A Currency Union for the Caribbean
Author: DeLisle Worrell

Contributor Notes

Author’s E-Mail Address: dworrell@imf.org

The experiences of Caribbean Economic Community countries show that exchange rate depreciation in these countries is inflationary, and that, while changes in the relative prices of tradables may affect exports, tourism, and imports, nominal exchange rate changes have no predictable effect on those relative prices. Under these circumstances, economic literature indicates that a fixed exchange rate regime is optimal, and Caribbean countries with (quasi-) currency boards have been successful in maintaining durable exchange rate pegs. Commitment to a currency board is a potentially vital step in achieving a currency union for the Caribbean.

Abstract

The experiences of Caribbean Economic Community countries show that exchange rate depreciation in these countries is inflationary, and that, while changes in the relative prices of tradables may affect exports, tourism, and imports, nominal exchange rate changes have no predictable effect on those relative prices. Under these circumstances, economic literature indicates that a fixed exchange rate regime is optimal, and Caribbean countries with (quasi-) currency boards have been successful in maintaining durable exchange rate pegs. Commitment to a currency board is a potentially vital step in achieving a currency union for the Caribbean.

I. Introduction

Forty years ago, the region that is now the Caribbean Community (Caricom)2 enjoyed monetary and exchange rate stability with currencies issued by two central banks (the National Bank of Haiti and the Central Bank of Suriname) and four currency boards, one of which covered almost a dozen British colonies and territories in the Eastern Caribbean. Today there are nine different currencies in the region, in two categories. Four currencies - in The Bahamas, Barbados, Belize, and the Eastern Caribbean Central Bank (ECCB)3 area - are managed as quasi-currency boards4, pegged to the U.S. dollar, at rates which have remained unchanged since the 1970s. The remainder officially maintains floating exchange rate regimes,5 with or without management by their central banks. Monetary union is the explicit objective of Caricom, but an agreement for its introduction, reached in 1992, has not been implemented. This paper argues that its implementation is an important development initiative, and suggests a way forward.

This paper argues that a currency union among countries that meet defined criteria is the best option for Caribbean countries to secure a stable means of payment and a nondepreciating store of value. The criteria set down in the agreement for the Caricom currency union are, with one exception which is explained in the paper, sufficient. The common Caribbean currency should be pegged to the U.S. dollar via a quasi-currency board rule, following the existing model of the Eastern Caribbean Central Bank.

The next section provides background on the evolution of exchange rate policies in Caricom countries. The third section explains why a stable, nondepreciating money in the Caribbean is best attained via an exchange rate peg. The fourth section gives the rationale for a peg to the U.S. dollar, in the form of a quasi-currency board, or dollarization involving the abolition of domestic currencies, and explains why the quasi-currency board is the better option. The fifth section makes the argument for a common currency for the region – in preference to individual quasi-currency boards – and suggests a strategy for achieving currency unification.

II. The Evolution of Exchange Rate Arrangements in Caricom

At the beginning of the 1960s the countries that now constitute the Caricom group all had stable currencies, managed, in the case of English-speaking countries, by currency boards, and denominated in sterling. The British Caribbean Currency Board (BCCB), headquartered in Port of Spain, Trinidad, issued a currency which circulated from British Guiana (now Guyana) in the south through the eastern Caribbean to St. Kitts, Nevis, and Anguilla in the north. The Jamaican currency board issued a currency which also circulated in the Cayman Islands, and there were separate boards for The Bahamas and British Honduras (now Belize). Although the majority of the region’s trade and financial links were with the United States - sugar and banana exports being the main exceptions—the sterling peg occasioned no market uncertainty in a world of fixed parities. However, in 1967 the devaluation of sterling against the U.S. dollar brought an inflationary shock, as all Caribbean currency boards followed suit.

During the 1960s and 1970s countries replaced currency boards with central banks as they moved from the status of self-governing colonies to independent political units, beginning with the establishment of the Bank of Jamaica in 1961. Guyana, Trinidad and Tobago, and Barbados opted out of the BCCB to set up central banks and issue their own currencies, leaving the members of the Organization of Eastern Caribbean States as the surviving members of the common currency arrangement, which was converted to a central bank in 1983. With the break up of the global parity system in the early 1970s the sterling peg generated uncertainty in Caribbean-U.S. dollar exchange rates, and the currencies were all switched to U.S. dollar pegs, albeit at different rates, depending on the sterling-U.S. dollar rate on the day of the switch (Clarke and Danns, 1997).6

Up to the mid-1970s the new central banks behaved very much like their currency board predecessors, issuing currency in exchange for foreign exchange and maintaining an asset portfolio composed mainly of foreign exchange, with only modest lending to government and domestic financial institutions. Although not legally bound to do so, they have consistently followed a quasi-currency board strategy, maintaining levels of foreign reserves, relative to the liabilities of the central bank, such that they were always able to intervene successfully in support of the exchange rate peg. The exchange rates of The Bahamas, Barbados, Belize, and the Eastern Caribbean dollar have been managed in this way ever since. While the ECCB is the only central bank to have retained a legal prescription on the maximum holding of domestic assets, the central banks in this group have maintained reserves in excess of 50 percent of their liabilities (see Figure 1)7. On two occasions when there were sustained losses of foreign reserves (Belize in 1985 and Barbados in 1991–92) major fiscal adjustment (with supporting structural measures) was undertaken to sustain the exchange rate anchor.

