## _wp0971

## Source details

**Canonical URL:** [_wp0971](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp0971.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp0971.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp0971.pdf.json)

---

### Introduction and Objectives
- Close economic linkages between the United States and Caribbean countries; the Eastern Caribbean Currency Union (ECCU) shifted its peg from the pound sterling to the U.S. dollar in 1976.
- Tourism is the dominant channel of U.S. influence on tourism-dependent ECCU countries: about one third of stayover tourists to the ECCU are from the U.S.
- The U.S. accounts for less than 5 percent of ECCU exports.
- Foreign direct investment to the ECCU averaged about 21 percent of GDP over the last five years, with a significant portion coming from the U.S., concentrated in the tourism sector.
- Paper objectives:
  - Quantify effects of U.S. business cycles on ECCU economies and identify spillover channels.
  - Use two empirical procedures:
    - “Common trends and common cycles” decomposition (Vahid and Engle (1993)) of real GDP into trend and cycle.
    - Standard VAR analysis (along lines of Bayoumi and Swiston (2008)).

### Business Cycles, Stylized Facts, and Proposed Spillover Channels
- Literature and stylized facts:
  - Caribbean research has emphasized trend growth over fluctuations; short time series and data inadequacies constrain analysis.
  - Cashin (2006) findings:
    - Strong co-movement between Canadian and Caribbean classical business cycles; less synchronization with U.S. and U.K. classical cycles.
    - Caribbean classical cycles are asymmetric: long expansions and short, sharp contractions; growth cycles more symmetric.
    - Classical cycles are longer than in other middle-income and developed countries.
    - ECCU growth cycles more synchronized with Canadian growth cycles (development assistance, Canadian banks, remittances).
  - Kandil (2009) finds (annual real GDP 1975–2006):
    - Output increases more slowly in expansionary phases and contracts sharply in recessions.
    - Prices rise more quickly in expansions and deflation is less during contractions, consistent with a kinked supply curve and rigidities.
- Proposed U.S. → ECCU spillover channels:
  - Trade (primarily tourism services): U.S. is the largest source of tourist arrivals (about one third of stay-over arrivals).
  - Commodity prices: ECCU are net commodity importers; commodity prices unlikely a major U.S.-driven channel.
  - Financial markets / monetary policy: U.S. dollar peg transmits U.S. monetary policy; influence likely small given low elasticity of investment and consumption to interest rate changes. Grenade and Moore (2008) show long-run convergence between U.S. and ECCU interest rates; short-run Fed Funds changes have an almost immediate effect on ECCU interest rates.
  - Remittances: ECCU Fund members among world’s top-twenty in migration rates; large proportion of migrants live in the U.S. Remittances typically countercyclical; net effect depends on motives (altruism, exchange, insurance).
  - Official Development Assistance (ODA): ODA to ECCU has declined and the U.S. has not been a significant ODA source; ODA likely a minor channel.

### Data and Econometric Methodology
- Data:
  - Annual real GDP, 1963–2007, for Belize, Barbados, Guyana, Jamaica, six ECCU Fund members (Antigua and Barbuda, Dominica, Grenada, St. Kitts and Nevis, St. Lucia, St. Vincent and the Grenadines), Trinidad and Tobago, and the U.S., from IMF WEO database.
  - All real GDP series are I(1); their first differences are I(0).
  - ECCU analyzed both as an aggregated currency union (six Fund members aggregated) and at country level in VARs.
- Common trends and common cycles approach:
  - Based on Vahid and Engle (1993): cointegration representation, decomposition into permanent (trend) and transitory (cycle) components when n = r + s.
  - For first-order cointegration systems (k = 1) all such systems have common cycles.
  - Growth elasticities estimated by OLS following Roache (2008) using:
    - Δyi_cycle = αi + βi_cycle ΔyUS_cycle + γi_cycle ΔyUS_trend + ε
    - Δyi_trend = αi + βi_trend ΔyUS_cycle + γi_trend ΔyUS_trend + ε
- VAR analysis:
  - Basic VAR relates ECCU country growth Δyi,t to U.S. growth ΔyUS,t and lagged terms; Cholesky ordering assumes U.S. dominance.
  - Channels identified by augmenting VAR with potential channel variable j and computing channel contribution ji = ri - rji, where ri is impulse response from basic VAR and rji from VAR with channel j included.
  - Channels tested: trade (annual growth of stayover arrivals from the U.S.), financial (U.S. three-month T-bill rate and ten-year government bond yield), commodity prices (annual percent changes of WEO fuel and nonfuel commodity indexes). Remittances excluded due to data limitations.

