---
title: Japan's 3-Year Graduate Turnover, Taken Apart | Nobu Sato
description: About 3 in 10 Japanese graduates leave their first employer within three years, at an even pace each year. The real spread is by industry and company size.
image: https://nobusato.jp/hubfs/japan-graduate-turnover-3-years-cover-en.png
---

[Nobu Sato](https://nobusato.jp/en?hsLang=en)

[Home](https://nobusato.jp/en?hsLang=en) [Profile](https://nobusato.jp/en/profile?hsLang=en) [Articles](https://nobusato.jp/en/blog?hsLang=en) [Contact](https://nobusato.jp/en/contact?hsLang=en)

[日本語](https://nobusato.jp/ja/blog/japan-graduate-turnover-3-years)[English](https://nobusato.jp/en/blog/japan-graduate-turnover-3-years)[中文](https://nobusato.jp/zh/blog/japan-graduate-turnover-3-years)

[Japan by the Numbers](https://nobusato.jp/en/blog/tag/japan-by-the-numbers)

# Japan's 3-Year Graduate Turnover, Taken Apart

[Nobu Sato](https://nobusato.jp/en/profile?hsLang=en)

![Japanese university graduates leaving their first workplace: 12.1% in year one, 11.9% in year two, 9.9% in year three](https://nobusato.jp/hs-fs/hubfs/japan-graduate-turnover-3-years-cover-en.png?width=1200&height=630&name=japan-graduate-turnover-3-years-cover-en.png)

Japan has a phrase for it: *san-nen san-wari*, three in ten within three years. It is the share of new university graduates who leave their first employer within three years of joining. The latest figure from the Ministry of Health, Labour and Welfare (MHLW), released in October 2026, is 31.2% for the class of March 2023, down 2.6 points from the year before. Every autumn the number makes the news, usually as proof that young people lack staying power or that lifetime employment is ending.

For anyone hiring in Japan, the headline total is not very useful. What helps is its shape once you take it apart: which year people leave in, and which industries and company sizes they leave from.

## How the number is counted

The data comes from employment insurance records. When a workplace enrolls a new hire, it reports a date of birth and a start date. The ministry infers education level from the birth date. Anyone enrolled as a new graduate between March and June 2022 and born on or before April 1, 2000 counts as a university graduate. If that person later leaves that workplace, for another job, for graduate school or to move home, they count as having left.

So the university group includes postgraduates and late graduates, and leaving is not the same as entering the job market. The number measures who left their first workplace. That distinction matters later.

## Not a first-year cliff

Year 1Year 2Year 3Class of 202112.3%12.3%10.3%Total 34.9%Class of 202212.1%11.9%9.9%Total 33.8%Source: MHLW, Turnover of new graduates

**Fig. 1**About the same share leaves each year. Class of 2022: 12.1% in year one, 11.9% in year two, 9.9% in year three. The published total is 33.8%; the parts differ slightly because of rounding.

Early turnover sounds like a first-year problem. The data says otherwise. For the class of 2022, 12.1% left in year one, 11.9% in year two and 9.9% in year three. The year before had the same shape: 12.3%, 12.3% and 10.3%.

That makes the *dai-ni shinsotsu* pool, graduates on their second job within a few years of starting, less a seasonal batch and more a steady stream. In each of the first three years after graduation, roughly the same share leaves the first employer. The most recent first-year rate is 10.1% for the class of 2024. It is drifting down, but the stream has not stopped.

Over a longer view, the three-year rate for university graduates has hovered around 30% for three decades, from just under 29% in low years to nearly 37% in high ones. The phrase describes something old. It is not evidence that lifetime employment is suddenly collapsing.

## The spread is by industry and size

Hospitality & food55.4%Personal services54.7%Education44.2%Healthcare & welfare40.8%Retail40.4%All industries33.8%IT & communications27.2%Finance & insurance25.0%Manufacturing21.2%Utilities12.4%University graduates, class of 2022, three-year turnover. Source: MHLW

**Fig. 2**Same three years, very different outcomes: 55.4% in hospitality, 12.4% in utilities. The dark bars are the moderate-rate, large-base segment discussed below.

