Journal Issue
Share
Article

Jamaica

Author(s):
International Monetary Fund
Published Date:
June 2008
Share
  • ShareShare
Show Summary Details

I. Jamaica and the External Environment1

A. Introduction

1. This chapter examines the impact on Jamaica’s economy of external economic slowdowns and credit crunches. The knock-on macroeconomic implications of the unfolding global credit crisis and slowdown in U.S. economic growth can potentially have large spillover effects for Jamaica. Jamaica’s economy is highly open, with external trade amounting to almost 100 percent of GDP (Figure 1). Remittances from the diaspora (mainly in North America and the United Kingdom) amount to a further17 percent of GDP annually. Moreover, Jamaica is highly reliant on external financing, given its large current account deficit (15½ percent of GDP) and high public debt (128 percent of GDP). Jamaica is, therefore, significantly exposed to the risks of weakened external economic conditions and to deteriorating financial market sentiments.

Figure 1.Jamaica: Global Links and Vulnerabilities

Sources: Bloomberg; JP Morgan; country authorities; and Fund staff estimates.

2. The study is purely statistical and based on historical relationships between variables. The paper derives econometric estimations of the behavior of various U.S. variables as well as of historical links between U.S. and Jamaican variables. The estimations are then used to evaluate the possible impact on Jamaica of various scenarios for the U.S. variables by shocking the latter in the estimated equations. While it identifies the direction and magnitude of the relationships between different variables, a study of this type, by itself, does not explicitly account for the economic rationale for such relationships. In most instances, the identified relationships do, however, make intuitive economic sense and the nature of the relationships are conjectured but they remain to be unambiguously established.

3. Furthermore, the relationships that existed in the past may not hold going forward. Jamaica is a shock-prone economy. The sheer number, variety, and frequency of shocks impacting on it make it difficult to accurately capture all the dynamic relationships between variables and, at the same time, retain econometric tractability. The possible existence of structural breaks limits the model’s predictive powers. Also, it is possible that the estimated parameters are not stable. For example, with global financial integration, financial flows to Jamaica may be affected by external shocks differently than they have in the past. Indeed, the transmission of the U.S. subprime debacle to other financial systems is unprecedented and unpredictable.

4. The study can, nevertheless, be useful in a number of ways. First and foremost, it can provide broad indications of the possible direction and perhaps even magnitude of some potential shocks currently. It can, therefore, help policy-makers identify areas to especially watch out for. Second, by providing insight into how variables have responded in the past to a deterioration of the external environment, the study may help identify how the transmission mechanism to domestic variables may function, going forward.

5. The analysis finds that the impact of U.S. macroeconomic stress is small on Jamaica’s GDP but larger on the budget, exports and capital flows. The historical data shows that while the impact of U.S. slowdowns on Jamaican growth has been limited, they have tended to be associated with a worsening primary fiscal balance. In addition, previous episodes of external slowdowns and credit crunches have been associated with a persistent reduction in capital inflows to Jamaica. The shocks have reduced remittances and export receipts but also imports, thereby limiting the overall impact on the current account.

6. The remainder of the chapter is organized in three sections. The following section (B) examines the impact of U.S. economic slowdowns and credit crunches on Jamaica’s balance of payments. The next section (C) extends the analysis to Jamaica’s GDP and fiscal variables. Section D concludes.

B. Impact on Jamaica’s Balance of Payments

7. The empirical strategy consists of measuring the difference in value that a selected group of Jamaican variables take under two alternative scenarios for the U.S. economy. The first scenario assumes no shocks to the U.S. variables. The second assumes negative shocks, of various pre-specified magnitudes, to specific U.S. variables. The difference in the Jamaican variables between the two scenarios provides the estimated impact on Jamaica of the change in external conditions.

