I. The Region the Concept Was Waiting For
Nine papers have tested the megaregion against conditions it was not designed for: primacy, water, distance, absence, aridity, saturation, archipelago. East Asia is the first region since North America and Europe where the concept fits without adjustment — and it fits at a scale that makes the founding cases look modest.
The Yangtze River Delta, on its official designation, covers Shanghai and three provinces holding roughly 237 million people. The Pearl River Delta and its Greater Bay Area extension holds about 86 million. Beijing–Tianjin–Hebei holds about 110 million. Chengdu–Chongqing holds close to 100 million. Greater Tokyo, at roughly 37 million, is the largest single metropolitan agglomeration on earth, sitting within a Pacific belt of 80 to 100 million. The Seoul Capital Area holds 26.1 million, more than half of South Korea. Taiwan’s western corridor holds roughly 20 million of the island’s 23 million along 350 kilometres.
These are not emergent, contested, or prospective formations. They are consolidated, officially designated in most cases, served by the densest high-speed rail networks ever built, and planned for as units.
Which produces a finding worth stating at the outset, before the details. East Asia requires almost nothing new from the typology. After nine papers and a running total of nineteen binding agents and fifteen morphologies, the region containing the world’s largest and most integrated megaregions adds one of each — and both additions come from its edges rather than its cores: one from the demographic condition of Japan, Korea, and Taiwan, and one from the constitutional peculiarity of Hong Kong and Macau.
The categories in this framework were built by peripheries. The Caribbean, Oceania, MENA, Africa, South Asia, and Southeast Asia supplied nearly all of them. The cores — North America, Europe, East Asia — needed the fewest, because the concept was built from cores in the first place. That is worth knowing about the framework, and it is the sort of thing only a tenth paper can observe.
II. The Span
A. China
China’s roughly 1.41 billion people are about two-thirds urban, and national planning designates nineteen city clusters (chengshiqun) in a formal hierarchy — the only official, national-level megaregional planning framework in the world. Five carry national-strategy status.
The Yangtze River Delta. Shanghai (roughly 25 million) with Jiangsu, Zhejiang, and Anhui: on the official integration designation, roughly 237 million people across about 358,000 square kilometres. The functional core — Shanghai, Suzhou, Wuxi, Changzhou, Nanjing, Hangzhou, Ningbo, Hefei — is a genuine Type A commuter field at distances that would be Type B anywhere else, because Shanghai–Suzhou is twenty-five minutes by high-speed rail. This is, by population and by output, the largest megaregion on earth.
The Greater Bay Area. Guangzhou, Shenzhen, Dongguan, Foshan, Zhuhai, Zhongshan, Huizhou, Jiangmen, Zhaoqing, plus Hong Kong and Macau: roughly 86 to 87 million. It is bound physically by the Hong Kong–Zhuhai–Macau Bridge, the Shenzhen–Zhongshan Link opened in 2024, and the Guangzhou–Shenzhen–Hong Kong express rail — and divided institutionally by three legal systems, three currencies, and three customs territories, all inside one sovereignty. Section IV gives this its own morphology, because nothing like it appears in the preceding nine papers.
Beijing–Tianjin–Hebei (Jing-Jin-Ji). Roughly 110 million, with Beijing at about 22 million and Tianjin at about 14 million, under a coordinated development plan since 2015. Its centrepiece is Xiong’an New Area, established in 2017 across 1,770 square kilometres about 100 kilometres from Beijing, intended to receive functions relieved from the capital. Its progress is instructive and is treated in Section V.
Chengdu–Chongqing. Close to 100 million in the upper Yangtze basin, elevated to national strategy in 2021 — the only interior formation of comparable weight, and the test of whether megaregional development can be produced away from the coast.
The middle Yangtze (Wuhan–Changsha–Nanchang), the Central Plains (Zhengzhou), the Shandong peninsula, Guanzhong (Xi’an), and the Northeast (Shenyang–Dalian–Harbin) complete the principal set. The Northeast is China’s contraction case, losing population steadily to the coastal clusters — the first large Chinese formation to which the shrinkage vocabulary developed in Europe applies.
