## Mind the Gaps: Caribbean Trade Patterns and the Connectivity Constraint — Introduction

## Source details

**Canonical URL:** [Mind the Gaps: Caribbean Trade Patterns and the Connectivity Constraint — Introduction](https://www.imf.org/-/media/files/publications/wp/2026/english/wpiea2026095-source-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2026/english/wpiea2026095-source-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2026/english/wpiea2026095-source-pdf.pdf.json)

---

### Context and motivation
- Amid intensifying geoeconomic fragmentation and rising geopolitical tensions, Caribbean economies face heightened exposure to external shocks.
- Evidence indicates a broad-based decline in potential growth in the Caribbean over the past two decades due to declining contributions of total factor productivity and human capital as well as to resource misallocation (Amundsen et al., 2025).
- Strengthening economic resilience is increasingly important given vulnerabilities in goods and services trade, notably food imports and tourism travel exports.

### Stylized facts on trade structure and concentration
- Goods trade:
  - Goods exports account for only a small share of GDP in most Caribbean countries, with the important exception of oil exporters.
  - Caribbean countries rely heavily on goods imports—particularly for food—implying vulnerability to foreign exporter supply disruptions.
  - Caribbean import-source concentration is high; 13 out of 14 CARICOM countries covered rank well above the 50th percentile globally in import source concentration (Herfindahl-Hirschman Index, HHI).
- Tourism:
  - Travel exports dominate foreign exchange earnings for many Caribbean countries; services analysis focuses on tourism exports.
  - Stayover arrivals from the United States have accounted for roughly 60 percent of total arrivals to CARICOM countries since the mid-1990s, with a further 30 percent originating from Europe and Canada, and the rest from all other countries combined.
  - Tourism accounts for over 20 percent of GDP and about 70 percent of service exports across CARICOM countries (Figure 8 summary).
  - Cruise passengers comprised nearly two-thirds of total arrivals in 2024, but stayover tourists generate substantially higher revenue (example: Dominica estimated tourism receipts per cruise passenger under $US 200 in 2023 vs. roughly $US 2,000 per stayover visitor).

### Key analytical approaches and data
- Goods trade:
  - Gravity-model based analysis using bilateral merchandise exports from CEPII’s BACI database (FOB basis).
  - Controls: CEPII Gravity database variables (great-circle distance, contiguity, common language, colonial ties), Regional Trade Agreements (Mario Larch’s database), effective bilateral tariff rates (Teti, 2024; Ahn and Tan, 2025), and robustness checks (UN Comtrade re-exports, IMF DOTS).
  - Estimation method: Poisson Pseudo Maximum Likelihood (PPML) with exporter–year and importer–year fixed effects; clustering at country-pair level.
- Tourism:
  - Bilateral country-level tourism data from UNWTO (arrivals by origin–destination); monthly regional data for 23 destinations from the Caribbean Tourism Organization (CTO).
  - Flight-level passenger capacity data from FlightRadar24 constructed into total inbound passenger seat capacity (global flights since November 2018 covering 264 origin-destination country pairs).
  - Gravity framework with an instrumental variable (shift-share/Bartik-style) using origin-level supply shocks to flight capacity.
- Connectivity:
  - Maritime connectivity: UNCTAD Liner Shipping Connectivity Index (LSCI) and UNCTAD Experimental Transport and Trade Dataset freight costs (average transport expenditure per kilogram, CIF−FOB definition, coverage 2016–2021).
  - Financial connectivity: BIS Locational Banking Statistics (LBS) on outstanding claims and liabilities (captures ~95 percent of cross-border banking activity) used to construct creditor–debtor bilateral banking-flow variables; shift-share IV using frozen creditor shares and creditor-country outward-claim shocks.

### Major empirical findings — Goods trade and maritime connectivity
- Gravity estimation results:
  - Caribbean countries trade relatively less with countries outside the Caribbean region than predicted by fundamentals, while trading more within the region than the model suggests.
  - The principal driver of this pattern is a disproportionately large adverse effect of geographical distance for CARICOM destinations.
  - Re-exports and geographic distance together account for most of the discrepancy between actual and model-predicted trade values.
- Connectivity evidence:
  - UNCTAD LSCI and freight-cost measures show CARICOM countries face relatively weak shipping connectivity and systematically higher shipping costs than global averages.
  - All CARICOM countries, except Jamaica, incur shipping costs above the global average; Antigua and Barbuda experiences rates that are three times higher than the world average (authors’ characterization).
  - CARICOM destinations exhibit significantly lower bilateral shipping connectivity after controlling for geographic fundamentals and origin-country factors (simple OLS bilateral-level connectivity analysis; Tables 2–3 referenced).
  - Hub-and-spoke structure: regional hubs (notably Jamaica and The Bahamas for CARICOM; Antigua and Barbuda for the ECCU) act as re-export/transshipment nodes; three CARICOM members rank among the world’s top 15 re-exporting hubs (UN Comtrade, 2025).