Model

8. The analysis consists of four distinct steps (Figure 2). These are: (i) estimating a purely U.S. economy dynamic model to generate two forecasts for U.S. variables—one with no further shocks and one with an initial negative shock to a specified U.S variable; (ii) an intermediate technical step to calculate principal components—the principal components retain the statistical information derived in (i) but overcome the problem of multicolinearity; (iii) regressing various Jamaican balance of payments variables on the principal components from the U.S. model to obtain two sets of forecasts for the Jamaican variables (one based on no shocks to the U.S. economy and one with the initial shocks); and, (iv) calculating the difference between the two sets of forecasts for the Jamaican variables derived in (iii).

Figure 2.Empirical Strategy

9. In the first step, the forecasts for the U.S. economy are derived from a VAR(3) equation. It takes the form

where yt is a vector of endogenous U.S. variables including GDP, money market interest rate, money stock, stock market index, private consumption, bank credit to the private sector, wages, imports of goods and services, as of year t. All variables are in real terms and expressed in logarithms, except for the interest rate which is not in logs. αi (i=0, …,3) are matrices of coefficients.εt is a vector of Gaussian error terms. This set of variables was chosen according to two criteria. Either they are variables meriting particular consideration in the current global environment (measures of U.S. economic activity and of credit market conditions) or variables that are á priori likely to be the main channels of transmission of shocks to Jamaica (for example U.S. imports of goods and services and U.S, real wages, which could affect remittances to Jamaica from the diaspora). Appendix I shows tests that support the econometric specification used, which assumes the existence of a long-run relation (cointegration) among the U.S. variables.

10. The U.S. VAR model is then used to produce two five-year dynamic forecasts. The forecasts span the period 2008–12.2 The first forecast, assumes there are no further shocks. In the second forecast we apply a negative shock to one of the U.S. indicators for 2008 and let the shock work its way through the other variables in the dynamic model. The size of the shock is set to be equal to two-standard deviations of the growth rate of the variable shocked. The VAR framework ensures that after one variable is shocked, other variables move along according to the historical patterns, including depth, persistence, and co-movement.

11. The second step of the process produces forecasts of the U.S. principal components under both the “no shocks” and shocked scenarios. This intermediate step is purely technical. The high degree of correlation among the U.S. variables results in a significant multicolinearity problem, which precludes using them directly as explanatory variables in equations for the Jamaican variables in step three. In particular, the multicolinearity leads to estimated coefficients being highly unstable, and therefore, rendering the estimated impact on the Jamaican variables as unreliable. To tackle this problem, we decompose the eight U.S. variables into their eight principal components. The principal components capture the common movements of the different U.S. indicators and are uncorrelated by construction, therefore avoiding the multicolinearity problem. However, critically, they still retain the information embedded in the U.S. indicators.

12. The third step links important variables in Jamaica’s balance of payments to the U.S. forecasts. The Jamaican balance of payments variables considered are net capital inflows, the current account balance, exports of goods and services (in real terms), gross current transfers from abroad and foreign direct investment.3 This set of indicators captures both current and capital account flows and includes specific line items that feature importantly in Jamaica’s balance of payments (FDI and transfers from abroad).

13. The link is established by estimating equations that assume the existence of long-term relationships between each Jamaican variable and the principal components of the U.S. variables.4 Statistically, the hypothesis of the existence of long-run relationships, or cointegration, could not be rejected, therefore, providing empirical support to our assumption. Intuitively, this finding implies that the U.S. and Jamaican indicators are not expected to drift too far apart from each other over time. Econometrically, the existence of cointegration is important to ensure that the estimated equations linking the Jamaican variables to the developments in the U.S variables are valid, i.e., the relationships are not spurious. Appendix I also presents a residual-based test that supports this specification.5 The existence of long-run relationships can also be rationalized on theoretical grounds. For example, if the number of Jamaican immigrants in the U.S. remains stable over the long term, then transfers should be expected to grow in line with U.S. GDP. Similarly, if the demand for tourism services in Jamaica by Americans grows along with real U.S. GDP in the long term, then FDI into the tourism sector in Jamaica can also be expected to grow in line with U.S. GDP.

