Japan vs. Korea Long-Term Care Insurance: Opposite Spending

When a country starts designing an elderly care system, it usually looks at Japan. Japan built public long-term care insurance in 2000, and it still works. Korea copied much of the design and launched its own version in 2008.

Then the two systems diverged in a way nobody planned. Researchers comparing them found that adding care institutions pushed spending per beneficiary down in Japan and up in Korea.¹ Same policy instrument, opposite result.

What the comparison actually found

The study was published in the International Journal of Health Policy and Management in November 2022. Hongsoo Kim, Nan-He Yoon, Youjin Hahn and Hideki Hashimoto compared regional administrative data from both countries.¹

Their sample covered 43 Japanese prefectures and 16 Korean provinces over three years, from 2013 to 2015. That produced 177 regional observations.¹

Two headline differences emerged. In Japan, 18.2% of people aged 65 and over used long-term care services. In Korea, 6.8% did.¹

Japan also spent more on each person it covered, roughly 1.4 times the Korean figure per recipient.¹

The finding that matters is the direction of the relationship. Institutional supply was negatively related to expenditure per beneficiary in Japan and positively related in Korea.¹ More supply meant cheaper care per person in one country and costlier care per person in the other.

The researchers were careful about what they could explain. Observable differences accounted for 89.4% of the gap in how many people used care, mostly differences in supply.¹ The spending gap was different. They attributed it largely to country-specific factors they could not observe in the data.¹

That last point is the honest limit of the study. It measured the inversion. It did not identify what caused it.

Where Korea's extra spending goes

Korea did not copy one thing: Japan's decision to put all institutional elderly care inside a single funding stream.

Korea runs two. Nursing homes are welfare facilities funded by long-term care insurance. Long-term care hospitals are physician-staffed medical institutions funded by the National Health Insurance system.⁹ The two serve older people with similar needs, under different rules.

The payment method drove the result. Long-term care hospitals are paid a flat daily rate per patient. Hyuk Ga, writing in the Ewha Medical Journal, describes a model that "guaranteed stable and predictable revenue regardless of patient acuity or service intensity".⁹ These hospitals then "proliferated at an extraordinary pace".

The supply figures show how far that went. Korea had about 60 long-term care beds per 1,000 older people, against an OECD average of 46.⁵ Of those beds, 59% sat in long-term care hospitals rather than care homes.

Length of stay followed. The average long-term care hospital stay ran to 165 days in 2019.⁵ A more recent national study of people who died between 2014 and 2023 found a median cumulative stay of 79 days, with 22.2% staying a year or longer.¹⁰

Two independent sources agree that Korean spending outran Korean caseloads. The Korea Economic Institute of America found long-term care spending growing 15.2% a year between 2009 and 2018, while recipient numbers grew 9.3% a year.⁵ National Health Insurance Service figures show benefit payments rising from 7.73 trillion won in 2019 to 14.76 trillion won in 2024, close to a doubling in five years.⁸

What Japan has that Korea did not copy

Japan built a gatekeeper into the system. Every user of Japanese long-term care insurance reaches benefits through a care manager, who assesses need and assembles the package.¹¹

That role does two things. It steers people toward home and community services where those will do, and it puts one person in charge of a plan.

Ga's review is direct about the contrast. Japan's system, "centrally organized around the care manager role, has achieved measurable success in reducing avoidable hospitalizations and supporting aging in place".⁹ Korea, by comparison, "lacked effective structures for integration".⁹

That supplies a plausible reading of the inversion. In Japan, building more capacity gives a care manager cheaper options to route people into, so average spending per person falls. In Korea, building more capacity adds beds that fill themselves, so average spending per person rises.

Plausible is not proven. The study behind this article did not test that mechanism, and said so.¹

Japan is not a finished model either

Treating Japan as the answer has its own problem. Japan's system is holding, but it is under visible strain.

Bankruptcies among Japanese medical and nursing care providers hit 478 in the year to March 2026, the highest count since 1988, according to the credit research agency Tokyo Shoko Research.¹² Small operators with four or fewer employees made up 72.1% of them.