Figure 1
Figure 1

Ratio of Central Bank Net Foreign Assets to Reserve Money

Citation: IMF Working Papers 2003, 035; 10.5089/9781451845372.001.A001

Source: IMF, International Financial Statistics

The quasi-currency board strategy was abandoned by the Bank of Guyana and the Bank of Jamaica in the mid-1970s, and by the Central Bank of Trinidad and Tobago in the mid-1980s (sees Worrell, 1987).8 Central bank credit to the government rose as foreign exchange reserves declined and were exhausted, and the exchange rate could no longer be sustained by foreign currency intervention. For some time an official rate was maintained, at least for some transactions, but an ever larger proportion of all transactions migrated to the parallel market over time, in spite of restrictions and foreign currency rationing. Restrictions were removed, starting in Guyana in 1988, and the exchange rates were allowed to depreciate. In the 1990s the Bank of Jamaica and the Central Bank of Trinidad and Tobago attempted to stabilize the exchange rate through interest rate policy, using indirect instruments. The Bank of Guyana has been inactive, although the pattern of exchange rate movement — not the level of the rate - in Guyana is no different from that observed in Jamaica and Trinidad and Tobago.

Caricom central banks have maintained collaboration and cooperation among themselves, almost from their inception. However, their initiatives have not been markedly successful. A multilateral scheme for intra-Caricom transfers (the Caricom Multilateral Clearing Facility, CMCF), which was set up in 1977 and operated successfully for several years, was wound up in 1983 when Guyana accumulated large arrears as a result of a prolonged balance of payments crisis. A scheme for regional travelers checks, denominated in Trinidad and Tobago dollars, never gained wide acceptance because the depreciation of the Trinidad and Tobago dollar rendered it unacceptable outside that country. The most recent initiative, the implementation of a monetary union in two phases, was agreed in 1992, but implementation remains stalled. Just when it appeared that countries comprising more than 50 percent of aggregate regional GDP9 had met the criteria for implementing Phase One of the agreement, Caricom’s largest economy, Trinidad and Tobago, failed the crucial exchange rate stability test.

III. Effects of Exchange Rate Changes

In the Caricom experience of the past thirty years, it has been possible to maintain stable, low inflation in economies with sustained growth only in the context of exchange rates pegged to the dollar under quasi currency board rules, either de jure or de facto. The nominal exchange rate has proven ineffective as a shock absorber or adjustment policy tool, because neither the external commodity nor the external financial market has a market-equilibrating mechanism based on the exchange rate. An increase in the price of foreign exchange does not provoke an increase in the net supply of foreign exchange, nor does a fall in the price of foreign exchange induce an increase in the net demand for foreign exchange, in either market. To see why, let us consider separately the goods and services and financial markets.

A. The Market for Traded Goods and Services

Conventional theory leads us to expect that an increase in the relative price of tradables induces an increase in the domestic supply of tradables, a fall in the domestic demand for tradables, a fall in the domestic supply of nontradables, and an increase in the domestic demand for nontradables, if not in the short run, then in the medium or long run. A useful summary of the conventional view may be found in Isard and Faruqee (1998). In their model the current account balance is a decreasing function of the real effective exchange rate, implying that a (real) depreciation improves the current account balance. Both exports and imports are affected by exchange rate changes, as well as by foreign and domestic real incomes, respectively.

However, consider the composition of the tradable sector in the typical Caribbean country: it comprises one or two of the following: tourism, a major agricultural staple (bananas, sugar, or rice), a mineral (oil or bauxite), and the manufacture of beverages (mainly rum) (see Table 1). In each of these activities supply is constrained by domestic capacity, while domestic demand absorbs nothing (in the case of bauxite) or a very small proportion of production (see Table 2). The effect of changes in relative prices on net domestic supply is therefore negligible in the short run. The effect of relative price changes in the medium and long term must be considered along with other factors that affect investment decisions: changes in technology, changes in taste, marketing initiatives, and changes in the international ownership of domestic capital. Let us consider each export sector in turn.

Table 1.

Exports of Goods and Services, By Sector

article image

Where data on enclave manufacturing are not available, the conservative assumption is made that two-thirds of all manufacturing—including enclave manufacturing—is consumed domestically.

Where data are not available, the international financial service sector is aggregated with tourism.

Discrepancies due to rounding.

2000.

Sources: IMF Staff Country Reports, various issues; Central Bank of Barbados, Annual Statistical Digest, 2001; and Bank of Guyana, Half Year Report and Statistical Bulletin
Table 2.

Exports as a Percentage of Production of Exportables, 2001 1/

article image

Exports of goods and services. The notation …. indicates that the country does not produce that commodity.

Where actual data are not available, a conservative assumption is made, that only one-third of the production of miscellaneous exportables is exported.

International financial services centers.

2000.

Sources: IMF Staff Country Reports, various issues; Central Bank of Barbados. Annual Statistical Digest, 2001; and Bank of Guyana, Half Year Report and Statistical Bulletin

Investment in hotels in the Caribbean in recent times has been influenced by cycles in the overall growth of resort tourism,10 the increasing popularity of cruise tourism, the increasing preference for prepaid “all inclusive” arrangements, the effects of airline deregulation, the impact of computerization and the internet on travel patterns and industry organization, and the emergence of new resorts in the Caribbean and in competing tropical destinations (Poon, 1993). Investment in the sugar industry has been affected by changes in protective export arrangements, changes in agrarian technology and practice in the Caribbean and in competing sugar exporting countries, ecological and health concerns affecting production processes and consumption patterns, and investment in associated industries, especially rum production. Funding for investment in the banana industry has been provided by the European Union, in anticipation of changes in negotiated agreements with the Community. The main factors in investment in the rice industry were the liberalization of investment and trade in Guyana in the 1980s and the provisions of the Lomé Agreement with the European Union. In addition to the impact of product-specific investment, the supply of all agricultural products was affected by weather (including hurricanes, flood, and drought) and other natural phenomena.