B. Japan
The Taiheiyō Belt. Gottmann himself identified the Tokaido corridor as a megalopolis alongside his American case. Greater Tokyo at roughly 37 million, Keihanshin (Osaka–Kobe–Kyoto) at roughly 19 million, and Chukyo (Nagoya) at roughly 9 million anchor a belt running from the Kanto plain to northern Kyushu that holds the large majority of Japan’s 123 million.
The explicit policy ambition is to fuse the three into a single labour market. The Chuo Shinkansen is a 438-kilometre maglev line between Tokyo and Osaka designed for 505 km/h, intended to cut Shinagawa–Nagoya to about 40 minutes and Shinagawa–Osaka to about 67 minutes — reductions of between 46 minutes and over an hour against the existing Tokaido Shinkansen, which is ageing and has reached the limits of its frequency and passenger volume.
Its history is a case study in what blocks such projects. Construction began in 2014 with a 2027 target for the Tokyo–Nagoya section. A single 8.9-kilometre stretch in Shizuoka Prefecture — out of 294.5 kilometres — blocked the project from 2017, on concerns that tunnelling would reduce flow in the Oi River, which supplies drinking water to roughly 620,000 people. The prefecture prohibited boring within its territory in late 2019. In January 2026 the governor and JR Central signed a letter of agreement under which the company will compensate for any reduction in Oi River flow regardless of proof of causation; in late March 2026 the prefecture’s expert committee approved all 28 environmental conservation measures; and on 7 July 2026 Governor Yasutomo Suzuki approved resumption of boring, with a formal agreement signed on 18 July. Boring is expected to take about ten years, putting the earliest opening at 2036 or 2037. Costs have risen across two revisions from an original ¥5.5 trillion to roughly ¥11 trillion.
Nine years of delay, on 3 percent of the route, over a water question. It is the clearest illustration in this series that a megaregional project is only as fast as its most reluctant jurisdiction.
And Japan is shrinking. Population peaked around 2008 at roughly 128 million and now stands near 123 million, falling by hundreds of thousands annually, with something on the order of nine million vacant dwellings nationally. Tokyo’s share of the national population continues to rise while the country contracts.
C. South Korea
The Seoul Capital Area. Seoul Special City, Gyeonggi Province, and Incheon hold approximately 26.1 million residents — more than half the national population of about 51.5 million — on roughly a tenth of the land. This is the most extreme concentration of any developed country.
The demography beneath it is without precedent. Total fertility sits at approximately 0.75 in 2026, slightly recovered from the 0.72 low of 2024 but still the lowest in the world; the monthly rate reached 0.99 in January 2026 as births and marriages improved. Deaths have exceeded births for 75 consecutive months since November 2019. National population peaked at 51.7 million in 2024 and is now declining. Median age is about 45.5 years, roughly 19 percent of residents are 65 or older, and the government has spent more than 320 trillion won on demographic policy since 2006 without arresting the decline.
The spatial pattern within this is the striking part. Seoul city proper has been losing population for three decades — from a peak of 10.6 million in 1992 to 9.4 million in 2024, an 11 percent fall driven by young families relocating to more affordable satellite cities in Gyeonggi, with a housing price-to-income ratio around 18 to 1. Meanwhile the capital area as a whole keeps gaining, and provinces including Jeolla and Gangwon have lost population for decades to both natural decline and migration toward Seoul. Sejong, established in 2012 as an administrative capital intended to relieve concentration, has grown quickly and remains small relative to Seoul.
So: a nation shrinking, a capital region growing, and the capital city itself deconcentrating outward within it. Section IV names the mechanism.
D. Taiwan, North Korea, Mongolia
Taiwan’s western corridor runs from Taipei–Keelung (roughly 9 million) through Taoyuan, Hsinchu, Taichung, and Tainan to Kaohsiung — roughly 20 million of 23 million along 350 kilometres served by high-speed rail since 2007, and containing the world’s most concentrated semiconductor manufacturing capacity. Fertility is among the world’s lowest and the population is declining.
