Three Decades of Wage Stagnation — The Structural Mechanisms Behind Japan's Plateau Since the 1997 Peak

"Average annual income in 1997: ¥4.67 million. In 2025: ¥4.65 million. Nearly unchanged for 30 years. Meanwhile, prices went up, social insurance premiums went up, taxes went up. Before we even noticed, Japan had become a country that's getting poorer." — via Threads

"I have a simple question: why can't you just live a normal life on the salary you earn from working hard?" — via Threads

Behind these voices lies a structure that data can illuminate.

The 2025 spring wage round settled at a weighted average of ¥16,356, or 5.25%, across the 5,162 unions that secured an answer on the average-wage basis, against ¥15,281 and 5.10% the year before[1]. Rengo's own summary describes this as exceeding the previous year, which had been the first settlement above 5% since 1991's 5.66% — a gap of 33 years. Yet workers' lived experience tells a different story.

Real wages in 2025 fell 1.3% from the previous year, a fourth consecutive annual decline[2]. Total cash earnings came to ¥355,919, up 2.3%[2], but that fell short of the 3.7% rise in the consumer price index excluding imputed rent, which is the deflator used for this series[2]. Wage increases arrived, but prices outpaced them. The real purchasing power of take-home pay has continued to fall.

This pattern did not begin in 2025. Japan's nominal wage peak was ¥4.67 million in annual income in 1997[3]. That was followed by more than 25 years of stagnation. Nominal wages did reach a new record of ¥4.78 million in 2024, but in real terms, they remain below the 1997 level.

International comparisons highlight the abnormality of this stagnation. Taking 1996 as 100, nominal productivity per worker reached 230 to 240 in the UK and the US and about 160 in France and Germany, while Japan stayed essentially flat; nominal wages per worker rose roughly in step with productivity everywhere except Japan, where they fell by about 4%[4].

The question being asked is not "why aren't wage increases happening?" The spring offensive has delivered 5% wage increases. The problem is the structural question of why wage increases are not translating into higher real purchasing power for take-home pay.

The Starting Point of 30 Years of Stagnation — The Bubble Collapse and "Three Excesses"

After the collapse of the asset bubble in the early 1990s, Japanese companies were consumed with unwinding the "three excesses" — excess employment, excess capacity, and excess debt. Labor cost compression became the top priority for corporate restructuring, and wages began to fall from their 1997 peak.

The choice companies made during this period was to restrain regular employment and expand non-regular employment. The non-regular employment ratio rose from 16.4% in the early 1990s to 36.8% in 2024[5]. The average wage of non-regular workers is roughly 60% that of regular workers. The rising non-regular ratio has a statistical effect of pushing down the average wage.

However, the expansion of non-regular employment alone does not explain 30 years of stagnation. Wages of regular employees have also failed to rise.

Two Decades of a Deflationary Mindset

For approximately 15 years from 1998 to 2012, Japan was in deflation. In an environment where prices continue to fall, companies cannot pass on costs and cannot secure the resources for wage increases.

The Bank of Japan's Outlook Report repeatedly records that firms, smaller ones above all, say passing wage increases through to prices is not easy. Fear of losing customers by raising prices gave companies a persistent reason to avoid wage increases. A figure comparing service-sector pass-through rates in Japan and the United States does not appear in the Bank's published material, so it has been dropped.

After 2013, the Bank of Japan's unprecedented monetary easing policy allowed Japan to escape deflation, but the behavioral pattern — "don't raise prices, don't raise wages" — remained embedded in corporate culture as a deflationary mindset.

¥637 Trillion in Retained Earnings and the Declining Labor Share

Productivity-Wage Correlation

1970–1994

0.99

Correlation: 0.99 (near-perfect)

1995–2021

0.36

Correlation: 0.36 (divergence widens)

Wage increase per +1% productivity

Japan+0.4%

United States+1%

Corporate Internal Reserves (Retained Earnings)

FY2012304T¥

FY2016406T¥

FY2020484T¥

FY2024 (record)637T¥

13 consecutive years of record highs. Profits flow to shareholder returns and reserves — not wages.

Corporate retained earnings (accumulated profits) reached ¥637 trillion[6] in FY2024, a 13th consecutive record high. Companies are generating profits. The problem is that those profits are not flowing to wages.

The mid-1990s, when productivity and wages parted

The disconnect between productivity and wages emerged in the mid-1990s. The Ministry of Health, Labour and Welfare "Analysis of the Labour Economy" puts the correlation coefficient between nominal productivity and nominal wages at 0.99 for 1970–1994 and 0.36 for 1995–2021[7]. What had moved almost as one now retains barely a third of that linkage. The white paper states that from the late 1990s onward, "a situation persisted in which wages did not rise as much as productivity." The same white paper notes that a 1% rise in nominal productivity per worker in the United States is matched by a rise of roughly the same 1% in nominal wages, whereas in Japan nominal wages rise by only about 0.4%[4].

