Section XXVI: XXVIThe Air-Naval Transition

There is a moment in the development of every continental land power when economic dominance reaches the threshold at which it must either be projected globally through air-naval capacity or remain bounded within the continental theater. This is the air-naval transition, the moment that has destroyed nearly every continental power that has attempted it, and the moment Beijing is in.

The pattern is consistent across the past four centuries. Britain became globally dominant by becoming a naval power in the seventeenth and eighteenth centuries, and was able to sustain the position because it was structurally an island state whose productive base was always going to require maritime extension. Wilhelmine Germany attempted the transition before its productive base could sustain a direct challenge to the dominant naval power, triggered a balance-of-power coalition against itself, and was destroyed. Imperial Japan attempted the transition with a productive base too small to sustain global power projection, was structurally constrained by imported energy and raw materials, and was destroyed. The Soviet Union attempted to match American capability across all military domains under conditions where the productive base could not sustain the cost, exhausted the productive base in the attempt, and was destroyed. The United States made the transition successfully starting in 1898 by becoming functionally something other than a continental land power. The US dominated through global air-naval power for 80 years after 1945 and is now exhausting the productive base that supported the projection.

On the seats

Germany in 1914 was the previously ascendant power rather than the rising one, arrested by competitors who saw the writing on the wall, making the now-or-never bid against encirclement and then making it again to self-destruction, which is Japan in 1941 and Russia since 2022, the cornered power lunging against slow strangulation. The confident rising productive hegemon was the United States: isolationist, biding, entering late and decisively, inheriting without a military bid, which is China’s seat now. The violent declining incumbent before the rupture was Britain, the Boer War, the naval race, the imperial enforcement, which is the United States at Venezuela, Iran, and Hormuz. The roles map across with one refinement: in this repetition the military violence is concentrated in the declining powers, the incumbent strangling at the chokepoints and Russia lunging in Ukraine, while the rising power declines the bid. And the repetition sharpens the only question rather than softening it, because the United States in the rising seat was bourgeois and the arc continued. Whether China in that seat is bourgeois or not is the whole of it.

The mapping assigns each state one seat, and the record assigns more than one. Germany in 1914 held two: it was rising against Britain and peaking against Russia, whose rearmament was due to finish in 1917, and the now-or-never bid was made against the second clock. China holds two and the outline of a third. In form it is the Britain of the second transition, the mixed formation carrying the new mode inside the old one, which is Section XXI’s claim. On the ledger it is the America of 1914, the base already moved and the creditor biding. In strategic geography it has Germany’s outline, a continental power building a fleet against the maritime incumbent with a treaty ring along its sea frontier, and little of Germany’s condition. Germany and Austria were ringed on land by hostile great powers. China’s land frontiers run along Russia, Pakistan, Central Asia, and North Korea, with India the exception behind the Himalaya, and pipelines and rail cross them that no navy can stop. The Entente was three great powers with their own armies and treasuries. The ring from Japan to Australia is one power and its dependents, and it stands for as long as the hub does. Germany never out-launched Britain, and the successor out-builds the incumbent many times over. The solvent removes the Schlieffen option from both sides. What remains of the third seat is a test of form. A war in Germany’s seat would have to be chosen, and the pressure to choose it comes from the ledger, the overcapacity that finds in arms the one buyer needing no realization. The refusal is the variable, and Taiwan inside the window is where it is tested.

The incumbent holds more than one seat as well, and neither is Britain’s of 1910. Edwardian Britain was the world’s creditor, its overseas portfolio a third of its national wealth, its yards launching most of the world’s tonnage, its budget near balance, and its partners in the Entente were great powers with armies of their own. It reached the present incumbent’s condition only after 1918, when the portfolio had been sold. The United States enters its contest as the world’s largest debtor, out-built more than 200 to one in yard capacity, its forward base in the Gulf wrecked by a regional power in the first days of the war, and its partners in the Pacific are dependents that rely on it for their defense. The nearer seats are two. On the ledger it is the Dutch Republic after 1780, the financial center whose capital has nowhere productive to go. In its fisc it is Bourbon France: a military power that spent itself contesting the rising center in every theater, servicing a debt that takes a third of its income-tax receipts, with a privileged estate it cannot tax. For the manner of its ending Section XVII offers a third seat, the late Soviet Union, and the two agree on the point that matters here: neither was beaten in the field. The Bourbon monarchy’s last war was a victory, in America, and the bill for it called the Estates-General. Section XXVIII is that seat’s sequel: a state inside the doom loop cannot pay for its own recuperation.³¹³