Investment in the petroleum and bauxite/alumina industries has responded to changes in world prices, the cost of extraction of domestic deposits, and the changing nature of international competition in these industries. Investment in mineral and agricultural export sectors has been seriously inhibited by market uncertainty. Except for petroleum and related sectors, there has been little or no expansion of capacity. Investment has financed rehabilitation and productivity upgrades to maintain competitiveness, sustain output levels, or slow the pace of output loss.

Product prices, relative to costs and competing supplies of similar quality, have undoubtedly been a factor in investment decisions in the sectors discussed above. However, it is impossible to predict the importance of relative price changes, in view of the complexity of decision processes just described. It may be that relative prices have an effect comparable to that of the nonprice factors, but they may be less significant or negligible in comparison, or they may have perverse effects, in cases where price is seen as a reflection of quality.

The typical composition of the nontradable sector in Caricom countries is wholesale and retail services, government services, business and personal services, and public utilities. None of these is substitutable for tradable goods, in production or consumption. The resources – skills, machines, and structures – used in the production of these nontradables may not be costlessly reallocated to the production of the tradables mentioned above: “defenders of floating exchange rates … point to the fact that flexible exchange rates make the adjustment of relative prices less costly, because equilibrium changes can be accommodated by a higher or a lower exchange rate with little effect on output and employment… However, in a realistic economy there are several distinct goods, each with a distinct labor market: gauchos cannot be quickly retrained as nuclear physicists, and vice versa” (Calvo and Reinhart, 1999, page 21). Major initiatives for retraining were incorporated in adjustment programs for Trinidad and Tobago (1985-86), Guyana (1988–89), Barbados (1991–92) and Jamaica, on several occasions since the mid-1970s, none of which may be considered an unqualified success.

B. Nominal and Real Exchange Rates

The effects discussed above depend on changes in the relative prices of tradables. However, in the Caribbean no predictable relationship has been established between the nominal exchange rate, the policy instrument, and relative prices. A simple linear relationship would obtain only if: (a) there were no change in the fiscal deficit or government’s financial requirements as a result of the exchange rate change; (b) there were no interest rate response; (c) there were no induced capital flows; and (d) there were no wealth effects on aggregate demand, credit, and financial accumulation. Since none of these restrictions holds, the relationship between nominal exchange rate changes and changes in relative prices is complex. It will reflect the interrelationships of the variables, both contemporaneously and after allowing for lagged effects. No one has attempted an exchange rate model that allows for such interrelationships for Caribbean countries.

However, there are reasons, based on the structure of Caribbean economies, why nominal devaluations are not expected to produce large changes in real exchange rates, except in the short term. Devaluation drives a wedge between domestic (i.e., nontradable) producer prices and foreign (i.e., tradable) prices, to the extent that nontradables do not require imported inputs. In the Caribbean the imported content of nontradables is very high, for all commodities produced, and the staples of the diet and other basic consumption goods are also imported. In these circumstances, the pressures to adjust nominal wages to exchange rate changes are especially strong and persistent.

Nevertheless, unless labor markets anticipate devaluations, wage costs catch up with prices only with a lag. As Lewis (1972) anticipated, this appears to have provided a temporary profit windfall for tradables producers which was used for a one-shot investment injection, on at least one occasion.11 This is not a strategy that can be repeated: unless labor markets are convinced that the new exchange rate will persist, devaluation expectations may become entrenched, causing overreaction and an eventual appreciation in the real exchange rate. Moreover, evidence is accumulating, world wide, about the disincentive effects of exchange rate uncertainty on investment (Pindyck, 1991), exports (Arize, 1996), and the long-term growth of capital markets (Reinhart, 2001), though the interpretation of this evidence has occasioned controversy (see, for example, Böhm and Funke, 2001).12

Unfortunately, much of the empirical literature on the effects of exchange rate changes uses inadequate measures of the real exchange rate. For countries whose exports are too small a proportion to affect world market prices (or prices in the target consumption markets), the most representative measure of relative prices is the relative price of tradables to nontradables. Marsh and Tokarick (1994) and Bynoe-Mayers (1997) discuss the disparity between this statistic and the usual empirical measure, the nominal exchange rate adjusted by indices of relative consumer or wholesale prices at home and abroad. (A similar discrepancy is highlighted in Masters and Ianchovichina, 1998.)

C. The Financial Market

Mechanisms exist in Caribbean financial markets for full interest rate arbitrage, allowing domestic interest rates to adjust to the level of comparable U.S. interest rates, the expected exchange rate change, and expected changes in the domestic rate itself (if, for example, the fiscal deficit or a decline in foreign exchange reserves is deemed unsustainable by market participants). These mechanisms include: foreign ownership of domestic banks and other financial institutions, foreign correspondent links by domestic financial institutions, shifts in trade credit (equivalent to 50 percent of exports of goods and services in some Caricom countries), changes in the patterns of remittances (for example, substitution between services in kind and foreign exchange transfers), and the use of informal channels of financial transfer. Through interest rate arbitrage, an expected change in the nominal exchange rate can be discounted in the domestic financial market, and it therefore provokes no change in the net supply of foreign exchange.13

Exchange rate changes tend to be offset by interest rate changes, without any foreign exchange supply response, and the increased exchange rate uncertainty that comes with a flexible exchange rate brings with it a higher interest cost. Figure 2 illustrates the interest premiums which resulted from exchange rate uncertainty in a selection of Caribbean countries. It may be seen that the domestic treasury bill rates in Jamaica and Guyana were very much higher than the U.S. treasury bill rate, after their exchange regimes were liberalized in 1991 and 1988, respectively. The premium which the Government of Trinidad and Tobago paid on treasury bills, compared to U.S. rates, was also higher after the exchange rate was allowed to float officially in 1993, although the differential is less pronounced than for Jamaica and Guyana. It may be seen that the premium on the domestic treasury bill is much higher for Jamaica and Guyana than for The Bahamas, which maintained a fixed exchange rate. The premium on the Trinidad and Tobago rate is also above that for The Bahamas, though much lower than for Jamaica and Guyana. It mirrors the relative stability of the Trinidad and Tobago dollar, which has changed much less in value since the exchange rate was floated than for the other two countries.

