A recurring argument on social media goes like this:
“You retired, the company asked you to stay, and your salary dropped? Come on. That is just politeness. They really want you to read the room and leave.”
That interpretation would make sense in a world where firms had spare workers, young recruits appeared on demand, and thirty years of company-specific know-how automatically uploaded itself to the cloud on retirement day.
That is not the world Japanese employers are operating in.
In July 2026, 51.3% of surveyed companies told Teikoku Databank that they were short of regular employees. That is not “nearly every company,” but it is a national majority.[1]
Manufacturing makes the problem even clearer. A 2026 JILPT survey found that 91.1% of manufacturers placed importance on skill succession. The most common measure—54.8%—was keeping older employees working through reemployment or extended service.[2]
So sometimes the problem is not “we secretly want this person gone.”
It is:
“If this person leaves, a private database walks out of the building with them.”
1. A veteran is not just one unit of labor
Long-serving workers do not only know the official procedure.
They know which machine makes a suspicious noise in winter, which supplier becomes unreliable at month-end, which defect usually points to which cause, which customer cares about an unwritten detail, and which “new” failure actually happened ten years ago.
JILPT found that firms wanted to transfer more than precision work. They also valued the ability to handle unexpected trouble (65.2%), apply methods creatively (59.7%), and use judgment or tacit know-how to adjust work appropriately (52.6%).[2]
That is why replacing a retiree with one newly hired person restores headcount, but not necessarily capability.
When the veteran leaves, the employee badge goes out—and so does the exception-handling engine.
2. “Teach the next generation before you go” is already a huge job
The value of reemployment is not limited to continuing the same production task.
There is a backlog of knowledge work to do: OJT, incident reviews, failure libraries, photo-based procedures, video manuals, customer and supplier quirks, equipment tips, inspection judgment, and a post-handover escalation channel.
In the JILPT survey, 31.2% of firms were working to make inherited skills visible through text, manuals, or digitization. Among firms using digital tools for succession, 57.2% used digital procedures or e-manuals, 41.7% used video manuals, and 35.1% used AI. Reported benefits included visualization and standardization of skills (77.8%) and reduced training time (55.4%).[2]
So before a veteran disappears, management effectively has one last request:
“Please export every file that currently exists only in your head.”
The problem is that not every file can be exported. Keeping the person available even a few days a week can therefore be valuable.
3. Why succession fails: not enough young workers, not enough time
Among firms saying skill succession was not going well, 62.5% cited an insufficient supply of young employees and 50.0% cited lack of time.[2]
That produces an almost comic sequence:
Management: “Train the young staff.”
Workplace: “We do not have enough young staff.”
Management: “Then make time for training.”
Workplace: “We are already overloaded.”
Veteran: “My retirement date is here.”
Everyone: “Wait.”
Reemployment can therefore function not merely as an older-worker policy but as a way to buy additional transfer time.
4. Why does pay fall if the veteran is so valuable?
This is where people often confuse value with price.
According to a 2026 Ministry of Health, Labour and Welfare reference document using 2024 data, post-retirement wages were most commonly 60–70% of pre-retirement pay (45.3% of firms). Another 24.2% paid 80–90%, while 15.4% paid roughly the same amount.[3]
No one loses 40% of their accumulated knowledge overnight because they had a birthday.
What can change are employment category, hours, responsibility, promotion expectations, contract horizon, and bargaining position.
The key distinction is simple:
Value to the firm and the price the firm pays are not the same variable.
If a worker prevents losses worth far more than their salary, the company still has an incentive to hire them at the lowest sustainable market price.
That logic is not unique to older workers.
5. Pay is partly a procurement price for labor
Companies do not set wages in isolation.
They look at competing employers, nearby employers, pay for similar jobs, applicant flow, offer rejections, turnover, replacement cost, internal grades, and ability to pay.
In Teikoku Databank’s 2026 wage survey, 63.5% of companies expected wage improvements. The leading reason was retention and securing labor at 74.3%. Recruiting strength was cited by 36.3%, and wage trends at industry peers by 28.4%.[4]
That does not prove every firm runs a spreadsheet called “the minimum salary before people escape.”