The objection that seats China in Germany’s chair has authors, and this work’s rule is to name them. Hal Brands and Michael Beckley argue in Danger Zone (2022) that China is a peaking power, its workforce shrinking and its growth slowing, and that peaking powers lunge while they still can. The argument fails on both legs. The slowing growth is the world market’s, the falling rate Section XXIV records arriving in Shenzhen as it arrived in Santa Clara, and a formation that redirects its surplus by plan can move out from under it. The shrinking workforce is being answered on the factory floor. China installed 54% of the world’s new industrial robots in 2024, more than 2 million now run there, its robot density passed Germany’s and Japan’s the year before, and a formation that does not need the machine to pay for itself in profit can set it against a labor shortage where the incumbent cannot. The two readings also predict different tapes. A peaking power raises military pressure as its window narrows. PLA sorties into Taiwan’s air defense zone in 2026 have run at less than half their 2024 and 2025 level, in the year the incumbent’s deployable carriers were committed to the Gulf, and the seat test in Section XXIX lets the next four years score it.³¹⁴

On the anomaly

The transition underway has a name inside the framework this work runs on, and the name marks an anomaly. Arrighi wrote in 1994, in the Epilogue of The Long Twentieth Century, that the bifurcation between American military power and East Asian economic power visible since the 1980s had “no precedent in the annals of capitalist history,” and he and Silver restated it in 1999 in Chaos and Governance in the Modern World System: in every prior hegemonic transition, financial and military power had migrated together to the rising center.³¹⁵ Amsterdam, London, and New York each financed the coalitions and fielded the decisive force, so the crucible’s verdict and the ledger’s verdict arrived as one. The present transition split them. Productive and financial ascendancy moved to East Asia while military supremacy stayed with the declining hegemon, and they flagged the split as the reason this transition could run more violently than its precedents, because a declining hegemon with intact military primacy can resist the adjustment Britain could not.

Arrighi’s later formulation in Adam Smith in Beijing caught the configuration exactly: a power no longer hegemonic but still militarily preeminent, holding leverage through a balance of terror that links its economic policies to its competitors and financiers. What this section documents is that anomaly closing, by a route the framework did not list. The primacy stays where it was and loses its purchase, because the capacity to build and to replace has moved to the productive side of the ledger. The Edwardian precedent sharpens the point by inversion. Britain in its autumn retained exactly one productive capability at world scale, shipbuilding, which is why it could still out-launch Wilhelmine Germany nearly two to one in the dreadnought race even as the rest of its industrial base migrated to its rivals. The United States retained the inverse. Its autumn capability is the reserve currency and the carrier fleet the currency financed, and the naval race has accordingly inverted: on the incumbent’s own naval intelligence estimate the successor’s yards hold more than 200 times its building capacity.³¹⁶ The one thing Britain kept was the thing its transition was fought with. The one thing America kept is the thing its transition is priced in.

On the mode of war

The closure has an agency history, and the history is the socialist movement’s material core told as a weapons ledger. The asymmetry opened in 1492 was closed three times, each time at a lower price and by a more general method. The first closure was parity by plan: the Soviet Union met the core’s best army at Stalingrad and Kursk with an industrial base built in a decade by the instrument this work’s second half examines, and the price was a revolution, three five-year plans, and 27 million dead. The second closure was denial by sacrifice: Vietnam defeated the core itself with the periphery’s bodies, and the price was measured in millions again. The third closure is running now, and its two operators change the theorem. Ukraine has stalemated Russia and Iran has stalemated America, neither of them a socialist state, both of them running the same solution: cheap precision in quantity, drones and missiles built from commodity electronics and solid fuel, the denial capability that once required a planned economy or a people’s war now available as a manufacturing recipe.

