I. The Short Answer, and Why It Is Not the End of the Matter
On the strict test, the Caribbean has no megaregions.
The test requires either a coalescence of adjacent metropolitan labor markets or a transactional corridor joining several centers of national weight. The Caribbean supplies neither, and the reason is arithmetic. The largest urban agglomeration in the insular Caribbean is Santo Domingo, at roughly three and a half million — smaller than Portland–Vancouver, smaller than Guadalajara, smaller than Curitiba. No island metropolis reaches four million. Havana is around two million and falling. San Juan’s metropolitan area is about two million and falling. Kingston with Spanish Town and Portmore is under one and a half million. Port-au-Prince before its current disintegration was perhaps three million. There is no Caribbean São Paulo, no Caribbean Toronto, and there is no pair of Caribbean cities close enough to grow into one another, because between almost any two of them lies open sea.
That answer is correct and unhelpful. It stops at the point where the interesting question begins, which is: what does the megaregional scale look like in a place where land contiguity is impossible?
Three things exist in the Caribbean at megaregional weight. The first is Hispaniola, an island holding roughly twenty-three million people — more than Chile, more than Ecuador, more than the Netherlands — divided by the steepest income gradient in the hemisphere and by a border that is currently being hardened for people and simultaneously formalized for freight. The second is a set of island-scale metropolitan systems — Puerto Rico, Trinidad, Jamaica, western Cuba — that are functionally single labor markets covering an entire national territory, which is a different thing from a metropolis and behaves differently. The third, and the most consequential, is a discontiguous formation bound by air routes, shipping lanes, and remittance flows, whose command functions sit not in the Caribbean at all but in South Florida, New York, Toronto, and a handful of European capitals.
That third formation is the real subject of this paper. It requires two additions to the typology developed in the North and South American companion papers, and it produces a comparative finding that neither of those papers could generate alone.
II. The Span
The insular Caribbean holds roughly 44 million people across some 7,000 islands, of which about 30 are inhabited at any scale. Land area is under 240,000 square kilometers — smaller than the state of Michigan. Figures below are approximate, and in two cases they are worse than approximate: Cuba and Haiti have not conducted reliable enumerations in over a decade.
A. Hispaniola
The only landmass in the region with megaregional demographic weight, and the only one containing a genuine international seam.
The Dominican Ozama–Cibao system. Santo Domingo and its ring (roughly 3.5 million), San Cristóbal, Santiago de los Caballeros (roughly 1.2 million), La Vega, San Francisco de Macorís, and Puerto Plata — joined by a modern autopista network across the Cordillera Central, with the eastern tourism enclave of Punta Cana–Bávaro–Higüey functioning as a fourth pole. Perhaps 7.5 million within the corridor, in a country of roughly 11.4 million. This is the closest thing in the insular Caribbean to a Type B transactional corridor, and it is the only Caribbean formation currently consolidating rather than holding steady or shrinking.
The Haitian west. Port-au-Prince and its plain, the Artibonite valley, and Cap-Haïtien on the north coast: nominally 7 to 8 million of Haiti’s roughly 11.5 million. This formation is not consolidating; it is coming apart. Armed groups controlled roughly 90 percent of the capital as of October 2025, with violence that had displaced 1.4 million people. By February 2026, over 1.45 million people were internally displaced and 6.4 million — more than half the national population — required humanitarian assistance, levels the UN’s humanitarian coordination office described as approaching those after the 2010 earthquake. By mid-June 2026 the figure had reached 1.5 million displaced nationally, with more than 300,000 displaced within Port-au-Prince alone. Election preparations have accelerated, with a $120 million budget agreed and hopes for a first round before the end of 2026 and an elected president in February 2027, though credible voting depends on a security improvement that has not yet occurred, with armed groups still controlling neighborhoods and major roads.
Port-au-Prince deserves a specific typological label. It is the clearest case in the hemisphere of megaregional disaggregation: not depopulation, not economic decline, but the physical partition of a metropolitan region’s internal circulation by armed control of its roads. A metropolitan area is defined by the ability to cross it. Where that ability is revoked neighborhood by neighborhood, the region ceases to exist as a functional unit while its buildings and its people remain in place.