### Empirical Results — Common Trends and Common Cycles
- Model selection and cointegration:
  - VAR lag selection: two of four criteria indicate lag order of one.
  - Trace and maximum eigenvalue tests suggest 3 cointegration vectors → implies 4 common trends among 7 GDP series.
  - First-order system with 3 cointegration vectors implies 4 cofeature vectors → 3 common cycles.
- Decomposition:
  - Each of the seven countries’ real GDP series decomposed into trend (4 common trends) and cycle (3 common cycles); results compared to HP filter counterparts.
- Growth elasticity estimates (OLS; selected figures preserved exactly as reported):
  - ECCU:
    - Simple growth correlation with the U.S. = 0.40***
    - Estimated cyclical elasticity to U.S. cycle = 0.92***
    - Trend elasticity to U.S. trend = 0.94***
  - Belize:
    - Simple growth correlation = 0.15
    - Trend elasticity to U.S. trend = 1.09***
  - Barbados:
    - Simple growth correlation = 0.21
    - Cycle elasticity to U.S. cycle = 0.21
    - Trend elasticity to U.S. trend = 1.64***
    - Additional reported entries: U.S. Cycle 1.03**, U.S. Trend -1.35***
  - Guyana:
    - Simple growth correlation = 0.06
    - Cycle elasticity to U.S. cycle = 1.25***
  - Jamaica:
    - Simple growth correlation = 0.10
    - Cycle elasticity to U.S. cycle = 0.10
    - Trend elasticity to U.S. trend = 0.36***
  - Trinidad and Tobago:
    - Simple growth correlation = -0.18
    - Cycle elasticity to U.S. cycle = -2.00**
    - Trend elasticity to U.S. trend = 1.10*
    - Additional reported trend elasticity = -1.46**
- Diagnostics (selected):
  - ECCU cycle equation: R-square 0.51, DW-stat 1.35, LM p-value 0.16 (examples of reported diagnostics across countries).
- Interpretation:
  - ECCU growth trend and cycle synchronize closely with U.S. movements (elasticities near 1).
  - Heterogeneity across countries:
    - Barbados and Trinidad and Tobago significantly affected by both U.S. trend and cycle, with varying directions.
    - Guyana and Barbados show cycle elasticities > 1 (amplified effects).
    - Jamaica relatively inelastic to U.S. fluctuations.
  - Contrast with Roache (2008) for Central America: in the Caribbean, some countries’ trends appear decoupled (Guyana, Jamaica), possibly due to domestic crises and policies.

### Empirical Results — VAR Evidence on ECCU Spillovers
- VAR setup and samples:
  - Annual data with lag 1.
  - Three sample periods: 1963–2007, 1976–2007 (post-peg), 1989–2007 (when stayover tourist arrivals from U.S. available).
- Magnitude and evolution of spillovers (ECCU aggregate response to a 1 percent U.S. growth shock):
  - 1963–2007: 0.4 percentage point in the first year.
  - 1976–2007: 0.7 percentage point in the first year.
  - 1989–2007: 0.8 percentage point in the first year.
  - Interpretation: impact of U.S. business cycles on ECCU strengthened over time.
- Individual ECCU country first-year responses to 1 percent U.S. growth shock:
  - 1963–2007: range from 0.3 to 0.7 percentage point.
  - 1976–2007: increased to 0.4–1.4 percentage point for individual countries.
  - 1989–2007: with the exception of St. Vincent and the Grenadines, estimated responses are higher than earlier periods.
- Channel identification:
  - Adding financial indicators or world commodity prices to VARs does not materially change impulse responses (ji ≈ i), suggesting financial and commodity channels account for little of the spillovers — consistent with limited financial integration despite the dollar peg.
  - Trade/tourism channel: limited evidence overall; for Antigua and Barbuda, annual growth of U.S. tourist arrivals explains about half of Antigua and Barbuda’s first-year response to a 1 percent U.S. growth shock. For the other five ECCU Fund member countries, adding U.S. tourist arrivals growth does not significantly change impulse responses.