Staffing is tighter still. Japan's health ministry expects demand for care workers to reach 2.72 million by 2040. The workforce stood at about 2.15 million in fiscal 2022, which leaves a projected shortfall of 570,000 people.¹³

So the care manager model that holds Japanese costs down depends on there being enough workers to staff the services a care manager routes people into. Korea is adopting that model at a point when its originator cannot fully staff it.

What has changed since the data was collected

The study's window closed in 2015. Both systems have moved since, and mostly in ways that narrow the gap.

Korea's coverage has climbed steeply. In 2024, 10.8% of Koreans aged 65 and over were certified for long-term care, and 31.9% of those aged 80 and over.⁴ That is up from 6.8% during the study period.¹

Japan barely moved over the same stretch. Its certification rate went from 18.2% in the study period to 19.4% in 2024, when 7.1 million of 35.9 million insured people aged 65 and over held a certification.³

Korea also crossed a demographic line. The share of Koreans aged 65 and over passed 20% in December 2024 and reached 22.1% by August 2026.⁴ Korea made the journey from an ageing society to a super-aged one in about 24 years, against 35 years for Japan.⁴

Then there is the policy response. Korea passed the Integrated Community Care Support Act in 2024, with full implementation set for 2026. It requires local government coordination and introduces a certified care manager workforce.⁹ Korea is, in other words, now copying the part of Japan's system it left out the first time.

Earlier community care pilots between 2019 and 2022 reported reductions in unnecessary hospitalisations and emergency department visits.⁹

What the headline numbers leave out

The core data is a decade old. The comparison rests on 2013 to 2015 administrative records.¹ Anyone citing the inversion as a current fact is citing a historical measurement.

The sample is small for a country comparison. Fifty-nine regional units over three years gives 177 observations.¹ That is enough for the regional analysis the authors ran, not enough to settle a national causal question.

Correlation is not the mechanism. The authors explained 89.4% of the utilisation gap with observable factors but attributed the spending gap to unobserved country-specific ones.¹ The dual-funding explanation in this article comes from separate literature on Korea's system, not from the comparison itself.⁹

The two countries were never actually identical. Korea's premium rate is roughly 0.92% of income against about 1.5% in Japan.⁵⁶ A scheme collecting a little over half as much per worker is not the same scheme.

Korea's problem is not too few beds. It has more long-term care beds per older person than the OECD average.⁵ The Korea Economic Institute also reports that about 30% of older people in long-term care institutions need hospital care and do not get it.⁵ Mismatch runs in both directions.

Money is not the only thing at stake. Among Korean care recipients, 67.5% said they would prefer to die at home, while 59.6% of family caregivers preferred hospitals.⁹ A Korean national study found 68.5% of people who used long-term care hospitals and then died did so during a hospital stay.¹⁰

Old-age poverty shapes what families choose. In Korea, 39.7% of people aged 66 and over live on less than half the national median income, according to OECD data.⁴ Cheap institutional care is not a neutral option when the alternative is unaffordable.

How to use a cross-country care comparison

Check what the copy left out before crediting the design. Korea adopted Japan's insurance scheme and not its care manager gatekeeping. The second turned out to matter for cost, which is not visible in a description of the first.

Look at where the beds are funded, not how many there are. Korea's bed supply is above the OECD average. Most of that supply sits in a payment system that rewards keeping people longer.⁵⁹ A bed count alone would have told you the opposite of the truth.

Date the finding before quoting it. The inversion is a real published result from 2013 to 2015 data. Korea's certification rate has risen by more than half since then and the coordinating law arrived in 2024.⁴⁹ The finding is evidence about how systems drift, not a current scorecard.

For a country writing its own scheme, that drift is the transferable part. Japan and Korea show that the financing rules around a care system determine its cost more reliably than the care system's own design does.