The seam. The 392-kilometer border carries four principal crossings — Dajabón–Ouanaminthe, Jimaní–Malpasse, Elías Piña–Belladère, Pedernales–Anse-à-Pitres — with binational market days at each and roughly a billion dollars of annual bilateral trade, much of it informal. Two policy movements are reshaping it in opposite directions at once. A border wall begun in 2022 was designed to cover nearly half the boundary, and was extended further in 2025 toward roughly 176 kilometers, with more than 276,000 expulsions of Haitians recorded in the preceding year. At the same time, in February 2026 the Dominican president announced a network of dry ports along the Haitian border under a free trade zone regime, a private investment described as exceeding US$300 million and characterized as the most important logistics project ever conceived for the area.
This combination — a wall for people and a free zone for goods — is the exact inverse of the North American seam pattern, where labor and freight both cross in volume. Section VII returns to what that inversion implies.
B. The Island-Metropolitan Systems
Western Cuba. Havana, Mayabeque, Artemisa, Matanzas, and Varadero: perhaps 3.5 million, in a country whose population figures are now genuinely disputed. Cuba has held no census since 2012; the operation planned for 2022 was repeatedly postponed for lack of resources, and results from a newer effort are not expected before 2027. Official data recorded 251,221 emigrants in 2024, while the demographer Juan Carlos Albizu-Campos has argued the true figure exceeded 545,000 and puts the country’s effective population near eight million — far below official estimates. Fertility stands around 1.29, and more than a quarter of Cubans are now 60 or older, with deaths running near double births. UN-adjusted estimates place the 2026 population near 10.9 million, with roughly two-thirds of the last decade’s decline attributable to migration rather than natural causes. Whatever the true number, the direction is not in doubt: this is a contracting formation, and the contraction is concentrated in working-age cohorts.
Puerto Rico. San Juan’s metropolitan area of roughly two million within an island population near 3.2 million, with Ponce, Mayagüez, and Aguadilla as secondary poles connected by a highway ring. Functionally a single labor market covering an entire territory. Population has fallen substantially since 2006 through sustained migration to Florida, New York, and Texas. Puerto Rico’s economic integration runs almost entirely northward rather than to its Caribbean neighbors — it is inside the U.S. customs territory, uses the dollar, and its residents move freely to the mainland.
Jamaica. Kingston–Spanish Town–Portmore (roughly 1.4 million) and Montego Bay, joined by the north-coast tourism belt and the Highway 2000 corridor: a bipolar national system of about 2.8 million, with Kingston Container Terminal functioning as one of the region’s principal transshipment hubs.
Trinidad and Tobago. The East–West Corridor from Port of Spain through Arima, plus Chaguanas and San Fernando and the Point Lisas industrial estate: roughly 1.2 million along a genuine linear corridor. This is the only heavily industrialized formation in the insular Caribbean, built on natural gas, ammonia, methanol, and LNG rather than on tourism or agriculture, and it behaves accordingly — it is the region’s clearest Type D case.
The Lesser Antilles. Barbados, the Windward and Leeward islands, the French departments, the Dutch constituent countries, and the Bahamas: perhaps 3.5 million altogether across more than twenty jurisdictions, none exceeding 400,000. Individually far below any megaregional threshold. Collectively — and only collectively — they constitute something worth analyzing, which is the point of Section IV.
C. The Rim and the Diaspora
The formation that actually operates at megaregional scale is not confined to the sea. Its components are:
- South Florida (roughly 6.2 million), which functions as the Caribbean’s banking center, wholesale market, airlift hub, specialized medical destination, and media capital while being politically part of the United States.
- The mainland littoral: Panama City and Colón (roughly 2 million, with the canal and the Colón Free Zone), Cartagena and Barranquilla (roughly 4 million), and the Cancún–Riviera Maya corridor (roughly 1.5 million).