### Conclusions and Policy Implications
- Main findings:
  - ECCU economies are very sensitive to both temporary (cycle) and permanent (trend) movements in the U.S. economy; linkages have strengthened over time.
  - U.S. monetary policy appears not to be an important transmission channel to the ECCU, reflecting stable ECCU interest rates and low spending elasticity to interest rate changes.
  - Tourism is a statistically important channel for only one ECCU country (Antigua and Barbuda).
  - Remittance and ODA channels likely limited or ambiguous given data and donor patterns.
- Policy implications and recommendations:
  - Fiscal policy must bear the burden of cyclical adjustment given high sensitivity to U.S. fluctuations and the ECCU’s hard peg to the U.S. dollar (limiting monetary independence).
  - Need for continued fiscal consolidation to reduce high public debt levels and create room for countercyclical fiscal policy in the future; current high debt and tight fiscal positions, exacerbated by food and fuel shocks, may constrain fiscal space.
  - Structural reforms to increase economic flexibility and reduce supply-side rigidities:
    - Address inflexible labor markets, product market imperfections, and investment-climate inefficiencies to improve supply response and reduce business cycle asymmetries.
  - Heterogeneous responses across Caribbean economies imply challenges for deeper regional monetary integration:
    - Proposed monetary unions including Trinidad and Tobago and ECCU members could be challenging; monetary policy under a common currency would need to account for disparate responses and likely require fiscal policy complements.

*Source: _wp0971*

### 1.  Summary Statistics of Real GDP Growth..........................................................................9

### 1.  Summary Statistics of Real GDP Growth..........................................................................9

### Major sections
- 1.  Summary Statistics of Real GDP Growth..........................................................................9
- 2.  VAR Lag Order Selection................................................................................................10
- 3.  Tests for the Number of Cointegrating Vectors...............................................................10
- 4.  Growth Elasticities in the Caribbean ...............................................................................12
- 5.  Diagnostics of Growth Elasticity Models ........................................................................13

### Figures (listed)
- 1.  Three Common Cycles ....................................................................................................16
- 2.  Four Common Trends ......................................................................................................17
- 3.  Caribbean Countries: Cyclical Components of Real GDP  .............................................18
- 4.  Caribbean Countries: Trend Components of Real GDP ..................................................19
- 5.  ECCU: Responses to One Percent U.S. Growth Shock...................................................20
- 6.  ECCU: Country Responses to One Percent U.S. Growth Shock.....................................21
- 7.  Antigua and Barbuda: Responses to One Percent U.S. Growth Shock ...........................22

*Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp0971.pdf*

### References .............................................................................................................

### _wp0971 - References .............................................................................................................

### Introduction
- Close economic linkages exist between the United States and Caribbean countries; the Eastern Caribbean Currency Union (ECCU) shifted its peg from the pound sterling to the U.S. dollar in 1976.
- Tourism is the dominant channel of U.S. influence on tourism-dependent ECCU countries: about one third of stayover tourists to the ECCU are from the U.S.
- The U.S. accounts for less than 5 percent of ECCU exports.
- Foreign direct investment to the ECCU averaged about 21 percent of GDP over the last five years, with a significant portion coming from the U.S., concentrated in the tourism sector.
- The paper’s objectives:
  - Quantify effects of U.S. business cycles on ECCU economies and identify spillover channels.
  - Use two empirical procedures: the “common trends and common cycles” (Vahid and Engle (1993)) decomposition of real GDP into trend and cycle, and standard VAR analysis (along lines of Bayoumi and Swiston (2008)).