### Major empirical findings — Tourism and air connectivity
- Sector importance and concentration:
  - Tourism accounts for over 20 percent of GDP and about 70 percent of service exports across CARICOM countries.
  - Tourism source-market concentration is high in CARICOM countries (most rank above the 50th percentile of HHI distribution).
- Elasticities and connectivity effects:
  - Table 6: average elasticity of tourist arrivals with respect to source-market GDP is around 0.8; for U.S. to CARICOM flows, estimated elasticity rises to around 2.
  - Total inbound passenger seat capacity is substantially lower for Caribbean destinations after controlling for income and remoteness (Table 7).
  - Decomposition explains ~94 percent of variation in total capacity: number of origin countries contributes ~40 percent and flight frequency ~54 percent (Table 8).
  - Gravity regressions (global, Table 9): flight capacity coefficient around 0.52–0.66 (p< 0.01), implying an elasticity of around 0.5 to 0.6 — a 10 percent increase in flight capacity associated with roughly a 5–6 percent increase in tourist arrivals.
  - Monthly ECCU results (Table 10): flight-capacity coefficient around 0.20; excluding intra-Caribbean flows raises elasticity to ~0.22.
- Tourism and banking link (Table 12 summary):
  - Baseline 2SLS regression results for bilateral tourism flows report:
    - Bilateral Bank Flows: 0.486*** (Specification 1), 0.487*** (Specification 2), 0.488*** (Specification 3)
    - Distance: -0.701*** (Specification 1), -0.645*** (Specification 2), -0.655*** (Specification 3)
    - Border: 1.478*** (Specification 1), 1.499*** (Specification 2), 1.495*** (Specification 3)
    - Language: 0.248* (Specification 1), 0.309** (Specification 2), 0.308** (Specification 3)
    - Colonial Ties: -1.976*** (Specification 1), -1.969*** (Specification 2), -1.946*** (Specification 3)
    - Bilateral Bank Flows x CARICOM: -0.563*** (Specification 2), -0.558*** (Specification 3)
    - Observations: 24,327 (Specification 1), 22,507 (Specification 2), 22,507 (Specification 3)
    - Origin FE: Yes (all)
    - Destination-Year FE: Yes (all)
    - Origin-Destination Clustered SEs: Yes (all)
    - R-squared: 0.522 (Specification 1), 0.537 (Specification 2), 0.536 (Specification 3)
  - Interpretation:
    - A one percent increase in cross-border banking flows increases tourism linkages by 0.486 percent (worldwide elasticity).
    - The combined effect for CARICOM economies is slightly negative at -0.076 percent and essentially zero.
    - For ECCU economies the interaction coefficient is not significantly negative and remains indistinguishable from zero.

### Major empirical findings — Financial connectivity and trade–finance linkages
- Worldwide banking–trade elasticity:
  - The worldwide banking–trade elasticity increases sharply to 0.92 percent, indicating that a one percent increase in cross-border banking linkages increases bilateral trade almost one to one.
- Banking–trade causal estimates:
  - Global IV estimate: elasticity of trade with respect to bilateral bank flows is 0.580 (IV), statistically significant at the 1 percent level; OLS comparable estimate is 0.443.
  - CARICOM heterogeneity: the CARICOM interaction coefficient is −0.437 (highly significant), implying the banking–trade elasticity for CARICOM destinations equals 0.580 − 0.437 = 0.143 (about 75 percent weaker than for non-CARICOM economies).
  - Robustness: result persists excluding The Bahamas, excluding oil exporters, excluding crisis years; first-stage F-statistic close to conventional thresholds though instrument strength declines for CARICOM interactions and when financial centers are excluded.
- Crisis dynamics:
  - During the Global Financial Crisis (GFC), CARICOM banking–trade elasticity collapsed to 0.01; world elasticity fell to 0.50.
  - Post-GFC recovery: CARICOM elasticity recovered to 0.23 (still below global levels), suggesting a persistent structural gap in how banking relationships facilitate trade in the region.
- Structural explanations for muted CARICOM banking–trade link:
  - Post-GFC withdrawal of correspondent banking relationships disproportionately affected the Caribbean.
  - Many CARICOM economies export commodities or rely heavily on services (tourism), which may need less traditional bank trade finance.
  - Connectivity constraints, small market scale, and alternative financing mechanisms (cash-in-advance, open-account terms) reduce reliance on interbank trade finance captured by BIS LBS.
  - BIS LBS do not capture cash-in-advance or open-account trade-financing mechanisms used in the region; these data gaps limit assessment of trade-financing resilience.