- The diaspora cities: New York, Toronto, Montreal, Boston, Orlando, London, Amsterdam, Paris, and Madrid, holding several million people of Caribbean origin whose transfers finance household consumption across the islands.
Counted this way, the Caribbean formation comprises somewhere between 60 and 70 million people, a majority of them resident outside the Caribbean Sea. Whether that count is legitimate is the central methodological question of this paper, and Section VI treats the objection seriously.
III. What Makes Them Cohere
Ten mechanisms, of which four have no counterpart in the North or South American inventories.
1. Airlift topology. The Caribbean’s internal geography is defined by route maps, and those maps are hub-and-spoke through points outside most of the region: Miami above all, with San Juan and Panama City as secondary switching points and Bridgetown and Port of Spain as sub-regional ones. Two Eastern Caribbean islands 150 kilometers apart are frequently connected only by a flight through a third country. The collapse and partial reconstitution of the region’s principal intra-island carrier over the past several years removed connections that have not been fully replaced. The practical effect is that the cost of moving between two Caribbean islands often exceeds the cost of moving from either to Florida — which shapes where business relationships form.
2. Transshipment ports serving global rather than regional flows. Kingston, Caucedo, Freeport, Colón, and Cartagena are among the busiest container facilities in the Americas, and almost none of their throughput is intra-Caribbean. They are relay points on Asia–Europe and Asia–U.S. East Coast strings, made viable by the Panama Canal. This is integration into the world economy that produces very little integration with the neighbors. The ports bind the region outward, not inward.
3. Remittances. The single most important economic binding agent, and it runs to North America and Europe rather than within the sea. For Haiti, transfers have run at or above a fifth of national output; for Jamaica the share is comparable; for the Dominican Republic the absolute volume is very large. Household consumption, school fees, roof repairs, and small business capital across much of the region are financed from Brooklyn, Miami, and Toronto. Any account of Caribbean economic geography that stops at the coastline is describing the wrong system.
4. Tourism enclaves and cruise itineraries. The itinerary is a real spatial structure. Multi-island circuits departing Florida ports connect islands that have no other functional link, and they do so on a schedule that determines which quays, taxi ranks, and craft markets receive traffic on which days. Separately, the all-inclusive enclave — Punta Cana, Varadero, Montego Bay’s north coast, Cancún — constitutes a distinct settlement type: a coastal strip of employment with a labor shed reaching deep inland and almost no forward or backward linkage into the host economy beyond wages.
5. Currency and monetary integration. The Eastern Caribbean dollar has been shared by eight members of the Organisation of Eastern Caribbean States under a common central bank since 1983 and pegged to the U.S. dollar since 1976. Several other economies are formally dollarized or tightly pegged. This is deeper monetary integration than anything achieved in North or South America at any scale, and it exists among the smallest political units on the continent — a pattern taken up in Section V.
6. Regulatory-arbitrage clustering. Offshore financial services, ship registries, and investment-migration programs are not incidental to Caribbean economies; for several jurisdictions they are the economy. Occupying the same niche in the world system produces coordination, because pressure from external standard-setters arrives collectively. In September 2025 five investment-migration states established a unified regulatory authority, ECCIRA, expected to become operational in 2026, harmonizing due diligence across programs that had previously competed on laxity. Cohesion here is defensive and it is produced by outside pressure rather than by internal complementarity.
7. Shared catastrophic exposure. Hurricanes, seismic risk, drought, and sargassum influx affect the entire basin on a common calendar, and the institutions built around that exposure are among the region’s most functional: a regional disaster management agency and a multi-country parametric insurance facility, the first of its kind anywhere, which pays out on measured wind speed and rainfall rather than on assessed loss. This is a binding agent with no analogue in the companion papers. North American and South American megaregions cohere around what they produce together; a good deal of Caribbean cohesion arises from what happens to them together.