### Business Cycles and Spillovers
- Literature and stylized facts:
  - Caribbean literature has focused more on trend growth than fluctuations; short time series and data inadequacies are constraints.
  - Cashin (2006) using an ideal band-pass filter found:
    - Strong co-movement between Canadian and Caribbean classical business cycles; less synchronization with U.S. and U.K. classical cycles.
    - Caribbean classical cycles are asymmetric: long expansions and short, sharp contractions; growth cycles are more symmetric.
    - Classical cycles are longer than in other middle-income and developed countries.
    - ECCU growth cycles more synchronized with Canadian growth cycles (development assistance, Canadian banks, remittances).
  - Kandil (2009) (annual real GDP 1975–2006) finds:
    - Output increases more slowly in expansionary phases and contracts sharply in recessions.
    - Prices rise more quickly in expansions and deflation is less during contractions, consistent with a kinked supply curve and rigidities.
- Proposed spillover channels from the U.S. to the ECCU:
  - Trade (primarily tourism services): U.S. is the single largest source of tourist arrivals (about one third of stay-over arrivals).
  - Commodity prices: ECCU are net commodity importers; major exports historically to the U.K.; commodity prices unlikely to be a major U.S.-driven channel.
  - Financial markets / monetary policy: peg to the U.S. dollar transmits U.S. monetary policy; however, influence likely small given low elasticity of investment and consumption to interest rate changes. Grenade and Moore (2008) show long-run convergence between U.S. and ECCU interest rates; short-run Fed Funds changes have an almost immediate effect on ECCU interest rates.
  - Remittances: ECCU Fund members among world’s top-twenty in migration rates; large proportion of migrants live in the U.S. Remittances are typically countercyclical; net effect depends on motives (altruism, exchange, insurance).
  - Official Development Assistance (ODA): ODA to ECCU has declined and U.S. has not been a significant ODA source; thus ODA likely a minor channel.

### Econometric Methodology and Data
- Common trends and common cycles approach:
  - Based on Vahid and Engle (1993): cointegration representation, decomposition into permanent (trend) and transitory (cycle) components when n = r + s.
  - For first-order cointegration systems (k = 1) all such systems have common cycles.
  - Growth elasticities estimated by OLS following Roache (2008) using equations:
    - Δyi_cycle = αi + βi_cycle ΔyUS_cycle + γi_cycle ΔyUS_trend + ε
    - Δyi_trend = αi + βi_trend ΔyUS_cycle + γi_trend ΔyUS_trend + ε
- VAR analysis:
  - Basic VAR relates ECCU country growth Δyi,t to U.S. growth ΔyUS,t and lagged terms; Cholesky ordering assumes U.S. dominance.
  - Channels identified by augmenting VAR with potential channel variable j and computing channel contribution ji = ri - rji, where ri is impulse response from basic VAR and rji from VAR with channel j included.
- Data:
  - Annual real GDP data for 1963–2007 for Belize, Barbados, Guyana, Jamaica, six ECCU Fund members (Antigua and Barbuda, Dominica, Grenada, St. Kitts and Nevis, St. Lucia, St. Vincent and the Grenadines), Trinidad and Tobago, and the U.S., from IMF WEO database.
  - All real GDP series are I(1) and their first differences are I(0).
  - ECCU analyzed both as aggregated currency union (six Fund members aggregated) and at country level in VARs.