### Implications and policy-relevant insights
- Primary constraints:
  - Physical connectivity constraints (maritime and air) are the main binding factors limiting diversification of goods import sources and tourism source markets; financial connectivity matters globally but is substantially weaker as a trade-enhancing channel for CARICOM countries.
- Diversification and resilience:
  - High concentration of import and tourism sources increases vulnerability to idiosyncratic exporter or source-country shocks; diversifying trading partners and tourism source markets could reduce exposure.
- Connectivity-focused policy priorities:
  - Improve maritime connectivity and reduce freight costs: address port infrastructure quality, customs and port administration efficiency, promote shipping competition, and reduce reliance on transshipment where feasible.
  - Expand air connectivity: route development, increased flight frequency, airport capacity projects to raise extensive margins (number of origin countries and flight frequency).
  - Rebuild and deepen correspondent banking and broader financial linkages where feasible to restore trade-finance channels, recognizing physical connectivity remains a core constraint.
- Data gaps and future work:
  - Address bilateral data gaps on alternative trade-financing mechanisms (cash-in-advance and open-account) to better assess resilience of trade financing in the Caribbean.

*Source: IMF Working Paper — Introduction section of “Mind the Gaps: Caribbean Trade Patterns and the Connectivity Constraint” (source PDF: wpiea2026095-source-pdf).*

### Introduction ...........................................................................................................

### Introduction

### Major themes covered
- Goods Trade
- Tourism
- Connectivity
- Bank flows and trade–finance linkages
- Data

### Structural components in the Introduction
- Background
- Concentration
- Gravity-based approach
- Connectivity (multiple subsections: Goods Trade and Connectivity; Tourism and Connectivity; Finance and Connectivity)
- Conclusion
- References

### Figures included in the content unit
- Figure 1. Goods Exports and Imports
- Figure 2. Import Source Concentration by Destination
- Figure 3. Correlation between Import Source Concentration and External Shock Volatility
- Figure 4. Maritime Connectivity and Average Freight Cost in the Caribbean
- Figure 5. Direct Cargo Shipping Routes in the Caribbean Basin
- Figure 6. The Pattern of Re-exports
- Figure 7. Deviations Between Actual and Predicted Trade Patterns in CARICOM
- Figure 8. Overview of Caribbean Tourism Sector
- Figure 9. Concentration of Caribbean Tourism Source Countries
- Figure 10. Stylized Representation of Global Banking-Trade Nexus
- Figure 11. Banking Linkages in the Caribbean Over Time
- Figure 12. Concentration Patterns of Caribbean Banking Positions

### Tables included in the content unit
- Table 1. OLS Regression of Import Concentration Index
- Table 2. OLS Regression of Connectivity Index
- Table 3. OLS Regression of Shipping Cost
- Table 4. PPML Estimation of the Gravity Model
- Table 5. OLS Estimation of Tourism Source Market Concentration
- Table 6. OLS Estimation of Tourism Elasticity
- Table 7. OLS Estimation of Flight Capacity
- Table 8. OLS Regression on Flight Capacity Decomposition
- Table 9. Gravity Regressions of Global Bilateral Tourism Flows
- Table 10. Gravity Regressions of Monthly Tourism Flows to the ECCU
- Table 11. Estimation Results for Effects of Banking Relationships on Goods Trade
- Table 12. Estimation Results for Effects of Banking Relationships on Tourism Flows

*IMF WORKING PAPERS Mind the Gaps: Caribbean Trade Patterns and the Connectivity Constraint — Introduction (pages listing and figures/tables inventory).*

### Introduction

### Introduction

### Context and motivation
- Amid intensifying geoeconomic fragmentation and rising geopolitical tensions, Caribbean economies face heightened exposure to external shocks.
- Evidence indicates a broad-based decline in potential growth in the Caribbean over the past two decades due to declining contributions of total factor productivity and human capital as well as to resource misallocation (Amundsen et al., 2025).
- Strengthening economic resilience of Caribbean countries is increasingly important given vulnerabilities in goods and services trade (notably food imports and tourism travel exports).