Under 557.5%5–2952.0%30–9941.9%100–49933.9%500–99931.5%1,000+27.0%University graduates, class of 2022, by number of staff at the workplace. Source: MHLW

**Fig. 3**The smaller the workplace, the higher the turnover: 57.5% under five staff, 27.0% at 1,000 or more.

Split by industry, the gap is wide. Hospitality and food service 55.4%, personal services and entertainment 54.7%, education 44.2%, against manufacturing at 21.2% and utilities at 12.4%. By workplace size, it runs from 57.5% at sites with fewer than five staff to 27.0% at sites with 1,000 or more. The class of 2023 shows the same pattern, with the largest employers down to 24.2%.

The national 30% is an average of very different groups. To judge whether a candidate's early exit is normal, look at the industry and size they came from.

## Which part of the stream matters to a foreign company

Foreign companies rarely win head-on against large Japanese employers for new graduates. As I wrote in [the post on the 98% graduate employment rate](https://nobusato.jp/en/blog/japan-graduate-employment-rate?hsLang=en), the annual new-graduate hiring system absorbs most of each cohort first. The second-job pool is the first group to move again after passing through that system.

Not all of it is relevant. The highest-turnover sectors supply many people, but most of them are far from a typical foreign B2B role. For most foreign companies the more useful segment is probably the one with moderate rates and a large base. Even at workplaces with 1,000 or more staff, about one in four leave within three years. IT and communications sits at 27.2%, finance and insurance at 25.0%. These people have been through a large company's training and are not yet locked into seniority pay.

A rough sense of scale: Japan had about 580,000 university graduates in March 2025, and 77% of them went into jobs. If around 30% leave within three years, that is well over a hundred thousand people per cohort. The exact count depends on definitions, but that is the order of magnitude.

On a [talent map](https://nobusato.jp/en/blog/talent-map-is-not-a-list?hsLang=en), I would take three things from this:

- Do not mark a candidate down just for leaving a first employer after a year or two. A tenth of each cohort does it every year. Ask why they left and what they want next.
- Draw the pool by source, not by national average. Narrow it to the industries and company sizes closest to your role, then look at that segment's rate and base.
- The window is years one to three after graduation, and it is open all year. Plan for year-round contact rather than a single April intake.

That the movers are mostly young also matches what the [job-change rate data](https://nobusato.jp/en/blog/japan-job-change-rate?hsLang=en) shows.

First, education is inferred from birth dates, not records. Leaving also includes graduate school and moves home, so the number who actually enter the mid-career market is smaller.

Second, the year-by-year split is from the class of 2022, published in 2025. For the class of 2023 I could only confirm the total and the main industry and size figures, so whether the structure has shifted will need the detailed tables.

Third, the count is by workplace. The documents do not say how transfers within a group or between sites are treated.

The second-job candidates you met recently: which industries did most of them come from?

Sources

- [MHLW, Turnover of new graduates (class of March 2022), press release (Japanese)](https://www.mhlw.go.jp/stf/houdou/0000177553_00010.html)
- [MHLW, Annex 1: three-year turnover by education level over time, split by year one, two and three](https://www.mhlw.go.jp/content/11805001/001318971.pdf)
- [MHLW, Annex 3: university graduates by industry and workplace size](https://www.mhlw.go.jp/content/11805001/001318973.pdf)
- [MHLW, Data source and calculation method](https://www.mhlw.go.jp/content/11805001/001318979.pdf)
- [Kyodo News via Daily Sports, October 23, 2026: class of March 2023 at 31.2%, 1,000+ staff at 24.2%](https://www.daily.co.jp/society/national/2026/10/23/0020959011.shtml)
- [MEXT, School Basic Survey FY2025: 584,304 university graduates, 77.0% employed](https://www.mext.go.jp/content/20260327-mxt_chousa01-000048552_10.pdf)

---

Shorter notes that don't make it into a full article go on LinkedIn. [Follow me there](https://www.linkedin.com/in/realnobusato/).