14. The cointegration equations are used to compute the two sets of forecasts for each of the Jamaican balance of payments variables of interest. Five-year estimates of each of the Jamaican variables of interest (notionally covering 2008–13) are produced by using either the “no-shocks” or shock-based forecasts for the U.S. economy in the cointegration equations.

15. The final step is to take the difference in the two sets of forecasts for the Jamaican variables of interest. The impact on the Jamaican variable of interest of a shock on any given U.S. indicator is, thus, simply this difference between the two forecasts. Note however, that the forecasted levels of either the U.S. or Jamaican variables are not critical for the results—only their difference is. Furthermore, the difference reflects entirely the impact of the shock and nothing else but the shock. In other words, had the projected path for the U.S. economy been lower in the “no shock” scenario, then the deviation from that path under the shocked projection would have also been lower by the same magnitude. Thus, the realism of the projected levels is not relevant to the results as long as the difference accurately captures the shock, which the four-step process described above helps ensure.

16. That is not to say, however, that the results do not have any limitations. The framework is unlikely to capture completely the dynamics under the current global financial conditions and their potential impact on the real economy. This is because the model is entirely based on the average historical behavior of the various series. However, (i) the world has experienced significant changes over the sample period—in other words, there may have been structural breaks, and, therefore, the true values of the parameters may now be different from those estimated; (ii) there are other sources of fluctuations in the data than those examined here and, therefore, the dynamics of the historical data might not be representative of those generated by the specific source of current fluctuations; and (iii) there might be nonlinearities in the relations studied that are not accounted for in our estimations (for example, if the size of the shock matters for the expected response of the variables). Finally, there are the usual problems of small sample size and model specification.

Results

17. Negative shocks to the U.S. economy are expected to reduce capital inflows, exports, and current transfers to Jamaica. This pattern is consistent irrespective of where the initial shock occurs, whether for example to U.S. GDP growth or to credit to the private sector in the United States. The impact on Jamaica, however, differs in magnitude from shock to shock (Figures 36).

Figure 3.Jamaica: Impact of Deceleration in U.S. Growth and Private Consumption

Sources: International Financial Statistics, IMF; calculations and estimates.

Figure 4.Jamaica: Impact of Deceleration in U.S. Wages and Imports

Sources: International Financial Statistics, IMF; country authorities; and Fund staff calculations and estimates.

Figure 5.Jamaica: Impact of Deceleration in U.S. Credit to the Private Sector and Money Balances

Sources: International Financial Statistics, calculations and estimates.

Figure 6.Jamaica: Impact of Decline in U.S. Interest Rate and Asset Prices

Sources: International Financial Statistics, IMF; country authorities; and Fund staff calculations and estimates.

18. Lower U.S. growth leads to lower capital inflows but the various impacts on the current account, while each significant, broadly cancel out in the aggregate. Lower U.S. growth, by 2 standard deviations, leads capital inflows to Jamaica to decline in the model by about US$170 million cumulatively over a five-year period, mostly due to lower FDI flows. The current account, however, changes only marginally. The model does predict a sharp decline in exports as well as in current transfers from abroad. The latter, however, along with the decline in capital inflows, likely reduces available financing for imports, thereby explaining the model’s prediction of a minimal impact on the current account. This is an intuitive result, as current transfers from abroad (mostly workers’ remittances) can be expected to finance consumer imports, while FDI inflows to Jamaica have historically had a large import component. Furthermore, the price of oil, a major import item for Jamaica, can be expected to weaken with the U.S. slowdown, further reducing the latter’s adverse impact on Jamaica’s current account.

19. Declines in U.S. private consumption and imports of goods and services have a broadly similar impact, but of a lower magnitude, on Jamaica’s balance of payments. In the case of private consumption, the lesser impact (compared to the impact of the U.S. growth slowdown) is likely due to the more stable nature of U.S. consumption along the business cycle. In the case of U.S. imports, the results may reflect factors not included in the model. For example, under the current global conditions, a decline in U.S. demand for Jamaican goods (income effect) may be counterbalanced by the depreciation of the Jamaican dollar vis-à-vis currencies other than the U.S. dollar, which has made Jamaica more competitive to tourists from Canada and Europe (substitution effect).