At the asymmetry’s height the periphery could buy the weapon and never the production system behind it. Ka’iana carried muskets back across the Pacific in 1788, and Kamehameha unified the islands with Western guns and Western gunners, and none of it slowed the flag that followed, because the asymmetry lived in the factories and never in the barrels. The recipe now travels with the factory attached, and the deep source of the components is the industrial base of the formation the gunboats once opened. The two theaters are not symmetric, and the difference completes the theorem rather than weakening it. Both run it on the successor’s components, as Russia does. Ukraine also needs what only the incumbent makes, the interceptors and the shells, and that magazine is exhausted, so it may yet lose. Iran needs nothing the incumbent makes and may yet win, with the memorandum’s terms already priced accordingly.

The strait has seen the pairing before. Hormuz fell in 1622 to a Persian army carried on the English company’s ships, the regional power with the rising sea power behind it. This time the rising power sends the factory and keeps its ships at home. The recipe is general and the hinterland is particular. Denial has been democratized; sustained denial still belongs to whoever holds the production system, which returns the closure to the formation this work centers, its arsenal rather than its operator.

Five centuries of the asymmetry ended without a rival to inherit it. The French naval school of the 1880s, the Jeune École, said as much of the battleship, and was right about the vulnerability and wrong about the conclusion. The carrier may have a future. It belongs to whoever can replace one, the monopoly on that has closed, and the rest of this section counts yards for that reason. The 11-carrier monopoly is the asymmetry’s terminal form, the largest concentration of the old capability ever assembled, priced by the cost-asymmetry flip of Section X against the new one, kept out of the strait it was built to hold open.

The Loire perfected, set out with the transitions, is the model the present war is testing again, and the roles are the ones the lineage predicts: China refuses the Wilhelmine path, Russia takes the Japanese one, and the United States takes the strangling incumbent’s, which is the one the pattern says loses last. The incumbent fights the plan Sedan taught and Halder executed: the strait declared clear inside a fortnight of February 28, the campaign declared won, and the second phase formed anyway. The regeneration it met had been rehearsed: after Israel’s 12-day campaign of June 2025 struck a third of Iran’s launchers by its fourth day, some 200 were rebuilt within eight months and the missile stock stood at roughly twice its June 2025 level by February 2026 on Israeli estimates.

The adversaries fight the Loire with the institution built: Russia’s 30,000 recruits a month and 7 million rounds a year against 400,000 in 2021. Ukraine’s 200,000 drones a month. And on the incumbent’s side a Patriot line that delivered 620 of its newest interceptor in 2025, with a seven-year plan to reach 2,000, against a war whose first three days fired about 800 Patriots on Kyiv’s count and 1,000 of every variant in 40 days.³¹⁷ In 1870 the levée was improvised and lost. In 1941 it was a standing institution and won. Since 2022 it has moved into the factory, and the belligerent that institutionalized production in peacetime is the one positioned to win the attrition that follows the opening blow, whoever lands the blow.

The rule Section IX states, that the crucible goes to whoever regenerates, has a limit, and the third transition holds it. Germany in 1914 had institutionalized production better than any belligerent in Europe and lost on reproduction: the blockade cut the nitrates, the fodder, and the fats, the turnip winter of 1916 followed, and the home front dissolved in 1918 with the army still on foreign soil. Regeneration decides a crucible only for the formation whose own reproduction the enemy cannot reach. The Hormuz closure ran that test on the successor before any blockade was declared. Gulf crude was about a third of China’s supply, the state drew down a stockpile near a hundred days of import cover and declined to bid, and an electricity system built on domestic coal, wind, and sun was never exposed. A closure of every strait at once is untested for the successor and for everyone else, the incumbent first among them, since it would cut the incumbent off from the manufactures, the pharmaceutical inputs, and the rare-earth magnets this section has already counted in its arsenal. The successor’s exposure is in calories, hydrocarbons, and ore. The incumbent’s is in everything made from them, which is what Section XI means when it says the two-oceans buffer fails for a net consumer of a substrate it has set on fire.

