Japan really did use pension-related and employment-insurance funds for resort and welfare facilities

Where this applies: This article explains how things work in Japan. Rules, amounts and procedures may be different where you live. Amounts are in Japanese yen (JPY).

How reading tools work

Listen reads the article aloud. Speed read shows phrases in sequence at your chosen pace. Language practice compares available translations. Save keeps a bookmark in this browser; find it in the player’s bookmarks.

Share this article
Advertisement
Advertisement

The story is not an internet myth. Japan once used pension-related and employment-insurance resources to build and support large numbers of recreation, lodging, sports and worker-welfare facilities.

The best-known example is Greenpia, a network of 13 large pension recreation complexes. By 2005, all 13 had been transferred. In a 2007 Diet session, the health minister stated that the construction cost was ¥195.3 billion, while total transfer revenue was ¥4.8 billion.[1]

Those two numbers look like someone walked into a resale shop carrying the world's largest receipt.

But subtracting ¥4.8 billion from ¥195.3 billion and calling the result the exact loss would also be wrong. The facilities operated for years, buildings depreciated, and some transfers for continued public use were intentionally discounted from appraised market value.

The real policy question is simpler: why was money tied to social insurance used to own and operate resort-like assets in the first place?

1. What Greenpia was

Greenpia was created by the former Pension Welfare Service Public Corporation for pensioners and insured workers. Thirteen large complexes were built across Japan, often combining lodging, sports, leisure and health-recreation functions.[2]

As private alternatives expanded and social demand changed, the model became harder to justify. A 2001 government rationalization plan called for Greenpia to be abolished by fiscal 2005, with facilities unable to cover even operating costs from their own revenue to be closed earlier.[2]

The final transfer, Greenpia Miki, was completed in December 2005.[1][3]

So “almost all went bankrupt” is not the precise description. The more accurate statement is that the government withdrew from the policy and the facilities were transferred or closed. Bankruptcy and policy exit are different things.

2. Employment insurance had its own large welfare-facility program

Pension funds were not the only source.

Japan's employment-insurance system once included an “employment welfare program” that supported educational, cultural, sports, recreation and other worker-welfare facilities.

A 2006 Ministry of Health, Labour and Welfare document covering 1,976 facilities scheduled for transfer reported original construction costs of ¥440.7 billion, an appraised value of ¥72.2 billion, and transfer prices totaling ¥12.73 billion.[4]

Again, the headline comparison is dramatic. But the ministry noted that age-related depreciation, facilities not designed to generate profits, leased land, and demolition obligations all reduced appraised values.[4]

The numbers are real. The interpretation still needs accounting context.

3. Is the problem that welfare facilities lose money, or that the wrong wallet paid for them?

A public or welfare service does not always exist to earn a standalone profit. Libraries, emergency shelters and vocational training would make little sense if every unit had to produce a commercial return.

So the strongest principle is not “every welfare facility must be profitable.”

It is: money should remain tied to the purpose for which it was collected.

Pension reserves exist to support future pension benefits. Employment insurance exists to address unemployment, reemployment, employment stability and skills. If the government wants to build a resort, promote a region or run a broader welfare program, using a separately authorized budget makes the cost and responsibility easier to see.

Otherwise social insurance becomes a universal wallet.

A hot spring here. A gym there. Regional revitalization next.

Do not give the wallet a side quest.

4. The system was pulled back sharply in the 2000s

On the employment-insurance side, new welfare-facility construction was already being stopped around the end of the 1990s, and existing facilities were disposed of. In 2007, the legal category of the employment welfare program itself was abolished.[5][6]

During Diet deliberations, the health minister explicitly stated that the remaining Employment Stability Program and Human Resources Development Program could not be used to resume the old-style worker welfare facilities.[5]

That did not eliminate facilities directly connected to employment policy. Vocational training facilities and support for education and training remain within the current Human Resources Development Program.[7]

A resort complex and a vocational training center are both “facilities,” but their distance from the purpose of employment insurance is very different.

5. Today's pension reserves cannot be invested for unrelated policy goals

Current Japanese law requires pension reserves to be managed solely for the benefit of insured persons, from a long-term perspective, safely and efficiently.[8]

The ministry further explains that investment cannot be driven by other government policy goals or policy measures — a restriction often described as prohibiting “consideration of unrelated matters.”[8]

That matters.

“Regional development would benefit.” “It would create jobs.” “People would like the facility.”

Those arguments alone are not valid reasons to deploy pension reserves.

The pension wallet now has a much higher fence around it.

6. GPIF is not a giant NISA, but the philosophy is much more boring

Since April 2025, the Government Pension Investment Fund's basic portfolio has been:

  • 25% domestic bonds
  • 25% foreign bonds
  • 25% domestic equities
  • 25% foreign equities.[9]

As of the end of June 2026, GPIF managed ¥317.7596 trillion. Cumulative investment income since market-based management began in fiscal 2001 was ¥221.0201 trillion.[10]

GPIF is not NISA. NISA is a tax-advantaged framework for individuals; GPIF is an institutional investor managing public pension reserves. It also uses alternative assets and active strategies.

But the philosophy has clearly shifted away from “own facilities” toward diversify, set rules, hold for the long term, and review the portfolio.

Modern pension management requires less hotel-management talent and more deliberately boring asset allocation.

