The economy of Japan: thirty years of deflation and the strange recovery after it

What broke in the first place

The Japanese economy of the 1980s ended in a deliberate policy decision. Trying to deflate speculation and hold down inflation, the Bank of Japan raised inter-bank lending rates sharply in late 1989. The bubble burst, the stock market crashed, equity and asset prices fell, and banks and insurers that had lent heavily against them were left holding bad debt.

What followed had the shape of a banking crisis and the duration of an era. Bank credit growth stagnated, institutions were bailed out with capital injections, and prices began to fall. Under deflation the value of cash rises, so Japanese firms cut wages, cut research and development, cut investment, and held cash instead. That behaviour is rational for each firm and disastrous in aggregate, and it ran for a generation.

It took twelve years for Japanese GDP to return to its 1995 level. Per capita output tells the same story more sharply: nominal GDP per head was 44,210 dollars in 1995 and 34,713 dollars in 2025. The stock market took until February 2024 to regain its 1989 level.

The corporate consequences were as visible as the macroeconomic ones. Japanese firms that had dominated their industries from the 1960s through the 1990s spent the 2000s fending off competitors from South Korea and China, and the conspicuous consumption of the 1980s never came back.

What was tried

Japan has run a fiscal deficit since 1991, and the accumulated stimulus is the largest reason the debt is what it is.

Monetary policy went further than anywhere else had. The Bank of Japan introduced Quantitative and Qualitative Monetary Easing in 2013 and a negative policy rate of −0.1% in 2016, which together produced mild inflation of around 0 to 1% by the late 2010s. Structural reform was attempted alongside it from 2012 under a programme of monetary, fiscal and structural arrows.

What finally ended the deflation was not domestic policy. The global inflation surge from 2021 to 2023 pushed Japanese inflation above 2%, and wage growth followed. A country that had spent thirty years trying to create inflation got it from outside.

Where the economy is now

The current position is unusual enough to be worth setting out precisely, because most of the individual numbers look benign and the combination does not.

Inflation has eased through 2026 to roughly half the 2025 average, and nearly every core measure sits below the central bank's 2% target. On the usual reading that would argue against tightening. The Bank of Japan raised rates anyway, and is expected to go further.

Three things sit underneath the headline number and all of them push the other way. Government subsidies on fuel, utilities, childcare and education are holding measured inflation down, to the point that energy prices in June 2026 were 0.1% below a year earlier. Meanwhile the producer price index was up more than 7% and import prices up 29.7% in yen terms, because the yen fell more than 11% over the year. Coordinated intervention by the Japanese finance ministry and the United States Treasury pushed the currency up sharply at the start of August 2026, and a higher policy rate is the tool that would keep it there.

So the tightening is a currency decision presented as an inflation decision, and the underlying price pressure it is aimed at is real even though the published figure does not show it yet.

Why the debt does not behave like a crisis

Japanese government debt has been above 200% of GDP for years, reaching around 240% by 2013 and standing at about 207% in 2025, with the International Monetary Fund expecting it to fall to 193% by 2031.

That is the largest debt burden of any developed economy and it has not produced a crisis, for reasons worth understanding rather than dismissing.

Net interest payments came to just 0.2% of GDP in 2025, projected to reach 1.7% by 2031. The comparison that makes the point is the United States, whose net interest payments were 3.7% of GDP in 2025 on a far smaller debt ratio. Japan has been able to carry an enormous stock of debt because it has been paying almost nothing to service it, and because most of that debt is held domestically.

The conditions are shifting. The ten-year government bond yield reached 2.88% on 9 July 2026, the highest since 1996. Most of that rise reflects inflation rather than a risk premium, and the inflation-adjusted yield on the five-year bond was still negative in June 2026, which means the government continues to borrow at a real cost below zero. A debt crisis in the next few years is improbable. A steadily rising interest bill on 200% of GDP is close to certain.

Fiscal policy is meanwhile expanding rather than contracting. Fuel and utility subsidies, free high school tuition and free elementary school lunches all support consumption, food items are due to become tax-exempt for two years, and military spending rose more than 13% in 2025.

What is actually driving growth

Two things, and one of them is new.