### Empirical Results
A. Caribbean Common Trends and Common Cycles
- Model selection and cointegration:
  - VAR lag selection: two of four criteria indicate lag order of one.
  - Trace and maximum eigenvalue tests suggest 3 cointegration vectors → implies 4 common trends among 7 GDP series.
  - First-order system with 3 cointegration vectors implies 4 cofeature vectors → 3 common cycles.
- Decomposition:
  - Real GDP series for each of the seven countries decomposed into trend (4 common trends) and cycle (3 common cycles); plots compare to HP filter counterparts.
- Growth elasticities (OLS estimates; Table 4 summarized):
  - ECCU: simple growth correlation with the U.S. = 0.40***; estimated cyclical elasticity to U.S. cycle = 0.92***; trend elasticity to U.S. trend = 0.94*** (elasticities close to 1).
  - Belize: simple growth correlation 0.15; trend elasticity to U.S. trend = 1.09***.
  - Barbados: simple growth correlation 0.21; cycle elasticity to U.S. cycle = 0.21; trend elasticity to U.S. trend = 1.64***; additional reported: U.S. Cycle 1.03**, U.S. Trend -1.35*** (table shows multiple entries by row).
  - Guyana: simple growth correlation 0.06; cycle elasticity to U.S. cycle = 1.25***.
  - Jamaica: simple growth correlation 0.10; cycle elasticity to U.S. cycle = 0.10; trend elasticity to U.S. trend = 0.36***.
  - Trinidad and Tobago: simple growth correlation -0.18; cycle elasticity to U.S. cycle = -2.00**; trend elasticity to U.S. trend = 1.10*; additional reported trend elasticity -1.46** (table notation indicates mixed signs and significance).
  - Diagnostics (Table 5): R-square and DW-statistics reported for cycle and trend equations; selected p-values for LM autocorrelation tests listed (e.g., ECCU cycle R-square 0.51, DW-stat 1.35, LM p-value 0.16).
- Interpretation:
  - ECCU growth trend and cycle synchronize closely with U.S. movements (elasticities near 1).
  - Heterogeneity across Caribbean countries: Barbados and Trinidad and Tobago significantly affected by both U.S. trend and cycle, with varying directions; Guyana and Barbados show cycle elasticities > 1 (amplified effects); Jamaica remains relatively inelastic to U.S. fluctuations.
  - Results contrast with Roache (2008) for Central America where cycles are U.S.-driven but trends are regional; in the Caribbean some countries’ trends appear decoupled (Guyana, Jamaica), possibly due to domestic crises and policies.

B. Spillovers from the U.S. to the ECCU (VAR evidence)
- VAR setup:
  - Annual data with lag 1 due to lack of quarterly data.
  - Three sample periods estimated to assess evolution: 1963–2007, 1976–2007 (post-peg), and 1989–2007 (from when stayover tourist arrivals from U.S. are available).
  - Channels tested: trade (captured by annual growth of stayover arrivals from the U.S.), financial (U.S. three-month T-bill rate and ten-year government bond yield), commodity prices (annual percent changes of WEO fuel and nonfuel commodity indexes). Remittances excluded due to data limitations.
- Magnitude and evolution of spillovers:
  - ECCU aggregate response to a 1 percent U.S. growth shock:
    - 1963–2007: 0.4 percentage point in the first year.
    - 1976–2007: 0.7 percentage point in the first year.
    - 1989–2007: 0.8 percentage point in the first year.
    - Interpretation: impact of U.S. business cycles on ECCU strengthened over time.
  - Individual ECCU country responses (first-year responses to 1 percent U.S. growth shock):
    - 1963–2007: range from 0.3 to 0.7 percentage point.
    - 1976–2007: increased to 0.4–1.4 percentage point for individual countries.
    - 1989–2007: with the exception of St. Vincent and the Grenadines, estimated responses are higher than earlier periods.
- Channel identification results:
  - Adding financial indicators or world commodity prices to VARs does not materially change impulse responses (ji ≈ i), suggesting financial and commodity channels account for little of the spillovers—consistent with limited financial integration despite the dollar peg.
  - Trade/tourism channel: limited evidence overall; for Antigua and Barbuda (largest ECCU economy), annual growth of U.S. tourist arrivals explains about half of Antigua and Barbuda’s first-year response to a 1 percent U.S. growth shock. For the other five ECCU Fund member countries, adding U.S. tourist arrivals growth does not significantly change impulse responses.

