I. The Question Oceania Answers
Each region in this series has tested the megaregion concept against a different structural condition. North America supplied the case the concept was built for. South America tested it against primacy. The Caribbean tested it against water and small size. Oceania tests it against something the first three papers never fully isolated: distance without obstacle.
Australia’s cities are not separated by mountains, forests, borders, or sea. They are separated by nothing at all — by eight hundred kilometers of open, flat, legally unified, English-speaking, single-currency country in which almost no one lives. Sydney and Melbourne are farther apart than Boston and Washington, with roughly the same combined population and roughly a hundredth of the settlement between them. Nothing prevents that space from filling. It has simply never filled, and after two centuries it is not going to.
This produces the purest available instance of a formation the North American paper had to invent a category for: the transactional corridor, where two or more centers are economically joined at the level of firms, capital, and executive labor while sharing no daily labor market whatsoever. In Texas the type was mixed with freight and energy. In Australia it is nearly pure, and the binding agent is almost entirely a single air route.
Oceania also supplies, in its Pacific island states, a stronger version of the Caribbean’s central finding — with one decisive difference. The Caribbean has a legal right to free movement that almost nobody exercises. The Pacific has quota-limited, deliberately designed, government-to-government mobility schemes that are heavily oversubscribed. Comparing them isolates what actually governs circulation between a small state and a large neighbor, and the answer is not law.
And Oceania contains one thing no other region in this series does: a formal international treaty contemplating the eventual relocation of an entire national population, on climate grounds, with a ballot already run twice.
II. The Span
A. Australia
Approaching 28 million people on a continent of 7.7 million square kilometers, with roughly two-thirds of the population in the capital cities of the eastern and southwestern coastal fringes. Australia is among the most urbanized countries on earth and among the least densely settled, which is the whole story in one sentence.
The Southeast Corridor (Sydney–Canberra–Melbourne). The principal candidate. Greater Sydney and Greater Melbourne are each in the range of 5.3 to 5.6 million, and which is larger depends on the boundary convention: on the Significant Urban Area measure Melbourne is larger, at roughly 5.33 million against Sydney’s 5.22 million, while on the Greater Capital City Statistical Area measure — built primarily around capital city labour markets — Sydney remains larger, chiefly because Greater Sydney includes the Central Coast and the Sydney urban area does not. Adding Canberra (roughly 475,000), Newcastle, the Central Coast, Wollongong, Geelong, and Albury–Wodonga yields something in the range of 13 to 14 million — about half of Australia — spread along an axis of roughly 900 kilometers with essentially nothing in its middle.
The Sydney Basin extended (the former “Six Cities Region”). Sydney with the Central Coast, the Lower Hunter and Newcastle, and the Illawarra–Shoalhaven: roughly 7 million within about 200 kilometers of coast, and, for a period, the only formation anywhere in the world legislated as a megaregion by name. Section V treats what happened to it.
South East Queensland. Brisbane (roughly 2.8 million), the Gold Coast, the Sunshine Coast, Ipswich, Logan, Moreton Bay, and Toowoomba: a genuinely coalescent conurbation of roughly 4 million along 200 kilometers, and the clearest Type A case on the continent. Brisbane’s 2.7 percent annual growth in 2023–24 puts it among the fastest-growing large urban regions in the developed world, and the 2032 Olympic Games are functioning as a forcing device for infrastructure sequencing across three tiers of government.
Perth–Peel. Roughly 2.4 million, with the highest capital city growth rate in the country at 3.1 percent in 2023–24, and no other city of comparable size within about 2,100 kilometers. Perth is the most geographically isolated large city on the planet, and the isolation is not incidental — it determines the city’s function, which is discussed under morphology below.
Adelaide (roughly 1.5 million), Hobart, Darwin, and the resource regions of the Pilbara, the Bowen Basin, and the Goldfields — the last of these being formations of real economic weight and minimal resident population, for reasons taken up in Section III.
B. New Zealand
Roughly 5.4 million, with a single candidate formation.