But the economic pressure is obvious:
Pay too much and cost rises. Pay too little and the worker leaves.
The acceptable band moves when the external market moves.
6. “I would take the veteran over a new graduate” can make sense in the short run
The time horizon matters.
A new graduate is an investment in the future. The firm must recruit, train, expose them to failures, build internal networks, and let them accumulate company-specific knowledge.
A reemployed veteran already knows the product, equipment, people, customers, suppliers, history, and weird exceptions.
So if the question is “who can solve tomorrow morning’s problem?”, some departments may rationally value the veteran more.
If the veteran also costs less after reemployment, the short-run cost-performance case can look extremely attractive.
But keeping only veterans and stopping entry hiring creates another succession crisis five years later.
The graduate is the future pipeline. The veteran knows where today’s emergency stop is.
The strongest design gives them overlapping time.
7. Staying after retirement is not an exotic exception
As of June 2025, 99.9% of Japanese firms covered by the reporting system had implemented measures to secure employment through age 65. Of these, 65.1% used continued-employment systems. Measures to secure opportunities through age 70 had been implemented by 34.8%.[5]
The age-65 figure is partly driven by legal obligations, so it does not mean every company passionately wants every older worker.
But it does mean that continued employment is a standard part of Japanese workforce architecture, not some rare case of a retiree failing to understand a hint.
8. “Keep working as long as possible” should not mean “keep the same physical load until failure”
There is a real limit to the veteran-positive story: health and safety risk rises with age.
MHLW data show occupational accident frequency rates increasing from ages 60–64 to 65–69: from 1.71 to 2.23 for men and from 2.32 to 3.37 for women. Expected absence after injury also tends to become longer with age.[3]
So the sensible model is not “same heavy work forever.”
It is to reduce physically demanding, dangerous, night, or prolonged tasks as needed while shifting value toward coaching, judgment, troubleshooting, review, standardization, and escalation support.
Do not use up the veteran’s body. Use the veteran’s knowledge.
9. “They need you” and “they are getting you cheaply” can both be true
This is the core of the entire debate.
Company: “We need your experience.”
Worker: “So I keep my old salary?”
Company: “That is a separate meeting.”
Harsh, but not logically inconsistent.
Employers may genuinely need a person and still prefer not to pay more than necessary. Workers may genuinely be needed and still reject a price that is too low.
A pay cut does not by itself prove “they want me gone.”
Being asked to stay does not by itself prove “they value me above everyone else.”
A better model is:
How much the company needs the person and how much it must pay to retain the person are separate variables.
Conclusion: the company is not buying “an old worker”; it is trying to keep disappearing knowledge at a sustainable price
Labor shortages.
Too few young workers.
Too little training time.
Aging experts.
Skill succession.
Recruiting competition.
Pressure to control labor cost.
Put those forces in one room and post-retirement reemployment becomes much easier to understand.
The corporate position is often neither “please leave” nor “you are priceless.”
It is more painfully corporate:
“Please stay.
Please teach the younger people.
Please document what you know.
Please help when weird failures happen.
If you still want to work, excellent.
...and if possible, we would also like to keep the labor bill down.”
Then the outside market gets a vote.
Market: “At that price, they will leave.”
Company: “Fine. Raise it.”
Wages are not love.
But labor scarcity has a way of turning necessity into money.
Sources
- Teikoku Databank, “Corporate Trends on Labor Shortages, July 2026,” 2026-08-17 tdb.co.jp
- JILPT, “Survey on Securing and Retaining Human Resources and Skill Succession in Manufacturing,” 2026-05-27 jil.go.jp
- MHLW, reference materials for the Basic Policy on Employment Stability Measures for Older Persons, 2026-03-27 mhlw.go.jp
- Teikoku Databank, “Corporate Survey on Wage Trends for FY2026,” 2026-02-24 tdb.co.jp
- MHLW, “2025 Report on Employment Status of Older Persons,” 2025-12-19 mhlw.go.jp