Figure 2
Figure 2

Treasury Bill Interest Differntials, Caricom Minus US, and Exchange Rate Changes

Citation: IMF Working Papers 2003, 035; 10.5089/9781451845372.001.A001

Source: IMF, International Financial Statistics

While expected exchange rate movements tend to be offset by changes in the interest rate, unexpected exchange rate changes have had asymmetrical effects. Unexpected revaluation of the exchange rate failed to induce any reaction in Jamaica in 1993 and 1996,14 as market participants waited to see whether the revaluation would be sustained. The reaction to unexpected devaluation has been capital flight, in Guyana, Jamaica, and Trinidad and Tobago, as market participants sought to insure against the possibility of future devaluations.

The Caricom experience of exchange rates and the capital account follows a well-established pattern. A devaluation15, rather than increasing the supply of foreign exchange on the financial market, leads to capital flight in the short term. Moreover, exchange rate expectations are heavily influenced by previous exchange rate experience. Devaluation typically is followed by a period of high domestic interest rates, relative to U.S. rates, reflecting expectations of further depreciation. In an environment of high interest rates, the yield curve often becomes inverted, finance tends toward the short term, and the investment/GDP ratio declines. In the conventional theory of exchange rate adjustment, in an economy with rational expectations the real exchange rate would converge on its long-term equilibrium value, and at this rate there would be a balance of saving and investment, determined by economic fundamentals such as income per capita, the fiscal deficit, the extent of spare capacity, real wages, and the ratio of dependents to the total population (Isard and Faruqee, 1998). In practice, the persistence of high interest rates may cause a financial position that would otherwise be viable to explode into crisis (Hausmann and others, 1999). A deliberate policy of fighting high inflation with high interest rates may also be destabilizing, in terms of actual output (Calvo and Vegh, 1995).

D. The Empirical Evidence

The Raw Data

A preliminary examination was made to see whether the raw data reveal an apparent balance of payments response to exchange rate changes, beginning with a comparison of exchange rate changes and inflation. To the extent that changes in the exchange rate are correlated with inflation, any relative price advantage the country may have gained from a devaluation is eroded. Figure 3 compares exchange rate changes and changes in the CPI, for three countries with flexible exchange rates (Haiti, Jamaica, and Trinidad and Tobago), Suriname, where the unofficial rate soared in the 1990s, and countries with fixed exchange rates. Jamaica and Suriname offer the most noticeable cases where domestic inflation is correlated with exchange rate depreciation, for Jamaica in 1978–79, 1984, and 1992–96, and for Suriname in 1994–95 and 1999. In Trinidad and Tobago, devaluations in 1985, 1989, and 1993 show less marked correlation with changes in the CPI; in the case of the 1985 devaluation, the corresponding price increase seems to have occurred about one year later. In Haiti high volatility of the exchange rate appears to be correlated with high inflation, in the 1990s.

Figure 3
Figure 3

Exchange Rate Changes and CPI Inflation

Citation: IMF Working Papers 2003, 035; 10.5089/9781451845372.001.A001

Source: International Financial Statistics

Figure 4 compares nominal exchange rate changes with changes in the real exchange rate, measured as the ratio of the deflator for tradable goods to the deflator for nontradables. The series of devaluations in Guyana in the second half of the 1980s correlates with the depreciation of the relative price of tradables in that country, though it is the nominal exchange rate change which lags the change in relative prices. In Barbados, Jamaica, and Trinidad and Tobago there is no correlation between the trends in relative prices and in exchange rates.

Figure 4
Figure 4

Trends in Relative Price Deflators (Left Scale) and Exchange Rates (Right Scale)

Citation: IMF Working Papers 2003, 035; 10.5089/9781451845372.001.A001

Sources: IMF, International Financial Statistics and Central Bank of Barbados, Annual Statistical Digest 2001.

Figure 5 shows the relationship between change in the relative prices of tradables and changes in exports of goods and services (proportionate changes, based on exports and services measured in U.S. dollars), for Barbados, Guyana, Jamaica, and Trinidad and Tobago. In no case is there an obvious correlation: observations are scattered apparently at random around the fitted lines, and there in no convincing trend in any of the fitted lines. The same is true for changes in imports for Barbados, Jamaica, and Trinidad and Tobago (based on imports measured in U.S. dollars), shown in Figure 6, and for current account balances (in U.S. dollars), shown in Figure 7. Only in Barbados is there a noticeable trend in Figure 7, for an improvement in the current account as the relative price of tradables falls, that is, as the country becomes less competitive, a counterintuitive result.

Figure 5
Figure 5

Exchange Rate Changes and Export Changes

Citation: IMF Working Papers 2003, 035; 10.5089/9781451845372.001.A001

Source: IMF, International Financial Statistics
Figure 6
Figure 6

Exchange Rate Changes and Import Changes

Citation: IMF Working Papers 2003, 035; 10.5089/9781451845372.001.A001

Source: IMF, International Financial Statistics
Figure 7
Figure 7

Exchange Rate Changes and the Current Account

Citation: IMF Working Papers 2003, 035; 10.5089/9781451845372.001.A001

Source: IMF, International Financial Statistics

This first look at the raw data is merely suggestive, because it takes no account of multivariate influences on prices, exports, services, imports, or the current account, or of possible lagged effects and indirect and simultaneous influences.