### Stylized facts on trade structure and concentration
- Goods exports account for only a small share of GDP in most Caribbean countries, with the important exception of oil exporters.
- Caribbean countries rely heavily on goods imports—particularly for food—implying vulnerability to foreign exporter supply disruptions.
- Travel exports dominate foreign exchange earnings for many Caribbean countries; the paper focuses its services analysis on tourism exports.
- Caribbean countries remain heavily concentrated in their sources of imported goods and tourism demand:
  - Stylized goods-trade facts: Caribbean import-source concentration is high; 13 out of 14 CARICOM countries covered rank well above the 50th percentile globally in import source concentration (Herfindahl-Hirschman Index, HHI).
  - Tourism source concentration: stayover arrivals from the United States have accounted for roughly 60 percent of total arrivals to CARICOM countries since the mid-1990s, with a further 30 percent originating from Europe and Canada, and the rest from all other countries combined.

### Key analytical approaches and data
- Goods trade:
  - Gravity-model based analysis using bilateral merchandise exports from CEPII’s BACI database (FOB basis).
  - Controls: CEPII Gravity database variables (great-circle distance, contiguity, common language, colonial ties), Regional Trade Agreements (Mario Larch’s database), effective bilateral tariff rates (Teti, 2024; Ahn and Tan, 2025), and robustness checks (UN Comtrade re-exports, IMF DOTS).
  - Estimation method: Poisson Pseudo Maximum Likelihood (PPML) with exporter–year and importer–year fixed effects; clustering at country-pair level.
- Tourism:
  - Bilateral country-level tourism data from UNWTO (arrivals by origin–destination), monthly regional data for 23 destinations from the Caribbean Tourism Organization (CTO).
  - Flight-level passenger capacity data from FlightRadar24 constructed into total inbound passenger seat capacity (global flights since November 2018 covering 264 origin-destination country pairs).
  - Gravity framework with an instrumental variable (shift-share/Bartik-style) that uses origin-level supply shocks to flight capacity to isolate exogenous variation.
- Connectivity:
  - Maritime connectivity: UNCTAD Liner Shipping Connectivity Index (LSCI) and UNCTAD Experimental Transport and Trade Dataset freight costs (average transport expenditure per kilogram, CIF−FOB definition, coverage 2016–2021).
  - Financial connectivity: BIS Locational Banking Statistics (LBS) on outstanding claims and liabilities (captures ~95 percent of cross-border banking activity) used to construct creditor–debtor bilateral banking-flow variables; shift-share IV using frozen creditor shares and creditor-country outward-claim shocks.

### Major empirical findings — Goods trade and maritime connectivity
- Gravity estimation results:
  - Caribbean countries trade relatively less with countries outside the Caribbean region than predicted by fundamentals, while trading more within the region than the model suggests.
  - The principal driver of this pattern is a disproportionately large adverse effect of geographical distance for CARICOM destinations.
  - Re-exports and geographic distance together account for most of the discrepancy between actual and model-predicted trade values.
- Connectivity evidence:
  - UNCTAD LSCI and freight-cost measures show CARICOM countries face relatively weak shipping connectivity and systematically higher shipping costs than global averages.
  - All CARICOM countries, except Jamaica, incur shipping costs above the global average; Antigua and Barbuda experiences rates that are three times higher than the world average (authors’ characterization).
  - CARICOM destinations exhibit significantly lower bilateral shipping connectivity after controlling for geographic fundamentals and origin-country factors (simple OLS bilateral-level connectivity analysis; Tables 2–3 referenced).
  - Hub-and-spoke structure: regional hubs (notably Jamaica and The Bahamas for CARICOM; Antigua and Barbuda for the ECCU) act as re-export/transshipment nodes; three CARICOM members rank among the world’s top 15 re-exporting hubs (UN Comtrade, 2025).