The Upper North Island “Golden Triangle.” Auckland (roughly 1.7 million), Hamilton and the Waikato (roughly 200,000 urban within a larger regional catchment), and Tauranga with the Bay of Plenty (roughly 165,000): about 2.5 million, or close to half the national population, within a triangle of roughly 250 kilometers on its longest side. Ports of Auckland and the Port of Tauranga together handle the great majority of the country’s container freight, and the completed Waikato Expressway together with the Hamilton–Auckland passenger rail service has, for the first time, made the Auckland–Hamilton relationship a partial commuter question rather than purely a freight one.
This is a formation of two and a half million people. By the standards of the first paper in this series it would not register. By the standards of its own national system it is overwhelming: it holds half the country, nearly all its immigration, and the entire upper end of its service economy.
Wellington–Manawatū (roughly 600,000) and Christchurch–Canterbury (roughly 700,000) complete the national urban system without approaching megaregional weight.
C. Melanesia — and the Instructive Absence
Papua New Guinea holds roughly 11 to 12 million people, making it the second most populous country in Oceania and larger than any Caribbean state. It has no megaregion, and it has no metropolis. Port Moresby is around 400,000, Lae around 120,000, and the country remains roughly 85 percent rural. The Highlands Highway is the only corridor of national consequence, and it is a supply line rather than a settlement axis.
PNG is the most useful negative case in this entire series. It demonstrates that megaregional formation is not a function of population, nor of land area, nor of resource endowment — PNG has all three in abundance. It is a function of whether urbanization occurred, and urbanization in PNG has been suppressed by land tenure arrangements, terrain, linguistic fragmentation across some 800 languages, and the enclave structure of the extractive economy, which delivers revenue without delivering cities.
Fiji (roughly 925,000) contains the Suva–Nausori corridor and the Nadi–Lautoka western axis, and functions as the region’s institutional hub: the Pacific Islands Forum secretariat, the University of the South Pacific’s main campus, and the principal international air gateway. Solomon Islands (roughly 800,000) and Vanuatu (roughly 330,000) complete Melanesia.
D. Polynesia, Micronesia, and the Territories
Roughly 2.5 million across some twenty jurisdictions if Hawaii is included, under 1.2 million if it is not. Samoa (roughly 220,000), Tonga (roughly 105,000), Kiribati (roughly 130,000), New Caledonia (roughly 270,000), French Polynesia (roughly 280,000), Guam (roughly 170,000), the Federated States of Micronesia, the Marshall Islands, Palau, Nauru, and Tuvalu (roughly 11,000).
Individually these are below any threshold. Collectively they raise the same question the Eastern Caribbean did, and they answer it more sharply, because several of them have already crossed a line the Caribbean has only approached: a majority of the nation lives somewhere else. The Cook Islands has perhaps 15,000 residents and several times that number of Cook Islanders in New Zealand and Australia. Niue has fewer than 2,000 residents and roughly 30,000 Niueans in New Zealand. Tokelau’s resident population is around 1,500. All three are in free association with New Zealand, and their people hold New Zealand citizenship, which is why the distribution looks the way it does.
E. The Diaspora and the Rim
Auckland is the largest Polynesian city in the world. Sydney, Brisbane, and Melbourne hold large and growing Pacific populations. Honolulu, the U.S. West Coast, Utah, and Springdale, Arkansas — the last through Marshallese migration under the Compacts of Free Association — hold substantial Micronesian and Polynesian communities. As in the Caribbean, the functional region includes cities outside the region, and as in the Caribbean, the transfers run inward while the command functions sit outward.
The difference is that Oceania’s outward links are channelized by treaty and by quota, and that changes their character entirely.
III. What Makes Them Cohere
1. A single air route. Sydney–Melbourne is consistently among the busiest city pairs on earth by passenger volume and has been for decades. It is not a commuter route; it is a business-day route, flown by people who will return the same evening. This one connection, together with Sydney–Brisbane and Melbourne–Brisbane, does the work that rail and road do in the Northeast Corridor. Australian megaregional integration is aviation integration, and it is unusually concentrated: a small number of city pairs carry the great majority of the interaction.
