Japan's Labor Share: 47% at Big Firms, 82% at the Smallest

Japanese companies paid out 64.2% of the value they created as labor costs in fiscal 2024. That is the official labor share from the Ministry of Finance's corporate statistics, and it is the number a head office will find if it asks what share of earnings normally goes to people in Japan. It is also close to useless as a budget benchmark. Split it by company size and firms with capital of 1 billion yen or more sit at 47.3%, while firms under 10 million yen sit at 81.5%. Same country, same year, a 34-point gap. Among manufacturing, wholesale and retail, in a trade ministry survey of larger firms, the gap is about 3 points.

What is the labor share a share of?

Labor costs divided by value added. In the Ministry of Finance's Financial Statements Statistics of Corporations, value added is the sum of five items: labor costs, interest paid, rent on movable and real property, taxes and public dues, and net operating profit. Labor costs are directors' pay and bonuses, employees' pay and bonuses, and welfare costs including the employer's social insurance. This is a different number from labor costs over sales. Over sales, Japanese companies spent 13.6% on people in fiscal 2024, by the Japan Institute for Labour Policy and Training's calculation from the same survey. When head office asks "what share of earnings," the first thing to settle is whether earnings means revenue or value added.

The survey covers for-profit corporations other than banks and insurers. It is a sample survey; the population in fiscal 2024 was 3,044,511 companies. Only 4,622 of them had capital of 1 billion yen or more, and 2,152,605 had under 10 million yen. The 64.2% is a ratio of summed yen, not an average across companies, so the large firms that generate most of the value added pull it down.

One detail matters more than it looks: directors' pay is in the numerator. In a company run by its owner, much of the profit is paid out as the owner's salary. That counts as labor cost and pushes the labor share up. Part of the reason small companies show a high share is this definition.

Three government numbers, one name

For the same fiscal year, Japan publishes three official labor shares. The Ministry of Finance survey gives 64.2%. The Cabinet Office's national accounts give 69.5%, as compensation of employees over national income. The Ministry of Economy, Trade and Industry's Basic Survey of Business Structure and Activities gives 46.6%. Highest to lowest, that is 23 points.

The differences are in the denominator and the population. The national accounts cover the whole economy, including sole proprietors and government. The METI survey covers only companies with 50 or more employees and capital of at least 30 million yen; its numerator is gross wages and its value added includes depreciation. Count only larger firms and widen the denominator, and the ratio falls. None of the three is wrong. But a headline that says "labor share" without naming the ministry cannot even be compared with last year's.

Size moves the number more than industry does

By capital size, the fiscal 2024 labor share in the Ministry of Finance survey was 47.3% for 1 billion yen and over, 62.8% for 100 million to 1 billion, 74.4% for 10 million to 100 million, and 81.5% for under 10 million. The 64.2% for all sizes is a point somewhere between large and small companies, and describes neither.

Industry differences look smaller, at least where I could measure them. The METI survey breaks its fiscal 2024 total of 46.6% into manufacturing 45.9%, wholesale 43.7% and retail 46.7%. Three points. It covers only companies with 50 or more employees, and only these three industries line up, so it cannot show the full industry spread. Within that range, whether a company makes things or sells them matters far less than whether its capital is 10 million yen or 1 billion.

The time series tells you how to read a single year. For all sizes the share was 67.5% in fiscal 2015, 68.6% in 2019, jumped to 71.5% in the pandemic year 2020, then fell four years running: 68.9%, 67.5%, 65.1%, 64.2%. The labor share rises in a downturn because value added shrinks first and payroll does not. So the fall to 64.2% in a year of the biggest wage rounds in three decades means value added grew faster than labor costs, which is the version of a falling labor share you want.

By size, firms at 1 billion yen and over went from 57.6% in fiscal 2020 to 47.3% in 2024, a drop of 10 points. Firms under 10 million yen peaked at 91.0% in fiscal 2021 before settling at 81.5%. For two pandemic years (86.5% in fiscal 2020, 91.0% in 2021) the smallest companies in Japan paid out about nine-tenths of what they created.

Why the smallest companies run the highest share

The answer is in the denominator. Small companies do not pay unusually high wages. They keep unusually little profit. The Small and Medium Enterprise Agency's 2026 White Paper puts net operating profit at 36.0% of value added for large companies and 9.5% for SMEs in fiscal 2024, and concludes that SMEs have less room than large firms for further wage increases. When profit is thin, whatever is left is mostly payroll.

Value added per person differs too. The same white paper gives 19.35 million yen per employee at large companies and 6.66 million yen at SMEs (capital of 10 to 100 million yen) in fiscal 2024, nearly three to one. One is the economics of plant and systems, the other of people's hours. When a company earns with hours, payroll grows in step with sales and the share stays high by construction.

Then there is the directors' pay point from above. Most of the 2.15 million companies under 10 million yen in capital are small, and it is common for the owner to take profit as salary. By definition all of that is labor cost, so a fair amount of what sits inside 81.5% was really profit. A high labor share at a small company is no evidence that its staff are well paid.

Which number a Japan subsidiary should budget against

Compute the labor share of a new foreign subsidiary in Japan and for the first two or three years it usually comes out above 100%. Value added is tiny or net operating profit is negative. If head office says "the Japanese average is 64%, you are too high," it is comparing against the wrong population. In the launch phase the useful numbers are the month people arrive and the month revenue starts, which I went through in Why your first-year budget in Japan runs over by half.

Once revenue is real, benchmark against companies your size. For a subsidiary with capital between 10 million and 100 million yen that is 74.4%; above 100 million it is 62.8%. Those two numbers are closer to your reality than any industry average. The 64.2% is pulled down by the weight of large companies; a 20-person entity that lands there is either unusually profitable or running short-handed.

And line up the numerator. The Ministry of Finance's labor costs include welfare costs, which means the employer's statutory social insurance. Put only gross salaries in your numerator and your share looks about 10 points lower than it is. The employer's add-on costs I set out in What one hire in Japan really costs: 1.2x to 1.5x salary belong inside. Directors' pay too: if head office pays the Japan country manager, the Japanese entity's share comes out below that of a comparable Japanese company. Before anyone at head office celebrates how efficient Japan is, add it back.

The labor share is an outcome, and it distorts when you target it. The only way to lower it without growing value added is to cut payroll. The share fell in a year of wage increases because the denominator grew faster than the numerator. What a Japan budget should show is whether next year's value added is likely to outrun next year's payroll growth. If it is not, fiddling with the ratio will not help. The problem is on the selling side.