Where has the profit gone? To dividends and share buybacks — that is, shareholder returns. This tendency accelerated from the 2010s onward, when corporate governance reform shifted companies toward prioritizing return on equity (ROE). The accumulation of retained earnings is also the result of corporate defensive behavior — "saving for uncertainty about the future."

The Structural Weakening of Labor Unions

The power of labor unions — the primary agents of wage demands — has weakened historically.

Japan's labor union membership rate is 16.1%[8] (2024, a record low). At its peak in 1949, it was 55.8%. More than 70% of union members are concentrated in large corporations; at small and medium-sized enterprises, the overwhelming majority of workplaces have no union at all.

Furthermore, Japan's labor unions are predominantly "enterprise-based unions." This structure is fundamentally different from Germany's industry-wide collective agreements (represented by industry-wide unions such as IG Metall, which negotiate wages for entire industries) or Nordic solidarity wage systems. Enterprise-based unions tend to restrain wage demands citing deteriorating company performance.

The 2025 spring offensive achieved over +5% at large corporations, but the gap with small and medium-sized enterprises exceeded 1 percentage point (approximately ¥6,500 per month). For workers at SMEs with low union membership rates, the benefits of the spring offensive are structurally difficult to reach.

Social Insurance Premiums as a "Hidden Extraction"

Another reason wage increases are not reflected in take-home pay is the increase in social insurance premiums.

An analysis by Daiwa Institute of Research (January 2025) reveals the transformation in household burdens over 35 years. The combined tax and social insurance burden rate rose from 20.6% in 1988 to 25.9% in 2023[9] (+5.3 percentage points). Nearly all of the increase is attributable to higher social insurance premiums.

As a result, real disposable income in 2023 was ¥11,000 per month lower than in 1988. Even when nominal wages rise, if withholdings for social insurance and taxes increase, take-home pay does not rise. The fruits of wage increases being absorbed by social insurance premiums — this structure is the true source of the feeling that "wages rise but life doesn't get richer."

The same structure operates on the corporate side. Since social insurance premiums are shared equally between employer and employee, premium burdens on companies also increase as wages rise. The structure in which part of the money for a pay rise goes to social insurance premiums holds down corporate willingness to raise wages. The share of statutory non-wage labor costs in total wages between 2001 and 2020 has been dropped: the Keidanren page cited reports year-end bonus settlements and carries no such figure, and Keidanren's own survey of welfare benefit costs ended in 2019.

Lessons from South Korea — Is Rapid Minimum Wage Increase the Answer?

"Substantial minimum wage increases" are frequently discussed as a prescription for wage stagnation. South Korea implemented this approach, carrying out minimum wage increases of +16.4% in 2018 and +10.9% in 2019.

The results were mixed. Wages of workers near the minimum wage rose, but the non-regular employment ratio increased by 3.4 percentage points in a single year, and youth unemployment also rose. Rapid minimum wage increases carry the risk of sacrificing employment quality.

The model Japan should consider may be Germany's industry-wide collective agreements or Nordic-style solidarity wage systems. In these systems, unions negotiate wage levels across entire industries, resulting in relatively smaller wage gaps between firms of different sizes. However, realizing the preconditions for these systems to function — high union membership rates (Sweden: 70%; Germany: approximately 17%, though collective agreement coverage is approximately 50%) — in Japan would require a fundamental rebuilding of the labor movement itself.

---

Even a 5 percent rise leaves real wages negative

The +5.25% in the 2025 spring offensive is a historically notable achievement. But what this figure reveals is not that "Japan's wages have started to rise." It reveals the severity of a structure in which "even a 5% increase still leaves real wages in the negative."

The structures that produced 30 years of stagnation — profit accumulation as retained earnings, the weakening of labor unions, the expansion of non-regular employment, the deflationary mindset, and the growth of social insurance premiums — cannot be resolved simply by raising spring-offensive wage rates. What is needed is a redesign of the mechanisms that link productivity improvements to wages.

For the structural challenges of the labor market, see "The Structural Transformation of 21 Million Non-Regular Employees（このサイトの記事）"; for the details of social insurance premium burdens, see "The Structure Behind the ¥1.06 Million Wall Reform（このサイトの記事）."

References

2025 Spring Struggle Summary — Japanese Trade Union Confederation (Rengo) (2025)

Trends in Household Tax and Social Insurance Burden Since the Heisei Era — Daiwa Institute of Research (Koeda Shungo and Hiraishi Ryuta) (2025)

Monthly Labour Survey, preliminary results for 2025 — Ministry of Health, Labour and Welfare (2026)

Corporate Enterprise Statistics Survey — Ministry of Finance (2024)

Basic Survey on Labour Unions, FY2024 — Ministry of Health, Labour and Welfare (2024)

Recommended Reading