In 1274, a Mongol-Korean invasion force landed at Hakata Bay in northern Kyushu. Japan’s samurai, trained for single combat, were routed by massed formations wielding gunpowder weapons they had never seen. A storm forced the invaders to withdraw, but the lesson was plain: the existing military model would not survive a second attempt. The Kamakura shogunate spent the next seven years preparing. It built a 20km stone seawall along Hakata Bay, developed small-boat night-raiding tactics and established Japan’s first permanent coastal defence force. When a far larger invasion arrived in 1281, the seawall held. The Mongol fleet sat offshore for weeks, unable to land. Then a typhoon struck and finished what Japanese preparation had started.
Western histories remember the “divine wind.” They miss the point. The typhoon destroyed a fleet that was already failing because Japan had spent seven years rebuilding its defences. Fortune favours the prepared.
Japan in 2026 is living through its own inter-invasion period. Chinese carriers circle Okinawa. North Korean missiles fly over the archipelago. Russian bombers probe the Sea of Japan. American military assets are burning in a Middle Eastern war that brings Japan no benefit: an AWACS radar plane destroyed at Prince Sultan Air Base, THAAD radar components hit in the Gulf, KC-135 tankers damaged on the flight line. These are the same categories of equipment that Japan’s ballistic missile defence depends on. The seawall is being built: a ¥9.04 trillion ($60 billion) defence budget, offensive missiles with 1,000km range, constitutional revision to formally recognise the armed forces. The question is the same as in 1274: will the preparation be complete before the next wave arrives?
There is, however, a second question the Mongol crisis did not pose. The Kamakura shogunate collapsed after 1281, and not because of foreign enemies. A defensive war produces no conquered territory, so there was no land with which to reward the warriors who had built and manned the seawall. The shogunate could not restructure its patronage system because the beneficiaries blocked every attempt. Half a century later the regime fell apart from within.
Modern Japan’s binding constraint is not money. One of the world’s largest net creditors, with ¥533 trillion ($3.7 trillion) in net overseas assets as of 2024, can afford a ¥9 trillion defence budget. The constraint is the postwar ecosystem of vested interests that extracts ¥12-15 trillion ($80-100 billion) a year from the Japanese economy. That system costs more to maintain than national defence.
Japan’s postwar bureaucracy built a three-step mechanism that hums along in every ministry, in every prefecture. Regulation creates a captive market. An incorporated association is set up to administer it. Retired officials from the ministry that wrote the regulation are placed on the association’s board. The Japanese call this amakudari, literally “descent from heaven.” The heaven in question is Kasumigaseki, Tokyo’s bureaucratic quarter.
The National Police Agency offers the most vivid case. To get a driving licence in Japan, you must attend a designated driving school at a cost of ¥200,000-350,000 ($1,300-2,300). You can theoretically take the test without attending, but the practical exam pass rate for self-study candidates is set so low as to make this a fiction. The federation overseeing Japan’s 1,200 designated driving schools is chaired by a retired police official. The traffic safety association that collects fees at every licence renewal is headed by a former Commissioner-General of Police. The road traffic information centre, the traffic accident research institute, the Japan Automobile Federation: all are landing pads for retired police bureaucrats. One investigation found that police-linked foundations paid retired officers a cumulative ¥10 billion in severance.
The pachinko industry takes the arrangement further. Gambling for cash is illegal in Japan. Pachinko parlours (a $200 billion industry that once generated more revenue than Las Vegas, Macau and Singapore combined) stay legal through a fiction called the “three-shop system.” The player wins steel balls at the parlour and exchanges them at the counter for “special prizes,” small tokens of negligible value. He then carries these tokens to a buy-back window, typically an unmarked slot in a wall around the corner, and sells them for cash. The buy-back shop sells the tokens to a wholesaler, which sells them back to the parlour. Because three “separate” businesses are involved, the parlour never hands cash directly to the player. Therefore, the logic goes, it is not gambling. Ask a parlour employee where the buy-back window is and he will not tell you. Every regular knows anyway. The police regulate this arrangement and choose not to enforce gambling laws against it. In return, the pachinko industry employs large numbers of retired police officers. The regulator and the regulated share personnel.
The Ministry of Land, Infrastructure, Transport and Tourism is the largest single source of amakudari placements: 911 retired officials sent to the private sector in five years. Each of its eight regional bureaux has an affiliated construction association. One alone, in central Japan, employed 233 former ministry officials and received ¥9.6 billion a year in government contracts. The justice ministry’s Civil Legal Affairs Association received ¥17.4 billion a year for outsourced registry work (84% of its total income) while employing 144 former officials.