8. Free movement — formally real, practically thin. Under a new Enhanced Cooperation Chapter of the Revised Treaty of Chaguaramas, which permits groups of at least three member states to advance integration among themselves where the Community as a whole cannot do so within a reasonable period, four states agreed to grant their nationals the right to enter, leave, re-enter, move freely, reside, and work in one another’s territories, with access to emergency and primary health care and to public primary and secondary education. Barbados, Belize, Dominica, and Saint Vincent and the Grenadines implemented this on 1 October 2025, eliminating the previous six-month limit and granting indefinite stay on arrival — the Community’s first unrestricted mobility arrangement extending beyond temporary economic purposes.
The uptake is the striking part. When Belize tabled its implementing legislation in February 2026, the responsible minister disclosed that only two people were residing in Belize on indefinite stay, with 83 individuals having entered and exited under the mechanism since it began. And as of July 2026, with the first anniversary approaching, no additional member states had publicly announced joining the original four, despite a regional private sector call for others to come aboard by the end of 2025. A right that costs nothing to exercise and is exercised by almost nobody is telling us something: the barrier to Caribbean labor mobility was never primarily legal. It is the cost and infrequency of the flight, the absence of jobs at the destination, and the fact that the return on migration is far higher toward North America than within the basin.
9. Language and constitutional partition as anti-binding agents. Four language blocs — Spanish, English, French and Kreyòl, Dutch — and a continuing constitutional attachment of many territories to metropoles outside the region. Guadeloupe and Martinique are in the European Union. Puerto Rico and the U.S. Virgin Islands are inside U.S. customs and currency. Curaçao and Aruba are constituent countries of the Kingdom of the Netherlands. These arrangements route trade, education, medical care, migration, and legal appeal outward along four separate spokes. The Caribbean is the only region in this series where the dominant institutional forces are actively centrifugal.
10. Circulation rather than settlement. Caribbean migration is characteristically circular and multi-generational rather than one-way: people return, retire, remit, build houses they occupy seasonally, and maintain dual residence. This produces a population that is functionally part of two labor markets and fully counted in neither, and it is the mechanism that makes the diaspora a component of the region rather than a subtraction from it.
Growth
Every major Caribbean formation except the Dominican corridor and the mainland rim is either flat or shrinking. Fertility is below replacement across most of the region. Cuba is in outright demographic contraction. Puerto Rico has lost population for two decades. Haiti’s population is growing but its metropolitan region is dispersing under violence. The Eastern Caribbean states have exported working-age cohorts for two generations. Where growth is occurring — the Dominican east, the Cancún corridor, parts of Panama — it is tourism-and-construction growth, which is Type E in the South American scheme, and it draws its labor from Haiti, Venezuela, and rural interiors rather than from natural increase.
The Caribbean is thus the first region in this series where contraction, not expansion, is the baseline condition, and any typology built here has to accommodate that.
IV. A Typology
The axes remain binding agent and morphology. Four binding agents and two morphologies must be added; two North American types are absent entirely.
Axis One: Binding Agent
Type A — Commuter-Field. Present only at island scale: Puerto Rico, Trinidad, the Santo Domingo ring, greater Kingston. These are single metropolitan labor markets that happen to cover an entire national or territorial jurisdiction, which gives them megaregional political salience without megaregional size.
Type B — Transactional Corridor. The Dominican Ozama–Cibao system, weakly. Little else.
Type C — Production-Sharing. Effectively absent. The apparel and medical-device free zones of the Dominican Republic and the assembly plants of Haiti’s north participate in North American supply chains but not in Caribbean ones. The region’s manufacturing links run outward.
Type D — Extractive-Logistical. Trinidad’s gas and petrochemical complex; Jamaican bauxite; the transshipment port system; Guyanese and Surinamese offshore oil on the southern rim.
Type E — Amenity-Migration. Present and important, but in an unusual form: the enclave rather than the region. Punta Cana, Varadero, the Riviera Maya, and Montego Bay’s north coast generate employment lattices without generating integrated settlement.
