Background: A chat with a friend made me go dig through official statistics on foreign workers' pay. This isn't a judgment of any individual. Details that could identify someone (nationality, employer, region, friendships) have been generalized. Amounts are before tax and social insurance unless noted.
1. The 5-second answer: roughly a three-stage rocket, but the rules don't say "cheap is fine"
In the Ministry of Health, Labour and Welfare's 2025 Basic Survey on Wage Structure, average scheduled monthly pay (base pay, excluding overtime; Japanese statistics quote pay in units of 10,000 yen, so "19.03 man yen" means 190,300 yen) was 19.03 man yen for Technical Intern Trainees, 22.14 man yen for Specified Skilled Workers, and 31.32 man yen for the "professional and technical fields" category excluding Specified Skilled Workers.
The 2024 Survey on the Employment of Foreign Nationals lets us isolate the "Engineer/Specialist in Humanities/International Services" visa (people usually shorten it to "Gijinkoku," so let's call it the specialist work visa). Its scheduled pay was 27.83 man yen, and its total regular cash pay, which includes overtime, was 31.12 man yen.
If you want a rough rule of thumb:
Technical intern about 19 man yen → Specified Skilled Worker about 22 man yen → specialist work visa about 28 man yen.
It really does look like a three-stage rocket. But no Japanese law or program says "trainees are trainees, so minimum wage is fine." For all three statuses, the system is built around the idea that pay must be equal to or higher than what a comparable Japanese worker would get.
In short: on paper, equal pay; in the labor market, layered pay. If you don't separate the two, the discussion gets lost.
2. The three statuses aren't the same "foreigner category"
The Technical Intern Training Program was designed to transfer skills. Specified Skilled Worker is a residence status for people with a certain level of skill working in industries short of labor. The specialist work visa covers professional, white-collar-type work in technology, the humanities, or international business.
So it's less "foreigners are paid differently" and more that you're comparing groups with different jobs, required skills, experience, company sizes and industries.
It's like pricing a supermarket's "bento boxes," "sushi" and "frozen tuna" as one big "seafood aisle" average. Of course the number comes out weird.
3. Average monthly pay: the easiest comparison is without overtime
| Status | Survey year | Average scheduled pay | How to read it |
|---|---|---|---|
| Technical intern | 2025 | 19.03 man yen | Lots of young people in hands-on jobs, so it's low |
| Specified Skilled Worker | 2025 | 22.14 man yen | A step above technical interns |
| Professional and technical fields (excl. Specified Skilled Workers) | 2025 | 31.32 man yen | Also mixes in other highly skilled people |
| Specialist work visa only | 2024 | 27.83 man yen | The only figure for this visa on its own |
Calling 31.32 man yen the "average for the specialist work visa" is risky. Highly skilled professionals are included, which pushes it up. If you only want the specialist work visa, 27.83 man yen is the more honest number.
The feeling that "specialist work visa = about a new Japanese graduate's pay" is roughly right in direction, but the statistics include experienced people too, not just new grads. So the safe way to put it is not "exactly new-grad pay" but "close to the pay band of young professionals and white-collar staff."
4. Median: if you only look at averages, the pay tables start playing tricks
The median is the value for the person in the middle, so it's less affected by a few high earners pulling the average up.
In the 2024 Survey on the Employment of Foreign Nationals, regular Specified Skilled Workers had scheduled pay of 21.30 man yen on average and 20.60 man yen at the median. Cash pay including overtime was 25.03 man yen on average and 25.30 man yen at the median. Annual bonuses and the like were 10.77 man yen on average, but 0 man yen at the median.
You see an average bonus of about 10 man yen and think, "Well, I guess they get about 10 man yen." Then the 0 man yen median taps you on the shoulder. The bonus column is a mirage.
For technical interns, the 2023 median scheduled pay in the Basic Survey on Wage Structure was 17.54 man yen, for Specified Skilled Workers 19.01 man yen, and for the wider "professional and technical fields (excl. Specified Skilled Workers)" 25.03 man yen.
The important point is that published statistics don't slice the data the same way every time. A median for the specialist work visa alone and a median for the whole professional and technical field are not the same thing. Don't fill in missing numbers by vibes. "Probably around this much" is a banned card in the statistics world.