The wars of 2022 to 2026 have superseded the combined-arms breakthrough of Isserson, the Soviet theorist of deep operations, only in part and on conditions. The new center of gravity is the contest for sight: the transparent battlefield is contested, and the side that can see and deny sight decides where the deep operation is possible again. The error to avoid is the Jeune École’s, which saw the torpedo boat sink the battleship and concluded the battleship was finished. The commentators who pronounce the tank dead and the drone supreme are making that error, and the Loire perfected is the corrective: the decisive institution is the one that regenerates what the transparent battlefield consumes.

The best practitioner of the successor’s mode of war fights for the incumbent’s coalition, and the fact is evidence rather than irony. Ukraine turned the drone into the artillery of a nation that had lost its artillery: more than 4 million airframes in 2025 from some 500 makers, 70 to 80% of the casualties on both sides by the estimates of Western officials and Ukrainian commanders, a deep-strike campaign that reached roughly a third of Russia’s large refineries, the transparent battlefield mastered by the belligerent with the least money.³¹⁸ The airframes are built from Chinese components, the mode is the levée in the factory that the challenger’s productive base can run and the incumbent’s cannot, and the Western contribution to the masterclass was the expensive systems that ran out, on the interceptor arithmetic the Loire paragraph gives. The masterclass is pinned to the wrong side, and that is the clearest evidence that the mode is open to any alliance, and the production system behind it decides. By September the diffusion had names on every continent: Shahed technology in North Korea by way of Russia, a Turkish maker assembling one-way drones in Nigeria, Poland building its own at a spending level near $7 billion, and Japan putting drones at the head of a record defense request.³¹⁹ In September the patron asked it to stop. With American diesel past $6 a gallon the president told Kyiv to leave Russian diesel alone and announced a truce neither capital had agreed to, and Ukraine struck the Syzran refinery the same night.³²⁰

The challenger’s version is the peacetime line that never stopped: the shipyards counted above, the missile and drone plants that Iran and Russia draw their components from, an industrial base that regenerates at war’s tempo without a war.

On the constituencies

Eisenhower, in his 1961 farewell address, named the structural risk that this trajectory produces. The military-industrial complex, once built, creates institutional concentrations whose interests favor its continuation and expansion. Defense contractors, military procurement bureaucracies, the political classes whose careers depend on the continuation of military spending, the regional economies built around military bases: all of these become constituencies for the continuation and expansion past the point where the system is producing strategic value. The constituencies eventually capture the political process. The American case played out approximately as he described it, on a longer timescale than he may have anticipated.

The Soviet Union failed at this with particular brutality. The Brezhnev-era military buildup, from approximately 1965 through the early 1980s, expanded the Soviet defense burden to roughly 15 to 20% of GDP,³²¹ a scale of military commitment that the productive base could not sustain. The military-industrial complex absorbed resources that would otherwise have gone to consumer goods, agriculture, and the technological development of the civilian sector. By the time Gorbachev attempted reform, the productive base had been hollowed by the military commitment, the military-industrial complex had captured the political framework’s capacity to direct it, and reform produced the destruction. The Soviet failure was a textbook Eisenhower outcome.

China is attempting the transition with several structural advantages the prior cases did not have. The productive base holds a share of world manufacturing near 30%, integrated into the world’s supply chains as the Soviet base never was.³²² The transition is being undertaken during a period of American imperial overextension rather than against an American capability at its peak. The civilian-military fusion that the CCP has institutionalized through commercial port operations, COSCO’s global shipping presence, the Belt and Road Initiative infrastructure backbone, and the dual-use technology architecture provides extension capacity short of bases, without requiring the dedicated military-logistics build-out that the USSR could not sustain. The shipbuilding industrial base has reached the point where one Chinese shipbuilder launched more commercial tonnage in 2024 than the entire American industry has built since 1945.³²³ The fifth-generation aircraft trajectory, with the J-20 operational and the J-35 carrier-deployed on the Fujian, is constructed on a domestic supply chain that has substantially reduced the foreign-component vulnerability the Soviet aerospace sector struggled with.