Type F — Administrative-Primate. Havana, Kingston, San Juan, Port-au-Prince, Santo Domingo, and every Eastern Caribbean capital. Universal in the region, given that most units are small enough that the capital simply is the country.
Type G — Maritime-Networked. (New.) Cohesion produced by sea and air links across water, where land contiguity is impossible in principle rather than merely absent in fact. The nodes are ports and airports; the corridor is a route, not a road; and connectivity is set by carrier decisions and vessel strings rather than by settlement patterns. A route can be cancelled in a way that a highway cannot, which makes Type G cohesion unusually reversible and unusually dependent on decisions taken by a small number of firms.
Type H — Remittance-Diaspora. (New.) Cohesion produced by transfers and circulation between resident and expatriate populations. The functional region includes cities on another continent, and the direction of dependence runs outward. Type H is the dominant binding agent for Haiti, Jamaica, and much of the Eastern Caribbean, and it is invisible to any method that delineates regions by commuting, freight, or contiguity.
Type I — Shared-Hazard. (New.) Cohesion produced by common catastrophic exposure and the pooled institutions built to manage it. Economically weak as a binder; institutionally strong. The region’s most functional multi-country bodies exist because storms do not respect maritime boundaries, and pooling is cheaper than self-insuring.
Type J — Regulatory-Arbitrage. (New.) Cohesion produced by shared occupation of a niche in the world regulatory order — offshore finance, ship registries, investment migration — and by the collective external pressure that niche attracts. Coordination here is defensive, episodic, and driven by deadlines set elsewhere.
Axis Two: Morphology
M2 — Primate-Radial. The regional default, at national scale.
M3 — Linear Corridor. Trinidad’s East–West Corridor; the Dominican Ozama–Cibao axis; Jamaica’s north coast.
M5 — Seam. The Hispaniola border, and nothing else. The region’s only land international boundary of any consequence.
M7 — Discontiguous Archipelago. (New.) Nodes separated by water, with connectivity purely nodal and no possibility of intervening settlement. The Eastern Caribbean chain is the type case. In an M7 formation, the “region” exists only as a schedule.
M8 — Extraterritorial-Cored. (New.) A formation whose command functions — finance, wholesale distribution, airlift control, specialized medicine, higher education, media, and capital supply — sit outside its own territory and under another sovereign’s jurisdiction. South Florida performs for the Caribbean roughly what São Paulo performs for Brazil, without being in it or accountable to it. This has no North or South American parallel, and it is the most consequential structural fact about the region.
The Cross-Classification
| Formation | Binding Agent | Morphology | Consolidation |
|---|---|---|---|
| Dominican Ozama–Cibao | A / B / E | M3 | Consolidating |
| Haitian west (Port-au-Prince) | F / H | M2 | Disaggregating |
| Hispaniola as a whole | H / D | M5 | Divided and hardening |
| Western Cuba | F | M2 | Contracting |
| Puerto Rico | A / F / H | M2 | Contracting |
| Jamaica (Kingston–Montego Bay) | F / D / H | M3 | Flat |
| Trinidad East–West Corridor | D | M3 | Flat |
| Eastern Caribbean (OECS) | G / H / I / J | M7 | Institutionally deep, demographically flat |
| Bahamas–Turks and Caicos | G / J / E | M7 | Growing modestly |
| Circum-Caribbean networked formation | G / H | M8 | Consolidating |
Four observations follow.
First, the region’s most consolidated formation is the one that does not appear on a map of the Caribbean. The networked formation with its core in South Florida is growing, deepening, and adding functions, while nearly every territorial formation inside the sea is flat or shrinking.
Second, Type H dominance means the region’s prosperity is decoupled from its own productivity. A place financed by transfers from elsewhere is exposed to the immigration policy, labor market, and exchange rate of the sending country in a way that no amount of domestic reform addresses. This is a different vulnerability from Type C policy-contingency in North America or Type F state-dependence in South America, and it is arguably more severe, because the decisive policies are made by governments in which the affected population has no vote.