Footnotes

1. https://pmc.ncbi.nlm.nih.gov/articles/PMC10125049/ — Kim H, Yoon NH, Hahn Y, Hashimoto H, "Explaining Variations in Long-term Care Use and Expenditures Under the Public Long-term Care Insurance Systems: A Case Study Comparison of Korea and Japan", International Journal of Health Policy and Management (published Nov 19, 2022). Accessed Sep 17, 2026. ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹ ↩¹² ↩¹³ ↩¹⁴ ↩¹⁵ ↩¹⁶ ↩¹⁷ ↩¹⁸ ↩¹⁹

2. https://pmc.ncbi.nlm.nih.gov/articles/PMC11467513/ — Ga H, "The Korean Long-Term Care System: 2024 Update", Annals of Geriatric Medicine and Research (published Sep 24, 2024). Accessed Sep 17, 2026. ↩

3. https://webapps.ilo.org/globalcare/south-4-care/24 — International Labour Organization, Global Care Policy Portal, "Long-term care insurance (Kaigo Hoken) in Japan". Accessed Sep 17, 2026. ↩ ↩²

4. https://medicalxpress.com/news/2026-09-south-korea-tackling-superaging-problem.html — Medical Xpress, "How South Korea is tackling its super-aging problem and what other aging nations can learn from its experience" by Jeongmin Park and Alexis R. Santos, republished from The Conversation (published Sep 15, 2026), citing Korean government and OECD data. Accessed Sep 17, 2026. ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶

5. https://keia.org/the-peninsula/improving-koreas-long-term-care-for-the-elderly/ — Korea Economic Institute of America, "Improving Korea's Long-Term Care for the Elderly" by Randall S. Jones (published Mar 2, 2023), citing OECD and Korean Ministry of Health and Welfare data. Accessed Sep 17, 2026. ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹

6. https://view.asiae.co.kr/en/article/2025110417144138401 — Asia Economy Daily, "Long-Term Care Insurance Premiums to Rise Next Year: Average Monthly Increase of 517 Won Per Household" (published Nov 4, 2025), reporting a Ministry of Health and Welfare announcement. Accessed Sep 17, 2026. ↩ ↩²

7. https://www.nippon.com/en/japan-data/h02170/ — Nippon.com, "Expenditures on Nursing Care in Japan Reach ¥11.5 Trillion" (published Nov 1, 2024), citing Ministry of Health, Labour and Welfare statistics. Accessed Sep 17, 2026. ↩

8. https://www.koreaherald.com/article/10424725 — The Korea Herald, "Long-term care spending for seniors doubles in last 5 years" by Yoon Min-sik (published Feb 20, 2025), citing National Health Insurance Service data. Accessed Sep 17, 2026. ↩ ↩²

9. https://pmc.ncbi.nlm.nih.gov/articles/PMC12611424/ — Ga H, "Shifting from a provider-centered to a person-centered model of long-term care for older patients in Korea: a narrative review", Ewha Medical Journal (published Oct 10, 2025). Accessed Sep 17, 2026. ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰

10. https://pmc.ncbi.nlm.nih.gov/articles/PMC13399523/ — Kim H, Shin S, "Long-term care hospitals as end-of-life care settings in South Korea: A nationwide analysis of utilization patterns and clinical trajectories", PLoS One (published Jul 24, 2026). Accessed Sep 17, 2026. ↩ ↩²

11. https://japanhpn.org/en/longtermcare/ — Health and Global Policy Institute, "Japan Health Policy NOW: Long-term Care Insurance". Accessed Sep 17, 2026. ↩

12. https://medicalbuyer.co.in/japan-records-478-bankruptcies-among-medical-nursing-care-providers/ — Medical Buyer, "Japan records 478 bankruptcies among medical, nursing care providers" (published May 2026), citing Tokyo Shoko Research data released May 4, 2026. Accessed Sep 17, 2026. ↩

13. https://isvd.or.jp/en/columns/2026-03-09-care-worker-shortage-structure — Institute for Social Vision Design, "The Structure of Japan's Care Worker Crisis: The 'Invisible Roadmap' to 2040" by Naoya Yokota (published Mar 9, 2026), citing the Ministry of Health, Labour and Welfare's "Required Number of Care Workers Based on the 9th Long-Term Care Insurance Business Plan" (2024). Accessed Sep 17, 2026. ↩

Show all 14 sources