### Major empirical findings — Tourism and air connectivity
- Sector importance:
  - Tourism accounts for over 20 percent of GDP and about 70 percent of service exports across CARICOM countries (Figure 8 summary).
  - Cruise passengers comprised nearly two-thirds of total arrivals in 2024, but stayover tourists generate substantially higher revenue (example: Dominica estimated tourism receipts per cruise passenger under $US 200 in 2023 vs. roughly $US 2,000 per stayover visitor).
- Concentration and elasticity:
  - Tourism source-market concentration is high in CARICOM countries (most rank above the 50th percentile of HHI distribution).
  - Table 6: average elasticity of tourist arrivals with respect to source-market GDP is around 0.8; for U.S. to CARICOM flows, estimated elasticity rises to around 2.
- Air connectivity impact:
  - Total inbound passenger seat capacity is substantially lower for Caribbean destinations after controlling for income and remoteness (Table 7).
  - Decomposition: extensive margin (number of origin countries with direct routes and flight frequency) explains ~94 percent of variation in total capacity; number of origin countries contributes ~40 percent and flight frequency ~54 percent (Table 8).
  - Gravity regressions (global, Table 9): flight capacity coefficient around 0.52–0.66 (p< 0.01), implying an elasticity of around 0.5 to 0.6 — a 10 percent increase in flight capacity associated with roughly a 5–6 percent increase in tourist arrivals.
  - Monthly ECCU results (Table 10): flight-capacity coefficient around 0.20; excluding intra-Caribbean flows raises elasticity to ~0.22.

### Major empirical findings — Financial connectivity and trade–finance linkages
- Banking-trade nexus:
  - Cross-sectional scatter: positive association between bilateral banking positions and trade, but relationship appears weaker for Caribbean destinations.
  - IV strategy: Bartik-style instrument using frozen creditor shares (2008) and creditor-country outward-claim shocks; two-stage estimation framework for causal inference.
  - Global effect: estimated elasticity of trade with respect to bilateral bank flows is 0.580 (IV), statistically significant at the 1 percent level; OLS comparable estimate is 0.443.
  - CARICOM heterogeneity: the CARICOM interaction coefficient is −0.437 (highly significant), implying the banking–trade elasticity for CARICOM destinations equals 0.580 − 0.437 = 0.143 (about 75 percent weaker than for non-CARICOM economies).
  - Robustness: result persists excluding The Bahamas, excluding oil exporters, excluding crisis years; first-stage F-statistic close to conventional thresholds though instrument strength declines for CARICOM interactions and when financial centers are excluded.
- Crisis dynamics and persistence:
  - During the Global Financial Crisis (GFC), CARICOM banking–trade elasticity collapsed to 0.01; world elasticity fell to 0.50.
  - Post-GFC recovery: CARICOM elasticity recovered to 0.23 (still below global levels), suggesting a persistent structural gap in how banking relationships facilitate trade in the region.
- Structural explanations for muted CARICOM banking-trade link:
  - Post-GFC withdrawal of correspondent banking relationships disproportionately affected the Caribbean.
  - Many CARICOM economies export commodities or rely heavily on services (tourism), which may need less traditional bank trade finance.
  - Connectivity constraints, small market scale, and alternative financing mechanisms (cash-in-advance, open-account terms) reduce reliance on interbank trade finance captured by BIS LBS.

### Implications and policy-relevant insights
- Primary constraints:
  - Physical connectivity constraints (maritime and air) are the main binding factors limiting diversification of goods import sources and tourism source markets; financial connectivity matters globally but is substantially weaker as a trade-enhancing channel for CARICOM countries.
- Diversification and resilience:
  - High concentration of import and tourism sources increases vulnerability to idiosyncratic exporter or source-country shocks; diversifying trading partners and tourism source markets could reduce exposure.
  - Policies aimed at strengthening shipping and air connectivity could facilitate diversification of goods and services trade partners.
- Connectivity-focused policy priorities signaled by the evidence:
  - Improve maritime connectivity and reduce freight costs (address port infrastructure quality, customs and port administration efficiency, promote shipping competition, and reduce reliance on transshipment where feasible).
  - Expand air connectivity (route development, increased flight frequency, airport capacity projects) to raise extensive margins (number of origin countries and flight frequency).
  - Rebuild and deepen correspondent banking and broader financial linkages where feasible to restore trade-finance channels, recognizing physical connectivity remains a core constraint.
- Data gaps and future work:
  - BIS LBS do not capture cash-in-advance or open-account trade-financing mechanisms used in the region; addressing bilateral data gaps on alternative trade-financing mechanisms is critical for assessing resilience of trade financing in the Caribbean.

*Source: IMF Working Paper — Introduction section of “Mind the Gaps: Caribbean Trade Patterns and the Connectivity Constraint” (source PDF: wpiea2026095-source-pdf).*

### 2015. The worldwide banking-trade elasticity increases sharply to 0.92 percent, indicating that a one percent increase i

### Mind the Gaps: Caribbean Trade Patterns and the Connectivity Constraint

### Worldwide banking–trade elasticity
- The worldwide banking–trade elasticity increases sharply to 0.92 percent, indicating that a one percent increase in cross-border banking linkages increases bilateral trade almost one to one.

