projections by prefecture, this article compares 2030–2050 changes: Tokyo +2.8%, Osaka ▲10.6%, Akita ▲26.3%. If occupancy tracks households, a 9% gross-yield property in Akita loses about 36% of NOI in 20 years, narrowing the gap with Tokyo.

Tenants rent per household, not per person. Where households keep growing, rental demand does not shrink immediately even if population has started to fall; where households decline, even a high gross yield erodes long-term NOI (net operating income). The Institute of Population and Social Security Research (IPSS) 2024 projection shows how far apart regions will be between 2030 and 2050: Tokyo +2.8%, Osaka ▲10.6%, Akita ▲26.3%.
For investment properties financed over 20 to 35 years, that gap affects both exit price and debt-service capacity. This article uses IPSS's prefectural projections to compare regional household outlooks and estimates the change in NOI if occupancy tracks household numbers. Projections are not certainties; different assumptions give different results.
National Households Peak in 2030
According to IPSS, private households nationwide rise from 55.705 million in 2020 to 57.732 million in 2030, then fall to 56.080 million in 2040 and 52.607 million in 2050. That is ▲8.9% from 2030 to 2050. Average household size falls from 2.21 in 2020 to 1.92 in 2050; smaller households are what keep the count up.
| National | 2020 | 2030 | 2040 | 2050 |
|---|---|---|---|---|
| Private households (10,000) | 5,570.5 | 5,773.2 | 5,608.0 | 5,260.7 |
| Average household size | 2.21 | 2.02 | 1.94 | 1.92 |
In other words, rental demand stays supported for a few more years, but from the 2030s total demand shrinks nationwide. A property bought in 2030 and held for 20 years spends its second half in a headwind even at the national average.
Households by Prefecture: The 2030–2050 Gap
Private households (thousands) in selected prefectures and the change versus 2030 are below (IPSS, 2024 projection, Table 12-47).
| Prefecture | 2020 | 2030 | 2040 | 2050 | 2050 vs 2030 |
|---|---|---|---|---|---|
| Tokyo | 7,217 | 7,710 | 7,955 | 7,928 | +2.8% |
| Okinawa | 613 | 666 | 686 | 670 | +0.6% |
| Kanagawa | 4,210 | 4,442 | 4,428 | 4,274 | ▲3.8% |
| Saitama | 3,158 | 3,381 | 3,349 | 3,226 | ▲4.6% |
| Chiba | 2,768 | 2,977 | 2,954 | 2,848 | ▲4.3% |
| Aichi | 3,233 | 3,453 | 3,430 | 3,289 | ▲4.7% |
| Fukuoka | 2,318 | 2,426 | 2,389 | 2,260 | ▲6.8% |
| Hyogo | 2,399 | 2,467 | 2,368 | 2,192 | ▲11.1% |
| Osaka | 4,127 | 4,215 | 4,041 | 3,767 | ▲10.6% |
| Hokkaido | 2,469 | 2,435 | 2,264 | 2,015 | ▲17.2% |
| Akita | 384 | 369 | 326 | 272 | ▲26.3% |
Only Tokyo and Okinawa stay above their 2030 level in 2050. The three prefectures around Tokyo peak near 2030 and then fall roughly 4–5% by 2050. Osaka and Hyogo peak around 2030 and decline by more than 10% by 2050. Akita's households were already falling from 2020, and by 2050 stand at about three-quarters of the 2030 level.
Estimate: Occupancy Tracking Households
Household change does not translate one-for-one into occupancy, but shrinking markets tend to see more vacancy and stronger downward pressure on rents. For simplicity, assume a whole building bought in 2030 for ¥100 million, 93% occupancy in 2030, operating costs (management, repairs, property tax and so on) fixed at 25% of full-occupancy rent, and 2050 occupancy proportional to the 2050/2030 household ratio (an illustrative assumption; rents held constant).