2. Corporate and regulatory centralization on two cities. Financial services, professional services, media, and head offices cluster in Sydney and Melbourne to a degree that makes the rest of the continent, in economic terms, a set of branch operations and resource provinces. Canberra adds the federal regulatory function as a third node. The result is a Type B corridor with a Type F overlay — the South American administrative-primate pattern, split across two capitals instead of one.
3. Rostered air commuting to extraction sites. This deserves separate treatment because nothing quite like it appears in the earlier papers. Australia’s iron ore, gas, and coal regions are worked substantially by fly-in fly-out labor: workers resident in Perth, Brisbane, Cairns, or Adelaide who fly 1,500 to 2,000 kilometers to a camp, work a two-and-two or eight-and-six roster, and fly home. The Pilbara generates a very large share of national export revenue with a modest permanent population, because its labor shed is a city on the other side of the state.
This is a commuter field with a two-week period and an aviation link instead of a daily period and a road. It satisfies the functional definition of metropolitan integration — the same people, the same employers, the same household budgets — while violating every spatial assumption behind it. Section IV gives it a type.
4. Housing cost gradients and interstate migration. Housing affordability in Sydney and Melbourne, where median prices reach income multiples comparable to Hong Kong and London, has driven internal migration toward Queensland, South Australia, and regional areas. Capital cities collectively lost 34,600 people to internal migration in 2023–24 even while growing by 427,800 overall — the growth was overseas migration, and the internal flow ran outward. South East Queensland is the principal beneficiary, which is why Australia’s fastest-consolidating megaregion is not one of its two largest.
5. Immigration as the near-sole growth engine. Of the 427,800 people added to Australia’s capitals in 2023–24, overseas migration supplied 373,000 and natural increase 89,500. Australian and New Zealand megaregional growth is, to a first approximation, a policy variable set in Canberra and Wellington. This is a materially different growth mechanism from anything in the Americas, and it makes these formations unusually responsive to political cycles.
6. Institutionalized circular labor mobility. The Pacific’s binding mechanism to Australia and New Zealand is not informal migration but designed programs. Australia’s PALM scheme allows employers to recruit workers from Fiji, Kiribati, Nauru, Papua New Guinea, Samoa, Solomon Islands, Timor-Leste, Tonga, Tuvalu, and Vanuatu for seasonal placements of up to nine months or longer-term placements of one to four years, and the government is working toward integrating New Caledonia with an initial cohort of up to 100 workers over two years. Roughly 32,000 people were working in Australia under the scheme as of late 2025, down from a peak of 34,830 in September 2023, with the long-term stream steady while short-term recruitment declined as employers reverted to backpacker labor for seasonal work. New Zealand’s Recognised Seasonal Employer scheme has operated on similar principles since 2007.
For Tonga, Samoa, and Vanuatu, the resulting remittance flows are among the highest in the world as a share of national output. This is the Caribbean’s Type H binding agent, but administered rather than emergent — with quotas, employer approval, country caps, and bilateral agreements, and with the corresponding vulnerability that a program can be reduced by a decision no Pacific government participates in.
7. Permanent migration pathways, balloted. Alongside temporary work, both countries operate lottery-based permanent residence routes for Pacific nationals — New Zealand’s Pacific Access Category of long standing, and Australia’s newer Pacific Engagement Visa, whose 2026–27 ballot for Pacific island and Timor-Leste nationals ran from 1 July to 29 July 2026. These convert a portion of circulation into settlement, and they are the mechanism by which Auckland became the largest Polynesian city in the world.
8. Climate exposure, and the first treaty response. The Caribbean paper identified shared hazard as a binding agent and flagged that it might, past a threshold, become a dissolving agent. In Oceania that threshold has been crossed for the atoll states, and there is now an instrument. The Australia–Tuvalu Falepili Union treaty, signed in 2023 and in force since August 2024, created a pathway for up to 280 Tuvaluans annually to receive permanent residence in Australia, with the right to work, study, and access health, education, and family support. Demand has been extraordinary: the first ballot opened on 16 June 2025 and attracted 3,125 applicants in its first four days; by the close, 8,750 people — more than half of Tuvalu’s citizens — had applied. The full allocation of 280 places was granted for 2025–26 and Australia proceeded to a second ballot, which opened on 1 May 2026 and ran to 1 June, after more than 100 Tuvaluans had relocated under the first round.