### Definition of financial centers used in analysis
- The study considers the United States, the United Kingdom, Switzerland, Germany, France, Japan, Singapore, the People's Republic of China—Hong Kong Special Administrative Region, Luxembourg, and the Netherlands as financial centers.

### Effects of banking relationships on tourism flows (estimation results)
- Baseline 2SLS regression results for bilateral tourism flows (Table 12) report:
  - Bilateral Bank Flows: 0.486*** (Specification 1), 0.487*** (Specification 2), 0.488*** (Specification 3)
  - Distance: -0.701*** (Specification 1), -0.645*** (Specification 2), -0.655*** (Specification 3)
  - Border: 1.478*** (Specification 1), 1.499*** (Specification 2), 1.495*** (Specification 3)
  - Language: 0.248* (Specification 1), 0.309** (Specification 2), 0.308** (Specification 3)
  - Colonial Ties: -1.976*** (Specification 1), -1.969*** (Specification 2), -1.946*** (Specification 3)
  - Bilateral Bank Flows x CARICOM: -0.563*** (Specification 2), -0.558*** (Specification 3)
  - Observations: 24,327 (Specification 1), 22,507 (Specification 2), 22,507 (Specification 3)
  - Origin FE: Yes (all)
  - Destination-Year FE: Yes (all)
  - Origin-Destination Clustered SEs: Yes (all)
  - R-squared: 0.522 (Specification 1), 0.537 (Specification 2), 0.536 (Specification 3)
- Note: Robust standard errors clustered at origin-destination country pair level are reported in parentheses. Significance: *** 1 percent; ** 5 percent; * 10 percent. The first-stage results of the 2SLS regressions are not reported for brevity.

### Interpretation of tourism results and elasticities
- Worldwide effect:
  - A one percent increase in cross-border banking flows increases tourism linkages by 0.486 percent (worldwide elasticity).
  - The positive worldwide effect likely reflects banking relationships supporting business-oriented travel (e.g., trade finance, corporate lending, government borrowing, FDI-related travel).
- CARICOM channel:
  - The combined effect for CARICOM economies is slightly negative at -0.076 percent and essentially zero.
  - Excluding commodity exporters (Guyana, Suriname, and Trinidad and Tobago) does not materially change the results.
  - For the subset of ECCU economies the interaction coefficient is not significantly negative and remains indistinguishable from zero; estimated imprecisely with a large standard error.
- Mechanisms explaining weaker CARICOM link:
  - A large share of tourism expenditures is pre-paid through intermediaries in origin countries and retained by foreign airlines, tour operators, and hotel chains before reaching destination economies (UNCTAD, 2010).
  - All-inclusive packages and cruise transactions processed through international operators; in-destination spending often occurs in cash or via international credit cards settled through major financial centers.
  - Cash transactions remain more prominent in the Caribbean compared to other regions (Survey of International Air Travelers and IDB, 2022).
  - Diaspora-related travel and remittance flows in the Caribbean are less reliant on formal interbank channels.

### Broader conclusions on connectivity constraints
- Physical connectivity as a central constraint:
  - Gravity model results suggest Caribbean goods trade is disproportionately sensitive to geographic distance.
  - Caribbean countries trade less with partners outside the region and more within the region than predicted by economic fundamentals alone.
  - The disproportionate effect of geographic distance suggests limited direct shipping routes and high freight costs.
  - High import dependence from concentrated source markets amplifies exposure to external shocks and shifts in global trade policies.
- Tourism and air connectivity:
  - Tourism flows in the Caribbean are sensitive to physical connectivity.
  - Controlling for standard demand factors, inbound flight capacity plays a quantitatively significant role in determining arrivals.
  - Supply-side constraints—small airports, limited carrier networks, dependence on a few connection hubs—materially affect tourism outcomes.
- Financial connectivity plays a more limited role:
  - Banking relationships facilitate trade globally but the link is weaker for Caribbean economies, consistent with the region’s export structure.
  - The channel by which banking relationships support tourism is muted for CARICOM economies due to the region’s tourism transaction structure.
  - Relative to financial linkages, physical connectivity plays a more prominent role in shaping Caribbean trade patterns and external vulnerabilities.

*Mind the Gaps: Caribbean Trade Patterns and the Connectivity Constraint — Working Paper No. WP/2026/95*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2026/english/wpiea2026095-source-pdf.pdf_