Third, Type G reversibility is underappreciated. A route can be cancelled. When a carrier withdraws from a city pair, two economies that were connected on Monday are not on Tuesday, and no infrastructure investment on either island can restore the link. Land-based megaregions do not face this.
Fourth, the region contains the hemisphere’s only current instance of megaregional disaggregation, and it should be studied as such rather than only as a humanitarian emergency. Port-au-Prince demonstrates that the internal circulation on which metropolitan function depends is a public good that can be withdrawn, and that its withdrawal dissolves the region without moving a single building.
V. The Governance Finding
Here the Caribbean produces the most transferable result in the whole series.
The Organisation of Eastern Caribbean States operates a common currency and central bank, a shared supreme court, a joint civil aviation authority, a joint telecommunications regulator, common pharmaceutical procurement, shared diplomatic representation, and free movement of nationals among its members. By any reasonable measure this is the deepest political and economic integration in the Western Hemisphere. Its combined population is under a million people.
CARICOM, at roughly eighteen million including Haiti, achieves far less: a partially implemented single market, a court whose appellate jurisdiction most members have not accepted, and a free movement regime that four small states have implemented while the large ones have not. The Community’s internal economics work against it — income per head ranges from Haiti at the bottom to the Bahamas at the top, Haiti accounts for more than half the Community’s population without having completed internal processes to fully implement the single market, and the Bahamas does not participate in it at all.
Hispaniola, at twenty-three million, has effectively no functioning binational governance despite being one island with one watershed system, one disease environment, one hurricane track, and one deeply interdependent labor market. The binational commission has been intermittent. What is being built instead is a wall and, alongside it, a free trade zone.
The pattern is an inverse relationship between the size of the units and the depth of their integration. The smallest states integrate most deeply because they cannot supply a central bank, a court of appeal, an air safety regulator, or a drug procurement office at national scale — the minimum efficient scale of these institutions exceeds the country. Larger units can supply them nationally, so they do, and integration stops.
This finding travels. It explains why Mercosur underdelivers relative to its treaty text while the Eastern Caribbean overdelivers relative to its obscurity. It explains why North American megaregional coordination succeeds only where a single physical asset is too large for one jurisdiction to own — a harbor, a river basin, an airshed. And it reframes the megaregional governance problem generally: coordination does not follow from functional interdependence, which exists everywhere. It follows from the impossibility of unilateral supply. Where a jurisdiction can provide a function alone, it will, however inefficiently. Institutional design that hopes to produce megaregional cooperation should therefore start by identifying which functions genuinely cannot be supplied at the existing scale, and build there, rather than exhorting cooperation on functions each party can perform badly by itself.
VI. Objections
This may be regional studies wearing a borrowed label. The strongest objection. The megaregion concept has now been stretched twice — once with Type F for South American primacy, and again here with four new binding agents and two new morphologies. At some point a category that requires redefinition at every stop is not describing a common phenomenon; it is providing a common vocabulary. The defense offered is that a common vocabulary is itself useful, and that the comparative findings in Section VII are ones neither Caribbean area studies nor the North American megaregion literature would have produced independently. A critic may reasonably judge that insufficient.
Counting the diaspora is a contestable move. Including Miami, New York, and Toronto in a Caribbean formation could be called a category error: those cities are components of other regions, fully counted there, and their Caribbean-origin populations are not a separable unit. The response is that Type H integration is measurable — transfer volumes, air passenger flows, dual residence, and return migration are all countable — and that excluding it produces an account of Caribbean economies that cannot explain how their households are financed. But the boundary is genuinely undrawable, and figures given here should be read as indicative.
The data are the worst in the series. Cuba has not held a census since 2012 and results from the current effort are not expected before 2027. Haiti’s enumeration is far older still and its recent attempts have been disrupted by the security situation. Regional statistical capacity is thin, informality is high, and circular migration is systematically undercounted at both ends.
Contraction may make the exercise moot. If most of the region’s territorial formations are shrinking and the growth is in the extraterritorial core, then the analytically interesting object may not be a Caribbean megaregion at all but a Caribbean-origin population in North American megaregions — which returns us to the previous paper’s frame and dissolves this one.