The treaty’s other limb is what makes it a genuinely new instrument. Its security provision requires that any partnership, arrangement, or engagement Tuvalu enters with another state or entity on security and defence matters be agreed with Australia — the migration pathway is the consideration on the other side of that bargain. Concerns about the effect on Tuvalu itself are already being raised: Australia’s own development plan for Tuvalu counts visas delivered, settlement experiences, and increases in remittances among its indicators, without comparable targets for retaining Tuvaluan health workers, teachers, engineers, or civil servants, or for measuring whether essential services deteriorate after each intake.
9. Shared institutions built on shared resources. The Parties to the Nauru Agreement — eight small states controlling the waters where much of the world’s skipjack tuna is caught — operate a pooled vessel-day scheme that sets a collective cap on fishing effort and allocates tradeable days among members. It has raised the members’ share of the fishery’s value by a large multiple. Alongside it sit the Pacific Islands Forum, the Pacific Community, the Forum Fisheries Agency, and the University of the South Pacific, which is jointly owned by twelve countries. Section V returns to what these have in common.
10. Currency by adoption, not by pooling. Nauru, Kiribati, and Tuvalu use the Australian dollar; the Cook Islands, Niue, and Tokelau use the New Zealand dollar; Palau, the Federated States of Micronesia, and the Marshall Islands use the U.S. dollar; the French territories use the CFP franc. Monetary integration in Oceania is thus as deep as the Eastern Caribbean’s in effect and entirely different in structure: it is adoption of a large neighbor’s currency rather than creation of a shared one. The islands get monetary stability and no monetary voice. The Eastern Caribbean states pooled and got both.
Growth
Australia and New Zealand grow through immigration and would otherwise be near stationary. Melanesia grows through natural increase and is young and rural. Polynesia and Micronesia are broadly flat or shrinking at home while growing abroad. The Pacific is therefore the second region in this series, after the Caribbean, where the resident populations of several formations are declining while the national populations, diaspora included, are not.
IV. A Typology
The axes remain binding agent and morphology. One binding agent and one morphology are added; several established types apply with unusual clarity.
Axis One: Binding Agent
Type A — Commuter-Field. South East Queensland; Auckland and its immediate ring; Perth–Peel; the Sydney Basin extended. Australia’s genuine coalescent cases, all of them intra-state.
Type B — Transactional Corridor. Sydney–Canberra–Melbourne, in the purest form this type takes anywhere. The Auckland–Hamilton–Tauranga triangle, partially transitioning toward Type A.
Type C — Production-Sharing. Absent. Australia and New Zealand have deeply integrated trade and labor markets with each other under long-standing arrangements, but the trans-Tasman relationship produces no shared manufacturing geography; it produces migration and services trade instead.
Type D — Extractive-Logistical. The Pilbara, the Bowen Basin, the Goldfields, PNG’s mining and LNG enclaves, New Caledonian nickel, and Nauru’s exhausted phosphate. Economically dominant in Western Australia and Queensland.
Type E — Amenity-Migration. The Gold Coast and Sunshine Coast, coastal New South Wales, and the Bay of Plenty. A real and growing driver of Australian internal migration, and the reason South East Queensland consolidates faster than the corridor that feeds it.
Type F — Administrative-Primate. Canberra as federal node; Wellington; every Pacific island capital, where the capital typically holds a fifth to a half of the national population.
Type G — Maritime-Networked. The island Pacific, over distances that dwarf the Caribbean’s. Fiji and Nadi function as the region’s switching point in the way San Juan and Miami do for the Caribbean, but the interstitial distances are an order of magnitude greater, and the route networks correspondingly thinner and more fragile.
Type H — Remittance-Diaspora. Tonga, Samoa, Tuvalu, Kiribati, and the free-association states above all. Distinguished from the Caribbean version by being administered: the flows exist because programs exist.
Type I — Shared-Hazard. Cyclones, seismic risk, and — uniquely in this series — sea level rise as an existential rather than episodic threat. In the Caribbean, hazard produces regional institutions. In Oceania it also produces treaties contemplating relocation.