20. Initial shocks applied to U.S. wages have a limited impact on Jamaica’s capital account, but lead to more of a deterioration in the current account than do other shocks. This makes intuitive sense as wage-specific U.S. shocks are most likely to affect remittances to Jamaica. On the other hand, a direct adverse impact on the capital account is unlikely. As the impulse works its way through other U.S. variables, however, some impact from those variables to capital inflows into Jamaica can also be expected.

21. When the initial shock is to credit, the decline in Jamaica’s capital inflows appear larger. Shocks to credit to the private sector in the U.S. show the largest impact, with an estimated decline in capital inflows to Jamaica by about US$500 million (roughly three times larger than the impact of GDP shock) cumulatively over five years. While significant, this figure, however, has to be compared against annual capital inflows into Jamaica in the range of US$1.5–1.9 billion in recent years. The impact on Jamaica’s current account of a shock on credit is in the aggregate similar in magnitude to that of a shock to U.S. GDP growth. However, there are larger impacts on individual components of the current account—significantly larger declines in current transfers from abroad (of about US$1.3 billion cumulatively) and exports (with an estimated deceleration of 7 percentage points), are broadly offset by a decline in imports also of a larger magnitude.

22. Declines in U.S. money market interest rates exhibit a relatively minor impact on Jamaica’s external sector. The results indicate a deceleration in net capital inflows of less than US$100 million over five years (the subcomponent—FDI flows—increase). The estimation also indicates almost no impact on the current account, exports and current transfers.

23. The historical data suggests that declines in asset prices in the US do not appear to have had a strong impact on Jamaica. Shocks to the U.S. stock market index, which can be interpreted as a proxy for asset prices in the U.S. in general, lead to a relatively low deceleration in net capital inflows to Jamaica. Specifically, capital inflows fall by an estimated US$124 million cumulatively over a five-year period, mostly due to lower FDI flows. The results also indicate that there is almost no impact on the current account, current transfers, and exports.

24. The above-mentioned relation between U.S. asset prices and Jamaica, however, needs to be interpreted with caution. Consumption responses to sharp asset price declines experienced by those who can afford to invest in stock markets are likely to be different from those whose only substantial asset is their home. Currently, there is an asset deflation in both the real estate and stock markets in the United States, but that has not always been the case in the sample period for this study. Thus, using the stock market index as a proxy for asset prices may not be a good predictor of the impact of the current U.S. economic stress.

C. Impact on Jamaica’s GDP and Fiscal Accounts

25. This section extends the analysis to the impact on Jamaica’s GDP and government finances of changes in the external environment. The same U.S. variables as in the preceding section are used as representative of external conditions, based on the same arguments. The external indicators to be shocked are chosen according to the same criteria. However, a richer model that also captures the dynamic relations between different Jamaican variables is used. As Jamaica’s external variables are the first set of variables to be affected by external developments, it made sense above to regress each variable on external conditions separately. Developments in Jamaica’s GDP and the fiscal accounts, however, reflect a complex set of interlinkages, not only with external variables but also between the Jamaican variables. The model used in this section seeks to capture that complexity.

Model

26. The empirical strategy is to compare two forecasts for the Jamaican economy from the following VAR(1) model:

where Y is a vector of endogenous Jamaican variables, X is a vector of exogenous variables that includes indicators of the U.S. economy as well as Jamaica’s inflation, ε is an error term, and the alphas are matrices of coefficients. Sub-index t indicates year. The first set of forecasts assumes no shock to the U.S. variables whereas the second set does. The Jamaican endogenous variables included in Y are GDP, real exchange rate, the treasury bill interest rate spread over the U.S. treasury bill, government primary expenditure and government revenues, expressed in real terms and in logarithms (except for the interest rate spread that is expressed in levels). Introducing the Jamaican variables in levels requires cointegration among them, a hypothesis supported by econometric testing (see Appendix II). The lag structure of the VAR allows the endogenous variables to provide feedback to each other.