The carrier count reads the same way from the incumbent’s side. The Nimitz was to leave service in May 2026 and take the fleet from 11 to 10, below the floor the statute sets. In March the Navy extended it to March 2027 because the Ford-class replacement was late, and the ship spent July tied up at Staten Island as the centerpiece of a naval review while two carriers fought the war, two sat in the Pacific, and six were in maintenance. The Ford came home from the Gulf after a 326-day deployment, the longest on record. The Lincoln is at sea behind it: out of San Diego on November 21, 2025 for seven months, past 300 days by the end of September with no relief announced and no port call in more than 200 of them, the longest run at sea on record, its acting secretary writing to Congress that eight sailors of its strike group had attempted suicide on the deployment, and the president saying of the deployment that it was not nearly long enough.³²⁴ The statutory 11 is a number kept by not decommissioning a ship, and the deployable fleet in the fourth transition’s war was four.³²⁵

The institutional-analytical class in the United States has begun to acknowledge the inversion. In early May 2026, Senators Ted Budd and Jeanne Shaheen, Republican and Democrat, both members of the Senate Armed Services Committee, published a bipartisan op-ed naming the air-superiority gap. The Chinese sixth-generation J-36 and J-50 first flew in 2024. The American F-47 is not expected to fly until 2028 and will not enter operational service until the mid-2030s. The Air Force fighter fleet has fallen from 4,100 aircraft in 1990 to 2,000 in 2024, with only 1,300 combat-coded, a record low. The fighter fleet averages 27 years old at a 50% operational rate. KC-135 tankers are over 60 years old. China’s defense spending has grown thirteenfold across 30 years. Chinese fighter production is projected to outpace American production by nearly 200% by 2027. The senators’ proposed remedies, multiyear procurement authorization and a 50% increase in fighter fleets by 2035 and retention bonuses for pilots, would address the symptoms without addressing the cause, which is the productive base from which the fleets are constructed and the political-economic conditions under which the productive base has hollowed.³²⁶

The defense-industrial base’s dependence on Chinese-processed inputs sits underneath the production-rate gap. The F-35 program depends on Chinese-processed rare earth elements for the permanent magnet motors that drive its actuator and propulsion systems. Each Virginia-class submarine requires approximately 9,200 pounds of rare earth materials, on a Defense Department estimate the Congressional Research Service recorded in 2013. The Tomahawk’s fin actuators run on samarium-cobalt magnets, and the magnets are made in Pennsylvania and New York from samarium that is processed almost nowhere but China: when Beijing restricted the metal in April 2025, the contractors were carried on a cache of samarium nitrate that had sat in a French factory since the 1970s. The interagency report of September 2018 on the defense industrial base, published unclassified with a classified annex, found that China’s industrial policies had already eliminated the American capability to process rare earth elements, called China a significant and growing risk to the supply of strategic materials, and left the department’s own recommendations in a classified action plan. The Government Accountability Office counted about $439 million awarded from 2020 to 2024 to rebuild domestic rare earth supply chains, against the department’s own assessment that new domestic sources can take a decade or longer, and outside estimates run to 15 to 20 years for the full chain. The arsenal of democracy cannot produce its own arsenals without Chinese inputs.³²⁷ The week the kingdom asked for interceptors the State Department cleared 48 F-35s for it at $24.3 billion, years from delivery, with magnets processed in the successor’s refineries. Four days before the September visit the president of the Peterson Institute printed the conclusion as advice: Beijing holds escalation dominance, the tariff threats were folly, decoupling under two administrations made the chains longer and darker, and the incumbent should now do what lesser powers have done toward it for 80 years, negotiate from weakness, import and stockpile from China for as long as it can under multiyear purchase agreements paid for in export-control relief, and invite Chinese plants onto its soil as the modern hostage exchange.³²⁸