North Korea holds roughly 26 million with Pyongyang at about 3 million, and is the most closed formation examined in this series. The Korean Demilitarized Zone is the most heavily militarized severed seam in the world, and Seoul’s northern suburbs sit within roughly 50 kilometres of it. The Kaesong Industrial Complex, operating from 2004 until its closure in 2016, was a designed-complementarity arrangement of exactly the kind the ninth paper identified in Southeast Asia — and its fate is the counter-case to Johor–Singapore.
Mongolia holds roughly 3.5 million with Ulaanbaatar at about 1.6 million: extreme primacy in an isolate condition, with no comparable city within a very long distance in any direction.
III. What Makes Them Cohere
1. High-speed rail at a density that changes the type. China operates over 48,000 kilometres of high-speed line; Japan, Korea, and Taiwan have long-established networks. The effect is not merely faster travel — it converts relationships that would be Type B transactional corridors elsewhere into Type A commuter fields. Beijing–Tianjin is thirty minutes; Shanghai–Suzhou twenty-five; Taipei–Hsinchu half an hour. The 100-to-300-kilometre band, which is a business-travel band in North America and Australia, is a commuting band here. The Chuo Shinkansen’s premise is that the 300-to-500-kilometre band can be converted too.
2. Manufacturing agglomeration at unmatched depth. The Pearl River Delta’s electronics ecosystem, the Yangtze Delta’s machinery and chemicals complex, Japan’s Chukyo automotive cluster, and Taiwan’s semiconductor corridor represent Type C production sharing at densities of specialization no other region approaches. Taiwan’s western corridor in particular concentrates a share of global advanced chip fabrication in 350 kilometres that has no parallel in any industry anywhere.
3. Official designation as a planning act. China’s nineteen city clusters, the Yangtze Delta integration plan, the Greater Bay Area outline plan of 2019, and the Jing-Jin-Ji coordinated development framework are national policy documents with budgets, institutions, and infrastructure programmes attached. Elsewhere, megaregions are analytical constructs that governments ignore; here they are administrative categories that governments fund.
4. Administratively constrained internal migration. China’s household registration system means that a large share of the workforce in the coastal clusters holds registration elsewhere, with consequences for schooling, healthcare, and housing eligibility. Xiong’an illustrates the mechanism in its purest form: most local registrations there are granted to staff of large state-owned enterprises, graduates of top universities, and high-skilled workers in fields including artificial intelligence, biotechnology, and fintech, while those who do not qualify may enter a lottery conferring some local benefits and free public transport that expire after five years.
The eighth paper named Type R — circular internal migration — for India and flagged China as the obvious comparison. The comparison holds, with an important refinement given in Section IV.
5. Absorptive concentration under national decline. Japan, Korea, and Taiwan are all shrinking nationally while their principal megaregions grow. Section IV names this.
6. Grid, water, and airshed systems. North China’s water scarcity produced the South–North Water Transfer Project, among the largest hydraulic works ever built, and the Jing-Jin-Ji airshed drives coordinated emissions control across three provincial-level units. Type O and Type I operate here in their scarcity mode, as in MENA rather than as in Southeast Asia.
7. Seismic and typhoon exposure. Tokyo, Osaka, Taipei, and much of the Chinese coast face hazard at a level that shapes building codes, redundancy planning, and infrastructure siting — the Chuo Shinkansen’s tunnel-heavy design is justified partly on earthquake resilience grounds.
8. State-directed relief cities. Xiong’an for Beijing, Sejong for Seoul, and historically the Japanese capital-function relocation debate: the recurring instrument for a megaregion judged too concentrated is to build somewhere else. Its record is discussed below.
9. Internal borders with external characteristics. The Greater Bay Area’s Hong Kong and Macau components maintain separate customs, currency, immigration, and legal systems within one sovereignty — producing a formation where infrastructure can be compelled by a single state while movement across it remains controlled.
10. Diaspora and cross-strait commercial networks. Taiwanese investment in the Yangtze and Pearl deltas, Hong Kong intermediation, and the historical Japanese and Korean corporate networks throughout the region bind these formations to each other in ways that political relations do not always reflect.