Type J — Regulatory-Arbitrage. Present but thinner than in the Caribbean: some offshore finance, the Marshall Islands ship registry, and Vanuatu’s citizenship program.
Type K — Rostered-Commute Extraction. (New.) Cohesion produced by long-distance air commuting on multi-week rosters between a residential metropolis and an extraction site with no meaningful permanent population. The labor shed is discontiguous by design and the site is deliberately not developed as settlement. Wages, household spending, and demographic weight all accrue to a city hundreds or thousands of kilometers from where the value is produced. Cases: Perth–Pilbara; Brisbane and Cairns to the Bowen Basin; parts of the PNG resource sector.
Type K is worth adding retrospectively as well. Alberta’s oil sands camps and the Chilean Atacama’s shift-work mining follow the same logic, and both papers treated them as ordinary Type D. They are not: a Type D region has a resident population that grows with the industry, while a Type K region is engineered so that it does not. The policy questions differ accordingly — Type D regions need housing, schools, and hospitals, and Type K regions are structured precisely to avoid providing them.
Axis Two: Morphology
M2 — Primate-Radial. Auckland; Perth; Wellington; every Pacific capital.
M3 — Linear Corridor. The Australian east coast in aggregate; the Suva–Nausori axis.
M4 — Constellational. Sydney–Canberra–Melbourne; the Upper North Island triangle.
M7 — Discontiguous Archipelago. The island Pacific, at oceanic scale.
M8 — Extraterritorial-Cored. The free-association states, in a form more complete than anything in the Caribbean: for the Cook Islands, Niue, and Tokelau, a clear majority of the national population resides in the metropole, holds its citizenship, and returns seasonally or not at all.
M9 — Isolate. (New.) A metropolitan formation with no comparable city within roughly a thousand kilometers, which must therefore supply the full range of urban functions internally — teaching hospitals, universities, capital markets, air freight, professional services, cultural infrastructure — at a scale that would ordinarily be shared across several centers. Cases: Perth; Auckland; Honolulu; Port Moresby in a degraded form. M9 formations are more self-sufficient and more expensive than their size would predict, and they are unusually vulnerable to the loss of any single function, because there is no neighbor to fall back on.
The Cross-Classification
| Formation | Binding Agent | Morphology | Consolidation |
|---|---|---|---|
| Sydney–Canberra–Melbourne | B / F | M4 | Consolidated, air-bound |
| Sydney Basin extended | A | M3 | Consolidated; governance repealed |
| South East Queensland | A / E | M3 | Consolidating fastest |
| Perth–Peel | A / F | M2 / M9 | Consolidating |
| Pilbara / Bowen Basin | K | M8 (labor) | Functional, near-unpopulated |
| Upper North Island triangle | B → A | M4 / M9 | Consolidating |
| Suva–Nausori / western Viti Levu | F / G | M3 | Modest, regional hub |
| Papua New Guinea | D / F | — | No megaregion; not forming |
| Polynesia / Micronesia | G / H / I | M7 / M8 | Resident populations flat or falling |
| Cook Islands, Niue, Tokelau | H | M8 | Majority resident abroad |
Four observations follow.
First, every consolidated Australian formation sits inside a single state. South East Queensland is in Queensland; the Sydney Basin is in New South Wales; Perth–Peel is in Western Australia. The one formation that crosses a state line — Sydney–Melbourne — has no coordinating body, no shared plan, and no rail project that has survived three consecutive governments. This reproduces the North American finding about Ontario’s Golden Horseshoe exactly.
Second, Type K should change how extraction regions are assessed generally. If the labor is rostered in from elsewhere, then the region’s economic output and its demographic weight are located in different places, and standard regional accounting will attribute prosperity to a hinterland that receives very little of it.
Third, M9 isolation is a cost structure, not a curiosity. Perth and Auckland must each carry a full urban functional stack for a population that in Europe or North America would share one across four cities. This partly explains why both are expensive, and why both are more exposed than their size suggests to the loss of a single air link or institution.
Fourth, the free-association states have completed the trajectory the Caribbean is on. When a majority of a nation lives in the metropole, the question of whether the diaspora “counts” as part of the region resolves itself: it is the region, and the islands are its origin point and its periodic destination.