The demonstration was run in 2025, before the war. By April 10 Washington’s tariff on Chinese goods stood at 145%; on April 4 Beijing put seven of the 17 rare earths and the magnets made from them under license and answered nothing else. Magnet exports fell by half in April and by half again in May, to 1,238 tons, the lowest since 2020, and the 46 tons that reached the United States that month were a tenth of the year before; Ford stopped the Explorer line in Chicago for a week for want of them. The tariff came down to 30% at Geneva on May 12 with the licenses still closed, and the licenses opened after London in June for the price of the incumbent’s own countermeasures, the design-software ban, the ethane licenses and the engine parts. Six weeks against 40 years of tariff talk: the valve beat the tariff because a tariff is paid at the border and a magnet is missing from the line.³²⁹ When Washington answered in October with a rule reaching every firm the sanctioned owned, Beijing answered with a rule reaching every product that carried a trace of its rare earths, and both were suspended for a year at Busan on October 30. The license desk is what remained of the exchange, and it is the instrument the war has now tested.³³⁰ The embargo the dependence invites has not come, and the form the restraint took tells more than an embargo would have. Beijing licensed rather than cut. The April 2025 licensing regime on the seven heavy elements and the magnets made from them stayed in force through the war, and the extraterritorial controls announced in October 2025 were suspended to November 10, 2026 and the American-specific ones to November 27. The flow was set by the license desk, uneven by month and by buyer, with European imports rebounding and American imports never returning to their 2024 level: 17 tons of yttrium shipped to the United States in the eight months after April against 333 in the eight months before.³³¹ On the other side of the valve the buyer has forbidden itself to accept what comes through it: the procurement rule reaching, with exceptions, the Chinese mining and processing behind the chain’s magnets, tantalum and tungsten takes effect on January 1, 2027, against a supplier that separates 90% of the world’s rare-earth oxides and nearly all of its samarium metal. A valve turned by license, priced by the month, against a buyer barred by its own rule from the source that holds nearly all of it, is a slower instrument than an embargo and a more exact one, and its two expiry dates fall in the weeks after the midterms, on the outgoing Congress. This work’s claim is that the license desk persists past those dates, renewed or replaced, and that no embargo comes. An embargo would count against the refusal thesis. The yuan rail is the same instrument in money, capacity built and held idle as a deterrent, and traffic on it would count the same way.

Plekhanov noted in 1885 that a backward country can at first, until its home market is glutted, shut out the insuperable competition of more advanced neighbours with a customs system. The incumbent’s tariffs are that instrument in the hands of the country that once wrote the rules against it, and their use is a confession of which side of the comparison it now stands on.³³²

The controls leak at their price. In August prosecutors in Taiwan indicted nine people, among them a manager at the chip seller’s Taiwan office, for routing 74 of 130 of its newest servers to buyers in China through Indonesia, Japan and Hong Kong after a staged audit of a data hall that could not have held them, servers that change hands in China for about $1 million apiece, close to double the American price. Exporting the chips to China was not itself a crime under Taiwanese law, and the charges were forgery and breach of trust.³³³

The Eisenhower risk is real anyway. The Central Military Commission’s authority over the military-industrial complex has tightened under Xi’s consolidation, both of its vice chairmen expelled within a year, He Weidong in October 2025 and Zhang Youxia, Xi’s closest military ally, on September 21, 2026, three days before the state visit and weeks before the October plenum, with the chief of the Joint Staff Department beside him.³³⁴ The political-economic resources flowing into the air-naval transition are creating constituencies inside the Party-state whose interests are the continuation and expansion of the transition. This is the dynamic Eisenhower warned about, the dynamic that the Soviet system could not contain, the dynamic that Xi is purging inside the Chinese military system through the anti-corruption campaign that has reached deep into the PLA leadership. Whether the party can contain it is another live experiment of phase two, and a purge cannot show it, since a purge is an event and capture is a drift. The series that can is the one this section names below, the share of the surplus routed to arms against the share routed to the household.

The bet Beijing is making is that the air-naval transition can be conducted with limits, the way the protracted-war stalemate-to-counter-offensive is being conducted with limits. The naval expansion is calibrated to deny American Pacific operations and support Chinese maritime claims and global trade. The carrier program is the place to watch the calibration. Three are in service and a fourth is building, and the Pentagon’s report of December 2025 assesses six more by 2035, for nine, which is inside the band of 8 to 12 that global projection takes. That is the incumbent’s count of the successor’s intentions, made by a department that needs the threat funded, and Beijing has confirmed none of it.³³⁵ It is also the count that would prove this section wrong. The strategic doctrine remains anti-access and area denial rather than offensive sea control. This is the calibration that, if maintained, fits the post-shift cost-asymmetry environment described earlier. China is building the defensive deterrent layer that supports the economic-diplomatic track without trying to substitute it with global military power projection. The hulls this work opened on are the calibration’s test: a third carrier and a drone carrier fit a denial fleet, and a sixth carrier with a network of declared bases abroad would be the security conundrum in the successor’s own yards. The one combat test of the successor’s aircraft and missiles against a Western fighter so far came in Pakistan’s hands in May 2025, and it went the successor’s way on the account of American officials, a missile fired from about 200 kilometers against pilots who had misjudged its reach.³³⁶