## More in this column

- ### [One Company, Four Female Manager Ratios](https://nobusato.jp/en/blog/same-company-different-numbers?hsLang=en)
  
  Hitachi's FY2025 securities report lists four female manager ratios: 17.4%, 22.7%, 8.9% and 6.8%. All correct, ...
- ### [Is LinkedIn Useful in Japan? What the Numbers Say](https://nobusato.jp/en/blog/is-linkedin-useful-in-japan?hsLang=en)
  
  LinkedIn reported 3 million members in Japan in 2022, about 4% of a 70 million labor force. Small, but dense in foreign ...
- ### [How Good Is English in Japan, Really?](https://nobusato.jp/en/blog/japan-english-proficiency?hsLang=en)
  
  Japan ranks 96th on the EF English index, yet a quarter of TOEIC takers score 745+. Neither measures the workplace. ...

[← All articles](https://nobusato.jp/en/blog?hsLang=en)

## Writing about launching in Japan, in three languages.

[Profile](https://nobusato.jp/en/profile?hsLang=en) [Get in touch](https://nobusato.jp/en/contact?hsLang=en)

Nobu Sato

Writing about launching in Japan, in three languages.

[info@nobusato.jp](mailto:info@nobusato.jp) [LinkedIn](https://www.linkedin.com/in/realnobusato/) [X](https://x.com/nobusato5888)

© 2026 Nobu Sato. All rights reserved.

```json
{
  "@context" : "https://schema.org",
  "@type" : "BreadcrumbList",
  "itemListElement" : [ {
    "@type" : "ListItem",
    "item" : "https://nobusato.jp/en",
    "name" : "Home",
    "position" : 1
  }, {
    "@type" : "ListItem",
    "item" : "https://nobusato.jp/en/blog",
    "name" : "Articles",
    "position" : 2
  }, {
    "@type" : "ListItem",
    "item" : "https://nobusato.jp/en/blog/japan-graduate-turnover-3-years",
    "name" : "Japan's 3-Year Graduate Turnover, Taken Apart",
    "position" : 3
  } ]
}
```

```json
{
  "@context" : "https://schema.org",
  "@id" : "https://nobusato.jp/en/blog/japan-graduate-turnover-3-years",
  "@type" : "WebPage",
  "author" : {
    "@id" : "https://nobusato.jp/#person"
  },
  "inLanguage" : "en",
  "isPartOf" : {
    "@id" : "https://nobusato.jp/en/blog",
    "@type" : "Blog",
    "name" : "Articles"
  },
  "name" : "Japan's 3-Year Graduate Turnover, Taken Apart",
  "url" : "https://nobusato.jp/en/blog/japan-graduate-turnover-3-years"
}
```

```json
{
  "@context" : "https://schema.org",
  "@id" : "https://nobusato.jp/#person",
  "@type" : "Person",
  "name" : "Nobu Sato",
  "sameAs" : [ "https://www.linkedin.com/in/realnobusato/", "https://x.com/nobusato5888" ],
  "url" : "https://nobusato.jp/en/profile"
}
```

```json
{
  "@context" : "https://schema.org",
  "@type" : "BlogPosting",
  "author" : {
    "@type" : "Person",
    "name" : "Nobu Sato",
    "url" : "https://nobusato.jp/en/blog/author/nobu-sato"
  },
  "dateModified" : "2026-10-24T03:07:54.502Z",
  "datePublished" : "2026-10-24T03:07:54.000Z",
  "headline" : "Japan's 3-Year Graduate Turnover, Taken Apart",
  "image" : [ "https://nobusato.jp/hubfs/japan-graduate-turnover-3-years-cover-en.png" ],
  "mainEntityOfPage" : {
    "@id" : "https://nobusato.jp/en/blog/japan-graduate-turnover-3-years",
    "@type" : "WebPage"
  },
  "publisher" : {
    "@type" : "Organization",
    "logo" : {
      "@type" : "ImageObject",
      "url" : "https://nobusato.jp/hubfs/nobusatojp-1.jpg"
    },
    "name" : "nobusato.jp"
  }
}
```