A survey of empirical studies

There appears to have been no attempt to estimate the effects of exchange rate policy using a global approach that allows for time lags, and simultaneous and indirect effects of the exchange rate on the demand and supply of foreign exchange. Structural models exist that might have been adapted for that purpose, but none has been used to simulate the effects of exchange rate changes. All the evidence we have, therefore, relates to the partial impact of exchange rate changes, and must be approached with the caveat that these effects may be negated by other consequences of exchange rate changes which also impact the balance of payments, or that the effects may wear off with the passage of time.

In a study of the export performance of Barbados, Costa Rica, the Dominican Republic, Jamaica, and Trinidad and Tobago, Harris (1994) found that depreciation of the exchange rate (adjusted for weighted changes in domestic prices relative to prices in trading partners) generated increases in exports, with an elasticity of 0.17 (a 10 percent devaluation induces a 1.7 percent increase in exports), on average for this group of countries. (Other factors contributing to export changes included capital inflows, the ratio of arable land to population, and the ratio of investment to GDP.) McIntyre (1995) found that the elasticity of clothing exports with respect to the real effective exchange rates (REER), measured as above, was greater than unity, for Barbados, Costa Rica, and the Dominican Republic. For total exports, the REER elasticity was statistically significant only for Trinidad and Tobago, with a value of 0.75. McIntyre notes that the REER may be adjusted by the use of incentives, as well as by nominal exchange rate changes. (In McIntyre’s tests the national capital stock also appears as an important explanatory variable, especially for clothing exports.) Worrell, Boamah, and Campbell (1996) observed that repeated devaluations by the Dominican Republic, Guyana, and Jamaica reduced their prices, wages and unit labor costs, relative to those for Barbados. In tourism, Barbados’ earnings grew more slowly than those of the Dominican Republic and Jamaica, and Barbados’ agricultural exports also grew more slowly than for the three comparator countries. In manufacturing, a minor export for Barbados, the country’s exports grew more quickly than Guyana’s and Jamaica’s, and kept pace with growth in the Dominican Republic.

Inferences about the possible impact of a devaluation on tourism may be drawn from the price elasticity estimates to be found in Clarke, Wood and Worrell (1986), Rosenweig (1988), and Whitehall and Greenidge (2000). In the most detailed tourism demand study reported for any Caribbean country, Clarke, Wood and Worrell estimated equations for Barbados, separately by tourists’ country of origin, type of accommodation, and season of travel. The variable that captures competition from other Caribbean destinations – hotel rates in Antigua - has no statistically significant impact on any segment of tourism demand. In contrast, Rosensweig found significant price elasticities of substitution among Caribbean countries (1.33), between the Caribbean and Mexico (1.85), and between Europe and the Caribbean (1.78), all for visitors from the United States. Elasticities of substitution among Caribbean countries and between the Caribbean and Mexico were even higher for the worldwide tourism market. Whitehall and Greenidge found a much smaller value, 0.68, for the elasticity of Barbados’ tourism with respect to relative prices, measured as the Barbados GDP deflator relative to the U.S. CPI.

Inferences about the direct, first-round impact of devaluation may be drawn from the estimated coefficients of import and export equations in structural models, even though the models were not simulated for the effects of exchange rate changes. Watson and Teelucksingh (1997) found a significant import decline in response to an increase in the unit value of imports relative to domestic prices in Trinidad and Tobago. The coefficient of export changes with respect to changes in the unit value of exports relative to a world price index is not significant. World price changes have a significant impact on domestic inflation. In a later study - Watson and Teelucksingh (2000) - the same authors estimate a significant increase in exports in response to a fall in the domestic price of exports relative to the local currency equivalent of U.S. industrial prices. Worrell (1987) found that the price of tradables relative to that of nontradables (measured by deflators) had no significant effect on the output of nontradables in Barbados or Jamaica, or on imports to Jamaica. However, an increase in the relative price of tradables significantly depressed Barbados’ imports. In a later study, Worrell (1992) completed estimates for the Dominican Republic, Guyana, and Trinidad and Tobago as well, with additional observations. The relative price impact on Jamaica’s imports and on the output of nontradables for Barbados was statistically significant, but all others were not. Increases in the price of tradables elicited a statistically significant increase in the output of tradables only for Jamaica.

A majority of studies indicates that changes in nominal exchange rates strongly affect domestic prices, and therefore can have only weak effects on relative prices. In the most comprehensive estimation of inflation formation for Barbados, Jamaica, and Trinidad and Tobago, Holder and Worrell (1985) estimate one-to-one impact of foreign price changes on the domestic price of tradables, affecting about one-third of production in each country, plus significant effects on foreign prices on the prices of nontradables in Barbados and Trinidad and Tobago (for Jamaica the coefficient measuring this impact was not statistically significant), as well as an inflationary impact of changes in the relative price of tradables on domestic consumer prices in Barbados. Lattie (2000) obtains a statistically significant coefficient of 0.3 for the effect of lagged exchange rate changes on domestic inflation in Jamaica, and Cumberbatch (1997) estimates a high elasticity of 4.0 for the impact of foreign prices on Barbados’ consumer price index. Other studies which estimate a significant impact of foreign prices on domestic prices for Barbados include Coppin (1993) and Downes, Holder, and Leon (1990), with elasticities of 0.09 and 0.25, respectively.

Overall, the empirical results available in the literature indicate a vigorous domestic price response to exchange rate changes, though estimated elasticities vary from one-third to much greater than unity. In other respects, the results are a mixed bag. While a majority of studies do indicate some influence of changes in relative prices on exports of goods and services and on imports, the results are not consistent across studies, and the approaches are too diverse to have produced robust inferences.