V. The Governance Findings
Oceania both confirms the Caribbean’s governance result and sharpens it.
The confirmation
The deepest functional pooling in the region is among its smallest states. The vessel-day scheme lets eight small countries act as a single seller into a global tuna market and capture a far larger share of the resource rent than any of them could alone. Twelve countries jointly own a university because none could sustain one. Nine or more use another country’s currency because none can run a central bank. In every case the driver is the same: the minimum efficient scale of the institution exceeds the state, so the state cannot supply it unilaterally and therefore pools.
The sharpening
Australia supplies the control condition, and it is damning. New South Wales did something no other jurisdiction in this series has done: it legislated a megaregion by name. The Greater Cities Commission Act 2022 replaced the Greater Sydney Region with a defined Six Cities Region comprising the Eastern Harbour City, Central River City, Central Coast City, Lower Hunter and Greater Newcastle City, Western Parkland City, and Illawarra-Shoalhaven City, and amended the planning act to require district strategic plans within it to include targets for net additional dwellings. In June 2023 the incoming government announced that the commission’s staff would be folded back into the planning department, and the Greater Cities Commission Repeal Act 2023 repealed the 2022 Act with effect from 1 January 2024, striking the definition of the Six Cities Region from the planning legislation.
A statutory megaregion was created and abolished inside twenty months, without any change in the underlying functional geography — the seven million people and their overlapping labor markets are exactly where they were. What changed was the government.
The lesson is not that the idea was bad. It is that a megaregional body attached to no asset and no revenue stream is a discretionary line item, and discretionary line items do not survive elections. Compare the durable cases across all four papers: the Port Authority of New York and New Jersey, the Itaipú binational entity, the Eastern Caribbean Central Bank, the vessel-day scheme. Every one of them owns something, sells something, or issues something. The Greater Cities Commission wrote plans. Plans are repealable in a way that a harbor, a dam, a currency, and a fishing quota are not.
Australia’s federal structure explains why the failure took this form. Each state can supply planning, transport, health, and education at its own scale, so each does, and nothing at the corridor scale is unavailable to any of them. There is no function that Victoria and New South Wales cannot each perform badly by themselves — which, on the Caribbean’s finding, is precisely the condition under which cooperation does not occur.
New Zealand ran the opposite experiment and it held: the 2010 amalgamation of Auckland’s councils into a single unitary authority created a body coextensive with a metropolitan labor market, and it has persisted through changes of government because it is the general-purpose government of a place, not a coordinating overlay on top of several.
VI. Objections
Australia may not have a megaregion at all. Nine hundred kilometers of near-empty country between two cities is a great deal to ask a category to absorb. If Type B can accommodate Sydney and Melbourne, it can accommodate almost any two cities with a busy air route, at which point the concept has stopped constraining anything. The response — that firm networks, capital allocation, executive labor, and regulatory relationships are measurably interlocked in ways they are not between, say, Sydney and Auckland — is a real one, but the boundary is soft.
The Pacific formations are national systems, not megaregions. A country of 105,000 people has an urban system; calling it a component of a megaregional network is an analytical convenience. The defense is comparative rather than intrinsic: placing them on the same axes surfaces Type K, M9, and the completed M8 trajectory, none of which would emerge from treating each as a separate national case.
The diaspora accounting is even more contestable here than in the Caribbean. For the free-association states it may be the only sensible accounting. For Tonga and Samoa it is defensible. For Fiji and PNG it is not, and the paper should not be read as claiming otherwise.
PALM and RSE may be doing harm this framework does not capture. Concerns about workforce depletion in small states, about employer control over tied visas, and about family separation are serious and are not resolved by pointing at remittance volumes. The scheme’s own trajectory — participation down from its 2023 peak as employers revert to backpacker labor — also suggests the flows are less durable than the diplomatic framing implies. The same caution applies to the Falepili arrangement, whose retention effects on Tuvalu are being questioned in real time.