| Item | Tokyo (5.0% gross) | Osaka (6.5% gross) | Akita (9.0% gross) |
|---|---|---|---|
| Full-occupancy rent (per year) | ¥5.0 million | ¥6.5 million | ¥9.0 million |
| Household ratio (2050 vs 2030) | 1.028 | 0.894 | 0.737 |
| Occupancy (2030 → 2050) | 93% → 95.6% | 93% → 83.1% | 93% → 68.5% |
| NOI in 2030 (per year) | ¥3.40 million | ¥4.42 million | ¥6.12 million |
| NOI in 2050 (per year) | ¥3.53 million | ¥3.78 million | ¥3.92 million |
| NOI change | +3.8% | ▲14.5% | ▲36.0% |
| NOI yield (2030 → 2050) | 3.4% → 3.5% | 4.4% → 3.8% | 6.1% → 3.9% |
Tokyo's gross yield is 5.0% against Akita's 9.0%, a 4-point gap, yet 20 years later the NOI yields are 3.5% and 3.9%, nearly level. The result depends on the assumptions; capping Tokyo occupancy at 97%, for example, would change it. Still, it shows that what a high-yield regional property offers up front is easily clawed back by lower occupancy later.
What Prefectural Averages Hide
Prefectural projections are averages, and demand varies widely across municipalities and station areas. Within Tokyo, some wards gain households while some Tama cities lose them; in regional prefectures, the prefectural capital's center and its outskirts can move in opposite directions. Use the prefectural figure as an entry point, then check:
- IPSS municipal projections (households, share of single-person households): is the target area stronger or weaker than the prefectural average?
- Observed rental demand: time from listing to contract, nearby asking-rent trends, vacancy rates
- Whether tenant types (singles, students, transferees, families) match the groups that are growing
- Dependence on specific demand anchors such as a university, factory or hospital
Working Back from Loan Term and Exit
Because households in many regions peak around 2030, that collides with long loan terms. For an older building bought with a loan of nearly 35 years, the second half of repayment falls in the declining period. Three tests apply:
- Does NOI that allows for household change over the loan term keep the debt service coverage ratio (DSCR) at 1.2–1.3?
- Compared on NOI yield in a downside case with occupancy a few points lower, not on gross yield, does the region still hold up?
- Does the exit price allow for buyers (individual investors, lenders' appraisals) finding it harder to obtain financing in shrinking regions?
Tokyo and Okinawa, where households grow, carry a different risk: yields are already low and purchase prices high. Ultimately, comparing yield, households and price in one table is the starting point for choosing a region.
Frequently Asked Questions
QShould rental demand in Japan be measured by population or by households?
Households are the more practical measure, because units are rented per household. IPSS's 2024 projection has Japan's private households rising from 55.71 million in 2020 to 57.73 million in 2030, then falling to 52.61 million in 2050. Smaller household sizes delay the peak in households relative to population.
QWill the number of households in Tokyo rise or fall by 2050?
In IPSS's 2024 projection, Tokyo's households rise from 7.22 million in 2020 to 7.96 million in 2040 and stand at 7.93 million in 2050. That is +2.8% versus 2030, and Tokyo and Okinawa are the only two of Japan's 47 prefectures with a positive change from 2030.
QWhat happens to household numbers in Osaka and Fukuoka by 2050?
In IPSS's 2024 projection, Osaka's households fall 10.6% from 4.22 million in 2030 to 3.77 million in 2050, and Fukuoka's fall 6.8% from 2.43 million to 2.26 million. Both peak in 2030, and the decline accelerates after 2040.
QWill a high-gross-yield rural property in Japan still be attractive in 20 years?
Not necessarily. In an illustrative model with a ¥100 million price, 25% operating costs and 93% occupancy in 2030, an Akita property at a 9% gross yield sees NOI yield fall from 6.1% in 2030 to 3.9% in 2050 if occupancy tracks households. A Tokyo property at 5% moves from 3.4% to 3.5%, so the two converge.
QHow can investors in Japan use household projections when choosing a property?
Check the change in households over a period matching the loan term, then compare yields using NOI that allows for lower occupancy. Prefectural averages hide municipal differences, so candidate areas should also be checked against IPSS municipal projections and observed data such as vacancy periods and rent trends.
Sources
Talk to us even if you don't know where to start
Buying, selling, financing, property management and inheritance planning. Practitioners will propose an approach tailored to your situation.






Comments
Comment guidelinesMembers can comment on this article. Every comment is checked by AI before it is published.
Sign in to comment