Japan's Gen Z and the Trades: What Actually Gets Paid

Japan's young people are not abandoning office jobs for the trades, whatever the headlines say. The "blue-collar billionaire" story crossed the Pacific from the US, and Japanese media picked it up fast. The numbers tell a narrower story. What's changing in Japan is which skills get paid, and that line doesn't follow the color of anyone's collar.

Is Gen Z in Japan really moving into skilled trades?

Not in any large way. Most young job seekers in Japan still won't consider blue-collar work.

In April 2026 UZUZ Holdings, a Tokyo firm that helps people in their twenties find jobs, surveyed 457 people in their twenties who had booked career consultations with it. Asked about blue-collar jobs, 64.3% said they would not consider them at all, whatever the terms. Another 21.0% said they might if pay and hours were good. Only 11.6% said jobs that only humans can do, even as AI improves, have a future. On construction site management specifically, the share who said they'd want it even with a path to ¥5 million a year in their twenties was zero.

That sample leans toward people looking for office work, so the result is not surprising. But it's hard to square with the idea of a generation rushing to the job site.

There is movement on the other side, though. Levages, a staffing company, surveyed 724 people currently working blue-collar jobs, mostly in manufacturing, construction and logistics, in March 2026. 20.4% had come from white-collar work. Among those in their twenties, 14.5% said they chose the job partly because it felt safe from AI.

The same survey asked what turned out worse than expected. The top answer, at 39.9%, was pay.

So some people are moving, and AI anxiety is part of why. But the people who actually made the move are the first to say the money story was oversold.

What do the job numbers show?

Labor shortages in Japan track how hard a skill is to acquire, not whether the work is manual or clerical.

Japan's Ministry of Health, Labour and Welfare publishes the jobs-to-applicants ratio by occupation every month, based on the public employment service, Hello Work. A ratio of 2 means two openings for every job seeker. In August 2026, across all occupations, the unadjusted ratio was 1.06.

Construction framework work (formwork, rebar, scaffolding) was at 8.25. Architectural, civil engineering and surveying technicians were at 5.71. Electrical work was at 3.23.

Now look at the low end. Hauling, cleaning and packing is manual work too, and it was at 0.66, below average. General clerical work was at 0.29, meaning roughly three and a half applicants for every opening. Meanwhile the surveying and engineering technicians, who sit in an office category, were at 5.71.

Japan's shortages follow skill, and both collars include jobs with too many applicants

Jobs-to-applicants ratio by occupation, regular jobs including part-time, not seasonally adjusted, August 2026. Ministry of Health, Labour and Welfare. All occupations: 1.06

Needs specialist skill or a licenseLow barrier to entry

Site and manual work

• Construction frameworkformwork, rebar, scaffolding
  8.25
• Electrical work
  3.23
• Hauling, cleaning, packing
  0.66

Office and technical work

• Civil engineering and surveying technicians
  5.71
• IT and telecom engineers
  1.33
• General clerical
  0.29

The question is whether anyone could do the job next week. If they could, the collar won't save the wage.

Wages tell the same story. Frog, a company that analyzes job-posting data, looked at full-time and contract postings on doda, Mynavi Tenshoku, en Tenshoku and Hello Work. From March 2020 to March 2026, the average advertised monthly salary rose 17.26% for blue-collar postings and 25.06% for white-collar ones. Postings offering ¥700,000 a month or more numbered 1,575 for white-collar jobs and 225 for blue-collar jobs, a gap of about seven times. These are advertised figures, not pay slips. Still, Japan is nowhere near the point where the trades out-earn the office.

The occupation-level data has its quirks, including weak coverage of IT jobs. I went through them in an earlier piece on job ratios by occupation.

Why Japan's version of the story is different

In the US, the trades pitch works partly because four-year college is expensive and apprenticeships pay while you learn. Japan's setup is different. Large employers hire new graduates in bulk for general roles, without asking for a specific skill, and train them inside the company. The office job still comes with a training path attached. That makes "skip college and learn a trade" a harder sell, even when the job data favors some trades.

It also means the office jobs that are oversupplied in Japan are the generic ones. General clerical work at 0.29 is the clearest case. It was oversupplied before generative AI arrived, so I wouldn't blame AI for the number. But data entry, checking and filling in standard reports are exactly the tasks easiest to write down as steps.

What AI threatens

The IMF estimated in January 2024 that almost 40% of global employment is exposed to AI, and about 60% in advanced economies. Roughly half of the exposed jobs may benefit through higher productivity. The other half may see lower demand, wages and hiring. The IMF also noted that younger workers may find it easier to use the opportunities, while older workers could struggle to adapt.

I read that as a split inside occupations more than between them. In the same job, one person loses half their tasks to AI and another doubles their output with it. The difference is whether the work follows steps or requires judgment.

The trades follow the same rule. Electrical work in Japan pays because work on homes and small shops has to be done by a licensed electrician under the Electricians Act, and because every site is different. Hauling and packing pays less because anyone can be trained in days and swapped in.

How to judge whether a skill is worth betting on

If a young person in Japan asked me, I'd suggest four checks.

Look at the specific occupation, over several years. "Construction" includes jobs at 8 and jobs near average. One strong year could be the cycle. Several years above 2 is a structural shortage.

Look at the barrier to entry. Is there a license? How many years until you work on your own? If anyone can learn it in a week, wages won't rise much however short-staffed the industry is. Count the cost of getting over the barrier too: the years, the fees, and the income you give up while training.

Ask whether it travels. Would another company pay for it? Another city? Another country? A skill that only has value inside one employer belongs to the employer.

Ask which part AI eats. Break the job into tasks. The parts you can write as steps go first. What stays is judgment on site, dealing with people, and owning the result. I used a similar breakdown in After AI, which layer of you loses value first?

None of this says white-collar work is finished. The engineering technicians at 5.71 are the best counterexample: technical skill plus a specific industry is very hard to replace. That combination is what I'd look for, in an office or on a site.

For foreign companies hiring in Japan, there's a practical point too. If you're competing for young candidates, "we'll train you" is no longer enough. Tell them what they'll be able to do in three years, and whether anyone else will pay for it. More on moving jobs in Japan is in the career guide.

---