91 Home-Visit Care Failures Against 45 in Day Services, Down a Fifth — In the Same Fee Revision, One Set of Units Fell and the Other Rose

According to Tokyo Shoko Research, 176 long-term care providers failed in 2025[1], a second consecutive record.

The breakdown does not move together. Home-visit care accounted for 91, up 12.3% from 81 the year before[1] — the highest for a third straight year. Yet day services and short stays came to 45, down 19.6% from 56[1]. Paid nursing homes also fell, from 18 to 16.

In one year, under one insurance system, visiting rose and gathering fell.

The size of the failed providers is skewed as well. 142 (80.6%) had fewer than ten employees, 128 (72.7%) had capital under five million yen, and 141 (80.1%) had debts under 100 million yen[1]. Weak revenue was the cause in 140 cases, 79.5% of the total[1], and labour shortage was cited in 29, up 45.0% on the year.

The point of divergence sits in April 2024, when the fiscal 2024 fee revision reset the units for each service.

Open the before-and-after document the ministry put to the subcommittee, and the home-visit column reads as follows. For care centred on personal assistance, under 20 minutes went from 167 to 163 units, 20 to 30 minutes from 250 to 244, and 30 to 60 minutes from 396 to 387[2]. For care centred on living assistance, 20 to 45 minutes went from 183 to 179 units and 45 minutes or more from 225 to 220[2]. Cuts of a little over two percent throughout.

The day service column in the same document faces the other way. For standard-scale service of three to four hours, care level 1 went from 368 to 370 units, care level 3 from 477 to 479, and care level 5 from 585 to 588[2].

One revision, decided on one day, pointing in two directions. That failure counts diverged over the following two years is not unrelated.

Long-term care fees are set in units. Units are set by service type and duration. So far, visiting and gathering have the same shape.

What differs is the cost that sits outside that duration.

Home-visit care goes house to house. Travel from the first home to the second, finding somewhere to park, climbing the stairs. None of it counts toward the duration. Payment measures from entering the door to finishing the care; getting there is carried by the provider, as is the fuel. Suppose a 30-minute visit takes 40 minutes of driving in total: the provider is tied up for 70 minutes and paid for 30. No statistic on travel time appears in the sources cited here; the relationship follows from how the units are defined.

Day services bring people to one place. There is transport, but one vehicle carries several people, and staff are already on site, so travel per person is small against the time in service. The same two percent move in units lands on a different base.

That difference shows up in the size breakdown of the failures. That 80.6% had fewer than ten employees means what fell were the providers unable to absorb travel inefficiency through headcount. With twenty home helpers you can divide a district and assign it; with five you cannot.

The causes point the same way. Weak revenue leads at 79.5%, but weak revenue in home-visit care does not mean customers stopped coming. The ceiling on visits per day arrives first. Labour shortage rising 45.0% as a stated cause follows from that ceiling being set by headcount.

Seen from the policy side, this is a problem of measurement. Payment is measured by the time care was delivered; cost accrues over the time the worker is held, travel included. The gap between the two varies by the shape of the business. Applying one uniform rate of reduction to a business with a wide gap does not produce a uniform effect.

The same shape appeared in freight. Fair cost in trucking（このサイトの記事） examined the problem of waiting and loading time not being carried in the freight rate. Who bears travel and waiting is decided by where the unit of pricing is placed.

That failures concentrate among small providers is not only a matter of how many firms remain. Home-visit care exists only if someone arrives at the door. In districts where providers disappear, what follows is what the prefectural gap in public assistance take-up（このサイトの記事） described: the system exists, and the route by which it arrives does not. The method for counting this non-take-up is set out in policy exclusion and non-take-up（このサイトの記事）. A provider closing lowers the rate at which a benefit arrives without changing the benefit at all.

Further Reading

• 『介護格差』 (Care Inequality)（外部サイト、新しいタブで開きます） (Yasuhiro Yuki, Iwanami Shoten). An account, drawn from practice, of what separates those who can obtain care services from those who cannot. Useful for thinking through what happens in districts where providers close.

References

Care Provider Failures in 2025 Reach a Record 176 — Tokyo Shoko Research (2026). TSR Data Insight

Fiscal 2024 Long-Term Care Fee Revision, Draft Fee Schedule (Appendix 1-1, Standards for Calculating Costs of Designated In-Home Services) — Ministry of Health, Labour and Welfare, Health and Welfare Bureau for the Elderly (2024). Social Security Council, Long-Term Care Benefit Subcommittee, 239th Session

Principal Items in the Fiscal 2024 Long-Term Care Fee Revision — Ministry of Health, Labour and Welfare, Health and Welfare Bureau for the Elderly (2024). Social Security Council, Long-Term Care Benefit Subcommittee, 239th Session, Document 1

Report on the Fiscal 2024 Long-Term Care Fee Revision — Ministry of Health, Labour and Welfare (2023). Ministry of Health, Labour and Welfare