What do these people do? “The reality is, they don’t have to do anything,” a former senior education ministry official told Gendai, a Japanese news outlet, in an investigation into amakudari compensation. “Staff on the ground see amakudari officials as incompetent old men getting high salaries for doing nothing.” A public-interest corporation chairman earns up to ¥1.65 million a month (about $130,000 a year). Bank advisers who never come to the office collect over ¥10 million a year. The worst practice is watari (“crossing”), in which a retiree hops from one amakudari post to another every two or three years, collecting a retirement bonus at each stop. A vice-minister receives about ¥63.4 million ($410,000) in severance on leaving government, before starting the circuit.
The most sophisticated form runs through international institutions. The Ministry of Finance pays Japan’s contributions to the IMF and sends retired officials to staff it. The IMF then warns, on cue, that Japan’s government debt has reached 236% of GDP. That figure counts only liabilities. Subtract the government’s financial assets (the ¥259 trillion ($1.7 trillion) GPIF pension fund, foreign exchange reserves, loans to public corporations) and net debt falls to about 134% of GDP, unremarkable by G7 standards. The IMF’s own public-sector balance-sheet data showed Japan’s public sector in net asset-positive territory (¥48 trillion, or 9% of GDP) as of 2020. That dataset attracts little attention. Robin Brooks, former chief economist of the Institute of International Finance, put it bluntly: “The reason this hasn’t been done yet is that there’s vested interests that want to keep managing the various financial assets Japan’s government owns.” The loop is circular: the finance ministry provides data to the IMF; the IMF publishes alarming gross-debt headlines; the ministry cites the IMF as independent corroboration; the fiscal-crisis narrative justifies consumption-tax increases; the tax revenue funds the public corporations where ministry retirees sit. Showing only one side of a balance sheet has a name in accounting. It is not a kind one.
Vehicle inspection crystallises the absurdity. Japan makes the world’s most reliable cars but requires every vehicle to pass a test costing ¥100,000-200,000 every two years. The actual fee at a government centre is about ¥28,000; the rest is the garage’s markup and unnecessary “recommended” repairs. With 62 million registered vehicles, the inspection industry extracts ¥1-2 trillion a year. Financial-planner examinations run by a body with deep finance ministry ties, boat-licence courses at designated schools, the technical intern programme that imports cheap foreign labour through amakudari-staffed intermediaries: the pattern repeats across every sector. Total the figures and you reach ¥12-15 trillion a year: ¥1.6 trillion in grants to 86 independent administrative agencies, ¥6 trillion in public works, ¥2.3 trillion in agricultural subsidies, ¥1-2 trillion in contracts to public-interest corporations, ¥2-3 trillion in indirect costs imposed on citizens by regulation-created monopolies. The sum is larger than the defence budget. The cost of maintaining the postwar system exceeds the cost of defending the country.
Why, then, might this time be different? The Kamakura shogunate failed to reform after 1281 because the external pressure was temporary. The Mongols planned a third invasion but never mounted it. When the threat receded, the case for change evaporated. The pressures bearing down on Japan in 2026 are not temporary. China deployed two aircraft carriers near Iwo Jima simultaneously in June 2025. North Korea continues missile tests. Russia maintains military deployments on the disputed Kuril Islands. Three nuclear-armed neighbours pressing Japan at the same time is without precedent in the postwar era.
America’s deterrent is being consumed in the Middle East. The loss of an AWACS, THAAD radars, tankers and drones in the Iran war degrades the same capabilities Japan’s missile defence relies on. Japan’s security establishment is watching irreplaceable American equipment burn in a war from which Japan gains nothing. The energy shock exposes Japan’s alliance burden: the closure of the Strait of Hormuz cost Japan roughly a fifth of its LNG supply and a third of its crude. The Philippines bought 2.48 million barrels of Russian oil and received emergency fuel from China. Vietnam signed a nuclear deal with Moscow. Japan, bound by its alliance commitments, cannot take any of these exits. It pays the highest cost of any ally and has the fewest escape routes.
The yen sits at multi-decade lows on the Bank for International Settlements’ real effective exchange rate index. The BOJ has begun hiking. Thirty-year JGB yields above 3.5% meet life insurers’ needs for the first time in two decades. Institutional capital is coming home. Yet Japanese retail investors are simultaneously sending roughly ¥1 trillion ($7 billion) a month abroad through NISA tax-free accounts into unhedged foreign equity funds. In the year to April 2025, an S&P 500 index fund returned -0.15% in yen as a 9.1% yen appreciation ate the entire equity gain. When that pattern holds long enough, the retail flow will reverse. None of these pressures will resolve on their own. There is no moment at which the vested interests can say “the crisis has passed.” That is what distinguishes 2026 from 1281.