Growth
East Asia is the first region in this series where every major economy is at or heading into population decline. China’s population has begun falling; Japan has been falling since 2008; Korea peaked in 2024; Taiwan has peaked. Only Mongolia and North Korea are still growing appreciably.
Megaregional growth is therefore, throughout the region, growth by absorption: the clusters gain what the interior, the periphery, and the northeast lose. This is the same dynamic the Caribbean and Europe papers identified as contraction, seen from the receiving end — and it is now the condition of a quarter of humanity.
IV. A Typology
Following the discipline of the seventh through ninth papers, this adds one binding agent, one morphology, and one refinement.
Axis One: Binding Agent — One Addition
Type T — Absorptive Concentration. (Labour family.) Cohesion and growth produced by sustained internal migration into a megaregion from a nationally declining population, such that the formation expands solely at the expense of its own country’s periphery, with no aggregate national increase. The megaregion is bound to its hinterland by draining it.
Distinguished from Type F, administrative primacy, which explains why a capital became large but not why it keeps growing while the nation shrinks; from Type R, circular internal migration, because Type T migrants settle permanently and do not return; and from Type E, amenity migration, because the pull is employment, education, and marriage-market access rather than climate or cost.
Cases: the Seoul Capital Area; Greater Tokyo; Taipei and the northern Taiwanese corridor; the Chinese coastal clusters relative to the Northeast and the interior. Prospectively: Milan and Madrid relative to their national peripheries, and Bangkok, as those countries’ declines deepen.
Type T has a distinctive internal signature, visible most clearly in Korea. The core city may itself decline while the region grows, as cost pushes settlement outward faster than in-migration fills the centre — Seoul falling from 10.6 to 9.4 million while its capital area gained. Analysts tracking the core city alone will read a Type T formation as shrinking when it is concentrating.
Type T also has a policy signature: the standard response is to build a relief city or relocate administrative functions, and the record of that response across five regions is now long enough to assess. It is assessed in Section V.
Axis Two: Morphology — One Addition
M16 — Internal Seam. Two or more urbanized areas separated by a customs, currency, and legal boundary that lies inside a single sovereignty, so that infrastructure across it can be compelled by one authority while movement across it remains controlled by two or more regimes.
Distinguished from M5, the open international seam, where two governments must agree and either may refuse; from M12, the severed seam; and from M13, the unbridged dyad, where the barrier is capital rather than law.
Cases: the Greater Bay Area, where Hong Kong and Macau maintain distinct systems within Chinese sovereignty.
M16 completes the seam taxonomy this series has been assembling since the first paper, and it fills the one empty cell. The other three combine two variables — legal permission and physical connection — held by two sovereigns. M16 splits the variables between levels: the connection is orderable from above, the permission is not. The result is visible in the Bay Area’s infrastructure record, which is among the most impressive anywhere, set against a movement regime that remains checkpointed. You can build a bridge by decree and still need a visa to cross it.
A Refinement
Type R now has two modes. The eighth paper defined circular internal migration as voluntary and economically driven, with families remaining at origin and return expected. China supplies the second mode: administratively constrained, where the non-settlement is enforced by a registration system that ties schooling, healthcare, and housing eligibility to a place of origin rather than a place of work.
The distinction matters because the remedies differ entirely. Voluntary-circular migration is addressed by portability — making entitlements travel with the worker, as the eighth paper recommended for India. Administratively constrained migration is addressed by registration reform, which is a different政策 question with different constituencies. Both produce the same surface phenomenon: a labour force enumerated somewhere other than where it works.
The Cross-Classification
| Formation | Binding Agent | Morphology | Consolidation |
|---|---|---|---|
| Yangtze River Delta | A / C / R | M1 | Consolidated; largest on earth |
| Greater Bay Area | C / A / P | M16 | Consolidated; movement checkpointed |
| Jing-Jin-Ji | F / O / P | M1 / M4 | Consolidated; relief city underfilled |
| Chengdu–Chongqing | C / F | M4 | Consolidating |
| Northeast China | D / F | M1 | Contracting |
| Taiheiyō Belt / Tokaido | A / B / C | M1 / M3 | Consolidated; maglev to 2036+ |
| Greater Tokyo | A / T | M2 | Growing within a shrinking nation |
| Seoul Capital Area | T / F | M2 | Growing; core city declining |
| Taiwan western corridor | C / T | M3 | Consolidated; hyper-specialized |
| Korean DMZ | — | M12 | Severed; Kaesong closed 2016 |
| Ulaanbaatar | F | M2 / M9 | Isolate |
| Xiong’an | P | — | Building; 1.41M against a 5M target |
Five observations follow.