27. The treatment of the real exchange rate, sovereign spread, and inflation merits some discussion. The real exchange rate and the sovereign spread are included as endogenous because they are both important channels of transmission from abroad and they reflect feedback effects from GDP and the fiscal accounts; i.e., the Jamaican variables of interest. A strong argument can be made that inflation should also be treated as endogenous. However, treating it as exogenous allows the model to control for periods of above-average volatility and instability that is not related to developments in the global economy, and which introduce noise in the estimation. Two main sources of such instability in Jamaica are natural disasters (which are very frequent, lead to agricultural damage and a spike in the heavily weighted food component of inflation) and episodes of speculative pressure related to domestic, rather than external, developments.

28. The impact on Jamaica of external shocks is the difference, one and two years forward, between the results of the two forecasts produced with the above equation. In the first set of forecasts, the U.S. variable of interest is kept at a predetermined path and in the second, it is set at a negative two-standard deviation of its growth rate. The one-step-ahead difference provides a measure of the contemporaneous impact on Jamaica of the external shock while the two-step-ahead forecast allows for feedback effects within the Jamaican economy from one Jamaican variable to the next.6 Note critically again that like in the preceding section, the results are not affected by either the variables’ forecasted levels or by the assumed paths of exogenous variables. This is because the impact effect is estimated as the difference between the two forecasts. Again, we take data and/or WEO estimates through 2007 as given and the forecasts are for the following two years.

Results

29. The model suggests that U.S. economic stress has generally been associated with a deterioration of Jamaica’s primary fiscal balance. The primary balance worsens in almost all cases. This is the case, whether the initial shock is applied to U.S. real variables (GDP, consumption, imports) or to financial variables such as (credit to the private sector, stock market, and interest rates).7Figures 7 and 8 visually display the size of the deterioration in Jamaica’s GDP, real revenues, real primary expenditures; and the primary balance stemming from shocks to different U.S. variables. The deterioration for GDP, real revenues and real primary expenditures are measured in terms of growth rates, in percent. For example, a bar value of -1 for real revenues indicates that it grows at a rate that is 1 percent lower in the scenario with a shock than in the one with no shock. Similarly, a value of -1 for GDP growth indicates that real GDP growth is 1 percentage point lower in Jamaica with a shock than without. The impact on the primary balance is, however, measured in terms of percentage points of GDP.

Figure 7.Impact on Jamaican GDP and Fiscal Sectors I

Sources: Country authorities; and Fund staff calculations.

Figure 8.Impact on Jamaican GDP and Fiscal Sectors II

Sources: Country authorities; and Fund staff calculations.

30. The deterioration in the primary balance is caused in large part by lower revenues. In all but one case (that of a shock to the stock market), revenues decline in Jamaica. The model does not explicitly account for the channels of transmission, as it relies purely on observed co-movements in the historical data. However, this result is not entirely surprising. While Jamaica’s tourism sector enjoys numerous tax exemptions, the direct and indirect impact on revenues of a decline in tourism that could be caused by economic stress in the U.S. is still likely to be significant. Some of this impact could show up on taxes on imports. A quarter of tax collections in Jamaica are based on imports, which the preceding section suggests declines with macroeconomic stress in the U.S. Also, revenues from bauxite could decline if U.S. slowdowns cause global commodity prices to decline.

31. The model generally predicts increases in primary expenditures in response to shocks, but the pattern is less clear than in the case of revenues. The contemporaneous impact (the first year) is generally an increase in expenditures. This is, however, not true for shocks to imports and U.S. real wages. Furthermore, in some other instances, the direction of impact switches from the first year to the second. In sum, while the statistical analysis suggests that expenditure policy has generally been countercyclical, at least at the outset of an external shock, it is not entirely clear why. That this has been a policy choice to limit the impact on the domestic economy of the shocks is very likely.