5. Differences by industry: construction is strong for Specified Skilled Workers, manufacturing sits in the low 20s (man yen)
The 2025 Basic Survey on Wage Structure shows scheduled pay for Specified Skilled Workers by industry.
| Industry | Specified Skilled Workers, scheduled pay | Approx. headcount in the sample |
|---|---|---|
| Construction | 26.59 man yen | 21,750 |
| Healthcare and welfare | 22.50 man yen | 31,550 |
| Accommodation and food services | 21.83 man yen | 9,110 |
| Wholesale and retail | 21.39 man yen | 5,770 |
| Manufacturing | 20.73 man yen | 72,660 |
| Real estate and goods rental | 22.43 man yen | 550 |
Among the major industries with a decent headcount, construction (26.59 man yen) and manufacturing (20.73 man yen) differ by 5.86 man yen a month. Over a year, that's about 70 man yen in simple math.
There are even lower published rows, such as transport and postal at 17.42 man yen and personal services and entertainment at 18.40 man yen, but the headcounts behind them are tiny, at 220 and 90 people. If you shout "worst industry ranking!" from those, the sample-size police will show up.
Also, this survey doesn't cover agriculture, forestry or fishing. Specified Skilled Worker status does exist in agriculture and fishing, but you can't compare them with this table.
And for technical interns and the specialist work visa, there's little in the way of a published table that lines up the median for all industries in the same year, with the same definitions and the same level of detail. Rather than forcing the table to be full, the right approach is to show only what can be compared.
6. Converting to annual income: "monthly pay × 12" is not the same as "realistic annual income"
First, simply multiplying 2025 scheduled pay by 12 gives:
| Status | Monthly pay | Simple annual figure |
|---|---|---|
| Technical intern | 19.03 man yen | about 228.4 man yen |
| Specified Skilled Worker | 22.14 man yen | about 265.7 man yen |
| Professional and technical fields overall | 31.32 man yen | about 375.8 man yen |
| Specialist work visa only (2024) | 27.83 man yen | about 334.0 man yen |
This is a mechanical ×12 with no overtime and no bonus.
Next, using the 2024 Survey on the Employment of Foreign Nationals, I built an annual-income-like reference figure: monthly cash pay including overtime × 12, plus bonuses and the like from the previous year.
| Status | Cash pay per month | Annual bonuses etc. | Reference annual figure |
|---|---|---|---|
| Technical intern | 21.00 man yen | 3.94 man yen | about 255.9 man yen |
| Specified Skilled Worker | 25.03 man yen | 10.77 man yen | about 311.1 man yen |
| Specialist work visa | 31.12 man yen | 35.73 man yen | about 409.2 man yen |
An easy way to remember it: about 260 → 310 → 410 man yen.
But this isn't strictly "actual income for the same twelve months." Monthly pay is for the survey month, while bonuses cover the previous year, so the periods differ. It's a stand-in calculation for annual income, not the sum of twelve actual pay slips.
7. "Trainees = minimum wage" is wrong under the rules, but understandable as an impression of reality
Article 9 of the Technical Intern Training Act makes it an approval standard that a trainee's pay be equal to or higher than what a Japanese person doing the same work would receive.
The Immigration Services Agency's Q&A also requires Specified Skilled Worker pay to be equal to or higher than that of Japanese workers doing equivalent work.
The specialist work visa likewise has a standard that pay must be at least equal to what a Japanese person would receive.
So under the rules:
"They're foreigners, so cheap is fine" is not allowed.
Still, one reason the technical intern average sits in the 19 man yen range is that the comparison group itself skews toward young people, hands-on jobs and low-wage industries. Length of service, region and company size differ too. "Equal to a Japanese worker doing the same job at a similar skill level," as the rules say, is a different thing from "equal to the national average for all workers."
So it isn't "literally minimum wage," but the observation that they tend to be concentrated in workplaces with pay close to minimum wage is understandable.
8. Look at remittances and the same 1,000 yen feels heavier
In the 2024 Survey on the Employment of Foreign Nationals, the share sending money to family in their home country was 45.1% for the specialist work visa, 81.6% for Specified Skilled Workers and 83.5% for technical interns. Among those who did send money, the yearly average was 115.3 man yen for the specialist work visa, 123.3 man yen for Specified Skilled Workers and 106.3 man yen for technical interns.