First, Type T reverses the framework’s assumptions about what growth means. Every earlier paper treated a growing megaregion as evidence of economic vitality and a shrinking one as evidence of trouble. In Type T conditions both readings are wrong: the megaregion grows because the country is failing to reproduce itself, and its growth is a symptom of the same process that is emptying the provinces. Seoul’s capital area is the most successful megaregion in Korea and the clearest indicator of the national demographic emergency, simultaneously.
Second, M16 demonstrates that the two border variables can be held at different levels of government, which no other case in this series does. It also suggests where else to look: internal customs boundaries elsewhere — Puerto Rico’s relationship to the United States, the Canary Islands’ to Spain, Svalbard’s to Norway — may be weaker members of the same class.
Third, East Asia has the world’s best megaregional infrastructure and its most constrained megaregional labour mobility. Europe’s arrangement binds households and lets the infrastructure lag; China’s builds the infrastructure and constrains the households. These are opposite solutions to the same problem, and the comparison is the single sharpest contrast the series has produced.
Fourth, the Chuo Shinkansen shows that a single jurisdiction can stop a national project for nine years even in a unitary state with a strong planning tradition, and that the resolution came through compensation rather than compulsion — JR Central agreeing to make good any reduction in river flow regardless of proof of causation. That is a transferable model for corridor disputes elsewhere.
Fifth, the Northeast is the world’s largest contracting industrial megaregion, and the shrinkage vocabulary developed for the Ruhr and Silesia applies to a formation several times their size.
V. The Governance Findings
Relief cities: five regions, one verdict
This series has now examined state-built capitals and relief cities in five regions: Brasília and, adjacently, Astana; Egypt’s New Administrative Capital; Naypyidaw; Nusantara; Sejong; and Xiong’an. The record is consistent enough to generalize.
Xiong’an is the best-resourced instance. Established in 2017 across 1,770 square kilometres, it had a permanent population of approximately 1.41 million by March 2026 against a target of 5 million by 2035; roughly 169,000 residents had been relocated by late 2025. More than 400 branches of centrally administered state-owned enterprises have moved there, along with campuses of several Beijing universities and more than 100,000 relocated students; over 4,000 Beijing-origin companies have established operations; the Zhongguancun science park there had attracted over 260 high-technology enterprises by March 2026; and 767 innovative companies have signed agreements to settle. Construction targets from the 2018 master plan have slipped, central Xiong’an remains sparsely populated relative to its built capacity, and the 5 million goal now looks optimistic. Xi Jinping’s March 2026 inspection — his fourth since 2017 — used notably urgent language, calling on officials to throw themselves into implementation. The area is deliberately closed to foreign investment, positioned as an engine for state-owned and domestic firms rather than the export-and-foreign-capital model that built Shenzhen.
Set beside Nusantara — core district nearly complete, budget cut roughly 85 percent, just over a thousand authority employees resident, designation shifted from “national capital” to “political capital” — and Sejong, which grew quickly and remains small relative to Seoul, the pattern is clear:
Relief cities get built. They get populated slowly. Administrative functions move because they can be ordered to. Economic functions do not, because they cannot. The old capital keeps the economy.
The reason is not mysterious. A capital’s concentration is produced by agglomeration in private markets — finance, professional services, headquarters, universities, media — and none of those can be relocated by decree. A government can move ministries and state enterprises, and it does. Everything else stays where the customers, colleagues, and competitors are.
The implication for the regions where this instrument is currently being reached for is direct. Tehran cannot solve water bankruptcy by moving the presidency; Jakarta cannot solve subsidence by moving the parliament; Beijing has not solved congestion by moving four state-enterprise headquarters. In each case the population that constitutes the problem remains where it was.