Treating climate as a binding agent understates it. For Kiribati, Tuvalu, and the Marshall Islands, sea level rise is not a shared hazard producing shared institutions; it is a terminal condition for territorial habitation on a horizon shorter than the planning cycles this series has been discussing. Any framework that files it under “cohesion” has miscategorized it, and this paper’s Type I is best read as a placeholder for something the vocabulary does not yet handle.
Data quality in Melanesia is poor. PNG’s census history is troubled, and its population estimates carry very wide error bars. Statements about PNG in this paper are directional only.
VII. Comparative Observations Across Four Regions
- Australia settles the question the Texas Triangle raised. Type B is a real category, not a rescue operation for weak commuting data. Where two centers of comparable weight are connected by one of the world’s busiest air routes and by dense corporate interlock across 900 kilometers of emptiness, the integration is genuine and the contiguity is simply irrelevant. The concept’s founding emphasis on coalescence was a description of one path, not a definition.
- Type K reorganizes the extraction cases across all four papers. Alberta, the Atacama, the Pilbara, and the Bowen Basin belong together, and they are not the same object as the Gulf Coast or Trinidad’s Point Lisas, where the workforce lives beside the plant. The distinction is whether the region is designed to hold people.
- Small-state integration depth is confirmed on a second continent, and the mechanism is now clear. The Eastern Caribbean and the Pacific island states pool deeply; CARICOM, Mercosur, and the Australian federation do not. In every case the variable is whether the function can be supplied unilaterally, not political will, cultural affinity, or treaty language.
- Durable megaregional institutions own assets; discretionary ones write plans. New South Wales legislated a megaregion and repealed it in twenty months. Nothing comparable has happened to a port authority, a binational dam, a currency board, or a fishing quota pool. This is now the clearest actionable finding in the series, and it argues for building megaregional governance around a revenue-generating asset even when the asset is not the point.
- Mobility outcomes are governed by cost and program design, not by legal right. The Caribbean grants four states’ nationals an unrestricted right to live and work in one another’s countries, and almost nobody moves. The Pacific offers heavily rationed, expensive, employer-mediated places and is massively oversubscribed — more than half of Tuvalu’s citizens entered a ballot for 280 permanent visas. The binding constraint on small-state mobility is the wage gradient at the destination and the price of getting there, and free movement among peers supplies neither.
- Contraction and existential exposure now appear in two of four regions. The megaregional literature is built for growth. It has nothing to say about Caracas, Havana, San Juan, Port-au-Prince, the Cook Islands, or Tuvalu. This gap is no longer a Caribbean peculiarity; it is a structural hole in the framework.
VIII. Conclusion
Oceania contains two megaregions on the ordinary reading — the Australian southeast corridor and South East Queensland — one strong candidate at a tenth of the usual scale in New Zealand’s Upper North Island, a set of extraction regions whose workforces live a thousand kilometers away, an instructive absence in Papua New Guinea, and an island system whose defining feature is that a large and growing share of its people live in someone else’s country by arrangement.
Taken with the three preceding papers, it completes a set of findings that none of them produced alone. Contiguity turns out not to matter; Australia’s corridor is bound by aircraft and Perth’s mining hinterland by rosters, and both are as integrated as anything in the Americas. Institutional depth turns out to run inversely to unit size, for a reason now clearly identified. And durability turns out to belong to institutions that own something, which is why a legislated megaregion in Sydney lasted twenty months while eight small states have run a shared tuna quota for over a decade.
The practical implications differ by type and are not interchangeable. Type B corridors need aviation policy, and — where the political system can manage it — rail; Australia’s inability to build the latter across a state line is a governance failure, not an engineering one. Type A formations need the ordinary metropolitan toolkit at a scale that already exists in Queensland and used to exist in New South Wales. Type K regions need honest accounting of where their prosperity actually lands, and a policy conversation about whether the deliberate non-settlement of high-value territory serves anyone besides the operators. Type H island economies need the mobility programs improved rather than merely defended, with retention measured as carefully as remittances.
And the atoll states need something for which this series has no category and no instrument, beyond noting that one government has now signed a treaty acknowledging the problem and that more than half of one nation’s citizens have entered a ballot for the exit. Whatever the right framework for that turns out to be, it is not a megaregional one — but it began in the same place, with people, distance, water, and the question of where a population can live.