Of course, not everyone sends that much, and whether someone sends money and what their family situation is varies from person to person. Still, for someone with lower pay who also sends money home, 1,000 or 2,000 yen for a meal out may not mean the same thing it does to us.
1,000 yen is physically 1,000 yen. But the load it puts on the household-budget CPU isn't necessarily the same.
9. So, are you a terrible person if you don't treat a technical-intern friend? No, don't set up a Ministry of Finance inside your friendship
From here on it's not statistics, it's human relationships.
Say your friend is a technical intern who probably earns less than you. Once you know that, a feeling can creep in: "Splitting the bill every time... am I being a bit awful?"
But an income gap does not mean the higher earner has to pay.
A friend isn't a welfare program, and if you introduce progressive income tax into friendship, an audit starts out of nowhere.
- Splitting the bill: perfectly normal
- "I've got this one today": kind
- They pay for you: fine to accept with thanks
- You pay every time: generous, but not an obligation
What matters isn't "I'll pay because they earn little" but the sense of agency: "I want to treat this person today, so I will."
10. The moment treating becomes an obligation, it turns from "kindness" into a "mysterious tax"
Treating feels good because you get to choose.
"Today was fun, I'll get this" is a gift.
"They earn less, so I'm a failure as a person if I don't pay every time" is already policy design. Don't open an Income Redistribution Agency inside your friendship uninvited.
On the flip side, if being treated makes you happy, the other person may feel the same. Some people feel only "I'm sorry" when someone treats them, and others just say "Yay, thanks for the meal!" and enjoy it.
Relationships aren't double-entry bookkeeping where every transaction has to come out 50:50.
11. The easiest way to run it: "Pay when you want to; accept gladly when someone pays for you"
Between friends, this approach is the least likely to go wrong.
① You invited them to an expensive place → it's natural to pay somewhat more.
② An ordinary meal → splitting the bill is perfectly fine.
③ You feel like treating today → treat. No explanation needed.
④ They treated you → if they aren't stretching themselves, just say thank you and accept.
⑤ Don't turn it into a fixed rule → it's harder for a sense of obligation or hierarchy to form.
The point is not to install a friendship ERP system that says "I'm owed 1, you're owed 1, and 324 yen of fried chicken is unsettled."
12. What the statistics can tell you, and what they can't
What they show is that pay levels differ quite a lot by residence status, and that even among Specified Skilled Workers, industry makes a big difference. They also show that the rules don't allow pay to be pushed unfairly low just because of nationality or residence status.
What they can't show is the financial situation of the one person in front of you.
Even among technical interns, dorm fees, overtime, remittances and savings goals differ. Even among specialist work visa holders, company, job type and years of experience change everything. Jumping straight to "poor because of this status" or "comfortable because of this status" is far too crude.
Statistics are a map for looking at people, not the people themselves.
13. Reading the numbers: don't mix averages, medians, overtime and bonuses
- Scheduled pay: pre-tax pay excluding overtime, late-night and holiday work pay.
- Total regular cash pay: scheduled pay plus overtime pay and similar. It's generally higher than scheduled pay.
- Annual bonuses and other special pay: some surveys report this for a different period.
- Average: can be pulled up by high earners.
- Median: the middle value. But the categories in published tables don't always match those in the average tables.
- By industry: rows with small samples swing a lot.
- 2025 wage survey: agriculture, forestry and fishing are outside the survey.
So "28 man yen a month," "31 man yen total monthly pay" and "409 man yen a year" are not measured on the same ruler. Before putting numbers side by side, check what each one includes.
14. Summary: pay is a three-stage rocket, but friendship doesn't need rocket science
As a feel for the going rate,
Technical intern: about 19 man yen a month → Specified Skilled Worker: about 22 man yen a month → specialist work visa: about 28 man yen a month (scheduled pay, as a guide).
As annual, income-like total pay reference figures,
Technical intern: about 256 man yen → Specified Skilled Worker: about 311 man yen → specialist work visa: about 409 man yen.
The first impression that "trainees are close to minimum wage and the specialist work visa is close to young Japanese professionals" holds up pretty well in the statistics. But under the rules, none of the three statuses is a system where "they're foreigners, so cheap is fine."
And seeing these numbers doesn't mean you have to feel, "Then I should treat the lower-paid friend every time."
Look at pay through statistics. Decide treating by feeling.
Keep the two separate, and both your wallet and your friendships stay pretty peaceful.