32. In contrast to the fiscal position, external shocks have generally had only a limited impact on GDP growth in Jamaica. Historically, Jamaica’s GDP has shown little variability. Since 1993, Jamaica has not grown faster than 2½ percent. On the other hand, the economy has not shrunk on a year-to-year basis either, except during the banking crisis years of 1997–98 when annual contractions were limited to 1¼ percent. What variability there is in growth appears due to natural disasters. Simple growth correlations with the rest of the Caribbean, Latin America and the United States for 1991–2006 are statistically insignificant. Figures 7 and 8 show that with the exception of those cases where shocks are applied to credit to the private sector or real wages, GDP growth in Jamaica would most likely not be expected to contract very sharply in response to U.S. macroeconomic stress. The adverse effect on Jamaican growth of negative shocks to the U.S. variables may also have been limited by the countercyclical nature of Jamaica’s fiscal policies, as outlined in the preceding paragraphs.

33. A decline in credit in the U.S. economy or a decline in U.S. wages have a particularly sharp negative impact on Jamaica. In the scenario of a credit decline (third chart in Figure 7), the model predicts lower revenue growth of more than 7 percentage points in the first year. Combined with increased expenditure, this leads to a deterioration of the primary balance by about 2 percentage points of GDP. The adverse impact on growth is also more pronounced than in other cases. As regards real wages, the model suggests that lower U.S. wage growth would result in the loss of more than 1½ percentage points of GDP growth in Jamaica and revenues lower by 4 percent. Lower U.S. wages are likely to affect Jamaica in two ways, explaining the larger magnitude of the impact. First, there is the direct impact on remittances and, hence, on aggregate demand in Jamaica. Second, lower U.S. wages will reduce U.S. demand for foreign goods and services, including from Jamaica, and thereby, adversely affect Jamaican growth.

D. Concluding Remarks

34. The historical data show statistical relationships suggesting that U.S. economic stress and global financial turmoil can adversely affect Jamaica’s balance of payments and fiscal accounts. U.S. slowdowns have been associated with a worsening of the primary fiscal balance in Jamaica but have not had a very significant economic growth impact. However, a decline in credit to the economy in the U.S. does appear to affect growth in Jamaica in a significant way. Furthermore, credit crunches, as well as slowdowns in U.S. GDP growth, are associated with a persistent reduction in capital inflows. The results indicate that while export receipts and remittances can decline sharply, the impact on the overall current account balance is likely to be limited.

35. There are a number of limitations to the study but the results are broadly intuitive. As indicated above, the econometric models could be misspecified; possible nonlinear effects may not be accounted for; and, the coefficients’ stability may be an issue, given the changing nature of the global financial structure and the degree of integration of Jamaica with the rest of the world. It is, therefore, important to exercise caution in interpreting the results. Importantly, the purely statistical nature of the models means that they do not explicitly account for the economic relationships and links among the variables (the results are based purely on observed historical co-movements in the data). In most cases, one can discern, from theoretical considerations, why the variables move in the fashion they do. This is, however, not possible in all instances, given the complexity of the real world and the limited number of variables included in the study for tractability reasons. Future research should focus on delving more deeply into why some of the variables move the way they do.

Appendix I. Section B Econometric Tables

These tables provide support to some of the assumptions in Section B. Table 1A tests for cointegration among the U.S. VAR model. Table 1B is a residual-based test for cointegration between the U.S. economy principal components and the Jamaican balance of payments indicators.

Table 1A:U.S. Economy Indicators Cointegration TestsUnrestricted Cointegration Rank Test (Trace)
Hypothesized No. of CE(s)EigenvalueTrace Statistic0.05 Critical ValueProbability**
None *0.84306.61159.530.00
At most 1 *0.68216.28125.620.00
At most 2 *0.65160.4595.750.00
At most 3 *0.58109.2469.820.00
At most 4 *0.5067.1647.860.00
At most 5 *0.3233.2829.800.02
At most 60.2314.2415.490.08
At most 70.031.313.840.25
Trace test indicates 6 cointegrating eqn(s) at the 0.05 level.

Denotes rejection of the hypothesis at the 0.05 level.