Skepticism about an active exchange rate strategy for Caricom countries is reinforced by growing sentiment, reflected in the literature, that exchange rate flexibility is inappropriate for small open economies such as Caricom member states. Caricom members satisfy most of the criteria suggested by a survey summarized in IMF (1997) for countries for which a pegged exchange rate might be advantageous: small size, economies that are very open to international trade and finance, a high degree of export concentration,16 an overwhelming proportion of external transactions conducted in a single foreign currency (the U.S. dollar), low inflation (except for Jamaica and Suriname), and a relatively high incidence of domestic nominal shocks. Dornbusch (2001) makes the case for more widespread use of currency boards, including for countries much larger than Caricom members. In their advice to countries seeking greater flexibility in exchange rate management, Eichengreen et al. (1998) recognize that tourism-based economies, and economies that trade with large neighbors, gain little from independent monetary policy, and therefore recommend exchange rate pegs in these cases. Moreover, there is evidence that many small countries around the world do in fact peg their currencies by intervention, either directly or via financial markets, irrespective of the formal status of their exchange rate regimes (Calvo and Reinhart, 2000).

IV. Currency Board and Full Dollarization

The foregoing points firmly in the direction of a fixed exchange rate regime for Caricom member countries: the market for traded goods and services is so structured that nominal exchange rates do not have significant expenditure switching effects, either in the short run (on the demand side) or in the long run (on the supply side); the relationship between nominal and real exchange rate changes, if it exists, is tenuous and unpredictable; the financial market is quite open, neutralizing expected exchange rate effects and penalizing unexpected changes with high investment costs; and the empirical evidence is fully consistent with these inferences.

The choice of an exchange rate peg is also straightforward, for Caricom member countries, because foreign transactions, both financial and on the current account, are overwhelmingly denominated in U.S. dollars. The entire Caribbean tourism product is priced in U.S. dollars, and all private foreign capital flows are dollar-denominated. Of countries with significant extra-regional exports other than tourism, all except Suriname earn 50 percent or more of their income in U.S. dollars. One-third to half of imports come from the United States, and much of the remainder is priced in U.S. dollars (see Worrell, 1993). With the decline in agricultural exports throughout the region, changes in the U.S. dollar value of sterling and the euro have relatively little impact, and no other currency is of importance in the international transactions of Caricom countries.17

The risk of an exchange rate crisis, resulting in catastrophic devaluation, is the only real cost of the pegged exchange rate in Caricom countries. This follows from the absence of predictable exchange rate effects of nominal exchange rate changes, implying that surrendering the exchange rate tool entails no loss in the armory of effective stabilization policies. In practice, only those Caricom currencies which were managed as quasi-currency boards have managed to avoid exchange rate crises. With relatively open financial markets, high levels of reserve backing appear to be an essential element in insuring against currency risk, even when fiscal and macroeconomic policies are fully consistent with the exchange rate.

The alternative, which eliminates currency risk altogether, is full dollarization. There are two well-established arguments for preferring the currency board over dollarization: the former allows for limited seigniorage accruing to the currency issuing authority,18 and that authority may serve as lender of last resort for domestic financial institutions, which have no recourse beyond the interbank market if the economy is fully dollarized.19 This puts domestic financial institutions at a disadvantage, in competition with banks which have recourse to an overseas head office.

There is a third and possibly stronger argument for the quasi-currency board: it lends credibility to domestic policy, especially fiscal policy (Baliño and others, 1997), because it is a very transparent mechanism. In the Caribbean, this is demonstrated by the recent economic history of The Bahamas, Barbados, Belize, and the OECS member states. Central banks in these countries have all followed quasi-currency board rules, de facto (or, in the case of member countries of the ECCB, de jure). On the whole, they maintained ratios of foreign exchange reserves close to or exceeding the value of their currency liabilities. Both Belize (in 1985) and Barbados (in 1991–92) came close to exhausting foreign reserves completely, but restored adequate levels by means of programs to reduce aggregate expenditure, thereby demonstrating credible commitment to the exchange rate peg. In the central banking era in the Caribbean - from the late 1960s to the present - these countries have experienced structural changes and economic shocks similar to those experienced by Jamaica and the Dominican Republic, countries where central banks did not operate de facto as quasi-currency boards.20 The governments of the quasi-currency board countries have enjoyed ongoing credibility for fiscal policies, in contrast to the experiences of Jamaica and the Dominican Republic, and have been able to borrow at interest rates which show only a small premium over comparable U.S. dollar interest rates (see Figure 8).21

Figure 8
Figure 8

Interest Rate Premiums, Fixed Rate Countries Percentages

Citation: IMF Working Papers 2003, 035; 10.5089/9781451845372.001.A001

Source: IMF, International Financial Statistics

In a fully dollarized economy there is no yardstick for measuring the efficiency of fiscal policy which gives as timely and unambiguous a signal of the need for corrective action. There is a need for measures of fiscal performance which are sufficiently prominent in the public mind – as is the rate of inflation in the popular targeting framework – that expected deviations from the target will attract public sanction and induce government to take corrective action before macroeconomic disequilibrium becomes acute. Under the quasi-currency board regimen, the level of foreign reserves in relation to central bank liabilities, and in relation to foreign obligations, the two benchmarks that define the currency board, perform the tracking function, and give early warning of the need for policy adjustment.