Designation as a governance instrument
China’s nineteen city clusters constitute the only official national megaregional planning hierarchy in existence, and the framework should ask honestly whether it works.
On the evidence: partially, and where it owns something. The Yangtze Delta Integrated Demonstration Zone, established across the boundary of three provincial-level units in 2019, is a jointly governed area with its own institutions and land — an asset-holding cross-boundary body of exactly the kind this series has repeatedly identified as the durable form. Cross-cluster rail, grid, and airshed coordination have real budgets and real enforcement. Where the designation is only a designation, results are correspondingly thinner.
This is the same finding as everywhere else, which is itself notable: a political system with unusual capacity to compel still gets better results from bodies that hold assets than from bodies that hold plans.
Jurisdictional redraw: the second and third cases
The ninth paper identified Vietnam’s 2025 consolidation of 63 provinces into 34 as the first instance in this series of the remedy the framework kept recommending, and observed that its precondition was a system in which subnational politicians could not refuse.
Japan supplies the democratic version. The Heisei municipal mergers reduced Japan’s municipalities from roughly 3,200 to roughly 1,700 between 1999 and 2010 — a comparable proportional consolidation, achieved not by one legislative act but by a decade of financial inducement, transitional bond authority, and sustained central pressure, with local referendums and considerable resistance along the way.
So the amended finding: jurisdictional redraw is available in democracies too, but it costs a decade and money rather than a session and a vote. That is a materially more encouraging conclusion than the ninth paper reached, and it means the remedy is not confined to single-party systems.
Korea offers the counter-case at megaregional scale. Sejong and the innovation-city programme were an attempt to redistribute functions rather than redraw jurisdictions, and the Seoul Capital Area’s share has continued to rise throughout.
Kaesong and the limits of designed complementarity
The ninth paper introduced Type S — designed complementarity — on the strength of Singapore–Johor, and warned that such arrangements fail unless something physical is built. Kaesong extends the warning. The complex operated from 2004 to 2016, combining South Korean capital and management with North Korean land and labour, with real infrastructure and tens of thousands of workers. It was closed by political decision, and nothing about its economic logic prevented that.
The lesson is that Type S arrangements are as durable as the political relationship that authorized them, and no more. Johor–Singapore has lasted because the relationship has; Kaesong did not because the relationship did not. When the ninth paper called the rail link the test of the Johor arrangement, that was only half right — the infrastructure is necessary, and the relationship is what makes the infrastructure worth having.
VI. Objections
Chinese figures may reflect designation rather than function. The Yangtze Delta’s 237 million is an administrative boundary covering four provincial-level units, including large rural interiors of Anhui that are not integrated into Shanghai’s economy in any meaningful sense. Comparing that number to a commuting-defined figure elsewhere overstates the formation considerably. The functional core is perhaps 100 to 120 million — still the largest in the world, but not by the margin the official number implies.
Type T may be premature. Korea, Japan, and Taiwan have been in this condition for a decade or less, and it is possible that concentration will reverse as the periphery empties and costs at the core become prohibitive. The type describes a current dynamic that may prove transitional.
M16 has one case. A morphology built on Hong Kong and Macau rests on an arrangement with a defined term and a contested trajectory. If the internal boundaries converge with the mainland’s, the morphology loses its only member.
The paper understates political variables. Cross-strait relations, the situation in Hong Kong, and the Korean peninsula are treated here purely as inputs to regional geography. Readers seeking analysis of those questions should look elsewhere; this paper deliberately does not offer it, and that restraint omits factors that plainly bear on how these formations will develop.
North Korea receives a paragraph. Twenty-six million people and the world’s most militarized boundary are treated in passing, for evidentiary reasons that do not fully excuse the omission.
The “cores need no new categories” observation may be circular. If the framework was built from North America and Europe, of course East Asia fits it — East Asian megaregions resemble Western ones because both are dense, industrialized, rail-served, and well-documented. The finding may say more about the framework’s origins than about the world.