Seawall and sword
Prime Minister Sanae Takaichi won 316 seats in the February 2026 election, the largest ruling-party majority in postwar history. 93% of winning candidates support constitutional revision. The LDP-Ishin coalition established an Article 9 review committee in November 2025.
What does constitutional revision actually mean? Not what Beijing and Moscow claim. Japan has maintained a de facto military since 1954. Its navy is larger than Britain’s or France’s. The constitution says armed forces “will never be maintained” while the country plainly maintains them. The legal fiction causes real damage: no military courts, ambiguous rules of engagement, perpetual political controversy over offensive capability, and, in a shrinking labour market, difficulty recruiting because constitutional ambiguity undermines the prestige of service. Carnegie’s James Brown of Temple University offers the sober assessment: Takaichi is “making reasonable defensive reforms to adapt Japan to a deteriorating security situation,” not returning to the 1930s.
The defence budget hit ¥9.04 trillion, the 12th consecutive record. The 2% of GDP target was reached two years ahead of schedule. Missiles with 1,000km+ range are being deployed. Defence exports have been liberalised; SIPRI reports that Japan’s top five defence firms saw a 40% revenue increase in 2024. The government resumed domestic LNG carrier construction after a seven-year gap, to reduce dependence on Chinese shipyards. Corporate governance reform, driven by the Tokyo Stock Exchange’s push on price-to-book improvement, has produced buybacks, cross-shareholding unwinds and a 25% rise in the Topix in 2025.
These are the visible reforms. The invisible part has barely moved. Vehicle inspections still cost six times the actual fee. Driving schools still hold their monopoly. Construction associations still employ hundreds of former ministry officials. The finance ministry still shows only one side of the balance sheet to the IMF.
The test is not whether Article 9 gets revised. That path is far from clear: Takaichi holds two-thirds of the lower house, but the LDP-Ishin coalition commands only 119 of the upper house’s 248 seats, well short of the 166 needed to put a constitutional amendment to referendum. The next upper house election is in 2028.
The test is whether the ¥12-15 trillion apparatus gets restructured. Demographics sharpen the question: Japan’s working-age population has fallen 16% since 1995 and is projected to shrink a further 31% by 2060. A declining workforce cannot sustain both a ¥9 trillion defence budget and a ¥12-15 trillion patronage system alongside ballooning social security costs. If Takaichi uses her supermajority for the constitution and defence alone while leaving the domestic extraction system untouched, Japan risks the Kamakura pattern: a country that arms itself against external threats but decays from within.
Japan’s postwar history is littered with half-finished reform. The Hashimoto reorganisation of 1996-98 reshuffled ministries but left amakudari intact. Koizumi focused on postal privatisation. The Democratic Party’s “project screening” in 2010 briefly exposed wasteful spending; many of the programmes that were cut reappeared under new names. Each attempt stalled because external pressure was weak enough for vested interests to outlast the reformers. The external pressure in 2026 is of a different order. But pressure alone does not produce reform. The Kamakura shogunate faced the largest external threat in Japanese history and still failed to restructure internally.
Where the chisel falls
The Japan that most foreign investors carry in their heads (pacifist, deflationary, zero-rate, export-dependent, ageing quietly) is the Japan of 2012-2023. That Japan is ending. Monetary policy is normalising. The military is seeking constitutional recognition. Corporate governance is improving. The defence industry is entering the export market.
If reform reaches the ¥12-15 trillion layer, the capital released will be reallocated domestically. If the finance ministry starts showing both sides of the balance sheet to international bodies, the change in narrative alone would affect JGB spreads and the yen’s valuation. The yen’s direction has already turned: BOJ hikes, institutional repatriation and carry-trade unwinds all point toward strength. In 2025 the Topix rose 25% while the most popular all-country foreign equity fund was flat in yen. The same FX move that disappoints Japanese retail creates opportunity for foreign capital. The yen at multi-decade lows prices none of this.
Japan’s reformations have always been forced, resisted and underestimated. In 1853, American warships under Commodore Matthew Perry sailed into Edo Bay and demanded that Japan open its ports, ending two centuries of isolation. The shogunate’s critics dismissed the threat as temporary. Within 15 years Japan had a new government, a modern army and an industrialisation programme. The oil shocks of the 1970s were supposed to destroy Japanese industry; instead they produced the world’s most efficient manufacturers. The investor who sees only the weak yen and the energy shock is making the error that outside observers have made at every turning point in Japan’s modern history. But past reformations have not always succeeded. The honest position is to hold both the optimism and the warning: watch whether Takaichi spends those 316 lower-house seats on the constitutional fight alone, or turns them toward the domestic extraction system. If she is still battling for upper-house votes in 2028 while the ¥15 trillion apparatus hums along untouched, the Kamakura pattern will have repeated itself.