V. A Currency Union for the Caribbean

A. The Case for a Single Investment Market

The adoption of currency board rules may be an essential step in achieving a currency union for Caricom, and a vital element in the creation of a regional capital market of sufficient size for the development of internationally competitive firms. This capital market would be a single economic space where investors from all member countries would jointly conceive, plan, and implement investment projects in the activities in which the Caribbean has already demonstrated a comparative advantage – tourism, export agriculture and mineral production (and manufacturing, in the case of Trinidad and Tobago only) – and those service exports where there seems to be as yet unrealized potential – information services and entertainment.

The case for integration of the regional capital market lies with the gains from eliminating transactions costs and overcoming market distortions and barriers that inhibit cross border investment within Caricom, and reduce overall levels of potential investment.22 There are also potential dynamic gains from economies of scale, technology transfer, and learning by doing, in integrating the region more firmly in the U.S. dollar area, discussed in Burki, Perry, and Calvo (1997). In the most complete analysis of trade and development in Caricom countries to date, McIntyre (1995) concludes that Caricom should be constituted as a single economic space for the development of exports and competitive import substitutes.

For the most part, Caribbean enterprises are far too small to be competitive internationally. Of companies quoted on the stock exchanges of Barbados, Jamaica and Trinidad and Tobago, or traded over the counter elsewhere, only 21 had turnover in excess of US$100 million in their fiscal year 1999 (Medianet Ltd., 2000), although the largest mineral and tourism enterprises are not quoted. Of those 21, at least 15 operate in more than one Caricom country. The average annual turnover of the three largest quoted Caribbean companies (US$480 million) is 3 percent of the average turnover of the three largest companies in Latin America, as reported by the Financial Times “FT 500: The World’s Largest Companies,” May 11, 2001. (Latin America had three firms in the list of the largest 500.)

A strategy of export diversification, to cushion the effects of adverse external shocks, is unlikely in the absence of strong indigenous firms of regional scope. Foreign investment has tended to follow well trodden paths of demonstrated profitability, leading to concentration on single products or single markets by most foreign investors. Because information is costly, foreign investors are often not aware of potentially profitable investment opportunities in emerging sectors. Foreign investors are frequently unwilling to share product and market development costs on which they will not realize full or immediate returns, and they are sometimes unwilling to ride out downturns in the economic cycle. “The key to investment strategy is to get domestic entrepreneurs excited about the home economy. Encouraging foreign investment or liberalizing everything and then waiting for things to improve does not work.” (Rodrik, 2002).

A regional capital market has begun to emerge, with indigenous firms that compete in extra-regional markets, and firms that compete effectively with international firms in the domestic market, having investments across Caribbean countries, in tourism, banking, insurance, and wholesale and retail services. However, in the absence of currency union, it remains small in relation to total investment in Caricom economies.

B. The Role of Currency Union

The cost of exchange rate uncertainty among regional currencies is a major barrier to unification of the Caricom investment market. A characteristic of Caricom currencies is that, unlike “core” currencies such as the U.S. dollar, the euro, the yen or sterling, private and public agencies in the issuing countries cannot borrow abroad in the domestic currency (Bordo and Flandreau, 2001). This remains true of Caricom member countries for borrowing among themselves. Although arrangements have been in place for cross-border transactions and quotations on the securities exchanges of Barbados, Jamaica, and Trinidad and Tobago for more than a decade, no major investment has been financed in this way, in the domestic currencies. Cross-border investment within Caricom, for example recent borrowings from Trinidad and Tobago banks by governments of eastern Caribbean countries, have been denominated in U.S. dollars.

A common currency would eliminate these currency conversion costs and risks and would reduce transactions costs on products and services that are shared across the enterprise.23 It would also add to the seigniorage revenue of the regional quasi-currency board, by replacing the U.S. dollar, the present common unit of account and vehicle for intra-regional settlement, with the Caribbean currency.24 The introduction of a common currency, linked to the U.S. dollar via a regional quasi-currency board, would therefore be a landmark step in the development of the regional capital market – that is, the market for the funding of fixed investment – though other measures, notably the implementation of agreements for the free movement of persons, would be necessary to complement the capital market integration.

C. Potential Cost of Currency Union

The potential costs due to the diversion of trade within the currency union are trivial. As empirical studies have shown (Gondwe and Griffith, 1989; World Bank, 1990; Lewis-Bynoe and Webster, 2001), the potential for intraregional trade growth is minimal. The implication of the arguments developed in Section III is that the nominal exchange rate does not function as a shock absorber, since it has no expenditure switching effect. There is therefore no cost to giving up this tool.

The main cost of joining the currency union is political. Member countries which have maintained stable quasi-currency board regimes are reluctant to join a new regime, however carefully designed, simply because it is untried, in comparison with their own well-established track record. The incentive for them is their poor growth prospects under current arrangements, in an environment of increasing trade and financial liberalization. Countries that have maintained flexible regimes do not appear ready to make the long-term fiscal and structural commitments which adherence to a pegged rate regime would imply. However, the failure to make such a commitment continues to extract a severe penalty in terms of high interest rates, an uncertain climate for investment and low growth potential.

D. Fiscal Discipline, Financial Constraints, and Convergence

The issues of fiscal discipline and economic convergence among the members of the monetary union were fully explored in the design of the existing unconsummated arrangements for Caricom monetary union. (Much of this background material appears in Farrell and Worrell, 1994.) Under these arrangements, fiscal discipline is achieved via an eligibility criterion for accession to the monetary union, instrument independence for the central bank of the monetary union, and rules governing central bank lending to member governments, as follows:

  • A sustainable external debt-service ratio, no higher than 15 percent of current account receipts, is one eligibility criterion for joining the monetary union;

  • The central bank of the monetary union would have independence in the use of monetary instruments, including open market operations;

  • The central bank will be constitutionally forbidden to lend to governments, except by way of securities issued on the open market.