VII. Comparative Observations Across Ten Regions
- The typology was built by peripheries. Nineteen binding agents and sixteen morphologies, and the great majority were added by the Caribbean, Oceania, MENA, Africa, South Asia, and Southeast Asia. The three regions with the largest and best-integrated megaregions supplied the fewest. A framework derived from cores describes cores adequately and required substantial extension everywhere else — which is the ordinary fate of exported analytical categories and worth stating plainly at the halfway point of a survey.
- The seam taxonomy is complete. Open (Basel, Johor–Singapore), goods-only (Detroit–Windsor), asymmetric (Dajabón), severed (Wagah, Oujda–Maghnia, the DMZ), unbridged (Kinshasa–Brazzaville), repaired (Fergana), and internal (the Greater Bay Area). Seven configurations, one underlying pair of variables — legal permission and physical connection — and no case in ten regions where geography explained the outcome.
- Relief cities do not relieve. Five regions, six cases, one result: administrative functions move, economic functions stay, and the megaregion that prompted the project retains its population. Any government currently proposing this instrument should be asked what distinguishes its case from Brasília, Naypyidaw, Sejong, Nusantara, and Xiong’an.
- Jurisdictional redraw works and is achievable in democracies. Vietnam did it in a session; Japan did it over a decade with money. Both worked. The framework’s central governance prescription — one general-purpose government covering the formation — is not confined to authoritarian systems, and every jurisdiction that has instead created a coordinating body has less to show for it.
- Infrastructure and mobility are separable, and the region that leads on one may lag on the other. Europe binds households and builds slowly; China builds fast and binds households less. Neither has both. The combination — European mobility rights with East Asian construction capacity — does not exist anywhere, and describing what it would produce is perhaps the most useful counterfactual this series can offer.
- Growth is no longer evidence of health. Type T establishes that a megaregion can grow precisely because its country is failing to reproduce itself. With China, Japan, Korea, Taiwan, most of Europe, Thailand, and increasingly India at or below replacement, this will be the dominant condition of megaregional growth worldwide within a generation, and the framework’s entire apparatus was built to explain growth that meant something else.
- The most consequential 350 kilometres on earth is in Taiwan, and this series has now examined roughly forty megaregional formations without finding another whose disruption would have comparable global effect. Concentration of that depth is a megaregional phenomenon and deserves treatment as one rather than only as a geopolitical one.
VIII. Conclusion
East Asia holds the largest megaregions in the world, plans for them officially, connects them with the best infrastructure ever built at that scale, and is entering population decline across every major economy in the region.
What it contributes to this framework is less a set of new categories than a set of confirmations and one uncomfortable insight. It confirms that designated planning bodies work when they own something and not otherwise, in a political system with unusual power to compel. It confirms that jurisdictional redraw is the effective remedy, and — through Japan — that democracies can accomplish it given a decade and a budget. It confirms, across six cases on three continents, that building a relief capital moves ministries and nothing else.
The uncomfortable insight is Type T. A megaregion that grows while its nation shrinks is not succeeding in the way the literature has always assumed growth indicates. Seoul’s capital area gains because Jeolla and Gangwon lose; Tokyo gains because Japan contracts; the Chinese coastal clusters gain partly because the Northeast empties. The concentration is real, the dynamism is real, and both are being produced by a demographic process that will eventually reach the core as well. Korea’s own projections have Seoul at a fraction of its current size by the end of the century.
The practical program follows. For Type T formations, the agenda is not managing growth but managing the succession of two problems: the periphery’s emptying now, and the core’s own contraction later, which the periphery will not be there to cushion. For M16 seams, the lesson is that infrastructure without movement rights produces impressive bridges and modest integration. For Type R in its administratively constrained mode, registration reform is the lever, and it is a different lever from portability. And for anyone contemplating a relief city, the honest advice after ten papers is to spend the money on the megaregion that exists rather than on the one that has been drawn.
Ten regions in, the single most consistent finding remains the one the first paper stumbled on and every subsequent paper has confirmed: megaregions are made by decisions about boundaries, jurisdictions, and who may cross what. Geography sets the board. It has not once determined the outcome.
