Japan's Union Rate Is 16%. Why the Shunto Still Sets Pay

Only 16.0% of Japan's employees belong to a union, a record low. Yet the 2026 spring wage round, the shunto, produced an average raise of 5.01%, the third straight year above 5%, and companies with no union at all moved with it. How does an institution that covers one worker in six set pay for a whole country? Start with what the 16% actually counts.

What is 16% a share of?

Union members as reported by the unions, divided by all employees in the government's Labour Force Survey. Every year the Ministry of Health, Labour and Welfare runs its Basic Survey on Labour Unions as of 30 June. For 2025 it counted 22,244 unions and 9,927,000 members. The denominator is the June Labour Force Survey figure of 62.05 million employees. Divide one by the other and you get an estimated union density of 16.0%.

Two things about the yardstick. The numerator includes public-sector employee associations, so this is not a private-sector figure. The denominator includes every employee: part-timers, temps, dispatched workers. A wide denominator pulls the rate down.

The trend is long and one-directional. In 1949 density was 55.8%, more than half of all employees. It fell below 20% for the first time in 2003, at 19.6%. The 2025 figure of 16.0% is the lowest the survey has recorded.

But membership in 2025 actually rose, by 15,000. The rate fell because the denominator grew faster: 660,000 more employees in a year. Unions didn't shrink. The workforce grew, and most of the new employees never joined.

Who is in the 16%?

Regular employees at large companies, plus a growing number of part-timers.

By company size in the private sector: at firms with 1,000 or more employees, density is 38.7%, and their 5,967,000 members are 68.3% of all private-sector members. At 100 to 999 employees, 9.9%. At 99 or fewer, 0.7%. Japan's companies are 99% small and medium-sized, and in that 99% unions barely exist.

By industry: 52.5% in compound services (mainly postal and cooperative organizations), 49.0% in electricity, gas, heat and water, 46.4% in finance and insurance, 28.1% in public service, 26.4% in manufacturing, 16.4% in wholesale and retail, 11.0% in hotels and restaurants, 5.3% in health care and welfare. Utilities, banks, factories and government offices. That is most of the 16%.

By status: part-time members number 1,494,000, up 2.1% on the year and 15.1% of all members, both records. Density among part-timers themselves is 8.8%. Women members number 3,545,000, a density of 12.3%.

And by banner: Rengo, the largest federation, has 6,822,000 members, 68.7% of all union members. Divide that by 62.05 million employees and you get about 11%. The organization that sets the demand and counts the answers every spring represents roughly one employee in nine.

16% is an average: nearly 40% at companies with 1,000+ staff, under 1% below 100

Estimated union density (%, union members divided by Labour Force Survey employees). Size bands are private sector. Ministry of Health, Labour and Welfare, Basic Survey on Labour Unions 2025, as of 30 June 2025

Where most foreign subsidiaries in Japan sitOther

All employees

• Total9.93m ÷ 62.05m
  16.0%

By company size　private sector

• 1,000+ employees68.3% of members are here
  38.7%
• 100 to 999
  9.9%
• 99 or fewer
  0.7%

By status

• Women
  12.3%
• Part-time workers15.1% of members, a record
  8.8%

Unions are a large-company institution. Below 100 employees, the pay benchmark still arrives from the shunto.

Why does the shunto still set pay at 16% density?

Because the shunto was built from the start to set a going rate for people who have no union, and that design hasn't changed in seventy years.

It began in 1955, when eight industrial federations of the Sohyo labor confederation agreed to bargain at the same time each spring. Strong industries settle first, and their number travels outward. The Nipponica encyclopedia describes the "shunto rate" as having served a social function: the benchmark by which low-density small companies and regional industries decided their own raises. Unions negotiate their own pay, and in doing so they publish a price for the whole market.

How does that price travel today? Three routes.

First, straight to the owner's desk as the going rate. The labor ministry's 2025 survey on wage revisions covers private firms with 100 or more regular employees. Only 22.0% of them have a union, 15.7% in the 100-to-299 band. Yet 91.5% raised pay, by 4.4% on average. Asked what mattered most, 41.7% said company performance, 17.0% said hiring and retention, 7.7% said the going rate, 5.9% said what the parent or group company did. Behind "performance" and "retention" stand employees and candidates who have read the shunto number in the news.

Second, into the minimum wage. For fiscal 2026 the national weighted average minimum wage rose 56 yen to 1,177 yen an hour, about 5.0%. At the central minimum wage council the labor side argued from "three straight years above 5%," and the employer side watched the survey of actual raises at small firms. The shunto sets a number in March; the minimum wage follows in the summer; hourly pay follows in October.

Third, down through corporate groups. The settlements that 38.7% density extracts at large companies flow into subsidiaries and affiliates through group pay systems. The 5.9% who answered "the parent company" are on this route.

So 16% and 5% don't contradict each other. 16% is the share of people at the bargaining table. 5% is how far the result reaches. They measure different things, which is the one point I keep making in Japan by the Numbers.

Your Japan subsidiary has no union. Which strings are attached to it?

Candidates, the minimum wage, and the parent company.

Most foreign subsidiaries in Japan have fewer than 100 employees, so by the numbers above the chance of a union is under 1%. Nobody comes to negotiate. The numbers walk in anyway. After the big companies' answer day in mid-March, candidates arrive with a new floor in their heads, and current staff have seen the same headlines. At the end of July the minimum wage council issues its guideline; it takes effect from October, prefecture by prefecture, and moves hourly pay in shops and warehouses. If your parent has other entities in Japan, the group's revision rate lands on your desk before your budget does.

For 2027, Rengo is asking 7% for fixed-term, part-time and contract workers, above the minimum wage increase. Your part-timers may not know that figure. The companies competing with you for part-timers do.

My advice is dull. Watch what comes out in July, every year. Early July: Rengo's final tally, both the overall figure and the small-company figure. End of July: the labor ministry's settlements at major companies. Also end of July: the minimum wage guideline. With those in hand, you have next year's pay range by August. No need to guess in December.

One thing I can't settle. The shunto's anchor is the large-company union, and that anchor is slipping. Density at companies with 1,000 or more employees fell from 40.0% to 38.7% in a single year, 1.3 points, and membership growth came from part-timers. Unions are turning into a combination of large-company regulars and part-time workers. Whether that group can still price labor for the whole country in ten years, I don't know. What I do know is that for 2027 they are asking for 6%, and that number is already on its way into budgets at companies that have never seen a union.

---