Climate may be the decisive variable and is not modeled here. Sea level rise, hurricane intensification, coral loss, and sargassum influx bear directly on the two sectors — tourism and coastal settlement — on which the region’s territorial economies depend. A binding agent that produced institutions (Type I) may, past a threshold, become a dissolving agent. This paper treats hazard as a source of cohesion. On a longer horizon that is likely to be the wrong framing.
VII. Comparative Observations Across the Three Regions
- The dominant binding agent differs by continent and predicts the dominant vulnerability. North America: labor markets and production sharing — vulnerable to trade policy. South America: commodity corridors and state authority — vulnerable to price cycles and state capacity. The Caribbean: routes and remittances — vulnerable to carrier decisions and to the immigration policy of countries where its population does not vote.
- Seams behave in three distinct ways. North American seams are open to both goods and labor and are dense on both counts. South American seams barely exist, because the megaregions do not touch. The Hispaniola seam is being deliberately opened to goods and closed to people simultaneously. This third pattern is the one to watch, because it is the arrangement a wealthy jurisdiction adopts toward a poorer neighbor when it wants the labor’s product without the labor’s presence, and it is not confined to Hispaniola.
- Integration depth varies inversely with unit size. Established in Section V. The Eastern Caribbean, the smallest units in the hemisphere, have the deepest integration; the largest units have the least. The operative variable is not political will but whether a function can be supplied unilaterally.
- Contraction is a megaregional condition, not merely a national one, and the literature has no vocabulary for it. Caracas, Havana, San Juan, and Port-au-Prince are all shrinking or dispersing formations. Every existing megaregional framework, including the two companion papers, is built around growth — where the next fifty million will live, how corridors consolidate. Nothing in that apparatus tells a planner what to do with a megaregion that is emptying. Given below-replacement fertility across most of the hemisphere, this gap will matter well beyond the Caribbean within a generation.
- The extraterritorial core is probably not unique. Having identified M8 here, one can ask where else it appears in weaker form — Central American economies cored on Los Angeles and Houston, Andean economies increasingly cored on Madrid and Santiago, Venezuelan networks now cored on Bogotá and Miami. The Caribbean may simply be the case where the pattern is most complete and therefore most visible.
VIII. Conclusion
The Caribbean has no megaregions in the sense that produced the term, and it will not acquire any, because the arithmetic of island populations forbids it. What it has instead is one large divided island, a set of national-scale metropolitan systems, an archipelago whose smallest states have built the hemisphere’s deepest institutions, and a networked formation of sixty million or more people bound by air routes, shipping strings, and money orders, whose commanding functions are performed in a city that belongs to another country.
Reading that formation as a megaregion is a stretch of the concept. Refusing to read it as anything is worse, because the alternative treatments — twenty-odd national economic analyses, or a humanitarian frame for Haiti and a tourism frame for everyone else — miss the fact that these places are systematically connected and connected in ways that determine their prospects.
The practical implications follow the typology. For Type G formations, the binding constraint is route economics, and the instruments are air service agreements, ferry subsidies, and shipping arrangements — unglamorous and decisive. For Type H, the relevant policy is made in Washington, Ottawa, and Brussels, which argues for diaspora-facing institutions and for treating remittance costs as a first-order development variable rather than a technicality. For Type I, the pooled facilities already work and deserve capitalization rather than reinvention. For the Hispaniola seam, the current trajectory — hardening for people, opening for freight — should at minimum be examined for what it will produce over twenty years, which on present evidence is a labor force that supplies a neighboring economy without being able to live in it.
And the governance finding stands on its own, beyond this region. Cooperation at the megaregional scale is not produced by interdependence, which is everywhere, nor by exhortation, which has been tried at every scale on three continents with little to show. It is produced where a needed function cannot be supplied alone. Eight small islands share a central bank because none of them can run one. That is the whole mechanism, and it is the one worth building on.