The first is the familiar one. A weak yen makes Japanese exports cheap, manufacturers' sentiment has been strong on the back of it, and export gains have carried the economy through an energy price shock that a net energy importer would normally feel severely.

The second is semiconductors and the demand behind them. More than a third of year-on-year real GDP growth in the first quarter of 2026 came from electronic components and devices alone, and integrated circuit exports were up 74% year on year in May 2026, against growth of 6.1% across 2025. Investment in artificial intelligence elsewhere in the world is showing up in Japanese export figures, because Japan makes the components and the equipment that make the chips.

Domestic demand has also recovered. The real consumer activity index rose 2.7% in May 2026, the strongest in more than three years, with some of that pulled forward ahead of rate rises but the growth broad-based across categories.

What the economy is made of

The postwar structure is still recognisable in the current one.

Manufacturing built the growth of the 1950s to the 1980s: vehicles, electronics, machine tools, precision instruments, steel and chemicals, sold overseas by firms organised for export and supported by close cooperation between government and industry. Japan became the largest manufacturer of automobiles in the world and a leading producer of electronics during that period, and several of those firms still lead their categories.

What changed was where the value sits. Japanese companies that dominated consumer electronics from the 1960s through the 1990s lost that ground to South Korean and Chinese competitors through the 2000s, and the response was to move upstream. The strength now is in components, materials, and the machines that make other people's products: semiconductor manufacturing equipment, specialty chemicals, optical and precision parts, the industrial robots that populate other countries' factories.

That repositioning is the reason the artificial intelligence boom shows up in Japanese export data at all. A country that no longer sells the finished device sells the equipment and the components that the device requires, and demand for those is less visible and considerably harder to displace.

Services employ more people than manufacturing, as in every developed economy, and tourism has become a substantial earner in its own right since the borders reopened. The travel article goes into what that has meant for the places absorbing the visitors.

Why wages stayed flat for thirty years

The most consequential feature of the Japanese labour market is not unemployment, which has been low throughout, but the price of labour, which did not move for a generation.

Lifetime employment underpinned the postwar system: firms retained a loyal and experienced workforce with the guarantee of a job for a career, and workers accepted seniority-based pay in exchange for security. That arrangement held through the growth decades and largely collapsed after 1989, when firms cutting costs under deflation turned to non-regular contracts instead.

The consequence was a two-tier workforce and a wage level that stopped rising. With prices flat or falling, holding wages down cost employers nothing in real terms and cost workers little in immediate purchasing power, so the situation persisted for decades without producing the pressure that would normally end it.

The break came from the outside. Imported inflation from 2021 gave workers a real-terms pay cut for the first time in years, and a labour market with too few workers in it gave them the standing to demand more. Nominal wages have grown above 3% for four consecutive months, against an average of 2.2% across 2025, and real wages turned positive at 1.6% in May 2026 after falling throughout the previous year.

Whether that holds is the open question in Japanese economics. It depends on whether a shrinking workforce keeps the labour market tight enough to sustain it, which is the one variable nobody has to forecast, because the people who will be of working age in twenty years have already been born.

What the demography does to all of it

Every figure above sits on a population that is falling and ageing fast.

Japan's population peaked at 128.1 million in October 2008. The total fertility rate has been below the replacement level of 2.1 since 1974, reached a low of 1.26 in 2005, and stood at 1.41 in 2016. Life expectancy was 85.1 years in 2016, at 81.7 for men and 88.5 for women.

The age structure follows arithmetically. The share of the population aged 65 and over rose from 25.9% in 2014 to 29.1% in 2022, and projections on current fertility put it at 40% by 2060, with the total population falling from 128 million in 2010 to about 87 million by 2060.

What that means for the economy is contested at the level of policy and not at the level of arithmetic. A shrinking working-age population raises questions about the workforce, about potential growth and about the solvency of the pension and healthcare systems, and the government has responded with measures to raise fertility and to keep older people economically active.

It also puts the current numbers in a different light. Non-farm employment grew 0.5% year on year in June 2026, which is unremarkable anywhere else and notable in a country with a shrinking labour force. Unemployment at 2.5% is not slack absorbed but a labour market with nobody left in it, which is why wages are finally rising after thirty years in which they did not. The tightest labour market in the developed world is a demographic fact before it is an economic achievement.