MacKinnon-Haug-Michelis (1999) p-values.

Trace test indicates 6 cointegrating eqn(s) at the 0.05 level.

Denotes rejection of the hypothesis at the 0.05 level.

MacKinnon-Haug-Michelis (1999) p-values.

Unrestricted Cointegration Rank Test(Maximum Eigenvalue)
Hypothesized No. of CE(s)EigenvalueMax-Eigen Statistic0.05 Critical ValueProbability**
None *0.8490.3352.360.00
At most 1 *0.6855.8446.230.00
At most 2 *0.6551.2040.080.00
At most 3 *0.5842.0933.880.00
At most 4 *0.5033.8827.580.01
At most 50.3219.0421.130.10
At most 60.2312.9314.260.08
At most 70.031.313.840.25
Max-eigenvalue test indicates 5 cointegrating eqn(s) at the 0.05 level.

Denotes rejection of the hypothesis at the 0.05 level.

MacKinnon-Haug-Michelis (1999) p-values.

Notes:Cointegration tests for the following US variables: Real GDP, stock market index, Fed Funds rate, money, credit to the private sector, imports, private consumption and wages.All variables in real terms and expressed in logarithms, except for the interest rate which is not in logs.Trend assumption: Linear deterministic trend.Included observations: 49 after adjustments.Sample (adjusted): 1958–2006.
Max-eigenvalue test indicates 5 cointegrating eqn(s) at the 0.05 level.

Denotes rejection of the hypothesis at the 0.05 level.

MacKinnon-Haug-Michelis (1999) p-values.

Notes:Cointegration tests for the following US variables: Real GDP, stock market index, Fed Funds rate, money, credit to the private sector, imports, private consumption and wages.All variables in real terms and expressed in logarithms, except for the interest rate which is not in logs.Trend assumption: Linear deterministic trend.Included observations: 49 after adjustments.Sample (adjusted): 1958–2006.
Table 1B.Residuals-Based Tests of Jamaican External Sector Cointegration with the U.S.
ADF test statistic 1/2/
Capital inflows–6.76
Current account balance–5.72
Exports of goods and services–4.74
FDI–4.92
Current Transfers–4.92

Augmented Dickey-Fuller (ADF) test statistics for the residuals of the Jamaican external sector indicators on the U.S. economy principal components.

All ADF test statistics indicate rejection of the hypothesis of a unit root.

Augmented Dickey-Fuller (ADF) test statistics for the residuals of the Jamaican external sector indicators on the U.S. economy principal components.

All ADF test statistics indicate rejection of the hypothesis of a unit root.

Appendix II. Section C Econometric Tables

This table provides support to the assumption of cointegration between the Jamaican economy model in Section C and each of the U.S. indicators used as exogenous variables. It is a residual-based test that looks at the stationarity of the residuals of each model.

Table 2.U.S. Variables Included as Exogenous in the Jamaica VAR
ADF test statistic 1/2/
U.S. GDP–4.46
U.S. Credit to Private Sector–4.52
Stock Market–4.91
U.S. Imports–4.52
U.S. Private Consumption–4.49
U.S. Wages–5.04
Fed Funds Rate–4.54
Mortgage Rate–4.49
LT Bond–4.52
MT Bond–4.52
T-Bill–4.57
Discount Rate–4.57

Augmented Dickey-Fuller (ADF) test statistics for the residuals of the Jamaican economy model with U.S. indicators as exogeneous variables.

All ADF test statistics indicate rejection of the hypothesis of a unit root.

Augmented Dickey-Fuller (ADF) test statistics for the residuals of the Jamaican economy model with U.S. indicators as exogeneous variables.

All ADF test statistics indicate rejection of the hypothesis of a unit root.

References

    AroraVivek and VamvakidisAthanasios2006“The Impact of U.S. Growth on the Rest of the World: How Much Does it Matter?”Journal of Economic IntegrationVol. 21No. 1 (March) pp. 2139.