### Conclusions and Policy Implications
- Main findings:
  - ECCU economies are very sensitive to both temporary (cycle) and permanent (trend) movements in the U.S. economy; linkages have strengthened over time.
  - United States monetary policy appears not to be an important transmission channel to the ECCU, reflecting stable ECCU interest rates and low spending elasticity to interest rate changes.
  - Tourism is a statistically important channel for only one ECCU country (Antigua and Barbuda).
  - Remittance and ODA channels likely limited or ambiguous given data and donor patterns.
- Policy implications and recommendations:
  - Given high sensitivity to U.S. fluctuations and the ECCU’s hard peg to the U.S. dollar (limiting monetary independence), fiscal policy must bear the burden of cyclical adjustment.
  - Need for continued fiscal consolidation to reduce high public debt levels and create room for countercyclical fiscal policy in the future, noting that current high debt and tight fiscal positions, exacerbated by food and fuel shocks, may constrain fiscal space.
  - Structural reforms to increase economic flexibility and reduce supply-side rigidities are important:
    - Address inflexible labor markets, product market imperfections, and investment-climate inefficiencies to improve supply response and reduce business cycle asymmetries.
  - Heterogeneous responses across Caribbean economies imply challenges for deeper regional monetary integration:
    - Proposed monetary unions including Trinidad and Tobago and ECCU members could be challenging; monetary policy under a common currency would need to account for disparate responses and likely require fiscal policy complements.

*Source: _wp0971 - References .............................................................................................................*

### References

### References

### Cited works
- Cashin, P., 2006, “Key Features of Caribbean Business Cycles,” in R. Sahay, D. Robinson and P. Cashin (eds.), The Caribbean: From Vulnerability to Sustainability, Washington DC: International Monetary Fund, pp. 83–121.
- Craigwell, R. and A. Maurin, 2002, “Production and Unemployment Cycles in the Caribbean: The Case of Barbados and Trinidad and Tobago,” (unpublished, Research Department, Central Bank of Barbados).
- Grenade, K. and W. Moore, 2008. "Co-Movements Between Foreign and Domestic U.S. Interest Rates In a Fixed Exchange Rate Regime: The Case Of The ECCU and U.S.," Applied Econometrics and International Development, Euro-American Association of Economic Development, Vol. 8(1), pp. 119–30.
- Hernandez, J., 2004, “Business Cycles in Mexico and the United States: Do They Share Common Movements” Journal of Applied Economics, Vol. VII, No. 2, pp. 303–22.
- Kandil, M., 2009, “Does Demand Volatility Lower Growth and Raise Inflation? Evidence from the Caribbean,” forthcoming IMF Working Paper.
- Kluyev, V., 2008, “Real Implications of Financial Linkage between Canada and the United States,” IMF Working Paper, WP/08/23.
- Mamingi, N., 1999, “Testing for Convergence and Common Features in International Output: The Case of the Eastern Caribbean Countries,” Journal of Eastern Caribbean Studies, Vol. 24, pp. 15–40.
- Murray, A., 2007, “Modeling the Jamaican Business Cycle: A Structural Vector Autoregression Approach.” (Unpublished paper presented at the 38th Conference of the Caribbean Centre for Money and Finance).
- Mishra, P., 2006 “Emigration and Brain-Drain: Evidence from the Caribbean,” IMF Working Paper, WP/06/25.
- Rappoport, H., and F. Docquier, 2004, “The Economics of Migrant Remittances,” Stanford Center for International Development Working Paper, No. 236.
- Roache, S., 2008, “Central America’s Regional Trends and U.S. Cycles” IMF Working Paper, WP/08/50.
- Swinston, A., and T. Bayoumi, 2008, “Spillovers Across NAFTA,” IMF Working Paper, WP/08/3.
- Vahid, F. and R. F. Engle, 1993, “Common Trends and Common Cycles”, Journal of Applied Econometrics, Vol. 8, No. 4, pp. 341–60.

*Source: _wp0971 - References*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp0971.pdf_