The G7 imperial core is still trying to maintain the calibration that worked in the pre-shift environment, projecting expensive offensive platforms forward against asymmetric defenders, and failing at it.

This section opened on three failure models, the Wilhelmine, the Japanese and the Soviet. The closure of Hormuz ran the Japanese test on the successor’s energy before any bid was made, as the paragraphs on the Loire record, and the two that remain are the Wilhelmine and the Soviet. The path between them is to build air-naval power on a productive base that exceeds the dominant power’s, calibrate the transition to regional capability with global influence rather than global power projection, conduct it during the dominant power’s overextension rather than at its peak, and use the institutional memory of both Soviet ossification and Cultural Revolution disruption to prevent the military-industrial constituencies from politically capturing the transition.

The Wilhelmine failure mode has an economics, and the falling rate supplies it. Overcapacity with no external market seeks demand in arms, because the state is the one buyer whose purchases need no realization: the shell is consumed by the state that ordered it, the steel and the chips are absorbed, and the rate of profit is restored by fiat on the order book. That is the war economy as the recomposition’s exit, the one the closed export route makes attractive, and it is the mechanism by which the military-industrial constituencies this section fears would capture the transition without anyone deciding that they should. The Taiwan buildup read as military Keynesianism is that exit in its opening phase, and the test of whether it is being taken is the share of the surplus the plan routes to arms against the share it routes to the household.

Taiwan is where the refusal is tested, and the mode of war this section describes dictates the form. An opposed crossing of a transparent strait is the combined-arms breakthrough at its least feasible, so the available operation is the bypass applied to an island: a quarantine of a polity that imports 96% of its energy and holds about 11 days of gas when the gas is arriving. Since February it has arrived only at a premium. Qatar supplied a third of the island’s gas and most of its helium before Ras Laffan went offline, and the island has held its 11-day floor, as Section XI records, by outbidding Europe and Karachi for spot cargoes, with no act from Beijing at all. The clocks are unequal by an order of magnitude. The island’s runs in weeks, the successor’s in months on the Hormuz evidence, and the incumbent cannot win a siege race on the island’s clock. It can break in, which means winning the contest for sight inside the launcher envelope, against sensors the successor’s deterrent shares. It can blockade from a distance, which the magazine and the ledger both forbid. Or it can accept. The successor’s interest is to present no discrete act at all and let the island’s price be set the way the strait’s was, by accretion: an arms package held in review before a summit, an energy guarantee offered in exchange for sovereignty, the patron’s credibility marked down with each month of the Gulf war. By September the accretion had a year’s tape: no large-scale exercise around the island since the autumn of 2025, while the incumbent fought in the Gulf; a $14 billion package on the president’s desk unapproved after he had called arms sales a bargaining chip with Beijing and the island a war 9,500 miles away; the opposition’s chair received in Beijing in April, weeks before the president was; and outbound investment to the mainland under a tenth of the island’s total from four fifths in 2010.³³⁷ The discrete act is the one thing that could put China in Germany’s seat, which is why the incumbent’s planners have invited it since 2017, and why the refusal is the variable.³³⁸

Whether the equipment is sufficient is what the transition will test. The CCP is better-prepared than any prior continental land power that attempted the air-naval transition, which is the most that can be said before the test plays itself out. The argument that this preparation is sufficient is conditional on Beijing maintaining the calibration of regional capability rather than global power projection. The argument that it is not sufficient, that the military-industrial constituencies will capture the Party-state on a trajectory the Soviet experience prefigures, is one I can’t rule out and the reader shouldn’t either.