    BaksKlaas and CharlesKramer1999“Global Liquidity and Asset Prices: Measurement, Implications and Spillovers,”IMF Working Paper 99/168 (Washington: International Monetary Fund).

    CalvoGuillermo1998“Capital Flows and Capital Market Crises: The Simple Economics of Sudden Stops,”Journal of Applied EconomicsVol. 1 (November) pp. 3554.

    CanovaFabio2003“The Transmission of U.S. Shocks to Latin America” CEPR Discussion Paper No. 3963 (London: Center for Economic Policy Research.)

    CardareliRoberto and AlessandroRebucci2007“Exchange Rates and the Adjustment of External Imbalances” Chapter 3 World Economic Outlook (April) (Washington: International Monetary Fund).

    CashinPaul and DyczewskiPawel (2006). “Government Responses to Natural Disasters in the Caribbean.” In The Caribbean: from vulnerability to sustained growth. (Washington: International Monetary Fund).

    HamiltonJames D1994“Time Series Analysis” (New Jersey: Princeton University Press).

    HelblingThomasPeterBerezinMichaelKumhofDougLaxtonNikolaSpatafora2007“Decoupling the Train? Spillovers and Cycles in the Global Economy” Chapter 4 World Economic Outlook (April) (Washington: International Monetary Fund).

    JansenW. Jos and Ad.C.J. Stockman2004“Foreign Direct Investment and International Business Cycle Comovement,”ECB Working Paper No. 401 (Frankfurt: European Central Bank).

    KaminskyGraciela and CarmenReinhart2003“The Center and the Periphery: The Globalization of Financial Turmoil.”NBER Working Paper No. 9479 (Cambridge,Massachusetts: National Bureau of Economic Research).

    Kaminsky and CarlosVegh2004“When It Rains It Pours: Procyclical Capital Flows and Macroeconomic Policies,”NBER Working Paper No. 10780 (CambridgeMassachusetts: National Bureau of Economic Researc).

    KoseM. AyhanEdwardPrasad and MarcoTerrones2003“How Does Globalization Affect the Synchronization of Business Cycles?”American Economic Review Papers and Proceedings.Vol. 93 (May) pp. 5662.

    NadalDe SimoneFrancisco2002“Common and Idiosyncratic Components in Real Output: Further International Evidence,”IMF Working Paper 02/229 (Washington: International Monetary Fund).

    RasmussenTobias2006“Natural Disasters and Their Macroeconomic Implications,” In The Caribbean: from vulnerability to sustained growth (Washington: International Monetary Fund).

    StockJames H. and MarkW. Watson2002“Forecasting Using Principal Components from a Large Number of Predictors,”Journal of the American Statistical AssociationVol. 97 (December) pp. 116779.

    StockJames2005“Understanding Changes in International Business Cycle Dynamics,”Journal of the European Economic AssociationVol. 3 (September) pp. 9681006.

Prepared by Alejandro Guerson and Christopher Faircloth.

For 2007, we use WEO estimates.

All Jamaican external sector indicators are measured in US$ millions except for the two measures for exports. In theory, introducing variables in U.S. dollars could introduce estimation bias, as these values tend to drift upwards with U.S. inflation. However, testing based on the same variables in real terms resulted in similar results. The decision to use variables in U.S. dollars is based on the fact that results are easier to interpret, especially when compared to the current status of the Jamaican balance of payments.

Only the principal components that turn out to be statistically significant are used in the equations.

The test supports the assumption of existence of a long-run relation (cointegration) between the U.S. principal components and Jamaica’s external sector indicators.

The fact that the Jamaican model is based on annual data implies that the feedback effect from one Jamaican variable to the next is picked up only in year two of the forecast. However, had a long time series with higher frequency been available, the feedback effect would have been observed earlier than a year after the shock.

Lowering of interest rates in the United States is usually associated with declining economic activity in the United States. Therefore, the impact on Jamaica should broadly be similar to that of negative shocks to U.S. GDP and consumption, which figure y generally shows to be the case.

Other Resources Citing This Publication