7. “Never lose money” is impossible; “do not gamble outside the mandate” can be enforced

Market investing can produce losses over a quarter or a year. If an investor accepts equity and bond risk to earn long-run returns, a rule requiring zero negative periods is unrealistic.

So a negative return does not automatically mean failed management.

What can be controlled are things such as:

  • using pension money for unrelated policy objectives,
  • taking risks that cannot be explained by the mandate,
  • hiding costs or outcomes,
  • weakening accountability.

You cannot prevent markets from falling. You can prevent the pension system from suddenly deciding it wants to become a resort operator.

8. “The pension system ran out of money because of waste” is also wrong

Criticism of historical facility programs should be separated from the current finances of the pension system.

Japan's public pension system is financed mainly from current contributions and other public financing, while reserves provide part of long-term funding and stabilization.[8]

So even if Greenpia and similar programs deserve criticism, it is inaccurate to say that they caused today's pension reserves to be exhausted.

As of June 2026, GPIF still managed more than ¥300 trillion, with strongly positive cumulative investment income since 2001.[10]

Demography also still matters. The financial balance of a largely pay-as-you-go pension system depends not only on investment returns but also on the number of contributors relative to beneficiaries.

Historical waste and population aging are separate issues. One does not erase the other.

9. Employment insurance today works differently from pension investment

Employment insurance is not just an investment pool.

Japan still has two employment-insurance programs: one for employment stability and one for human-resources development. According to the ministry, the cost of these two programs is financed entirely by premiums paid by employers, not by the employee portion used for unemployment-related benefits.[7]

Their statutory purposes are also more directly tied to employment: preventing unemployment, expanding job opportunities, vocational training and education support.

The practical rule has moved closer to: if employment-insurance money is used, show the connection to employment.

10. What the “make politicians pay for failures” reaction is really about

When a large public project goes badly, people often object to the asymmetry: decision-makers may not personally bear the financial loss while contributors do.

Cutting legislators' pay, however, would not automatically reimburse pension reserves. Political responsibility and accounting restitution are separate mechanisms.

For prevention, more direct safeguards are:

  • legally narrow the permitted purpose,
  • prohibit unrelated policy considerations,
  • disclose objectives and risks before spending,
  • publish costs and performance,
  • require independent audit and ex-post evaluation,
  • create an exit rule before a program becomes permanent.

Punishment after the fact is one tool. A stronger tool is a checkout system that says, “this wallet cannot buy that.”

11. Conclusion: this is not an argument against welfare; it is an argument for labeled wallets

If a welfare facility is genuinely needed, fund it as welfare policy. If regional development is needed, fund it as regional policy. If vocational training is needed, show why it serves employment insurance. And if pension reserves are being managed, manage them for pensions.

“It helps the public” is not a magic phrase that erases the label on the wallet.

Japan's reforms in the 2000s did not make public finance perfect. But they did substantially strengthen this boundary.

The old social-insurance wallet somehow picked up a side job running resorts. The modern version has much stricter parental controls.


References (10)

  1. 衆議院 第166回国会 厚生労働委員会 第20号(2007-05-16)。グリーンピア13基地、建設費1,953億円、譲渡収入48億円。 shugiin.go.jp
  2. 厚生労働省 第9回社会保障審議会年金部会資料「年金福祉事業団が行っていた融資・施設事業について」(2002)。Greenpiaの設置目的・13基地・撤退方針。 mhlw.go.jp
  3. 厚生労働省「三木大規模年金保養基地〔グリーンピア三木〕の譲渡について」(2005-12-20)。13基地すべての譲渡完了。 mhlw.go.jp
  4. 厚生労働省「勤労者福祉施設の譲渡価格について」(2006)。1,976施設、当初建設費4,407億円、時価722億円、譲渡価格127.3億円。 mhlw.go.jp
  5. 衆議院 第166回国会 厚生労働委員会 第5号(2007-03-16)。雇用福祉事業廃止、雇用安定・能力開発事業では従来型勤労者福祉施設を行えない旨の答弁。 shugiin.go.jp
  6. 厚生労働省・雇用保険法(雇用福祉事業廃止に伴う経過措置)。 mhlw.go.jp
  7. 厚生労働省「雇用保険二事業の概要」。雇用安定事業、能力開発事業、二事業費用は事業主負担保険料で賄う。 mhlw.go.jp
  8. 厚生労働省「年金制度の仕組みと考え方 第14 年金資金運用」。専ら被保険者の利益、安全かつ効率的な長期運用、他事考慮の禁止。 mhlw.go.jp
  9. 年金積立金管理運用独立行政法人「基本ポートフォリオの考え方」。2025年4月以降、国内債券・外国債券・国内株式・外国株式を各25%。 gpif.go.jp
  10. 年金積立金管理運用独立行政法人「2026年度の運用状況」。2026年度第1四半期末の運用資産額317兆7,596億円、2001年度以降の累積収益額221兆201億円。 gpif.go.jp

AdBooks on this topic

  • Books on insurance

    Book search results for insurance, the topic of this article.

This article contains affiliate links (ads). About advertising As an Amazon Associate I earn from qualifying purchases.

Read this today

Each one answers a question readers of this article tend to ask next.

Browse all articlesMore on Public services and programs

Advertisement

Find other articles

All articles

Mendoi-chan

Who runs this site

Mendoi-chan

She turns friction at work and in everyday life into clear structure and practical next steps.