These stipulations are sufficient, in an open economy in a world with well-informed international financial markets, to ensure fiscal discipline and eliminate pressure for money creation. An intuitive explanation is that government finance requirements, should they exceed the domestic supply of funds, will spill over into the foreign market, where government is constrained by solvency requirements. However, the international financial market is not always fully informed, and may continue to lend to governments for some time after they have exceeded prudent limits of borrowing; therefore, a stipulation is made on the external debt service ratio.25 With open financial markets, a separate limit on domestic debt is redundant.

The combination of an exchange rate anchor and the above stipulations for fiscal discipline ensure convergence of inflation and interest rates, to levels prevailing in the United States. However, additional screening will be necessary of each potential member at the point of admission, to ensure there are no other sources of imbalance. Such a stipulation would be required with respect to Jamaica’s accession to the monetary union under current circumstances. The Bank of Jamaica has kept the currency stable, maintained prudent fiscal balances, and contained external debt service, but has been forced to sell government securities at high real rates of interest, which cannot be sustained. In this case, the domestic financial requirement has not spilled over into the international financial market because of the high interest premium on domestic currency. The reduction of this premium would have to be a precondition for Jamaican accession to the monetary union.

Efficient and transparent financial regulation and supervision are essential for the credibility of the exchange rate peg. Financial institutions should be subject to prudential limits on their net open positions in foreign currencies, and limits on exposures via currency mismatches, in the prevailing climate of worldwide exchange rate volatility. Arrangements for government borrowing from financial institutions should be fully transparent, so that the price of credit responds to any excessive demand for finance from government. In particular, there should be no directed credit, either for state-owned or private financial institutions.

Monetary union will not, of itself, promote convergence of output growth, employment growth, productivity growth, or a sustainable balance of external payments. The direct impact of monetary union on these variables is neutral, precisely because exchange rate policy is not an effective allocation tool for Caricom countries. Henry and Downes (1994) found that there would be little immediate or short term effect on labor markets, and that the longer-term effects, while largely speculative, might well be positive. Fiscal policy could not be the same everywhere because of structural differences among Caricom members, most obviously in the case of Trinidad and Tobago, the group’s only oil exporter. (Fiscal implications are discussed in Theodore, 1994.)

The long-term viability of the monetary union is unaffected by potential divergences in the evolution of the equilibrium real exchange rates of members, precisely because there is little relation between the nominal value of the currency and the relative price — or cost of production – of tradables and nontradables. The equilibrium “real effective exchange rates,” the most common measure of the “real” exchange rate, will converge, as the inflation rates of member countries converge on the U.S. inflation rate. Policies other than nominal exchange rate changes, for example labor market reforms and other structural policies, must be used to ensure convergence in the equilibrium paths of relative prices of tradables and nontradables.

E. The Path to Currency Union

The member countries of the Caribbean Economic Community (Caricom) in 1992 signed an agreement for monetary union,26 with the following criteria for accession to the union: the prospective member must have

  • Maintained an unchanged U.S. dollar value of its currency for at least 36 consecutive months;

  • Maintained a minimum of foreign exchange reserves equivalent to three months’ of imports, for at least 12 consecutive months;

  • Recorded a ratio of external debt service to exports of goods and services of no more than 15 percent.

These criteria remain sufficient to assure a path towards Caribbean monetary union, with one exception: experience has proven the exchange rate criterion to be inadequate. The principal reason the agreement was not implemented was the devaluation of the Trinidad and Tobago dollar in 1995, after a little over three years when that currency’s exchange rate had not changed. That experience indicated that, contrary to the implicit assumption underlying the agreed Caricom criterion, three years of an unchanged exchange rate, even when backed with reserve cover above the stipulated limit and a moderate external debt service ratio, was not enough to reduce the expected exchange rate change to zero.

The currency board rule offers an alternative, and effective, means of reducing expected exchange rate changes to zero during the transition to the currency union. The recommendation of this paper, therefore, is that the agreed criteria for Caricom monetary union should be supplemented by the requirement that potential members commit to the currency board rule.27 For additional credibility in the foundation years of the monetary union, the currency board requirement could be a strict one – involving a minimum of 100 percent foreign exchange cover of domestic liabilities – for any country that has recorded an exchange rate change in the past decade.

VI. Summary

From a theoretical point of view, changes in the exchange rates of small, very open economies do not result in switches in expenditure, towards production for export and consumption of import substitutes, and they either have no effect on financial inflows and outflows (if the exchange rate change is anticipated) or they precipitate capital flight (in cases of unanticipated devaluation). These theoretical considerations, as well as experience over the past three decades, indicate that a fixed exchange rate regime is preferable for Caricom countries. The experiences of Caribbean countries show that exchange rate depreciation is inflationary, and that, while changes in the relative prices of tradables may affect exports, tourism, and imports, nominal exchange rate changes have no predictable effect on those relative prices.

A quasi-currency board, which ensures the stability of the exchange rate by maintaining a supply of foreign exchange reserves for its defense, is the only regimen that has produced a sustainable fixed exchange rate in the Caricom region. It has advantages over full dollarization: some seigniorage accrues to the domestic monetary economy, the monetary authority may act as a lender of last resort for domestic banks, and a commitment to a fixed exchange rate lends credibility to appropriate fiscal policies.

The Caricom region has as an objective a single market for investment, to achieve the scale and capture the synergies needed to produce internationally competitive firms. A common currency is potentially a vital pillar in the unification of the Caribbean investment market. It may be achieved, under the conditions already agreed by Caricom heads of government, if all members commit to currency board regimes before acceding to the union.

A Currency Union for the Caribbean
Author: Mr. Rupert D Worrell